0
We're unable to sign you in at this time. Please try again in a few minutes.
Retry
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
Retry
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Review | Clinician's Corner

Treatment of Depression in Patients With Alcohol or Other Drug Dependence:  A Meta-analysis FREE

Edward V. Nunes, MD; Frances R. Levin, MD
[+] Author Affiliations

Author Affiliations: Depression Evaluation Service (Dr Nunes) and Division on Substance Abuse (Drs Nunes and Levin), New York State Psychiatric Institute and Department of Psychiatry, Columbia University (Drs Nunes and Levin), New York.


JAMA. 2004;291(15):1887-1896. doi:10.1001/jama.291.15.1887.
Text Size: A A A
Published online

Context Depression and substance abuse are common and costly disorders that frequently co-occur, but controversy about effective treatment for patients with both disorders persists.

Objective To conduct a systematic review and meta-analysis to quantify the efficacy of antidepressant medications for treatment of combined depression and substance use disorders.

Data Sources PubMed, MEDLINE, and Cochrane database search (1970-2003), using the keywords antidepressant treatment or treatment depressed in conjunction with each of the following alcohol dependence, benzodiazepine dependence, opiate dependence, cocaine dependence, marijuana dependence, and methadone; a search of bibliographies; and consultation with experts in the field.

Study Selection Among inclusion criteria used for study selection were prospective, parallel group, double-blind, controlled clinical trials with random assignment to an antidepressant medication or placebo for which trial patients met standard diagnostic criteria for current alcohol or other drug use and a current unipolar depressive disorder. Of the more than 300 citations extracted, 44 were placebo-controlled clinical trials, 14 of which were selected for this analysis and included 848 patients: 5 studies of tricyclic antidepressants, 7 of selective serotonin re-uptake inhibitors, and 2 from other classes

Data Extraction We independently screened the titles and abstracts of each citation, identified placebo-controlled trials of patients with both substance dependence and depression, applied the inclusion criteria, and reached consensus. Data on study methods, sample characteristics, and depression and substance use outcomes were extracted. The principal measure of effect size was the standardized difference between means on the Hamilton Depression Scale (HDS).

Data Synthesis For the HDS score, the pooled effect size from the random-effects model was 0.38 (95% confidence interval, 0.18-0.58). Heterogeneity of effect on HDS across studies was significant (P <.02), and studies with low placebo response showed larger effects. Moderator analysis suggested that diagnostic methods and concurrent psychosocial interventions influenced outcome. Studies with larger depression effect sizes (>0.5) demonstrated favorable effects of medication on measures of quantity of substance use, but rates of sustained abstinence were low.

Conclusions Antidepressant medication exerts a modest beneficial effect for patients with combined depressive- and substance-use disorders. It is not a stand-alone treatment, and concurrent therapy directly targeting the addiction is also indicated. More research is needed to understand variations in the strength of the effect, but the data suggest that care be exercised in the diagnosis of depression–either by observing depression to persist during at least a brief period of abstinence or through efforts by clinical history to screen out substance-related depressive symptoms.

Figures in this Article

Depression and substance dependencies are among the most prevalent and costly disorders confronting the health care system, and they tend to co-occur,1,2 leading to greater overall severity and worse health-related outcomes,3 including increased risk of suicide.46 A wide range of initiatives have been developed to improve screening and diagnosis of depressive and drug- or alcohol-use disorders in primary medical care settings where they commonly present, as well as in mental health or addiction treatment programs.7 However, controversy about how to treat patients with the concurrent disorders persists. One of the most basic questions is whether to treat depression in the setting of ongoing substance abuse. Recent recommendations suggest concurrent substance abuse should not be a barrier to treating depression.8 However, clinicians working with substance-dependent patients are often reluctant, for a variety of reasons, to initiate specific antidepressant treatment, including concerns about confusing substance-induced depressive symptoms with true depressive disorders.9 Evidence is needed to help guide treatment.

We therefore undertook a meta-analysis of placebo-controlled trials to examine whether depression responds to antidepressant medication treatment in substance-dependent patients and the impact of such treatment on concurrent substance abuse. Reviews of a prior generation of studies found little consistent evidence of efficacy but revealed methodological problems, including weak diagnostic methods wherein patients were selected only with scales measuring current depressive symptoms1012 Our meta-analysis, therefore, focuses on more recent studies in which depressive disorders were identified by clinical history using established diagnostic criteria. The heterogeneity of study results is critically examined to better understand factors, such as methods of diagnosing depression, that may contribute to medication efficacy in depressed substance abusers.

Search Strategy and Study Inclusion Criteria

PubMed, MEDLINE, and Cochrane databases from 1970 through December 2003 were searched using the keywords antidepressant treatment or treatment depressed in conjunction with each of the following words: alcohol dependence, benzodiazepine dependence, opiate dependence, cocaine dependence, marijuana dependence, and methadone. No language restrictions were applied. In a further effort to locate both published and unpublished studies, the reference lists of key publications and related reviews were examined, and 11 experts in the field were consulted, 9 in the United States and 2 in Europe. The 2 authors (E.V.N. and F.R.L.), both clinical investigators, independently screened the titles and abstracts of each citation, identified placebo-controlled trials involving patients with both substance dependence and depression, applied the inclusion criteria, and reached consensus.

Quiz Ref IDStudies were included if they met the following criteria: (1) Prospective, parallel groups, double-blind, controlled clinical trial with random assignment to an antidepressant medication or placebo; (2) Patients met standard diagnostic criteria set forth in the Diagnostic and Statistical Manual of Mental Disorders third, revised third, or fourth edition (DSM-III,13DSM-III-R,14DSM-IV15) for both a current drug (opiate, cocaine, other stimulant, or sedative-hypnotic) or alcohol-use disorder and a current unipolar depressive disorder (major depression, dysthymia, or depression not otherwise specified), diagnoses made by clinical or structured diagnostic interview; (3) Outcome of depressive symptoms is reported; and (4) For trials entering patients both with and without current depressive disorders, the depressed subgroup is an a priori stratification factor in the design. Secondary analyses were excluded because of concern that they may capitalize on chance and bias in favor of medication treatment.16 Fourteen studies1730 ultimately met these criteria.

Data Extraction

Data were extracted from reports by 1 of the authors (E.V.N.), checked by the second (F.R.L.), and discrepancies resolved by consensus. The principal outcome measure was the Hamilton Depression Scale (HDS) score.31 The measure of effect size was the standardized difference between means (Cohen d), which was calculated from the F test for analysis of covariance on the difference between means for each treatment at study end, covaried for the baseline score or when analysis of covariance was not available from means and SDs or significance level. Data from the sample of all randomly assigned patients, or the most complete sample reported, were used. We also compiled categorical measures of depression response, reflecting the proportion of patients meeting a criterion of clinically significant improvement.

For continuous measures of substance use outcome, consisting of various measures (mainly self-reported) of frequency or quantity of use, standardized differences between means were calculated from raw means and SDs, or estimated from graphs of the data.17,23,28 When several such measures were reported within a study, the effect sizes were averaged to obtain a summary effect. We also compiled categorical measures, representing the proportions of patients meeting various criteria of sustained abstinence or remission.

We compiled a panel of moderator variables, including principal substance-use disorder in each study, placebo response (percentage decrease in the HDS between baseline and at end of study in the placebo group), features of the depression diagnoses and diagnostic methods, sample demographics, class of antidepressant medication, and whether the psychosocial background treatment was manual guided. Compliance with medication was not reported consistently, but the percentage of patients retained in the trial was available for all studies.

