Skip to main content

Visions Journal

A reminder that this article from our magazine Visions was published more than 1 year ago. It is here for reference only. Some information in it may no longer be current. It also represents the point of the view of the author only. See the author box at the bottom of the article for more about the contributor.

Exploring the relationship between alcohol use, depression, anxiety and gender

Sarah, Hamid-Balma

Reprinted from "Men's" issue of Visions Journal, 2005, 2 (5), p. 11-12

Self-medication: the theory

One of the theories I always expect to hear in discussions of how non help-seeking men cope with depression or anxiety is “they self-medicate with alcohol.” The self-medication theory—that people use substances (alcohol, tobacco, other drugs) to relieve some of the distress associated with symptoms of an underlying mental illness—is a popular one among media, academics, service providers, mental health consumers and families. I wanted to look at what some of the research has to say about this theory in the particular area of alcohol use, gender, depression and anxiety.

Problematic use of alcohol occurs frequently among people with mood and/or anxiety disorders— that is certainly not in dispute. In a Canadian survey, 15% of men meeting the criteria for ‘alcohol abuse’ or ‘dependence’ had a coexisting mood disorder and 32% an anxiety disorder; for women, the numbers rose to 40% and 54%, respectively.

This strong relationship (what researchers call a correlation) has been found in both clinical (treatment) and community (non-treatment) settings. The self-medication theory attempts to explain the relationship by saying that the depression or anxiety came first and the alcohol use was a direct coping response to the symptoms of the depression or anxiety.

But self-medication theory does not explain all cases, nor does it provide a water-tight explanation, because of the complex interactions involved. In addition to direct self-medication with the aim of alleviating mood, there are at least five other possible ‘pathways’:

  • For some people, mood and anxiety disorders can be consequences of problem drinking, dependence and/or withdrawal

  • Even if mood and anxiety disorders precede alcohol use, it may not be self-medication at work, but rather the mood and anxiety problems may have occurred at an earlier age (often in childhood and younger adolescence), before alcohol entered the picture

  • Alcohol may be used to cope with other physical symptoms of these disorders or their treatments (e.g., sleep problems or bodily complaints such as headaches) rather than to self-medicate feelings of sadness or anxiety

  • A common factor such as a personality disorder or trait may lead a person to be vulnerable to both alcohol use and depression or anxiety

  • In some cases, the issues may occur independently and then interact with each other later on

State of the evidence

Evidence of self-medication has, to date, been indirect at best. In both community and clinical settings, research strategies make it easier to demonstrate correlation (condition A and B occur together) than they do causation (condition A causes condition B). For example, it’s easier to show that depression coexists with alcohol use than it is to show that someone with depression clearly chooses to drink alcohol because they are feeling ‘down’— as opposed to some other reason.

A further complexity is that people who consume alcohol are not a homogeneous population. Some use alcohol in beneficial or non-problematic ways, while others engage in patterns of use that are clearly harmful. The social, emotional and contextual influences for a person in the first group may be very different from the characteristics and risk factors of an individual in the latter group.2 To be meaningful, research must distinguish between these influences and anxiety or depression as causative factors in drinking behaviour.

Another issue is that studies depending on self-reported data may overstate the case for self- medication. In one study, for example, most of the male patients reported that they used alcohol to cope with lower-back pain, implying that self-medication was the cause of their alcoholism. When carefully examined, however, 82% of the patients had alcohol use disorders predating the onset of their pain. One can’t self-medicate a condition that doesn’t yet exist.

An anxiety connection

Research published in 2000 by French and American researchers Swendsen, Tennen, Carney et al.4 explored the self-medication theory as it applies to depression/anxiety (states, not diagnoses) and alcohol use. They studied the brief cycles of mood and anxiety states, patterns of drinking of alcohol (and desire to drink) and the effects of alcohol use on mood and anxiety states. Novel strategies were used to attempt to separate out risk factors other than mood that may also influence alcohol consumption. (The full methods and procedures for running the study, selecting participants, and limitations of approaches and findings are described in the published article.)

The study found that only anxiety-related states (not sadness or other negative moods) were found to predict increases in alcohol consumption. It also showed that the higher the anxiety scores, the higher the effects of alcohol on lowering levels of nervousness. The authors state that the study does not discount a potential relationship between negative mood and alcohol use, but rather that the nature of the relationship may be different than for anxiety. (One piece of the depression-alcohol relationship we do know is that depression can be a consequence of problem drinking.)

One must remember that the Swendsen et al. study looked at “states” and “behaviours,” not disorders. Feeling depressed or anxious and choosing to drink are all like distant cousins of, respectively, major clinical depression, anxiety disorders or alcoholism; they’re related, yet significantly different.

The Swendsen et al. results—even with the focus just on states—do mirror those of an earlier international study that looked at these three phenomena when they were diagnosable; that is, they studied mood and anxiety disorders in relation to alcohol use disorders. In that study of six countries, including Canada, the onset of anxiety disorders was more likely to precede that of substance disorders in all countries2—whereas such a timeline for mood disorders was not found.

Timelines are important given that the self-medication theory cannot work unless alcohol use problems or disorders occur after the development of the mental disorder.

Two other researchers who examined those with anxiety disorders found that the self-medication theory is particularly relevant for those with phobias or social anxiety and those suffering from traumas such as sexual molestation.8,9 Their findings indicate that individuals coping with these types of anxiety problems may be at higher risk for later substance use problems.

