Meeting the Challenge of Depression in HIV

Both doctors and patients often assume that people with serious and progressive illnesses are likely to be clinically depressed, especially when they become very ill. However, findings from both cross-sectional and longitudinal studies of people with HIV illness contradict the notion that clinical depression is an inevitable or even likely concomitant of potentially fatal illness. Depressive disorder is not "only to be expected," and it is not common at any stage of HIV illness.

With the advent of protease inhibitors, combination therapy and the sense of hope for "a second life," many people who thought their time was limited are beginning to think about issues they never expected to consider: returning to work; managing their depleted finances; reconsidering current relationships; housing; and most of all, deciding whether to return to their prior life activities or to make significant changes in their life trajectory. Others, who may have spent much of their lives unemployed and who are now receiving multiple government benefits, may eventually encounter bureaucratic expectations that they give up their benefits and seek employment. While these changes clearly are exciting, they also can provoke considerable conflict, anxiety and uncertainty, particularly in the absence of scientific data to assure long-term efficacy of the new treatments. Whether these psychological ramifications are best handled in newly formed support groups, or privately with a psychotherapist or vocational counselor, depends on individual preferences and financial options.

There will be others whose hopes are raised but who fail to achieve or maintain benefit from the new combination therapies. (See lead article of this issue.) Mental health interventions may be particularly useful to help address failed hopes and renewed apprehensions about illness progression.

Diagnosing Depression

The term "depression" is used to describe various conditions, ranging from transient blue moods to clinical psychiatric disorders. Everyone feels sad now and then, and in the midst of the AIDS epidemic, there are many kinds of losses that induce unhappiness, grief and heartbreak. A defining characteristic of clinical disorder is persistence of sadness, as well as a clustering of characteristic-related problems (symptoms).

Ordinary sadness may last for days at a time, but one retains the capacity for positive feelings as well. When the duration is weeks, feeling sad most of the day nearly every day, and is accompanied by other changes such as social withdrawal and suspension of usual activites, this is beyond the limits of "normal." Clinical depression is defined as a cluster of symptoms that occurs together for an extended period of time (at least two weeks). The major forms of clinical depressive disorder are major depression, which can be episodic but recurrent, and "dysthymia," which is the milder, chronic condition.

The diagnosis of major depression, as currently defined by psychiatrists,1 requires the presence of at least five of nine specific symptoms (see below) occurring together, including one of the first two. They must occur most of the day, nearly every day, during a period lasting two weeks or more:

  1. Depressed mood.
  2. Markedly diminished interest or pleasure in all, or almost all, activities.
  3. Significant unintentional weight loss (or gain).
  4. Insomnia (or oversleeping).
  5. Psychomotor retardation or agitation observable by others.
  6. Fatigue or loss of energy.
  7. Feelings of worthlessness or excessive or inappropriate guilt.
  8. Diminished ability to think or concentrate.
  9. Recurrent thoughts of death, suicidal ideation or attempt, or a specific plan for committing suicide.

These symptoms should "count" toward a depressive diagnosis except when they are clearly and fully accounted for by medical illness (e.g., weight loss associated with cryptosporidiosis-related diarrhea).

The milder depressive condition known as "dysthymia" is technically defined as presence of sad mood most of the day, more days than not, during the past two years. Dysthymia is also accompanied by two or more of the following symptoms (not clearly due to a concurrent medical condition): poor appetite, disrupted sleep, fatigue, low self-esteem, poor concentration, and/or feelings of hopelessness.

Although diagnosing depression is usually straightforward with HIV-positive people, diagnosis may be complicated at times. Some criteria for diagnosing depression such as loss of sexual desire, loss of appetite, insomnia, slowed movements and impaired concentration, low energy and fatigue, are also caused by HIV-related illnesses, HIV-related pain, or the side effects of HIV medications. This problem is partially addressed by making a distinction between the "somatic" symptoms of depression listed above, and "cognitive" or "affective" symptoms such as pervasive loss of interest in formerly enjoyable activities, loss of self-esteem and thoughts that one is a bad person, and the feeling that life is not worth living, with recurrent thoughts of suicide. (Note that the wish to die, expressed by a patient in a terminal illness episode, is not necessarily included here.) These symptoms, occurring together, are more clearly indices of depressive disorder and cannot be accounted for by the presence of medical conditions.

