Hypercaloric 16:8 time-restricted eating during 8 weeks of resistance exercise in well-trained men and women

ABSTRACT

Background

The effectiveness and practicality of time-restricted eating (TRE) when trying to maximize muscle mass and strength is unclear. Thus, we examined the effects of a hypercaloric 16:8 TRE approach during supervised progressive resistance exercise.

Methods

Seventeen healthy and well-trained men (n = 10) and women (n = 7) were randomly assigned to TRE or control (FED). Both groups consumed a 10% hypercaloric high-protein (2.2 g/kg/d) diet and performed supervised resistance exercise 4× per wk for 8 wk. TRE consumed all calories within an 8 h window starting at least 1 h post-exercise, while FED consumed the same number of calories throughout the day.

Results

Eating windows were significantly different (TRE: 7.9 ± 0.1 h vs. FED: 13.2 ± 0.6 h). Calorie, carbohydrate, fat, and protein intake did not differ statistically between groups. Total exercise volume was significantly lower in TRE than FED (6,960 ± 287 vs. 7,334 ± 289 repetitions), as were subjective daily energy ratings (week 4 = -1.41; p = 0.04, week 8 = -1.04; p = 0.06). Both groups increased maximal upper and lower body strength (1RM) and muscular endurance (ME); however, gains in squat 1RM were 4.0 ± 1.9 kg lower in TRE (p = 0.05). Both groups increased fat-free mass similarly (TRE: 2.67 kg; FED: 1.82 kg, p = 0.04), but FED added 1.4 ± 0.6 kg more fat mass (p = 0.04). Subjective mood and sleep ratings did not change in either group.

Conclusions

16:8 TRE is viable during periods of muscle size, strength, and endurance development in well-trained young men and women when engaging in progressive resistance exercise and eating in a caloric surplus with adequate protein. However, the differences in total training volume, squat 1RM, fat mass accumulation, and energy are notable and practically relevant. These findings should be considered within the broader context of an individual’s goals, lifestyle, preferences, and exercise demands.

1. Introduction

Time-restricted eating (TRE) is a term for dietary strategies whereby energy intake is restricted to a specific window [1–5]. The 16:8 TRE method requires all caloric consumption to occur daily during an 8 h window. Most research on this popular approach has focused on the body composition, inflammation, cardiovascular health, and immune function implications when using TRE to restrict caloric intake [1–13]. A 2021 systematic review and meta-analysis reported that TRE is effective for reducing body mass (BM) and body fat (BF%) while preserving fat-free mass (FFM), relative to non-fasting control diets when combined with resistance exercise (RE) [14]. A prior review by Keenan and colleagues also reported similar findings with various TRE formats in sedentary and sub-elite populations [15]. Tinsley’s group and others showed comparable findings when examined across multiple participant groups, demographics, periods, and TRE formats [1–3,5,10]. As public health tools, effective tactics for decreasing fat mass (FM) while preserving FFM are needed. However, resistance exercise (RE) and nutritional methodologies for maintaining muscle mass and strength differ from those aimed at maximally gaining them. This distinction matters as skeletal muscle size and strength are critical to successful health, aging, and athletic performance [16,17]. Yet, to date, no studies have examined the effectiveness of a 16:8 TRE approach when attempting to maximize RE-induced muscle hypertrophy and strength.

Numerous nutritional factors, such as total energy and macronutrient balance, distribution, and timing, influence the rate of muscle mass and strength gains [18]. A hypercaloric diet facilitates recovery and maximizes the anabolic response to RE [19]. Dietary protein is needed to deliver the essential amino acids involved in muscle protein synthesis. Current recommendations for protein intake during RE-induced muscle hypertrophy suggest dosing between 20 and 40 g every 3–4 h distributed evenly across the day [20–22]. Carbohydrates provide the primary substrate for energy during RE while potentially promoting anabolism via blunting muscle protein breakdown [23–25]. Yet, the notion that protein and carbohydrate ingestion must occur immediately post-RE to maximize the anabolic response has long been dispelled, with more contemporary data indicating that total intake across the 24 h period is more important [26].

Few studies have tested such a conclusion in the presence of caloric restriction across long windows (i.e. 16 h of daily fasting), highly trained individuals, whole-body measures (i.e. not molecular markers), and a longitudinal timeframe (i.e. not acute) [2,5,10,13,27]. Tinsley and Moro concluded RE with TRE did not reduce strength and muscle mass in trained males and females in interventions ranging from 8 wk to 12 mo [2,5,10,28]. In 2019, Tinsley reported that total body FFM and muscle thickness increased with TRE when consuming an isonitrogenous (1.6 g/kg) and isocaloric diet compared to the control [5]. However, all training occurred in the fed state with the primary aims of reducing fat mass and improving cardiometabolic health, rather than solely maximizing muscular growth [2,5,10]. Currently, the study by Stratton et al. [28] is the only investigation in which diet was controlled with a specific 25% caloric deficit. Conversely, Trabelsi et al. [13] reported hypertrophy was not affected in Turkish bodybuilders while training in either a fasted or fed state during the 29 days of Ramadan. While participants fasted for 15 h, interpretability was limited as exercise performance was not reported and body composition variables were predicted from skinfold assessments [13]. Thus, the long-term implications of performing resistance training in the fasted state remain unknown. Collectively, the effectiveness and practicality of a 16:8 TRE approach in a caloric surplus to maximally increase muscle mass and strength remains unexplored.

2. Methods and materials

2.1. Study design

Seventeen well-trained and healthy men and women (Table 1) performed 8 weeks (4× per wk; 32 total bouts) of progressively overloaded RE while consuming a hypercaloric (10% surplus) and macronutrient-controlled diet (protein = 2.2 g/kg/d and fat = 25% of the total energy intake) in either a traditional (FED) or 16:8 TRE fashion between summer of 2019 and winter of 2020 prior to the COVID-19 pandemic. The group assignment was randomized via a random sequence generator. The research team was segmented to ensure blinding. Performance testing was conducted by the same research team member for all participants pre- and post-intervention, and that person was not involved in daily dietary support or post-study data analysis. Similarly, the dieticians worked remotely with participants and were not involved in recruitment, training, testing, or data analysis. The California State University, Fullerton, institutional review board approved all procedures, and informed consent was obtained before starting any data collection.

Table 1.

Participant characteristics at baseline.

