ABSTRACT BACKGROUND:

Associations between behaviors and individual chronic diseases have been demonstrated. However, the relationship between time spent on sedentary behavior and multimorbidity remains less clear.

OBJECTIVE:

To identify the predictive power of various intensities of physical activity versus sedentary behavior, as discriminatory factors for cardiometabolic multimorbidity (cardiovascular diseases and diabetes) in the elderly.

DESIGN AND SETTING:

Cross-sectional study in different residential census tracts and residential households in Florianópolis (SC).

METHODS:

The participants were 425 elderly people (65% women) from the EpiFloripa Aging study in 2014. Sociodemographic variables and self-reported incidence of cardiovascular diseases and diabetes were obtained via a questionnaire. Light physical activity (LPA), moderate-to-vigorous physical activity (MVPA) and sedentary behavior (SB) were measured using accelerometers. The analyses were stratified according to sex and included a diagnosis for interpretation. Behaviors were taken into consideration if their predictive power in terms of area under the receiver operating characteristic (ROC) curve was greater than 0.50. The time cutoff point was defined from sensitivity and specificity.

RESULTS:

For older adult men with diabetes, the predictive value of MVPA for absence of multi-morbidity was an area of 0.75 (95% confidence interval, CI: 0.538-0.962), and a cutoff of 17 minutes per day. Older adult women with diabetes had an area of 0.71 (95% CI: 0.524-0.866) and a cutoff of 10 minutes per day. LPA and SB did not present predictive values.

CONCLUSION:

The time spent on MVPA is a predictor of absence of multimorbidity in elderly people with diabetes, for both sexes.