Even "psychological" symptoms like loss of interest in usually pleasurable activities may at times be difficult to evaluate in advanced illness as when someone has been hospitalized for weeks, or is housebound with no money, few visitors, and nothing to do. Still, it is possible with focused queries to identify depressed mood or loss of interest: Does the person answer the telephone? Open the mail? Can they think of things that they'd like to do if they had the resources and physical capacity? With advanced illness, symptoms can usually be clearly attributed to HIV in debilitated, apathetic patients who deny depression, whereas depressed mood stands out in a minority of very sick patients who are also depressed. Experienced clinicians familiar with both clinical depression and the varieties of HIV illnesses and medications can reliably diagnose depression in the context of symptomatic HIV illness. Experts in only one domain may have difficulty with the diagnostic challenge.

Although a diagnosis of depressive disorder is not given if the disturbance is an "expectable response" to the death of a loved one, protracted distress may warrant treatment just the same. This may take the form of counseling, bereavement groups, or antidepressant medication or medication for insomnia which may occur during such times.

The absence of a clinical depressive disorder does not preclude the experience of unhappiness, anger, grief or diminished future expectations. These negative states are, however, often accompanied or intermingled with positive emotions, a sense of purpose and "fighting spirit." The distinction between "normative" negative emotions and psychiatric disorders is important, both in understanding behavior and in attempting to alleviate distress on one's own or with professional intervention.

Prevalence of Depression in HIV-Infected People

Several community-based studies of the natural history of HIV illness and associated psychological adjustment were initiated in the mid- to late-1980s. Investigators conducted formal clinical diagnostic evaluations of a range of psychiatric disorders. Most of these "cohort" studies consisted predominantly or exclusively of gay men, and most were asymptomatic or mildly symptomatic at study entry.2 Since then, at least one longitudinal study of intravenous drug users has been conducted3,4 and there is an ongoing longitudinal study at Cornell following a group of men with symptomatic HIV illness or AIDS.5 All of these studies included HIV-negative comparison groups from the same communities as the HIV-positive groups. In the studies of gay men and non-intravenous-drug using women, the findings cumulatively show that rates of current (past month) major depression are in the range of 5 - 10% in both the HIV-positive and HIV-negative groups.

Cross-sectional studies of patients at different stages of illness have yielded mixed findings regarding rates of depression in patients with symptomatic HIV/AIDS, compared to those who are asymptomatic. In the few longitudinal studies following the same people as they get sicker, no increases in rates of depressive disorders over time have been observed.4,5,6,7,8

In our study of long-term AIDS survivors conducted at GMHC in 1990,9 53 men participated in a standardized psychiatric assessment of current mood disorders, psychiatric distress, future outlook, quality of life and physical functioning using several clinician and self-rated scales and a semi-structured interview. Almost all of the men had experienced more than one episode of life-threatening illness, and many had been led to believe at the time of their AIDS diagnosis (in 1987 or earlier) that they had only months to live. Nevertheless, we did not find increased rates of clinical depression compared to rates for asymptomatic HIV-positive gay men or general population rates.10

Psychopharmacologic Treatments for Depression

Standard Treatments

Although published research on antidepressant treatments for HIV-infected patients is limited, results to date endorse the efficacy and safety of tricyclic antidepressants (such as desipramine [Norpramin], amitriptyline [Elavil] and imipramine [Tofranil]) as well as serotonin reuptake inhibitors (fluoxetine [Prozac] or sertraline [Zoloft]), prescribed at standard doses. Ongoing studies of newer antidepressants such as paroxetine (Paxil) for HIV-positive patients suggest they are likely to be equally safe and effective. Two placebo-controlled clinical trials of imipramine, including almost 150 patients in all, found that patients with HIV illness respond to imipramine at the same (high) rate as medically healthy depressed patients.9,11,12