	TRE		FED
	M (n = 6)	F (n = 4)	M (n = 4)	F (n = 3)
Age (y)	25.8 ± 3.9	24.2 ± 1.4	25.1 ± 3.0	22.8 ± 2.2
Body mass (kg)	84.6 ± 7.6	59.6 ± 4.4	82.9 ± 5.7	64.3 ± 2.6
Height (cm)	180.8 ± 5.5	162.0 ± 5.4	176.4 ± 3.7	162.5 ± 3.1
Body mass index (kg/m2)	25.9 ± 2.0	22.6 ± 1.7	26.6 ± 2.0	24.4 ± 0.3
Fat-free mass index (kg/m2)	20.6 ± 1.6	18.1 ± 1.0	23.1 ± 1.2	17.2 ± 1.2
Body fat (%)	20.2 ± 2.7	20.2 ± 1.8	13.3 ± 5.5	29.4 ± 4.7
Bench press strength (kg/kg body mass)	1.2 ± 0.1	0.8 ± 0.2	1.4 ± 0.1	0.6 ± 0.1
Squat strength (kg/kg body mass)	1.6 ± 0.2	1.5 ± 0.2	1.6 ± 0.3	1.3 ± 0.1

Values displayed in mean ± SD.

We defined well trained as having performed RE ≥3× per wk (including ≥1× per wk of lower body-specific RE) for the previous 12 months and able to achieve a minimum level of maximal strength (1RM) standards for back squat (1.5× bodyweight for men, 1.0× bodyweight for women) and bench press (1.0× bodyweight for men, 0.55× bodyweight for women). Additional qualification criteria and controls included the following:

1. Consumed a non-specialized (e.g. not vegan, ketogenic, Mediterranean) mixed macronutrient diet for the previous 3 months.

2. <35% body fat, measured via air displacement plethysmography (ADP).

3. A lifestyle that enabled a reasonably consistent 7+ h of sleep per night.

4. Free of any metabolic disorders and any current joint, musculoskeletal, or neuromuscular injuries.

5. Reported no current or past use of any performance-enhancing drugs.

6. No use of any recreational drugs, prescription or non-prescription, legal, or illegal, chronically or in the past month.

7. No consumption of antibiotics in the last 6 months.

The first visit included an introduction to the research team and informed consent, medical, exercise, nutrition, mood, sleep, and menstrual cycle history questionnaires. Mood, sleep, and menstrual cycle questionnaires were then repeated at mid (4 wk) and post-time points (8 wk). Nutrition tracking occurred daily, starting 7 days before pre-testing through the final day of testing. Body composition testing occurred at pre-, mid, and post-time points. Performance testing occurred only at pre- and post-time points. All data collection visits were 48–72 h apart from their prior visit or training session. All participants were required to refrain from caffeine and food but were permitted to consume water, during the 12-h period before body composition testing. FED was permitted to consume food between body composition and performance testing, while TRE remained fasted.

TRE (n = 10; 6 males, 4 females) performed RE before noon and consumed all daily macronutrients within an 8 h eating window at least 1 hour after training. FED (n = 7; 4 males, 3 females) also performed RE before noon but consumed all daily macronutrients throughout the day at their discretion. They were required to consume at least one full meal before training. However, the total number of meals per day was not regulated. All participants, regardless of group, consumed 0.35 g/kg of whey protein (Legion Athletics) post-RE. Therefore, the total amount of whey protein provided to each participant was adjusted weekly if BM changed. FED consumed their whey protein during or immediately after each RE session, while TRE did not ingest theirs until their eating window started later in the day, at least 1 h post-RE. Participants were required to refrain from utilizing any supplements beyond the whey protein provided by the research team and from engaging in any additional physical exercise (outside of activities of daily living). All RE sessions were performed in the laboratory and overseen by certified strength and conditioning specialists. Body mass and nutritional adherence were collected before, and RPE (rating of perceived exertion, 1–10) was recorded at the end of each RE session. Participants were blinded to results in order to reduce behavior changes based on body weight.

2.2. Body composition testing

Body composition was tested five times throughout the study (2× Pre, 1× Mid, and 2× Post). Participants were instructed not to eat within 12 h of testing and to be hydrated, which was analyzed using urine-specific gravity (<1.030). If under-hydrated, water was consumed until hydration was sufficient. One failed test occurred; the participant passed soon after hydrating. Height was measured using a stadiometer (Sec, Ontario, CA) and anthropometric BM through an electronic scale (ES200L; Ohaus Corporation, Pinebrook, NJ). Body volume was assessed via a BODPOD 250 (Indianapolis, IN). The BODPOD was calibrated by manufacturer specifications before each measurement. Participants were asked to wear tight-fitting compression and a provided swim cap. They were required to sit quietly and still in the BODPOD for ~2 min. Lung volumes were estimated via the internal software. Bioelectrical Impedance Spectroscopy (BIS) (Impedimed SFB7, Carlsbad, CA) was performed to account for total body water in FM calculation. The BIS device emits 256 measurement frequencies ranging from 4 to 1000 kHz using cole modeling [29] and mixture theories [30] to predict body fluids rather than regression equations in other impedance technologies [31]. Participants were instructed to lay supine on a nonmetallic surface with shoes, socks, and jewelry removed. Lead sites (left hand and foot) were cleaned with an alcohol swab and let dry for 30 s. Electrodes were then attached 5 cm apart per manufacturer recommendations. Participants were instructed to remain calm and still, while measurements were taken. Values were extracted and then utilized in a three-compartment model using the Siri equation [32] (FM (kg) = 2.118 (body volume) − 0.78 (total body water) − 1.351 (body mass)). Body composition at pre and post were performed in duplicate and values were averaged for analysis. The same technician performed all calibration and measurements.

2.3. Maximal strength (1RM) and muscular endurance (ME)

Physical performance tests for maximal strength (1RM) and muscular endurance (ME) were completed at pre and post 8 weeks of RE after the duplicate body composition measurement. 1RM tests were performed in order of back squat, bench press, and leg extension. After 1 RM testing, leg press and bench press ME tests were performed. All maximal tests were administered by the same researcher and by methods previously described by our research group [33]. All loads were adjusted by the researcher and not discussed with the participants.

For all 1RM testing, participants began with 5 min of aerobic exercise followed by a 10-min standardized whole-body dynamic warm-up routine. Participants then completed 4 exercise-specific warm-up sets: 10 repetitions at 50% of estimated 1RM, 8 repetitions at 60%, 5 repetitions at 75%, and 1 repetition at 90%, with 2–3 min in between each set [33]. Participants performed 1 repetition 5–10 lb less than their predicted 1RM. 5–10 lb was then added and a single repetition was attempted until failure. Failure was deemed as the inability to complete a repetition through the full range of motion without assistance. Three minutes of rest were given between each 1RM attempt and 5 min were given as rest in between each test. All testing parameters (shoes, accessories, food consumed before (FED only), grip width, stance width) were recorded and standardized during post-testing.

