Time-Restricted Eating Plus Exercise Cuts Fat Mass, Network Meta-Analysis Finds

Time-restricted eating appears to give people an extra edge when they exercise, but the type of workout matters less than many fitness enthusiasts might assume. A new systematic review and network meta-analysis published in Frontiers in Nutrition finds that confining daily eating to a window of twelve hours or less added an average fat-mass reduction of roughly 1.85 kilograms beyond what the same exercise programme achieved on its own. The analysis, which pooled fifteen unique randomised trials, also ranked the combined approaches against one another and found that concurrent training, high-intensity interval training, and aerobic exercise held the highest probabilities of producing the largest fat losses, while resistance training stood out for a different reason: it had the most repeated evidence for preserving muscle mass and strength.

The research team, led by Hui Li of Dalian Polytechnic University with colleagues from Liaoning Normal University and Hanyang University, searched six databases covering the period from 2016 to 20 July 2026. Their initial sweep returned 5,749 records. After removing 1,880 duplicates and screening nearly 3,900 titles and abstracts, the reviewers assessed thirty-one full reports and ultimately included seventeen reports representing fifteen distinct randomised trials. Careful trial linkage prevented double-counting: the Moro 2021 long-term follow-up was tied to the original Moro 2016 cohort, and the TREHIIT trial’s primary report and two-year follow-up were linked under a single identifier, so that only one prespecified time point from each cohort entered any quantitative model.

Eligible trials enrolled adults aged eighteen or older who followed a daily eating window of twelve hours or less for at least two weeks alongside a prospectively prescribed exercise programme. The populations ranged widely, from resistance-trained men and elite male cyclists to women with overweight or obesity and inactive women with obesity. Ramadan fasting, alternate-day fasting, 5:2 regimens, and acute fasted-versus-fed experiments were excluded, as were quasi-randomised studies and conference abstracts lacking sufficient data. The reviewers compared seven intervention nodes: usual diet or control, time-restricted eating alone, exercise alone, and time-restricted eating combined with aerobic exercise, resistance training, high-intensity interval training, or mixed and concurrent training.

The headline pairwise result came from eight independent trials that contributed compatible fat-mass contrasts. Time-restricted eating plus exercise reduced fat mass by a mean difference of 1.85 kilograms more than the same exercise without the eating window, with a 95 percent confidence interval of 2.34 to 1.35 kilograms. Heterogeneity was essentially absent: Cochran’s Q yielded a p-value of 0.822, I-squared was zero percent, and tau-squared was 0.00. The 95 percent prediction interval, spanning 2.46 to 1.23 kilograms, suggested that a new trial would very likely show a benefit in the same direction. The estimate proved robust when the reviewers varied the assumed within-person correlation used to reconstruct change-score variances from 0.30 to 0.70, with pooled estimates shifting only between 1.86 and 1.84 kilograms.

The network meta-analysis then placed all seven nodes on a single comparative scale, using usual diet or control as the reference. Against that baseline, time-restricted eating plus concurrent training showed the largest estimated fat-mass reduction at 3.32 kilograms, followed by the combination with high-intensity interval training at 3.10 kilograms, aerobic exercise at 3.03 kilograms, and resistance training at 2.79 kilograms. Time-restricted eating alone was estimated at 1.51 kilograms and exercise alone at 1.17 kilograms. P-scores, a frequentist ranking metric, placed concurrent training first at 0.838, followed by HIIT at 0.757, aerobic exercise at 0.742, and resistance training at 0.650. Residual network heterogeneity was not detected, with Q equal to 3.29 on six degrees of freedom and a p-value of 0.772.

Those rankings, however, come with a substantial caveat that the authors themselves emphasise. The confidence intervals for the active combined modalities overlapped, meaning the data cannot distinguish cleanly among them. No trial in the network directly randomised participants between two different time-restricted-eating-plus-exercise combinations, so all cross-modality comparisons rest on indirect evidence. With only eight trials spread across seven nodes, the P-score ordering is indicative rather than definitive. The reviewers rated the certainty of these comparative rankings as very low using the GRADE and CINeMA frameworks, citing serious within-study bias concerns, very serious indirectness, and serious imprecision.

