| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a Department of Psychiatry, University of Connecticut Health Center, Farmington
Correspondence: Nancy M. Petry, PhD, Department of Psychiatry, University of Connecticut Health Center, 263 Farmington Avenue, Farmington, CT 06030-3944. E-mail: petry{at}psychiatry.uchc.edu.
Decision Editor: Laurence G. Branch, PhD
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Pathological gambling • Older adults • Gender • Treatment
Pathological gambling affects about 1.6% of the adult population (Shaffer, Hall, and Vander Bilt 1999
). The personal and social effects of this disorder include significant financial losses, family problems, legal and employment difficulties, and psychological distress, including suicide (Petry and Armentano 1999). However, relatively little is known about the antecedents or correlates of pathological gambling in older adults.
Older age is often associated with lower rates of pathological gambling (National Research Council 1999; Shaffer et al. 1999). In general population surveys, prevalence rates of pathological gambling in older adults are quite low. For example, in the National Opinion Research Center's (NORC; 1999) telephone survey, only 0.4% of individuals aged older than 65 years met diagnostic criteria for pathological gambling.
Although these data may suggest that gambling is not a significant public health concern among older adults, this conclusion must be drawn with caution. Many studies do not break out prevalence rates by age groups, and usually less than 20% of individuals in telephone surveys are in the oldest cohorts. In the NORC (1999) study, for example, less than 400 respondents were 65 years or older. This sample size may not be large enough to derive accurate estimates of prevalence rates for a disorder that occurs at a relatively low frequency.
Only one known published report focused exclusively on gambling in older adults. McNeilly and Burke 2000 surveyed a nonrandom sample of 315 older adults in Nebraska. Ninety-one participants were recruited from gaming facilities (commercial and charitable bingo facilities or in a shuttle bus day-trip to a casino), and 224 were recruited from senior and retirement centers of American Association of Retired Persons chapter members. Among the participants surveyed at gaming venues, 11% were classified as pathological gamblers. About 3% of those surveyed from other community events were pathological gamblers. This study, albeit of a sample of convenience, suggests a clinically significant rate of disordered gambling in older adults, especially among those with recent participation in gambling activities.
The expansion of legalized gambling opportunities in the past 15 years may be associated with a rise in gambling participation, and this increase is particularly pronounced in older adults (NORC 1999). In a national telephone survey of gambling conducted in 1975, only 35% of individuals older than 65 years reported gambling in their lifetimes (Kallick, Suits, Dielman, and Hybels 1976). In contrast, 80% of individuals older than 65 years who were interviewed in 1998 reported lifetime participation in gambling activities (NORC 1999). Past-year gambling participation also increased dramatically in this oldest age group, with only 23% of older adults reporting past-year gambling in 1975 versus 50% in 1998.
A rise in disordered gambling behaviors seems to parallel the spread of legalized gambling opportunities. In a meta-analysis of all prevalence studies conducted in North America, Shaffer and colleagues 1999 found a statistically significant increase in the percentages of adults classified with pathological gambling in studies conducted since 1993 compared with those conducted earlier. Therefore, older adults may begin to experience gambling problems at higher rates as gambling participation expands in this age group.
Because of the paucity of clinical and research attention to gambling in older adults, the National Gambling Impact Study Commission 1999 recommended further investigation of this issue. This study is one step in this direction. Using data collected from treatment-seeking pathological gamblers, I compared gambling and related psychosocial problems across age groups, specifically focusing on gamblers in the oldest age cohort.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Procedure
The Addiction Severity Index (ASI; McLellan, Luborsky, Cacciola, and Griffith 1985) was administered to all participants at intake to treatment. The ASI assesses severity of medical, psychiatric, employment, family/social, legal, alcohol, and drug problems experienced in the past month. Composite scores are derived from responses to items within each of these problem areas. Responses are standardized and summed to produce a mathematical estimate of status in each area and range from 0.00 to 1.00, with higher scores indicative of more severe problems. A number of studies have demonstrated the reliability and validity of this instrument in a variety of substance-abusing populations (Kosten, Rounsaville, and Kleber 1983; McLellan, Alterman, Cacciola, Metzger, and O'Brien 1992; McLellan et al. 1985), and the ASI is one of the most commonly used instruments for both clinical and research purposes in addiction populations.