ME tests were performed following a short warm-up of 3–7 repetitions of the participant’s testing load. The bench press test required performing as many repetitions as possible at 50% of 1RM until failure. The leg press ME test required participants to perform repetitions to failure with a load relative to their back squat 1RM. Males performed the test with a load of 150% of 1 RM while females used 100% of back squat 1RM. Pre- and post-testing loads were related to the 1RM loads achieved at each time point. A metronome was set to 60 bpm, and participants were instructed to complete the concentric and eccentric phases across 3 beats. Participants were allowed to practice completing the lifts on tempo before test administration during their warm-up. Failure was defined as the inability to complete the repetition within two consecutive beat intervals [33]. The number of successful repetitions was multiplied by the absolute load (kg) to generate a final composite volume load score.

2.4. Diet

Energy intake was calculated as basal metabolic rate (BMR) [34], multiplied by 1.15 for a 15% thermic effect of food, a pre-set physical activity level factor of 1.25, and then multiplied by 1.1 to achieve a 10% energy surplus. BMR was estimated using the Katch McArdle equation [34]. The intensity factor of 1.25 (moderately active) was chosen to account for the active lifestyles of the participants and ensure a caloric surplus was achieved. Participants were assigned to a remote dietitian and communicated weekly to ensure adherence, provide professional consultation, and weekly progress-based adjustments. Each dietitian was available as needed including food choices, recipes, and participant-specific coaching weekly. Participants were instructed to record all energy intake with specific time stamps using food tracking software (MyFitnessPal, San Francisco, CA). Groups were prescribed the same total macronutrient intakes relative to their body composition and provided education about measuring and recording food intake within the software a week before participating. Participants were instructed to weigh and track their intake immediately upon consumption. Further resources were provided to aid in estimating amounts when a food scale was not readily available. Nutritional adherence was calculated as the number of days with complete data entry, as defined by having all meals and snacks logged with appropriate time stamps.

Kcals and macronutrients were adjusted each week relative to the participant’s body weight recorded prior to each training session. If a participant’s average weekly body weight did not increase by >1%, energy intake was increased by 10% before beginning the next week of training. Energy intake remained constant until body weight did not increase 1% within each training week. Fat intake was set at 25% of the total energy intake, protein intake was set at 2.2 g/kg/d, and carbohydrate intake was matched to fulfill the remaining energy intake [24]. All macronutrients and energy prescriptions were communicated and updated weekly relative to BM gain or lack thereof. TRE was instructed to delay all energy intake until 1 h after training to initiate the defined 8 h eating window. FED was instructed to distribute their macronutrient intake approximately equally over four meals distributed across the day.

2.5. Training program

Participants completed four training sessions per wk for 8 wk, alternating between two workouts (A-B-A-B). Training sessions occurred at similar times of the day, within 3–4 h of waking, and separated by 24–48 h. Each session began with 5 min of low-intensity cardiorespiratory exercise, followed by 10 min of a standardized dynamic warm-up protocol including 10 repetitions (per direction and/or per limb) of arm circles, inchworms, walking high knees, walking hamstring stretch, and walking lunges. Additionally, two warm-up sets were performed before each exercise’s work sets. The first warm-up set was performed at 70% of the estimated 1RM for five repetitions and the second at 80% estimated 1RM for three repetitions. During working sets, participants were instructed to perform as many repetitions as possible without failing any repetitions, and a tempo of 1–2 sec per concentric and eccentric phase was implemented to ensure relative velocity profiles. Initial sessions were limited to an RPE of 6/10 until participants were calibrated to the protocol. Subsequent training loads were adjusted based on the number of repetitions performed during the work sets using an autoregulatory technique referred to as autoregulatory progressive resistance exercise (APRE). If a participant exceeded the repetition range in any set by two repetitions, the load was increased by 2.5%; if the participant fell below the repetition range, the weight was decreased by 10%. All loads were rounded to the nearest 5 or 10 lb increment. Session RPE was recorded after each training session on a scale of 1–10. A certified strength and conditioning specialist was present during each training session to validate proper technique and complete movements. The training program is displayed in Table 2.

Table 2.

Resistance exercise program.

	Exercise		Sets	Repetition range	Rest interval (minutes)
Workout A		1) Romanian deadlift	4	8–12	3
		2) Leg extension	5	8–12	3
		3) Barbell bench press	4	6–10	3
		4) Neutral grip pulldown	4	8–12	3
		5) Seated calf raises	5	8–12	2
		6) Cable lateral raise	3	8–12	2
Workout B		1) High-bar squat	4	5–8	3
		2) Lying leg curl	5	8–12	3
		3) Barbell overhead press	3	6–10	3
		4) Seated row	3	8–12	3
		5) Cable chest fly	5	6–10	3
		6) Triceps pushdown	2	8–12	2
		7) Seated dumbbell curl	2	8–12	2

2.6. Questionnaires

Medical, exercise, and dietary history questionnaires were administered during the intake process to establish participants who were within the inclusion and exclusion criteria and equipped to comply with the rigors of the study design. Mood, sleep, eating, and menstrual cycle questionnaires (Supplemental Table S1) were administered in Pre, Mid, and Post to assess the effects of the intervention on daily living and experience beyond the laboratory. Nutritional difficulty was measured on a subjective rating scale of 1–10 in which higher values represented a higher degree of difficulty in adhering to the dietary strategy and recommended energy intake. A single researcher distributed and collected the responses. All other researchers were blinded until analysis.

2.7. Statistical analysis

Data were analyzed in R (v. 4.2.1). Outcomes with multiple values per group were evaluated using linear mixed-effects models (nlme package [35], v. 3.1–157), supplemented by analysis of covariance (car package [36], v. 3.1–0) with baseline values included as a covariate. For linear mixed-effects models, a random intercept for participant and a first-order autoregressive (AR1) variance-covariance matrix were used. These models were fitted by maximizing the restricted log-likelihood (REML). In all models, the reference group for the condition was FED, the reference group for time was the baseline/pre-time point, and the reference group for sex was female. Model assumptions were examined through graphical methods (i.e. residuals vs. fitted plots and quantile–quantile plots). Model coefficients were summarized and visualized using the sjPlot [37] package (v. 2.8.14).

Outcomes with a single value per group were compared using either Welch’s independent sample t-tests or Wilcoxon rank sum tests as appropriate. When outliers or normality violations were present, the Wilcoxon rank sum test was used rather than Welch’s t-test. Both tests were performed using the rstatix package [38] (v. 0.7.0), and appropriate effect sizes (Cohen’s d or Wilcoxon r) were generated. Individual responses were visualized using the ggplot2 [39] package (v. 3.4.0). Values are presented in mean ± SD unless otherwise noted. Statistical significance was accepted at p <0.05.

