Applied Evidence

Fifth and sixth diseases are frequently encountered viral exanthems in family medicine. This article delineates the unique clinical characteristics of these disorders, describes rare but serious sequelae of each, and offers recommendations to guide your practice.

Fifth disease

Parvovirus B19, an infectious agent found worldwide, is the cause of fifth disease, also known as slapped cheek syndrome or erythema infectiosum. It is transmitted via respiratory droplets, most commonly in late winter and early spring. The peak incidence of parvovirus B19 infection is in children ages 5 to 15 years.¹ Approximately 20% of parvovirus B19 infections remain subclinical.^1,2 An observational study of children in the United Kingdom who were 6 months to 16 years of age and had been immunized for measles and rubella revealed that parvovirus B19 was the number one identifiable cause of febrile rash, responsible for 17% of cases.³ Seroprevalence increases with age, and 40% to 60% of adults test positive for prior infection.¹

Clinical presentation: Not necessarily limited to fever and rash


Associated arthritis. Parvovirus B19 may also cause a symmetric polyarthritis of the hands, wrists, knees, or ankles, particularly in adult females. The course of arthritis usually lasts 1 to 3 weeks, but up to 20% may evolve into a chronic arthritis.⁶ In addition, numerous case studies suggest that parvovirus B19 may, in rare cases, cause a viral myocarditis in infants and children.⁷

Hemolytic complications. The target of parvovirus B19 is the erythroid blood cell line.¹ Consequently, immunocompromised patients and those with chronic hemolytic conditions (eg, sickle cell disease, thalassemia, spherocytosis, or pyruvate kinase deficiency) may develop hematologic complications such as aplastic crisis, chronic anemia, thrombocytopenia, neutropenia, or pancytopenia. Patients with hemolytic complications can be quite ill, presenting with fever, malaise, tachycardia, tachypnea, and profound anemia.

Parvovirus B19 may cause
a symmetric polyarthritis, particularly in adult females, usually lasting one to 3 weeks. But it can evolve into a chronic arthritis.Perinatal perils. Approximately one-third of pregnant mothers are at risk for parvovirus B19 infection, and having children at home, a severe medical condition, or stressful employment have been shown to increase their risk of active infection.⁸ The annual incidence of symptomatic parvovirus B19 during pregnancy is 1.5%, increasing to 13% during epidemics.⁹ Such infection can cause significant morbidity and mortality for the fetus. Mothers newly infected during the first trimester have experienced a 71% increased risk of intrauterine fetal demise (fetal loss <20 weeks gestation) when compared with baseline risk of fetal loss.⁹ In one prospective observational study, fetal death was only observed when mothers were infected prior to 20 weeks of gestation.¹⁰ Intrauterine B19 infection during any trimester carries a 4% overall risk of hydrops fetalis, thought to be due to high output cardiac failure secondary to severe anemia.¹⁰

Rely on clinical findings to diagnose; restrict serologic testing


The characteristic “slapped cheek” rash usually distinguishes fifth disease from other causes of febrile rash. Differential diagnosis includes measles, scarlet fever, roseola infantum, enterovirus, and adenovirus. A diagnostic tool (TABLE) can help differentiate fifth disease from other viral exanthems.

In most cases of suspected parvovirus B19 infection, serologic testing is not indicated. However, consider serologic testing for pregnant women with known exposure to the virus, immunocompromised patients, patients with chronic hemolytic conditions, or patients with severe or persistent arthropathy. Serum immunoglobulin M can usually be detected 10 days after infection and can persist for 3 months, while serum immunoglobulin G is produced 2 weeks after inoculation and presumably lasts for life.¹¹

Treat supportively

No specific treatment exists for parvovirus B19 infection. Management is supportive and the infection is usually mild and self-limiting. A nonsteroidal anti-inflammatory agent may be sufficient for associated arthritis; if needed, a low-dose oral corticosteroid can be used without prolonging the viral illness.⁶ Refer for hematologic consultation any immunocompromised patient with confirmed parvovirus who develops a hematologic complication, which may require intravenous immunoglobulin treatment or, in severe cases, bone marrow transplantation.

Clinical recommendations

Parvovirus B19 is communicable only during the nonspecific prodromal period—the 4 to 21 days of incubation in which the patient seems to have a common cold, with coryza, sore throat, and headache. With the appearance of the “slapped cheek” rash (an immune-mediated, postinfectious sequela), a child with erythema infectiosum is no longer infectious. At this stage, exclusion from school or child care is unnecessary.¹

Perform serologic testing to determine immunity for all pregnant women with documented exposure to parvovirus B19.¹² Retest women who are initially nonimmune after 3 to 4 weeks. Patients who seroconvert should undergo serial ultrasounds for 10 weeks to evaluate for hydrops fetalis or growth restriction. Repeat testing is unwarranted for those who do not seroconvert. There is no evidence to suggest that seronegative pregnant women should avoid work environments during endemic periods of infection.¹³

Sixth disease

Human herpesvirus 6 (HHV-6) causes sixth disease, also known as roseola infantum or exanthem subitum. Ninety percent of children have been infected by 2 years of age, with peak incidence occurring between 9 and 21 months of age.¹⁴ HHV-6 is most likely transmitted via the saliva of healthy individuals and enters the body via a mucosal surface. One percent of HHV-6 infection is acquired congenitally without known sequelae, similar to the transmission rate of cytomegalovirus.¹⁵

Clinical presentation: 
Only 20% may exhibit a rash


Complications. Fifteen percent of infected children have febrile seizures.¹ Based on several case reports, HHV-6 infection has been associated with meningoencephalitis, acute disseminated demyelination, hepatitis, and myocarditis.¹⁷ It is unknown whether seizures increase the risk of these complications. Long-term sequelae from these manifestations of HHV-6 infection include developmental disorders and autism-spectrum disorders.^18,19

Treat supportively

Patients with primary HHV-6 infection usually require antipyretics and frequent hydration. Reserve antivirals such as ganciclovir, foscarnet, and cidofovir for immunocompromised patients or those with HHV-6 encephalitis.²⁰

Clinical recommendations

Most patients infected with HHV-6 have fever and rhinorrhea,
 and are fussy. Only 20% exhibit the rash of roseola. Treat seizures associated with HHV-6 infection as you would any other febrile seizure, giving an antiepileptic (diazepam, lorazepam, or midazolam) if the seizure lasts >5 minutes. Risk of seizure recurrence with HHV-6 is equivalent to that seen with other causes of febrile seizure.¹

Because of the ubiquitous prevalence of HHV-6 infection, there are no effective preventive measures. Little is known about the effect of HHV-6 exposure during pregnancy because most pregnant mothers are immune to the virus.²¹ Exclusion from school or child care is not recommended because of the prolonged shedding of the virus.^16,22

CORRESPONDENCE
Jason S. O’Grady, MD, Department of Family Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905; [email protected]

ACKNOWLEDGEMENT
The author thanks Anne Mounsey, MD, Department of Family Medicine, University of North Carolina at Chapel Hill, for her invaluable assistance in editing this manuscript.

Fifth and sixth diseases are frequently encountered viral exanthems in family medicine. This article delineates the unique clinical characteristics of these disorders, describes rare but serious sequelae of each, and offers recommendations to guide your practice.

Fifth disease

Parvovirus B19, an infectious agent found worldwide, is the cause of fifth disease, also known as slapped cheek syndrome or erythema infectiosum. It is transmitted via respiratory droplets, most commonly in late winter and early spring. The peak incidence of parvovirus B19 infection is in children ages 5 to 15 years.¹ Approximately 20% of parvovirus B19 infections remain subclinical.^1,2 An observational study of children in the United Kingdom who were 6 months to 16 years of age and had been immunized for measles and rubella revealed that parvovirus B19 was the number one identifiable cause of febrile rash, responsible for 17% of cases.³ Seroprevalence increases with age, and 40% to 60% of adults test positive for prior infection.¹

Clinical presentation: Not necessarily limited to fever and rash


Associated arthritis. Parvovirus B19 may also cause a symmetric polyarthritis of the hands, wrists, knees, or ankles, particularly in adult females. The course of arthritis usually lasts 1 to 3 weeks, but up to 20% may evolve into a chronic arthritis.⁶ In addition, numerous case studies suggest that parvovirus B19 may, in rare cases, cause a viral myocarditis in infants and children.⁷

Hemolytic complications. The target of parvovirus B19 is the erythroid blood cell line.¹ Consequently, immunocompromised patients and those with chronic hemolytic conditions (eg, sickle cell disease, thalassemia, spherocytosis, or pyruvate kinase deficiency) may develop hematologic complications such as aplastic crisis, chronic anemia, thrombocytopenia, neutropenia, or pancytopenia. Patients with hemolytic complications can be quite ill, presenting with fever, malaise, tachycardia, tachypnea, and profound anemia.

Parvovirus B19 may cause
a symmetric polyarthritis, particularly in adult females, usually lasting one to 3 weeks. But it can evolve into a chronic arthritis.Perinatal perils. Approximately one-third of pregnant mothers are at risk for parvovirus B19 infection, and having children at home, a severe medical condition, or stressful employment have been shown to increase their risk of active infection.⁸ The annual incidence of symptomatic parvovirus B19 during pregnancy is 1.5%, increasing to 13% during epidemics.⁹ Such infection can cause significant morbidity and mortality for the fetus. Mothers newly infected during the first trimester have experienced a 71% increased risk of intrauterine fetal demise (fetal loss <20 weeks gestation) when compared with baseline risk of fetal loss.⁹ In one prospective observational study, fetal death was only observed when mothers were infected prior to 20 weeks of gestation.¹⁰ Intrauterine B19 infection during any trimester carries a 4% overall risk of hydrops fetalis, thought to be due to high output cardiac failure secondary to severe anemia.¹⁰

Rely on clinical findings to diagnose; restrict serologic testing


The characteristic “slapped cheek” rash usually distinguishes fifth disease from other causes of febrile rash. Differential diagnosis includes measles, scarlet fever, roseola infantum, enterovirus, and adenovirus. A diagnostic tool (TABLE) can help differentiate fifth disease from other viral exanthems.

In most cases of suspected parvovirus B19 infection, serologic testing is not indicated. However, consider serologic testing for pregnant women with known exposure to the virus, immunocompromised patients, patients with chronic hemolytic conditions, or patients with severe or persistent arthropathy. Serum immunoglobulin M can usually be detected 10 days after infection and can persist for 3 months, while serum immunoglobulin G is produced 2 weeks after inoculation and presumably lasts for life.¹¹

Treat supportively

No specific treatment exists for parvovirus B19 infection. Management is supportive and the infection is usually mild and self-limiting. A nonsteroidal anti-inflammatory agent may be sufficient for associated arthritis; if needed, a low-dose oral corticosteroid can be used without prolonging the viral illness.⁶ Refer for hematologic consultation any immunocompromised patient with confirmed parvovirus who develops a hematologic complication, which may require intravenous immunoglobulin treatment or, in severe cases, bone marrow transplantation.

Clinical recommendations

Parvovirus B19 is communicable only during the nonspecific prodromal period—the 4 to 21 days of incubation in which the patient seems to have a common cold, with coryza, sore throat, and headache. With the appearance of the “slapped cheek” rash (an immune-mediated, postinfectious sequela), a child with erythema infectiosum is no longer infectious. At this stage, exclusion from school or child care is unnecessary.¹

Perform serologic testing to determine immunity for all pregnant women with documented exposure to parvovirus B19.¹² Retest women who are initially nonimmune after 3 to 4 weeks. Patients who seroconvert should undergo serial ultrasounds for 10 weeks to evaluate for hydrops fetalis or growth restriction. Repeat testing is unwarranted for those who do not seroconvert. There is no evidence to suggest that seronegative pregnant women should avoid work environments during endemic periods of infection.¹³

Sixth disease

Human herpesvirus 6 (HHV-6) causes sixth disease, also known as roseola infantum or exanthem subitum. Ninety percent of children have been infected by 2 years of age, with peak incidence occurring between 9 and 21 months of age.¹⁴ HHV-6 is most likely transmitted via the saliva of healthy individuals and enters the body via a mucosal surface. One percent of HHV-6 infection is acquired congenitally without known sequelae, similar to the transmission rate of cytomegalovirus.¹⁵

Clinical presentation: 
Only 20% may exhibit a rash


Complications. Fifteen percent of infected children have febrile seizures.¹ Based on several case reports, HHV-6 infection has been associated with meningoencephalitis, acute disseminated demyelination, hepatitis, and myocarditis.¹⁷ It is unknown whether seizures increase the risk of these complications. Long-term sequelae from these manifestations of HHV-6 infection include developmental disorders and autism-spectrum disorders.^18,19

Treat supportively

Patients with primary HHV-6 infection usually require antipyretics and frequent hydration. Reserve antivirals such as ganciclovir, foscarnet, and cidofovir for immunocompromised patients or those with HHV-6 encephalitis.²⁰

Clinical recommendations

Most patients infected with HHV-6 have fever and rhinorrhea,
 and are fussy. Only 20% exhibit the rash of roseola. Treat seizures associated with HHV-6 infection as you would any other febrile seizure, giving an antiepileptic (diazepam, lorazepam, or midazolam) if the seizure lasts >5 minutes. Risk of seizure recurrence with HHV-6 is equivalent to that seen with other causes of febrile seizure.¹

Because of the ubiquitous prevalence of HHV-6 infection, there are no effective preventive measures. Little is known about the effect of HHV-6 exposure during pregnancy because most pregnant mothers are immune to the virus.²¹ Exclusion from school or child care is not recommended because of the prolonged shedding of the virus.^16,22

CORRESPONDENCE
Jason S. O’Grady, MD, Department of Family Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905; [email protected]

ACKNOWLEDGEMENT
The author thanks Anne Mounsey, MD, Department of Family Medicine, University of North Carolina at Chapel Hill, for her invaluable assistance in editing this manuscript.

Fifth and sixth diseases are frequently encountered viral exanthems in family medicine. This article delineates the unique clinical characteristics of these disorders, describes rare but serious sequelae of each, and offers recommendations to guide your practice.

Fifth disease

Parvovirus B19, an infectious agent found worldwide, is the cause of fifth disease, also known as slapped cheek syndrome or erythema infectiosum. It is transmitted via respiratory droplets, most commonly in late winter and early spring. The peak incidence of parvovirus B19 infection is in children ages 5 to 15 years.¹ Approximately 20% of parvovirus B19 infections remain subclinical.^1,2 An observational study of children in the United Kingdom who were 6 months to 16 years of age and had been immunized for measles and rubella revealed that parvovirus B19 was the number one identifiable cause of febrile rash, responsible for 17% of cases.³ Seroprevalence increases with age, and 40% to 60% of adults test positive for prior infection.¹

Clinical presentation: Not necessarily limited to fever and rash


Associated arthritis. Parvovirus B19 may also cause a symmetric polyarthritis of the hands, wrists, knees, or ankles, particularly in adult females. The course of arthritis usually lasts 1 to 3 weeks, but up to 20% may evolve into a chronic arthritis.⁶ In addition, numerous case studies suggest that parvovirus B19 may, in rare cases, cause a viral myocarditis in infants and children.⁷

Hemolytic complications. The target of parvovirus B19 is the erythroid blood cell line.¹ Consequently, immunocompromised patients and those with chronic hemolytic conditions (eg, sickle cell disease, thalassemia, spherocytosis, or pyruvate kinase deficiency) may develop hematologic complications such as aplastic crisis, chronic anemia, thrombocytopenia, neutropenia, or pancytopenia. Patients with hemolytic complications can be quite ill, presenting with fever, malaise, tachycardia, tachypnea, and profound anemia.

Parvovirus B19 may cause
a symmetric polyarthritis, particularly in adult females, usually lasting one to 3 weeks. But it can evolve into a chronic arthritis.Perinatal perils. Approximately one-third of pregnant mothers are at risk for parvovirus B19 infection, and having children at home, a severe medical condition, or stressful employment have been shown to increase their risk of active infection.⁸ The annual incidence of symptomatic parvovirus B19 during pregnancy is 1.5%, increasing to 13% during epidemics.⁹ Such infection can cause significant morbidity and mortality for the fetus. Mothers newly infected during the first trimester have experienced a 71% increased risk of intrauterine fetal demise (fetal loss <20 weeks gestation) when compared with baseline risk of fetal loss.⁹ In one prospective observational study, fetal death was only observed when mothers were infected prior to 20 weeks of gestation.¹⁰ Intrauterine B19 infection during any trimester carries a 4% overall risk of hydrops fetalis, thought to be due to high output cardiac failure secondary to severe anemia.¹⁰

Rely on clinical findings to diagnose; restrict serologic testing


The characteristic “slapped cheek” rash usually distinguishes fifth disease from other causes of febrile rash. Differential diagnosis includes measles, scarlet fever, roseola infantum, enterovirus, and adenovirus. A diagnostic tool (TABLE) can help differentiate fifth disease from other viral exanthems.

In most cases of suspected parvovirus B19 infection, serologic testing is not indicated. However, consider serologic testing for pregnant women with known exposure to the virus, immunocompromised patients, patients with chronic hemolytic conditions, or patients with severe or persistent arthropathy. Serum immunoglobulin M can usually be detected 10 days after infection and can persist for 3 months, while serum immunoglobulin G is produced 2 weeks after inoculation and presumably lasts for life.¹¹

Treat supportively

No specific treatment exists for parvovirus B19 infection. Management is supportive and the infection is usually mild and self-limiting. A nonsteroidal anti-inflammatory agent may be sufficient for associated arthritis; if needed, a low-dose oral corticosteroid can be used without prolonging the viral illness.⁶ Refer for hematologic consultation any immunocompromised patient with confirmed parvovirus who develops a hematologic complication, which may require intravenous immunoglobulin treatment or, in severe cases, bone marrow transplantation.

Clinical recommendations

Parvovirus B19 is communicable only during the nonspecific prodromal period—the 4 to 21 days of incubation in which the patient seems to have a common cold, with coryza, sore throat, and headache. With the appearance of the “slapped cheek” rash (an immune-mediated, postinfectious sequela), a child with erythema infectiosum is no longer infectious. At this stage, exclusion from school or child care is unnecessary.¹

Perform serologic testing to determine immunity for all pregnant women with documented exposure to parvovirus B19.¹² Retest women who are initially nonimmune after 3 to 4 weeks. Patients who seroconvert should undergo serial ultrasounds for 10 weeks to evaluate for hydrops fetalis or growth restriction. Repeat testing is unwarranted for those who do not seroconvert. There is no evidence to suggest that seronegative pregnant women should avoid work environments during endemic periods of infection.¹³

Sixth disease

Human herpesvirus 6 (HHV-6) causes sixth disease, also known as roseola infantum or exanthem subitum. Ninety percent of children have been infected by 2 years of age, with peak incidence occurring between 9 and 21 months of age.¹⁴ HHV-6 is most likely transmitted via the saliva of healthy individuals and enters the body via a mucosal surface. One percent of HHV-6 infection is acquired congenitally without known sequelae, similar to the transmission rate of cytomegalovirus.¹⁵

Clinical presentation: 
Only 20% may exhibit a rash


Complications. Fifteen percent of infected children have febrile seizures.¹ Based on several case reports, HHV-6 infection has been associated with meningoencephalitis, acute disseminated demyelination, hepatitis, and myocarditis.¹⁷ It is unknown whether seizures increase the risk of these complications. Long-term sequelae from these manifestations of HHV-6 infection include developmental disorders and autism-spectrum disorders.^18,19

Treat supportively

Patients with primary HHV-6 infection usually require antipyretics and frequent hydration. Reserve antivirals such as ganciclovir, foscarnet, and cidofovir for immunocompromised patients or those with HHV-6 encephalitis.²⁰

Clinical recommendations

Most patients infected with HHV-6 have fever and rhinorrhea,
 and are fussy. Only 20% exhibit the rash of roseola. Treat seizures associated with HHV-6 infection as you would any other febrile seizure, giving an antiepileptic (diazepam, lorazepam, or midazolam) if the seizure lasts >5 minutes. Risk of seizure recurrence with HHV-6 is equivalent to that seen with other causes of febrile seizure.¹

Because of the ubiquitous prevalence of HHV-6 infection, there are no effective preventive measures. Little is known about the effect of HHV-6 exposure during pregnancy because most pregnant mothers are immune to the virus.²¹ Exclusion from school or child care is not recommended because of the prolonged shedding of the virus.^16,22

CORRESPONDENCE
Jason S. O’Grady, MD, Department of Family Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905; [email protected]

ACKNOWLEDGEMENT
The author thanks Anne Mounsey, MD, Department of Family Medicine, University of North Carolina at Chapel Hill, for her invaluable assistance in editing this manuscript.

Most physicians do a satisfactory job in choosing when and how to do a skin biopsy, but there is always room for improvement. The 9 pointers we provide here are based on standard of care practices and literature when available, and also on our collective experiences as a pathologist/dermatologist (JM), dermatopathologist (DZ), primary care physician (BR), and dermatologist/Mohs surgeon (EB).

1. Choose your biopsy type wisely.

Using the appropriate type of biopsy can have the greatest effect on a proper diagnosis. The decision of which biopsy type to use is not always easy. The most common biopsy types are shave, punch, excisional, and curettage. Several reference articles detail each type of biopsy commonly used in primary care and how to perform them.^1,2 (For a series of how-to videos that illustrate how to perform some of these biopsies, visit The Journal of Family Practice Multimedia Library at http://www.jfponline.com/multimedia/video.html.)

Each type of biopsy has inherent advantages and disadvantages. In general, the shave biopsy is most commonly used for lesions that are solitary, elevated, and give the impression that a sufficient amount of tissue can be sampled using this technique. The punch biopsy is the biopsy of choice for most “rashes” (inflammatory skin disorders).² Excisional biopsy is used to remove melanocytic neoplasms or larger lesions. And curettage, while still used by some clinicians for melanocytic lesions because of its speed and simplicity, should almost never be used for diagnostic purposes. Each of these techniques is described in greater detail in the tips that follow.

2. When performing a shave biopsy, avoid obtaining a sample that's too superficial.

The advantage of the shave biopsy is that it is minimally invasive and quick to perform. If kept small while not compromising the amount of sample retrieved, the scars left by shave biopsies have the potential to blend well. The major disadvantage of the shave biopsy is that occasionally, if the shave is not deep enough, an insufficient amount of tissue is obtained, which can make it challenging to establish an accurate diagnosis.

An advantage 
of the punch biopsy is that patients are left with linear scars, rather than the round scars that are often associated with shave biopsy.Balancing the need to obtain adequate tissue while minimizing scarring takes skill and experience. Taking a biopsy that is inadequate is a common occurrence. At times, the physician’s clinical impression may be that a biopsy has obtained adequate tissue, but histologically only the superficial part of the skin surface has been sampled. This often is because of thickening of the superficial skin, whether as a manifestation of the anatomic site (eg, acral skin) or the disease process itself.

Unfortunately, this superficial skin often is nondiagnostic when unaccompanied by underlying epidermis and dermis. It is important to keep this in mind when obtaining a skin biopsy, especially when dealing with lesions that are very scaly or keratinized. An equivocal biopsy wastes time, energy, and money, and can negatively impact patient care.³ It can be difficult to balance practical aspects of the biopsy (ie, optimizing cosmetic outcomes, minimizing scarring and wound size) with the need to obtain sufficient tissue sampling (FIGURE 1).

3. Choose punch over shave biopsy for rashes.

In a punch biopsy, a disposable metal cylinder with a sharpened edge is used to “punch” out a piece of skin that can be examined under the microscope. Punch biopsy is the preferred technique for almost all inflammatory skin conditions (rashes) because the pathologist is able to examine both the superficial and deep portions of the dermis⁴ (FIGURE 2).

Pathologists use the pattern of inflammation, in conjunction with epidermal changes, to distinguish different types of inflammatory processes. For example, lichen planus is typically associated with superficial inflammation, while lupus is known to have prominent superficial and deep inflammation.

An inadequate punch biopsy sample can hinder histological assessment of inflammatory skin disorders that involve both the superficial and deep portions of the dermis and can make arriving at a definitive diagnosis more challenging. The diameter of a punch cylinder ranges from 1 to 8 mm. Smaller punch biopsies often create diagnostic challenges because they provide so little sample. A punch biopsy size of 4 mm is commonly used for rashes.

An advantage of the punch biopsy is that patients are left with linear scars rather than round, potentially dyspigmented (darker or lighter) scars that are often associated with shave biopsy. A well-sutured punch biopsy can be cosmetically elegant, particularly if closure is oriented along relaxed skin tension lines. For this reason, punch biopsies are well suited for cosmetically sensitive locations such as the face, although shave biopsies are also often performed on the face.

4. Choose an excisional biopsy for a melanocytic neoplasm, when possible.

The purpose of an excisional biopsy (which typically includes a 1 to 3 mm rim of normal skin around the lesion) is to completely remove a lesion. The excisional biopsy generally is the preferred technique for clinically atypical melanocytic neoplasms (lesions that are not definitively benign).^4-8

When suspicion for melanoma is high, excisional biopsies should be performed with minimal undermining to preserve the accuracy of any future sentinel lymph node biopsy surgeries. Excisional biopsy is the most involved type of biopsy and has the largest potential for cosmetic disfigurement if not properly planned and performed. While guidelines from the American Academy of Dermatology state that “narrow excisional biopsy that encompasses [the] entire breadth of lesion with clinically negative margins to ensure that the lesion is not transected” is preferred, they also acknowledge that partial sampling (incisional biopsy) is acceptable in select clinical circumstances,⁹ such as when a lesion is large or on a cosmetically sensitive site such as the face.¹⁰

While a larger punch biopsy (6 or 8 mm) or even deep shave/saucerization may function as an excisional biopsy for very small lesions, this approach can be problematic. For one thing, these biopsies are more likely than an excisional biopsy to leave a portion of the lesion in situ. Another concern is that a shave biopsy of a melanocytic lesion can lead to error or difficulty in obtaining the correct diagnosis on later biopsy.¹¹ For pathologists, smaller or incomplete samples make it challenging to establish an accurate diagnosis.¹² Among melanomas seen at a tertiary referral center, histopathological misdiagnosis was more common with a punch or shave biopsy than with an excisional biopsy.⁹

It has been shown that partial biopsy for melanoma results in more residual disease at wide local excision and makes it more challenging to properly stage the lesion.^13,14 If a shave biopsy is used to sample a suspected melanocytic neoplasm, it is imperative to document the specific site of the biopsy, indicate the size of the melanocytic lesion on the pathology requisition form, and ensure that all (or nearly all) of the clinically evident lesion is sampled. Detailing the location of the lesion in the chart is not only essential in evaluating the present lesion, but it will serve you well in the future. Without knowing the patient’s clinical history, benign nevi that recur after a prior biopsy can be difficult to histologically distinguish from melanoma (FIGURE 3). For more on this, see tip #7.

