The consequences of exercise-induced weight loss on food reinforcement. A randomized controlled trial.

BACKGROUND: Obesity remains a primary threat to the health of most Americans, with over 66% considered overweight or obese with a body mass index (BMI) of 25 kg/m2 or greater. A common treatment option many believe to be effective, and therefore turn to, is exercise. However, the amount of weight loss from exercise training is often disappointingly less than expected with greater amounts of exercise not always promoting greater weight loss. Increases in energy intake have been prescribed as the primary reason for this lack of weight loss success with exercise. Research has mostly focused on alterations in hormonal mediators of appetite (e.g.: ghrelin, peptide YY, GLP-1, pancreatic polypeptide, and leptin) that may increase hunger and/or reduce satiety to promote greater energy intake with exercise training. A less understood mechanism that may be working to increase energy intake with exercise is reward-driven feeding, a strong predictor of energy intake and weight status but rarely analyzed in the context of exercise.
DESIGN: Sedentary men and women (BMI: 25-35 kg/m2, N = 52) were randomized into parallel aerobic exercise training groups partaking in either two or six exercise sessions/week, or sedentary control for 12 weeks.
METHODS: The reinforcing value of food was measured by an operant responding progressive ratio schedule task (the behavioral choice task) to determine how much work participants were willing to perform for access to a healthy food option relative to a less healthy food option before and after the exercise intervention. Body composition and resting energy expenditure were assessed via DXA and indirect calorimetry, respectively, at baseline and post testing.
RESULTS: Changes in fat-free mass predicted the change in total amount of operant responding for food (healthy and unhealthy). There were no correlations between changes in the reinforcing value of one type of food (healthy vs unhealthy) to changes in body composition.
CONCLUSION: In support of previous work, reductions in fat-free mass resulting from an aerobic exercise intervention aimed at weight loss plays an important role in energy balance regulation by increasing operant responding for food.

RCT Entities:   Population Interventions Outcomes

Introduction

The prevalence of obesity and its comorbidities, including cardiovascular disease, hypertension, diabetes, dyslipidemia, metabolic syndrome, and certain cancers, are still plaguing the nation today [1-4]. Exercise is a long-standing remedy for nearly all of obesity’s comorbidities and often recommended as an economical and health-promoting option for weight loss and weight loss maintenance [5]. Unfortunately, as supported by the ever-escalating prevalence of obesity, exercise is often only marginally successfully at reducing body weight [6]. The lack of weight-loss success with exercise has been blamed on a coordinated set of compensatory mechanisms the human body uses to maintain energy balance, thereby resisting the sustained negative energy balance required for weight loss [7]. These compensatory responses may be physiological, and include reductions in resting energy expenditure and metabolic efficiency in attempts to return the body back to energy balance [7]. However, the most prevalent mechanism responsible for maintaining energy homeostasis during an exercise program is increases in energy intake, largely due to the fact that the rate of energy intake far exceeds the rate of energy expenditure [7-9]. For instance, it takes most people between 40 and 60 minutes to expend 500 kcal through exercise, which is about the energy content of a fast-food cheeseburger that can be consumed in 5–10 minutes.

There has been a great deal of research on exercise’s effects on energy intake, specifically by assessing how hormonal mediators of appetite, lab-based food intake, and hunger/satiety scales change with exercise [10-14]. However, many have demonstrated single bouts of exercise do not result in changes in appetite, food intake, or appetite-regulating hormones [15-18]. Some have actually demonstrated greater perceptions of hunger and fullness persist 24 hours after exercise [19], and long-term exercise improves the satiety response to a meal [13, 20]. These results are at odds with the apparent compensatory response individuals display, therefore necessitating novel and innovative research to better understand the causes behind exercise-induced increases in energy intake. Empirical hypotheses postulate that individuals who exercise for weight loss have distorted portion control, seek rewards for exercising in the form of food, and derive greater pleasure from high-fat, high-sugar, energy-dense food [7, 21, 22], all of which may be independent of hunger. These behaviors may be explained by a neuro-behavioral cross-talk that is in play with exercise and eating, as both are products of the central dopamine system [23, 24]. There is evidence of such direct cross-talk between eating and exercise reward in rats, whereas a small dose of wheel running that produces a low-level dopaminergic response facilitates eating [25]. A similar neuro-behavioral cross-talk between the reinforcing effects of different drugs is also known, specifically involving endogenous opioid and cannabinoid systems, whereas administration of one increases the reinforcing value of the other [26]. The rewarding aspects of food and eating are important volitional behavioral responses, assessed by defining one’s relative reinforcing value of food (RRVfood) or food reinforcement [27, 28]. Food reinforcement is a measure of how much an individual wants to engage in a particular eating behavior, as someone who finds food highly reinforcing “wants” food to a greater degree. RRVfood is the reinforcing value of food relative to an alternative, which may be a non-eating activity or a different type of food (high vs. low energy density for example). Importantly, food reinforcement is a more robust predictor of food intake than the hedonic value (liking) of the food [28] and is a strong predictor of body weight and energy intake [27, 29]. It therefore seems plausible that the reinforcing value of food, is increased by exercise to promote food intake, although very little research has been done on this issue. Understanding these volitional behavioral responses could inform future weight loss and exercise recommendations by adding important considerations when exercising for weight control. The present study hypothesized that a 12-week aerobic exercise for weight loss intervention would increase food reinforcement, specifically for the high energy-dense foods provided (assessed by an operant responding task) among sedentary, overweight to obese humans. It was further hypothesized that a greater dose of exercise would evoke a greater effect on food reinforcement and that changes in body composition would be correlated with changes in food reinforcement.

Materials and methods

Participants

A total of 80 participants aged 18 to 40 years volunteered and were enrolled into the study. Of these, 52 completed all baseline tests and were randomized into one of three groups (six exercise sessions per week, two sessions per week, and sedentary control) during this longitudinal, randomized, controlled trial on a 1:1:1 ratio. Of these 52 randomized participants, 44 completed the study (32 female), with six (four female) withdrawing for personal reasons and two females being excluded for non-compliance (did not complete the required 85% of exercise sessions assigned per month). A consort diagram is depicted in Fig 1. All participants had a body mass index (BMI) ranging from 25–35 kg/m2 and were inactive (not engaging in any form of exercise), determined during screening where participants were asked of their exercise behaviors. Participants were also free of medical issues that would be a contraindication to exercise, without metabolic or cardiovascular disease, not engaging in a weight loss diet, were weight stable (not lost or gained 5% of their current body weight in the previous 6 months), not taking medications that may influence appetite, and not pregnant or nursing. Recruitment began in the winter of 2018 and continued until recruitment goals were met (spring of 2019) in and around Lexington, Kentucky. Participants were a sample who responded to recruitment media including printed brochures and flyers and online advertisements placed on University of Kentucky’s Center for Clinical and Translational Science (CCTS) website. This study was approved by the University of Kentucky Institutional Review Board. The present analysis is a secondary outcome of a trial aimed at assessing mechanisms of energy compensation at different doses of exercise, ClinicalTrials.gov identifier: NCT03413826.

Fig 1

Consort diagram.

Figure depicting recruitment, retention, and randomization of present trial.