Assessment of Study Quality

Each of us independently rated the quality of each study, using a 15-item scale (1 point per item for a maximum score of 15) developed by Detsky and colleagues,32 which addresses fundamental aspects of the methods and reporting of clinical trials, including randomization and blinding procedures, sample selection, therapeutic regimen, outcome measures, and statistical analysis (mean [SD] difference in total scores between raters, 0.57 [51] points; intraclass correlation coefficient, 0.74). In addition, each study was rated according to the guidelines set forth by Schulz and colleagues33 as adequate, unclear, or inadequate with respect to concealment of treatment allocation from clinical and research staff at the time of randomization in order to avoid selection bias, method of generating the sequence of treatment allocation, and blinding. Supplemental descriptions of these procedures were provided by each of the lead authors. These features have been associated with bias in clinical trials in the medical literature and suggested as important quality indicators.33

Statistical Analysis

Effect sizes and pooled estimates of effect across studies were calculated with the Comprehensive Meta-Analysis software package,34 using analysis of variance models for standardized mean differences (Cohen d), and the inverse variance model for differences between response rates. Studies were entered as random effects in all models. For the primary outcome measure, the fixed effects model is also reported for comparison. Publication bias was addressed by inspection of the funnel plot on the primary outcome measure and by application of the sensitivity analysis developed by Copas and Shi.35 Heterogeneity was evaluated with the Q statistic, and the I2 statistic, a transformation of Q that estimates the percentage of the variation in effect sizes that is due to heterogeneity.36 Categorical moderators (eg, alcohol vs other drug-dependent sample) were entered as grouping variables in the analysis of variance model. Continuous moderators (eg, placebo response) were first regressed against effect size using a weighted regression37 and then dichotomized by median split and analyzed as grouping variables in the analysis of variance; results were similar both ways, and only the dichotomized analyses are reported for ease of presentation and comparison among moderators.

Search Findings

The search strategy yielded more than 300 citations, from which 44 placebo-controlled clinical trials involving drug or alcohol-use disorders and depression were targeted for detailed review. Fourteen trials1730 met the inclusion criteria for the meta-analysis and included 848 patients randomly assigned to receive medication or placebo (Table 1). Sixteen trials3954 were excluded because they did not select patients with current syndromal depression—having mainly applied a cutoff score on a cross-sectional scale measuring depressive symptoms. Seven excluded reports were secondary analyses, four5558 examining redundant subsamples of 1 of the included trials,22 and three5961 reporting positive effects of medication in small depressed subgroups from larger trials. Two62,63 were preliminary reports of trials later published with complete samples.19,22 One trial,64 a trend-positive study, was excluded because it was a placebo-controlled discontinuation trial rather than a prospective, parallel groups trial. Four trials could not be included because they are being worked up or submitted for publication—preliminary reports indicating that 1 showed a significant antidepressant effect (David McDowell, MD, written communication, August 29, 2003), 2 showed no difference between medication and placebo (Patrick J. McGrath MD, written communication, August 29, 2003; Susan Tross, PhD, written communication, September 3, 2003), and 1 showed a small nonsignificant trend.65

Table Graphic Jump LocationTable 1. Placebo-Controlled Trials of Antidepressants in Substance-Dependent Patients With Depression: Methodology, Placebo Response, and Effect Size on the Hamilton Depression Scale
Characteristics of Included Studies and Quality Assessment

Quiz Ref IDEight of the trials meeting inclusion criteria recruited alcohol-dependent patients,1719,2225,30 4 recruited methadone maintained opiate-dependent patients,20,26,27,29 and 2 recruited cocaine-dependent patients.21,28 Mean sample ages ranged from 29.0 to 44.5 years (median, 37.9). There was substantial representation of women (median sample, 40% women; range, 8%-55%) and minority patients (median sample, 31% minority; range, 1%-53%). Across samples, large proportions of patients were unmarried (median, 73%; range, 40%-88%), unemployed (median, 47%; range, 30%-49%), or having other indicators of low socioeconomic status such as low educational attainment (median years of education, 13.2; range, 11-15 years). Mean baseline HDS scores ranged from 11.7 to 28.9 (median, 19.5), indicating moderate to severe levels of depressive symptoms. Hamilton Depressive Scale scores that are in the range of 10 to 20 indicate mild to moderate depression and scores higher than 20, severe depression; the higher the score the greater the severity of depression.

Methodological features of the trials are summarized in Table 1. In addition to being randomized, double-blind, and placebo-controlled, the designs of the included trials were uniform in using DSM-III, DSM-III-R, or DSM-IV criteria for diagnosing depressive disorders; in providing adequate medication dosages; and in having trial durations of at least 6 weeks and, for most, 12 or more weeks, which is sufficient time for antidepressant effects to occur. Outcomes on the HDS were evaluated either on the sample of all randomized patients or the sample of patients returning for at least 1 visit after randomization so that 827 (97%) of 848 patients were evaluated. For secondary outcomes (categorical response measures and continuous measures of substance use), in several studies data were not available or smaller samples were evaluated, resulting in smaller overall numbers of patients. All studies used independent staff members to carry out the randomization and to prepare blind medication, and all studies were rated as adequate34 with respect to concealment of treatment allocation, method of generating the sequence of treatment allocation, and blinding. The overall study quality ratings on the scale of Detsky et al32 were moderately high, with a mean (SD) of 12.25 (1.00), out of a total possible score of 15 (range, 10-14). Taken together these data suggest the studies contained most essential features of well-designed antidepressant trials.

Depression Outcome and Moderators of Antidepressant Effect

Depression outcome data are summarized in Table 1 and Table 2, where studies are listed in order of depression effect size from greatest to least. Eight of the studies1724 were reported as demonstrating significant or trend significant antidepressant effects, and 6 studies2530 were negative (Table 1). For the principal outcome, the HDS score, the pooled effect size (standardized mean difference between medication and placebo groups) from the random effects model is 0.38 (95% confidence interval [CI], 0.18-0.58). For comparison, the fixed effects model yielded a similar point estimate (0.38) with a narrower CI (95% CI, 0.24-0.51). The raw rates of depression response were 52.1% for those taking medication and 38.1% for those taking placebo, and the pooled random effects estimate of difference in response rates between treatments was 16.8% (95% CI, 6.9%-26.7%; Table 2).

Table Graphic Jump LocationTable 2. Effects of Antidepressant Medication Treatment on Rates of Depression Response, and Abstinence or Sustained Remission From Substance Abuse*

There is significant heterogeneity of effect on the HDS across studies (Q13, 26.47; P<.02; I2 = 51% of variability in effect sizes due to heterogeneity). The analysis of moderators is presented in Table 3, and the underlying data can be inspected in Table 1. As can be seen, there is a trend for the medication effect to be larger in the studies of patients dependent on alcohol than in studies of patients dependent on other drugs, but this explains only 7% of the variance and substantial heterogeneity remains within both groupings. Quiz Ref IDPlacebo response (percentage reduction in the HDS in the placebo group) is a powerful predictor of antidepressant effect, explaining 71% of the variance in effect sizes across studies and eliminating variance due to heterogeneity; placebo response of more than 25% is associated with a pooled estimate of effect size near 0 while low placebo response (<25%) yields a pooled estimate of effect in the range between medium and large. This can be confirmed by inspecting Table 1, wherein the effect sizes and placebo responses are tabulated and by the Forrest plot (Figure 1), wherein studies are grouped by placebo response. Similarly, in Table 2, an orderly relationship is observed between placebo response, measured herein by percentage of patients in the placebo group meeting a standard depression response criterion, and difference in response rates between medication and placebo.

