Clinical Inquiries

Evidence summary

The US Preventive Services Task Force considers screening adults for hypertension a grade A recommendation because it’s known to improve patient outcomes through early diagnosis, treatment, and prevention of serious cardiovascular complications.¹

The Fourth Task Force of the National High Blood Pressure Education Program Working Group, endorsed by the American Academy of Pediatrics, states that maintaining a large national database of BP values throughout childhood allows physicians to recognize children and adolescents with elevated BP.² Data indicate that, in this population, the prevalence of prehypertension is 10% and the prevalence of hypertension is 4%.³

The Task Force suggests that detecting and treating childhood hypertension should be important because of increasing childhood obesity, the risk of developing left ventricular hypertrophy, and other intermediate cardiovascular effects in undiagnosed and untreated children. The Task Force acknowledges, however, that prospective longitudinal outcome studies in untreated children and adolescents are lacking.

Hypertensive children often grow up to be hypertensive adults
A prospective cohort study showed that children with elevated BP had a greater likelihood of adult hypertension than children with normal BP. Investigators followed 2445 children 7 to 18 years of age to determine whether elevated BP in childhood correlated with increased BP in adulthood.

Investigators obtained BP, height, and weight measurements biennially during the children’s school years and when they were young adults between 20 and 30 years of age. Twelve to 13 years later, 24% of children with BP above the 90th percentile still had BP above the 90th percentile (relative risk [RR]=2.4; P<.001) and 39% had BP above the 80th percentile (RR=1.9; P<.001). Ninety-four percent of children with more than 3 normal readings during the study were normotensive as young adults. Children with one or 2 abnormal readings had a 17% and 24% chance, respectively, of having hypertension as adults (P<.001).⁴

High childhood systolic BP may predict CAD in adulthood
A retrospective study evaluated 126 children 10 to 17 years of age who were admitted to the hospital for an elective surgical procedure between 1950 and 1967. Children with documented BP readings at admission were eligible for the study; children with preexisting cardiac and renal disease were excluded. Investigators reassessed patients as adults (age range 42-68 years); the mean follow-up period was 42 years.

Mean BP was 125/80 mm Hg at admission and 133/75 mm Hg at follow-up. Univariate logistic regression analysis showed a significant association between systolic BP in childhood and coronary artery disease at follow-up (odds ratio [OR]=1.052; 95% confidence interval [CI], 1.005-1.101; P=.027). Children with systolic BP at or above the 95th percentile had a 4-fold increase in coronary artery disease at follow-up compared with children whose systolic BP was below the 95th percentile (29% vs 7%, P=.03). Investigators also found an association between elevated BP in childhood and a diagnosis of hypertension at follow-up (P=.007).

Limitations of the study included small sample size, selection bias, changes in the definition of hypertension during the 4 decades since the study began, and limited childhood BP data (a single measurement at admission for surgery).⁵

Parents of hypertensive children are likely to be hypertensive themselves
Screening BP in children has the potential to identify families at increased risk for cardiovascular disease. A case series found a high incidence of hypertension among the parents of children with elevated BP. Investigators measured several risk factors, including BP in 141 children (mean age 10.5±3.4 years) and 108 parents (at least one a biological parent, mean age 38.5±7.5 years). They obtained 2 BP readings 15 to 30 minutes apart.

Parents of children with BPs at or above the 95th percentile had a 15-fold greater likelihood of hypertension themselves (OR=14.7; 95% CI, 3.02-71.56; P=.009, positive predictive value=75%; negative predictive value=81%).⁶ Limitations of the study included small sample size, high prevalence of obesity and black ethnicity in the study population (a population with a greater incidence of hypertension), and only 2 BP measurements in the same day, which isn’t diagnostic for hypertension.

Recommendations

The American College of Obstetricians and Gynecologists recommends screening girls for hypertension between 13 and 15 years of age.⁷

The American Academy of Family Physicians concludes that the evidence is insufficient to recommend for or against routine screening for hypertension in children and adolescents to reduce the risk of cardiovascular disease.⁸

The European Society of Hypertension and European Society of Cardiology recommend that children older than 3 years have auscultatory BP measurements at each clinic visit.⁹

Evidence summary

The US Preventive Services Task Force considers screening adults for hypertension a grade A recommendation because it’s known to improve patient outcomes through early diagnosis, treatment, and prevention of serious cardiovascular complications.¹

The Fourth Task Force of the National High Blood Pressure Education Program Working Group, endorsed by the American Academy of Pediatrics, states that maintaining a large national database of BP values throughout childhood allows physicians to recognize children and adolescents with elevated BP.² Data indicate that, in this population, the prevalence of prehypertension is 10% and the prevalence of hypertension is 4%.³

The Task Force suggests that detecting and treating childhood hypertension should be important because of increasing childhood obesity, the risk of developing left ventricular hypertrophy, and other intermediate cardiovascular effects in undiagnosed and untreated children. The Task Force acknowledges, however, that prospective longitudinal outcome studies in untreated children and adolescents are lacking.

Hypertensive children often grow up to be hypertensive adults
A prospective cohort study showed that children with elevated BP had a greater likelihood of adult hypertension than children with normal BP. Investigators followed 2445 children 7 to 18 years of age to determine whether elevated BP in childhood correlated with increased BP in adulthood.

Investigators obtained BP, height, and weight measurements biennially during the children’s school years and when they were young adults between 20 and 30 years of age. Twelve to 13 years later, 24% of children with BP above the 90th percentile still had BP above the 90th percentile (relative risk [RR]=2.4; P<.001) and 39% had BP above the 80th percentile (RR=1.9; P<.001). Ninety-four percent of children with more than 3 normal readings during the study were normotensive as young adults. Children with one or 2 abnormal readings had a 17% and 24% chance, respectively, of having hypertension as adults (P<.001).⁴

High childhood systolic BP may predict CAD in adulthood
A retrospective study evaluated 126 children 10 to 17 years of age who were admitted to the hospital for an elective surgical procedure between 1950 and 1967. Children with documented BP readings at admission were eligible for the study; children with preexisting cardiac and renal disease were excluded. Investigators reassessed patients as adults (age range 42-68 years); the mean follow-up period was 42 years.

Mean BP was 125/80 mm Hg at admission and 133/75 mm Hg at follow-up. Univariate logistic regression analysis showed a significant association between systolic BP in childhood and coronary artery disease at follow-up (odds ratio [OR]=1.052; 95% confidence interval [CI], 1.005-1.101; P=.027). Children with systolic BP at or above the 95th percentile had a 4-fold increase in coronary artery disease at follow-up compared with children whose systolic BP was below the 95th percentile (29% vs 7%, P=.03). Investigators also found an association between elevated BP in childhood and a diagnosis of hypertension at follow-up (P=.007).

Limitations of the study included small sample size, selection bias, changes in the definition of hypertension during the 4 decades since the study began, and limited childhood BP data (a single measurement at admission for surgery).⁵

Parents of hypertensive children are likely to be hypertensive themselves
Screening BP in children has the potential to identify families at increased risk for cardiovascular disease. A case series found a high incidence of hypertension among the parents of children with elevated BP. Investigators measured several risk factors, including BP in 141 children (mean age 10.5±3.4 years) and 108 parents (at least one a biological parent, mean age 38.5±7.5 years). They obtained 2 BP readings 15 to 30 minutes apart.

Parents of children with BPs at or above the 95th percentile had a 15-fold greater likelihood of hypertension themselves (OR=14.7; 95% CI, 3.02-71.56; P=.009, positive predictive value=75%; negative predictive value=81%).⁶ Limitations of the study included small sample size, high prevalence of obesity and black ethnicity in the study population (a population with a greater incidence of hypertension), and only 2 BP measurements in the same day, which isn’t diagnostic for hypertension.

Recommendations

The American College of Obstetricians and Gynecologists recommends screening girls for hypertension between 13 and 15 years of age.⁷

The American Academy of Family Physicians concludes that the evidence is insufficient to recommend for or against routine screening for hypertension in children and adolescents to reduce the risk of cardiovascular disease.⁸

The European Society of Hypertension and European Society of Cardiology recommend that children older than 3 years have auscultatory BP measurements at each clinic visit.⁹

Evidence summary

The US Preventive Services Task Force considers screening adults for hypertension a grade A recommendation because it’s known to improve patient outcomes through early diagnosis, treatment, and prevention of serious cardiovascular complications.¹

The Fourth Task Force of the National High Blood Pressure Education Program Working Group, endorsed by the American Academy of Pediatrics, states that maintaining a large national database of BP values throughout childhood allows physicians to recognize children and adolescents with elevated BP.² Data indicate that, in this population, the prevalence of prehypertension is 10% and the prevalence of hypertension is 4%.³

The Task Force suggests that detecting and treating childhood hypertension should be important because of increasing childhood obesity, the risk of developing left ventricular hypertrophy, and other intermediate cardiovascular effects in undiagnosed and untreated children. The Task Force acknowledges, however, that prospective longitudinal outcome studies in untreated children and adolescents are lacking.

Hypertensive children often grow up to be hypertensive adults
A prospective cohort study showed that children with elevated BP had a greater likelihood of adult hypertension than children with normal BP. Investigators followed 2445 children 7 to 18 years of age to determine whether elevated BP in childhood correlated with increased BP in adulthood.

