Abstract
Background
Pathological gambling (PG) is an important public health problem that is prevalent, costly to society, and associated with substance misuse, depression, domestic violence, crime, and suicide. Despite these challenges, little is known about the physical health and medical correlates of PG. The goal of this project was to assess self-reported chronic medical conditions, medication usage, lifestyle choices, health care utilization, quality of life variables, and body mass index (BMI) in persons with and without PG.
Methods
Subjects with PG and community controls were systematically assessed for their medical health, lifestyle choices, medication usage, and health care utilization. We administered the Medical Outcome Study Short-Form 36 Health Survey to assess perceived health and quality of life. BMI was calculated for all subjects. Obesity was defined as having a BMI ≥ 30 kg/m2.
Results
We compared 95 subjects with DSM-IV PG (South Oaks Gambling Screen [SOGS] score ≥ 5) and 91 control subjects without PG (SOGS ≤ 2) selected through random digit dialing from the general community. PG subjects and controls were similar in age and gender. Persons with PG had more medical and mental health conditions than controls, and were more likely to avoid regular exercise, smoke ≥ 1 pack/day, drink ≥ 5 servings of caffeine daily, and watch television ≥ 20 hours/week. They had more emergency department visits for physical and mental health conditions, were more likely to have been psychiatrically hospitalized in the past year, and were more likely to take psychotropic medication. They were less likely to have had regular dental visits and were more likely to put off medical care due to financial problems. Severity of gambling was positively correlated with number of medical conditions. Persons with PG had poorer self-reported health perceptions on all but one SF-36 subscale. Importantly, persons with PG had a higher BMI than controls and were more likely to be obese.
Conclusions
PG is associated with obesity, chronic medical conditions, poor lifestyle choices, worse quality of life, and the use of costly forms of medical care. Pathological gamblers are less likely to receive regular dental care and are more likely to be unable to pay for medical care. The implications of the findings are discussed.
Keywords: pathological gambling, health perceptions, obesity, health care utilization, lifestyle
Pathological gambling (PG) is prevalent, costly to society, and associated with substance misuse, depression, domestic violence, crime, and suicide.1–7 Nearly 90% of the general adult population participates in some form of gambling, 8 and an estimated 1.2%-3.4% develop PG, the most severe form of disordered gambling.1,2 Problem (“at risk”) gambling may be even be more common with lifetime prevalence estimates of 3.5%-5.1%.9 Additionally, PG is associated with mood and anxiety disorders, substance use disorders, and antisocial behavior.10 Despite these challenges, little is known about the physical health and medical correlates of PG.
Several clinical and epidemiologic studies have reported an association between disordered gambling and adverse health consequences beginning with small descriptive studies of both problem gamblers and their spouses that documented depression, insomnia, intestinal disorders, headaches and other stress-related disorders, as well as high rates of sick leave and poorer self-appraisal of health. 11–13 Following on the heels of these reports have been a spate of epidemiologic studies that have linked disordered gambling with adverse health conditions or behaviors. Morasco et al. 14 reported that among more than 43,000 persons assessed as part of the National Epidemiologic Survey of Alcohol and Related Conditions (NESARC) study, persons with PG were more likely to have symptoms of tachycardia, angina, cirrhosis, or other liver disease; they also reported greater health care utilization, including emergency department visits. In a survey of 1051 primary care patients, Pasternak and Fleming 15 found more alcohol and tobacco use, worse self-reported health, and more symptoms of heartburn and backache among the 6.2% with a gambling disorder than among non-gamblers. In adults seen in an urban primary care clinic, Morasco et al. 16 found that increasing gambling severity was associated with worse physical and emotional health. Associations between worse health indices and disordered gambling have been reported in adolescents, elderly people, persons receiving disability, dental clinic patients, and persons followed in a methadone clinic.17–21
These studies indicate that disordered gambling is associated with poor health outcomes, yet there have been no studies in which the medical health status of carefully assessed subjects with PG has been compared to that of an appropriate control group. We now report data from a comparison of subjects with PG and controls. We hypothesize that persons with PG are more likely than controls to have worse selfreported health and quality of life, to report more medical symptoms and syndromes, to have greater use of emergency and other medical services, to use more medications, and to be obese.
Methods
Subjects
Subjects were recruited in the course of a family study of PG through a study registry, advertisements, meetings of Gamblers Anonymous, and word of mouth. They were interviewed between February 2005 and June 2010. Controls were recruited through random digit dialing methods by the Center of Social and Behavioral Research at the University of Northern Iowa (Cedar Falls, IA) and were approximately matched to PG subjects for sex and age.
Subjects with PG were required to have a South Oaks Gambling Score (SOGS) 22 ≥ 5 and a NORC DSM Screen for Gambling Problems (NODS) 6 score ≥ 5; they also had to meet DSM-IV PG criteria. 23 All subjects were at least 18 years, spoke English, and they could not have a psychotic, cognitive, or a chronic neurological disorder (e.g., Parkinson’s disease). Controls were required to have a SOGS score ≤ 2 and a NODS score of 0. Written, informed consent was obtained from all subjects according to procedures approved by University of Iowa Institutional Review Board.
