It’s been a while since I wrote a full-length blog post. Almost a whole year, in fact. The reasons are twofold: 1) My priorities have been on my work in British Athletics and my PhD; 2) Nothing had made me want to devote a significant amount of time to write about a topic. As I had some time off over Xmas, I didn’t want to completely disengage my brain. Furthermore, the topic of fasted vs. fed cardio with the aim of fat loss has been a trending topic recently, both in the fitness industry as well as academia. As such, the aim of this article is to highlight our understanding to date surrounding this issue, dispel some key misconceptions regarding it, and conclude with my practical recommendations. It is worth mentioning that though the potential benefits of manipulating food intake around exercise involve factors other than just fat loss (e.g. training adaptations, markers of health and disease etc.), such topics are beyond the scope of this article.
Rationale for fasted cardio for fat loss
A common strategy among those competing in aesthetic sports (e.g. bodybuilders, fitness competitors etc.) and those competing in weight class sports (e.g. boxing, wrestling, judo etc.) is to perform cardiovascular exercise after an overnight fast, waiting until after the exercise bout to consume breakfast. The basic premise for this practise is that low levels of glycogen (and/or glycogen depletion during the exercise bout itself) and insulin, shift energy utilisation away from carbohydrate for fuel, thereby allowing greater mobilisation of stored fat that can be used for fuel (fat oxidation).
What does the research say?
Acute effects of fasted vs. fed exercise on fat oxidation
During the exercise bout
Results from several acute studies (linked here, here and here) support the idea that exercise in a fed state can result in a decrease in fat oxidation during the exercise bout due to a reduced entry of long chain fatty acids into the mitochondria. These observations in a fed state have been attributed to several factors, including the already mentioned insulin-mediated attenuation of adipose tissue lipolysis, an increased glycolytic flux and decreased expression of genes involved in fatty acid transport and oxidation.
That said, things aren’t so straightforward, as several studies have demonstrated that exercising in a fed state has little to no effect on fat oxidation during the exercise bout. Indeed Coyle’s research group has twice demonstrated (Links here and here) that carbohydrate ingestion during moderate-intensity (65-75% VO2max) does not reduce fat oxidation during the first 120 min of exercise in trained men. Furthermore, Horowitz and colleagues studied the fat burning response of six moderately trained individuals in a fed vs. fasted state to different training intensities. The participants each underwent four exercise conditions that involved two hours of cycling. During two of the trials, they consumed a high-glycemic carbohydrate meal at 30, 60, and 90 minutes of exercise, once at a low intensity (25% peak oxygen consumption) and once at a moderate intensity (68% peak oxygen consumption). Subjects performed the other two conditions at the same exercise intensities but this time following a 12-14 hour fast. Results in the low-intensity trials showed that although lipolysis was suppressed by 22% in the fed state compared with the fasted state, fat oxidation remained similar between groups until the 80–90 minute mark (after this point greater fat oxidation was observed when fasted). Conversely, during moderate-intensity cycling, fat oxidation was not affected by during exercise feeding at any time. This occurred despite a 20–25% reduction in lipolysis and plasma free fatty acid (FFA) concentration.
In another study involving endurance-trained subjects, Febbraio et al looked at the effect of pre-exercise and during-exercise carbohydrate consumption on fat oxidation. In a crossover design, seven participants cycled for 120 minutes at approximately 63% of peak power output (the intensity at which maximum fat oxidation occurred). Results showed no evidence of impaired fat oxidation associated with consumption of carbohydrate either before or during exercise, despite elevated insulin levels in the fed conditions.
It seems that such discrepancies in findings are largely due to the training status of the subjects involved in each trial, with the effects of fed status on fat oxidation diminishing in higher trained individuals. It is possible that in endurance-trained individuals, significantly more fat is broken down than that the body can use for fuel. FFAs that are not oxidised ultimately become re-esterified in adipose tissue (basically meaning that the unused fatty acids are reincorporated into fat tissue), thereby nullifying any lipolytic benefits afforded by pre-exercise fasting.
Though data from the above studies give some insight into answering the question of this article, the effects of feeding or not on subsequent substrate use during the exercise bout are limited to the exercise bout. Rather, looking at substrate use over a 24-h period around the exercise bout would give us a greater insight into the effects of pre-exercise feeding on fat loss.
24-h fat oxidation
A popular study cited by those when this argument comes up is a paper by Paoli et al titled ExercisingFastingorFed to Enhance Fat Loss. In this study, the authors compared the effect on oxygen consumption (VO2) and substrate utilisation (estimated by the respiratory exchange ratio (RER)), in eight men who performed the same moderate-intensity training session (36 min of running on treadmill at 65% maximum heart rate) in the morning in either a fed or fasted state. Cardio after breakfast increased both VO2 and RER significantly (4.21 vs. 3.74 and 0.96 vs. 0.84, respectively). Twelve hours after the cardio session, VO2 was still higher in the fed condition, whereas RER was significantly lower, indicating greater fat utilisation. This shift towards greater lipid use in the fed condition was still significant at the 24-h mark. A notable limitation of this study, however, was that dietary intake wasn’t strictly controlled, though individual subjects food diaries “demonstrated substantial similarity” between trials. Furthermore, having VO2 and RER measures at more regular intervals may have painted a better picture of substrate use over a 24-h period.