Table Graphic Jump LocationTable 3. Moderators of Effect of Antidepressant Medication on Outcome of the Hamilton Depression Scale in Substance-Dependent Patients With Depressive Disorders
Figure 1. Effect of Antidepressant Medication on Outcome of Depression (Hamilton Depression Scale)
Graphic Jump Location
Effect sizes are Cohen d, error bars represent the 95% confidence intervals (CIs). Studies are stratified into groups with low (<25%) vs high (>25%) categories of placebo response, measured as the percentage decrease in the Hamilton Depression Scale between baseline and end-of-study in the placebo group.

Four other significant moderators explained smaller proportions of the variance. Diagnosis of depression after at least a week of abstinence in an inpatient or outpatient setting was associated with greater antidepressant effect while a larger proportion of women in the sample, selective serotonin reuptake inhibitor (SSRI) medication, and a concurrent psychosocial intervention that was manual guided, such as cognitive behavioral therapy or relapse prevention, were associated with smaller effect sizes. Type of depression diagnosis (major depression vs mixed diagnoses), severity of depression at baseline (HDS score), minority representation in the sample, age, proportion of patients completing the study, and study quality score (Detsky et al32 scale) were not significant moderators.

Substance Use Outcome

The pooled effect size on continuous measures substance use, representing mainly self-reported quantity of use, is 0.25 (95% CI, 0.08-0.42), and the test for heterogeneity nears significance (Q12, 17.12; P<.15; I2 = 30%). As shown in Figure 2, substance use effect sizes were related to depression effect size. Among studies1722 with depression effect sizes greater than 0.5, the substance use effect size was 0.56 (95% CI, 0.33-0.79) while among studies with lower depression effect sizes,2330 the pooled substance use effect size was near 0 (0.07; 95% CI, −0.11 to 0.25; between strata Q1, 13.3; P<.001). Quiz Ref IDHowever, as shown in Table 2, rates of sustained abstinence or substance remission were low across studies and treatment groups, with only a small pooled difference between medication and placebo.

Figure 2. Effect of Antidepressant Medication on Outcome of Substance Use*
Graphic Jump Location
Effect sizes are Cohen d on outcomes reflecting quantity of substance use (mainly by self-report). Error bars represent the 95% confidence intervals (CIs). Studies are stratified into 2 groups based on the size of the effect of medication on depression outcome (Cohen d for the Hamilton Depression Scale) effect size lower than 0.50 (low); and effect size greater than 0.50 (high). The study by Roy18 did not have the data necessary for analysis for this figure.
Assessment of Publication Bias

The funnel plot was inspected and found to be roughly symmetric, suggesting absence of publication bias. The sensitivity analysis35 models the robustness of the pooled effect size with respect to a putative population of all studies (published and unpublished) as a function of parameters reflecting different patterns of selection bias. The model estimates the pooled effect sizes and checks the fit of each to the funnel plot over a range of increasing selection severities (from small to large proportions of missing studies). For our sample, fit to the funnel plot was adequate (P>.05) throughout the range of selection biases modeled. However, caution is suggested, since as the number of studies missing increased to 8 or more, the pooled estimate of effect size approached 0.20 and its 95% CI expanded to include 0.

Overall, this meta-analysis finds antidepressant medication effective for treatment of depressive syndromes among patients with alcohol or drug dependence. There was also significant heterogeneity in effect across studies, which was strongly related to placebo response. Six studies, all with high placebo response, yielded effect sizes near 0 while 6 studies, all with low placebo response, yielded effect sizes greater than 0.5. The combined effect size must be interpreted with caution in the setting of heterogeneity, but its magnitude (0.38; 95% CI, 0.18-0.58) represents an effect in the range between small and medium,66 similar to the effect size (0.43) found in a recent meta-analysis of antidepressant trials in depressed outpatients.67 Reviews of a prior generation of antidepressant trials in alcoholics and opiate-dependent patients found limited evidence of antidepressant efficacy and identified several methodological problems, including short trial lengths, inadequate antidepressant doses, and selection of patients with cross-sectional measures of depressive symptoms.1012 The findings in our meta-analysis suggest that antidepressant medications can be effective among dually diagnosed patients when used at adequate doses, for at least 6 weeks, and for depressive syndromes (ie, major depression or dysthymia) identified by clinical history and established diagnostic criteria.

The findings also suggest that when medication is effective in treating depression (effect size >0.5), it helps diminish quantity of substance use. However, the proportions of patients achieving sustained abstinence or remission were low, even among studies with robust antidepressant effects. Thus, the effect of treating depression on substance abuse outcome appears limited, and among such dually diagnosed patients, therapies directly targeting the addiction are also needed.

High placebo response likely results either from selection of patients with transient depressions that improve without specific treatment or a concurrent psychosocial intervention that exerts substantial antidepressant effects, obscuring effects of medication. Depressive symptoms among substance-abusing patients are often transient, representing toxic or withdrawal effects that resolve in response to abstinence6872 or to entry into substance abuse treatment.73,74Quiz Ref IDStudies that diagnosed depression after at least a week of abstinence,1719,22 mainly on inpatient units, yielded larger effect sizes and low placebo response, suggesting the abstinence requirement succeeded in screening out transient symptoms and in selecting primary or independent depressive disorders for treatment.

Thus, our findings are consistent with traditional recommendations that co-occurring depression be diagnosed and treated when it can be observed to persist after at least a brief period of abstinence from substance use.68,69,75 Hospitalization is less an option in the era of managed care, nor is it acceptable to many patients. But abstinence can be initiated among outpatients with psychosocial methods, such as brief physician's advice to quit,76 motivational interviewing,77 or voucher incentives.78 Furthermore, some studies20,21,23,24 that enrolled patients who were actively using alcohol or other drugs observed medication effects. These studies imposed additional diagnostic criteria consistent with the DSM-IV15 constructs of primary and substance induced depression (eg, requiring that depression have antedated the onset of substance abuse, have persisted during past periods of abstinence, or have been of long duration in the current episode). However, several of the negative studies also used such critera.25,26,28,29 Future research should refine such diagnostic methods in order to better distinguish outpatient substance abusers who would benefit from antidepressant treatment.

Inclusion of a manual-guided psychosocial intervention as background treatment was associated with lower effect sizes, and of the 5 studies23-25,28,30 in this grouping, 4 of them23,25,28,30 had high-placebo response. Such interventions may improve mood by reducing substance use or may have intrinsic antidepressant effects. For example, cognitive behavioral relapse prevention approaches include modules on coping with dysphoric symptoms, and 2 of the studies we reviewed25,28 actually augmented this aspect of their interventions. Cognitive behavior therapy also showed promise in a study of depressed alcoholics.79 The findings suggest a clinical approach that begins with an evidence-based psychosocial intervention, followed by antidepressant medication if depression does not improve.

The moderating effect of sex should be approached with caution because it was mainly accounted for by predominantly male samples in 3 of the studies that required initial abstinence and had large effects.1719 One study25 that examined sex as a moderator within its sample found the opposite pattern—a significant antidepressant effect of sertraline among women but not men.

Selective serotonin reuptake inhibitors performed less well overall than tricyclics or other classes of antidepressants. This finding too must be viewed with caution, since all 5 SSRI studies that were negative25,26,2830 had high-placebo response rates. The 2 SSRI studies with low placebo response18,22 yielded substantial medication effects. Selective serotonin reuptake inhibitors remain an attractive first-line treatment among substance-dependent patients because of their tolerability, low toxic effects, and minimal sedating effects. If an SSRI trial fails, consideration should be given to an agent with noradrenergic or other mechanisms.

