RESEARCH QUESTION
What factors contribute to the risk for pathological gambling?

PURPOSE
To examine potential risk factors for pathological gambling in three domains: gambling behaviours, substance abuse and other problem behaviours, and sociodemographic factors.

HYPOTHESES
None stated.

PARTICIPANTS
Two thousand six hundred thirty one US residents aged 18 or older.

PROCEDURE
A national random-digit-dial telephone survey was conducted between August 1999 and October 2000. A total of 4338 eligible households from across the US were contacted. A total of 2631 households agreed to participate and the adult with a birthday closest to the date of the survey was interviewed.

MAIN OUTCOME MEASURES
Gambling participation in 15 types of gambling was assessed for frequency played during the previous year (from never to every day) and money won/lost the last time they engaged in the activity. The 15 activities assessed were raffles/charity, lottery, pulltabs, internet gambling, casino, horse or dog track, off-track, gambling machines, cards, games of skill, video lottery, bingo, dice, sports betting, and other. Gambling volume was calculated by multiplying frequency of gambling by absolute value of latest win/loss. Pathological gambling was assessed with the 13-item Diagnostic Interview Schedule (DIS)-IV which measures 10 criteria (e.g., preoccupation with gambling). Ethanol consumption was assessed with questions regarding the frequency and quantity of consumption of 5 types of alcohol (i.e., beer/malt liquor, fortified wine, wine, liquor, and wine cooler). Drug use was assessed with questions regarding frequency of use of each of four types of drugs (i.e., marijuana, cocaine, barbiturates and sedatives, narcotics and analgesics). Alcohol and drug abuse/dependence were assessed with the DIS for DSM-IV. Number of criminal offenses was assessed on the basis of self-reported involvement in 12 common crimes (e.g., larceny and bad cheques). Socioeconomic Status (SES) was assessed as the average of family income, years of education and occupational prestige.

KEY RESULTS
Results are presented in terms of incidence rate ratios (IRR) which is the factor by which pathological gambling is multiplied for every unit increase in the independent variable under consideration (e.g., casino gambling has an IRR of 1.40 which means that for every $1000/year of casino gambling, gambling pathology symptoms increase by 40%).

Of the 15 types of gambling investigated, only five (casino, pulltabs, cards, bingo and sports betting) were found to significantly contribute to individual gambling. Casino, lottery and bingo were the most common gambling activities engaged in and were found to significantly contribute to gambling pathology in the entire sample.

Gambling frequency, average win/loss, and number of types of gambling games played each predicted gambling pathology. After these factors were controlled for, alcohol abuse/dependence was the only ‘problem behaviour’ that predicted gambling pathology. After controlling for gambling behaviours and substance use/criminal offending, ethnicity and SES predicted gambling pathology.

LIMITATIONS
None stated.

CONCLUSIONS
Diagnoses of pathological and problem gambling may have complex causes beyond mere frequent gambling or making large bets. Risk for pathological gambling is related to gambling versatility, alcohol pathology, and membership in at-risk sociodemographic groups.