Procedures

During the initial screening and consenting visit, participants provided their written informed consent and were screened of eligibility criteria, completing a physical activity readiness questionnaire (PARQ), health history questionnaire, and screened on their dieting, weight loss history, and physical activity behaviors. Participants also completed a taste-test of the study foods and were scheduled for other assessments including rate of energy expenditure (resting and during exercise), body composition, and food reinforcement (all detailed below).

Study design

The study was a randomized, controlled trial that included a 12-week exercise intervention of either six sessions (days) per week, two sessions per week, or a sedentary control group (no exercise). Men and women were randomized separately, with separate allocation sequences generated and maintained by the study statistician. Participants were randomized upon completion of all baseline assessments with no blinding of intervention assignments. 13 of the 19 randomized participants to the six-session group, 17 of the 20 randomized to the two-session group, and 11 of the 14 randomized to the control group were female. Participants were assessed for outcome measures at baseline and after the intervention. Food reinforcement and body composition were assessed 24 to 48 hours after the participant completed their final exercise session of the 12-week intervention while resting energy expenditure was assessed 48 to 72 hours after completing the final exercise session.

Exercise intervention

Participants were provided a Polar A-300 heart rate monitor (watch and chest strap, Kempele, Finland) for the duration of the 12-week intervention and instructed to exercise either two or six times per week on their own and were provided access to a fitness center. Participants were instructed to engage in only aerobic exercise either indoors (treadmill, elliptical or cycle ergometer) or outdoors (walking, running, biking) as long as they attained a heart rate at least in heart rate reserve (HRR) zone 1. Participants in the control group were instructed to remain sedentary and return for post-testing 12 weeks later, receiving the exercise intervention after post-testing if they desired. Those in the exercise groups returned to the lab weekly to meet a researcher and download their exercise sessions using the PolarFlow software, which allowed research staff to monitor and track compliance. If a participant was not 85% compliant (completed 85% of expected exercise sessions per month) they were dropped from the study. The downloaded exercise session reports provided the amount of time spent in each heart rate zone, which allowed for the calculation of total energy expended during each exercise session based off individual rates of energy expenditure averaged across each heart rate zone calculated from the graded exercise test with indirect calorimetry performed at baseline and week six. Participants in the two-day per week group were instructed to perform two long exercise sessions per week and encouraged to try to expend 1,000 kcal per session. Participants in the six-day per week group were instructed to keep their sessions to 400 kcal per session. Although most participants in the two-day per week group were not able to attain the 1,000 kcal goal, they still expended significantly greater kcal per session compared to the six-day group. Participants received personalized heart-rate based exercise prescriptions that, if followed, would result in them expending the assigned energy per exercise session. Participants were also provided feedback each week on their energy expenditure of each session of the prior week so they could tailor future exercise sessions. All participants were instructed not to purposely change dietary habits during the intervention, i.e., not begin an energy-restricted diet.

Assessments

Liking of foods and hunger (secondary outcome measure)

Participants’ liking (hedonic value) of the food options used in the food reinforcement test were assessed using a 10-point scale (1 = “do not like at all” and 10 = “like very much”). Participants sampled three “healthy” snack foods and three “unhealthy” snack foods (Table 1). The most liked food of each category was used during the food reinforcement testing session, completed on a separate day. All participants rated at least a moderate liking (liking score of ≥5) for at least one of the healthy and unhealthy food options. This test was completed once, prior to the baseline food reinforcement test. Hunger and satiety scales were presented prior to each food reinforcement test and used as an additional variable in regression models when predicting changes in food reinforcement. A copy of the VAS scales are provided in the supplementary data.

Table 1

Foods used in the progressive ratio schedules of reinforcement task to measure food reinforcement.

		Kcal/gram	%CHO1	%Fat2	%Pro3	%Sugar4
“Unhealthy” food options	Snickers	4.9	50.2	43.6	6.2	41.2
	Nacho Cheese Doritos	5.40	48.0	48.0	5.2	2.7
	Oreo cookies	4.7	60.2	38.0	4.7	35.0
“Healthy” food options	Nutri-grain bar	3.2	73.3	20.6	6.1	80.0
	Dried Banana	5.2	42.8	55.7	0.02	6.8
	Mixed nuts	6.1	12.5	75.2	12.3	0.03

1%CHO: percent of total kcal derived from carbohydrates

2%Fat: percent of total kcal derived from dietary fat

3%Protein: percent of total kcal derived from protein

4%Sugar: percent of total kcal derived from added sugars.

Food reinforcement (primary outcome measure)

Reinforcing value of food was assessed by evaluating the amount of operant responding (mouse button presses) a participant performed to gain access to each alternative (healthy and unhealthy snack foods, Table 1) [29-31]. The testing environment included two workstations with computers in the same room. One computer had a game that was set-up for participants to earn points towards their highest liked healthy food while the other computer had the same game that participants could use to earn points toward the highest liked unhealthy food. Participants could switch between stations as much as they chose. The computer programs presented a game that mimics a slot machine; a point is earned each time the shapes match. For every five points, a session is completed and the participant received an approximately 60-kcal portion of the reinforcer that was earned (either healthy or unhealthy food). The game was performed until the participant no longer wished to work for access to either food. At first, points were delivered after every four presses, but then the schedule of reinforcement doubled (4, 8, 16, 32, […] 1024) each time five points were earned. For instance, the participant initially had to click the mouse four times to earn each point for schedule one. After the first five points were earned, schedule one was complete and the participant earned a portion of food for the reinforcer earned (healthy or unhealthy food). Then eight clicks were required to earn each of the next five points for schedule two before another portion of the reinforcer was earned. Schedule three required 16 clicks to earn one point, schedule four required 32 clicks to earn one point, and so on [29, 30]. Participants received the food earned after completing the game, which ended when the participant no longer wished to earn points for eating either type of food. This progressive ratio schedule task (Behavioral Choice Task) software was developed at the University of Buffalo, reviewed in [27] and used previously by the current investigative team [31, 37]. Participants were not allowed to take food out of the lab with them, but were not required to consume all of the food they earned in the event a participant earned more food than they could eat in one sitting, although this occurred on only two occasions throughout the study. The food reinforcement tests were conducted two to four hours post-prandial between usual lunch and dinner times when snack foods are likely to be consumed. Similar button pressing tasks have been shown to be valid predictors the reinforcing value of eating behaviors [29, 32–35]. The breakpoint, or Pmax, [30] was the total number of schedules completed for the healthy or unhealthy food. The Relative Reinforcing Value of healthy foods (RRVhealthy) was the proportion of Pmax for healthy food compared to the unhealthy food calculated as (Pmax healthy)/ (Pmax healthy + Pmax unhealthy). As constructed, an RRVhealthy score over 0.5 indicates a greater reinforcing value for the healthy snack food option relative to unhealthy option, while a RRVunhealthy score of over 0.5 represents a greater reinforcing value of unhealthy food relative to healthy [29, 30]. As a measure of total food reinforcement, Pmax healthy and Pmax unhealthy were added together, reflecting the total number of schedules completed for food, or PmaxTotal. Thus, PmaxTotal reflects an overall score of total food reinforcement, irrespective of type (healthy or unhealthy). Changes in Pmax healthy and Pmax unhealthy were also assessed independently.

