Abstract and Introduction
Background and Objectives: Evidence relates moderate alcohol consumption in later life to fewer depressive symptoms and greater functional ability. This study evaluated social interaction as a mediator of these outcomes.
Research Design and Methods: Data included older adults in the Health and Retirement Study. In Study 1, cross-sectional mediation analyses evaluated social interaction as a mechanism through which moderate alcohol use was related to decreased depressive symptomatology (Model 1; N = 2,294) and less functional limitation (Model 2; N = 2,357). In Study 2, longitudinal cross-lagged path analyses further evaluated the impact of moderate alcohol use and social interaction on future depressive symptoms and functional limitation.
Results: In Study 1, results from Model 1 indicated that moderate alcohol use was associated with decreased depressive symptomatology indirectly via greater social interaction. In Model 2, moderate alcohol use was associated with less functional limitation indirectly via greater social interaction. In Study 2, significant indirect effects corroborated findings from Study 1. In Model 3, moderate alcohol use in 2012 inversely predicted depressive symptomatology in 2014 via greater levels of social interaction in 2012. In Model 4, moderate alcohol use in 2012 predicted less functional limitation in 2014 via greater social interaction in 2012.
Discussion and Implications: Social interaction is essential to the seemingly beneficial effect of moderate alcohol use on depressive symptomatology and functional ability. Clinically, this suggests caution in attributing health benefits to consumption of alcohol itself and identifies social interaction as a treatment target for improved health outcomes in later life.
Alcohol use in the Older Adult Population
The 2015 National Survey on Drug Use and Health indicated that approximately 42.7% of adults aged 65 years or older consumed alcohol in the past month (Center for Behavioral Health Statistics and Quality, 2016). Although heavy use of alcohol has been consistently related to adverse health outcomes (Wood et al., 2018), evidence suggests that moderate drinkers may be more likely to experience less difficulty with instrumental activities of daily living (IADLs; Cawthon et al., 2007), and improved quality of life, mood (Chan, von Mühlen, Kritz-Silverstein, & Barrett-Connor, 2009), and well-being (Dear, Henderson, & Korten, 2002). There remain important questions as to the nature of potential health benefits of moderate drinking in later life, and if present, how such benefits emerge.
The Elusive Health Benefits of Moderate Alcohol use
Contradictory findings continue to emerge regarding the impact of alcohol use on various health outcomes, ranging from conclusions that moderate alcohol use precipitates cortical shrinkage (Topiwala et al., 2017) and some cancers (Boissoneault, Lewis, & Nixon, 2016), to other findings that moderate alcohol use reduces all-cause mortality risk (Rehm, Gutjahr, & Gmel, 2001). With respect to later life health outcomes, a line of work emerged that astutely raised the question of how selection bias influences the appearance of a relationship between alcohol use and health benefits. Specifically, it was suggested that moderate drinkers may appear healthier than nondrinkers because as older adults face increasing morbidity, recreational alcohol use is likely to decline (Fillmore, Stockwell, Chikritzhs, Bostrom, & Kerr, 2007; Naimi et al., 2017; Shaper, Wannamethee, & Walker, 1988). Thus, research samples of abstinent older adults that are based on cross-sectional data are likely to include individuals who stopped drinking alcohol in response to emerging health concerns or medical contraindications. Inadequate identification of and control for third variables represent an additional source of error in this developing literature. For example, moderate drinkers tend to be better educated and wealthier than nondrinkers (Towers, Philipp, Dulin, & Allen, 2016), and so moderate drinkers may also have better access to healthcare and other resources that facilitate healthy aging. When taking these sources of bias into consideration, previously robust health benefits associated with moderate alcohol use are relatively weaker, yet remain statistically and clinically significant (Di Castelnuovo et al., 2006).
