Psychosocial factors associated with changes in cognition among middle‐aged and older Hispanics/Latinos: Findings from the HCHS/SOL and the sociocultural and SOL‐Investigation of Neurocognitive Aging (SOL‐INCA) ancillary studies.

Background: Psychosocial factors including intrapersonal and interpersonal resources and social stressors are associated with changes in cognition in older non‐Hispanic whites and Blacks. Few studies have examined these associations in Hispanics/Latinos, the largest ethnic/racial group in the US, or evaluated ethnically‐based psychosocial factors that may be more relevant to this population. We examined whether psychosocial factors predict seven‐year cognitive changes among US Hispanics/Latinos. Method: Data from the parent HCHS/SOL, and Sociocultural and SOL‐Investigation of Neurocognitive Aging ancillary studies were used (n= 2,158; ages 45+). Baseline psychosocial factors included: intrapersonal factors (optimism; purpose in life), interpersonal factors (family cohesion; social network embeddedness; social support), social stressors (loneliness; subjective social status), and ethnically‐based factors (ethnic identity; familism; perceived ethnic discrimination). Outcomes were changes in individual test scores for episodic learning and memory (B‐SEVLT‐Sum and SEVLT‐Recall), word fluency (WF), and processing speed (Digit Symbol Substitution, DSS) derived using regressions adjusting for time between assessments. Survey linear regression models were used to examine the associations between each psychosocial factor (z‐scores) and 7‐year changes in cognitive outcomes adjusting for age, sex, education, income, Hispanic/Latino background, language preference, and depressive symptoms. Result: Mean age was 55.9±7.9 years; females 54% of target population. Higher (baseline) levels of familism [(βSEVLT‐Sum (SE))= ‐0.08 (0.03), P<0.05)] predicted seven‐year decline in verbal learning. Higher optimism [βSEVLT‐Recall= 0.09 (0.03), P<0.01], social support [βSEVLT‐Recall= 0.07 (0.03), P<0.05], and ethnic identity [βSEVLT‐Recall= 0.06 (0.03), P<0.05] each predicted increase in memory. Higher familism [βSEVLT‐Recall= ‐0.08 (0.03), P<0.05] and loneliness [βSEVLT‐Recall= ‐0.06 (0.03), P<0.05] each predicted decline in memory. Higher purpose in life [βWF= 0.07 (0.03), P<0.05] predicted increase in word fluency. Higher family cohesion [βDSS= ‐0.08 (0.03, P<0.01) predicted decline in processing speed. Social network embeddedness, subjective social status, and perceived ethnic discrimination were not associated with any outcome. Conclusion: Psychosocial factors differentially related to changes in cognition. Results confirm previous studies in non‐Hispanic whites on associations of optimism, purpose in life, social support, and loneliness with select cognitive tests. They also highlight that ethnic identity and familism are ethnically‐based psychosocial factors that should be considered when examining healthy cognitive aging in US Hispanics/Latinos. [ABSTRACT FROM AUTHOR]