Assessments
We modified the Medical History and Services Utilization Interview created by Frankenburg and Zanarini 24 to collect data on self-reported medical care and treatment, chronic medical conditions, health-related lifestyle choices, and medical services utilization, including emergency department visits, physician and dentist visits, hospitalizations, and psychotropic drugs taken. The instrument is reported to have moderate to excellent reliability.24 We also administered the Medical Outcome Short Form-36 (SF-36) Health Survey 25 to gather objective data about medical health and quality of life. Body Mass Index (BMI) was calculated for all subjects using their self-reported weight and height. BMI was calculated by dividing weight in kilograms by the height in meters squared. Obesity was defined as having a BMI of ≥ 30 kg/m2.
Statistical Analysis
PG and control subjects were compared on social and demographic characteristics using the Chi-square test for categorical variables and the Mann-Whitney test for dimension variables. Logistic regression modeling was used to test for differences in health-related lifestyle choices between PG subjects and controls. 26 Sex, age, race (Caucasian vs. other), and years of schooling were used as covariates. To account for multiple outcomes being studied, the difference between groups was tested using a multivariate score test, analogous to MANOVA for continuous outcomes. 27 The dichotomous health-related lifestyle choices were grouped as healthy lifestyle choices (e.g., regular exercise, taking a multi-vitamin) and unhealthy habits and behaviors (e.g., tobacco use, drinking alcohol). In addition to the multivariate tests, odds ratio estimates and 95% confidence intervals are reported for the group differences. The same logistic regression methodology was used to test for differences between PG subjects and controls for health care utilization and chronic medical conditions. Dichotomous health care utilization outcomes were grouped as regular medical care, financial barriers to medical care, psychotropic medication usage, and urgent medical care. Eleven of the PG subjects were recruited when they were in a clinical setting (e.g., inpatient, outpatient). To remove the natural bias caused by recruitment in these settings, the subjects were removed from the analyses for the outcomes grouped under urgent medical care and psychotropic medication usage. Under chronic medical conditions, outcomes were grouped as chronic physical health conditions and chronic mental health conditions.
MANOVA was used to test for group differences in quality of life measures (SF-36 results) and other dimensional measures of health. Sex, age, race, and years of schooling were used as covariates. In addition to the MANOVA test, estimates of mean group differences are reported, along with their confidence intervals and p values. Among subjects with PG, we also tested correlations between gambling severity (NODS and SOGS scores) and medical outcomes (BMI, total number of chronic medical conditions, and SF-36 results). Statistical significance was set at p<.05.
Results
Table 1 shows the social and demographic characteristics of 95 subjects with PG and 91 controls. As can be seen, the two groups were similar in age and gender. Overall, 90% of the subjects were Caucasian, 6% African-American, 2% Hispanic/Latino, and 2% American Indian. Using a dichotomous variable for race (Caucasian vs. other), the PG group had a smaller proportion of minority subjects. PG subjects had fewer years of education. Occupational status was assessed by asking subjects to check each category that applied during the last 12 months. Therefore, subjects could be classified as both employed and unemployed, for example. Subjects with PG were less likely to be homemakers or retired, but were more likely to be disabled. More subjects with PG were unemployed in the past year, but the difference was not significant. Subjects with PG were more likely to be separated or divorced. Among subjects with PG, the mean (SD) SOGS total score was 13.4 (3.7).
Table 1.
Demographic characteristics of PG and Control Subjects
| Characteristic | PG (n=95) |
Control (n=91) |
χ2, df | P-value |
|---|---|---|---|---|
| Female (%) | 58% | 63% | 0.44, 1 | .509 |
| Age, mean, years (SD) | 45.6 (12.8) | 49.4 (16.0) | 2.331, 1 | .127 |
| Caucasian (%) | 85% | 95% | 4.33, 1 | .038 |
| Occupational status (%) | ||||
| Employed | 77% | 75% | 0.11, 1 | .736 |
| Unemployed | 18% | 11% | 1.79, 1 | .181 |
| Student | 17% | 9% | 2.68, 1 | .102 |
| Homemaker | 4% | 15% | 6.64, 1 | .010 |
| Retired | 9% | 21% | 4.73, 1 | .030 |
| Disabled | 21% | 5% | 9.67, 1 | .002 |
| Any children (%) | 72% | 92% | 13.37, 1 | <.001 |
| Marital status (%) | FET | <.001 | ||
| Married | 35% | 80% | ||
| Divorced/separated | 36% | 8% | ||
| Widowed | 3% | 5% | ||
| Single | 26% | 7% | ||
| Years of school (mean, SD) | 14.1 (1.9) | 15.2 (2.4) | 7.371, 1 | .007 |
Mann-Whitney Test, FET=Fisher’s Exact Test
Table 2 shows the prevalence of health-related lifestyle choices. The two groups had differences in healthy lifestyle choices and unhealthy habits and behaviors (p<.001 for both tests). Specifically, PG subjects were more likely to avoid regular exercise (OR=2.38, p=.007), to drink alcoholic beverages while pregnant (OR=4.88, p=.041), to smoke ≥ one pack of cigarettes per day (or ≥ one can of chewing tobacco) (OR=6.51, p<.001), to drink ≥ 5 servings of caffeine daily (OR=4.08, p=.002), and to watch ≥ 20 hours of television weekly (OR=3.28, p=.003).