Rate of energy expenditure (secondary outcome measure)

A graded exercise treadmill test was used to determine each participant’s rate of energy expenditure at five different heart-rate zones. Oxygen consumed and CO2 produced were analyzed by indirect calorimetry (VMAX Encore Metabolic Cart, Vyaire Medical, Mettawa, IL) which included an integrated 12 lead ECG for monitoring heart rate and used in conjunction with the Trackmaster TMX428 Metabolic cart interfaced treadmill. Upon completion of a five-minute warm-up walking at 0% grade, 3.0 mph, the treadmill grade increased to 2.5% for three minutes. The treadmill grade was then increased every three minutes to produce an approximately 10 beat per minute increase in heart rate from the previous stage with the speed fixed at 3.0 mph. The test continued until a heart rate of 85% HRR was attained or the participant felt they could no longer continue. Energy expenditure (kcal per minute) was determined from the amount of oxygen consumed and CO2 expired using the Weir equation [35]. The average rate of energy expenditure during the last 30 seconds of each stage of the graded exercise test was regressed against the heart rate averaged over the last 30 seconds of the corresponding stage to calculate the rate of energy expenditure at different heart rates. Heart rate zones were calculated using the HRR formula as (220-age)-resting HR * zone % + resting HR [36]. Heart rate Zone 1 ranged from 0% to 25% HRR, Zone 2 corresponded to 26–40% HRR, Zone 3 was 41–58% HRR, Zone 4 was 59–75% HRR, and Zone 5 was 76–90%. Energy expenditure in kcal/min was then averaged across each heart rate zone for determination of energy expenditure per minute for each zone. This test was completed at baseline and week six to recalculate rates of energy expenditure to account for improvements in fitness.

Resting energy expenditure (secondary outcome measure)

Resting Energy Expenditure (REE) was measured using indirect calorimetry (Quark RMR; Cosmed USA, Chicago, IL) with a ventilated canopy. Before each test, calibrations were performed on the flow meter using a 3.0-L syringe and on the gas analyzers using verified gases of known concentrations. After 30 minutes of quiet rest in the supine position in a dimly lit, temperature-controlled room between 22 and 24 C, REE was measured for 30 minutes. The test was monitored to ensure participants remained awake and between 0.8 and 1.2% feCO2. Criteria for a valid REE was a minimum of 15 minutes of steady state, determined as a <10% fluctuation in oxygen consumption and <5% fluctuation in respiratory quotient. The Weir equation (36) was used to determine REE from the measured oxygen consumption and CO2 production. Participants completed the baseline REE assessment prior to the exercise test and 36–72 hours after their final exercise session of the intervention.

Body composition (primary outcome measure)

Body composition was measured using a GE Lunar iDXA machine prior to the exercise test. The iDXA technique allows the non-invasive assessment of soft tissue composition by region with a precision of 1–3% [37]. A total body scan was conducted with participants lying supine on the table and arms positioned to the side. Most scans were completed using the thick mode suggested by the software as participants were overweight to obese. All scans were analyzed using GE Lunar enCORE Software (13.60.033). Automatic edge detection was used for scan analyses. The machine was calibrated before each scanning session, using the GE Lunar calibration phantom. Outcome measures included kg of body fat mass (FM), kg of fat-free mass (FFM)- which included body water, bone mineral content, and dry lean mass, and percent FM.

Analytic plan

Differences in changes in FM, FFM, REE, RRVhealthy, RRVunhealthy, Pmax healthy, Pmax unhealthy and total food reinforcement (PmaxTotal) between groups were determined via one-way ANOVA. Change scores were also tested if significantly different from zero via T-tests for each group, separately. Shapiro-Wilk tests and inspection of histograms revealed a non-normal distribution for changes in PmaxTotal, thus differences in PmaxTotal between groups were also assessed via Kruskal Wallis Test. Spearmen correlation analysis was performed between changes in PmaxTotal, hunger, FM, FFM, REE, and exercise variables including exercise frequency (group) and exercise energy expenditure per week. Significant correlations were further assessed with quantile regression, using change in PmaxTotal as the dependent variable. All analyses were performed in IBM SPSS Version 26 (IBM corporation, Armonk, New York).

Power analysis

The present analysis is a secondary outcome of a study originally designed to test effects of exercise dose on energy compensation, powered to detect significant differences in body fat loss between groups to draw conclusions regarding energy compensation in a clinically relevant scenario, that is, a scenario in which individuals who are overweight to obese lose a significant amount of body weight. We previously demonstrated significant differences (1.7 kg) in body fat loss in groups exercising to expend 3,000 kcal per week vs. 1,500 kcal per week for 12 weeks [37]. Using an 80% power and 95% confidence level, 13 participants per group were needed to detect a significant change in body fat loss from baseline to post intervention with a standard deviation of 2.3.

Results

Baseline measures of food measures, BMI, body composition, REE, and demographics (% female) are presented in Table 2. Change scores (post value minus baseline value) for outcome variables and exercise energy expenditure are presented in Table 3. The six-day per week group expended more energy than the two-day per week group, which contributed to only the six-day group significantly decreasing percent body weight and FM. Change in FM was greater in both intervention groups compared to control with no differences in FFM change. There were no changes in PmaxTotal, Pmax healthy, Pmax unhealthy, REE, or RRV of healthy or unhealthy food over time or between groups (Table 3). All variables were of normal distribution with the exception of changes in PmaxTotal (Shapiro-Wilk P<0.05, Skewness: 1.24, SE = 0.37; Kurtosis: 9.26, SE = 0.72; histogram included in supplementary data). Thus, a Kruskal-Wallis Test was used to confirm no difference between groups for changes in PmaxTotal (Table 4). Spearmen correlation analysis indicated that PmaxTotal negatively correlated with changes in FFM (P = 0.03), with no correlations between changes in FM, REE, hunger or exercise parameters (weekly energy expenditure, exercise group). Positive correlations (both when assessed via Pearson or Spearmen) were observed between FM change, exercise energy expenditure per week, and exercise group (all P<0.1). Pearson correlations demonstrated a positive correlation between changes in REE and changes in hunger (P = 0.02). Quantile regression analysis predicting changes in PmaxTotal is presented in Table 5, demonstrating changes in FFM to be the only significant predictor for changes in PmaxTotal when controlling for weekly exercise energy expenditure, exercise frequency per week (group), and changes in FM, REE and hunger. A reduced model is also presented, removing independent variables that were correlated with each other to account for collinearity, thus leaving changes in FFM, FM and REE as the independent variables with only changes in FFM remaining a significant predictor. Mediation analysis did not reveal significant mediation effects for changes in REE on changes in FFM when predicting changes in PmaxTotal (Sobel test statistic: <0.01, SE = <0.01, P = 0.99).

Table 2

Demographics, body composition, and food reinforcement measures for all randomized participants at baseline.