Open studies of serotonin reuptake inhibitors (SRIs) have been conducted by our group and others, with similar findings: The response rates among depressed HIV-positive patients are equivalent to those of medically healthy depressed patients. Even HIV-positive patients with advanced illness do as well, usually have no more side effects, and tolerate the same doses as medically healthy patients. SRIs may have several advantages over tricyclics: milder and more transient side effects, less risk of toxicity in overdose and none of the "anticholinergic" effects (e.g., drowsiness, postural hypotension, constipation, blurred vision, tremor, muscle pain) that are common with tricyclic antidepressants and that may be particularly difficult for HIV-positive patients. Common side effects of SRIs include a feeling of overstimulation, increased gastrointestinal activity and headache. All antidepressants may induce sexual dysfunction (erectile failure, anorgasmia) which may or may not be transient. Switching to another antidepressant helps often but not always.

Unconventional Treatments

Research examining psychostimulants indicates they have antidepressant activity. Psychostimulants include dextroamphetamine (Dexedrine), methylphenidate (Ritalin) and pemoline (Cylert), in descending order of potency. Acute antidepressant effects have been shown among patients with depression and low energy, depressed patients with cognitive impairment, and depressed medically ill patients. In HIV research, we and others have conducted open trials of Dexedrine for patients with late-stage HIV illness and the accompanying problem of severe fatigue, with very good results.13 Our group is now conducting a brief double-blind study of Dexedrine for patients with AIDS, depressed mood and significant fatigue (If interested in participating, call Dr. Glenn Wagner at 212/543-5331).

The primary advantage of using psychostimulants is the speed of response. In contrast to the SRIs, which usually take two to six weeks or even longer to produce a sustained effect, psychostimulants commonly induce a response in a few days. The primary concern among both patients and doctors is risk of abuse, and for this reason we and others do not offer this treatment to patients with a history of significant stimulant abuse or dependence. However, there have been no reports of addiction among either HIV-positive or HIV-negative patients treated with psychostimulants under medical supervision.14

In the last days of life, patients who are bedridden with home-care nursing can still benefit from stimulant medication when they want to be alert, as when family and friends visit. We have spoken with several nurses who care for such patients; they tell us that such patients are noticeably more interactive, even when on a morphine drip.


Testosterone is a naturally occurring hormone that has both "androgenic" (virilizing) and anabolic (energizing and muscle-building) effects. In the past few years it has been increasingly recognized that endocrine (hormone) disorders are common in late-stage HIV illness, and the most common is a decline in testosterone level.15 Symptoms of low testosterone include loss of sexual desire, almost always accompanied by one or more of these problems: low mood, low energy, low appetite or loss of weight and muscle mass. We have conducted testosterone therapy studies with over 150 men so far, and have found striking improvement reported by up to 90% of those we treated, whose primary problem was low libido.16

We also observed weight gain in some, but not all, particularly among men who exercised (Wagner et al., submitted for publication, 1997). It should be noted that our study was limited to men with symptomatic HIV illness, CD4 cell counts under 400 (over half had counts under 50), and deficient or low normal serum testosterone levels. Women were not included, since testosterone is not approved by the FDA for use with women, and the side effects women may experience (facial hair, deepening of voice, enlarged clitoris) are irreversible.

A newly marketed steroid, oxandrolone, may be useful to treat low libido, low mood and wasting for women, but results of ongoing studies are not yet available. We are currently conducting a study of DHEA for HIV-positive men and women with low mood and fatigue. These are the two symptoms that appear most likely to respond to DHEA based on a review of the medical literature,17 but we have no results yet.

An unanticipated finding was that testosterone appears to be an effective antidepressant, improving mood for those who were feeling low, slowed down and lacking zest for their life. We also found striking improved mood in those with clinical depressive disorders, we recently received federal funding to compare the antidepressant effects of testosterone and Prozac (for information, call Dr. Rabkin at 212/543-5762).

As we and others have noted, symptoms of clinical hypogonadism (diminished libido plus low mood, low energy, low appetite/weight loss) and deficient serum testosterone levels are most commonly observed among men with later-stage HIV illness. Whether or not these deficiencies will be corrected spontaneously with combination antiviral treatment (including protease inhibitors), as viral activity is reduced, remains to be seen.

