Associations of low-carbohydrate and low-fat intakes with all-cause mortality in subjects with prediabetes with and without insulin resistance.

Summary Background & aims We investigated the associations of low-carbohydrate and low-fat diets with all-cause mortality in people with prediabetes according to insulin resistance status using data from the National Health and Nutrition Examination Survey (NHANES). Methods We analyzed the NHANES participants with prediabetes from 2005 to 2008, and their vital status was linked to the National Death Index through the end of 2011. Low-carbohydrate and low-fat diets were defined as ≦40% and ≦30% of calories from carbohydrate and fat, respectively. The homeostasis model assessment of insulin resistance (HOMA-IR) was used to determine insulin resistance. Weighted Cox proportional hazards regression models were used to compare the hazard ratios for the associations of low-carbohydrate and low-fat diets with all-cause mortality. Results Among the 1687 participants with prediabetes, 96 of them had died after a median follow-up of 4.5 years. Participants with a HOMA-IR >3.0 had an increase in all-cause mortality compared with those who had a HOMA-IR ≦3.0 (HR 1.797, 95% CI 1.110 to 2.909, p = 0.019). Participants with ≦40% of calories from carbohydrate and >30% from fat (3.75 per 1000 person-years) had a lower all-cause mortality rate compared with those who had >40% from carbohydrate and >30% from fat (10.20 per 1000 person-years) or >40% from carbohydrate and ≦30% from fat (8.09 per 1000 person-years), with statistical significance observed in those who had a HOMA-IR ≦3.0. Conclusions A low-carbohydrate intake (≦40%) was associated with a lower all-cause mortality rate in people with prediabetes.