The new england journal of medicine

clinical practice

Late-Life

Jürgen Unützer, M.D., M.P.H.

This Journal feature begins with a case vignette highlighting a common clinical problem. Evidence supporting various strategies is then presented, followed by a review of formal guidelines, when they exist. The article ends with the author’s clinical recommendations.

A 71-year-old man, whose wife died 6 months previously, presents with foot pain from diabetic neuropathy, poor sleep, lack of energy, and increasing frustration about his inability to “keep his diabetes under control.” On examination, he also notes lack of interest in usual activities, decreased appetite, a weight loss of 4.5 kg (10 lb) over the past 3 months, and intermittent thoughts that he would be better off dead. How should his case be managed?

The Clinical Problem

As many as 10% of adults 65 years of age or older who are seen in primary care From the Department of Psychiatry and settings have clinically significant depression.1 Depression is particularly common Behavioral Sciences, University of Wash- 2 ington, Seattle. Address reprint requests in women, in patients with chronic medical disorders or persistent , and to Dr. Unützer at Box 356560, Seattle, WA in patients who have experienced stressful life events (e.g., the loss of a spouse), 98195, or at [email protected]. functional decline, or .3 Criteria for the diagnosis of an episode of N Engl J Med 2007;357:2269-76. major depression are summarized in Table 1. Copyright © 2007 Massachusetts Medical Society. Late-life depression is often undetected or undertreated in primary care,4 espe- cially in men and members of racial and ethnic minority groups.5 Reasons for undertreatment include stigma associated with depression6 and the belief that depression is a normal part of aging.7 Patients and providers may correctly associ- ate depression with the loss of a loved one; however, if symptoms of major depres- sion persist for more than 2 months after a loss, treatment for depression should be strongly considered. Coexisting problems, such as chronic medical disorders, pain, cognitive impairment (which can be associated with depression or ), and alcohol or substance misuse, may also complicate the diagnosis and treatment of depression. Late-life depression that is untreated can last for years and is associated with a poor quality of life, difficulty with social and physical functioning, poor adherence to treatment, worsening of chronic medical problems,2 and increased morbidity and mortality from and other causes. Older men have the highest rates of completed suicide (with the use of firearms in most cases).8 Recognizing and treating depression9,10 and reducing access to firearms may be the most important things primary care providers can do to reduce the risk of suicide.

Strategies and Evidence

Screening The Patient Health Questionnaire 2 (PHQ-2), a two-item screening instrument that asks about depressed mood and (loss of interest and pleasure) in the previous 2 weeks, is easily administered by an office staff member or a physician during a primary care visit. This questionnaire is useful in identifying patients at

n engl j med 357;22 www.nejm.org november 29, 2007 2269 The New England Journal of Medicine Downloaded from nejm.org at UNIVERSITY OF VIRGINIA on June 26, 2012. For personal use only. No other uses without permission. Copyright © 2007 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

Table 1. Criteria for the Diagnosis of an Episode of Major Depression.*

Five or more of the following nine symptoms nearly every day during the same 2-wk period (with at least one of the symp- toms being either depressed mood or diminished interest or pleasure): Depressed mood most of the day nearly every day Markedly diminished interest or pleasure in all or almost all activities Clinically significant weight loss in the absence of dieting or weight gain (e.g., a change of more than 5% in body weight in a month) or a decrease in appetite Insomnia or hypersomnia Observable or retardation or loss of energy Feelings of worthlessness or excessive or inappropriate guilt Diminished ability to think or concentrate, or indecisiveness Recurrent thoughts of death, recurrent without a specific plan, a specific plan for committing suicide, or a suicide attempt

* These criteria are adapted from the Diagnostic and Statistical Manual of Mental Disorders, 4th edition. These symptoms cause significant distress or impairment in social, occupational, or other important areas of functioning, and they are not due to the direct physiological effects of a substance or a general medical condition (e.g., hypothyroidism). These symp- toms are not better accounted for by bereavement (i.e., after the loss of a loved one), and they persist for longer than 2 months or are characterized by marked functional impairment, morbid preoccupation with worthlessness, suicidal ideation, psychotic symptoms, or . If the patient has symptoms of depression as well as a his- tory of mania or symptoms of mania, he or she may not have a diagnosis of depression but may have .

