Characterising the distributions of height and body-mass index and their interrelationship

Abstract

Height and weight are indicators of healthy versus unhealthy nutrition. Body-mass index (BMI) measures weight gain beyond what is expected from height gain. Having short stature or too little weight for one’s height, represented by low BMI, as well as having excessive weight for one’s height, represented by high BMI, increases the risk of morbidity and mortality. My thesis aimed to provide summary statistics that characterise the distributions of height and BMI and their interrelationship. Data were collated via the NCD Risk Factor Collaboration (NCD-RisC) network, from population-based studies that had anthropometrics measurements from 1985 to 2019 for a total of 2,896 studies with 187 million participants. Of these, 1,282 surveys provided over 11 million participants aged 20 to 79 years, with 1,021 surveys of over 1.4 million women aged 40-49 years and 815 surveys of over 870,000 men aged 40-49 years. I used a hierarchical Bayesian model to estimate mean, variance and skewness of height and BMI, as well as their correlation coefficient; I also used a regression model to estimate the contribution of the change in mean BMI to the change of prevalence in underweight and obesity. In 2019 versus 1985, mean and variance of both the height and BMI distributions increased in most countries and sexes; skewness of the height distribution remained around zero for both women and men in most countries, while skewness of the BMI distribution, although it was a positive number for both sexes, it decreased in women from most countries and increased in men from all countries. Changes in the prevalence of underweight and total obesity, and to a lesser extent severe obesity, were largely driven by shifts in the distribution of BMI, with smaller contributions from changes in the shape of the distribution. The correlation coefficient between height and BMI did not change significantly from zero for most countries and sexes. Considering that the height gain was not proportional to the BMI increase, segments of the two distributions were affected heterogeneously so policy makers and health practitioners need to tackle the double burden of malnutrition.