Table 2.
Health-related Lifestyle Choices in persons with PG and Controls
| Lifestyle Choice | Prevalence | Adjusted OR (95% CI) |
P- value |
|
|---|---|---|---|---|
| PG (n=95) |
Control (n=91) |
|||
| Score test for overall group difference in healthy lifestyle choices: χ2=28.9, df=5, p<.001 | ||||
| Lacks regular exercise | 57% | 36% | 2.38 (1.27, 4.47) | .007 |
| Takes a multi-vitamin | 37% | 60% | 0.55 (0.29, 1.06) | .075 |
| Takes natural remedies | 20% | 32% | 0.66 (0.32, 1.37) | .268 |
| High caffeine intake (≥ 5 servings daily) | 28% | 9% | 4.08 (1.67, 9.94) | .002 |
| Television use (≥ 20 hrs/week) | 38% | 18% | 3.28 (1.49, 7.21) | .003 |
| Score test for overall group difference in unhealthy habits and risky behaviors: χ2=28.9, df=9, p<.001 | ||||
| Drinks alcohol | 55% | 56% | 1.04 (0.55, 1.94) | .914 |
| Drink daily | 19% | 12% | 1.53 (0.62, 3.77) | .354 |
| Drank while pregnant1 | 17% | 6% | 4.88 (1.07, 22.26) | .041 |
| Ever used tobacco consistently | 68% | 46% | 2.43 (1.29, 4.56) | .006 |
| Currently use tobacco consistently | 51% | 19% | 3.70 (1.87, 7.35) | <.001 |
| Smokes ≥1 pack/day | 29% | 5% | 6.51 (2.32, 18.24) | <.001 |
| Used tobacco while pregnant1 | 35% | 18% | 2.12 (0.80, 5.63) | .130 |
| Injected illicit drug | 5% | 1% | .2122 | |
| Vehicular accident causing bodily injury | 45% | 33% | 1.68 (0.89, 3.17) | .107 |
OR=odds ratio, CI=confidence interval,
Only women who were pregnant included in analysis,
Fisher’s Exact Test
Table 3 compares health care utilization for PG and control subjects. The two groups were significantly different with respect to regular medical care (p=.003). Subjects with PG were more likely to have seen a specialist in the past year and to have received a magnetic resonance imaging/computerized tomographic scan, but were less likely to have seen a dentist for regular checkups (OR=0.40, p=.015). The two groups were also different with respect to financial barriers to medical care (p=.005). Subjects with PG were more likely to have delayed seeing a doctor because of insurance or other financial problems (OR=2.15, p=.048) and were more likely to be unable to pay for medical attention or to have gone into debt due to illness (OR=3.45, p=.002). Large group differences were observed for utilization of urgent medical care and psychotropic medications (p<.001). Subjects with PG were more likely to have at least one emergency department visit in the last year for a physical disorder or mental health/substance use disorder. Subjects with PG were more likely to have ≥ 1 hospitalization in the last year for mental health purposes. The groups were similar for hospitalizations for physical health reasons. Subjects with PG were much more likely to be taking psychotropic medications (OR=8.1, p<.001), with 37% of the PG subjects taking an antidepressant. Significant differences were observed for all classes of psychotropics, with the exception of antipsychotics and “other” (e.g., pregabalin).
Table 3.
