The surprisingly satiating effects of fasted cardio

Despite exercise increasing daily energy expenditure, and thus, in theory, creating negative energy balance, interventions that have focused on using exercise as a tool for weight loss have revealed^[1] lower weight loss than predicted based on the amount of extra energy expended. This is mostly due to the fact that some^[2] people compensate for the energy expended during exercise by consuming more calories, thus negating any effect of exercise on energy balance. Another compensatory behaviour that can occur^[3] in response to exercise is a reduction in non-exercise physical activity, meaning that people move less in their daily lives. These responses limit the ability of exercise to cause negative energy balance and weight loss. On the other side of the equation, skipping breakfast has been shown^[4] to promote a lower energy intake compared to eating breakfast (see Study Deep Dives #53, Volume 1, “Will eating breakfast keep you lean?”), but it appears to reduce morning physical activity compared to eating breakfast in lean^[5] and obese^[6] participants.

It has been proposed^[7] that the body’s glucose stores, in particular hepatic glycogen (the primary source of blood glucose between meals), plays an important role in regulating energy intake (serving as a kind of “sensor” of energy availability). As depicted in Figure 1, if glucose oxidation is too high, and hepatic glycogen levels are reduced, this could trigger a compensatory increase in energy intake to maintain levels of glycogen above certain levels. Accordingly, there is some evidence^[8] in humans showing that higher levels of glucose oxidation during exercise are associated with higher energy intake after exercise. Therefore, increasing the proportion of energy derived from fat oxidation during exercise (and reducing glucose oxidation), like what happens during fasted exercise^[9], might prevent the compensatory increase in energy intake.

The main goal of the study under review was to assess the role of carbohydrate availability during morning exercise sessions on 24-hour energy balance, in normal weight men.

Weight loss interventions that have used exercise as a weight loss tool have shown mixed results as a consequence of either energy intake compensation or a subsequent reduction in non-exercise physical activity. There is some evidence that suggests how much glucose one uses during exercise is associated with how many calories one consumes after exercise. As performing exercise in the fasted state reduces the relative contribution of glucose to the overall energy demands of exercise, exercising before breakfast, or “fasted cardio,” might prevent or at least ameliorate compensatory increases in energy intake.

This study was preregistered^[10] and had a randomized cross-over design. Twelve young, healthy and physically active males (23 years old, BMI 23.6, 14% body fat) participated in three trials (separated by seven to 30 days): breakfast followed by rest (BR), breakfast followed by exercise (BE) and overnight fasting (12-14 hours) followed by exercise (FE). Trials were done in a random and counterbalanced order. In order to standardize pre-trial food consumption, participants ate the same meal at 8 p.m. the night before the trials. Power analysis based on previous research^[11] showed that 12 participants were needed to detect a 378 kilocalorie difference in energy balance with an 80% chance. To have a complete picture of the day, some data was collected while the participants were in the laboratory and some after they went home.

The basics of the study design are laid out in Figure 2. Participants either consumed 431 kcal breakfast (72 grams of instant refined oats and 360 milliliters of semi-skimmed milk; providing 65 grams of carbohydrate, 11 grams of fat and 19 grams of protein) or rested for two hours before cycling for one hour at 50% peak power output. In the BR condition, after breakfast, participants continued resting. After the exercise or rest, a two-hour oral glucose tolerance test (OGTT) was performed. After the OGTT, participants were given an ad libitum lunch in the laboratory and a weighted food package for consumption over the remaining part of the day (free-living). Daily energy expenditure was calculated via indirect calorimetry (within-laboratory) and estimated based on heart rate with accelerometry (free living).

Blood samples were collected at 40 and 50 minutes of exercise (or rest), and every 10 minutes during the two-hour OGTT. Expired gases were collected every 15 minutes during exercise and hourly during the OGTT. The infusion of labeled glucose, which was used to monitor the substrate utilization by the individuals, was initiated before exercise and continued during the trial. The primary outcomes were the rate of appearance of exogenous glucose during the OGTT and energy balance over 24 hours.

This was a randomized cross-over trial in which 12 normal-weight participants underwent three different trials: breakfast and rest, breakfast followed by exercise, or exercise in the fasted state. Energy intake, energy expenditure, and substrate utilization were measured during exercise and over 24 hours.

The details for the main outcome of the study are shown in the graphs in Figure 3.

FE led to significantly more negative 24-hour energy balance (-400 kcal) than BR (+492 kcal) and BE (+7 kcal). These differences were accounted for by greater exercise energy expenditure in the exercise groups and limited compensation via increased energy intake in the FE group. In other words, participants in the FE group did not eat enough during the rest of the day to compensate for the fact that they had skipped breakfast and exercised.