A gambling section of the ASI has also been created (Lesieur and Blume 1991a). It has adequate-to-excellent reliability and validity in assessing gambling problems in substance abusers who also present with gambling problems, as well as among individuals with a primary diagnosis of pathological gambling (Lesieur and Blume 1991a, Lesieur and Blume 1992; Petry 2001).
All participants also completed the South Oaks Gambling Screen (SOGS; Lesieur and Blume 1987). Scores of greater than or equal to 5 are indicative of a diagnosis of pathological gambling, and all participants included in this report scored over 5. SOGS scores were not a criterion for treatment entry, but these scores suggest that all individuals seeking gambling treatment were likely to meet diagnostic criteria for pathological gambling.
Data Analysis
Participants were divided into age groups that, broadly defined, covered young adulthood (18 to 35 years), middle age (36 to 55 years), and older adulthood (56 years and older). Basic demographics were compared across the groups using chi-square tests for categorical data and analyses of variance for continuous data. Variables that were nonnormally distributed were transformed when possible. For variables that could not be normally distributed even after transformation, nonparametric tests were used to evaluate differences among the groups.
Because more women were in the middle and older age groups compared with the younger group, subsequent age-related comparisons controlled for gender. First, multivariate analysis of covariance (MANCOVA) was used to evaluate differences among the groups, with gender and age categories as fixed factors and ASI composite index scores as the dependent variables. Age emerged as a significant predictor in the overall analysis, as well as in six of the eight domains. For domains in which overall F tests were significant, Dunnett's post-hoc tests compared each age group to the others. Further, when an overall test for a domain was significant, subsequent analyses evaluating differences with respect to specific variables making up these domains were protected against multiple comparisons (Tabachnick and Fidell 1996). Analyses of covariance then evaluated differences among the groups on specific items associated with each problem area. Gender was included as a fixed covariate in these analyses. For dichotomous variables associated with domains that differed among the age groups, chi-square analyses were used.
Data analyses were conducted on SPSS (SPSS Inc., Chicago, IL, 1997). The alpha value was .05, and all tests were two-tailed. Missing data, composing less than 5% of the total data, were not interpolated, so degrees of freedom vary slightly depending on how many participants responded to each item.
Because age is a continuous variable, these analyses were repeated using age as a continuous, rather than a categorical, variable. The overall MANCOVA produced similar results to those reported herein, and therefore, for ease of presentation, this article presents data with age groups trichotomized. Individuals in the middle age range demonstrated very similar patterns in ASI scores, regardless of whether ages were broken down by decades (e.g., 30–39, 40–49, 50–59), or according to the simpler three-group classification presented within this article. Finally, these same analyses were conducted classifying only adults older than age 60 years in the oldest age cohort. Again, similar results were obtained, but only 8% of the sample was aged older than 60 years, so data are presented with the larger group of older gamblers including those aged older than 55 years.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
2 (2, N = 343) = 18.24, p < .001. The three age groups also differed in terms of marital status, 2(8, N = 343) = 71.6, p < .001, with the oldest age group most likely to be married and the youngest age group most likely to be single. No other demographic differences were noted among the groups. The preferred forms of gambling are also shown across the three age groups. Slot machine gambling was the most popular form of gambling among the middle and older age groups. These age groups also contained higher proportions of women. Preferences for slot machine gambling differed significantly between genders, with only 18% of men compared with 61% of women preferring slot machine gambling, 2(1, N = 343) = 62.51, p < .001. Preferences for slot machine gambling also differed significantly across the age groups, 2(2, N = 343) = 37.94, p < .001.
