Low-carbohydrate vs. Low-fat diets for Weight Loss: New Evidence

A new high-profile study compared the weight loss and cardiovascular effects of a low-carbohydrate diet vs. a low-fat diet.  Although many studies have done this before, this one is novel enough to add to our current understanding of diet and health.  Unlike most other studies of this nature, diet adherence was fairly good, and carbohydrate restriction produced greater weight loss and cardiovascular risk factor improvements than fat restriction at the one-year mark.  Yet like previous studies, neither diet produced very impressive results.

The Study

Lydia A. Bazzano and colleagues at Tulane University randomly assigned 148 obese men and women without cardiovascular disease into two groups (1):

1. Received instructions to eat less than 40 grams of carbohydrate per day, plus one low-carbohydrate meal replacement per day.  No specific advice to alter calorie intake.  Met regularly with dietitians to explain the dietary changes and maintain motivation.
2. Received instructions to eat less than 30 percent of calories from fat, less than 7 percent of calories as saturated fat, and 55 percent of calories from carbohydrate, plus one low-fat meal replacement per day.  No specific advice to alter calorie intake.  This is based on NCEP guidelines, which are actually designed for cardiovascular risk reduction and not weight loss.  Met regularly with dietitians to explain the dietary changes and maintain motivation.

Participants were followed up for one year, with data reported for 3 month, 6 month, and 12 month timepoints.  This study actually measured body fat percentage, but unfortunately did so using bioelectrical impedance (like on some bathroom scales), which is essentially meaningless in this context.

Results

About 20 percent of people in each group dropped out before the end of the study.  Over the course of a year, they followed their respective diets well enough to make the study informative.

The low-carbohydrate group didn't achieve the target of less than 40 grams of carbohydrate a day, however participants did reduce carbohydrate intake substantially, from 242 g at baseline to 97, 93, and 127 g at 3, 6, and 12 months, respectively.  Even at the 12-month timepoint, this group was only eating about half its initial carbohydrate intake.  This is a low-carb diet any way you slice it, although not "very low carb" or particularly ketogenic.  This relatively high level of diet adherence sets this study apart from most other low-carb studies.

The low-fat diet group reduced fat intake from 81 g per day at baseline to 45, 46 and 52 g per day at 3, 6, and 12 months, respectively.  This qualifies as a moderately low-fat diet, although not "very low fat".

What happened to body weight and fatness?  At one year, body weight was down by 12 lbs (5.3 kg) in the low-carb group and 4 lbs (1.8 kg) in the low-fat group, with the difference being statistically significant.  This is notable because most studies show no difference in weight loss between these two diets at 1-2 year timepoints.  This could be because the low-fat diet group didn't receive calorie advice, because the low-fat diet wasn't specifically designed for weight loss, and/or because diet adherence was fairly good in the LC group.  It's also notable that both groups maintained their weight loss fairly well over time, which also isn't typical.  This is probably because the daily meal replacement that was provided facilitated a higher level of diet adherence over time.

Waist circumference decreased by about the same amount in both groups, indicating similar reductions in harmful abdominal fat.  The change in body composition they reported was tiny, but again the method they used to measure it is unreliable.

Improvements in cardiovascular risk factors were greater in the low-carb group.  Total cholesterol and LDL ("bad") cholesterol remained the same in both groups, despite substantial differences in saturated fat intake.  HDL ("good") cholesterol increased in the low-carb group, and the ratio of total:HDL cholesterol consequently improved.  C-reactive protein, an inflammatory risk factor, was lower on the low-carb diet.  The Framingham Risk Score, a measure used to estimate the overall risk of having a heart attack over the next 10 years, improved on the low-carb diet and didn't change on the low-fat diet.  The FRS is based on age, total cholesterol, HDL cholesterol, smoking status, and blood pressure.

Interestingly, there were no differences in fasting glucose or insulin between groups, and if anything there was a trend for both to be lower in the low-fat group.

Summary

The low-fat diet had modest beneficial effects on weight and health.  The low-carb diet caused greater weight loss that was fairly well sustained over a one-year follow-up, however the degree of weight loss was modest, amounting to a 5.5 percent reduction in body weight.  This is important because the low-carb diet group reduced carbohydrate intake by 48-60 percent and sustained it for a year.  

The low-carb diet did improve estimated cardiovascular risk somewhat, but the effect was modest.  

Perspective

The low-carb vs. low-fat weight loss debate is getting somewhat stale, but this study was novel enough to warrant a post.  This study is one of the minority of trials that demonstrated a long-term (1-2 year) weight loss advantage for a low-carb diet over a low-fat diet, however it's consistent with all previous studies suggesting that neither approach is very effective for fat loss in the long term.  Both diet groups went from obese to slightly less obese.

I view the low-carbohydrate diet as a tool in the fat loss toolbox, but one that's most effective as part of a broader overall fat loss strategy.  In the Ideal Weight Program, both of our fat loss diets are lower in carbohydrate and high in protein.  Our approach is to incorporate multiple diet and lifestyle factors that have been shown to impact food intake and body weight, rather than limiting ourselves to a single-factor intervention.

However, this study does have a substantial silver lining.  Both diet groups, but particularly the low-carb group, were able to sustain their weight loss over a one-year period.  This is in contrast to most other diet studies.  I've always wondered how much of this regain is due to decreasing diet adherence over time, and how much is due to physiological and behavioral adaptation to the diets.  Adherence was pretty good in this study, probably due to the fact that the investigators provided daily meal replacements.  This argues that weight regain over time may have more to do with relaxed adherence than to adaptation, which is good news.

This study also adds to the evidence that low-fat high-carbohydrate diets can cause weight loss.  Even though the degree of weight loss is very modest, and possibly not significant from a clinical standpoint, this further undermines the argument that the carbohydrate-centric USDA dietary recommendations caused the obesity epidemic.  It turns out, when you put people on a diet that's similar to the USDA guidelines, they don't generally gain weight, and they often lose a little bit.

A key unanswered question is whether low-carbohydrate diets are compatible with cardiovascular health.  This study adds to the evidence suggesting that they are, although we still need long-term studies with hard cardiovascular outcomes to be certain.

A key virtue of low-carbohydrate and low-fat diets is that they're easy for people to understand.  "Carbs/fats are bad; avoid them" is a simple heuristic that's easily internalized and put into practice.  This advice can lead to weight loss and apparent improvements in cardiovascular health, and in some individuals it can be quite effective, but it's not a silver bullet.