Assessing the Linkages Between a Processing-Based Dietary Index and Nutrient Intakes, Obesity, and the Gut Microbiome Among Bolivian Women in a Region Undergoing the Nutrition Transition

Abstract

Low- and middle-income countries (LMICs) are undergoing a rapid dietary transition from traditional foods to highly processed Western diets. Conventional analyses of dietary patterns do not account for the level of processing of foods in the diet, yet methods of food processing and ingredients introduced into foods to make them more palatable, nutrient-dense, and/or shelf-stable have been hypothesized to increase individuals’ risk for obesity, in part through disruptions to the gut microbiome.

This dissertation implemented the use of two food processing-based indices, the Processed Food Dietary Index (PFDI) and NOVA, to characterize the level of processing of diets among women recruited from the baseline participants of a three-year longitudinal cohort study in Montero, Bolivia, a country experiencing a rapid dietary transition, to further examine the associations between the extent of processing of diets, obesity, and gut microbiome composition. With the linkages between the nutrition transition in LMICs and obesity best established in women, we collected three 24-hr dietary recalls, anthropometric measurements, and two fecal samples from 160 women of reproductive age (18-49 years) based on the extent of processing in the diet using data from the aforementioned baseline assessment data of the longitudinal study.

The first aim utilized both the PFDI and NOVA as a single measure of diet quality to examine and compare relationships between nutrient intakes and the dietary share of processing level groups; it also examined the distribution of nutrient means across quintiles of the dietary share of ultra-processed foods and drinks (UPFDs) and unprocessed/minimally processed foods. We found statistically significant linear relationships between various nutrient means and PFDI and NOVA scores, although not always in the hypothesized directions. There were also a number of nutrients typically associated with processing that did not exhibit statistically significant linear relationships with processing levels (i.e., saturated fat, trans fat, sodium).

The second aim examined the association between the processing level of the diet and measures of obesity (i.e., body mass index (BMI), waist circumference (WC), and waist-to-hip ratio (WHR)) using the PFDI and NOVA as a single measure of overall dietary quality. Although daily caloric intake increased across quintiles of PFDI and NOVA scores, there were no differences found in BMI, WC, or WHR between quintile pairs. Neither processing level of the diet nor the consumption of UPFDs were associated with obesity.

The third aim evaluated the association between the processing level of the diet, obesity, and the gut microbiome. We found that the processing level of the diet, as measured by the PFDI, may influence the Firmicutes/Bacteroidetes (F/B) ratio and that the proportion of UPFD intake in the diet is associated with the diversity of microbiota present in the gut. We also observed that F/B ratios were not different between obese and lean individuals; however, obese individuals had a less diverse microbiome than those who were lean.

These studies provide information regarding the use of the PFDI and NOVA classification systems in future studies assessing the impact of food processing on human health, as well as novel insight into the relationships between the level of processing in the diet, obesity, and the composition of the gut microbiome.