Associations Between Gambling and Cognitive Function among U.S. Chinese Older Adults

Abstract

This study aims to examine the associations between gambling participation, problem gambling, and cognition. Data were derived from the PINE study. Gambling participation was measured by Modified South Oaks Gambling Screen. Problem gambling was assessed with Problem Gambling Severity Index. Cognition was measured by East Boston Memory Test, Digit Span Backward, Symbol Digit Modalities Test, and Mini-Mental State Examination. Of all participants, 41.940% were male. Higher gambling participation was significantly associated with higher global cognition (b = .008, SE = .003, p < .05), executive function (b = .118, SE = .049, p < .05), and episodic memory (b = .009, SE = .004, p < .05). Higher problem gambling was significantly related to lower Mini-Mental State Examination (b = −.105, SE = .031, p < .001). Guiding older adults’ gambling behaviors and intervening in problem gambling timely might be crucial for their cognitive function.

Introduction

Cognitive impairment features for recession in memory, thinking, attention, comprehension, calculation, learning capacity, language, and judgment (World Health Organization, 2015). Cognitive function is considered one important aspect of successful aging (Baltes & Baltes, 1990). It includes various dimensions including executive function, episodic memory, and working memory. Specifically, executive function is a wide range of cognitive processes, such as the novel problem solving, evaluating the intermediate outcome, and modifying the plan as needed (Carpenter et al., 2000). Episodic memory is a neurocognitive system that enables individuals to recollect the past experiences (Tulving, 2002). Working memory refers to the temporary storage and manipulation of new information (Baddeley & Hitch, 1974).

Cognitive performance appears to decline after reaching the peak in early adulthood (Lyketsos et al., 1999). Compared with younger adults, older adults might be at higher risks of experiencing varying levels of cognitive function decline (Li & Hsu, 2015; Reuter-Lorenz & Park, 2014) as the features commonly associated with transition to older adulthood may expose them to increased vulnerability to illness (Baltes & Baltes, 1990). Recent years have witnessed a rapidly expanding literature documenting correlates of cognitive function. Both protective factors and risk factors for cognitive function were identified in literature. Higher income and education were associated with better cognitive function (Crowe et al., 2013; Lyketsos et al., 1999; Lyu & Burr, 2016; Stern, 2009). Unmarried female older adults reported worse cognitive function compared with their married and male counterparts (Hakansson et al., 2009; Oksuzyan et al., 2018). Despite increasing number of empirical studies, the associations between gambling behaviors (e.g., gambling participation and problem gambling) and cognitive function have rarely been discussed.

Gambling Participation and Health Outcomes

Gambling participation is mainly about the engagement of older adults in various kinds of gambling activities (Chiesi et al., 2013). Gambling industry, as one of the major recreational industries in the United States, experienced remarkable growth in recent decades (Horváth & Paap, 2012). The proliferation of gambling industry fuels growth of gambling participation (Welte et al., 2015). Older adults are regarded as potential consumers by gambling industry as they have substantial time for entertainment (Bilt et al., 2004; Desai et al., 2004; van der Maas et al., 2017a).

Increasing number of older adults spend their leisure time in gambling activities in the United States (McNeilly & Burke, 2002). Older adults are more likely to take gambling as a leisure activity that help them to escape from boring time after retirement as it provides an opportunity to socialize, experience excitement, and win money (Pattinson & Parke, 2017; van der Maas et al., 2017b). Specifically, gambling as a social activity brings people together, makes them well connected with each other, and improves individuals’ well-being through increasing their opportunities of building social networks and enhancing social integration (Alberghetti & Collins, 2015; Bilt et al., 2004). In addition, certain kinds of gambling activities are associated with improved coping strategies by strengthening older adults’ skills through game tactics, mathematical calculation, concentration, and hand-eye cooperation (Shaffer & Korn, 2002), which may influence their cognitive function. In a study examining perceptions about aging across various racial/ethnic groups in the United States, Laditka and associates (2009) found that except African Americans and Vietnamese, older adults in the other racial/ethnic groups (Caucasian, American Indian, Chinese, and Hispanic) considered that taking part in cognitive activities, such as gambling, was helpful to keep the brain active.