|
Table 2 shows gambling-related variables. Because of the gender difference across the age groups, values are presented separately for men and women within each age category. ASI gambling composite scores differed significantly across the age categories, F(2,337) = 3.07, p < .05, with the middle age group showing the most severe gambling difficulties on this index of recent (past month) gambling problems. Gambling ASI scores did not differ by gender.
|
Older gamblers were likely to begin gambling later in life, F(2,329) = 21.29, p < .001, as were women, F(1,329) = 86.63, p < .001. An Age x Gender effect was noted, F(2,329) = 9.63, p < .001, with older women not initiating gambling until an average age of 42 years. This pattern was similar with respect to the age at which participants began gambling regularly, and the effects of age, gender, and Age x Gender were all significant, with F(2,333) = 58.85, F(1,333) = 84.15, and F(2,333) = 9.78, ps < .001. The total number of years of gambling problems also differed among the three age groups, F(2,333) = 6.95, p < .01, as well as between genders, F(1,333) = 29.90, p < .01. The Age x Gender interaction was significant as well, F(2,333) = 4.59, p = .01. Before entering treatment, men (especially older men) experienced a longer duration of gambling problems than women did.
Days of gambling in the past month differed among the three age groups, F(2,336) = 4.07, p < .05, with the middle age group gambling most frequently. Gender and the interaction between gender and age were not related to days of gambling. Amount wagered in the past month did not differ between genders. However, the Age x Gender interaction effect was significant, F(2,336) = 4.60, p < .05, with older women wagering the largest amounts. Amount gambled as a percentage of monthly income also varied across the age groups, F(2,323) = 3.00, p < .05, as well as between genders, F(1,323) = 4.57, p < .05. Women gambled large proportions of their monthly incomes prior to entering treatment, and this effect was especially pronounced in older women, who gambled in excess of 200% of their incomes. No significant differences in gambling debt or previous gambling treatment were noted across groups.
Employment, social, and legal variables are shown in Table 3 . ASI employment composite scores differed among the three age groups, F(2,322) = 8.70, p < .05, with the middle age group having the least severe problems. Average responses to some of the items that are included in the employment section are shown in Table 3 , and, as a reference point, employment composite scores in drug-abusing patients usually range from 0.6 to 0.8, substantially higher than scores in these pathological gamblers. Not surprisingly, fewer older participants were employed full time compared with younger participants, 2(2, N = 343) = 30.48, p < .001, but only 9.1% of the men and 18.5% of the women in the oldest cohort were retired (data not shown). Past-month income differed across the age groups, F(2,322) = 5.18, p < .01, with the middle age group having the highest monthly income.
|
2(2, N = 343) = 6.05, p < .05.
Controlling for gender, ASI legal scores also differed across the age groups, F(2,328) = 3.62, p < .05, with most severe legal problems in the youngest age group. No age effects emerged with respect to the percentages ever incarcerated in their lifetimes. However, the youngest age group was more likely to be on probation or parole, 2(2, N = 343) = 12.02, p < .01. About 10% of the sample was awaiting legal charges, trial, or sentencing, and these percentages did not differ among the age groups. In the past month, significantly more of the younger participants reported engaging in illicit activities than did the older participants, 2(2, N = 343) = 15.91, p < .001.
Drug and alcohol variables are shown in Table 4 . Controlling for gender, ASI composite scores were significantly different across the three age groups, F(2,332) = 5.18 and 3.89, ps < .05, for alcohol and drugs, respectively. The youngest age group had the highest ASI scores, whereas the two older age groups did not differ from one another. Younger gamblers were more likely to smoke cigarettes than older gamblers were, 2(2, N = 343) = 10.25, p < .01. The majority of participants in the younger and middle age groups had smoked marijuana in their lifetimes, although marijuana use was less prevalent in the older age group, 2(2, N = 343) = 37.35, p < .001. A similar age effect was noted with respect to cocaine use, 2(2, N = 343) = 10.87, p < .01. No age-related differences emerged in terms of percentages of the participants who had ever been treated for a substance-use disorder. Relative to the youngest age group, the older participants were also less likely to have used illicit drugs in the month prior to entering treatment for gambling, 2(2, N = 343) = 24.45, p < .001.
