Evidence-Based Reviews

Many catatonia cases respond to benzodiazepines—especially lorazepam—but up to 30% do not. Electroconvulsive therapy (ECT) can be effective, but what’s the next step when ECT is unavailable or inappropriate for your patient?

To help you solve this dilemma, we describe our diagnosis and treatment decisions for a patient we call Mr. C. We explain how our process was guided by recent understandings of an abnormal neural circuit that appears to cause catatonia’s complex motor and behavioral symptoms.

This article describes that neurologic pathology and answers common questions about the clinical workup and treatment of catatonia.

CASE: TROUBLE IN TV LAND

Mr. C, age 69, caused a disturbance at a local TV station, demanding that they broadcast a manuscript he had written. Police took him to a local hospital, where he was stabilized and then transferred to a neuropsychiatric hospital for evaluation.

The psychiatric interview revealed that he had developed insomnia, excessive activity, and delusional thinking 2 weeks before admission. His medical history included coronary artery disease (CAD), hypertension, and hypothyroidism. Medications included thyroid hormone replacement therapy, furosemide, potassium, ranitidine, simvastatin, metoprolol, and lisinopril. CAD treatment included stent placement and nitroglycerin as needed.

He had been hospitalized in his 30s and treated with ECT for what he called “bad thoughts.” He said he improved after 1 month and had no subsequent psychiatric history. He denied drug or alcohol abuse.

Shortly after admission, he refused to eat or drink and after 1 week became dehydrated. He also showed mutism, immobility, and stupor. He was transferred to the medical service for IV rehydration.

MANY SCENARIOS AND SIGNS

Mr. C’s symptoms suggest possible catatonia, a neuropsychiatric syndrome of motor dysregulation found in up to 10% of acutely ill psychiatric inpatients.^1,2 A movement disorder,^1,2 catatonia occurs with general medical conditions and psychiatric disorders (Table 1).

Pathophysiology. Catatonic signs develop when aberrant signals from neurochemical abnormalities trigger a neural circuit that affects the medial gyrus of the orbital frontal lobe, the lateral gyrus, caudate nucleus, globus pallidus, and thalamus (Box).^3-5

Presentation. A focused exam is required because patients with catatonia often do not provide a comprehensive or reliable history.² They show mutism, characteristic postures, rigidity, aberrant speech, negativism, and stereotyped behaviors.^1,2 They may present in an excited or retarded state:

• Excited patients may injure themselves or others and develop hyperthermia, tachycardia, and elevated blood pressure from excessive motor activity.
• Patients in a retarded state may present with bradykinesia and poor self-care. They may be unresponsive to external stimuli, develop catatonic stupor, and refuse to eat or drink.

Mr. C’s earlier insomnia, excessive activity, and delusional thinking (such as the TV station incident) may have signaled an excited catatonia. On admission to the medical service, however, he presented in a retarded state.

Signs. Part of the challenge with detecting catatonia’s signs is that there are so many; some rating scales list more than 20. Not all signs need to be present to make the diagnosis, however, and if you find one, others usually turn up in the examination.

A mnemonic from the Bush-Francis Catatonia Screening Instrument (Table 2) represents diagnostic signs in patients with the excited or retarded forms.² We recommend that you review an authoritative text (see Related resources) to understand catatonia’s psychopathology.²

Table 1

Common diagnoses of patients with catatonia

Psychiatric	Schizophrenia Mood disorder (depression, bipolar disorder [manic, mixed, depressed]) Other psychoses Personality or conversion (somatoform) disorders Developmental disorders (autism)
Organic	Due to a general medical condition Drug-induced Idiopathic

Box

 

Table 2

WIRED `N MIRED: Mnemonic for detecting catatonia

Waxy flexibility/catalepsy
Immobility/stupor
Refusal to eat or drink
Excitement
Deadpan staring
Negativism/negative symptoms
Mutism
Impulsivity
Rigidity
Echolalia/echopraxia
Direct observation

CASE CONTINUED: MAKING THE DIAGNOSIS

In the medical unit, Mr. C was found to be in a catatonic stupor, with immobility, mutism (monosyllabic speech), catalepsy, intermittent waxy flexibility, withdrawal (refusal to eat and drink), automatic obedience, and mitgehen (exaggerated movements in response to light finger pressure, despite instructions to stay still). ECT workup was started, along with a trial of lorazepam, 1 mg tid.

Laboratory studies revealed high BUN/creatinine (80/2.0) that returned to normal range (BUN 7 to 21 mg/dL; creatinine 0.5 to 1.2 mg/dL) after 3 days of hydration. Because of Mr. C’s earlier excited symptoms and delusional thinking, we considered a diagnosis of bipolar disorder with catatonia. However, his symptoms did not improve with a trial of valproic acid (serum level 64 mcg/mL).

Head CT showed generalized atrophy and EEG showed delta slowing. Single-photo emission computer tomography (SPECT) showed areas of decreased perfusion in the cortex, with no perfusion in the left posterior parietal area (PP).

Mental status exam found Mr. C disoriented with poor short-term memory and unable to complete the Mini Mental State Examination (MMSE). His Bush-Francis Catatonia Rating Scale score was 28 and included many catatonic signs that would not be seen a patient with simple dehydration.

The workup supported a diagnosis of catatonia due to general medical condition (vascular dementia) and ruled out schizophrenia with catatonic features, bipolar disorder, or major depression with catatonia.

EVALUATION AND DIAGNOSIS

Medical causes. A careful history and thorough physical examination are essential for making an accurate diagnosis and ruling out medical conditions that could present with or mimic catatonia (Table 3). Medications that can induce catatonia include antipsychotics, corticosteroids, and disulfiram at therapeutic doses. Drug abuse (such as with phencyclidine), use of the general anesthetic ketamine, and benzodiazepine withdrawal may also lead to catatonia.

Head CT or MRI is indicated for patients being considered for ECT or for localizing neurologic findings. EEG can be useful when patients present with features of seizure activity—such as tongue biting, incontinence, or stupor—or with catatonia as a manifestation of delirium or dementia.

A history of head injury or neurologic disease warrants further neurologic investigation. Also consider a neurology consult when the patient has prolonged stupor or does not respond to initial drug therapy.

Psychiatric causes. The clinical setting may suggest the most likely primary psychiatric disorders to consider, such as:

• bipolar or major depression in acute inpatient psychiatric units
• autism and pervasive developmental disorders (PDD) in pediatric or PDD units
• catatonic schizophrenia in chronic psychotic patients
• somatoform or factitious disorders in forensic settings.

These generalizations are not clinically exclusive, of course, but may provide a starting point for the treatment team confronted with limited history and exam information.

Table 3

Catatonia workup: Recommended lab tests

Test	Recommendation
Complete blood count with WBC differential	Look for leukocytosis
Serum chemistries	Look for electrolyte imbalances
Serum iron	May be low in NMS
Serum creatine kinase	If NMS is suspected
Brain MRI or CT	If structural lesion is suspected
Electroencephalography	If seizure disorder or brain abnormality is suspected
Lumbar puncture	If encephalitis or meningits is suspected
NMS: neuroleptic malignant syndrome

Initial treatment. Catatonia related to medical and psychiatric causes has been shown to respond to lorazepam and to ECT.^6,7 Lorazepam is preferred because of its specificity for the GABAa receptor and ease of administration (oral, IM, or IV). Other agents that act on GABA—including amobarbital and zolpidem—have also been used. Catatonia’s hallmark features such as mutism and immobility have been shown to respond to lorazepam.^8,9

ECT is a first-line treatment for catatonia with life-threatening conditions and should be considered for refractory cases.

Lorazepam. The starting dosage is usually 1 mg tid for healthy adults; 0.5 mg tid can be used for children and the elderly. Observe the patient for improvement in catatonic signs after the first dose and before giving the second. Dosages of up to 16 mg/d have been used.

In many cases, lorazepam can be tapered off after adequate treatment of the primary psychiatric condition. In severe cases, however—such as when patients refuse to eat or drink—lorazepam may be continued for as long as 1 year. Weigh the risk of benzodiazepine tolerance, dependence, and misuse versus the possibility of relapse and rehospitalization.

Medical catatonias and neuroleptic malignant syndrome (NMS) have responded favorably to ECT.⁸ Addressing the medical cause itself usually does not resolve catatonia, with the possible exception of seizure-induced (“ictal”) catatonia, which may respond to anticonvulsants and lorazepam.^6,7

ECT. An ECT workup can begin as soon as a patient presents with catatonia. If lorazepam produces no response within 24 hours, consider ECT.

 

CASE CONTINUED: PERSISTENT SYMPTOMS

After three 1-mg doses of lorazepam, Mr. C became more alert and oriented but his catatonia symptoms persisted, as indicated by a Bush-Francis score of 23, significant grasp reflex, and gegenhalten (automatic rather than willful resistance to passive limb movement in proportion to the strength of the stimulus). An attempt to gradually increase lorazepam to 2 mg tid produced delirium. He remained confused even when lorazepam was reduced to 0.5 mg tid, so the drug was discontinued.

Mr. C’s neurologist added amantadine, 100 mg tid, and carbidopa/levodopa, 10/100 mg tid, to treat his parkinsonian rigidity.

WHAT NEXT? OTHER OPTIONS

Antipsychotics have been investigated as a possible treatment for catatonia. The literature suggests that conventional antipsychotics may cause catatonia and atypical antipsychotics may improve it. Conventional antipsychotics are best avoided in catatonia because they:

• appear less effective than other treatments in resolving catatonic symptoms^8,10
• are associated with catatonic-like side effects, such as rigidity, akinesia, and staring¹⁰
• appear to increase NMS risk in patients with catatonic symptoms.^11,12

Atypicals appear more effective in treating catatonia and less likely to cause NMS. Case reports^13,14 indicate many of these agents can be effective and well tolerated in treating catatonic symptoms, although this was not the case for Mr. C.

Anticonvulsants such as valproate¹⁵ and carbamazepine, 600 to 1200 mg/d,¹⁶ may take longer to work than lorazepam but may be options for patients who do not respond to benzodiazepines.^8,9

Amantadine, an N-methyl-D-aspartate (NMDA) antagonist, has been used with some success for catatonia that does not unrespond to lorazepam.¹⁷ However, amantidine’s dopamine agonist activity could worsen underlying psychosis.

Memantine—another NMDA antagonist—differs from amantadine despite having a similar chemical structure. Memantine is a noncompetitive antagonist at the NMDA receptor, without affinity for dopamine, norepinephrine, serotonin, or muscarinic receptors.¹⁸

Although no published data support using memantine in patients with catatonia, it might be considered for those who are not candidates for lorazepam or ECT. For instance, a double-blind, placebo-controlled study found that lorazepam was not effective for catatonic schizophrenia.¹⁹ We have found memantine to help in some patients with catatonic schizophrenia.

CASE CONTINUED: TRIAL OF MEMANTINE

Mr. C remained in a catatonic stupor, but we decided against ECT because he resumed eating and drinking and was not medically at risk. Quetiapine, 100 to 300 mg/d, was tried to address his dementia symptoms, confusion, and poor mentation. This trial was discontinued after Mr. C fell and was readmitted to the medical unit. We then added memantine, 5 mg bid.

In the first week after beginning memantine, Mr. C’s MMSE score was 21, consistent with vascular dementia, but he remained immobile and staring. Motor signs also persisted, including automatic obedience, ambitendency, and a grasp reflex.

The next week, we increased memantine to 10 mg bid. Mr. C was oriented to person, place, and time, and his affect was blunted. His MMSE score increased to 25, showing improved cognition and memory. His Bush-Francis scale score was 6, showing reduced catatonic signs, with remaining mild immobility, bradykinesia, speech-prompt mutism, staring, and grasp reflex.

He maintained this improvement on carbidopa/levodopa, 10/100 tid; amantadine, 100 mg tid; and memantine, 10 mg bid, and was discharged from the nursing home unit.

IMPROVEMENT WITH MEMANTINE

Memantine may reduce excess glutamate at the NMDA receptor in the parietal-SMA-frontal cortical circuit. It may help to increase GABA and dopamine, which are deficient in catatonia. Our patient with vascular dementia had a severe ischemic deficit in the posterior parietal area, as seen on SPECT.

Amantadine, another NMDA receptor antagonist, acts on dopamine neurons and may have anticholinergic-like side effects, whereas memantine does not. Although both drugs share antagonism at the NMDA glutamate receptor, noncompetitive binding is weak for amantadine and moderate for memantine. Memantine has some serotonin (5-HT3) antagonism, but neither agent has direct GABA activity.

Memantine can improve function in vascular dementia.²⁰ Thus, Mr. C’s improvement may have been caused by the drug’s effect on his vascular dementia, the primary neuropsychiatric illness. However, his catatonic signs improved without antipsychotics, cholinesterase inhibitors, benzodiazepines, or ECT. No anticoagulation treatment or cerebral perfusion procedures account for his improved mental status.

CASE CONCLUSION

Mr. C went to live with his son’s family. Although he has problems with calculation, he shows good selfcare. When asked why he did not respond during his catatonic stupor, Mr. C stated that he believed the physician was an Internal Revenue Service agent asking him about serious tax problems. Upon reflection, he said he no longer believes this.

Related resources

• Fink M, Taylor MA. Catatonia: a clinician’s guide to diagnosis and treatment. Cambridge, UK: Cambridge University Press, 2003.
• Caroff SN, Mann SC, Francis A, Fricchione GE. Catatonia: from psychopathology to neurobiology. Washington, DC: American Psychiatric Publishing, 2004.
• Mann SC, Caroff SN, Keck PE Jr, Lazarus A. Neuroleptic malignant syndrome and related conditions (2nd ed). Washington, DC: American Psychiatric Publishing, 2003.
• Neuroleptic Malignant Syndrome Information Service. www.NMSIS.org.

 

Drug brand names

• Amantadine • Symmetrel
• Amobarbital • Amytal sodium
• Carbamazepine • Carbatrol, Equetro
• Carbidopa/levodopa • Sinemet
• Disulfiram • Antabuse
• Divalproex • Depakote
• Furosemide • Lasix
• Lisinopril • Prinivil, Zestril
• Lorazepam • Ativan
• Memantine • Namenda
• Metoprolol • Lopressor
• Ranitidine • Zantac
• Simvastatin • Zocor
• Valproic acid • Depakene
• Zolpidem • Ambien

Disclosure

Dr. Carroll and Dr. Hawkins are speakers for Forest Laboratories. The other authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Acknowledgment

The authors thank Dr. Niraj Ahuja, consultant psychiatrist and honorary clinical lecturer (psychiatry), Newcastle, North Tyneside and Northumberland Mental Health Trust, UK, for assistance with the figure.

Many catatonia cases respond to benzodiazepines—especially lorazepam—but up to 30% do not. Electroconvulsive therapy (ECT) can be effective, but what’s the next step when ECT is unavailable or inappropriate for your patient?

To help you solve this dilemma, we describe our diagnosis and treatment decisions for a patient we call Mr. C. We explain how our process was guided by recent understandings of an abnormal neural circuit that appears to cause catatonia’s complex motor and behavioral symptoms.

This article describes that neurologic pathology and answers common questions about the clinical workup and treatment of catatonia.

CASE: TROUBLE IN TV LAND

Mr. C, age 69, caused a disturbance at a local TV station, demanding that they broadcast a manuscript he had written. Police took him to a local hospital, where he was stabilized and then transferred to a neuropsychiatric hospital for evaluation.

The psychiatric interview revealed that he had developed insomnia, excessive activity, and delusional thinking 2 weeks before admission. His medical history included coronary artery disease (CAD), hypertension, and hypothyroidism. Medications included thyroid hormone replacement therapy, furosemide, potassium, ranitidine, simvastatin, metoprolol, and lisinopril. CAD treatment included stent placement and nitroglycerin as needed.

He had been hospitalized in his 30s and treated with ECT for what he called “bad thoughts.” He said he improved after 1 month and had no subsequent psychiatric history. He denied drug or alcohol abuse.

Shortly after admission, he refused to eat or drink and after 1 week became dehydrated. He also showed mutism, immobility, and stupor. He was transferred to the medical service for IV rehydration.

MANY SCENARIOS AND SIGNS

Mr. C’s symptoms suggest possible catatonia, a neuropsychiatric syndrome of motor dysregulation found in up to 10% of acutely ill psychiatric inpatients.^1,2 A movement disorder,^1,2 catatonia occurs with general medical conditions and psychiatric disorders (Table 1).

Pathophysiology. Catatonic signs develop when aberrant signals from neurochemical abnormalities trigger a neural circuit that affects the medial gyrus of the orbital frontal lobe, the lateral gyrus, caudate nucleus, globus pallidus, and thalamus (Box).^3-5

Presentation. A focused exam is required because patients with catatonia often do not provide a comprehensive or reliable history.² They show mutism, characteristic postures, rigidity, aberrant speech, negativism, and stereotyped behaviors.^1,2 They may present in an excited or retarded state:

• Excited patients may injure themselves or others and develop hyperthermia, tachycardia, and elevated blood pressure from excessive motor activity.
• Patients in a retarded state may present with bradykinesia and poor self-care. They may be unresponsive to external stimuli, develop catatonic stupor, and refuse to eat or drink.

Mr. C’s earlier insomnia, excessive activity, and delusional thinking (such as the TV station incident) may have signaled an excited catatonia. On admission to the medical service, however, he presented in a retarded state.

Signs. Part of the challenge with detecting catatonia’s signs is that there are so many; some rating scales list more than 20. Not all signs need to be present to make the diagnosis, however, and if you find one, others usually turn up in the examination.

A mnemonic from the Bush-Francis Catatonia Screening Instrument (Table 2) represents diagnostic signs in patients with the excited or retarded forms.² We recommend that you review an authoritative text (see Related resources) to understand catatonia’s psychopathology.²

Table 1

Common diagnoses of patients with catatonia

Psychiatric	Schizophrenia Mood disorder (depression, bipolar disorder [manic, mixed, depressed]) Other psychoses Personality or conversion (somatoform) disorders Developmental disorders (autism)
Organic	Due to a general medical condition Drug-induced Idiopathic

Box

 

Table 2

WIRED `N MIRED: Mnemonic for detecting catatonia

Waxy flexibility/catalepsy
Immobility/stupor
Refusal to eat or drink
Excitement
Deadpan staring
Negativism/negative symptoms
Mutism
Impulsivity
Rigidity
Echolalia/echopraxia
Direct observation

CASE CONTINUED: MAKING THE DIAGNOSIS

In the medical unit, Mr. C was found to be in a catatonic stupor, with immobility, mutism (monosyllabic speech), catalepsy, intermittent waxy flexibility, withdrawal (refusal to eat and drink), automatic obedience, and mitgehen (exaggerated movements in response to light finger pressure, despite instructions to stay still). ECT workup was started, along with a trial of lorazepam, 1 mg tid.

Laboratory studies revealed high BUN/creatinine (80/2.0) that returned to normal range (BUN 7 to 21 mg/dL; creatinine 0.5 to 1.2 mg/dL) after 3 days of hydration. Because of Mr. C’s earlier excited symptoms and delusional thinking, we considered a diagnosis of bipolar disorder with catatonia. However, his symptoms did not improve with a trial of valproic acid (serum level 64 mcg/mL).

Head CT showed generalized atrophy and EEG showed delta slowing. Single-photo emission computer tomography (SPECT) showed areas of decreased perfusion in the cortex, with no perfusion in the left posterior parietal area (PP).

Mental status exam found Mr. C disoriented with poor short-term memory and unable to complete the Mini Mental State Examination (MMSE). His Bush-Francis Catatonia Rating Scale score was 28 and included many catatonic signs that would not be seen a patient with simple dehydration.

The workup supported a diagnosis of catatonia due to general medical condition (vascular dementia) and ruled out schizophrenia with catatonic features, bipolar disorder, or major depression with catatonia.

EVALUATION AND DIAGNOSIS

Medical causes. A careful history and thorough physical examination are essential for making an accurate diagnosis and ruling out medical conditions that could present with or mimic catatonia (Table 3). Medications that can induce catatonia include antipsychotics, corticosteroids, and disulfiram at therapeutic doses. Drug abuse (such as with phencyclidine), use of the general anesthetic ketamine, and benzodiazepine withdrawal may also lead to catatonia.

Head CT or MRI is indicated for patients being considered for ECT or for localizing neurologic findings. EEG can be useful when patients present with features of seizure activity—such as tongue biting, incontinence, or stupor—or with catatonia as a manifestation of delirium or dementia.

A history of head injury or neurologic disease warrants further neurologic investigation. Also consider a neurology consult when the patient has prolonged stupor or does not respond to initial drug therapy.

Psychiatric causes. The clinical setting may suggest the most likely primary psychiatric disorders to consider, such as:

• bipolar or major depression in acute inpatient psychiatric units
• autism and pervasive developmental disorders (PDD) in pediatric or PDD units
• catatonic schizophrenia in chronic psychotic patients
• somatoform or factitious disorders in forensic settings.

These generalizations are not clinically exclusive, of course, but may provide a starting point for the treatment team confronted with limited history and exam information.

Table 3

Catatonia workup: Recommended lab tests

Test	Recommendation
Complete blood count with WBC differential	Look for leukocytosis
Serum chemistries	Look for electrolyte imbalances
Serum iron	May be low in NMS
Serum creatine kinase	If NMS is suspected
Brain MRI or CT	If structural lesion is suspected
Electroencephalography	If seizure disorder or brain abnormality is suspected
Lumbar puncture	If encephalitis or meningits is suspected
NMS: neuroleptic malignant syndrome

Initial treatment. Catatonia related to medical and psychiatric causes has been shown to respond to lorazepam and to ECT.^6,7 Lorazepam is preferred because of its specificity for the GABAa receptor and ease of administration (oral, IM, or IV). Other agents that act on GABA—including amobarbital and zolpidem—have also been used. Catatonia’s hallmark features such as mutism and immobility have been shown to respond to lorazepam.^8,9

ECT is a first-line treatment for catatonia with life-threatening conditions and should be considered for refractory cases.

Lorazepam. The starting dosage is usually 1 mg tid for healthy adults; 0.5 mg tid can be used for children and the elderly. Observe the patient for improvement in catatonic signs after the first dose and before giving the second. Dosages of up to 16 mg/d have been used.

In many cases, lorazepam can be tapered off after adequate treatment of the primary psychiatric condition. In severe cases, however—such as when patients refuse to eat or drink—lorazepam may be continued for as long as 1 year. Weigh the risk of benzodiazepine tolerance, dependence, and misuse versus the possibility of relapse and rehospitalization.

Medical catatonias and neuroleptic malignant syndrome (NMS) have responded favorably to ECT.⁸ Addressing the medical cause itself usually does not resolve catatonia, with the possible exception of seizure-induced (“ictal”) catatonia, which may respond to anticonvulsants and lorazepam.^6,7

ECT. An ECT workup can begin as soon as a patient presents with catatonia. If lorazepam produces no response within 24 hours, consider ECT.

 

CASE CONTINUED: PERSISTENT SYMPTOMS

After three 1-mg doses of lorazepam, Mr. C became more alert and oriented but his catatonia symptoms persisted, as indicated by a Bush-Francis score of 23, significant grasp reflex, and gegenhalten (automatic rather than willful resistance to passive limb movement in proportion to the strength of the stimulus). An attempt to gradually increase lorazepam to 2 mg tid produced delirium. He remained confused even when lorazepam was reduced to 0.5 mg tid, so the drug was discontinued.

Mr. C’s neurologist added amantadine, 100 mg tid, and carbidopa/levodopa, 10/100 mg tid, to treat his parkinsonian rigidity.

WHAT NEXT? OTHER OPTIONS

Antipsychotics have been investigated as a possible treatment for catatonia. The literature suggests that conventional antipsychotics may cause catatonia and atypical antipsychotics may improve it. Conventional antipsychotics are best avoided in catatonia because they:

• appear less effective than other treatments in resolving catatonic symptoms^8,10
• are associated with catatonic-like side effects, such as rigidity, akinesia, and staring¹⁰
• appear to increase NMS risk in patients with catatonic symptoms.^11,12

Atypicals appear more effective in treating catatonia and less likely to cause NMS. Case reports^13,14 indicate many of these agents can be effective and well tolerated in treating catatonic symptoms, although this was not the case for Mr. C.

Anticonvulsants such as valproate¹⁵ and carbamazepine, 600 to 1200 mg/d,¹⁶ may take longer to work than lorazepam but may be options for patients who do not respond to benzodiazepines.^8,9

Amantadine, an N-methyl-D-aspartate (NMDA) antagonist, has been used with some success for catatonia that does not unrespond to lorazepam.¹⁷ However, amantidine’s dopamine agonist activity could worsen underlying psychosis.

Memantine—another NMDA antagonist—differs from amantadine despite having a similar chemical structure. Memantine is a noncompetitive antagonist at the NMDA receptor, without affinity for dopamine, norepinephrine, serotonin, or muscarinic receptors.¹⁸

Although no published data support using memantine in patients with catatonia, it might be considered for those who are not candidates for lorazepam or ECT. For instance, a double-blind, placebo-controlled study found that lorazepam was not effective for catatonic schizophrenia.¹⁹ We have found memantine to help in some patients with catatonic schizophrenia.

CASE CONTINUED: TRIAL OF MEMANTINE

Mr. C remained in a catatonic stupor, but we decided against ECT because he resumed eating and drinking and was not medically at risk. Quetiapine, 100 to 300 mg/d, was tried to address his dementia symptoms, confusion, and poor mentation. This trial was discontinued after Mr. C fell and was readmitted to the medical unit. We then added memantine, 5 mg bid.

In the first week after beginning memantine, Mr. C’s MMSE score was 21, consistent with vascular dementia, but he remained immobile and staring. Motor signs also persisted, including automatic obedience, ambitendency, and a grasp reflex.

The next week, we increased memantine to 10 mg bid. Mr. C was oriented to person, place, and time, and his affect was blunted. His MMSE score increased to 25, showing improved cognition and memory. His Bush-Francis scale score was 6, showing reduced catatonic signs, with remaining mild immobility, bradykinesia, speech-prompt mutism, staring, and grasp reflex.

He maintained this improvement on carbidopa/levodopa, 10/100 tid; amantadine, 100 mg tid; and memantine, 10 mg bid, and was discharged from the nursing home unit.

IMPROVEMENT WITH MEMANTINE

Memantine may reduce excess glutamate at the NMDA receptor in the parietal-SMA-frontal cortical circuit. It may help to increase GABA and dopamine, which are deficient in catatonia. Our patient with vascular dementia had a severe ischemic deficit in the posterior parietal area, as seen on SPECT.