5. Be careful with curettage.

6. Remember the importance of proper fixation and processing.

As obvious as it may sound, it is important to remember to promptly place sampled tissue in an adequate amount of formalin so that the tissue is submersed in it in the container.¹⁵ Failure to do so can result in improper fixation and will make it difficult to render an appropriate diagnosis. Conventionally, a 10:1 formalin volume to tissue volume ratio is recommended. If the “cold time”—the amount of time a tissue sample is out of formalin—is long enough (greater than a few hours), an appropriate assessment can be impossible.

Appropriate fixation and fixation times are important because molecular testing is being increasingly used to make pathological diagnoses.¹⁶ Additionally, aggressively manipulating a biopsy sample while extracting it or placing it in formalin can cause “crush” artifact, which can limit interpretability (FIGURE 5).

7. Properly photograph and document the biopsy location.

When performing a biopsy of a suspicious neoplasm, physicians often remove all of the lesion’s superficial components, which means that at the patient’s follow-up appointment and subsequent treatments, only a well-healed biopsy site will remain. The biopsy site may be so well healed that it blends seamlessly into the surrounding skin and is nearly impossible for the physician to identify. This problem is seen most often when patients present for surgical excision or Mohs micrographic surgery.¹⁷

To properly record the site of a biopsy for future dermatologic exams, take pictures of the lesion at the time of biopsy. The photographs should clearly document the lesion in question, and should be taken far enough from the site that surrounding lesions and/ or other anatomic landmarks are also visible. Biangulation or triangulation (taking a series of 2 or 3 measurements, respectively, from the site of the lesion to nearby anatomic landmarks) can be used in conjunction with photographs.

When using measurements, be as specific and accurate as possible with anatomic terms. For example, measuring the distance from the “ear” is not helpful. It would be more helpful to measure the distance from the “tragus” or the “root of the helix.” Without a properly photographed and documented biopsy site, surgical treatment may need to be delayed until the location can be confirmed.

8. Give the pathologist a pertinent history.

Providing the pathologist with a sufficient history, including the distribution and appearance of the lesion, and how long the patient has had it, can be essential in narrowing the diagnosis or making the differential diagnoses. Things like medication use or new exposures to perfumes, lotions, or plants can be especially helpful and are often overlooked when filling out the pathology requisition form.

When warranted, phone calls are helpful. You might, for example, call the pathologist and give him or her a more detailed physical examination description or additional pertinent history that was discovered after the requisition was filled out. Providing a good history can make the difference between a specific diagnosis and a broad differential.

9. Know when to refer.


There is no shame in asking for a second opinion when it comes to evaluating a skin issue, especially in regards to melanocytic neoplasms, where the stakes can be high, or skin eruptions that do not respond to conventional therapy. Remember, many cases are difficult, even for experts, and require a careful balance of clinical and histopathological judgment.¹⁸

CORRESPONDENCE
Jayson Miedema, MD, Department of Dermatology, University of North Carolina at Chapel Hill, 410 Market Street, Suite 400, Chapel Hill, NC 27516; [email protected]

Most physicians do a satisfactory job in choosing when and how to do a skin biopsy, but there is always room for improvement. The 9 pointers we provide here are based on standard of care practices and literature when available, and also on our collective experiences as a pathologist/dermatologist (JM), dermatopathologist (DZ), primary care physician (BR), and dermatologist/Mohs surgeon (EB).

1. Choose your biopsy type wisely.

Using the appropriate type of biopsy can have the greatest effect on a proper diagnosis. The decision of which biopsy type to use is not always easy. The most common biopsy types are shave, punch, excisional, and curettage. Several reference articles detail each type of biopsy commonly used in primary care and how to perform them.^1,2 (For a series of how-to videos that illustrate how to perform some of these biopsies, visit The Journal of Family Practice Multimedia Library at http://www.jfponline.com/multimedia/video.html.)

Each type of biopsy has inherent advantages and disadvantages. In general, the shave biopsy is most commonly used for lesions that are solitary, elevated, and give the impression that a sufficient amount of tissue can be sampled using this technique. The punch biopsy is the biopsy of choice for most “rashes” (inflammatory skin disorders).² Excisional biopsy is used to remove melanocytic neoplasms or larger lesions. And curettage, while still used by some clinicians for melanocytic lesions because of its speed and simplicity, should almost never be used for diagnostic purposes. Each of these techniques is described in greater detail in the tips that follow.

2. When performing a shave biopsy, avoid obtaining a sample that's too superficial.

The advantage of the shave biopsy is that it is minimally invasive and quick to perform. If kept small while not compromising the amount of sample retrieved, the scars left by shave biopsies have the potential to blend well. The major disadvantage of the shave biopsy is that occasionally, if the shave is not deep enough, an insufficient amount of tissue is obtained, which can make it challenging to establish an accurate diagnosis.

An advantage 
of the punch biopsy is that patients are left with linear scars, rather than the round scars that are often associated with shave biopsy.Balancing the need to obtain adequate tissue while minimizing scarring takes skill and experience. Taking a biopsy that is inadequate is a common occurrence. At times, the physician’s clinical impression may be that a biopsy has obtained adequate tissue, but histologically only the superficial part of the skin surface has been sampled. This often is because of thickening of the superficial skin, whether as a manifestation of the anatomic site (eg, acral skin) or the disease process itself.

Unfortunately, this superficial skin often is nondiagnostic when unaccompanied by underlying epidermis and dermis. It is important to keep this in mind when obtaining a skin biopsy, especially when dealing with lesions that are very scaly or keratinized. An equivocal biopsy wastes time, energy, and money, and can negatively impact patient care.³ It can be difficult to balance practical aspects of the biopsy (ie, optimizing cosmetic outcomes, minimizing scarring and wound size) with the need to obtain sufficient tissue sampling (FIGURE 1).

3. Choose punch over shave biopsy for rashes.

In a punch biopsy, a disposable metal cylinder with a sharpened edge is used to “punch” out a piece of skin that can be examined under the microscope. Punch biopsy is the preferred technique for almost all inflammatory skin conditions (rashes) because the pathologist is able to examine both the superficial and deep portions of the dermis⁴ (FIGURE 2).

Pathologists use the pattern of inflammation, in conjunction with epidermal changes, to distinguish different types of inflammatory processes. For example, lichen planus is typically associated with superficial inflammation, while lupus is known to have prominent superficial and deep inflammation.

An inadequate punch biopsy sample can hinder histological assessment of inflammatory skin disorders that involve both the superficial and deep portions of the dermis and can make arriving at a definitive diagnosis more challenging. The diameter of a punch cylinder ranges from 1 to 8 mm. Smaller punch biopsies often create diagnostic challenges because they provide so little sample. A punch biopsy size of 4 mm is commonly used for rashes.

An advantage of the punch biopsy is that patients are left with linear scars rather than round, potentially dyspigmented (darker or lighter) scars that are often associated with shave biopsy. A well-sutured punch biopsy can be cosmetically elegant, particularly if closure is oriented along relaxed skin tension lines. For this reason, punch biopsies are well suited for cosmetically sensitive locations such as the face, although shave biopsies are also often performed on the face.

4. Choose an excisional biopsy for a melanocytic neoplasm, when possible.

The purpose of an excisional biopsy (which typically includes a 1 to 3 mm rim of normal skin around the lesion) is to completely remove a lesion. The excisional biopsy generally is the preferred technique for clinically atypical melanocytic neoplasms (lesions that are not definitively benign).^4-8

When suspicion for melanoma is high, excisional biopsies should be performed with minimal undermining to preserve the accuracy of any future sentinel lymph node biopsy surgeries. Excisional biopsy is the most involved type of biopsy and has the largest potential for cosmetic disfigurement if not properly planned and performed. While guidelines from the American Academy of Dermatology state that “narrow excisional biopsy that encompasses [the] entire breadth of lesion with clinically negative margins to ensure that the lesion is not transected” is preferred, they also acknowledge that partial sampling (incisional biopsy) is acceptable in select clinical circumstances,⁹ such as when a lesion is large or on a cosmetically sensitive site such as the face.¹⁰

While a larger punch biopsy (6 or 8 mm) or even deep shave/saucerization may function as an excisional biopsy for very small lesions, this approach can be problematic. For one thing, these biopsies are more likely than an excisional biopsy to leave a portion of the lesion in situ. Another concern is that a shave biopsy of a melanocytic lesion can lead to error or difficulty in obtaining the correct diagnosis on later biopsy.¹¹ For pathologists, smaller or incomplete samples make it challenging to establish an accurate diagnosis.¹² Among melanomas seen at a tertiary referral center, histopathological misdiagnosis was more common with a punch or shave biopsy than with an excisional biopsy.⁹

It has been shown that partial biopsy for melanoma results in more residual disease at wide local excision and makes it more challenging to properly stage the lesion.^13,14 If a shave biopsy is used to sample a suspected melanocytic neoplasm, it is imperative to document the specific site of the biopsy, indicate the size of the melanocytic lesion on the pathology requisition form, and ensure that all (or nearly all) of the clinically evident lesion is sampled. Detailing the location of the lesion in the chart is not only essential in evaluating the present lesion, but it will serve you well in the future. Without knowing the patient’s clinical history, benign nevi that recur after a prior biopsy can be difficult to histologically distinguish from melanoma (FIGURE 3). For more on this, see tip #7.

5. Be careful with curettage.

6. Remember the importance of proper fixation and processing.

As obvious as it may sound, it is important to remember to promptly place sampled tissue in an adequate amount of formalin so that the tissue is submersed in it in the container.¹⁵ Failure to do so can result in improper fixation and will make it difficult to render an appropriate diagnosis. Conventionally, a 10:1 formalin volume to tissue volume ratio is recommended. If the “cold time”—the amount of time a tissue sample is out of formalin—is long enough (greater than a few hours), an appropriate assessment can be impossible.

Appropriate fixation and fixation times are important because molecular testing is being increasingly used to make pathological diagnoses.¹⁶ Additionally, aggressively manipulating a biopsy sample while extracting it or placing it in formalin can cause “crush” artifact, which can limit interpretability (FIGURE 5).

7. Properly photograph and document the biopsy location.

When performing a biopsy of a suspicious neoplasm, physicians often remove all of the lesion’s superficial components, which means that at the patient’s follow-up appointment and subsequent treatments, only a well-healed biopsy site will remain. The biopsy site may be so well healed that it blends seamlessly into the surrounding skin and is nearly impossible for the physician to identify. This problem is seen most often when patients present for surgical excision or Mohs micrographic surgery.¹⁷

To properly record the site of a biopsy for future dermatologic exams, take pictures of the lesion at the time of biopsy. The photographs should clearly document the lesion in question, and should be taken far enough from the site that surrounding lesions and/ or other anatomic landmarks are also visible. Biangulation or triangulation (taking a series of 2 or 3 measurements, respectively, from the site of the lesion to nearby anatomic landmarks) can be used in conjunction with photographs.

When using measurements, be as specific and accurate as possible with anatomic terms. For example, measuring the distance from the “ear” is not helpful. It would be more helpful to measure the distance from the “tragus” or the “root of the helix.” Without a properly photographed and documented biopsy site, surgical treatment may need to be delayed until the location can be confirmed.

8. Give the pathologist a pertinent history.

Providing the pathologist with a sufficient history, including the distribution and appearance of the lesion, and how long the patient has had it, can be essential in narrowing the diagnosis or making the differential diagnoses. Things like medication use or new exposures to perfumes, lotions, or plants can be especially helpful and are often overlooked when filling out the pathology requisition form.

When warranted, phone calls are helpful. You might, for example, call the pathologist and give him or her a more detailed physical examination description or additional pertinent history that was discovered after the requisition was filled out. Providing a good history can make the difference between a specific diagnosis and a broad differential.

9. Know when to refer.


There is no shame in asking for a second opinion when it comes to evaluating a skin issue, especially in regards to melanocytic neoplasms, where the stakes can be high, or skin eruptions that do not respond to conventional therapy. Remember, many cases are difficult, even for experts, and require a careful balance of clinical and histopathological judgment.¹⁸

CORRESPONDENCE
Jayson Miedema, MD, Department of Dermatology, University of North Carolina at Chapel Hill, 410 Market Street, Suite 400, Chapel Hill, NC 27516; [email protected]

Most physicians do a satisfactory job in choosing when and how to do a skin biopsy, but there is always room for improvement. The 9 pointers we provide here are based on standard of care practices and literature when available, and also on our collective experiences as a pathologist/dermatologist (JM), dermatopathologist (DZ), primary care physician (BR), and dermatologist/Mohs surgeon (EB).

1. Choose your biopsy type wisely.

Using the appropriate type of biopsy can have the greatest effect on a proper diagnosis. The decision of which biopsy type to use is not always easy. The most common biopsy types are shave, punch, excisional, and curettage. Several reference articles detail each type of biopsy commonly used in primary care and how to perform them.^1,2 (For a series of how-to videos that illustrate how to perform some of these biopsies, visit The Journal of Family Practice Multimedia Library at http://www.jfponline.com/multimedia/video.html.)

Each type of biopsy has inherent advantages and disadvantages. In general, the shave biopsy is most commonly used for lesions that are solitary, elevated, and give the impression that a sufficient amount of tissue can be sampled using this technique. The punch biopsy is the biopsy of choice for most “rashes” (inflammatory skin disorders).² Excisional biopsy is used to remove melanocytic neoplasms or larger lesions. And curettage, while still used by some clinicians for melanocytic lesions because of its speed and simplicity, should almost never be used for diagnostic purposes. Each of these techniques is described in greater detail in the tips that follow.

2. When performing a shave biopsy, avoid obtaining a sample that's too superficial.

The advantage of the shave biopsy is that it is minimally invasive and quick to perform. If kept small while not compromising the amount of sample retrieved, the scars left by shave biopsies have the potential to blend well. The major disadvantage of the shave biopsy is that occasionally, if the shave is not deep enough, an insufficient amount of tissue is obtained, which can make it challenging to establish an accurate diagnosis.

An advantage 
of the punch biopsy is that patients are left with linear scars, rather than the round scars that are often associated with shave biopsy.Balancing the need to obtain adequate tissue while minimizing scarring takes skill and experience. Taking a biopsy that is inadequate is a common occurrence. At times, the physician’s clinical impression may be that a biopsy has obtained adequate tissue, but histologically only the superficial part of the skin surface has been sampled. This often is because of thickening of the superficial skin, whether as a manifestation of the anatomic site (eg, acral skin) or the disease process itself.

Unfortunately, this superficial skin often is nondiagnostic when unaccompanied by underlying epidermis and dermis. It is important to keep this in mind when obtaining a skin biopsy, especially when dealing with lesions that are very scaly or keratinized. An equivocal biopsy wastes time, energy, and money, and can negatively impact patient care.³ It can be difficult to balance practical aspects of the biopsy (ie, optimizing cosmetic outcomes, minimizing scarring and wound size) with the need to obtain sufficient tissue sampling (FIGURE 1).

3. Choose punch over shave biopsy for rashes.

In a punch biopsy, a disposable metal cylinder with a sharpened edge is used to “punch” out a piece of skin that can be examined under the microscope. Punch biopsy is the preferred technique for almost all inflammatory skin conditions (rashes) because the pathologist is able to examine both the superficial and deep portions of the dermis⁴ (FIGURE 2).

Pathologists use the pattern of inflammation, in conjunction with epidermal changes, to distinguish different types of inflammatory processes. For example, lichen planus is typically associated with superficial inflammation, while lupus is known to have prominent superficial and deep inflammation.

An inadequate punch biopsy sample can hinder histological assessment of inflammatory skin disorders that involve both the superficial and deep portions of the dermis and can make arriving at a definitive diagnosis more challenging. The diameter of a punch cylinder ranges from 1 to 8 mm. Smaller punch biopsies often create diagnostic challenges because they provide so little sample. A punch biopsy size of 4 mm is commonly used for rashes.

An advantage of the punch biopsy is that patients are left with linear scars rather than round, potentially dyspigmented (darker or lighter) scars that are often associated with shave biopsy. A well-sutured punch biopsy can be cosmetically elegant, particularly if closure is oriented along relaxed skin tension lines. For this reason, punch biopsies are well suited for cosmetically sensitive locations such as the face, although shave biopsies are also often performed on the face.

4. Choose an excisional biopsy for a melanocytic neoplasm, when possible.

The purpose of an excisional biopsy (which typically includes a 1 to 3 mm rim of normal skin around the lesion) is to completely remove a lesion. The excisional biopsy generally is the preferred technique for clinically atypical melanocytic neoplasms (lesions that are not definitively benign).^4-8

When suspicion for melanoma is high, excisional biopsies should be performed with minimal undermining to preserve the accuracy of any future sentinel lymph node biopsy surgeries. Excisional biopsy is the most involved type of biopsy and has the largest potential for cosmetic disfigurement if not properly planned and performed. While guidelines from the American Academy of Dermatology state that “narrow excisional biopsy that encompasses [the] entire breadth of lesion with clinically negative margins to ensure that the lesion is not transected” is preferred, they also acknowledge that partial sampling (incisional biopsy) is acceptable in select clinical circumstances,⁹ such as when a lesion is large or on a cosmetically sensitive site such as the face.¹⁰

While a larger punch biopsy (6 or 8 mm) or even deep shave/saucerization may function as an excisional biopsy for very small lesions, this approach can be problematic. For one thing, these biopsies are more likely than an excisional biopsy to leave a portion of the lesion in situ. Another concern is that a shave biopsy of a melanocytic lesion can lead to error or difficulty in obtaining the correct diagnosis on later biopsy.¹¹ For pathologists, smaller or incomplete samples make it challenging to establish an accurate diagnosis.¹² Among melanomas seen at a tertiary referral center, histopathological misdiagnosis was more common with a punch or shave biopsy than with an excisional biopsy.⁹

It has been shown that partial biopsy for melanoma results in more residual disease at wide local excision and makes it more challenging to properly stage the lesion.^13,14 If a shave biopsy is used to sample a suspected melanocytic neoplasm, it is imperative to document the specific site of the biopsy, indicate the size of the melanocytic lesion on the pathology requisition form, and ensure that all (or nearly all) of the clinically evident lesion is sampled. Detailing the location of the lesion in the chart is not only essential in evaluating the present lesion, but it will serve you well in the future. Without knowing the patient’s clinical history, benign nevi that recur after a prior biopsy can be difficult to histologically distinguish from melanoma (FIGURE 3). For more on this, see tip #7.

5. Be careful with curettage.

6. Remember the importance of proper fixation and processing.

As obvious as it may sound, it is important to remember to promptly place sampled tissue in an adequate amount of formalin so that the tissue is submersed in it in the container.¹⁵ Failure to do so can result in improper fixation and will make it difficult to render an appropriate diagnosis. Conventionally, a 10:1 formalin volume to tissue volume ratio is recommended. If the “cold time”—the amount of time a tissue sample is out of formalin—is long enough (greater than a few hours), an appropriate assessment can be impossible.

Appropriate fixation and fixation times are important because molecular testing is being increasingly used to make pathological diagnoses.¹⁶ Additionally, aggressively manipulating a biopsy sample while extracting it or placing it in formalin can cause “crush” artifact, which can limit interpretability (FIGURE 5).

7. Properly photograph and document the biopsy location.

When performing a biopsy of a suspicious neoplasm, physicians often remove all of the lesion’s superficial components, which means that at the patient’s follow-up appointment and subsequent treatments, only a well-healed biopsy site will remain. The biopsy site may be so well healed that it blends seamlessly into the surrounding skin and is nearly impossible for the physician to identify. This problem is seen most often when patients present for surgical excision or Mohs micrographic surgery.¹⁷

To properly record the site of a biopsy for future dermatologic exams, take pictures of the lesion at the time of biopsy. The photographs should clearly document the lesion in question, and should be taken far enough from the site that surrounding lesions and/ or other anatomic landmarks are also visible. Biangulation or triangulation (taking a series of 2 or 3 measurements, respectively, from the site of the lesion to nearby anatomic landmarks) can be used in conjunction with photographs.

When using measurements, be as specific and accurate as possible with anatomic terms. For example, measuring the distance from the “ear” is not helpful. It would be more helpful to measure the distance from the “tragus” or the “root of the helix.” Without a properly photographed and documented biopsy site, surgical treatment may need to be delayed until the location can be confirmed.

8. Give the pathologist a pertinent history.

Providing the pathologist with a sufficient history, including the distribution and appearance of the lesion, and how long the patient has had it, can be essential in narrowing the diagnosis or making the differential diagnoses. Things like medication use or new exposures to perfumes, lotions, or plants can be especially helpful and are often overlooked when filling out the pathology requisition form.

When warranted, phone calls are helpful. You might, for example, call the pathologist and give him or her a more detailed physical examination description or additional pertinent history that was discovered after the requisition was filled out. Providing a good history can make the difference between a specific diagnosis and a broad differential.

9. Know when to refer.


There is no shame in asking for a second opinion when it comes to evaluating a skin issue, especially in regards to melanocytic neoplasms, where the stakes can be high, or skin eruptions that do not respond to conventional therapy. Remember, many cases are difficult, even for experts, and require a careful balance of clinical and histopathological judgment.¹⁸

CORRESPONDENCE
Jayson Miedema, MD, Department of Dermatology, University of North Carolina at Chapel Hill, 410 Market Street, Suite 400, Chapel Hill, NC 27516; [email protected]

CASE › Bob F, age 54, seeks care for chronic low back pain. The conservative treatments you have prescribed, including physical therapy, regular exercise, and an over-the-counter nonsteroidal anti-inflammatory drug, have provided minimal pain relief. Mr. F is reluctant to take a prescription pain medication and has expressed interest in trying a complementary and alternative medicine (CAM) therapy, such as acupuncture or yoga. What should you tell him?

Almost 40% of Americans use CAM modalities to treat specific conditions or for overall well-being,¹ and these practices are increasingly becoming a part of our approach to health care, as evidenced by the nearly 50 of facilities across the country that boast integrative health care programs, which combine CAM modalities with conventional medicine.² Emerging evidence suggests several integrative practices may offer health benefits, and primary care physicians must become well-versed in these modalities to effectively communicate potential benefits and harms to patients. In this article, we present evidence from Cochrane reviews and other studies of 8 commonly used CAM therapies, including dietary interventions, a psychotherapeutic modality, and other treatments (TABLE).^1,3-30 And while motivational interviewing technically is not a form of CAM, we also review this modality, which has proven useful in the treatment of patients for substance use. (See “Motivational interviewing for substance abuse.”)

Fish oil for hypertriglyceridemia

High triglyceride levels are a risk factor for cardiovascular disease and a component of metabolic syndrome.⁸ A 2008 review of 47 randomized controlled trials (RCTs) that included 16,511 participants found that omega-3 fatty acid (fish oil) supplements significantly reduced triglyceride levels compared to placebo.⁷ The American Heart Association recommends 2 to 4 g/d of eicosapentaenoic acid (EPA) plus docosahexaenoic acid (DHA) to lower triglyceride levels.⁸

Most studies have found that fish oil supplements are associated with few adverse effects; gastrointestinal (GI) complaints are most common. However, these supplements should be discontinued following an acute bleeding event, such as hemorrhagic stroke, due to their anticoagulant properties.⁹ Some evidence suggests that the risk for prostate cancer is increased in men with high blood levels of omega-3 fatty acids.¹⁰

Glucosamine for osteoarthritis

The AHA recommends 2 to 4 g/d 
of EPA plus DHA to lower triglyceride levels.Glucosamine is an amino sugar that is a building block of cartilage proteoglycans. Although it occurs naturally in the body, the glucosamine used in supplements is typically harvested from seashells. Glucosamine stimulates the metabolism of synovial cells and chondrocytes in articular cartilage and may delay joint degeneration.^31,32

Glucosamine is widely used in the United States as a dietary supplement, most often as glucosamine sulfate but also as n-acetyl glucosamine and glucosamine hydrochloride, although there is limited evidence of effectiveness for the latter formulations.³³

Most studies have examined the effects of oral glucosamine sulfate, 500 mg taken 3 times a day for 30 to 90 days. Once-a-day dosing as high as 1500 mg also has been used.

A Cochrane review of 25 studies with 4963 patients concluded that oral glucosamine sulfate may reduce osteoarthritis (OA) pain and improve functionality, without many adverse effects.¹¹ A 2-year double-blind RCT compared the effects of glucosamine hydrochloride 500 mg tid, chondroitin sulfate 400 mg tid, glucosamine plus chondroitin, celecoxib 200 mg/d, or placebo in 662 patients with knee OA.¹² While all groups experienced early and sustained symptomatic relief, the odds of achieving a 20% reduction in pain and improved functioning were highest with celecoxib and glucosamine.

Oral glucosamine sulfate can cause mild GI effects, but drowsiness, skin reactions, and headache also have been reported. Shellfish allergy also is a concern; however, shellfish allergies occur due to the proteins in the meat, and not from the shell from which glucosamine is derived. Glucosamine may increase glucose levels and the anticoagulant effects of warfarin.¹³

Probiotics to prevent antibiotic-associated diarrhea

Antibiotic-associated diarrhea (AAD) is a common problem.²¹ Probiotics—microorganisms found in oral supplements, yogurt, and other food—are commonly used to help maintain the balance of intestinal flora.³⁴ A recent Cochrane review of 16 RCTs that included approximately 3400 patients found evidence that probiotics can prevent AAD.²² A 2012 systematic review and meta-analysis of 63 RCTs with more than 11,000 participants concluded that probiotics lowered the relative risk of developing diarrhea compared to control groups.³⁵ The American Academy of Pediatrics supports the use of probiotics, citing results from a meta-analysis that found probiotics reduced the risk of developing diarrhea from 28.5% to 11.9% compared to placebo.³⁶

Exact dosages for probiotics have not been established, and recommendations range from 5 billion to 40 billion colony-forming units/d.²² The most commonly used probiotics are from the Lactobacillus and Saccharomyces genera; relatively little evidence supports other genera.^21,22,35,36

Probiotics are considered relatively safe, but are not recommended for patients who are immunocompromised or have an indwelling medical device.²³ Adverse effects are rare, but may include flatulence, vomiting, rash, chest pain, and increased phlegm.²¹

For a review of the latest evidence on using probiotics to reduce crying in infants with colic, see "Probiotics for colic? A PURL update."