Gender differences

The other major finding of the Swendsen et al. study was that the rate of men self-medicating nervousness with alcohol was higher than for women. Men were more likely than women to use alcohol after a rise in anxious states. Men were also more likely to report that they could have “really used a drink” when they had previously been nervous. The authors state that the findings don’t indicate a lack of possible self-medication in women, but that “the effect size is larger or more consistent for men, a finding similar to past investigations."

While the reasons for the gender differences are complex and more research needs to be done in this area, the authors point to both biological and psychological factors. For example, alcohol is, in fact, less consistent in reducing anxiety in women,10 and men may have greater expectations for having tension reduced through alcohol use.11,12 Another study found an opposite pattern for depression: men perceive alcohol as providing less relief for depressive symptoms than women do (56% vs. 89%, respectively).


As just one of the mechanisms that could explain the high rate of co-occurring mental disorders and substance use, the self-medication theory is surprisingly pervasive in the academic literature. This article has summarized a few findings in just one small slice of that literature: alcohol and depression/ anxiety. But much more study is still needed.

For numerous reasons, it appears that self-medicating, in general, has much less scientific support overall than would be predicted from the lay media, but that there is a more solid base of support for self-medicating with alcohol following anxiety states, particularly for men. The evidence does not appear to be there for a similar relationship to depression.

And just to complicate things further, the Swendsen et al. study also found that for both men and women, feeling pleasant or happy was also predictive of alcohol consumption,4 showing that social reasons play an important role even if self-medication is true some of the time for some people.

It behooves the practitioner to look at timelines and motives and to test the claim of “I’m drinking to cope with my symptoms” by investigating all the possible alternatives and contexts. The latter may be true sometimes, but it may just as easily be an excuse or a wrong self assumption.

What do clinicians in Canada think?

Based on opinion research by the Centre for Addiction and Mental Health:

  • Social workers and counsellors view self-medication as a significantly more relevant issue in alcoholism than do physicians.

  • Non-psychiatrist physicians tend to view most psychopathology (i.e., psychological and behavioural problems) in alcoholics as the direct result of using alcohol as opposed to a possible coexisting mental illness

  • On treating alcohol use problems, all clinician groups agree on the value of cognitive-behavioural treatment, the need for medications to be combined with psychosocial interventions, and the importance of teaching self-monitoring and coping skills

About the Author

Sarah is Director of Public Education and Communications at Canadian Mental Health Association, BC Division, and Production Editor for Visions



  1. Ross, H. (1995). DSM-III-R alcohol abuse and dependence and psychiatric co-morbidity in Ontario: Results from the Mental Health Supplement to the Ontario Health Survey. Drug and Alcohol Dependence, 39, 111-128

  2. Merikangas, K.R., Mehta, R.L., Molnar, B.E. et al. (1998). Comorbidity of substance use disorders with mood and anxiety disorders: Results of the International Consortium in Psychiatric Epidemiology. Addictive Behaviors, 23(6), 893-907.

  3. Atkinson, J.H., Slater, M.A., Patterson, T.L. et al. (1991). Prevalence, onset, and risk of psychiatric disorders in men with chronic low back pain: A controlled study. Pain, 45, 111-121.

  4. Swendsen, J.D., Tennen, H., Carney, M.A. et al. (2000). Mood and alcohol consumption: An experience sampling test of the self-medication hypothesis. Journal of Abnormal Psychology, 109(2), 198-204

  5. For a review, see Raimo, E.B. & Shuckit, M.A. (1998). Alcohol dependence and mood disorders. Addictive Behaviors, 23(6), 933-946.

  6. Harrington, R., Fundge, H., Rutter, M. et al. (1990). Adult outcomes of childhood and adolescent depression. Archives of General Psychiatry, 47, 465-473.

  7. Rao, U., Ryan, N.D., Birmaher, B. et al. (1995). Unipolar depression in adolescents: Clinical outcome in adulthood. Journal of the American Academy of Child and Adolescent Psychiatry, 34,566-578.

  8. Baker, C.R. (2001). Drug use and mental disorders: An examination of the self-medication hypothesis. Dissertation Abstracts International: Section B: The Sciences & Engineering, 62(3-B), 1310.

  9. Randall, C.L. (2003). Self-medication in social phobia: A review of the alcohol literature. Addictive Behaviors, 28(2), 269-284.

  10. Wilson, G.T., Brick, J. Adler, J. et al. (1989). Alcohol and anxiety reduction in female social drinkers. Journal of Studies on Alcohol, 50, 226-235.

  11. Cox, B.J., Swinson, R.P., Shulman, I.D. et al. (1993). Gender effects and alcohol use in panic disorder with agoraphobia. Behaviour Research and Therapy, 31, 413-416.

  12. Kushner, M.G., Sher, K.J., Wood, M.D. et al. (1994). Anxiety and drinking behavior: Moderating effects of tension-reduction alcohol outcome expectancies. Alcohol Clinical and Experimental Research, 18, 852-860.

  13. Walsh, S.M. (2001). The effect of gender on selected factors of dual diagnoses. Dissertation Abstracts International: Section B: The Sciences & Engineering, 61(3-B), 6726.

  14. Meza, E.D., Cunningham, J.A., el-Guebaly, N. et al. (2001). Alcoholism: Beliefs and attitudes among Canadian alcoholism treatment practitioners. Canadian Journal of Psychiatry, 46, 167-172.

Stay Connected

Sign up for our various e-newsletters featuring mental health and substance use resources.