Despite these possible positive effects of gambling on cognitive function, there is a paucity of empirical research regarding the relationship between gambling participation and cognitive function. Existing studies mainly focused on physical and mental health outcomes of gambling participation (Desai et al., 2004; Kerber et al., 2015). The evidence is mixed regarding the relationships between gambling participation and physical and mental health (Chen & Dong, 2015; Desai et al., 2004; Okunna et al., 2016; Pietrzak et al., 2007). Specifically, Chen and Dong (2015) found that older adults who gambled had better health status than the non-gamblers. Desai et al. (2004) also found that recreational gamblers reported better health status than non-gamblers. The negative effects of gambling on mental and physical health were also reported in previous studies. Okunna et al. (2016) found that recreational gambling was significantly associated with poor mental health. In the study among U.S. older adults, Pietrzak et al. (2007) found that recreational gambler had significantly elevated rates of anxiety and obesity. The inconsistencies might be partially explained by the fact that different measurements were used, and the research was conducted among various populations.

Although previous studies indicated a possible link between gambling participation and cognitive function, comparable empirical work lags behind. To the best of our knowledge, there exists only one such study which examined the association between gambling participation and cognitive function. Bilt and colleagues (2004) found that higher levels of gambling were significantly associated with better cognitive function. The pioneering work contributes to this topic. However, in that study, gambling participation was measured by a single question of whether the participants left their home to gamble or play bingo which may not capture various gambling behaviors among older adults. Given the importance of this topic, a test of this issue awaits more empirical research.

Problem Gambling and Health Outcomes

Problem gambling is defined as the gambling behavior which causes disruptions in major aspects of life (National Council on Problem Gambling, 2014). The extended leisure time after retirement puts older adults at a high risk of experiencing problem gambling (Subramaniam et al., 2015).

Problem gambling among older adults is associated with physical and mental health deterioration, as well as financial and social problems (Medeiros et al., 2017; Pietrzak et al., 2005, 2007; Tse et al., 2012). A systematic review study indicated that gambling disorder was associated with negative health outcomes (Tse et al., 2012). A study among old adults from senior centers and outpatient medical clinics in America showed that older adults with gambling disorders had poorer health status and more chronic illnesses in comparison with a sample without disordered gambling behaviors (Pietrzak et al., 2005). Previous studies also suggested that older problem gamblers reported more mental health problems, such as depression and suicidality (Pietrzak & Petry, 2006; Potenza et al., 2006) and anxiety (Pietrzak & Petry, 2006). Moreover, due to their fixed income and limited savings after retirement, older adults have more difficulties in recovering from the worse health consequences of problem gambling (Fong, 2005; Subramaniam et al., 2015; Van der Maas et al., 2017). With increasing availability of gambling, the potential destructive effects of problem gambling raised public health concern (Medeiros et al., 2015).

Limited empirical research on the relationship between gambling and cognitive function is available, especially regarding the U.S. Chinese older adults. Asian Americans are the fastest growing population in the United States, and Chinese Americans comprise the largest component of Asian American population (U.S. Census Bureau, 2016). Although the number of Chinese Americans has increased rapidly, their health needs for cognitive interventions are not given enough attention by the mainstream society (Li et al., 2017). As reported in previous research, older Chinese Americans are at a high risk for delayed diagnosis of dementia (Woo, 2017). While engaging in cognitive activities, such as gambling, was considered a protective factor for cognitive function (Laditka et al., 2009), no empirical study was conducted to examine such relationship among U.S. Chinese older adults. Moreover, the relationship between problem gambling and cognitive function is still unknown. Therefore, an improved understanding of cognitive function correlates of gambling has significant research and health policy implications. To fill this research gap, this study aims to examine cognitive function in relation to gambling participation and problem gambling among U.S. Chinese older adults. Specifically, this study aims to test: (1) the associations between gambling participation and cognitive function; and (2) the associations between problem gambling and cognitive function.

