who also had gambling problems were twice as likely as those without gambling problems to have more drug overdoses, greater past treatment for alcohol and for drugs, and more drug use in the past month (Steinberg et al., 1992). They were 1.5 times as likely as those without gambling problems to use opiates and solvents. Studies have found that persons admitted to chemical dependence treatment programs are three to six times more likely to be problem gamblers than people from the general population (Lesieur and Heineman, 1988; Lesieur et al., 1986; Steinberg et al., 1992; Lesieur and Rosenthal, 1998). Natural history studies of gambling and substance abuse are rare. Ramirez and colleagues found that substance use predated the onset of gambling problems in their study addressing the age of onset of these behaviors (Ramirez et al., 1983). The results might have been different if the onset of drug or alcohol problems, rather than use only, had been evaluated.
Given the reported high prevalence rates of alcoholism among American Indians, one concern since passage of the Indian Gaming Regulatory Act of 1988 is the suspected increase in the comorbidity of alcoholism and gambling for this population. In the first study exploring this relationship (Elia and Jacobs, 1993), researchers, using the South Oaks Gambling Screen among a small sample of 85 patients on an alcohol treatment ward of the Ft. Meade Veterans Administration hospital, found that American Indians compared with whites had a higher rate of probable pathological gambling (22 compared with 7.3 percent) and had more problems from gambling (41 compared with 21.3 percent).
Studies of pathological and problem gambling among general population samples are needed to minimize the bias inherently attributable to treatment samples. The St. Louis component of the ECA study showed that, after adjusting for the effects of a number of variables, gambling without problems and problem gambling were associated with substance use, abuse, and dependence (Cunningham-Williams et al., 1998). In another study, a random sample of Texans was interviewed by telephone about both their gambling and their substance use behaviors (Feigelman et al., 1998). The study found that persons with both conditions were more likely than individuals with a single disorder to be psychosocially dysfunctional. These results demonstrate a new