Research metodology

British Journal of Addiction (1988), 83, 1373-1380

Alcoholism, Anxiety and Depression

MARC A. SCHUCKIT, M.D.i & MARISTELA G. MONTEIRO, M.D., Ph.D.2

^Professor of Psychiatry, University of California, San Diego School of Medicine and Director, Alcohol Research Center, San Diego Veterans Administration Center (116A), 3350 La Jolla Village Drive, San Diego, CA 92161, U.S.A. and ^Associate Professor of Psychopharmacology, Escola Paulista de Medicina, Sao Paulo, Brazil

Summary This paper discusses diagnostic and treatment approaches for dealing with patients who present with two or more psychiatric disorders. We emphasize the relationship between depressive and anxiety syndromes on one hand, and alcohol abuse on the other. Some reasons for diagnostic confusion are noted, such as the need to distinguish between drinking and alcoholism, sadness and depression, and anxiety feelings and major anxiety disorders. Because symptoms of sadness or anxiety and excessive drinking frequently overlap, a working hierarchy must be established—the approach suggested here is the determination of the primary disorder based on the chronology of development of symptoms. The authors point out some clinical guidelines for use in evaluating patients with primary alcohol abuse and secondary anxiety or depression. Using the data from the literature as well as clinical experience, it is concluded that alcoholism, major affective disorder and major anxiety disorder are distinct illnesses with different prognoses and treatments. Some implications for clinical practice are discussed.

1 Introduction Patients, present to health care professionals re- questing help. Ideally, the clinician then evaluates the results of the examination and relevant labora- tory tests and selects a course of action likely to do the most good with the least potential for harm. The information gathered from the patient and the resource person determines the decision of whether to intervene and, if so, how to choose the optimal treatment.''^

In the mental health and substance abuse fields, this decision tree rests primarily with the mental status evaluation and the accurate determination of the clinical history. Lacking sensitive laboratory tests, the mental health professional gathers infor-

Reprint requests should be addressed to Marc A. Schuckit at the VA Medical Center.

mation from the patient and an appropriate resource person in an effort to determine a diagnostic label that can help estimate the problems likely to appear in the future and the most probable response to various interventions. To be useful, such labels must be based upon carefully worded diagnostic schemes that have been applied to large numbers of individuals who are then followed over time to demonstrate that their disorder was not just a variant or a prodromal phase of another problem, and that there is a somewhat predictable course.-*

Even when careful and well studied labels are employed, however, we still face a problem when a patient fulfills criteria for two or more disorders. In this instance, some working hierarchy must be established in order to guide the clinician on the most appropriate intervention technique. One ap- proach for dealing with a patient with multiple

1373

1374 Marc A. Schuckit & Maristela G. Monteiro

labels is to distinguish between primary and secon- dary illness. While these words have been applied to several different schemes, a paradigm especially relevant to the alcohol and substance abuse field is the determination of a primary disorder based on the chronology of development of symptoms.

Using an analogy from medicine, when a patient presents with pneumonia that developed after the onset of an immune deficiency syndrome, it is likely that the overall clinical course will be predicted by the first-appearing or primary disorder, although immediate clinical interventions will have to con- sider both the primary illness and secondary pneu- monia. On the other hand, to recognize only pneumonia and ignore the primary disorder is likely to lead to an erroneous estimate of the prognosis and inappropriate therapeutic interventions.

The primary/secondary dichotomy is especially relevant to the substance abuse and alcoholism fields.'-''"* All of these substances of abuse have direct actions on the central nervous system (CNS) and each is capable of producing symptoms similar to those observed in major psychiatric disorders.'-'" The two groups of drugs most relevant to mental disorders are the CNS depressants (ethanol, all prescription sleeping pills, and most prescription antianxiety drugs) and the CNS stimulants (e.g. amphetamines, cocaine, and many weight-reducing products). For example, in the context of primary stimulant abuse people are likely to develop drama- tic secondary psychoses characterized by paranoid delusions and auditory hallucinations occurring without insight. During these episodes, the psycho- tic symptoms resemble schizophrenia. However, the patient with a primary stimulant abuse and secon- dary psychosis is likely to return to a state of proper reality orientation within days to weeks of absti- nence from the drug, while the primary schizo- phrenic will probably remain severely psychiatri- cally impaired for months or years despite social and pharmacological interventions. The distinction be- tween primary and secondary disorders in this situation is based on the chronology of development of the syndromes, with the primary problem being the one that appeared first.

This paper focuses on the application of the primary/secondary scheme to periods of depression or anxiety observed in the context of substance misuse. Reflecting the rather extensive literature on the topic, we will emphasize the data on the relationship between depressive and anxiety syndromes on the one hand and alcoholism on the other.

