I don't know if you noticed, but a new study on intermittent fasting recently made headlines in several media. The paper basically found that alternate-day fasting (or "on-off fasting" as it was named in some articles) helped obese adults to lose weight.
That's not a huge surprise, really. If you're obese, it means you eat too much of the wrong foods and probably too often. Fasting every other day means you'll at least be eating them less often, if not less per se. However, the conclusions that have been drawn in various articles from the study seemed a little suspicious to me, so I decided to read the whole paper.
Indeed, the authors themselves appear to be somewhat confused as to what really caused the weight loss. So without further due, let's take a closer look at what the study actually found (link).
Study design
12 obese women and 4 obese men were recruited for the study. Mean age was 46 years, mean body weight was 96.8 kg, and mean BMI 33.8. Not exactly the featherweight league.
The study consisted of three phases. The first one was a 2-week control phase, during which the subjects were told to maintain their usual weight by eating and exercising as they normally would. In the second phase, which lasted for 4 weeks, all participants were given a standard menu containing 25% of their baseline energy needs on the fast days. On the feeding days they could eat as much as they wanted. The third phase, also 4 weeks, was similar to the second one. The only difference was that the participants could choose what they wanted to eat on their fasting days, as long as they only ate 25% of their baseline needs.
So, the first thing that sets this study apart from how most people do intermittent fasting is that they didn't consume zero calories during their fast. They just ate significantly less. The second thing is that the fasts began and ended at midnight. Most people (including me) start and end their fast sometime during the day, because it allows them to eat at least once a day. If the participants went to sleep before midnight, their "fasts" were significantly longer than 24 hours.
Their standard diets were also less than optimal in my opinion. Things like chicken fettucini, vegetarian pizza, chicken enchilada, cookies, and crackers aren't exactly paleolithic foods consistent with the idea of intermittent fasting. But then, this wasn't a paleo study, which explains why they were fed high-carb, low-fat foods. For the third phase, they were given diet tips by a registered dietitian:
On the ad libitum food intake day, subjects were instructed to limit fat intake to <30%> dairy options.
So more carbs and less fat once again. They probably took some of this advice and applied for their feeding days as well, which makes me wonder if they wouldn't have lost even more weight had they opted for low-carb foods instead. Still, as you can see from the results below, they did manage to lose weight even with this diet.
Results
During the first phase, there was no weight loss. This is unsurprising, since all the participants just kept on eating whatever made them obese in the first place. During the eight weeks of alternate-day fasting they did lose weight, however.
In the second and third phase the subjects lost weight at a rate of ~0.7 kg per week. At the end of the study, they had lost about 5.6 kg, most of which was fat. Mean BMI decreased from 33.7 to 29.9, while body fat percentage dropped from ~45 to ~42%.
Cholesterol levels were also reduced as a result of alternate-day fasting. HDL remained the same, but LDL decreased by almost 25%. This to me is a more impressive result than the weight loss, which I think could've been greater with proper food choices.
Systolic blood pressure was lowered by 4.4 mm Hg, but diastolic blood pressure remained the same. Heart rates varied throughout the study, but at the end, they were about 4 beats per minute lower than in the beginning.
Discussion
Okay, so everyone lost some weight and improved their LDL/HDL ratio on this modified version of intermittent fasting. But what exactly is behind these results? The first thing that came to my mind as I was reading the paper was: how much were these people eating on their feeding days? Surely that would have a drastic effect on weight loss.
Unfortunately, there's no mention of this in the paper. Yes, they were told they could eat ad libitum, but apparently they were not told to keep a record of what or how much calories they ate when they were not fasting. I know food diaries are generally unreliable (people tend to underestimate how much they eat), but it would've been better than nothing. The authors seem surprised that the subjects lost as much weight as they did:
We predicted that subjects would lose a total of 4.5 kg fat mass after 8 wk (on the basis of a 75% decrease in energy intake on the fast day, with no change in energy intake on the feed day). The actual fat mass lost (5.4 kg) exceeded our predictions.
With no change in energy intake on the feed day? I'm not sure where the authors got the idea that when you eat very little on one day, the next day you won't be extra hungry and compensate. That's certainly not my experience. On the contrary, I fully compensate for any missed calories by just eating twice as much the next day. Certainly not compensating can be done if one wants to, which may be what happened here:
This indicates that these subjects were also limiting their energy intake on the feed day, which may have occurred because the subjects knew they were enrolled in a weight loss trial.
On the other hand, some articles in the media (such as this one) have reported that there was a slight compensation going on:
On non-fasting days people typically ate between 100 percent or 125 percent of their calorie needs.
This statement is probably from a press release by the authors, but the paper itself is silent on the issue. In any case, if you're eating 75% less on one day and only 25% more the next, you're still 50% short.
That, in effect, makes this a study on caloric restriction, not intermittent fasting. How is it surprising in any way that people who eat half (or even less) than what they're used to manage to lose weight? Isn't that obvious? The amount of weight they lost is pretty much what you'd expect on a low-fat, calorie-restricted diet. And based on other studies, if they'd eaten more calories but restricted carbohydrates, they'd have lost more weight.
The whole point of intermittent fasting is that you don't have to restrict your energy intake, you just don't eat all the time. If energy intake is the same and yet there are health benefits, then we can conclude that it's the fasting that is behind them. If, at the same time, calorie intake is restricted, there's no way of knowing whether it's the reduced calories or the fasting that is the cause. And of course, if intermittent fasting leads to a voluntary reduction in energy intake, that tells us something too.
I have a feeling that eating the small meal every other day magically resulted in no hunger in this study. They were probably consciously limiting their food intake even though they were told they could eat as much as they want. The fact that the authors actually expected the participants to eat only their usual calories on feeding days makes me even more suspicious.
Try eating 500 kcal on one day and then seeing how you feel the next day. If you typically eat 2,000 kcal, somehow I don't think 2,500 kcal is going to cut it. I know it doesn't for me, at least not in the long term. As for this diet being much easier to follow than old-school calorie restriction (which the paper seems to suggest), I doubt it.
For more information on intermittent fasting and caloric restriction, see these posts:
Intermittent Fasting: Understanding the Hunger Cycle
Slowing Down Aging with Intermittent Protein Restriction
Caloric Restriction Improves Memory in the Elderly
Intermittent Fasting Reduces Mitochondrial Damage and Lymphoma Incidence in Aged Mice