Amantadine, another NMDA receptor antagonist, acts on dopamine neurons and may have anticholinergic-like side effects, whereas memantine does not. Although both drugs share antagonism at the NMDA glutamate receptor, noncompetitive binding is weak for amantadine and moderate for memantine. Memantine has some serotonin (5-HT3) antagonism, but neither agent has direct GABA activity.

Memantine can improve function in vascular dementia.²⁰ Thus, Mr. C’s improvement may have been caused by the drug’s effect on his vascular dementia, the primary neuropsychiatric illness. However, his catatonic signs improved without antipsychotics, cholinesterase inhibitors, benzodiazepines, or ECT. No anticoagulation treatment or cerebral perfusion procedures account for his improved mental status.

CASE CONCLUSION

Mr. C went to live with his son’s family. Although he has problems with calculation, he shows good selfcare. When asked why he did not respond during his catatonic stupor, Mr. C stated that he believed the physician was an Internal Revenue Service agent asking him about serious tax problems. Upon reflection, he said he no longer believes this.

Related resources

• Fink M, Taylor MA. Catatonia: a clinician’s guide to diagnosis and treatment. Cambridge, UK: Cambridge University Press, 2003.
• Caroff SN, Mann SC, Francis A, Fricchione GE. Catatonia: from psychopathology to neurobiology. Washington, DC: American Psychiatric Publishing, 2004.
• Mann SC, Caroff SN, Keck PE Jr, Lazarus A. Neuroleptic malignant syndrome and related conditions (2nd ed). Washington, DC: American Psychiatric Publishing, 2003.
• Neuroleptic Malignant Syndrome Information Service. www.NMSIS.org.

 

Drug brand names

• Amantadine • Symmetrel
• Amobarbital • Amytal sodium
• Carbamazepine • Carbatrol, Equetro
• Carbidopa/levodopa • Sinemet
• Disulfiram • Antabuse
• Divalproex • Depakote
• Furosemide • Lasix
• Lisinopril • Prinivil, Zestril
• Lorazepam • Ativan
• Memantine • Namenda
• Metoprolol • Lopressor
• Ranitidine • Zantac
• Simvastatin • Zocor
• Valproic acid • Depakene
• Zolpidem • Ambien

Disclosure

Dr. Carroll and Dr. Hawkins are speakers for Forest Laboratories. The other authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Acknowledgment

The authors thank Dr. Niraj Ahuja, consultant psychiatrist and honorary clinical lecturer (psychiatry), Newcastle, North Tyneside and Northumberland Mental Health Trust, UK, for assistance with the figure.

Many catatonia cases respond to benzodiazepines—especially lorazepam—but up to 30% do not. Electroconvulsive therapy (ECT) can be effective, but what’s the next step when ECT is unavailable or inappropriate for your patient?

To help you solve this dilemma, we describe our diagnosis and treatment decisions for a patient we call Mr. C. We explain how our process was guided by recent understandings of an abnormal neural circuit that appears to cause catatonia’s complex motor and behavioral symptoms.

This article describes that neurologic pathology and answers common questions about the clinical workup and treatment of catatonia.

CASE: TROUBLE IN TV LAND

Mr. C, age 69, caused a disturbance at a local TV station, demanding that they broadcast a manuscript he had written. Police took him to a local hospital, where he was stabilized and then transferred to a neuropsychiatric hospital for evaluation.

The psychiatric interview revealed that he had developed insomnia, excessive activity, and delusional thinking 2 weeks before admission. His medical history included coronary artery disease (CAD), hypertension, and hypothyroidism. Medications included thyroid hormone replacement therapy, furosemide, potassium, ranitidine, simvastatin, metoprolol, and lisinopril. CAD treatment included stent placement and nitroglycerin as needed.

He had been hospitalized in his 30s and treated with ECT for what he called “bad thoughts.” He said he improved after 1 month and had no subsequent psychiatric history. He denied drug or alcohol abuse.

Shortly after admission, he refused to eat or drink and after 1 week became dehydrated. He also showed mutism, immobility, and stupor. He was transferred to the medical service for IV rehydration.

MANY SCENARIOS AND SIGNS

Mr. C’s symptoms suggest possible catatonia, a neuropsychiatric syndrome of motor dysregulation found in up to 10% of acutely ill psychiatric inpatients.^1,2 A movement disorder,^1,2 catatonia occurs with general medical conditions and psychiatric disorders (Table 1).

Pathophysiology. Catatonic signs develop when aberrant signals from neurochemical abnormalities trigger a neural circuit that affects the medial gyrus of the orbital frontal lobe, the lateral gyrus, caudate nucleus, globus pallidus, and thalamus (Box).^3-5

Presentation. A focused exam is required because patients with catatonia often do not provide a comprehensive or reliable history.² They show mutism, characteristic postures, rigidity, aberrant speech, negativism, and stereotyped behaviors.^1,2 They may present in an excited or retarded state:

• Excited patients may injure themselves or others and develop hyperthermia, tachycardia, and elevated blood pressure from excessive motor activity.
• Patients in a retarded state may present with bradykinesia and poor self-care. They may be unresponsive to external stimuli, develop catatonic stupor, and refuse to eat or drink.

Mr. C’s earlier insomnia, excessive activity, and delusional thinking (such as the TV station incident) may have signaled an excited catatonia. On admission to the medical service, however, he presented in a retarded state.

Signs. Part of the challenge with detecting catatonia’s signs is that there are so many; some rating scales list more than 20. Not all signs need to be present to make the diagnosis, however, and if you find one, others usually turn up in the examination.

A mnemonic from the Bush-Francis Catatonia Screening Instrument (Table 2) represents diagnostic signs in patients with the excited or retarded forms.² We recommend that you review an authoritative text (see Related resources) to understand catatonia’s psychopathology.²

Table 1

Common diagnoses of patients with catatonia

Psychiatric	Schizophrenia Mood disorder (depression, bipolar disorder [manic, mixed, depressed]) Other psychoses Personality or conversion (somatoform) disorders Developmental disorders (autism)
Organic	Due to a general medical condition Drug-induced Idiopathic

Box

 

Table 2

WIRED `N MIRED: Mnemonic for detecting catatonia

Waxy flexibility/catalepsy
Immobility/stupor
Refusal to eat or drink
Excitement
Deadpan staring
Negativism/negative symptoms
Mutism
Impulsivity
Rigidity
Echolalia/echopraxia
Direct observation

CASE CONTINUED: MAKING THE DIAGNOSIS

In the medical unit, Mr. C was found to be in a catatonic stupor, with immobility, mutism (monosyllabic speech), catalepsy, intermittent waxy flexibility, withdrawal (refusal to eat and drink), automatic obedience, and mitgehen (exaggerated movements in response to light finger pressure, despite instructions to stay still). ECT workup was started, along with a trial of lorazepam, 1 mg tid.

Laboratory studies revealed high BUN/creatinine (80/2.0) that returned to normal range (BUN 7 to 21 mg/dL; creatinine 0.5 to 1.2 mg/dL) after 3 days of hydration. Because of Mr. C’s earlier excited symptoms and delusional thinking, we considered a diagnosis of bipolar disorder with catatonia. However, his symptoms did not improve with a trial of valproic acid (serum level 64 mcg/mL).

Head CT showed generalized atrophy and EEG showed delta slowing. Single-photo emission computer tomography (SPECT) showed areas of decreased perfusion in the cortex, with no perfusion in the left posterior parietal area (PP).

Mental status exam found Mr. C disoriented with poor short-term memory and unable to complete the Mini Mental State Examination (MMSE). His Bush-Francis Catatonia Rating Scale score was 28 and included many catatonic signs that would not be seen a patient with simple dehydration.

The workup supported a diagnosis of catatonia due to general medical condition (vascular dementia) and ruled out schizophrenia with catatonic features, bipolar disorder, or major depression with catatonia.

EVALUATION AND DIAGNOSIS

Medical causes. A careful history and thorough physical examination are essential for making an accurate diagnosis and ruling out medical conditions that could present with or mimic catatonia (Table 3). Medications that can induce catatonia include antipsychotics, corticosteroids, and disulfiram at therapeutic doses. Drug abuse (such as with phencyclidine), use of the general anesthetic ketamine, and benzodiazepine withdrawal may also lead to catatonia.

Head CT or MRI is indicated for patients being considered for ECT or for localizing neurologic findings. EEG can be useful when patients present with features of seizure activity—such as tongue biting, incontinence, or stupor—or with catatonia as a manifestation of delirium or dementia.

A history of head injury or neurologic disease warrants further neurologic investigation. Also consider a neurology consult when the patient has prolonged stupor or does not respond to initial drug therapy.

Psychiatric causes. The clinical setting may suggest the most likely primary psychiatric disorders to consider, such as:

• bipolar or major depression in acute inpatient psychiatric units
• autism and pervasive developmental disorders (PDD) in pediatric or PDD units
• catatonic schizophrenia in chronic psychotic patients
• somatoform or factitious disorders in forensic settings.

These generalizations are not clinically exclusive, of course, but may provide a starting point for the treatment team confronted with limited history and exam information.

Table 3

Catatonia workup: Recommended lab tests

Test	Recommendation
Complete blood count with WBC differential	Look for leukocytosis
Serum chemistries	Look for electrolyte imbalances
Serum iron	May be low in NMS
Serum creatine kinase	If NMS is suspected
Brain MRI or CT	If structural lesion is suspected
Electroencephalography	If seizure disorder or brain abnormality is suspected
Lumbar puncture	If encephalitis or meningits is suspected
NMS: neuroleptic malignant syndrome

Initial treatment. Catatonia related to medical and psychiatric causes has been shown to respond to lorazepam and to ECT.^6,7 Lorazepam is preferred because of its specificity for the GABAa receptor and ease of administration (oral, IM, or IV). Other agents that act on GABA—including amobarbital and zolpidem—have also been used. Catatonia’s hallmark features such as mutism and immobility have been shown to respond to lorazepam.^8,9

ECT is a first-line treatment for catatonia with life-threatening conditions and should be considered for refractory cases.

Lorazepam. The starting dosage is usually 1 mg tid for healthy adults; 0.5 mg tid can be used for children and the elderly. Observe the patient for improvement in catatonic signs after the first dose and before giving the second. Dosages of up to 16 mg/d have been used.

In many cases, lorazepam can be tapered off after adequate treatment of the primary psychiatric condition. In severe cases, however—such as when patients refuse to eat or drink—lorazepam may be continued for as long as 1 year. Weigh the risk of benzodiazepine tolerance, dependence, and misuse versus the possibility of relapse and rehospitalization.

Medical catatonias and neuroleptic malignant syndrome (NMS) have responded favorably to ECT.⁸ Addressing the medical cause itself usually does not resolve catatonia, with the possible exception of seizure-induced (“ictal”) catatonia, which may respond to anticonvulsants and lorazepam.^6,7

ECT. An ECT workup can begin as soon as a patient presents with catatonia. If lorazepam produces no response within 24 hours, consider ECT.

 

CASE CONTINUED: PERSISTENT SYMPTOMS

After three 1-mg doses of lorazepam, Mr. C became more alert and oriented but his catatonia symptoms persisted, as indicated by a Bush-Francis score of 23, significant grasp reflex, and gegenhalten (automatic rather than willful resistance to passive limb movement in proportion to the strength of the stimulus). An attempt to gradually increase lorazepam to 2 mg tid produced delirium. He remained confused even when lorazepam was reduced to 0.5 mg tid, so the drug was discontinued.

Mr. C’s neurologist added amantadine, 100 mg tid, and carbidopa/levodopa, 10/100 mg tid, to treat his parkinsonian rigidity.

WHAT NEXT? OTHER OPTIONS

Antipsychotics have been investigated as a possible treatment for catatonia. The literature suggests that conventional antipsychotics may cause catatonia and atypical antipsychotics may improve it. Conventional antipsychotics are best avoided in catatonia because they:

• appear less effective than other treatments in resolving catatonic symptoms^8,10
• are associated with catatonic-like side effects, such as rigidity, akinesia, and staring¹⁰
• appear to increase NMS risk in patients with catatonic symptoms.^11,12

Atypicals appear more effective in treating catatonia and less likely to cause NMS. Case reports^13,14 indicate many of these agents can be effective and well tolerated in treating catatonic symptoms, although this was not the case for Mr. C.

Anticonvulsants such as valproate¹⁵ and carbamazepine, 600 to 1200 mg/d,¹⁶ may take longer to work than lorazepam but may be options for patients who do not respond to benzodiazepines.^8,9

Amantadine, an N-methyl-D-aspartate (NMDA) antagonist, has been used with some success for catatonia that does not unrespond to lorazepam.¹⁷ However, amantidine’s dopamine agonist activity could worsen underlying psychosis.

Memantine—another NMDA antagonist—differs from amantadine despite having a similar chemical structure. Memantine is a noncompetitive antagonist at the NMDA receptor, without affinity for dopamine, norepinephrine, serotonin, or muscarinic receptors.¹⁸

Although no published data support using memantine in patients with catatonia, it might be considered for those who are not candidates for lorazepam or ECT. For instance, a double-blind, placebo-controlled study found that lorazepam was not effective for catatonic schizophrenia.¹⁹ We have found memantine to help in some patients with catatonic schizophrenia.

CASE CONTINUED: TRIAL OF MEMANTINE

Mr. C remained in a catatonic stupor, but we decided against ECT because he resumed eating and drinking and was not medically at risk. Quetiapine, 100 to 300 mg/d, was tried to address his dementia symptoms, confusion, and poor mentation. This trial was discontinued after Mr. C fell and was readmitted to the medical unit. We then added memantine, 5 mg bid.

In the first week after beginning memantine, Mr. C’s MMSE score was 21, consistent with vascular dementia, but he remained immobile and staring. Motor signs also persisted, including automatic obedience, ambitendency, and a grasp reflex.

The next week, we increased memantine to 10 mg bid. Mr. C was oriented to person, place, and time, and his affect was blunted. His MMSE score increased to 25, showing improved cognition and memory. His Bush-Francis scale score was 6, showing reduced catatonic signs, with remaining mild immobility, bradykinesia, speech-prompt mutism, staring, and grasp reflex.

He maintained this improvement on carbidopa/levodopa, 10/100 tid; amantadine, 100 mg tid; and memantine, 10 mg bid, and was discharged from the nursing home unit.

IMPROVEMENT WITH MEMANTINE

Memantine may reduce excess glutamate at the NMDA receptor in the parietal-SMA-frontal cortical circuit. It may help to increase GABA and dopamine, which are deficient in catatonia. Our patient with vascular dementia had a severe ischemic deficit in the posterior parietal area, as seen on SPECT.

Amantadine, another NMDA receptor antagonist, acts on dopamine neurons and may have anticholinergic-like side effects, whereas memantine does not. Although both drugs share antagonism at the NMDA glutamate receptor, noncompetitive binding is weak for amantadine and moderate for memantine. Memantine has some serotonin (5-HT3) antagonism, but neither agent has direct GABA activity.

Memantine can improve function in vascular dementia.²⁰ Thus, Mr. C’s improvement may have been caused by the drug’s effect on his vascular dementia, the primary neuropsychiatric illness. However, his catatonic signs improved without antipsychotics, cholinesterase inhibitors, benzodiazepines, or ECT. No anticoagulation treatment or cerebral perfusion procedures account for his improved mental status.

CASE CONCLUSION

Mr. C went to live with his son’s family. Although he has problems with calculation, he shows good selfcare. When asked why he did not respond during his catatonic stupor, Mr. C stated that he believed the physician was an Internal Revenue Service agent asking him about serious tax problems. Upon reflection, he said he no longer believes this.

Related resources

• Fink M, Taylor MA. Catatonia: a clinician’s guide to diagnosis and treatment. Cambridge, UK: Cambridge University Press, 2003.
• Caroff SN, Mann SC, Francis A, Fricchione GE. Catatonia: from psychopathology to neurobiology. Washington, DC: American Psychiatric Publishing, 2004.
• Mann SC, Caroff SN, Keck PE Jr, Lazarus A. Neuroleptic malignant syndrome and related conditions (2nd ed). Washington, DC: American Psychiatric Publishing, 2003.
• Neuroleptic Malignant Syndrome Information Service. www.NMSIS.org.

 

Drug brand names

• Amantadine • Symmetrel
• Amobarbital • Amytal sodium
• Carbamazepine • Carbatrol, Equetro
• Carbidopa/levodopa • Sinemet
• Disulfiram • Antabuse
• Divalproex • Depakote
• Furosemide • Lasix
• Lisinopril • Prinivil, Zestril
• Lorazepam • Ativan
• Memantine • Namenda
• Metoprolol • Lopressor
• Ranitidine • Zantac
• Simvastatin • Zocor
• Valproic acid • Depakene
• Zolpidem • Ambien

Disclosure

Dr. Carroll and Dr. Hawkins are speakers for Forest Laboratories. The other authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Acknowledgment

The authors thank Dr. Niraj Ahuja, consultant psychiatrist and honorary clinical lecturer (psychiatry), Newcastle, North Tyneside and Northumberland Mental Health Trust, UK, for assistance with the figure.

Why do some people stay married for decades while others jump from one relationship to the next? If humans are like rodents, recent research suggests that neuropeptides may help forge the ties that bind.

FOLLOWING VOLE BEHAVIOR

The vole, which inhabits many grasslands in the United States, is a rodent resembling a mouse but related to the lemming. Males in some vole species (such as meadow voles) show no partner preference whereas males in other species prefer one partner, share the same nest with her, and help care for their offspring.¹

The neuropeptides oxytocin and arginine vasopressin are mediators of pair bonding. The brain of a prairie vole—a species in which males tend to bond with female partners—has more vasopressin receptors than that of the solitary meadow vole (Figure 1).

Figure 1 More vasopressin in bonding voles

Ventral forebrain autoradiograms show greater vasopressin (VP) expression in prairie voles (top) than meadow voles.

Reprinted with permission. © 2004, Nature Publishing Group.Lim et al, however, found that meadow voles were more likely to bond with a single female after the males’ vasopressin receptors were increased.² The researchers isolated and replicated the gene sequence responsible for the vasopressin receptor in the prairie vole. Then, using a viral vector, they injected the gene into the ventral pallidum of 11 male meadow voles. Eleven other voles received placebo.

Two weeks later, each sexually naïve male vole was housed for 24 hours with a sexually receptive female vole. Then, each male was placed for 3 hours in a three-chamber apparatus with the partner female in one chamber and a novel female in another. The time spent huddling with each female was recorded.

Across 3 hours, the placebo group voles spent 10 to 15 minutes with either female—normal behavior for this species. By contrast, the voles that received the vasopressin receptors spent approximately 40 minutes with their partners but only about 5 minutes with the novel voles, thus showing more affiliative behavior and a clear preference for their mates (Figure 2). The findings suggest that the researchers may have produced a profound change in social behavior by altering one gene.

IMPLICATIONS FOR HUMANS

How this research relates to humans is unknown, as there is no sound evidence of a link between vole and human pair bonding. Likewise, the influence of the higher cortical areas in orchestrating human behaviors cannot be underestimated.

Neuroimaging, however, has shown that brain regions rich with oxytocin and vasopressin receptors are activated while a person views pictures of loved ones.³ Additionally, mens’ vasopressin levels have been shown to increase when they are sexually aroused.⁴ Whether these findings one day lead to a medicine that promotes monogamous behavior in men remains to be seen.⁵

Figure 2 Male meadow voles’ interactions with females after vasopressin or placebo treatment

Source: Adapted from reference 2

Why do some people stay married for decades while others jump from one relationship to the next? If humans are like rodents, recent research suggests that neuropeptides may help forge the ties that bind.

FOLLOWING VOLE BEHAVIOR

The vole, which inhabits many grasslands in the United States, is a rodent resembling a mouse but related to the lemming. Males in some vole species (such as meadow voles) show no partner preference whereas males in other species prefer one partner, share the same nest with her, and help care for their offspring.¹

The neuropeptides oxytocin and arginine vasopressin are mediators of pair bonding. The brain of a prairie vole—a species in which males tend to bond with female partners—has more vasopressin receptors than that of the solitary meadow vole (Figure 1).

Figure 1 More vasopressin in bonding voles

Ventral forebrain autoradiograms show greater vasopressin (VP) expression in prairie voles (top) than meadow voles.

Reprinted with permission. © 2004, Nature Publishing Group.Lim et al, however, found that meadow voles were more likely to bond with a single female after the males’ vasopressin receptors were increased.² The researchers isolated and replicated the gene sequence responsible for the vasopressin receptor in the prairie vole. Then, using a viral vector, they injected the gene into the ventral pallidum of 11 male meadow voles. Eleven other voles received placebo.

Two weeks later, each sexually naïve male vole was housed for 24 hours with a sexually receptive female vole. Then, each male was placed for 3 hours in a three-chamber apparatus with the partner female in one chamber and a novel female in another. The time spent huddling with each female was recorded.

Across 3 hours, the placebo group voles spent 10 to 15 minutes with either female—normal behavior for this species. By contrast, the voles that received the vasopressin receptors spent approximately 40 minutes with their partners but only about 5 minutes with the novel voles, thus showing more affiliative behavior and a clear preference for their mates (Figure 2). The findings suggest that the researchers may have produced a profound change in social behavior by altering one gene.

IMPLICATIONS FOR HUMANS

How this research relates to humans is unknown, as there is no sound evidence of a link between vole and human pair bonding. Likewise, the influence of the higher cortical areas in orchestrating human behaviors cannot be underestimated.

Neuroimaging, however, has shown that brain regions rich with oxytocin and vasopressin receptors are activated while a person views pictures of loved ones.³ Additionally, mens’ vasopressin levels have been shown to increase when they are sexually aroused.⁴ Whether these findings one day lead to a medicine that promotes monogamous behavior in men remains to be seen.⁵

Figure 2 Male meadow voles’ interactions with females after vasopressin or placebo treatment

Source: Adapted from reference 2

Why do some people stay married for decades while others jump from one relationship to the next? If humans are like rodents, recent research suggests that neuropeptides may help forge the ties that bind.

FOLLOWING VOLE BEHAVIOR

The vole, which inhabits many grasslands in the United States, is a rodent resembling a mouse but related to the lemming. Males in some vole species (such as meadow voles) show no partner preference whereas males in other species prefer one partner, share the same nest with her, and help care for their offspring.¹

The neuropeptides oxytocin and arginine vasopressin are mediators of pair bonding. The brain of a prairie vole—a species in which males tend to bond with female partners—has more vasopressin receptors than that of the solitary meadow vole (Figure 1).

Figure 1 More vasopressin in bonding voles

Ventral forebrain autoradiograms show greater vasopressin (VP) expression in prairie voles (top) than meadow voles.

Reprinted with permission. © 2004, Nature Publishing Group.Lim et al, however, found that meadow voles were more likely to bond with a single female after the males’ vasopressin receptors were increased.² The researchers isolated and replicated the gene sequence responsible for the vasopressin receptor in the prairie vole. Then, using a viral vector, they injected the gene into the ventral pallidum of 11 male meadow voles. Eleven other voles received placebo.

Two weeks later, each sexually naïve male vole was housed for 24 hours with a sexually receptive female vole. Then, each male was placed for 3 hours in a three-chamber apparatus with the partner female in one chamber and a novel female in another. The time spent huddling with each female was recorded.

Across 3 hours, the placebo group voles spent 10 to 15 minutes with either female—normal behavior for this species. By contrast, the voles that received the vasopressin receptors spent approximately 40 minutes with their partners but only about 5 minutes with the novel voles, thus showing more affiliative behavior and a clear preference for their mates (Figure 2). The findings suggest that the researchers may have produced a profound change in social behavior by altering one gene.

IMPLICATIONS FOR HUMANS

How this research relates to humans is unknown, as there is no sound evidence of a link between vole and human pair bonding. Likewise, the influence of the higher cortical areas in orchestrating human behaviors cannot be underestimated.

Neuroimaging, however, has shown that brain regions rich with oxytocin and vasopressin receptors are activated while a person views pictures of loved ones.³ Additionally, mens’ vasopressin levels have been shown to increase when they are sexually aroused.⁴ Whether these findings one day lead to a medicine that promotes monogamous behavior in men remains to be seen.⁵

Figure 2 Male meadow voles’ interactions with females after vasopressin or placebo treatment

Source: Adapted from reference 2

Because 40% of individuals with a psychotic prodrome develop schizophrenia, detecting and preventing this transition could improve many patients’ lives. Unfortunately:

• psychotic prodrome lacks clear-cut symptoms and is difficult to identify
  
• little evidence exists to help clinicians select psychotropics and decide how long to use them
  
• treating all prodromal patients would expose those who never develop psychosis to the risk of psychotropics’ side effects.
  

How, then, can psychiatrists help patients who present with possible prodromal symptoms? Based on research and our experience, this article describes the psychotic prodrome and offers a pragmatic, evidence-based approach to diagnosis and treatment.

WHAT CAUSES PSYCHOTIC CONVERSION?

Reduced gray matter volumes in certain brain regions may be associated with conversion to psychosis (Box 1). Stress also may play a role; elevated stress-reactive cortisol levels are associated with positive symptom severity in the prodrome.¹ Other factors being investigated include obstetric complications at birth, maternal age >30, premorbid schizotypal personality disorder, and impaired olfaction.

Symptoms. Nearly 80% of patients with schizophrenia experience a psychotic prodrome that lasts a few months to several years.² Common features include:

• gradual worsening of perceptual disturbance
  
• referential thinking
  
• paranoia
  
• mild cognitive deficits
  
• mood lability
  
• impulsivity
  
• suicidality
  
• declining social function and academic performance.^3,4
  

Patients with these symptoms may be at imminent risk; if untreated, an estimated 40% progress to schizophrenia within 1 year.⁵

A premorbid phase often precedes the prodrome, with symptoms such as impaired attention, soft neurologic signs, and subtle social deficits. These changes may be harbingers of the prodrome but are too nonspecific to be diagnostic. Other functional impairments—including anxiety, depression, drug abuse, and psychosocial factors such as school stress—may mimic schizophrenic prodrome.

Prognosis. Studies of patients’ first schizophrenia episodes suggest that prodrome duration may predict outcome. A longer prodrome is thought to indicate a poor prognosis,⁶ such as in patients who wait a year before seeking treatment.⁷ A review of 22 studies of first-episode psychosis found early psychosocial and pharmacologic interventions improved long-term prognosis, and medication discontinuation predicted more-severe and chronic disease.⁸

Box 1

Genetic risk. Schizophrenia has a strong genetic predisposition, although not everyone in the genetic high-risk group develops schizophrenia. Persons with a family history of schizophrenia have a 10% to 20% risk of psychotic conversion.⁹

Pioneering work by McGorry et al¹⁰ identified an “ultra high-risk group” with a psychotic conversion rate of 40% to 60%. These patients present with three symptom patterns:

• attenuated positive symptoms
  
• brief intermittent psychotic episodes
  
• genetic risk and recent deterioration syndrome (Table 1).
  

