Alcohol dependence frequently presents with comorbid depression. As pharmacological treatment for alcoholism becomes more common, with the advent of novel medications such as naltrexone and acamprosate, the issue of combined pharmacotherapy for comorbid depression in these patients will arise more frequently. There has been significant concern in the past about the use of antidepressant medication for depressed patients with comorbid alcohol dependence, and research has not always supported the efficacy of such treatment (1, 2). Coupled with the etiological focus of treating psychiatrists on whether the depression associated with alcoholism was primary or secondary, based on whether the alcoholism preceded or followed the depressive symptoms, there was considerable reluctance among treating physicians to aggressively medicate that depression, even "primary" depression. More recent efforts to distinguish among affective syndromes associated with alcoholism have incorporated the idea of substance-induced major depression (in which the depression occurs in the context of substance use or withdrawal) and substance-independent major depression (in which the depression either occurs before the onset of alcoholism or occurs during a period of abstinence from alcohol) (3).
Research is now showing the benefits of pharmacologically treating the depression associated with alcoholism, secondary as well as primary (4, 5). Given these advances, as well as advances in pharmacological treatment of alcoholism (6–8), it is inevitable that some patients will benefit from two treatment approaches: management of depression with an antidepressant and management of alcohol consumption and craving with naltrexone or acamprosate. To illustrate these important clinical and research issues and to highlight some of the current thinking about primary versus secondary depression and the more recent proposed dichotomy of substance-dependent and substance-independent depression (3), we present the following case from the Substance Abuse Treatment Unit at Yale University.
A 53-year-old married Caucasian woman came to the clinic in response to an advertisement recruiting patients with alcohol problems who were seeking treatment. She had two children, both of whom lived away from home, and she was living with her second husband. She was working as a temporary administrative assistant.
At the time of presentation, the patient was drinking a half-gallon of vodka per week. She drank at home, alone or with her husband, and believed she had had a problem for at least 8 years, with a constant steady intake over that time. She suffered from shaking and sweating at times but had never had delirium tremens or required treatment for alcohol withdrawal. Having first tried alcohol at age 16 in the setting of a family home where alcohol was always around, she was drinking regularly by her early 30s, often drinking vodka in the evenings. By her late 30s she had begun drinking more heavily, and she described herself as drinking to relieve any trouble or unhappiness. Her assessment of her own drinking pattern was that she drank excessively first and became depressed later. By her late 40s she was drinking a half-pint of vodka daily in the evenings, and drinking had become a noticeable problem for her. Her tolerance increased, and life began to revolve around alcohol. Her drinking problem was made worse by various personal problems, including the breakup of her first marriage, the loss of her employment owing to downsizing, and her declaration of personal bankruptcy. She had smoked up to 2 packs of cigarettes per day since her early 20s but had no other substance abuse history.
The patient grew up in a caring and loving family, the second of three children born to a sales representative and a homemaker. Her parents had a good marriage, and she had a good relationship with her parents. She had one older sister, with whom she was not close owing to a significant age gap, and a younger brother, with whom she got on well. She was an excellent student, and she graduated with high honors from both grade school and high school. Her time at school was pleasant, and she had a large number of friends and participated in various extracurricular activities. While at college in her late teens and early 20s she worked as a waitress to supplement her income. At 19 she married a firefighter and moved out of her parents’ house. She graduated from college with a bachelor’s degree in marketing in her mid-20s. She had two children, one boy and one girl, and had an excellent relationship with both. While the children were young she worked sporadically, and later she went to work full-time at a local manufacturing company as an administrator. She was promoted initially and had some success as a manager, but she was laid off when the business failed, and she declared personal bankruptcy for tax purposes. At this time she began to have problems in her marriage, and she divorced her husband soon afterward. She lived alone for a while before meeting and marrying her second husband, the manager of a small business. After losing her job, she worked occasionally as an administrative assistant, signing up with a temporary-employment agency.
The patient had received psychiatric treatment for depression on one occasion 5 years earlier, at which time she had been admitted to a psychiatric ward with significant suicidal ideation. Her drinking had been escalating for a number of years, and she had previous bouts of untreated depression. Her first marriage and her job were both in trouble at the time of her admission. She was treated initially as an inpatient; she was then treated as an outpatient with the antidepressant fluoxetine and experienced significant improvement in her mood. After about a year she felt better and stopped both psychiatric treatment and medication. She was drinking during this period but not to the same extent as previously or subsequently.
While there were no immediate family members with alcohol problems, she recalled that alcohol had always been around her parents’ house and freely available. However, she reported that her paternal uncle was a heavy drinker and that both her mother’s parents were alcohol dependent. Her first husband drank heavily, but neither of her children had alcohol-related problems. There was no history of depression in her family.