Health Care Utilization in persons with PG and Controls
| Health Care Utilization | Prevalence | Adjusted OR (95% CI) |
P- value |
|
|---|---|---|---|---|
| PG (n=95) |
Control (n=91) |
|||
| Score test for overall group difference in regular medical care: χ2=19.7, df=6, p=.003 | ||||
| Has a primary care physician | 83% | 89% | 1.11 (0.38, 3.22) | .846 |
| Has seen doctor in last year | 77% | 82% | 1.19 (0.48, 2.96) | .702 |
| Dental visits for regular checkups | 63% | 85% | 0.40 (0.19, 0.84) | .015 |
| Has received MRI/brain scan | 62% | 41% | 2.44 (1.31, 4.55) | .005 |
| See any specialists | 60% | 48% | 2.09 (1.10, 3.97) | .024 |
| Has seen specialist in last year | 55% | 43% | 1.97 (1.05, 3.70) | .035 |
| Score test for overall group difference in financial barriers to medical care: χ2=10.7, df=2, p=.005 | ||||
| Delayed seeing doctor (insurance/finances) |
34% | 15% | 2.15 (1.01, 4.57) | .048 |
| Unable to pay/debt for physical problem | 36% | 12% | 3.45 (1.57, 7.60) | .002 |
| Score test for overall group difference in urgent medical care: χ2=39.8, df=8, p<.001 | ||||
| ED visits since age of 18 | 87% | 69% | 2.56 (1.12, 5.86) | .026 |
| ED visits last year | 38% | 19% | 2.40 (1.14, 5.51) | .022 |
| ED visits last year for physical health | 38% | 19% | 2.40 (1.14, 5.51) | .022 |
| ED visits last year for mental health | 10% | 0% | .0022 | |
| Hospitalization since age of 18 | 88% | 84% | 2.22 (0.79, 6.62) | .155 |
| Hosp. last year | 21% | 15% | 1.52 (0.66, 3.47) | .323 |
| Hosp. last year for physical health | 18% | 15% | 1.25 (0.53, 2.95) | .610 |
| Hosp. last year for mental health | 7% | 0% | .0112 | |
| Score test for overall group difference in psychotropic medication usage: χ2=36.9, df=7, p<.001 | ||||
| Psychotropic medications | ||||
| Any1 | 55% | 20% | 8.05 (3.65, 17.76) | <.001 |
| Antidepressants | 37% | 15% | 3.92 (1.80, 8.50) | <.001 |
| Mood stabilizers | 14% | 1% | 20.9 (2.56, 171.9) | .005 |
| Benzodiazepines | 17% | 7% | 4.19 (1.43, 12.26) | .009 |
| Sedative/hypnotics | 13% | 2% | 8.43 (1.67, 42.54) | .010 |
| Psychostimulants | 8% | 2% | .0052 | |
| Antipsychotics | 5% | 0% | .0512 | |
| Other | 5% | 0% | .0512 | |
OR=odds ratio, CI=confidence interval,
Not included in overall score test because outcome is derived from other outcomes,
Fisher’s Exact Test, MRI=magnetic resonance imaging
Table 4 compares chronic medical conditions of PG and control subjects, with the overall tests revealing significant group differences in both chronic physical (p<.001) and mental (p<.001) health conditions. Subjects with PG were more likely to be obese (OR=2.79, p=.004). With the exception of “cancer other than skin cancer” and diabetes, prevalence of other chronic medical conditions was higher in the PG subjects. The largest differences were observed for heartburn/stomach conditions (OR=4.98, p=.001), headaches (OR=2.84, p=.012), sleep disorders (OR=3.14, p=.002), mood/emotional concerns (OR=9.27, p<.001), anxiety, tension, or stress (OR=9.01, p<.001), and having a head injury with loss of consciousness (OR=3.84, p=.003). Subjects with PG were more likely to have ≥ 4 of the conditions assessed (OR=6.80, p<.001).
Table 4.
Chronic Medical Conditions of PG and Control Subjects
| Condition | Prevalence | Adjusted OR (95% CI) |
P- value |
|
|---|---|---|---|---|
| PG (n=95) |
Control (n=91) |
|||
| Score test for overall group difference in chronic physical health conditions: χ2=51.9, df=16, p<.001 | ||||
| Obesity | 49% | 28% | 2.79 (1.42, 5.48) | .004 |
| Arthritis/rheumatism | 21% | 20% | 1.69 (0.76, 3.90) | .190 |
| Diabetes | 7% | 8% | 0.94 (0.29, 3.00) | .693 |
| High blood pressure | 32% | 31% | 1.57 (0.75, 3.31) | .211 |
| High cholesterol | 22% | 15% | 2.30 (0.93, 5.70) | .117 |
| Chronic back pain | 29% | 25% | 1.26 (0.64, 2.48) | .470 |
| Asthma/chronic lung disease | 17% | 10% | 2.12 (0.85, 5.32) | .099 |
| Heart disease/coronary heart disease | 11% | 10% | 1.57 (0.53, 4.67) | .401 |
| Heartburn/stomach condition | 28% | 10% | 4.98 (1.97, 12.57) | .001 |
| Enlarged prostate/urinary condition | 16% | 14% | 1.57 (0.65, 3.80) | .299 |
| Headaches | 26% | 14% | 2.84 (1.23, 6.53) | .012 |
| Cancer (other than skin cancer) | 8% | 11% | 0.86 (0.31, 2.40) | .807 |
| Thyroid disease | 16% | 12% | 1.99 (0.80, 4.99) | .133 |
| Low sex drive | 11% | 4% | 2.89 (0.83, 10.04) | .089 |
| Sleep disorders | 37% | 18% | 3.14 (1.52, 6.52) | .002 |
| Head injury with loss of consciousness | 25% | 9% | 3.84 (1.56, 9.44) | .003 |
| Score test for overall group difference in chronic mental health conditions: χ2=60.8, df=2, p<.001 | ||||
| Mood/emotional concerns | 64% | 24% | 9.27 (4.36, 19.71) | <.001 |
| Anxiety, tension, or stress | 49% | 15% | 9.01 (4.01, 20.23) | <.001 |
| ≥ 4 conditions2 | 66% | 36% | 6.80 (3.14, 14.71) | <.001 |
| Unemployed for health reasons1, 2 | 50% | 40% | 2.07 (0.38, 11.22) | .400 |
OR=odds ratio, CI=confidence interval, SD=standard deviation
Analysis restricted to subjects who are unemployed,
Not included in overall score test.