2 Sadness, Affective Disorders, Drinking and Alcoholism 2.1 Reasons for Confusion There are a variety of reasons why one might expect a complex interaction between drinking, alcohol- related life problems, and various forms of sadness. First, as a typical CNS depressant, alcohol is capable of inducing feelings of sadness. This is especially likely to be observed at falling blood alcohol concentrations (BACs) or during periods of sus- tained intoxication. In fact, as few as two or three drinks are likely to have an effect on mood with feelings of sadness or irritability that at times can carry over to the next day.""'"* While the degree of impairment is different for different people, it is likely to be longer lasting and more profound following higher BACs and longer periods of drinking. Several different laboratories have docu- mented apparent intense levels of sadness, coming close to or actually fulfilling diagnostic criteria for depressive disorders, during protracted periods of drinking carried out in an alcohol research set- ting."-"

A second reason for the interaction between sadness and heavy drinking relates to the pattern of life problems likely to be observed in the course of alcoholism. The natural history of alcohol depen- dence and abuse includes the probability of alterna- tions between periods of abstinence, relatively brief periods of controlled drinking, and times when both intake and alcohol related difficulties rapidly esca- late.^-'*'" It makes sense that alcoholics are more likely to come in for treatment at a time when their ability to cope is compromised. Therefore, not only can alcohol itself be expected to induce feelings of sadness which at times can be intense, but life problems inherent in the course of alcoholism are also expected to induce affective disturbances. Alcoholics who appear in our offices seeking help are, thus, likely to also be in relatively severe difficulties and could be expected to demonstrate a mood consistent with their situation.

A third, but related, phenomenon is the probabil- ity that the coexistence of two or more symptom patterns is more likely to bring an individual into treatment than would be true if only one problem were apparent." Thus, surveys of individuals who have entered treatment have documented that alcoholics under care are more likely to demonstrate multiple symptom patterns than alcoholics identi- fied through community surveys."-^*

Fourth, it is held as 'common knowledge' that people might be expected to increase their drinking

Alcohol, Anxiety and Depression 1375

in the context of an independent depressive episode. Thus, many clinicians expect that patients with primary major depressive disorders, for example, might attempt to 'self-medicate' by increasing their alcohol intake. However, despite this belief, there is little evidence to support this 'self-medication' hypothesis during depression. While an increase in all behaviors is observed during manic episodes,^''^^ 70%-80% of patients with severe depression do not escalate their alcohol intake.^^'^'' Most patients with major depressive disorders either decrease their ethanol consumption or maintain a similar pattern of drinking, perhaps because they recognize that their symptoms intensify with alcohol. When criteria for actual life impairment related to alcohol are invoked, only 5%-10% of patients with primary major depressive disorders actually developed sec- ondary

2.2 Rates of Depressive Symptoms Among Alcoholics In the final analysis, there are multiple reasons why heavy drinking and sadness often go together. The cross-sectional symptoms alone, however, give us little useful information about the probable short- and long-term prognoses, nor are they likely to give us enough information to make most relevant treatment decisions. Overall, the rate of severe depression in the course of alcohol abuse or dependence is at least 30%-40%, and when less stringent criteria are applied the figure may be 70% or higher.^'-'" As is true in the population in general, depressive symptoms are more likely to be reported by female alcoholics than male alcoholics."-'^ Re- gardless of sex, the quality of depressions seen in the course of heavy drinking can be similar to the symptom profile noted during primary affective episodes.'-'-'' Although periods of sadness are common in alcoholics, they usually do not last the required 2-4 weeks of intense depression on a daily basis to meet research diagnostic criteria for actual affective disorders.^'-'^-" Even using this restric- tion, however, in analysing 577 consecutive male alcoholics who were evaluated with face-to-face structured interviews with themselves and two relatives for each, our laboratory found that more than one-quarter of the carefully defined primary alcoholics had at least one secondary affective episode that did last the required period of time at sometime during their heavy drinking years.^'

Taking this reasoning a step further, while temporary sadness and even intense and persistent depressions appear to be relatively common in the

context of heavy drinking and associated with- drawal, there is little evidence demonstrating that alcoholics are more likely than the general popula- tion to demonstrate severe depressive episodes either before the onset of heavy drinking or in the context of periods of abstinence lasting 3 or more months. The study of 577 alcoholics carried out by our research group was able to document a rate of independent major depressive disorder or bipolar manic depressive disease in only 5% or less of our carefully evaluated subjects.^' Similarly, another evaluation of 565 patients carried out at five Veterans Administration Medical Centers reported rates of independent depressive episodes in the same range.'*