Investigators obtained BP, height, and weight measurements biennially during the children’s school years and when they were young adults between 20 and 30 years of age. Twelve to 13 years later, 24% of children with BP above the 90th percentile still had BP above the 90th percentile (relative risk [RR]=2.4; P<.001) and 39% had BP above the 80th percentile (RR=1.9; P<.001). Ninety-four percent of children with more than 3 normal readings during the study were normotensive as young adults. Children with one or 2 abnormal readings had a 17% and 24% chance, respectively, of having hypertension as adults (P<.001).⁴

High childhood systolic BP may predict CAD in adulthood
A retrospective study evaluated 126 children 10 to 17 years of age who were admitted to the hospital for an elective surgical procedure between 1950 and 1967. Children with documented BP readings at admission were eligible for the study; children with preexisting cardiac and renal disease were excluded. Investigators reassessed patients as adults (age range 42-68 years); the mean follow-up period was 42 years.

Mean BP was 125/80 mm Hg at admission and 133/75 mm Hg at follow-up. Univariate logistic regression analysis showed a significant association between systolic BP in childhood and coronary artery disease at follow-up (odds ratio [OR]=1.052; 95% confidence interval [CI], 1.005-1.101; P=.027). Children with systolic BP at or above the 95th percentile had a 4-fold increase in coronary artery disease at follow-up compared with children whose systolic BP was below the 95th percentile (29% vs 7%, P=.03). Investigators also found an association between elevated BP in childhood and a diagnosis of hypertension at follow-up (P=.007).

Limitations of the study included small sample size, selection bias, changes in the definition of hypertension during the 4 decades since the study began, and limited childhood BP data (a single measurement at admission for surgery).⁵

Parents of hypertensive children are likely to be hypertensive themselves
Screening BP in children has the potential to identify families at increased risk for cardiovascular disease. A case series found a high incidence of hypertension among the parents of children with elevated BP. Investigators measured several risk factors, including BP in 141 children (mean age 10.5±3.4 years) and 108 parents (at least one a biological parent, mean age 38.5±7.5 years). They obtained 2 BP readings 15 to 30 minutes apart.

Parents of children with BPs at or above the 95th percentile had a 15-fold greater likelihood of hypertension themselves (OR=14.7; 95% CI, 3.02-71.56; P=.009, positive predictive value=75%; negative predictive value=81%).⁶ Limitations of the study included small sample size, high prevalence of obesity and black ethnicity in the study population (a population with a greater incidence of hypertension), and only 2 BP measurements in the same day, which isn’t diagnostic for hypertension.

Recommendations

The American College of Obstetricians and Gynecologists recommends screening girls for hypertension between 13 and 15 years of age.⁷

The American Academy of Family Physicians concludes that the evidence is insufficient to recommend for or against routine screening for hypertension in children and adolescents to reduce the risk of cardiovascular disease.⁸

The European Society of Hypertension and European Society of Cardiology recommend that children older than 3 years have auscultatory BP measurements at each clinic visit.⁹

Evidence summary

A double-blind RCT found that probiotic vaginal suppositories reduce the incidence of recurrent BV. Investigators randomized 120 Chinese women, 18 to 55 years of age with a history of 2 or more episodes of BV in the previous year, to use suppositories containing either probiotics (Lactobacillus rhamnosus,L acidophilus, and Streptococcus thermophilus, total of 8×10⁹ colony-forming units [cfu]) or placebo.¹ All the women used suppositories daily for a week, stopped for a week, and then used them for another week.

Fewer women who used probiotic suppositories had recurrences of BV on examination during the following 2 months than women who used placebo (16% vs 45%; P<.001; number needed to treat [NNT]=3.4), and fewer reported recurrences in telephone interviews 2 to 11 months after treatment (11% vs 28%; P<.05; NNT=5.8). Interviewers recorded two-thirds fewer complaints of discharge and malodor among women who used probiotics than among women who used placebo (P<.05 for both comparisons).

But another RCT finds no effect on recurrent BV or VVC
Another RCT treated 95 women 18 to 45 years of age with clindamycin ovules (for BV) or clotrimazole suppositories (for VVC) and, after 5 days, randomized them to use probiotic suppositories (Lactobacillus species, 10⁸-10¹⁰ cfu) or placebo for 5 more days.²

Probiotic suppositories after treatment didn’t reduce clinician-diagnosed recurrences of either BV or VVC compared with placebo (7% vs 17% after 2-3 days; 22% vs 29% after the first menstrual cycle; P=not significant for both). Probiotics did reduce self-reported malodorous discharge, however (P=.03). Probiotics didn’t produce adverse effects.

Probiotic yogurt decreases recurrent BV but not VVC in an RCT
An RCT that randomized 46 women, 20 to 39 years of age with a history of 4 or more episodes of BV or VVC in the previous year, to eat L acidophilus-enriched yogurt (10⁸ cfu) or pasteurized yogurt daily for 2 months found that consuming probiotic-containing yogurt reduced the incidence of recurrent BV but not VVC.³

Women who ate L acidophilus yogurt had fewer episodes of clinician-diagnosed BV at 1 month than women who ate pasteurized yogurt (24% vs 53%; P<.05) and also at 2 months (4% vs 36%; P<.05). However, they didn’t have significantly fewer episodes of VVC (43% vs 37% at 1 month, 21% vs 29% at 2 months; P=not significant for both). Investigators reported no adverse effects.

Small, flawed trial finds fewer episodes of VVC with yogurt
An unblinded crossover trial found that daily consumption of probiotic yogurt reduced VVC recurrences in women with a history of the infection. Investigators randomized 33 women 24 to 50 years of age to eat either 8 ounces a day of yogurt (with L acidophilus, 10⁸ cfu) or a yogurt-free diet.⁴ After 6 months, the groups switched. Investigators saw all patients monthly.

Women who ate yogurt had fewer episodes of VVC than women who didn’t (0.4 vs 2.5 over 6 months; P<.001) and reported no adverse effects. The study was flawed by small size and high attrition rates (only 13 women completed the trial).

Recommendations

The World Health Organization says some clinical evidence suggests that oral and vaginal administration of lactobacilli can eradicate asymptomatic and symptomatic BV. Supporting evidence for prevention of recurrent BV or VVC by probiotics is limited.⁵

A literature review by the Natural Standard Research Collaboration states that insufficient evidence exists to recommend probiotics for treating or preventing bacterial vaginosis and that preventing or treating vaginal yeast infections with probiotics hasn’t been adequately studied.⁶

Evidence summary

A double-blind RCT found that probiotic vaginal suppositories reduce the incidence of recurrent BV. Investigators randomized 120 Chinese women, 18 to 55 years of age with a history of 2 or more episodes of BV in the previous year, to use suppositories containing either probiotics (Lactobacillus rhamnosus,L acidophilus, and Streptococcus thermophilus, total of 8×10⁹ colony-forming units [cfu]) or placebo.¹ All the women used suppositories daily for a week, stopped for a week, and then used them for another week.

Fewer women who used probiotic suppositories had recurrences of BV on examination during the following 2 months than women who used placebo (16% vs 45%; P<.001; number needed to treat [NNT]=3.4), and fewer reported recurrences in telephone interviews 2 to 11 months after treatment (11% vs 28%; P<.05; NNT=5.8). Interviewers recorded two-thirds fewer complaints of discharge and malodor among women who used probiotics than among women who used placebo (P<.05 for both comparisons).

But another RCT finds no effect on recurrent BV or VVC
Another RCT treated 95 women 18 to 45 years of age with clindamycin ovules (for BV) or clotrimazole suppositories (for VVC) and, after 5 days, randomized them to use probiotic suppositories (Lactobacillus species, 10⁸-10¹⁰ cfu) or placebo for 5 more days.²

Probiotic suppositories after treatment didn’t reduce clinician-diagnosed recurrences of either BV or VVC compared with placebo (7% vs 17% after 2-3 days; 22% vs 29% after the first menstrual cycle; P=not significant for both). Probiotics did reduce self-reported malodorous discharge, however (P=.03). Probiotics didn’t produce adverse effects.

Probiotic yogurt decreases recurrent BV but not VVC in an RCT
An RCT that randomized 46 women, 20 to 39 years of age with a history of 4 or more episodes of BV or VVC in the previous year, to eat L acidophilus-enriched yogurt (10⁸ cfu) or pasteurized yogurt daily for 2 months found that consuming probiotic-containing yogurt reduced the incidence of recurrent BV but not VVC.³

Women who ate L acidophilus yogurt had fewer episodes of clinician-diagnosed BV at 1 month than women who ate pasteurized yogurt (24% vs 53%; P<.05) and also at 2 months (4% vs 36%; P<.05). However, they didn’t have significantly fewer episodes of VVC (43% vs 37% at 1 month, 21% vs 29% at 2 months; P=not significant for both). Investigators reported no adverse effects.

Small, flawed trial finds fewer episodes of VVC with yogurt
An unblinded crossover trial found that daily consumption of probiotic yogurt reduced VVC recurrences in women with a history of the infection. Investigators randomized 33 women 24 to 50 years of age to eat either 8 ounces a day of yogurt (with L acidophilus, 10⁸ cfu) or a yogurt-free diet.⁴ After 6 months, the groups switched. Investigators saw all patients monthly.

