Finally, the first study on intermittent fasting and weight training has arrived. Thank God. I have been waiting since 2006 to find out. My body is ready. 

The title of the study is Time-restricted feeding in young men performing resistance training: A randomized controlled trialI previously mentioned it in Intermittent Fasting: Where Are We Now? and now that the paper is published, I will explore the results and lessons within.

In this study, researchers sought to determine the effects of weight training and time-restricted feeding (TRF), which is essentially another term for intermittent fasting, on nutrient intake body composition and strength.

Briefly, here’s how the study was done:

The researchers tracked two groups of males who trained and ate for 8 weeks. Both groups did an alternating upper/lower body split three times per week, using 4 sets of 8-12 reps to failure in movements like the bench press, lat pulldown and squat.

Their diet regimens differed radically on 4 out of 7 days. One group ate as usual – ND (Normal Diet) on all days of the week. The other group did TRF (time-restricted feeding) for 4 out of 7 days. They ate as usual on training days, just like the other group (ND), but restricted their food intake to a 4-hour window between 4 p.m. and midnight on their rest days.

ND-Group: Ate anything they wanted on all days.
TRF-Group: Ate anything they wanted on training days (3 days) and anything they wanted within a 4-hour window on rest days (4 days).

Participants reported their food intake via diet logs during the first, 4th and 8th week. These records were subsequently analysed to pinpoint calorie- and macronutrient intakes for each respective group, and it’s important to note that the researchers did not influence the diet in any way or form beyond the implementation of time-restrictions on the one group.

Beyond that, participants were free to eat whatever they wanted, which is to say that these folks basically followed the Standard American Diet throughout the process – think of how your non-lifting friends eat and you get the idea.

That said, how did the results compare? Let’s find out.


Unsurprisingly perhaps, participants in the TRF-group consumed 667 kcal less on fasting days compared to normal days. More surprising is the fact that they didn’t compensate this deficit by eating more on normal days; they ate 1631 kcal on fasting days and 2318 kcal on normal days, which for the average participant (87.4 kg) means that they were in marked caloric deficit on fasting/rest days and a slight deficit on normal/training days.

As the study goes on, average intake in TRF drops to 2207 and 1370 kcal in the 4th week, and then to 2150 and 1674 kcal in the 8th week…

*Ehum* Spoiler alert, I might revisit these numbers later.

In ND, average intake is 2642 kcal in the first week, 2715 kcal in the 4th week and 2106 kcal in the last week. You don’t need to be a rocket scientist to figure out that the average intake in ND was a lot higher than in TRF throughout the study. With such a long period of under eating, you’d expect significant weight loss in the TRF-group, and some weight gain in ND, but what actually happened?
Here are the results.

For the RT-TRF, per cent changes for individual participants ranged from −5.5% to +2.6% for body weight, −22.1% to +4.5% for fat mass, −4.0% to +4.6% for lean body mass, +4.4% to +22.7% for bench press 1-RM, and +13.7% to +48.1% for hip sled 1-RM.

For the RT-ND group, per cent changes ranged from −1.4% to +2.1% for body weight, −13.5% to +12.6% for fat mass, −2.5% to +3.9% for lean body mass, +4.7% to +12.2% for bench press 1-RM, and +13.6% to +31.5% for hip sled 1-RM.

To say that “substantial variability in outcomes was observed in both groups” is an understatement, because the numbers are all over the place. You got one guy losing a ton of fat (-22.1%) on TRF, yet another one gaining a bit (+4.5%). Another one gained a good chunk of muscle (+4.6%), while another one lost some (-4%), etc.

Performance wise, TRF clearly outperformed ND. Low-responders in both groups increased their bench press by merely 4% in both groups, with the highest number reaching +12.2% in ND, and an impressive +22.7% in TRF. Same goes for hip sled 1-RM. Both groups had their lowest increase at 13%, while TRF had its highest at 48.1% compared to 31.5% in ND.

Dietary analysis reveals that average calorie and carbohydrate intake on fasting days appeared to be related to the increase in hip sled 1-RM which is somewhat interesting…simply for the fact that it’s one of the few things that makes sense, because none of the rest really does:

There were no significant correlations between overall energy or macronutrient intake and body composition or strength changes in either group.

Let that sink in for a minute. Do you understand how strange that statement is? Based on the dietary analysis, the researchers could not find any relationship between what participants ate, how they performed or how it affected their body composition. Imagine if you had one group claiming to eat 4000 kcal and another one claiming to eat 1000 kcal, and yet looking at the results, you couldn’t tell one from the other.

