In recent years, one of the most popular diet interventions has been fasting, in a variety of different forms. These have included intermittent fasting diets in which you’re supposed to fast for a few days per week, such as the 5:2 diet, or a few days per month. They’ve also included various forms of time restricted eating, such as the 16:8 diet, where you’re supposed to get all your calories within an eight hour window each day, and the more extreme warrior diet, in which you’re supposed to get all your calories in a four hour window. But there is still little clarity on how effective these modifications are in terms of weight loss. And up to this point, pretty much all the evidence in support of fasting comes from animal studies, which are notoriously unreliable.
A study has just been published in JAMA (the Journal of the American Medical Association) that looks at the 16:8 diet in particular. It recruited 141 people and followed them over a twelve week period, during which they were randomized to either time restricted eating in the form of a 16:8 diet or a control diet. The authors of the study have financial ties to companies that promote ketogenic diets and fasting, so they were probably hoping for a positive study, which is something to be aware of before we get in to the data.
The 16:8 group was supposed to consume all its calories between 12:00 pm and 20:00 pm, and was not allowed to consume anything containing calories outside this window. The control group was instructed to eat three square meals per day, but was also allowed to snack as much as desired between meals (presumably the reason the researchers didn’t just tell them to do whatever they wanted was so they wouldn’t realize that they were in the control group).
Neither group had any restrictions placed on what they could eat or on the total number of calories they were allowed to consume, which is good, because we want to know if the time restriction in itself results in weight loss, even without further restrictions.
All participants received a bluetooth connected scale with which they were supposed to weigh themselves at home on a daily basis, and they downloaded an app which provided daily reminders in order to help them stick with their prescribed diet. They also filled in whether they were successfully following the diet on a daily basis in the app.
Of 141 randomized participants, 105 followed through to the end of the 12 week intervention. Of the 36 individuals who dropped out, 25 never took part in the study in any way other than signing up for it. Since no weight measurements were ever recorded for these 25 individuals, they weren’t included in the subsequent analysis, which I think is reasonable. So there were actually 116 individuals who provided at least one weight measurement, and were included in the analysis.
Of the 116 participants, 59 were randomized to time restricted eating, while 57 were randomized to the control group. 40% of participants were female, and 60% were male. The average age of the participants was 47 years. The average weight was 99 kilograms (218 pounds) and the average BMI (Body Mass Index) was 33 (the cut-off for overweight is 25, and for obesity is 30).
So, what were the results?
Participants in the control group adhered to the prescribed diet of three solid meals per day (with freedom to snack whenever desired) 92% of the time. Participants in the time restricted eating group adhered to the 16:8 requirement 84% of the time. It makes sense that the time restricted group would have a harder time sticking to their diet than the control group, since they were being expected to fast for a significant portion of the day. But 83% adherence is still pretty good, suggesting that most people are able to stick with a 16 hour fast most of the time without too much difficulty, at least over a 12 week period.
In the control group, the average weight loss over 12 weeks was 0,68 kilograms (1,5 pounds). In the time restricted eating group, the average weight loss was 0,94 kilograms (2,1 pounds). This was an unimpressive 0,26 kilogram (0,6 pound) difference in favor of time restricted eating. And the difference wasn’t anywhere close to being statistically significant.
Unfortunately, things gets worse. The researchers gathered a lot of additional data from a subgroup of 50 participants (25 from each group) that lived close enough to the research lab in San Francisco to come and have measurements taken in person. One of the things they looked at was muscle mass in the arms and legs (a measure known as “appendicular lean mass”).
In the sub-groups, the control group lost 0,17 kilograms (0,4 pounds) of arm and leg muscle out of a total weight loss of 0,6 kilograms (1,3 pounds). The time restricted eating group lost 0,64 kilograms (1,4 pounds) of arm and leg muscle out of a total weight loss of 1,7 kilograms (3,7 pounds). While the total weight loss was not statistically significant between the groups, there was a statistically significant greater loss of muscle mass in the time restricted eating group.
That’s not good. Of the small amount of weight lost in the time restricted eating group, 37% was muscle in the arms and legs. In the control group, the proportion of weight loss that was was arm and leg muscle was 28% .
So, what conclusions can we draw?
The 16:8 diet does not seem to work as a method for losing weight. The weight loss over twelve weeks was marginal. Of course, there is a big weakness with this study, and that is that it only followed participants for twelve weeks. It is possible that a statistically significant difference would have been seen between time restricted eating and control if the study had gone on for a full year.