	Six-day per week group N = 19	Two-day per week group N = 20	Control N = 14
Sex (% female)	68.4	85.0	78.8
BMI1	29.0 ± 2.87	30.51 ± 3.47	29.36 ± 2.87
FM2	31.25 ± 8.58	35.58 ± 6.55	30.27 ± 6.42
FFM3	48.50 ± 9.68	48.91 ± 9.35	43.77 ± 5.92
Body Fat %	37.96 ± 6.87	41.29 ± 4.41	39.68 ± 4.11
REE4	1505.8 ± 48.13	1666.9 ± 76.55	1546.40 ± 50.59
RRVHealthy5	0.54 ± 0.32	0.62 ± 0.29	0.62 ± 0.31
RRVUnhealthy6	0.46 ± 0.32	0.38 ± 0.30	0.38 ± 0.31
PmaxTotal7	65.00 ± 73.77	78.25 ± 97.59	102.00 ± 108.97

Data are mean ± SD

1Body Mass Index, kg/m2

2FM: Fat Mass, kg

3FFM: Fat Free Mass, kg

4REE: Resting energy expenditure (Kcal/24 hours)

5RRVHealthy: Relative Reinforcing value of healthy food, assessed via progressive ratio schedule of reinforcement task and calculated as last schedule completed for healthy food / total schedules completed

6RRVUnhealthy: Relative Reinforcing value of unhealthy food, assessed via progressive ratio schedule of reinforcement task and calculated as the last schedule completed for unhealthy food / total schedules completed

7PmaxTotal: Total schedules completed for food, both healthy and unhealthy, in the progressive ratio schedule of reinforcement task

Table 3

Resulting data from the 12-week exercise intervention between groups that exercised either six or two days per week and a sedentary control group.

	Six-day per week group N = 15	Two-day per week group N = 17	Control N = 11
ExEE/week1	2,753.5 ± 561.2*	1,490.7 ± 503.4*	0*
Percent weight loss2	-1.48 ± 2.48^	-0.84 ± 2.72	+1.45 ± 3.71
ΔFM3	-1.82 ± 1.51^#	-0.64 ± 0.95*	0.98 ± 2.62^*
ΔFFM4	0.38 ± 1.39	-0.04 ± 0.23	-0.06 ± 1.36
ΔREE5	39.18 ± 151.5	-38.0 ± 246.5	-1.13 ± 259.0
ΔRRV healthy food6	0.12 ± 0.35	-0.05 ± 0.41	-0.01 ± 0.20
ΔRRV junk food7	-0.12 ± 0.35	0.11 ± 1.86	0.13 ± 0.36
ΔPmaxTotal8	20.36 ± 141.7	-39.38 ± 93.31	-14.55 ± 78.87
ΔPmaxhealthy9	33.21 ± 131.07	-36.47 ± 78.78	-33.25 ± 56.83
ΔPmaxunhealthy10	-12.86 ± 54.94	-0.88 ± 23.84	21.36 ± 51.17

Data are mean ± SD

*, ^: like letters indicate significant differences between groups, P ≤ 0.05.

#Significant change over time (change different from zero) P ≤ 0.05.

1ExEE/week: Exercise energy expenditure (in kilocalories) per week.

2Percent weight loss: kg of weight change (12-week value minus baseline value) / baseline body weight in kg

3ΔFM: kg of fat mass change: (12-week kg of body fat mass minus baseline kg body fat)

4ΔFFM: kg of fat-free mass change (12-week kg of fat-free mass minus baseline kg of fat free mass)

5ΔREE: Changes in REE (12-week value minus baseline value) in kcal/24 hrs.

6ΔRRVhealthy: changes in relative reinforcing value of healthy food assessed via progressive ratio schedule of reinforcement task, calculated as 12-week RRVhealthy−baseline RRVhealthy

7ΔRRVunhealthy: changes in relative reinforcing value of unhealthy food assessed via progressive ratio schedule of reinforcement task, calculated as 12-week RRVunhealthy−baseline RRVunhealthy

8ΔPmaxTotal: changes in total schedules completed for food (healthy + unhealthy) during the progressive ratio schedule of reinforcement task.

9ΔPmaxhealthy: changes in total schedules completed for healthy food option during the progressive ratio schedule of reinforcement task.

10ΔPmaxunhealthy: changes in total schedules completed for unhealthy food option during the progressive ratio schedule of reinforcement task.

Table 4

Kruskal Wallis Test for changes total food reinforcement (ΔPmaxTotal, non-normally distributed) between groups that exercised either six or two days per week and a sedentary control group.

Data are mean rank.

	Six-day per week group N = 15	Two-day per week group N = 17	Control N = 11	χ2(2)
ΔPmaxTotal1	24.46	18.75	1.86	1.872

P = 0.392

1ΔPmaxTotal: changes in total schedules completed for food (healthy + unhealthy) assessed during the progressive ratio schedule of reinforcement task

Table 5

Quantile regression models predicting changes in total food reinforcement (PmaxTotal) among participants who exercised for 12 weeks at a frequency of either two or six days per week.

Effect	β	SE	P
Full model of all predictors
Intercept	-161.7	151.3	0.32
ExEE/week1	-0.05	0.06	0.43
Exercise Frequency2	68.8	75.4	0.39
ΔFM3	-26.6	21.4	0.25
ΔFFM4	-72.6	22.7	0.01
ΔREE5	-0.04	0.18	0.84
ΔHunger6	1.69	1.83	0.38
Reduced model
Intercept	-36.2	21.7	0.11
ΔFM3	-7.01	11.2	0.54
ΔFFM4	-46.2	18.5	0.02
ΔREE5	0.06	0.11	0.54

1ExEE/week: Exercise energy expenditure (in kilocalories) per week

2Exercise Frequency: Participants were randomly assigned to exercise 6 days per week or 2 days per week.

3ΔFM: kg of fat mass change: (12-week kg of body fat mass minus baseline kg body fat)

4ΔFFM: kg of fat-free mass change (12-week kg of fat-free mass minus baseline kg of fat free mass)

5ΔREE: Changes in REE (12-week value minus baseline value) in kcal/24 hrs.

6ΔHunger: Hunger assessed prior to each progressive ratio schedules of reinforcement task (1–10 scale).

Discussion

Most people have an enormous capacity to increase their energy expenditure to promote a negative energy balance. Depending on an individual's aerobic fitness, exercise intensities can be maintained for prolonged periods at two- to 16-fold above resting rates of energy expenditure. As such, 250 to 2500 kcal can be expended during a single exercise session resulting in an acute energy deficit that can be repeated across days, prompting many individuals to turn to exercise for obesity treatment [38, 39]. This has led the American College of Sports Medicine to issue separate recommendations to either maintain health [40] or support weight loss through exercise (5). For these reasons, exercise has become the most common weight loss approach amongst those attempting to lose weight, with a 65% prevalence rate [41]. As the obesity epidemic continues to escalate in America and other developed countries, the number of individuals seeking to lose weight is also expected to increase, thereby increasing the number of individuals using exercise as their weight loss method of choice. This is despite the lack of consistency in weight loss outcomes from exercise interventions and lack of long-term weight loss maintenance [42]. For this reason, additional research is needed to determine factors influencing weight loss from an exercise program and methods to improve exercise’s utility as a weight loss treatment.