Women who ate yogurt had fewer episodes of VVC than women who didn’t (0.4 vs 2.5 over 6 months; P<.001) and reported no adverse effects. The study was flawed by small size and high attrition rates (only 13 women completed the trial).

Recommendations

The World Health Organization says some clinical evidence suggests that oral and vaginal administration of lactobacilli can eradicate asymptomatic and symptomatic BV. Supporting evidence for prevention of recurrent BV or VVC by probiotics is limited.⁵

A literature review by the Natural Standard Research Collaboration states that insufficient evidence exists to recommend probiotics for treating or preventing bacterial vaginosis and that preventing or treating vaginal yeast infections with probiotics hasn’t been adequately studied.⁶

Evidence summary

A double-blind RCT found that probiotic vaginal suppositories reduce the incidence of recurrent BV. Investigators randomized 120 Chinese women, 18 to 55 years of age with a history of 2 or more episodes of BV in the previous year, to use suppositories containing either probiotics (Lactobacillus rhamnosus,L acidophilus, and Streptococcus thermophilus, total of 8×10⁹ colony-forming units [cfu]) or placebo.¹ All the women used suppositories daily for a week, stopped for a week, and then used them for another week.

Fewer women who used probiotic suppositories had recurrences of BV on examination during the following 2 months than women who used placebo (16% vs 45%; P<.001; number needed to treat [NNT]=3.4), and fewer reported recurrences in telephone interviews 2 to 11 months after treatment (11% vs 28%; P<.05; NNT=5.8). Interviewers recorded two-thirds fewer complaints of discharge and malodor among women who used probiotics than among women who used placebo (P<.05 for both comparisons).

But another RCT finds no effect on recurrent BV or VVC
Another RCT treated 95 women 18 to 45 years of age with clindamycin ovules (for BV) or clotrimazole suppositories (for VVC) and, after 5 days, randomized them to use probiotic suppositories (Lactobacillus species, 10⁸-10¹⁰ cfu) or placebo for 5 more days.²

Probiotic suppositories after treatment didn’t reduce clinician-diagnosed recurrences of either BV or VVC compared with placebo (7% vs 17% after 2-3 days; 22% vs 29% after the first menstrual cycle; P=not significant for both). Probiotics did reduce self-reported malodorous discharge, however (P=.03). Probiotics didn’t produce adverse effects.

Probiotic yogurt decreases recurrent BV but not VVC in an RCT
An RCT that randomized 46 women, 20 to 39 years of age with a history of 4 or more episodes of BV or VVC in the previous year, to eat L acidophilus-enriched yogurt (10⁸ cfu) or pasteurized yogurt daily for 2 months found that consuming probiotic-containing yogurt reduced the incidence of recurrent BV but not VVC.³

Women who ate L acidophilus yogurt had fewer episodes of clinician-diagnosed BV at 1 month than women who ate pasteurized yogurt (24% vs 53%; P<.05) and also at 2 months (4% vs 36%; P<.05). However, they didn’t have significantly fewer episodes of VVC (43% vs 37% at 1 month, 21% vs 29% at 2 months; P=not significant for both). Investigators reported no adverse effects.

Small, flawed trial finds fewer episodes of VVC with yogurt
An unblinded crossover trial found that daily consumption of probiotic yogurt reduced VVC recurrences in women with a history of the infection. Investigators randomized 33 women 24 to 50 years of age to eat either 8 ounces a day of yogurt (with L acidophilus, 10⁸ cfu) or a yogurt-free diet.⁴ After 6 months, the groups switched. Investigators saw all patients monthly.

Women who ate yogurt had fewer episodes of VVC than women who didn’t (0.4 vs 2.5 over 6 months; P<.001) and reported no adverse effects. The study was flawed by small size and high attrition rates (only 13 women completed the trial).

Recommendations

The World Health Organization says some clinical evidence suggests that oral and vaginal administration of lactobacilli can eradicate asymptomatic and symptomatic BV. Supporting evidence for prevention of recurrent BV or VVC by probiotics is limited.⁵

A literature review by the Natural Standard Research Collaboration states that insufficient evidence exists to recommend probiotics for treating or preventing bacterial vaginosis and that preventing or treating vaginal yeast infections with probiotics hasn’t been adequately studied.⁶

Evidence summary

A prospective RCT found that medical therapy, radioablation with iodine-131 (¹³¹I), and surgery produced similar control of Graves’ hyperthyroidism in 179 patients.¹ Investigators stratified patients by age, assigning younger patients (20-34 years; N=60) to antithyroid medication (methimazole and a β-blocker) for 18 months or subtotal thyroidectomy and older patients (35-55 years; N=119) to 18 months of antithyroid medication, subtotal thyroidectomy, or ¹³¹I radioablation.

After 6 weeks, all therapies produced serum triiodothyronine levels of less than 2.5 nmol/L (data extracted from table; no comparison statistic given). Patients were followed for 48 to 121 months (average follow-up time not given). Investigators found no significant differences in sick leave (72 vs 83 days for medical treatment compared with radioablation; no comparison statistic given) or patient satisfaction (95% for both medical treatment and radioablation; no comparison statistic given).

Medication (initially methimazole) was changed in 16% of patients because of adverse effects. More than a third of patients relapsed after medications were stopped (time to relapse 1-57 months); 21% relapsed after a single ¹³¹I treatment (time to relapse 5-16 months).

In another study, radioablation outperforms medication
A retrospective case series found that radioablation resolved hyperthyroidism more often than medical therapy among 194 consecutive Saudi Arabian patients (mean age 32 years) diagnosed with Graves’ disease and followed for an average of 50 months.² One dose of radioiodine (13-15 mCi) cured hyperthyroidism in 83% of patients, whereas 18 months of medical therapy produced remission lasting at least 6 months past the end of therapy in only 26% of patients (no comparison statistic given).

The presence of TAO at diagnosis increased the likelihood of radioablation failure (odds ratio for failure to respond to single dose of radioiodine=6.4; 95% confidence interval [CI], 1.51-24.4; P<.01). A major weakness of the study was that the investigators didn’t describe the medication therapy clearly.

More patients develop TAO after radioablation than medical therapy
An RCT found that radioablation is more commonly associated with development of TAO than medical therapy.³ When investigators randomized 313 patients to receive ¹³¹I radioablation or medical therapy for 18 months and followed them for as long as 4 years, more patients receiving radioablation developed TAO (38% compared with 18% for medical therapy, using intention-to-treat analysis; P<.001; number needed to harm [NNH]=5). Twenty-five percent of patients initially receiving medical therapy later underwent radioablation, but these patients didn’t develop TAO at a higher rate.

An earlier meta-analysis of 2 RCTs (N=189) also found an increased risk of TAO with radioablation compared with medical therapy.⁴ Patients receiving radioablation were more likely to develop TAO (18% vs 4%; relative risk [RR]=4.2; 95% CI, 2.0-8.8; NNH=7) and more likely to develop severe TAO (10% vs 1.6%; RR=4.4; 95% CI, 1.3-15; NNH=12). Adjunctive use of steroids with radioablation didn’t alter the risk of new TAO. However, steroid prophylaxis in patients with preexisting TAO significantly reduced the risk of progression after radioablation (RR=0.03; 95% CI, 0.00-0.24). The authors of the meta-analysis didn’t evaluate the quality of the RCTs.

Despite low neoplasia risk, radioablation isn’t for young children
Expert guidelines state that the goal of radioablation is to induce lifelong hypothyroidism, which is managed with thyroid hormone replacement.⁵ The risk of neoplasia after radioablation is believed to be low with appropriate dosing. However, based on “theoretical concerns,” experts don’t recommend using radioiodine in children younger than 5 years and advise limited use in children 5 to 10 years of age.⁵

Medication adverse effects include rashes, transient agranulocytosis
A Cochrane review with 7 RCTs (N=620) describing withdrawal rates for patients receiving medication for Graves’ disease found that 9% to 16% of patients discontinued treatment because of adverse effects.⁶ Rashes were the most common adverse effect (6%-10% of patients), but as many as 3% of patients developed transient agranulocytosis. In addition, patients on medication need frequent blood tests to monitor for thyroid activity and potential toxicity.

Recommendations

The guidelines of the American Thyroid Association and the American Association of Clinical Endocrinologists state that overt Graves’ hyperthyroidism may be treated with any of the following: ¹³¹I radioablation, antithyroid medication, or thyroidectomy.³ Patient characteristics (pregnancy, mild disease, goiter compression symptoms) should help determine the appropriate option in any given case.

Evidence summary

A prospective RCT found that medical therapy, radioablation with iodine-131 (¹³¹I), and surgery produced similar control of Graves’ hyperthyroidism in 179 patients.¹ Investigators stratified patients by age, assigning younger patients (20-34 years; N=60) to antithyroid medication (methimazole and a β-blocker) for 18 months or subtotal thyroidectomy and older patients (35-55 years; N=119) to 18 months of antithyroid medication, subtotal thyroidectomy, or ¹³¹I radioablation.