Table 1

3 patient groups considered at ‘ultra high risk’ to develop schizophrenia

Patients with…	Symptoms
Attenuated psychotic symptoms	Overvalued ideas, perceptual disorders
	Present at least 1 week; not >5 years
	At least 1 symptom several times a week
Brief intermittent psychotic episodes	Frank psychotic features
	Resolve spontaneously within 7 days
	Can be drug-induced
Genetic risk and recent deterioration syndrome	Psychotic disorder in a first-degree relative
	Schizotypal personality disorder
	Present at least 1 month; not >5 years
	Significant functional decline
Source: Adapted from reference 10

Early identification of these high-risk individuals with perceptual distortions, frank psychotic symptoms, family history of psychosis, and schizotypal personality disorder may aid in early recognition and treatment.

The Edinburgh High Risk Study of 162 individuals ages 16 to 25 showed more marked psychopathology in those with at least two close relatives with schizophrenia, compared with control groups. A direct correlation was seen between genetic liability and poor neurocognitive performance.¹¹

PRODROME RATING SCALES

Researchers are using outcome measures to diagnose prodromal symptoms and assess their severity. Operational, validated assessment tools include:

• Bonn Scale for the Assessment of Basic Symptoms (BSABS): captures subtle changes in thinking, feeling, and perception.
  
• Schizophrenia Prediction Instrument for Adults (SPI-A): defines prepsychotic deviations and rates symptoms that are subjectively experienced by the patient.
  
• Comprehensive Assessment of At Risk Mental State (CAARMS): defines ultra high-risk criteria and incorporates eight dimensions of psychopathology.
  
• Scale of Prodromal Symptoms (SOPS): rates psychosis severity. When embedded within the Structured Interview for Prodromal Syndromes (SIPS), the SOPS determines the presence or absence of psychosis and predicts progression to psychopathology.
  
• Criteria for Prodromal Symptoms (COPS): defines ultra high-risk categories.
  
• Presence of Psychosis Scale (POPS): rates severity, intensity, and duration of positive prodromal symptoms.¹²
  

 

These instruments may identify prodromal symptoms in psychiatric practice, but further validation of clinical criteria is needed before they could be recommended for routine patient assessment.

PROPHYLACTIC ANTIPSYCHOTICS?

Atypical antipsychotics may be the standard of care for patients with a first psychotic episode, but this intervention is based on few double-blind controlled trials. Not surprisingly, only a handful of studies have examined antipsychotic therapy for the prodrome’s less clear-cut symptoms.

Risperidone. An 8- to 12-week open-label study in adolescents with first- and second-degree relatives with schizophrenia¹³ included four prodromal and six first-episode psychosis patients who met criteria for a cluster A personality disorder. Risperidone, 1.0 mg/d and 1.8 mg/d, respectively, improved thought disorder and attention symptoms, as measured with the Child Behavior Checklist. Verbal memory improved minimally, and no medication side effects were reported.

An open-label observational study¹⁴ identified four middle-aged subjects with a genetic risk of schizophrenia who reported negative symptoms and neurocognitive deficits. Risperidone, started at 0.25 mg/d and gradually increased to a maximum of 2 mg/d, improved negative symptoms, attention, and working memory. Mild side effects including tremors, sedation, dry mouth, and anxiety symptoms were reported.

An open-label, randomized, comparator trial¹⁵ examined psychotic transition rates in 59 subjects (mean age, 20) who met ultra high-risk criteria. They received:

• a needs-based intervention (NBI) comprising case management, psychotropics excluding antipsychotics, and supportive psychotherapy
  
• or a specific preventive intervention (SPI) that included risperidone, 1 to 2 mg/d, and a modified cognitive-behavioral therapy (CBT).
  

Table 2

Should you intervene with patients in suspected psychotic prodrome?

Arguments for:
Early treatment may prevent psychosocial decline Treatment may delay or ameliorate psychosis onset Treatment may improve patient awareness and acceptance of diagnosis Antipsychotics are effective for symptoms and may be neuroprotective Medications may improve overall outcome Neuroimaging findings may predict psychosis Outcome scales have improved diagnosis Treatment may reduce prodrome duration and improve prognosis
Arguments against:
Treatment would likely be given to persons who would not develop psychosis Treatment would unnecessarily stigmatize individuals who do not have schizophrenia Exposing patients with uncertain diagnoses to treatment risks is an ethical dilemma Antipsychotics are associated with side effects Psychotic prodrome studies are inconclusive, with small sample sizes and short follow-up No biological markers exist to predict psychosis

After 6 months, 10 of 28 (36%) in the NBI group had converted to a first episode of psychosis, compared with 3 of 31 (10%) in the SPI group (P = 0.03). At this point, risperidone was stopped, and all patients were offered NBI for 6 more months.

At 12-months’ follow-up, another 3 SPI patients who had been partially adherent or non-adherent to antipsychotic therapy had converted to psychosis. For adherent SPI patients, protection against conversion appeared to persist for 6 months after risperidone therapy ended. All medication side effects were mild and transient.

Table 3

Psychotic prodrome: Unanswered clinical questions

Does prodrome reflect a vulnerability for progression to psychosis? Can prodrome progress to psychosis in the absence of early interventions? Does duration of untreated psychosis predict prognosis? Do treatments reduce the risk of conversion to psychosis? Do interventions alter disease severity and prevent relapses? Can early interventions prevent cognitive and functional impairment?

Olanzapine. One double-blind, randomized, placebo-controlled trial has been published using olanzapine in patients with a prodromal syndrome.¹⁶ Sixty patients received olanzapine, 5 to 15 mg/d, or placebo. The olanzapine group showed significant reductions in positive, negative, and disorganization subscale scores and total SOPS and Positive and Negative Syndrome Scale scores, compared with the placebo group.

In the first year, 11 of 29 placebo-group patients and 5 of 31 receiving olanzapine converted to psychosis. Among patients receiving no treatment in the second year, 2 of 8 former placebo patients and 3 of 9 former olanzapine patients converted to psychosis.

Discontinuation rates were 35% and 28%, respectively. Compared with the placebo group, patients taking olanzapine experienced greater weight gain, suggesting that risks associated with antipsychotic therapy may exceed unproven benefits in this population.

Discussion. Little information exists on using quetiapine, ziprasidone, or aripiprazole in prodromal patients. As cited above, preliminary studies with risperidone and olanzapine suggest that these agents may improve several domains of psychotic prodrome. The evidence does not support firm conclusions, however, given the trials’ small sample sizes and brief duration.

The prevalence of obesity and metabolic syndrome in patients with schizophrenia and the added metabolic risks associated with atypical antipsychotics make their use during the prodrome controversial. Weighing the potential advantages and disadvantages (Table 2), we consider antipsychotics to be the last resort after psychosocial interventions have failed to improve prodromal symptoms.

Low-dose atypical antipsychotics may be warranted for some patients, but their use requires stringent monitoring of:

 

• weight and waist circumference
  
• vital signs
  
• metabolic parameters such as fasting blood glucose and lipid profile
  
• abnormal involuntary movements
  
• prolactin elevations.
  

Lifestyle modification—including diet and exercise to counteract the risk of weight gain—must be part of the treatment plan. Because the exact risks of antipsychotics are unknown—particularly their weight-gain potential among children and adolescents—we recommend specialist consultation (see Related resources) and careful documentation of all treatment decisions and discussions.

OTHER THERAPIES

Antidepressants. Researchers are also exploring the efficacy of using antidepressants and anxiolytics in the prodromal phase. The only published naturalistic study of adolescents found antidepressants alone or in combination with mood stabilizers or anxiolytics to be as effective as atypical antipsychotics in treating prodromal symptoms.¹⁷ A more substantial study is ongoing.

Psychotherapy. For patients with a suspected psychotic prodrome, nondrug strategies may help minimize functional and cognitive impairments, ease distress, and improve coping skills.

CBT has been shown to reduce psychotic progression over 12 months.¹⁸ Use CBT to help patients cope with the illness while focusing on:

• symptom monitoring
  
• premorbid and present functioning
  
• establishing a therapeutic alliance
  
• assessing the patient’s experience of psychosis and any thought distortions.
  

Also assess and treat co-occurring conditions such as alcohol and drug abuse. Involving the family in early intervention has been shown to improve prognosis.¹⁹

‘REAL WORLD’ EARLY INTERVENTION

Patients with prodromal symptoms are often referred to psychiatrists by family members, primary care physicians, or other mental health professionals. They tend to be young adults, and a few may present in their teens. Most are experiencing behavioral changes such as social isolation, feeling suspicious, perceptual disturbances, depression, and/or anxiety symptoms that seem abnormal but fall short of DSM-IV criteria for schizophrenia diagnosis.

Many clinical questions about schizophrenia’s prodromal phase remain unanswered (Table 3). Our primary aim is to adequately assess these patients and provide treatment and follow-up, taking into account:

• the individual’s presentation
  
• risks and benefits of available interventions.
  

Begin by educating patients and their families and providing social and emotional support to alleviate their distress that a mental illness may be developing (Box 2). Psychosocial and pharmacologic treatment options are based on presenting symptoms and other patient-specific variables.

Box 2

Psychotherapy should emphasize coping strategies, education about warning signals of psychotic conversion, establishing a therapeutic alliance, assessing thought distortions, and monitoring premorbid and present functioning.

Consider atypical antipsychotics for patients with distressing psychotic symptoms, rapidly deteriorating function, increased agitation, and safety risks. Consider antidepressant and/or anxiolytic therapy for depression and anxiety, respectively.

Discuss at length with patients and families the risks and benefits of pharmacologic treatments. When clinically appropriate, cautiously discontinue or taper any medication with patients’ consent, while monitoring for side effects and symptoms.

Related resources

• Issue devoted to early prodrome research. SchizophrBull 2003;29(4):621-879.
  
• Diagnostic and therapeutic intervention during psychotic prodrome. CNS Spectrums 2004;9(8):578-606.
  
• PRIME (Prevention through Risk Identification, Management & Education) Research Clinic. Department of Psychiatry, Yale University. http://info.med.yale.edu/psych/prime/pintro.html.
  
• Youth Mental Health Update. Schizophrenia: New strategies for early detection and treatment. RAPP Clinic, Zucker Hillside Hospital, Glen Oaks, NY. http://schoolnet.lij.edu/eshare/files/rapp.html
  

Drug brand names

• Aripiprazole • Abilify
  
• Olanzapine • Zyprexa
  
• Quetiapine • Seroquel
  
• Risperidone • Risperdal
  
• Ziprasidone • Geodon
  

Disclosures

Dr. Narasimhan receives research support from Eli Lilly and Co. and Janssen Pharmaceutica and is a speaker or consultant for Eli Lilly and Co., Pfizer Inc., and Abbott Laboratories.

Dr. Buckley receives research support from AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly and Co., Janssen Pharmaceutica, Novartis Pharmaceuticals Corp., Pfizer Inc., and Solvay Pharmaceuticals. He is a consultant to and/or speaker for Abbott Laboratories, Alamo Pharmaceuticals, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly and Co., Janssen Pharmaceutica, Novartis Pharmaceuticals Corp., Pfizer Inc., and Pharmstar.

Because 40% of individuals with a psychotic prodrome develop schizophrenia, detecting and preventing this transition could improve many patients’ lives. Unfortunately:

• psychotic prodrome lacks clear-cut symptoms and is difficult to identify
  
• little evidence exists to help clinicians select psychotropics and decide how long to use them
  
• treating all prodromal patients would expose those who never develop psychosis to the risk of psychotropics’ side effects.
  

How, then, can psychiatrists help patients who present with possible prodromal symptoms? Based on research and our experience, this article describes the psychotic prodrome and offers a pragmatic, evidence-based approach to diagnosis and treatment.

WHAT CAUSES PSYCHOTIC CONVERSION?

Reduced gray matter volumes in certain brain regions may be associated with conversion to psychosis (Box 1). Stress also may play a role; elevated stress-reactive cortisol levels are associated with positive symptom severity in the prodrome.¹ Other factors being investigated include obstetric complications at birth, maternal age >30, premorbid schizotypal personality disorder, and impaired olfaction.

Symptoms. Nearly 80% of patients with schizophrenia experience a psychotic prodrome that lasts a few months to several years.² Common features include:

• gradual worsening of perceptual disturbance
  
• referential thinking
  
• paranoia
  
• mild cognitive deficits
  
• mood lability
  
• impulsivity
  
• suicidality
  
• declining social function and academic performance.^3,4
  

Patients with these symptoms may be at imminent risk; if untreated, an estimated 40% progress to schizophrenia within 1 year.⁵

A premorbid phase often precedes the prodrome, with symptoms such as impaired attention, soft neurologic signs, and subtle social deficits. These changes may be harbingers of the prodrome but are too nonspecific to be diagnostic. Other functional impairments—including anxiety, depression, drug abuse, and psychosocial factors such as school stress—may mimic schizophrenic prodrome.

Prognosis. Studies of patients’ first schizophrenia episodes suggest that prodrome duration may predict outcome. A longer prodrome is thought to indicate a poor prognosis,⁶ such as in patients who wait a year before seeking treatment.⁷ A review of 22 studies of first-episode psychosis found early psychosocial and pharmacologic interventions improved long-term prognosis, and medication discontinuation predicted more-severe and chronic disease.⁸

Box 1

Genetic risk. Schizophrenia has a strong genetic predisposition, although not everyone in the genetic high-risk group develops schizophrenia. Persons with a family history of schizophrenia have a 10% to 20% risk of psychotic conversion.⁹

Pioneering work by McGorry et al¹⁰ identified an “ultra high-risk group” with a psychotic conversion rate of 40% to 60%. These patients present with three symptom patterns:

• attenuated positive symptoms
  
• brief intermittent psychotic episodes
  
• genetic risk and recent deterioration syndrome (Table 1).
  

Table 1

3 patient groups considered at ‘ultra high risk’ to develop schizophrenia

Patients with…	Symptoms
Attenuated psychotic symptoms	Overvalued ideas, perceptual disorders
	Present at least 1 week; not >5 years
	At least 1 symptom several times a week
Brief intermittent psychotic episodes	Frank psychotic features
	Resolve spontaneously within 7 days
	Can be drug-induced
Genetic risk and recent deterioration syndrome	Psychotic disorder in a first-degree relative
	Schizotypal personality disorder
	Present at least 1 month; not >5 years
	Significant functional decline
Source: Adapted from reference 10

Early identification of these high-risk individuals with perceptual distortions, frank psychotic symptoms, family history of psychosis, and schizotypal personality disorder may aid in early recognition and treatment.

The Edinburgh High Risk Study of 162 individuals ages 16 to 25 showed more marked psychopathology in those with at least two close relatives with schizophrenia, compared with control groups. A direct correlation was seen between genetic liability and poor neurocognitive performance.¹¹

PRODROME RATING SCALES

Researchers are using outcome measures to diagnose prodromal symptoms and assess their severity. Operational, validated assessment tools include:

• Bonn Scale for the Assessment of Basic Symptoms (BSABS): captures subtle changes in thinking, feeling, and perception.
  
• Schizophrenia Prediction Instrument for Adults (SPI-A): defines prepsychotic deviations and rates symptoms that are subjectively experienced by the patient.
  
• Comprehensive Assessment of At Risk Mental State (CAARMS): defines ultra high-risk criteria and incorporates eight dimensions of psychopathology.
  
• Scale of Prodromal Symptoms (SOPS): rates psychosis severity. When embedded within the Structured Interview for Prodromal Syndromes (SIPS), the SOPS determines the presence or absence of psychosis and predicts progression to psychopathology.
  
• Criteria for Prodromal Symptoms (COPS): defines ultra high-risk categories.
  
• Presence of Psychosis Scale (POPS): rates severity, intensity, and duration of positive prodromal symptoms.¹²
  

 

These instruments may identify prodromal symptoms in psychiatric practice, but further validation of clinical criteria is needed before they could be recommended for routine patient assessment.

PROPHYLACTIC ANTIPSYCHOTICS?

Atypical antipsychotics may be the standard of care for patients with a first psychotic episode, but this intervention is based on few double-blind controlled trials. Not surprisingly, only a handful of studies have examined antipsychotic therapy for the prodrome’s less clear-cut symptoms.

Risperidone. An 8- to 12-week open-label study in adolescents with first- and second-degree relatives with schizophrenia¹³ included four prodromal and six first-episode psychosis patients who met criteria for a cluster A personality disorder. Risperidone, 1.0 mg/d and 1.8 mg/d, respectively, improved thought disorder and attention symptoms, as measured with the Child Behavior Checklist. Verbal memory improved minimally, and no medication side effects were reported.

An open-label observational study¹⁴ identified four middle-aged subjects with a genetic risk of schizophrenia who reported negative symptoms and neurocognitive deficits. Risperidone, started at 0.25 mg/d and gradually increased to a maximum of 2 mg/d, improved negative symptoms, attention, and working memory. Mild side effects including tremors, sedation, dry mouth, and anxiety symptoms were reported.

An open-label, randomized, comparator trial¹⁵ examined psychotic transition rates in 59 subjects (mean age, 20) who met ultra high-risk criteria. They received:

• a needs-based intervention (NBI) comprising case management, psychotropics excluding antipsychotics, and supportive psychotherapy
  
• or a specific preventive intervention (SPI) that included risperidone, 1 to 2 mg/d, and a modified cognitive-behavioral therapy (CBT).
  

Table 2

Should you intervene with patients in suspected psychotic prodrome?

Arguments for:
Early treatment may prevent psychosocial decline Treatment may delay or ameliorate psychosis onset Treatment may improve patient awareness and acceptance of diagnosis Antipsychotics are effective for symptoms and may be neuroprotective Medications may improve overall outcome Neuroimaging findings may predict psychosis Outcome scales have improved diagnosis Treatment may reduce prodrome duration and improve prognosis
Arguments against:
Treatment would likely be given to persons who would not develop psychosis Treatment would unnecessarily stigmatize individuals who do not have schizophrenia Exposing patients with uncertain diagnoses to treatment risks is an ethical dilemma Antipsychotics are associated with side effects Psychotic prodrome studies are inconclusive, with small sample sizes and short follow-up No biological markers exist to predict psychosis

After 6 months, 10 of 28 (36%) in the NBI group had converted to a first episode of psychosis, compared with 3 of 31 (10%) in the SPI group (P = 0.03). At this point, risperidone was stopped, and all patients were offered NBI for 6 more months.

At 12-months’ follow-up, another 3 SPI patients who had been partially adherent or non-adherent to antipsychotic therapy had converted to psychosis. For adherent SPI patients, protection against conversion appeared to persist for 6 months after risperidone therapy ended. All medication side effects were mild and transient.

Table 3

Psychotic prodrome: Unanswered clinical questions

Does prodrome reflect a vulnerability for progression to psychosis? Can prodrome progress to psychosis in the absence of early interventions? Does duration of untreated psychosis predict prognosis? Do treatments reduce the risk of conversion to psychosis? Do interventions alter disease severity and prevent relapses? Can early interventions prevent cognitive and functional impairment?

Olanzapine. One double-blind, randomized, placebo-controlled trial has been published using olanzapine in patients with a prodromal syndrome.¹⁶ Sixty patients received olanzapine, 5 to 15 mg/d, or placebo. The olanzapine group showed significant reductions in positive, negative, and disorganization subscale scores and total SOPS and Positive and Negative Syndrome Scale scores, compared with the placebo group.

In the first year, 11 of 29 placebo-group patients and 5 of 31 receiving olanzapine converted to psychosis. Among patients receiving no treatment in the second year, 2 of 8 former placebo patients and 3 of 9 former olanzapine patients converted to psychosis.

Discontinuation rates were 35% and 28%, respectively. Compared with the placebo group, patients taking olanzapine experienced greater weight gain, suggesting that risks associated with antipsychotic therapy may exceed unproven benefits in this population.

Discussion. Little information exists on using quetiapine, ziprasidone, or aripiprazole in prodromal patients. As cited above, preliminary studies with risperidone and olanzapine suggest that these agents may improve several domains of psychotic prodrome. The evidence does not support firm conclusions, however, given the trials’ small sample sizes and brief duration.

The prevalence of obesity and metabolic syndrome in patients with schizophrenia and the added metabolic risks associated with atypical antipsychotics make their use during the prodrome controversial. Weighing the potential advantages and disadvantages (Table 2), we consider antipsychotics to be the last resort after psychosocial interventions have failed to improve prodromal symptoms.

Low-dose atypical antipsychotics may be warranted for some patients, but their use requires stringent monitoring of:

 

• weight and waist circumference
  
• vital signs
  
• metabolic parameters such as fasting blood glucose and lipid profile
  
• abnormal involuntary movements
  
• prolactin elevations.
  

Lifestyle modification—including diet and exercise to counteract the risk of weight gain—must be part of the treatment plan. Because the exact risks of antipsychotics are unknown—particularly their weight-gain potential among children and adolescents—we recommend specialist consultation (see Related resources) and careful documentation of all treatment decisions and discussions.

OTHER THERAPIES

Antidepressants. Researchers are also exploring the efficacy of using antidepressants and anxiolytics in the prodromal phase. The only published naturalistic study of adolescents found antidepressants alone or in combination with mood stabilizers or anxiolytics to be as effective as atypical antipsychotics in treating prodromal symptoms.¹⁷ A more substantial study is ongoing.

Psychotherapy. For patients with a suspected psychotic prodrome, nondrug strategies may help minimize functional and cognitive impairments, ease distress, and improve coping skills.

CBT has been shown to reduce psychotic progression over 12 months.¹⁸ Use CBT to help patients cope with the illness while focusing on:

• symptom monitoring
  
• premorbid and present functioning
  
• establishing a therapeutic alliance
  
• assessing the patient’s experience of psychosis and any thought distortions.
  

Also assess and treat co-occurring conditions such as alcohol and drug abuse. Involving the family in early intervention has been shown to improve prognosis.¹⁹

‘REAL WORLD’ EARLY INTERVENTION

Patients with prodromal symptoms are often referred to psychiatrists by family members, primary care physicians, or other mental health professionals. They tend to be young adults, and a few may present in their teens. Most are experiencing behavioral changes such as social isolation, feeling suspicious, perceptual disturbances, depression, and/or anxiety symptoms that seem abnormal but fall short of DSM-IV criteria for schizophrenia diagnosis.

Many clinical questions about schizophrenia’s prodromal phase remain unanswered (Table 3). Our primary aim is to adequately assess these patients and provide treatment and follow-up, taking into account:

• the individual’s presentation
  
• risks and benefits of available interventions.
  

Begin by educating patients and their families and providing social and emotional support to alleviate their distress that a mental illness may be developing (Box 2). Psychosocial and pharmacologic treatment options are based on presenting symptoms and other patient-specific variables.

Box 2

Psychotherapy should emphasize coping strategies, education about warning signals of psychotic conversion, establishing a therapeutic alliance, assessing thought distortions, and monitoring premorbid and present functioning.

Consider atypical antipsychotics for patients with distressing psychotic symptoms, rapidly deteriorating function, increased agitation, and safety risks. Consider antidepressant and/or anxiolytic therapy for depression and anxiety, respectively.

Discuss at length with patients and families the risks and benefits of pharmacologic treatments. When clinically appropriate, cautiously discontinue or taper any medication with patients’ consent, while monitoring for side effects and symptoms.

Related resources

• Issue devoted to early prodrome research. SchizophrBull 2003;29(4):621-879.
  
• Diagnostic and therapeutic intervention during psychotic prodrome. CNS Spectrums 2004;9(8):578-606.
  
• PRIME (Prevention through Risk Identification, Management & Education) Research Clinic. Department of Psychiatry, Yale University. http://info.med.yale.edu/psych/prime/pintro.html.
  
• Youth Mental Health Update. Schizophrenia: New strategies for early detection and treatment. RAPP Clinic, Zucker Hillside Hospital, Glen Oaks, NY. http://schoolnet.lij.edu/eshare/files/rapp.html
  

Drug brand names

• Aripiprazole • Abilify
  
• Olanzapine • Zyprexa
  
• Quetiapine • Seroquel
  
• Risperidone • Risperdal
  
• Ziprasidone • Geodon
  

Disclosures

Dr. Narasimhan receives research support from Eli Lilly and Co. and Janssen Pharmaceutica and is a speaker or consultant for Eli Lilly and Co., Pfizer Inc., and Abbott Laboratories.

Dr. Buckley receives research support from AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly and Co., Janssen Pharmaceutica, Novartis Pharmaceuticals Corp., Pfizer Inc., and Solvay Pharmaceuticals. He is a consultant to and/or speaker for Abbott Laboratories, Alamo Pharmaceuticals, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly and Co., Janssen Pharmaceutica, Novartis Pharmaceuticals Corp., Pfizer Inc., and Pharmstar.

Because 40% of individuals with a psychotic prodrome develop schizophrenia, detecting and preventing this transition could improve many patients’ lives. Unfortunately:

• psychotic prodrome lacks clear-cut symptoms and is difficult to identify
  
• little evidence exists to help clinicians select psychotropics and decide how long to use them
  
• treating all prodromal patients would expose those who never develop psychosis to the risk of psychotropics’ side effects.
  

How, then, can psychiatrists help patients who present with possible prodromal symptoms? Based on research and our experience, this article describes the psychotic prodrome and offers a pragmatic, evidence-based approach to diagnosis and treatment.

WHAT CAUSES PSYCHOTIC CONVERSION?

Reduced gray matter volumes in certain brain regions may be associated with conversion to psychosis (Box 1). Stress also may play a role; elevated stress-reactive cortisol levels are associated with positive symptom severity in the prodrome.¹ Other factors being investigated include obstetric complications at birth, maternal age >30, premorbid schizotypal personality disorder, and impaired olfaction.

Symptoms. Nearly 80% of patients with schizophrenia experience a psychotic prodrome that lasts a few months to several years.² Common features include:

• gradual worsening of perceptual disturbance
  
• referential thinking
  
• paranoia
  
• mild cognitive deficits
  
• mood lability
  
• impulsivity
  
• suicidality
  
• declining social function and academic performance.^3,4
  

Patients with these symptoms may be at imminent risk; if untreated, an estimated 40% progress to schizophrenia within 1 year.⁵

A premorbid phase often precedes the prodrome, with symptoms such as impaired attention, soft neurologic signs, and subtle social deficits. These changes may be harbingers of the prodrome but are too nonspecific to be diagnostic. Other functional impairments—including anxiety, depression, drug abuse, and psychosocial factors such as school stress—may mimic schizophrenic prodrome.

