We are what we eat? Eating ‘against the grain’ may not be as beneficial to performance and ‘economy’ in endurance athletes

During prolonged, submaximal exercise skeletal muscle uses both carbohydrates (CHO) and lipids as substrates to synthesize adenosine triphosphate (ATP), the fundamental unit of energy. The relative contribution of these substrates to fuel pathways depends upon several factors including mode, intensity and duration of exercise, the athlete's training status, and both acute and prolonged dietary intake. With regard to the latter, optimal performance by the athlete is greatly influenced by the combination of adequate fuel stores in relation to the demands of his/her event as well as ‘metabolic flexibility’, defined in the context of sports performance as the ability to rapidly and efficiently utilize these pathways to maximize ATP synthesis.

Seminal work conducted by Coyle's lab in the early 1990s showed that CHO‐dense diets were more ergogenic than fat‐rich diets for aerobic performance (Coggan & Coyle, 1991). These findings are supported by the notion that intramuscular glycogen stores are depleted far in advance of intramuscular adipose tissue; therefore, boosting glycogen depositions improves performance beyond increases in adipose tissue, a paradigm which holds strong amongst athletes and coaches today. However, these effects remain equivocal with some athletes responding well to such nutritional strategies whereas others are less responsive. A growing number of ultra‐endurance athletes have reported experimenting with alternative nutritional strategies, such as low‐CHO diets, experiencing a spectrum of self‐perceived benefits. Such data have led to the concept of ‘training low, but competing high’, whereby selected training sessions are completed in conditions of reduced CHO availability with CHO reserves restored immediately prior to competition. The intent of this strategy, termed the low‐CHO high fat diet (LCHF) or the ‘keto‐diet’, is to increase the utilization of intramuscular fat as fuel elevating circulating ketone levels (Phinney et al. 1983; Burke, 2015; Volek et al. 2016). Thus, acutely manipulating substrate availability via diet modification (i.e. LCHF) may influence intramuscular substrate stores and patterns of fuel metabolism – a potential consequence of this being an improved ‘economy’ and aerobic performance. Despite the growing interest of LCHF diets, athletes and coaches remain hindered by scarce scientific data and frequently resort to anecdotal evidence as the primary argument for such nutritional‐training strategies. Our understanding of nutritional strategies (i.e. the ketogenic diet) and endurance sports performance is developing but remains limited by tightly controlled studies. Although evidence has shown that LCHF allows individuals to metabolize fat more favourably (compared to glycogen) (Volek et al. 2016), the beneficial effects on endurance sport performance remain unknown. Therefore, an article in this issue of The Journal of Physiology by Burke and colleagues (Burke et al. 2017) attempted to address some of these long‐awaited issues by investigating whether a low‐carbohydrate, ketogenic diet is equal or superior in performance to the popular high‐carbohydrate diet in endurance athletes.

Burke et al. recruited 21 Olympic‐calibre male race walkers to participate in one of three dietary interventions, either: (1) high carbohydrate (HCHO), periodized low/high carbohydrate (PCHO), or low carbohydrate/high fat (LCHF) over 3 weeks, during two separate training camps. Interventions were calorically equivalent and both training camps were of similar structure (i.e. modes and volume of exercise). Assessments occurred during 3 days prior to, and following, both training camps; all assessment days utilized identical protocols.

Day 1's assessment consisted of a fasted, maximal exercise test to determine peak O₂ uptake (${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$). During the second test day, subjects completed a 10 km race walk, and on the final test day, subjects completed a 25 km competitive race walk. All experimental groups experienced improvements in ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$; however, only the HCHO and PCHO groups improved their 10 km race walk time. Additionally, a reduction in respiratory exchange ratio at maximal exertion was noted in the LCHF group despite an increased ${\dot{V}}_{{\mathrm{O}}_2\mathrm{peak}}$ alluding to a dietary‐induced shift in substrate metabolism favouring β‐oxidation during near‐maximal exercise. The LCHF group was also the sole group to increase their ${\dot{V}}_{{\mathrm{O}}_2}$ values during submaximal stages of the maximal exercise test following training. Collectively, these findings suggest that training‐induced improvements in submaximal running economy and endurance performance are compromised with a LCHF diet.