3. Results

3.1. Nutritional intake

The average eating windows (Table 3) were 7.9 ± 0.1 h (median ± IQR) for TRE and 13.2 ± 0.6 h for FED (p <0.001, Wilcoxon r 0.83 [large]). Tracking adherence did not differ between groups (TRE 94.1 ± 4.7%, FED 97.0 ± 2.3%; p = 0.11, Cohen’s d 0.79 [moderate]). From the linear mixed-effects model, there was no effect of group (p = 0.40), nor interaction of group and time (p = 0.64), for nutrition difficulty ratings. However, a main effect of time was present (p <0.001). Follow-up pairwise comparisons indicated that nutrition difficulty ratings were higher at mid (4.6 ± 2.7, p = 0.04) and post (5.5 ± 2.8, p = 0.001) as compared to pre (2.6 ± 2.0), regardless of group.

Table 3.

Nutrition data is shown as (median ± IQR).

	Group	Value	P	Effect Size
Average eating window (h)	TRE	7.9 ± 0.1	<0.001***	0.83r
	FED	13.2 ± 0.6
Tracking Adherence (%)	TRE	94.1 ± 4.7	0.11	0.79d
	FED	97.0 ± 2.3
Calories (kcal/day)	TRE	3,090 ± 643	0.94	0.04d
	FED	3,115 ± 629
Calorie Gap (kcal/day)	TRE	−90 ± 359	0.60	0.14r
	FED	−49 ± 97
Carbohydrate (g/d)	TRE	374 ± 78	0.64	0.24d
	FED	396 ± 99
Carbohydrate Gap (g/d)	TRE	−64 ± 63	0.07	0.45r
	FED	−20 ± 39
Fat (g/d)	TRE	106 ± 48	0.48	0.19r
	FED	97 ± 7
Fat Gap (g/d)	TRE	7 ± 24	0.81	0.07r
	FED	1 ± 16
Protein (g/d)	TRE	172 ± 35	0.68	0.21d
	FED	165 ± 32
Protein Gap (g/d)	TRE	6 ± 18	0.26	0.56d
	FED	−3 ± 14

*0.01 < p <0.05; ** 0.001 < p <0.01; ***p <0.001. Effect sizes are shown as either Wilcoxon ^ror Cohen’s ^ddependent upon normality.

There was no difference between groups for energy consumed during the intervention (TRE 3,090 ± 643 kcal/d, FED 3,115 ± 629 kcal/d; p = 0.94, Cohen’s d 0.04 [negligible]). Similarly, there was no difference in the kilocalorie gap (i.e. the difference between prescribed and reported energy intake) between groups (TRE −90 ± 359 kcal/d [median ± IQR]; FED −49 ± 97 kcal/d; p = 0.60, Wilcoxon r 0.14 [small]). Carbohydrate intake did not differ between groups (TRE 374 ± 78 g/d, FED 396 ± 99 g/d; p = 0.64, Cohen’s d 0.24 [small]). However, a trend and moderate effect size did exist in the carbohydrate gap (i.e. the difference between prescribed and reported carbohydrate intake) between groups (TRE −64 ± 63 g/d [median ± IQR], FED −20 ± 39 g/d; p = 0.07, Wilcoxon r 0.45 [moderate]). Fat intake did not differ between groups (TRE 106 ± 48 g/d [median ± IQR], FED 97 ± 7 g/d; p = 0.48, Wilcoxon r 0.19 [small]). Nor was a difference observed for the fat gap (i.e. the difference between prescribed and reported fat intake) between groups (TRE 7 ± 24 g/d [median ± IQR], FED 1 ± 16 g/d; p = 0.81, Wilcoxon r 0.07 [small]). Protein intake did not differ between groups (TRE 172 ± 35 g/d, FED 165 ± 32 g/d; p = 0.68, Cohen’s d 0.21 [small]). There was also no difference in the protein gap (i.e. the difference between prescribed and reported protein intake) between groups (TRE 6 ± 18 g/d, FED −3 ± 14 g/d; p = 0.26, Cohen’s d 0.56 [moderate]).

3.2. Resistance exercise training

The number of days on the training program did not differ between groups (TRE 59 ± 2 days [median ± IQR], FED 59 ± 4 days; p = 0.59). Total RPE throughout the intervention did not differ between groups (TRE 53.6 ± 4.2, FED 52.7 ± 4.0; p = 0.66). Similarly, there were no significant effects of group (p = 0.60), sex (p = 0.20), or the interaction of group and time (p = 0.56) for weekly RPE. However, a main effect of time was present (p <0.001), with mixed model coefficients indicating an increase in RPE over time (0.22 units/wk [95% CI: 0.10 to 0.34]; p = 0.001).

Total volume across the 8 wk training program was lower in TRE (6,960 ± 287 repetitions) than FED (7,334 ± 289 repetitions; p = 0.02, Cohen’s d 1.3 [large]). Regarding weekly total volume, a significant interaction of group and time was observed (p = 0.02), along with significant effects of group (p = 0.02) and time (p <0.001) but not sex (p = 0.33). Mixed model coefficients indicated a decrease of 16.2 (95% CI: 7.4 to 25.0; p < 0.001) repetitions per week across groups, along with an additional decrease of 13.4 (95% CI: 2.0 to 24.9; p = 0.02) repetitions per week in TRE (Figure 1(a,b)).

Figure 1.

Resistance exercise volume during the intervention. Coefficients from linear mixed-effects models are displayed in panels a, c, and e, with positive coefficients displayed in blue and negative coefficients displayed in red. Individual responses are displayed in panels b, d, and f, with males displayed in blue and females displayed in red. *0.01 < p <0.05; ** 0.001 < p <0.01; ***p <0.001.

For total lower body volume across the training program, a significant difference between conditions was observed (TRE 2,896 ± 133 repetitions, FED 3,028 ± 98 repetitions; p = 0.03, Cohen’s d 1.13 [large]). Regarding weekly lower body volume, a significant interaction of group and time was observed (p = 0.02), along with significant effects of group (p = 0.04) and time (p <0.001) but not sex (p = 0.68). Mixed model coefficients indicated a decrease of 8.3 (95% CI: 4.6 to 12.0; p < 0.001) repetitions per week across groups, along with an additional decrease of 5.9 (95% CI: 1.0 to 10.7; p = 0.02) repetitions per week in TRE (Figure 1(c,d)).

For total upper body volume across the training program, a significant difference between conditions was observed (TRE 4,064 ± 171 repetitions, FED 4,306 ± 210 repetitions; p = 0.03, Cohen’s d 1.27 [large]). Regarding weekly upper body volume, a significant interaction of group and time was observed (p = 0.04), along with significant effects of group (p = 0.02) and time (p <0.001) but not sex (p = 0.13). Mixed model coefficients indicated a decrease of 8.4 (95% CI: 3.2 to 13.5; p = 0.002) repetitions per week across groups, along with an additional decrease of 6.9 (95% CI: 0.2 to 13.6; p = 0.04) repetitions per week in TRE (Figure 1(e,f)).

