(P13-016-25) Exploring Dose-Response Relationships Between Free Sugars and Metabolic Health: The Impact of Sugar Source

Objectives: This study expands upon the European Food Safety Authority’s (EFSA) meta-analytic assessment of the Tolerable Upper Intake Level for Dietary Sugars, which identified positive pooled mean associations (low vs. high dose) between free sugars (FS) intake and metabolic health markers. Using the same dataset, we aim to explore how these relationships vary by dose and sugar source (liquid versus solid sugars).

Methods: We utilized the randomized controlled trials from EFSA’s report to perform mixed-effects linear meta-regression to evaluate dose-response effects of FS intake, with interactions by source—beverages (B), foods (F), and mixed sources (Mx). Outcomes modeled included fasting lipids, glucose, and insulin, with exposure defined as the difference in FS intake between study arms (% energy).

Results: Among lipids, there was an overall positive dose-response relationship at FS intake differences ranging from 6-30%E for total cholesterol (TC), LDL-C, and triglycerides (TG) (no dose-response noted for HDL-C). Every 1%E increase in FS intake was associated with an average increase of: 2.4 mg/dL in TC (95% CI: 0.64, 4.16), 1.1 mg/dL (0.29, 1.84) in LDL-C, and 1.5 mg/dL in TG (0.98, 2.10). However, there was evidence of interaction by source for TC, LDL-C, and TG. For example, the dose-response relationship for TG was strongest for beverages (2.8 mg/dL increase per 1%E increase in FS intake) compared to mixed sources (1.7) and foods (0.49) (p=0.001). For both fasting insulin and glucose, there was an overall positive dose-response relationship but insufficient evidence of interaction by source. Residual heterogeneity remained high across all models, suggesting unmeasured factors contribute to the variability in these outcomes.

Conclusions: Identified dose-response relationships between FS intake and some metabolic outcomes vary by sugars source. For example, for lipids, we found stronger associations among beverages compared to solid or mixed sources. Despite significant heterogeneity and imprecision in the data, these findings suggest potential food matrix effects in the metabolic impacts of free sugars. Further research is warranted to clarify the underlying mechanisms and contextual factors influencing these relationships to inform more precise dietary guidance on free sugars consumption.

Funding Sources: World Sugar Research Organisation