By: Dr. Geoffrey Modest
The annals of int med published a small randomized controlled trial of a low-carb vs low-fat diet for weight loss in September (see doi:10.7326/M14-0180). 148 people 22-75 yo with BMI 30-45 but free of cardiovascular disease, diabetes, or kidney disease were recruited at Tulane University, with mean age 46.8 yo, 88% female, 51% African-American and 45% white. Average cholesterol 202, HDL 55 and LDL 123. only 30% on antihypertensives, 14% on lipid lowering agents, and median 10-yr Framingham risk score of 4%. Followed 12 months.
Details:
–73 in low-fat diet group (<30% of daily energy intake from fats, <7% from saturated fats, and 55% from carbs). 75 in low-carb group (digestible carbs – total carbs minus total fiber– less than 40g/d). Neither diet had a reduced-calorie goal and all patients were told to continue the same exercise they had been doing. All were given a handbook with recipes, sample menus, food lists, meal planners, as well as 1 meal replacement (bar or shake) per day. Weekly 1 hr individual visits to dietitian for 4 weeks, then small group counseling sessions every two weeks for next five months, then monthly for last 6 months. dietary recall questionnaires provided every 3 months. 80% completed the 12 month intervention.
–Total energy intake similar between groups (1527 kcal in low-fat, 1448 in low carb) with significant differences in food composition (but well above the targets) in carbs (198g in low-fat and 127 in low-carb) and fat (52g in low-fat and 69 in low-carb). Similar physical activity in the 2 groups.
–At 6 months (comparing low-fat to low-carb):
–Body wt dec 2.3 vs 5.6 kg, with diff of 3.2 kg (signif at <.001 level)
–Waist circumference dec 4 vs 5.9 cm (signif)
–HDL no change vs inc 0.10 mmol/L (3.86 mg/dl); chol/HDL ratio inc 0.07 vs dec 0.25; triglycerides dec 0.01 mmol/L (0.9 mg/dl) vs dec 0.22 (19.5 mg/dl). No signif change in total chol, LDL (though all trended to be better with low-carb diet)
–Framingham risk score inc 0.4 vs dec of 0.7 (signif)
–CRP (C-reactive protein) level inc 6.7 vs dec 4.8 (signif)
–At 12 months (comparing low-fat to low-carb):
–Body wt dec 1.8 vs 5.3 kg, with diff of 3.5 kg (signif)
–Waist circumf dec 5 vs 6.7 cm (nonsignif)
–HDL inc 0.06 mmol/L (2.31 mg/dl) vs inc 0.24 (9.27 mg/dl); chol/HDL ratio dec 0.05 vs dec 0.49; TG dec 0.07 mmol/L (6.2 mg/dl) vs dec 0.23 (20.4 mg/dl). No signif change in total chol, LDL (though all trended to be better with low-carb diet)
–Framingham risk score inc 0.4 vs dec 1.0 (signif)
–CRP level inc 8.6 vs dec 6.7 (signif)
–Only difference in adverse events was increase of headaches at 3 months in the low-fat group
So, pretty unusually healthy group to start with (given their exclusions, which limit generalizability), in one center in Louisiana (generalizable to people eating different, and probably less tasty foods in other geographical areas). Pluses of the study include that there was significant representation of African-Americans (unusual in these studies), the results were impressive after 12 months (most studies show decrease in efficacy of diet after first 6 months or so), the dietary changes actually made were hardly extreme (which may go along with the better adherence), the intervention was pretty easy/cheap, and the results are consistent with several other studies finding that low-carb (or low glycemic index diets) are associated with decreased appetite/less food intake than low-fat diets (even without calorie restriction), more profound weight loss (esp early on, related to the finding that low glycemic index diets depress resting and total energy expenditure less than low fat diets), decreased CRP levels, and improvement in chol/HDL ratio and TG. also more improvement in A1c in diabetics.