Strengths and Limitations

Treatment studies on comorbidity are challenging to carry out because with inclusion criteria requiring 2 current disorders, recruitment is slow. Hence, the single-site studies gathered for this meta-analysis offered, for the most part, small sample sizes and limited statistical power. Combining these studies resulted in a total sample of more than 800 patients and greater power to estimate treatment effects. This meta-analysis was also facilitated by relatively consistent methods and quality indicators across studies although variations in substance use outcome measures may have limited precision to measure treatment effects in that domain. The funnel plot and sensitivity analysis suggest only limited risk of publication bias, consistent with studies in this field being mainly investigator-initiated and thus likely to get published, whether positive or negative.

The variation in sample characteristics across studies, ranging from inpatient or outpatient alcoholics to opiate- or cocaine-dependent patients, can be viewed as a strength with respect to the generalizability of the findings, but it also raises concerns. The neuropharmacology of depression may differ between various substance-dependent populations, suggesting more treatment research is needed particularly among drug-dependent groups. The moderators were identified retrospectively, so that unmeasured confounders could explain the findings, and the moderators often overlapped. Because this is a study-level rather than a patient-level meta-analysis, moderators that reflect patient characteristics, such as sex, must be viewed with particular caution, and the underlying distributions of outcome or moderator variables cannot be checked. Future studies should isolate and examine factors such as diagnostic methods, medication type, and concurrent psychosocial interventions prospectively, likely requiring larger sample sizes and multisite designs.

The meta-analysis was restricted to studies on treatment of depressive disorders. A number of excluded studies selected substance abusers with depressive symptoms measured by cross-sectional scales, some43,44,46,49 with positive results. Screening instruments80 have effectively identified treatable depression in primary care settings. Future research should similarly examine brief screens as part of the effort to develop diagnostic approaches that are cost effective and widely disseminable.

Conclusions

Little more than a decade ago, the treatment of depression among substance dependent-patients was discouraged by most clinicians, given evidence on the transient nature of depressive symptoms in the setting of substance abuse,68,69,7174 equivocal findings of earlier treatment studies,1012 and concern that focus on depression not distract attention from treatment of the addiction. Recent years have witnessed increased acceptance of such treatment although the likelihood that depression will be recognized among substance-dependent patients continues to vary by treatment setting,81 and, as in the general population, undertreatment is still probably the rule.82,83 The studies reviewed herein lend empirical support to recent recommendations that drug or alcohol abuse not be a barrier to treatment of depression8 but suggest that care is needed in diagnosis either to observe patients during at least a brief period of abstinence prior to diagnosis and treatment of depression or to make efforts to distinguish treatable depression by history.

Antidepressant treatment may have a limited impact on alcohol or other drug use, and attention is needed in the treatment plan to specific psychosocial or pharmacologic interventions targeting addiction itself. At the same time, this should not detract from the importance of treating depression, given growing evidence of its adverse prognostic implications in a variety of health domains.8,83