The current study is an initial investigation on how exercise may influence eating behaviors through a mechanism separate from hunger (reward-driven feeding), an alternative approach that has received much less attention, although still justifiable, as research on exercise’s effects on hunger are inconsistent with many concluding little to no effects [15-18]. Reward-driven feeding may therefore have a greater influence on the increased drive to eat while in an energy-deficit induced by exercise. However, very little research has focused on the effects exercise may have on food reinforcement and how this may alter eating behaviors, which is a plausible hypothesis in that food and exercise reinforcement are both driven by the meso-accumbal dopamine system. The central dopamine system is essential for experiencing reward and assimilating information about energy balance (i.e. the dopamine hypothesis of reward), originally proposed to explain drug addiction [23, 24, 43]. Indeed, genetic polymorphisms that control dopamine uptake and transport have been linked to both exercise and eating reinforcement [32, 44–46], possibly explaining the cross-talk observed between eating and exercise reward in rats [25]. Research on this topic in humans, however, is sparse, with one study demonstrating no changes in food reward after an acute bout of exercise [47] and another demonstrating decreases in the reinforcing value of high-energy dense foods with increases in the reinforcing value of low-energy dense foods after two weeks of aerobic exercise [48]. These findings may indicate longer duration exercise interventions are needed to increase food reinforcement and thus influence eating behavior. This would be in line with previous research demonstrating changes in exercise reinforcement occur only after large-dose exercise interventions [49, 50]. It is also plausible that a large energy deficit needs to be created (larger than what can be created in one exercise bout or a two-week intervention) to increase reward-driven feeding. In this light, a yet to be determined metabolic signal that feeds back on the central dopamine system to instill greater food reinforcement when energy stores are low may exist. The notion that the reinforcing value of low energy dense foods increased after short-term exercise owe to the possibility that individuals may be more motivated to eat only certain (healthier) foods when partaking in exercise, which is in line with research pointing to one health behavior change having a spillover effect on other behaviors [48, 51]. For this reason, the present study analyzed the reinforcing value of two different types of food independently and constructed a measure of total food reinforcement (total number of schedules completed for all food). In this scenario, one may decrease their reinforcing value of unhealthy foods and increase their reinforcing value of healthy foods to result in a gross increase in food reinforcement if the increase in healthy foods was greater than the decrease in unhealthy. This appeared to have occurred in the present study among the six-day per week group, although the increases in this group did not reach statistical significance due to large variability (minimum and maximum values for changes in PmaxTotal ranged from -340 to 440). Being a secondary aim of the original trial (assessing mechanisms of energy compensation using different doses of exercise) this analysis may not have been adequately powered to detect changes in total responses for food and thus would be the logical next step for future interventions. This is also evident when focusing on the changes in Pmax healthy and Pmax unhealthy, where the six-day per week group completed over 12 schedules less for unhealthy foods and 33 schedules more for healthy food post-intervention, although none of these changes were deemed significant.

Despite the lack of significant changes in PmaxTotal after the exercise intervention, regression analysis revealed that changes in FFM were a significant independent predictor for changes in PmaxTotal, whereas greater losses in FFM produced more responding for food. This held when controlling for exercise energy expenditure (kcal per week), exercise frequency (sessions per week), and changes in FM, REE and hunger. In this full model, several variables were correlated with each other, presenting the likelihood of collinearity causing an inflation of regression coefficients. Specifically, changes in FM, weekly exercise energy expenditure, and exercise group were all correlated. Thus, we removed both weekly exercise energy expenditure and exercise group while keeping changes in FM in the reduced model as FM change was likely the result of exercising at greater frequencies and expending more energy per week. Interestingly, changes in REE were positively correlated with changes in hunger, whereas those who had greater increases in hunger had greater increases in REE. Such correlations have been demonstrated before, with the notion that FFM determines REE to influence hunger and food intake [52]. Although others have concluded REE, and not FFM, influences hunger and food intake [53, 54], the present analysis indicate that FFM is the primary driver of food reinforcement, thus we included changes in REE in the final model due to the known link between FFM, REE, and food intake regulation [55]. Therefore, the reduced model included changes in FM, FFM, and REE as predictors of changes in total food reinforcement (PmaxTotal), where changes in FFM remained the only significant predictor, indicating changes in FFM after a 12-week exercise for weight loss intervention influences food reinforcement to the greatest degree. This is supported by recent findings, indicating FFM is associated with several brain regions involved in the homeostatic control of food intake, suggesting a centrally mediated mechanism whereby FFM influences eating behaviors [56]. This has also been observed in the classic Minnesota Starvation Experiments, where the FFM deficit, independently of FM deficit, predicted the degree of hyperphagia that occurred during post-starvation ad libitum refeeding, which continued until FFM was repleted, often well after FM was restored [57, 58].

The present findings, demonstrating FFM deficits were the root cause in the increase in operant responding for food after 12-weeks of exercise, extend previous research by suggesting a direct mechanism that may be prompting increases in energy intake when FFM is depleted. This is an important consideration for present weight-loss recommendations, as many individuals are utilizing aerobic exercise to induce the energy deficit required for weight loss. This aerobic exercise-induced weight loss, especially when coupled with energy restriction, almost always produces concurrent losses in FFM [55, 59]. Although decreases in FFM, if accompanied by a larger decrease in FM, is often overlooked as many markers of cardiometabolic health are often improved. However, as indicated by the present study and supported by others, declines in FFM may have the unintended consequence of increasing food reinforcement, causing overeating to promote the return to baseline FFM. Unfortunately, when returning to baseline FFM levels, “fat overshooting” often occurs, where FM is increased beyond baseline levels [55, 60]. Therefore, one can argue that exercise-for weight loss interventions should include resistance exercise to promote, or at least maintain, FFM when attempting weight loss [61]. Maintaining FFM, in this light, would theoretically attenuate the increase in food reinforcement and improve long-term weight-loss success with exercise. Future interventions assessing the effects of resistance exercise and food reinforcement following weight loss may add to current exercise recommendations to increase the effectiveness of exercise as a weight loss strategy. It may also be interesting to investigate how the reinforcing value of food changes during an exercise intervention to determine how may exercise sessions, how large of an energy deficit, or how much FFM needs to be lost to render food more rewarding. It may also be interesting to investigate how long it may take for food reinforcement to return to baseline levels, or if it remains elevated long after ceasing exercise and returning to energy balance.

This study is not without limitations. Liking of the test foods was not assessed as baseline liking scores did not influence food reinforcement, but if may have been interesting to see how or if liking of these tests foods changed as a result of the exercise intervention. Ab libitum energy intake would be best assessed in an inpatient feeding design where all of participants’ meals and snacks are consumed in a controlled environment and recorded by research staff to prevent the known under-reporting that often occurs with self-reported dietary intake [62]. Without this, the present study does not have the exact amount of energy consumed, how dietary intake changed during the exercise program, and how these changes aligned with the changes in food reinforcement. Due to the known influence restrictive diets play on food reinforcement [37], participants were inscructed not to purposely change their habitual diets (i.e. start a weight loss diet). Due to the modest decrease in weight loss, we do not believe participants purposely restricted their energy intake. Contrary to that, it is likely that participants consumed more food to compensate for the energy expended, although we do not know this for sure without rigorous dietary intake assessment methods. As previously noted, the analysis, being a secondary aim of a larger study, was likely underpowered to detect significant differences in change scores (post value minus baseline value) in food reinforcement and likely affected our mediation analysis, an issue future studies must address. The duration of the intervention (12 weeks) is similar to most exercise interventions in this area of research, although a longer period may have been required to detect significant changes in food reinforcement and changes in percent weight loss among the two-day per week group. We also did not monitor adherence of the control group, although we believe each remained sedentary as they did not significantly change body weight after 12 weeks. Although without accelerometry data, we cannot be sure those in the control group did not begin exercising on their own. Participants were also only provided their most liked food out of five choices for each the healthy and unhealthy options for the RRV assessment. If participants were provided with their favorite food choice they may have responded differently. Additionally, out of the 44 participants who completed the current study, 40 were white, thus limiting the generalizability to other race/ethnic groups. This study also was not designed to detect sex differences and included an unbalanced sample of females; thus, sex effects cannot be drawn.

