Access Type

Open Access Dissertation

Date of Award

January 2015

Degree Type


Degree Name




First Advisor

John L. Woodard


The goal of this dissertation is to examine the effect of cognitive and social activities on cognitive performance and health conditions in a national sample of older adults from the Health and Retirement Study (HRS). This dissertation first aimed to identify longitudinal relations between activity frequency and cognitive functioning. Two hypotheses were tested, baseline activity frequency predicts change in cognitive functioning over time, and baseline cognitive performance predicts change in activity frequency over time. The dissertation’s second aim was to identify links between activity frequency and cognitive trajectories. The third aim was to identify longitudinal relations between activity frequency and overall health. Changes in cognitive functioning and health were also compared simultaneously over three time points, to identify causal relations. The sample included 3,397 respondents aged ≥ 60 years old from the Health and Retirement Study’s 2008, 2010, and 2012 waves. Respondents completed brief cognitive tests and items regarding their health during each wave, as well as items ranking frequency of engagement in cognitive and social activities in 2008 and 2012. A series of structural equation models were implemented to test the aforementioned aims. A paucity of significant findings precluded a comparison between the two hypothesized models on activity frequency and cognitive functioning, as well as between the two contrasting models for activity frequency and overall health. Activity frequency did not significantly predict rate of change in cognitive performance or health conditions over time. Activity frequency also did not significantly change over time. However, frequency of baseline cognitive activity was associated with initial level of episodic memory. Further, a lower frequency of cognitive activities was associated with a higher number of health conditions at baseline, when education was not included in the model. Relations between health and cognition were not consistently indicated over time, suggesting that health conditions may not have strong causal effects on age-related changes in cognitive functioning. Significant associations between baseline activity engagement and initial level of episodic memory and health conditions does not allow one to rule out the protective effect of activity engagement on cognition and overall health.