Regier DA, Farmer ME, Rae DS, Locke BZ, Keith SJ, Judd LL, Goodwin FK. Comorbidity of mental disorders with alcohol and other drug abuse: results from the Epidemiologic Catchment Area (ECA) Study.  JAMA.1990;264:2511-2518.
PubMed
Kessler RC, Crum RM, Warner LA, Nelson CB, Schulenberg J, Anthony JC. Lifetime co-occurrence of DSM-III-R alcohol abuse and dependence with other psychiatric disorders in the National Comorbidity Survey.  Arch Gen Psychiatry.1997;54:313-321.
PubMed
Hasin DS, Nunes EV. Comorbidity of alcohol, drug, and psychiatric disorders: epidemiology. In: Dual Diagnosis and Treatment: Substance Abuse and Psychiatric Disorders. Kranzler HR, Rounsaville BJ, eds. New York, NY: Marcel Decker Inc; 1988.
Murphy GE, Wetzel RD, Robins E, McEvoy L. Multiple risk factors predict suicide in alcoholism.  Arch Gen Psychiatry.1992;49:459-463.
PubMed
Levy JC, Deykin EY. Suicidality, depression, and substance abuse in adolescence.  Am J Psychiatry.1989;146:1462-1467.
PubMed
Tondo L, Baldessarini RJ, Hennen J.  et al.  Suicide attempts in major affective disorder patients with comorbid substance use disorders.  J Clin Psychiatry.1999;60(suppl 2):63-69.
PubMed
Levin FR, Gunderson EW, Levounis P. Medical education. In: Galanter M, Kleber H, eds. Textbook of Subsantance Abuse Treatment, 3rd ed. Washington, DC: American Psychiatric Press. In press.
Whooley MA, Simon GE. Managing depression in medical outpatients.  N Engl J Med.2000;343:1942-1950.
PubMed
Weiss RD, Najavits LM, Hennesy G. Overview of treatment modalities for dual diagnosis patients: pharmacotherapy, psychotherapy, and 12-step programs. In: Kranzler HR, Tinsley JA, eds, Dual Diagnosis and Treatment: Substance Abuse and Comorbid Disorders, 2nd ed. New York, NY: Marcel Dekker. In press.
Ciraulo DA, Jaffe JH. Tricyclic antidepressants in the treatment of depression associated with alcoholism.  J Clin Psychopharmacol.1981;1:146-150.
PubMed
Liskow BI, Goodwin DW. Pharmacologic treatment of alcohol intoxication, withdrawal, and dependence: a critical review.  J Stud Alcohol.1987;48:356-370.
PubMed
Nunes E, Quitkin F, Brady R, Post-Koenig T. Antidepressant treatment in methadone maintenance patients.  J Addict Dis.1994;13:13-24.
PubMed
American Psychiatric Association.  Diagnostic and Statistical Manual of Mental Disorders, Third Edition. Washington, DC: American Psychiatric Association; 1980.
American Psychiatric Association.  Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition. Washington, DC: American Psychiatric Association; 1987.
American Psychiatric Association.  Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric Association; 1994.
Yusuf S, Wittes J, Probstfield J, Tyroler HA. Analysis and interpretation of treatment effects in subgroups of patients in randomized clinical trials.  JAMA.1991;266:93-98.
PubMed
Altamura AC, Mauri MC, Girardi T, Panetta B. Alcoholism and depression: a placebo controlled study with viloxazine.  Int J Clin Pharmacol Res.1990;10:293-298.
PubMed
Roy A. Placebo-controlled study of sertraline in depressed recently abstinent alcoholics.  Biol Psychiatry.1998;44:633-637.
PubMed
Mason BJ, Kocsis JH, Ritvo EC, Cutler RB. A double-blind, placebo-controlled trial of desipramine for primary alcohol dependence stratified on the presence or absence of major depression.  JAMA.1996;275:761-767.
PubMed
Nunes EV, Quitkin FM, Donovan SJ.  et al.  Imipramine treatment of opiate-dependent patients with depressive disorders: a placebo-controlled trial.  Arch Gen Psychiatry.1998;55:153-160.
PubMed
Nunes EV, McGrath PJ, Quitkin FM.  et al.  Imipramine treatment of cocaine abuse: possible boundaries of efficacy.  Drug Alcohol Depend.1995;39:185-195.
PubMed
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.
PubMed
Roy-Byrne PP, Pages KP, Russo JE.  et al.  Nefazodone treatment of major depression in alcohol-dependent patients: a double-blind, placebo-controlled trial.  J Clin Psychopharmacol.2000;20:129-136.
PubMed
McGrath PJ, Nunes EV, Stewart JW.  et al.  Imipramine treatment of alcoholics with primary depression: a placebo-controlled clinical trial.  Arch Gen Psychiatry.1996;53:232-240.
PubMed
Moak DH, Anton RF, Latham PK, Voronin KE, Waid RL, Durazo-Arvizu R. Sertraline and cognitive behavioral therapy for depressed alcoholics: results of a placebo-controlled trial.  J Clin Psychopharmacol.2003;23:553-562.
PubMed
Carpenter KM, Brooks AC, Vosburg SK, Nunes EV. The effect of sertraline and environmental context on treating depression and illicit substance use among methadone maintained opiate dependent patients: a controlled clinical trial.  Drug Alcohol Depend.In press.
Kleber HD, Weissman MM, Rounsaville BJ, Wilber CH, Prusoff BA, Riordan CE. Imipramine as treatment for depression in addicts.  Arch Gen Psychiatry.1983;40:649-653.
PubMed
Schmitz JM, Averill P, Stotts AL, Moeller FG, Rhoades HM, Grabowski J. Fluoxetine treatment of cocaine-dependent patients with major depressive disorder.  Drug Alcohol Depend.2001;63:207-214.
PubMed
Petrakis I, Carroll KM, Nich C, Gordon L, Kosten T, Rounsaville B. Fluoxetine treatment of depressive disorders in methadone-maintained opiate addicts.  Drug Alcohol Depend.1998;50:221-226.
PubMed
Pettinati HM, Volpicelli JR, Luck G, Kranzler HR, Rukstalis MR, Cnaan A. Double-blind clinical trial of sertraline treatment for alcohol dependence.  J Clin Psychopharmacol.2001;21:143-153.
PubMed
Hamilton MA. A rating scale for depression.  J Neurol Neurosurg Psychiatry.1960;23:56-62.
PubMed
Detsky AS, Naylor CD, O'Rourke K, McGeer AJ, L'Abbe KA. Incorporating variations in the quality of individual randomized trials into meta-analysis.  J Clin Epidemiol.1992;45:255-265.
PubMed
Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias: dimensions of methodological quality associated with estimates of treatment effects in controlled trials.  JAMA.1995;273:408-412.
PubMed
Borenstein M, Borenstein H. Comprehensive Meta-AnalysisEnglewood, NJ: Biostat Inc; 1999.
Copas JB, Shi JQ. A sensitivity analysis for publication bias in systematic reviews.  Stat Methods Med Res.2001;10:251-265.
PubMed
Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis.  Stat Med.2002;21:1539-1558.
PubMed
Lipsey MW, Wilson DB. Practical Meta-AnalysisThousand Oaks, Calif: Sage Publications; 2003. Applied Social Research Methods Series, vol 49.
Raskin A, Schulterbrandt J, Reatig N.  et al.  Replication of factors of psychopathology in interview, ward behavior and self-report ratings of hospitalized depressives.  J Nerv Ment Dis.1969;148:87-98.
PubMed
Wilson IC, Alltop LB, Riley L. Tofranil in the treatment of post alcoholic depressions.  Psychosomatics.1970;11:488-494.
PubMed
Butterworth AT. Depression associated with alcohol withdrawal: imipramine therapy compared with placebo.  Q J Stud Alcohol.1971;32:343-348.
PubMed
Shaw JA, Donley P, Morgan DW, Robinson JA. Treatment of depression in alcoholics.  Am J Psychiatry.1975;132:641-644.
PubMed
Feldman HS. Antidepressant drug therapy: addicts vs nonaddicts.  Psychosomatics.1972;13:41-48.
PubMed
Woody GE, O'Brien CP, Rickels K. Depression and anxiety in heroin addicts: a placebo-controlled study of doxepin in combination with methadone.  Am J Psychiatry.1975;132:447-450.
PubMed
Titievsky J, Seco G, Barranco M, Kyle EM. Doxepin as adjunctive therapy for depressed methadone patients: a double-blind study.  J Clin Psychiatry.1982;43:454-456.
PubMed
Dean AJ, Bell J, Mascord DJ, Parker G, Christie MJ. A randomized, controlled trial of fluoxetine in methadone maintenance patients with depressive symptoms.  J Affect Disord.2002;72:85-90.
PubMed
Krupitsky EM, Burakov AM, Ivanov VB.  et al.  Baclofen administration for the treatment of affective disorders in alcoholic patients.  Drug Alcohol Depend.1993;33:157-163.
PubMed
Siris SG, Bermanzohn PC, Mason SE, Rifkin A, Alvir JM. Adjunctive imipramine for dysphoric schizophrenic patients with past histories of cannabis abuse.  Prog Neuropsychopharmacol Biol Psychiatry.1992;16:539-547.
PubMed
Giannini AJ, Malone DA, Giannini MC, Price WA, Loiselle RH. Treatment of depression in chronic cocaine and phencyclidine abuse with desipramine.  J Clin Pharmacol.1986;26:211-214.
PubMed
Carroll KM, Nich C, Rounsaville BJ. Differential symptom reduction in depressed cocaine abusers treated with psychotherapy and pharmacotherapy.  J Nerv Ment Dis.1995;183:251-259.
PubMed
Powell BJ, Campbell JL, Landon JF.  et al.  A double-blind, placebo-controlled study of nortriptyline and bromocriptine in male alcoholics subtyped by comorbid psychiatric disorders.  Alcohol Clin Exp Res.1995;19:462-468.
PubMed
Goldstein, McBride, Westy.  Cited by: Woody GE, O'Brien CP, McLellan AT, Marcovici M, Evans BD. The use of antidepressants with methadone in depressed maintenance patients.  Ann N Y Acad Sci.1982;398:120-127.
PubMed
Batki S, Wheeler S, Jones R, Sorensen J, Brennan K. Doxepin Treatment of Depressed Opiate Addicts Undergoing Methadone DetoxificationWashington, DC: National Institute on Drug Abuse; 1987:292. Research Monograph Series, No. 81.
McLean PC, Ancill RJ, Szulecka TK. Mianserin in the treatment of depressive symptoms in alcoholics: a double-blind placebo controlled study using a computer delivered self-rating scale.  Psychiatr Pol.1986;20:417-427.
PubMed
Gonzalez G, Feingold A, Oliveto A, Gonsai K, Kosten TR. Comorbid major depressive disorder as a prognostic factor in cocaine-abusing buprenorphine-maintained patients treated with desipramine and contingency management.  Am J Drug Alcohol Abuse.2003;29:497-514.
PubMed
Cornelius JR, Salloum IM, Thase ME.  et al.  Fluoxetine vs placebo in depressed alcoholic cocaine abusers.  Psychopharmacol Bull.1998;34:117-121.
PubMed
Cornelius JR, Salloum IM, Ehler JG.  et al.  Double-blind fluoxetine in depressed alcoholic smokers.  Psychopharmacol Bull.1997;33:165-167.
PubMed
Cornelius JR, Salloum IM, Haskett RF.  et al.  Fluoxetine vs placebo for the marijuana use of depressed alcoholics.  Addict Behav.1999;24:111-114.
PubMed
Cornelius JR, Salloum IM, Haskett RF. Fluoxetine vs placebo in depressed alcoholics: a 1-year follow up study.  Addict Behav.2000;25:307-310.
PubMed
Kranzler HR, Burleson JA, Korner P.  et al.  Placebo-controlled trial of fluoxetine as an adjunct to relapse prevention in alcoholics.  Am J Psychiatry.1995;152:391-397.
PubMed
Ziedonis DM, Kosten TR. Depression as a prognostic factor for pharmacologic treatment of cocaine dependence.  Psychopharmacol Bull.1991;27:337-343.
PubMed
Margolin A, Kosten TR, Avants SK.  et al.  A multicenter trial of bupropion for cocaine dependence in methadone-maintained patients.  Drug Alcohol Depend.1995;40:125-131.
PubMed
Mason BJ, Kocsis JH. Desipramine treatment of alcoholism.  Psychopharmacol Bull.1991;27:155-161.
PubMed
Cornelius JR, Salloum IM, Cornelius MD.  et al.  Preliminary report: double-blind, placebo-controlled study of fluoxetine in depressed alcoholics.  Psychopharmacol Bull.1995;31:297-303.
PubMed
Nunes EV, McGrath PJ, Quitkin FM.  et al.  Imipramine treatment of alcoholism with comorbid depression.  Am J Psychiatry.1993;150:963-965.
PubMed
Kranzler H.for the Sertraline for Depression in Alcoholism Study Group.  Sertraline treatment of major depression in alcohol dependence. In: Program and abstracts of the 26th Annual meeting of the Research Society on Alcoholism; July 1, 2002; San Francisco, Calif.
Cohen J. Statistical Power Analysis for the Behavioral Sciences2nd ed. Hillsdale, New Jersey: Lawrence Erlbaum Associates; 1988.
Walsh BT, Seidman SN, Sysko R, Gould M. Placebo response in studies of major depression; variable, substantial, and growing.  JAMA.2002;287:1840-1847.
PubMed
Schuckit MA. Genetic and clinical implications of alcoholism and affective disorder.  Am J Psychiatry.1986;143:140-147.
PubMed
Brown SA, Schuckit MA. Changes in depression among abstinent alcoholics.  J Stud Alcohol.1988;49:412-417.
PubMed
Liappas J, Paparrigopoulos E, Tzavellas G, Christodoulou G. Impact of alcohol detoxification on anxiety and depressive symptoms.  Drug Alcohol Depend.2002;68:215-220.
PubMed
Weddington WW, Brown BS, Haertzen CA.  et al.  Changes in mood, craving, and sleep during short-term abstinence reported by male cocaine addicts; a controlled residential study.  Arch Gen Psychiatry.1990;47:861-868.
PubMed
Satel SL, Price LH, Palumbo JM.  et al.  Clinical phenomenology and neurobiology of cocaine abstinence: a prospective inpatient study.  Am J Psychiatry.1991;148:1712-1716.
PubMed
DeLeon G, Skodol A, Rosenthal MS. Phoenix House: changes in psychopathological signs of resident addicts.  Arch Gen Psychiatry.1973;28:131-135.
PubMed
Rounsaville BJ, Kosten TR, Kleber HD. Long-term changes in current psychiatric diagnoses of treated opiate addicts.  Compr Psychiatry.1986;27:480-498.
PubMed
Brady KT, Halligan P, Malcolm RJ. Dual diagnosis. In: Galanter M and Kleber HD, eds. Textbook of Substance Abuse Treatment. 2nd ed. Washington, DC: American Psychiatric Press; 1975:475-483.
Fleming MF, Barry KL, Manwell LB, Johnson K, London R. Brief physician advice for problem alcohol drinkers: a randomized controlled trial in community-based primary care practices.  JAMA.1997;277:1039-1045.
PubMed
Miller WR, Rollnick S. Motivational Interviewing: Preparing People for Change2nd ed. New York, NY: Guilford Press; 2002.
Higgins ST, Budney AJ, Bickel WK, Hughes JR, Foerg F, Badger G. Achieving cocaine abstinence with a behavioral approach.  Am J Psychiatry.1993;150:763-769.
PubMed
Brown RA, Evans DM, Miller IW, Burgess ES, Mueller TI. Cognitive-behavioral treatment for depression in alcoholism.  J Consult Clin Psychol.1997;65:715-726.
PubMed
Spitzer RL, Kroenke K, Williams JBW.for the Patient Health Questionnaire Primary Care Study Group.  Validity and utility of a self-report version of PRIME-MD: the PHQ Primary Care Study.  JAMA.1999;282:1737-1744.
PubMed
Petrakis IL, Leslie D, Rosenheck R. The use of antidepressants in alcohol dependent veterans.  J Clin Psychiatry.2003;64:865-870.
PubMed
Young AS, Klap R, Sherbourne CD, Wells KB. The quality of care for depressive and anxiety disorders in the United States.  Arch Gen Psychiatry.2001;58:55-61.
PubMed
Kessler RC, Berglund P, Demler O.  et al.  The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R).  JAMA.2003;289:3095-3105.
PubMed