Prognosis. Studies of patients’ first schizophrenia episodes suggest that prodrome duration may predict outcome. A longer prodrome is thought to indicate a poor prognosis,⁶ such as in patients who wait a year before seeking treatment.⁷ A review of 22 studies of first-episode psychosis found early psychosocial and pharmacologic interventions improved long-term prognosis, and medication discontinuation predicted more-severe and chronic disease.⁸

Box 1

Genetic risk. Schizophrenia has a strong genetic predisposition, although not everyone in the genetic high-risk group develops schizophrenia. Persons with a family history of schizophrenia have a 10% to 20% risk of psychotic conversion.⁹

Pioneering work by McGorry et al¹⁰ identified an “ultra high-risk group” with a psychotic conversion rate of 40% to 60%. These patients present with three symptom patterns:

• attenuated positive symptoms
  
• brief intermittent psychotic episodes
  
• genetic risk and recent deterioration syndrome (Table 1).
  

Table 1

3 patient groups considered at ‘ultra high risk’ to develop schizophrenia

Patients with…	Symptoms
Attenuated psychotic symptoms	Overvalued ideas, perceptual disorders
	Present at least 1 week; not >5 years
	At least 1 symptom several times a week
Brief intermittent psychotic episodes	Frank psychotic features
	Resolve spontaneously within 7 days
	Can be drug-induced
Genetic risk and recent deterioration syndrome	Psychotic disorder in a first-degree relative
	Schizotypal personality disorder
	Present at least 1 month; not >5 years
	Significant functional decline
Source: Adapted from reference 10

Early identification of these high-risk individuals with perceptual distortions, frank psychotic symptoms, family history of psychosis, and schizotypal personality disorder may aid in early recognition and treatment.

The Edinburgh High Risk Study of 162 individuals ages 16 to 25 showed more marked psychopathology in those with at least two close relatives with schizophrenia, compared with control groups. A direct correlation was seen between genetic liability and poor neurocognitive performance.¹¹

PRODROME RATING SCALES

Researchers are using outcome measures to diagnose prodromal symptoms and assess their severity. Operational, validated assessment tools include:

• Bonn Scale for the Assessment of Basic Symptoms (BSABS): captures subtle changes in thinking, feeling, and perception.
  
• Schizophrenia Prediction Instrument for Adults (SPI-A): defines prepsychotic deviations and rates symptoms that are subjectively experienced by the patient.
  
• Comprehensive Assessment of At Risk Mental State (CAARMS): defines ultra high-risk criteria and incorporates eight dimensions of psychopathology.
  
• Scale of Prodromal Symptoms (SOPS): rates psychosis severity. When embedded within the Structured Interview for Prodromal Syndromes (SIPS), the SOPS determines the presence or absence of psychosis and predicts progression to psychopathology.
  
• Criteria for Prodromal Symptoms (COPS): defines ultra high-risk categories.
  
• Presence of Psychosis Scale (POPS): rates severity, intensity, and duration of positive prodromal symptoms.¹²
  

 

These instruments may identify prodromal symptoms in psychiatric practice, but further validation of clinical criteria is needed before they could be recommended for routine patient assessment.

PROPHYLACTIC ANTIPSYCHOTICS?

Atypical antipsychotics may be the standard of care for patients with a first psychotic episode, but this intervention is based on few double-blind controlled trials. Not surprisingly, only a handful of studies have examined antipsychotic therapy for the prodrome’s less clear-cut symptoms.

Risperidone. An 8- to 12-week open-label study in adolescents with first- and second-degree relatives with schizophrenia¹³ included four prodromal and six first-episode psychosis patients who met criteria for a cluster A personality disorder. Risperidone, 1.0 mg/d and 1.8 mg/d, respectively, improved thought disorder and attention symptoms, as measured with the Child Behavior Checklist. Verbal memory improved minimally, and no medication side effects were reported.

An open-label observational study¹⁴ identified four middle-aged subjects with a genetic risk of schizophrenia who reported negative symptoms and neurocognitive deficits. Risperidone, started at 0.25 mg/d and gradually increased to a maximum of 2 mg/d, improved negative symptoms, attention, and working memory. Mild side effects including tremors, sedation, dry mouth, and anxiety symptoms were reported.

An open-label, randomized, comparator trial¹⁵ examined psychotic transition rates in 59 subjects (mean age, 20) who met ultra high-risk criteria. They received:

• a needs-based intervention (NBI) comprising case management, psychotropics excluding antipsychotics, and supportive psychotherapy
  
• or a specific preventive intervention (SPI) that included risperidone, 1 to 2 mg/d, and a modified cognitive-behavioral therapy (CBT).
  

Table 2

Should you intervene with patients in suspected psychotic prodrome?

Arguments for:
Early treatment may prevent psychosocial decline Treatment may delay or ameliorate psychosis onset Treatment may improve patient awareness and acceptance of diagnosis Antipsychotics are effective for symptoms and may be neuroprotective Medications may improve overall outcome Neuroimaging findings may predict psychosis Outcome scales have improved diagnosis Treatment may reduce prodrome duration and improve prognosis
Arguments against:
Treatment would likely be given to persons who would not develop psychosis Treatment would unnecessarily stigmatize individuals who do not have schizophrenia Exposing patients with uncertain diagnoses to treatment risks is an ethical dilemma Antipsychotics are associated with side effects Psychotic prodrome studies are inconclusive, with small sample sizes and short follow-up No biological markers exist to predict psychosis

After 6 months, 10 of 28 (36%) in the NBI group had converted to a first episode of psychosis, compared with 3 of 31 (10%) in the SPI group (P = 0.03). At this point, risperidone was stopped, and all patients were offered NBI for 6 more months.

At 12-months’ follow-up, another 3 SPI patients who had been partially adherent or non-adherent to antipsychotic therapy had converted to psychosis. For adherent SPI patients, protection against conversion appeared to persist for 6 months after risperidone therapy ended. All medication side effects were mild and transient.

Table 3

Psychotic prodrome: Unanswered clinical questions

Does prodrome reflect a vulnerability for progression to psychosis? Can prodrome progress to psychosis in the absence of early interventions? Does duration of untreated psychosis predict prognosis? Do treatments reduce the risk of conversion to psychosis? Do interventions alter disease severity and prevent relapses? Can early interventions prevent cognitive and functional impairment?

Olanzapine. One double-blind, randomized, placebo-controlled trial has been published using olanzapine in patients with a prodromal syndrome.¹⁶ Sixty patients received olanzapine, 5 to 15 mg/d, or placebo. The olanzapine group showed significant reductions in positive, negative, and disorganization subscale scores and total SOPS and Positive and Negative Syndrome Scale scores, compared with the placebo group.

In the first year, 11 of 29 placebo-group patients and 5 of 31 receiving olanzapine converted to psychosis. Among patients receiving no treatment in the second year, 2 of 8 former placebo patients and 3 of 9 former olanzapine patients converted to psychosis.

Discontinuation rates were 35% and 28%, respectively. Compared with the placebo group, patients taking olanzapine experienced greater weight gain, suggesting that risks associated with antipsychotic therapy may exceed unproven benefits in this population.

Discussion. Little information exists on using quetiapine, ziprasidone, or aripiprazole in prodromal patients. As cited above, preliminary studies with risperidone and olanzapine suggest that these agents may improve several domains of psychotic prodrome. The evidence does not support firm conclusions, however, given the trials’ small sample sizes and brief duration.

The prevalence of obesity and metabolic syndrome in patients with schizophrenia and the added metabolic risks associated with atypical antipsychotics make their use during the prodrome controversial. Weighing the potential advantages and disadvantages (Table 2), we consider antipsychotics to be the last resort after psychosocial interventions have failed to improve prodromal symptoms.

Low-dose atypical antipsychotics may be warranted for some patients, but their use requires stringent monitoring of:

 

• weight and waist circumference
  
• vital signs
  
• metabolic parameters such as fasting blood glucose and lipid profile
  
• abnormal involuntary movements
  
• prolactin elevations.
  

Lifestyle modification—including diet and exercise to counteract the risk of weight gain—must be part of the treatment plan. Because the exact risks of antipsychotics are unknown—particularly their weight-gain potential among children and adolescents—we recommend specialist consultation (see Related resources) and careful documentation of all treatment decisions and discussions.

OTHER THERAPIES

Antidepressants. Researchers are also exploring the efficacy of using antidepressants and anxiolytics in the prodromal phase. The only published naturalistic study of adolescents found antidepressants alone or in combination with mood stabilizers or anxiolytics to be as effective as atypical antipsychotics in treating prodromal symptoms.¹⁷ A more substantial study is ongoing.

Psychotherapy. For patients with a suspected psychotic prodrome, nondrug strategies may help minimize functional and cognitive impairments, ease distress, and improve coping skills.

CBT has been shown to reduce psychotic progression over 12 months.¹⁸ Use CBT to help patients cope with the illness while focusing on:

• symptom monitoring
  
• premorbid and present functioning
  
• establishing a therapeutic alliance
  
• assessing the patient’s experience of psychosis and any thought distortions.
  

Also assess and treat co-occurring conditions such as alcohol and drug abuse. Involving the family in early intervention has been shown to improve prognosis.¹⁹

‘REAL WORLD’ EARLY INTERVENTION

Patients with prodromal symptoms are often referred to psychiatrists by family members, primary care physicians, or other mental health professionals. They tend to be young adults, and a few may present in their teens. Most are experiencing behavioral changes such as social isolation, feeling suspicious, perceptual disturbances, depression, and/or anxiety symptoms that seem abnormal but fall short of DSM-IV criteria for schizophrenia diagnosis.

Many clinical questions about schizophrenia’s prodromal phase remain unanswered (Table 3). Our primary aim is to adequately assess these patients and provide treatment and follow-up, taking into account:

• the individual’s presentation
  
• risks and benefits of available interventions.
  

Begin by educating patients and their families and providing social and emotional support to alleviate their distress that a mental illness may be developing (Box 2). Psychosocial and pharmacologic treatment options are based on presenting symptoms and other patient-specific variables.

Box 2

Psychotherapy should emphasize coping strategies, education about warning signals of psychotic conversion, establishing a therapeutic alliance, assessing thought distortions, and monitoring premorbid and present functioning.

Consider atypical antipsychotics for patients with distressing psychotic symptoms, rapidly deteriorating function, increased agitation, and safety risks. Consider antidepressant and/or anxiolytic therapy for depression and anxiety, respectively.

Discuss at length with patients and families the risks and benefits of pharmacologic treatments. When clinically appropriate, cautiously discontinue or taper any medication with patients’ consent, while monitoring for side effects and symptoms.

Related resources

• Issue devoted to early prodrome research. SchizophrBull 2003;29(4):621-879.
  
• Diagnostic and therapeutic intervention during psychotic prodrome. CNS Spectrums 2004;9(8):578-606.
  
• PRIME (Prevention through Risk Identification, Management & Education) Research Clinic. Department of Psychiatry, Yale University. http://info.med.yale.edu/psych/prime/pintro.html.
  
• Youth Mental Health Update. Schizophrenia: New strategies for early detection and treatment. RAPP Clinic, Zucker Hillside Hospital, Glen Oaks, NY. http://schoolnet.lij.edu/eshare/files/rapp.html
  

Drug brand names

• Aripiprazole • Abilify
  
• Olanzapine • Zyprexa
  
• Quetiapine • Seroquel
  
• Risperidone • Risperdal
  
• Ziprasidone • Geodon
  

Disclosures

Dr. Narasimhan receives research support from Eli Lilly and Co. and Janssen Pharmaceutica and is a speaker or consultant for Eli Lilly and Co., Pfizer Inc., and Abbott Laboratories.

Dr. Buckley receives research support from AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly and Co., Janssen Pharmaceutica, Novartis Pharmaceuticals Corp., Pfizer Inc., and Solvay Pharmaceuticals. He is a consultant to and/or speaker for Abbott Laboratories, Alamo Pharmaceuticals, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb Co., Eli Lilly and Co., Janssen Pharmaceutica, Novartis Pharmaceuticals Corp., Pfizer Inc., and Pharmstar.

Compulsive hoarding behavior is considered notoriously difficult to treat, but targeting its characteristic symptoms with medication and psychotherapy can be successful. This article provides a guide for the psychiatrist—alone or with a cognitive-behavioral therapist—to diagnose compulsive hoarding syndrome and help patients overcome the anxieties that fuel its symptoms.

WHAT IS COMPULSIVE HOARDING?

Hoarders acquire and are unable to discard items that others consider of little use or value.¹ They most often save newspapers, magazines, old clothing, bags, books, mail, notes, and lists. Hoarding and saving behaviors occur in nonclinical populations and with other neuropsychiatric disorders—schizophrenia, dementia, eating disorders, mental retardation—but are most often found in persons with obsessive-compulsive disorder (OCD).

OCD is a heterogeneous clinical entity with several major symptom domains:^2,3

• aggressive, sexual, and religious obsessions with checking compulsions
  
• symmetry/order obsessions with ordering, arranging, and repeating compulsions
  
• contamination obsessions with washing and cleaning compulsions
  
• hoarding and saving symptoms.
  

Box

Among OCD patients, 18% to 42% have hoarding and saving compulsions.^4,5 Hoarding and saving can be part of a broader clinical syndrome that includes indecisiveness, perfectionism, procrastination, difficulty organizing tasks, and avoiding routine daily activities.^6,7 The 1 to 2 million Americans whose most prominent and distressing OCD symptom is hoarding and saving and who show these other associated symptoms are considered to have “compulsive hoarding syndrome.”^7,8 Evidence suggests that this syndrome may be a neurobiologically distinct OCD variant (Box).^9-16

ASSESSMENT AND TREATMENT PLANNING

To manage compulsive hoarding syndrome, begin with a thorough neuropsychiatric evaluation:

• Rule out primary psychotic disorders, dementia, and other cognitive impairments and neurologic disorders.
  
• Rule out primary major depression, as clutter and self-neglect may be caused by amotivation, low energy, or hopelessness.
  
• Determine if the patient has OCD.
  

After making a compulsive hoarding diagnosis (Table 1),⁶ visit the patient’s home or view photographs to assess his or her environment and behaviors (Table 2).

Amount of clutter. Living areas may be so cluttered that sleeping in a bed, sitting on chairs, or preparing food on a kitchen counter are impossible. How much of the home is cluttered? How much floor and counter space is usable? Are rooms unusable or inaccessible because of clutter? Can the patient use the laundry, prepare food in the kitchen, use the shower, toilet, etc.?

Health or safety hazards. Huge piles of papers can be a fire hazard. Clutter may be blocking the exits. Collected items may extend beyond patients’ homes to their cars, garages, storage lockers, and even storage areas owned by friends and family.

Beliefs about possessions. Compulsive hoarders often have distorted feelings about their possessions. They may over-buy or impulsively purchase items they feel have emotional or monetary value. They may consider the items extensions of themselves and suffer grief-like loss when discarding things.⁷

Some collect free items—flyers, coupons, newspapers, discarded goods—hoping to save money or be prepared “just in case” the item is ever needed. This may represent unattainable expectations of perfection, needing to maintain preparedness for every possible contingency. Hoarders often believe they have poor memory and have catastrophic fears of what might happen if they forget something. Thus, their desire to keep their possessions in sight is strong.¹⁷

 

Information processing deficits. Because of anxieties about making mistakes, most hoarders have great difficulty making decisions.¹⁸ It is easier to not decide than to suffer the consequences of a “wrong” decision. To gauge this behavior, ask patients how long routine decisions take them and which decisions they procrastinate or avoid.

Compulsive hoarders often have trouble categorizing possessions;^6,7 because every item feels unique, they create a special category for each one and resist storing items together.

Table 1

Proposed criteria to diagnose obsessive-compulsive hoarding*

Patient acquires and fails to discard a large number of possessions that appear useless or of limited value
Clutter prevents patient from using living or work spaces for activities for which they were designed
Hoarding behavior causes significant distress or functional impairment
* Proposed by Frost and Hartl, reference 6.

Many hoarders also report marked distractibility and inattention, jumping from one task to the next without completing any of them. Their communication style is often as cluttered and disorganized as their homes, with tangential, circumstantial, and over-inclusive descriptions.

Avoidance behaviors are a hallmark of the compulsive hoarding syndrome. To avoid deciding to discard items, they put them in a box, garage, rented storage facility, etc. They may also avoid routine decision-making tasks that could lead to making a mistake.

Daily functioning. Hoarders may take a long time to do even small chores, such as taking an hour to pay one bill. An inordinate amount of time may be spent “churning”—moving items from one pile to another but never discarding any item or establishing a consistent system or organization.

Medication compliance. Compulsive hoarders often forget to take medications or take them at inappropriate times. They may lose their medications in the clutter.

Insight. Hoarders often have little awareness of how their behavior and clutter affect their lives.¹⁹ They minimize the clutter in their homes and its health and safety risks. Insight can fluctuate over time and needs to be assessed repeatedly during treatment.

Table 2

Assessing a patient with compulsive hoarding symptoms

Domain	Useful questions or strategies
Amount of clutter	Visit home and/or see pictures
Hazards relating to clutter	Ask: What precautions do you take to reduce risk of fire? Have you ever had a problem with rodent or insect infestation as a result of the clutter? Have neighbors complained about the risks of fire or infestation that the clutter might impose on their homes?
Beliefs about loss of possessions	Ask: What is the worst thing that would happen if you threw this item away? If you did not have this, what do you think would happen?
Information-processing deficits	Ask: How long do routine decisions take you? Which decisions do you procrastinate or avoid?
Decision-making and organizational skills	Ask: How do you pay and store your bills?
Avoidance behaviors	Ask: Do you avoid other things (sorting mail, returning calls, doing dishes, or paying bills, rent, or taxes)?
Daily functioning	Ask: Do you get everything done that you want to do? Are you often late? Do you have difficulty starting or finishing tasks? Describe a typical day.
Insight	Ask: Do you think this amount of clutter is normal? Do you think having this clutter has caused problems in your life?
Motivation for treatment	Ask: What brings you into therapy now? Do you think you have a problem with excessive hoarding/saving? If it was not for your family, would you come for help?
Social and occupational functioning	Ask: How has your clutter affected your personal relationships? When was the last time you had someone come to your home? What prevents you from working right now? Are you working to your full potential?
Support from friends and family	Ask: What does your family say about your clutter? Do your friends or family understand what is going on?
Treatment compliance	Ask: How long does it typically take before you renew your prescriptions when you run out of medications?

Table 3

Cognitive behavioral therapy for compulsive hoarding

Treatment sequence	Methods and goals
Educate patient about hoarding	Help improve insight and motivation
Set up treatment	With patient, select target area of clutter
	Assess items together, creating a hierarchy of least to most difficult areas to sort and items to discard
	Create realistic categories and a storage system
Begin discarding	Patient must decide to keep or discard each item and permanently remove it from pile
	Patient must store saved items appropriately
	Continue until area is clear, then move to next area
Plan and implement appropriate use of space
Stop incoming clutter	Cancel subscriptions
	Address compulsive buying and acquisition
Provide organization training	Organize possessions, time, tasks, etc.
Prevent relapse	Replace hoarding with healthier behaviors to prevent clutter from re-accumulating
Source: Adapted from reference 23.

Social and occupational functioning. Many compulsive hoarders have very little family or social support. They frequently are too embarrassed by their clutter to have people come to their homes, sometimes for years. The syndrome frequently impairs work performance.

 

²⁰

Motivation. Like insight, motivation can fluctuate over time. Patients usually must work tremendously hard to adhere to treatment. To support these efforts, we periodically review with patients compulsive hoarding’s negative effects and the activities they would enjoy—such as improved relationships, greater work capacity, hobbies—if overcoming this behavior allowed them more time and space.

Rating scales. The symptom checklist of the Yale-Brown Obsessive-Compulsive Scale (YBOCS)²¹ contains two items for hoarding obsessions and compulsions but none for avoidance behaviors, which are prominent with compulsive hoarding. The Saving Inventory-Revised²² is a validated, 23-item self-report measure of clutter, difficulty discarding, and excessive acquisition, which distinguishes compulsive hoarders, nonhoarding OCD patients, and normal controls.

TREATMENT

The compulsive hoarder’s problems will not be solved by someone else throwing away or organizing his or her possessions. These actions often anger patients, who see them as intrusive and a loss of control.

In our experience, family members’ attempts to intervene can disrupt relationships and worsen hoarders’ social withdrawal. “Taking over” also does not help the patient create a sustainable system for keeping clutter-free.

Algorithm

Medication treatment for compulsive hoarding*

Start with SSRIs, as for nonhoarding OCD (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline)	High doses and 12-week trials Some compulsive hoarders will respond well to SSRIs Other OCD symptoms usually improve as well Comorbid depression and other anxiety symptoms may respond If ineffective, may need to do 3 or 4 full trials of different SSRIs, clomipramine, or venlafaxine
Treat comorbid conditions	Mood disorders, other anxiety disorders, ADHD, psychotic disorders, etc.
Use adjunctive medications if SSRIs give only partial response	Atypical antipsychotics (risperidone, olanzapine, quetiapine) Stimulants Mood stabilizers (for comorbid bipolar disorder, cyclothymia, or impulsivity)
* Combine medication treatment with cognitive-behavioral therapy
SSRI: selective serotonin reuptake inhibitor
OCD: obsessive-compulsive disorder
ADHD: attention-deficit/hyperactivity disorder

We find that combining cognitive-behavioral therapy (CBT) and medication is optimal treatment for compulsive hoarding,²³ although no controlled studies have compared this combination with each treatment alone. One controlled trial²⁴ and three uncontrolled trials^8,25,26 have shown some benefit of CBT for compulsive hoarding, although with poorer response and higher dropout rates than for nonhoarding OCD patients.

Psychotherapy. Exposure and response prevention (ERP) focuses on preventing incoming clutter, discarding, organizing, and relapse prevention (Table 3).²³ Start treatment by explaining compulsive hoarding syndrome to patients as having problems with information processing, obsessional anxiety, and avoiding decisions.

Preventing incoming clutter. Before you focus on discarding, patients must stop incoming clutter; otherwise, it will come in as fast as it goes out. We ask patients to keep a daily log of every item they acquire or buy to build their awareness of what triggers their behavior.

Discarding. To desensitize over time, we repeatedly expose patients to the anxiety, sadness, or anger they feel when discarding items and making decisions. We encourage them to provoke their anxiety by throwing away as many items as possible, keeping only necessary items.

We support ERP with cognitive restructuring, prompting patients to reframe their obsessive fears about losing something necessary or valuable. By thinking through the consequences of discarding their clutter, they challenge their erroneous beliefs that dire consequences will occur.

Organizing. When patients decide they must keep an item, ask them to immediately identify a specific place and deadline to store it. After an area is cleared, patients must keep it clear and use it for its intended purpose. Most patients need training in time management, scheduling, and prioritizing.

Relapse prevention. Replace hoarding behaviors with more-adaptive, healthy behaviors. Teach patients to create a realistic schedule that includes time for chores, eating and sleeping, CBT homework, and recreation. Treatment goals are to:

• extinguish obsessional fears and compulsive saving behaviors
  
• teach lasting organizational and decision-making skills, thereby reducing relapse risk.
  

Practical matters. Sorting and discarding a houseful of clutter takes time, although the patient does this primarily as homework. To control costs and your time commitment, consider collaborating with a CBT therapist trained in working with compulsive hoarders. Use your sessions with the patient to create hierarchies, go over assignments, do brief exposures, and monitor drug therapy.

Medications. No controlled studies have examined whether any medications are effective for compulsive hoarding syndrome. The treatment strategies and algorithm described here are based on our clinical experience, controlled trials of OCD patients, and limited OCD studies secondarily examining hoarders’ specific treatment responses.

Selective serotonin reuptake inhibitors (SSRIs) may be less effective for compulsive hoarding than for other compulsive behaviors.^14,27 Nevertheless, SSRIs may help alleviate hoarders’ core symptoms, other OCD symptoms, depression, and anxiety. For hoarding treatment to be effective, comorbid disorders must be treated and stabilized.

 

Several studies of SSRI use in OCD patients have shown modest improvements in compulsive hoarders:

• In a descriptive study of patients with compulsive hoarding, 1 of 18 (6%) patients had a “marked” response to at least one SSRI trial. The others showed a partial response,¹⁷ with YBOCS scores decreasing by at least 25% in approximately 50% of this group.
  
• When 17 OCD patients with hoarding symptoms were treated with paroxetine, CBT, or placebo, 18% responded to active treatment. Response was defined as a 40% reduction on YBOCS scores and “very much” or “much” improved on the Clinical Global Impression Scale (CGI).¹⁴
  

Based on these findings and our clinical experience, an effective approach to treating compulsive hoarding with medications—as with other OCD patients—is to start with SSRIs²⁸ (Algorithm). If adequate SSRI trials fail to improve a patient’s hoarding/saving symptoms, adjunctive medications can be added.

Atypical antipsychotics may be effective for OCD symptoms that do not respond adequately to SSRIs.²⁹ Conventional antipsychotics are also effective adjuncts to SSRIs—particularly for patients with coexisting tic or psychotic disorders³⁰—but consider the potential for extrapyramidal side effects and tardive dyskinesia.

We find that stimulants help some compulsive hoarders, particularly those with comorbid ADHD, other attentional problems, low motivation, or lethargy. Mood stabilizers are necessary to treat comorbid bipolar disorder, cyclothymia, and impulsivity.

Related resources

• Saxena S, Maidment K. Treatment of compulsive hoarding. J Clin Psychol 2004;60(11):1143-54.
  
• UCLA OCD Intensive Treatment Program. For information on research studies contact Karron Maidment (310) 794-7305 or visit www.mentalhealth.ucla.edu/projects/anxiety/ocdintensivetreatment.htm.
  
• Compulsive hoarding project. Institute of Living Anxiety Disorders Center, Hartford Hospital. http://instituteofliving.org/adc/compulsive_hoarding.htm
  
• Obsessive Compulsive Foundation compulsive hoarding Web site. http://www.ocfoundation.org/1005/index.html
  
• Neziroglu F, Bubrick J, Yaryura-Tobias JA. Overcoming compulsive hoarding. Oakland, CA: New Harbinger Publications, 2004.
  

Drug brand names

• Citalopram • Celexa
  
• Clomipramine • Anafranil
  
• Escitalopram • Lexapro
  
• Fluoxetine • Prozac
  
• Fluvoxamine • Luvox
  
• Olanzapine • Zyprexa
  
• Paroxetine • Paxil
  
• Quetiapine • Seroquel
  
• Risperidone • Risperdal
  
• Sertraline • Zoloft
  
• Venlafaxine • Effexor
  

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Compulsive hoarding behavior is considered notoriously difficult to treat, but targeting its characteristic symptoms with medication and psychotherapy can be successful. This article provides a guide for the psychiatrist—alone or with a cognitive-behavioral therapist—to diagnose compulsive hoarding syndrome and help patients overcome the anxieties that fuel its symptoms.