Table 5 compares the two groups on BMI, number of psychotropic medications, total number of chronic medical conditions, medical outcomes, and quality of life. The overall MANOVA test revealed significant group differences (p<.001). For all scales of the SF-36, PG subjects scored lower than control subjects. Compared with controls, PG subjects scored particularly low on Emotional Role, Social Functioning, Mental Summary, and Vitality. Subjects with PG had a much higher BMI (difference = 3.6, p =.001), and had more chronic medical conditions (difference = 2.4, p<.001).
Table 5.
Quality of Life Measures for PG and Control Subjects
| Measure | Mean (SD) | Adjusted Diff. (95% CI) |
P- value |
|
|---|---|---|---|---|
| PG (n=95) |
Control (n=91) |
|||
| MANOVA test for overall group difference in quality of life measures: F=6.69; df=12, 143; p<.001 | ||||
| BMI | 31.5 (8.4) | 27.9 (5.4) | 3.6 (1.4, 5.8) | .001 |
| Number of psychotropics1 | 1.1 (1.4) | 0.3 (0.7) | 0.9 (0.6, 1.2) | <.0012 |
| Number of chronic conditions2 | 4.7 (2.9) | 2.8 (2.4) | 2.3 (1.6, 3.1) | <.001 |
| SF-36 Results | ||||
| Physical Functioning | 73.5 (27.0) | 84.3 (22.0) | −10.9 (−17.9, −3.9) | .002 |
| Physical Role | 69.9 (38.7) | 77.0 (35.6) | −6.3 (−17.9, −5.2) | .273 |
| Emotional Role | 61.3 (44.6) | 87.6 (26.8) | −25.8 (−37.4, −14.2) | <.001 |
| Vitality | 45.1 (20.1) | 60.9 (17.5) | −16.2 (−22.1, −10.4) | <.001 |
| Mental Health | 60.1 (20.9) | 78.4 (12.8) | −17.8 (−23.1, −12.4) | <.001 |
| Social Functioning | 68.1 (29.9) | 88.2 (18.7) | −20.3 (−27.9, −12.6) | <.001 |
| Bodily Pain | 66.1 (28.3) | 75.3 (23.8) | −9.7 (−17.7, −1.7) | .016 |
| General Health | 56.5 (21.4) | 73.0 (19.3) | −16.8 (−23.2, −10.3) | <.001 |
| Physical Summary3 | 67.2 (23.8) | 78.0 (21.2) | −10.6 (−17.5, −3.8) | .002 |
| Mental Summary3 | 59.0 (25.2) | 79.2 (14.2) | −20.7 (−27.0, −14.4) | <.001 |
Correlations between gambling severity (SOGS, NODS) and medical outcomes including SF-36 subscales were calculated (data not shown). This analysis was restricted to subjects with PG. As gambling severity increased, all measures of medical outcomes worsened, but not all correlations were significant. The most robust correlations were for SF-36 subscales Mental Summary (r=−0.363, −0.507), Social Functioning (r=−0.319, −0.487), and Emotional Role (r=−0.376, −0.398). SOGS total score was positively correlated with total number of chronic medical conditions (r=0.211, p=.040).
Discussion
The results show that persons with PG are at increased risk for chronic medical conditions, obesity, increased health service utilization, and poor lifestyle choices, such as avoiding exercise. They also have a worse perception of their health status. Our study adds to the literature by confirming these differences in a well characterized sample of persons with PG and in appropriate controls. Importantly, among pathological gamblers, increasing severity measured with the SOGS and NODS was positively correlated with worse self-reported health perceptions and total number of chronic medical conditions. This work shows that even within this narrowly defined subject group, we are able to demonstrate high risk for adverse medical consequences.
The findings raise questions about the strength and direction of these associations for which there are no ready explanations. Does PG contribute to poor health and obesity, or do persons who are obese and have worse health become pathological gamblers? Gambling could contribute to an inactive lifestyle because it is largely sedentary. The physical setting for gambling may itself contribute to poor health status, for example, sitting for long periods in a smoke filled casino, eating more (or less) frequently than normal, and being offered free or discounted alcoholic beverages. 28 These behaviors may be particularly harmful for subsets of pathological gamblers, such as obese persons or the elderly, groups at risk for diabetes, cardiopulmonary disease, or circulatory problems. For example, there is some evidence that the stress of casino gambling can contribute to cardiovascular disease. 29 On the other hand, it may be that persons who are obese or in poor health gamble because it is an easy form of recreation for them. It takes little physical ability or stamina to play slots or video poker, and for that reason these games are within reach to persons unable to participate in more physically demanding activities. A remote possibility is that persons with PG are simply more likely to over report medical and physical complaints, yet are otherwise physically healthy. There is some evidence that gamblers tend to somatize so this possibility cannot be entirely discounted, though it does not explain their obesity. 30,31 Another possibility is that some health-related symptoms could result from gambling withdrawal. Rosenthal and Lesieur 32 report that nearly two-thirds of persons with PG experience physical symptoms with abstinence such as headaches, insomnia, upset stomach, and physical weakness.
