RESEARCH QUESTION
How prevalent is problem gambling behavior among Australian veterans suffering from posttraumatic stress disorder (PTSD)?
PURPOSE
There is a possible relationship between substance use and problematic gambling, a known association of PTSD and substance abuse, and possibly a shared mechanism of self-medication underlying both substance abuse and gambling. It is therefore important to describe actual gambling behaviors and to identify the predictors of those behaviors in veterans, particularly treatment-seeking veterans with combat-related PTSD.
HYPOTHESIS
None stated.
PARTICIPANTS
A total of 194 Australian veterans who were consecutively admitted into group therapy treatment at Department of Veterans’ Affairs-funded PTSD treatment programs were invited to participate. Of those veterans, 153 (100% male; average age = 54 years) chose to participate.
PROCEDURE
Prior to admission, all veterans were clinically assessed to ensure PTSD diagnosis. On admission, all participants completed a routine questionnaire that included measures of PTSD, alcohol use, depression, and anxiety. A gambling activities survey was also included.
MAIN OUTCOME MEASURES
PTSD symptoms were assessed via the Posttraumatic Stress Disorder Check List. Alcohol use was measured by the Alcohol Use Disorders Identification Test, the World Health Organization’s screening test for identifying early alcohol problems. To measure depression and anxiety the Hospital Anxiety and Depression Scale was used. It comprised two scales that provided incremental severity measures of emotional disorder. The self-report South Oaks Gambling Screen (SOGS) assessed gambling severity (scores = 5 were indicative of probable problem gambling). The SOGS was a lifetime measure: respondents indicated whether they had ever engaged in the endorsed behavior. The DSM-IV gambling scale probed the extent to which respondents experienced a range of psychological, behavioural, and psychosocial features of problem gambling, including preoccupation, development of tolerance, irritability, and gambling as an escape (a cut-off point of 5 was used to identify problem gamblers). The timeframe for the DSM-IV was indeterminate; some questions related to past experiences, whereas others probed present behaviors. The SOGS and the DSM-IV had four items in common: chasing losses, control of gambling, borrowing resources, and lying about the extent of gambling behavior. Finally, participants were asked whether, in the last year, they had participated in any of 11 popular gambling activities and, if so, how many times per week, month, or year.
KEY RESULTS
The proportion of veterans (84%) who reported gambling activity in the past year was similar to the general community (82%). Type of Gambling. Participants were twice as likely to gamble on lotto, 50% more likely to gamble with scratch tickets, over 70% more likely to gamble on both electronic gaming machines (EGMs) and horses, and 4 times more likely to gamble on keno than the general community. Gambling Frequency. Fifty one percent reported gambling on lotto weekly or more often, as did 29% for EGMs, 16% for scratch tickets, and 10% for horses. Twenty-two percent of participants gambled on one type of gambling weekly or more often, as did 38% on two or more different types of gambling. Gambling Severity. Twenty nine percent met criteria for probable problem gambling on the SOGS and 17% on the DSM-IV gambling scale. These rates compared with 3% for the Australian community generally. Study participants were 34 to 64 times more likely to be problem gamblers when compared with the matched (50-64 year-old) Australian age-group. Veterans who endorsed gambling to escape were 23 times more likely to be classified as a probable problem gambler. Predictors. The key predictors of problem gambling were gambling on EGMs weekly or more often, gambling at casino tables, and living alone. There was no significant relationship between problem gambling, PTSD, anxiety, depression, or alcohol use.
LIMITATIONS
A convenience sample of PTSD treatment-seeking veterans was used and the results are subject to limitations associated with non-probability sampling. The sample size required the data be simplified to avoid violating statistical assumptions. The lack of relationships may have been due both to the fact that all study participants were diagnosed with PTSD, and that there was multiple psychiatric symptomatology within the sample.
CONCLUSIONS
Overall, the present study found support for gambling among PTSD treatment-seeking veterans. Moreover, the study identified an entrenched gambling culture among PTSD treatment-seeking veterans.