Clinical observations of alcoholics with secondary affective disorders occurring only in the context of heavy drinking or withdrawal indicate that the severe depressions occurring in this context rarely run the course of primary affective episodes. Although the intensity of symptoms can be severe, the affective disturbances are likely to be transitory, showing great improvement or disappearance of symptoms within several days or weeks of absti- nence.^''"-''*-" Also, several longer term follow-ups of alcoholics who were able to maintain abstinence showed little evidence of severe or prolonged depression during the follow-up periods."-" Finally, comparisons of alcoholics with and without secondary affective disorders revealed that the former more closely resemble alcoholics than pa- tients with primary affective disorder. This is true for sociodemographic characteristics,'*-'*''"-" for family histories of alcoholism,"-^' and for early life

social problems as well as for the adult course of alcoholism.'

2.3 Some Clinical Suggestions As has been described in other publications, it is possible to take this information and develop some clinical guidelines for dealing with patients who present with severe depressive symptoms and recent histories of heavy alcohol intake. These clinical conclusions are based on the information outlined above, clinical experience, and common sense. The guidelines might be viewed as a series of stages.

First, any patient who is so depressed that he is thinking of suicide, has a plan for killing himself, and the means to carry oUt such a plan should be considered as an emergency situation, no matter what the primary diagnosis. Therefore, in an office, clinic, or emergency room setting the primary/

1376 Marc A. Schuckit & Maristela G. Monteiro

secondary distinction is not an essential part of the decision regarding hospitalization of the suicidal individual.

Second, the primary diagnosis does have an impact on the initial treatment plan. It is often useful at the time of hospitalization to be able to project whether a locked facility with suicidal precautions is likely to be required for only several days to a week or so (as in the case of primary alcoholism with secondary depressive episodes), or if the depressive symptoms are likely to persist for many weeks (as would be expected with primary affective disorder and secondary alcoholism). In addition, while it is important to observe a patient for a period of days to weeks before beginning antidepressant medications regardless of the pri- mary diagnosis, it is often useful to be able to predict whether after several weeks antidepressants will be required.

Third, the final treatment offered rests with both the establishment of a primary diagnosis and observation of the clinical course while the patient is under care. The primary diagnosis requires that a careful history of the chronology of the develop- ment of symptoms be taken from both the patient and a resource person who has observed the patient over a period of time (e.g. often a relative or close friend). The approximate age of onset of alcoholism is established by reviewing the major life problems that have occurred in the course of heavy drinking and establishing the age of first major life problem. This would include the first time that alcohol had actually interfered in a major way with a relation- ship, the age of a second arrest related to alcohol, the first time and individual received treatment for alcoholism or was told by a physician that alcohol had harmed health, the first time alcoholic with- drawal had been observed, and so on. After deter- mining the approximate age of onset of alcoholism, it is then important to establish any periods of abstinence of perhaps 3 or more months that have occurred since the onset of severe problematic drinking. Next, episodes of depression lasting all day, every day, for a period of 2 or more weeks and interfering with functioning should be deter- mined—not just feelings of sadness or even intense sadness, but severe enough depressive episodes to meet criteria for major depressive disorder. When depressive episodes have developed independent of heavy drinking (e.g. before the onset of alcoholism or during periods of extended abstinence) it makes sense that the patient might have an independent major depressive disorder. In this instance, one

might expect that the depression would be less likely to disappear with abstinence, suicidal ideas might be expected to last for a longer period of time, and even after several weeks of abstinence, antidepressant medications might be required.

Finally, it is important to not forget the impor- tance of observing a patient's condition over time. For example, if an individual appears from the history to have primary alcoholism with secondary depressive disorder but after 2-4 weeks of absti- nence he is still so depressed on a daily basis that he cannot function, the possibility of an independent depressive disorder should be considered. If the depression is intense enough, treatment with antide- pressant medications might be required. To place these comments into perspective, however, when a patient meets criteria for alcoholism and also demonstrates severe depression, for 95% or so of men and probably 85% or so of such women the depressive symptoms are likely to be secondary to the heavy alcohol intake and are likely to improve markedly within several weeks of absti-

3 Drinking, Anxiety and Anxiety Disorders Similar lines of reasoning can be applied to the relationship between drinking, alcohol-related life problems, and anxiety. To avoid excessive redun- dancy, these issues are only briefly reviewed below.

