Classification of obesity among South African female adolescents: comparative analysis of logistic regression and random forest algorithms

Abstract

Previously considered to affect well-developed and high-income countries, obesity is a public health threat that is increasing in low- and middle-income countries (LMICs). The global adult obesity rate has almost tripled in the past four decades. Currently, 39% of adults globally are overweight and 13% obese. Overweight among children and adolescents aged 5-19 years increased from 4% in 1975 to over 18% in 2016 [1]. Like other middle-income countries, South Africa is currently facing a triple burden of malnutrition, highlighted by a co-existence of overweight and underweight and micronutrient deficiencies [2]. This has led to the acceleration of obesity rates in children and adolescents in South Africa, as identified in national surveys.