Statistical analysis of the potential risk factors for adult obesity in the United States

Open Access
Farde, Amey Madhukar
Graduate Program:
Industrial Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
Committee Members:
  • Andris Freivalds, Thesis Advisor
  • Dr Samuel Oyewole, Thesis Advisor
  • Dr Paul Griffin, Thesis Advisor
  • Obesity
  • risk factors
  • obesity model
Obesity is a chronic physical condition which is characterized by high body fat, is accompanied by health problems such cardiovascular disorders, musculoskeletal injuries, diabetes, high blood pressure, high blood cholesterol, heart disease and stroke. In the United States, over 60% of the population is at least overweight. Approximately $147 – $168 billion is spent annually on obesity-related conditions and treatments. There have been studies related to the measure of obesity based on Body mass index (BMI), weight, body fat, waist hip ratio and waist circumference (WC). In this research, obesity was analyzed among men and women of varied age groups, race, family income and education level using BMI cutoffs recommended by WHO. The data used in the analysis was collected from National Health and Nutrition Examination Survey (NHANES) from January 2009 to December 2010. From the 8,379 individuals used in the study, a model was developed to predict the BMI of an individual using inverse transformation. According to the analysis, 6.16% of the total female population (n=4184) were extremely obese compared to 3.14% of male population. It was evident that the prevalence of obesity was higher in female population than that in male. Amongst all the races, Non-Hispanic blacks were in greater percentages in class I (16.72%), class II (8.17%) and class III (8.30%) obesity. Class I and Class II obesity was more evident among individuals aged greater than 60 years. The analysis indicates rise in obesity with increasing age for the population with BMI greater than 25. A predictable trend was visible in class I obesity, with the level of obesity inversely related to the education level. No significant implications could be pointed out from the analysis of the family income of the population. This research could provide a new approach in the design of methods to prevent obesity and promote physical activity. The analysis could also be used for health-care planning to estimate the cardiovascular risks among the population and their costs implications.