3.1 Reasons for Confusion There are a number of reasons to expect a possible relationship between anxiety and drinking related problems. First, the 'tension reduction hypothesis' is based at least in part on upon observations of moderate drinkers for whom experiences of feelings of relaxation are likely to be observed after imbibing one or two drinks at a party or after a difficult day. This clinical observation is consistent with the theory that if light drinking might decrease mild tension, then alcoholics might consume large amounts of alcohol to cope with severe anxiety problems.''^"'''' However, most studies indicate that even modest doses of alcohol are associated with increases, not decreases, in physiological measures associated with tension as well as with subjective feelings of anxiety.''^"'" Therefore, there is little objective evidence to corroborate the theory that a substantial proportion of alcoholics might develop their disorder in an attempt to seek out possible

Alcohol, Anxiety and Depression 1377

tension reduction properties of a brain depressant such as alcohol.

A second and more impressive correlation be- tween alcoholism and severe anxiety syndromes comes from the observation of anxiety in the context of alcoholic withdrawal. After the development of physical dependence to any brain depressant, the acute abstinence syndrome, lasting for 4-5 days or more, is likely to include tremors, feelings of tension, restlessness, and insomnia.'" This is often followed by a secondary or protracted abstinence syndrome lasting for months and characterized by atixiety, emotional instability, autonomic overactiv- ity, restlessness, and sleep impairment.'"-'""" While numerous studies have documented high levels of atixiety syndromes among alcoholics, almost all of these involve symptom profiles observed during the first several weeks of abstinence.*'-'^"'' Thus, while anxiety symptoms and even severe syndromes are frequently observed during acute and protracted abstinence from alcohol, there is additional data that these symptoms are likely to decrease with absti- nence, indicating that they may not represent an independent anxiety disorder.

Third, in a manner similar to that described for depressive episodes, it is likely that alcoholics demonstrating secondary anxiety syndromes would be more likely to enter care than those without such concomitant anxiety associated with withdrawal."-^" This in turn could indicate that estimates of the prevalence of anxiety syndromes in the context of withdrawal may be artificially inflated by the focus of most studies upon alcoholics actually in treatment.

A fourth area of confusion emanates from studies documenting drinking practices among patients with bonafide primary anxiety disorders. Thus, Marks and colleagues" documented a 10% rate of substance abuse in 38 phobic and anxious patients followed for 5 years, Quitkin et al.^'' reported a similar 10% rate in phobics, and Bibb & Chambliss'*^ diagnosed DSM III alcoholism in 10%-20% of outpatients with agoraphobia. Cloninger et a/.'* have also noted that among 32 patients with probable primary panic disorder, 9% (three indivi- duals) fulfilled criteria for chronic alcoholism and an additional 6% had alcoholism in the past which appeared to be in remission. On the other hand, these results are not very impressive when one considers the probable rate of alcoholism in the general population. Reflecting the high rate of drinking in general,"' perhaps 30%-40% of men develop temporary alcohol-related difficulties in- volving an alcoholic blackout, arguments with

friends, missing some time from school or work, or a single drunk driving arrest.'"-"-'" Even when more rigorous criteria are used and actual alcoholism is diagnosed, the lifetime risk for this disorder is at least 10% among men and 3%-5% among women,'" with even higher rates (perhaps as high as 20% or more) reported from the recent Epidemiologic Catchment Area (ECA) national and international surveys."-'^ As a consequence of this high general population prevalence, it should also be expected that a high proportion of men and women with true anxiety disorders drink and that some will develop alcohol related problems of a temporary nature. However, there is little impressive evidence of an actual higher rate of alcoholism among these pa- tients.

In summary, the literature does not clearly document a close association between primary major anxiety disorders and alcoholism. Many patients with anxiety disorders drink, but it is not clear whether they are more likely to develop alcoholism than the general population; most alcoholics experi- ence severe anxiety, but this is usually in the context of acute or protracted abstinence syndromes and it is not clear whether these men and women are more likely to have major anxiety syndromes either before the onset of their alcoholism or during periods of protracted abstinence; and the tension reduction hypothesis of heavy drinking has little consistent data to support its validity. While it is still possible that there is a clinically relevant relationship between usual course of alcoholism and independent symptoms of anxiety, the present data can be used to develop some clinical guidelines.

3.2 Some Clinical Suggestions As clinicians we should expect our alcoholic patients to present with symptoms of anxiety. When these are observed in the context of recent abstinence from drinking, these might be intense symptons including panic attacks, phobias, or generalized anxiety. It is probably wise not to assume that this symptom pattern indicates the same long-term prognosis and treatment needs likely to be observed for primary panic disorder, agoraphobia, and so on. It is more likely that with education and reassurance the symptoms will decrease fairly rapidly over the first month, after which continued slow improve- ment is to be expected.

