We know that ketogenic diets promote fat and weight loss. We also know that some lean body mass loss (not necessarily muscle loss) must coincide with fat loss.
However, there’s been much debate as to whether hypocaloric ketogenic diets cause muscle loss, or, at least when compared to hypocaloric non-ketogenic diets.
In this article, Vincent looks at what the scientific literature on ketogenic diets and muscle mass says.
- Calorie- and protein-matched ketogenic and non-ketogenic diets lead to similar rates of fat loss.
- Ketogenic diets lead to larger weight losses, with greater water and *protein* losses than balanced macronutrient diets (both in terms of percentage of weight loss, and in absolute terms).
- It’s not clear if the protein losses during ketogenic diets are a result of losses in muscle tissue or in other protein-containing tissues (e.g. skin, organs, etc.).
- Issues with nitrogen balance preclude us from drawing any firm conclusions from trials using nitrogen balance to measure protein losses with ketogenic diets.
- Keto/lower-carbohydrate diets present a useful (short-term) tool for powerlifters looking to improve their Wilks score.
- Keto is best suited for those with weight/fat loss goals (likely short-term, but perhaps long-term) than for those with muscle building/weight gaining goals.
Relevant research on keto and muscle loss
Let’s start by looking at some relevant research on ketogenic diets and muscle loss or lean body mass loss.
Kephart et al. (2018) took recreationally trained Crossfitters, put some on a ketogenic diet, and left others to their conventional diets. Lean body mass and fat mass were assessed (via DXA) after 12 weeks of crossfit training. The ketogenic diet group lost more body fat than the conventional diet group, but they also lost (leg) lean mass, while the conventional diet group did not.
The significantly larger loss in body fat in the ketogenic diet group suggests that the subjects were in a bigger caloric deficit, which may explain the larger losses in lean mass relative to the non-ketogenic diet group.
Greene et al. (2018) employed a similar study protocol (among intermediate-advanced olympic lifters), but also incorporated a crossover design. Here, subjects alternated between 3 months of their typical diet (>250 grams of carbohydrate) and a ketogenic diet (<50 grams of carbohydrate per day). The keto group lost more fat mass, as well as lower body and total lean mass. Nonetheless, lifting performance was roughly similar between diet conditions.
Again, there was more fat loss during the ketogenic diet treatment, which could explain (at least partly) why the ketogenic diet resulted in more lean mass loss.
Of note, Vargas et al. (2018) also assessed body composition changes in ketogenic vs. non-ketogenic diets (in trained lifters) undergoing resistance training. However, despite being prescribed a caloric surplus, the ketogenic diet group *lost* weight (which indicates a caloric deficit). Thus, lean mass changes between the ketogenic and non-ketogenic diet groups could not be compared, since the non-keto group gained mass. However, the ketogenic diet group did not gain lean mass (on average) throughout the 8-week resistance training program, and other research indicates that trained lifters can gain lean mass while losing weight.
Otherwise, Chatterton et al. (2017) conducted a case study (in intermediate-level olympic lifters), in which subjects consumed 1g/kg of carbohydrate (borderline ketogenic diet). In this study, all subjects lowered their carbohydrate intake. Two out of five olympic lifters increased their strength similarly on the ketogenic and higher-carbohydrate diets. Alas, two lifters hit plateaus on the lower-carbohydrate diet, and another lifter actually lost (both relative and absolute) strength (however, this subject reduced carb intake the most).
This suggests that responses to ketogenic dieting may vary between individuals, but that keto probably doesn’t offer any advantages for strength gain. Moreover, this study *weakly* supports the notion that ketogenic diets induce muscle loss, since none of the trainees increased quadriceps or triceps muscle thickness, but one subject experienced a 10.8% decrease in triceps muscle thickness.
Summary of the research on keto and muscle loss
The available research comparing the effects of non-ketogenic vs ketogenic diets on muscle loss in trained athletes/lifters indicates consistently larger lean mass losses with ketogenic diets.
- this does not imply muscle loss, especially considering “keto flush” (in which substantial water losses occur at the onset of a very low-carb diet).