high risk for depression, and it has a sensitivity of things that others thought were excessive or risky, 100%, a specificity of 77%, and a positive predic- or spending too much money.13 tive value of 14%.11 (The full questionnaire is avail- Clinicians should also obtain a medical history able at www.aafp.org/afp/20040915/1101.html.) and perform a physical examination and labora- Positive results should prompt a clinical evalua- tory tests as clinically indicated to assess medical tion for major depression. conditions or medications that may be contribut- ing to depression. Measurement of thyrotropin Evaluation levels is recommended in patients with symptoms Evaluation of the patient should include a review of hypothyroidism (e.g., fatigue, weight gain, or of the nine symptoms of major depression (Ta- cold intolerance), but the value of routine thyro- ble 1), including thoughts of suicide. This evalu- tropin screening in patients with depression is ation can be facilitated with the use of a brief, not well established.14 Additional screening is rec- nine-item self-rating scale called the Patient Health ommended for patients with cognitive impair- Questionnaire 9 (PHQ-9).12 (The full questionnaire ment or alcohol or substance misuse (including is available at www.depression-primarycare.org/ misuse of prescription drugs) because these con- clinicians/toolkits/materials/forms/phq9/.) Clini- ditions can complicate the management of de- cians should assess the duration of the patient’s pression. current depressive episode, associated functional impairment, and history of and treatment for de- Management pression. Effective treatment of late-life depression has been Providers should ask whether the patient has associated with improved emotional, social, and a history of bipolar disorder, or manic depression, physical functioning and quality of life. It has also which may be misdiagnosed as unipolar depres- been associated with better self-care for chronic sion. Screening can be facilitated by the use of medical conditions and reduced mortality.9,10,15,16 a brief questionnaire called the Mood Disorder In primary care, medications Questionnaire, which asks about symptoms and are the most commonly used treatments for major behaviors suggestive of mania. Screening ques- depression, but there are several other evidence- tions for symptoms of mania include questions based treatments, including structured psycho- about periods of excess energy or talkativeness, therapies.17 A recent meta-analysis of random- racing thoughts, being much more active than ized, controlled trials indicated that antidepressant usual, needing much less sleep than usual, doing medications and structured have

2270 n engl j med 357;22 www.nejm.org november 29, 2007 The New England Journal of Medicine Downloaded from nejm.org at UNIVERSITY OF VIRGINIA on June 26, 2012. For personal use only. No other uses without permission. Copyright © 2007 Massachusetts Medical Society. All rights reserved. clinical practice roughly equivalent efficacies in older adults.18 participants have a substantial reduction in symp- A combination of pharmacotherapy and psycho- toms of depression as compared with 25 to 30% therapy is recommended for severe or chronic of controls.17,23 A physical-exercise program could forms of depression.19,20 Other evidence-based be a first-line strategy for patients with mild-to- treatments include electroconvulsive therapy (ECT) moderate depression who prefer this approach, and physical-exercise programs. Specific thera- but it may be difficult for patients with depression pies are discussed below. to engage in such a program, and additional treat- Treatment plans for late-life depression should ment with or may take into account the patient’s preferences, treat- be needed. ment history (focusing on treatments that have been helpful in the past), and coexisting medical Pharmacologic Management and psychiatric conditions. Treatment availability More than 20 antidepressants have been approved should also be considered. Before initiating treat- by the Food and Drug Administration (FDA) for ment, clinicians should address common patient the treatment of depression in older adults. Selec- concerns about side effects. They should reassure tive serotonin-reuptake inhibitors (SSRIs) are most patients that dependence is not a realistic con- often used as first-line treatments (Table 2). The cern with antidepressant medications and that most common side effect is gastrointestinal irrita- medications will not inhibit normal emotional tion (dyspepsia or nausea), which usually resolves reactions such as bereavement.21 Careful listen- within 7 to 10 days. Serotonin–norepinephrine ing, education, and reassurance can help address reuptake inhibitors (SNRIs) may be particularly such concerns. useful for patients with coexisting pain, particu- larly if it is neuropathic. Side effects include nau- Psychotherapy sea, agitation, insomnia, and hypertension, es- Several forms of psychotherapy have been shown pecially at high doses. Although head-to-head in randomized, controlled trials to be effective for comparisons are scarce, some research suggests late-life depression, including cognitive behavioral that SNRIs may be less well tolerated by frail older therapy (which helps patients correct negative adults than SSRIs.26 , an antidepres- thoughts associated with depression), interperson- sant with serotonergic and noradrenergic proper- al psychotherapy (which focuses on interpersonal ties, is associated with sedation, increased appetite, causes of depression), and problem-solving ther- and weight gain; thus, it may be particularly useful apy (which helps patients learn strategies for solv- for patients with insomnia or weight loss. Bupro- ing everyday problems associated with depres- pion may cause jitteriness and insomnia, and it sion).22 Such structured psychotherapies, which may be particularly useful in patients with leth- can be delivered by trained therapists in 6 to 12 argy, daytime sedation, or fatigue. Neither mir- sessions in mental health or primary care settings, tazapine nor has sexual side effects. should be strongly considered if antidepressant Trazodone is not recommended as a primary an- treatment is not preferred or not effective in a pa- tidepressant because of sedation and orthostatic tient. The efficacy of such evidence-based psycho- hypotension at therapeutic doses, but in low doses therapies is roughly similar to that of antidepres- (e.g., 25 to 50 mg) it is useful for insomnia as- sant medications, with 45 to 70% of patients sociated with depression.27 Priapism is a rare but treated with psychotherapy having substantial im- potentially serious side effect.28 provement in depression (at least a 50% reduction Tricyclic antidepressants are effective but are in symptoms of depression) as compared with 25 no longer considered to be first-line treatments to 35% of controls.18,22 because of their side effects (Table 2) and be- cause of cardiotoxic effects in patients who take Exercise Programs an overdose. They should be considered in pa- Several randomized, controlled trials suggest that tients who have previously had a good response short-term (e.g., 12-week), supervised, group-based to tricyclic antidepressants or who have depres- physical-exercise programs involving walking or sion that does not improve with other antidepres- other forms of aerobic exercise can reduce de- sants. Tricyclic antidepressants are contraindicat- pression in older adults; 45 to 65% of program ed in patients with a recent history (e.g., within