In a manner similar to that described for the relationship between alcoholism and depressive disorders, it is important to gather the best informa-

1378 Marc A. Schuckit & Maristela G. Monteiro

tion possible from both the patient and a resource person. Using a similar timeline type of approach, the absence of major anxiety syndromes antedating the onset of alcoholism or during periods of protracted abstinence is likely to indicate a high probability of spontaneous improvement in anxiety. However, it is also important to observe the clinical condition over time and to consider the institution of more aggressive behavior modification ap- proaches or even the appropriate limited use of medications if the anxiety increases rather than decreases or if no improvement is observed after a period of perhaps 4-8 weeks of abstinence.'"

4 Some General Conclusions This paper has presented one type of approach to establish the probable clinical course and treatment needs for a group of patients with complex clinical pictures. The information offered above is a blend of clinical data and clinical experience to arrive at a logical course of action. The basic premise is that all interventions can be harmful, and the best way to decide whether to treat and what intervention to use can be based on diagnosis.

This approach to diagnosis becomes problematic when a patient presents with two or more syn- dromes. This is especially likely to be a problem with substances of abuse because intoxication and withdrawal from these agents can mimic many psychiatric disorders. It is also probable that the abuse of any of these drugs is likely to increase the intensity of symptoms of pre-existing primary psychiatric disorders.

The clinical scheme described above requires that a history be taken from both the patient and a resource person, documenting the approximate age of occurrence of major syndromes. It is important to establish the age of probable onset of alcoholism or drug abuse as well as periods of protracted absti- nence. When major psychiatric syndromes have only been documented in the context of drinking or in the first several months following withdrawal, it can be expected that the psychiatric symptoms will mark- edly improve spontaneously within several weeks of abstinence. Both common sense and clinical experi- ence indicate that when major depressions or anxiety syndromes have been observed in the absence of substance misuse, the clinical psychiatric syndrome is less likely to disappear with time alone. To be maximally effective and relevant to our patient's needs, it is also important to observe their clinical course over the first month or so of

abstinence, keeping an open mind to the possibility that two independent disorders are present.

Much less data is available on patients with severe primary psychiatric disorders and secondary abuse of substances. For example, the patient with an onset in the late teens of social withdrawal and a bland affect accompanied by hallucinations and delusions (e.g. schizophrenia) who then begins to use and abuse brain depressants and stimulants is likely to demonstrate a severe intensification of the psychotic symptoms.^-*' Most clinicians would re- cognize the need for a focus on the major interven- tion techniques through psychiatry—attempting to stabilize the psychiatric picture in the context of abstinence. However, decreasing the probability that the patient will return to misuse of substances with a subsequent 'revolving door syndrome' of the repeated need for inpatient care, is no simple matter. While the focus of this paper has been on primary substance abuse with secondary anxiety or depression, the clinical needs of patients with primary manic depressive disease or primary schizo- phrenia, for example, along with secondary alcohol- ism or other substances of abuse are also clinically important. In this situation, it is probably best to offer maximal care for the major psychiatric dis- order along with consultation or referral to educa- tion/counseling programs aimed at helping the patient to understand the need to abstain from drugs and aiding in his or her quest to maintain a substance free lifestyle.

As clinicians we try to do the best we can with the tools available to us. Unfortunately, psychiatry, psychology, and behavioral medicine must depend upon an imperfect system. We can only utilize the clinical history and course of symptoms during treatment as our major guidelines, because few, if any, reliable biological tests are available to us. The algorithm for approaching patients with dual diagnoses outlined in this clinically oriented manu- script must also be seen as an imperfect approach that takes advantage of available data. It is hoped that in the near future greater levels of understand- ing of psychological, sociological, and biological aspects of psychiatric syndromes and abuse of substances will allow us to develop a clinical approach to these complex patients that is both easier to implement and more reliable. Until that time, as both clinicians and researchers we find a great deal of clinical usefulness to the primary/ secondary dichotomy in substance abusing patients who also fulfill criteria for major psychiatric dis- orders.

Alcohol, Anxiety and Depression 1379

Acknowledgements Parts of this paper were adapted from: Schuckit, M. A. (1986) Genetic and clinical implications of alcoholism and affective disoidei, American Journal of Psychiatry, 143, pp. 140-147; and Schuckit, M. A., Irwin, M. & Brown, S. (forthcoming) The history of anxiety symptoms among 171 primary alcoholics, Jowma/ of Studies on Alcohol. This work was supported by the Veterans Administration Research Service, NIAAA Grant No. 05526, FA- PESP Fellowship from Sao Paulo Brazil and the Alcoholic Beverage Medical Research Foundation.

