Effects of a Ketogenic Diet on Body Composition in Healthy, Young, Normal-Weight Women: A Randomized Controlled Feeding Trial

There has been a profound increase in interest in ketogenic diets over the past decade, attracting attention from both researchers and the general public. Initially seen as a method to combat obesity and diabetes, the therapeutic effects of ketogenic diets are now recognized across a broad range of conditions, including epilepsy, cancer, inflammation, and neurodegenerative diseases [1]. In addition, one of these new research areas concerns whether a ketogenic diet can improve body composition and/or physical performance among athletes as well as everyday exercisers. In a recently published systematic review, Kang et al. [2] examined the potential ergogenic effects of a ketogenic diet on healthy normal-weight individuals who adhered to a ketogenic diet for a minimum of three weeks. According to the 14 studies that reported on body composition, it seems that fat-free body mass (i.e., muscle) can be preserved, while fat mass decreases. However, as the authors of the systematic review highlighted, the majority (~85%) of participants in these studies were men. This gender disparity in research on the effect of diets on body composition is furthermore emphasized in the International Society of Sports Nutrition (ISSN) position stand [3]. It is a well-known fact that there are some biological differences between men and women. Sex chromosomes and sex-specific hormones, including androgens in males and estrogens and progesterone in females, affect energy metabolism [4]. For example, there appears to be a difference in substrate utilization when adapting to a ketogenic diet. A study of CrossFit-trained athletes showed that males appear to shift towards greater fat utilization during submaximal-intensity exercises under a ketogenic diet compared to female athletes [5]. Consequently, one cannot extrapolate data from men to women. The majority of the studies in Kang’s systematic review did not include women at all, and there was only one study [6] where the number of participating women was more than five individuals. However, the data were not broken down by sex in that study. Another recent systematic review by Murphy et al. [7] on ketogenic diets (≥two weeks) and performance in non-obese healthy individuals with more objectively defined ketosis (carbohydrate intake ˂ 50 g/day or circulating β-hydroxybutyrate ≥ 0.5 mM) found nine studies that reported on body composition. However, only a handful of these studies included women (with no more than ten women in any single study), and the authors suggested that sex may be one of several factors that contribute to discrepancies in the outcomes between studies. In addition, the authors call for future studies where food intake is carefully controlled in order to minimize differences in body mass change and thereby eliminate a potential confounding variable.

In summary, there is a lack of body composition data from well-controlled and randomized feeding studies on young, healthy, normal-weight women who eat a truly ketogenic diet. In light of this, an explorative, randomized, and controlled feeding trial with a crossover design was conducted on young, healthy, normal-weight women following a non-energy-restricted ketogenic low-carbohydrate high-fat (LCHF) diet. A Swedish National Food Agency (NFA)-recommended diet was used as a control diet. The primary outcome was LDL cholesterol [8]. Here, we report the effects of a 4-week non-energy-restricted ketogenic LCHF diet on body composition in young, healthy, normal-weight women.

In this randomized controlled feeding trial with a crossover design among normal-weight, young, healthy women, we demonstrate that four weeks of a truly ketogenic LCHF diet reduces the lean mass twice as much as fat mass in absolute terms. Our overall conclusion is that normal-weight women adhering to a ketogenic LCHF diet should consider engaging in deliberate strength training to minimize the risk of the undesired loss of skeletal muscle.

The ketogenic LCHF diet has been proposed as an effective dietary strategy for improving body composition. Advocates of this diet emphasize that it leads to a reduction in body fat while preserving muscle mass. These effects are attributed to the metabolic shift that occurs when the body no longer has access to dietary carbohydrates. In the absence of glucose, glucagon, epinephrine, and cortisol enhance lipolysis, leading to the loss of fat mass [20]. In addition, there will be a reduction in fat mass due to the reduced fasting levels of insulin, an anti-lipolytic hormone. Furthermore, this hormonal profile, characterized by a high ratio of glucagon to insulin, will favor not only lipolysis but also gluconeogenesis and ketogenesis. The mobilization of free fatty acids from adipose tissue provides the body with fuel, not only in the form of free fatty acids but also in the form of glycerol, a gluconeogenic precursor [21]. Glycerol is one of the substrates for glucose synthesis; however, amino acids released from skeletal muscle tissue are used as substrates as well. The hormonal profile will not only activate gluconeogenetic enzymes but also augment the transfer of amino acids to the liver, which will negatively impact lean mass. Therefore, to prevent extensive loss of lean mass, it is most important that the brain, a significant consumer of glucose, switches its energy source from glucose to ketone bodies produced through increased fat oxidation [22]. Ketone bodies not only function as an alternative energy substrate but can also activate the enzyme AMP-activated protein kinase (AMPK), which increases the mitochondrial oxidation of fatty acids while simultaneously inhibiting the mechanistic target of rapamycin complex 1 (mTORC1) in skeletal tissue [23]. mTORC1 plays a crucial role in integrating growth and metabolism by responding to nutrient levels and growth signals [24]. The inhibition of mTORC1 aims to suppress energy-consuming protein synthesis in skeletal muscle tissue during glucose scarcity [23]. In summary, following a ketogenic LCHF diet impacts both fat and lean mass through hormonal changes and the molecular effects of elevated ketone bodies.