Figures

Figure 1. Effect of Antidepressant Medication on Outcome of Depression (Hamilton Depression Scale)
Graphic Jump Location
Effect sizes are Cohen d, error bars represent the 95% confidence intervals (CIs). Studies are stratified into groups with low (<25%) vs high (>25%) categories of placebo response, measured as the percentage decrease in the Hamilton Depression Scale between baseline and end-of-study in the placebo group.
Figure 2. Effect of Antidepressant Medication on Outcome of Substance Use*
Graphic Jump Location
Effect sizes are Cohen d on outcomes reflecting quantity of substance use (mainly by self-report). Error bars represent the 95% confidence intervals (CIs). Studies are stratified into 2 groups based on the size of the effect of medication on depression outcome (Cohen d for the Hamilton Depression Scale) effect size lower than 0.50 (low); and effect size greater than 0.50 (high). The study by Roy18 did not have the data necessary for analysis for this figure.

Tables

Table Graphic Jump LocationTable 1. Placebo-Controlled Trials of Antidepressants in Substance-Dependent Patients With Depression: Methodology, Placebo Response, and Effect Size on the Hamilton Depression Scale
Table Graphic Jump LocationTable 2. Effects of Antidepressant Medication Treatment on Rates of Depression Response, and Abstinence or Sustained Remission From Substance Abuse*
Table Graphic Jump LocationTable 3. Moderators of Effect of Antidepressant Medication on Outcome of the Hamilton Depression Scale in Substance-Dependent Patients With Depressive Disorders