Performing exercise and skipping breakfast only resulted in partial compensation of calories during lunch (166 kcal), an amount which was lower than the energy that the omitted breakfast would have contained (431 kcal), and there was no further compensation during the day. Therefore, total 24-hour energy intake was significantly lower in FE than BE (-393 kcal), and lower in FE than BR (-148 kcal), although the latter comparison didn’t reach statistical significance.

There was a positive correlation between plasma glucose utilization (representing hepatic sources) and energy intake compensation at lunch in FE (r = 0.62). In other words, the more hepatic carbohydrate used during the fasted exercise, the higher the energy compensation after the exercise.

Energy expenditure during exercise was similar between BE and FE, although carbohydrate utilization was higher in BE and whole-body lipid utilization was higher in FE. No differences in carbohydrate or fat utilization during the OGTT were observed. Free-living physical activity was similar for all trials, and thus daily energy expenditure did not differ between BE and FE.

Omission of breakfast before exercise produced a lower daily energy balance, predominantly due to a lower within-lab fat balance, compared to BE. Conversely, the lower within-lab energy balance in BE compared to BR was entirely due to a difference in carbohydrate balance.

Performing low-intensity exercise without prior breakfast led to lower daily energy intake than performing the same exercise after eating breakfast. There was a positive correlation between hepatic glucose utilization during fasted exercise and energy intake compensation at lunch. Overall, performing exercise before breakfast led to a negative energy balance due to a negative fat balance.

The main takeaway from this study is that, in the short term, performing low intensity exercise and skipping breakfast does not result in complete energy compensation later in the day, and thus, induces a more negative 24-hour energy balance compared to exercising after eating breakfast.

By now, there is plenty of evidence^[12] showing that performing exercise in the fasted state leads to higher fatty acid oxidation during exercise than consuming a meal containing carbohydrate before exercising. This is because the meal consumed before an exercise bout will influence^[13] the proportion of energy used from a specific macronutrient. Thus, a carbohydrate-rich meal (like the one used in the study under review) will increase glucose oxidation during exercise, and this will displace fat oxidation. In the fasted state, the body is primarily using fatty acids, so exercising during this period will increase the utilization of energy derived from fat stores. At the same time, there will be less glucose oxidized. Effectively, this would create a temporary negative fat balance. However, it is important to take into account that this does not necessarily equate to more fat mass lost over time, as oxidation of carbohydrate and fat after the exercise during the day might change and any difference could be balanced out^[14][15] through increased energy intakes and the person achieving a neutral energy balance, irrespective of where this energy comes from. If the omission of breakfast and exercise before breakfast is not fully compensated later in the day, the increased energy expenditure will, as it has been demonstrated in the study under review, induce a more negative fat balance, compared to exercising in the fed state acutely.

As of now, there are no long-term studies addressing if these differences would translate to significant differences in fat loss, though. One study^[16] showed no difference in fat loss between the results of performing low intensity exercise in the fed versus fasted state. However, this study had major limitations, such as a small sample size, a duration of only four weeks, lack of diet control, exercise was performed only three times per week, and there was no control for the menstrual cycle. Another^[17] six-week study in women (again with a small sample size and no control for the menstrual cycle) didn’t find any difference in fat loss between performing high intensity cardio fed or fasted. Despite these limitations, these studies suggest that pragmatically, the recommendation to perform fasted cardio seems to lead to no better fat loss results over four to six weeks than doing it in the fed state. Other tightly controlled studies performed in men^[18] and women^[19], consuming eucaloric diets, have shown similar results to the study under review: fasted cardio leads to a more negative 24-hour energy balance, due to a negative fat balance, even when energy intake is matched to expenditure. In these studies, the increase in fat oxidation was correlated with the transient carbohydrate deficit. Future studies should test the effects of fasted cardio over longer time periods and the effects of specific dietary patterns, such as low-carbohydrate diets.

A very interesting finding of the study under review is the correlation between carbohydrate utilization from liver glycogen during fasted exercise and feeding compensation during the day. This suggests that the higher the glucose oxidation during the exercise, and potentially the greater the depletion of liver glycogen, the larger the compensation in energy intake after the exercise. Potentially, increasing fat oxidation during exercise and therefore reducing liver glycogen depletion could reduce compensatory energy intake and hence aid weight loss. There is some evidence that some genes involved in lipid metabolism and insulin sensitivity in adipose tissue are affected^[20] by performing fasted versus fed exercise. Over time, fasted exercise could promote higher fat oxidation rates and reduce compensatory energy intake. Nevertheless, as these data are all acute, it is not known how much time is needed to observe these effects.

Finally, it is worth noting that this study was performed in healthy lean men (14% body fat) and measurements were only done on one day. It is not known how much difference would be observed in other populations (like people with obesity) or if these differences would change over time. Longer duration studies are needed to determine the relevance of the findings over time and other populations, as well as their real-world effects on body composition, i.e. the amount of stored body fat.