The total average load summed across weeks did not differ between groups (TRE 438 ± 112 kg, FED 441 ± 138 kg; p = 0.97). No interaction of group and time (p = 0.37), nor effect of group (p = 0.91), was observed for the weekly total average load. However, the effects of time (p <0.001) and sex (p <0.001) were identified. Mixed model coefficients indicated that the average load increased by 3.2 kg per week (95% CI: 2.6 to 3.7; p <0.001) across groups. Additionally, the load was 27.6 kg (95% CI: 22.1 to 33.0 kg; p <0.001) higher in males as compared to females (Figure 2(a,b)).

Figure 2.

Resistance exercise average load during the intervention. Coefficients from linear mixed-effects models are displayed in panels a, c, and e, with positive coefficients displayed in blue and negative coefficients displayed in red. Individual responses are displayed in panels b, d, and f, with males displayed in blue and females displayed in red. *0.01 < p <0.05; ** 0.001 < p <0.01; ***p <0.001.

Average lower body load summed across weeks did not differ between groups (TRE 659 ± 149 kg, FED 661 ± 172 kg; p = 0.98). No interaction of group and time (p = 0.39), nor effect of group (p = 0.99), was observed for weekly average lower body load. However, the effects of time (p <0.001) and sex (p <0.001) were identified. Mixed model coefficients indicated that the average lower body load increased by 5.4 kg per week (95% CI: 4.5 to 6.4; p < 0.001) across groups. Additionally, the load was 34.8 kg (95% CI: 25.7 to 43.8 kg; p <0.001) higher in males as compared to females (Figure 2(c,d)).

Average upper body load summed across weeks did not differ between groups (TRE 365 ± 157 kg, FED 329 ± 121 kg; p = 0.60). A trend for an interaction of group and time was observed (p = 0.07) for weekly average upper body load. No effect of group was observed (p = 0.48). However, the effects of time (p <0.001) and sex (p <0.001) were noted. Mixed model coefficients indicated the average upper body load increased by 2.0 kg per week (95% CI: 1.3 to 2.7; p <0.001) across groups. Additionally, the load was 28.1 kg (95% CI: 16.8 to 39.4 kg; p <0.001) higher in males than females. A trend (p = 0.07) was present for a greater increase in load for TRE compared to FED (0.8 kg per week, 95% CI: −0.1 to 1.7 kg; Figure 2(e,f)).

Total tonnage (i.e. volume × load) across the 8 wk training did not significantly differ between groups (TRE 400,479 ± 106,816 kg, FED 426,906 ± 137,108 kg; p = 0.68, Cohen’s d 0.22 [small]). No interaction of group and time (p = 0.22) nor effect of group (p = 0.16) was found for weekly total tonnage. However, significant effects of time (p <0.001) and sex (p <0.001) were noted. Mixed model coefficients indicated that, across groups, weekly tonnage increased by 2,274 kg (95% CI: 1,286 to 3,262; p <0.001). Additionally, tonnage values for males were 25,721 kg (95% CI: 19592 to 31,850; p <0.001) greater than for females (Figure 3(a,b)).

Figure 3.

Resistance exercise tonnage during the intervention. Coefficients from linear mixed-effects models are displayed in panels a, c, and e, with positive coefficients displayed in blue and negative coefficients displayed in red. Individual responses are displayed in panels b, d, and f, with males displayed in blue and females displayed in red. *0.01 < p <0.05; ** 0.001 < p <0.01; ***p <0.001.

No significant difference between groups was observed for total lower body tonnage across the training program (TRE 230,415 ± 56,386 kg, FED 243,763 ± 64,947 kg; p = 0.67). No interaction of group and time (p = 0.24), nor effect of group (p = 0.26), was observed for weekly lower body tonnage. However, significant effects of time (p <0.001) and sex (p <0.001) were observed. Mixed model coefficients indicated that, across groups, weekly lower body tonnage increased by 1,481 kg (95% CI: 872 to 2,090; p <0.001). Additionally, tonnage for males was 12,239 kg (95% CI: 8,546 to 15,931; p <0.001) greater than for females (Figure 3(c,d)).

No significant difference between groups was observed for total upper body tonnage across the training program (TRE 170,064 ± 52,136 kg, FED 183,143 ± 72,875 kg; p = 0.69). No interaction of group and time (p = 0.23), nor effect of group (p = 0.17), was observed for weekly upper body tonnage. However, significant effects of time (p <0.001) and sex (p <0.001) were observed. Mixed model coefficients indicated that, across groups, weekly upper body tonnage increased by 712 kg (95% CI: 383 to 1,040; p <0.001). Additionally, tonnage for males was 13,811 kg (95% CI: 11136 to 16,486; p <0.001) greater than for females (Figure 3(e,f)).

3.3. Resistance exercise performance

In the linear mixed-effects model for squat 1RM, there was a significant interaction of the TRE group and the post-time point (p = 0.04; Figure 4(a,b)), along with significant effects of the post-time point (p <0.001) and male sex (p <0.001), but not the TRE group (p = 0.75). Model coefficients indicated both groups improved squat 1RM; however, the TRE group improved less at the post-time point (−4.2 kg, 95% CI: −8.2 to −0.2 kg). Similarly, analysis of covariance accounting for baseline values indicated a trend for a difference in final squat 1RM at the end of the study (TRE – FED: −4.0 ± 1.9 kg [mean ± SE], p = 0.058).

Figure 4.

Resistance exercise performance changes. Coefficients from linear mixed-effects models are displayed in panels a, c, e, g, and i, with positive coefficients displayed in blue and negative coefficients displayed in red. Individual responses are displayed in panels b, d, f, h, and j, with males displayed in blue and females displayed in red. *0.01 < p <0.05; ** 0.001 < p <0.01; ***p <0.001.

For leg extension 1RM, significant effects of the post-time point (p = 0.003) and the male sex (p <0.001) were observed (Figure 4(g,h)), without effects of the TRE group (p = 0.66) or the interaction of the TRE group and the post-time point (p = 0.26). Additionally, analysis of covariance accounting for baseline values indicated no difference in final leg extension 1RM between groups (TRE – FED: −3.3 ± 2.6 kg [mean ± SE], p = 0.22).

In the model for leg press ME, a significant effect of the post-time point was observed (p = 0.04; Figure 4(i,j)), without effects of male sex (p = 0.20), the TRE group (p = 0.64) or the interaction of the TRE group and the post-time point (p = 0.84). Similarly, analysis of covariance accounting for baseline values indicated no difference between final leg press ME between groups (TRE – FED: −150.4 ± 590.4 kg [mean ± SE], p = 0.80).