A more recent, strictly controlled study by Iwayama and colleagues sought to determine whether the time of day that one exercises affects 24-h fat oxidation in a group of nine, young endurance athletes. The subjects performed 100 minutes of treadmill running in a metabolic chamber under three conditions: 1) before breakfast [AM] 2) after lunch [PM] 3) 50 minutes before breakfast and 50 minutes after lunch [AM/PM]. In addition, meals were provided to meet individuals’ energy balance. Building on the findings of their earlier work, the authors observed that although 24-h energy expenditure was similar between trials (signifying no changes in energy balance), 24-h fat oxidation was higher in the AM group compared with the AM/PM and PM groups (1142, 809 and 608 kcal, respectively).
According to the authors, this higher fat oxidation in the AM group compared with the PM group was due to a transient energy deficit during the AM trial. In other words, in the three hours following exercise, the AM group burned more Calories than the PM group (330 vs. 280 kcal, respectively). Subsequently, because of the greater glycogen depletion in the AM group, it seems likely that this led to greater levels of fat oxidation. That said, the greater initial energy deficit was nullified by the 24-h mark, demonstrating that there was no residual beneficial effect of fasted training on energy balance. These data are interesting and in stark contrast to previous research that has either found no difference, or a more beneficial effect, for fed exercise on subsequent energy expenditure, probably due to a greater thermic effect of feeding on subsequent exercise (links here, here, here and here). Even if we choose to ignore the weight of the evidence regarding an increase in exercise thermogenesis in a fed state, the fact that was a transient energy deficit observed for fasted cardio in the study of Iwayama et al is of little significance given a lack of difference over 24-h.
Why the discrepancies in acute fat oxidation?
These unique observations from Iwayama’s lab (i.e. greater fat oxidation for fasted vs. fed cardio) may be explained by several factors. Firstly, the exercise time of 100 minutes is longer than in the other studies cited. As such, it is possible that the increased training volume, and its associated greater glycogen depletion, would have been sufficient to highlight such differences in fat oxidation. Moreover, these observed differences across studies may also be explained by the greater endurance training experience of the participants in the Iwayama study (although, subjects in the Paoli study were said to be trained, the subjects’ mean resting heart rate of 66.3 beats/min would suggest otherwise). Interestingly, it is well documented that intramuscular triglyceride (IMTG) can contribute to increases in fat oxidation, with these increases being substantially more pronounced in endurance-trained individuals. For example, in untrained subjects, roughly 50-70% of fat oxidised comes from plasma FFAs with the remaining coming from IMTG. In endurance-trained individuals, however, it is estimated that the contribution of nonplasma fatty acids to total fat oxidation is approximately double that of untrained individuals. One study by Hurley and colleagues reported that in trained individuals the contribution of IMTG stores to total fat oxidation was approximately 80% during 120 minutes of moderate-intensity activity. As I’m struggling to put it better myself, the following paragraph from a paper by Brad Schoenfeld perfectly highlights my thoughts on the relevance of the oxidation of IMTG for fat loss:
As a final point, though fat oxidation over 24-h gives us a better insight into body composition changes that may occur compared with that during the exercise bout alone, it still may not translate to actual body composition changes over the long term (i.e. fat loss over weeks and months). Indeed, in the Iwayama paper – the one which people cite to “show” that fasted cardio is better for fat loss – the authors state the following:
The point the authors are getting at is that as long as total energy balance remains the same, fat loss will be very similar, as generally, the more carbohydrate you “burn” during exercise, you inevitably “burn” more fat at a later point in time. Another point supporting the idea that 24-h energy balance is more meaningful than acute changes in fat oxidation can be found by observing the literature surrounding high-intensity interval-training (HIIT) and fat loss. Indeed, on a minute for minute basis, HIIT has been shown to be a superior method for maximising fat loss when compared with moderate-intensity exercise (links here, here & here). Interestingly, superior fat loss occurs when performing HIIT despite it being shown that attenuation in blood flow to adipose tissue occurs as exercise intensity increases. Still, despite lower fat oxidation rates during exercise for individuals performing HIIT, fat loss is greater over time.
Due to these factors, it is of my opinion that only long-term studies examining meaningful endpoints (i.e. actual fat loss) can get us close to answering this article’s question.
Do acute changes in fat oxidation relate to long-term body composition changes (i.e. fat loss)?