Conclusions

The present analysis offers an initial look into an alternative explanation as to why exercise programs are often only marginally effective for weight loss. Our findings indicate that there is great variability in individuals’ change in food reinforcement after a 12-week aerobic exercise intervention, and those who did increase their food reinforcement were also those who lose the greatest amount of FFM post-intervention, even when controlling for other variables such as the energy expended during the intervention or the frequency of exercise (sessions per week). From these findings we can draw two primary conclusions. First, it appears a large sample size is required to better elucidate food reinforcement changes after exercise, potentially taking sex effects into account. Second, it appears preventing the loss in FFM would be a valuable piece to a weight loss program, hinting at the potential for resistance training or increasing dietary protein intake as important adjunct therapy that may be the topic of future research.

(PDF)

Click here for additional data file.

PARQ, physical activity readiness questionnaire.

Provided to participants at baseline to ensure they were healthy enough to engage in exercise without physician consent.

(PDF)

Click here for additional data file.

Healthy history questionnaire.

Provided to participants at baseline to ensure the met certain inclusion criteria regarding current or recent health issues.

(PDF)

Click here for additional data file.

VAS scales, visual analog scales.

Scale used to assess hunger and desire to eat prior to each RRV test.

(PDF)

Click here for additional data file.

Liking of food.

Questionnaire used to assess liking of study foods to determine which will be used in the RRV test.

(PDF)

Click here for additional data file.

Study protocol.

IRB-approved protocol for present trial.

(PDF)

Click here for additional data file.

Simple histogram for delta PmaxTotal.

Histogram to depict changes in PmaxTotal.

(PDF)

Click here for additional data file.

Transfer Alert

This paper was transferred from another journal. As a result, its full editorial history (including decision letters, peer reviews and author responses) may not be present.

1 Apr 2020

PONE-D-20-01440

The Consequences of Exercise-Induced Weight Loss on Food Reinforcement, A Randomized, Controlled Trial

PLOS ONE

Dear Dr. Flack,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The Reviewers have several minor comments especially pertinent to Methods and Discussion of the manuscript that need detailed responses and clarifications.

We would appreciate receiving your revised manuscript by May 16 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Maciej S. Buchowski

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

1. Please explain any discrepancy between the trial protocol and the manuscript; for example, we note that, while the manuscript states that participants aged 18- 49  were included, in the registry and in the protocol the age limit is set at 40 years old .

2. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information. Moreover, please include more details on how the questionnaire was pre-tested, and whether it was validated.

3. Please note that according to our submission guidelines (http://journals.plos.org/plosone/s/submission-guidelines), outmoded terms and potentially stigmatizing labels should be changed to more current, acceptable terminology. For example: “Caucasian” should be changed to “white” or “of [Western] European descent” (as appropriate).

4. Our internal editors have looked over your manuscript and determined that it is within the scope of our Determinants, Consequences and Management of Obesity  Call for Papers. This collection of papers is headed by a team of Guest Editors for PLOS ONE:Rachel Nugent and Pratibha V. Nerurkar. Additional information can be found on our announcement page: https://collections.plos.org/s/obesity-one.

If you would like your manuscript to be considered for this collection, please let us know in your cover letter and we will ensure that your paper is treated as if you were responding to this call. If you would prefer to remove your manuscript from collection consideration, please specify this in the cover letter.

5.  Thank you for including the following funding information within your acknowledgements section; "The project described was supported by the NIH National Center for Advancing 410 Translational Sciences through grant number UL1TR001998"

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

"The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

6.  We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

7. Please upload a copy of Figure 1, to which you refer in your text on page 13. If the figure is no longer to be included as part of the submission please remove all reference to it within the text.

8. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Overall

This paper builds upon both theoretical considerations, the impact of physical activity upon dietary reinforcement, while also offering practical advice for those interested in losing weight via physical activity, maintain lean mass via resistance training. The overall paper is well written and adds to the literature. The additions I would suggest are as follows:

Introduction:

Line 40: add that exercise also essential to maintaining healthy weight after loss. Considerable research evidence supports this.

Line 62: “seek rewards FOR exercising …”

Methods:

Line 84: Can you say “increase food reinforcement” if measuring 2 food types?

Line 146: Did any participants report an atypical diet (keto, Atkins etc.) diet or that they were currently dieting to lose weight at baseline?

Line: 191 spell-check of “unhealthy”

Line 186-191: Pmax can be calculated for healthy and unhealthy separately. Specify that Pmax_total is for both added together. Also, why did you only examine Pmax total and not the separate Pmaxs for each food group?

Mention that participants’ favorite food (healthy or unhealthy) may not have been offered and they would respond differently to different options

Did you examine proportion of food earned during the RRV task to food consumed afterwards? Those results may be interesting.

Results

Line 237-238: Because changes in body composition are so important for study results, please include body fat percentage, and kg of fat mass and lean mass in Tables 2.

Line 264: Unhealthy food, to keep terminology consistent

Discussion:

RRV task and DXA took place very shortly after end of a very intense training regime. How might results change if participants were given opportunity to taper and/or rest before post-intervention results?

Any indication whether or not participants change their diets, either intentionally or unintentionally?

Reviewer #2: This is a well-crafted study that examines a novel and important topic. I enjoyed reading it. I have listed some suggestions, by section, below. While I believe you can handle these issues relatively easily, I do think they will strengthen the manuscript and should be addressed. If these issues are addressed, I believe the manuscript warrants publication. Well done!

Introduction

“The most prevalent mechanism responsible for maintaining energy homeostasis during an exercise program is increases in energy intake, largely due to the fact that the rate of energy intake far exceeds the rate of energy expenditure” while I agree with this point, it may be worthwhile to discuss metabolic changes that can occur with weight loss. At least mention adaptive thermogenesis as a possible mechanism complicating weight loss from exercise. It’s possible that alterations in food reinforcement may be part of the adaptive thermogenesis response.

Methods

“… 52 completed all baseline tests and were randomized into one of three groups…” I think it’s a good idea to list the three groups parenthetically here. Revie to “… 52 completed all baseline tests and were randomized into one of three groups (six exercise sessions per week, two weekly sessions, sedentary control)…”

“The study was a randomized, controlled trial that included a 12-week exercise intervention of either six sessions (days) per week, two sessions per week, or a sedentary control group (no exercise) blocked on gender” list the number of males and females in each group.

How was compliance of the sedentary group assessed?

What were some of the exercise options available top your exercise groups? Did they only do cardio (e.g., treadmill, bike) or were they able to do resistance training?

Why was liking only assessed at baseline? Changes in hedonics as a result of the intervention would have been interesting to look at as well as RRV.

Interesting uses of Pmax for the RRV task. I like how you quantified RRV but, out of curiosity, why not just use raw Pmax scores for healthy versus unhealthy foods as your measure of RRV?

“Change scores were also tested if significantly different from zero via T-tests.” Were these done for each group separately?

