Study: a handful of average-bodied Crossfitters in their mid-30’s were recruited and told to either: 1) keep doing what they’re doing; or 2) go full keto.  Crossfit 4x/week.   Strength testing before and after 6 weeks (Gregory et al., 2017).

I’ll start with the best part: KETOADAPTATION IS A REAL TRUE THING THAT WORKS (P<0.05).  Otherwise, this group’s performance would’ve plummeted.  It is known.

The performance test was time to complete a 500-meter row, 40 body weight squats, 30 abdominal mat sit-ups, 20 hand release pushups, and 10 pull-ups.

Tl;dr: both groups knocked about a half a minute off their time!

The key here is duration: 6 weeks of ketogenic dieting is adequate to restore performance back to baseline.  <3 weeks is not.




Here’s the downside (sort of):


body comp

Basically, the keto group dropped carbs and failed to compensate by upping other calories.  I know I know, spontaneous ad lib appetite reduction, but this is a study on PHYSICAL PERFORMANCE.



And in further support of “muscle growth sans carbs,” keto dieters upped protein by 15% and this still wasn’t enough to compensate for the reduction in carbs/insulin: they still lost a bit of lean mass (NS).  Imagine if they hadn’t increased the brotein? yikes




so basically, they lost body fat because CICO and retained lean mass because exercise and protein haha jk


Admittedly, it was cool to see the body comp changes, but we know fat loss eventually plateaus and people start eating maintenance calories again (maybe a bit more if Ebbeling can be believed).  And this is where they remain for the rest of their lives (hopefully).  So at 6 weeks, they were still losing weight, nowhere near where they’re going to be for the rest of their lives, but THAT’s where I’d like to see performance testing (ie, at a stable body weight).  Don’t get me wrong, I hate myself in advance for making this critique: the researchers should’ve pushed more calories in the keto dieters bc this is a confounder in a study on PHYSICAL PERFORMANCE…  but this doesn’t really matter in the big scheme of things because Blackburn’s group did that and showed the results were the same haha



On another note, I don’t think people should expect an additional performance boost from being more ketoadapted (or more fat-adapted or whatever), primarily because whether the study is 3 weeks or 6, performance never really gets better than baseline in experienced athletes.  With more advanced training techniques, sure (and I think this is common), but not more keto- or fat-adaptation.

For full access to all articles and much more (or if you just like what I do and want to support it), become a Patron! It’s three bucks a month and there are many other options. Sign up soon because there are only a limited number of spots left at the $3 level. It’s ad-free and you can cancel if it sucks ????

Also, I’m open to suggestions, so please don’t hesitate to leave a comment or contact me directly at

Affiliate discounts: if you’re still looking for a pair of hot blue blockers, Carbonshade  is offering 15% off with the coupon code LAGAKOS and Spectra479 is offering 15% off HERETrueDark is running a pretty big sale HEREIf you have no idea what I’m talking about, read this then this.

20% off some delish stocks and broths from Kettle and Fire HERE

If you want the benefits of  ‘shrooms but don’t like eating them, Real Mushrooms makes great extracts. 10% off with coupon code LAGAKOS.


calories proper

Become a Patron!



Be Sociable, Share!
  • Wab Mester

    Too bad we can’t read the paper till next year, but thanks for the cool preview! I was surprised by the higher BF% in the keto group until I noticed they were almost all female. It would have been interesting to see body comp changes in a no-exercise keto group. Biggest surprise to me is the lack of body comp improvement in the non-keto group. Crossfit 4X per week for 6 weeks = no gains?

    • “Biggest surprise to me is the lack of body comp improvement in the non-keto group”

      not surprising, because: “Study: a handful of average-bodied Crossfitters in their mid-30’s were
      recruited and told to either: 1) keep doing what they’re doing”

      in trained athletes, progress comes in years, not weeks

    • ” It would have been interesting to see body comp changes in a no-exercise keto group.”

      They were exercisers to begin with, so this would be a “stop exercise + start keto” group. IOW, no one does this.

  • Ben

    The calories seem to be really low for both groups. I’d expect somewhat trained persons to be eating more kcal + protein.

    • these weren’t particularly large or muscular people: ~160lbs and 30%BF via DEXA, but I see your point

  • John Lushefski

    My carbs are very low for my activity level, and I’ve never noticed any sort of issue, whether it be mood, sleep, performance, etc, that others always mention. I do dancing/ballet 4-7 days per week and weightlifting 4 days. I am 5’9, 180lbs, with very low body fat.

    I don’t know what 30% bf looks like on a female, but I think that performance studies on amateurs/casuals should be treated with caution, which is how I would describe crossfitters with that bf%.

    What are they actually eating? I wish there were more guidelines than “very low carb.”

    • good points

      -there were more women than men, so an average BF of 30% for the whole group isn’t really high

      -actual meal plans, recommended foods, and the workouts are in the Appendix. Ketones tested positive in the keto group, and the workouts looked pretty good imo

      -the participants were recruited from a bona fide crossfit gym and their performance measures weren’t too bad, so might not be prudent to assume “amateurs/casuals”

  • Man

    Would have been interesting to test the cross-fitters on heavy weight lifting, the kind that requires good glycogen levels.

    Note: it was an awfully low amount of calories …

    • They tested “the kind that requires good glycogen levels,” no difference between the groups.

  • Matt Huston

    Any significance to the time-drop difference (41.2 vs 55.8 seconds) between the control and LCKD groups?

    • significantly different from baseline, but not between groups