WHAT IS COMPULSIVE HOARDING?

Hoarders acquire and are unable to discard items that others consider of little use or value.¹ They most often save newspapers, magazines, old clothing, bags, books, mail, notes, and lists. Hoarding and saving behaviors occur in nonclinical populations and with other neuropsychiatric disorders—schizophrenia, dementia, eating disorders, mental retardation—but are most often found in persons with obsessive-compulsive disorder (OCD).

OCD is a heterogeneous clinical entity with several major symptom domains:^2,3

• aggressive, sexual, and religious obsessions with checking compulsions
  
• symmetry/order obsessions with ordering, arranging, and repeating compulsions
  
• contamination obsessions with washing and cleaning compulsions
  
• hoarding and saving symptoms.
  

Box

Among OCD patients, 18% to 42% have hoarding and saving compulsions.^4,5 Hoarding and saving can be part of a broader clinical syndrome that includes indecisiveness, perfectionism, procrastination, difficulty organizing tasks, and avoiding routine daily activities.^6,7 The 1 to 2 million Americans whose most prominent and distressing OCD symptom is hoarding and saving and who show these other associated symptoms are considered to have “compulsive hoarding syndrome.”^7,8 Evidence suggests that this syndrome may be a neurobiologically distinct OCD variant (Box).^9-16

ASSESSMENT AND TREATMENT PLANNING

To manage compulsive hoarding syndrome, begin with a thorough neuropsychiatric evaluation:

• Rule out primary psychotic disorders, dementia, and other cognitive impairments and neurologic disorders.
  
• Rule out primary major depression, as clutter and self-neglect may be caused by amotivation, low energy, or hopelessness.
  
• Determine if the patient has OCD.
  

After making a compulsive hoarding diagnosis (Table 1),⁶ visit the patient’s home or view photographs to assess his or her environment and behaviors (Table 2).

Amount of clutter. Living areas may be so cluttered that sleeping in a bed, sitting on chairs, or preparing food on a kitchen counter are impossible. How much of the home is cluttered? How much floor and counter space is usable? Are rooms unusable or inaccessible because of clutter? Can the patient use the laundry, prepare food in the kitchen, use the shower, toilet, etc.?

Health or safety hazards. Huge piles of papers can be a fire hazard. Clutter may be blocking the exits. Collected items may extend beyond patients’ homes to their cars, garages, storage lockers, and even storage areas owned by friends and family.

Beliefs about possessions. Compulsive hoarders often have distorted feelings about their possessions. They may over-buy or impulsively purchase items they feel have emotional or monetary value. They may consider the items extensions of themselves and suffer grief-like loss when discarding things.⁷

Some collect free items—flyers, coupons, newspapers, discarded goods—hoping to save money or be prepared “just in case” the item is ever needed. This may represent unattainable expectations of perfection, needing to maintain preparedness for every possible contingency. Hoarders often believe they have poor memory and have catastrophic fears of what might happen if they forget something. Thus, their desire to keep their possessions in sight is strong.¹⁷

 

Information processing deficits. Because of anxieties about making mistakes, most hoarders have great difficulty making decisions.¹⁸ It is easier to not decide than to suffer the consequences of a “wrong” decision. To gauge this behavior, ask patients how long routine decisions take them and which decisions they procrastinate or avoid.

Compulsive hoarders often have trouble categorizing possessions;^6,7 because every item feels unique, they create a special category for each one and resist storing items together.

Table 1

Proposed criteria to diagnose obsessive-compulsive hoarding*

Patient acquires and fails to discard a large number of possessions that appear useless or of limited value
Clutter prevents patient from using living or work spaces for activities for which they were designed
Hoarding behavior causes significant distress or functional impairment
* Proposed by Frost and Hartl, reference 6.

Many hoarders also report marked distractibility and inattention, jumping from one task to the next without completing any of them. Their communication style is often as cluttered and disorganized as their homes, with tangential, circumstantial, and over-inclusive descriptions.

Avoidance behaviors are a hallmark of the compulsive hoarding syndrome. To avoid deciding to discard items, they put them in a box, garage, rented storage facility, etc. They may also avoid routine decision-making tasks that could lead to making a mistake.

Daily functioning. Hoarders may take a long time to do even small chores, such as taking an hour to pay one bill. An inordinate amount of time may be spent “churning”—moving items from one pile to another but never discarding any item or establishing a consistent system or organization.

Medication compliance. Compulsive hoarders often forget to take medications or take them at inappropriate times. They may lose their medications in the clutter.

Insight. Hoarders often have little awareness of how their behavior and clutter affect their lives.¹⁹ They minimize the clutter in their homes and its health and safety risks. Insight can fluctuate over time and needs to be assessed repeatedly during treatment.

Table 2

Assessing a patient with compulsive hoarding symptoms

Domain	Useful questions or strategies
Amount of clutter	Visit home and/or see pictures
Hazards relating to clutter	Ask: What precautions do you take to reduce risk of fire? Have you ever had a problem with rodent or insect infestation as a result of the clutter? Have neighbors complained about the risks of fire or infestation that the clutter might impose on their homes?
Beliefs about loss of possessions	Ask: What is the worst thing that would happen if you threw this item away? If you did not have this, what do you think would happen?
Information-processing deficits	Ask: How long do routine decisions take you? Which decisions do you procrastinate or avoid?
Decision-making and organizational skills	Ask: How do you pay and store your bills?
Avoidance behaviors	Ask: Do you avoid other things (sorting mail, returning calls, doing dishes, or paying bills, rent, or taxes)?
Daily functioning	Ask: Do you get everything done that you want to do? Are you often late? Do you have difficulty starting or finishing tasks? Describe a typical day.
Insight	Ask: Do you think this amount of clutter is normal? Do you think having this clutter has caused problems in your life?
Motivation for treatment	Ask: What brings you into therapy now? Do you think you have a problem with excessive hoarding/saving? If it was not for your family, would you come for help?
Social and occupational functioning	Ask: How has your clutter affected your personal relationships? When was the last time you had someone come to your home? What prevents you from working right now? Are you working to your full potential?
Support from friends and family	Ask: What does your family say about your clutter? Do your friends or family understand what is going on?
Treatment compliance	Ask: How long does it typically take before you renew your prescriptions when you run out of medications?

Table 3

Cognitive behavioral therapy for compulsive hoarding

Treatment sequence	Methods and goals
Educate patient about hoarding	Help improve insight and motivation
Set up treatment	With patient, select target area of clutter
	Assess items together, creating a hierarchy of least to most difficult areas to sort and items to discard
	Create realistic categories and a storage system
Begin discarding	Patient must decide to keep or discard each item and permanently remove it from pile
	Patient must store saved items appropriately
	Continue until area is clear, then move to next area
Plan and implement appropriate use of space
Stop incoming clutter	Cancel subscriptions
	Address compulsive buying and acquisition
Provide organization training	Organize possessions, time, tasks, etc.
Prevent relapse	Replace hoarding with healthier behaviors to prevent clutter from re-accumulating
Source: Adapted from reference 23.

Social and occupational functioning. Many compulsive hoarders have very little family or social support. They frequently are too embarrassed by their clutter to have people come to their homes, sometimes for years. The syndrome frequently impairs work performance.

 

²⁰

Motivation. Like insight, motivation can fluctuate over time. Patients usually must work tremendously hard to adhere to treatment. To support these efforts, we periodically review with patients compulsive hoarding’s negative effects and the activities they would enjoy—such as improved relationships, greater work capacity, hobbies—if overcoming this behavior allowed them more time and space.

Rating scales. The symptom checklist of the Yale-Brown Obsessive-Compulsive Scale (YBOCS)²¹ contains two items for hoarding obsessions and compulsions but none for avoidance behaviors, which are prominent with compulsive hoarding. The Saving Inventory-Revised²² is a validated, 23-item self-report measure of clutter, difficulty discarding, and excessive acquisition, which distinguishes compulsive hoarders, nonhoarding OCD patients, and normal controls.

TREATMENT

The compulsive hoarder’s problems will not be solved by someone else throwing away or organizing his or her possessions. These actions often anger patients, who see them as intrusive and a loss of control.

In our experience, family members’ attempts to intervene can disrupt relationships and worsen hoarders’ social withdrawal. “Taking over” also does not help the patient create a sustainable system for keeping clutter-free.

Algorithm

Medication treatment for compulsive hoarding*

Start with SSRIs, as for nonhoarding OCD (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline)	High doses and 12-week trials Some compulsive hoarders will respond well to SSRIs Other OCD symptoms usually improve as well Comorbid depression and other anxiety symptoms may respond If ineffective, may need to do 3 or 4 full trials of different SSRIs, clomipramine, or venlafaxine
Treat comorbid conditions	Mood disorders, other anxiety disorders, ADHD, psychotic disorders, etc.
Use adjunctive medications if SSRIs give only partial response	Atypical antipsychotics (risperidone, olanzapine, quetiapine) Stimulants Mood stabilizers (for comorbid bipolar disorder, cyclothymia, or impulsivity)
* Combine medication treatment with cognitive-behavioral therapy
SSRI: selective serotonin reuptake inhibitor
OCD: obsessive-compulsive disorder
ADHD: attention-deficit/hyperactivity disorder

We find that combining cognitive-behavioral therapy (CBT) and medication is optimal treatment for compulsive hoarding,²³ although no controlled studies have compared this combination with each treatment alone. One controlled trial²⁴ and three uncontrolled trials^8,25,26 have shown some benefit of CBT for compulsive hoarding, although with poorer response and higher dropout rates than for nonhoarding OCD patients.

Psychotherapy. Exposure and response prevention (ERP) focuses on preventing incoming clutter, discarding, organizing, and relapse prevention (Table 3).²³ Start treatment by explaining compulsive hoarding syndrome to patients as having problems with information processing, obsessional anxiety, and avoiding decisions.

Preventing incoming clutter. Before you focus on discarding, patients must stop incoming clutter; otherwise, it will come in as fast as it goes out. We ask patients to keep a daily log of every item they acquire or buy to build their awareness of what triggers their behavior.

Discarding. To desensitize over time, we repeatedly expose patients to the anxiety, sadness, or anger they feel when discarding items and making decisions. We encourage them to provoke their anxiety by throwing away as many items as possible, keeping only necessary items.

We support ERP with cognitive restructuring, prompting patients to reframe their obsessive fears about losing something necessary or valuable. By thinking through the consequences of discarding their clutter, they challenge their erroneous beliefs that dire consequences will occur.

Organizing. When patients decide they must keep an item, ask them to immediately identify a specific place and deadline to store it. After an area is cleared, patients must keep it clear and use it for its intended purpose. Most patients need training in time management, scheduling, and prioritizing.

Relapse prevention. Replace hoarding behaviors with more-adaptive, healthy behaviors. Teach patients to create a realistic schedule that includes time for chores, eating and sleeping, CBT homework, and recreation. Treatment goals are to:

• extinguish obsessional fears and compulsive saving behaviors
  
• teach lasting organizational and decision-making skills, thereby reducing relapse risk.
  

Practical matters. Sorting and discarding a houseful of clutter takes time, although the patient does this primarily as homework. To control costs and your time commitment, consider collaborating with a CBT therapist trained in working with compulsive hoarders. Use your sessions with the patient to create hierarchies, go over assignments, do brief exposures, and monitor drug therapy.

Medications. No controlled studies have examined whether any medications are effective for compulsive hoarding syndrome. The treatment strategies and algorithm described here are based on our clinical experience, controlled trials of OCD patients, and limited OCD studies secondarily examining hoarders’ specific treatment responses.

Selective serotonin reuptake inhibitors (SSRIs) may be less effective for compulsive hoarding than for other compulsive behaviors.^14,27 Nevertheless, SSRIs may help alleviate hoarders’ core symptoms, other OCD symptoms, depression, and anxiety. For hoarding treatment to be effective, comorbid disorders must be treated and stabilized.

 

Several studies of SSRI use in OCD patients have shown modest improvements in compulsive hoarders:

• In a descriptive study of patients with compulsive hoarding, 1 of 18 (6%) patients had a “marked” response to at least one SSRI trial. The others showed a partial response,¹⁷ with YBOCS scores decreasing by at least 25% in approximately 50% of this group.
  
• When 17 OCD patients with hoarding symptoms were treated with paroxetine, CBT, or placebo, 18% responded to active treatment. Response was defined as a 40% reduction on YBOCS scores and “very much” or “much” improved on the Clinical Global Impression Scale (CGI).¹⁴
  

Based on these findings and our clinical experience, an effective approach to treating compulsive hoarding with medications—as with other OCD patients—is to start with SSRIs²⁸ (Algorithm). If adequate SSRI trials fail to improve a patient’s hoarding/saving symptoms, adjunctive medications can be added.

Atypical antipsychotics may be effective for OCD symptoms that do not respond adequately to SSRIs.²⁹ Conventional antipsychotics are also effective adjuncts to SSRIs—particularly for patients with coexisting tic or psychotic disorders³⁰—but consider the potential for extrapyramidal side effects and tardive dyskinesia.

We find that stimulants help some compulsive hoarders, particularly those with comorbid ADHD, other attentional problems, low motivation, or lethargy. Mood stabilizers are necessary to treat comorbid bipolar disorder, cyclothymia, and impulsivity.

Related resources

• Saxena S, Maidment K. Treatment of compulsive hoarding. J Clin Psychol 2004;60(11):1143-54.
  
• UCLA OCD Intensive Treatment Program. For information on research studies contact Karron Maidment (310) 794-7305 or visit www.mentalhealth.ucla.edu/projects/anxiety/ocdintensivetreatment.htm.
  
• Compulsive hoarding project. Institute of Living Anxiety Disorders Center, Hartford Hospital. http://instituteofliving.org/adc/compulsive_hoarding.htm
  
• Obsessive Compulsive Foundation compulsive hoarding Web site. http://www.ocfoundation.org/1005/index.html
  
• Neziroglu F, Bubrick J, Yaryura-Tobias JA. Overcoming compulsive hoarding. Oakland, CA: New Harbinger Publications, 2004.
  

Drug brand names

• Citalopram • Celexa
  
• Clomipramine • Anafranil
  
• Escitalopram • Lexapro
  
• Fluoxetine • Prozac
  
• Fluvoxamine • Luvox
  
• Olanzapine • Zyprexa
  
• Paroxetine • Paxil
  
• Quetiapine • Seroquel
  
• Risperidone • Risperdal
  
• Sertraline • Zoloft
  
• Venlafaxine • Effexor
  

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Compulsive hoarding behavior is considered notoriously difficult to treat, but targeting its characteristic symptoms with medication and psychotherapy can be successful. This article provides a guide for the psychiatrist—alone or with a cognitive-behavioral therapist—to diagnose compulsive hoarding syndrome and help patients overcome the anxieties that fuel its symptoms.

WHAT IS COMPULSIVE HOARDING?

Hoarders acquire and are unable to discard items that others consider of little use or value.¹ They most often save newspapers, magazines, old clothing, bags, books, mail, notes, and lists. Hoarding and saving behaviors occur in nonclinical populations and with other neuropsychiatric disorders—schizophrenia, dementia, eating disorders, mental retardation—but are most often found in persons with obsessive-compulsive disorder (OCD).

OCD is a heterogeneous clinical entity with several major symptom domains:^2,3

• aggressive, sexual, and religious obsessions with checking compulsions
  
• symmetry/order obsessions with ordering, arranging, and repeating compulsions
  
• contamination obsessions with washing and cleaning compulsions
  
• hoarding and saving symptoms.
  

Box

Among OCD patients, 18% to 42% have hoarding and saving compulsions.^4,5 Hoarding and saving can be part of a broader clinical syndrome that includes indecisiveness, perfectionism, procrastination, difficulty organizing tasks, and avoiding routine daily activities.^6,7 The 1 to 2 million Americans whose most prominent and distressing OCD symptom is hoarding and saving and who show these other associated symptoms are considered to have “compulsive hoarding syndrome.”^7,8 Evidence suggests that this syndrome may be a neurobiologically distinct OCD variant (Box).^9-16

ASSESSMENT AND TREATMENT PLANNING

To manage compulsive hoarding syndrome, begin with a thorough neuropsychiatric evaluation:

• Rule out primary psychotic disorders, dementia, and other cognitive impairments and neurologic disorders.
  
• Rule out primary major depression, as clutter and self-neglect may be caused by amotivation, low energy, or hopelessness.
  
• Determine if the patient has OCD.
  

After making a compulsive hoarding diagnosis (Table 1),⁶ visit the patient’s home or view photographs to assess his or her environment and behaviors (Table 2).

Amount of clutter. Living areas may be so cluttered that sleeping in a bed, sitting on chairs, or preparing food on a kitchen counter are impossible. How much of the home is cluttered? How much floor and counter space is usable? Are rooms unusable or inaccessible because of clutter? Can the patient use the laundry, prepare food in the kitchen, use the shower, toilet, etc.?

Health or safety hazards. Huge piles of papers can be a fire hazard. Clutter may be blocking the exits. Collected items may extend beyond patients’ homes to their cars, garages, storage lockers, and even storage areas owned by friends and family.

Beliefs about possessions. Compulsive hoarders often have distorted feelings about their possessions. They may over-buy or impulsively purchase items they feel have emotional or monetary value. They may consider the items extensions of themselves and suffer grief-like loss when discarding things.⁷

Some collect free items—flyers, coupons, newspapers, discarded goods—hoping to save money or be prepared “just in case” the item is ever needed. This may represent unattainable expectations of perfection, needing to maintain preparedness for every possible contingency. Hoarders often believe they have poor memory and have catastrophic fears of what might happen if they forget something. Thus, their desire to keep their possessions in sight is strong.¹⁷

 

Information processing deficits. Because of anxieties about making mistakes, most hoarders have great difficulty making decisions.¹⁸ It is easier to not decide than to suffer the consequences of a “wrong” decision. To gauge this behavior, ask patients how long routine decisions take them and which decisions they procrastinate or avoid.

Compulsive hoarders often have trouble categorizing possessions;^6,7 because every item feels unique, they create a special category for each one and resist storing items together.

Table 1

Proposed criteria to diagnose obsessive-compulsive hoarding*

Patient acquires and fails to discard a large number of possessions that appear useless or of limited value
Clutter prevents patient from using living or work spaces for activities for which they were designed
Hoarding behavior causes significant distress or functional impairment
* Proposed by Frost and Hartl, reference 6.

Many hoarders also report marked distractibility and inattention, jumping from one task to the next without completing any of them. Their communication style is often as cluttered and disorganized as their homes, with tangential, circumstantial, and over-inclusive descriptions.

Avoidance behaviors are a hallmark of the compulsive hoarding syndrome. To avoid deciding to discard items, they put them in a box, garage, rented storage facility, etc. They may also avoid routine decision-making tasks that could lead to making a mistake.

Daily functioning. Hoarders may take a long time to do even small chores, such as taking an hour to pay one bill. An inordinate amount of time may be spent “churning”—moving items from one pile to another but never discarding any item or establishing a consistent system or organization.

Medication compliance. Compulsive hoarders often forget to take medications or take them at inappropriate times. They may lose their medications in the clutter.

Insight. Hoarders often have little awareness of how their behavior and clutter affect their lives.¹⁹ They minimize the clutter in their homes and its health and safety risks. Insight can fluctuate over time and needs to be assessed repeatedly during treatment.

Table 2

Assessing a patient with compulsive hoarding symptoms

Domain	Useful questions or strategies
Amount of clutter	Visit home and/or see pictures
Hazards relating to clutter	Ask: What precautions do you take to reduce risk of fire? Have you ever had a problem with rodent or insect infestation as a result of the clutter? Have neighbors complained about the risks of fire or infestation that the clutter might impose on their homes?
Beliefs about loss of possessions	Ask: What is the worst thing that would happen if you threw this item away? If you did not have this, what do you think would happen?
Information-processing deficits	Ask: How long do routine decisions take you? Which decisions do you procrastinate or avoid?
Decision-making and organizational skills	Ask: How do you pay and store your bills?
Avoidance behaviors	Ask: Do you avoid other things (sorting mail, returning calls, doing dishes, or paying bills, rent, or taxes)?
Daily functioning	Ask: Do you get everything done that you want to do? Are you often late? Do you have difficulty starting or finishing tasks? Describe a typical day.
Insight	Ask: Do you think this amount of clutter is normal? Do you think having this clutter has caused problems in your life?
Motivation for treatment	Ask: What brings you into therapy now? Do you think you have a problem with excessive hoarding/saving? If it was not for your family, would you come for help?
Social and occupational functioning	Ask: How has your clutter affected your personal relationships? When was the last time you had someone come to your home? What prevents you from working right now? Are you working to your full potential?
Support from friends and family	Ask: What does your family say about your clutter? Do your friends or family understand what is going on?
Treatment compliance	Ask: How long does it typically take before you renew your prescriptions when you run out of medications?

Table 3

Cognitive behavioral therapy for compulsive hoarding

Treatment sequence	Methods and goals
Educate patient about hoarding	Help improve insight and motivation
Set up treatment	With patient, select target area of clutter
	Assess items together, creating a hierarchy of least to most difficult areas to sort and items to discard
	Create realistic categories and a storage system
Begin discarding	Patient must decide to keep or discard each item and permanently remove it from pile
	Patient must store saved items appropriately
	Continue until area is clear, then move to next area
Plan and implement appropriate use of space
Stop incoming clutter	Cancel subscriptions
	Address compulsive buying and acquisition
Provide organization training	Organize possessions, time, tasks, etc.
Prevent relapse	Replace hoarding with healthier behaviors to prevent clutter from re-accumulating
Source: Adapted from reference 23.

Social and occupational functioning. Many compulsive hoarders have very little family or social support. They frequently are too embarrassed by their clutter to have people come to their homes, sometimes for years. The syndrome frequently impairs work performance.

 

²⁰

Motivation. Like insight, motivation can fluctuate over time. Patients usually must work tremendously hard to adhere to treatment. To support these efforts, we periodically review with patients compulsive hoarding’s negative effects and the activities they would enjoy—such as improved relationships, greater work capacity, hobbies—if overcoming this behavior allowed them more time and space.

Rating scales. The symptom checklist of the Yale-Brown Obsessive-Compulsive Scale (YBOCS)²¹ contains two items for hoarding obsessions and compulsions but none for avoidance behaviors, which are prominent with compulsive hoarding. The Saving Inventory-Revised²² is a validated, 23-item self-report measure of clutter, difficulty discarding, and excessive acquisition, which distinguishes compulsive hoarders, nonhoarding OCD patients, and normal controls.

TREATMENT

The compulsive hoarder’s problems will not be solved by someone else throwing away or organizing his or her possessions. These actions often anger patients, who see them as intrusive and a loss of control.

In our experience, family members’ attempts to intervene can disrupt relationships and worsen hoarders’ social withdrawal. “Taking over” also does not help the patient create a sustainable system for keeping clutter-free.

Algorithm

Medication treatment for compulsive hoarding*

Start with SSRIs, as for nonhoarding OCD (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline)	High doses and 12-week trials Some compulsive hoarders will respond well to SSRIs Other OCD symptoms usually improve as well Comorbid depression and other anxiety symptoms may respond If ineffective, may need to do 3 or 4 full trials of different SSRIs, clomipramine, or venlafaxine
Treat comorbid conditions	Mood disorders, other anxiety disorders, ADHD, psychotic disorders, etc.
Use adjunctive medications if SSRIs give only partial response	Atypical antipsychotics (risperidone, olanzapine, quetiapine) Stimulants Mood stabilizers (for comorbid bipolar disorder, cyclothymia, or impulsivity)
* Combine medication treatment with cognitive-behavioral therapy
SSRI: selective serotonin reuptake inhibitor
OCD: obsessive-compulsive disorder
ADHD: attention-deficit/hyperactivity disorder

We find that combining cognitive-behavioral therapy (CBT) and medication is optimal treatment for compulsive hoarding,²³ although no controlled studies have compared this combination with each treatment alone. One controlled trial²⁴ and three uncontrolled trials^8,25,26 have shown some benefit of CBT for compulsive hoarding, although with poorer response and higher dropout rates than for nonhoarding OCD patients.

Psychotherapy. Exposure and response prevention (ERP) focuses on preventing incoming clutter, discarding, organizing, and relapse prevention (Table 3).²³ Start treatment by explaining compulsive hoarding syndrome to patients as having problems with information processing, obsessional anxiety, and avoiding decisions.

Preventing incoming clutter. Before you focus on discarding, patients must stop incoming clutter; otherwise, it will come in as fast as it goes out. We ask patients to keep a daily log of every item they acquire or buy to build their awareness of what triggers their behavior.

Discarding. To desensitize over time, we repeatedly expose patients to the anxiety, sadness, or anger they feel when discarding items and making decisions. We encourage them to provoke their anxiety by throwing away as many items as possible, keeping only necessary items.

We support ERP with cognitive restructuring, prompting patients to reframe their obsessive fears about losing something necessary or valuable. By thinking through the consequences of discarding their clutter, they challenge their erroneous beliefs that dire consequences will occur.

Organizing. When patients decide they must keep an item, ask them to immediately identify a specific place and deadline to store it. After an area is cleared, patients must keep it clear and use it for its intended purpose. Most patients need training in time management, scheduling, and prioritizing.

Relapse prevention. Replace hoarding behaviors with more-adaptive, healthy behaviors. Teach patients to create a realistic schedule that includes time for chores, eating and sleeping, CBT homework, and recreation. Treatment goals are to:

• extinguish obsessional fears and compulsive saving behaviors
  
• teach lasting organizational and decision-making skills, thereby reducing relapse risk.
  

Practical matters. Sorting and discarding a houseful of clutter takes time, although the patient does this primarily as homework. To control costs and your time commitment, consider collaborating with a CBT therapist trained in working with compulsive hoarders. Use your sessions with the patient to create hierarchies, go over assignments, do brief exposures, and monitor drug therapy.

Medications. No controlled studies have examined whether any medications are effective for compulsive hoarding syndrome. The treatment strategies and algorithm described here are based on our clinical experience, controlled trials of OCD patients, and limited OCD studies secondarily examining hoarders’ specific treatment responses.

Selective serotonin reuptake inhibitors (SSRIs) may be less effective for compulsive hoarding than for other compulsive behaviors.^14,27 Nevertheless, SSRIs may help alleviate hoarders’ core symptoms, other OCD symptoms, depression, and anxiety. For hoarding treatment to be effective, comorbid disorders must be treated and stabilized.

 

Several studies of SSRI use in OCD patients have shown modest improvements in compulsive hoarders:

• In a descriptive study of patients with compulsive hoarding, 1 of 18 (6%) patients had a “marked” response to at least one SSRI trial. The others showed a partial response,¹⁷ with YBOCS scores decreasing by at least 25% in approximately 50% of this group.
  
