The impact of prolonged fasting on 24h energy metabolism and its 24h rhythmicity in healthy, lean males: A randomized cross-over trial

Objective: Human energy expenditure and substrate oxidation are under circadian control and food intake is a time cue for the human biological clock, leading to 24h feeding–fasting cycles in energy and substrate metabolism. In recent years, (intermittent) fasting protocols have also become popular to improve metabolic health. Here, we aimed to investigate the impact of food intake on the 24h patterns of energy metabolism as well as to provide data on the timeline of changes in energy metabolism that occur upon an extended period of fasting. Research design and methods: In a randomized, cross-over design, twelve healthy males underwent a 60h fast which was compared to a 60h fed condition. In the fed condition meals were provided at energy balance throughout the study. Conditions were separated by a two week period of habitual diet. Volunteers resided in a respiration chamber for the entire 60h to measure energy expenditure and substrate oxidation hour by hour. Volunteers performed a standardized activity protocol while in the chamber. Blood samples were drawn after 12, 36 and 60h. Results: Immediately following the breakfast meal (in the fed condition), fat oxidation became higher in the fasted condition compared to the fed condition and remained elevated throughout the study period. The initial rapid increase in fat oxidation corresponded with a decline in the hepatokine activin A (r = -0.86, p = 0.001). The contribution of fat oxidation to total energy expenditure gradually increased with extended abstinence from food, peaking after 51h of fasting at 160 mg/min. Carbohydrate oxidation stabilized at a low level during the second day of fasting and averaged around 60 mg/min with only modest elevations in response to physical activity. Although 24h energy expenditure was significantly lower with prolonged fasting (11.0 ± 0.4 vs 9.8 ± 0.2 and 10.9 ± 0.3 vs 10.3 ± 0.3 MJ in fed vs fasting, day 2 and 3 respectively, p < 0.01), the 24h fluctuations in energy expenditure were comparable between the fasted and fed condition. The fluctuations in substrate oxidation were, however, significantly (p < 0.001 for both carbohydrate and fat oxidation) altered in the fasted state, favouring fat oxidation. Conclusions: Energy expenditure displays a day–night rhythm, which is independent of food intake. In contrast, the day–night rhythm of both carbohydrate and fat oxidation is mainly driven by food intake. Upon extended fasting, the absolute rate of fat oxidation rapidly increases and keeps increasing during a 60h fast, whereas carbohydrate oxidation becomes progressively diminished. Trial registration: www.trialregister.nl NTR 2042.