Estimating the heterogeneous effect of life satisfaction on cognitive functioning among older adults: Evidence of US and UK national surveys

Abstract

BackgroundsThe emerging field of positive psychology suggests higher life satisfaction, a form of psychological well-being, may improve cognitive functioning. Although evidence exists for population-average associations between psychological well-being and better cognitive function, little is known about how the relationship varies across individuals. MethodsWe analyzed a national sample of US and UK adults aged ≥50 from the Health and Retirement Study (HRS) (n = 10,650) and the English Longitudinal Survey of Aging (ELSA) (n = 5514). We assessed life satisfaction at baseline using the Satisfaction with Life Scale. Cognitive functioning was assessed using a modified version of the Telephone Interview for Cognitive Status score after 4-year follow-up. We estimated the population-average association between life satisfaction and cognitive functioning in each sample via doubly-robust targeted maximum likelihood estimation with SuperLearning. To assess effect heterogeneity, we estimated conditional average effects via a causal forest algorithm. ResultsWe did not find reliable evidence of a population-average association between life satisfaction and higher cognitive functioning in HRS (HRS: β = -0.12; 95%CI: -0.30, 0.06) and ELSA (ELSA: β = 0.39; 95%CI: -0.00, 0.79). Our machine-learning-based approach for estimating effect heterogeneity discovered the effect of life satisfaction on cognitive function can substantially vary across individuals. Life satisfaction appeared less beneficial, or even detrimental, among individuals with lower socioeconomic status, poor health status, and more negative psychological conditions, both in the US and UK samples. ConclusionsFurther research is needed to uncover mechanisms underlying the heterogeneous effects of life satisfaction on cognitive function, as it may have unintended adverse consequences among some subgroups.