Does that mean that a calorie isn’t a calorie and that we’ve just debunked the laws of thermodynamics? Of course not. It comes down to this, the dietary record keeping in this study is bullshit, more so in one group (TRF) than the other.

Let me show you how much it’s off, because it’s not by a small margin. According to my calculations,  the average calorie intake for the entire duration of the study was 1955 kcal in TRF. Over a duration of 8 weeks, this would yield a net loss of approximately 5.5 kg (12 lbs) for a 87.4 kg male. Now how much was actually lost? 1 kg.


I would have preferred if this was fact was illuminated in the discussion, but that’s too much to ask. When fishing for an explanation for the discrepancy between reported energy intake and lack of changes in body composition, it is briefly skimmed over and bundled in with a bunch of far fetched bullshit like metabolic adaptation.

Although the reported energy intake was substantially lower in the TRF group, the lack of changes in body weight and body fat, as well as the very small effect sizes for these parameters, indicates that there could have been spontaneous reductions in energy expenditure (e.g. decreased non-exercise activity thermogenesis), substantial misreporting of dietary intake, or metabolic adaptations in the TRF group that conserved energy, thereby minimizing weight loss (Byrne, Wood, Schutz, & Hills, 2012; Müller et al., 2015).

Substantial misreporting…you think?

I’ve been critical of many studies on intermittent fasting in the past, most of them in fact, and this is probably not going to be the last. It’s not the researchers fault that people misreport and underestimate their food intake, that’s actually to be expected, but in this study, the misreporting is off the charts. I saw it the minute I laid eyes on the data. I mean, in week 4, participants in the TRF-group claim to be eating an average of 1370 kcal on rest days and that’s just absurd. These guys aren’t any lightweights either at 87 kg.

There’s more accurate ways of tracking food intake than the method chosen here, which consisted of filling out a 4-day diet log week 1, 4 and 8, and I find it really disappointing to see that the error is not acknowledged in the discussion.


Summing it up, what conclusions can we draw from this paper then? Well, that depends on who you ask, but the researchers concluded that TRF had a positive effect on performance:

Interestingly, effect size data indicate that the RT-TRF group had greater improvements in lower body strength and endurance, as well as upper body endurance, as compared to the RT-ND group.

…But that low protein intake on TRF may have limited muscle growth:

When calculated relative to body weight, the average daily protein intake in the RT-ND group was 1.4 g/kg body weight/day, whereas the intake in the RT-TRF group, taking both TRF and non- TRF days into account, was 1.0 g/kg. While this level of intake in the RT-TRF group is similar to both the European Food Safety Authority rec- ommendation and the US Recommended Dietary Allowance of 0.8 g/kg, it is likely suboptimal for muscular hypertrophy during weight training and lean mass retention during weight-loss diets (Morton, McGlory, & Phillips, 2015; Phillips, 2014).

…So to grow more, you need to eat more:

Specifically, individuals who consumed more calories, carbohydrate, and protein on TRF days tended to have greater improvements in maximal lower body strength. Future research should examine the impact of total caloric and macronutrient intake during TRF days on RT performance, as well as employ higher protein intakes in individuals undergoing intermittent fasting programmes to determine if this promotes greater lean mass accretion.

Higher protein intakes to promote greater lean mass accretion? Let me save you some time on that future research.

How about this big ass chicken salad? Throw a pound of sliced chicken on some lettuce and pour dressing on it, there’s your meal. There’s more protein in there than what these guys ate for an entire day on TRF. No wonder they didn’t grow.

And lastly, participants rated TRF 3.6 out of 10 in terms of difficulty, which means that it’s a feasible diet approach for just about anyone.

This information, coupled with the relatively low ratings of difficulty of adherence, could indicate that young males are able to adhere to a TRF programme, although long-term evidence is not readily available.

If you’ve spent any amount of time on my site, you already know this shit. The real lesson of this study is this. If you tell Average Joe to lift weights and occasionally fast a few days a week, he’ll lean out spontaneously over the course of a few weeks, eating whatever he wants. And judging by this data, he’ll short-change his results by eating the same shit he usually eats in smaller quantities. This is a less positive spin on the results, but it’s the reality as far as I see it.

Adding to that, there’s a few lessons here about food reporting and eating behaviour, but I think anyone interested in that aspect of science, has already figured it out.

That’s all for today, folks. I found the study fairly disappointing. You and me both, I guess. There’s good times ahead though, guaranteed, so stay tuned! 🙂