• When 17 OCD patients with hoarding symptoms were treated with paroxetine, CBT, or placebo, 18% responded to active treatment. Response was defined as a 40% reduction on YBOCS scores and “very much” or “much” improved on the Clinical Global Impression Scale (CGI).¹⁴
  

Based on these findings and our clinical experience, an effective approach to treating compulsive hoarding with medications—as with other OCD patients—is to start with SSRIs²⁸ (Algorithm). If adequate SSRI trials fail to improve a patient’s hoarding/saving symptoms, adjunctive medications can be added.

Atypical antipsychotics may be effective for OCD symptoms that do not respond adequately to SSRIs.²⁹ Conventional antipsychotics are also effective adjuncts to SSRIs—particularly for patients with coexisting tic or psychotic disorders³⁰—but consider the potential for extrapyramidal side effects and tardive dyskinesia.

We find that stimulants help some compulsive hoarders, particularly those with comorbid ADHD, other attentional problems, low motivation, or lethargy. Mood stabilizers are necessary to treat comorbid bipolar disorder, cyclothymia, and impulsivity.

Related resources

• Saxena S, Maidment K. Treatment of compulsive hoarding. J Clin Psychol 2004;60(11):1143-54.
  
• UCLA OCD Intensive Treatment Program. For information on research studies contact Karron Maidment (310) 794-7305 or visit www.mentalhealth.ucla.edu/projects/anxiety/ocdintensivetreatment.htm.
  
• Compulsive hoarding project. Institute of Living Anxiety Disorders Center, Hartford Hospital. http://instituteofliving.org/adc/compulsive_hoarding.htm
  
• Obsessive Compulsive Foundation compulsive hoarding Web site. http://www.ocfoundation.org/1005/index.html
  
• Neziroglu F, Bubrick J, Yaryura-Tobias JA. Overcoming compulsive hoarding. Oakland, CA: New Harbinger Publications, 2004.
  

Drug brand names

• Citalopram • Celexa
  
• Clomipramine • Anafranil
  
• Escitalopram • Lexapro
  
• Fluoxetine • Prozac
  
• Fluvoxamine • Luvox
  
• Olanzapine • Zyprexa
  
• Paroxetine • Paxil
  
• Quetiapine • Seroquel
  
• Risperidone • Risperdal
  
• Sertraline • Zoloft
  
• Venlafaxine • Effexor
  

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Women with depression and polycystic ovary syndrome (PCOS) can be trapped in a vicious cycle of hormonal dysregulation. Treating these patients appropriately—with or without antidepressants—requires an understanding of their underlying metabolic disorder.

This article presents a case¹ that exemplifies the association between depression and PCOS (Box 1).^2-4 Based on our research and clinical experience, we offer recommendations to help you manage depression in patients with PCOS.

CHRONIC DEPRESSION WITH AMENORRHEA

Ms. K, age 30, presented for psychiatric evaluation of treatment-resistant recurrent major depression. Her symptoms included sad mood, sleep disturbance, decreased energy, anhedonia, poor concentration, and feelings of guilt and worthlessness. Her score of 28 on the Hamilton Rating Scale for Depression (HAM-D-21) indicated severe depression.

Box 1

This was Ms. K’s third episode of major depression, with the first two occurring at ages 24 and 26. She reported similar symptoms each time. Previous psychiatrists had tried a variety of antidepressants with no therapeutic benefit before she responded to citalopram, 40 mg/d, at age 27. With this selective serotonin reuptake inhibitor, her depressive symptoms resolved for >1 year, except for mild irritability and low mood the week before menses.

Her depressive symptoms returned, however, and her previous psychiatrist treated her for approximately 13 months with extended-release venlafaxine, 225 mg/d. When she remained depressed, she was referred to our clinic for evaluation. Based on her history, we could not determine whether she had become euthymic at age 27 or her symptoms had improved but still met criteria for a major depressive episode.

Ms. K reported having PCOS symptoms from age 19. These included amenorrhea since menarche, hirsutism, hair loss, alopecia, central fat distribution characteristic of PCOS, and male-pattern hair growth on her abdomen and thighs. She was not obese—with a body mass index (BMI) of approximately 25—but had gained 10 to 15 lbs while taking venlafaxine.

Ms. K had never been treated for PCOS but reported that her desire to become pregnant made her more concerned about her symptoms and possible infertility.

Diagnosis. PCOS is usually diagnosed using endocrinologic, clinical, and ultrasonographic criteria (Table 1). Obesity is not a presenting symptom in all women with PCOS. As with Ms. K, about 50% of patients have normal BMI.

Causes of depression in PCOS. Depressed mood in PCOS may be both physiologic and psychological:

• Hypothalamic, pituitary, and other end-organ system dysregulation occurs in both PCOS and affective disorders, which share clinical and biochemical markers including insulin resistance, obesity, and hyperandrogenism.
• PCOS’ clinical sequelae—hirsutism, acne, obesity, hormonal disturbances, fear of infertility, and psychological distress—may damage their self-esteem and female identity.

PCOS’ physical symptoms alone apparently do not account for patients’ worsened mood states. Weiner et al⁵ found that women with PCOS and free testosterone (FT) of 10 to 26 pg/mL (just above normal range) were more depressed than women without PCOS (FT <10 pg/mL) and women with PCOS and FT >26 pg/mL. Women with PCOS with the lower and higher FT levels had similar demographic profiles, but those with the highest FT levels were not the most depressed.

Similarly, in a study of 32 women with PCOS, we found no association between depression and other possibly distressing PCOS symptoms, including hirsutism, irregular menses, acne, or alopecia.⁴

Testosterone and mood disorders. Women such as Ms. K with hyperandrogenic syndromes are at increased risk for mood disorders.⁶ Many metabolic changes associated with PCOS—insulin resistance, obesity, and hyperandrogenism—are also described in patients with affective disorders. Our investigation of reproductive status in women with bipolar disorder found no clinical or biochemical evidence of PCOS with mood-stabilizer treatment. We did find that menstrual disturbances were common, however, and sometimes preceded bipolar disorder onset.⁷

Insulin resistance and depression. Others have linked insulin resistance and depression;⁸ depression has been shown to be associated with impaired insulin sensitivity and hyperinsulinemia.⁹ Depressed persons also tend to eat more sweets, drink more alcohol, exercise less, and sleep fewer hours than the nondepressed—all of which contribute to insulin sensitivity and insulin resistance.¹⁰

 

WHAT LINKS ENDOCRINE, MOOD DISORDERS?

Insulin and serotonin. Insulin affects central serotonin (5-HT) levels.¹¹ Thus, central 5-HT system dysregulation that causes depression might also affect peripheral insulin sensitivity—or vice versa.⁹

Tryptophan, a serotonin precursor, competes with other large neutral amino acids for access to the transport system that moves them across the blood-brain barrier. Insulin stimulates uptake of the competing amino acids—but not tryptophan—into muscle tissue. The resulting increased tryptophan ratio in plasma affords it greater access to the transport system to contribute toward serotonin synthesis. Insulin also promotes central catecholaminergic activity, perhaps by suppressing norepinephrine reuptake and prolonging its residence in the synaptic cleft.¹⁰

Table 1

Diagnostic signs of PCOS

Endocrine
Chronically elevated plasma free and total testosterone levels
Increased luteinizing hormone (LH) secretion because of enhanced pituitary sensitivity to GnRH stimulation
Low or normal plasma follicle-stimulating hormone (FSH) levels
Hyperinsulinemia and peripheral insulin resistance
Metabolic abnormalities of hyperinsulinemia and peripheral insulin resistance, including obesity and increased susceptibility to type 2 diabetes and cardiovascular disease
Clinical
Hirsutism
Acne
Infertility
Anovulation or menstrual abnormalities amenorrhea oligomenorrhea dysfunctional uterine bleeding
Morphologic
Ovaries may appear to have multiple cysts 8 to 10 mm in diameter

Several studies have found that insulin resistance and hyperinsulinemia can resolve after depression recovery.^9,12 Because insulin resistance is a cardinal feature in PCOS pathophysiology,² insulin resistance may be a common link between depression and PCOS.

Weight gain and obesity also have been described in patients with affective disorders.¹³ Not known is whether the weight gain precedes the psychopathology or is caused by long-term exposure to drugs used to treat affective disorders.

Box 2

We reported a correlation between depression and BMI in women with PCOS.⁴ Depression might be independently associated with BMI, as weight gain and obesity are distressing symptoms associated with depression.¹⁴ However, we found no association between depression and other possibly distressing PCOS symptoms. Thus, the correlation between BMI and depression might more likely reflect the relationship between depression and insulin resistance, as degree of insulin resistance is known to correlate with BMI.

Elevated cortisol. Clearly, other factors—such as hypothalamus-pituitary-adrenal (HPA) axis dysfunction—are known to link affective and endocrine disorders. Hypercortisolemia can lead to both insulin resistance and obesity. Cortisol is one of the glucocorticoids the body secretes in response to stress to mobilize energy by increasing blood glucose levels. Early life stress and chronic emotional stress:

• can impair the negative feedback system that limits cortisol production during stress
• are associated with depression.

Approximately one-half of individuals with depression have elevated serum cortisol.¹⁰ Epidemiologic data show a positive correlation between cortisol levels and insulin resistance,¹⁵ and an association between HPA dysfunction and obesity has been described.¹⁶ Insulin can trigger androgen production by enhancing adrenal sensitivity to adrenocorticotrophic hormone (ACTH).¹⁷

CASE: IMPROVING INSULIN RESISTANCE

We referred Ms. K to an endocrinologist for PCOS evaluation and treatment. Her serum glucose and insulin levels were 83 mg/dL (normal range 70 to 125 mg/dL) and 19.0 uIU/mL (normal range <10 uIU/mL), respectively. These values indicated insulin resistance as determined by the homeostasis model assessment (HOMA) ratio (fasting insulin x fasting glucose/22.5). Values >3.2 indicate insulin resistance, and Ms. K had a HOMA ratio of 3.9. The endocrinologist recommended:

• metformin, starting at 850 mg/d and gradually increased to 2,550 mg/d
• spironolactone, 100 mg/d.

Metformin, a biguanide approved for treating for type 2 diabetes, inhibits hepatic glucose production and increases peripheral insulin sensitivity, but it does not modify pancreatic insulin secretion. It may decrease insulin resistance by reducing gut absorption of glucose, improving glucose uptake by tissues, and/or increasing the number of insulin receptors.¹⁸ In treating PCOS, metformin can:

 

• restore ovulation¹⁹
• decrease insulin resistance, acne, hirsutism, total and bioavailable testosterone, BMI, and waist-hip ratio.^20,21

Although the link between insulin resistance and depression is unclear, insulin is known to contribute to 5-HT synthesis by promoting tryptophan influx into the brain.²² Therefore, drugs used to treat insulin resistance—such as metformin and alpha lipoic acid²³—might be useful in treating depression.

Spironolactone, a mineralocorticoid receptor (MR) antagonist, reduces hirsutism in women with PCOS.²⁴ It also can decrease insulin resistance and fasting insulin levels in PCOS patients and reduce serum testosterone.²⁵

Evidence on treating mood disorders with hormonal agents such as spironolactone is scarce, although treatment-resistant depression has been reported to resolve with antiglucocorticoid use (Box 2).^25-31 Modulating HPA axis activity to treat affective disorders has been investigated.

CASE: GOING ANTIDEPRESSANT-FREE

At first, Ms. K said she wanted to continue taking venlafaxine with the PCOS treatment. After 2 weeks of combined therapy, however, she chose to stop the antidepressant after her depressive symptoms persisted, and her HAM-D-21 score remained at 28.

During the next 4 weeks, as we tapered off the venlafaxine, Ms. K’s HAM-D-21 score dropped to 7, indicating depressive symptom resolution. Despite slow venlafaxine titration, withdrawal symptoms of excessive crying, not feeling “present,” and tingling sensations occurred. Three days of fluoxetine, 20 mg/d, alleviated these symptoms.

Table 2

Insulin-sensitizing medications used to treat PCOS

Medication	Normal dosage	Common side effects
Metformin	500 mg tid	Headache, GI effects (nausea, diarrhea, flatulence) at start of therapy, weight loss, taste disturbances
Pioglitazone	15 to 45 mg once daily	Swelling, headache, respiratory infection, abdominal discomfort, muscle soreness
Rosiglitazone	4 to 8 mg once daily	Headache, mild weight gain

We continued Ms. K’s treatment without antidepressants, and her mood continued to improve with metformin, 2,550 mg/d, and spironolactone, 100 mg/d.

TREATMENT RECOMMENDATIONS

PCOS therapy may take up to 6 months to resolve symptoms such as anovulation or hirsutism, but affective symptoms may improve during the first 6 weeks, as this case shows. Choosing medications to treat depression in patients such as Ms. K depends on whether their PCOS is being treated when they present for psychiatric evaluation.

Patient not being treated for PCOS. Refer her to an endocrinologist for PCOS treatment with an insulin-sensitizing medication, such as metformin (Table 2). Treating the insulin resistance associated with PCOS may also resolve the depression. PCOS drug therapy may also include antiandrogens such as spironolactone to treat hirsutism, male-pattern baldness, and acne. We suggest that spironolactone’s antiandrogen effects may help reduce depressive symptoms.

If your patient is not taking an antidepressant at presentation, try PCOS treatment first. If depressive symptoms persist after 3 months, consider adding an antidepressant. If your patient is taking an antidepressant at presentation but continues to be depressed, offer two options:

• taper off the antidepressant while starting PCOS treatment
• continue the antidepressant while starting PCOS treatment.

The first option allows you to try PCOS treatment alone. If the depression is caused by a common underlying pathophysiology—such as insulin resistance—treating PCOS alone may alleviate her depressive symptoms. Monitor for withdrawal symptoms, which may be minimized with a short-term SSRI, such as fluoxetine.

The second option may help patients who have responded to antidepressants in the past. Adjunctive PCOS treatment may “jump-start” the antidepressant response without withdrawal symptoms, but you will not be sure whether the antidepressant response was caused by PCOS therapy alone or the combination therapy.

Patient being treated for PCOS. Treat her with antidepressants, and consult with her endocrinologist to consider more-aggressive insulin-sensitizing medications, especially if she exhibits high levels of insulin resistance.

Other interventions that increase insulin sensitivity and improve glycemic control—such as improving dietary management and sleep habits, reducing alcohol consumption, and increasing physical activity—might have an antidepressant effect. Therefore, recommend these health practices to patients as adjuncts to drug therapies.

CASE: DEPRESSION IN REMISSION

At the 3-month follow-up visit, Ms. K scored zero on the HAM-D-21 scale. With metformin treatment, she had lost approximately 10 lbs and resumed menstruating approximately every 33 days. She reported experiencing low mood, decreased energy, and irritability during the week before her periods, but these symptoms resolved with menses onset.

Serum glucose was within normal range at 89 mg/dL, and serum insulin was 15.0 uIU/Ml. Her HOMA ratio had dropped to 2.8 (below the 3.2 cut-off for insulin resistance).

Ms. K’s endocrinologist monitored her spironolactone and metformin therapy for approximately 1 year, when she became pregnant.

Discussion. PCOS treatment duration depends on the patient’s response and her goals for therapy. Whether or not she continues PCOS treatment, her primary care physician or endocrinologist should continue to monitor her for insulin resistance’s metabolic consequences, including increased risk of type 2 diabetes and cardiovascular disease.

 

Related resources

• Legro RS, Winans EA. Overview of polycystic ovary syndrome and its relation to psychiatric illness and treatment. Current Psychiatry 2004;3(Dec[suppl]):12-20.
• The Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004 Jan; 81(1):19-25.
• Polycystic ovary syndrome (patient information). The National Women’s Health Information Center. U.S. Department of Health and Human Services, Office on Women’s Health. Available at: http://www.4woman.gov/faq/pcos.htm. Accessed Jan. 13, 2005.

Drug brand names

• Citalopram • Celexa
• Fluoxetine • Prozac
• Ketoconazole • Nizoral
• Metformin • Glucophage
• Metyrapone • Metopirone
• Mifepristone • Mifiprex
• Pioglitazone • Actos
• Rosiglitazone • Avandia
• Spironolactone • Aldactone
• Venlafaxine • Effexor XR

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Women with depression and polycystic ovary syndrome (PCOS) can be trapped in a vicious cycle of hormonal dysregulation. Treating these patients appropriately—with or without antidepressants—requires an understanding of their underlying metabolic disorder.

This article presents a case¹ that exemplifies the association between depression and PCOS (Box 1).^2-4 Based on our research and clinical experience, we offer recommendations to help you manage depression in patients with PCOS.

CHRONIC DEPRESSION WITH AMENORRHEA

Ms. K, age 30, presented for psychiatric evaluation of treatment-resistant recurrent major depression. Her symptoms included sad mood, sleep disturbance, decreased energy, anhedonia, poor concentration, and feelings of guilt and worthlessness. Her score of 28 on the Hamilton Rating Scale for Depression (HAM-D-21) indicated severe depression.

Box 1

This was Ms. K’s third episode of major depression, with the first two occurring at ages 24 and 26. She reported similar symptoms each time. Previous psychiatrists had tried a variety of antidepressants with no therapeutic benefit before she responded to citalopram, 40 mg/d, at age 27. With this selective serotonin reuptake inhibitor, her depressive symptoms resolved for >1 year, except for mild irritability and low mood the week before menses.

Her depressive symptoms returned, however, and her previous psychiatrist treated her for approximately 13 months with extended-release venlafaxine, 225 mg/d. When she remained depressed, she was referred to our clinic for evaluation. Based on her history, we could not determine whether she had become euthymic at age 27 or her symptoms had improved but still met criteria for a major depressive episode.

Ms. K reported having PCOS symptoms from age 19. These included amenorrhea since menarche, hirsutism, hair loss, alopecia, central fat distribution characteristic of PCOS, and male-pattern hair growth on her abdomen and thighs. She was not obese—with a body mass index (BMI) of approximately 25—but had gained 10 to 15 lbs while taking venlafaxine.

Ms. K had never been treated for PCOS but reported that her desire to become pregnant made her more concerned about her symptoms and possible infertility.

Diagnosis. PCOS is usually diagnosed using endocrinologic, clinical, and ultrasonographic criteria (Table 1). Obesity is not a presenting symptom in all women with PCOS. As with Ms. K, about 50% of patients have normal BMI.

Causes of depression in PCOS. Depressed mood in PCOS may be both physiologic and psychological:

• Hypothalamic, pituitary, and other end-organ system dysregulation occurs in both PCOS and affective disorders, which share clinical and biochemical markers including insulin resistance, obesity, and hyperandrogenism.
• PCOS’ clinical sequelae—hirsutism, acne, obesity, hormonal disturbances, fear of infertility, and psychological distress—may damage their self-esteem and female identity.

PCOS’ physical symptoms alone apparently do not account for patients’ worsened mood states. Weiner et al⁵ found that women with PCOS and free testosterone (FT) of 10 to 26 pg/mL (just above normal range) were more depressed than women without PCOS (FT <10 pg/mL) and women with PCOS and FT >26 pg/mL. Women with PCOS with the lower and higher FT levels had similar demographic profiles, but those with the highest FT levels were not the most depressed.

Similarly, in a study of 32 women with PCOS, we found no association between depression and other possibly distressing PCOS symptoms, including hirsutism, irregular menses, acne, or alopecia.⁴

Testosterone and mood disorders. Women such as Ms. K with hyperandrogenic syndromes are at increased risk for mood disorders.⁶ Many metabolic changes associated with PCOS—insulin resistance, obesity, and hyperandrogenism—are also described in patients with affective disorders. Our investigation of reproductive status in women with bipolar disorder found no clinical or biochemical evidence of PCOS with mood-stabilizer treatment. We did find that menstrual disturbances were common, however, and sometimes preceded bipolar disorder onset.⁷

Insulin resistance and depression. Others have linked insulin resistance and depression;⁸ depression has been shown to be associated with impaired insulin sensitivity and hyperinsulinemia.⁹ Depressed persons also tend to eat more sweets, drink more alcohol, exercise less, and sleep fewer hours than the nondepressed—all of which contribute to insulin sensitivity and insulin resistance.¹⁰

 

WHAT LINKS ENDOCRINE, MOOD DISORDERS?

Insulin and serotonin. Insulin affects central serotonin (5-HT) levels.¹¹ Thus, central 5-HT system dysregulation that causes depression might also affect peripheral insulin sensitivity—or vice versa.⁹

Tryptophan, a serotonin precursor, competes with other large neutral amino acids for access to the transport system that moves them across the blood-brain barrier. Insulin stimulates uptake of the competing amino acids—but not tryptophan—into muscle tissue. The resulting increased tryptophan ratio in plasma affords it greater access to the transport system to contribute toward serotonin synthesis. Insulin also promotes central catecholaminergic activity, perhaps by suppressing norepinephrine reuptake and prolonging its residence in the synaptic cleft.¹⁰

Table 1

Diagnostic signs of PCOS

Endocrine
Chronically elevated plasma free and total testosterone levels
Increased luteinizing hormone (LH) secretion because of enhanced pituitary sensitivity to GnRH stimulation
Low or normal plasma follicle-stimulating hormone (FSH) levels
Hyperinsulinemia and peripheral insulin resistance
Metabolic abnormalities of hyperinsulinemia and peripheral insulin resistance, including obesity and increased susceptibility to type 2 diabetes and cardiovascular disease
Clinical
Hirsutism
Acne
Infertility
Anovulation or menstrual abnormalities amenorrhea oligomenorrhea dysfunctional uterine bleeding
Morphologic
Ovaries may appear to have multiple cysts 8 to 10 mm in diameter

Several studies have found that insulin resistance and hyperinsulinemia can resolve after depression recovery.^9,12 Because insulin resistance is a cardinal feature in PCOS pathophysiology,² insulin resistance may be a common link between depression and PCOS.

Weight gain and obesity also have been described in patients with affective disorders.¹³ Not known is whether the weight gain precedes the psychopathology or is caused by long-term exposure to drugs used to treat affective disorders.

Box 2

We reported a correlation between depression and BMI in women with PCOS.⁴ Depression might be independently associated with BMI, as weight gain and obesity are distressing symptoms associated with depression.¹⁴ However, we found no association between depression and other possibly distressing PCOS symptoms. Thus, the correlation between BMI and depression might more likely reflect the relationship between depression and insulin resistance, as degree of insulin resistance is known to correlate with BMI.

Elevated cortisol. Clearly, other factors—such as hypothalamus-pituitary-adrenal (HPA) axis dysfunction—are known to link affective and endocrine disorders. Hypercortisolemia can lead to both insulin resistance and obesity. Cortisol is one of the glucocorticoids the body secretes in response to stress to mobilize energy by increasing blood glucose levels. Early life stress and chronic emotional stress:

• can impair the negative feedback system that limits cortisol production during stress
• are associated with depression.

Approximately one-half of individuals with depression have elevated serum cortisol.¹⁰ Epidemiologic data show a positive correlation between cortisol levels and insulin resistance,¹⁵ and an association between HPA dysfunction and obesity has been described.¹⁶ Insulin can trigger androgen production by enhancing adrenal sensitivity to adrenocorticotrophic hormone (ACTH).¹⁷

CASE: IMPROVING INSULIN RESISTANCE

We referred Ms. K to an endocrinologist for PCOS evaluation and treatment. Her serum glucose and insulin levels were 83 mg/dL (normal range 70 to 125 mg/dL) and 19.0 uIU/mL (normal range <10 uIU/mL), respectively. These values indicated insulin resistance as determined by the homeostasis model assessment (HOMA) ratio (fasting insulin x fasting glucose/22.5). Values >3.2 indicate insulin resistance, and Ms. K had a HOMA ratio of 3.9. The endocrinologist recommended:

• metformin, starting at 850 mg/d and gradually increased to 2,550 mg/d
• spironolactone, 100 mg/d.

Metformin, a biguanide approved for treating for type 2 diabetes, inhibits hepatic glucose production and increases peripheral insulin sensitivity, but it does not modify pancreatic insulin secretion. It may decrease insulin resistance by reducing gut absorption of glucose, improving glucose uptake by tissues, and/or increasing the number of insulin receptors.¹⁸ In treating PCOS, metformin can:

 

• restore ovulation¹⁹
• decrease insulin resistance, acne, hirsutism, total and bioavailable testosterone, BMI, and waist-hip ratio.^20,21

Although the link between insulin resistance and depression is unclear, insulin is known to contribute to 5-HT synthesis by promoting tryptophan influx into the brain.²² Therefore, drugs used to treat insulin resistance—such as metformin and alpha lipoic acid²³—might be useful in treating depression.

Spironolactone, a mineralocorticoid receptor (MR) antagonist, reduces hirsutism in women with PCOS.²⁴ It also can decrease insulin resistance and fasting insulin levels in PCOS patients and reduce serum testosterone.²⁵

Evidence on treating mood disorders with hormonal agents such as spironolactone is scarce, although treatment-resistant depression has been reported to resolve with antiglucocorticoid use (Box 2).^25-31 Modulating HPA axis activity to treat affective disorders has been investigated.

CASE: GOING ANTIDEPRESSANT-FREE

At first, Ms. K said she wanted to continue taking venlafaxine with the PCOS treatment. After 2 weeks of combined therapy, however, she chose to stop the antidepressant after her depressive symptoms persisted, and her HAM-D-21 score remained at 28.

During the next 4 weeks, as we tapered off the venlafaxine, Ms. K’s HAM-D-21 score dropped to 7, indicating depressive symptom resolution. Despite slow venlafaxine titration, withdrawal symptoms of excessive crying, not feeling “present,” and tingling sensations occurred. Three days of fluoxetine, 20 mg/d, alleviated these symptoms.

Table 2

Insulin-sensitizing medications used to treat PCOS

Medication	Normal dosage	Common side effects
Metformin	500 mg tid	Headache, GI effects (nausea, diarrhea, flatulence) at start of therapy, weight loss, taste disturbances
Pioglitazone	15 to 45 mg once daily	Swelling, headache, respiratory infection, abdominal discomfort, muscle soreness
Rosiglitazone	4 to 8 mg once daily	Headache, mild weight gain

We continued Ms. K’s treatment without antidepressants, and her mood continued to improve with metformin, 2,550 mg/d, and spironolactone, 100 mg/d.

TREATMENT RECOMMENDATIONS

PCOS therapy may take up to 6 months to resolve symptoms such as anovulation or hirsutism, but affective symptoms may improve during the first 6 weeks, as this case shows. Choosing medications to treat depression in patients such as Ms. K depends on whether their PCOS is being treated when they present for psychiatric evaluation.

