Pharmacologic Treatment of Comorbid Depression


Numerous epidemiological surveys have found high rates of depression among substance abusers. This comorbidity contributes to poorer motivation and treatment compliance, lower and poorer out-comes. Dr. Greenfield presented an overview of the co-occurrence of alcohol use and depressive disorders and evidence for negative consequences of not treating the depressive disorder. Dr. Cornelius presented data demonstrating that fluoxetine is effective in decreasing but not eliminating the depressive symptoms and the level of alcohol consumption. Dr. Nunes reviewed a series of studies examining the evaluation and treatment of depression in substance dependent patients—for alcohol, co-aine and opiates.

Symposium Chair: Richard N. Rosenthal, M.D., Albert Einstein College of Medicine, Beth Israel Medical Center, New York, NY

Epidemiology and Course of Co-Occurring Alcohol Use and Depressive Disorders

Shelly F. Greenfield, MD, MPH, McLean Hospital, Belmont, MA

Co-occurring depression and alcohol dependence presents difficult clinical issues including: How do co-occurring alcohol and depressive disorders affect each other? How does one diagnose depression in someone with alcohol dependence? How does one treat co-occurring depression and alcohol dependence? To ad-dress these questions, it is necessary to consider individual patients, clinical studies, and population-based samples.

Case One is a 30-year-old man with longstanding dysthymia and alcohol dependence, abstinent for 12-months, who now has six months of symptoms consistent with depressive disorder. Case Two is a 32-year-old woman who presents with a six-month history of symptoms consistent with depressive disorder. Assessment reveals a pattern of drinking that has increased from nightly consumption of one glass of wine up to three to four glasses in the last year. The patient says she needs the wine to get to sleep. Case Three is a 50-year-old woman with recurrent major depression now resistant to antidepressants and psychotherapy. She denies symptoms consistent with alcohol abuse or dependence and agrees to abstinence from her usual one to two drinks per night. Six months later she shows no improvement in depression. A family interview reveals that she has never stopped drinking and daily intake is more extensive than treaters or family had known.

An epidemiological overview needs to address these questions: 1) What is the epidemiology of co-occurrence of alcohol and depressive disorders? 2) What is the relationship of these disorders to one another? 3) How is depression ascertained in individuals with alcohol dependence? 4) What is the course of these two dis-orders when they co-occur? Population-based studies were used to address the first two questions because they are not biased in terms of treatment seeking, where we know the comorbidity is high. Clinical studies, including one at McLean Hospital, were used to address the last two questions.

The National Co-morbidity Survey (NCS)1, showed that the lifetime prevalence of alcohol abuse in the population is 9.4% and the 12-month prevalence is 2.5%; lifetime prevalence of alcohol dependence is 14.1% and the 12-month prevalence is 4.4%. Men are more likely than women to have lifetime alcohol abuse and

dependence. The NCS found that a majority of those with alcohol use disorders (AUDs) have at least one psychiatric disorder and there was stronger co-occurrence in those with alcohol dependence (AD) rather than abuse. Co-occurrence stronger among women and anxiety and affective disorders were the largest proportion of co-occurring disorders in women. Substance, conduct and antisocial personality disorders were the largest proportion of co-occurring disorders in men.

The NCS also reported the prevalence of depression as follows: among individuals with lifetime history of alcohol abuse, 9% of men and 30.1% of women had lifetime history of depression; among lifetime alcohol dependence, 24.3% of men and 48.5% of women had lifetime history of depression. In addition, the NCS reported on risk, which means being compared to ones neighbor who is similar, but doesn’t have AD. Compared to men without AD, men with AD are three times as likely to have major depressive disorder (MDD) in their lifetime. Compared to women without AD, women with AD are four times as likely to have MDD in their lifetime. Generally, MDD was found to precede AD.

Several factors influence the relationship of AUDs and MDD. Neither the risk nor the temporal onset can imply causality, where one disorder causes the other. The disorders can be independent despite a high rate of comorbidity. It is possible that a prior depression can influence the onset of AUD and vice versa. Also the onset of both disorders can be influenced by common factors such as community context, stress, lack of social support, and genetic predisposition.

Family studies2,3 examined genetic pre-disposition and found increased alcoholism in relatives of individuals (MDD). The studies also found in-creased alcoholism in relatives of individuals with AD and MDD compared to individuals with alcoholism alone. Another study showed increased risk for depression among relatives of individuals with alcohol abuse or dependence. In female twin studies, Kendler4 found overlapping liability for major depression and AD attributable to genetic factors shared by the two disorders. Another study of twin pairs showed that individuals with major depression were at increased risk for AD. It noted that there was sex-specific genetic and environ-mental transmission, and the causes of depression and alcoholism overlap, but are not the same. In addition, 51-60% are attributable to shared genetic factors, but there was no evidence that common environmental factors contribute to alcoholism-depression comorbidity.