To be fair, most persons gamble responsibly and do not develop problematic gambling behavior. For these persons, gambling can serve as a community activity that brings people together and fosters social interactions. Desai et al. 33 found that elderly gamblers were healthier than comparable non-gamblers and had better social adjustment. Thus, not all persons who gamble report poor health and lifestyle choices, and a subset of gamblers may actually benefit from the recreational and social nature of gambling. Importantly, Desai et al. were not describing persons with problematic gambling behavior.
There are several important findings to highlight. First, pathological gamblers weigh more than controls and are more likely to be obese. Obesity has become epidemic in the United States,34 and has been associated with many poor health outcomes, including arthritis, diabetes, hypertension, gastroesophageal reflux, and sleep apnea. For that reason, obesity may help explain some of the poor health outcomes reported by pathological gamblers in our sample. The finding is also important because physicians have a responsibility to diagnose obesity and counsel their patients on healthy eating habits, exercise, and weight loss. The high rate of obesity in persons with PG could result from poor lifestyle choices that encourage increased food consumption and a sedentary lifestyle. Pathological gamblers are also more likely to make poorer health-related lifestyle choices, including smoking cigarettes, avoiding exercise, watching television ≥ 20 hours/week, and having a high caffeine intake. The fact that the gamblers did not report greater alcohol consumption may have to do with the wording of the questionnaire which asked about recent, not past, drinking behavior (“Do you currently drink any alcohol?). 10, 35
Pathological gamblers were more likely to use expensive medical treatments including emergency department visits, hospitalization, and use of computerized tomographic/magnetic resonance imaging scans. This pattern of use may reflect a lack of means to seek regular preventive care, or perhaps that pathological gamblers are underinsured. This is clearly an important yet underappreciated contributor to the overall economic and social cost of PG. Also, persons with PG are less likely to have regular dental care, perhaps because health insurance often does not cover dental care.
Self-reported medical health status and quality of life was impaired in persons with PG, as indicated by results from the SF-36 health survey. 36 Pathological gamblers reported impaired physical and emotional role functioning, but also acknowledged bodily pain, impaired social functioning, and low vitality. The findings are in accord with results from an earlier pilot study, 37 and from the report of Morasco et al. 14 based on findings from the NESARC sample. The fact that pathological gamblers are obese and report many medical and physical conditions may help explain the SF-36 findings. For example, the impairment in social functioning may stem from the fact that persons with PG are gambling rather than engaging in social activities with family or friends, while the fact that gamblers report low vitality indicates a more general problem with energy and pep. The bodily pain item indicates that the pathological gamblers have more pain complaints than controls that limit their physical activity. Pathological gamblers are much more likely to take psychotropic medications, and their usage is significantly greater than the controls for nearly all categories. This may be explained in part by the considerable psychiatric comorbidity found in persons with PG. 10
There are several methodologic limitations to acknowledge. First, persons with PG were mainly recruited through a study registry, advertising or participation in treatment programs, and not through epidemiologic sampling methods; thus, they may not be representative of persons with PG as a whole. Eleven were recruited from clinical settings (i.e., inpatient or outpatient), and while they were excluded from analyses involving psychotropic usage and medical service utilization, they may have skewed other results. Next, data were largely obtained through self-report, and it is possible that a subject’s perception of their medical health and level of physical activity is inaccurate because of denial or exaggeration. For example, people sometimes underestimate their weight, and overestimate their amount of exercise. Nonetheless, research generally supports the accuracy of self-reports of health and levels of exercise.38, 39 Pathological gamblers did not report a higher rate of alcohol use, but this was likely a quirk of our questionnaire which did not ask if subjects had a past drinking problem. However, there was some evidence of higher rates of problem drinking as women with PG reported they were more likely to have used alcohol while pregnant. Another methodological limitation is the large number of outcomes tested, resulting in a higher risk or a Type 1 error. Using MANOVA (for continuous outcomes) and an analogous score test for dichotomous outcomes, we grouped similar outcomes together and then tested the joint (multivariate) significance of differences between PG and controls. Because there were fewer multivariate tests and they all indicated highly significant differences between PG subjects and controls, we believe the grouped outcome results can be generalized to other settings.