Soy for hyperlipidemia

Soybeans are a species of legume that contain significant amounts of protein, fiber, potassium, and iron. Although soy has been used to prevent or treat cancer, osteoporosis, and menopausal symptoms, current evidence is unfavorable or inconclusive for such conditions. Some RCTs have found soy has small, favorable effects on serum levels of low-density lipoprotein and total cholesterol,²⁴ while others have shown modest improvements in triglyceride levels without significant improvements in other lipid levels.²⁵

A 2011 meta-analysis of 10 RCTs that included 268 participants found that a diet high in non-soy legume products, such as alfalfa, lentils, and other beans, also improved lipid levels.³⁷ A review of 136 studies that described 22 dietary interventions concluded that among other helpful dietary approaches to controlling hyperlipidemia, dietary soy—which contains fiber and polyunsaturated fats—is favored over supplementation of soy protein alone.³⁸

Use caution when recommending soy for patients with thyroid dysfunction or hormone-sensitive cancers because some evidence suggests soy may interfere with absorption of levothyroxine and increase the risk of developing clinical symptoms of hypothyroidism.³⁹

Soy also contains phytoestrogens, and prolonged use of soy supplements may increase the risk of endometrial hyperplasia.²⁴ This risk has been documented only in the use of soy supplements, and not from dietary soy. GI disturbances and rare allergic reactions also have been reported.²⁴

St. John’s wort for depression

Hypericum perforatum (St. John’s wort), a perennial herb, has been used to treat mood disorders and other ailments for more than 2000 years.^40,41 Commercial preparations typically are alcohol extracts with an herb-to-extract ratio of 4:1 to 8:1.²⁶ The normal dose ranges from 900 to 1500 mg/d in 2 to 3 divided doses of the alcohol extract standardized to 0.3% hypericin and/or 3% to 5% hyperforin.

St. John’s wort has been studied extensively as a treatment for depressive disorders. A 2001 double-blind RCT conducted in 11 US academic medical centers and community clinics between 1998 and 2000 that included 200 patients found that St. John’s wort was not effective for moderately severe major depression; a trend toward a positive effect was noted in both the placebo and St. John’s wort groups.²⁶

However, a 2009 Cochrane review of 29 international studies (5489 patients) concluded that St. John’s wort may be better than placebo and as effective as antidepressants for mild to moderate major depression,²⁷ and appeared to have fewer side effects than antidepressants. This review, conducted in German-speaking countries where medical professionals have long prescribed St. John’s wort, reported more positive results than those conducted in other countries.

St. John’s wort interacts with many medications, including antidepressants, oral contraceptives, cyclosporine, digoxin, indinavir, phenytoin, phenobarbital, warfarin, and others. It induces cytochrome P450 (CYP450) enzymes, and therefore can potentially reduce the efficacy of any medication that is metabolized by a CYP450 enzyme. When used in high doses in combination with antidepressants, St. John’s wort may cause serotonin syndrome. Other side effects include photosensitivity, GI complaints, fatigue, and increased risk of cataracts. Due to a lack of clinical data, St. John’s wort is contraindicated in women who are pregnant or breastfeeding.⁴²

Guided imagery for anxiety 
and pain


Guided imagery is a relaxation technique that involves visualizing positive outcomes to reduce one’s reaction to anxiety-provoking or painful experiences.⁴³ It can be practiced independently or under the direction of an instructor. One RCT of 96 women with newly diagnosed breast cancer found that adding relaxation and guided imagery to standard breast cancer treatment protocols positively affected mood and quality of life.¹⁴ While this study saw no change in pathologic responses to chemotherapy,¹⁴ a more recent RCT concluded that such biochemical advantages may be possible.⁴⁴ Guided imagery has been linked to decreased anxiety in diverse studies of students, women in labor, individuals suffering from nightmares, and in occupational settings such as training for pilots and surgeons.^15-18

A review of 9 RCTs of guided imagery for decreasing musculoskeletal pain involving 201 patients found 8 studies reported positive results, though the methodological quality of the studies was low.¹⁹ Of 6 high-quality studies included in a 2012 systematic review, 5 supported the use of guided imagery for postoperative, abdominal, and other nonmusculoskeletal pain.²⁰ This initial evidence is promising, but additional research of high methodological quality is needed to validate the use of guided imagery for anxiety and pain.

Acupuncture for pain

From 2002 to 2007, the use of acupuncture significantly increased in the United States, primarily for the treatment of pain.¹ A 2012 meta-analysis of 29 RCTs that included almost 18,000 participants evaluated the clinical usefulness of acupuncture for back, neck, and shoulder pain, OA, and headache.³ Compared to no treatment, both acupuncture and sham acupuncture significantly improved pain scores. The authors of this meta-analysis found that acupuncture offered a small but significant advantage over sham acupuncture, and concluded that the benefits of acupuncture were not due to a placebo effect.

In 2007, the American College of Physicians (ACP) and the American Pain Society (APS) issued a joint statement indicating that acupuncture should be considered for patients with chronic low back pain who do not respond to conventional therapies.⁴ The North American Spine Society also supports acupuncture, stating that it provides “...better short-term pain relief and functional improvement than no treatment and the addition of acupuncture to other treatment modalities provides a greater benefit than those treatments alone.”⁵ Additional evidence found acupuncture for chronic low back pain improves function and serves as an adjunct therapy.⁶

Reported adverse effects of acupuncture include—but may not be limited to—infection, skin irritation, hematoma, pneumothorax, and spontaneous needle migration.^1,3-6

Yoga for low back pain

An ACP/APS statement indicates that acupuncture should be considered for patients with chronic back pain who don't respond to conventional therapies.Back pain is the most common reason patients use CAM therapies.¹ A systematic review of 10 RCTs that included 967 participants with chronic low back pain found strong evidence for the short-term effectiveness and moderate evidence for the long-term effectiveness of yoga.²⁸ A review of 17 studies that included 1626 patients concluded that yoga improves both pain and functionality; this review did not recommend a specific type of yoga practice.²⁹ In a recent study of 95 minority adults with moderate-to-severe chronic low back pain, once-weekly and twice-weekly yoga for 12 weeks were similarly effective for reducing pain and improving functionality.³⁰

Guidelines from the ACP and the APS recommend yoga as part of an intensive interdisciplinary rehabilitation program for patients with chronic or subacute low back pain who do not improve using other self-care options.⁴ This recommendation is specifically for Viniyoga, a practice in which the instructor recommends modifications to body positioning for each individual based on past injuries and overall physical condition. (For more information on therapeutic uses of yoga, see “Yoga as therapy: When is it helpful?”)

CORRESPONDENCE
Roger Zoorob, MD, MPH, FAAFP; Department of Family and Community Medicine, Baylor College of Medicine, 3701 Kirby Drive, Suite 600, Houston, TX 77098; [email protected]

CASE › Bob F, age 54, seeks care for chronic low back pain. The conservative treatments you have prescribed, including physical therapy, regular exercise, and an over-the-counter nonsteroidal anti-inflammatory drug, have provided minimal pain relief. Mr. F is reluctant to take a prescription pain medication and has expressed interest in trying a complementary and alternative medicine (CAM) therapy, such as acupuncture or yoga. What should you tell him?

Almost 40% of Americans use CAM modalities to treat specific conditions or for overall well-being,¹ and these practices are increasingly becoming a part of our approach to health care, as evidenced by the nearly 50 of facilities across the country that boast integrative health care programs, which combine CAM modalities with conventional medicine.² Emerging evidence suggests several integrative practices may offer health benefits, and primary care physicians must become well-versed in these modalities to effectively communicate potential benefits and harms to patients. In this article, we present evidence from Cochrane reviews and other studies of 8 commonly used CAM therapies, including dietary interventions, a psychotherapeutic modality, and other treatments (TABLE).^1,3-30 And while motivational interviewing technically is not a form of CAM, we also review this modality, which has proven useful in the treatment of patients for substance use. (See “Motivational interviewing for substance abuse.”)

Fish oil for hypertriglyceridemia

High triglyceride levels are a risk factor for cardiovascular disease and a component of metabolic syndrome.⁸ A 2008 review of 47 randomized controlled trials (RCTs) that included 16,511 participants found that omega-3 fatty acid (fish oil) supplements significantly reduced triglyceride levels compared to placebo.⁷ The American Heart Association recommends 2 to 4 g/d of eicosapentaenoic acid (EPA) plus docosahexaenoic acid (DHA) to lower triglyceride levels.⁸

Most studies have found that fish oil supplements are associated with few adverse effects; gastrointestinal (GI) complaints are most common. However, these supplements should be discontinued following an acute bleeding event, such as hemorrhagic stroke, due to their anticoagulant properties.⁹ Some evidence suggests that the risk for prostate cancer is increased in men with high blood levels of omega-3 fatty acids.¹⁰

Glucosamine for osteoarthritis

The AHA recommends 2 to 4 g/d 
of EPA plus DHA to lower triglyceride levels.Glucosamine is an amino sugar that is a building block of cartilage proteoglycans. Although it occurs naturally in the body, the glucosamine used in supplements is typically harvested from seashells. Glucosamine stimulates the metabolism of synovial cells and chondrocytes in articular cartilage and may delay joint degeneration.^31,32

Glucosamine is widely used in the United States as a dietary supplement, most often as glucosamine sulfate but also as n-acetyl glucosamine and glucosamine hydrochloride, although there is limited evidence of effectiveness for the latter formulations.³³

Most studies have examined the effects of oral glucosamine sulfate, 500 mg taken 3 times a day for 30 to 90 days. Once-a-day dosing as high as 1500 mg also has been used.

A Cochrane review of 25 studies with 4963 patients concluded that oral glucosamine sulfate may reduce osteoarthritis (OA) pain and improve functionality, without many adverse effects.¹¹ A 2-year double-blind RCT compared the effects of glucosamine hydrochloride 500 mg tid, chondroitin sulfate 400 mg tid, glucosamine plus chondroitin, celecoxib 200 mg/d, or placebo in 662 patients with knee OA.¹² While all groups experienced early and sustained symptomatic relief, the odds of achieving a 20% reduction in pain and improved functioning were highest with celecoxib and glucosamine.

Oral glucosamine sulfate can cause mild GI effects, but drowsiness, skin reactions, and headache also have been reported. Shellfish allergy also is a concern; however, shellfish allergies occur due to the proteins in the meat, and not from the shell from which glucosamine is derived. Glucosamine may increase glucose levels and the anticoagulant effects of warfarin.¹³

Probiotics to prevent antibiotic-associated diarrhea

Antibiotic-associated diarrhea (AAD) is a common problem.²¹ Probiotics—microorganisms found in oral supplements, yogurt, and other food—are commonly used to help maintain the balance of intestinal flora.³⁴ A recent Cochrane review of 16 RCTs that included approximately 3400 patients found evidence that probiotics can prevent AAD.²² A 2012 systematic review and meta-analysis of 63 RCTs with more than 11,000 participants concluded that probiotics lowered the relative risk of developing diarrhea compared to control groups.³⁵ The American Academy of Pediatrics supports the use of probiotics, citing results from a meta-analysis that found probiotics reduced the risk of developing diarrhea from 28.5% to 11.9% compared to placebo.³⁶

Exact dosages for probiotics have not been established, and recommendations range from 5 billion to 40 billion colony-forming units/d.²² The most commonly used probiotics are from the Lactobacillus and Saccharomyces genera; relatively little evidence supports other genera.^21,22,35,36

Probiotics are considered relatively safe, but are not recommended for patients who are immunocompromised or have an indwelling medical device.²³ Adverse effects are rare, but may include flatulence, vomiting, rash, chest pain, and increased phlegm.²¹

For a review of the latest evidence on using probiotics to reduce crying in infants with colic, see "Probiotics for colic? A PURL update."

Soy for hyperlipidemia

Soybeans are a species of legume that contain significant amounts of protein, fiber, potassium, and iron. Although soy has been used to prevent or treat cancer, osteoporosis, and menopausal symptoms, current evidence is unfavorable or inconclusive for such conditions. Some RCTs have found soy has small, favorable effects on serum levels of low-density lipoprotein and total cholesterol,²⁴ while others have shown modest improvements in triglyceride levels without significant improvements in other lipid levels.²⁵

A 2011 meta-analysis of 10 RCTs that included 268 participants found that a diet high in non-soy legume products, such as alfalfa, lentils, and other beans, also improved lipid levels.³⁷ A review of 136 studies that described 22 dietary interventions concluded that among other helpful dietary approaches to controlling hyperlipidemia, dietary soy—which contains fiber and polyunsaturated fats—is favored over supplementation of soy protein alone.³⁸

Use caution when recommending soy for patients with thyroid dysfunction or hormone-sensitive cancers because some evidence suggests soy may interfere with absorption of levothyroxine and increase the risk of developing clinical symptoms of hypothyroidism.³⁹

Soy also contains phytoestrogens, and prolonged use of soy supplements may increase the risk of endometrial hyperplasia.²⁴ This risk has been documented only in the use of soy supplements, and not from dietary soy. GI disturbances and rare allergic reactions also have been reported.²⁴

St. John’s wort for depression

Hypericum perforatum (St. John’s wort), a perennial herb, has been used to treat mood disorders and other ailments for more than 2000 years.^40,41 Commercial preparations typically are alcohol extracts with an herb-to-extract ratio of 4:1 to 8:1.²⁶ The normal dose ranges from 900 to 1500 mg/d in 2 to 3 divided doses of the alcohol extract standardized to 0.3% hypericin and/or 3% to 5% hyperforin.

St. John’s wort has been studied extensively as a treatment for depressive disorders. A 2001 double-blind RCT conducted in 11 US academic medical centers and community clinics between 1998 and 2000 that included 200 patients found that St. John’s wort was not effective for moderately severe major depression; a trend toward a positive effect was noted in both the placebo and St. John’s wort groups.²⁶

However, a 2009 Cochrane review of 29 international studies (5489 patients) concluded that St. John’s wort may be better than placebo and as effective as antidepressants for mild to moderate major depression,²⁷ and appeared to have fewer side effects than antidepressants. This review, conducted in German-speaking countries where medical professionals have long prescribed St. John’s wort, reported more positive results than those conducted in other countries.

St. John’s wort interacts with many medications, including antidepressants, oral contraceptives, cyclosporine, digoxin, indinavir, phenytoin, phenobarbital, warfarin, and others. It induces cytochrome P450 (CYP450) enzymes, and therefore can potentially reduce the efficacy of any medication that is metabolized by a CYP450 enzyme. When used in high doses in combination with antidepressants, St. John’s wort may cause serotonin syndrome. Other side effects include photosensitivity, GI complaints, fatigue, and increased risk of cataracts. Due to a lack of clinical data, St. John’s wort is contraindicated in women who are pregnant or breastfeeding.⁴²

Guided imagery for anxiety 
and pain


Guided imagery is a relaxation technique that involves visualizing positive outcomes to reduce one’s reaction to anxiety-provoking or painful experiences.⁴³ It can be practiced independently or under the direction of an instructor. One RCT of 96 women with newly diagnosed breast cancer found that adding relaxation and guided imagery to standard breast cancer treatment protocols positively affected mood and quality of life.¹⁴ While this study saw no change in pathologic responses to chemotherapy,¹⁴ a more recent RCT concluded that such biochemical advantages may be possible.⁴⁴ Guided imagery has been linked to decreased anxiety in diverse studies of students, women in labor, individuals suffering from nightmares, and in occupational settings such as training for pilots and surgeons.^15-18

A review of 9 RCTs of guided imagery for decreasing musculoskeletal pain involving 201 patients found 8 studies reported positive results, though the methodological quality of the studies was low.¹⁹ Of 6 high-quality studies included in a 2012 systematic review, 5 supported the use of guided imagery for postoperative, abdominal, and other nonmusculoskeletal pain.²⁰ This initial evidence is promising, but additional research of high methodological quality is needed to validate the use of guided imagery for anxiety and pain.

Acupuncture for pain

From 2002 to 2007, the use of acupuncture significantly increased in the United States, primarily for the treatment of pain.¹ A 2012 meta-analysis of 29 RCTs that included almost 18,000 participants evaluated the clinical usefulness of acupuncture for back, neck, and shoulder pain, OA, and headache.³ Compared to no treatment, both acupuncture and sham acupuncture significantly improved pain scores. The authors of this meta-analysis found that acupuncture offered a small but significant advantage over sham acupuncture, and concluded that the benefits of acupuncture were not due to a placebo effect.

In 2007, the American College of Physicians (ACP) and the American Pain Society (APS) issued a joint statement indicating that acupuncture should be considered for patients with chronic low back pain who do not respond to conventional therapies.⁴ The North American Spine Society also supports acupuncture, stating that it provides “...better short-term pain relief and functional improvement than no treatment and the addition of acupuncture to other treatment modalities provides a greater benefit than those treatments alone.”⁵ Additional evidence found acupuncture for chronic low back pain improves function and serves as an adjunct therapy.⁶

Reported adverse effects of acupuncture include—but may not be limited to—infection, skin irritation, hematoma, pneumothorax, and spontaneous needle migration.^1,3-6

Yoga for low back pain

An ACP/APS statement indicates that acupuncture should be considered for patients with chronic back pain who don't respond to conventional therapies.Back pain is the most common reason patients use CAM therapies.¹ A systematic review of 10 RCTs that included 967 participants with chronic low back pain found strong evidence for the short-term effectiveness and moderate evidence for the long-term effectiveness of yoga.²⁸ A review of 17 studies that included 1626 patients concluded that yoga improves both pain and functionality; this review did not recommend a specific type of yoga practice.²⁹ In a recent study of 95 minority adults with moderate-to-severe chronic low back pain, once-weekly and twice-weekly yoga for 12 weeks were similarly effective for reducing pain and improving functionality.³⁰

Guidelines from the ACP and the APS recommend yoga as part of an intensive interdisciplinary rehabilitation program for patients with chronic or subacute low back pain who do not improve using other self-care options.⁴ This recommendation is specifically for Viniyoga, a practice in which the instructor recommends modifications to body positioning for each individual based on past injuries and overall physical condition. (For more information on therapeutic uses of yoga, see “Yoga as therapy: When is it helpful?”)

CORRESPONDENCE
Roger Zoorob, MD, MPH, FAAFP; Department of Family and Community Medicine, Baylor College of Medicine, 3701 Kirby Drive, Suite 600, Houston, TX 77098; [email protected]

CASE › Bob F, age 54, seeks care for chronic low back pain. The conservative treatments you have prescribed, including physical therapy, regular exercise, and an over-the-counter nonsteroidal anti-inflammatory drug, have provided minimal pain relief. Mr. F is reluctant to take a prescription pain medication and has expressed interest in trying a complementary and alternative medicine (CAM) therapy, such as acupuncture or yoga. What should you tell him?

Almost 40% of Americans use CAM modalities to treat specific conditions or for overall well-being,¹ and these practices are increasingly becoming a part of our approach to health care, as evidenced by the nearly 50 of facilities across the country that boast integrative health care programs, which combine CAM modalities with conventional medicine.² Emerging evidence suggests several integrative practices may offer health benefits, and primary care physicians must become well-versed in these modalities to effectively communicate potential benefits and harms to patients. In this article, we present evidence from Cochrane reviews and other studies of 8 commonly used CAM therapies, including dietary interventions, a psychotherapeutic modality, and other treatments (TABLE).^1,3-30 And while motivational interviewing technically is not a form of CAM, we also review this modality, which has proven useful in the treatment of patients for substance use. (See “Motivational interviewing for substance abuse.”)

Fish oil for hypertriglyceridemia

High triglyceride levels are a risk factor for cardiovascular disease and a component of metabolic syndrome.⁸ A 2008 review of 47 randomized controlled trials (RCTs) that included 16,511 participants found that omega-3 fatty acid (fish oil) supplements significantly reduced triglyceride levels compared to placebo.⁷ The American Heart Association recommends 2 to 4 g/d of eicosapentaenoic acid (EPA) plus docosahexaenoic acid (DHA) to lower triglyceride levels.⁸

Most studies have found that fish oil supplements are associated with few adverse effects; gastrointestinal (GI) complaints are most common. However, these supplements should be discontinued following an acute bleeding event, such as hemorrhagic stroke, due to their anticoagulant properties.⁹ Some evidence suggests that the risk for prostate cancer is increased in men with high blood levels of omega-3 fatty acids.¹⁰

Glucosamine for osteoarthritis

The AHA recommends 2 to 4 g/d 
of EPA plus DHA to lower triglyceride levels.Glucosamine is an amino sugar that is a building block of cartilage proteoglycans. Although it occurs naturally in the body, the glucosamine used in supplements is typically harvested from seashells. Glucosamine stimulates the metabolism of synovial cells and chondrocytes in articular cartilage and may delay joint degeneration.^31,32

Glucosamine is widely used in the United States as a dietary supplement, most often as glucosamine sulfate but also as n-acetyl glucosamine and glucosamine hydrochloride, although there is limited evidence of effectiveness for the latter formulations.³³

Most studies have examined the effects of oral glucosamine sulfate, 500 mg taken 3 times a day for 30 to 90 days. Once-a-day dosing as high as 1500 mg also has been used.

A Cochrane review of 25 studies with 4963 patients concluded that oral glucosamine sulfate may reduce osteoarthritis (OA) pain and improve functionality, without many adverse effects.¹¹ A 2-year double-blind RCT compared the effects of glucosamine hydrochloride 500 mg tid, chondroitin sulfate 400 mg tid, glucosamine plus chondroitin, celecoxib 200 mg/d, or placebo in 662 patients with knee OA.¹² While all groups experienced early and sustained symptomatic relief, the odds of achieving a 20% reduction in pain and improved functioning were highest with celecoxib and glucosamine.

Oral glucosamine sulfate can cause mild GI effects, but drowsiness, skin reactions, and headache also have been reported. Shellfish allergy also is a concern; however, shellfish allergies occur due to the proteins in the meat, and not from the shell from which glucosamine is derived. Glucosamine may increase glucose levels and the anticoagulant effects of warfarin.¹³

Probiotics to prevent antibiotic-associated diarrhea

Antibiotic-associated diarrhea (AAD) is a common problem.²¹ Probiotics—microorganisms found in oral supplements, yogurt, and other food—are commonly used to help maintain the balance of intestinal flora.³⁴ A recent Cochrane review of 16 RCTs that included approximately 3400 patients found evidence that probiotics can prevent AAD.²² A 2012 systematic review and meta-analysis of 63 RCTs with more than 11,000 participants concluded that probiotics lowered the relative risk of developing diarrhea compared to control groups.³⁵ The American Academy of Pediatrics supports the use of probiotics, citing results from a meta-analysis that found probiotics reduced the risk of developing diarrhea from 28.5% to 11.9% compared to placebo.³⁶

Exact dosages for probiotics have not been established, and recommendations range from 5 billion to 40 billion colony-forming units/d.²² The most commonly used probiotics are from the Lactobacillus and Saccharomyces genera; relatively little evidence supports other genera.^21,22,35,36

Probiotics are considered relatively safe, but are not recommended for patients who are immunocompromised or have an indwelling medical device.²³ Adverse effects are rare, but may include flatulence, vomiting, rash, chest pain, and increased phlegm.²¹

For a review of the latest evidence on using probiotics to reduce crying in infants with colic, see "Probiotics for colic? A PURL update."

Soy for hyperlipidemia

Soybeans are a species of legume that contain significant amounts of protein, fiber, potassium, and iron. Although soy has been used to prevent or treat cancer, osteoporosis, and menopausal symptoms, current evidence is unfavorable or inconclusive for such conditions. Some RCTs have found soy has small, favorable effects on serum levels of low-density lipoprotein and total cholesterol,²⁴ while others have shown modest improvements in triglyceride levels without significant improvements in other lipid levels.²⁵

A 2011 meta-analysis of 10 RCTs that included 268 participants found that a diet high in non-soy legume products, such as alfalfa, lentils, and other beans, also improved lipid levels.³⁷ A review of 136 studies that described 22 dietary interventions concluded that among other helpful dietary approaches to controlling hyperlipidemia, dietary soy—which contains fiber and polyunsaturated fats—is favored over supplementation of soy protein alone.³⁸

Use caution when recommending soy for patients with thyroid dysfunction or hormone-sensitive cancers because some evidence suggests soy may interfere with absorption of levothyroxine and increase the risk of developing clinical symptoms of hypothyroidism.³⁹

Soy also contains phytoestrogens, and prolonged use of soy supplements may increase the risk of endometrial hyperplasia.²⁴ This risk has been documented only in the use of soy supplements, and not from dietary soy. GI disturbances and rare allergic reactions also have been reported.²⁴

St. John’s wort for depression

Hypericum perforatum (St. John’s wort), a perennial herb, has been used to treat mood disorders and other ailments for more than 2000 years.^40,41 Commercial preparations typically are alcohol extracts with an herb-to-extract ratio of 4:1 to 8:1.²⁶ The normal dose ranges from 900 to 1500 mg/d in 2 to 3 divided doses of the alcohol extract standardized to 0.3% hypericin and/or 3% to 5% hyperforin.

St. John’s wort has been studied extensively as a treatment for depressive disorders. A 2001 double-blind RCT conducted in 11 US academic medical centers and community clinics between 1998 and 2000 that included 200 patients found that St. John’s wort was not effective for moderately severe major depression; a trend toward a positive effect was noted in both the placebo and St. John’s wort groups.²⁶

However, a 2009 Cochrane review of 29 international studies (5489 patients) concluded that St. John’s wort may be better than placebo and as effective as antidepressants for mild to moderate major depression,²⁷ and appeared to have fewer side effects than antidepressants. This review, conducted in German-speaking countries where medical professionals have long prescribed St. John’s wort, reported more positive results than those conducted in other countries.

St. John’s wort interacts with many medications, including antidepressants, oral contraceptives, cyclosporine, digoxin, indinavir, phenytoin, phenobarbital, warfarin, and others. It induces cytochrome P450 (CYP450) enzymes, and therefore can potentially reduce the efficacy of any medication that is metabolized by a CYP450 enzyme. When used in high doses in combination with antidepressants, St. John’s wort may cause serotonin syndrome. Other side effects include photosensitivity, GI complaints, fatigue, and increased risk of cataracts. Due to a lack of clinical data, St. John’s wort is contraindicated in women who are pregnant or breastfeeding.⁴²

Guided imagery for anxiety 
and pain


Guided imagery is a relaxation technique that involves visualizing positive outcomes to reduce one’s reaction to anxiety-provoking or painful experiences.⁴³ It can be practiced independently or under the direction of an instructor. One RCT of 96 women with newly diagnosed breast cancer found that adding relaxation and guided imagery to standard breast cancer treatment protocols positively affected mood and quality of life.¹⁴ While this study saw no change in pathologic responses to chemotherapy,¹⁴ a more recent RCT concluded that such biochemical advantages may be possible.⁴⁴ Guided imagery has been linked to decreased anxiety in diverse studies of students, women in labor, individuals suffering from nightmares, and in occupational settings such as training for pilots and surgeons.^15-18

A review of 9 RCTs of guided imagery for decreasing musculoskeletal pain involving 201 patients found 8 studies reported positive results, though the methodological quality of the studies was low.¹⁹ Of 6 high-quality studies included in a 2012 systematic review, 5 supported the use of guided imagery for postoperative, abdominal, and other nonmusculoskeletal pain.²⁰ This initial evidence is promising, but additional research of high methodological quality is needed to validate the use of guided imagery for anxiety and pain.

