ABSTRACT
INTRODUCTION: In this study, we measured neurocognitive function, perceived stress, quality of life (QOL), and psychopathology in community-dwelling adults, with data contrasted across tertiles of exercise frequency. METHODS: A group of 998 adults (age 18-85 yr) was measured for neurocognitive function using a computerized neuropsychological test from CNS Vital Signs (Morrisville, NC). They also completed the Brief Symptom Inventory (BSI), which measures psychopathology, as well as the World Health Organization QOL questionnaire and the Perceived Stress Scale. General linear modeling was used to examine relationships between exercise frequency and neurocognitive function, BSI, QOL, and the Perceived Stress Scale. Backward selection in the GLMSELECT procedure in SAS (version 9.1.3; SAS Institute, Inc., Cary, NC) was used to identify confounding variables including age, gender, body mass index, marital status, education level, stress level, alcohol, smoking, and chronic disease. A contrast to test linear trend was performed after adjusting for confounders. Pairwise comparisons were performed across exercise frequency tertiles using the Tukey-Kramer method. RESULTS: P values for trend tests and pairwise comparisons were nonsignificant for all five cognition function domains across exercise frequency tertiles after adjustment for confounders. Age and education level emerged as the best correlates of neurocognitive function. P values for trend were significant for all BSI domains and indices, QOL, and perceived stress, across exercise frequency tertiles. CONCLUSIONS: In conclusion, nine BSI psychopathology domains, perceived stress, and QOL but not five neurocognitive function domains were modestly but significantly associated with aerobic exercise frequency in a heterogeneous group of community-dwelling adults after adjustment for demographic and lifestyle factors.