Drug Interactions: Antidepressants and Protease Inhibitors

As protease inhibitors become widely used, the possibility of drug interactions becomes increasingly salient. The four marketed protease inhibitors, ritonavir, indinavir, saquinavir and nelfinavir, are all metabolized in the liver via the P450 enzyme system, primarily the 3A3/4 isoform, with the 2D6 isoform being a secondary metabolic pathway. Many psychotropic drugs share these metabolic pathways, so that the simultaneous use of both classes of drugs may alter the clearance rate of one or the other. For example, according to pharmacists at the Merck (indinavir) hotline, nefazodone (Serzone) increases the serum level of indinavir, thereby increasing the risk of indinavir side effects unless the indinavir dose is adjusted accordingly.

In other instances, the serum level of the psychotropic drug is raised, as in the combination of ritonavir with benzodiazepines or bupropion (Wellbutrin). On a presumptive basis, the manufacturer of ritonavir has listed all benzodiazepines and bupropion as contraindicated, and SRIs, anticonvulsants, most antipsychotics and most tricyclic antidepressants as relatively contraindicated. There are no systematic studies of such combinations, however, and simply abandoning an effective antidepressant or anxiolytic (antianxiety) treatment without substituting an equally effective alternative is not a satisfactory answer. Research in this area is sorely needed. Until then, physicians should consider undertaking concurrent use of antidepressants and protease inhibitors with greater surveillance of effects and side effects of both.


Research has shown that brief, focused psychotherapy, such as "interpersonal" therapy or cognitive-behavioral therapy, is demonstrably helpful for depressed HIV-positive (as well as HIV-negative) patients.12,18 Psychotherapy can be conducted successfully at virtually all stages of HIV illness, although treatment goals may differ. The sense of limited time often makes HIV-positive patients particularly highly motivated to make changes in their lives, leading to sometimes remarkable transformations in values, relationships and life goals as well as mood. However, psychotherapists need to be much more flexible than usual in terms of scheduling, willingness to visit patients when they are ill, staying in telephone contact when that is the only option, working with friends and family who are at the bedside, and "being there" at the end of life, even if their academic training has not prepared them for such exigencies.

Many support groups for people with AIDS are designed as peer counseling rather than group therapy. In such settings, the stress of living with AIDS can be acknowledged with others in similar situations, and advice and solutions to common problems can be shared. However, this is not the appropriate forum in which to seek help for persistent depression. (See box for studies on professionally led support groups.)

Choosing the Type of Treatment

Factors that may influence the choice of treatment modality for depression include personal preference, as well as stage of HIV illness, and the associated considerations of stamina, mobility and insurance coverage. Psychotherapy may be more feasible for depressed asymptomatic or mildly symptomatic HIV patients, who can attend weekly sessions with a separate provider without difficulty. Patients who are sicker may lack the energy and endurance to make extra medical visits. Some patients may already be taking low doses of antidepressant medication for neuropathy, so that the most efficient strategy in treating their depression is to raise the dose to a therapeutic antidepressant level, as side effects permit. Primary care physicians often can successfully prescribe and manage SRIs for depressed patients who are medically ill, once depression is brought to their attention. The "ideal" treatment of combined psychotherapy and antidepressant medication simply may not be feasible for many symptomatic HIV-positive patients in terms of financial and practical considerations.

HIV-related symptoms also may influence choice of treatment. For example, some medications such as amitriptyline are quite sedating, which is useful at bedtime for a patient with insomnia, but distressing if taken during the day by a patient already experiencing fatigue and low energy. SRIs such as fluoxetine may stimulate gastrointestinal activity. They would not be a good choice for someone with chronic diarrhea, but may be particularly suitable for patients taking opiate analgesics, which often induce constipation.

A Psychiatric Success Story

Both antidepressant medications (conventional or not) and psychotherapy are effective in treating HIV-positive patients who are depressed. Depressed patients with late-stage HIV infection respond to treatment as well as medically healthy patients do. We have observed no negative interactions among antidepressants and a wide range of concomitant HIV medications (although, as noted, there has been little information on protease inhibitors to date). Depression in the context of HIV illness is not an inevitable consequence to which patients must adapt; it is not common or "expectable." Treating depression is one of psychiatry's success stories, and HIV-positive patients are no exception. Depression is definitely treatable, and such treatment can dramatically improve quality of life at all stages of illness.