Acupuncture for pain

From 2002 to 2007, the use of acupuncture significantly increased in the United States, primarily for the treatment of pain.¹ A 2012 meta-analysis of 29 RCTs that included almost 18,000 participants evaluated the clinical usefulness of acupuncture for back, neck, and shoulder pain, OA, and headache.³ Compared to no treatment, both acupuncture and sham acupuncture significantly improved pain scores. The authors of this meta-analysis found that acupuncture offered a small but significant advantage over sham acupuncture, and concluded that the benefits of acupuncture were not due to a placebo effect.

In 2007, the American College of Physicians (ACP) and the American Pain Society (APS) issued a joint statement indicating that acupuncture should be considered for patients with chronic low back pain who do not respond to conventional therapies.⁴ The North American Spine Society also supports acupuncture, stating that it provides “...better short-term pain relief and functional improvement than no treatment and the addition of acupuncture to other treatment modalities provides a greater benefit than those treatments alone.”⁵ Additional evidence found acupuncture for chronic low back pain improves function and serves as an adjunct therapy.⁶

Reported adverse effects of acupuncture include—but may not be limited to—infection, skin irritation, hematoma, pneumothorax, and spontaneous needle migration.^1,3-6

Yoga for low back pain

An ACP/APS statement indicates that acupuncture should be considered for patients with chronic back pain who don't respond to conventional therapies.Back pain is the most common reason patients use CAM therapies.¹ A systematic review of 10 RCTs that included 967 participants with chronic low back pain found strong evidence for the short-term effectiveness and moderate evidence for the long-term effectiveness of yoga.²⁸ A review of 17 studies that included 1626 patients concluded that yoga improves both pain and functionality; this review did not recommend a specific type of yoga practice.²⁹ In a recent study of 95 minority adults with moderate-to-severe chronic low back pain, once-weekly and twice-weekly yoga for 12 weeks were similarly effective for reducing pain and improving functionality.³⁰

Guidelines from the ACP and the APS recommend yoga as part of an intensive interdisciplinary rehabilitation program for patients with chronic or subacute low back pain who do not improve using other self-care options.⁴ This recommendation is specifically for Viniyoga, a practice in which the instructor recommends modifications to body positioning for each individual based on past injuries and overall physical condition. (For more information on therapeutic uses of yoga, see “Yoga as therapy: When is it helpful?”)

CORRESPONDENCE
Roger Zoorob, MD, MPH, FAAFP; Department of Family and Community Medicine, Baylor College of Medicine, 3701 Kirby Drive, Suite 600, Houston, TX 77098; [email protected]

CASE › Carla S is a 64-year old African American whom you’re seeing for the first time. Her health has been excellent over the last 10 years, she reports, with one caveat: She has “borderline hypertension,” but has never been treated for it and denies any symptoms. Her blood pressure (BP) today is 154/82 mm Hg. A physical exam is unremarkable. Blood tests reveal a normal blood count and normal renal function, and a nonfasting glucose level of 145 mg/dL. You ask Ms. S to return in a week for a repeat BP and fasting lab work.

Hypertension is the most common condition seen by physicians in primary care,¹ and a major risk factor for cardiovascular disease (CVD) and the morbidity and mortality associated with it. US treatment costs are an estimated $131 billion per year.^2-4 With this in mind, the Joint National Committee on Hypertension (JNC) released its eighth report (JNC 8) in December 2013¹—the first update in a decade.

In many ways, JNC 8 guidelines are simpler than those of JNC 7,² with more evidence-based recommendations and less reliance on expert opinion. The JNC has eliminated definitions such as stage 1, 2, and 3 hypertension, and focuses on outcomes instead. At the heart of the recommendations are 3 key questions:

    1. At what BP should treatment be initiated to improve outcomes?
    2. What should the target BP be for those undergoing treatment?
    3. Which medications are best?

Answers to the first 2 questions, of course, go hand in hand. In other words, if the threshold for treatment is a systolic BP ≥140 mm Hg (more on that in a moment), then the target of treatment is a systolic BP of <140 mm Hg. In answer to the third question, JNC 8 offers guidance but gives physicians greater discretion in determining which type of drug to use when initiating treatment.¹

In the text, algorithm, and table that follow, we present an overview of JNC 8. We also discuss the optimal treatment of hypertension in patients with heart failure (HF) and coronary artery disease (CAD)—populations JNC 8 does not address.

Age-based recommendations are a bit less stringent

JNC 8 does not address the optimal treatment 
of hypertension in patients with heart failure and coronary artery disease, but the ACC/AHA guidelines we report on here can help.60 years and older. Unlike JNC 7, which recommended initiating treatment for otherwise healthy patients of all ages with a BP ≥140/90 mm Hg,² JNC 8 clearly delineates its recommendations by age. It calls for treating patients ages 60 or older with systolic BP ≥150 mm Hg and/or diastolic BP ≥90.¹

The change is evidence-based: Moderate- to high-quality randomized controlled trials (RCTs) have found a reduced incidence of stroke, HF, and coronary heart disease when BP was treated to <150/90, but no additional benefit from a systolic BP target of <140 mm Hg for patients in this age group.^5,6 Notably, JNC 8 does not recommend a change in medication for patients 60 years or older for whom the more stringent target is being maintained without adverse effects.¹

18 to 59 years. For adults younger than 60, JNC 8 recommends treating systolic BP ≥140 and diastolic BP ≥90 mm Hg.¹ The systolic BP guideline is based on expert opinion, however, as there is no high-quality evidence for a systolic threshold in this age group. This is largely because most patients younger than 60 who have systolic BP ≥140 also have diastolic BP ≥90, making it difficult to study the treatment of systolic BP alone. High-quality trials have shown improved health outcomes when patients ages 30 to 59 years were treated for diastolic BP ≥90, however.^7-12 For patients younger than 30, the recommendation for treatment of diastolic pressure is based on expert consensus, as no sufficiently high-quality evidence exists.

Targets for patients with CKD and diabetes

Chronic kidney disease (CKD). JNC 8 recommends treating patients ages 18 to 69 years who have CKD and BP ≥140/90 mm Hg. JNC 7’s more stringent recommendation—treating such patients with BP ≥130/80 mm Hg²—was relaxed because there is little evidence of a lower mortality rate or cardiovascular or cerebrovascular benefits as a result of tighter control. In patients younger than 70, CKD is defined as an estimated (or measured) glomerular filtration rate (GFR) <60 mL/min/1.73 m² or albuminuria (>30 mg of albumin per g of creatinine).¹

The new guidelines include 3 methods of dosing antihypertensive medications; all stress the importance of avoiding ACEI and ARB combinations. It is important to note that this goal does not apply to individuals who have CKD and are 70 years or older. This is due to insufficient evidence, as well as uncertainty about the accuracy of an estimated GFR in this patient population. JNC 8 recommends that treatment of BP in patients 70 or older be based on comorbidities, including albuminuria, among other patient-specific considerations.¹

Diabetes. JNC 8 recommends treating patients age 18 years or older who have diabetes and BP ≥140/90 mm Hg, as JNC 7 did.² This is based largely on expert opinion.

Studies suggest that adults with both hypertension and diabetes have a reduction in mortality and improved cardiovascular and cerebrovascular outcomes when systolic BP is <150 mm Hg,^13-15 but no strong data support a goal of <140/90 mm Hg. The Action to Control Cardiovascular Risk in Diabetes (ACCORD) BP trial, for example, showed comparable outcomes in patients with systolic BP of 150 or 140 mm Hg.¹⁶ The use of expert opinion vs well-designed studies in this instance seems at odds with JNC 8’s general policy of placing greater emphasis on evidence.

CASE › On her second visit, Ms. S’s BP is 144/82 mm Hg and her cholesterol levels are within the normal range. Her fasting glucose level is 104 mg/dL and glycated hemoglobin (HbA1c) is 6%. At a repeat visit one month later, her BP is 146/76 mm Hg. Given these 2 acceptable readings (<150/90 mm Hg for individuals age 60 and older who do not have diabetes), you do not initiate antihypertensive treatment.

However, you explain to the patient that her fasting glucose and HbA1c are evidence of insulin resistance. Although a diagnosis of diabetes is not warranted, you arrange for Ms. S to meet with a diabetes nurse educator for help in improving her diet and following an exercise regimen.

Pharmacotherapy: JNC 8 offers wider latitude

Like its predecessor, JNC 8 stresses the importance of diet and exercise. (See “Controlling hypertension starts with lifestyle modification”¹⁷ in this article.) It diverges from JNC 7, however, in its recommendations for initiating treatment (ALGORITHM).¹ The earlier version recommended thiazide diuretics as first-line therapy but included multiple indications for initiating therapy with other drug classes. JNC 8 guidelines are less specific.

Starting therapy with a thiazide diuretic, angiotensin-converting enzyme inhibitor (ACEI), angiotensin receptor blocker (ARB), or calcium channel blocker (CCB)—all of which have high-quality evidence of improved outcomes^18-20—is recommended for most patients, including those with diabetes. (Blacks and patients with CKD are exceptions.) The recommended doses of these medications, summarized in the TABLE,^1,21 are similar to those used in RCTs. Other types of drugs are not recommended, either because they were shown to be inferior to another class of antihypertensive or because there is insufficient evidence of their efficacy.

For most blacks... JNC 8 recommends thiazide diuretics and CCBs as first-line therapy—a recommendation that is evidence-based. The Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT)²² revealed that black patients taking thiazide diuretics had fewer cerebrovascular and cardiovascular events and a lower rate of HF compared with those taking ACEIs, whether or not they had diabetes. Diuretics were more effective than CCBs in preventing HF, but no difference in rates of cerebrovascular and cardiovascular events, kidney disease, or overall mortality was found.²²

For patients with CKD and proteinuria, regardless of race, JNC 8 calls for either an ACEI or an ARB as first-line agent to prevent progression to end-stage renal disease. This recommendation is based on expert consensus, and intended to prevent progression to end-stage renal disease.^1,23

The optimal first-line agent for patients who have CKD without proteinuria is less clear. For such patients, JNC 8 notes, any of the 4 recommended drug classes can be used for initial therapy.¹

Guidance on starting—and titrating—therapy


JNC 7 guidelines featured a complex means of diagnosing and monitoring hypertension.² JNC 8 has simplified the recommendations, which call for patients to be reassessed within a month of initiating therapy.

The new guidelines include 3 distinct methods of dosing antihypertensive medications, none of which has demonstrated better outcomes than any other. All call for replacing one type of drug with another if the first trial is ineffective or results in adverse effects. And all stress the importance of avoiding ACEI and ARB combinations due to increases in serum creatinine and hyperkalemia and the need for monitoring. Note, however, that Method 3 is recommended for patients with more severe hypertension.¹

Method 1. Initiate one medication from any of the 4 classes of antihypertensives recommended for initial treatment, and titrate to the maximum effective dose. If the BP goal is not achieved at maximum dose, add a medication from a second class and titrate that drug to the maximum effective dose, as well. If the goal is still not reached, add a medication from a third class and titrate up as needed.

Method 2. Initiate one medication, then add a second agent from a different drug class, if necessary, and titrate until both are at the maximum effective dose. If the goal still has not been reached, add a third agent and titrate that until BP is well controlled.

Method 3. Initiate 2 medications from 2 different classes of drugs simultaneously. If BP is not at goal after a reasonable trial, add a third agent and titrate to maximum effective dose. (Use this approach for patients who have systolic BP >160 mm Hg and/or diastolic BP >100 mm Hg or systolic BP >20 mm Hg above goal and/or diastolic BP >10 mm Hg above goal.)

As a general rule, a trial with monotherapy should be considered if BP is ≤160/100; a 2-agent combination is recommended as first-line therapy for pressure that exceeds that threshold. If a patient’s BP target is not reached even with the above strategies, a consultation with a hypertension specialist may be needed.

Treating patients with cardiovascular comorbidities

As noted earlier, JNC 8 offers no guidance in treating patients with HF or CAD and multiple comorbidities. In such cases, we turn to the American College of Cardiology (ACC) and American Heart Association (AHA).²⁴

Recent ACC/AHA guidelines recommend a beta-blocker and ACEI for patients with a history of symptomatic stable HF and a left ventricular ejection fraction (EF) ≤40%, unless contraindications exist.²⁴ Beta-blockers and an ACEI or an ARB should be used to prevent HF in patients with a history of myocardial infarction (MI) or acute coronary syndrome and a reduced EF. Beta-blockers with evidence to support their use in such cases include carvedilol, bisoprolol, and sustained-release metoprolol succinate.²⁴

For symptomatic patients with dyspnea or other mild fluid retention, a loop diuretic or a thiazide diuretic can be used. Nondihydropyridine CCBs should be avoided in post-MI patients with low left ventricular EF due to the medication’s negative inotropic effects.²⁴ The optimal drug regimen for secondary stroke prevention is not clear due to a lack of studies comparing drug regimens, but data suggest that a diuretic or a diuretic-ACEI combination is beneficial.²⁵

Evaluating treatment-resistant hypertension


When a patient presents with treatment-resistant hypertension—elevated BP that is not controlled with a 3-drug regimen, all at maximum doses—start by asking several questions.²⁶ Is the patient:

• having difficulty following a drug regimen that calls for multiple daily doses?
• drinking excessive amounts of alcohol?
• failing to adhere to a low-salt dietary regimen?
• 
taking any other medications or supplements that might elevate BP (eg, nonsteroidal anti-inflammatory agents, pseudoephedrine, ephedra, or licorice)?
• unable to afford all the drugs prescribed?

If no such issues are identified, consider a referral to a specialist for further evaluation and to rule out disorders associated with treatment-resistant hypertension, including CKD, renal artery stenosis, hyperaldosteronemia, sleep apnea, and coarctation of the aorta.²⁶

CASE › When Ms. S returns 3 months later, her BP is 140/70 mm Hg, her fasting glucose is 94 mg/dL, and her HbA1c is 5.7%. You encourage her to continue her new dietary and exercise regimen and schedule a follow-up visit in 6 months.

CORRESPONDENCE
Tracy D. Mahvan, PharmD, University of Wyoming, School of Pharmacy, Health Sciences Center, Room 292, 1000 East University Avenue, Department 3375, Laramie, WY 82071; [email protected]

CASE › Carla S is a 64-year old African American whom you’re seeing for the first time. Her health has been excellent over the last 10 years, she reports, with one caveat: She has “borderline hypertension,” but has never been treated for it and denies any symptoms. Her blood pressure (BP) today is 154/82 mm Hg. A physical exam is unremarkable. Blood tests reveal a normal blood count and normal renal function, and a nonfasting glucose level of 145 mg/dL. You ask Ms. S to return in a week for a repeat BP and fasting lab work.

Hypertension is the most common condition seen by physicians in primary care,¹ and a major risk factor for cardiovascular disease (CVD) and the morbidity and mortality associated with it. US treatment costs are an estimated $131 billion per year.^2-4 With this in mind, the Joint National Committee on Hypertension (JNC) released its eighth report (JNC 8) in December 2013¹—the first update in a decade.

In many ways, JNC 8 guidelines are simpler than those of JNC 7,² with more evidence-based recommendations and less reliance on expert opinion. The JNC has eliminated definitions such as stage 1, 2, and 3 hypertension, and focuses on outcomes instead. At the heart of the recommendations are 3 key questions:

    1. At what BP should treatment be initiated to improve outcomes?
    2. What should the target BP be for those undergoing treatment?
    3. Which medications are best?

Answers to the first 2 questions, of course, go hand in hand. In other words, if the threshold for treatment is a systolic BP ≥140 mm Hg (more on that in a moment), then the target of treatment is a systolic BP of <140 mm Hg. In answer to the third question, JNC 8 offers guidance but gives physicians greater discretion in determining which type of drug to use when initiating treatment.¹

In the text, algorithm, and table that follow, we present an overview of JNC 8. We also discuss the optimal treatment of hypertension in patients with heart failure (HF) and coronary artery disease (CAD)—populations JNC 8 does not address.

Age-based recommendations are a bit less stringent

JNC 8 does not address the optimal treatment 
of hypertension in patients with heart failure and coronary artery disease, but the ACC/AHA guidelines we report on here can help.60 years and older. Unlike JNC 7, which recommended initiating treatment for otherwise healthy patients of all ages with a BP ≥140/90 mm Hg,² JNC 8 clearly delineates its recommendations by age. It calls for treating patients ages 60 or older with systolic BP ≥150 mm Hg and/or diastolic BP ≥90.¹

The change is evidence-based: Moderate- to high-quality randomized controlled trials (RCTs) have found a reduced incidence of stroke, HF, and coronary heart disease when BP was treated to <150/90, but no additional benefit from a systolic BP target of <140 mm Hg for patients in this age group.^5,6 Notably, JNC 8 does not recommend a change in medication for patients 60 years or older for whom the more stringent target is being maintained without adverse effects.¹

18 to 59 years. For adults younger than 60, JNC 8 recommends treating systolic BP ≥140 and diastolic BP ≥90 mm Hg.¹ The systolic BP guideline is based on expert opinion, however, as there is no high-quality evidence for a systolic threshold in this age group. This is largely because most patients younger than 60 who have systolic BP ≥140 also have diastolic BP ≥90, making it difficult to study the treatment of systolic BP alone. High-quality trials have shown improved health outcomes when patients ages 30 to 59 years were treated for diastolic BP ≥90, however.^7-12 For patients younger than 30, the recommendation for treatment of diastolic pressure is based on expert consensus, as no sufficiently high-quality evidence exists.

Targets for patients with CKD and diabetes

Chronic kidney disease (CKD). JNC 8 recommends treating patients ages 18 to 69 years who have CKD and BP ≥140/90 mm Hg. JNC 7’s more stringent recommendation—treating such patients with BP ≥130/80 mm Hg²—was relaxed because there is little evidence of a lower mortality rate or cardiovascular or cerebrovascular benefits as a result of tighter control. In patients younger than 70, CKD is defined as an estimated (or measured) glomerular filtration rate (GFR) <60 mL/min/1.73 m² or albuminuria (>30 mg of albumin per g of creatinine).¹

The new guidelines include 3 methods of dosing antihypertensive medications; all stress the importance of avoiding ACEI and ARB combinations. It is important to note that this goal does not apply to individuals who have CKD and are 70 years or older. This is due to insufficient evidence, as well as uncertainty about the accuracy of an estimated GFR in this patient population. JNC 8 recommends that treatment of BP in patients 70 or older be based on comorbidities, including albuminuria, among other patient-specific considerations.¹

Diabetes. JNC 8 recommends treating patients age 18 years or older who have diabetes and BP ≥140/90 mm Hg, as JNC 7 did.² This is based largely on expert opinion.

Studies suggest that adults with both hypertension and diabetes have a reduction in mortality and improved cardiovascular and cerebrovascular outcomes when systolic BP is <150 mm Hg,^13-15 but no strong data support a goal of <140/90 mm Hg. The Action to Control Cardiovascular Risk in Diabetes (ACCORD) BP trial, for example, showed comparable outcomes in patients with systolic BP of 150 or 140 mm Hg.¹⁶ The use of expert opinion vs well-designed studies in this instance seems at odds with JNC 8’s general policy of placing greater emphasis on evidence.

CASE › On her second visit, Ms. S’s BP is 144/82 mm Hg and her cholesterol levels are within the normal range. Her fasting glucose level is 104 mg/dL and glycated hemoglobin (HbA1c) is 6%. At a repeat visit one month later, her BP is 146/76 mm Hg. Given these 2 acceptable readings (<150/90 mm Hg for individuals age 60 and older who do not have diabetes), you do not initiate antihypertensive treatment.

However, you explain to the patient that her fasting glucose and HbA1c are evidence of insulin resistance. Although a diagnosis of diabetes is not warranted, you arrange for Ms. S to meet with a diabetes nurse educator for help in improving her diet and following an exercise regimen.

Pharmacotherapy: JNC 8 offers wider latitude

Like its predecessor, JNC 8 stresses the importance of diet and exercise. (See “Controlling hypertension starts with lifestyle modification”¹⁷ in this article.) It diverges from JNC 7, however, in its recommendations for initiating treatment (ALGORITHM).¹ The earlier version recommended thiazide diuretics as first-line therapy but included multiple indications for initiating therapy with other drug classes. JNC 8 guidelines are less specific.

Starting therapy with a thiazide diuretic, angiotensin-converting enzyme inhibitor (ACEI), angiotensin receptor blocker (ARB), or calcium channel blocker (CCB)—all of which have high-quality evidence of improved outcomes^18-20—is recommended for most patients, including those with diabetes. (Blacks and patients with CKD are exceptions.) The recommended doses of these medications, summarized in the TABLE,^1,21 are similar to those used in RCTs. Other types of drugs are not recommended, either because they were shown to be inferior to another class of antihypertensive or because there is insufficient evidence of their efficacy.

For most blacks... JNC 8 recommends thiazide diuretics and CCBs as first-line therapy—a recommendation that is evidence-based. The Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT)²² revealed that black patients taking thiazide diuretics had fewer cerebrovascular and cardiovascular events and a lower rate of HF compared with those taking ACEIs, whether or not they had diabetes. Diuretics were more effective than CCBs in preventing HF, but no difference in rates of cerebrovascular and cardiovascular events, kidney disease, or overall mortality was found.²²

For patients with CKD and proteinuria, regardless of race, JNC 8 calls for either an ACEI or an ARB as first-line agent to prevent progression to end-stage renal disease. This recommendation is based on expert consensus, and intended to prevent progression to end-stage renal disease.^1,23

The optimal first-line agent for patients who have CKD without proteinuria is less clear. For such patients, JNC 8 notes, any of the 4 recommended drug classes can be used for initial therapy.¹

Guidance on starting—and titrating—therapy


JNC 7 guidelines featured a complex means of diagnosing and monitoring hypertension.² JNC 8 has simplified the recommendations, which call for patients to be reassessed within a month of initiating therapy.

The new guidelines include 3 distinct methods of dosing antihypertensive medications, none of which has demonstrated better outcomes than any other. All call for replacing one type of drug with another if the first trial is ineffective or results in adverse effects. And all stress the importance of avoiding ACEI and ARB combinations due to increases in serum creatinine and hyperkalemia and the need for monitoring. Note, however, that Method 3 is recommended for patients with more severe hypertension.¹

Method 1. Initiate one medication from any of the 4 classes of antihypertensives recommended for initial treatment, and titrate to the maximum effective dose. If the BP goal is not achieved at maximum dose, add a medication from a second class and titrate that drug to the maximum effective dose, as well. If the goal is still not reached, add a medication from a third class and titrate up as needed.

Method 2. Initiate one medication, then add a second agent from a different drug class, if necessary, and titrate until both are at the maximum effective dose. If the goal still has not been reached, add a third agent and titrate that until BP is well controlled.

Method 3. Initiate 2 medications from 2 different classes of drugs simultaneously. If BP is not at goal after a reasonable trial, add a third agent and titrate to maximum effective dose. (Use this approach for patients who have systolic BP >160 mm Hg and/or diastolic BP >100 mm Hg or systolic BP >20 mm Hg above goal and/or diastolic BP >10 mm Hg above goal.)

As a general rule, a trial with monotherapy should be considered if BP is ≤160/100; a 2-agent combination is recommended as first-line therapy for pressure that exceeds that threshold. If a patient’s BP target is not reached even with the above strategies, a consultation with a hypertension specialist may be needed.

Treating patients with cardiovascular comorbidities

As noted earlier, JNC 8 offers no guidance in treating patients with HF or CAD and multiple comorbidities. In such cases, we turn to the American College of Cardiology (ACC) and American Heart Association (AHA).²⁴

Recent ACC/AHA guidelines recommend a beta-blocker and ACEI for patients with a history of symptomatic stable HF and a left ventricular ejection fraction (EF) ≤40%, unless contraindications exist.²⁴ Beta-blockers and an ACEI or an ARB should be used to prevent HF in patients with a history of myocardial infarction (MI) or acute coronary syndrome and a reduced EF. Beta-blockers with evidence to support their use in such cases include carvedilol, bisoprolol, and sustained-release metoprolol succinate.²⁴

For symptomatic patients with dyspnea or other mild fluid retention, a loop diuretic or a thiazide diuretic can be used. Nondihydropyridine CCBs should be avoided in post-MI patients with low left ventricular EF due to the medication’s negative inotropic effects.²⁴ The optimal drug regimen for secondary stroke prevention is not clear due to a lack of studies comparing drug regimens, but data suggest that a diuretic or a diuretic-ACEI combination is beneficial.²⁵

Evaluating treatment-resistant hypertension


When a patient presents with treatment-resistant hypertension—elevated BP that is not controlled with a 3-drug regimen, all at maximum doses—start by asking several questions.²⁶ Is the patient:

• having difficulty following a drug regimen that calls for multiple daily doses?
• drinking excessive amounts of alcohol?
• failing to adhere to a low-salt dietary regimen?
• 
taking any other medications or supplements that might elevate BP (eg, nonsteroidal anti-inflammatory agents, pseudoephedrine, ephedra, or licorice)?
• unable to afford all the drugs prescribed?

If no such issues are identified, consider a referral to a specialist for further evaluation and to rule out disorders associated with treatment-resistant hypertension, including CKD, renal artery stenosis, hyperaldosteronemia, sleep apnea, and coarctation of the aorta.²⁶

CASE › When Ms. S returns 3 months later, her BP is 140/70 mm Hg, her fasting glucose is 94 mg/dL, and her HbA1c is 5.7%. You encourage her to continue her new dietary and exercise regimen and schedule a follow-up visit in 6 months.

CORRESPONDENCE
Tracy D. Mahvan, PharmD, University of Wyoming, School of Pharmacy, Health Sciences Center, Room 292, 1000 East University Avenue, Department 3375, Laramie, WY 82071; [email protected]

CASE › Carla S is a 64-year old African American whom you’re seeing for the first time. Her health has been excellent over the last 10 years, she reports, with one caveat: She has “borderline hypertension,” but has never been treated for it and denies any symptoms. Her blood pressure (BP) today is 154/82 mm Hg. A physical exam is unremarkable. Blood tests reveal a normal blood count and normal renal function, and a nonfasting glucose level of 145 mg/dL. You ask Ms. S to return in a week for a repeat BP and fasting lab work.