References 1. GOODWIN, D . W . A. & GUZE, S. B . (1984) Psychiatric

Diagnosis, 3rd edn (New York, Oxford University Press).

2. SCHUCKrr, M. A. (1984) Drug and Alcohol Abuse: clinical guide to diagnosis and treatment (New York, Plenum Press).

3. GUZE, S. B . (1970) The need for toughmindedness in psychiatric thinking. Southern Medical Journal, 63, pp. 662-667.

4. WINOKUR, G . (1972) Family histories studies, VIII; secondary is alive and well, a n d . . . , Diseases of Nervous System, 33, pp. 94-99.

5. GUZE, S. B., WOODRUFF, R . A. & CLAYTON, P. J. (1974) Secondary affective disorder: a study of 95 cases. Psychology Medicine, 1, pp. 426-428.

6. SCHUCKIT, M . A. (1979) Alcoholism and affective disorder: diagnostic confusion, in: D. W. GOODWIN & C. K. ERICKSON (Eds) Alcoholism and Affective Disorders pp. 99-108 (New York, SP Medical & Scientific Books).

7. MARTIN, R . L . , CLONINGER, C . R . & GUZE, S. B . (1985) Alcohol misuse and depression in women criminals. Journal of Studies on Alcohol, 46, pp. 65-71.

8. MERIKANGAS, K . R . , LECKMAN, J. R., PRUSOFF, B . A., PAULS, D . L . & WEISSMAN, M . M (1985) Pamilial transmission of depression and alcoholism. Archives of General Psychiatry, 42, pp. 367-372.

9. SCHUCKIT, M . A. (1983) Alcoholism and other psychiatric disorders. Hospital Community Psychia- try, 34, pp. 1022-1027.

10. BALDESSARINI, R . J. (1980) Drugs and the treatment of Psychiatric Disorders in: A. G. GOODMAN, L . S. GOODMAN & A. GILMAN (Eds) The Pharmacological Basis of Therapeutics pp. 391-447 (New York, MacMillan).

11. GIBSON, S. & BECKER,. J. (1973) Changes in alcohol- ics' self-reported depression. Quarterly Journal of Studies on Alcohol, 34, pp. 829-836.

12. MAYFIBLD, D . G . (1968) Psychopharmacology of alcohol: affective change with intoxication, drinking behavior, and affective state, J'oMma/ of Nervous and Mental Diseases, 146, pp. 314-321.

13. TAMERIN, J . S., WEINER, S. & MENDELSON, J. H. (1970) Alcoholics' expectancies and recall of experi- ences during intoxication, American Journal of Psy- chiatry, 126, pp. 1697-1704.

14. BiRNBAUM, I., TAYLOR, T . & PARKER, E . (1983) Alcohol and sober mood state in female social drinkers. Alcoholism: Clinical and Experimental Re- search, 1, pp. 362-368.

15. VAILLANT, G . E . (1982) Natural history of male alcoholism. Archives of General Psychiatry, 39, pp. 127-133.

16. VAILLANT, G . E . (1983) The Natural History of Alcoholism (Cambridge, Ma, Harvard University Press).

17. LuDWiG, A. M. (1985) Cognitive processes associ- ated with 'spontaneous' recovery from alcoholism. Journal of Studies on Alcohol, 46, pp. 53-58.

18. LUDWIG, A. M . (1972) On and off the wagon. Reasons for drinking and abstaining by alcoholics. Quarterly Journal of Studies on Alcohol, 33, pp. 91-96.

19. WEISSMAN, M . M . (1984) Alcoholism and depression: separate entities? Presented at the Cambridge Hospi- tal 7th Annual Alcoholism Symposium, 3 March, Boston, MA [unpublished].

20. BERKSON, J . (1946) Limitations of the application of the 4-fold table analyses to hospital data. Biometrics, 2, pp. 47-53.

21. LiSKOW, B., MAYFIELD, D . 8E THIELE, J. (1982) Alcohol and affective disorder: assessment and treat- ment. Journal of Clinical Psychiatry, 43, pp. 144-147.

22. MORRISON, J. R. (1974) Bipolar affective disorder and aicohoViiia, American Journal of Psychiatry, 131, pp. 1130-1133.

23. SCHUCKIT, M . A. (1986) Genetic and clinical impli- cations of alcoholism and affective disorder, American Journal of Psychiatry, 143, pp. 140-147.

24. MAYFIELD, D . G . (1979) Alcohol and affect: experi- mental studies, in: D. W. GOODWIN & C. K. ERICKSON (Eds) Alcoholism and Affective Disorders pp. 9-20 (New York, SP Medical & Scientific Books).