Clinical studies examine how these disorders affect one another and what to do about it. 1) What is the effect of AD on the course of recovery from depression? One ten- year prospective study5 found that depressed individuals with no history of AD or with past but not current AD, were twice as likely to recover from MDD than those with current AD. 2) What is the effect of depression on AD? There have been many studies that have shown that, in general, psychiatric comorbidity and severity negatively affect drinking outcomes. However, there are conflicting results from these studies of the effect of depression on AD. Some show no effect of MDD on drinking and others show that MDD worsens drinking outcomes. Some of the differing results reflect methodological differences that include the type of samples used, the follow-up intervals, use of lifetime or current depression, and measurement of depression by symptom scale or by diagnosis.

The following is a study at McLean Hospital.6 This report is within the context of a larger study that looked at whether women and men differ with respect to predictors of return-to-drinking and psychosocial adjustment during the 12-months following entry into treatment. The focus was on how depression affected recovery from the AD and whether there were gender differences. It controlled for a variety of possible predictors of treatment outcome, e.g., education, marital status. This was a prospective follow-up study of men and women with AD, recruited from the McLean inpatient drug/alcohol unit, that were followed monthly for one year after discharge. Depression was measured by both the Beck and the DSM criteria. There were 101 subjects—41 women and 60 men. A current diagnosis of MDD was significantly related to time-to-relapse. Gender and MDD were not significantly related to relapse. Median time to re-lapse was 150 days for nondepressed subjects and 41 days for depressed subjects.

What does this mean for assessment and treatment? 1) Be alert for the likelihood that mood and alcohol use disorders co-occur. 2) Take a full history of onset, duration, and development of AUD. 3) Take a full history of onset and progression of mood disorders. 4) Clarify if symptoms meet diagnostic criteria. 5) Clarify the relationship between the two disorders over time.

In summary, MDD and AUDs are highly prevalent and often co-occur. There is evidence that the presence of each disorder worsens the other. Careful diagnostic assessment is necessary to clarify co-occurring disorders. And, there is mounting evidence that effective treatment of each disorder may be helpful in recovery from the other.


  1. Kessler RC. Lifetime co-occurrence of DSM-IIIR alcohol abuse and dependence with other psychiatric disorders in the national comorbidity survey. Arch Gen Psychiatry. 1997;54:313-321.
  2. Grant BF, Hasin DS, Dawson DA The relationship between DSM-IV alcohol use disorders and DSM-IV major depression: Examination of the primary- secondary distinction in a general population sample. J Affect Disord 1996;38:113-128.
  3. Dawson DA, Grant BF. Family history of alcohol-ism and gender: Their combined effects on DSM-IV alcohol dependence and major depression. J Stud Alcohol. 1998;59:97-106.
  4. Kendler KS, et al. Alcoholism and major depression in women: A twin study of the causes of comorbidity. Arch Gen Psychiatry. 1993;50:690- 698.
  5. Mueller TI, et al. Prognostic effect of the variable course of alcoholism on the 10-year course of depression. Am J Psychiatry. 1994;151:701-706.
  6. Greenfield SF, Weiss RD, Muenz LR, Vagge LM, Kelly JF, Bello LR, Michael J: The effect of depression on return to drinking: A prospective study. Arch Gen Psychiatry 1998;55:259-265.


Fluoxetine and Comorbid Depression in Substance Abuse

Jack R. Cornelius, MD, MPH, University of Pittsburgh, Pittsburgh, PA

Results of several pharmacotherapy studies involving comorbid depression and substance abuse are presented. These studies were all conducted in Pittsburgh over the last 10 years. The first study was an open label pilot study involving 12 patients with comorbid MDD and AD, using the SSRI medication fluoxetine. The major depression symptoms of these patients at baseline were severe, reflecting the fact that all were recruited from the dual diagnosis unit of an inpatient psychiatric hospital. Fluoxetine was used for that study because it was the only SSRI antidepressant on the market. The results of that pilot study suggested efficacy for fluoxetine for decreasing both the depressive symptoms and the drinking of that comorbid population.

The pilot study results led to the funding of a double-blind, placebo-controlled study of fluoxetine in patients with comorbid major depressive disorder (MDD) and alcohol dependence (AD). Patients with substance dependence other than AD were excluded from the study, but patients with comorbid substance abuse were not excluded. The results demonstrated efficacy of fluoxetine for decreasing both the depressive symptoms and the drinking of this population. The fluoxetine group also avoided pathological drinking (5 or more drinks per day) significantly longer than the placebo group. However, by week six of the study, even the fluoxetine group had demonstrated on average at least one day of pathological drinking. Thus, the fluoxetine was significantly decreasing the drinking and the depressive symptoms of this treatment group, but substantial residual symptoms generally persisted at the end of the trial.