Hypertension is the most common condition seen by physicians in primary care,¹ and a major risk factor for cardiovascular disease (CVD) and the morbidity and mortality associated with it. US treatment costs are an estimated $131 billion per year.^2-4 With this in mind, the Joint National Committee on Hypertension (JNC) released its eighth report (JNC 8) in December 2013¹—the first update in a decade.

In many ways, JNC 8 guidelines are simpler than those of JNC 7,² with more evidence-based recommendations and less reliance on expert opinion. The JNC has eliminated definitions such as stage 1, 2, and 3 hypertension, and focuses on outcomes instead. At the heart of the recommendations are 3 key questions:

    1. At what BP should treatment be initiated to improve outcomes?
    2. What should the target BP be for those undergoing treatment?
    3. Which medications are best?

Answers to the first 2 questions, of course, go hand in hand. In other words, if the threshold for treatment is a systolic BP ≥140 mm Hg (more on that in a moment), then the target of treatment is a systolic BP of <140 mm Hg. In answer to the third question, JNC 8 offers guidance but gives physicians greater discretion in determining which type of drug to use when initiating treatment.¹

In the text, algorithm, and table that follow, we present an overview of JNC 8. We also discuss the optimal treatment of hypertension in patients with heart failure (HF) and coronary artery disease (CAD)—populations JNC 8 does not address.

Age-based recommendations are a bit less stringent

JNC 8 does not address the optimal treatment 
of hypertension in patients with heart failure and coronary artery disease, but the ACC/AHA guidelines we report on here can help.60 years and older. Unlike JNC 7, which recommended initiating treatment for otherwise healthy patients of all ages with a BP ≥140/90 mm Hg,² JNC 8 clearly delineates its recommendations by age. It calls for treating patients ages 60 or older with systolic BP ≥150 mm Hg and/or diastolic BP ≥90.¹

The change is evidence-based: Moderate- to high-quality randomized controlled trials (RCTs) have found a reduced incidence of stroke, HF, and coronary heart disease when BP was treated to <150/90, but no additional benefit from a systolic BP target of <140 mm Hg for patients in this age group.^5,6 Notably, JNC 8 does not recommend a change in medication for patients 60 years or older for whom the more stringent target is being maintained without adverse effects.¹

18 to 59 years. For adults younger than 60, JNC 8 recommends treating systolic BP ≥140 and diastolic BP ≥90 mm Hg.¹ The systolic BP guideline is based on expert opinion, however, as there is no high-quality evidence for a systolic threshold in this age group. This is largely because most patients younger than 60 who have systolic BP ≥140 also have diastolic BP ≥90, making it difficult to study the treatment of systolic BP alone. High-quality trials have shown improved health outcomes when patients ages 30 to 59 years were treated for diastolic BP ≥90, however.^7-12 For patients younger than 30, the recommendation for treatment of diastolic pressure is based on expert consensus, as no sufficiently high-quality evidence exists.

Targets for patients with CKD and diabetes

Chronic kidney disease (CKD). JNC 8 recommends treating patients ages 18 to 69 years who have CKD and BP ≥140/90 mm Hg. JNC 7’s more stringent recommendation—treating such patients with BP ≥130/80 mm Hg²—was relaxed because there is little evidence of a lower mortality rate or cardiovascular or cerebrovascular benefits as a result of tighter control. In patients younger than 70, CKD is defined as an estimated (or measured) glomerular filtration rate (GFR) <60 mL/min/1.73 m² or albuminuria (>30 mg of albumin per g of creatinine).¹

The new guidelines include 3 methods of dosing antihypertensive medications; all stress the importance of avoiding ACEI and ARB combinations. It is important to note that this goal does not apply to individuals who have CKD and are 70 years or older. This is due to insufficient evidence, as well as uncertainty about the accuracy of an estimated GFR in this patient population. JNC 8 recommends that treatment of BP in patients 70 or older be based on comorbidities, including albuminuria, among other patient-specific considerations.¹

Diabetes. JNC 8 recommends treating patients age 18 years or older who have diabetes and BP ≥140/90 mm Hg, as JNC 7 did.² This is based largely on expert opinion.

Studies suggest that adults with both hypertension and diabetes have a reduction in mortality and improved cardiovascular and cerebrovascular outcomes when systolic BP is <150 mm Hg,^13-15 but no strong data support a goal of <140/90 mm Hg. The Action to Control Cardiovascular Risk in Diabetes (ACCORD) BP trial, for example, showed comparable outcomes in patients with systolic BP of 150 or 140 mm Hg.¹⁶ The use of expert opinion vs well-designed studies in this instance seems at odds with JNC 8’s general policy of placing greater emphasis on evidence.

CASE › On her second visit, Ms. S’s BP is 144/82 mm Hg and her cholesterol levels are within the normal range. Her fasting glucose level is 104 mg/dL and glycated hemoglobin (HbA1c) is 6%. At a repeat visit one month later, her BP is 146/76 mm Hg. Given these 2 acceptable readings (<150/90 mm Hg for individuals age 60 and older who do not have diabetes), you do not initiate antihypertensive treatment.

However, you explain to the patient that her fasting glucose and HbA1c are evidence of insulin resistance. Although a diagnosis of diabetes is not warranted, you arrange for Ms. S to meet with a diabetes nurse educator for help in improving her diet and following an exercise regimen.

Pharmacotherapy: JNC 8 offers wider latitude

Like its predecessor, JNC 8 stresses the importance of diet and exercise. (See “Controlling hypertension starts with lifestyle modification”¹⁷ in this article.) It diverges from JNC 7, however, in its recommendations for initiating treatment (ALGORITHM).¹ The earlier version recommended thiazide diuretics as first-line therapy but included multiple indications for initiating therapy with other drug classes. JNC 8 guidelines are less specific.

Starting therapy with a thiazide diuretic, angiotensin-converting enzyme inhibitor (ACEI), angiotensin receptor blocker (ARB), or calcium channel blocker (CCB)—all of which have high-quality evidence of improved outcomes^18-20—is recommended for most patients, including those with diabetes. (Blacks and patients with CKD are exceptions.) The recommended doses of these medications, summarized in the TABLE,^1,21 are similar to those used in RCTs. Other types of drugs are not recommended, either because they were shown to be inferior to another class of antihypertensive or because there is insufficient evidence of their efficacy.

For most blacks... JNC 8 recommends thiazide diuretics and CCBs as first-line therapy—a recommendation that is evidence-based. The Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT)²² revealed that black patients taking thiazide diuretics had fewer cerebrovascular and cardiovascular events and a lower rate of HF compared with those taking ACEIs, whether or not they had diabetes. Diuretics were more effective than CCBs in preventing HF, but no difference in rates of cerebrovascular and cardiovascular events, kidney disease, or overall mortality was found.²²

For patients with CKD and proteinuria, regardless of race, JNC 8 calls for either an ACEI or an ARB as first-line agent to prevent progression to end-stage renal disease. This recommendation is based on expert consensus, and intended to prevent progression to end-stage renal disease.^1,23

The optimal first-line agent for patients who have CKD without proteinuria is less clear. For such patients, JNC 8 notes, any of the 4 recommended drug classes can be used for initial therapy.¹

Guidance on starting—and titrating—therapy


JNC 7 guidelines featured a complex means of diagnosing and monitoring hypertension.² JNC 8 has simplified the recommendations, which call for patients to be reassessed within a month of initiating therapy.

The new guidelines include 3 distinct methods of dosing antihypertensive medications, none of which has demonstrated better outcomes than any other. All call for replacing one type of drug with another if the first trial is ineffective or results in adverse effects. And all stress the importance of avoiding ACEI and ARB combinations due to increases in serum creatinine and hyperkalemia and the need for monitoring. Note, however, that Method 3 is recommended for patients with more severe hypertension.¹

Method 1. Initiate one medication from any of the 4 classes of antihypertensives recommended for initial treatment, and titrate to the maximum effective dose. If the BP goal is not achieved at maximum dose, add a medication from a second class and titrate that drug to the maximum effective dose, as well. If the goal is still not reached, add a medication from a third class and titrate up as needed.

Method 2. Initiate one medication, then add a second agent from a different drug class, if necessary, and titrate until both are at the maximum effective dose. If the goal still has not been reached, add a third agent and titrate that until BP is well controlled.

Method 3. Initiate 2 medications from 2 different classes of drugs simultaneously. If BP is not at goal after a reasonable trial, add a third agent and titrate to maximum effective dose. (Use this approach for patients who have systolic BP >160 mm Hg and/or diastolic BP >100 mm Hg or systolic BP >20 mm Hg above goal and/or diastolic BP >10 mm Hg above goal.)

As a general rule, a trial with monotherapy should be considered if BP is ≤160/100; a 2-agent combination is recommended as first-line therapy for pressure that exceeds that threshold. If a patient’s BP target is not reached even with the above strategies, a consultation with a hypertension specialist may be needed.

Treating patients with cardiovascular comorbidities

As noted earlier, JNC 8 offers no guidance in treating patients with HF or CAD and multiple comorbidities. In such cases, we turn to the American College of Cardiology (ACC) and American Heart Association (AHA).²⁴

Recent ACC/AHA guidelines recommend a beta-blocker and ACEI for patients with a history of symptomatic stable HF and a left ventricular ejection fraction (EF) ≤40%, unless contraindications exist.²⁴ Beta-blockers and an ACEI or an ARB should be used to prevent HF in patients with a history of myocardial infarction (MI) or acute coronary syndrome and a reduced EF. Beta-blockers with evidence to support their use in such cases include carvedilol, bisoprolol, and sustained-release metoprolol succinate.²⁴

For symptomatic patients with dyspnea or other mild fluid retention, a loop diuretic or a thiazide diuretic can be used. Nondihydropyridine CCBs should be avoided in post-MI patients with low left ventricular EF due to the medication’s negative inotropic effects.²⁴ The optimal drug regimen for secondary stroke prevention is not clear due to a lack of studies comparing drug regimens, but data suggest that a diuretic or a diuretic-ACEI combination is beneficial.²⁵

Evaluating treatment-resistant hypertension


When a patient presents with treatment-resistant hypertension—elevated BP that is not controlled with a 3-drug regimen, all at maximum doses—start by asking several questions.²⁶ Is the patient:

• having difficulty following a drug regimen that calls for multiple daily doses?
• drinking excessive amounts of alcohol?
• failing to adhere to a low-salt dietary regimen?
• 
taking any other medications or supplements that might elevate BP (eg, nonsteroidal anti-inflammatory agents, pseudoephedrine, ephedra, or licorice)?
• unable to afford all the drugs prescribed?

If no such issues are identified, consider a referral to a specialist for further evaluation and to rule out disorders associated with treatment-resistant hypertension, including CKD, renal artery stenosis, hyperaldosteronemia, sleep apnea, and coarctation of the aorta.²⁶

CASE › When Ms. S returns 3 months later, her BP is 140/70 mm Hg, her fasting glucose is 94 mg/dL, and her HbA1c is 5.7%. You encourage her to continue her new dietary and exercise regimen and schedule a follow-up visit in 6 months.

CORRESPONDENCE
Tracy D. Mahvan, PharmD, University of Wyoming, School of Pharmacy, Health Sciences Center, Room 292, 1000 East University Avenue, Department 3375, Laramie, WY 82071; [email protected]

CASE › Carl L, a 57-year-old obese patient (body mass index [BMI] >40) who had not been to a doctor in a decade, comes to see you after a health fair screening revealed dyslipidemia (low-density lipoprotein [LDL] cholesterol, 188 mg/dL; high-density lipoprotein cholesterol [HDL], 32 mg/dL; total cholesterol, 240 mg/dL; triglycerides, 100 mg/dL). His blood pressure (BP) is 146/90 mm Hg and his fasting glucose is 101 mg/dL. Labs drawn that day reveal a glycated hemoglobin (HbA1c) of 5.9%, alanine aminotransferase (ALT) of 45 U/L, and aspartate aminotransferase (AST) of 62 U/L. In taking his history, you discover that Mr. L also has a prominent family history of heart disease.

Mr. L agrees to take a low-dose statin, and you prescribe atorvastatin 10 mg and a thiazide diuretic. You advise the patient to contact you immediately if he develops significant myalgia, jaundice, dark urine, or symptoms of hyperglycemia such as excessive thirst or urination, and schedule a follow-up visit in 8 weeks.

Long recognized as the bedrock of hyperlipidemia therapy, statins achieved even greater prominence when the American College of Cardiology/American Heart Association (ACC/AHA) issued a new cholesterol guideline1 late last year. The ACC and AHA now recommend statins for a wider range of patients, often at a higher starting dose. (To read about the controversy the recommendations generated, see “The new cholesterol guideline: Beyond the headlines,” J Fam Pract. 2013;62:730.)

Based on the new recommendations, the use of statins is likely to rise.² (A statin [rosuvastatin] is already the nation’s most widely prescribed medication.²) Thus, it is more important than ever for physicians to be knowledgeable about the risks associated with statins and able to assess the benefits of therapy for individual patients.

A 2013 retrospective cohort study of >100,000 patients on statins found that 17% developed adverse effects (AEs). Therapy was withheld, at least temporarily, for 10% of study participants (60% of those experiencing AEs).³ At the same time, the authors of a large meta-analysis (135 randomized controlled trials [RCTs] and >240,000 patients) reported that AEs associated with statins as a class were uncommon. The meta-analysis also found that the overall discontinuation rate for statin users—5.7%—was not significantly different from that of patients on placebo.⁴

Such discrepancies regarding particular risks as well as the overall incidence of AEs and discontinuation rates make the evidence difficult to sort out. We created this update with that in mind.

Musculoskeletal symptoms
 are most common


Studies indicate that exercise increases the risk of statin-induced myalgia and that patients taking statins are more prone to exercise-related injury.Skeletal muscle symptoms are the most common AEs reported by patients taking statins.⁵ These range from muscle weakness, fatigue, and pain to (rarely) rhabdomyolysis—a life-threatening condition characterized by severe muscle pain, muscle weakness, a 10-fold increase in creatine kinase (CK), and increased serum creatinine, often with myoglobinuria.⁵

Patients with myopathy—an umbrella term for any muscle disease—may report stiffness, weakness, tenderness, soreness, cramping, or heaviness. Symptoms usually are symmetrical and often involve the proximal limbs and trunk.6 Studies indicate that exercise increases the risk of statin-induced myalgia—muscle pain or weakness without an increase in CK—and that patients taking statins are more prone to exercise-related injury.^7,8

A baseline CK is recommended for patients with an increased risk of muscular disorders.¹ Risk factors include a personal or family history of statin intolerance or muscle disease, age >75 years, low levels of vitamin D, and concomitant use of medications that may increase the risk of myopathy (TABLE 1).¹ Routine monitoring of CK is not recommended, but CK levels should be obtained for those who exhibit muscle symptoms while on statin therapy.¹

What the studies show

The incidence of myalgia reported in clinical studies is highly variable, ranging from <1% to 20%.^1,9,10 The ACC/AHA guideline reports only one additional case of myopathy per 10,000 statin users compared with those on placebo and cites a rhabdomyolysis occurrence rate of <.06% over 5 years.¹

A 2006 systematic review estimated the absolute risk of rhabdomyolysis to be 3.4 per 100,000 person-years, but the incidence was 10 times higher for patients taking both a statin and gemfibrozil.¹¹ (See TABLE 2^12,13 for more on drug-drug interactions.) But both the meta-analysis cited earlier⁴ and an earlier systematic review¹⁴ (35 RCTs and >74,000 patients) found that statins as a class do not increase the incidence of myalgia or rhabdomyolysis.

Differences in the way muscular disorders are defined has been suggested as one reason for the discrepancies.¹⁰ In addition, many clinical trials exclude patients at higher risk of statin-associated AEs, such as those with renal or hepatic insufficiency, prior muscular complaints, poorly controlled diabetes, or potential drug-drug interactions.¹

An FDA advisory. In a safety communication last updated in February 2012, the US Food and Drug Administration (FDA) cautioned against starting patients on the highest dose of simvastatin (80 mg).¹⁵ The warning is based on a large study (N=12,064) that found an increased risk of myopathy (0.9%) and rhabdomyolysis (0.2%) in patients on the 80-mg dose vs those taking 20 mg (0.02% and 0%, respectively).¹⁶

With the ACC/AHA now recommending intensive therapy (atorvastatin 40-80 mg or rosuvastatin 20-40 mg) to achieve an LDL reduction >50% for many patients,¹ it is important to be aware that this risk is specific to simvastatin. A recent meta-analysis of studies directly comparing patients receiving intensive statin therapy with those on low to moderate doses did not find any increased risk in rhabdomyolysis associated with more intensive therapy when those taking 80 mg simvastatin were excluded.¹⁷

THE BOTTOM LINE: Although rhabdomyolysis is rare, its severity—a fatality rate of 10%¹¹—makes it critical to educate patients about the disorder and instruct them to stop taking the statin and call the office immediately if they develop severe muscle pain or weakness.

Recommend CoQ10
 for statin-induced myopathy


Although the exact mechanism of statin-induced myopathy is unknown, the most likely explanation is a depletion of coenzyme Q10 (CoQ10), which has negative effects on mitochondrial energy production.¹⁸ While studies using CoQ10 to treat this AE have been small and had mixed results, the overall evidence suggests that it decreases the development and/or severity of symptoms.^18-20

In fact, CoQ10 supplementation is the only treatment that has shown promise in treating statin-induced muscle symptoms.^18-20 Doses of about 100 mg bid have been found to be beneficial and safe; no clinically relevant AEs have been seen with doses <300 mg/d.^18,20,21 A large placebo-controlled study is currently evaluating a 600 mg/d dose of CoQ10 in patients with statin-induced myopathy.¹⁹

CASE › On his next visit, Mr. L reports a new ache in his left shoulder and upper back, which he describes as mild, but annoying. He also tells you his memory seems to be getting worse and that he has developed an odd tingling in his hands. These symptoms began about a month after he started the medications, Mr. L says. He also began a new exercise program, but his BMI is unchanged.

On examination, you find the affected shoulder and upper back modestly and diffusely tender to palpation, but with no decline in strength. Mr. L’s BP has fallen to 134/84 mm Hg and his fasting glucose is 105 mg/dL. Lab tests reveal an LDL of 144 mg/dL and HDL of 36 mg/dL, HbA1c of 6.1%, ALT of 105 U/L, AST of 61 U/L, and a normal CK.

You recommend 100 mg CoQ10 bid. Because it is available only over the counter, you advise the patient to look for a product whose purity and potency have been verified by an external source, such as the US Pharmacopeial Convention. You also prescribe metformin 500 mg bid for insulin resistance, refer the patient to a nutritionist and diabetes specialist, and order tests to evaluate his other symptoms.

Hepatic effects are rare

CoQ10 supplementation is the only treatment that has shown promise in treating statin-induced muscle symptoms.Historically, statins have been linked to potential hepatotoxicity, with case reports of serum transaminase elevation, cholestasis, hepatitis, and acute liver failure. It is now recognized that hepatic AEs are rare and that statins are not associated with a risk of acute or chronic liver failure.^1,11 In patients with coronary heart disease, the incidence of hepatotoxicity with statin use is reported to be <1.5% over the course of 5 years, and appears to be dose-dependent.¹

In 2012, the FDA revised the labeling for most statins, relaxing its earlier recommendations for monitoring of liver function, clarifying the risk of myopathy, and providing additional information about drug-drug interactions.¹³

Checking transaminase levels before initiating therapy is recommended by both the ACC/AHA and FDA.^1,13 Routine monitoring is not necessary, the ACC/AHA guideline states, because RCTs have found little evidence of ALT/AST elevation.¹ But here, too, evidence varies. An older meta-analysis (13 trials and nearly 50,000 participants) concluded that as a class, statins have no greater risk of transaminase elevations than placebo.²² But the 135-RCT meta-analysis⁴ found otherwise: Statins did increase the risk of transaminase elevation (odds ratio [OR], 1.51; 95% confidence interval [CI], 1.24-1.84) compared with placebo, with differences associated with particular drugs and higher doses associated with more clinically significant elevations.⁴ It is important to note, however, that there was significant heterogeneity among the studies and no consistent definition of clinical significance.

THE BOTTOM LINE: Statins have been shown in multiple prospective studies to be safe for patients with chronic liver disease.^22,23

Statin use and diabetes:
 Is there a link?


Recent studies have found an increased risk of new-onset type 2 diabetes in statin users, with a greater risk associated with higher potency statins, including rosuvastatin and atorvastatin.^4,24 Although the exact mechanism is not known, statins may modify insulin signaling in peripheral tissues or directly impair insulin secretion.

The ACC/AHA guideline reports an excess rate of diabetes of one per 1000 patient-years for moderate-intensity therapy and 3 per 1000 patient years for high-intensity therapy.¹ The 2013 meta-analysis found that the elevated risk of diabetes was relatively small (OR=1.09; 95% CI, 1.02-1.16).⁴ No difference among various statins was found.

In another meta-analysis—this one encompassing 17 RCTs and >110,000 patients—no statistically significant difference in the incidence of new-onset diabetes was seen based on either the specific statin being taken or the intensity of therapy (high vs moderate).²⁴

THE BOTTOM LINE: Physicians should monitor patients taking statins for signs and symptoms of hyperglycemia.

Statins may be renoprotective

Statin use has been found to be associated with an increased risk of tubular proteinuria—an effect that is both dose- and potency-dependent.²⁵ Nonetheless, it has been suggested that statins may be a rare example of a drug class that is renoprotective in the long term, despite having an increased rate of proteinuria in the short term.²⁵

The evidence? In prospective studies, statin therapy has been shown to slow the progression of kidney disease in diverse patient populations, including renal transplant recipients and those with chronic kidney disease (CKD).^26,27

The Kidney Expert Panel of the National Lipid Association (NLA) has concluded that statins do not appear to cause significant proteinuria or acute kidney injury. The panel does not recommend routine monitoring for proteinuria or kidney function in statin users unless otherwise indicated, but does recommend a lower dose for patients with CKD.²⁸

THE BOTTOM LINE: Kidney Disease Improving Global Outcomes guidelines recommend that patients who have CKD, but are not on dialysis, be treated with statin therapy. Statins are contraindicated for patients on dialysis, as clinical trials have failed to show significant cardiovascular benefit.²⁹

Intracerebral hemorrhage: 
Statins increase recurrence risk


Routine monitoring 
of transaminase levels is not necessary, according to the ACC/AHA guideline.In recent years there has been considerable concern about a statin-induced increased risk for intracerebral hemorrhage (ICH). In a major prospective study in which patients were put on high-dose statin therapy or placebo after an acute ischemic or hemorrhagic stroke, the overall incidence of a recurrent stroke was significantly lower in the statin group.30 Among those who’d had an ICH, however, the recurrence rate was 73% higher for patients taking statins.

A subanalysis that looked only at patients who’d had a hemorrhagic stroke as their initial event (n=93) found that the absolute risk of recurrent ICH was 15.6% for patients randomized to atorvastatin vs 4.2% for those on placebo.³¹ Despite being based on a small subset of the original study group, the increased risk was statistically significant in multivariate analysis (hazard ratio [HR]=1.69; 95% CI, 1.1-2.6).

A subsequent decision analysis study based on these results proposed that patients with a history of spontaneous deep ICH would need an exceedingly high 10-year cardiovascular event risk (>40%) for the benefits of statin therapy to outweigh the risk.³² The risk is particularly high for those with a history of lobar ICH, which has an extremely high recurrence rate. However, subsequent retrospective and observational studies have found that patients who were already on statins when the ICH occurred had less severe strokes and more favorable outcomes, with a lower mortality rate at 90 days post-ICH.^33-35

A 2010 ICH guideline from the AHA/American Stroke Association states that there is “insufficient data to recommend restrictions on use of statin agents” for patients who have had an ICH.³⁶

THE BOTTOM LINE: Physicians should carefully evaluate the anticipated cardiovascular risk for patients who have had a hemorrhagic stroke to determine whether statin therapy would be beneficial.

Other serious adverse effects: Which reports are accurate?


Statin use has been associated with a number of other serious AEs. Some reports appear to be accurate; others do not hold up after a close look at the evidence.

Malignancy. A potential link between statins and an increased risk of malignancy has been considered for years. A large trial (N=5804) from 2002 found a correlation between pravastatin and an increased risk of new cancer diagnoses compared with placebo (HR=1.25; 95% CI, 1.04-1.51; P=.02).³⁷ But a 10-year follow-up did not substantiate this finding, and it is now believed that the original result may have been due to chance.³⁸ Numerous other meta-analyses and systematic reviews have found no link between statin use and malignancy.^39-41

Cataracts. Potential ocular effects have been widely studied and debated in recent years. Observational studies reporting an association between statin use and cataracts have had conflicting results, with some showing statins as protective^42-45 and others finding an increased risk.^46,47 However, a recent propensity-score matched analysis found that statin users do indeed have an increased risk of developing cataracts.⁴⁸ The authors concluded that for primary prevention, the risk-benefit equation for statin use should include this added risk.⁴⁸

In addition, a review of the databases of the National Registry of Drug-Induced Ocular Side Effects, the World Health Organization, and the FDA from 1987 to 2008 indicates that statin therapy may also cause diplopia, ptosis, and ophthalmoplegia.⁴⁹

Peripheral neuropathy. Despite case reports of statin-induced peripheral neuropathy, the NLA’s Neurology Expert Panel states that statins do not appear to cause this condition. If a patient receiving statin therapy develops peripheral neuropathy, a full work-up for other causes should be initiated before considering a modification of statin therapy, the panel advises.²⁸

Statins have also been linked to headache and dizziness, respiratory symptoms, gastrointestinal problems, and rash, among other AEs (TABLE 3).⁵⁰

Which drug? Potential differences in statins


The meta-analysis with >240,000 participants evaluated patients taking 7 different statins (atorvastatin, fluvastatin, lovastatin, pravastatin, pitavastatin, rosuvastatin, and simvastatin), looking at AEs of the drugs both collectively and individually.4 As noted earlier, the overall discontinuation rate due to AEs for all statins was 5.7%. Discontinuation rates for each agent were not reported.⁴

A recent propensity-score matched analysis found that statin users have an increased risk of developing cataracts.The researchers did report, however, that atorvastatin and rosuvastatin had the highest discontinuation rates; atorvastatin and fluvastatin had the highest incidence of transaminase elevations (OR, 2.6 and 5.2, respectively); and pravastatin and simvastatin appeared to be the best tolerated and safest statins, with the lowest discontinuation rates. However, higher doses of simvastatin (>40 mg/d) significantly increased the risk of CK and transaminase elevations (OR, 4.1 and 2.8, respectively),⁴ as well as the risk of rhabdomyolysis at the highest dose.^15,16

Are statins safe for these patients?