- since ketogenic diets tend to result in a decrease in caloric intake, the larger reductions in lean body mass/muscle mass may be due to the resulting caloric deficit rather than due to the ketogenic diets per se.
Anyway, on average, weightlifters roughly maintain strength performance with keto, and keto’s effects on strength differ between people.
Taken together, keto has been shown to produce similar (average) strength gains, *and* greater lean mass loss. This paints a murky picture around keto for muscle growth, but a (potentially crucial) bit of data from 1976 adds another piece to the puzzle.
The 1976 keto vs non-keto trial
In a metabolic ward crossover trial (in 6 obese inactive males), Yang and Van Itallie alternated subjects between 3 (10-day) diet conditions:
- keto (800 calories; 10 grams of carbs),
- starvation (0 calories), and
- mixed-macronutrient dieting (800 calories; 90 g carbs),
with (5-day) 1200 calorie breaks in-between.
Subjects drank some combination of corn oil, sucrose, and sodium caseinate (for fat, carbohydrate, and protein, respectively) throughout the study.
The researchers observed that 10 days of keto elicited ~4.7 kg of weight loss, on average. Of this weight loss, ~2.88 kg (61.2%) was water, ~1.64 kg (35%) was fat and ~.179 kg (3.8%) was protein. Compare this to the (isocaloric) mixed macronutrient diet treatment, which caused a body weight loss of ~2.8 kg, of which ~1.04 kg (37.1%) was water, ~1.67 kg (59.5%) was fat, and ~.095 kg (3.4%) was protein.
Importantly, the keto group lost 179 grams of protein, but the balanced group only lost 95 grams of protein. To be clear, an 84 gram difference in protein loss is minuscule, but we’d expect protein losses to be small (because subjects dieted for just 10 days) under each condition.
*However*, it’s noteworthy that the keto group lost 88% more protein over this time period. Indeed, the keto group lost almost twice as much protein, despite having similar energy, protein, and fat intakes. What’s more, both diet groups lost similar amounts of fat, thus *body composition change was better* in the higher-carbohydrate (mixed-macronutrient) diet condition.
Here are a few other things worth noting:
- Upon returning to 5 days at 1200 calories, the keto group lost another ~.013 kg of protein and ~.63 kg of fat, while the balanced group lost another ~.03 kg of protein and ~.6 kg of fat. Thus, 5 days at higher carbohydrate intake slowed protein loss on keto.
- During this 5-day period, the keto group re-gained ~.9 kg of water weight (for a net weight gain of ~.26 kg). Meanwhile, the balanced diet group continued losing weight overall (for an additional weight loss of ~.82 kg). Taken together, this suggests that while energy balance dictates fat loss, carbohydrate intake can alter body weight and protein losses.
- The starvation diet produced significantly greater nitrogen losses than the keto and mixed diets. This highlights that keto won’t produce as much protein loss as prolonged fasting would.
- Since this was a crossover trial, the same subjects underwent all diet conditions. Thus, when the same people followed a calorie and protein-matched ketogenic diet, they lost more protein (on average) than they did on a higher-carb diet.
So, while this is just one study, a metabolic ward trial with crossover design is very high quality. That said, this study’s results can’t necessarily be generalized to athletes, since just 6 obese individuals (with no resistance training) were examined. Also, the results can’t be generalized to the longer-term, since each diet condition only lasted 10 days.
Ketogenic diets and powerlifting
Considering the above, a very low-carbohydrate diet probably induces greater protein losses than a calorie-matched, higher-carb diet. This, plus the fact that keto has been shown to decrease lower-body lean mass and (possibly) triceps muscle thickness, supports the idea that keto can decrease muscle size (perhaps in skeletal muscle, in addition to/rather than organ tissue).
The key word here is *can*, as individual responses seem to vary. Nonetheless, if your goal is maximal hypertrophy, you should probably consume greater than 1 gram of carbohydrate per kilogram of body weight.