“Differences in changes in…” revise to “Differences in change scores (post value minus baseline value) in…”

Discussion

“The present findings, demonstrating FFM deficits are the root cause in the increase…” this language is too definitive for my liking. Revise to “The present findings, demonstrating FFM deficits were the root cause in the increase…” or something similar.

“Our findings indicate that there is great variability in individuals’ change in food reinforcement after a 12-week aerobic exercise intervention, and those who do increase their food reinforcement are those who lose the greatest amount of FFM post-intervention…” Again this reads as too definitive. I think it’s important to describe the results only within the context of the present study. Stating “those who do increase their food reinforcement are those who lose the greatest amount of FFM post-intervention” sounds like you are concluding this is what happens with weight loss for everyone. While that may be true, you cannot conclude that from this single study. I know this seems like a small detail, but it is important to remain tentative. Revise to “Our findings indicate that there was great variability in individuals’ change in food reinforcement after a 12-week aerobic exercise intervention, and those who did increase their food reinforcement were also those who lost the greatest amount of FFM post-intervention…” This way you are only talking about your results.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Jacob E Barkley

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

14 Apr 2020

Response to Reviewers

We would like to thank the reviewers and Dr. Buchowski for their time and effort into providing their important feedback. We have carefully addressed each comment and feel that this manuscript has improved significantly. Specific comments are addressed below:

1. Please explain any discrepancy between the trial protocol and the manuscript; for example, we note that, while the manuscript states that participants aged 18- 49 were included, in the registry and in the protocol the age limit is set at 40 years old.

This was incorrectly stated in the manuscript; we have now corrected this on line 91.

2. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information. Moreover, please include more details on how the questionnaire was pre-tested, and whether it was validated.

There were no survey/questionnaires used to gather data on primary or secondary outcomes. Participants completed questionnaires at baseline to demonstrate they were healthy enough to participate in the exercise intervention. Participants also rated their liking of the study foods at baseline and rated their hunger and desire to eat on visual analog scales prior to each RRV test. A copy of each is included as supporting information as requested.

3. Please note that according to our submission guidelines (http://journals.plos.org/plosone/s/submission-guidelines), outmoded terms and potentially stigmatizing labels should be changed to more current, acceptable terminology. For example: “Caucasian” should be changed to “white” or “of [Western] European descent” (as appropriate).

We changed “Caucasian” to “white” on line 487 and removed other potentially stigmatizing labels in this section.

4. Our internal editors have looked over your manuscript and determined that it is within the scope of our Determinants, Consequences and Management of Obesity Call for Papers. This collection of papers is headed by a team of Guest Editors for PLOS ONE:Rachel Nugent and Pratibha V. Nerurkar. Additional information can be found on our announcement page: https://collections.plos.org/s/obesity-one.

If you would like your manuscript to be considered for this collection, please let us know in your cover letter and we will ensure that your paper is treated as if you were responding to this call. If you would prefer to remove your manuscript from collection consideration, please specify this in the cover letter.

We have specified in the cover letter that we would like to include this manuscript in the Determinants, Consequences and Management of Obesity call.

5. Thank you for including the following funding information within your acknowledgements section; "The project described was supported by the NIH National Center for Advancing 410 Translational Sciences through grant number UL1TR001998"

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

"The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

We have removed the funding-related text from the manuscript. We would like our funding statement to read: “The project described was supported by the NIH National Center for Advancing Translational Sciences through grant number UL1TR001998. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

6. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

There are no legal or ethical restrictions on sharing our data, thus we have deposited our data in a public repository and provided the DOI on Line 511 of the manuscript.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

7.Please upload a copy of Figure 1, to which you refer in your text on page 13. If the figure is no longer to be included as part of the submission please remove all reference to it within the text.

Our figure has been uploaded; we apologize for overlooking this in the initial submission.

8. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

We have included captions for our supporting information as requested at the end of our manuscript, lines 674-682

Reviewer #1: Overall

This paper builds upon both theoretical considerations, the impact of physical activity upon dietary reinforcement, while also offering practical advice for those interested in losing weight via physical activity, maintain lean mass via resistance training. The overall paper is well written and adds to the literature. The additions I would suggest are as follows:

Introduction:

Line 40: add that exercise also essential to maintaining healthy weight after loss. Considerable research evidence supports this.

We certainly agree with this statement and included “and weight loss maintenance” in this statement as our reference (#5) also backs this (line 42).

Line 62: “seek rewards FOR exercising …”

This has been edited as suggested, now line 63.

Methods:

Line 84: Can you say “increase food reinforcement” if measuring 2 food types?

This is a fine point and we further clarified that we hypothesized food reinforcement for specifically high energy dense foods available would increase (line 85-86).

Line 146: Did any participants report an atypical diet (keto, Atkins etc.) diet or that they were currently dieting to lose weight at baseline?

Participants were not dieting to lose weight at baseline. This was an inclusion criterion, along with being weight stable. We realize this was not included in the participant characteristics section and now include both of these criteria (line 101-102)

Line: 191 spell-check of “unhealthy”

This has been corrected, now line 210.

Line 186-191: Pmax can be calculated for healthy and unhealthy separately. Specify that Pmax_total is for both added together. Also, why did you only examine Pmax total and not the separate Pmaxs for each food group?

The last two sentences of this section state “Pmax healthy and Pmax unhealthy were added together, reflecting the total number of schedules completed for food, or PmaxTotal. Thus, PmaxTotal reflects an overall score of total food reinforcement, irrespective of type (healthy or unhealthy)” There were no differences in changes in Pmax healthy or Pmax unhealthy, we specified this in Table 3, results (line 284), and touched on it in the discussion (lines 409-412).

Mention that participants’ favorite food (healthy or unhealthy) may not have been offered and they would respond differently to different options

We have included this point when addressing limitations of the present study, lines 484-486..

Did you examine proportion of food earned during the RRV task to food consumed afterwards? Those results may be interesting.

There were only two occasions where participants did not consume all the food they earned in the RRV task, thus these analyses were not preformed. We have stated this more specifically on line 202.

Results

Line 237-238: Because changes in body composition are so important for study results, please include body fat percentage, and kg of fat mass and lean mass in Tables 2.

We have now included baseline values for body fat %, FM and FFM in table 2.

Line 264: Unhealthy food, to keep terminology consistent

This change has been made, now line 285.

Discussion:

RRV task and DXA took place very shortly after end of a very intense training regime. How might results change if participants were given opportunity to taper and/or rest before post-intervention results?

This is an interesting question and we have expanded on this regarding when RRV testing is done in relation to the exercise intervention, lines 456-461

Any indication whether or not participants change their diets, either intentionally or unintentionally?

We have included, lines 470-475, a couple sentences about how we believe they did not intentionally engage in restrictive eating behaviors but likely increased their food intake in response to the exercise intervention.

Reviewer #2: This is a well-crafted study that examines a novel and important topic. I enjoyed reading it. I have listed some suggestions, by section, below. While I believe you can handle these issues relatively easily, I do think they will strengthen the manuscript and should be addressed. If these issues are addressed, I believe the manuscript warrants publication. Well done!

Introduction

“The most prevalent mechanism responsible for maintaining energy homeostasis during an exercise program is increases in energy intake, largely due to the fact that the rate of energy intake far exceeds the rate of energy expenditure” while I agree with this point, it may be worthwhile to discuss metabolic changes that can occur with weight loss. At least mention adaptive thermogenesis as a possible mechanism complicating weight loss from exercise. It’s possible that alterations in food reinforcement may be part of the adaptive thermogenesis response.

