ABSTRACT
OBJECTIVES: To determine the strength of the association between objectively and subjectively measured physical activity (PA) and cardiovascular risk factors (CVRF) in a sample of older adults to assess which measure of PA is most appropriate for use in this population. STUDY DESIGN AND OUTCOME MEASURES: The study was a cross-sectional analysis of a community-dwelling sample aged 60 years or more (61 women and 34 men). Data collection included anthropometric measurements, blood analytics, and objective and subjective measurements of PA from accelerometers and Yale PA Survey questionnaires, respectively. RESULTS: Although the questionnaire indicated a higher summary index (total units) of PA in men than in women, objective measurements of total PA (mean counts per minute and steps per day) did not differ by sex. However, we obtained different results between women and men when we analyzed the relation between objective and subjective PA parameters and CVRF. In women, parameters that indicate cardiovascular risk, such as body mass index (BMI), waist circumference (WC), glucose, and total cholesterol, negatively correlated with objective PA parameters, while HDLc positively correlated with objective PA parameters and negatively with subjective PA parameters. In men, BMI and WC were negatively correlated with objective PA parameters, while blood cholesterol, LDLc, and triglycerides negatively correlated with subjective PA parameters. CONCLUSIONS: In women, it seems more appropriate to use objective methods to measure PA, while in men, objective and subjective methods seem to be complementary. Both methods of measurement, accelerometers and questionnaires, should be used simultaneously in research studies in older populations to establish precise relationships between PA and cardiovascular risk.