Methods

Population and Settings

Data used in this study were derived from the Population Study of Chinese Elderly in Chicago (PINE). The PINE study is in response to limited data on U.S. Chinese older adults’ health and well-being. Following a community-based participatory research (CBPR) approach, the PINE study recruited U.S. Chinese older adults in the Greater Chicago area to complete the survey (Dong, 2014). This study embodied a close collaboration among social service agencies, community centers, and research organizations. Out of 3542 eligible older adults, 3157 agreed to participate in the research with a response rate of 91.9%. All participants were interviewed by well-trained research assistants. They interviewed participants with English, Cantonese, mandarin, or other Chinese dialects based on the interviewee’s language preference. The PINE study was approved by the Institutional Review Board of the Rush University Medical Center. The written informed consent was obtained from all participants.

Measurements

Cognitive Function.

The cognitive measures included global cognitive function, executive function, episodic memory, working memory, and Mini-Mental State Examination (MMSE). Executive function was assessed with Symbol Digit Modalities Test (SDMT) ranging from 0 to 80. Episodic memory was the averaged z-score of East Boston Memory Test-Immediate Recall (EBMT) and the East Boston Memory Test-Delayed Recall (EBDR) (Albert et al., 1991). Working memory was measured using Digit Span Backward (DB) (Wechsler, 1987). Raw data of the five indicators including MMSE, SDMT, EBMT, EBDR, and DB were transformed to z-score and then averaged to represent global cognitive function with higher scores indicating better global cognition. The composite scores are widely used in previous studies on cognitive function (Bretsky et al., 2003; Weuve et al., 2005; Zhang et al., 2009).

Gambling Participation

Gambling participation was measured using 5-item scale Modified South Oaks Gambling Screen (MSOGS) which included type and frequency of gambling. MSOGS was derived from South Oaks Gambling Screen (SOGS) (Lesieur & Blume, 1987). It focused on gambling participation (Chiesi et al., 2013). Participants reported frequency of participation in the following five different types of activities: (1) purchasing state lottery tickets; (2) betting money on mah-jong or card games; (3) betting money on sports games; (4) going to the casino; and (5) playing slot machines or betting money on video poker. Response options ranged from 1 (once per 2 years) to 6 (once per week). We calculated scores by summing the items. MSOGS ranged from 0 to 30 with higher scores reflecting higher levels of gambling participation.

Problem Gambling

Problem gambling was measured with the Problem Gambling Severity Index (PGSI). It focused on emotional and psychological correlates of pathological gambling and economic and social problems directly associated with gambling (Ferris & Wynne, 2001).

Participants were asked to answer a series of questions regarding the frequency of taking part in the activities, such as betting more than one could really afford to lose and gambling with larger amounts of money to get the same feeling of excitement. Each item was given a 3-point scale response ranging from 0 (never) to 3 (almost always). For the purpose of this study, we used the continuous PGSI score which has also been used in previous studies (Elton-Marshall et al., 2018; MacLaren et al., 2012) . The PGSI ranged from 0 to 27 with higher scores indicating severer gambling problem.

Control Variables

Several socio-demographic and other factors that could possibly account for the association between gambling and cognitive function were controlled for. Specifically, the confounding factors included age, gender, education, income, marital status, number of children, years in the United States, and physical function. Age was a continuous variable. We operationalized gender as a dichotomous measure (1 = female and 0 = male). Education and income were treated as categorical variables. Number of children alive and years in the United States were continuous variables. Marital status was captured by a dummy variable contrasting married with other marital status. Physical function was measured by Katz Index of Activities of Daily Living (ADL) with higher scores indicating poorer physical function (Katz et al., 1970).