n engl j med 357;22 www.nejm.org november 29, 2007 2271 The New England Journal of Medicine Downloaded from nejm.org at UNIVERSITY OF VIRGINIA on June 26, 2012. For personal use only. No other uses without permission. Copyright © 2007 Massachusetts Medical Society. All rights reserved. The new england journal of medicine ------Comments bined with certain drugs§ certain with bined com if syndrome serotonin of risk drugs§ certain with bined seizures for risk creased seizures for risk creased Risk of if com if syndrome serotonin of Risk interactions drug–drug few Relatively interactions drug–drug few Relatively interactions, drug–drug few Relatively coated) (enteric broken be not Should interactions drug–drug few Relatively in at patients in Contraindicated in at patients in Contraindicated - -

Side Effects sexual dysfunction, jitteriness, insomnia, insomnia, jitteriness, dysfunction, sexual hyponatremia‡ hyponatremia‡ sexual somnolence, insomnia, itation, dysfunction gain weight or effects side gain weight or effects side or vision, blurry constipation, retention, impairment hypotension, thostatic conduction cardiac of Nausea, dyspepsia, , tremors, anxiety, anxiety, tremors, anorexia, dyspepsia, Nausea, diarrhea stools, Loose gain, weight fatigue, drowsiness, mouth, Dry ag dizziness, mouth, dry sweating, Nausea, Hypertension Hypertension effects side sexual no gain, weight Sedation, sexual no jitteriness; agitation, Insomnia, sexual no jitteriness; agitation, Insomnia, urinary mouth, dry gain, weight Sedation, Fatigue agitation Insomnia, Long Short Short Short Short Short Short Short Short Short Short Half-Life† Ultrashort Ultrashort 25 daily Therapeutic Dose 10–60 mg once daily once mg 10–60 daily once mg 10–30 20–60 mg once daily once mg 20–60 daily once mg 20–50 daily once mg 20–60 15–45 mg every night every mg 15–45 50–200 mg once daily once mg 50–200 daily once mg 75–300 25–150 mg twice daily twice mg 25–150 75–125 mg every night every mg 75–125 100–200 mg once daily once mg 100–200 100–150 mg twice daily twice mg 100–150 75–150 mg twice or thrice or twice mg 75–150 hyponatremia developed in 9% of older adults in whom paroxetine treatment was initiated. Hyponatremia appeared to be particularly problematic Starting Dose 24 10 mg once daily once mg 10 daily once mg 10 daily once mg 10 daily once mg 10 25 mg once daily once mg 25 daily once mg 25 daily once mg 30 daily once mg 75 daily once mg 25 15 mg every night every mg 15 night every mg 10 100 mg once daily once mg 100 37.5 mg once daily once mg 37.5 Depression.* Major for Antidepressants Used Commonly 2. Table Medication SSRIs (Prozac)¶ Fluoxetine (Zoloft)¶ Sertraline (Celexa)¶ Citalopram (Lexapro) Escitalopram (Paxil)¶ Paroxetine SNRIs (Effexor)¶ Venlafaxine XR) (Effexor XR Venlafaxine (Cymbalta) Duloxetine antidepressants newer Other (Remeron)¶ Mirtazapine (Wellbutrin)¶ Bupropion SR) (Wellbutrin SR Bupropion antidepressants ∥ Tricyclic (Pamelor)¶ Nortriptyline (Norpramin)¶ Desipramine Short-acting antidepressants (particularly paroxetine) have been associated with an influenza-like discontinuation syndrome with symptoms such as insomnia, nausea, imbalance, sen sory disturbances, and hyperarousal when the drug is discontinued abruptly or tapered too rapidly. The symptoms are usually mild and last 1 to 2 weeks, and they can be minimized by slowly tapering the antidepressant (e.g., over a 2-to-4-week period). In a trial reported by Fabian et al., for patients with a lower body-mass index and a baseline plasma sodium level of less than 138 mmol/liter. The serotonin syndrome is a potentially lethal complication manifested as altered mental status, myoclonus, tremors, hyperreflexia, fever, and autonomic changes. This syndrome is as sociated with combinations of serotonergic antidepressants (SSRIs or SNRIs) with certain drugs (e.g., meperidine, tricyclic antidepressants, sumatriptan, tramadol, linezolid, and monoamine oxidase inhibitors). Data are from Boyer and Shannon. cardiac infarction, myocardial of 2 weeks) within (e.g., history a recent with patients in contraindicated are antidepressants Tricyclic imipramine). and doxepin, amitriptyline, (e.g., pressants reported conduction been defects, has orthostatic which hypotension, urinary retention, prostatic hypertrophy, nortriptyline, cognitive for impairment, or exists narrow-angle level glaucoma. Measurement blood of serum levels a therapeutic of for tricyclic antidepres evidence best The effects. side excessive or depression persistent for patients evaluate to useful be may sants to have a therapeutic window of 50 to 150 ng per milliliter. An electrocardiogram should be obtained before initiation of treatment and after dose increases with tricyclic antidepressants. Generic preparations of this medication are available. Secondary amine tricyclic antidepressants (e.g., nortriptyline and desipramine) are preferred in older adults because of a lower side-effect burden than that of tertiary amine tricyclic antide The list of medications is not comprehensive. SSRIs denotes selective serotonin-reuptake inhibitors, and SNRIs serotonin–norepinephrine reuptake inhibitors.