Until recently, only anecdotal evidence has supported the contention that the LCHF diet improves aerobic performance. Indeed, the applicability of performance assessment protocols clouds our understanding of these studies in relation to ultra‐endurance events. Burke and colleagues utilized a clever design to assess performance by mimicking 10 km and 20 km race walk competition scenarios to which athletes received monetary compensation based on performance. This design allowed the athletes to maximize their effort without the underlying bias/influence of dietary status. To this end, the results provide first‐hand evidence to suggest that a LCHF diet blunts performance which is likely derived from a reduced economy at submaximal work rates. The authors are to be applauded for these significant and novel contributions to nutrition and exercise science; however, several notable considerations may warrant attention when interpreting these results as they may provide additional clarity to findings and may stimulate new ideas for future work.

It should also be noted that while the population investigated in Burke's study adds to the study's novelty, cross‐sectional evidence (Volek et al. 2016) suggests that athletes competing in events >50 miles (or >2 h in duration) exhibit greater benefits from LCHF diets compared to athletes competing in events <2 h. Extended endurance races (ultramarathons) occur at much slower paces and at greater durations, which stimulates a greater reliance on fat stores as opposed to muscle glycogen. Additionally, Burke's sample population competed exclusively in race walking whereas Volek's sample population included runners and triathletes (Volek et al. 2016). Although the calibre of athletes between the studies of Volek et al. (2016) and Burke et al. (2017) appear similar, their events and respective training regimens are notably different. Future work aimed at investigating event‐specific metabolic demands among individual athletes would further advance our understanding of nutritional strategies, namely the efficacy of LCHF diets during prolonged (>2 h) exercise.

As the Burke et al. (2017) and Phinney et al. (1983) studies are the only works to date assessing ketogenic LCHF, translation of these results into practice should be cautioned as relatively low sample populations were employed. Additionally, Burke et al. (2017) reported fat oxidation rates among select participants exceeding what has been previously reported, further colluding with our understanding of LCHF diets (Phinney et al. 1983; Volek et al. 2016). Furthermore, given the large variability in performance times, it is tempting to consider the possibility that interindividual differences may exist (i.e. responders and non‐responders) with regard to the LCHF dietary regimen and 20 km race walking performance. Nevertheless, the mixed modelling statistical approach utilized by Burke and colleagues (2017) enabled the authors to take several covariates into consideration and evaluate the typical outcome of the dietary regimens. Potential solutions which may clarify the previously mentioned assumptions are: (1) incorporate a specific genotypic analysis of the athletes where genetic commonalities can be illustrated, and/or (2) the inclusion of a correlational analysis linking physiological biomarkers with performance changes to interrogate individual outcomes.

Although the mechanisms associated with the compromised performance in this study remain unclear, and hypotheses have been proposed in previous works (Coggan & Coyle, 1991; Burke, 2015), we propose an additional potential mechanism. Oxidative stress and inflammation induced by high fat diets impair vascular endothelial function, effectively reducing blood flow during exercise. Although acute and chronic endurance exercise have been shown to attenuate postprandial surges in endothelial dysfunction, it remains unknown whether LCHF diets negate the positive effects of long‐term aerobic exercise. Lastly, it remains unknown whether the LCHF‐adapted athletes who ‘carbo‐loaded’ just prior to competition experience a postprandial insulin spike, a factor which would likely influence oxidative stress and reactive oxygen species (ROS), thus altering performance. Investigating inflammatory (ICAM‐1, TNF‐α, IL‐6) and oxidative stress markers (ROS, isoprostanes, GPX‐1) may contribute to our understanding of the underlying mechanisms linking LCHF and vascular function in elite athletes.

In conclusion, Burke et al. (2017) advance our understanding of dietary strategies to enhance performance in elite endurance athletes by showing that training‐induced improvements in submaximal running economy and endurance performance are compromised with a LCHF diet. Nevertheless, many questions remain to be answered prior to redefining nutritional guidelines regarding LCHF for elite, endurance athletes. Clearly, there is an ongoing need for appropriately designed and conducted studies investigating long‐term, low‐carbohydrate diets on performance in the new and intriguing area of research.

Additional information

Competing interests

None declared.