Patient not being treated for PCOS. Refer her to an endocrinologist for PCOS treatment with an insulin-sensitizing medication, such as metformin (Table 2). Treating the insulin resistance associated with PCOS may also resolve the depression. PCOS drug therapy may also include antiandrogens such as spironolactone to treat hirsutism, male-pattern baldness, and acne. We suggest that spironolactone’s antiandrogen effects may help reduce depressive symptoms.

If your patient is not taking an antidepressant at presentation, try PCOS treatment first. If depressive symptoms persist after 3 months, consider adding an antidepressant. If your patient is taking an antidepressant at presentation but continues to be depressed, offer two options:

• taper off the antidepressant while starting PCOS treatment
• continue the antidepressant while starting PCOS treatment.

The first option allows you to try PCOS treatment alone. If the depression is caused by a common underlying pathophysiology—such as insulin resistance—treating PCOS alone may alleviate her depressive symptoms. Monitor for withdrawal symptoms, which may be minimized with a short-term SSRI, such as fluoxetine.

The second option may help patients who have responded to antidepressants in the past. Adjunctive PCOS treatment may “jump-start” the antidepressant response without withdrawal symptoms, but you will not be sure whether the antidepressant response was caused by PCOS therapy alone or the combination therapy.

Patient being treated for PCOS. Treat her with antidepressants, and consult with her endocrinologist to consider more-aggressive insulin-sensitizing medications, especially if she exhibits high levels of insulin resistance.

Other interventions that increase insulin sensitivity and improve glycemic control—such as improving dietary management and sleep habits, reducing alcohol consumption, and increasing physical activity—might have an antidepressant effect. Therefore, recommend these health practices to patients as adjuncts to drug therapies.

CASE: DEPRESSION IN REMISSION

At the 3-month follow-up visit, Ms. K scored zero on the HAM-D-21 scale. With metformin treatment, she had lost approximately 10 lbs and resumed menstruating approximately every 33 days. She reported experiencing low mood, decreased energy, and irritability during the week before her periods, but these symptoms resolved with menses onset.

Serum glucose was within normal range at 89 mg/dL, and serum insulin was 15.0 uIU/Ml. Her HOMA ratio had dropped to 2.8 (below the 3.2 cut-off for insulin resistance).

Ms. K’s endocrinologist monitored her spironolactone and metformin therapy for approximately 1 year, when she became pregnant.

Discussion. PCOS treatment duration depends on the patient’s response and her goals for therapy. Whether or not she continues PCOS treatment, her primary care physician or endocrinologist should continue to monitor her for insulin resistance’s metabolic consequences, including increased risk of type 2 diabetes and cardiovascular disease.

 

Related resources

• Legro RS, Winans EA. Overview of polycystic ovary syndrome and its relation to psychiatric illness and treatment. Current Psychiatry 2004;3(Dec[suppl]):12-20.
• The Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004 Jan; 81(1):19-25.
• Polycystic ovary syndrome (patient information). The National Women’s Health Information Center. U.S. Department of Health and Human Services, Office on Women’s Health. Available at: http://www.4woman.gov/faq/pcos.htm. Accessed Jan. 13, 2005.

Drug brand names

• Citalopram • Celexa
• Fluoxetine • Prozac
• Ketoconazole • Nizoral
• Metformin • Glucophage
• Metyrapone • Metopirone
• Mifepristone • Mifiprex
• Pioglitazone • Actos
• Rosiglitazone • Avandia
• Spironolactone • Aldactone
• Venlafaxine • Effexor XR

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Women with depression and polycystic ovary syndrome (PCOS) can be trapped in a vicious cycle of hormonal dysregulation. Treating these patients appropriately—with or without antidepressants—requires an understanding of their underlying metabolic disorder.

This article presents a case¹ that exemplifies the association between depression and PCOS (Box 1).^2-4 Based on our research and clinical experience, we offer recommendations to help you manage depression in patients with PCOS.

CHRONIC DEPRESSION WITH AMENORRHEA

Ms. K, age 30, presented for psychiatric evaluation of treatment-resistant recurrent major depression. Her symptoms included sad mood, sleep disturbance, decreased energy, anhedonia, poor concentration, and feelings of guilt and worthlessness. Her score of 28 on the Hamilton Rating Scale for Depression (HAM-D-21) indicated severe depression.

Box 1

This was Ms. K’s third episode of major depression, with the first two occurring at ages 24 and 26. She reported similar symptoms each time. Previous psychiatrists had tried a variety of antidepressants with no therapeutic benefit before she responded to citalopram, 40 mg/d, at age 27. With this selective serotonin reuptake inhibitor, her depressive symptoms resolved for >1 year, except for mild irritability and low mood the week before menses.

Her depressive symptoms returned, however, and her previous psychiatrist treated her for approximately 13 months with extended-release venlafaxine, 225 mg/d. When she remained depressed, she was referred to our clinic for evaluation. Based on her history, we could not determine whether she had become euthymic at age 27 or her symptoms had improved but still met criteria for a major depressive episode.

Ms. K reported having PCOS symptoms from age 19. These included amenorrhea since menarche, hirsutism, hair loss, alopecia, central fat distribution characteristic of PCOS, and male-pattern hair growth on her abdomen and thighs. She was not obese—with a body mass index (BMI) of approximately 25—but had gained 10 to 15 lbs while taking venlafaxine.

Ms. K had never been treated for PCOS but reported that her desire to become pregnant made her more concerned about her symptoms and possible infertility.

Diagnosis. PCOS is usually diagnosed using endocrinologic, clinical, and ultrasonographic criteria (Table 1). Obesity is not a presenting symptom in all women with PCOS. As with Ms. K, about 50% of patients have normal BMI.

Causes of depression in PCOS. Depressed mood in PCOS may be both physiologic and psychological:

• Hypothalamic, pituitary, and other end-organ system dysregulation occurs in both PCOS and affective disorders, which share clinical and biochemical markers including insulin resistance, obesity, and hyperandrogenism.
• PCOS’ clinical sequelae—hirsutism, acne, obesity, hormonal disturbances, fear of infertility, and psychological distress—may damage their self-esteem and female identity.

PCOS’ physical symptoms alone apparently do not account for patients’ worsened mood states. Weiner et al⁵ found that women with PCOS and free testosterone (FT) of 10 to 26 pg/mL (just above normal range) were more depressed than women without PCOS (FT <10 pg/mL) and women with PCOS and FT >26 pg/mL. Women with PCOS with the lower and higher FT levels had similar demographic profiles, but those with the highest FT levels were not the most depressed.

Similarly, in a study of 32 women with PCOS, we found no association between depression and other possibly distressing PCOS symptoms, including hirsutism, irregular menses, acne, or alopecia.⁴

Testosterone and mood disorders. Women such as Ms. K with hyperandrogenic syndromes are at increased risk for mood disorders.⁶ Many metabolic changes associated with PCOS—insulin resistance, obesity, and hyperandrogenism—are also described in patients with affective disorders. Our investigation of reproductive status in women with bipolar disorder found no clinical or biochemical evidence of PCOS with mood-stabilizer treatment. We did find that menstrual disturbances were common, however, and sometimes preceded bipolar disorder onset.⁷

Insulin resistance and depression. Others have linked insulin resistance and depression;⁸ depression has been shown to be associated with impaired insulin sensitivity and hyperinsulinemia.⁹ Depressed persons also tend to eat more sweets, drink more alcohol, exercise less, and sleep fewer hours than the nondepressed—all of which contribute to insulin sensitivity and insulin resistance.¹⁰

 

WHAT LINKS ENDOCRINE, MOOD DISORDERS?

Insulin and serotonin. Insulin affects central serotonin (5-HT) levels.¹¹ Thus, central 5-HT system dysregulation that causes depression might also affect peripheral insulin sensitivity—or vice versa.⁹

Tryptophan, a serotonin precursor, competes with other large neutral amino acids for access to the transport system that moves them across the blood-brain barrier. Insulin stimulates uptake of the competing amino acids—but not tryptophan—into muscle tissue. The resulting increased tryptophan ratio in plasma affords it greater access to the transport system to contribute toward serotonin synthesis. Insulin also promotes central catecholaminergic activity, perhaps by suppressing norepinephrine reuptake and prolonging its residence in the synaptic cleft.¹⁰

Table 1

Diagnostic signs of PCOS

Endocrine
Chronically elevated plasma free and total testosterone levels
Increased luteinizing hormone (LH) secretion because of enhanced pituitary sensitivity to GnRH stimulation
Low or normal plasma follicle-stimulating hormone (FSH) levels
Hyperinsulinemia and peripheral insulin resistance
Metabolic abnormalities of hyperinsulinemia and peripheral insulin resistance, including obesity and increased susceptibility to type 2 diabetes and cardiovascular disease
Clinical
Hirsutism
Acne
Infertility
Anovulation or menstrual abnormalities amenorrhea oligomenorrhea dysfunctional uterine bleeding
Morphologic
Ovaries may appear to have multiple cysts 8 to 10 mm in diameter

Several studies have found that insulin resistance and hyperinsulinemia can resolve after depression recovery.^9,12 Because insulin resistance is a cardinal feature in PCOS pathophysiology,² insulin resistance may be a common link between depression and PCOS.

Weight gain and obesity also have been described in patients with affective disorders.¹³ Not known is whether the weight gain precedes the psychopathology or is caused by long-term exposure to drugs used to treat affective disorders.

Box 2

We reported a correlation between depression and BMI in women with PCOS.⁴ Depression might be independently associated with BMI, as weight gain and obesity are distressing symptoms associated with depression.¹⁴ However, we found no association between depression and other possibly distressing PCOS symptoms. Thus, the correlation between BMI and depression might more likely reflect the relationship between depression and insulin resistance, as degree of insulin resistance is known to correlate with BMI.

Elevated cortisol. Clearly, other factors—such as hypothalamus-pituitary-adrenal (HPA) axis dysfunction—are known to link affective and endocrine disorders. Hypercortisolemia can lead to both insulin resistance and obesity. Cortisol is one of the glucocorticoids the body secretes in response to stress to mobilize energy by increasing blood glucose levels. Early life stress and chronic emotional stress:

• can impair the negative feedback system that limits cortisol production during stress
• are associated with depression.

Approximately one-half of individuals with depression have elevated serum cortisol.¹⁰ Epidemiologic data show a positive correlation between cortisol levels and insulin resistance,¹⁵ and an association between HPA dysfunction and obesity has been described.¹⁶ Insulin can trigger androgen production by enhancing adrenal sensitivity to adrenocorticotrophic hormone (ACTH).¹⁷

CASE: IMPROVING INSULIN RESISTANCE

We referred Ms. K to an endocrinologist for PCOS evaluation and treatment. Her serum glucose and insulin levels were 83 mg/dL (normal range 70 to 125 mg/dL) and 19.0 uIU/mL (normal range <10 uIU/mL), respectively. These values indicated insulin resistance as determined by the homeostasis model assessment (HOMA) ratio (fasting insulin x fasting glucose/22.5). Values >3.2 indicate insulin resistance, and Ms. K had a HOMA ratio of 3.9. The endocrinologist recommended:

• metformin, starting at 850 mg/d and gradually increased to 2,550 mg/d
• spironolactone, 100 mg/d.

Metformin, a biguanide approved for treating for type 2 diabetes, inhibits hepatic glucose production and increases peripheral insulin sensitivity, but it does not modify pancreatic insulin secretion. It may decrease insulin resistance by reducing gut absorption of glucose, improving glucose uptake by tissues, and/or increasing the number of insulin receptors.¹⁸ In treating PCOS, metformin can:

 

• restore ovulation¹⁹
• decrease insulin resistance, acne, hirsutism, total and bioavailable testosterone, BMI, and waist-hip ratio.^20,21

Although the link between insulin resistance and depression is unclear, insulin is known to contribute to 5-HT synthesis by promoting tryptophan influx into the brain.²² Therefore, drugs used to treat insulin resistance—such as metformin and alpha lipoic acid²³—might be useful in treating depression.

Spironolactone, a mineralocorticoid receptor (MR) antagonist, reduces hirsutism in women with PCOS.²⁴ It also can decrease insulin resistance and fasting insulin levels in PCOS patients and reduce serum testosterone.²⁵

Evidence on treating mood disorders with hormonal agents such as spironolactone is scarce, although treatment-resistant depression has been reported to resolve with antiglucocorticoid use (Box 2).^25-31 Modulating HPA axis activity to treat affective disorders has been investigated.

CASE: GOING ANTIDEPRESSANT-FREE

At first, Ms. K said she wanted to continue taking venlafaxine with the PCOS treatment. After 2 weeks of combined therapy, however, she chose to stop the antidepressant after her depressive symptoms persisted, and her HAM-D-21 score remained at 28.

During the next 4 weeks, as we tapered off the venlafaxine, Ms. K’s HAM-D-21 score dropped to 7, indicating depressive symptom resolution. Despite slow venlafaxine titration, withdrawal symptoms of excessive crying, not feeling “present,” and tingling sensations occurred. Three days of fluoxetine, 20 mg/d, alleviated these symptoms.

Table 2

Insulin-sensitizing medications used to treat PCOS

Medication	Normal dosage	Common side effects
Metformin	500 mg tid	Headache, GI effects (nausea, diarrhea, flatulence) at start of therapy, weight loss, taste disturbances
Pioglitazone	15 to 45 mg once daily	Swelling, headache, respiratory infection, abdominal discomfort, muscle soreness
Rosiglitazone	4 to 8 mg once daily	Headache, mild weight gain

We continued Ms. K’s treatment without antidepressants, and her mood continued to improve with metformin, 2,550 mg/d, and spironolactone, 100 mg/d.

TREATMENT RECOMMENDATIONS

PCOS therapy may take up to 6 months to resolve symptoms such as anovulation or hirsutism, but affective symptoms may improve during the first 6 weeks, as this case shows. Choosing medications to treat depression in patients such as Ms. K depends on whether their PCOS is being treated when they present for psychiatric evaluation.

Patient not being treated for PCOS. Refer her to an endocrinologist for PCOS treatment with an insulin-sensitizing medication, such as metformin (Table 2). Treating the insulin resistance associated with PCOS may also resolve the depression. PCOS drug therapy may also include antiandrogens such as spironolactone to treat hirsutism, male-pattern baldness, and acne. We suggest that spironolactone’s antiandrogen effects may help reduce depressive symptoms.

If your patient is not taking an antidepressant at presentation, try PCOS treatment first. If depressive symptoms persist after 3 months, consider adding an antidepressant. If your patient is taking an antidepressant at presentation but continues to be depressed, offer two options:

• taper off the antidepressant while starting PCOS treatment
• continue the antidepressant while starting PCOS treatment.

The first option allows you to try PCOS treatment alone. If the depression is caused by a common underlying pathophysiology—such as insulin resistance—treating PCOS alone may alleviate her depressive symptoms. Monitor for withdrawal symptoms, which may be minimized with a short-term SSRI, such as fluoxetine.

The second option may help patients who have responded to antidepressants in the past. Adjunctive PCOS treatment may “jump-start” the antidepressant response without withdrawal symptoms, but you will not be sure whether the antidepressant response was caused by PCOS therapy alone or the combination therapy.

Patient being treated for PCOS. Treat her with antidepressants, and consult with her endocrinologist to consider more-aggressive insulin-sensitizing medications, especially if she exhibits high levels of insulin resistance.

Other interventions that increase insulin sensitivity and improve glycemic control—such as improving dietary management and sleep habits, reducing alcohol consumption, and increasing physical activity—might have an antidepressant effect. Therefore, recommend these health practices to patients as adjuncts to drug therapies.

CASE: DEPRESSION IN REMISSION

At the 3-month follow-up visit, Ms. K scored zero on the HAM-D-21 scale. With metformin treatment, she had lost approximately 10 lbs and resumed menstruating approximately every 33 days. She reported experiencing low mood, decreased energy, and irritability during the week before her periods, but these symptoms resolved with menses onset.

Serum glucose was within normal range at 89 mg/dL, and serum insulin was 15.0 uIU/Ml. Her HOMA ratio had dropped to 2.8 (below the 3.2 cut-off for insulin resistance).

Ms. K’s endocrinologist monitored her spironolactone and metformin therapy for approximately 1 year, when she became pregnant.

Discussion. PCOS treatment duration depends on the patient’s response and her goals for therapy. Whether or not she continues PCOS treatment, her primary care physician or endocrinologist should continue to monitor her for insulin resistance’s metabolic consequences, including increased risk of type 2 diabetes and cardiovascular disease.

 

Related resources

• Legro RS, Winans EA. Overview of polycystic ovary syndrome and its relation to psychiatric illness and treatment. Current Psychiatry 2004;3(Dec[suppl]):12-20.
• The Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004 Jan; 81(1):19-25.
• Polycystic ovary syndrome (patient information). The National Women’s Health Information Center. U.S. Department of Health and Human Services, Office on Women’s Health. Available at: http://www.4woman.gov/faq/pcos.htm. Accessed Jan. 13, 2005.

Drug brand names

• Citalopram • Celexa
• Fluoxetine • Prozac
• Ketoconazole • Nizoral
• Metformin • Glucophage
• Metyrapone • Metopirone
• Mifepristone • Mifiprex
• Pioglitazone • Actos
• Rosiglitazone • Avandia
• Spironolactone • Aldactone
• Venlafaxine • Effexor XR

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Children with developmental problems and serious psychopathologies often do not fit neatly into DSM diagnoses.^1,2 These “diagnostically homeless” children—handicapped by hyperactivity, volcanic rages, extreme anxieties, and other complex problems—need assessment and treatment that address four domains of dysfunction:

• mood/anxiety problems
  
• possible psychosis
  
• language/thought disorder
  
• relationship/socialization problems.
  

This article offers snapshots of four children with undetermined diagnoses, explores the dilemma of treating such patients without knowing what they really have, and recommends a treatment approach to help them function better in school and at home.

WHO ARE THE ‘DIAGNOSTICALLY HOMELESS’?

Devon is 5. He is extremely hyperactive and impulsive, with a normal IQ but significant language delay. He exhibits little but not absent interest in peers and rages when changes are imposed on him.

Table 1

Criteria describing impairments in ‘diagnostically homeless’ children

Domain	Multiple complex developmental disorder (MCDD)*	Multidimensionally impaired (MDI) syndrome†	Schizotypal personality disorder
Anxiety symptoms	Intense generalized anxiety, diffuse tension or irritability; unusual fears and phobias, peculiar in content or intensity; recurrent panic episodes, terror, or flooding with anxiety	Unspecified	Excessive social anxiety associated with paranoid fears
Affect regulation	Significant, wide, emotional variability out of proportion to precipitants	Nearly daily periods of emotional lability disproportionate to precipitants	Inappropriate or constricted affect
Psychotic-like symptoms	Magical thinking; illogical confusion between reality and fantasy; grandiose fantasies of special powers	Poor ability to separate reality from fantasy	Ideas of reference; unusual perceptual experiences; suspicious; eccentric
Thought/language disorder	Thought problems including irrationality, sudden intrusions on normal thought process, neologisms or nonsense words repeated over and over; blatantly illogical, bizarre ideas	Thought disorder specifically excluded	Odd thinking; vague, circumstantial, metaphorical speech, overelaborate or stereotyped
Problems with social functioning	Social disinterest, detachment; instrumental relatedness; high degrees of ambivalence to adults, manifested by clinging, overly controlling, needy behavior and/or aggressive, oppositional behavior; limited capacity to empathize	Impaired interpersonal skills despite desire to initiate social interactions with peers	Lack of close friends or confidants other than relatives
* PDD NOS (pervasive developmental disorder, not otherwise specified) is the closest DSM-IV-TR designation.
† Psychosis NOS is the closest DSM-IV-TR designation.
Source: References 1, 3, 8-13.

Devon says he can run faster than cars and swim across the ocean. He has “more than ADHD,” less than autism/pervasive developmental disorder (PDD). Were he older, his grandiosity might seem manic but his age and language delay make this suspect.

Steven is 11, referred “to rule out bipolar disorder” and to evaluate hyperactivity, explosiveness, and nightmares. He didn’t speak until he was 22 months old. He worries that bad people are chasing him, fears skeletons under his bed, has nightmares of vampires, and believes that cartoon characters are real and that Sponge Bob is his protector. He says he sees “scary stuff” out of the corner of his eyes. He does not have a thought disorder; psychotic symptoms are more than an overactive imagination or anxiety.

Lauren, age 12, has been diagnosed with attention-deficit/hyperactivity disorder (ADHD) but now presents with withdrawn, depressed, and defiant behaviors. She is described as a “loner” who has never related well to other children. Lauren speaks about being tortured by her peers to the point of sounding paranoid. Her conversation is extremely circumstantial and rambling.

Richard, age 8, has motor coordination, attachment, and disinhibition problems. He hears voices telling him to do bad things, such as hurt people, steal things, and “break stuff.” He doesn’t mind the voices much, and they don’t pervade his life the way hallucinations do in schizophrenia.

Children such as these are common, and it is unclear whether they have a developmental disorder, the prodrome of a psychotic or mood disorder, or idiosyncratic personalities. They don’t meet criteria for many disorders, including autism, bipolar disorder, schizophrenia, and obsessive-compulsive disorder (OCD). They have more-extensive difficulties than those seen in ADHD, generalized anxiety disorder (GAD), or OCD.

Clinically, they are either forced into a category someone thinks they resemble (such as mania in Devon’s case) or are given a “not otherwise specified” (NOS) label (such as PDD NOS, psychosis NOS, or mood disorder NOS), the severity of which goes unacknowledged.

Problems with ‘NOS.’ Some might consider “NOS” a less-severe problem than a specific diagnosis, but these children are very impaired. They are excluded from treatment studies because they do not meet formal criteria for the designated disorder or they get included erroneously because the structured diagnostic interview doesn’t assess what they really have.

Meaningful psychoeducation for their parents is impossible because no Web site or book exists to help them help their child. Finally, no follow-up studies have been done of this group of children because no one can agree on a diagnosis. Small studies have addressed some of these concerns, but outcomes—not surprisingly—are wide-ranging.^3-6

 

NOS diagnoses also don’t adequately address children with marked anxiety, unusual fears, and perseverative behaviors who are socially clumsy but manage reciprocal social interaction. These children are substantially disabled by:

• attention difficulties
  
• mood dysregulation (including anxiety and/or manic symptoms)
  
• trouble with transitions/change
  
• motor problems (not infrequently)
  
• pragmatic language/social difficulties.
  

Few tests exist for pragmatic language skills, which include being able to maintain a reciprocal conversation, stay on topic, understand the listener’s needs, and use correct body language and voice tone. Children with PDD, ADHD, and other language disorders are most often disabled in this area of communication.

Diagnostic terms that have tried to classify these children (Table 1) include:

• childhood-onset PDD, described in DSM-III. This category was dropped in DSM III-R to be included in PDD, then largely ignored in DSM-IV when autism criteria were refined.
  
• multiple complex developmental disorder (MCDD),^7-9 which appears to describe children within the autism spectrum (such as PDD NOS)
  
• multidimensionally-impaired (MDI) syndrome, whose atypical psychosis has been called “psychosis NOS”^10-11
  
• schizotypal personality disorder, which addresses similar symptoms (although mental health professionals are loathe to use a personality disorder diagnosis in a child).¹²
  

These designations all include psychopathology in four domains: anxiety, affect regulation, communication, psychosis, and relatedness.

At this time, however, diagnostic conclusions about this heterogeneous group of children are premature. Our classification system does not do them justice, and we need to study them for what they have, rather than forcing them into our current alternatives.

Prevalence. To find out how many patients in our university-based, tertiary-care clinic do not fit DSM-IV-TR nosology, we examined data from faculty evaluations of 624 children and adolescents.¹³ These included semi-structured interviews of parent and child, rating scales from parents and teachers, and testing information from the schools in two-thirds of cases.

The result: nearly 25% of our child and adolescent psychiatry outpatients are “diagnostically homeless.” Like the rest of our patient population, these children are:

• 80% male
  
• 60% under age 12
  
• 86% Caucasian
  
• 85% living with their biological mothers.
  

These children are referred to psychiatrists for many reasons:

• ADHD (16%). They have great difficulty with executive functions, such as paying attention, inhibiting impulsive responses, planning and organizing, making transitions from one activity to another, and controlling emotion. Their problems, however, go much beyond ADHD.
  
• Bipolar disorder (15%) or depression/anxiety (16%). They have catastrophic anxiety and/or frightening rages triggered by apparently trivial circumstances. They balk or “shut down” when people want them to move or act faster than they can move or act.
  
• To “rule out autism” (19%). More than one-half (56%) of these children have a diagnosable speech or language disorder, compared with 35% among our other child psychiatry outpatients.
  
• For educational assessment (23%). School systems request guidance for educational interventions because these children are possibly psychotic and disturbing to teachers and children. They may be unable to execute homework assignments and fail their courses but surprisingly do grade-level work on achievement tests.
  

ASSESSING FOUR DOMAINS

We can consolidate the domains needing assessment into mood/anxiety problems, possible psychosis, language/thought disorder, and relationship/socialization problems. Although evaluating and treating some of these domains may be beyond the psychiatrist’s purview, we must make sure that other professionals attend to them.

Anxiety and mood. Understanding these children’s anxieties is important. A routine fear of bees is a simple phobia, whereas catastrophic anxiety over a highly unlikely impending tornado and perseverative interest in the weather may be more common in a PDD spectrum disorder. Anxiety about going to sleep because a monster is going to suck out one’s brains does not easily fit into the rubric of generalized anxiety.¹⁴

Irritability is these youngsters’ most disabling mood symptom. Volcanic anger and rage that prompts referral occurs in numerous conditions, including mania. Many of the children described in Ross Greene’s book, The Explosive Child,¹⁵ have conditions other than bipolar disorder. Although parents and teachers often describe these events as occurring without provocation, a good functional behavioral assessment will usually reveal a precipitant.

