the DSM-III (Lesieur and Klein, 1987). In a school newspaper survey of over 1,100 students at an inner-city, largely minority Atlantic City high school, 62 percent reported having gambled at area casinos, and 9 percent reported gambling at least once a week (Arcuri et al., 1985). In another study among students from six colleges and universities in New York, New Jersey, Oklahoma, Texas, and Nevada (Lesieur et al., 1991), using the South Oaks Gambling Screen (SOGS), the lifetime gambling rate was found to be 85 percent, the rate of problem gambling was 15 percent, and the rate of probable pathological gambling was 5.5 percent (Lesieur and Blume, 1987).2 Comparable lifetime gambling rates were found in a Minnesota study of 1,094 youths ages 15-18 (including 684 from a random telephone sample and 410 from a school sample) in which the rate of problem gambling was found to be 6.3 percent and the age of onset for over half the subjects was reported to be before or during the sixth grade (Winters et al., 1993a). Finally, in a recent review of 12 U.S. and 5 Canadian adolescent gambling studies, Jacobs found that in the past 10 years the number of teenagers ages 12 to 17 reporting serious gambling problems has increased from 50 to 66 percent. The age of onset for gambling has dropped so that now, throughout America, the majority of 12-year-olds have already gambled (Jacobs, in press).
Studies of those who seek help for themselves or others indicate that gambling severity and frequency varies by age. A recent analysis of problem gambling help-line calls in Texas revealed that the frequency of calls increased with age, peaked at ages 35 to 44, and declined for callers age 45 and older (Cox, 1998). In fact, adults age 55 and older who called about their own gambling problems (14 percent of all callers) were comparable in frequency to those age 18 and younger who called about their own gambling problems (13 percent of all callers). The percentage of calls about a problem gambler from a friend, family member, or other concerned person followed a similar age pattern. Although interesting and clinically meaningful, these help-line data alone do