After 6 weeks, all therapies produced serum triiodothyronine levels of less than 2.5 nmol/L (data extracted from table; no comparison statistic given). Patients were followed for 48 to 121 months (average follow-up time not given). Investigators found no significant differences in sick leave (72 vs 83 days for medical treatment compared with radioablation; no comparison statistic given) or patient satisfaction (95% for both medical treatment and radioablation; no comparison statistic given).

Medication (initially methimazole) was changed in 16% of patients because of adverse effects. More than a third of patients relapsed after medications were stopped (time to relapse 1-57 months); 21% relapsed after a single ¹³¹I treatment (time to relapse 5-16 months).

In another study, radioablation outperforms medication
A retrospective case series found that radioablation resolved hyperthyroidism more often than medical therapy among 194 consecutive Saudi Arabian patients (mean age 32 years) diagnosed with Graves’ disease and followed for an average of 50 months.² One dose of radioiodine (13-15 mCi) cured hyperthyroidism in 83% of patients, whereas 18 months of medical therapy produced remission lasting at least 6 months past the end of therapy in only 26% of patients (no comparison statistic given).

The presence of TAO at diagnosis increased the likelihood of radioablation failure (odds ratio for failure to respond to single dose of radioiodine=6.4; 95% confidence interval [CI], 1.51-24.4; P<.01). A major weakness of the study was that the investigators didn’t describe the medication therapy clearly.

More patients develop TAO after radioablation than medical therapy
An RCT found that radioablation is more commonly associated with development of TAO than medical therapy.³ When investigators randomized 313 patients to receive ¹³¹I radioablation or medical therapy for 18 months and followed them for as long as 4 years, more patients receiving radioablation developed TAO (38% compared with 18% for medical therapy, using intention-to-treat analysis; P<.001; number needed to harm [NNH]=5). Twenty-five percent of patients initially receiving medical therapy later underwent radioablation, but these patients didn’t develop TAO at a higher rate.

An earlier meta-analysis of 2 RCTs (N=189) also found an increased risk of TAO with radioablation compared with medical therapy.⁴ Patients receiving radioablation were more likely to develop TAO (18% vs 4%; relative risk [RR]=4.2; 95% CI, 2.0-8.8; NNH=7) and more likely to develop severe TAO (10% vs 1.6%; RR=4.4; 95% CI, 1.3-15; NNH=12). Adjunctive use of steroids with radioablation didn’t alter the risk of new TAO. However, steroid prophylaxis in patients with preexisting TAO significantly reduced the risk of progression after radioablation (RR=0.03; 95% CI, 0.00-0.24). The authors of the meta-analysis didn’t evaluate the quality of the RCTs.

Despite low neoplasia risk, radioablation isn’t for young children
Expert guidelines state that the goal of radioablation is to induce lifelong hypothyroidism, which is managed with thyroid hormone replacement.⁵ The risk of neoplasia after radioablation is believed to be low with appropriate dosing. However, based on “theoretical concerns,” experts don’t recommend using radioiodine in children younger than 5 years and advise limited use in children 5 to 10 years of age.⁵

Medication adverse effects include rashes, transient agranulocytosis
A Cochrane review with 7 RCTs (N=620) describing withdrawal rates for patients receiving medication for Graves’ disease found that 9% to 16% of patients discontinued treatment because of adverse effects.⁶ Rashes were the most common adverse effect (6%-10% of patients), but as many as 3% of patients developed transient agranulocytosis. In addition, patients on medication need frequent blood tests to monitor for thyroid activity and potential toxicity.

Recommendations

The guidelines of the American Thyroid Association and the American Association of Clinical Endocrinologists state that overt Graves’ hyperthyroidism may be treated with any of the following: ¹³¹I radioablation, antithyroid medication, or thyroidectomy.³ Patient characteristics (pregnancy, mild disease, goiter compression symptoms) should help determine the appropriate option in any given case.

Evidence summary

A prospective RCT found that medical therapy, radioablation with iodine-131 (¹³¹I), and surgery produced similar control of Graves’ hyperthyroidism in 179 patients.¹ Investigators stratified patients by age, assigning younger patients (20-34 years; N=60) to antithyroid medication (methimazole and a β-blocker) for 18 months or subtotal thyroidectomy and older patients (35-55 years; N=119) to 18 months of antithyroid medication, subtotal thyroidectomy, or ¹³¹I radioablation.

After 6 weeks, all therapies produced serum triiodothyronine levels of less than 2.5 nmol/L (data extracted from table; no comparison statistic given). Patients were followed for 48 to 121 months (average follow-up time not given). Investigators found no significant differences in sick leave (72 vs 83 days for medical treatment compared with radioablation; no comparison statistic given) or patient satisfaction (95% for both medical treatment and radioablation; no comparison statistic given).

Medication (initially methimazole) was changed in 16% of patients because of adverse effects. More than a third of patients relapsed after medications were stopped (time to relapse 1-57 months); 21% relapsed after a single ¹³¹I treatment (time to relapse 5-16 months).

In another study, radioablation outperforms medication
A retrospective case series found that radioablation resolved hyperthyroidism more often than medical therapy among 194 consecutive Saudi Arabian patients (mean age 32 years) diagnosed with Graves’ disease and followed for an average of 50 months.² One dose of radioiodine (13-15 mCi) cured hyperthyroidism in 83% of patients, whereas 18 months of medical therapy produced remission lasting at least 6 months past the end of therapy in only 26% of patients (no comparison statistic given).

The presence of TAO at diagnosis increased the likelihood of radioablation failure (odds ratio for failure to respond to single dose of radioiodine=6.4; 95% confidence interval [CI], 1.51-24.4; P<.01). A major weakness of the study was that the investigators didn’t describe the medication therapy clearly.

More patients develop TAO after radioablation than medical therapy
An RCT found that radioablation is more commonly associated with development of TAO than medical therapy.³ When investigators randomized 313 patients to receive ¹³¹I radioablation or medical therapy for 18 months and followed them for as long as 4 years, more patients receiving radioablation developed TAO (38% compared with 18% for medical therapy, using intention-to-treat analysis; P<.001; number needed to harm [NNH]=5). Twenty-five percent of patients initially receiving medical therapy later underwent radioablation, but these patients didn’t develop TAO at a higher rate.

An earlier meta-analysis of 2 RCTs (N=189) also found an increased risk of TAO with radioablation compared with medical therapy.⁴ Patients receiving radioablation were more likely to develop TAO (18% vs 4%; relative risk [RR]=4.2; 95% CI, 2.0-8.8; NNH=7) and more likely to develop severe TAO (10% vs 1.6%; RR=4.4; 95% CI, 1.3-15; NNH=12). Adjunctive use of steroids with radioablation didn’t alter the risk of new TAO. However, steroid prophylaxis in patients with preexisting TAO significantly reduced the risk of progression after radioablation (RR=0.03; 95% CI, 0.00-0.24). The authors of the meta-analysis didn’t evaluate the quality of the RCTs.

Despite low neoplasia risk, radioablation isn’t for young children
Expert guidelines state that the goal of radioablation is to induce lifelong hypothyroidism, which is managed with thyroid hormone replacement.⁵ The risk of neoplasia after radioablation is believed to be low with appropriate dosing. However, based on “theoretical concerns,” experts don’t recommend using radioiodine in children younger than 5 years and advise limited use in children 5 to 10 years of age.⁵

Medication adverse effects include rashes, transient agranulocytosis
A Cochrane review with 7 RCTs (N=620) describing withdrawal rates for patients receiving medication for Graves’ disease found that 9% to 16% of patients discontinued treatment because of adverse effects.⁶ Rashes were the most common adverse effect (6%-10% of patients), but as many as 3% of patients developed transient agranulocytosis. In addition, patients on medication need frequent blood tests to monitor for thyroid activity and potential toxicity.

Recommendations

The guidelines of the American Thyroid Association and the American Association of Clinical Endocrinologists state that overt Graves’ hyperthyroidism may be treated with any of the following: ¹³¹I radioablation, antithyroid medication, or thyroidectomy.³ Patient characteristics (pregnancy, mild disease, goiter compression symptoms) should help determine the appropriate option in any given case.

Evidence summary

A Cochrane review of antibiotics for the common cold that included 5 RCTs with a total of 772 participants with purulent nasal discharge found no benefit from antibiotics.¹ The relative risk (RR) for persistent acute purulent rhinitis with antibiotics compared with placebo was 0.63 (95% confidence interval [CI], 0.38-1.07; P=.087). The antibiotic groups showed an increase in adverse effects, with an RR of 1.46 (95% CI, 1.01-1.94; P=.047).

Benefits of antibiotics tempered by adverse effects
A meta-analysis of 6 RCTs with more than 1400 subjects showed persistent nasal discharge at 5 to 8 days, on average, in 23% of patients who received antibiotics compared with 46% of patients who received placebo (RR of benefits=1.18; 95% CI, 1.05-1.33; P=.05).² Most subjects were between 12 and 50 years of age; 2 of the trials included children between 2 months and 16 years of age. All subjects had symptoms for fewer than 10 days.

The adverse effects of antibiotic treatment, primarily rash and diarrhea, were also addressed (RR of adverse effects=1.46; 95% CI, 1.10-1.94; P=.028). Given the overlap of the number needed to treat (7-15) and number needed to harm (12-78), the authors concluded that most patients get better without antibiotics, supporting “no antibiotic as first line” treatment advice.