References

Regier DA, Farmer ME, Rae DS, Locke BZ, Keith SJ, Judd LL, Goodwin FK. Comorbidity of mental disorders with alcohol and other drug abuse: results from the Epidemiologic Catchment Area (ECA) Study.  JAMA.1990;264:2511-2518.
PubMed
Kessler RC, Crum RM, Warner LA, Nelson CB, Schulenberg J, Anthony JC. Lifetime co-occurrence of DSM-III-R alcohol abuse and dependence with other psychiatric disorders in the National Comorbidity Survey.  Arch Gen Psychiatry.1997;54:313-321.
PubMed
Hasin DS, Nunes EV. Comorbidity of alcohol, drug, and psychiatric disorders: epidemiology. In: Dual Diagnosis and Treatment: Substance Abuse and Psychiatric Disorders. Kranzler HR, Rounsaville BJ, eds. New York, NY: Marcel Decker Inc; 1988.
Murphy GE, Wetzel RD, Robins E, McEvoy L. Multiple risk factors predict suicide in alcoholism.  Arch Gen Psychiatry.1992;49:459-463.
PubMed
Levy JC, Deykin EY. Suicidality, depression, and substance abuse in adolescence.  Am J Psychiatry.1989;146:1462-1467.
PubMed
Tondo L, Baldessarini RJ, Hennen J.  et al.  Suicide attempts in major affective disorder patients with comorbid substance use disorders.  J Clin Psychiatry.1999;60(suppl 2):63-69.
PubMed
Levin FR, Gunderson EW, Levounis P. Medical education. In: Galanter M, Kleber H, eds. Textbook of Subsantance Abuse Treatment, 3rd ed. Washington, DC: American Psychiatric Press. In press.
Whooley MA, Simon GE. Managing depression in medical outpatients.  N Engl J Med.2000;343:1942-1950.
PubMed
Weiss RD, Najavits LM, Hennesy G. Overview of treatment modalities for dual diagnosis patients: pharmacotherapy, psychotherapy, and 12-step programs. In: Kranzler HR, Tinsley JA, eds, Dual Diagnosis and Treatment: Substance Abuse and Comorbid Disorders, 2nd ed. New York, NY: Marcel Dekker. In press.
Ciraulo DA, Jaffe JH. Tricyclic antidepressants in the treatment of depression associated with alcoholism.  J Clin Psychopharmacol.1981;1:146-150.
PubMed
Liskow BI, Goodwin DW. Pharmacologic treatment of alcohol intoxication, withdrawal, and dependence: a critical review.  J Stud Alcohol.1987;48:356-370.
PubMed
Nunes E, Quitkin F, Brady R, Post-Koenig T. Antidepressant treatment in methadone maintenance patients.  J Addict Dis.1994;13:13-24.
PubMed
American Psychiatric Association.  Diagnostic and Statistical Manual of Mental Disorders, Third Edition. Washington, DC: American Psychiatric Association; 1980.
American Psychiatric Association.  Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition. Washington, DC: American Psychiatric Association; 1987.
American Psychiatric Association.  Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric Association; 1994.
Yusuf S, Wittes J, Probstfield J, Tyroler HA. Analysis and interpretation of treatment effects in subgroups of patients in randomized clinical trials.  JAMA.1991;266:93-98.
PubMed
Altamura AC, Mauri MC, Girardi T, Panetta B. Alcoholism and depression: a placebo controlled study with viloxazine.  Int J Clin Pharmacol Res.1990;10:293-298.
PubMed
Roy A. Placebo-controlled study of sertraline in depressed recently abstinent alcoholics.  Biol Psychiatry.1998;44:633-637.
PubMed
Mason BJ, Kocsis JH, Ritvo EC, Cutler RB. A double-blind, placebo-controlled trial of desipramine for primary alcohol dependence stratified on the presence or absence of major depression.  JAMA.1996;275:761-767.
PubMed
Nunes EV, Quitkin FM, Donovan SJ.  et al.  Imipramine treatment of opiate-dependent patients with depressive disorders: a placebo-controlled trial.  Arch Gen Psychiatry.1998;55:153-160.
PubMed
Nunes EV, McGrath PJ, Quitkin FM.  et al.  Imipramine treatment of cocaine abuse: possible boundaries of efficacy.  Drug Alcohol Depend.1995;39:185-195.
PubMed
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.
PubMed
Roy-Byrne PP, Pages KP, Russo JE.  et al.  Nefazodone treatment of major depression in alcohol-dependent patients: a double-blind, placebo-controlled trial.  J Clin Psychopharmacol.2000;20:129-136.
PubMed
McGrath PJ, Nunes EV, Stewart JW.  et al.  Imipramine treatment of alcoholics with primary depression: a placebo-controlled clinical trial.  Arch Gen Psychiatry.1996;53:232-240.
PubMed
Moak DH, Anton RF, Latham PK, Voronin KE, Waid RL, Durazo-Arvizu R. Sertraline and cognitive behavioral therapy for depressed alcoholics: results of a placebo-controlled trial.  J Clin Psychopharmacol.2003;23:553-562.
PubMed
Carpenter KM, Brooks AC, Vosburg SK, Nunes EV. The effect of sertraline and environmental context on treating depression and illicit substance use among methadone maintained opiate dependent patients: a controlled clinical trial.  Drug Alcohol Depend.In press.
Kleber HD, Weissman MM, Rounsaville BJ, Wilber CH, Prusoff BA, Riordan CE. Imipramine as treatment for depression in addicts.  Arch Gen Psychiatry.1983;40:649-653.
PubMed
Schmitz JM, Averill P, Stotts AL, Moeller FG, Rhoades HM, Grabowski J. Fluoxetine treatment of cocaine-dependent patients with major depressive disorder.  Drug Alcohol Depend.2001;63:207-214.
PubMed
Petrakis I, Carroll KM, Nich C, Gordon L, Kosten T, Rounsaville B. Fluoxetine treatment of depressive disorders in methadone-maintained opiate addicts.  Drug Alcohol Depend.1998;50:221-226.
PubMed
Pettinati HM, Volpicelli JR, Luck G, Kranzler HR, Rukstalis MR, Cnaan A. Double-blind clinical trial of sertraline treatment for alcohol dependence.  J Clin Psychopharmacol.2001;21:143-153.
PubMed
Hamilton MA. A rating scale for depression.  J Neurol Neurosurg Psychiatry.1960;23:56-62.
PubMed
Detsky AS, Naylor CD, O'Rourke K, McGeer AJ, L'Abbe KA. Incorporating variations in the quality of individual randomized trials into meta-analysis.  J Clin Epidemiol.1992;45:255-265.
PubMed
Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias: dimensions of methodological quality associated with estimates of treatment effects in controlled trials.  JAMA.1995;273:408-412.
PubMed
Borenstein M, Borenstein H. Comprehensive Meta-AnalysisEnglewood, NJ: Biostat Inc; 1999.
Copas JB, Shi JQ. A sensitivity analysis for publication bias in systematic reviews.  Stat Methods Med Res.2001;10:251-265.
PubMed
Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis.  Stat Med.2002;21:1539-1558.
PubMed
Lipsey MW, Wilson DB. Practical Meta-AnalysisThousand Oaks, Calif: Sage Publications; 2003. Applied Social Research Methods Series, vol 49.
Raskin A, Schulterbrandt J, Reatig N.  et al.  Replication of factors of psychopathology in interview, ward behavior and self-report ratings of hospitalized depressives.  J Nerv Ment Dis.1969;148:87-98.
PubMed
Wilson IC, Alltop LB, Riley L. Tofranil in the treatment of post alcoholic depressions.  Psychosomatics.1970;11:488-494.
PubMed
Butterworth AT. Depression associated with alcohol withdrawal: imipramine therapy compared with placebo.  Q J Stud Alcohol.1971;32:343-348.
PubMed
Shaw JA, Donley P, Morgan DW, Robinson JA. Treatment of depression in alcoholics.  Am J Psychiatry.1975;132:641-644.
PubMed
Feldman HS. Antidepressant drug therapy: addicts vs nonaddicts.  Psychosomatics.1972;13:41-48.
PubMed
Woody GE, O'Brien CP, Rickels K. Depression and anxiety in heroin addicts: a placebo-controlled study of doxepin in combination with methadone.  Am J Psychiatry.1975;132:447-450.
PubMed
