When testing the effects of different diets, it helps to ensure that the diets are truly different

Share This Post

Over the last few years there has been intense debate in nutrition circles about what the ideal diet is for weight loss. There are lots of different dietary approaches that may be advocated, but the most contentious debate has been about whether a diet low in fat, or one low is carbohydrate, is best if the goal is weight loss. There are now quite a few studies that have pitted low fat against low carb regimes. I’m not aware of any that show low fat to be superior in terms of weight loss. There are, however, several studies which have found low carb outperformed low fat.

Some of you may have read about a study published last week which was published in the New England Journal of Medicine that is said to prove that for weight loss, it’s not dietary composition, but merely calorie intake, that’s important. The study tested four diets of varying composition over a 2-year period. The diets varied in terms of their content of fat, protein and carbohydrate [1].

The percentage if calories contributed by fat/protein/carbohydrate for the four diets was as follows:

Diet 1. 20/15/65
Diet 2. 20/25/55
Diet 3. 40/15/45
Diet 4 40/25/35

At the end of 2 years, there was no significant difference in weight loss between the groups, and average weight loss was 4 kg (9 lbs). The conclusion of this study is that Reduced-calorie diets result in clinically meaningful weight loss regardless of which macronutrients they emphasize. This study has been used to vindicate the conventional line that a calorie is a calorie: it doesn’t matter what form calories take, their impact on weight is the same.

But is this really so. For start, while study investigators and their assistants may prescribe specific diets to the participants, there is no assurance that they’ll stick to them. Compliance issues are common in dietary study, and so the results of such an experiment are ideally assessed not on the basis of the prescribed diet, but the diets that were actually eaten.

Here is the actual nutritional make-up of the diets eaten by the individuals in the four groups when they were assessed at the end of the study (percentages have been rounded to nearest whole number):

Diet 1. 27/20/53
Diet 2. 28/21/51
Diet 3. 33/20/49
Diet 4. 35/21/43

It’s clear from these figures that the diets were, overall, much more similar in make-up than the prescribed diets. For example, the percentage of calories contributed by carbohydrate in diets 1 and 4 should have been 30 per cent (63 minus 35 percent). In reality, the difference was 10 per cent. Those eating diet 1 should have been eating almost twice as much carbohydrate as those eating diet 4. In reality, they ended up eating 23 per cent more carb.

Because the diets were so similar in composition, it seems inappropriate to use it to vindicate the ‘calorie is a calorie’ concept.

And what this study certainly can’t be used for (though some have tried) is to suggest that low-carb diets have no particular merit.

One reason for this is that diet 4 – the ‘low-carb’ diet ” was most certainly not low in carb. Individuals in this group were getting about 43 per cent of their calories from carb. Let’s for argument sake say a low-carb diet contains no more than 40 g of carb a day (the induction phase of the Atkins diet allows just half this amount). This 40 g equates to about 10 per cent of calorie intake. That’s less than a quarter of the percentage of calories contributed by carb in diet 4.

Also, those eating diet 4 were consuming, on average, 152 grams of carbohydrate a day (almost four times as much as someone eating a true low-carb diet containing 40 g of carb a day).

While the authors of this study mention low-carb diets several times, it seems they were not necessarily keen to test the effectiveness of such a diet. If they had been, then they would have done well to include a diet that was actually low in carb.

The other thing to consider is that while weight loss may indeed be a laudable aim, it is perhaps worthwhile considering that different diets have been shown to have different effects on biomarkers of illness and health including blood fat levels. Here, low-carb diets have been found to be generally superior to low fat ones .

Also, there is emerging evidence that protein-rich diets can promote weight loss in a way that cannot be explained by mere calories. A year ago I blogged about a study which pitted a high protein diet against a high carb one in women. The effects on weight were essentially the same. But, as is often the way, there had been compliance issues in this study. Analysis revealed that the more protein women ate, the more weight they lost. If you take a look at the figures from the NEJM study again, you’ll see that individuals in the four diet groups ended up eating, essentially, the same amount of protein (20/21 per cent of calories in all four groups).

So, my advice to the authors of the recent NEJM study is this: if you want to test the effect of different diets on weight or any other aspect of health, then it helps to make sure that these diets are truly different.


1. Sacks FM, et al Comparison of Weight-Loss Diets with Different Compositions of Fat, Protein, and Carbohydrates. NEJM 9;360:859-873

More To Explore

Walking versus running

I recently read an interesting editorial in the Journal of American College of Cardiology about the relative benefits of walking and running [1]. The editorial

We uses cookies to improve your experience.