Alcohol use and Depressive Symptoms in Older Adults
It is estimated that 8%–16% of community dwelling older adults in the United States experience clinically significant depressive symptoms (Blazer, 2003). Among older adults, depression is associated with substantial functional impairment and negative health outcomes. For example, greater depressive symptomatology in later life is associated with increased risk for disability in activities of daily living (ADLs; Bruce, Seeman, Merrill, & Blazer, 1994), more visits to a medical doctor, and more days spent sick in bed (Yochim, Mast, & Lichtenberg, 2003). Given the imminent increase in the proportion of older adults in the United States (Colby & Ortman, 2015), there is good cause to explore modifiable factors associated with detrimental health outcomes in later life (e.g., depression).
The relationship between alcohol use and mood in later life is often characterized as a J- or U-shaped curve, such that moderate drinkers tend to report better subjective well-being and fewer depressive symptoms than alcohol abstainers and excessive drinkers (Chan et al., 2009; Coulson et al., 2014; Dear et al., 2002). For instance, in a population-based sample of Australian men, the odds of lifetime depression were greater for abstainers (odds ratio [OR] = 2.50) and for individuals consuming more than three drinks per day (OR = 1.45) when compared to individuals drinking up to two drinks per day (Coulson et al., 2014).
Moderate Alcohol use and Functional Ability Among Older Adults
Moderate alcohol use has been posited as a significant predictor of functional ability among older adults, raising questions regarding how alcohol use may promote independence throughout aging. Several studies have associated moderate alcohol use with better management of IADLs (Cawthon et al., 2007; Reid, Concaio, Towle, Williams, & Tinetti, 1999; Zimmerman, McDougall, & Becker, 2004). For example, older adults who drink moderately reported lower functional limitation considering difficulty with meal preparation, shopping, housework, walking two-to-three blocks, and/or climbing 10 stairs compared to abstainers (Cawthon et al., 2007). Further, a longitudinal study found that older adult moderate drinkers demonstrated decreased rates of functional decline as defined by ADLs, IADLs, and physical function (Wang, Van Belle, Kukull, & Larson, 2002) compared to nondrinkers.
Moderate Alcohol use: The Search for a Mechanism
There is limited work examining potential mechanisms of action for the posited beneficial effect of moderate alcohol use on health outcomes. If, in fact, health benefits are associated with moderate alcohol use, it is important to understand why these relationships exist. Such efforts may inform both alcohol use guidelines and interventions to improve health outcomes in ways that may not increase risk for incidental iatrogenic outcomes, such as alcohol use disorders, for older adults.
One promising mechanism through which alcohol use and biopsychosocial outcomes may be related is social interaction. Indeed, social interaction is related both to alcohol consumption patterns and health factors. For instance, socially isolated retirement community residents were more likely to drink less or be nondrinkers than were those with active social lives (Alexander & Duff, 1988). A similar study found a positive association among drinkers between frequency of contact with close friends and family and frequency of alcohol use (Adams, 1996). Integration of these results indicates a link between social interaction and degree of alcohol consumption. In addition, several cross-sectional studies have found that social support and participation are associated with fewer depressive symptoms (e.g., Burnette et al., 2017; Lam, Cervantes, & Lee, 2014), better IADL functioning (James, Boyle, Buchman, & Bennett, 2011; Tomioka, Kurumatani, & Hosoi, 2016), and lower risk for disability (de Leon, Gold, Glass, Kaplan, & George, 2001; James et al., 2011). In combination, these findings suggest a mechanistic hypothesis regarding the role of social interaction in the relationship between moderate alcohol use and health outcomes—depression and functional limitation in particular. The goal of these studies is to elucidate this network of relationships. In Study 1, we used cross-sectional analyses to evaluate the proposed models. In Study 2, we evaluated these networks longitudinally. Specifically, it was hypothesized that frequency of social interaction would mediate the relationship between (a) level of alcohol use (moderate vs abstaining) and depressive symptomatology, and (b) level of alcohol use and functional limitation (via IADLs) in a large, demographically representative sample of older adults.
Gerontologist. 2020;60(1):80-88. © 2020 Oxford University Press