Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

School of Aging Studies

Major Professor

Brent J. Small, Ph.D.

Committee Member

Jerri D. Edwards, Ph.D.

Committee Member

Heather S. Jim, Ph.D.

Committee Member

Cathy L. McEvoy, Ph.D.

Committee Member

Stacey Scott, Ph.D.


cognition, genetic association, aging, multileve l modeling


The current dissertation used a gene x environment (G x E) approach to examine the independent and interactive effects of specific genetic variants and participation in physical and cognitive/social activities (PA and CSA) on cognitive performance in 4,764 participants of the Health and Retirement Study. Using three-wave data, three sets of multi-level growth models were conducted to examine baseline, longitudinal, and interactive effects of genotype (i.e., ApoE, COMT, and BDNF) and CSA/PA on performance across five cognitive measures: immediate, delayed and total word recall, and serial 7s and backwards counting.

At baseline, the ApoE ε4 allele predicted worse performance in all measures except backwards counting, and the BDNF Met allele predicted better recall scores. The effect of COMT genotype was not significant. Higher CSA/PA predicted better performance on almost all measures. One significant G x E interaction was found between COMT x CSA for backwards counting. Longitudinally, participation in CSA moderated the effect of time on word recall in the ApoE and BDNF models. These results support the idea that genetic and environmental factors are mechanisms of cognitive aging, but also exemplify the variability seen in genetic association studies. Further research is needed to translate such findings into clinically relevant criteria that can be used to identify individual susceptibility to cognitive decline.