Acknowledgements
The study was funded through a grant RO1DA021361 from the National Institute on Drug Abuse (Dr. Black). Dr. Black has received research support from AstraZeneca and Psyadon. He receives royalties from American Psychiatric Publishing, Inc. and Oxford University Press. Ms. Shaw, Mr. McCormick, and Dr. Allen report no conflicts. We are grateful to Jo Ann Franklin and Rebecca Hansel for data collection and data entry.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Kessler RC, Hwang I, LaBrie R, et al. DSM-IV pathological gambling in the National Comorbidity Survey Replication. Psych Medicine. 38:1351–1360. doi: 10.1017/S0033291708002900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Petry NM, Stinson FS, Grant BF. Comorbidity of DSM-IV pathological gambling and other psychiatric disorders: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry. 2005;66:564–574. doi: 10.4088/jcp.v66n0504. [DOI] [PubMed] [Google Scholar]
- 3.Grinols EL. Gambling in America: Costs and Benefits. New York: Cambridge University Press; 2004. [Google Scholar]
- 4.Shaw M, Forbush K, Schlinder J, et al. The effect of pathological gambling on families, marriages, and children. CNS Spectrums. 2007;12:615–622. doi: 10.1017/s1092852900021416. [DOI] [PubMed] [Google Scholar]
- 5.Petry NM, Kiluk BD. Suicidal ideation and suicide attempts in treatment-seeking pathological gamblers. J Nerv Ment Dis. 2002;190:462–469. doi: 10.1097/01.NMD.0000022447.27689.96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.National Opinion Research Center at the University of Chicago (NORC) Gambling Impact and Behavior Study, Report to the National Gambling Impact Study Commission. 1999 Apr 1;
- 7.Lesieur HR, Custer RL. Pathological gambling: Roots, phases, and treatment. Ann Am Acad Political Soc Sci. 1984;474:147–156. [Google Scholar]
- 8.Welte JW, Barnes M, Wieczorek WF, Tidwell GM, Parker J. Gambling participation in the US – Results from a national survey. J Gambling Stud. 2002;18:313–337. doi: 10.1023/a:1021019915591. [DOI] [PubMed] [Google Scholar]
- 9.Volberg RA. Prevalence studies of problem gambling in the United States. J Gambling Stud. 1996;12:111–128. doi: 10.1007/BF01539169. [DOI] [PubMed] [Google Scholar]
- 10.Argo T, Black DW. The characteristics of pathological gambling. In: Grant J, Potenza M, editors. Understanding and Treating Pathological Gambling. Washington DC: American Psychiatric Publishing Inc.; 2004. pp. 39–53. [Google Scholar]
- 11.Lorenz VC, Yaffee RA. Pathological gambling: Psychosomatic, emotional and marital difficulties as reported by the gambler. J Gambl Behav. 1986;2:40–49. [Google Scholar]
- 12.Lorenz VC, Yaffee RA. Pathological gambling: Psychosomatic, emotional and marital difficulties as reported by the spouse. J Gambl Behav. 1988;4:13–26. [Google Scholar]
- 13.Bergh C, Kulhorn E. Social, psychological, and physical consequences of pathological gambling in Sweden. J Gambl Stud. 1994;10:275–285. doi: 10.1007/BF02104968. [DOI] [PubMed] [Google Scholar]
- 14.Morasco BJ, Pietrzak RH, Blanco C, Grant BF, Hasin D, Petry NM. Health problems and medical utilization associated with gambling disorders: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Psychosom Med. 2006;68:976–984. doi: 10.1097/01.psy.0000238466.76172.cd. [DOI] [PubMed] [Google Scholar]
- 15.Pasternak AV, Fleming MF. Prevalence of gambling disorders in a primary care setting. Arch Fam Med. 1999;8:515–520. doi: 10.1001/archfami.8.6.515. [DOI] [PubMed] [Google Scholar]
- 16.Mosrasco BJ, vom Eigen KA, Petry NM. Severity of gambling is associated with physical and emotional health in primary care patients. Psychiatry and Primary Care. 2006;28:94–100. doi: 10.1016/j.genhosppsych.2005.09.004. [DOI] [PubMed] [Google Scholar]
- 17.Erickson L, Molina CA, Ladd GT, Pietrzak RH, Petry NM. Problems and pathological gambling are associated with poorer mental and physical health in older adult. Int J Geriatr Psychiatry. 2005;20:754–759. doi: 10.1002/gps.1357. [DOI] [PubMed] [Google Scholar]
- 18.Morasco BJ, Petry NM. Gambling problems and health functioning in adults receiving disability. Disability and Rehabilitation. 2006;28:619–623. doi: 10.1080/09638280500242507. [DOI] [PubMed] [Google Scholar]