On admission to the clinic the patient was well dressed and well groomed. She had a direct and pleasant manner, and her speech and thought patterns were normal, with no evidence of severe depressive rumination, hallucinations, or delusions. Although she reported a mild depression and feeling sad at times, with crying episodes about once a week, she showed no evidence of psychomotor retardation. She felt somewhat helpless about her career but not about the future in general. She reported feeling guilty and disappointed in herself but had no suicidal ideation or intent. There was occasional situational anxiety but no panic attacks or specific phobias. While discussing her career prospects, she became tearful; at other times she displayed a full range of affect. She was fully oriented, exhibited good concentration and memory, and had good insight into her problems. Although she did not meet the criteria for current major depression, she did meet the criteria for past major depressive episode.
The patient signed a consent form for participation in a 9-month study and after 5 days of sobriety was initiated into the initial 10-week phase. A complete physical examination revealed mild hepatomegaly. She had a routine full blood screen, including liver function tests, which showed a raised g-glutamyl transpeptidase (GGT) level, 103 U/liter (normal range=0–45). Her score on the Beck Depression Inventory(9)was 16 (range=0–64), and her score on the Obsessive Compulsive Drinking Scale (10)was 9 (range=0–48). This initial phase of the study consisted of weekly relapse-prevention group psychotherapy and pharmacotherapy with naltrexone, 50 mg/day. She continued naltrexone therapy through the 9-month trial.
The first few weeks of treatment went well; the patient remained completely abstinent, and her depressive symptoms cleared up; her score on the Beck Depression Inventory was 4 at the end of week 2. At the same time, she reported reduction in cravings for alcohol and had a score of 5 on the Obsessive Compulsive Drinking Scale. At the end of her 3rd week of treatment and 4th week of abstinence, her Beck Depression Inventory score was 2 and her Obsessive Compulsive Drinking Scale score was 4. However, she reported a sudden onset of severe depressive symptoms at the end of her 4th week of treatment and a parallel increase in alcohol cravings; her Beck Depression Inventory score rose to 22, and the Obsessive Compulsive Drinking Scale score increased to 16. There was no precipitating event for this depression, but she remarked that it felt like her previous depressive episodes. Feelings of hopeless, lack of ambition, irritability, and detachment were reported without insomnia or anorexia, and she was able to keep up her daily level of activity. She remained at her job but had difficulty in concentrating at work. It is important to note that the patient did not decrease or change her smoking habit at this time or at any other time during the study. After a week of depression, she agreed to see the study psychiatrist for a medication review.
At the psychiatric interview 1 week later, the patient reported increasing depression and onset of insomnia, with early-morning waking and initial insomnia. Her appetite was now affected, and her energy was low. She reported diurnal variation in mood, with tearfulness in the mornings, and general hopelessness about the future but no suicidality. The patient agreed to take the antidepressant fluoxetine, 20 mg/day, and reported that she had responded quickly in the past. Indeed, 2 weeks later she said that her depression had largely lifted, as had her alcohol cravings. She was sleeping well, her appetite and concentration were fine, and she had no tears or diurnal variation in mood. She did not feel helpless and did not fear the future. Her Beck depression score at this time was 3, her Obsessive Compulsive Drinking Scale score was 4, and she remained abstinent and compliant.
Over the next 2 weeks her mood deteriorated somewhat, necessitating an increase in fluoxetine dose to 30 mg/day, to which the patient responded within 10 days. She continued on the regimen of fluoxetine and naltrexone, with less frequent psychotherapy visits, as per the study protocol. By week 23 of treatment she felt that she did not need the fluoxetine and stopped taking it by herself. The psychiatrist met with her at week 29, when she reported a significant increase in depression, with tearfulness, initial insomnia and early-morning waking, diminished energy, poor concentration, and feelings of hopelessness. Having switched jobs to work as a logistics planner in a manufacturing plant, she was enjoying the job until she got depressed for the second time, again with no precipitating event. Although she was irritable and tense, she reported no increase in cravings for alcohol and remained abstinent. After discussion with the patient, fluoxetine at the 20-mg/day dose was reinitiated. After taking the medication for 3 weeks, the patient experienced a significant improvement in depressive symptoms and full resolution of all problems concerning energy, concentration, insomnia, and hopelessness. After completing the full 9-month program with naltrexone for alcoholism, the patient asked to continue taking fluoxetine and was discharged to private psychiatric care. She expressed satisfaction with the treatment she had received during the study and was pleased that she had been abstinent for the entire 9-month period, for the first time in many years.