25. LEWIS, C . E . , HELZER, J., CLONINGER, C . R . , CROUGHAN, J . & WHITMAN, B . Y . (1982) Psychiatric diagnostic predispositions to alcoholism. Comprehen- sive Psychiatry, 23, pp. 451-461.

26. KELLER, M . B., LAVORI, P. E., LEWIS, C . E . & KLERMAN, G . L . (1983) Predictors of relapse in major depressive disorder. Journal of the American Medical Association, 250, pp. 3299-3304.

27. DuNNER, D. L., HENSEL, B . M . & FIEVE, R . R . (1979) Bipolar illness: factors in drinking behavior, .<4mencan Journal of Psychiatry, 136, pp. 583-585.

28. ROBINS, E . , GENTRY, K . A., MLTNOZ, R . A. & MARTEN, S. (1977) A contrast of the three more common illnesses with the ten less common in a study and 18-month follow-up of 14 psychiatric emergency room patients, II: characteristics of patients with the three more common illnesses. Archives of General Psychiatry, 34, pp. 269-281.

29. SCHUCKIT, M . A. (1983) Alcoholic patients with secondary depression, American Journal of Psychia- try, 140, pp. 711-714.

30. CADORET, R . , & WINOKUR, G . (1974) Depression in alcoholism,/Inna/s of New York Academy of Sciences, 233, pp. 34-39.

31. ScHUCKrr, M. A. & WINOKUR, G . (1972) A short term follow-up of women alcoholics. Diseases of the Nervous System, 33, pp. 672-678.

1380 Marc A. Schuckit & Maristela G. Monteiro

32. HATStnovMl, D. & PiCKEUS, R. W. (1982) Posttreat- ment depression in an alcohol and drug abuse population, American Journal of Psychiatry, 139, pp. 1563-1566.

33. WEISSMAN, M . M . , POTTENGER, M . , KLEBER, H . , RUBEN, H . L . , WILLIAMS, D . & THOMPSON, W . D . (1977) Symptom patterns in primary and secondary depression. Archives of General Psychiatry, 34, pp. 854-862.

34. SCHUCKIT, M . A. (1985) The clinical implications of primary diagnosis groups among alcoholics. Archives of General Psychiatry, 42, pp. 1043-1049.

35. ScHUCWT, M. A., ZisoOK, S. & MORTOLA,J. (1985) Clinical implications of DSM III diagnoses for alcohol abuse and alcohol dependence, American Journal of Psychiatry, 142, pp. 1403-1408.

36. POWELL, B . J., READ, M . R . , PENICK, E . C , MILLER, N. S. & BiNGHAM, S. F. (1987) Primary and secondary depression in alcoholic men: an important distinction? Journal of Clinical Psychiatry, 46, pp. 98-101.

37. PETTINATI, H . M . , SUGERMAN, A. A. & MAURER, H . S. (1982) Four year MMPI changes in abstinent and drinking alcoholics. Alcoholism: Clinical and Experi- mental Research, 6, pp. 487-494.

38. WooDRtn'F, R. J. JR., MURPHY, G . E . & HERJANIC, M . (1967) The natural history of affective disorders: I: symptoms of 72 patients at the time of index hospital admission Journal of Psychiatric Research, 5, pp. 255-263.

39. WOODRUFF, R. A., GUZE, S. B., CLAYTON, P. J., & CARR, D . (1979) Alcoholism and depression, in: D. W. GOODWIN & C. K. ERICKSON (Eds) Alcoholism and Affective Disorders pp. 39-48 (New York, SP Medical & Scientific Books).

40. WOODRUFF, R . A., GUZE, S. B . , CLAYTON, P. J. & CARR, D . (1973) Alcoholism and depression, y^rcAioej of General Psychiatry, 28, pp. 97-100.

41 SCHUCKIT, M . A., IRWIN, M . & BROWN, S. (forth- coming) The history of anxiety among 171 primary alcoholics, J'ottma/ of Studies on Alcohol.

42. BIBB, J. L. & CHAMBLISS, D . L . (1986) Alcohol use and abuse among diagnosed agoraphobics. Behaviour Research and Therapy, 24, pp. 49-58.

43. ScHUCKrr, M. A. (1985) Etiologic theories on alcoholism. Chapter 2, in: N. EsTES & E. HEINEMAN (Eds) Alcoholism: Development, Consequences, and Interventions, 3rd edn, pp. 15-31 (St. Louis, Mo, C.V. Mosby).