Other studies show minimal benefit for antibiotics
A meta-analysis of 9 placebo-controlled RCTs (2640 adult subjects with rhinosinusitis-like complaints) found that antibiotics provided minimal benefit. For patients with visible purulent drainage in the pharynx, the NNT overlapped with the NNH; patients without visible purulent discharge showed even less benefit from antibiotics.³

Clinical improvement is insufficient to recommend antibiotic treatment
Three double-blinded RCTs studied patients older than 12 years who presented to a family practice clinic complaining of purulent rhinitis.^4-6 All 3 studies compared amoxicillin treatment with placebo; outcomes were based primarily on patient diaries that recorded symptoms, including nasal discharge.

The first study randomized 135 patients to either amoxicillin (n=67) or placebo (n=68) for 10 days.⁴ At the end of 2 weeks, both groups had similar rates of symptom improvement—although in a subgroup of 57 patients who had complete symptom resolution at 2 weeks, the median number of days until resolution of purulent nasal discharge was 8 in the amoxicillin group compared with 12 days for the placebo group (P=.039). The authors could not identify clinical characteristics favoring antibiotic treatment.

In the second study, 207 patients received amoxicillin and 209 placebo.⁵ After 10 days of therapy, symptom resolution rates were not significantly different (35% for amoxicillin vs 29% for placebo). However, patients in the amoxicillin group had quicker resolution of purulent nasal discharge (9 vs 14 days for 75% of patients to be free of that symptom; P=.007).⁵

The third study (240 adults) didn’t find a significant decrease in duration of purulent nasal discharge in the antibiotic group compared with the placebo group.⁶

Despite the findings of decreased duration of purulent nasal discharge in the first 2 studies, the authors of all 3 studies concluded that the clinical difference in improvement between antibiotic and placebo groups was not enough to recommend treatment with antibiotics. Although the trials didn’t measure adverse outcomes, the authors advised clinicians to consider the potential for adverse reactions before recommending antibiotic treatment.

Recommendations

Both the American Academy of Otolaryngology and the American Academy of Allergy, Asthma, and Immunology recommend watchful waiting without antibiotics for acute sinusitis with mild pain or temperature lower than 101°F and consideration of antibiotics only if symptoms worsen or fail to improve by 7 days after diagnosis. Neither group offers specific recommendations regarding patients with purulent discharge.^7,8

The Centers for Disease Control and Prevention recommend reserving antibiotic treatment of acute bacterial rhinosinusitis for patients with symptoms lasting longer than 7 days and patients who have unilateral symptoms with purulent nasal discharge.⁹

Evidence summary

A Cochrane review of antibiotics for the common cold that included 5 RCTs with a total of 772 participants with purulent nasal discharge found no benefit from antibiotics.¹ The relative risk (RR) for persistent acute purulent rhinitis with antibiotics compared with placebo was 0.63 (95% confidence interval [CI], 0.38-1.07; P=.087). The antibiotic groups showed an increase in adverse effects, with an RR of 1.46 (95% CI, 1.01-1.94; P=.047).

Benefits of antibiotics tempered by adverse effects
A meta-analysis of 6 RCTs with more than 1400 subjects showed persistent nasal discharge at 5 to 8 days, on average, in 23% of patients who received antibiotics compared with 46% of patients who received placebo (RR of benefits=1.18; 95% CI, 1.05-1.33; P=.05).² Most subjects were between 12 and 50 years of age; 2 of the trials included children between 2 months and 16 years of age. All subjects had symptoms for fewer than 10 days.

The adverse effects of antibiotic treatment, primarily rash and diarrhea, were also addressed (RR of adverse effects=1.46; 95% CI, 1.10-1.94; P=.028). Given the overlap of the number needed to treat (7-15) and number needed to harm (12-78), the authors concluded that most patients get better without antibiotics, supporting “no antibiotic as first line” treatment advice.

Other studies show minimal benefit for antibiotics
A meta-analysis of 9 placebo-controlled RCTs (2640 adult subjects with rhinosinusitis-like complaints) found that antibiotics provided minimal benefit. For patients with visible purulent drainage in the pharynx, the NNT overlapped with the NNH; patients without visible purulent discharge showed even less benefit from antibiotics.³

Clinical improvement is insufficient to recommend antibiotic treatment
Three double-blinded RCTs studied patients older than 12 years who presented to a family practice clinic complaining of purulent rhinitis.^4-6 All 3 studies compared amoxicillin treatment with placebo; outcomes were based primarily on patient diaries that recorded symptoms, including nasal discharge.

The first study randomized 135 patients to either amoxicillin (n=67) or placebo (n=68) for 10 days.⁴ At the end of 2 weeks, both groups had similar rates of symptom improvement—although in a subgroup of 57 patients who had complete symptom resolution at 2 weeks, the median number of days until resolution of purulent nasal discharge was 8 in the amoxicillin group compared with 12 days for the placebo group (P=.039). The authors could not identify clinical characteristics favoring antibiotic treatment.

In the second study, 207 patients received amoxicillin and 209 placebo.⁵ After 10 days of therapy, symptom resolution rates were not significantly different (35% for amoxicillin vs 29% for placebo). However, patients in the amoxicillin group had quicker resolution of purulent nasal discharge (9 vs 14 days for 75% of patients to be free of that symptom; P=.007).⁵

The third study (240 adults) didn’t find a significant decrease in duration of purulent nasal discharge in the antibiotic group compared with the placebo group.⁶

Despite the findings of decreased duration of purulent nasal discharge in the first 2 studies, the authors of all 3 studies concluded that the clinical difference in improvement between antibiotic and placebo groups was not enough to recommend treatment with antibiotics. Although the trials didn’t measure adverse outcomes, the authors advised clinicians to consider the potential for adverse reactions before recommending antibiotic treatment.

Recommendations

Both the American Academy of Otolaryngology and the American Academy of Allergy, Asthma, and Immunology recommend watchful waiting without antibiotics for acute sinusitis with mild pain or temperature lower than 101°F and consideration of antibiotics only if symptoms worsen or fail to improve by 7 days after diagnosis. Neither group offers specific recommendations regarding patients with purulent discharge.^7,8

The Centers for Disease Control and Prevention recommend reserving antibiotic treatment of acute bacterial rhinosinusitis for patients with symptoms lasting longer than 7 days and patients who have unilateral symptoms with purulent nasal discharge.⁹

Evidence summary

A Cochrane review of antibiotics for the common cold that included 5 RCTs with a total of 772 participants with purulent nasal discharge found no benefit from antibiotics.¹ The relative risk (RR) for persistent acute purulent rhinitis with antibiotics compared with placebo was 0.63 (95% confidence interval [CI], 0.38-1.07; P=.087). The antibiotic groups showed an increase in adverse effects, with an RR of 1.46 (95% CI, 1.01-1.94; P=.047).

Benefits of antibiotics tempered by adverse effects
A meta-analysis of 6 RCTs with more than 1400 subjects showed persistent nasal discharge at 5 to 8 days, on average, in 23% of patients who received antibiotics compared with 46% of patients who received placebo (RR of benefits=1.18; 95% CI, 1.05-1.33; P=.05).² Most subjects were between 12 and 50 years of age; 2 of the trials included children between 2 months and 16 years of age. All subjects had symptoms for fewer than 10 days.

The adverse effects of antibiotic treatment, primarily rash and diarrhea, were also addressed (RR of adverse effects=1.46; 95% CI, 1.10-1.94; P=.028). Given the overlap of the number needed to treat (7-15) and number needed to harm (12-78), the authors concluded that most patients get better without antibiotics, supporting “no antibiotic as first line” treatment advice.

Other studies show minimal benefit for antibiotics
A meta-analysis of 9 placebo-controlled RCTs (2640 adult subjects with rhinosinusitis-like complaints) found that antibiotics provided minimal benefit. For patients with visible purulent drainage in the pharynx, the NNT overlapped with the NNH; patients without visible purulent discharge showed even less benefit from antibiotics.³

Clinical improvement is insufficient to recommend antibiotic treatment
Three double-blinded RCTs studied patients older than 12 years who presented to a family practice clinic complaining of purulent rhinitis.^4-6 All 3 studies compared amoxicillin treatment with placebo; outcomes were based primarily on patient diaries that recorded symptoms, including nasal discharge.

The first study randomized 135 patients to either amoxicillin (n=67) or placebo (n=68) for 10 days.⁴ At the end of 2 weeks, both groups had similar rates of symptom improvement—although in a subgroup of 57 patients who had complete symptom resolution at 2 weeks, the median number of days until resolution of purulent nasal discharge was 8 in the amoxicillin group compared with 12 days for the placebo group (P=.039). The authors could not identify clinical characteristics favoring antibiotic treatment.

