A new meta-analysis shows that even modest increases in ultra-processed food intake can drive up type 2 diabetes risk, especially when consumption exceeds 300 grams per day.
Study: Ultra-Processed Food Intake and Risk of Type 2 Diabetes Mellitus: A Dose-Response Meta-Analysis of Prospective Studies. Image Credit: rafa jodar / Shutterstock
In a recent study published in the Diabetes and Metabolism Journal, researchers investigated the associations between ultra-processed food (UPF) intake and the risk of type 2 diabetes (T2D).
UPFs are industrial formulations made from fractions of whole foods, often modified and enhanced with food additives. UPFs are typically energy-dense, rich in sugar, sodium, and fat, and low in fiber. As such, UPF intake could lead to poor diet quality and excess energy intake, contributing to higher risks of obesity and other diseases. High UPF consumption may also impair insulin signaling and increase the risk of T2D. Food additives in UPFs, such as artificial sweeteners and emulsifiers, may also contribute by disrupting gut microbiota and promoting insulin resistance.
About the study
In the present study, researchers meta-analyzed the relationship between UPF intake and T2D risk. First, they identified studies from PubMed, Web of Science, and Embase databases, using relevant search terms. This meta-analysis was limited to prospective studies that reported relative risk (RR) estimates and 95% confidence intervals for the associations between T2D risk and UPF intake. Studies not following the Nova classification (of food groups) were excluded.
The following data were extracted from eligible studies: sample size, age, follow-up duration, dietary assessment methods, exposure range, covariates, relative risks (RRs), and timing of UPF assessment, among other variables. The Newcastle-Ottawa Scale and Oxford Centre for Evidence-Based Medicine criteria were used to assess the risk of bias and the level of evidence, respectively. DerSimonian-Laird random-effects models were used for the meta-analysis comparing the highest versus the lowest UPF intake and individual UPF subgroups.
Further, non-linear and linear dose-response meta-analyses were performed for different UPF units (serving/day, absolute grams (g)/day, and percentage of g/day), using the two-stage random-effects model with restricted cubic splines and the two-stage generalized least squares trend estimation method, respectively. Publication bias was assessed using Egger’s test and funnel plots.
Sensitivity analyses were performed by sequentially excluding each study from the meta-analysis to evaluate the robustness of the findings. Heterogeneity was examined using the I-squared statistic and Cochran’s Q test. Further, subgroup analyses were performed on methodological characteristics and potential effect modifiers to identify the sources of heterogeneity.
Findings
Overall, the study included 10 publications across 12 prospective cohort studies. Most studies were conducted in Europe (five) and North America (four). Seven publications were of high quality, while three had moderate quality. The meta-analysis of the highest vs. lowest UPF intake categories included 38,308 cases of diabetes from over 714,000 participants.
The summary RR was 1.48 for T2D risk, with moderate-to-high heterogeneity. Despite the heterogeneity in RR magnitude across studies, a consistent positive association was observed. No publication bias was noted. Among studies that reported associations for individual UPF subgroups, sugar- and artificially sweetened beverages, as well as processed meats, were positively associated with T2D risk.
Conversely, packaged sweet snacks and desserts, as well as packaged savory snacks and ultra-processed cereals and breads, were reported as inversely associated, although the inverse association for sweet snacks and desserts was not statistically significant (RR=0.92, 95% CI: 0.85–1.00). These inverse associations may be partly explained by factors such as flavanol content in chocolate-based products or the inclusion of whole-grain breads within the cereal subgroup.
In subgroup analyses by region, higher summary RRs were observed for European and North American studies compared to studies from other regions. This regional variation may be linked to higher intake of processed meats in Western populations, which was the UPF subgroup most strongly associated with T2D. The summary RRs for the highest versus the lowest UPF intake categories were comparable across studies using different UPF units.
Eight studies reported associations using the percentage of g/day as the unit of UPF intake. Each 10% g/day increment in UPF intake was significantly associated with a 14% higher risk of T2D, with no evidence of non-linearity. Four studies used absolute g/day as the unit of UPF intake. Each 100 g/day increase in UPF intake was associated with a 5% increased risk of T2D.
A non-linear relationship was observed only for absolute g/day intake, with a steeper increase in risk at intake levels exceeding 300 g/day. Likewise, four studies used servings per day as the unit of UPF intake. A one-serving-per-day increment in UPF intake significantly elevated the risk of T2D by 4%, and this association was linear.
Further, results remained robust in sensitivity analyses that excluded one study at a time.
Conclusions
In summary, the meta-analysis revealed a positive association between UPF intake and T2D risk. Higher UPF consumption was associated with a 48% increased risk of T2D. The association remained statistically significant after adjusting for body mass index (BMI), though the effect was partially attenuated, suggesting BMI may mediate part of the relationship.
The associations also held after adjustment for diet quality and total energy intake, suggesting an independent effect. A non-linear relationship was observed specifically for absolute g/day intake, with a steeper increase in T2D risk at intake levels exceeding 300 g/day of UPF. Given the growing surge in UPF intake, more efforts are needed to decrease UPF consumption by increasing access to minimally processed or unprocessed foods.