Statistical Analysis

We fitted multivariate linear regression models and quantile regression models to test the relationships between gambling participation, problem gambling, and cognitive function outcomes. More specifically, multivariate linear regression analyses were used to examine the associations of gambling participation and problem gambling with global cognitive function, executive function, episodic memory, and working memory. In addition, quantile regression was employed to examine the associations between gambling participation, problem gambling, and MMSE. Unstandardized coefficients were presented, indicating the magnitude of change in cognitive function with every 1-point increase in gambling participation or problem gambling.

The relationships between gambling participation, problem gambling, and cognitive function outcomes were tested after controlling for demographic variables, socio-economic status, immigration related variables, family related variables, and functional status. Listwise deletion was used to address missing data. All analyses were performed using SAS, version 9.2 (SAS Institute Inc., Cary, USA).

Results

Table 1 presents the demographic statistics and cognitive function outcomes of older adults with/without gambling in the past 12 months. Mean age of the participants was 72.815 years. Of all the participants, 58.060% were female and 41.940% were male. Non-gamblers reported significantly higher levels of education (Mean= 8.900 vs Mean = 7.820, p < .001), lower income (Mean = 1.920 vs Mean = 2.108, p < .001), less children (Mean = 2.838 vs Mean = 3.056, p < .01), less years in the United States (Mean = 19.142 vs Mean = 25.077, p < .001), and poorer physical function (Mean = .223 vs Mean = .069, p < .001) compared with gamblers.

Table 1.

Characteristics and Cognitive Function Outcomes Between Gamblers and Non-Gamblers.

Variable	All Participants		Gamblers	Non-gamblers	P
	Mean/n (%)	SD	Mean/n (%)	Mean/n (%)
Age	72.815	8.30	72.220	72.881	.112
Sex
Female	1823 (58.06)	—	236 (7.52)	1587 (50.54)	<.001
Male	1317 (41.94)	—	231(7.36)	1086 (34.58)
Education	8.72	5.05	7.820	8.900	<.001
Income	1.959	1.190	2.108	1.920	<.001
Marital status
Married	2223 (70.93)	—	326 (10.40)	1897 (60.53)	.619
Not married	911 (29.07)	—	140 (4.47)	771 (24.60)
Number of children alive	2.874	1.506	3.056	2.838	.003
Years in the United States	20.020	13.183	25.077	19.142	<.001
Activities of daily living	.200	.916	.069	.223	<.001
Global cognitive function	−.038	.821	−.021	−.038	<.001
Executive function	29.590	12.121	29.151	29.704	<.001
Episodic memory	−.044	.982	−.040	−.041	.002
Working memory	5.025	2.397	5.013	5.031	<.001
Mini-mental state examination	25.300	4.686	25.885	25.197	.086

Note. Standard deviation was abbreviated to SD.

We also applied t-test for global cognitive function, executive function, episodic memory, and working memory and Wilcoxon two-sample test for MMSE to examine the difference between gamblers and non-gamblers. Among the participants, 467 (14.873%) older adults reported gambling behaviors in the past 12 months. The non-gamblers reported significantly lower global cognitive function (Mean = −.038 vs Mean = −.021, p < .001) and episodic memory (Mean = −.041 vs Mean = −.040, p < .01) than gamblers. Non-gamblers reported significantly higher executive function (Mean = 29.704 vs Mean = 29.151, p < .001) and working memory (Mean = 5.031 vs Mean = 5.013, p < .001) than gamblers.

Table 2 and Table 3 show the tolerance and variance inflation factor (VIF) of the models in this study. These two indicators examined the collinearity of the models. The tolerance indicates the percent of variance in the predictor that cannot be explained by the other predictors. Values of tolerance less than 0.10 may need to be further studied (Nia et al., 2017). The VIF refers to the magnitude of variance of the coefficient estimate inflated by multicollinearity (Nia et al., 2017). If VIF is above 10, it means that one of the independent variables has associated with other variables (O’brien, 2007). As shown in Tables 2 and 3, the tolerance values for all of the independent variables were larger than 0.10, and VIF values for all the independent variables were below 10, suggesting multicollinearity is not a problem in the regressions.

Table 2.

Tolerance and Variance Inflation Factor of the Models Regarding the Association between Gambling Participation and Cognitive Function.