When considering statin therapy, there are some patient populations that warrant particular concern:

Women of childbearing age. Statins are contraindicated in women who are pregnant or breastfeeding,¹ and should not be initiated in women who are trying to conceive.

Children and adolescents (ages 8-18 years). Statins have been shown to be safe and effective for children and adolescents with familial hyperlipidemia. No effect on growth or maturation has been seen.⁵¹ As with adults, however, higher statin doses and the use of concomitant interacting drugs increase the risk of AEs.

Asians. The new ACC/AHA guideline suggests taking Asian ancestry into consideration when prescribing statins because Asians may be more sensitive to medications metabolized by the CYP450 system.¹ However, there are no reports of an increased risk of AEs in Asian patients on statins.⁵² (To read more about statin use in particular patient populations, see “Statin therapy: When to think twice,” J Fam Pract. 2013;62:726-732.)

Patient factors that increase risk

Risk factors for statin-induced AEs include:¹

• multiple and/or serious comorbidities (eg, hypothyroidism, impaired renal or hepatic function, rheumatic disorders)
• unexplained ALT elevation >3 times the 
upper limit of normal
• history of prior statin intolerance or concomitant use of drugs that affect statin 
metabolism
• age >75 years
• preexisting muscle disorders
• low vitamin D levels.


If a patient who would clearly benefit from statin therapy develops an AE requiring discontinuation, a retrial—with the same drug or a different statin—is generally recommended once the symptoms resolve.¹

Statins have been shown to be safe and effective for children and adolescents with familial hyperlipidemia.CASE › The risk of elevated serum transaminases, insulin resistance, cognitive impairment, and neuropathy associated with statin use is minimal, and further evaluation revealed that Mr. L’s recent symptoms had other causes. The elevated transaminases were due to fatty liver disease, the cognitive impairment was secondary to sleep apnea (both linked to his obesity), and the tingling in his hands was the result of carpal tunnel syndrome caused by his exercise regimen.

When he returns in 6 months, Mr. L reports that he has been working with both a nutritionist and an athletic trainer. He has sustained a 15-lb weight loss. He is still taking atorvastatin 10 mg; after he began taking CoQ10, his muscle pain resolved. The patient’s cholesterol and transaminase levels are normal, and the cognitive impairment and peripheral neuropathy he reported at his last visit have improved significantly.

CORRESPONDENCE
Tracy D. Mahvan, PharmD, University of Wyoming, School of Pharmacy Health Sciences Center, Room 292, 1000 E. University Avenue, Laramie, WY 82071; [email protected]

CASE › Carl L, a 57-year-old obese patient (body mass index [BMI] >40) who had not been to a doctor in a decade, comes to see you after a health fair screening revealed dyslipidemia (low-density lipoprotein [LDL] cholesterol, 188 mg/dL; high-density lipoprotein cholesterol [HDL], 32 mg/dL; total cholesterol, 240 mg/dL; triglycerides, 100 mg/dL). His blood pressure (BP) is 146/90 mm Hg and his fasting glucose is 101 mg/dL. Labs drawn that day reveal a glycated hemoglobin (HbA1c) of 5.9%, alanine aminotransferase (ALT) of 45 U/L, and aspartate aminotransferase (AST) of 62 U/L. In taking his history, you discover that Mr. L also has a prominent family history of heart disease.

Mr. L agrees to take a low-dose statin, and you prescribe atorvastatin 10 mg and a thiazide diuretic. You advise the patient to contact you immediately if he develops significant myalgia, jaundice, dark urine, or symptoms of hyperglycemia such as excessive thirst or urination, and schedule a follow-up visit in 8 weeks.

Long recognized as the bedrock of hyperlipidemia therapy, statins achieved even greater prominence when the American College of Cardiology/American Heart Association (ACC/AHA) issued a new cholesterol guideline1 late last year. The ACC and AHA now recommend statins for a wider range of patients, often at a higher starting dose. (To read about the controversy the recommendations generated, see “The new cholesterol guideline: Beyond the headlines,” J Fam Pract. 2013;62:730.)

Based on the new recommendations, the use of statins is likely to rise.² (A statin [rosuvastatin] is already the nation’s most widely prescribed medication.²) Thus, it is more important than ever for physicians to be knowledgeable about the risks associated with statins and able to assess the benefits of therapy for individual patients.

A 2013 retrospective cohort study of >100,000 patients on statins found that 17% developed adverse effects (AEs). Therapy was withheld, at least temporarily, for 10% of study participants (60% of those experiencing AEs).³ At the same time, the authors of a large meta-analysis (135 randomized controlled trials [RCTs] and >240,000 patients) reported that AEs associated with statins as a class were uncommon. The meta-analysis also found that the overall discontinuation rate for statin users—5.7%—was not significantly different from that of patients on placebo.⁴

Such discrepancies regarding particular risks as well as the overall incidence of AEs and discontinuation rates make the evidence difficult to sort out. We created this update with that in mind.

Musculoskeletal symptoms
 are most common


Studies indicate that exercise increases the risk of statin-induced myalgia and that patients taking statins are more prone to exercise-related injury.Skeletal muscle symptoms are the most common AEs reported by patients taking statins.⁵ These range from muscle weakness, fatigue, and pain to (rarely) rhabdomyolysis—a life-threatening condition characterized by severe muscle pain, muscle weakness, a 10-fold increase in creatine kinase (CK), and increased serum creatinine, often with myoglobinuria.⁵

Patients with myopathy—an umbrella term for any muscle disease—may report stiffness, weakness, tenderness, soreness, cramping, or heaviness. Symptoms usually are symmetrical and often involve the proximal limbs and trunk.6 Studies indicate that exercise increases the risk of statin-induced myalgia—muscle pain or weakness without an increase in CK—and that patients taking statins are more prone to exercise-related injury.^7,8

A baseline CK is recommended for patients with an increased risk of muscular disorders.¹ Risk factors include a personal or family history of statin intolerance or muscle disease, age >75 years, low levels of vitamin D, and concomitant use of medications that may increase the risk of myopathy (TABLE 1).¹ Routine monitoring of CK is not recommended, but CK levels should be obtained for those who exhibit muscle symptoms while on statin therapy.¹

What the studies show

The incidence of myalgia reported in clinical studies is highly variable, ranging from <1% to 20%.^1,9,10 The ACC/AHA guideline reports only one additional case of myopathy per 10,000 statin users compared with those on placebo and cites a rhabdomyolysis occurrence rate of <.06% over 5 years.¹

A 2006 systematic review estimated the absolute risk of rhabdomyolysis to be 3.4 per 100,000 person-years, but the incidence was 10 times higher for patients taking both a statin and gemfibrozil.¹¹ (See TABLE 2^12,13 for more on drug-drug interactions.) But both the meta-analysis cited earlier⁴ and an earlier systematic review¹⁴ (35 RCTs and >74,000 patients) found that statins as a class do not increase the incidence of myalgia or rhabdomyolysis.

Differences in the way muscular disorders are defined has been suggested as one reason for the discrepancies.¹⁰ In addition, many clinical trials exclude patients at higher risk of statin-associated AEs, such as those with renal or hepatic insufficiency, prior muscular complaints, poorly controlled diabetes, or potential drug-drug interactions.¹

An FDA advisory. In a safety communication last updated in February 2012, the US Food and Drug Administration (FDA) cautioned against starting patients on the highest dose of simvastatin (80 mg).¹⁵ The warning is based on a large study (N=12,064) that found an increased risk of myopathy (0.9%) and rhabdomyolysis (0.2%) in patients on the 80-mg dose vs those taking 20 mg (0.02% and 0%, respectively).¹⁶

With the ACC/AHA now recommending intensive therapy (atorvastatin 40-80 mg or rosuvastatin 20-40 mg) to achieve an LDL reduction >50% for many patients,¹ it is important to be aware that this risk is specific to simvastatin. A recent meta-analysis of studies directly comparing patients receiving intensive statin therapy with those on low to moderate doses did not find any increased risk in rhabdomyolysis associated with more intensive therapy when those taking 80 mg simvastatin were excluded.¹⁷

THE BOTTOM LINE: Although rhabdomyolysis is rare, its severity—a fatality rate of 10%¹¹—makes it critical to educate patients about the disorder and instruct them to stop taking the statin and call the office immediately if they develop severe muscle pain or weakness.

Recommend CoQ10
 for statin-induced myopathy


Although the exact mechanism of statin-induced myopathy is unknown, the most likely explanation is a depletion of coenzyme Q10 (CoQ10), which has negative effects on mitochondrial energy production.¹⁸ While studies using CoQ10 to treat this AE have been small and had mixed results, the overall evidence suggests that it decreases the development and/or severity of symptoms.^18-20

In fact, CoQ10 supplementation is the only treatment that has shown promise in treating statin-induced muscle symptoms.^18-20 Doses of about 100 mg bid have been found to be beneficial and safe; no clinically relevant AEs have been seen with doses <300 mg/d.^18,20,21 A large placebo-controlled study is currently evaluating a 600 mg/d dose of CoQ10 in patients with statin-induced myopathy.¹⁹

CASE › On his next visit, Mr. L reports a new ache in his left shoulder and upper back, which he describes as mild, but annoying. He also tells you his memory seems to be getting worse and that he has developed an odd tingling in his hands. These symptoms began about a month after he started the medications, Mr. L says. He also began a new exercise program, but his BMI is unchanged.

On examination, you find the affected shoulder and upper back modestly and diffusely tender to palpation, but with no decline in strength. Mr. L’s BP has fallen to 134/84 mm Hg and his fasting glucose is 105 mg/dL. Lab tests reveal an LDL of 144 mg/dL and HDL of 36 mg/dL, HbA1c of 6.1%, ALT of 105 U/L, AST of 61 U/L, and a normal CK.

You recommend 100 mg CoQ10 bid. Because it is available only over the counter, you advise the patient to look for a product whose purity and potency have been verified by an external source, such as the US Pharmacopeial Convention. You also prescribe metformin 500 mg bid for insulin resistance, refer the patient to a nutritionist and diabetes specialist, and order tests to evaluate his other symptoms.

Hepatic effects are rare

CoQ10 supplementation is the only treatment that has shown promise in treating statin-induced muscle symptoms.Historically, statins have been linked to potential hepatotoxicity, with case reports of serum transaminase elevation, cholestasis, hepatitis, and acute liver failure. It is now recognized that hepatic AEs are rare and that statins are not associated with a risk of acute or chronic liver failure.^1,11 In patients with coronary heart disease, the incidence of hepatotoxicity with statin use is reported to be <1.5% over the course of 5 years, and appears to be dose-dependent.¹

In 2012, the FDA revised the labeling for most statins, relaxing its earlier recommendations for monitoring of liver function, clarifying the risk of myopathy, and providing additional information about drug-drug interactions.¹³

Checking transaminase levels before initiating therapy is recommended by both the ACC/AHA and FDA.^1,13 Routine monitoring is not necessary, the ACC/AHA guideline states, because RCTs have found little evidence of ALT/AST elevation.¹ But here, too, evidence varies. An older meta-analysis (13 trials and nearly 50,000 participants) concluded that as a class, statins have no greater risk of transaminase elevations than placebo.²² But the 135-RCT meta-analysis⁴ found otherwise: Statins did increase the risk of transaminase elevation (odds ratio [OR], 1.51; 95% confidence interval [CI], 1.24-1.84) compared with placebo, with differences associated with particular drugs and higher doses associated with more clinically significant elevations.⁴ It is important to note, however, that there was significant heterogeneity among the studies and no consistent definition of clinical significance.

THE BOTTOM LINE: Statins have been shown in multiple prospective studies to be safe for patients with chronic liver disease.^22,23

Statin use and diabetes:
 Is there a link?


Recent studies have found an increased risk of new-onset type 2 diabetes in statin users, with a greater risk associated with higher potency statins, including rosuvastatin and atorvastatin.^4,24 Although the exact mechanism is not known, statins may modify insulin signaling in peripheral tissues or directly impair insulin secretion.

The ACC/AHA guideline reports an excess rate of diabetes of one per 1000 patient-years for moderate-intensity therapy and 3 per 1000 patient years for high-intensity therapy.¹ The 2013 meta-analysis found that the elevated risk of diabetes was relatively small (OR=1.09; 95% CI, 1.02-1.16).⁴ No difference among various statins was found.

In another meta-analysis—this one encompassing 17 RCTs and >110,000 patients—no statistically significant difference in the incidence of new-onset diabetes was seen based on either the specific statin being taken or the intensity of therapy (high vs moderate).²⁴

THE BOTTOM LINE: Physicians should monitor patients taking statins for signs and symptoms of hyperglycemia.

Statins may be renoprotective

Statin use has been found to be associated with an increased risk of tubular proteinuria—an effect that is both dose- and potency-dependent.²⁵ Nonetheless, it has been suggested that statins may be a rare example of a drug class that is renoprotective in the long term, despite having an increased rate of proteinuria in the short term.²⁵

The evidence? In prospective studies, statin therapy has been shown to slow the progression of kidney disease in diverse patient populations, including renal transplant recipients and those with chronic kidney disease (CKD).^26,27

The Kidney Expert Panel of the National Lipid Association (NLA) has concluded that statins do not appear to cause significant proteinuria or acute kidney injury. The panel does not recommend routine monitoring for proteinuria or kidney function in statin users unless otherwise indicated, but does recommend a lower dose for patients with CKD.²⁸

THE BOTTOM LINE: Kidney Disease Improving Global Outcomes guidelines recommend that patients who have CKD, but are not on dialysis, be treated with statin therapy. Statins are contraindicated for patients on dialysis, as clinical trials have failed to show significant cardiovascular benefit.²⁹

Intracerebral hemorrhage: 
Statins increase recurrence risk


Routine monitoring 
of transaminase levels is not necessary, according to the ACC/AHA guideline.In recent years there has been considerable concern about a statin-induced increased risk for intracerebral hemorrhage (ICH). In a major prospective study in which patients were put on high-dose statin therapy or placebo after an acute ischemic or hemorrhagic stroke, the overall incidence of a recurrent stroke was significantly lower in the statin group.30 Among those who’d had an ICH, however, the recurrence rate was 73% higher for patients taking statins.

A subanalysis that looked only at patients who’d had a hemorrhagic stroke as their initial event (n=93) found that the absolute risk of recurrent ICH was 15.6% for patients randomized to atorvastatin vs 4.2% for those on placebo.³¹ Despite being based on a small subset of the original study group, the increased risk was statistically significant in multivariate analysis (hazard ratio [HR]=1.69; 95% CI, 1.1-2.6).

A subsequent decision analysis study based on these results proposed that patients with a history of spontaneous deep ICH would need an exceedingly high 10-year cardiovascular event risk (>40%) for the benefits of statin therapy to outweigh the risk.³² The risk is particularly high for those with a history of lobar ICH, which has an extremely high recurrence rate. However, subsequent retrospective and observational studies have found that patients who were already on statins when the ICH occurred had less severe strokes and more favorable outcomes, with a lower mortality rate at 90 days post-ICH.^33-35

A 2010 ICH guideline from the AHA/American Stroke Association states that there is “insufficient data to recommend restrictions on use of statin agents” for patients who have had an ICH.³⁶

THE BOTTOM LINE: Physicians should carefully evaluate the anticipated cardiovascular risk for patients who have had a hemorrhagic stroke to determine whether statin therapy would be beneficial.

Other serious adverse effects: Which reports are accurate?


Statin use has been associated with a number of other serious AEs. Some reports appear to be accurate; others do not hold up after a close look at the evidence.

Malignancy. A potential link between statins and an increased risk of malignancy has been considered for years. A large trial (N=5804) from 2002 found a correlation between pravastatin and an increased risk of new cancer diagnoses compared with placebo (HR=1.25; 95% CI, 1.04-1.51; P=.02).³⁷ But a 10-year follow-up did not substantiate this finding, and it is now believed that the original result may have been due to chance.³⁸ Numerous other meta-analyses and systematic reviews have found no link between statin use and malignancy.^39-41

Cataracts. Potential ocular effects have been widely studied and debated in recent years. Observational studies reporting an association between statin use and cataracts have had conflicting results, with some showing statins as protective^42-45 and others finding an increased risk.^46,47 However, a recent propensity-score matched analysis found that statin users do indeed have an increased risk of developing cataracts.⁴⁸ The authors concluded that for primary prevention, the risk-benefit equation for statin use should include this added risk.⁴⁸

In addition, a review of the databases of the National Registry of Drug-Induced Ocular Side Effects, the World Health Organization, and the FDA from 1987 to 2008 indicates that statin therapy may also cause diplopia, ptosis, and ophthalmoplegia.⁴⁹

Peripheral neuropathy. Despite case reports of statin-induced peripheral neuropathy, the NLA’s Neurology Expert Panel states that statins do not appear to cause this condition. If a patient receiving statin therapy develops peripheral neuropathy, a full work-up for other causes should be initiated before considering a modification of statin therapy, the panel advises.²⁸

Statins have also been linked to headache and dizziness, respiratory symptoms, gastrointestinal problems, and rash, among other AEs (TABLE 3).⁵⁰

Which drug? Potential differences in statins


The meta-analysis with >240,000 participants evaluated patients taking 7 different statins (atorvastatin, fluvastatin, lovastatin, pravastatin, pitavastatin, rosuvastatin, and simvastatin), looking at AEs of the drugs both collectively and individually.4 As noted earlier, the overall discontinuation rate due to AEs for all statins was 5.7%. Discontinuation rates for each agent were not reported.⁴

A recent propensity-score matched analysis found that statin users have an increased risk of developing cataracts.The researchers did report, however, that atorvastatin and rosuvastatin had the highest discontinuation rates; atorvastatin and fluvastatin had the highest incidence of transaminase elevations (OR, 2.6 and 5.2, respectively); and pravastatin and simvastatin appeared to be the best tolerated and safest statins, with the lowest discontinuation rates. However, higher doses of simvastatin (>40 mg/d) significantly increased the risk of CK and transaminase elevations (OR, 4.1 and 2.8, respectively),⁴ as well as the risk of rhabdomyolysis at the highest dose.^15,16

Are statins safe for these patients?

When considering statin therapy, there are some patient populations that warrant particular concern:

Women of childbearing age. Statins are contraindicated in women who are pregnant or breastfeeding,¹ and should not be initiated in women who are trying to conceive.

Children and adolescents (ages 8-18 years). Statins have been shown to be safe and effective for children and adolescents with familial hyperlipidemia. No effect on growth or maturation has been seen.⁵¹ As with adults, however, higher statin doses and the use of concomitant interacting drugs increase the risk of AEs.

Asians. The new ACC/AHA guideline suggests taking Asian ancestry into consideration when prescribing statins because Asians may be more sensitive to medications metabolized by the CYP450 system.¹ However, there are no reports of an increased risk of AEs in Asian patients on statins.⁵² (To read more about statin use in particular patient populations, see “Statin therapy: When to think twice,” J Fam Pract. 2013;62:726-732.)

Patient factors that increase risk

Risk factors for statin-induced AEs include:¹

• multiple and/or serious comorbidities (eg, hypothyroidism, impaired renal or hepatic function, rheumatic disorders)
• unexplained ALT elevation >3 times the 
upper limit of normal
• history of prior statin intolerance or concomitant use of drugs that affect statin 
metabolism
• age >75 years
• preexisting muscle disorders
• low vitamin D levels.


If a patient who would clearly benefit from statin therapy develops an AE requiring discontinuation, a retrial—with the same drug or a different statin—is generally recommended once the symptoms resolve.¹

Statins have been shown to be safe and effective for children and adolescents with familial hyperlipidemia.CASE › The risk of elevated serum transaminases, insulin resistance, cognitive impairment, and neuropathy associated with statin use is minimal, and further evaluation revealed that Mr. L’s recent symptoms had other causes. The elevated transaminases were due to fatty liver disease, the cognitive impairment was secondary to sleep apnea (both linked to his obesity), and the tingling in his hands was the result of carpal tunnel syndrome caused by his exercise regimen.

When he returns in 6 months, Mr. L reports that he has been working with both a nutritionist and an athletic trainer. He has sustained a 15-lb weight loss. He is still taking atorvastatin 10 mg; after he began taking CoQ10, his muscle pain resolved. The patient’s cholesterol and transaminase levels are normal, and the cognitive impairment and peripheral neuropathy he reported at his last visit have improved significantly.

CORRESPONDENCE
Tracy D. Mahvan, PharmD, University of Wyoming, School of Pharmacy Health Sciences Center, Room 292, 1000 E. University Avenue, Laramie, WY 82071; [email protected]

CASE › Carl L, a 57-year-old obese patient (body mass index [BMI] >40) who had not been to a doctor in a decade, comes to see you after a health fair screening revealed dyslipidemia (low-density lipoprotein [LDL] cholesterol, 188 mg/dL; high-density lipoprotein cholesterol [HDL], 32 mg/dL; total cholesterol, 240 mg/dL; triglycerides, 100 mg/dL). His blood pressure (BP) is 146/90 mm Hg and his fasting glucose is 101 mg/dL. Labs drawn that day reveal a glycated hemoglobin (HbA1c) of 5.9%, alanine aminotransferase (ALT) of 45 U/L, and aspartate aminotransferase (AST) of 62 U/L. In taking his history, you discover that Mr. L also has a prominent family history of heart disease.

Mr. L agrees to take a low-dose statin, and you prescribe atorvastatin 10 mg and a thiazide diuretic. You advise the patient to contact you immediately if he develops significant myalgia, jaundice, dark urine, or symptoms of hyperglycemia such as excessive thirst or urination, and schedule a follow-up visit in 8 weeks.

Long recognized as the bedrock of hyperlipidemia therapy, statins achieved even greater prominence when the American College of Cardiology/American Heart Association (ACC/AHA) issued a new cholesterol guideline1 late last year. The ACC and AHA now recommend statins for a wider range of patients, often at a higher starting dose. (To read about the controversy the recommendations generated, see “The new cholesterol guideline: Beyond the headlines,” J Fam Pract. 2013;62:730.)

Based on the new recommendations, the use of statins is likely to rise.² (A statin [rosuvastatin] is already the nation’s most widely prescribed medication.²) Thus, it is more important than ever for physicians to be knowledgeable about the risks associated with statins and able to assess the benefits of therapy for individual patients.

A 2013 retrospective cohort study of >100,000 patients on statins found that 17% developed adverse effects (AEs). Therapy was withheld, at least temporarily, for 10% of study participants (60% of those experiencing AEs).³ At the same time, the authors of a large meta-analysis (135 randomized controlled trials [RCTs] and >240,000 patients) reported that AEs associated with statins as a class were uncommon. The meta-analysis also found that the overall discontinuation rate for statin users—5.7%—was not significantly different from that of patients on placebo.⁴

Such discrepancies regarding particular risks as well as the overall incidence of AEs and discontinuation rates make the evidence difficult to sort out. We created this update with that in mind.

Musculoskeletal symptoms
 are most common


Studies indicate that exercise increases the risk of statin-induced myalgia and that patients taking statins are more prone to exercise-related injury.Skeletal muscle symptoms are the most common AEs reported by patients taking statins.⁵ These range from muscle weakness, fatigue, and pain to (rarely) rhabdomyolysis—a life-threatening condition characterized by severe muscle pain, muscle weakness, a 10-fold increase in creatine kinase (CK), and increased serum creatinine, often with myoglobinuria.⁵

Patients with myopathy—an umbrella term for any muscle disease—may report stiffness, weakness, tenderness, soreness, cramping, or heaviness. Symptoms usually are symmetrical and often involve the proximal limbs and trunk.6 Studies indicate that exercise increases the risk of statin-induced myalgia—muscle pain or weakness without an increase in CK—and that patients taking statins are more prone to exercise-related injury.^7,8

A baseline CK is recommended for patients with an increased risk of muscular disorders.¹ Risk factors include a personal or family history of statin intolerance or muscle disease, age >75 years, low levels of vitamin D, and concomitant use of medications that may increase the risk of myopathy (TABLE 1).¹ Routine monitoring of CK is not recommended, but CK levels should be obtained for those who exhibit muscle symptoms while on statin therapy.¹

What the studies show

The incidence of myalgia reported in clinical studies is highly variable, ranging from <1% to 20%.^1,9,10 The ACC/AHA guideline reports only one additional case of myopathy per 10,000 statin users compared with those on placebo and cites a rhabdomyolysis occurrence rate of <.06% over 5 years.¹

A 2006 systematic review estimated the absolute risk of rhabdomyolysis to be 3.4 per 100,000 person-years, but the incidence was 10 times higher for patients taking both a statin and gemfibrozil.¹¹ (See TABLE 2^12,13 for more on drug-drug interactions.) But both the meta-analysis cited earlier⁴ and an earlier systematic review¹⁴ (35 RCTs and >74,000 patients) found that statins as a class do not increase the incidence of myalgia or rhabdomyolysis.

Differences in the way muscular disorders are defined has been suggested as one reason for the discrepancies.¹⁰ In addition, many clinical trials exclude patients at higher risk of statin-associated AEs, such as those with renal or hepatic insufficiency, prior muscular complaints, poorly controlled diabetes, or potential drug-drug interactions.¹

An FDA advisory. In a safety communication last updated in February 2012, the US Food and Drug Administration (FDA) cautioned against starting patients on the highest dose of simvastatin (80 mg).¹⁵ The warning is based on a large study (N=12,064) that found an increased risk of myopathy (0.9%) and rhabdomyolysis (0.2%) in patients on the 80-mg dose vs those taking 20 mg (0.02% and 0%, respectively).¹⁶

With the ACC/AHA now recommending intensive therapy (atorvastatin 40-80 mg or rosuvastatin 20-40 mg) to achieve an LDL reduction >50% for many patients,¹ it is important to be aware that this risk is specific to simvastatin. A recent meta-analysis of studies directly comparing patients receiving intensive statin therapy with those on low to moderate doses did not find any increased risk in rhabdomyolysis associated with more intensive therapy when those taking 80 mg simvastatin were excluded.¹⁷

THE BOTTOM LINE: Although rhabdomyolysis is rare, its severity—a fatality rate of 10%¹¹—makes it critical to educate patients about the disorder and instruct them to stop taking the statin and call the office immediately if they develop severe muscle pain or weakness.

Recommend CoQ10
 for statin-induced myopathy


Although the exact mechanism of statin-induced myopathy is unknown, the most likely explanation is a depletion of coenzyme Q10 (CoQ10), which has negative effects on mitochondrial energy production.¹⁸ While studies using CoQ10 to treat this AE have been small and had mixed results, the overall evidence suggests that it decreases the development and/or severity of symptoms.^18-20

In fact, CoQ10 supplementation is the only treatment that has shown promise in treating statin-induced muscle symptoms.^18-20 Doses of about 100 mg bid have been found to be beneficial and safe; no clinically relevant AEs have been seen with doses <300 mg/d.^18,20,21 A large placebo-controlled study is currently evaluating a 600 mg/d dose of CoQ10 in patients with statin-induced myopathy.¹⁹

CASE › On his next visit, Mr. L reports a new ache in his left shoulder and upper back, which he describes as mild, but annoying. He also tells you his memory seems to be getting worse and that he has developed an odd tingling in his hands. These symptoms began about a month after he started the medications, Mr. L says. He also began a new exercise program, but his BMI is unchanged.