As expected, fasted exercise produces higher fat oxidation levels compared to fed exercise. While acute, well-controlled studies have found that this produces a more negative fat balance per day, the albeit limited contemporary evidence seems to suggest that this will not translate to long-term benefits for fat loss. Moreover, in some cases, fat oxidation could increase later in the day to compensate for the lower fat oxidation during fed exercise. Overall, the study seems to confirm the hypothesis that the degree of energy compensation is correlated with the use of glucose during fasted exercise. However, this effect could change over longer periods of time.

Though much of the discussion on fasted exercise seems to be focused on fat loss, it is equally important to consider its potential consequences on metabolic health. Some studies have shown that performing fasted exercise improves glucose tolerance^[21] and upregulates^[22] enzymes involved in fatty acid oxidation. This would be a beneficial effect, as metabolic inflexibility, the inability to switch between using glucose and fatty acids, has been implicated^[23] in metabolic diseases.

Usually, when the body is fasted, body tissues primarily use fatty acids for energy. However, upon consuming carbohydrate, the use of fatty acids is rapidly reduced and the body switches to use glucose. The ability to make this switch appears to be impaired in people with obesity and metabolic syndrome. Therefore, besides weight loss, performing exercise in the fasted state could be beneficial^[24] for improving metabolic flexibility. However, long-term studies are needed to determine if this strategy would yield tangible health benefits. In this context, it is interesting to note that, in the current study, no differences in glucose tolerance were observed. It is possible that a longer study duration is needed to observe an effect on this parameter, or that the type and intensity of the exercise plays a bigger role.

The study at hand is a powerful reminder of the potential importance of how energy deficits are achieved, i.e. via a negative glucose/glycogen or fat balance. In other words, next to the commonly heard term “energy balance,” researchers can also define a “carbohydrate balance” (relating to the body’s carbohydrate stores) or “fat balance” (relating to the body’s fat stores). Ideally, one would like to manipulate the latter for fat loss and weight loss-related metabolic improvements. Fasted cardio could promote a more negative fat balance over time than cardio in the fed state. However, fat (and energy) intake form part of the other side of the equation, so changes in fat mass are derived from the balance between the two. For fat loss, increases in fat oxidation must be accompanied by decreases in fat storage. Based on this and previous data, it would then be beneficial to focus on manipulating fat stores (promoting utilization) and glucose stores (promoting storage instead of utilization) differentially. Combining different types of exercises, such as low intensity cardiovascular exercise in the fasted state and high intensity resistance exercise, could aid in this endeavour.

Irrespective of its effects on fat loss, fasted exercise could promote beneficial metabolic adaptations. However, much of the data currently available comes from short-term studies. Hence, longer term studies are needed to determine if the acute and short-term effects are clinically relevant in the long term. Current understanding of the interaction between exercise and food intake may benefit from a model that incorporates the role of specific energy stores (glucose or fat) and how their manipulation can contribute to overall health and beneficial changes in body composition.

Q. Do fasting duration, baseline diet, and meal/exercise timing influence the response to fasted training?

Most studies determining the effects of fasted training have used designs in which participants fast overnight (12-14 hours). However, fasting for longer time periods increases^[25] fat mobilization (longer than 16 hours). In theory, exercising during this period could promote an even higher rate of fat oxidation than exercising after an overnight fast. However, the amount of fat oxidized not only depends on the available free fatty acids in plasma but on the energy demands of the exercise. Moreover, overall fat oxidation during fasting at specific time points probably depend on the baseline diet. For example, shorter fasting time would be needed during a low carbohydrate diet, as hepatic glycogen is already low.

The effects of exercise on weight loss have been greatly reduced by energy intake compensation, as well as reduction in daily physical activity. There is some evidence suggesting that hepatic glycogen, which maintains blood glucose, might determine the compensatory increase in calories observed in some people.

This study aimed to determine the role of carbohydrate availability during exercise on net energy balance over 24 hours, using fasted and fed exercise, as well as glucose tolerance after exercise between these interventions.

Over the whole day, exercising in the fasted state led to a more negative energy balance, explained primarily by a more negative fat balance. When exercising in the fasted state, energy was not completely compensated later in the day. Moreover, in conjunction with previous research, the study under review is consistent with the hypothesis that glucose utilization during exercise and its downstream effects on hepatic glycogen stores is correlated with energy compensation after exercise.

These results suggest that fasted exercise might be beneficial for weight loss, as well as metabolic health, by increasing fat oxidation and daily energy balance. Over time, glucose utilization during exercise might decrease as an adaptation to fasted exercise and hence, subsequent energy intake could be reduced. However, the available short-term data suggests that any difference might wane over time. Long-term studies are needed to corroborate these findings and determine if this effect would translate to a clinically meaningful difference.