§ † ‡ ¶ ∥ *

2272 n engl j med 357;22 www.nejm.org november 29, 2007 The New England Journal of Medicine Downloaded from nejm.org at UNIVERSITY OF VIRGINIA on June 26, 2012. For personal use only. No other uses without permission. Copyright © 2007 Massachusetts Medical Society. All rights reserved. clinical practice

2 weeks) of myocardial infarction, cardiac con- older adults often require full adult doses for an duction defects, orthostatic hypotension, narrow- adequate response (Table 2). angle glaucoma, urinary retention, prostatic hy- Recent FDA analyses suggest that antide- pertrophy, or cognitive impairment. Monoamine pressant use may increase the risk of suicidal oxidase inhibitors have a narrow therapeutic in- thoughts in youths and adults younger than 25 dex and require special dietary and medication years of age. However, these drugs have a neutral restrictions. Their use, which is generally limited or protective effect against suicidal ideation or to patients in whom other antidepressants have behavior in older adults.36 failed or in patients who have had a previous re- Antidepressant treatment should be continued sponse to this class of drugs, should involve con- at full doses for at least 6 to 12 months after sultation with a physician who is experienced in patients are in remission because recurrence rates prescribing them. after earlier discontinuation are as high as 70%. There have been few head-to-head comparison In controlled trials involving older adults with studies of various antidepressants in older adults, depression who had a response to antidepressant but data from randomized, placebo-controlled treatment, patients who were randomly assigned trials of various agents in older adults suggest to receive such continuation treatment had a 60% similar efficacies.29 However, as compared with reduction in the risk of recurrence as compared SSRIs, tricyclic antidepressants may be associat- with patients who received placebo after discon- ed with higher dropout rates due to side effects.30 tinuation of antidepressant therapy.37,38 In addition to side-effect profiles, other consider- ations in the selection of antidepressants include Electroconvulsive Therapy the patient’s response to previous treatment, the Several randomized, controlled trials have estab- potential for drug interactions, the frequency of lished the efficacy of ECT for severe late-life de- dosing, the safety of the drug in overdose, and pression, with efficacy rates ranging from 60 to cost. The consideration of responses to treatment 80%.39-42 ECT is particularly indicated for pa- in close relatives with depression may also be tients with depression that is resistant to other helpful predicting a patient’s responsiveness, al- treatments and for patients at risk for serious though this correlation has not been well studied. harm because of psychotic depression, suicidal One commonly used approach involves initial ideation, or severe malnutrition. ECT is usually treatment with an SSRI, with a switch to a differ- administered as a series of 6 to 12 treatments in ent class according to the patient’s symptoms an inpatient psychiatric setting over a period of and the side-effect profile of the drug if the SSRI 2 to 4 weeks. Common side effects include head- is not effective or is poorly tolerated.31 ache that usually responds to analgesics and tem- Up to 12 weeks of treatment with antidepres- porary confusion or impairment. Less sant medications may be needed to elicit a full common side effects include memory loss for response. However, a recent study suggests that events during the period surrounding treatment a full response is expected in two thirds of pa- and falls immediately after treatment sessions. tients who have partial improvement after 4 weeks The mortality associated with ECT is less than of treatment, as compared with about one third 1 death in 10,000 patients. A successful course of of patients without a response at 4 weeks.32 Even ECT should be followed by maintenance phar- under the best of circumstances, only 40 to 65% macologic treatment because of high rates of of patients have an adequate response to any relapse. In a randomized trial involving patients given antidepressant, and trials of alternative with depression that had improved after ECT, antidepressants or combinations of antidepres- 6-month relapse rates were 84% among patients sants, with or without psychotherapy, are required receiving placebo, 60% among patients receiving in a substantial number of patients.33-35 nortriptyline, and 39% among patients receiving Antidepressant monotherapy is preferred in lithium plus nortriptyline.43 order to minimize side effects and drug interac- tions, reduce out-of-pocket costs, and enhance Treatment Follow-Up the likelihood of treatment adherence. To mini- During the initial 8 to 10 weeks of pharmaco- mize side effects, starting doses for older adults logic or nonpharmacologic treatments for depres- may be lower than those for younger adults, but sion, patients should be followed closely (e.g.,

