The association between glucose dynamics and energy intake in young, healthy women

Glucose has been implicated in the control of appetite and food intake. This study investigated whether glycaemic patterns relate to energy intake and whether appetite-related hormones mediate this relationship.

Thirty healthy young women (age: 25 ± 4 years; BMI: 21.4 ± 2.0 kg/m²) arrived at the laboratory at 8:00 AM after an overnight fast. Upon arrival, participants completed a 30 min resting period, during which the cannula was inserted, and all fasted glucose measurements were collected. Glucose was measured every 5 min using FreeStyle Libre 2™ continuous glucose monitors (CGM), and one venous plasma glucose (VPG) sample was obtained immediately after cannula insertion. Following the 30 min fasting measurement period, participants consumed a fixed breakfast and then remained in the laboratory for an additional 240 min. Glucose was monitored via CGM every 5 min and VPG was measured every 15 min after breakfast. Energy intake was assessed at 240 min using an ad-libitum homogeneous pasta meal. Subjective appetite ratings were collected fasted and every 15 min after breakfast. Appetite-related hormones, including insulin, acylated ghrelin (AG), total glucagon-like peptide-1 (GLP-1) and oxyntomodulin (OXM) were measured in the fasted state, immediately after breakfast (t = 0 min), and subsequently at 30 min intervals after breakfast until t = 240 mins. Menstrual cycle phase was recorded and included as a covariate in all analyses. Associations between glycaemic variables and satiety or energy intake were examined using generalised linear models with a gamma distribution and log link function, adjusting for BMI and menstrual cycle phase. Bootstrap-based causal mediation analyses were conducted to evaluate whether appetite-related hormones mediated any significant associations between glycaemic responses and satiety or energy intake.

CGM-derived glucose nadir (lowest concentration) and dip (deviation of nadir from baseline as a percentage) were significantly associated with subsequent energy intake. A higher glucose nadir was associated with lower energy intake (β = -0.17, p = 0.003), whereas a greater glucose dip was associated with higher energy intake (β = 0.007; p < 0.001). When standardised, a one standard deviation (SD) increases in glucose nadir corresponded to an approximately 13% reduction in energy intake, while a one-SD increase in glucose dip corresponded to an approximately 16% increase in energy intake. Glucose nadir (CGM: β = 0.23, p = 0.041; VPG: β = 0.26, p = 0.008) and dip (CGM: β = -0.01, p = 0.006; VPG: β = -0.012, p < 0.001) derived from both CGM and VPG were also significantly associated with overall satiety measured immediately before the ad libitum lunch. A one-SD increase in glucose nadir was associated with higher satiety (CGM: 20% increase; VPG: 20% increase), whereas a one-SD increase in glucose dip was associated with lower satiety (CGM: 15% reduction; VPG: 17% reduction). Causal mediation analyses provided no evidence that insulin, acylated ghrelin, total GLP-1, or oxyntomodulin mediated these associations (p ≥ 0.09).

Glucose dynamics including nadir and dip may influence satiety and subsequent energy intake. These data support for a role of glucose in models of appetite regulation.