Recently, a naturalistic open label one-year follow-up study was conducted involving the patients from the double-blind study described above. Findings indicated that the patients tended to continue the same treatment, which they had had during the trial, though they were free to use any treatment. Significant differ-ences between groups persisted on the Hamilton and Beck depression scales, global assessment scale, and drinking scales. The magnitude of these differences was similar to that which had been demonstrated at the end of the 12-week acute phase trial. The results of the one-year follow-up evaluation suggest persistent long-term (one-year) efficacy for fluoxetine for decreasing both the depressive symptoms and the drinking of patients with comorbid MMD and AD.

Some secondary analyses were then performed. The results suggested that fluoxetine appears to decrease the level of marijuana use among those patients who demonstrated marijuana abuse in addition to MMD and AD. Secondary analyses were also performed that focused on cigarette smoking. The fluoxetine group was found to have smoked 32% less than the placebo group. However, only one patient abstained from smoking, so the clinical significance of the decrease in smoking is un-clear. In addition, the presence of comorbid cocaine abuse among some subjects was assessed for whether it affected the outcome of the double-blind study. Those patients were found to show significantly less improvement on measures of depression and drinking than those without comorbid cocaine abuse. Thus, comorbid cocaine abuse served as a predictor of poor outcome for both the depressive symptoms and the drinking of patients with MMD and AD, regardless of their treatment group.

A recent pilot study was conducted to assess whether dual diagnosis patients who had had only a partial response to SSRIs might demonstrate an additional decrease in drinking if naltrexone were added to their treatment regimen. The data suggested that there was a significant decrease in level of drinking and a trend toward a decrease in depressive symptoms. The results suggest that double-blind, placebo-controlled studies with SSRI anti-depressants and naltrexone are warranted among patients with comorbid major depression and alcohol dependence.

An open label pilot study is also being con-ducted to assess whether the therapeutic benefits shown by fluoxetine in treating comorbid adults also extend to a younger age group of adolescents. The results suggest efficacy of fluoxetine for decreasing both the depressive symptoms and the drinking of that adolescent population. On the Clinical Global Impressions scale, all 13 patients demonstrated much or very much improvement in depressive symptoms at the end of the trial, but only seven of the 13 subjects showed much or very much improvement in symptoms of alcohol dependence. The results of that study suggest that fluoxetine may have a more robust affect on decreasing depressive symptoms than on decreasing drinking of comorbid adolescents, though both of these symptom domains appear to respond at least partially to treatment with fluoxetine. The study results also suggest that subsequent double-blind, placebo controlled studies are warranted to assess the efficacy of SSRI antidepressants in treating adolescents with comorbid major depression and an alcohol use disorder.

In addition, a double-blind, placebo-controlled study is currently being conducted to assess the efficacy of valproate plus lithium in contrast to lithium alone in treating a population with comorbid bipolar disorder and alcohol dependence. There are as yet no study findings.


  1. Cornelius JR, Salloum IM, Cornelius MD, et al. Fluoxetine trial in suicidal depressed alcoholics. Psychopharmacol Bull. 1993;29;195-199.
  2. Cornelius JR, Salloum IM, Ehler JG, et al. Fluoxetine in depressed alcoholics: A double-blind placebo-controlled trial. Arch Gen Psychiatry. 1997;54:700-705.
  3. Cornelius JR, Salloum IM, Thase ME, et al. Fluoxetine vs. placebo in depressed alcoholic cocaine abusers. Psychopharmacol Bull. 1998; 23:119-121.
  4. Salloum IM, Cornelius JR, Daley DC, et al. Naltrexone utility in patients with comorbid major depression and alcohol dependence. Psychopharmacol Bull. 1998;34:111-115.
  5. Cornelius JR Salloum IM, Haskett RF, et al. Fluoxetine vs placebo for the marijuana use of de-pressed alcoholics. Addict Behav. 1999;24:111-114.
  6. Cornelius JR, Perkins KA, Salloum IM, et al. Fluoxetine vs. placebo for the smoking of depressed alcoholics. J Clin Psychopharmacol. 1999;19:183-184.
  7. Cornelius JR, Salloum IM, Haskett RF, et al. Fluoxetine versus placebo in depressed alcoholics: A one-year follow-up study. Addict Behav. 2000;25: 307-310.
  8. Cornelius JR, Bukstein OG, Birmaher B, et al. Fluoxetine in adolescents with major depression and an alcohol use disorder: An open label trial. Addict Behav. In Press.