n engl j med 357;22 www.nejm.org november 29, 2007 2273 The New England Journal of Medicine Downloaded from nejm.org at UNIVERSITY OF VIRGINIA on June 26, 2012. For personal use only. No other uses without permission. Copyright © 2007 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

weekly or every other week, either in person or by with other medications or psychotherapy, are in- telephone) for side effects, drug interactions, and dicated. A recent trial involving older adults with worsening of depression (e.g., psychotic or manic depression showed that 50% of patients with an symptoms or suicidal ideation) and in order to incomplete response to the SSRI paroxetine had a adjust treatment as needed. Such close follow-up full response when a second antidepressant was during this initial treatment phase may reduce the added.35 ECT has been shown to be efficacious in rates of premature discontinuation of antidepres- patients in whom other treatments have failed, sant medication; such discontinuation has been but its use is frequently not considered in primary reported in up to 50% of patients within 4 weeks care. More research comparing ECT with other after initiating treatment. The use of rating scales active treatments is needed. Other strategies that such as the PHQ-944 can facilitate monitoring, may improve depression include aggressively treat- with the goal of treatment being a complete re- ing coexisting conditions such as pain48 and ad- mission of depression (e.g., a PHQ-9 score of <5 on dressing psychosocial stressors that contribute to a scale of 0 to 27, with higher scores indicating depression. However, such strategies have not been more depression). carefully studied. Consultation with a mental health specialist is More research is needed to guide the treat- recommended for patients who prefer nonphar- ment of older patients with depression and as- macologic treatments such as psychotherapy or sociated cognitive impairment or dementia. In the who have persistent depression after one or more absence of definitive evidence, many clinicians trials of antidepressants. Other reasons for re- prefer a stepwise treatment approach that starts ferral include , a history of mania or the with a trial of an antidepressant before consid- emergence of manic symptoms during treatment ering the addition of an antidementia agent such for depression, and concern about suicide. as a cholinesterase inhibitor or memantine.49 Several randomized trials have shown that, as More research is also needed regarding improved compared with usual care, systematic programs access to treatment for older adults who are at of care management for depression can signifi- high risk for undertreatment; these patients cantly increase patient and provider satisfaction include elderly men and members of minority and the effectiveness of treatment.9,10,15,45 In these groups. programs, a care manager (usually a nurse or social worker) supports the treating physician by Guidelines providing the patient with information about de- pression, proactively tracking depression with the Although they were not specifically developed for use of a scale such as the PHQ-9, monitoring treat- older adults, the Agency for Health Care Policy ment adherence and side effects, providing brief and Research guidelines for the treatment of de- evidence-based psychotherapy, and facilitating pression in primary care, which were first pub- consultation with a psychiatrist. lished in 199350 and were updated in 1998,51 are largely applicable to the treatment of late-life de- Areas of Uncertainty pression. A recent British practice guideline52 and several consensus statements53-55 focus on the There is controversy about the effectiveness of treatment of depression in older adults. The rec- psychotherapies and antidepressant medications ommendations in this article are generally con- in patients with depression that does not meet the cordant with those guidelines. full diagnostic criteria for major depression.46,47 Watchful waiting may be appropriate for such pa- Conclusions tients as long as the depression is carefully and Recommendations tracked and treatment is initiated if symptoms worsen. The patient described in the vignette has symp- The optimal care of patients with depression toms that are typical of major depression, although that is resistant to one or more antidepressants other causes of his depressive symptoms (e.g., remains uncertain. Studies involving younger colon as a potential cause of weight loss) adults33,34 suggest that several trials of antide- should also be investigated. Untreated, he would pressant medications, alone or in combination be at substantial risk for poor functioning, poor

