Gluten consumption may contribute to worldwide obesity prevalence

Abstract

Gluten consumption has been controversially associated with obesity in previous studies. We sought to examine this association at the worldwide level.

Country specific data were obtained from 168 countries. Scatter plots, bivariate, partial correlation and multiple linear regression models were used to explore and compare the coincidence between obesity prevalence and consumption of gluten, non-gluten cereal protein and total cereal protein respectively. The established risk factors of obesity: caloric intake, sedentary lifestyle, urbanization, socioeconomic status, meat protein intake and sugar consumption were included in analyses as potential confounders. The 168 countries were also stratified into developing and developed country groupings for further examination of the relationships.

Worldwide, bivariate correlation analyses revealed that the strength and direction of correlations between all variables (independent, dependent and potential confounders) were at similar levels. Obesity prevalence was positively correlated to gluten consumption but was negatively correlated to consumption of non-gluten cereal protein, and was in almost nil correlation to total cereal protein consumption. These relationships were similar across all countries (n= 168), developed country grouping (N=44) and devel­oping country grouping (n=124). When caloric intake, Gross Domestic Product at Purchasing Power Parity, sedentary lifestyle and urbanization were kept statistically constant in the partial correlation analysis, obesity was significantly correlated to gluten consumption in all countries, developed country grouping and developing country grouping, and was significantly but inversely and weakly correlated to non-gluten cereal protein in all countries and developing countries, and was in almost nil correlation to total cereal protein in all country groupings. Globally, stepwise multiple regression analysis, when all the independent variables and potential confounding factors were included, selected consumption of sugar as the variable having the greatest influence on obesity with R2 = 0.510, while gluten was placed second increasing R2 to 0.596.

Gluten consumption may have been emerging as an inconspicuous, but significant cause of obesity. While Westernization has driven the diet patterns worldwide to incorporate more gluten crops, obesity prevalence projection methods may estimate future obesity rates poorly if gluten consumption is not considered.