However, people who start a new diet usually see the biggest effect in the first few months. Even if we were to assume a linear weight loss over the course of a full year, the difference between the time restricted eating group and the control group would still only be 1,1 kilograms (2.4 pounds), which is tiny.
And what is worse is that that an unusually large proportion of the weight loss in the time restricted eating group came from muscle mass, and in particular skeletal muscle in the arms and legs. These muscles are critical to our ability to stay independent as we age, and we definitely don’t want to lose them any faster than necessary.
Of course, this is only one study, and a hundred odd patients isn’t that many, and they weren’t followed for very long, so we shouldn’t draw too far reaching conclusions. But the evidence from this study is going in the wrong direction when it comes to both weight loss and muscle loss. Based on that, I wouldn’t recommend time restricted eating to anyone.
As mentioned earlier, the authors had ties to companies that promote ketogenic diets and fasting. Based on that information, I would assume that they wanted the study to show a positive effect. If they wanted a positive study and got a negative one, then that suggests we can trust these results.
It’s possible that an even more extreme restriction, like 20:4 (“the warrior diet”) would still work, and I hope that is investigated in future studies, just as I hope we will see more studies on fasting in general and for longer time periods, so that we can truly find out what works and what doesn’t. When it comes to fasting interventions, there is a lot of guesswork at the moment, and a lot of recommendations are made based on weak evidence. It would be great if that could change in the near future.
You might also enjoy my article about whether exercise is an effective way to lose weight or my article about whether a low carb or low fat diet is better for weight loss.
If you haven’t already done so, I strongly recommend that you read my guide to scientific method in medical and health science.
Hi Sebastian, thanks for the post. I was wondering if any difference was observed in glycemic control between the groups? Or perhaps other studies looking at the effect of time-restricted eating on glycemic control. Thanks.
Hi Pedro, there were no statistically significant differences in fasting glucose or insulin between the groups.
I looked in vain for a credible hypothesis explaining WHY this diet would cause weight loss. Articles mention depletion of glucagon stores…but so what? The best I can come up with is that it stops you snacking for most of the day. On the other hand, there are credible (though not necessarily true) theories about the diet’s other supposed benefits, such as its effects on the immune system.
Is there a reasonable hypothesis/theory to explain how fat loss occurs or is it all anecdotal?
I am always suspicious of claims made purely on the basis of counting numbers, without an explanatory mechanism.
Good question. I think the main hypothesis is that it prevents you snacking partly by restricting when you can eat and partly by decreasing the number of peaks and troughs in blood sugar. But the underlying hypothesis is pretty moot since it doesn’t seem to work!
I feel that any ‘diet’ (which means rule) for losing weight is as good as a person will follow its rules.
[ I worked very closely with the best diabetes doctors in town ].
Just as any addict will cease to be an addict if it really wants to leave addiction.
When 40 years old, I loose 8kg in 3 months with the first Atkins scheme. My lipid profile changed remarkably: all markers went to very normal values.
Sebastian thanks for the info on this study. I have been doing a time restricted regiment of fasting 16:8 for over two years now along with regular exercise like walking and weightlifting. I feel very good and have list about 20 pounds and should lose about 10 more. I have found that body fat loss using this method takes quite a bit of time.
My starting premise was that my body fat was mostly a result of eating food that was unhealthy. In my case to many carbs and not enough protein and veg. The result being not only body fat gain but metabolic illness. Keto eating is first a method to heal your body of metabolic illness. Then body fat loss follows. Excessive body fat is a symptom of general metobolic ill-health.
So many people in the western world are metabolically sick. I believe there is a link to the initial slow body fat loss and the fact that keto eating and fasting initially starts to heal the metobolic systems of the body and then after some time body fat is diminished.
It seems to me that all that this study shows is that TRE does not work for those eating a normal Standard American Diet. IOW, cramming all that carbage into an eight hour window instead of the normal sixteen hours, would not be expected to result in weight loss. Why is anyone surprised by this?
Hi Garth,
A lot of the claims made around fasting are that you can keep eating whatever you want as long as you do some form of fasting.
And if we did a study showing that a combination of time restriction and diet modification worked, we would still need to do follow-up studies to see whether the benefit is coming from one or the other or both together.