Despite an apparent theoretical basis, acute observations of increased fat oxidation of fasted exercise have failed to translate to any benefit compared with exercise in a fed state. Though a few studies have looked at the effects of body composition changes associated with fasted vs. fed exercise (links here, here and here), to date, only a single study has looked at the chronic effects of fasted vs. fed cardio on body composition changes during energy restriction. In this paper, Schoenfeld and his usual partners in crime looked at the changes in fat mass and fat-free mass following four weeks of aerobic exercise (one hour of steady-state running at ~70% max heart rate, three times per week) in a group of 20 young women during a hypocaloric diet (~500 kcal deficit). Though both fed and fasted groups lost weight and body fat over the four weeks, there were no differences between groups, demonstrating that no body composition advantage occurred for any feeding strategy. Like all research, this study was not without its limitations. Firstly, the duration of testing was relatively short. Should the study have carried on for two or several months, body composition differences between groups may have had time to manifest. Secondly, although a trained and experienced dietitian (Aragon) attempted to control the participants’ dietary intakes, it is possible that inaccurate self-reporting of intakes may have confounded the results. Given that mean weight loss was in fact less than would be predicted by the energy deficit, underreporting of intakes in both groups seems very likely. Furthermore, the method of air displacement plethysmography (i.e. BodPod) to assess body composition has been shown to be highly variable in some populations. Nevertheless, there were still no differences between groups for weight loss. Finally, these results may only apply to women. It would be interesting to see whether these results would transfer to a male population.
In a less recent study, Gillen and colleagues compared the effects of fasted versus fed state high-intensity interval-training on body composition changes in a group of 16 overweight/obese women. Exercise consisted of ten 60-second cycling intervals at 90% maximal heart rate with a 1:1 work/recovery ratio performed three days per week for six weeks. Though dietary control was less stringent than in the Schoenfeld study (participants were given a standardised breakfast and were then instructed to maintain pre-intervention eating habits), body mass remained unchanged after six weeks in both groups. As assessed by DEXA (a more valid measure of body composition that BodPod), fat mass was reduced, and lean body mass was increased, to a similar extent in both groups.
It is worth mentioning that longer-term studies will never be as well controlled as acute interventions; doing so would be a logistical nightmare and prohibitively expensive in most instances. However, owing to their greater external validity (essentially, how well the data can be applied in real-world conditions), long-term studies are arguably more relevant in answering the question of this article. Therefore, taken together, these data from longer term studies signify that there is no difference between fasted vs. fed exercise on body composition changes. That said, the case is far from closed on the matter and more research is needed to give a more definitive answer. Still, with our understanding of human physiology, we can discuss the theoretical implications for the effects of fasted vs. fed cardio on body composition changes.
To paraphrase Krieger, if a study is designed whereby where people do fasted or non-fasted cardio to produce a 500 kcal daily energy deficit, the energy deficit is the same between the groups, and hence weight loss (not taking into account water fluctuations due to glycogen status) will also be the same. If indeed 24-hour fat oxidation makes a difference to fat loss in this example, then the fasted group should lose more fat. However, for greater fat loss to occur in the fasted group, they must also preserve more lean body mass (so that weight loss remains the same). As such, for a given energy deficit, for fasted cardio to result in more fat loss, it must also have a greater protein sparing effect. For this not to be the case would result in a violation of the first law of thermodynamics, which basically states that energy cannot be created nor destroyed.
Now let’s go through each of Krieger’s points and see what the research says regarding them.
(1) From the studies already discussed in this article, as well as other research I looked at in preparation for writing this post, there is no evidence that fasted cardio results in an increased 24-h energy expenditure relative to exercise performed in a fed state. So far, so good.
(2) I’m not aware of any data showing that performing cardio in a fasted state suppresses appetite to a greater degree than non-fasted cardio. In fact, in free-living conditions, the greater glycogen storage following fed cardio (due to the meal’s effect on the sparing of muscle glycogen) may inhibit subsequent food intake to a greater degree, since carbohydrate balance is a predictor of subsequent ad libitum food intake.
(3) To my knowledge, there is no evidence that exercising in a fasted state will have a greater protein sparing effect than that in a fed state, nor can think of a mechanism as to why this may be the case. Interestingly, nitrogen losses (~14 g of amino acids per hour) have been observed during 60 minutes of fasted exercise. That said, the source of the nitrogen losses were not identified so it is not possible to comment on whether this would have a negative effect on muscle tissue. It is likely that, providing dietary protein is adequate, any long-term differences in the quantity of muscle tissue after performing fasted or fed exercise would be negligible. If there were indeed a protein sparing effect associated with either strategy, it would be with performing cardio in a fed state.
Taken together, and to quote Krieger,
I’d agree with these conclusions and so would the weight of the research to date.
Conclusions & practical applications
In summary, despite a logical rationale for superior effects of fasted cardio for fat loss when compared to that performed in a fed state, acute oxidation data as well as longer-term trials have failed to support it. Instead it would appear that there is little to no difference between fasted vs. fed exercise for the goal of fat loss, a conclusion that is supported when looking at things from a theoretical standpoint. As such, if fat loss were the aim, my recommendations would simply be to perform the type of cardio that you prefer and can consistently do on a regular basis.