The literature shows a striking absence of studies where both activity patterns and dietary intake have been carefully controlled to evaluate the effect of a specific dietary approach on body composition. This applies even more to women. In a recent systematic review by Coleman and colleagues [25] on body composition changes in physically active populations consuming ketogenic diets, the authors discussed some limitations and risks of bias in the current literature. More specifically, they stressed the lack of randomization, the lack of a controlled food intake, and the use of insensitive techniques to measure body composition. In the current study, we have considered these common limitations in the study design. In addition, we also chose to utilize a crossover design with an extended washout period. The crossover design enhances statistical power by having participants serve as their own controls, and the extensive 15-week washout period reduces the risk of carry-over effects resulting from dietary interventions. The population used in the analysis of the present trial was the full analysis set, i.e., participants with measurements from two completed diets. In addition to the decrease in statistical power due to the loss of data, the dropout of seven participants increased the risk of bias when estimating the effect of the diets if the data are not missing completely at random (MCAR) [26]. However, considering the relatively small sample size and the complexity of the crossover data, with dependency between measurements on the same participants, we judged that multiplying imputing data from a model would not increase the validity of the results in this case.

Our current feeding study demonstrates that the ketogenic LCHF diet mediates a twofold reduction in lean mass (~1.4 kg) compared to fat mass (~0.7 kg). This occurred despite the following factors: (1) adherence to both dietary patterns was good (energy intake did not significantly differ between the two dietary interventions based on reported deviations in food intake) [14] and (2) physical activity patterns did not differ between the two dietary interventions [14]. There are several potential explanations for the increased loss of body mass during ketogenic diet intake, particularly during the first week of dietary transition. Although the body rapidly adapts to increased fat metabolism [27], there have been reports of a transient increase in energy expenditure (approximately 100 kcal/day) during the initial week of transitioning to a ketogenic diet [27], including elevated nitrogen losses and increased gluconeogenesis. Transitioning to a ketogenic diet also reduces the storage of both water and fuel in the form of glycogen [28,29,30,31]. When carbohydrate intake is minimized, muscle glycogen decreases, and consequently, the water bound to glycogen (water binds to glycogen in a ratio of approximately 3:1 [32]) is reflected as a reduced lean mass in DXA measurements. One way to minimize and account for this effect during DXA measurements would have been to introduce some form of ‘carbohydrate refeed’ before the DXA assessment [33]. However, since the intervention continued for a few more days after the DXA measurement, we chose not to compromise the nutritional ketosis with a ‘carbohydrate refeed’. An intriguing question arises: how much of the decrease in lean mass constitutes muscle mass, and how much is attributed to glycogen and glycogen-bound water? This inquiry has garnered significant interest within the field of ketogenic diets and exercise performance, particularly in studies involving men. Interestingly, the duration of adherence to a ketogenic diet appears to play a role. In a study of keto-adapted ultra-endurance runners (with an average of 20 months on the diet), Volek et al. [34] reported no significant differences in resting muscle glycogen content. Another study examined endurance-trained male cyclists who followed an LCHF diet for an extended period (>8 months), revealing a glycogen content of approximately 55–60% compared to the control group (mixed diet) [35]. Additionally, in a study where endurance-trained men followed an LCHF diet for the same duration as the women in our study (28 days), these men retained approximately 45% of muscle glycogen [28]. These measurements were conducted after an overnight fast. Assuming that the women in our study would lose 55% of total muscle glycogen (total muscle glycogen in the body is approximately 400 g), similar to Phinney and colleagues’ research [28], the actual loss of glycogen, along with glycogen-bound water, would constitute a significant portion (~600 g) of the reduction in lean mass among our female participants. Given the likelihood that the LCHF diet reduced glycogen content and, thus, total water content in the current investigation, caution is warranted when interpreting the results on lean mass. Incorporating measurements of urine osmolality using an osmometer and estimating total body water through bioelectrical impedance analysis (BIA) can provide additional insights to evaluate changes in body composition.

Even when considering the reduction in muscle glycogen, we observed an extensive decline in lean mass among the normal-weight, young women despite a high protein intake (approximately 1.8 g/kg bw/d). This decline in lean mass can even be considered surprisingly high, especially in physically active individuals and given that weight loss studies have shown that maintaining nitrogen balance can minimize the loss of lean mass [36]. In studies (predominantly involving men) with diet- and exercise-induced negative energy balance lasting up to four weeks, a protein intake of ≥1.6 g/kg bw/d supports greater retention of lean body mass [36,37,38]. However, despite our current feeding study design and the fact that reported energy intake and activity patterns did not differ significantly between the two diets, we still observed a significant decrease in both lean mass and fat mass, further emphasizing that ketogenic diets effectively reduce body weight in those who adhere to this dietary regimen. This is in line with a recent systematic review by Murphy et al. on ketogenic diets and performance in non-obese, healthy individuals (mainly men) with objectively defined ketosis, which reported a decline in body mass, fat mass, and fat-free mass [7].

