Psychosocial stress and central adiposity: A Brazilian study with a representative sample of the public health system users

Objective To assess the association between indicators of psychosocial stress and central adiposity in adult users of the Unified Health System (SUS) from Southeast of Brazil. Methods This cross-sectional study was conducted with 384 adults (20 to 59 years old) from the city of Alegre, Southeastern Brazil. The simple random sample represented the population using the public health system of the municipality. The prevalence of obesity was based on the Body Mass Index, and central adiposity (dependent variable) was measured by waist circumference in centimeters. The independent variables were the following indicators of psychosocial stress: food and nutrition insecurity (yes/no), serum cortisol (mu g/dL), symptoms suggestive of depression using the Beck Depression Inventory-II > 17 (yes/no), and altered blood pressure > 130/85 mmHg (yes/no). Univariate linear regression was performed between central adiposity and each stress indicator, and later the models were adjusted for socioeconomic, health, and lifestyle variables. All analyses were made separately by rural and urban location. Results The prevalence of weight excess, by the classification of the Body Mass Index > 25.0 kg/ m2, was 68.3% and, by waist circumference, 71.5% of individuals presented an increased risk for metabolic complications related to central adiposity. Mean waist circumference scores for the rural and urban population were 89.3 +/- 12.7 cm and 92.9 +/- 14.7 cm, respectively (p = 0.012). Indicators of stress that were associated with central adiposity were: cortisol in the rural population (beta = -0.60; 95% CI = -1.09;-0.11) and altered blood pres- Competing interests: The authors have declared sure in the urban population (beta = 6.66; 95% CI = 2.14;11.18). This occurred both in the raw analysis and in the models adjusted for confounding factors. Conclusion Central adiposity was inversely associated with cortisol in the rural population and directly associated with higher arterial blood pressure in the urban population, suggesting a local influence on how individuals react to stress.