Table 2

Assessing children’s social and language skills

Social assessment	Seen in…
Are the child’s social abilities delayed?	ADHD
Is he uninterested in social situations?	Autism
Is he clueless about social interaction?	Autism spectrum disorders including MCDD, MDI, PDD NOS, nonverbal learning disability
Are social interactions deviant?	Schizotypal personality disorder/schizophrenia
Does child appear shut down/behaviorally inhibited in unfamiliar settings, with greater comfort at home or with familiar people?	Social phobia
Language assessment (can be done by psychiatrist)
Age at first word use; age at first use of short sentences Early interest in language? Nonverbal communication? Communication for sharing?
Useful questions	Seen in…
Was communication delayed but then progressed “normally”?	Developmental language disorder
Did it begin normally and stop?	Autism
Was/is it egocentric and/or unidimensional?	Asperger’s disorder; nonverbal learning disability
Was/is it bizarre or paranoid?	Schizotypal personality disorder
Pragmatic language problems?	All of the above, MCDD, MDI, ADHD
Communication domains (may require speech pathologist assessment)
Expressive and receptive language
Pragmatic language (the child’s ability to communicate in the real world; see Table 3)
Written language
Audiology (hearing and auditory processing)
ADHD: attention-deficit/hyperactivity disorder
MCDD: multiple complex developmental disorder
MDI: multidimensionally impaired syndrome
PDD NOS: pervasive developmental disorder not otherwise specified

 

Possible psychosis. These children may have impaired reality testing that can be difficult to assess; thus, deciding whether the child is experiencing psychotic symptoms can be a challenge. The child may be intensely involved with fantasy characters or imaginary companions to such a degree that he or she insists the character is real.^16,17 Developmentally normal fears—as of the dark, monsters, or images from dreams—may preoccupy him or her during the day. Quasi-psychotic symptoms such as these are easily missed if:

• we don’t ask about them
  
• we assume the child is “just pretending” or has a “great imagination”
  
• the child does not volunteer the information spontaneously.¹⁸
  

Table 3

Communication skills children need to learn

Rules of conversation (for example, who is likely to be interested in what) Topic management (when to expand, shift, end a conversation) Awareness of nonverbal cues Social expectations in various settings Operational knowledge of the language of emotions and mental states (how to express feelings and the different ways we experience ourselves) How to monitor a listener’s relative interest The meaning of eye contact, voice tone, and voice inflection Awareness of how social settings affect communication, such as voice volume (whisper in the library, shout on the soccer field) and speech style (slang with peers, formal style for classroom recitation) Body proximity (how to avoid invading someone’s space) Decoding facial expression (such as what it means when someone rolls his eyes) Special instruction to help decipher nonliteral communication, including teasing, irony, sarcasm, emotional tones of speech

In assessing psychotic symptoms, the first goal is to get a detailed picture of unusual thoughts or images the child is experiencing in different settings, including school, home, and with peers. Then evaluate these symptoms in the broader clinical context of how the child is functioning in other domains.

Language/thought disorder. Parents may not recognize that their child has a thought or language disorder because they have filled in the blanks and interpreted for him or her for so long. Asking the child “yes” and “no” questions will not elucidate these disorders, either. The examiner must talk to the child to determine his or her ability to:

• sustain an extended narration that makes sense
  
• stay on the topic
  
• care whether the listener understands what the child is talking about
  
• make a point.
  

Distinguishing between a thought or and language disorder in a child is difficult, although the more illogical the communication, the less likely it is to be a language disorder. If the child connects ideas that don’t make sense, ask him or her to explain how the subject shifted or what he or she meant. Children with language disorders may have misunderstood the question or may have expected the examiner to make connections, but the explanation usually makes sense. When it doesn’t, we become more concerned that the child has a thought disorder.

Nonverbal communication realms include eye contact, appropriate hand gestures and facial expression, tone of voice, and vocal inflection. Other important areas of language to assess are summarized in Table 2.

Relationship/socialization problems. It is important to know whether the child has friends, wants friends, or prefers being with younger children. Peer relationships may be absent, delayed, or deviant.

Other assessments. The diagnostically homeless children we see have complicated family histories of psychopathology. Their first-degree relatives have a higher number of heritable disorders—including bipolar disorder, panic disorder, ADHD, learning disabilities, and “nervous breakdowns”—than do those of children with uncomplicated ADHD, bipolar disorder, or anxiety disorders. Ask about these conditions when taking the family history; if a family member is said to be bipolar, get a description of the person’s symptoms.

Table 4

Targeting drug therapies to treat children’s symptoms

Drug class	Efficacy by symptom domain
Atypical antipsychotics	Psychosis/thought disorder: Can reduce psychotic symptoms
	Anxiety symptoms: Can reduce extreme anxieties
	Affect regulation: Improved by mood-stabilizing effect
	Socialization problems: Appear to modify affective aggression, hyperactivity, and impulsivity, which can improve socialization and pragmatic communication
Mood stabilizers	Psychosis/thought disorder: Not primary area of effectiveness
	Anxiety symptoms: May be helpful; not primary area of effectiveness
	Affect regulation: Address mood dysregulation
	Socialization problems: May reduce aggressive outbursts and mood, which can improve socialization
Stimulants*	Psychosis/thought disorder: Can produce or intensify psychotic symptoms and agitation
	Anxiety: Usually do not improve anxiety; can intensify anxiety and agitation
	Affect regulation: Not a primary effect in severe cases; address impulsive aggression via mood stabilization
	Socialization problems: Can improve functioning via decreased impulsivity, inattention, and aggression
SSRI antidepressants†	Psychosis/thought disorder: Do not directly address
	Anxiety: Can be effective in decreasing anxiety
	Affect regulation: Can improve depressed mood
	Socialization problems: Can be improved as a result of improved mood and decreased anxiety
* Stimulants often increase agitation and disinhibition.
† Watch for behavioral disinhibition, possible increase in suicidality, with selective serotonin reuptake inhibitors (SSRIs).

 

A skilled psychologist or speech pathologist can help you determine the presence or absence of cognitive and language dysfunction and learning disabilities. Even before we interview the parents and child, we ask parents and teachers to rate the child’s attention, behavior, mood, PDD-like symptoms, and anxiety, using the Child/Adolescent Symptom Inventory (see Related resources). We use the youth version with children age 10 and older, then review the symptoms with the parents and child to make sure we understand all presenting comorbidities.

TREATMENT

Nonmedical interventions begin with an accurate diagnosis, where possible. Then the four steps of treatment are to:

• address each domain of dysfunction
  
• translate findings to parent, child, and teachers/school.
  
• provide settings and resources that allow the child to work most effectively
  
• develop a behavioral program for the most frequent problems, with consistent response by caretakers and educators.
  

The educational setting needs to be adapted for these children. This usually implies individualized attention in small classes or small work groups. Assigning an aide to the child may be effective in larger settings, but other support and expertise is needed. Otherwise, all the aide does is run interference for the child, which ultimately may be more isolating than a special education class.

A communication specialist interested in pragmatics is needed to make sure the child is understood and being understood in the classroom. Table 3, summarizes communications skills the child needs to learn. An educational specialist who serves a resource to other professionals may also help the child. Curriculum should be based on long-term goals rather than on inflexible credit schedules that teach worthless, unlearnable information and demoralize the student.

Finally, the education setting should provide opportunities for structured social interaction and less-structured but supervised—”bully-proofed”—interactions.

Medications. No systematic medical treatment data exist, as there is no way to classify these children. They are usually treated with multiple medications for their specific symptom cluster abnormalities (Table 4). Options include:

• atypical antipsychotics such as risperidone, quetiapine, aripiprazole, ziprasidone, or olanzapine
  
• mood stabilizers such as valproic acid, lithium, or lamotrigine
  
• stimulants such as methylphenidate, amphetamine salts, atomoxetine, or bupropion (a mild stimulant and an antidepressant)
  
• selective serotonin reuptake inhibitors, such as fluoxetine, sertraline, citalopram, paroxetine, or fluvoxamine.
  

Unfortunately, drug therapy may cause behavioral toxicity—tearfulness, irritability, disinhibition, activation, agitation, hallucinations and possibly even suicidal behavior. Stopping the medication usually reverses this kind of adverse effect.¹⁹

Medication side effects understandably frighten parents—who may be reluctant to have their children use any drug therapies. Counsel the parents in advance that side effects may occur.

Related resources

• Child/Adolescent Symptom Inventory. http://www.checkmateplus.com. Accessed Jan. 11, 2005.
  

Drug brand names

• Amphetamine • Adderall
  
• Aripiprazole • Abilify
  
• Atomoxetine • Strattera
  
• Bupropion • Wellbutrin
  
• Citalopram • Celexa
  
• Fluoxetine • Prozac
  
• Fluvoxamine • Luvox
  
• Lamotrigine • Lamictal
  
• Lithium carbonate • Lithobid, others
  
• Methylphenidate • Concerta, Ritalin
  
• Olanzapine • Zyprexa
  
• Paroxetine • Paxil
  
• Quetiapine • Seroquel
  
• Risperidone • Risperdal
  
• Sertraline • Zoloft
  
• Valproic acid • Depakote
  
• Ziprasidone • Geodon
  

Disclosures

Dr. Weisbrot receives grants from Pfizer Inc.

Dr. Carlson receives grants from or is a speaker for Janssen Pharmaceutica, Eli Lilly and Co., Shire Pharmaceuticals Groups, and Abbott Laboratories; is a consultant to Janssen Pharmaceutica and Eli Lilly and Co.; and is an advisor to Otsuka America Pharmaceutical, Pfizer Inc., and Ortho-McNeil Pharmaceutical.

Children with developmental problems and serious psychopathologies often do not fit neatly into DSM diagnoses.^1,2 These “diagnostically homeless” children—handicapped by hyperactivity, volcanic rages, extreme anxieties, and other complex problems—need assessment and treatment that address four domains of dysfunction:

• mood/anxiety problems
  
• possible psychosis
  
• language/thought disorder
  
• relationship/socialization problems.
  

This article offers snapshots of four children with undetermined diagnoses, explores the dilemma of treating such patients without knowing what they really have, and recommends a treatment approach to help them function better in school and at home.

WHO ARE THE ‘DIAGNOSTICALLY HOMELESS’?

Devon is 5. He is extremely hyperactive and impulsive, with a normal IQ but significant language delay. He exhibits little but not absent interest in peers and rages when changes are imposed on him.

Table 1

Criteria describing impairments in ‘diagnostically homeless’ children

Domain	Multiple complex developmental disorder (MCDD)*	Multidimensionally impaired (MDI) syndrome†	Schizotypal personality disorder
Anxiety symptoms	Intense generalized anxiety, diffuse tension or irritability; unusual fears and phobias, peculiar in content or intensity; recurrent panic episodes, terror, or flooding with anxiety	Unspecified	Excessive social anxiety associated with paranoid fears
Affect regulation	Significant, wide, emotional variability out of proportion to precipitants	Nearly daily periods of emotional lability disproportionate to precipitants	Inappropriate or constricted affect
Psychotic-like symptoms	Magical thinking; illogical confusion between reality and fantasy; grandiose fantasies of special powers	Poor ability to separate reality from fantasy	Ideas of reference; unusual perceptual experiences; suspicious; eccentric
Thought/language disorder	Thought problems including irrationality, sudden intrusions on normal thought process, neologisms or nonsense words repeated over and over; blatantly illogical, bizarre ideas	Thought disorder specifically excluded	Odd thinking; vague, circumstantial, metaphorical speech, overelaborate or stereotyped
Problems with social functioning	Social disinterest, detachment; instrumental relatedness; high degrees of ambivalence to adults, manifested by clinging, overly controlling, needy behavior and/or aggressive, oppositional behavior; limited capacity to empathize	Impaired interpersonal skills despite desire to initiate social interactions with peers	Lack of close friends or confidants other than relatives
* PDD NOS (pervasive developmental disorder, not otherwise specified) is the closest DSM-IV-TR designation.
† Psychosis NOS is the closest DSM-IV-TR designation.
Source: References 1, 3, 8-13.

Devon says he can run faster than cars and swim across the ocean. He has “more than ADHD,” less than autism/pervasive developmental disorder (PDD). Were he older, his grandiosity might seem manic but his age and language delay make this suspect.

Steven is 11, referred “to rule out bipolar disorder” and to evaluate hyperactivity, explosiveness, and nightmares. He didn’t speak until he was 22 months old. He worries that bad people are chasing him, fears skeletons under his bed, has nightmares of vampires, and believes that cartoon characters are real and that Sponge Bob is his protector. He says he sees “scary stuff” out of the corner of his eyes. He does not have a thought disorder; psychotic symptoms are more than an overactive imagination or anxiety.

Lauren, age 12, has been diagnosed with attention-deficit/hyperactivity disorder (ADHD) but now presents with withdrawn, depressed, and defiant behaviors. She is described as a “loner” who has never related well to other children. Lauren speaks about being tortured by her peers to the point of sounding paranoid. Her conversation is extremely circumstantial and rambling.

Richard, age 8, has motor coordination, attachment, and disinhibition problems. He hears voices telling him to do bad things, such as hurt people, steal things, and “break stuff.” He doesn’t mind the voices much, and they don’t pervade his life the way hallucinations do in schizophrenia.

Children such as these are common, and it is unclear whether they have a developmental disorder, the prodrome of a psychotic or mood disorder, or idiosyncratic personalities. They don’t meet criteria for many disorders, including autism, bipolar disorder, schizophrenia, and obsessive-compulsive disorder (OCD). They have more-extensive difficulties than those seen in ADHD, generalized anxiety disorder (GAD), or OCD.

Clinically, they are either forced into a category someone thinks they resemble (such as mania in Devon’s case) or are given a “not otherwise specified” (NOS) label (such as PDD NOS, psychosis NOS, or mood disorder NOS), the severity of which goes unacknowledged.

Problems with ‘NOS.’ Some might consider “NOS” a less-severe problem than a specific diagnosis, but these children are very impaired. They are excluded from treatment studies because they do not meet formal criteria for the designated disorder or they get included erroneously because the structured diagnostic interview doesn’t assess what they really have.

Meaningful psychoeducation for their parents is impossible because no Web site or book exists to help them help their child. Finally, no follow-up studies have been done of this group of children because no one can agree on a diagnosis. Small studies have addressed some of these concerns, but outcomes—not surprisingly—are wide-ranging.^3-6

 

NOS diagnoses also don’t adequately address children with marked anxiety, unusual fears, and perseverative behaviors who are socially clumsy but manage reciprocal social interaction. These children are substantially disabled by:

• attention difficulties
  
• mood dysregulation (including anxiety and/or manic symptoms)
  
• trouble with transitions/change
  
• motor problems (not infrequently)
  
• pragmatic language/social difficulties.
  

Few tests exist for pragmatic language skills, which include being able to maintain a reciprocal conversation, stay on topic, understand the listener’s needs, and use correct body language and voice tone. Children with PDD, ADHD, and other language disorders are most often disabled in this area of communication.

Diagnostic terms that have tried to classify these children (Table 1) include:

• childhood-onset PDD, described in DSM-III. This category was dropped in DSM III-R to be included in PDD, then largely ignored in DSM-IV when autism criteria were refined.
  
• multiple complex developmental disorder (MCDD),^7-9 which appears to describe children within the autism spectrum (such as PDD NOS)
  
• multidimensionally-impaired (MDI) syndrome, whose atypical psychosis has been called “psychosis NOS”^10-11
  
• schizotypal personality disorder, which addresses similar symptoms (although mental health professionals are loathe to use a personality disorder diagnosis in a child).¹²
  

These designations all include psychopathology in four domains: anxiety, affect regulation, communication, psychosis, and relatedness.

At this time, however, diagnostic conclusions about this heterogeneous group of children are premature. Our classification system does not do them justice, and we need to study them for what they have, rather than forcing them into our current alternatives.

Prevalence. To find out how many patients in our university-based, tertiary-care clinic do not fit DSM-IV-TR nosology, we examined data from faculty evaluations of 624 children and adolescents.¹³ These included semi-structured interviews of parent and child, rating scales from parents and teachers, and testing information from the schools in two-thirds of cases.

The result: nearly 25% of our child and adolescent psychiatry outpatients are “diagnostically homeless.” Like the rest of our patient population, these children are:

• 80% male
  
• 60% under age 12
  
• 86% Caucasian
  
• 85% living with their biological mothers.
  

These children are referred to psychiatrists for many reasons:

• ADHD (16%). They have great difficulty with executive functions, such as paying attention, inhibiting impulsive responses, planning and organizing, making transitions from one activity to another, and controlling emotion. Their problems, however, go much beyond ADHD.
  
• Bipolar disorder (15%) or depression/anxiety (16%). They have catastrophic anxiety and/or frightening rages triggered by apparently trivial circumstances. They balk or “shut down” when people want them to move or act faster than they can move or act.
  
• To “rule out autism” (19%). More than one-half (56%) of these children have a diagnosable speech or language disorder, compared with 35% among our other child psychiatry outpatients.
  
• For educational assessment (23%). School systems request guidance for educational interventions because these children are possibly psychotic and disturbing to teachers and children. They may be unable to execute homework assignments and fail their courses but surprisingly do grade-level work on achievement tests.
  

ASSESSING FOUR DOMAINS

We can consolidate the domains needing assessment into mood/anxiety problems, possible psychosis, language/thought disorder, and relationship/socialization problems. Although evaluating and treating some of these domains may be beyond the psychiatrist’s purview, we must make sure that other professionals attend to them.

Anxiety and mood. Understanding these children’s anxieties is important. A routine fear of bees is a simple phobia, whereas catastrophic anxiety over a highly unlikely impending tornado and perseverative interest in the weather may be more common in a PDD spectrum disorder. Anxiety about going to sleep because a monster is going to suck out one’s brains does not easily fit into the rubric of generalized anxiety.¹⁴

Irritability is these youngsters’ most disabling mood symptom. Volcanic anger and rage that prompts referral occurs in numerous conditions, including mania. Many of the children described in Ross Greene’s book, The Explosive Child,¹⁵ have conditions other than bipolar disorder. Although parents and teachers often describe these events as occurring without provocation, a good functional behavioral assessment will usually reveal a precipitant.

Table 2

Assessing children’s social and language skills

Social assessment	Seen in…
Are the child’s social abilities delayed?	ADHD
Is he uninterested in social situations?	Autism
Is he clueless about social interaction?	Autism spectrum disorders including MCDD, MDI, PDD NOS, nonverbal learning disability
Are social interactions deviant?	Schizotypal personality disorder/schizophrenia
Does child appear shut down/behaviorally inhibited in unfamiliar settings, with greater comfort at home or with familiar people?	Social phobia
Language assessment (can be done by psychiatrist)
Age at first word use; age at first use of short sentences Early interest in language? Nonverbal communication? Communication for sharing?
Useful questions	Seen in…
Was communication delayed but then progressed “normally”?	Developmental language disorder
Did it begin normally and stop?	Autism
Was/is it egocentric and/or unidimensional?	Asperger’s disorder; nonverbal learning disability
Was/is it bizarre or paranoid?	Schizotypal personality disorder
Pragmatic language problems?	All of the above, MCDD, MDI, ADHD
Communication domains (may require speech pathologist assessment)
Expressive and receptive language
Pragmatic language (the child’s ability to communicate in the real world; see Table 3)
Written language
Audiology (hearing and auditory processing)
ADHD: attention-deficit/hyperactivity disorder
MCDD: multiple complex developmental disorder
MDI: multidimensionally impaired syndrome
PDD NOS: pervasive developmental disorder not otherwise specified

 

Possible psychosis. These children may have impaired reality testing that can be difficult to assess; thus, deciding whether the child is experiencing psychotic symptoms can be a challenge. The child may be intensely involved with fantasy characters or imaginary companions to such a degree that he or she insists the character is real.^16,17 Developmentally normal fears—as of the dark, monsters, or images from dreams—may preoccupy him or her during the day. Quasi-psychotic symptoms such as these are easily missed if:

• we don’t ask about them
  
• we assume the child is “just pretending” or has a “great imagination”
  
• the child does not volunteer the information spontaneously.¹⁸
  

Table 3

Communication skills children need to learn

Rules of conversation (for example, who is likely to be interested in what) Topic management (when to expand, shift, end a conversation) Awareness of nonverbal cues Social expectations in various settings Operational knowledge of the language of emotions and mental states (how to express feelings and the different ways we experience ourselves) How to monitor a listener’s relative interest The meaning of eye contact, voice tone, and voice inflection Awareness of how social settings affect communication, such as voice volume (whisper in the library, shout on the soccer field) and speech style (slang with peers, formal style for classroom recitation) Body proximity (how to avoid invading someone’s space) Decoding facial expression (such as what it means when someone rolls his eyes) Special instruction to help decipher nonliteral communication, including teasing, irony, sarcasm, emotional tones of speech

In assessing psychotic symptoms, the first goal is to get a detailed picture of unusual thoughts or images the child is experiencing in different settings, including school, home, and with peers. Then evaluate these symptoms in the broader clinical context of how the child is functioning in other domains.

Language/thought disorder. Parents may not recognize that their child has a thought or language disorder because they have filled in the blanks and interpreted for him or her for so long. Asking the child “yes” and “no” questions will not elucidate these disorders, either. The examiner must talk to the child to determine his or her ability to:

• sustain an extended narration that makes sense
  
• stay on the topic
  
• care whether the listener understands what the child is talking about
  
• make a point.
  

Distinguishing between a thought or and language disorder in a child is difficult, although the more illogical the communication, the less likely it is to be a language disorder. If the child connects ideas that don’t make sense, ask him or her to explain how the subject shifted or what he or she meant. Children with language disorders may have misunderstood the question or may have expected the examiner to make connections, but the explanation usually makes sense. When it doesn’t, we become more concerned that the child has a thought disorder.

Nonverbal communication realms include eye contact, appropriate hand gestures and facial expression, tone of voice, and vocal inflection. Other important areas of language to assess are summarized in Table 2.

Relationship/socialization problems. It is important to know whether the child has friends, wants friends, or prefers being with younger children. Peer relationships may be absent, delayed, or deviant.

Other assessments. The diagnostically homeless children we see have complicated family histories of psychopathology. Their first-degree relatives have a higher number of heritable disorders—including bipolar disorder, panic disorder, ADHD, learning disabilities, and “nervous breakdowns”—than do those of children with uncomplicated ADHD, bipolar disorder, or anxiety disorders. Ask about these conditions when taking the family history; if a family member is said to be bipolar, get a description of the person’s symptoms.

Table 4

Targeting drug therapies to treat children’s symptoms

Drug class	Efficacy by symptom domain
Atypical antipsychotics	Psychosis/thought disorder: Can reduce psychotic symptoms
	Anxiety symptoms: Can reduce extreme anxieties
	Affect regulation: Improved by mood-stabilizing effect
	Socialization problems: Appear to modify affective aggression, hyperactivity, and impulsivity, which can improve socialization and pragmatic communication
Mood stabilizers	Psychosis/thought disorder: Not primary area of effectiveness
	Anxiety symptoms: May be helpful; not primary area of effectiveness
	Affect regulation: Address mood dysregulation
	Socialization problems: May reduce aggressive outbursts and mood, which can improve socialization
Stimulants*	Psychosis/thought disorder: Can produce or intensify psychotic symptoms and agitation
	Anxiety: Usually do not improve anxiety; can intensify anxiety and agitation
	Affect regulation: Not a primary effect in severe cases; address impulsive aggression via mood stabilization
	Socialization problems: Can improve functioning via decreased impulsivity, inattention, and aggression
SSRI antidepressants†	Psychosis/thought disorder: Do not directly address
	Anxiety: Can be effective in decreasing anxiety
	Affect regulation: Can improve depressed mood
	Socialization problems: Can be improved as a result of improved mood and decreased anxiety
* Stimulants often increase agitation and disinhibition.
† Watch for behavioral disinhibition, possible increase in suicidality, with selective serotonin reuptake inhibitors (SSRIs).

 

A skilled psychologist or speech pathologist can help you determine the presence or absence of cognitive and language dysfunction and learning disabilities. Even before we interview the parents and child, we ask parents and teachers to rate the child’s attention, behavior, mood, PDD-like symptoms, and anxiety, using the Child/Adolescent Symptom Inventory (see Related resources). We use the youth version with children age 10 and older, then review the symptoms with the parents and child to make sure we understand all presenting comorbidities.

TREATMENT

Nonmedical interventions begin with an accurate diagnosis, where possible. Then the four steps of treatment are to:

• address each domain of dysfunction
  
• translate findings to parent, child, and teachers/school.
  
• provide settings and resources that allow the child to work most effectively
  
• develop a behavioral program for the most frequent problems, with consistent response by caretakers and educators.
  

The educational setting needs to be adapted for these children. This usually implies individualized attention in small classes or small work groups. Assigning an aide to the child may be effective in larger settings, but other support and expertise is needed. Otherwise, all the aide does is run interference for the child, which ultimately may be more isolating than a special education class.

A communication specialist interested in pragmatics is needed to make sure the child is understood and being understood in the classroom. Table 3, summarizes communications skills the child needs to learn. An educational specialist who serves a resource to other professionals may also help the child. Curriculum should be based on long-term goals rather than on inflexible credit schedules that teach worthless, unlearnable information and demoralize the student.

Finally, the education setting should provide opportunities for structured social interaction and less-structured but supervised—”bully-proofed”—interactions.

Medications. No systematic medical treatment data exist, as there is no way to classify these children. They are usually treated with multiple medications for their specific symptom cluster abnormalities (Table 4). Options include:

• atypical antipsychotics such as risperidone, quetiapine, aripiprazole, ziprasidone, or olanzapine
  
• mood stabilizers such as valproic acid, lithium, or lamotrigine
  
• stimulants such as methylphenidate, amphetamine salts, atomoxetine, or bupropion (a mild stimulant and an antidepressant)
  
• selective serotonin reuptake inhibitors, such as fluoxetine, sertraline, citalopram, paroxetine, or fluvoxamine.
  

Unfortunately, drug therapy may cause behavioral toxicity—tearfulness, irritability, disinhibition, activation, agitation, hallucinations and possibly even suicidal behavior. Stopping the medication usually reverses this kind of adverse effect.¹⁹

Medication side effects understandably frighten parents—who may be reluctant to have their children use any drug therapies. Counsel the parents in advance that side effects may occur.

Related resources

• Child/Adolescent Symptom Inventory. http://www.checkmateplus.com. Accessed Jan. 11, 2005.
  

Drug brand names

• Amphetamine • Adderall
  
• Aripiprazole • Abilify
  
• Atomoxetine • Strattera
  
• Bupropion • Wellbutrin
  
• Citalopram • Celexa
  
• Fluoxetine • Prozac
  
• Fluvoxamine • Luvox
  
• Lamotrigine • Lamictal
  
• Lithium carbonate • Lithobid, others
  
• Methylphenidate • Concerta, Ritalin
  
• Olanzapine • Zyprexa
  
• Paroxetine • Paxil
  
• Quetiapine • Seroquel
  
• Risperidone • Risperdal
  
• Sertraline • Zoloft
  
• Valproic acid • Depakote
  
• Ziprasidone • Geodon
  

Disclosures

Dr. Weisbrot receives grants from Pfizer Inc.

Dr. Carlson receives grants from or is a speaker for Janssen Pharmaceutica, Eli Lilly and Co., Shire Pharmaceuticals Groups, and Abbott Laboratories; is a consultant to Janssen Pharmaceutica and Eli Lilly and Co.; and is an advisor to Otsuka America Pharmaceutical, Pfizer Inc., and Ortho-McNeil Pharmaceutical.