In the model for bench press 1RM, significant effects of the post-time point (p < 0.001) and male sex (p < 0.001) were observed (Figure 4(c,d)), without effects of the TRE group (p = 0.27) or the interaction of the TRE group and post-time point (p = 0.41). Additionally, analysis of covariance accounting for baseline values indicated no difference in final bench press 1RM between groups (TRE – FED: 1.2 ± 1.8 kg [mean ± SE], p = 0.51).

For bench press ME, a significant effect of male sex (p <0.001) was observed (Figure 4(e,f)), without effects of the TRE group (p = 0.99), the post-time point (p = 0.25), or the interaction of the TRE group and post-time point (p = 0.76). Similarly, analysis of covariance accounting for baseline values indicated no difference in final bench press ME between groups (TRE – FED: 23.5 ± 72.5 kg [mean ± SE], p = 0.75).

3.4. Body composition

In the linear mixed-effects model for BM, significant effects of the mid- and post-time points (p <0.001) and male sex (p <0.001) were observed relative to baseline, without significant effects of the TRE group (p = 0.73) or the interaction of the TRE group and the mid (p = 0.14) or post (p = 0.64) time points (Figure 5(a,b)). Additionally, analysis of covariance accounting for baseline values indicated no difference between groups in final BM values (TRE – FED: −0.5 ± 1.0 kg [mean ± SE], p = 0.64).

Figure 5.

Body composition changes. Coefficients from linear mixed-effects models are displayed in panels a, c, e, and g, with positive coefficients displayed in blue and negative coefficients displayed in red. Individual responses are displayed in panels b, d, f, and h, with males displayed in blue and females displayed in red. *0.01 < p <0.05; ** 0.001 < p <0.01; ***p <0.001.

In the model for FFM, significant effects of the mid (p = 0.01) and post (p = 0.01) time points and male sex (p <0.001) were observed (Figure 5(c,d)), with increases of 1.5 kg (95% CI: 0.4 to 2.6) and 1.8 kg (95% CI: 0.4 to 3.2) at the mid- and post-time points relative to baseline, respectively. No effects of the TRE group (p = 0.50) or the interaction of the TRE group and the mid (p = 0.96) or post (p = 0.34) time points were observed. Similarly, analysis of covariance accounting for baseline values indicated no difference in final FFM values between groups (TRE – FED: 0.7 ± 0.9 kg [mean ± SE], p = 0.46).

For FM, there was a significant interaction of the TRE group and the post-time point (p = 0.02; Figure 5(e,f)), along with significant effects of the post-time point (p <0.001), but not the mid-time point (p = 0.32), the TRE group (p = 0.80), male sex (p = 0.87), or the interaction of the TRE group and the mid-time point (p = 0.09). Model coefficients indicated significantly lower FM in the TRE group at the post-time point (−1.4 kg, 95% CI: −2.5 to −0.3 kg). Similarly, analysis of covariance accounting for baseline values indicated a difference in FM at the end of the study (TRE – FED: −1.4 ± 0.6 kg [mean ± SE], p = 0.047).

For BF%, there was a significant effect of the interaction of the TRE group and post-time point (p = 0.03; Figure 5(g,h)), along with significant effects of the post-time point (p = 0.04) and the male sex (p = 0.03). No significant effects were observed for the TRE group (p = 0.94), mid-time point (p = 0.73), or interaction of the TRE group and mid-time point (p = 0.08). Model coefficients indicated significantly lower BF% in the TRE group at the post-time point (−1.6%, 95% CI: −3.0 to −0.21 %). Analysis of covariance accounting for baseline values indicated a difference in BF% at the end of the study (TRE – FED: −1.6 ± 0.7% [mean ± SE], p = 0.04).

3.5. Mood, energy, and sleep

No significant effects were observed in the linear mixed-effects model for mood, including effects of the TRE group (p = 0.70), mid-time point (p = 0.94), post-time point (p = 0.23), male sex (p = 0.17), and interaction of the TRE group and the mid (p = 0.29) and post (p = 0.52) time points (Figure 6(a)). Similarly, analysis of covariance accounting for baseline values indicated no difference in mood at the end of the study (TRE – FED: −0.3 ± 0.4 [mean ± SE], p = 0.52).

Figure 6.

Mood, energy, and sleep changes. Coefficients from linear mixed-effects models are displayed, with positive coefficients displayed in blue and negative coefficients displayed in red. *0.01 < p <0.05; ** 0.001 < p <0.01; ***p <0.001.

For ratings of energy, a significant interaction of the TRE group and mid-time point was observed (p = 0.04; Figure 6(b)), without significant effects of the TRE group (p = 0.71), mid-time point (p = 0.36), post-time point (p = 0.16), male sex (p = 0.38), and interaction of TRE group and the post-time point (p = 0.06). The model coefficient indicated lower energy in the TRE group at the mid-time point (−1.4; 95% CI: −2.8 to −0.1). Additionally, analysis of covariance accounting for baseline values indicated a trend for less energy at the end of the study for TRE (TRE – FED: −0.9 ± 0.5 [mean ± SE], p = 0.09).

No significant effects were observed in the linear mixed-effects model for sleep duration, including effects of the TRE group (p = 0.48), mid-time point (p = 0.50), post-time point (p = 0.77), male sex (p = 0.31), and interaction of the TRE group and the mid (p = 0.52) and post (p = 0.88) time points (Figure 6(c)). Analysis of covariance accounting for baseline values indicated no difference in sleep duration at the end of the study (TRE – FED: −0.1 ± 0.4 [mean ± SE], p = 0.79).

For the number of sleep disturbances per night, a significant effect of the post-time point was observed (p = 0.02; Figure 6(d)), without significant effects of the TRE group (p = 0.22), mid-time point (p = 0.59), male sex (p = 0.74), or interaction of the TRE group and mid (p = 0.59) or post (p=0.13) time points. The model coefficient indicated higher sleep disturbances at the post-time point (0.7; 95% CI: 0.1 to 1.2). Analysis of covariance accounting for baseline values indicated no difference in sleep disturbances at the end of the study (TRE – FED: −0.7 ± 0.5 [mean ± SE], p = 0.13).

No significant effects were observed in the linear mixed-effects model for sleep latency, including effects of the TRE group (p = 0.37), mid-time point (p = 0.90), post-time point (p = 0.59), male sex (p = 0.65), and interaction of the TRE group and the mid (p = 0.78) and post (p = 0.56) time points (Figure 6(e)). Similarly, analysis of covariance accounting for baseline values indicated no difference in sleep latency at the end of the study (TRE – FED: 1.8 ± 1.1 [mean ± SE], p = 0.14).