In recent years, ketogenic diets have become more popular, especially among normal-weight adults and individuals pursuing aesthetic goals, including those engaged in resistance training [2]. Evidence suggests that within untrained populations, ketogenic diets may lead to a marginally higher loss of fat-free mass (FFM) compared to non-ketogenic dietary patterns [39,40,41]. Interestingly, a recent systematic review and meta-analysis examining the impact of resistance training combined with ketogenic diets on body composition also found a significant reduction in FFM among those assigned to a ketogenic diet compared to non-ketogenic counterparts [42]. The authors suggested that ketogenic diets may inhibit the mechanistic target of the rapamycin (mTOR) signaling pathway, possibly through heightened AMPK activity. Elevated AMPK activation has been shown to impede mTOR signaling, a pivotal regulator of muscle mass gains [43]. In essence, it appears that a ketogenic diet impedes anabolic pathways, although concurrent resistance exercise may help mitigate the loss of FFM. Consequently, the potential reduction in FFM remains a notable concern associated with ketogenic diets, even in the presence of concurrent resistance training [42]. Importantly, the risks associated with the loss of lean mass, primarily muscle mass, along with fat mass in female athletes, include decreased muscle strength and performance, increased risk of hormonal imbalance, menstrual irregularities, compromised immune function, sports injuries, decreased bone density, stress fractures, and long-term osteoporosis [44]. Therefore, female athletes should carefully consider these risks if adhering to a ketogenic LCHF diet.

In the most recent Position Stand on diets and body composition [3], the ISSN tried to provide clarity on the effects of various diets on body composition. The ISSN stresses that every method used to assess body composition comes with its own set of strengths and limitations. The DXA method used in the present investigation has numerous inherent strengths [3]. First, it operates as a non-invasive and expeditious technique, devoid of active subject performance, thereby streamlining the assessment process. Notably, its commendable precision and reproducibility across diverse age brackets underscore its reliability. Moreover, the method’s efficacy remains unhampered by the presence of pathological conditions or developmental anomalies, ensuring a robust and uninfluenced measurement output. In relation to the research question in the current study, the downside of DXA may be a potential limitation pertaining to its reliability in longitudinal studies involving subjects experiencing substantial fluctuations in glycogen or hydration levels between measurements. This limitation is relevant in relation to the present study, both in light of the preceding discussion regarding the effects of the ketogenic diet on muscle glycogen stores and when considering that the DXA measurements were scheduled one hour after lunch during the final week of the dietary intervention. Based on the study’s design (feeding trial), food and fluid intake in the hours before measurement was not standardized, which should be kept in mind when interpreting the results. The duration of this feeding trial was four weeks, which might be too short for the body to fully adapt to the ketogenic metabolic state. Additionally, our participants were young, normal-weight females who exercised weekly and were highly motivated to follow a strict ketogenic LCHF diet. To deepen our understanding of how this dietary regime affects body composition, future studies are warranted to investigate its long-term effects, as well as the interplay between age and lifestyle factors.

This unique feeding trial emphasizes the effectiveness of a strict ketogenic diet in inducing weight loss even in normal-weight healthy young women with an unaltered physical activity score. Notably, the decline in lean mass was twice as much as fat mass in absolute terms. To minimize the risk of undesired muscle mass reduction, women who adhere to a ketogenic LCHF diet should consider engaging in deliberate resistance training.

We thank all the women who participated in this study. We thank Ingegerd Karlsson, the research nurse at the Clinical Research Center, for excellent assistance with recruitment, screening, and guidance. We thank Cecilia Fall for the medical assessments during this study.

Conceptualization, J.B. and A.S.; methodology, J.B., M.S., and A.S.; software, P.L.; formal analysis, J.B., P.L., and A.S.; investigation, J.B. and A.S.; resources, J.B., M.S., P.L., and A.S.; data curation, J.B. and A.S.; writing—original draft preparation, J.B. and A.S.; writing—review and editing, J.B., M.S., P.L., and A.S.; visualization, J.B., P.L., and A.S.; supervision, J.B. and A.S.; project administration, J.B. and A.S.; funding acquisition, J.B. and A.S. All authors have read and agreed to the published version of the manuscript.

This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethical Review Board at Umeå University, Umeå, Sweden (Dnr 2014-361-31M: 8 January 2015 and Dnr 2015-45-32M: 19 February 2015).

Informed consent was obtained from all subjects involved in this study.

The data presented in this study are available upon request from the corresponding author.

The authors declare no conflicts of interest. The funders had no role in the design of this study, the collection, analyses, or interpretation of the data, the writing of the manuscript, or the decision to publish the results.

This research was funded by Dr P. Håkansson’s Foundation, Eslöv, Sweden, the Magnus Bergvall’s Foundation (grant number 2014-00411, Sweden), and the Department of Food, Nutrition and Culinary Science, Umeå University, Umeå, Sweden. The authors did not receive any funding or benefits from commercial organizations.