Abstract
BACKGROUND It has been suggested that low-energy sweeteners (LES) may be associated with an increased risk of metabolic diseases, possibly due to stimulation of glucose-responsive mechanisms. OBJECTIVE We conducted a systematic review and meta-analysis of human intervention studies examining the acute effect of LES intake on postprandial glucose (PPG) and postprandial insulin (PPI) responses, in order to comprehensively and objectively quantify these relations. METHODS We systematically searched the Medline, OVID FSTA, and SCOPUS databases until January 2020. Randomized controlled trials comparing acute postprandial effects on PPG and/or PPI after exposure to LES, either alone, with a meal, or with other nutrient-containing preloads to the same intervention without LES were eligible for inclusion. PPG and PPI responses were calculated as mean incremental area under the curve divided by time. Meta-analyses were performed using random effects models with inverse variance weighing. RESULTS Twenty-six papers (34 PPG trials and 29 PPI trials) were included. There were no reports of statistically significant differences in the effects of LES on PPG and PPI responses compared with control interventions. Pooled effects of LES intake on the mean change difference in PPG and PPI were -0.02 mmol/L (95% CI: -0.09, 0.05) and -2.39 pmol/L (95% CI: -11.83, 7.05), respectively. The results did not appreciably differ by the type or dose of LES consumed, cointervention type, or fasting glucose and insulin levels. Among patients with type 2 diabetes, the mean change difference indicated a smaller PPG response after exposure to LES compared with the control (-0.3 mmol/L; 95% CI: -0.53, -0.07). CONCLUSIONS Ingestion of LES, administered alone or in combination with a nutrient-containing preload, has no acute effects on the mean change in postprandial glycemic or insulinemic responses compared with a control intervention. Apart from a small beneficial effect on PPG (-0.3 mmol/L) in studies enrolling patients with type 2 diabetes, the effects did not differ by type or dose of LES, or fasting glucose or insulin levels. This review and meta-analysis was registered at PROSPERO as CRD42018099608.
