Building on findings from a study utilizing the U.S. Health and Retirement Study (HRS) (Du et al., 2023), the current research aimed to further investigate the relationship between perceived social support from children in later life and cognitive function. Additionally, C-reactive protein (CRP) and depressive symptoms were explored as possible mechanisms in this relationship due to past findings of the presence of inflammation in depressed individuals (Orsolini et al., 2023) and the effect of depression on cognitive function in older adults (Morimoto et. al, 2014). Perceived Social support may ameliorate some depressive symptoms, and therefore, help improve cognition in older adults.

The sample (N = 735) was gathered from the 2016 survey of the HRS. Respondents 65 years of age and older that had interpretable values for their CRP levels, scores on the CESD, answers to Perceived Social Support from their children, and valid scores on cognitive variables (Trail Making Test B, CERAD Word List Recall Delayed, Story Memory Delayed, Raven’s Standard Progressive Matrices, HRS Number Series, Retrieval Fluency, Letter Cancellation, Backward Count, CERAD Construction Praxis Delayed, and Symbol Digit Modalities Test) from the Harmonized Cognitive Assessment Protocol (HCAP) were included in the analysis. Perceived Social Support was determined by the mean of the response to the sum of three questions. A path analysis was run using Hayes’ Process macro for mediation analysis. The model investigated the relationship between an individual's perceived social support from their children and cognitive scores on HCAP measures. Additionally, mediation of this relationship by CRP levels and depressive symptom endorsement as measured by the CESD established the model. Models were run with and without covariates (age, gender, race, ethnicity, education, household income, ADL and IADLS, and medical conditions).

Models with covariates eliminated most partial and complete mediation effects noted in the analysis conducted without covariates. There was a direct effect between perceived social support and a single measure of processing speed (Symbol Digit Modalities Test) b = -1.12, SE = .48, CI = [-2.0635, -0.1759] when covariates were added to the model. However, the indirect effects were not significant. Depression endorsement b = .47, SE = .29, CI [ .011, 1.15] accounted for some of the relationship between perceived social support and a single measure of quantitative reasoning (HRS Number Series) when covariates were added to the model, and there is evidence for complete mediation as the total indirect effect was also significant b = .48, SE = .29, CI [.015, 1.17].  CRP does not appear to have an association with perceived social support in later life and cognitive outcomes, in models with and without covariates.

The study attempted to elucidate specific factors of cognition that may be implicated by perceived social support from children in later life. Based on this sample, there may be a relationship between perceived social support, depressive symptom endorsement, and some types of processing speed and quantitative reasoning. Further research is needed to determine the long-term effects of perceived social support from children on aspects of cognition.