A Low Carbohydrate Diet Is More Effective for Reducing Weight, % Body Fat and HbA1c Than Healthy Eating in Both Diabetic and Non-Diabetic Subjects
Abstract Number: 2574-PO
Authors: PAMELA A. DYSON, SUE BEATTY, DAVID R. MATTHEWS.
Institutions: Oxford, Oxfordshire, United Kingdom.
Results: Background: There is some evidence that low carbohydrate diets show significant weight loss and improvement in glycaemia in people with type 2 diabetes, but studies to date tend to be short term, lack a control group or include small numbers.
Aim: To assess the impact of a low carbohydrate diet on body weight, % body fat, HbA1c and lipid profiles in diabetic and non-diabetic subjects.
Subjects and methods: In a randomised clinical trial type 2 diabetic subjects on diet or Metformin (n=13) and non-diabetic subjects (n=14) were allocated either a low carbohydrate diet (LC) (<40g carbohydrate/day) or a low fat healthy eating diet (HE) and were seen monthly for 3 months. Baseline characteristics of subjects (25% male) were mean (SD) age 52 (9) years, weight 96.3 (16.6) kg, BMI 35.1 (7) kg/m2, HbA1c 6.6 (1.1)%, body fat 45.5 (8.5)%, total cholesterol 5.1 (1.1) mmol/l, HDL cholesterol 1.3 (0.4) mmol/l, triglycerides 1.5 (0.7) mmol/l and LDL cholesterol 3.1 (0.9) mmol/l.
Results: Analysis was by intention to treat with last observation carried forward. 20 subjects (77%) completed the study. Weight loss was greater (6.9 v 1.6 kg, p=0.0003) and change in HbA1c levels was larger (-0.3 v -0.1%, p=0.01) in the LC group with no difference in lipid profile. Weight loss was explained by a greater change in % body fat (-1.5 v -0.2, p=0.03) in the LC group and the reduction in HbA1c was concordant with changes in fasting plasma glucose (-0.8 v -0.3mmol/l, ns). Two hour plasma glucose levels by OGTT (-0.6 v -0.4mmol/l, ns) and changes in beta cell function (-13.3 v -10.6%, ns) and insulin sensitivity (35.6 v 13.7%, ns) assessed by HOMA showed no difference between the groups.
Conclusions: Weight loss, reductions in HbA1c and % body fat were greater in subjects on a low carbohydrate diet with no difference observed in lipid profiles.