The pattern of included studies also shaped what the network could say. Resistance-training trials primarily enrolled younger, trained adults, whereas the aerobic, HIIT, and mixed-modality trials more often recruited women with overweight or obesity. This uneven distribution of potential effect modifiers across the network means the apparent superiority of the higher-ranked modalities may partly reflect who was studied rather than a true biological difference between training types. The reviewers evaluated transitivity across age, sex, BMI status, training status, metabolic risk, eating-window duration, intervention duration, energy-intake control, exercise supervision, and body-composition measurement method, and reflected these imbalances directly in their certainty ratings and cautious interpretation.

Beyond fat mass, the review examined what happened to lean tissue, metabolism, and athletic performance. Fat-free mass was maintained or increased in most resistance-training trials: one study reported increases of roughly two to three percent across groups, while several others found no material loss attributable to time-restricted eating. One trial reported a decrease in fat-free mass with the eating window alone, an increase with resistance training alone, and maintenance with the combined intervention. Body-fat percentage and body weight generally declined when the eating window created an additional energy deficit, and one cycling study in elite athletes recorded a 1.19-percentage-point greater reduction in body-fat percentage under time-restricted eating. Not every combination added benefit, though: a randomised crossover trial in male recreational runners found no additional body-composition or aerobic-capacity advantage from the eating window.

Metabolic and performance outcomes were reported less consistently across trials. One HIIT trial in women with overweight or obesity found improved glycated hemoglobin in the combined arm, and a high-intensity functional training study in inactive women with obesity reported larger changes in fasting glucose, insulin, and HOMA-IR with the combination than with either component alone. A flexible eight-hour eating window paired with aerobic exercise in women aged forty to sixty improved fat mass and related metabolic measures. Performance results were modality-specific: resistance-training trials generally preserved or improved strength, while endurance trials showed either similar improvements under both dietary conditions or improved power relative to body weight after weight reduction. Glycemic and performance findings carried very low certainty because few trials used common outcome definitions.

The authors conclude that time-restricted eating adds a modest fat-mass reduction when layered onto an exercise programme, and that the choice of modality should follow the individual’s primary goal rather than the fat-mass league table alone. For adults with overweight or obesity prioritising fat loss and glycaemic improvement, the eating window combined with aerobic exercise, HIIT, or a mixed programme may be considered, subject to exercise tolerance, supervision, and medical risk. For trained adults focused on preserving lean mass and strength, resistance training holds the strongest repeated direct evidence, with adequate protein and total energy intake remaining important. The direct pooled fat-mass effect was rated low certainty because several variances required reconstruction and most trials had some risk-of-bias concerns, most often involving incomplete reporting of allocation concealment, attrition, or prespecified analysis plans. The protocol was prospectively registered in PROSPERO, and the authors note that fewer than ten independent studies contributed to the primary comparison, which precluded funnel-plot assessment of small-study effects.

Subject of Research: Comparative effects of time-restricted eating combined with different exercise modalities on body composition, metabolic health, and physical performance in adults

Article Title: Comparative efficacy of time-restricted eating combined with different exercise modalities for improving body composition, metabolic health, and physical performance in adults: a systematic review and fat-mass network meta-analysis

Article References: Li, H., Lu, C., Qin, G., & Li, D. (2026). Comparative efficacy of time-restricted eating combined with different exercise modalities for improving body composition, metabolic health, and physical performance in adults: a systematic review and fat-mass network meta-analysis. Frontiers in Nutrition, 13, 1966901. https://doi.org/10.3389/fnut.2026.1966901

Image Credits: AI Generated

DOI: 10.3389/fnut.2026.1966901

Keywords: time-restricted eating, intermittent fasting, exercise, fat mass, network meta-analysis, resistance training, HIIT, aerobic exercise, concurrent training, body composition, metabolic health, systematic review