The Evaluation and Treatment of Depression in Substance Dependent Patients

Edward V. Nunes, MD, New York State Psychiatric Institute, New York, NY

Most of what we know about opiate-dependent patients has been derived from methadone-maintained patients. Depression is the most common comorbid psychiatric disorder in opiate addicts along with antisocial personality disorder. Depression has been associated consistently with poor treatment outcomes. That leads to a treatment hypothesis that opiate dependence and depression are causally linked in some partial way, and treatment of depression will improve the treatment outcome of both mood and substance use disorders.

The relationship between depression and opiate dependence is complicated. Depression can be a primary independent disorder, a response to chronic drug toxicity or withdrawal, or a response to stress and loss. One of the problems in diagnosing depression in opiate addicts is how to distinguish treatable depression. Some of the issues are that drug abuse tends to have started early, it is chronic over time, and abstinent periods are rare.

About 10 years ago, early studies looked at treatment of depression with tricyclic antidepressants. The weakness of these studies was that depression was identified with cross-sectional symptom scales, not a history and a diagnosis. Two of the studies with modest sample sizes found medication was superior to placebo in terms of depression outcomes. Three showed that there was a very high placebo response rate. In fact, what a cross-sectional scale might be measuring in these depressed opiate addicts is transient toxicity or stress related depressive symptoms that may go away spontaneously without treatment and do not represent a true depressive disorder. None of these studies demonstrated any clear effect of antidepressant medication on drug use outcome.

A working approach to the diagnosis of depression in these patients was developed. It included a clinical psychiatric history and a parallel history of the time course of drug use. Subjects were asked for ages of onset and periods of abstinence. The time course of depression was related to the time course of drug use. Data was sought for depressions that antedate the onset of drug use, persist during an abstinence period or are chronic in the current episode—at least 3 months duration. A pilot trial of imipramine with 24 depressed methadone maintenance patients showed that of the 17 completed, nine showed improvement on both depression and drug use. This was an open label, uncontrolled study, but with promising data. Because there was no evidence that dysthymia or major depression respond differently, both were included in these studies.

This next study is a placebo-controlled trial in methadone maintenance clinics in Queens, New York. One-hundred thirty-seven patients were randomized and 84 completed at least a six-week trial. There were more dropouts due to side effects of the medication or medical problems. The results were encouraging, but with caveats. On the global response which required a substantial improvement in both mood and drug use, 57% were responders compared to 7% on placebo. On depression response alone there was 67% compared to 26% on placebo. Self-reported craving was reduced on imipramine. The measure of sustained abstinence (at least four weeks) by self-report confirmed by a clean urine was very low. There was a significant reduction in self-reported drug use. It was concluded that depression can be diagnosed by clinical history and treated in methadone maintenance patients. There was modest improvement in sub-stance use, but not abstinence. That suggested the need for better concurrent behavioral therapies in these patients. Imipramine is an effective agent, but limited by side effects.

Currently, there are two seratonin reuptake inhibitors (SRI) trials going on. One already shows a negative effect. Interim analysis of the ongoing sertraline trial is not demonstrating good outcome. The result may be either a negative study or one showing a very weak effect of antidepressant medication. Some of these nonresponders may need a medication that has more of an immediate soporific or antianxiety effect.

Some clinical recommendations include: 1) Scales like the Beck are acceptable for screening, but diagnosis should be made by clinical history. 2) Treat primary depression, and substance induced or secondary depression is likely to respond to the medication as well. 3) Conduct a medical workup including HIV, Hepatitis C, asthma, and hypertension. Medical comorbidity increases with age, particularly when a patient is on methadone maintenance. 4) Look at the patient’s level of stress and loss. These are the aspects of depression that a medication will not address. It is hoped that a combination of medication and psychotherapy will do better for issues such as unemployment, marital breakups, deaths of relatives, homelessness, and living in abusing relationships. 5) Avoid benzodiazepines at all costs as these patients tend to get into trouble with them. 6) Consider SRIs to be the first line treatment for depression because they are safer and better tolerated. If an SRI fails then consider a medication with a more potent noradminergic or dopaminergic activity. 7) Look for other disorders such as PTSD, ADHD, panic disorder, social phobia, conduct disorder, antisocial personality disorder, and anger problems. It may be that depression in these patients is a kind of common pathway. However, if one goes back to early develop-mental history, it is possible that some of these other disorders and problems which led to the substance use in the first place may be found. They may have helped create a more severe course that resulted in depression. Then treatment of other disorders needs to be pursued.



2000 Proceedings

2000 AAAP Annual Meeting Proceedings Copyright 2001 AAAP