Background: Low-carbohydrate diets decrease hemoglobin A1c among patients with type 2 diabetes mellitus. However, evidence on effects of low-carbohydrate diets on hemoglobin A1c among individuals with prediabetes is limited.

Methods: We conducted a randomized parallel-group trial of 150 adults aged 40-70 years with untreated hemoglobin A1c from 6.0-6.9%. Participants were assigned to a healthy low-carbohydrate diet (target <40 grams net carbohydrates months 1-3; <60 grams months 3-6) or usual diet. The low-carbohydrate arm received regular dietary counseling. Data on hemoglobin A1c and other metabolic risk factors were collected at 0, 3, and 6 months.

Results: Mean age was 59 years; 72% were female and 59% were Black. 95% completed the study. Participants in the low-carbohydrate arm had significantly greater 6-month decrease in hemoglobin A1c (net difference, -0.23% [95% CI, -0.32 to -0.14%]) , fasting plasma glucose (net difference, -10.3 mg/dL [95% CI, -15.7 to -4.9 mg/dL]) , and body weight (net difference, -5.9 kg [95% CI, -7.2 to -4.7 kg]) compared with usual diet arm. At 6 months, 53% in the low-carbohydrate arm and 32% in the usual diet arm had hemoglobin A1c <6.0% (p=0.01) .

Conclusion: A healthy low-carbohydrate diet led to improvements in glycemic and other cardiometabolic outcomes in adults with elevated hemoglobin A1c and could be recommended for prevention of type 2 diabetes mellitus.

Disclosure

K.S.Dorans: None. M.J.O'brien: None. G.I.Uwaifo: None. J.He: None. L.A.Bazzano: None. L.Qi: None. H.He: None. J.Chen: None. L.J.Appel: n/a. F.Hu: None. J.Samet: None. K.T.Mills: None.

Funding

National Institute of General Medical Sciences (1P20GM109036-01A1) . The California Walnut Commission (Folsom, CA) supported the study by providing walnuts. Shakes and bars were donated by The Simply Good Foods Company (Denver, CO) . Sugar replacement samples were provided by Swerve LLC (New Orleans, LA) .

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.