Variables	Tolerance	VIF
Age	.671	1.491
Female	.843	1.186
Education	.788	1.269
Income	.849	1.178
Married	.757	1.322
Number of children	.773	1.294
Years in the United States	.731	1.367
Activities of daily living	.927	1.079
Gambling participation	.937	1.067

Note. VIF = Variance Inflation Factor.

Table 3.

Tolerance and Variance Inflation Factor of the Models Regarding the Association between Problem Gambling and Cognitive Function.

Variables	Tolerance	VIF
Age	.697	1.434
Female	.854	1.170
Education	.795	1.258
Income	.847	1.181
Married	.773	1.294
Number of children	.789	1.267
Years in the United States	.756	1.323
Activities of daily living	.934	1.071
Problem gambling	.979	1.022

Note. VIF = Variance Inflation Factor.

The results of associations between gambling participation and cognitive function outcomes are presented in Table 4. In Model A, gambling participation was significantly associated with higher global cognitive function after controlling for potentially confounding variables. Every 1-point increase in gambling participation was associated with .008 units increase in global cognitive function (b = .008, SE = .003, p < .01). Age was negatively associated with global cognitive function (b = −.024, SE = .002, p < .001). Compared with male, female older adults reported lower score in global cognitive function (b = −.054, SE = .024, p < .05). Education (b = .079, SE = .002, p < .001) and income (b = .032, SE = .010, p < .005) and were positively associated with global cognitive function. Higher ADL scores were significantly associated with lower score in global cognitive function (b = −.181, SE = .013, p < .001). In Model B, gambling participation was significantly associated with higher executive function (b = .118, SE = .049, p < .05). In Model C, gambling participation was significantly associated with better episodic memory. Every 1-point increase in gambling participation was associated with .009 units increase in episodic memory (b = .009, SE = .004, p < .05). Gambling participation, together with the covariates, accounted for 47.1% of variance in global cognitive function, 42.3% in executive function, and 30.1% in episodic memory, respectively.

Table 4.

Gambling Participation and Cognitive Function Outcomes.

	Model A	Model B	Model C	Model D	Model E
	Global Cognitive Function	Executive Function	Episodic Memory	Working Memory	Mini-Mental State Examination
	b	b	b	B	b
Age Female Education	−.024*** (.002) −.054* (.024) .079*** (.002)	−.355*** (.027) −.416 (.400) 1.101*** (.041)	−.027*** (.002) .039 (.032) .077*** (.003)	−.037*** (.005) −.436*** (.080) .201*** (.008)	−.096*** (.010) −.365** (.132) .304*** (.013)
Income	.032** (.010)	.677*** (.168)	.032* (.014)	.068* (.034)	.073 (.042)
Married	.036 (.027)	.397 (.471)	.029 (.038)	.043 (.092)	.716*** (.163)
Number of children Years in the United States Activities of daily living Gambling participation	−.039*** (.008) .001 (.001) −.181*** (.013) .008** (.003)	−.854*** (.139) −.008 (.016) −1.991*** (.214) .118* (.049)	−.037** (.011) .001 (.001) −.146*** (.018) .009* (.004)	−.088** (.027) −.000 (.003) −.326*** (.043) .007 (.010)	−.172** (.056) .019*** (.006) −1.350*** (.209) .015 (.017)
Adjusted R2	.471	.423	.301	.294
F	299.900	204.650	145.67	141.980
Pr > F	<.0001	<.0001	<.0001	<.0001

Note. Standard errors are in the parentheses

^*p < .05

^**p < .01

^***p < .001.

Table 5 reveals the relationships between problem gambling and cognitive domains holding covariates constant. The associations between problem gambling and global cognitive function, executive function, episodic memory, and working memory were examined in Models A–D, respectively. However, these associations were not significant. Model E presented the significant negative association between problem gambling and MMSE. Every 1-point increase in problem gambling was associated with .105 units decrease in MMSE (b = −.105, SE = .031, p < .001). Stepwise regression models and standardized coefficients are shown in the Supplementary Appendixes A–F.