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
The older gamblers demonstrated different gambling histories and patterns than the younger gamblers did. The oldest group of gamblers had ASI gambling composite scores that were lower than the middle age group. Compared with middle-aged gamblers, the older gamblers wagered on fewer days. Age x Gender effects emerged in terms of amount wagered per month, with older female gamblers spending the most money, and the highest percentage of their income, gambling.
Age x Gender effects were noted with respect to development of gambling problems as well. Among the older female gamblers, age of gambling initiation and regular gambling did not occur until much later in life. An individual analysis of the female pathological gamblers older than 55 years of age revealed that 52% of these women had never gambled and 89% did not begin regular (weekly or more frequent) gambling until casinos became legalized on Native American reservations in Connecticut in 1992. Most of the women reported casino gambling, and slot machines in particular, to be their preferred form of gambling.
In contrast, the older men generally reported a lifelong history of gambling. Over 75% began gambling in their teenage years, and 77% began regular gambling prior to the age of 30 years. Because most states had no forms of legalized gambling in the 1950s to 1960s when they began gambling, these men probably were participating in illicit forms of gambling during their youth. The high rate of incarceration (32%) in these older male pathological gamblers suggests that they engaged in a variety of illegal behaviors when they were younger.
Prior to the widespread growth in legalized gaming, pathological gambling was almost exclusively a male disorder, and virtually no reports of female treatment-seeking pathological gamblers exist prior to the early 1990s (see Lesieur 1988, Lesieur 1993; Lesieur and Blume 1991b; Marks and Lesieur 1992). Although male gender is still considered a risk factor for development of gambling problems, rates of pathological gambling currently are estimated to be only 1.5 to 2 times that found in women (NORC 1999).
This gender gap may be narrowing with the widespread legalization of gambling. As the current generation grows up accustomed to legalized gambling, development of gambling problems in the later stages of life, as shown in this sample of women, may become a less common phenomenon. Now that women are exposed to gambling throughout life, those who may be more prone to develop gambling problems may do so at younger ages. As noted with substance-use initiation, opportunity to sample may be the greatest predictor of involvement and subsequent development of problems (Van Etten, Neumark, and Anthony 1999). The gender differences noted with respect to prevalence of substance-use disorders seem to be related to girls having fewer opportunities than boys do to sample illicit drugs. Girls who have had the opportunity to sample drugs use them at the same rates as boys do (Van Etten et al. 1999). Likewise, more and more women are gaining exposure to gambling opportunities, and therefore the gender differences in gambling participation seem to be decreasing (NORC 1999). More research is necessary to examine the prevalence and correlates of disordered gambling among women, especially as access to legalized gambling spreads.
In the present sample of treatment-seeking pathological gamblers, close to 50% were female. Increases in treatment seeking among women may be reflective of a well-known phenomenon that women seek mental health treatment services more often than men do, and they do so after experiencing symptoms for a shorter duration (e.g., Garfield 1994). Women entered treatment an average of 4 to 5 years after development of gambling problems, compared with an average of 11 years for men. These data are consistent with gender-specific treatment-seeking patterns in other mental health disorders (Horwitz 1977; Kessler, Brown, and Broman 1981).
Other differences in psychosocial problems were noted across the age groups as well. Family and social problems were less severe in the older treatment-seeking pathological gamblers. The older gamblers were more satisfied with their living situations than were the younger gamblers. However, older pathological gamblers scored higher on the ASI employment composite index. As expected, older adults were less likely to be employed full time, although few were retired. Their monthly incomes were lower than the middle age group's, yet they showed trends toward wagering larger amounts of money and increased debt. Older adults may have a more difficult time paying off gambling debts, because many receive limited incomes that are unlikely to increase over time. Therefore, financial counseling may be an important component of treatment for older gamblers.
