Food processing according to the NOVA classification is not associated with glycemic index and glycemic load: results from an analysis of 1995 food items

Abstract

BackgroundUltraprocessed foods (UPFs) comprise most calories in the United States diet. Glycemic index (GI) and glycemic load (GL) are measures of the quality and quantity of carbohydrates in food based on their effect on postprandial blood glucose. Diets high in UPFs and GI/GL are associated with chronic metabolic diseases but the relationship between them is unclear. ObjectivesOur objective was to examine the GI and GL of foods assigned to NOVA food processing groups. We hypothesized that GI and GL would be lowest in minimally processed foods (MPFs) compared to processed food (PRF) and UPF (with no difference between PRF and UPF). MethodsGI and GL values produced by healthy individuals for 1995 food items were collated from published sources. Food items were manually coded by processing levels according to NOVA classification. In addition, as the effects of processing on glycemic potential may vary between types of foods, food items were coded into 8 groups (beans, nuts, and seeds; beverages; dairy; fats and sweets; fruit; grains; fish, meat, and poultry; and vegetables). Multilevel linear modeling was used to determine significance with an α value of 0.05. ResultsThe effect of food processing on GI and GL was contrary to our hypothesis as means did not differ significantly across processing levels: GI—MPF: 54.1 ± 19.5, PRF: 53.2 ± 18.9, UPF: 49.3 ± 18.1 (P = 0.712); GL—MPF: 17.1 ± 10.3, PRF: 15.8 ± 12.4, UPF; 11.5 ± 7.9 (P = 0.890). Within food groups, there was no significant association between processing level and GI (P = 0.184), but GL was inversely associated with grains and vegetables (P < 0.001). ConclusionsAcross analyzed foods, GI and GL do not differ between processing levels, whereas GL was lower in ultraprocessed grains and vegetables than MPF. Any potential adverse outcomes associated with UPF are unlikely to be related to effects on glycemia.This project was preregistered at the Open Science Framework (OSF) Registries through the Center for Open Science as 10.17605/OSF.IO/PJWG9.