2274 n engl j med 357;22 www.nejm.org november 29, 2007 The New England Journal of Medicine Downloaded from nejm.org at UNIVERSITY OF VIRGINIA on June 26, 2012. For personal use only. No other uses without permission. Copyright © 2007 Massachusetts Medical Society. All rights reserved. clinical practice self-care, worsening of his chronic medical prob- weeks, and those who are at high risk for suicide, lems,2 and suicide. who have had previous mania or manic symp- Initial treatment may involve either a trial of toms while receiving antidepressant medications, an antidepressant or psychotherapy (e.g., inter- or who have psychotic symptoms should be re- personal therapy); the choice should be guided ferred for psychiatric consultation. For patients by the patient’s preferences, his treatment his- with depression that has improved, maintenance tory, and the costs and availability of treatments. treatment for at least 6 to 12 months should be If medication is used, I would generally start with considered in order to reduce the risk of relapse. a low dose of an SSRI and increase the dose Educational information about late-life depres- gradually over the next 2 to 4 weeks to a thera- sion is available from the National Institute of Men- peutic dose, as tolerated (Table 2). The patient tal Health (www.nimh.nih.gov/HealthInformation/ should be closely monitored for adherence to and Depressionmenu.cfm), the American Association the effectiveness and adverse effects of treatment, for Geriatric Psychiatry (www.aagpgpa.org), and and the medication should be adjusted if the de- the IMPACT Program for Late-Life Depression pression does not show improvement after 2 to (http://impact-uw.org). 4 weeks of treatment. Dr. Unützer reports receiving fees for a continuing medical education symposium sponsored by an unrestricted grant from Patients with depression that persists after one Eli Lilly. No other potential conflict of interest relevant to this or more trials of medication, each lasting 8 to 12 article was reported.