On the flip side, keto provides a useful tool for some powerlifters. For example, if you can make strength gains as usual on keto, then a ketogenic/lower-carbohydrate approach may allow you to increase your Wilks score.
This is because the ketogenic diet consistently reduces ad libitum energy intake (i.e. keto causes people to eat less, without trying), likely due to the diet’s hunger-blunting effect. This, combined with transient water loss, allows many lifters to improve their Wilks scores/relative strength (*at least* in the short-term). As such, the ketogenic diet is probably best seen as a short-term weight/fat loss tool.
For those with lesser performance/size goals, the ketogenic diet may be a viable long-term option, but it’s worth noting that many people find it hard to stick to keto in the long-term (i.e. for more than 3 months).
Lastly, I want to emphasize that it isn’t clear whether *ketone production* or *very low carb intake* causes the lean mass/protein losses seen on keto. As such, a targeted ketogenic diet (perhaps higher carbohydrate, with carbs biased around exercise) may allow you to reap the best of both worlds (i.e. any benefits of keto, with less/no muscle/performance loss).
A saving grace for keto?
In a conversation with Marty Kendall of Optimising Nutrition, it was brought to my attention that we’d actually expect to see more nitrogen loss with ketogenic diets, as some protein is used for gluconeogenesis (the production of glucose from non-carbohydrate sources).
By this logic, the keto group lost more protein/muscle because they used some for gluconeogenesis. However, if they compensated with greater protein intake, then nitrogen losses might have been similar in the keto group (for Van Itallie and Yang’s study).
In practice, this means that higher-carb diets (containing ~1.8 g/kg protein) and ketogenic diets (containing ~2.2 g/kg protein) could produce similar protein balances.
If true, this would mean that, with calories and protein matched, keto is still inferior for muscle growth. However, protein-enriched ketogenic diets might cause as much muscle growth as higher-carb diets. As such, body composition changes may be similar on keto, given higher protein intakes.
Nitrogen balance and changes in muscle mass
Ultimately, nitrogen balance is a very limited measurement tool. Protein researchers Jorn Trommelen and Stuart Phillips have each advised against drawing any conclusions based on nitrogen balance, given its constraints.
For example, your nitrogen balance can be positive despite clear muscle loss. Thus, I may have erroneously drawn conclusions from Van Itallie and Yang, since nitrogen balance indicates little about changes in skeletal muscle mass
Final thoughts and conclusions
All said, we don’t have sufficient evidence to say with certainty if a ketogenic diet causes more muscle loss compared to a non-ketogenic/higher-carbohydrate diet. However, we know enough to draw a few conclusions:
- Calorie- and protein-matched non-ketogenic and ketogenic diets cause similar rates of fat loss, but ketogenic diets result in more weight loss due to decreases in water. Since water is a component of lean body mass, research that assesses body composition with DEXA can’t differentiate between a decrease in water and a decrease in muscle mass.
- There is only one documented case of a decrease in muscle size on the ketogenic diet (one subject in the Chatterton et al. trial who lost triceps thickness). However, while triceps thickness is a generally good measure of muscle mass, it cannot account for changes in fluids inside the muscle. This means that a decrease in muscle glycogen levels (which could very well happen with a ketogenic diet) may result in a decrease in muscle thickness, without an accompanying decrease in actual contractile proteins.
- Strength gains/losses on ketogenic diets seem to be an individual thing – i.e. some people do better than others.
- While the ketogenic diet is (at best) on equal footing with other diets, it’s still a viable option for those seeking body composition improvements, especially for people struggling with hunger, or those who prefer to eat ad libitum (i.e. to satiety, without tracking calorie intake).
- The potential for performance and/or muscle loss might reasonably steer competitive bodybuilders and powerlifters away from keto, although, in the short-term, the water weight drop that accompanies ketogenic diets can improve powelifters’ Wilks score, since they lose more relative weight than they do strength performance.
That said, the vast majority of people don’t need to worry about truly *maximizing* lifting performance and muscle growth. The differences between gains on calorie- and protein-matched keto vs non-keto diets are likely small, if even present at all. So, if keto is working well for you, then you should probably stick with it!