There is a complex and often puzzling interaction between depressive disorders and alcohol dependence. Intoxication with alcohol can cause significant but often temporary depressive symptoms, even in people with no history of affective disorder (11, 12). Estimates of the co-occurrence of alcohol dependence and depression in clinical samples of patients with alcoholism have ranged from 16% to 68% (12). Given the high comorbidity of depression and alcoholism, there has been considerable research on how these syndromes are similar or distinct neurobiologically, genetically, and clinically (reviewed by Markou et al. ). For example, a serotonin deficiency has been proposed in alcohol dependence (14) and likewise in depression (15), and indeed there is some evidence for serotonergic involvement in depression and alcoholism in animal genetic models (16).
In order to guide clinical practice, efforts have been made to distinguish subtypes of comorbid depression and alcoholism. The most prominent conceptualization to date has been the distinction between primary and secondary depression; the temporal onset of disorders and factors such as family history of affective disorder and clinical course of depressive symptoms may distinguish the two subgroups. Abstinent alcohol-dependent patients with primary depressive disorder may experience slower remission of depressive symptoms than do patients with secondary depression (17), and they may have stronger family histories of affective disorder (18). Recent data from the National Longitudinal Alcohol Epidemiologic Survey (19) indicate that patients with a strong family history of alcoholism may also be at a heightened risk of major depression, irrespective of whether the depression is primary, concurrent with the onset of dependence, or secondary.
The temporal pathway of affective symptoms in relation to abstinence from alcohol varies; in a large proportion of patients these symptoms are relieved within a few days or weeks of abstinence from alcohol (20–22). This patient had mild depressive symptoms in the context of heavy drinking, and she did not meet the criteria for major depression at intake. These symptoms cleared up with abstinence and basic alcohol-relapse-prevention psychotherapy that was not targeted at depression. During this time she was receiving naltrexone, which did not interfere with the commonly observed pattern of relief of depressive symptoms in the early abstinence phase of recovery.
After 4 weeks of abstinence the patient experienced an acute onset of severe depression. There may be a number of causes of this depression. When the depression is the primary diagnosis, patients often have more frequent and longer episodes of depression than when the depression is secondary to alcohol dependence (18, 23). In this patient’s case, the temporal onsets of the disorders favor a diagnosis of secondary depression, since the patient began to drink heavily in her 30s and had her first psychiatric treatment for depression after the appearance of suicidal ideation when she was in her late 40s. As for many patients, the actual age at onset of alcohol dependence was unclear, but this patient reported that she was alcohol dependent when she was admitted with her suicide attempt. She reported episodes of feeling down during this period but not to the extent that she sought help or the feelings interfered with her daily activity.
Her initial episode of depression during the study should probably be regarded as substance independent, as she had been sober for 4 weeks at this time. It might be regarded as substance dependent, however, if other factors came into play. She could have been misrepresenting her alcohol consumption, and the rise in alcohol craving at the same time as the occurrence of her depressive symptoms might have indicated a relapse to heavy drinking that she did not report. In the opinions of the treating psychotherapist and the psychiatrist this was not likely, because the patient did confess to alcohol consumption before this time and did not appear to be lying on this occasion. The values on her liver function tests did not rise in the course of the study, and the GGT level, often an index of heavy alcohol consumption, fell from 103 U/liter at inception to 46 U/liter at week 4 and to 32 U/liter at week 10 (normal range=0–45 U/liter).
Another possible explanation for her depression is that naltrexone may have triggered its onset in this vulnerable patient. There have been occasional reports of dysphoria associated with naltrexone in healthy control subjects (24) and former opiate abusers (25). Controlled trials of naltrexone in healthy subjects have produced varying results. In one study, single doses of naltrexone given to healthy control subjects produced adverse effects, including depression (26), while a placebo-controlled 8-week trial of naltrexone with 36 healthy control subjects produced no overall increase in dysphoria in the naltrexone group, although one subject had to leave the study because of severe dysphoria (27). Among alcohol-dependent patients, however, naltrexone has not been shown to increase the frequency of depression over the rate for placebo in double-blind placebo-controlled studies (6, 7, 28)or the rate for an untreated comparison group (29). In the present case we cannot conclude whether naltrexone contributed to the onset of depression given that we did not discontinue and rechallenge the patient with naltrexone. Regardless of etiology, however, standard antidepressant therapy ameliorated the patient’s symptoms while naltrexone therapy was continued. In addition, the patient had no increase in side effects while taking both an antidepressant and naltrexone, and this is consistent with the results of the naltrexone in an alcoholism safety trial (30), in which equal numbers of adverse events were reported for subjects taking naltrexone and those taking naltrexone and an antidepressant. Taken with these data, this case study suggests that antidepressant therapy should be considered for patients who develop a major depression while taking naltrexone and that there is no need to stop administering the naltrexone itself. A related finding comes from a recent study of 14 depressed patients who had been maintained on antidepressant therapy but who continued to drink heavily (31); the addition of naltrexone therapy resulted in a reduction in drinking, no increase in adverse events, and some small improvement in depression.
