Frontiers in Nutrition · Nov 16, 2026
Association of Carbohydrate Intake with Glycemic Variability in Adults with Central or General Obesity: Distinguishing Within-and Between-Person Effects Using CGM
By Xiaoyuan Zhang, Jiayun Wang, Zhonghui He et al.
Background: The role of carbohydrate intake in glycemic control remains debated, specifically whether the proportion of energy from carbohydrates or the consistency of daily intake matters more. This study investigated the associations between carbohydrate intake and dynamic glucose metrics in adults with central or general obesity using continuous glucose monitoring (CGM), with a focus on distinguishing within-person from between-person effects. Methods: Ninety-nine sedentary adults with central or general obesity (mean age 25.55 years, mean body mass index (BMI) 30.04 kg/m²) were monitored for 7–14 days using CGM, dietary records, and accelerometers. Person-level regression models were used to evaluate the between-person associations between habitual carbohydrate energy percentage and glucose parameters. Additionally, linear mixed-effects models with person-mean centering were applied to disentangle within-person and between-person effects of carbohydrate intake, controlling for age, gender, BMI, moderate-to- vigorous physical activity,, and total caloric intake. Isocaloric substitution models were further constructed to assess the effects of replacing protein or fat with carbohydrates. Results: Habitual carbohydrate energy percentage showed no significant association with any glycemic parameter (all p>0.05). Conversely, daily within-person fluctuations in carbohydrate intake significantly predicted increased glycemic variability (standard deviation: β=0.004, P<0.001; coefficient of variation: β=0.060, P<0.001; largest amplitude of glycemic excursions: β=0.011, P=0.011) and lower minimum glucose (β=−0.005, P=0.028). These primary associations remained robust after false discovery rate correction. Time in range did not show a statistically significant change after full adjustment for daily fat and energy fluctuations (P=0.288). Isocaloric substitution models revealed that the adverse within-person effects were primarily driven by carbohydrate replacing protein, rather than replacing fat. Conclusion: Long-term habitual carbohydrate intake was not significantly associated with glycemic profiles, whereas daily fluctuations in carbohydrate intake robustly exacerbated glycemic variability and were associated with lower minimum glucose. These adverse within-person effects were primarily driven by carbohydrate displacing protein. Therefore, dietary interventions for adults with central or general obesity should prioritize maintaining day-to-day consistency in carbohydrate intake and preserving adequate protein consumption, rather than exclusively focusing on restricting the overall proportion of carbohydrates.