On examination, you find the affected shoulder and upper back modestly and diffusely tender to palpation, but with no decline in strength. Mr. L’s BP has fallen to 134/84 mm Hg and his fasting glucose is 105 mg/dL. Lab tests reveal an LDL of 144 mg/dL and HDL of 36 mg/dL, HbA1c of 6.1%, ALT of 105 U/L, AST of 61 U/L, and a normal CK.

You recommend 100 mg CoQ10 bid. Because it is available only over the counter, you advise the patient to look for a product whose purity and potency have been verified by an external source, such as the US Pharmacopeial Convention. You also prescribe metformin 500 mg bid for insulin resistance, refer the patient to a nutritionist and diabetes specialist, and order tests to evaluate his other symptoms.

Hepatic effects are rare

CoQ10 supplementation is the only treatment that has shown promise in treating statin-induced muscle symptoms.Historically, statins have been linked to potential hepatotoxicity, with case reports of serum transaminase elevation, cholestasis, hepatitis, and acute liver failure. It is now recognized that hepatic AEs are rare and that statins are not associated with a risk of acute or chronic liver failure.^1,11 In patients with coronary heart disease, the incidence of hepatotoxicity with statin use is reported to be <1.5% over the course of 5 years, and appears to be dose-dependent.¹

In 2012, the FDA revised the labeling for most statins, relaxing its earlier recommendations for monitoring of liver function, clarifying the risk of myopathy, and providing additional information about drug-drug interactions.¹³

Checking transaminase levels before initiating therapy is recommended by both the ACC/AHA and FDA.^1,13 Routine monitoring is not necessary, the ACC/AHA guideline states, because RCTs have found little evidence of ALT/AST elevation.¹ But here, too, evidence varies. An older meta-analysis (13 trials and nearly 50,000 participants) concluded that as a class, statins have no greater risk of transaminase elevations than placebo.²² But the 135-RCT meta-analysis⁴ found otherwise: Statins did increase the risk of transaminase elevation (odds ratio [OR], 1.51; 95% confidence interval [CI], 1.24-1.84) compared with placebo, with differences associated with particular drugs and higher doses associated with more clinically significant elevations.⁴ It is important to note, however, that there was significant heterogeneity among the studies and no consistent definition of clinical significance.

THE BOTTOM LINE: Statins have been shown in multiple prospective studies to be safe for patients with chronic liver disease.^22,23

Statin use and diabetes:
 Is there a link?


Recent studies have found an increased risk of new-onset type 2 diabetes in statin users, with a greater risk associated with higher potency statins, including rosuvastatin and atorvastatin.^4,24 Although the exact mechanism is not known, statins may modify insulin signaling in peripheral tissues or directly impair insulin secretion.

The ACC/AHA guideline reports an excess rate of diabetes of one per 1000 patient-years for moderate-intensity therapy and 3 per 1000 patient years for high-intensity therapy.¹ The 2013 meta-analysis found that the elevated risk of diabetes was relatively small (OR=1.09; 95% CI, 1.02-1.16).⁴ No difference among various statins was found.

In another meta-analysis—this one encompassing 17 RCTs and >110,000 patients—no statistically significant difference in the incidence of new-onset diabetes was seen based on either the specific statin being taken or the intensity of therapy (high vs moderate).²⁴

THE BOTTOM LINE: Physicians should monitor patients taking statins for signs and symptoms of hyperglycemia.

Statins may be renoprotective

Statin use has been found to be associated with an increased risk of tubular proteinuria—an effect that is both dose- and potency-dependent.²⁵ Nonetheless, it has been suggested that statins may be a rare example of a drug class that is renoprotective in the long term, despite having an increased rate of proteinuria in the short term.²⁵

The evidence? In prospective studies, statin therapy has been shown to slow the progression of kidney disease in diverse patient populations, including renal transplant recipients and those with chronic kidney disease (CKD).^26,27

The Kidney Expert Panel of the National Lipid Association (NLA) has concluded that statins do not appear to cause significant proteinuria or acute kidney injury. The panel does not recommend routine monitoring for proteinuria or kidney function in statin users unless otherwise indicated, but does recommend a lower dose for patients with CKD.²⁸

THE BOTTOM LINE: Kidney Disease Improving Global Outcomes guidelines recommend that patients who have CKD, but are not on dialysis, be treated with statin therapy. Statins are contraindicated for patients on dialysis, as clinical trials have failed to show significant cardiovascular benefit.²⁹

Intracerebral hemorrhage: 
Statins increase recurrence risk


Routine monitoring 
of transaminase levels is not necessary, according to the ACC/AHA guideline.In recent years there has been considerable concern about a statin-induced increased risk for intracerebral hemorrhage (ICH). In a major prospective study in which patients were put on high-dose statin therapy or placebo after an acute ischemic or hemorrhagic stroke, the overall incidence of a recurrent stroke was significantly lower in the statin group.30 Among those who’d had an ICH, however, the recurrence rate was 73% higher for patients taking statins.

A subanalysis that looked only at patients who’d had a hemorrhagic stroke as their initial event (n=93) found that the absolute risk of recurrent ICH was 15.6% for patients randomized to atorvastatin vs 4.2% for those on placebo.³¹ Despite being based on a small subset of the original study group, the increased risk was statistically significant in multivariate analysis (hazard ratio [HR]=1.69; 95% CI, 1.1-2.6).

A subsequent decision analysis study based on these results proposed that patients with a history of spontaneous deep ICH would need an exceedingly high 10-year cardiovascular event risk (>40%) for the benefits of statin therapy to outweigh the risk.³² The risk is particularly high for those with a history of lobar ICH, which has an extremely high recurrence rate. However, subsequent retrospective and observational studies have found that patients who were already on statins when the ICH occurred had less severe strokes and more favorable outcomes, with a lower mortality rate at 90 days post-ICH.^33-35

A 2010 ICH guideline from the AHA/American Stroke Association states that there is “insufficient data to recommend restrictions on use of statin agents” for patients who have had an ICH.³⁶

THE BOTTOM LINE: Physicians should carefully evaluate the anticipated cardiovascular risk for patients who have had a hemorrhagic stroke to determine whether statin therapy would be beneficial.

Other serious adverse effects: Which reports are accurate?


Statin use has been associated with a number of other serious AEs. Some reports appear to be accurate; others do not hold up after a close look at the evidence.

Malignancy. A potential link between statins and an increased risk of malignancy has been considered for years. A large trial (N=5804) from 2002 found a correlation between pravastatin and an increased risk of new cancer diagnoses compared with placebo (HR=1.25; 95% CI, 1.04-1.51; P=.02).³⁷ But a 10-year follow-up did not substantiate this finding, and it is now believed that the original result may have been due to chance.³⁸ Numerous other meta-analyses and systematic reviews have found no link between statin use and malignancy.^39-41

Cataracts. Potential ocular effects have been widely studied and debated in recent years. Observational studies reporting an association between statin use and cataracts have had conflicting results, with some showing statins as protective^42-45 and others finding an increased risk.^46,47 However, a recent propensity-score matched analysis found that statin users do indeed have an increased risk of developing cataracts.⁴⁸ The authors concluded that for primary prevention, the risk-benefit equation for statin use should include this added risk.⁴⁸

In addition, a review of the databases of the National Registry of Drug-Induced Ocular Side Effects, the World Health Organization, and the FDA from 1987 to 2008 indicates that statin therapy may also cause diplopia, ptosis, and ophthalmoplegia.⁴⁹

Peripheral neuropathy. Despite case reports of statin-induced peripheral neuropathy, the NLA’s Neurology Expert Panel states that statins do not appear to cause this condition. If a patient receiving statin therapy develops peripheral neuropathy, a full work-up for other causes should be initiated before considering a modification of statin therapy, the panel advises.²⁸

Statins have also been linked to headache and dizziness, respiratory symptoms, gastrointestinal problems, and rash, among other AEs (TABLE 3).⁵⁰

Which drug? Potential differences in statins


The meta-analysis with >240,000 participants evaluated patients taking 7 different statins (atorvastatin, fluvastatin, lovastatin, pravastatin, pitavastatin, rosuvastatin, and simvastatin), looking at AEs of the drugs both collectively and individually.4 As noted earlier, the overall discontinuation rate due to AEs for all statins was 5.7%. Discontinuation rates for each agent were not reported.⁴

A recent propensity-score matched analysis found that statin users have an increased risk of developing cataracts.The researchers did report, however, that atorvastatin and rosuvastatin had the highest discontinuation rates; atorvastatin and fluvastatin had the highest incidence of transaminase elevations (OR, 2.6 and 5.2, respectively); and pravastatin and simvastatin appeared to be the best tolerated and safest statins, with the lowest discontinuation rates. However, higher doses of simvastatin (>40 mg/d) significantly increased the risk of CK and transaminase elevations (OR, 4.1 and 2.8, respectively),⁴ as well as the risk of rhabdomyolysis at the highest dose.^15,16

Are statins safe for these patients?

When considering statin therapy, there are some patient populations that warrant particular concern:

Women of childbearing age. Statins are contraindicated in women who are pregnant or breastfeeding,¹ and should not be initiated in women who are trying to conceive.

Children and adolescents (ages 8-18 years). Statins have been shown to be safe and effective for children and adolescents with familial hyperlipidemia. No effect on growth or maturation has been seen.⁵¹ As with adults, however, higher statin doses and the use of concomitant interacting drugs increase the risk of AEs.

Asians. The new ACC/AHA guideline suggests taking Asian ancestry into consideration when prescribing statins because Asians may be more sensitive to medications metabolized by the CYP450 system.¹ However, there are no reports of an increased risk of AEs in Asian patients on statins.⁵² (To read more about statin use in particular patient populations, see “Statin therapy: When to think twice,” J Fam Pract. 2013;62:726-732.)

Patient factors that increase risk

Risk factors for statin-induced AEs include:¹

• multiple and/or serious comorbidities (eg, hypothyroidism, impaired renal or hepatic function, rheumatic disorders)
• unexplained ALT elevation >3 times the 
upper limit of normal
• history of prior statin intolerance or concomitant use of drugs that affect statin 
metabolism
• age >75 years
• preexisting muscle disorders
• low vitamin D levels.


If a patient who would clearly benefit from statin therapy develops an AE requiring discontinuation, a retrial—with the same drug or a different statin—is generally recommended once the symptoms resolve.¹

Statins have been shown to be safe and effective for children and adolescents with familial hyperlipidemia.CASE › The risk of elevated serum transaminases, insulin resistance, cognitive impairment, and neuropathy associated with statin use is minimal, and further evaluation revealed that Mr. L’s recent symptoms had other causes. The elevated transaminases were due to fatty liver disease, the cognitive impairment was secondary to sleep apnea (both linked to his obesity), and the tingling in his hands was the result of carpal tunnel syndrome caused by his exercise regimen.

When he returns in 6 months, Mr. L reports that he has been working with both a nutritionist and an athletic trainer. He has sustained a 15-lb weight loss. He is still taking atorvastatin 10 mg; after he began taking CoQ10, his muscle pain resolved. The patient’s cholesterol and transaminase levels are normal, and the cognitive impairment and peripheral neuropathy he reported at his last visit have improved significantly.

CORRESPONDENCE
Tracy D. Mahvan, PharmD, University of Wyoming, School of Pharmacy Health Sciences Center, Room 292, 1000 E. University Avenue, Laramie, WY 82071; [email protected]

Yoga is practiced by 15.8 million Americans,¹ and is often recommended as therapy for a variety of medical conditions. However, the scientific literature on yoga is limited in scope and quality. This article presents good evidence for yoga as treatment for chronic back pain, depression, and anxiety, and fair evidence for treating asthma, symptoms of menopause, hypertension, and mobility issues in the elderly.

Yoga’s rising popularity as therapy

Yoga is a system of movement and breathing exercises meant to foster mind-body connection. Its roots are in ancient Indian practices codified by the writer Patanjali in the first or second century BCE.²

The practice of yoga was introduced to the Western world by a series of popular gurus from the 1930s to 1970s and consists primarily of asanas, or postures, and breathing exercises known as pranayama. Since then, yoga has been further subdivided into different schools and brands (TABLE^1,2), some of which are extremely taxing and vigorous and should be performed only by fit and healthy individuals, while others are gentle and accessible to anyone. Yoga has steadily gained in popularity, and nearly half of those who practice it say they do so to improve their health.¹

How useful is the research on yoga therapy?

Yoga has been a subject of Western scientific inquiry for more than 100 years. It has been deemed effective for treating conditions from hypertension to epilepsy,³ but many claims are poorly substantiated. Most studies report on a single case or series. The few investigational studies are mainly very small, of short duration, and lacking in appropriate blinding.

Moreover, yoga practices used in the interventions vary markedly, making comparison of results difficult. Interventions range from a single 1-hour session to weekly sessions over several months to inpatient treatment that includes many lifestyle modifications. Some studies required subjects to practice physically demanding asanas, while others focused on pranayama or practices similar to guided relaxation.

Helping patients navigate
the yoga domain


The variability in practices described as “yoga” and the lack of a standardized credentialing for yoga teachers make it challenging for patients to find a source suitable for their particular needs. Although choosing a style of yoga appropriate to one’s fitness level and finding an experienced instructor are not straightforward undertakings, physicians familiar with the styles, risks, and benefits of yoga can help direct patients seeking this type of therapy.

The Yoga Alliance is the best-known credentialing organization; it offers a 200-hour and 500-hour curriculum covering anatomy, yoga philosophy, and hands-on practice, and grants credit for years of experience in teaching.⁴ However, the Yoga Alliance began its current credentialing project just 7 years ago, and it is far from ubiquitous in the industry. Some types of yoga, such as Iyengar and Bikram, have their own certification systems that teachers may preferentially use.

Therapy credentialing. The International Association of Yoga Therapists (IAYT) was founded in 1989 to define yoga therapy and to organize practitioners attempting to use yoga to treat health conditions. As of July 2012, it had published suggested curricula for yoga therapists requiring 800 hours of study.4 Clearly, it will take time for these standards to become disseminated through the industry. At this point, IAYT membership does not require any certification or credentials.4 Moreover, the broad and decentralized nature of yoga practice means that any type of teacher and therapist credentialing or licensure will be controversial and not universally accepted among practitioners. Because of the relative newness of teacher and therapist licensing programs, many experienced and well-respected instructors may lack formal credentials or certifications.

Patients should do extensive research before choosing a type of yoga and an instructor (see “Finding a yoga instructor”). They should choose a type of yoga suited to their fitness level and general health (TABLE^1,2) to avoid serious injury, which can include fractures, neuralgia, and arterial dissection.²

The evidence for yoga’s benefits for specific conditions


The promotion of yoga as medical treatment is rife with dubious claims, but there is solid evidence for its benefits in some common conditions. The evidence summaries that follow reflect searches on Medline, via PubMed, and the Cochrane Database using the phrase “yoga review.”

Back pain

Often a stress-related musculoskeletal problem, back pain seems an appropriate indication for treatment with yoga, and there is a large body of literature on the subject.⁵ In a systematic review, Chou and Huffman⁶ found only 3 studies meeting inclusion criteria on yoga’s effectiveness for subacute or chronic low back pain. One large study found 6 weeks of Viniyoga was superior to conventional exercise programs and a self-care booklet in reducing pain and “bothersomeness” scores, as well as reducing the need for analgesic medication.⁷ Physician visits for back pain were not reduced in the treatment group, however.⁷ Also included in the systematic review were 2 smaller studies of Iyengar yoga on low back pain; results did not rise to statistical significance.⁶

A review by Posadzki and Ernst⁸ included 4 randomized controlled trials (RCTs) not included in Chou and Huffman, although only one of these had >50 subjects. Yoga practices for the treatment groups were mostly Iyengar and Viniyoga and lasted for 12 to 24 weeks, although one study used a 7-day intensive inpatient treatment program. Yoga practitioners had lower pain scores and lower Roland Morris Disability scores.⁸ A 2004 Clinical Inquiry in The Journal of Family Practice found limited evidence to suggest yoga may speed healing for patients with chronic back pain.⁹

Most recently, Cramer et al¹⁰ found 12 studies meeting inclusion criteria that reported on Viniyoga, Iyengar, and Hatha yoga interventions. Ten of these studies were included in the meta-analysis, which strongly favored yoga over control interventions for reducing pain and disability scores.¹⁰

Depression and anxiety

Iyengar yoga and Viniyoga have been shown
 effective in reducing chronic low back pain.Yoga therapy for depression and anxiety has been commonly studied, given that aspects of mindfulness and relaxation are thought to be important parts of treatment. Moreover, patients uncomfortable with pharmacologic therapy for their disorders may be amenable to yoga treatment. In a recent Clinical Inquiry, Skowronek et al¹¹ found evidence (strength of recommendation [SOR] B) for yoga to treat depression and anxiety symptoms based on 3 recently published review articles that commented on a total of 23 RCTs.

A handful of additional review papers on this subject have selected slightly different groups of studies to include in their analyses, but all have found generally positive results.^12-14 Inclusion criteria varied: one review omitted breathing-only modalities such as Sudarshan Kriya yoga, while another included them.^12,14 One omitted Mindfulness-Based Stress Reduction (MBSR), which is a program developed in the United States based on several Eastern and Western methodologies including yoga.¹² MBSR already has a large body of literature supporting its use for anxiety and depression.¹²

One of these reviews,¹² which involved a meta-analysis of 9 studies regarding depression, also included a meta-analysis of 5 studies on yoga for anxiety. Pooled results for depression showed significant benefit for yoga over usual care, and smaller but still significant benefit for yoga over aerobic exercise or other relaxation techniques. For anxiety, pooled analysis showed yoga to be equal to usual care but superior to other relaxation modalities.¹² As with earlier reviews, study groups were heterogeneous and included young and older adults, caregivers for dementia patients, and those receiving inpatient treatment for alcohol dependency; symptoms of depression ranged from mild to severe.¹²

In a review focusing on anxiety disorders, Kirkwood et al¹⁵ located 8 trials, 6 of which were randomized. Many of these were published in the 1970s and 80s. The yoga interventions varied and included weekly Kundalini sessions, pranayama techniques, and savasana (a pose in which practitioners lie supine while focusing on breathing and muscle relaxation). These practices were compared with anxiolytic medication, progressive muscular relaxation, placebo capsule, and no treatment. All found a statistically significant reduction in anxiety indices in the yoga treatment groups, and the authors noted that the positive effects of yoga for those suffering from obsessive-compulsive disorders are particularly well documented.¹⁵ More recently, Li and Goldsmith¹⁶ reviewed 6 interventional studies that included some trials without randomization, blinding, or a control group. Subjects of the studies included cancer patients, postmenopausal women, pregnant women, and firefighters. Six of 9 trials showed improvement in externally validated anxiety indices such as the State-Trait Anxiety Inventory or Perceived Stress Scale.

Asthma

With its focus on awareness of breath and the mechanics of breathing, yoga would seem a natural adjunct to conventional asthma therapy. One systematic review found 4 trials (3 RCTs) that showed statistically significant improvements in spirometric measurements in patients with asthma who practiced yoga techniques.¹⁷ An additional 3 RCTs showed no improvements with yoga over conventional treatments.¹⁷ Overall, the reviewers noted that study quality was poor, although they said several studies were appropriately designed. Again, the interventions described as “yoga” varied considerably, from Iyengar-type classes to meditation-focused techniques to pranayama exercises. Follow-up ranged from 6 weeks to 6 months.¹⁷

A more recent and thorough review found 14 RCTs using yoga to treat asthma symptoms.¹⁸ The investigators performed pooled analysis despite significant heterogeneity in the studies. The analysis showed some improvement in the yoga group compared with usual therapy, but no difference in comparison with sham yoga or non-yoga breathing exercises.¹⁸

Symptoms of menopause

Studies have focused on alternative or adjunctive therapies for menopause symptoms, primarily hot flashes, since hormone replacement therapy and other conventional medical therapies have been found to have a high incidence of adverse effects. However, evidence that yoga can reduce hot flashes is sparse.

A Cochrane review examined the effects of exercise on hot flashes and found 2 RCTs using yoga as a treatment modality. Neither one found statistically significant differences between the yoga groups and conventional exercise groups.¹⁹ The authors concluded there was insufficient evidence to show yoga was more effective than other forms of exercise on vasomotor symptoms of menopause. However, a large RCT included in the Cochrane review did show lower stress levels and decreased overall symptoms in the yoga arm.²⁰

The yoga intervention in this study consisted of pranayama, sun salutation (a repetitive sequence of 12 yoga postures), and cyclic meditation.²⁰ Lee et al²¹ reviewed the 2 studies used in the Cochrane paper as well as 5 other studies. Two were RCTs showing that yoga intervention was not superior to a no-treatment control. Four studies showed favorable results for yoga interventions; however, one was a nonrandomized controlled trial and 3 lacked control groups.

Cramer et al²² attempted pooled analysis of 5 studies, including those in the Cochrane paper, with similar results: Yoga interventions were not efficacious for somatic, vasomotor, or urogenital symptoms of menopause. Yoga was somewhat efficacious for psychological symptoms associated with menopause.²² More recently, an RCT (N=249) found that yoga reduces vasomotor symptoms no more frequently than non-yoga exercise.²³

Hypertension

Yoga is often said to reduce blood pressure (BP), which would make sense given the emphasis put on relaxation by many schools of yoga. In the past 2 years, 3 review articles have been published, as well as 2 relevant RCTs not included in those reviews.

Hagins et al²⁴ found 17 RCTs using yoga to treat adults with hypertension and prehypertension. These included both blinded and unblinded studies, and yoga interventions were compared with usual treatment, education, or non-yoga exercise. The authors included only studies of asanas intervention, and excluded interventions using only breathing or relaxation techniques.²⁴ In meta-analysis, pooled data showed the yoga treatment decreased both diastolic BP (DBP) and systolic BP (SBP) by 3 to 4 mm Hg compared with usual treatment, but not when compared with other exercise therapies.²⁴ Reviewers concluded that yoga was likely as effective for lowering BP as other types of physical activity.²⁴

In a review without meta-analysis, Posadzki et al²⁵ also found 17 blinded RCTs using yoga to treat hypertension or prehypertension in adults. Eleven of the 17 studies favored yoga, with 8 showing a decrease in SBP and 5 in DBP.²⁵ All but 2 studies were found to be of poor quality, especially with regard to blinding.²⁵ The authors noted that studies using subjects with prehypertension or hypertension with comorbidities were more likely to show significant results, speculating that yoga may be more effective for these populations.²⁵

In an ambitious review article on yoga as treatment for a variety of risk factors for cardiovascular disease, Cramer et al²⁶ located 28 RCTs that addressed effects of yoga on BP. Seven of the studies in the Posadzki review²⁵ were included. Meta-analysis showed a statistically significant decrease in SBP of 5.85 mm Hg and in DBP of 4.12 mm Hg.²⁶ Although wide in scope, this meta-analysis included many studies of healthy patients without hypertension who could conceivably have differing neuroendocrine responses to yoga practice.

Search engines offered by the Yoga Alliance and the International Association of Yoga Therapists can help patients to locate a yoga instructor or therapist in their area.In a pilot RCT, Cohen et al²⁷ found a significant decrease in BP among subjects randomized into Iyengar yoga classes for 24 weeks compared with a control group educated about lifestyle modification. A larger study with 102 subjects is currently underway.²⁸ These studies were unique in that no subjects were currently being treated with antihypertensive medications^27,28; most other trials on this subject enrolled participants on antihypertensive medications if their regimens had been stable for some time.

In an RCT published recently by Hagins et al,²⁹ 68 subjects with pre- or stage I hypertension were randomized into Ashtanga yoga classes or non-aerobic exercise classes formulated to burn equivalent METs. After 12 weeks of treatment, the yoga subjects’ BP had significantly decreased from starting values, but was not improved compared with the exercise subjects.²⁹ This further supports the assertion that yoga is equivalent to other forms of physical activity in decreasing BP among hypertensive subjects.

Balance and stability in the elderly

With its emphasis on strength, balance, and body awareness, yoga would seem a helpful intervention for older patients at risk of injury from falls. Unfortunately this area of research lacks significant numbers of controlled trails. In a Cochrane review of exercise interventions for improving balance in the elderly, the reviewers were unable find any studies specifically using yoga that met their criteria.³⁰ Jeter et al³¹ attempted a review more recently, and found 15 studies meeting inclusion criteria, 5 of which were RCTs. Overall, however, the poor quality of the studies and variation in both the type of yoga used as intervention and measurements of balance precluded pooled analysis, although some studies did have positive results.

A small but well-designed pilot RCT was recently published showing that an Iyengar yoga intervention significantly improved timed one-leg balancing among community dwelling older adults.³² However, this study did not show a significant difference in a standardized fall risk survey after the intervention.³²

Cautioning against yoga in this context are several articles chronicling increased risks of some yoga exercises, especially for those with osteoporosis or other risks for fractures.³³ At this point, the well-documented risks of yoga practice in this group probably outweigh the unsubstantiated rewards.

CORRESPONDENCE
Genevieve Verrastro, MD, MAHEC Family Health Center at Biltmore, 123 Hendersonville Road, Asheville, NC 28803; [email protected]

Yoga is practiced by 15.8 million Americans,¹ and is often recommended as therapy for a variety of medical conditions. However, the scientific literature on yoga is limited in scope and quality. This article presents good evidence for yoga as treatment for chronic back pain, depression, and anxiety, and fair evidence for treating asthma, symptoms of menopause, hypertension, and mobility issues in the elderly.

Yoga’s rising popularity as therapy

Yoga is a system of movement and breathing exercises meant to foster mind-body connection. Its roots are in ancient Indian practices codified by the writer Patanjali in the first or second century BCE.²

The practice of yoga was introduced to the Western world by a series of popular gurus from the 1930s to 1970s and consists primarily of asanas, or postures, and breathing exercises known as pranayama. Since then, yoga has been further subdivided into different schools and brands (TABLE^1,2), some of which are extremely taxing and vigorous and should be performed only by fit and healthy individuals, while others are gentle and accessible to anyone. Yoga has steadily gained in popularity, and nearly half of those who practice it say they do so to improve their health.¹

How useful is the research on yoga therapy?