44. BLANE, H . & LEONARD, K . (1987) Psychological Theories of Drinking and Alcoholism (New York, Guilford Press).

45. WILLIAMS, A. F. (1966) Social drinking, anxiety and depression. Journal of Personal and Social Psycho- logy, 3, pp. 689-693.

46. LoGtre, P. B., GENTRY, W . D . , LINNOILA, M . & ERWIN, C . W . (1978) Effect of alcohol consumption on state anxiety changes in male and female non- alcoholics, American Journal of Psychiatry, 135, pp. 1079-1081.

47. SCHUCKIT, M . A. (1985) Overview: epidemiology of alcoholism. Chapter 1, in: M. A. SCHUCKIT (Ed.) Alcohol Patterns and Problems, pp. 1-42 (New Brunswick, NJ, Rutgers University Press).

48. ScHUCKrr, M. A. (1987) Guidelines for the treatment of alcoholic withdrawal. Fair Oaks Hospital Psychia- try Letter, 5, pp. 13-20.

49. ALLING, C , BALLDIN, J., BOKSTROM, K . , GOTTFRIES,

C. G., KARLSSON, 1. & LANGSTROM, G . (1982) Studies on duration of a late recovery period after chronic abuse of ethanol, Acta Psychiatrica Scandi- navica, 66, pp. 384-397.

50. D E SOTO, C . B . , O'DONNELL, W . E . , ALLRED, L . J. & LOPES, C . E . (1985) Symptomatology in alcoholics at various stages of abstinence. Alcoholism: Clinical and Experimental Research, 9, pp. 505-512.

51. KISSIN, B. (1979) Biological investigations in alcohol research, in: M. KELLER (Ed.) Research Priorities on Alcohol, (Suppl), 8, pp. 146-181.

52. MuLLANEY, J. A. & TRIFPETT, C . J. (1979) Alcohol dependence and phobias: clinical description and relevance, British Journal of Psychiatry, 135, pp. 565-573.

53. SMAIL, P . , STOCKWELL, T . , CANTER, S. & HODGON, R. (1984) Alcohol dependence and phobic anxiety states. I. A prevalence study, British Journal of Psychiatry, 144, pp. 53-57.

54. BOWEN, R. C , CiPYWNYK, D . , D ' A R C Y , C . & K E E G A N , D. (1984) Alcoholism, anxiety disorders, and agora- phobia. Alcoholism: Clinical Experimental Research, 8, pp. 48-50.

55. HALIKAS, J . A . , HERZOG, M . A., MIRASSON, M . M . & LYTTLE, M . D . (1981) Psychiatric diagnosis among female alcoholics. Currents in Alcoholism, 8, pp. 283-291.

56. MARKS, L M . , BIRLEY, J. R. T. & GELDER, M . G . (1966) Modified leucotomy in severe agoraphobia: a controlled serial inquiry, British Journal of Psychia- try, 112, pp. 757-769.

57. QtnTKiN, F. M., RIFKIN, A., KAPLAN, J., KLEIN, D . F . & OAKS, G . (1972) Phobic anxiety syndrome compli- cated by drug dependence and addiction. Archives of General Psychiatry, 27, pp. 159-162.

58. CLONINGER, C . R . , MARTIN, R . L . , CLAYTON, P. & GUZE, S. B . (1981) A blind follow-up and family study of anxiety neurosis: preliminary analysis of the St Louis 500, in: D. F. KLEIN & J. RABKIN (Eds) Anxiety: New Research and Changing Concepts, pp. 137-148 (New York, Raven Press).

59. CAHALAN, D . & ROOM, R . (1974) Problem Drinking Among American Men (New Brunswick, NJ, Rutgers Center of Alcohol Studies).

60. FiLLMORE, K. M. (1974) Drinking and problem drinking in early adulthood and middle age. Quarterly Journal of Studies on Alcohol, 35, pp. 819-840.

61. WEISSMAN, M . M . , MYERS, J. K. & HARDING, P. S. (1980) Prevalence and psychiatric heterogeneity of alcoholism in a tJnited States urban community. Journal of Studies on Alcohol, 41, pp. 672-681.

62. ROBINS, L . N . , HELZER, J. E. & G U Z E , S. B . (1984)

Lifetime prevjilence of specific psychiatric disorders in three sites Archives of General Psychiatry, 41, pp. 949-948.

63. SEGAL, D . S. & ScHUCKrr, M. A. (1983) Animal models of stimulant-induced psychosis, in: I. CREESE (Ed.) Stimulants: Neurochemical, Behavioral and Clinical Perspectives, pp. 131-168 (New York, Raven Press).