Association Between Ultra-Processed Food Intake and Metabolic Profile and Risk of Type 2 Diabetes and Cardiovascular Disease in U.S.-Residing Cohorts of Hispanic/Latino and White Adults

Abstract

Background: The global prevalence of type 2 diabetes (T2D) and cardiovascular disease (CVD) has increased alongside the rise of heavily industrialized food processing and stagnating dietary quality. Ultra-processed foods (UPF), characterized by poor nutritional composition and potentially harmful additives, have been described as major contributors to these epidemics. While several studies have associated UPF consumption with increased T2D risk, none have focused on Hispanic and Latino adults in the US. Moreover, the metabolome response to UPF intake in this population remains unexamined. Although several studies have reported a harmful association between UPF intake and CVD, only one has accounted for the nutritional heterogeneity within the UPF category through food group analyses, also with limited evidence on this matter for T2D.

Objectives: Analyzing data from US-residing prospective cohorts of Hispanic/Latino and White adults, this thesis aimed to assess the association between Nova-defined UPF intake, the risk of T2D, circulating metabolome response, and CVD risk. The aims of this thesis were: 1. To characterize the intake of UPF and estimate the associations between total and group-specific UPF intakes, diet quality, and T2D risk in The Hispanic Community Health Study/Study of Latinos (HCHS/SOL). 2. To identify a plasma metabolomics signature cross-sectionally reflective of high intake of UPF at baseline, and assess its association with incident T2D in the HCHS/SOL. 3. To estimate the association of total and group-specific UPF intakes with CVD, coronary heart disease (CHD), and stroke in the Nurses' Health Study (NHS), Nurses' Health Study II, and Health Professionals Follow-Up Study (HPFS). Also to conduct a systematic review and meta-analyses to update the evidence of associations of total UPF intake with these cardiovascular outcomes.

Methods: Chapter 1 consists of analyses including 8,010 Hispanic/Latino adults, assessing UPF intake via 24-hour dietary recalls and food propensity questionnaires. Survey-based regression analyses examined associations of baseline UPF intake with diet quality components (cross-sectionally) and incident T2D (prospectively; 6-year follow-up). Chapter 2 entailed a random sub-sample of the HCHS/SOL including 3,264 adults. Baseline 24-hour dietary recall-derived UPF intake was regressed on 650 serum metabolites using elastic net and leave-one-out cross-validation. A metabolomics signature was calculated as the weighted sum of selected metabolites, and survey-based logistic models estimated the odds ratio of T2D per 1-SD increase in the signature and selected individual metabolites, adjusting for multiple comparisons. Chapter 3 comprises analyses in the NHS (n=75,735), NHSII (n=90,813), and HPFS (n=40,409), where the cumulatively-averaged UPF intake was assessed via food frequency questionnaires. Prospective cohort-specific associations of UPF intake with CVD, CHD, and stroke risks (median follow-up of 26.6 to 31.9 years) were estimated using Cox regression and pooled through fixed-effect models. Random-effects meta-analyses combined prospective findings on UPF-CVD associations from Medline and Embase (up to April 5, 2024).

Results: 1. UPF was associated with a compromised diet quality and with a higher risk of T2D, particularly among Puerto Rican and US-born individuals. Significant divergent relationships with T2D risk were observed for different types of UPF within these study groups and among participants with low diet quality. Adverse associations were found for refined bread/cereals, sugar-sweetened beverages (SSBs), artificially-sweetened beverages (SSBs), processed poultry/fish, and hard liquors, while inverse associations were observed for sweet snacks, desserts, spreads, and savory snacks. 2. An increase of a standard deviation in a 204-metabolite signature reflective of higher UPF intake (an 18-percentage-point increase) predicted a greater risk of T2D at the 6-year follow-up among Puerto Ricans, US-born participants, participants with the lowest diet quality, and participants with excess weight. Compared with the total metabolite panel, the signature underrepresented lipids, overrepresented xenobiotics, and had an equal contribution from amino acids. The most common metabolites within this signature and with the strongest associations with T2D pertained to the branched-chain amino acid (BCCA), sphingomyelin, and aromatic amino acid (AAA) metabolic sub-pathways. 3. Higher total UPF intake was adversely associated with higher risk of CVD and CHD in the NHS, NHSII, and HPFS, and meta-analyses also suggested a small significant excess risk of stroke associated with the consumption of these products. Divergent associations were observed; SSBs, processed meats, and ASBs were associated with higher CVD and CHD risk. Conversely, savory snacks, cold cereals, and yoghurt/dairy-based desserts were inversely associated with CVD and CHD risk. Bread and cold cereals were associated with lower stroke risk, and hard liquors with lower CHD risk.

Conclusions: Widespread evidence regarding the adverse role of total UPF intake on T2D risk extends to US Hispanic/Latino adults. Puerto Rican and US-born people deserve special attention, as there may be acculturation factors facilitating higher intake of these products that promotes a higher T2D risk among them. The consumption of UPF appears to have a measurable yet heterogeneous impact on the serum metabolome, predicting a higher future risk of T2D among Puerto Ricans, US-born participants, participants with the lowest diet quality, and participants with excess weight. Although further investigation on biological mechanisms is warranted, the metabolic response associated with UPF in this thesis encompassed BCAAs, sphingomyelins, and AAAs, metabolic sub-pathways that may participate in impaired insulin homeostasis and other pathophysiological pathways preceding the incidence of T2D. Total UPF intake was adversely associated with CVD and CHD risk in three cohorts of US adults mostly of White race, corroborated by prospective studies from multiple countries, also suggesting a small excess stroke risk. Given the divergent patterns of associations across different types of UPF, nutritional guidance and food policies for T2D and cardiovascular health should account for the distinct impacts of specific UPF groups.