1. American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV). Washington, DC: American Psychiatric Association, 1994.
2. Rabkin JG. International Review of Psychiatry. 1996; 8:157-66.
3. Lipsitz J et al. American Journal of Psychiatry. Nov 1994; 151(11):1662-8.
4. Rabkin JG et al. AIDS. March 1997; 11(4):507-15.
5. Rabkin JG et al. Comprehensive Psychiatry. May 1997; 38(3):146-54.
6. Perry S et al. American Journal of Psychiatry. May 1993; 150(5):775-9.
7. Kessler RC et al. American Journal of Psychiatry. June 1991; 148(6):733-8.
8. Joseph J et al. J. Nervous and Mental Disease. 1990; 178:607-15.
9. Rabkin JG et al. Hospital and Community Psychiatry. Feb 1993; 44(2):162-7.
10. Rabkin JG et al. AIDS Care. 1993; 5(4):401-11.
11. Rabkin JG et al. American Journal of Psychiatry. April 1994; 151(4):516-23.
12. Markowitz J et al. AIDS. April 1994; 8(4):403-12.
13. Wagner GJ et al. J Psychosomatic Research. April 1997; 42(4):407-11
14. Satel SL, Nelson JC. J Clin Psychiatry. July 1989; 50(7):241-9.
15. Poretsky L et al. Metabolism. July 1995; 44(7):946-53.
16. Rabkin JG, Rabkin R. The AIDS Reader. 1995; 5:136-44.
17. Rabkin JG, Ferrando SJ. The AIDS Reader. Jan/Feb 1997; 28-36.
18. Markowitz JC et al. Amer J Psychiatry. Oct 1995; 152(10):1504-9.

Group Therapy for Depression

The National Institute of Mental Health is enrolling a study on women and depression in New York, New Jersey and Florida. The study, called the Smart/EST Women's Project, is designed to evaluate the benefit of supportive therapy interventions on quality of life and health status in a group of ethnically diverse women with AIDS. Four hundred fifty participants will be recruited and randomized to either an intervention or control format. The intervention group will attend ten weekly professionally led two-hour group sessions that focus on stress and risk management and relaxation therapy. The control group will attend an equal number of sessions during which time videotapes will be shown on coping with AIDS and stress management. Both groups will be provided with monetary compensation (up to $575), transportation and childcare services.

After the first ten weeks, all participants will be followed monthly during a three-month maintenance phase and semi-annually for one year thereafter. The study will look at quality of life and assess the effects of the interventions on morbidity and mortality. Anyone interested in enrolling in the Miami area may call 305/243-2103 and ask for Rosalind Mathis. In the New York/New Jersey area, call 212/255-3841 and ask for Anna Sanchez (for New York) or Yolene Gousse (for New Jersey).

In California last summer, both HIV-positive men and women began enrolling in a Stanford University School of Medicine study evaluating how group psychotherapy affects both quality of life and health-related behavior. Half of the volunteers are randomized to attend professionally led group meetings. The design of this study, also funded by the National Institute of Mental Health, is based on observations made by Dr. David Spiegal, the study's principal investigator. He previously found that participating in group therapy improved mood and reduced pain in women with breast cancer. Goals include identifying people who will benefit from group therapy, looking at effects on mood and coping and examining whether group therapy reduces risky behavior and improves treatment adherence. One hundred participants have enrolled so far and another 100 are being sought.

Michele Gill, the study recruiter, stated that since 70% of the current participants are men, study organizers want 70% of the next 100 to be women. She acknowledged the difficulty in recruiting HIV-positive women and attributed this in part to problems in contacting eligible women, who may not be as organized as, for instance, men in the gay community. There is also often the issue of family obligations, and provisions have been made for childcare in an effort to remove this obstacle. Anyone in the San Francisco Bay area who is interested in participating should call Michele Gill at 650/723-2661.