Background: Carbohydrate restriction may benefit β-cell function and glucose metabolism in type 2 diabetes (T2D) but also leads to weight loss which in itself is beneficial.

Methods: In order to determine the additional effect of carbohydrate restriction in addition to a fixed body weight loss, we randomly assigned 72 adults with T2D and obesity (mean ± SD HbA_1c 7.4 ± 0.7%, BMI 33 ± 5 kg/m²) to a carbohydrate-reduced high-protein diet (CRHP; energy percent from carbohydrate/protein/fat: 30/30/40) or an isocaloric conventional diabetes diet (CD; 50/17/33) for 6 weeks. All foods were provided free of charge and total energy intake was tailored individually, so both groups lost 6% of baseline body weight.

Results: Despite significantly greater reductions in HbA_1c (mean [95% CI] -1.9 [-3.5, -0.3] mmol/mol) after 6 weeks, the CRHP diet neither improved glucose tolerance, β-cell response to glucose, insulin sensitivity, during a 4-h oral glucose tolerance test, nor basal proinsulin secretion when compared to the CD diet, but increased C-peptide concentration and insulin secretion rate (area under the curve [AUC] and peak) significantly more (~10%, P ≤ 0.03 for all). Furthermore, compared with the CD diet, the CRHP diet borderline increased basal glucagon concentration (16 [-0.1, 34]%, P = 0.05), but decreased glucagon net AUC (-2.0 [-3.4, -0.6] mmol/L ×240 min, P < 0.01), decreased basal triglyceride and total AUC (~20%, P < 0.01 for both), and increased gastric inhibitory polypeptide total AUC (14%, P = 0.01).

Conclusion: A moderately carbohydrate-restricted diet for 6 weeks decreased HbA_1c but did not improve β-cell function or glucose tolerance beyond the effects of weight loss when compared with a conventional diabetes diet in people with T2D.

Clinical trials registration: www.Clinicaltrials.gov, Identifier: NCT02472951.