Physical Activity Clusters and Income Inequality in Brazilian Adults

Abstract

Background: Income is an important determinant of physical activity (PA) when analyzed in its different domains. Sociodemographic characteristics such as sex, age, education, and marital status reveal distinct population profiles when PA domains are analyzed in isolation. This study aimed to describe clusters of PA in domains within income inequalities and to investigate the associated sociodemographic characteristics of Brazilian adults. Methods: A secondary analysis of the National Health Survey was performed (N = 50,176). PA, sociodemographic characteristics, and family income were investigated. Low- (n = 9504) and high-income adults (n = 6330) were analyzed. Two-step cluster and Rao–Scott chi-square tests were employed. Results: High-income adults accumulated 1.06 times more PA in leisure time compared with low-income adults. Of the 3 clusters observed, the inactive cluster was more prevalent (low-income group: 65.9%; 95% confidence interval [CI], 64.1–67.5; high-income group: 84.5%; 95% CI, 82.9–86.0). Work/leisure activities (21.2%; 95% CI, 19.8–22.8) and commuting/household activities (12.9%; 95% CI, 11.8–14.1) characterized low-income adults. Work/household activities (10.9%; 95% CI, 9.6–12.3) and commuting/leisure activities (4.6%, 95% CI, 3.9–5.4) characterized high-income adults. Sex (P < .001), age (P < .001), and marital status (P = .0023) were associated with low-income clusters. Conclusion: PA clustering differs within income inequalities. PA in leisure differentiates the opportunities in low- and high-income groups, but it is representative of a very small portion of the wealth.