In the second study, 207 patients received amoxicillin and 209 placebo.⁵ After 10 days of therapy, symptom resolution rates were not significantly different (35% for amoxicillin vs 29% for placebo). However, patients in the amoxicillin group had quicker resolution of purulent nasal discharge (9 vs 14 days for 75% of patients to be free of that symptom; P=.007).⁵

The third study (240 adults) didn’t find a significant decrease in duration of purulent nasal discharge in the antibiotic group compared with the placebo group.⁶

Despite the findings of decreased duration of purulent nasal discharge in the first 2 studies, the authors of all 3 studies concluded that the clinical difference in improvement between antibiotic and placebo groups was not enough to recommend treatment with antibiotics. Although the trials didn’t measure adverse outcomes, the authors advised clinicians to consider the potential for adverse reactions before recommending antibiotic treatment.

Recommendations

Both the American Academy of Otolaryngology and the American Academy of Allergy, Asthma, and Immunology recommend watchful waiting without antibiotics for acute sinusitis with mild pain or temperature lower than 101°F and consideration of antibiotics only if symptoms worsen or fail to improve by 7 days after diagnosis. Neither group offers specific recommendations regarding patients with purulent discharge.^7,8

The Centers for Disease Control and Prevention recommend reserving antibiotic treatment of acute bacterial rhinosinusitis for patients with symptoms lasting longer than 7 days and patients who have unilateral symptoms with purulent nasal discharge.⁹

Evidence summary

The evidence linking nail changes with nutritional deficiencies and toxic exposures is sparse, dated, and of low quality. Conditions associated with nail changes include iron deficiency, B₁₂ deficiency, hypoalbuminemia, protein malnutrition, and PCB/PCDF exposures.

Iron deficiency is associated with koilonychia
A case-control study of 400 infants in a low-income well-baby clinic found 22 with koilonychia, a prevalence of approximately 5%. Randomly selected age-matched infants without koilonychia served as controls.

Infants with koilonychia had significantly lower hematocrit (30% vs 34%; P<.005), hemoglobin (9.4 vs 10.7 g/dL; P<.02), and serum iron (50 vs 84 mcg/dL; P<.001) than controls.¹

B₁₂ deficiency can discolor nails, but changes are reversible
Four articles describe 5 case reports of pigment changes to nails associated with B₁₂ deficiency, all of which resolved with B₁₂ therapy. Nail changes included brownish reticulate pigmentation,² longitudinal hyperpigmented streaks,³ bluish-black pigment of all nails with transverse longitudinal hyperpigmented streaks,⁴ and entirely blue nails.⁵ B₁₂ levels ranged from undetectable to 113 pg/mL.

Hypoalbuminemia linked to Muehrcke’s lines
A cross-sectional study of 72 patients selected on the basis of general cachectic appearance found 44 to have low serum albumin. Of those, 10 (23%) had Muehrcke’s lines. All of the patients with Muehreke’s lines had albumin levels <2.7 g/dL. None of the 28 patients with normal albumin had Muehrcke’s lines.⁶

Kwashiorkor can cause fingernail clubbing
In a cross-sectional study of 60 children 1 to 4 years of age diagnosed with kwashiorkor without evidence of tuberculosis infection, fingernail clubbing was found in 46 (76.7%). Clubbing was mild in 26 (43.3%) children, moderate in 19 (37.7%), and severe in one (1.7%).⁷ Kwashiorkor is extremely rare in children in the United States and other developed countries.

Transplacental chemical exposures found to deform and discolor nails
A case-control study compared finger- and toenail findings from more than 100 Taiwanese children exposed transplacentally to high levels of PCBs and PCDFs with nail findings from a comparable number of controls. The investigators looked at parental reports and physical examination results.⁸

Although the rates of nail abnormalities reported by parents of exposed children differed slightly from rates documented by physical examination, children exposed to PCBs and PCDFs had consistently higher rates of nail deformity than controls. The researchers examined 117 cases and 106 controls. They identified dystrophic fingernails in 15% of exposed children and 1% of controls (odds ratio [OR]=15.4; 95% confidence interval [CI], 2-119); dystrophic toenails occurred in 32% of exposed children and 18% of controls (OR=2.2; 95% CI, 1.2-4.1). The most common fingernail deformities were grooves and ridges. The most common toenail deformities were koilonychias, ridges, thickening, and pigmentation changes.⁸

PCB exposure is an issue in the United States and other developed countries, but at much lower levels than the accidental contamination. Whether lower levels of exposure cause nail changes isn’t known.

Recommendations

No recommendations are available. A dermatology textbook lists several nail changes associated with nutritional deficiencies and toxic exposures (TABLE).⁹

Table
Nail changes linked to nutrition deficits and toxins⁹

Evidence summary

The evidence linking nail changes with nutritional deficiencies and toxic exposures is sparse, dated, and of low quality. Conditions associated with nail changes include iron deficiency, B₁₂ deficiency, hypoalbuminemia, protein malnutrition, and PCB/PCDF exposures.

Iron deficiency is associated with koilonychia
A case-control study of 400 infants in a low-income well-baby clinic found 22 with koilonychia, a prevalence of approximately 5%. Randomly selected age-matched infants without koilonychia served as controls.

Infants with koilonychia had significantly lower hematocrit (30% vs 34%; P<.005), hemoglobin (9.4 vs 10.7 g/dL; P<.02), and serum iron (50 vs 84 mcg/dL; P<.001) than controls.¹

B₁₂ deficiency can discolor nails, but changes are reversible
Four articles describe 5 case reports of pigment changes to nails associated with B₁₂ deficiency, all of which resolved with B₁₂ therapy. Nail changes included brownish reticulate pigmentation,² longitudinal hyperpigmented streaks,³ bluish-black pigment of all nails with transverse longitudinal hyperpigmented streaks,⁴ and entirely blue nails.⁵ B₁₂ levels ranged from undetectable to 113 pg/mL.

Hypoalbuminemia linked to Muehrcke’s lines
A cross-sectional study of 72 patients selected on the basis of general cachectic appearance found 44 to have low serum albumin. Of those, 10 (23%) had Muehrcke’s lines. All of the patients with Muehreke’s lines had albumin levels <2.7 g/dL. None of the 28 patients with normal albumin had Muehrcke’s lines.⁶

Kwashiorkor can cause fingernail clubbing
In a cross-sectional study of 60 children 1 to 4 years of age diagnosed with kwashiorkor without evidence of tuberculosis infection, fingernail clubbing was found in 46 (76.7%). Clubbing was mild in 26 (43.3%) children, moderate in 19 (37.7%), and severe in one (1.7%).⁷ Kwashiorkor is extremely rare in children in the United States and other developed countries.

Transplacental chemical exposures found to deform and discolor nails
A case-control study compared finger- and toenail findings from more than 100 Taiwanese children exposed transplacentally to high levels of PCBs and PCDFs with nail findings from a comparable number of controls. The investigators looked at parental reports and physical examination results.⁸

Although the rates of nail abnormalities reported by parents of exposed children differed slightly from rates documented by physical examination, children exposed to PCBs and PCDFs had consistently higher rates of nail deformity than controls. The researchers examined 117 cases and 106 controls. They identified dystrophic fingernails in 15% of exposed children and 1% of controls (odds ratio [OR]=15.4; 95% confidence interval [CI], 2-119); dystrophic toenails occurred in 32% of exposed children and 18% of controls (OR=2.2; 95% CI, 1.2-4.1). The most common fingernail deformities were grooves and ridges. The most common toenail deformities were koilonychias, ridges, thickening, and pigmentation changes.⁸

PCB exposure is an issue in the United States and other developed countries, but at much lower levels than the accidental contamination. Whether lower levels of exposure cause nail changes isn’t known.

Recommendations

No recommendations are available. A dermatology textbook lists several nail changes associated with nutritional deficiencies and toxic exposures (TABLE).⁹

Table
Nail changes linked to nutrition deficits and toxins⁹

Evidence summary

The evidence linking nail changes with nutritional deficiencies and toxic exposures is sparse, dated, and of low quality. Conditions associated with nail changes include iron deficiency, B₁₂ deficiency, hypoalbuminemia, protein malnutrition, and PCB/PCDF exposures.

Iron deficiency is associated with koilonychia
A case-control study of 400 infants in a low-income well-baby clinic found 22 with koilonychia, a prevalence of approximately 5%. Randomly selected age-matched infants without koilonychia served as controls.

Infants with koilonychia had significantly lower hematocrit (30% vs 34%; P<.005), hemoglobin (9.4 vs 10.7 g/dL; P<.02), and serum iron (50 vs 84 mcg/dL; P<.001) than controls.¹

B₁₂ deficiency can discolor nails, but changes are reversible
Four articles describe 5 case reports of pigment changes to nails associated with B₁₂ deficiency, all of which resolved with B₁₂ therapy. Nail changes included brownish reticulate pigmentation,² longitudinal hyperpigmented streaks,³ bluish-black pigment of all nails with transverse longitudinal hyperpigmented streaks,⁴ and entirely blue nails.⁵ B₁₂ levels ranged from undetectable to 113 pg/mL.

Hypoalbuminemia linked to Muehrcke’s lines
A cross-sectional study of 72 patients selected on the basis of general cachectic appearance found 44 to have low serum albumin. Of those, 10 (23%) had Muehrcke’s lines. All of the patients with Muehreke’s lines had albumin levels <2.7 g/dL. None of the 28 patients with normal albumin had Muehrcke’s lines.⁶

Kwashiorkor can cause fingernail clubbing
In a cross-sectional study of 60 children 1 to 4 years of age diagnosed with kwashiorkor without evidence of tuberculosis infection, fingernail clubbing was found in 46 (76.7%). Clubbing was mild in 26 (43.3%) children, moderate in 19 (37.7%), and severe in one (1.7%).⁷ Kwashiorkor is extremely rare in children in the United States and other developed countries.