References 1. Lyness JM, Caine ED, King DA, Cox C, 11. Li C, Friedman B, Conwell Y, Fiscella K. nitive behavioral-analysis system of psy- Yoediono Z. Psychiatric disorders in older Validity of the Patient Health Question- chotherapy, and their combination for the primary care patients. J Gen Intern Med naire 2 (PHQ-2) in identifying major de- treatment of chronic depression. N Engl J 1999;14:249-54. pression in older people. J Am Geriatr Soc Med 2000;342:1462-70. [Erratum, N Engl 2. Katon WJ. Clinical and health services 2007;55:596-602. J Med 2001;345:232.] relationships between major depression, 12. Kroenke K, Spitzer RL, Williams JB. 20. Reynolds CF III, Frank E, Perel JM, et depressive symptoms, and general medi- The PHQ-9: validity of a brief depression al. Nortriptyline and interpersonal psycho- cal illness. Biol Psychiatry 2003;54:216-26. severity measure. J Gen Intern Med 2001; therapy as maintenance therapies for re- 3. Bruce ML. Psychosocial risk factors 16:606-13. current major depression: a randomized for depressive disorders in late life. Biol 13. Hirschfeld RM, Williams JB, Spitzer controlled trial in patients older than 59 Psychiatry 2002;52:175-84. RL, et al. Development and validation of years. JAMA 1999;281:39-45. 4. Klap R, Unroe KT, Unutzer J. Caring a screening instrument for bipolar spec- 21. Givens JL, Datto CJ, Ruckdeschel K, et for mental illness in the United States: trum disorder: the Mood Disorder Ques- al. Older patients’ aversion to antidepres- a focus on older adults. Am J Geriatr Psy- tionnaire. Am J Psychiatry 2000;157:1873- sants: a qualitative study. J Gen Intern chiatry 2003;11:517-24. 5. Med 2006;21:146-51. 5. Unützer J, Katon W, Callahan CM, et 14. Fraser SA, Kroenke K, Callahan CM, 22. Cuijpers P, van Straten A, Smit F. Psy- al. Depression treatment in a sample of Hui SL, Williams JW Jr, Unützer J. Low chological treatment of late-life depres- 1,801 depressed older adults in primary yield of thyroid-stimulating hormone test- sion: a meta-analysis of randomized con- care. J Am Geriatr Soc 2003;51:505-14. ing in elderly patients with depression. trolled trials. Int J Geriatr Psychiatry 6. Sirey JA, Bruce ML, Alexopoulos GS, Gen Hosp Psychiatry 2004;26:302-9. 2006;21:1139-49. Perlick DA, Friedman SJ, Meyers BS. Stigma 15. Unützer J, Katon W, Callahan CM, et 23. Sjösten N, Kivelä S-L. The effects of as a barrier to recovery: perceived stigma al. Collaborative care management of late- physical exercise on depressive symptoms and patient-rated severity of illness as pre- life depression in the primary care set- among the aged: a systematic review. Int J dictors of antidepressant drug adherence. ting: a randomized controlled trial. JAMA Geriatr Psychiatry 2006;21:410-8. Psychiatr Serv 2001;52:1615-20. 2002;288:2836-45. 24. Fabian TJ, Amico JA, Kroboth PD, et al. 7. Sarkisian CA, Lee-Henderson MH, 16. Gallo JJ, Bogner HR, Morales KH, Paroxetine-induced hyponatremia in older Mangione CM. Do depressed older adults Post EP, Lin JY, Bruce ML. The effect of adults: a 12-week prospective study. Arch who attribute depression to “old age” be- a primary care practice-based depression Intern Med 2004;164:327-32. lieve it is important to seek care? J Gen intervention on mortality in older adults: 25. Boyer EW, Shannon M. The serotonin Intern Med 2003;18:1001-5. a randomized trial. Ann Intern Med 2007; syndrome. N Engl J Med 2005;352:1112-20. 8. National strategy for suicide preven- 146:689-98. [Erratum, N Engl J Med 2007;356:2437.] tion: goals and objectives for action. Rock- 17. Frazer CJ, Christensen H, Griffiths 26. Oslin DW, Ten Have TR, Streim JE, et ville, MD: Public Health Service, 2001. KM. Effectiveness of treatments for de- al. Probing the safety of medications in 9. Unutzer J, Tang L, Oishi S, et al. Re- pression in older people. Med J Aust 2005; the frail elderly: evidence from a random- ducing suicidal ideation in depressed older 182:627-32. ized clinical trial of sertraline and venla- primary care patients. J Am Geriatr Soc 18. Pinquart M, Duberstein PR, Lyness JM. faxine in depressed nursing home resi- 2006;54:1550-6. Treatments for later-life depressive con- dents. J Clin Psychiatry 2003;64:875-82. 10. Bruce ML, Ten Have TR, Reynolds CF ditions: a meta-analytic comparison of 27. Nierenberg AA, Adler LA, Peselow E, III, et al. Reducing suicidal ideation and pharmacotherapy and psychotherapy. Am Zornberg G, Rosenthal M. Trazodone for depressive symptoms in depressed older J Psychiatry 2006;163:1493-501. antidepressant-associated insomnia. Am primary care patients: a randomized con- 19. Keller MB, McCullough JP, Klein DN, J Psychiatry 1994;151:1069-72. trolled trial. JAMA 2004;291:1081-91. et al. A comparison of nefazodone, the cog- 28. Spina E, Scordo MG. Clinically sig-

n engl j med 357;22 www.nejm.org november 29, 2007 2275 The New England Journal of Medicine Downloaded from nejm.org at UNIVERSITY OF VIRGINIA on June 26, 2012. For personal use only. No other uses without permission. Copyright © 2007 Massachusetts Medical Society. All rights reserved. clinical practice

