Obesity and ultra-processed products

Abstract

Introduction Obesity is a global health problem with a wide range of determinants. In recent years, the relationship between ultra-processed products and obesity has become cause for concern. However, there remains very limited understanding of the nature of this relationship globally. In order to develop policies to counteract the effects of ultra-processed foods, a greater understanding is required of the extent of these effects. Aim To investigate the association between trends in volume sales of ultra-processed products and trends in adult obesity, across countries at a global level. Methods The research was divided into three parts as follows:  An analysis was performed of the trends in volume sales of ultra-processed products across regions and income categories between 2002 and 2016 using sales data sourced from Euromonitor International.  Using the same sales data, a weighted least squares regression was performed for the association between ultra-processed products and BMI, with adjustments made for confounding using GDP, proportion of population in urban areas, and mean years of education.  Potential moderation of the association between ultra-processed products and BMI was investigated for the covariates GDP, proportion of population in urban areas, and mean years of education. Results Large variation was seen in the volume sales of ultra-processed products across regions globally during the period 2002-2016 with the highest levels of volume sales of ultra-processed products observed for North America and Australasia, and lowest for Asia. Across income categories, the high-income countries showed the largest levels of volume sales of ultra-processed products during 2002-2016, followed by the upper-middle-income countries, and the lowest levels found for the lower-middle-income countries. The adjusted model from the weighted least squares regression showed that an extra 100g/day of ultra-processed products would give an average of 1.05 kg extra weight over the population for men and 1.03kg for women. A significant positive association was found between volume sales of ultra-processed products and BMI for both men and women. Significant interaction effects for lnGDP, proportion of population in urban areas and mean years of education were also found for both sexes, with women showing stronger interaction effects for each. Conclusions Large variation was found between regions globally as well as income categories, for the trends in volume sales of ultra-processed products. Increasing intakes of ultra-processed products are a likely contributor to the increasing prevalence of BMI across countries and regions. There is a crucial need for the concept of ultra-processed products to be incorporated into strategies targeted towards obesity reduction. As well as this, further research is required around the effectiveness of different methods for reducing the consumption of ultra-processed products.