Transplacental chemical exposures found to deform and discolor nails
A case-control study compared finger- and toenail findings from more than 100 Taiwanese children exposed transplacentally to high levels of PCBs and PCDFs with nail findings from a comparable number of controls. The investigators looked at parental reports and physical examination results.⁸

Although the rates of nail abnormalities reported by parents of exposed children differed slightly from rates documented by physical examination, children exposed to PCBs and PCDFs had consistently higher rates of nail deformity than controls. The researchers examined 117 cases and 106 controls. They identified dystrophic fingernails in 15% of exposed children and 1% of controls (odds ratio [OR]=15.4; 95% confidence interval [CI], 2-119); dystrophic toenails occurred in 32% of exposed children and 18% of controls (OR=2.2; 95% CI, 1.2-4.1). The most common fingernail deformities were grooves and ridges. The most common toenail deformities were koilonychias, ridges, thickening, and pigmentation changes.⁸

PCB exposure is an issue in the United States and other developed countries, but at much lower levels than the accidental contamination. Whether lower levels of exposure cause nail changes isn’t known.

Recommendations

No recommendations are available. A dermatology textbook lists several nail changes associated with nutritional deficiencies and toxic exposures (TABLE).⁹

Table
Nail changes linked to nutrition deficits and toxins⁹

Evidence summary

Two open-label RCTs compared oral and intramuscular (IM) therapy for vitamin B₁₂ deficiency.^1,2 Both studies enrolled patients from hospital-based clinics—not primary care centers. Most patients (63 of 93 total) had conditions associated with intestinal malabsorption, including 7 patients with pernicious anemia and 3 with ileal resection. Both trials excluded patients with celiac and inflammatory bowel disease.

Oral therapy works as well as injections and costs less
One RCT compared the effects of oral B₁₂ with IM therapy in 60 patients (mean age 62 years) with B₁₂ deficiency and megaloblastic anemia.¹ Investigators gave patients in each group equivalent doses of cobalamin: 1000 mcg daily for 10 days, weekly for 4 weeks, and then monthly to complete a 90-day course.

The mean hemoglobin increased significantly in both the oral and IM groups (from 8.4 to 13.8 g/dL, P<.001 for oral therapy; from 8.3 to 13.7 g/dL, P<.001 for IM therapy), as did mean serum B₁₂ levels (from 73 to 214 pg/mL, P<.001, oral; and from 70 to 226 pg/mL, P<.001, IM). Neurologic symptoms (sensitive peripheral neuropathy, alteration of cognitive function, loss of sense of vibration) either cleared or improved markedly in both groups within one month (7 of 9 patients with oral therapy and 9 of 12 patients with IM treatment; P value not given).

Oral therapy cost less ($80 vs $220 per patient) and neither group reported adverse effects.¹

B₁₂ therapy changes hematologic parameters
The second RCT compared oral with IM B₁₂ therapy in 33 patients (mean age 72 years) with newly diagnosed B₁₂ deficiency.² Investigators randomized patients to receive either oral cyanocobalamin (2000 mcg daily) for 120 days or IM cobalamin (1000 mcg) on Days 1, 3, 7, 10, 14, 21, 30, 60, and 90.

At 4 months, both groups had improved significantly from baseline in all metabolite measures and achieved a normal serum cobalamin level. The higher-dose oral therapy raised cobalamin levels more than IM therapy (from 93 to 1005 pg/mL, P<.0005 with oral therapy vs from 95 to 325 pg/mL, P<.0005 with IM treatment). Oral therapy increased cobalamin levels above 300 pg/mL in all patients; only half the patients treated with injections reached that level.

B₁₂ therapy also significantly changed hematologic parameters from baseline even though the patients in the study didn’t have anemia. Mean corpuscular volume, for example, decreased from 100 to 90 fL with oral therapy and 102 to 97 fL with IM therapy (P<.005 for each). Neurologic symptoms (memory loss, paresthesias, ataxia) either cleared or improved markedly in all patients. Investigators compared all parameters against baseline values but didn’t directly compare oral with IM therapy. The trial didn’t assess safety outcomes.²

Recommendations

Canada’s British Columbia Medical Association and Ministry of Health recommend oral replacement of B₁₂ (1000-2000 mcg/d) for most cases of vitamin B₁₂ deficiency, including pernicious anemia. For patients with neurologic symptoms, they recommend an initial B₁₂ injection (1000 mcg IM) followed by oral replacement.³

The US Centers for Disease Control and Prevention recommends either oral (1000 mcg daily) or parenteral B₁₂ replacement. They advise giving parenteral therapy either subcutaneously (to reduce the burning sensation) or IM (1000 mcg per week for 8 weeks, then monthly for life).⁴

Evidence summary

Two open-label RCTs compared oral and intramuscular (IM) therapy for vitamin B₁₂ deficiency.^1,2 Both studies enrolled patients from hospital-based clinics—not primary care centers. Most patients (63 of 93 total) had conditions associated with intestinal malabsorption, including 7 patients with pernicious anemia and 3 with ileal resection. Both trials excluded patients with celiac and inflammatory bowel disease.

Oral therapy works as well as injections and costs less
One RCT compared the effects of oral B₁₂ with IM therapy in 60 patients (mean age 62 years) with B₁₂ deficiency and megaloblastic anemia.¹ Investigators gave patients in each group equivalent doses of cobalamin: 1000 mcg daily for 10 days, weekly for 4 weeks, and then monthly to complete a 90-day course.

The mean hemoglobin increased significantly in both the oral and IM groups (from 8.4 to 13.8 g/dL, P<.001 for oral therapy; from 8.3 to 13.7 g/dL, P<.001 for IM therapy), as did mean serum B₁₂ levels (from 73 to 214 pg/mL, P<.001, oral; and from 70 to 226 pg/mL, P<.001, IM). Neurologic symptoms (sensitive peripheral neuropathy, alteration of cognitive function, loss of sense of vibration) either cleared or improved markedly in both groups within one month (7 of 9 patients with oral therapy and 9 of 12 patients with IM treatment; P value not given).

Oral therapy cost less ($80 vs $220 per patient) and neither group reported adverse effects.¹

B₁₂ therapy changes hematologic parameters
The second RCT compared oral with IM B₁₂ therapy in 33 patients (mean age 72 years) with newly diagnosed B₁₂ deficiency.² Investigators randomized patients to receive either oral cyanocobalamin (2000 mcg daily) for 120 days or IM cobalamin (1000 mcg) on Days 1, 3, 7, 10, 14, 21, 30, 60, and 90.

At 4 months, both groups had improved significantly from baseline in all metabolite measures and achieved a normal serum cobalamin level. The higher-dose oral therapy raised cobalamin levels more than IM therapy (from 93 to 1005 pg/mL, P<.0005 with oral therapy vs from 95 to 325 pg/mL, P<.0005 with IM treatment). Oral therapy increased cobalamin levels above 300 pg/mL in all patients; only half the patients treated with injections reached that level.

B₁₂ therapy also significantly changed hematologic parameters from baseline even though the patients in the study didn’t have anemia. Mean corpuscular volume, for example, decreased from 100 to 90 fL with oral therapy and 102 to 97 fL with IM therapy (P<.005 for each). Neurologic symptoms (memory loss, paresthesias, ataxia) either cleared or improved markedly in all patients. Investigators compared all parameters against baseline values but didn’t directly compare oral with IM therapy. The trial didn’t assess safety outcomes.²

Recommendations

Canada’s British Columbia Medical Association and Ministry of Health recommend oral replacement of B₁₂ (1000-2000 mcg/d) for most cases of vitamin B₁₂ deficiency, including pernicious anemia. For patients with neurologic symptoms, they recommend an initial B₁₂ injection (1000 mcg IM) followed by oral replacement.³

The US Centers for Disease Control and Prevention recommends either oral (1000 mcg daily) or parenteral B₁₂ replacement. They advise giving parenteral therapy either subcutaneously (to reduce the burning sensation) or IM (1000 mcg per week for 8 weeks, then monthly for life).⁴

Evidence summary

Two open-label RCTs compared oral and intramuscular (IM) therapy for vitamin B₁₂ deficiency.^1,2 Both studies enrolled patients from hospital-based clinics—not primary care centers. Most patients (63 of 93 total) had conditions associated with intestinal malabsorption, including 7 patients with pernicious anemia and 3 with ileal resection. Both trials excluded patients with celiac and inflammatory bowel disease.

Oral therapy works as well as injections and costs less
One RCT compared the effects of oral B₁₂ with IM therapy in 60 patients (mean age 62 years) with B₁₂ deficiency and megaloblastic anemia.¹ Investigators gave patients in each group equivalent doses of cobalamin: 1000 mcg daily for 10 days, weekly for 4 weeks, and then monthly to complete a 90-day course.