nificant drug interactions with antidepres- tients: placebo-controlled study of main- K, Swick H. Watchful waiting for minor sants in the elderly. Drugs Aging 2002;19: tenance therapy. Br J Psychiatry 2002;181: depression in primary care: remission 299-320. 29-35. rates and predictors of improvement. Gen 29. Mottram P, Wilson K, Strobl J. Antide- 39. van der Wurff FB, Stek ML, Hoogen- Hosp Psychiatry 2006;28:205-12. pressants for depressed elderly. Cochrane dijk WJ, Beekman AT. The efficacy and 48. Thielke SM, Fan MY, Sullivan M, Unütz- Database Syst Rev 2006;1:CD003491. safety of ECT in depressed older adults: er J. Pain limits the effectiveness of col- 30. Wilson K, Mottram P. A comparison of a literature review. Int J Geriatr Psychiatry laborative care for depression. Am J Geri- side effects of selective serotonin reuptake 2003;18:894-904. atr Psychiatry 2007;15:699-707. inhibitors and tricyclic antidepressants in 40. Sackeim HA. Electroconvulsive ther- 49. Steffens DC, Potter GG. Geriatric de- older depressed patients: a meta-analysis. apy in late-life depression. In: Salzman C, pression and cognitive impairment. Psy- Int J Geriatr Psychiatry 2004;19:754-62. ed. Clinical geriatric psychopharmacology. chol Med (in press). 31. Kennedy GJ, Marcus P. Use of antide- 4th ed. Baltimore: Lippincott Williams & 50. Depression Guideline Panel. Depres- pressants in older patients with co-morbid Wilkins, 2004:385. sion in primary care: detection and diag- medical conditions: guidance from studies 41. Kujala I, Rosenvinge B, Bekkelund SI. nosis. Clinical practice guideline. No. 5. of depression in somatic illness. Drugs Ag- Clinical outcome and adverse effects of Rockville, MD: Agency for Healthcare Pol- ing 2005;22:273-87. electroconvulsive therapy in elderly psychi- icy and Research, 1993. (AHCPR publica- 32. Mulsant BH, Houck PR, Gildengers atric patients. J Geriatr Psychiatry Neurol tion no. 93-0550.) AG, et al. What is the optimal duration of 2002;15:73-6. 51. Schulberg HC, Katon W, Simon GE, a short-term antidepressant trial when 42. Dombrovski AY, Mulsant BH. The evi- Rush AJ. Treating major depression in pri- treating geriatric depression? J Clin Psy- dence for electroconvulsive therapy (ECT) mary care practice: an update of the chopharmacol 2006;26:113-20. in the treatment of severe late-life depres- Agency for Health Care Policy and Re- 33. Rush AJ, Trivedi MH, Wisniewski SR, sion: ECT: the preferred treatment for se- search Practice Guidelines. Arch Gen Psy- et al. Bupropion-SR, sertraline, or venla- vere depression in late life. Int Psychoge- chiatry 1998;55:1121-7. faxine-XR after failure of SSRIs for depres- riatr 2007;19:10-4, 24-35. 52. Baldwin RC, Anderson D, Black S, et sion. N Engl J Med 2006;354:1231-42. 43. Sackeim HA, Haskett RF, Mulsant BH, al. Guideline for the management of late- 34. Trivedi MH, Fava M, Wisniewski SR, et al. Continuation pharmacotherapy in life depression in primary care. Int J Geri- et al. Medication augmentation after the the prevention of relapse following elec- atr Psychiatry 2003;18:829-38. failure of SSRIs for depression. N Engl J troconvulsive therapy: a randomized con- 53. Lebowitz BD, Pearson JL, Schneider Med 2006;354:1243-52. trolled trial. JAMA 2001;285:1299-307. LS, et al. Diagnosis and treatment of de- 35. Dew MA, Whyte EM, Lenze EJ, et al. 44. Löwe B, Unützer J, Callahan CM, Per- pression in late life: consensus statement Recovery from major depression in older kins AJ, Kroenke K. Monitoring depres- update. JAMA 1997;278:1186-90. adults receiving augmentation of antide- sion treatment outcomes with the Patient 54. Alexopoulos GS, Katz IR, Reynolds CF pressant pharmacotherapy. Am J Psychia- Health Questionnaire-9. Med Care 2004; III, Carpenter D, Docherty JP, Ross RW. try 2007;164:892-9. 42:1194-201. Pharmacotherapy of depression in older 36. Kuehn BM. FDA panel seeks to bal- 45. Oxman TE, Dietrich AJ, Schulberg HC. patients: a summary of the expert consen- ance risks in warnings for antidepressants. Evidence-based models of integrated man- sus guidelines. J Psychiatr Pract 2001;7: JAMA 2007;297:573-4. agement of depression in primary care. 361-76. 37. Reynolds CF III, Dew MA, Pollock BG, Psychiatr Clin North Am 2005;28:1061-77. 55. Charney DS, Reynolds CF III, Lewis L, et al. Maintenance treatment of major de- 46. Lyness JM, Heo M, Datto CJ, et al. Out- et al. Depression and Bipolar Support Al- pression in old age. N Engl J Med 2006; comes of minor and subsyndromal de- liance consensus statement on the unmet 354:1130-8. pression among elderly patients in primary needs in diagnosis and treatment of mood 38. Klysner R, Bent-Hansen J, Hansen HL, care settings. Ann Intern Med 2006;144: disorders in late life. Arch Gen Psychiatry et al. Efficacy of citalopram in the preven- 496-504. 2003;60:664-72. tion of recurrent depression in elderly pa- 47. Hegel MT, Oxman TE, Hull JG, Swain Copyright © 2007 Massachusetts Medical Society.

COLLECTIONS OF ARTICLES ON THE JOURNAL’S WEB SITE The Journal’s Web site (www.nejm.org) sorts published articles into more than 50 distinct clinical collections, which can be used as convenient entry points to clinical content. In each collection, articles are cited in reverse chronologic order, with the most recent first.

2276 n engl j med 357;22 www.nejm.org november 29, 2007 The New England Journal of Medicine Downloaded from nejm.org at UNIVERSITY OF VIRGINIA on June 26, 2012. For personal use only. No other uses without permission. Copyright © 2007 Massachusetts Medical Society. All rights reserved.