For the number of naps per week, no significant effects were observed, including effects of the TRE group (p = 0.73), mid-time point (p = 0.06), post-time point (p = 0.44), male sex (p = 0.16), and interaction of the TRE group and the mid (p = 0.58) and post (p = 0.59) time points (Figure 6(f)). Analysis of covariance accounting for baseline values indicated no difference in naps at the end of the study (TRE – FED: 0.4 ± 0.4 [mean ± SE], p = 0.29).

4. Discussion

This study examined the effects of 8 weeks of 16:8 TRE while in a caloric surplus and engaging in resistance exercise designed to increase muscle size and strength. Both dietary approaches were executable and globally effective. Adherence to the nutrition (eating windows, calories, and protein) and exercise protocols were >94% in both groups and no differences were found between subjective markers such as how hard the diet was to follow, sleep quality and duration, and daily mood. The nutrition and exercise programs were effective as both groups significantly increased strength and FFM. However, TRE negatively impacted subjective daily energy, lower body strength development, and final total exercise volume, yet resulted in significantly less FM gains. These collective data provide novel and practically valuable insights into the role of nutrient timing and 16:8 TRE on physical performance and body composition.

Our data are generally consistent with previous research and indicate that TRE does not inherently augment nor eliminate the ability to improve strength [2,5,10,28]. Our participants got stronger in every metric, regardless of group, in an amount similar to other comparable research [40,41]. However, TRE blunted maximal strength gains in the lower body. Back squat 1RM increased by ~4.2 kg less in TRE (10.4 kg) than in FED (14.6 kg). LE 1RM followed a similar theme, both groups improved 1RM by 10.2 kg and the expected increase was ~4.3 kg greater in FED than TRE, though it did not reach statistical significance. BP 1RM was not impacted by TRE as it increased by 6.8 kg but displayed no statistical or practically relevant difference between the groups. A recent study found a similar phenomenon and reported that TRE increased strength in the upper, but not the lower body, relative to habitual eating [42]. Helms et al. [43] also reported that the amount of caloric surplus influenced the rate of muscle size and strength gains differently in the upper versus lower body in well-trained individuals after 8 wk of progressive RE.

The compromised strength development in TRE was likely caused by reduced daily energy and total exercise volume. Both groups experienced a progressive training load increase and volume decreased as intended in the program design. However, the volume reduction was more dramatic in TRE (−16.2 vs. −29.6 repetitions/wk), which led to significantly less total volume by the end of the study than FED (6,960 vs 7,334 repetitions). Volume accumulation is critical for stimulating adaptation and thus one of the most important factors in proper long-term program design and periodization [44]. The short study duration (8 wk) may have been insufficient to capture the effect of the volume bifurcation between groups on performance in the other strength tests. Daily subjective energy followed a similar rate of decline in TRE through the 8 wk, indicating individuals were more physiologically or psychologically fatigable or under-recovered and not able or wanting to perform the same amount of work as FED, signifying insufficient energy availability during acute exercise, the recovery period, or both.

Some studies indicate performance is reduced when engaging in acute RE while fasting, but most research finds no detriment. Drummond and colleagues [45] reported a single 12 or 16 h fast did not reduce the maximum number of repetitions in the back squat and leg press, maximal voluntary isometric contraction, nor CMJ performance in trained men. Regarding chronic fasting, a review on Ramadan fasting found RE in the fasted state reduced power but not aerobic performance, strength, jump height fatigue index, and total work [46]. The 16:8-specific research in trained [2,3,5,10,42,47] and untrained [5,28] participants generally shows no attenuation of strength gains. Tinsley and colleagues have repeatedly observed no difference in strength between TRE and traditional eating strategies in both men and women [2,3,5,10,28]. A recent review of 24 studies also concluded that no differences exist in physical abilities such as aerobic capacity, anaerobic capacity, strength, and power when engaging in IF and TRE [48].

However, the critical difference between the current data and these previous studies is that we are the first to analyze TRE during a hypercaloric state and when performance and hypertrophy improvements are the primary goal, not fat loss. Accordingly, our training program was more frequent and substantially higher in volume and intensity than most prior investigations. This distinction matters as Terada et al. [49] also reported a reduction in exercise volume, total work, and peak power output during 4 weeks of high-intensity sprint interval training in participants who consumed carbohydrates prior to exercise versus those who performed the same exercise after a 10 hr overnight fast. Thus, while caloric intake immediately prior to exercise may not always alter performance, it may do so over time (>4 wk) with highly rigorous training demands. This finding supports the general consensus that while protein and carbohydrate timing does not impact muscle growth [50], it does influence recovery and resulting day performance, particularly when engaging in maximum effort full body exercise on consecutive days [51].

Burke, Kiens, and Ivy covered the role of carbohydrate timing for exercise recovery in their comprehensive landmark review in 2004 [52]. Work since then has confirmed the critical role of carbohydrate intake and timing for performance during intense physical activity [53]. Current recommendations to maximize recovery for prolonged moderate to high-intensity exercise consist of consuming carbohydrates as quickly and frequently (i.e. 15–30 min) as possible post-exercise. Failure to do so may compromise muscle glycogen resynthesis rates, the central drive to exercise, muscle damage, metabolism during subsequent exercise bouts, or a combination of these mechanisms [54]. Last year, Henselmans et al. [51] were the first to systematically review the impact of carbohydrate intake on RE and found that higher acute ingestion only increased performance in 6 of 19 qualifying studies. Interestingly, the 6 studies which found a benefit were compared to fasting control groups and utilized higher exercise volumes (>10 sets per muscle group).

Another consideration for the blunted adaptations is mistakes (intentional or unintentional) in the total number of calories or carbohydrates consumed by TRE. No statistically significance differences existed between our groups in caloric intake. Tracking adherence did tend to be lower in TRE (94.1%) than FED (97.0%), with a moderate effect size (0.79). Furthermore, the inaccuracy of self-reported food intakes is worth considering [55], especially as TRE is known to facilitate spontaneous caloric restriction [10,12] with resulting maintenance of FFM and strength and a concurrent reduction in FM [2,3,5,10,28]. Differences in macronutrient distribution are also plausible as TRE consumed an average of ~44 g less carbohydrates and 9 g more fat per day than FED, which was not statistically significant but displayed a moderate effect size. The weekly caloric increase strategy also differed as TRE favored a more pronounced increase in fat and less carbohydrates than FED. A week 1 vs. week 8 comparison revealed TRE increased fat by 72 g and carbohydrates by 157 g, while FED increased fat by 28 g and carbohydrates by 176 g.