The mean hemoglobin increased significantly in both the oral and IM groups (from 8.4 to 13.8 g/dL, P<.001 for oral therapy; from 8.3 to 13.7 g/dL, P<.001 for IM therapy), as did mean serum B₁₂ levels (from 73 to 214 pg/mL, P<.001, oral; and from 70 to 226 pg/mL, P<.001, IM). Neurologic symptoms (sensitive peripheral neuropathy, alteration of cognitive function, loss of sense of vibration) either cleared or improved markedly in both groups within one month (7 of 9 patients with oral therapy and 9 of 12 patients with IM treatment; P value not given).

Oral therapy cost less ($80 vs $220 per patient) and neither group reported adverse effects.¹

B₁₂ therapy changes hematologic parameters
The second RCT compared oral with IM B₁₂ therapy in 33 patients (mean age 72 years) with newly diagnosed B₁₂ deficiency.² Investigators randomized patients to receive either oral cyanocobalamin (2000 mcg daily) for 120 days or IM cobalamin (1000 mcg) on Days 1, 3, 7, 10, 14, 21, 30, 60, and 90.

At 4 months, both groups had improved significantly from baseline in all metabolite measures and achieved a normal serum cobalamin level. The higher-dose oral therapy raised cobalamin levels more than IM therapy (from 93 to 1005 pg/mL, P<.0005 with oral therapy vs from 95 to 325 pg/mL, P<.0005 with IM treatment). Oral therapy increased cobalamin levels above 300 pg/mL in all patients; only half the patients treated with injections reached that level.

B₁₂ therapy also significantly changed hematologic parameters from baseline even though the patients in the study didn’t have anemia. Mean corpuscular volume, for example, decreased from 100 to 90 fL with oral therapy and 102 to 97 fL with IM therapy (P<.005 for each). Neurologic symptoms (memory loss, paresthesias, ataxia) either cleared or improved markedly in all patients. Investigators compared all parameters against baseline values but didn’t directly compare oral with IM therapy. The trial didn’t assess safety outcomes.²

Recommendations

Canada’s British Columbia Medical Association and Ministry of Health recommend oral replacement of B₁₂ (1000-2000 mcg/d) for most cases of vitamin B₁₂ deficiency, including pernicious anemia. For patients with neurologic symptoms, they recommend an initial B₁₂ injection (1000 mcg IM) followed by oral replacement.³

The US Centers for Disease Control and Prevention recommends either oral (1000 mcg daily) or parenteral B₁₂ replacement. They advise giving parenteral therapy either subcutaneously (to reduce the burning sensation) or IM (1000 mcg per week for 8 weeks, then monthly for life).⁴

Evidence summary

A systematic review of 55 studies that included 45 validating prospective studies (total 2579 patients, of whom 919 had inflammatory, liver, or neoplastic disease) found that a low serum ferritin level was the most sensitive and specific initial test to diagnose IDA, even in the presence of inflammation.¹ More than 80% of patients studied had confirmatory bone marrow aspiration.

Serum ferritin <45 ng/dL performed well for identifying IDA in all patients combined (sensitivity=85%; specificity=92%; positive likelihood ratio=11), although the authors didn’t calculate sensitivity and specificity for the subgroup of patients with an inflammatory disease process.

A serum ferritin level >100 ng/dL made IDA unlikely (sensitivity=5.5%; specificity=29%; negative likelihood ratio=0.08). Serum ferritin levels between 45 and 99 ng/dL weren’t useful in evaluating IDA. The subgroup analysis of the patients with inflammatory, liver, or neoplastic disease found that serum ferritin outperformed measurements of mean cell volume, transferrin saturation, red cell protoporphyrin, and red cell volume for diagnosing IDA.¹

Follow up with the sTfR-ferritin index if necessary
The sTfR-ferritin index is sensitive and specific for diagnosing IDA in patients with concomitant inflammation, including those with intermediate serum ferritin levels of 45 to 99 ng/dL. A systematic review of 9 prospective validating cohort studies (total 818 patients, of whom 678 had inflammatory disease) compared sTfR levels and sTfR-ferritin index (sTfR-to-log ferritin ratio) values against bone marrow biopsy for identifying IDA.²

A mean sTfR level ≥2.5 mg/L diagnosed IDA with a sensitivity of 68% to 97% and a specificity of 47% to 100%. In patients with acute and chronic inflammation, a mean sTfR-ferritin index ≥1.5 mg/L had a higher sensitivity (88%-100%) and specificity (93%-100%). The studies were limited by heterogenous populations, small sample sizes, and nonstandardized techniques and reference ranges for sTfR levels.

Recommendations

No major professional organizations in the United States have issued recommendations or clinical guidelines for diagnosing iron deficiency in patients with chronic inflammation.

The British Society of Gastroenterology suggests that a serum ferritin level <50 ng/dL is consistent with iron deficiency but lists the use of sTfR as promising, but unproven, in the clinical setting.³

Evidence summary

A systematic review of 55 studies that included 45 validating prospective studies (total 2579 patients, of whom 919 had inflammatory, liver, or neoplastic disease) found that a low serum ferritin level was the most sensitive and specific initial test to diagnose IDA, even in the presence of inflammation.¹ More than 80% of patients studied had confirmatory bone marrow aspiration.

Serum ferritin <45 ng/dL performed well for identifying IDA in all patients combined (sensitivity=85%; specificity=92%; positive likelihood ratio=11), although the authors didn’t calculate sensitivity and specificity for the subgroup of patients with an inflammatory disease process.

A serum ferritin level >100 ng/dL made IDA unlikely (sensitivity=5.5%; specificity=29%; negative likelihood ratio=0.08). Serum ferritin levels between 45 and 99 ng/dL weren’t useful in evaluating IDA. The subgroup analysis of the patients with inflammatory, liver, or neoplastic disease found that serum ferritin outperformed measurements of mean cell volume, transferrin saturation, red cell protoporphyrin, and red cell volume for diagnosing IDA.¹

Follow up with the sTfR-ferritin index if necessary
The sTfR-ferritin index is sensitive and specific for diagnosing IDA in patients with concomitant inflammation, including those with intermediate serum ferritin levels of 45 to 99 ng/dL. A systematic review of 9 prospective validating cohort studies (total 818 patients, of whom 678 had inflammatory disease) compared sTfR levels and sTfR-ferritin index (sTfR-to-log ferritin ratio) values against bone marrow biopsy for identifying IDA.²

A mean sTfR level ≥2.5 mg/L diagnosed IDA with a sensitivity of 68% to 97% and a specificity of 47% to 100%. In patients with acute and chronic inflammation, a mean sTfR-ferritin index ≥1.5 mg/L had a higher sensitivity (88%-100%) and specificity (93%-100%). The studies were limited by heterogenous populations, small sample sizes, and nonstandardized techniques and reference ranges for sTfR levels.

Recommendations

No major professional organizations in the United States have issued recommendations or clinical guidelines for diagnosing iron deficiency in patients with chronic inflammation.

The British Society of Gastroenterology suggests that a serum ferritin level <50 ng/dL is consistent with iron deficiency but lists the use of sTfR as promising, but unproven, in the clinical setting.³

Evidence summary

A systematic review of 55 studies that included 45 validating prospective studies (total 2579 patients, of whom 919 had inflammatory, liver, or neoplastic disease) found that a low serum ferritin level was the most sensitive and specific initial test to diagnose IDA, even in the presence of inflammation.¹ More than 80% of patients studied had confirmatory bone marrow aspiration.

Serum ferritin <45 ng/dL performed well for identifying IDA in all patients combined (sensitivity=85%; specificity=92%; positive likelihood ratio=11), although the authors didn’t calculate sensitivity and specificity for the subgroup of patients with an inflammatory disease process.

A serum ferritin level >100 ng/dL made IDA unlikely (sensitivity=5.5%; specificity=29%; negative likelihood ratio=0.08). Serum ferritin levels between 45 and 99 ng/dL weren’t useful in evaluating IDA. The subgroup analysis of the patients with inflammatory, liver, or neoplastic disease found that serum ferritin outperformed measurements of mean cell volume, transferrin saturation, red cell protoporphyrin, and red cell volume for diagnosing IDA.¹

Follow up with the sTfR-ferritin index if necessary
The sTfR-ferritin index is sensitive and specific for diagnosing IDA in patients with concomitant inflammation, including those with intermediate serum ferritin levels of 45 to 99 ng/dL. A systematic review of 9 prospective validating cohort studies (total 818 patients, of whom 678 had inflammatory disease) compared sTfR levels and sTfR-ferritin index (sTfR-to-log ferritin ratio) values against bone marrow biopsy for identifying IDA.²

A mean sTfR level ≥2.5 mg/L diagnosed IDA with a sensitivity of 68% to 97% and a specificity of 47% to 100%. In patients with acute and chronic inflammation, a mean sTfR-ferritin index ≥1.5 mg/L had a higher sensitivity (88%-100%) and specificity (93%-100%). The studies were limited by heterogenous populations, small sample sizes, and nonstandardized techniques and reference ranges for sTfR levels.

Recommendations

No major professional organizations in the United States have issued recommendations or clinical guidelines for diagnosing iron deficiency in patients with chronic inflammation.

The British Society of Gastroenterology suggests that a serum ferritin level <50 ng/dL is consistent with iron deficiency but lists the use of sTfR as promising, but unproven, in the clinical setting.³