[Study of obesity in a rural children population and its relationship with anthropometric variables].

GOALS: To know the prevalence of obesity in primary and secondary school students, and to evaluate the diagnostic accuracy of anthropometric variables for its detection.

DESIGN: Cross-sectional study.

LOCATION: Rural area of Córdoba. In the year 2014.

PARTICIPANTS: Student population. A stratified sampling was performed according to age, gender and educational centers. A total of 323 students from 6 to 16 years were included in the study, all parents had signed informed consent.

MAIN MEASUREMENTS: The prevalence of obesity was determined and sociodemographic, anthropometric, physical condition and dietary predictor variables were collected. A binary logistic regression was performed determining crude and adjusted Odds Ratio (OR) values, ROC curves were obtained and cut-off values were determined, calculating the sensitivity, specificity and Youden index.

RESULTS: The prevalence of overweight and obesity was 26.2% and 22.3%, respectively. Only 15.2% of school children had an optimal Mediterranean diet. The waist-height ratio (WtHR) was the predictive variable with the highest adjusted OR 7.1 (4.3-11.6) and the largest area under the curve 0.954 (0.928-0.979), from a global cut-off value to discriminate obesity of 0.507. This gave a sensitivity of 90% and specificity of 87.2%.

CONCLUSIONS: The high prevalence of obesity, the low-medium adherence to the Mediterranean diet and the low physical fitness make this population a priority target for the prevention of future cardiovascular events. The WtHR has been the best anthropometric predictor of obesity, recommending its use for the diagnosis of obesity in children at the expense of body mass index.