- 19.Pietrzak RH, Petry NM. Gambling severity and health functioning in adolescents recruited from urban primary care settings. J Adolesc Health. 2006;39:764–766. doi: 10.1016/j.jadohealth.2006.04.005. [DOI] [PubMed] [Google Scholar]
- 20.Pietrzak RH, Molina CA, Ladd GT, Kerins GJ, Petry NM. Heath and psychosocial correlated of disordered gambling in older adults. Am J Geri Psychiatry. 2005;13:510–519. doi: 10.1176/appi.ajgp.13.6.510. [DOI] [PubMed] [Google Scholar]
- 21.Weinstock J, Blanco C, Petry NM. Health correlates of pathological gambling in a methadone clinic. Exp Clin Psychopharmacol. 2006;14:87–93. doi: 10.1037/1064-1297.14.1.87. [DOI] [PubMed] [Google Scholar]
- 22.Lesieur HR, Blume SB. The South Oaks Gambling Screen (SOGS): a new instrument for identification of pathological gamblers. Am J Psychiatry. 1987;144:1184–1188. doi: 10.1176/ajp.144.9.1184. [DOI] [PubMed] [Google Scholar]
- 23.American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition Text Revision. Washington, DC: American Psychiatric Press; 2000. [Google Scholar]
- 24.Frankenburg FR, Zanarini MC. The association between borderline personality disorder and chronic medical illnesses, poor health-related lifestyle choices, and costly forms of health care utilization. J Clin Psychiatry. 2004;65:1660–1665. doi: 10.4088/jcp.v65n1211. [DOI] [PubMed] [Google Scholar]
- 25.Ware JE, Snow KK, Kosinski M, Gandek B. SF-36 Health Survey Manual and Interpretation Guide. Boston: New England Medical Center Health Institute; 1993. [Google Scholar]
- 26.SAS Institute Inc., SAS/STAT. 9.1 Users Guide. Cary, NC: SAS Institute, Inc.; 2004. [Google Scholar]
- 27.Agresti A, Klingenberg B. Multivariate tests comparing binomial probabilities, with application to safety studies for drugs. J Roy Stat Soc C-App. 2005;54:691–706. [Google Scholar]
- 28.Desai RJ. Older adults. In: Grant JE, Potenza M, editors. Pathological Gambling: A Clinical Guide to Treatment. Washington, DC: American Psychiatric Publishing, Inc.; 2004. pp. 83–96. [Google Scholar]
- 29.Meyer G, Hauffa BP, Schedlowski M, Pawlak C, Stadler MA, Exton MS. Casino gabmling increases heart rate and salivary cortisone in regular gamblers. Biol Psychiatry. 2000;48:948–953. doi: 10.1016/s0006-3223(00)00888-x. [DOI] [PubMed] [Google Scholar]
- 30.Blaszczynski AP, McConaghy N. SCL-90 assessed psychopathology in pathological gamblers. Psychol Rep. 1988;62:547–552. doi: 10.2466/pr0.1988.62.2.547. [DOI] [PubMed] [Google Scholar]
- 31.Petry NM. Psychiatric symptoms in problem gamblers and non-problem gambling substance abusers. Am J Addict. 2000;9:163–171. doi: 10.1080/10550490050173235. [DOI] [PubMed] [Google Scholar]
- 32.Rosenthal R, Lesieur H. Self reported withdrawal symptoms and pathological gambling. Am J Addict. 1992;102:13–18. [Google Scholar]
- 33.Desai RJ, Maciejewski PK, Dausey DJ, Caldarone BJ, Potenza MN. Health correlates of recreational gambling in older adults. Am J Psychiatry. 2004;161:1672–1679. doi: 10.1176/appi.ajp.161.9.1672. [DOI] [PubMed] [Google Scholar]
- 34.Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999–2008. JAMA. 2010;303:235–241. doi: 10.1001/jama.2009.2014. [DOI] [PubMed] [Google Scholar]
- 35.Grant JE, Kim SW, Odlaug BL, Potenza MN. Daily tobacco smoking in treatment-seeking pathological gamblers: Clinical correlates and co-occurring psychiatric disorders. J Addict Med. 2008;2:178–184. doi: 10.1097/ADM.0b013e3181878673. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.McHorney CA, Ware JE, Raczek AK. The MOS 36-Item Short Form Health Survey II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care. 1993;31:247–263. doi: 10.1097/00005650-199303000-00006. [DOI] [PubMed] [Google Scholar]
- 37.Black DW, Moyer T, Schlosser S. Quality of life and family history in pathological gambling. J Nerv Ment Dis. 2003;191:124–126. doi: 10.1097/01.NMD.0000050942.86352.47. [DOI] [PubMed] [Google Scholar]
- 38.Carnethon MR, Gidding SS, Nehgme R, Sidney S, Jacobs DR, Jr, Liu K. Cardiorespiratory fitness in young adulthood and the development of cardiovascular disease risk factors. JAMA. 2003;290:3092–3100. doi: 10.1001/jama.290.23.3092. [DOI] [PubMed] [Google Scholar]
- 39.Schmidt NB, Telch MJ. Nonpsychiatric medical comorbidity, health perceptions, and treatment outcome in patients with panic disorder. Health Psychol. 1997;16:114–122. doi: 10.1037//0278-6133.16.2.114. [DOI] [PubMed] [Google Scholar]