Yoga has been a subject of Western scientific inquiry for more than 100 years. It has been deemed effective for treating conditions from hypertension to epilepsy,³ but many claims are poorly substantiated. Most studies report on a single case or series. The few investigational studies are mainly very small, of short duration, and lacking in appropriate blinding.

Moreover, yoga practices used in the interventions vary markedly, making comparison of results difficult. Interventions range from a single 1-hour session to weekly sessions over several months to inpatient treatment that includes many lifestyle modifications. Some studies required subjects to practice physically demanding asanas, while others focused on pranayama or practices similar to guided relaxation.

Helping patients navigate
the yoga domain


The variability in practices described as “yoga” and the lack of a standardized credentialing for yoga teachers make it challenging for patients to find a source suitable for their particular needs. Although choosing a style of yoga appropriate to one’s fitness level and finding an experienced instructor are not straightforward undertakings, physicians familiar with the styles, risks, and benefits of yoga can help direct patients seeking this type of therapy.

The Yoga Alliance is the best-known credentialing organization; it offers a 200-hour and 500-hour curriculum covering anatomy, yoga philosophy, and hands-on practice, and grants credit for years of experience in teaching.⁴ However, the Yoga Alliance began its current credentialing project just 7 years ago, and it is far from ubiquitous in the industry. Some types of yoga, such as Iyengar and Bikram, have their own certification systems that teachers may preferentially use.

Therapy credentialing. The International Association of Yoga Therapists (IAYT) was founded in 1989 to define yoga therapy and to organize practitioners attempting to use yoga to treat health conditions. As of July 2012, it had published suggested curricula for yoga therapists requiring 800 hours of study.4 Clearly, it will take time for these standards to become disseminated through the industry. At this point, IAYT membership does not require any certification or credentials.4 Moreover, the broad and decentralized nature of yoga practice means that any type of teacher and therapist credentialing or licensure will be controversial and not universally accepted among practitioners. Because of the relative newness of teacher and therapist licensing programs, many experienced and well-respected instructors may lack formal credentials or certifications.

Patients should do extensive research before choosing a type of yoga and an instructor (see “Finding a yoga instructor”). They should choose a type of yoga suited to their fitness level and general health (TABLE^1,2) to avoid serious injury, which can include fractures, neuralgia, and arterial dissection.²

The evidence for yoga’s benefits for specific conditions


The promotion of yoga as medical treatment is rife with dubious claims, but there is solid evidence for its benefits in some common conditions. The evidence summaries that follow reflect searches on Medline, via PubMed, and the Cochrane Database using the phrase “yoga review.”

Back pain

Often a stress-related musculoskeletal problem, back pain seems an appropriate indication for treatment with yoga, and there is a large body of literature on the subject.⁵ In a systematic review, Chou and Huffman⁶ found only 3 studies meeting inclusion criteria on yoga’s effectiveness for subacute or chronic low back pain. One large study found 6 weeks of Viniyoga was superior to conventional exercise programs and a self-care booklet in reducing pain and “bothersomeness” scores, as well as reducing the need for analgesic medication.⁷ Physician visits for back pain were not reduced in the treatment group, however.⁷ Also included in the systematic review were 2 smaller studies of Iyengar yoga on low back pain; results did not rise to statistical significance.⁶

A review by Posadzki and Ernst⁸ included 4 randomized controlled trials (RCTs) not included in Chou and Huffman, although only one of these had >50 subjects. Yoga practices for the treatment groups were mostly Iyengar and Viniyoga and lasted for 12 to 24 weeks, although one study used a 7-day intensive inpatient treatment program. Yoga practitioners had lower pain scores and lower Roland Morris Disability scores.⁸ A 2004 Clinical Inquiry in The Journal of Family Practice found limited evidence to suggest yoga may speed healing for patients with chronic back pain.⁹

Most recently, Cramer et al¹⁰ found 12 studies meeting inclusion criteria that reported on Viniyoga, Iyengar, and Hatha yoga interventions. Ten of these studies were included in the meta-analysis, which strongly favored yoga over control interventions for reducing pain and disability scores.¹⁰

Depression and anxiety

Iyengar yoga and Viniyoga have been shown
 effective in reducing chronic low back pain.Yoga therapy for depression and anxiety has been commonly studied, given that aspects of mindfulness and relaxation are thought to be important parts of treatment. Moreover, patients uncomfortable with pharmacologic therapy for their disorders may be amenable to yoga treatment. In a recent Clinical Inquiry, Skowronek et al¹¹ found evidence (strength of recommendation [SOR] B) for yoga to treat depression and anxiety symptoms based on 3 recently published review articles that commented on a total of 23 RCTs.

A handful of additional review papers on this subject have selected slightly different groups of studies to include in their analyses, but all have found generally positive results.^12-14 Inclusion criteria varied: one review omitted breathing-only modalities such as Sudarshan Kriya yoga, while another included them.^12,14 One omitted Mindfulness-Based Stress Reduction (MBSR), which is a program developed in the United States based on several Eastern and Western methodologies including yoga.¹² MBSR already has a large body of literature supporting its use for anxiety and depression.¹²

One of these reviews,¹² which involved a meta-analysis of 9 studies regarding depression, also included a meta-analysis of 5 studies on yoga for anxiety. Pooled results for depression showed significant benefit for yoga over usual care, and smaller but still significant benefit for yoga over aerobic exercise or other relaxation techniques. For anxiety, pooled analysis showed yoga to be equal to usual care but superior to other relaxation modalities.¹² As with earlier reviews, study groups were heterogeneous and included young and older adults, caregivers for dementia patients, and those receiving inpatient treatment for alcohol dependency; symptoms of depression ranged from mild to severe.¹²

In a review focusing on anxiety disorders, Kirkwood et al¹⁵ located 8 trials, 6 of which were randomized. Many of these were published in the 1970s and 80s. The yoga interventions varied and included weekly Kundalini sessions, pranayama techniques, and savasana (a pose in which practitioners lie supine while focusing on breathing and muscle relaxation). These practices were compared with anxiolytic medication, progressive muscular relaxation, placebo capsule, and no treatment. All found a statistically significant reduction in anxiety indices in the yoga treatment groups, and the authors noted that the positive effects of yoga for those suffering from obsessive-compulsive disorders are particularly well documented.¹⁵ More recently, Li and Goldsmith¹⁶ reviewed 6 interventional studies that included some trials without randomization, blinding, or a control group. Subjects of the studies included cancer patients, postmenopausal women, pregnant women, and firefighters. Six of 9 trials showed improvement in externally validated anxiety indices such as the State-Trait Anxiety Inventory or Perceived Stress Scale.

Asthma

With its focus on awareness of breath and the mechanics of breathing, yoga would seem a natural adjunct to conventional asthma therapy. One systematic review found 4 trials (3 RCTs) that showed statistically significant improvements in spirometric measurements in patients with asthma who practiced yoga techniques.¹⁷ An additional 3 RCTs showed no improvements with yoga over conventional treatments.¹⁷ Overall, the reviewers noted that study quality was poor, although they said several studies were appropriately designed. Again, the interventions described as “yoga” varied considerably, from Iyengar-type classes to meditation-focused techniques to pranayama exercises. Follow-up ranged from 6 weeks to 6 months.¹⁷

A more recent and thorough review found 14 RCTs using yoga to treat asthma symptoms.¹⁸ The investigators performed pooled analysis despite significant heterogeneity in the studies. The analysis showed some improvement in the yoga group compared with usual therapy, but no difference in comparison with sham yoga or non-yoga breathing exercises.¹⁸

Symptoms of menopause

Studies have focused on alternative or adjunctive therapies for menopause symptoms, primarily hot flashes, since hormone replacement therapy and other conventional medical therapies have been found to have a high incidence of adverse effects. However, evidence that yoga can reduce hot flashes is sparse.

A Cochrane review examined the effects of exercise on hot flashes and found 2 RCTs using yoga as a treatment modality. Neither one found statistically significant differences between the yoga groups and conventional exercise groups.¹⁹ The authors concluded there was insufficient evidence to show yoga was more effective than other forms of exercise on vasomotor symptoms of menopause. However, a large RCT included in the Cochrane review did show lower stress levels and decreased overall symptoms in the yoga arm.²⁰

The yoga intervention in this study consisted of pranayama, sun salutation (a repetitive sequence of 12 yoga postures), and cyclic meditation.²⁰ Lee et al²¹ reviewed the 2 studies used in the Cochrane paper as well as 5 other studies. Two were RCTs showing that yoga intervention was not superior to a no-treatment control. Four studies showed favorable results for yoga interventions; however, one was a nonrandomized controlled trial and 3 lacked control groups.

Cramer et al²² attempted pooled analysis of 5 studies, including those in the Cochrane paper, with similar results: Yoga interventions were not efficacious for somatic, vasomotor, or urogenital symptoms of menopause. Yoga was somewhat efficacious for psychological symptoms associated with menopause.²² More recently, an RCT (N=249) found that yoga reduces vasomotor symptoms no more frequently than non-yoga exercise.²³

Hypertension

Yoga is often said to reduce blood pressure (BP), which would make sense given the emphasis put on relaxation by many schools of yoga. In the past 2 years, 3 review articles have been published, as well as 2 relevant RCTs not included in those reviews.

Hagins et al²⁴ found 17 RCTs using yoga to treat adults with hypertension and prehypertension. These included both blinded and unblinded studies, and yoga interventions were compared with usual treatment, education, or non-yoga exercise. The authors included only studies of asanas intervention, and excluded interventions using only breathing or relaxation techniques.²⁴ In meta-analysis, pooled data showed the yoga treatment decreased both diastolic BP (DBP) and systolic BP (SBP) by 3 to 4 mm Hg compared with usual treatment, but not when compared with other exercise therapies.²⁴ Reviewers concluded that yoga was likely as effective for lowering BP as other types of physical activity.²⁴

In a review without meta-analysis, Posadzki et al²⁵ also found 17 blinded RCTs using yoga to treat hypertension or prehypertension in adults. Eleven of the 17 studies favored yoga, with 8 showing a decrease in SBP and 5 in DBP.²⁵ All but 2 studies were found to be of poor quality, especially with regard to blinding.²⁵ The authors noted that studies using subjects with prehypertension or hypertension with comorbidities were more likely to show significant results, speculating that yoga may be more effective for these populations.²⁵

In an ambitious review article on yoga as treatment for a variety of risk factors for cardiovascular disease, Cramer et al²⁶ located 28 RCTs that addressed effects of yoga on BP. Seven of the studies in the Posadzki review²⁵ were included. Meta-analysis showed a statistically significant decrease in SBP of 5.85 mm Hg and in DBP of 4.12 mm Hg.²⁶ Although wide in scope, this meta-analysis included many studies of healthy patients without hypertension who could conceivably have differing neuroendocrine responses to yoga practice.

Search engines offered by the Yoga Alliance and the International Association of Yoga Therapists can help patients to locate a yoga instructor or therapist in their area.In a pilot RCT, Cohen et al²⁷ found a significant decrease in BP among subjects randomized into Iyengar yoga classes for 24 weeks compared with a control group educated about lifestyle modification. A larger study with 102 subjects is currently underway.²⁸ These studies were unique in that no subjects were currently being treated with antihypertensive medications^27,28; most other trials on this subject enrolled participants on antihypertensive medications if their regimens had been stable for some time.

In an RCT published recently by Hagins et al,²⁹ 68 subjects with pre- or stage I hypertension were randomized into Ashtanga yoga classes or non-aerobic exercise classes formulated to burn equivalent METs. After 12 weeks of treatment, the yoga subjects’ BP had significantly decreased from starting values, but was not improved compared with the exercise subjects.²⁹ This further supports the assertion that yoga is equivalent to other forms of physical activity in decreasing BP among hypertensive subjects.

Balance and stability in the elderly

With its emphasis on strength, balance, and body awareness, yoga would seem a helpful intervention for older patients at risk of injury from falls. Unfortunately this area of research lacks significant numbers of controlled trails. In a Cochrane review of exercise interventions for improving balance in the elderly, the reviewers were unable find any studies specifically using yoga that met their criteria.³⁰ Jeter et al³¹ attempted a review more recently, and found 15 studies meeting inclusion criteria, 5 of which were RCTs. Overall, however, the poor quality of the studies and variation in both the type of yoga used as intervention and measurements of balance precluded pooled analysis, although some studies did have positive results.

A small but well-designed pilot RCT was recently published showing that an Iyengar yoga intervention significantly improved timed one-leg balancing among community dwelling older adults.³² However, this study did not show a significant difference in a standardized fall risk survey after the intervention.³²

Cautioning against yoga in this context are several articles chronicling increased risks of some yoga exercises, especially for those with osteoporosis or other risks for fractures.³³ At this point, the well-documented risks of yoga practice in this group probably outweigh the unsubstantiated rewards.

CORRESPONDENCE
Genevieve Verrastro, MD, MAHEC Family Health Center at Biltmore, 123 Hendersonville Road, Asheville, NC 28803; [email protected]

Yoga is practiced by 15.8 million Americans,¹ and is often recommended as therapy for a variety of medical conditions. However, the scientific literature on yoga is limited in scope and quality. This article presents good evidence for yoga as treatment for chronic back pain, depression, and anxiety, and fair evidence for treating asthma, symptoms of menopause, hypertension, and mobility issues in the elderly.

Yoga’s rising popularity as therapy

Yoga is a system of movement and breathing exercises meant to foster mind-body connection. Its roots are in ancient Indian practices codified by the writer Patanjali in the first or second century BCE.²

The practice of yoga was introduced to the Western world by a series of popular gurus from the 1930s to 1970s and consists primarily of asanas, or postures, and breathing exercises known as pranayama. Since then, yoga has been further subdivided into different schools and brands (TABLE^1,2), some of which are extremely taxing and vigorous and should be performed only by fit and healthy individuals, while others are gentle and accessible to anyone. Yoga has steadily gained in popularity, and nearly half of those who practice it say they do so to improve their health.¹

How useful is the research on yoga therapy?

Yoga has been a subject of Western scientific inquiry for more than 100 years. It has been deemed effective for treating conditions from hypertension to epilepsy,³ but many claims are poorly substantiated. Most studies report on a single case or series. The few investigational studies are mainly very small, of short duration, and lacking in appropriate blinding.

Moreover, yoga practices used in the interventions vary markedly, making comparison of results difficult. Interventions range from a single 1-hour session to weekly sessions over several months to inpatient treatment that includes many lifestyle modifications. Some studies required subjects to practice physically demanding asanas, while others focused on pranayama or practices similar to guided relaxation.

Helping patients navigate
the yoga domain


The variability in practices described as “yoga” and the lack of a standardized credentialing for yoga teachers make it challenging for patients to find a source suitable for their particular needs. Although choosing a style of yoga appropriate to one’s fitness level and finding an experienced instructor are not straightforward undertakings, physicians familiar with the styles, risks, and benefits of yoga can help direct patients seeking this type of therapy.

The Yoga Alliance is the best-known credentialing organization; it offers a 200-hour and 500-hour curriculum covering anatomy, yoga philosophy, and hands-on practice, and grants credit for years of experience in teaching.⁴ However, the Yoga Alliance began its current credentialing project just 7 years ago, and it is far from ubiquitous in the industry. Some types of yoga, such as Iyengar and Bikram, have their own certification systems that teachers may preferentially use.

Therapy credentialing. The International Association of Yoga Therapists (IAYT) was founded in 1989 to define yoga therapy and to organize practitioners attempting to use yoga to treat health conditions. As of July 2012, it had published suggested curricula for yoga therapists requiring 800 hours of study.4 Clearly, it will take time for these standards to become disseminated through the industry. At this point, IAYT membership does not require any certification or credentials.4 Moreover, the broad and decentralized nature of yoga practice means that any type of teacher and therapist credentialing or licensure will be controversial and not universally accepted among practitioners. Because of the relative newness of teacher and therapist licensing programs, many experienced and well-respected instructors may lack formal credentials or certifications.

Patients should do extensive research before choosing a type of yoga and an instructor (see “Finding a yoga instructor”). They should choose a type of yoga suited to their fitness level and general health (TABLE^1,2) to avoid serious injury, which can include fractures, neuralgia, and arterial dissection.²

The evidence for yoga’s benefits for specific conditions


The promotion of yoga as medical treatment is rife with dubious claims, but there is solid evidence for its benefits in some common conditions. The evidence summaries that follow reflect searches on Medline, via PubMed, and the Cochrane Database using the phrase “yoga review.”

Back pain

Often a stress-related musculoskeletal problem, back pain seems an appropriate indication for treatment with yoga, and there is a large body of literature on the subject.⁵ In a systematic review, Chou and Huffman⁶ found only 3 studies meeting inclusion criteria on yoga’s effectiveness for subacute or chronic low back pain. One large study found 6 weeks of Viniyoga was superior to conventional exercise programs and a self-care booklet in reducing pain and “bothersomeness” scores, as well as reducing the need for analgesic medication.⁷ Physician visits for back pain were not reduced in the treatment group, however.⁷ Also included in the systematic review were 2 smaller studies of Iyengar yoga on low back pain; results did not rise to statistical significance.⁶

A review by Posadzki and Ernst⁸ included 4 randomized controlled trials (RCTs) not included in Chou and Huffman, although only one of these had >50 subjects. Yoga practices for the treatment groups were mostly Iyengar and Viniyoga and lasted for 12 to 24 weeks, although one study used a 7-day intensive inpatient treatment program. Yoga practitioners had lower pain scores and lower Roland Morris Disability scores.⁸ A 2004 Clinical Inquiry in The Journal of Family Practice found limited evidence to suggest yoga may speed healing for patients with chronic back pain.⁹

Most recently, Cramer et al¹⁰ found 12 studies meeting inclusion criteria that reported on Viniyoga, Iyengar, and Hatha yoga interventions. Ten of these studies were included in the meta-analysis, which strongly favored yoga over control interventions for reducing pain and disability scores.¹⁰

Depression and anxiety

Iyengar yoga and Viniyoga have been shown
 effective in reducing chronic low back pain.Yoga therapy for depression and anxiety has been commonly studied, given that aspects of mindfulness and relaxation are thought to be important parts of treatment. Moreover, patients uncomfortable with pharmacologic therapy for their disorders may be amenable to yoga treatment. In a recent Clinical Inquiry, Skowronek et al¹¹ found evidence (strength of recommendation [SOR] B) for yoga to treat depression and anxiety symptoms based on 3 recently published review articles that commented on a total of 23 RCTs.

A handful of additional review papers on this subject have selected slightly different groups of studies to include in their analyses, but all have found generally positive results.^12-14 Inclusion criteria varied: one review omitted breathing-only modalities such as Sudarshan Kriya yoga, while another included them.^12,14 One omitted Mindfulness-Based Stress Reduction (MBSR), which is a program developed in the United States based on several Eastern and Western methodologies including yoga.¹² MBSR already has a large body of literature supporting its use for anxiety and depression.¹²

One of these reviews,¹² which involved a meta-analysis of 9 studies regarding depression, also included a meta-analysis of 5 studies on yoga for anxiety. Pooled results for depression showed significant benefit for yoga over usual care, and smaller but still significant benefit for yoga over aerobic exercise or other relaxation techniques. For anxiety, pooled analysis showed yoga to be equal to usual care but superior to other relaxation modalities.¹² As with earlier reviews, study groups were heterogeneous and included young and older adults, caregivers for dementia patients, and those receiving inpatient treatment for alcohol dependency; symptoms of depression ranged from mild to severe.¹²

In a review focusing on anxiety disorders, Kirkwood et al¹⁵ located 8 trials, 6 of which were randomized. Many of these were published in the 1970s and 80s. The yoga interventions varied and included weekly Kundalini sessions, pranayama techniques, and savasana (a pose in which practitioners lie supine while focusing on breathing and muscle relaxation). These practices were compared with anxiolytic medication, progressive muscular relaxation, placebo capsule, and no treatment. All found a statistically significant reduction in anxiety indices in the yoga treatment groups, and the authors noted that the positive effects of yoga for those suffering from obsessive-compulsive disorders are particularly well documented.¹⁵ More recently, Li and Goldsmith¹⁶ reviewed 6 interventional studies that included some trials without randomization, blinding, or a control group. Subjects of the studies included cancer patients, postmenopausal women, pregnant women, and firefighters. Six of 9 trials showed improvement in externally validated anxiety indices such as the State-Trait Anxiety Inventory or Perceived Stress Scale.

Asthma

With its focus on awareness of breath and the mechanics of breathing, yoga would seem a natural adjunct to conventional asthma therapy. One systematic review found 4 trials (3 RCTs) that showed statistically significant improvements in spirometric measurements in patients with asthma who practiced yoga techniques.¹⁷ An additional 3 RCTs showed no improvements with yoga over conventional treatments.¹⁷ Overall, the reviewers noted that study quality was poor, although they said several studies were appropriately designed. Again, the interventions described as “yoga” varied considerably, from Iyengar-type classes to meditation-focused techniques to pranayama exercises. Follow-up ranged from 6 weeks to 6 months.¹⁷

A more recent and thorough review found 14 RCTs using yoga to treat asthma symptoms.¹⁸ The investigators performed pooled analysis despite significant heterogeneity in the studies. The analysis showed some improvement in the yoga group compared with usual therapy, but no difference in comparison with sham yoga or non-yoga breathing exercises.¹⁸

Symptoms of menopause

Studies have focused on alternative or adjunctive therapies for menopause symptoms, primarily hot flashes, since hormone replacement therapy and other conventional medical therapies have been found to have a high incidence of adverse effects. However, evidence that yoga can reduce hot flashes is sparse.

A Cochrane review examined the effects of exercise on hot flashes and found 2 RCTs using yoga as a treatment modality. Neither one found statistically significant differences between the yoga groups and conventional exercise groups.¹⁹ The authors concluded there was insufficient evidence to show yoga was more effective than other forms of exercise on vasomotor symptoms of menopause. However, a large RCT included in the Cochrane review did show lower stress levels and decreased overall symptoms in the yoga arm.²⁰

The yoga intervention in this study consisted of pranayama, sun salutation (a repetitive sequence of 12 yoga postures), and cyclic meditation.²⁰ Lee et al²¹ reviewed the 2 studies used in the Cochrane paper as well as 5 other studies. Two were RCTs showing that yoga intervention was not superior to a no-treatment control. Four studies showed favorable results for yoga interventions; however, one was a nonrandomized controlled trial and 3 lacked control groups.

Cramer et al²² attempted pooled analysis of 5 studies, including those in the Cochrane paper, with similar results: Yoga interventions were not efficacious for somatic, vasomotor, or urogenital symptoms of menopause. Yoga was somewhat efficacious for psychological symptoms associated with menopause.²² More recently, an RCT (N=249) found that yoga reduces vasomotor symptoms no more frequently than non-yoga exercise.²³

Hypertension

Yoga is often said to reduce blood pressure (BP), which would make sense given the emphasis put on relaxation by many schools of yoga. In the past 2 years, 3 review articles have been published, as well as 2 relevant RCTs not included in those reviews.

Hagins et al²⁴ found 17 RCTs using yoga to treat adults with hypertension and prehypertension. These included both blinded and unblinded studies, and yoga interventions were compared with usual treatment, education, or non-yoga exercise. The authors included only studies of asanas intervention, and excluded interventions using only breathing or relaxation techniques.²⁴ In meta-analysis, pooled data showed the yoga treatment decreased both diastolic BP (DBP) and systolic BP (SBP) by 3 to 4 mm Hg compared with usual treatment, but not when compared with other exercise therapies.²⁴ Reviewers concluded that yoga was likely as effective for lowering BP as other types of physical activity.²⁴

In a review without meta-analysis, Posadzki et al²⁵ also found 17 blinded RCTs using yoga to treat hypertension or prehypertension in adults. Eleven of the 17 studies favored yoga, with 8 showing a decrease in SBP and 5 in DBP.²⁵ All but 2 studies were found to be of poor quality, especially with regard to blinding.²⁵ The authors noted that studies using subjects with prehypertension or hypertension with comorbidities were more likely to show significant results, speculating that yoga may be more effective for these populations.²⁵

In an ambitious review article on yoga as treatment for a variety of risk factors for cardiovascular disease, Cramer et al²⁶ located 28 RCTs that addressed effects of yoga on BP. Seven of the studies in the Posadzki review²⁵ were included. Meta-analysis showed a statistically significant decrease in SBP of 5.85 mm Hg and in DBP of 4.12 mm Hg.²⁶ Although wide in scope, this meta-analysis included many studies of healthy patients without hypertension who could conceivably have differing neuroendocrine responses to yoga practice.

Search engines offered by the Yoga Alliance and the International Association of Yoga Therapists can help patients to locate a yoga instructor or therapist in their area.In a pilot RCT, Cohen et al²⁷ found a significant decrease in BP among subjects randomized into Iyengar yoga classes for 24 weeks compared with a control group educated about lifestyle modification. A larger study with 102 subjects is currently underway.²⁸ These studies were unique in that no subjects were currently being treated with antihypertensive medications^27,28; most other trials on this subject enrolled participants on antihypertensive medications if their regimens had been stable for some time.

In an RCT published recently by Hagins et al,²⁹ 68 subjects with pre- or stage I hypertension were randomized into Ashtanga yoga classes or non-aerobic exercise classes formulated to burn equivalent METs. After 12 weeks of treatment, the yoga subjects’ BP had significantly decreased from starting values, but was not improved compared with the exercise subjects.²⁹ This further supports the assertion that yoga is equivalent to other forms of physical activity in decreasing BP among hypertensive subjects.

Balance and stability in the elderly

With its emphasis on strength, balance, and body awareness, yoga would seem a helpful intervention for older patients at risk of injury from falls. Unfortunately this area of research lacks significant numbers of controlled trails. In a Cochrane review of exercise interventions for improving balance in the elderly, the reviewers were unable find any studies specifically using yoga that met their criteria.³⁰ Jeter et al³¹ attempted a review more recently, and found 15 studies meeting inclusion criteria, 5 of which were RCTs. Overall, however, the poor quality of the studies and variation in both the type of yoga used as intervention and measurements of balance precluded pooled analysis, although some studies did have positive results.

A small but well-designed pilot RCT was recently published showing that an Iyengar yoga intervention significantly improved timed one-leg balancing among community dwelling older adults.³² However, this study did not show a significant difference in a standardized fall risk survey after the intervention.³²

Cautioning against yoga in this context are several articles chronicling increased risks of some yoga exercises, especially for those with osteoporosis or other risks for fractures.³³ At this point, the well-documented risks of yoga practice in this group probably outweigh the unsubstantiated rewards.

CORRESPONDENCE
Genevieve Verrastro, MD, MAHEC Family Health Center at Biltmore, 123 Hendersonville Road, Asheville, NC 28803; [email protected]