the world and address criteria that meet a certain threshold of severity. In fact, because there is a perceived similarity between some of the symptoms of pathological gambling and affective disorders, the DSM-III-R states that, during a manic or hypomanic episode, loss of judgment and excessive gambling may follow the onset of the mood disturbance. When manic-like mood changes occur in pathological gamblers, they are generally related to winning streaks, and they are usually followed by depressive episodes because of subsequent losses. Periods of depression tend to increase as the disorder progresses. Although somewhat reasonable, the current understanding of this progression is informed only by anecdotal information and case histories of patients who have entered treatment.
In one of the first studies to distinguish whether a depression or a gambling disorder came first (i.e., whether pathological gambling was a primary or secondary disorder), McCormick et al. (1984) found that 76 percent of gamblers in treatment met the criteria for a major affective disorder. They also found that gambling preceded depression 86 percent of the time, and even the onset of pathological gambling preceded depression. In another study of gamblers both in and out of treatment, the investigators found, using a structured diagnostic interview, that gamblers in treatment, compared with untreated controls, were about three times more likely to meet criteria for major depression (Specker et al., 1996).
Perhaps the only general population study that has examined the relationship between problem gambling and depression was the ECA study described above. Unfortunately, of the five sites that were involved in the ECA study, only in St. Louis did investigators ask questions to assess pathological gambling. The study found that problem gamblers were at least three times as likely to meet criteria for depression, schizophrenia, alcoholism, and antisocial personality disorder than nongamblers (Cunningham-Williams et al., 1998; Cunningham-Williams, 1998). Because the diagnostic instrument used ascertains the age of onset of psychiatric symptoms, the investigators were able to determine that the depression preceded the gambling problems, unlike the Specker et al. study.
Studies have shown no association between problem gam-