The inflammatory potential of the diet as a link between food processing and low-grade inflammation: An analysis on 21,315 participants to the Moli-sani study

Background & aims: Food processing may adversely affect human health through a variety of mechanisms, including the development of a chronic pro-inflammatory state. In this study we aimed to test the hypothesis that an increasing degree of food processing is directly associated with low-grade inflammation, and evaluate to what extent this association is mediated by the inflammatory potential of highly processed foods. Methods: Cross-sectional analysis on 21,315 subjects (mean age 55 ± 3 y) from the Moli-sani Study with complete dietary data collected by a validated 188-item food frequency questionnaire. The NOVA classification was used to categorize foods on the basis of industrial processing as: 1) unprocessed/minimally processed foods; 2) processed culinary ingredients; 3) processed foods; 4) ultra-processed foods (UPF). The inflammatory potential of the diet was evaluated through the Energy-adjusted Dietary Inflammatory Index (E-DII™). Low-grade inflammation was assessed by a composite INFLA-score including C-reactive protein, leukocyte and platelet counts and the granulocyte to lymphocyte ratio. Results: In multivariable-adjusted linear regression models, the INFLA-score was positively associated with E-DII (β = 0.15; 0.10, 0.19), processed foods (β = 0.04; 0.01, 0.08) and UPF (β = 0.13; 0.07, 0.19), but inversely associated with minimally processed foods (β = −0.09; −0.13, −0.06). The E-DII score was inversely associated with minimally processed food (β = −0.40; 95%CI -0.41, −0.39 for 5% increment in the weight ratio) but directly with either processed culinary ingredients (β = 0.18; 0.15, 0.21 for 1% increment), processed food (β = 0.28; 0.27, 0.29 for 5% increment) or UPF (β = 0.34; 0.32, 0.36 for 5% increment). The inclusion of the E-DII into the multivariable model explained 88.5% of the association of processed food with the INFLA-score (p < 0.0001) and mitigated by 32.6% (p < 0.0001) the association with UPF. Conclusions: The association of UPF with low-grade inflammation is only partially explained by the high pro-inflammatory potential of these foods. Further studies are warranted to test whether the observed adverse relationship of UPF with low-grade inflammation could be triggered by mechanisms that are not directly related to the pro-inflammatory potential of highly processed food products.