Despite lower volume, strength, and perceived energy, TRE gained an equivalent amount of FFM (1.82 kg), but significantly less FM than FED (1.67 vs. 0.26 kg, respectively). The gain in FFM was larger than most, but not all, comparable TRE research, as we expected given the study design, goal, and population. We are the only study to report an increase in FM [2,3,5,28], but that was also anticipated, given our hypercaloric design. Non-TRE RE training studies of a similar length and population pool with caloric surplus targets from ~380–1,250 kcal/day and protein intakes between 2.4 and 4.4 g/kg/bw all report similar changes in FFM (0.5–2.0 kg) and FM (−1.5–1.5 kg) [56–60]. The pronounced difference in fat accumulation between groups most likely highlights the existence of accidental differences in nutritional intake. Although it seems peculiar that such small differences in macronutrient distribution could result in such meaningful changes in body composition, it has been reported in the literature [12,61]. One recent study outlined a statistically significant 3% increase in FM and 2% drop in FFM in response to a 50 kcal difference in caloric intake (~300 kcal reduction vs. 250 kcal reduction), and the caloric reduction came almost entirely from carbohydrate (65 g reduction vs. 31 g reduction) [12].

Hormonal status was not examined in our trial, and thus remains a plausible contributor to our findings, with or without errors in nutritional tracking. In the fasted state, plasma elevations of cortisol, catecholamines, and growth hormone (GH) increase with concomitant decreases in insulin and IGF-1. These changes promote adipose lipolysis and lean mass preservation, whereas increases in insulin and IGF-1 in the fed state promote adipose deposition and muscle protein synthesis, known as the feast–famine cycle [62,63]. While a GH increase does not impart an anabolic stimulus to skeletal muscle, it does facilitate increased lipolysis, liberating free fatty acids into the bloodstream to be used as an energy source, thereby sparring carbohydrate and protein oxidation in fasted states as a protective mechanism against starvation and net protein balance [64–66]. Lack of GH in fasting states increases protein loss and urea production (a marker of protein breakdown) by ~50% with a similar amount of muscle protein breakdown [64]. However, the GH reductions and increased lipolysis seem to be tightly linked to the abolishment of insulin secretion during prolonged fasting (2–3 days) and not overnight fasting in healthy, non-obese participants [67,68]. In the muscle, adiponectin interacts with AMPK, a key energy-sensing kinase that stimulates PGC-1α and in turn increases mitochondrial biogenesis, improving lipid oxidation. In the brain, adiponectin increases energy expenditure and causes fat loss [2,69]. Moro et al. [2] also reported TRE with resistance exercise training did significantly reduce glucose and insulin concentrations with concurrent IGF-1 and testosterone reductions. These data reduce the likelihood of insulin-receptor-mediated increases in the IGF-1 anabolic pathway leading to increases in FFM. Thus, while FM loss and utilization are starkly different, TRE and exercise-induced increases in GH and reductions in insulin due to periods of fasting and/or energy restriction could feasibly influence the rate of FM gain or loss. In the present study and potentially prior reports, the repeated intermittent periods of anabolic stimuli (resistance exercise, protein, carbohydrate) paired with periods of energy restriction (16:8 TRE, spontaneous reduction in energy intake) may have led to preferential manipulation of the feast–famine cycle leading to FFM accumulation with limited FM gain.

Another contributing factor could have been our method of achieving caloric surplus. Currently, there is no established ideal rate of FFM gain as skeletal muscle accumulation is a complex, multifactorial adaptive process. However, our approach (increasing kcal each week to remain in surplus; TRE = ~125 kcal/wk, FED = ~110 kcal/wk) resulted in a rate of body mass gain consistent with what is recommended to enable FFM gain while minimizing FM increase with a modest kcal surplus of 350–500 kcal/day [70]. A 2017 review of 25 overfeeding studies (21/25 utilized sedentary populations) indicated nearly all used either a specific caloric amount or a fixed percentage based upon habitual energy intake, which ranged from 8% to 70% (~400–1,200 kcal) [71]. Our energy intake based on weekly body mass testing was unique and may have avoided an excessive energy surplus and metabolic adaptation. Additional chrononutritional factors such as time of day, time between first and last meal, and nutrient-specific timing may also be relevant contributors, but more data are needed to understand their impact.

The limitations of our study include the free-living nature of the study design, use of a three-compartment body composition model, use of estimate equations for the determination of metabolic rate, and sample size. The latter was mostly driven by the COVID-19 pandemic. The final cohort was lost within a few weeks of post-testing as the laboratory was shut down. Restarting data collection after our facilities were re-opened 2 years later was not viable given it would require dependent variable testing, physical training, and nutrition coaching to occur with all new laboratory personnel, which would introduce potential errors and biases given the somewhat subjective nature of maximal performance testing. We do not consider enrollment a major limitation given that our final sample size was comparable to previous TRE research [2,3,12]. Direct assessment of muscle size via ultrasound was also eliminated after the availability of key laboratory personnel were lost too far into data collection, which would have added valuable additional information.

5. Conclusion

A 16:8 TRE approach is viable for increasing muscle size, strength, and endurance in well-trained young men and women when engaging in progressive resistance exercise and eating in a caloric surplus with adequate protein. However, the differences in total training volume, lower body 1RM, fat mass accumulation, and energy are notable and practically relevant. The reduced training volume is particularly noteworthy given its importance to strength and volume over longer interventions (i.e. >8 wk), but additional research is needed to confirm or deny the legitimacy of this potential concern. Both nutritional strategies (TRE and FED) appear to confer unique advantages and disadvantages. In practical terms, intermittent fasting diets may lend themselves better to fat loss focused diets rather than overfeeding diets to maximize muscular development, but these findings should be considered within the broader context of an individual’s goals, lifestyle, preferences, and exercise demands.

Funding Statement

Funding was received from the International Scientific Research Foundation for Fitness and Nutrition.

Disclosure statement

GMT is an inventor of the international patent “Compositions and methods of use of beta-hydroxy beta-methylbutyrate (HMB) associated with intermittent fasting.” Whey protein supplement was donated by Legion Athletics, Inc.

Recommended reviewers

Jose Antonio, Bill Campbell, Brad Schoenfeld, Steven Machek, and Mike Roberts.

Abbreviations

TRE

Time Restricted Eating, 16:8, trained fasted

FED

Control, 3–4 meals spread across the day, trained fed

BM

Body Mass

FM

Fat Mass

FFM

Fat-Free Mass

BF%

Body Fat Percentage

BIS

Bioelectrical Impedance Spectroscopy

1RM

1 Repetition Maximum

ME

Muscular Endurance

RE

Resistance Exercise

BMR

Basal Metabolic Rate

PA

Physical Activity

TEF

Thermic Effect of Food

MEDHQ

Medical, Exercise, Nutrition History Questionnaire

MSNMQ

Mood, Sleep, Nutrition, and Menstrual Cycle Questionnaire

RE

Resistance Exercise

APRE

Autoregulatory Progressive Resistance Exercise

Supplementary Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15502783.2025.2492184