Table 5.

Problem Gambling and Cognitive Function Domains.

	Model A	Model B	Model C	Model D	Model E
	Global CognitiveFunction	Executive Function	Episodic Memory	Working Memory	Mini-Mental State Examination
Age Female Education	−.025*** (.002) −.062** (.024) .079*** (.002)	−.366*** (.027) −.592 (.400) 1.092*** (.041)	−.027*** (.002) .032 (.032) .077*** (.003)	−.038*** (.005) −.445*** (.080) .201*** (.008)	−.099*** (.010) −.406** (.128) .305*** (.014)
Income	.031** (.010)	.664*** (.169)	.032* (.014)	.068* (.034)	.063 (.034)
Married	.034 (.027)	.328 (.472)	.029 (.038)	.042 (.093)	.675*** (.166)
Number of children Years in the United States Activities of daily living Problem gambling Adjusted R2	−.039*** (.008) .002 (.001) −.183*** (.013) −.006 (.014) .469	−.845*** (.139) .002 (.016) −2.031*** (.214) −.303 (.221) .421	−.036** (.011) .001 (.001) −.148*** (.018) .009 (.020) .299	−.087** (.028) −.001 (.003) −.328*** (.043) −.018 (.048) .293	−.177** (.055) .021*** (.006) −1.353*** (.207) −.105*** (.031)
F	302.830	207.750	147.440	299.000
Pr > F	<.0001	<.0001	<.0001	<.0001

Note. Standard errors are in the parentheses

^*p < .05

^**p < .01

^***p < .001.

Discussion

Our study investigated the relationships between gambling participation, problem gambling, and cognitive outcomes among U.S. Chinese older adults. Older individuals with higher levels of gambling participation had better cognitive function outcomes regarding global cognitive function, executive function, and episodic memory, whereas gambling participation was not significantly associated with working memory and MMSE. Additionally, we also examined the associations between problem gambling and cognitive function outcomes. We found that higher levels of problem gambling were significantly associated with lower scores of MMSE.

In this study, 14.873% of the participants reported the experience of gambling in the past 12 months. The prevalence rate was lower than that reported in a previous study (50.3%) among a representative sample of older adults in the United States (Desai et al., 2004). This lower gambling participation rate may be explained by the fact that many Chinese older adults are inactively engaged in traditional gambling activities (e.g., bingo) due to cultural barriers (Chen & Dong, 2015). Problem gambling was also lower in this study. In this sample, there were only 29 low-risk gamblers, 29 moderate risk gamblers, and seven problem gamblers which prohibited us to examine different subgroups of problem gamblers in the regression analysis. This finding was consistent with previous research which reported that the prevalence of problem gambling is lower among Asians compared with other minorities (Welte et al., 2015). According to a longitudinal study in the United States, problem gambling rates were lowest for whites and Asians and highest for blacks and Hispanics (Welte et al., 2015). Another empirical study also showed that Asian respondents comprised an unobservable portion of problem gamblers among Massachusetts adults (Mazar et al., 2018).

This study revealed that U.S. Chinese older adults with higher levels of gambling participation had higher scores in global cognitive function, executive function, and episodic memory. Previous research among rural older adults with low socio-economic status in Pennsylvania showed that the association between gambling participation and MMSE was not significant (Bilt et al., 2004). Compared with the native Americans, U.S. Chinese older adults, as an immigrant population, have restricted social networks (Dong & Chang, 2017). Gambling participation provided them with the opportunity to develop their social networks. As previous studies revealed, gambling participation acts as a way of social engagement and social participation which helps individuals to get involved into social networks and be well connected with each other (Alberghetti & Collins, 2015; Bilt et al., 2004; Shaffer & Korn, 2002). Using the same sample of this study, Li and Dong (2018) found that social networks were essential to maintain cognitive function for U.S. Chinese older adults. In addition, these positive associations between gambling participation and cognitive function outcomes may be partially explained by the fact that gambling participation, as a cognitive stimulation activity, helps individuals to slow down the process of cognitive function deterioration (Cabrera et al., 2020). Gambling participation may prevent older adults’ cognitive decline by improving their skills in memory enhancement and problem solving by exposing them to remembering and calculating numbers and making decisions (Shaffer & Korn, 2002).