Titievsky J, Seco G, Barranco M, Kyle EM. Doxepin as adjunctive therapy for depressed methadone patients: a double-blind study.  J Clin Psychiatry.1982;43:454-456.
PubMed
Dean AJ, Bell J, Mascord DJ, Parker G, Christie MJ. A randomized, controlled trial of fluoxetine in methadone maintenance patients with depressive symptoms.  J Affect Disord.2002;72:85-90.
PubMed
Krupitsky EM, Burakov AM, Ivanov VB.  et al.  Baclofen administration for the treatment of affective disorders in alcoholic patients.  Drug Alcohol Depend.1993;33:157-163.
PubMed
Siris SG, Bermanzohn PC, Mason SE, Rifkin A, Alvir JM. Adjunctive imipramine for dysphoric schizophrenic patients with past histories of cannabis abuse.  Prog Neuropsychopharmacol Biol Psychiatry.1992;16:539-547.
PubMed
Giannini AJ, Malone DA, Giannini MC, Price WA, Loiselle RH. Treatment of depression in chronic cocaine and phencyclidine abuse with desipramine.  J Clin Pharmacol.1986;26:211-214.
PubMed
Carroll KM, Nich C, Rounsaville BJ. Differential symptom reduction in depressed cocaine abusers treated with psychotherapy and pharmacotherapy.  J Nerv Ment Dis.1995;183:251-259.
PubMed
Powell BJ, Campbell JL, Landon JF.  et al.  A double-blind, placebo-controlled study of nortriptyline and bromocriptine in male alcoholics subtyped by comorbid psychiatric disorders.  Alcohol Clin Exp Res.1995;19:462-468.
PubMed
Goldstein, McBride, Westy.  Cited by: Woody GE, O'Brien CP, McLellan AT, Marcovici M, Evans BD. The use of antidepressants with methadone in depressed maintenance patients.  Ann N Y Acad Sci.1982;398:120-127.
PubMed
Batki S, Wheeler S, Jones R, Sorensen J, Brennan K. Doxepin Treatment of Depressed Opiate Addicts Undergoing Methadone DetoxificationWashington, DC: National Institute on Drug Abuse; 1987:292. Research Monograph Series, No. 81.
McLean PC, Ancill RJ, Szulecka TK. Mianserin in the treatment of depressive symptoms in alcoholics: a double-blind placebo controlled study using a computer delivered self-rating scale.  Psychiatr Pol.1986;20:417-427.
PubMed
Gonzalez G, Feingold A, Oliveto A, Gonsai K, Kosten TR. Comorbid major depressive disorder as a prognostic factor in cocaine-abusing buprenorphine-maintained patients treated with desipramine and contingency management.  Am J Drug Alcohol Abuse.2003;29:497-514.
PubMed
Cornelius JR, Salloum IM, Thase ME.  et al.  Fluoxetine vs placebo in depressed alcoholic cocaine abusers.  Psychopharmacol Bull.1998;34:117-121.
PubMed
Cornelius JR, Salloum IM, Ehler JG.  et al.  Double-blind fluoxetine in depressed alcoholic smokers.  Psychopharmacol Bull.1997;33:165-167.
PubMed
Cornelius JR, Salloum IM, Haskett RF.  et al.  Fluoxetine vs placebo for the marijuana use of depressed alcoholics.  Addict Behav.1999;24:111-114.
PubMed
Cornelius JR, Salloum IM, Haskett RF. Fluoxetine vs placebo in depressed alcoholics: a 1-year follow up study.  Addict Behav.2000;25:307-310.
PubMed
Kranzler HR, Burleson JA, Korner P.  et al.  Placebo-controlled trial of fluoxetine as an adjunct to relapse prevention in alcoholics.  Am J Psychiatry.1995;152:391-397.
PubMed
Ziedonis DM, Kosten TR. Depression as a prognostic factor for pharmacologic treatment of cocaine dependence.  Psychopharmacol Bull.1991;27:337-343.
PubMed
Margolin A, Kosten TR, Avants SK.  et al.  A multicenter trial of bupropion for cocaine dependence in methadone-maintained patients.  Drug Alcohol Depend.1995;40:125-131.
PubMed
Mason BJ, Kocsis JH. Desipramine treatment of alcoholism.  Psychopharmacol Bull.1991;27:155-161.
PubMed
Cornelius JR, Salloum IM, Cornelius MD.  et al.  Preliminary report: double-blind, placebo-controlled study of fluoxetine in depressed alcoholics.  Psychopharmacol Bull.1995;31:297-303.
PubMed
Nunes EV, McGrath PJ, Quitkin FM.  et al.  Imipramine treatment of alcoholism with comorbid depression.  Am J Psychiatry.1993;150:963-965.
PubMed
Kranzler H.for the Sertraline for Depression in Alcoholism Study Group.  Sertraline treatment of major depression in alcohol dependence. In: Program and abstracts of the 26th Annual meeting of the Research Society on Alcoholism; July 1, 2002; San Francisco, Calif.
Cohen J. Statistical Power Analysis for the Behavioral Sciences2nd ed. Hillsdale, New Jersey: Lawrence Erlbaum Associates; 1988.
Walsh BT, Seidman SN, Sysko R, Gould M. Placebo response in studies of major depression; variable, substantial, and growing.  JAMA.2002;287:1840-1847.
PubMed
Schuckit MA. Genetic and clinical implications of alcoholism and affective disorder.  Am J Psychiatry.1986;143:140-147.
PubMed
Brown SA, Schuckit MA. Changes in depression among abstinent alcoholics.  J Stud Alcohol.1988;49:412-417.
PubMed
Liappas J, Paparrigopoulos E, Tzavellas G, Christodoulou G. Impact of alcohol detoxification on anxiety and depressive symptoms.  Drug Alcohol Depend.2002;68:215-220.
PubMed
Weddington WW, Brown BS, Haertzen CA.  et al.  Changes in mood, craving, and sleep during short-term abstinence reported by male cocaine addicts; a controlled residential study.  Arch Gen Psychiatry.1990;47:861-868.
PubMed
Satel SL, Price LH, Palumbo JM.  et al.  Clinical phenomenology and neurobiology of cocaine abstinence: a prospective inpatient study.  Am J Psychiatry.1991;148:1712-1716.
PubMed
DeLeon G, Skodol A, Rosenthal MS. Phoenix House: changes in psychopathological signs of resident addicts.  Arch Gen Psychiatry.1973;28:131-135.
PubMed
Rounsaville BJ, Kosten TR, Kleber HD. Long-term changes in current psychiatric diagnoses of treated opiate addicts.  Compr Psychiatry.1986;27:480-498.
PubMed
Brady KT, Halligan P, Malcolm RJ. Dual diagnosis. In: Galanter M and Kleber HD, eds. Textbook of Substance Abuse Treatment. 2nd ed. Washington, DC: American Psychiatric Press; 1975:475-483.
Fleming MF, Barry KL, Manwell LB, Johnson K, London R. Brief physician advice for problem alcohol drinkers: a randomized controlled trial in community-based primary care practices.  JAMA.1997;277:1039-1045.
PubMed
Miller WR, Rollnick S. Motivational Interviewing: Preparing People for Change2nd ed. New York, NY: Guilford Press; 2002.
Higgins ST, Budney AJ, Bickel WK, Hughes JR, Foerg F, Badger G. Achieving cocaine abstinence with a behavioral approach.  Am J Psychiatry.1993;150:763-769.
PubMed
Brown RA, Evans DM, Miller IW, Burgess ES, Mueller TI. Cognitive-behavioral treatment for depression in alcoholism.  J Consult Clin Psychol.1997;65:715-726.
PubMed
Spitzer RL, Kroenke K, Williams JBW.for the Patient Health Questionnaire Primary Care Study Group.  Validity and utility of a self-report version of PRIME-MD: the PHQ Primary Care Study.  JAMA.1999;282:1737-1744.
PubMed
Petrakis IL, Leslie D, Rosenheck R. The use of antidepressants in alcohol dependent veterans.  J Clin Psychiatry.2003;64:865-870.
PubMed
Young AS, Klap R, Sherbourne CD, Wells KB. The quality of care for depressive and anxiety disorders in the United States.  Arch Gen Psychiatry.2001;58:55-61.
PubMed
Kessler RC, Berglund P, Demler O.  et al.  The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R).  JAMA.2003;289:3095-3105.
PubMed

Letters

CME


You need to register in order to view this quiz.

Multimedia

Some tools below are only available to our subscribers or users with an online account.

Web of Science® Times Cited: 226

Related Content

Customize your page view by dragging & repositioning the boxes below.

Articles Related By Topic
Related Collections
PubMed Articles