This study shows that absent other interventions, 16:8 doesn’t work. If a future study shows that 16:8 + diet modification works better than diet modification alone, then we should recommend it. Until then the evidence does not support it.
Agreed. Lets hope the future studies that you propose become a reality.
Thank you very much, Dr. Rushworth. Chris Kresser also posted this study a few days ago, so I’ve been thinking about it. I’ve been doing time-restricted eating for two or three years. About three months ago I switched to a mostly carnivore diet, and have noticed much improvement in sleep quality and overall feeling of well being. In recent weeks, though, I’ve begun to feel hypoglycemic in the morning. I do strenuous muscle-building exercise three or four days a week (formerly a long-distance runner), and I’ve always exercised in a fasted state, but I can no longer do that. On my most strenuous days I’ve begun to eat three meals. I certainly don’t want to lose muscle mass! I’ve spent the last seven years rebuilding the muscle mass I lost between young adulthood and age 65. Chris Kresser suggest those of us in my age group may need more protein, so I’ve added whey protein. Life is a life-long experiment.
It’s a wonderful article based on logic and evidence. Having said that, I would like to draw your attention to a very popular movement started by a doctor in the state of Maharashtra in India which advocates only two times eating during the window of 55 minutes each. This prevents insulin secretion at other times which would have converted in to fat and blood sugar. Thus it’s a form of intermittent fasting. And the results on very large sample of people is very successful in both weight loss as well as reduction in HbA1c.
What is your opinion on the above phenomenon?
Hi Suhas,
Without a randomized controlled trial, I can’t really comment, but I would love to see one.
The Devil is always in the ‘details’ and Comments have fleshed them out nicely. IMHO, that Study was a success. It proved that ‘time’ alone was not a strong significant factor, now focus is firmly on food choice / balance such as Carnivore cf Vegetarian styles.
Further Studies are indicated !!!
James DownUnder: Right you are. Only requires finding funding, which may be difficult. I suspect that, just as proper grazing patterns for ruminants in food production requires mimicking Nature, as Alan Savory puts it, looking at patterns of food consumption by our ancestors can inform what is metabolically healthiest for us. Surely we have evolved under times and plenty and times of scarcity, and our physiology is adapted to this. I miss joining the conversation on Dr. Kendrick’s blog, if only to say thanks, but my comments no longer post.
As per usual, great article, thanks for adding to the investigation. It is increasingly apparent that it is not just what we eat, but also the way we eat that is also important for maintaining good health over one’s lifetime.
Sebastian, I hope you are comfortable with me sharing some of my ideas with you and your readers?
I’ve been using the COVID Lockdowns of this year in NZ to do some personal research and experimentation, then written a series of articles on the topic which began a year back. While these articles have an “athlete” slant, the ideas in them apply equally well to anyone, including the sedentary and those struggling with chronic ill health.
Here is one lot: https://www.garymoller.com/blog-1/search/keto
And the other: https://www.garymoller.com/blog-1/search/intermittent
I have a request: I’m developing a dislike of the terms “keto” and “ketosis” because these typically refer to an extreme diet that few can sustain and which may have some health risks such as constipation. I’ve been using the terms “intermittent fasting” and “gentle keto” instead.
I like the idea of returning to the best of the way we used to eat before the influence of industrialisation.
“Time restricted eating” may be a better description. Can readers suggest a better term?
The study discussed here had a design that was guaranteed to make the time restricted eating regimen fail. Having the window from noon to 8:00 pm had people eating too late in the day. Late-in-the-day eating has been shown to have harmful metabolic effects as well as in promoting weight gain.
This video discussion between Drs. Satchin Panda and Rhonda Patrick provides a glimpse into how important circadian rhythms and chrono-biology are with respect to diet:
https://youtu.be/QKicol7566g
A follow-up article on the importance of taking into account circadian rhythm would be quite worthwhile.
Nick: Thank you very much for posting that conversation. Much food for thought.
Hello, thank you for your interesting post. I actually thought that the metabolism of ketone bodies and autophagy (what is thought to happen after hours of fasting) promotes health and weight loss? This would make sense taking into account that our ancestors were not constantly eating like the humans do it nowadays.
I think there is a lot to be said for fasting, but as yet no-one knows exactly how long you need to fast for, or how often, to gain the supposed benefits. Hopefully there will be a lot more research coming out over the next few years.