Although few of these individuals seeking treatment for pathological gambling were actively abusing drugs or alcohol in the month prior to entering treatment, the older gamblers had less serious current substance-abuse problems than the younger gamblers did. The older gamblers were less likely to have used illicit drugs in the month prior to entering treatment. About a third of the participants, regardless of age group, had a history of substance-abuse treatment, and these rates of substance-abuse problems are consistent with other studies of treatment-seeking pathological gamblers (Lesieur, Blume, and Zoppa 1986; Ramirez, McCormick, Russo, and Taber 1983). Illegal activities showed a similar trend, with the youngest age group showing more serious current problems, although lifetime problems were consistent across the age groups. This pattern may be reflective of a "maturing-out" phenomenon of illegal behaviors as individuals age (Winick 1962).
Several limitations to this study affect the generalization of these findings. First, only 13% of the sample was over age 55 years. Although very similar effects were noted when analyses were limited to even older adults (older than age 60 years), the sample size became quite small (8%, n = 28), especially when broken down by gender. These low percentages of older adults seeking gambling treatment may emerge because older age may be a protective factor against development of gambling problems (NORC 1999), or it may reflect a general reduction in treatment seeking for mental health problems in older age groups (Shapiro, Skinner, Kessler, et al. 1984).
Second, because this sample was drawn from a group of treatment-seeking pathological gamblers, these results cannot be generalized to the larger group of pathological gamblers who do not enter treatment programs. The results are further limited by fact that all the gamblers were from Connecticut. Whether these results can be generalized to treatment-seeking pathological gamblers in other areas of the country remains to be determined. Large prevalence studies of pathological gamblers in the general population will be necessary to better understand the association between gambling and psychosocial problems and how these variables may differ with age.
Despite these limitations, this is the first known study evaluating gambling and psychosocial problems of older adult pathological gamblers. A number of differences emerged across the age groups, and these differences may suggest that specific issues should be addressed to attract and retain older pathological gamblers in treatment programs. Integrated treatment programs for substance-abuse and pathological gambling, for example, may not be appropriate for older pathological gamblers, because so few have current substance-use problems. Treatments that address financial problems, especially as related to an older age group with fixed incomes, may be particularly useful for this population given their high monthly gambling expenditures and relatively low incomes. Finally, gender-specific treatments may be useful in older gamblers, as the onset of disordered gambling later in life for female gamblers is a very striking finding in this study. Further understanding of pathological gambling and its treatment in older adults is needed.
Practice Concepts
The Forum
Book Reviews
|
Acknowledgments |
|---|
JoAnne Boccuzzi, Jaime Kelley, and Cheryl Molina assisted in data collection and management, and the staff at the Compulsive Gambling Treatment Program, Bettor Choice Programs, and the Connecticut Council on Problem Gambling are thanked for their participation in this project. Dr. Richard Fortinsky provided helpful suggestions on the manuscript.
Received for publication May 11, 2001. Accepted for publication August 31, 2001.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  F. W. Preston, P. D. Shapiro, and J. R. Keene Successful Aging and Gambling: Predictors of Gambling Risk Among Older Adults in Las Vegas American Behavioral Scientist, September 1, 2007; 51(1): 102 - 121. [Abstract] [PDF]
|
|
|
|
 |
|
 A. N. Burge, R. H. Pietrzak, C. A Molina, and N. M. Petry Age of Gambling Initiation and Severity of Gambling and Health Problems Among Older Adult Problem Gamblers Psychiatr Serv, December 1, 2004; 55(12): 1437 - 1439. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Desai, P. K. Maciejewski, D. J. Dausey, B. J. Caldarone, and M. N. Potenza Health Correlates of Recreational Gambling in Older Adults Am J Psychiatry, September 1, 2004; 161(9): 1672 - 1679. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Grant and R. Grosz Pharmacotherapy Outcome in Older Pathological Gamblers: A Preliminary Investigation J Geriatr Psychiatry Neurol, March 1, 2004; 17(1): 9 - 12. [Abstract] [PDF]
|
|
|
|
|
|
O. Kausch Pathological Gambling among Elderly Veterans J Geriatr Psychiatry Neurol, March 1, 2004; 17(1): 13 - 19. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