Given the reported association between naltrexone and dysphoria in healthy subjects and opiate addicts, an interesting question is whether our patient’s course would have been similar had she been treated with acamprosate. Acamprosate is a novel pharmacotherapy approved for treatment of alcoholism in Europe and under study in the United States that is hypothesized to work through alteration of neuronal excitability at the N-methyl-d-aspartate (NMDA) receptors and other nonopioid effects (reviewed by Wilde and Wagstaff ). The European studies of the benefits of acamprosate in the treatment of alcohol dependence did not show a significant change in patients’ mood (8), and a controlled study of acamprosate for healthy volunteers revealed no change in mood after a 7-day trial (33). To our knowledge, studies of acamprosate for depressed alcohol-dependent patients that might further guide clinical practice have not been undertaken at this time.
It is also important to note that our patient did not decrease or change her smoking habit during the study, as decreases in nicotine intake can have profound mood-changing effects and may even trigger a depressive episode (34). If recovering alcoholics also cease smoking, there does not appear to be an increased rate of alcoholic relapse (35). However, cigarette smokers with a history of depression who succeed in stopping smoking are vulnerable to a new depressive episode even months later (34).
With regard to psychosocial precipitants of the patient’s depression, her assessments of her situation and of the problems caused in her life by alcohol may have been contributing factors. She was particularly concerned about her career situation and her career prospects in middle age, and she felt her current job as a temporary administrative assistant to be beneath her, especially as she had been a successful manager for many years. Alcohol may have formerly helped her avoid facing some of the disappointments in her life, and the recovery from depression after initial sobriety may have been only a temporary respite in an ongoing depressive episode.
The temporal association between depressive symptoms and craving for alcohol is of interest. The patient had a substantial increase in her craving for alcohol, as measured by the Obsessive Compulsive Drinking Scale, temporally corresponding to the onset of her first episode of depression in this treatment episode. If indeed she remained sober at this time (as the GGT evidence indicates), it shows a remarkable correlation between alcohol craving and depression (36). She did not in fact relapse, despite the craving, perhaps because of the support and concern of her psychotherapist, use of cognitive methods to manage craving, and her own motivation to remain sober. In addition, there is evidence that a high level of craving for alcohol does not always predict relapse in alcoholism (37). It is interesting that by the time of her depressive relapse at week 29, after a much longer period of sobriety than preceded the first episode of depression in the study, there was no increase in cravings for alcohol. This indicates that this patient’s craving for alcohol was related to mood only in the context of recent alcohol consumption or recent abstinence. Thus, craving for alcohol proved to be independent of mood when the period of sobriety was long enough.
Clinical trials for alcohol-dependent patients with either primary depression (4) or secondary depression (5) have shown response to effective doses of antidepressant medication, mainly in depressive symptoms. The lack of distinction in terms of response to antidepressants that is suggested by this research could undermine the significance of the primary-secondary distinction and may argue for the importance of the substance-dependent and substance-independent categorization of mood disorder recently suggested by Schuckit et al. (3). In this patient’s case, there appears to have been a substance-independent onset of depression, despite its being a "secondary" depression. Nonetheless, irrespective of the primary-secondary dichotomy, a large number of alcohol-dependent patients with histories of depression report a substance-induced depressive episode (3). The prognostic significance of the distinction between substance-independent and substance-induced for clinical course and response to pharmacotherapy needs to be resolved through additional research.
This case presents some of the difficulties inherent in treating comorbid depressive and substance abuse disorders. The patient appears to have had mild substance-induced depressive symptoms at intake that required only a period of abstinence to clear up, and she also had two episodes of a substance-independent depression, both requiring treatment with an antidepressant medication. Craving for alcohol was induced in one of the "independent" episodes of depression and not in the other. Naltrexone was effective and was tolerated well during the course of treatment, and it did not mitigate the effectiveness of the antidepressant. In summary, this case illustrates the complex interaction, etiologically and temporally, of alcoholism and depression, and it demonstrates the value of actively and comprehensively treating each disorder.
Received Oct. 8, 1998; revision received Feb. 12, 1999; accepted Feb. 23, 1999. From the Substance Abuse Treatment Unit, Department of Psychiatry, Yale University School of Medicine, New Haven, Conn. Address reprint requests to Dr. Farren, Division of Addiction Psychiatry, Bronx VA Hospital, 130 West Kingsbridge Rd., Bronx, NY 10468; firstname.lastname@example.org (e-mail). Supported in part by grants AA-00171 and AA-09538 from the National Institute on Alcohol Abuse and Alcoholism and grant DA-00167 from the National Institute on Drug Abuse. Naltrexone was supplied by Dupont Pharma.