In contrast to gambling participation, our findings showed that older adults with higher levels of problem gambling reported significantly lower scores in MMSE. The relationship between problem gambling and cognitive function outcomes was understudied in literature. The cognitive deterioration among U.S. Chinese problem gamblers might be attributable, in part, to the fact that problem gambling is an additive behavior which impairs individuals’ physical health (Pietrzak et al., 2005). Older adults with declined physical health might be more likely to stay at home and their opportunities to socialize with people decrease, resulting into the poor function. It could also possibly be explained by the fact that excessive gambling may contribute to the sedentary and unhealthy lifestyle (Pietrzak et al., 2005) which further leads to cognitive deterioration. Future study could test the mechanism through which problem gambling is associated with cognitive function.

This study should be interpreted with caution. We could not separate age/cohort effects and infer casual relationships between gambling behaviors and cognitive function as the data set is not longitudinal. Furthermore, data of older adults in this study were representative of older Chinese Americans in the Greater Chicago area (Simon et al., 2014), and it is not clear whether different associations could be found for other ethnic populations or Chinese older adults in other places. Notably, culture, values, and beliefs affect the gambling patterns and individuals’ responses to risk gambling behaviors (Raylu & Oei, 2004), which may make the results not generalizable. In addition, we only tested the associations between gambling participation, problem gambling, and cognitive function, yet the underlying mechanisms linking gambling participation and problem gambling with cognitive function remain largely unexamined. A test of the underlying mechanism awaits future research.

Despite these limitations, this study has important implications for social services regarding U.S. Chinese older adults. In mental health services, social service agencies could educate older adults through well-designed programs regarding gambling participation to help U.S. Chinese older adults to develop appropriate attitudes and motivations toward gambling. In addition, social service agencies could develop screening tools to identify early warning signs of problem gambling to prevent the recreational gambling participation from evolving into serious problem gambling behaviors among U.S. Chinese older adults. Furthermore, strategies improving the awareness of family practitioners might provide the precious opportunity for surveillance and more targeting preventive interventions. This study may also have clinical implications as it provided the empirical evidence regarding the associations between gambling behaviors and cognitive domains which makes the future clinical intervention more targeting. When the problem gambling is identified, cognitive behavioral therapy, cognitive therapy, and motivational interventions could be provided based on the evaluation of professionals from these specific areas.

Conclusion

The relationships between gambling behaviors and health status have been frequently studied in previous studies (Chen & Dong, 2015; Desai et al., 2004), whereas the gambling-cognition relationship is rarely investigated. This study reveals that higher levels of gambling participation were associated with higher global cognitive function, executive function, and episodic memory. However, higher levels of problem gambling were associated with lower scores in MMSE among U.S. Chinese older adults. Future studies could focus on the underlying mechanism between gambling behaviors and cognitive function among older adults. There is also an urgentneed to collect longitudinal data on the gambling behaviors and cognitive function that would enable us to explore the causal relationship between gambling behaviors and cognitive function.

Author Biographies

Nannan Zhang, PhD, is a postdoctoral fellow at the East China University of Science and Technology. Her research interests include cognitive impairment in older age, lifestyle and health, and community environment and health.

Mengting Li, PhD, is an Assistant Professor of Nursing at Rutgers Institute for Health, Health Care Policy and Aging Research. Her research focuses on intergenerational relationships, elder mistreatment, resilience, and cognitive function of older adults.

XinQi Dong, MD, MPH, is the Director of the Institute for Health, Health Care Policy, and Aging Research at Rutgers University as well as the inaugural Henry Rutgers Distinguished Professor of Population Health Sciences. His research interests include violence prevention, elder abuse, and population health.