We certainly agree with this point and have mentioned physiologic adaptations to an exercise-induced energy deficit that also work to return the body back to energy balance, lines 46-48)

Methods

“… 52 completed all baseline tests and were randomized into one of three groups…” I think it’s a good idea to list the three groups parenthetically here. Revie to “… 52 completed all baseline tests and were randomized into one of three groups (six exercise sessions per week, two weekly sessions, sedentary control)…”

We have made this addition as suggested, line 93.

“The study was a randomized, controlled trial that included a 12-week exercise intervention of either six sessions (days) per week, two sessions per week, or a sedentary control group (no exercise) blocked on gender” list the number of males and females in each group.

We have included this as suggested on lines 126-127

How was compliance of the sedentary group assessed?

We have included in our 481-484 that we did not assess the physical activity of the control group during the intervention but believe they remained sedentary as their body weight did not change. We included this in the section on the weaknesses of the study.

What were some of the exercise options available to your exercise groups? Did they only do cardio (e.g., treadmill, bike) or were they able to do resistance training?

Participants only engaged in aerobic exercise, this is now detailed on lines 135-138

Why was liking only assessed at baseline? Changes in hedonics as a result of the intervention would have been interesting to look at as well as RRV.

We agree this would have been interesting to include and noted this in the limitations section, lines 462-464.

Interesting uses of Pmax for the RRV task. I like how you quantified RRV but, out of curiosity, why not just use raw Pmax scores for healthy versus unhealthy foods as your measure of RRV?

We combined the Pmax scores to get an overall food reinforcement score, not relative to anything. But we were also interested if they may increase their reinforcing value of one type of food relative to the other, so maybe they wouldn’t change Pmax of healthy or unhealthy independently, but when put against each other, we may see a significant change in the proportion (maybe people find candy bars more reinforcing than dried banana after exercising). We were worried that assessing the reinforcing value of food relative to a non-food activity (reading or watching TV for example) may have been skewed as people may prefer doing a sedentary activity opposed to exercise if they were burnt out from the intervention. Thus, this would have been more of a test for the reinforcing value of sedentary activities than food and confounded our results. In response to reviewer 1, we provided the results on changes in Pmax for healthy and unhealthy foods in Table 3, the results section, and provided a bit on the discussion on this (lines 409-412)

“Change scores were also tested if significantly different from zero via T-tests.” Were these done for each group separately?

These were done for each group separately, now indicated on line 263-264..

“Differences in changes in…” revise to “Differences in change scores (post value minus baseline value) in…”

This has been revised as suggested, lines 476-477

Discussion

“The present findings, demonstrating FFM deficits are the root cause in the increase…” this language is too definitive for my liking. Revise to “The present findings, demonstrating FFM deficits were the root cause in the increase…” or something similar.

We have revised as suggested, line 439.

“Our findings indicate that there is great variability in individuals’ change in food reinforcement after a 12-week aerobic exercise intervention, and those who do increase their food reinforcement are those who lose the greatest amount of FFM post-intervention…” Again this reads as too definitive. I think it’s important to describe the results only within the context of the present study. Stating “those who do increase their food reinforcement are those who lose the greatest amount of FFM post-intervention” sounds like you are concluding this is what happens with weight loss for everyone. While that may be true, you cannot conclude that from this single study. I know this seems like a small detail, but it is important to remain tentative. Revise to “Our findings indicate that there was great variability in individuals’ change in food reinforcement after a 12-week aerobic exercise intervention, and those who did increase their food reinforcement were also those who lost the greatest amount of FFM post-intervention…” This way you are only talking about your results.

We agree with this statement and have made the changes to the conclusion as suggested, line 495.

Submitted filename: response to reviewers.pdf

Click here for additional data file.

22 May 2020

PONE-D-20-01440R1

The consequences of exercise-induced weight loss on food reinforcement. A randomized controlled trial

PLOS ONE

Dear Dr. Flack,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The manuscript was reviewed by a biostatitcian, who pointed several limitation of the manuscript including lack of an a priori power assessment and several problems with data analysis.

Please submit your revised manuscript by Jul 06 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Maciej S. Buchowski

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #3: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #3: I Don't Know

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: My previous revision requests have been addressed. One typo on line 471. Other than that, it is ready for publication.

Reviewer #3: Important note: This review pertains only to ‘statistical aspects’ of the study and so ‘clinical aspects’ [like medical importance, relevance of the study, ‘clinical significance and implication(s)’ of the whole study, etc.] are to be evaluated [should be assessed] separately/independently.

My first question is when title says “A randomized controlled trial” then why this study is classified as [Article Type:] ‘Research Article’? If it is really ‘A randomized controlled trial’ then it is expected to estimate ‘required sample size’. For the study on this topic a small effect size is generally assumed. Even for medium effect size [type I error 5%, Power=80%] according to ‘table-2 on page 158 of an old but classical paper “A power primer” in Psychological Bulletin, 1992, vol.:112, pp 155-159 the required sample size is 52 per group. However, in this study total sample size is only 52. [For small effect size (type I error 5%, Power=80%) it is as large as 322]. Note that in lines 498-500 it is said that “it appears a large sample size is required to better elucidate food reinforcement changes after exercise, potentially taking sex effects into account”. But a priori estimation of the required sample size is not found.

What exactly you want to convey by ‘blocked on gender’ [a sedentary control group (no exercise) blocked on gender], the phrase used (in line 123) while describing group-III? {If that means ‘matched’ why the wide differences (68.4%, 85%, 78.8%, in Gr.I,II,III respectively)? It cannot be usual meaning of ‘block’ because group sizes are so different (19, 20, 14 in Gr.I,II,III respectively)}. Please explain.

According to line 260 “Group differences in baseline study characteristics where tested via 1-way ANOVA” but P-values are not reported in table-2 displayed in lines 505- {which is in fact very good as it is often said that

To provide a description of baseline characteristics is entirely reasonable (since it is clearly important in assessing to whom the results of the trial can be applied), however, it does not require the division of baseline characteristics by treatment groups. Statistical comparison of baseline characteristics is not desirable at all [because even if P-value turns out to be significant (while comparing baseline characteristics despite random allocation), it is, by definition, a false positive] as you then are supposed to be testing ‘randomization’ then, which in any single trial may not balance all baseline characteristics because ‘randomization’ is a sort of ‘insurance’ and not a guarantee scheme}.

If the study was to observe effect of ‘progressive ratio’, then why [as said in lines 148-9] ‘Participants in the two-day per week group were instructed to perform two long exercise sessions per week and encouraged to try to expend 1,000 kcal per session’ is not understood. Will you please thro light on purpose [of that action]. ‘Power Analysis’ account given in lines 271 onwards is not very convincing though ‘was based on our previous study’.

In Table 3 [Resulting data from the 12-week exercise intervention between groups that exercised either six or two days per week and a sedentary control group] data are mean ± SE whereas “It may please be noted that ‘data always should be presented as means ± standard deviation’ [and never as means ± standard errors]. The standard deviation (SD) and standard error (SE) of the mean measures two very different things. The standard error depends heavily upon the sample size. [Note: standard error of the mean tells not about variability in the original population, as the standard deviation does, but about the certainty with which a sample mean estimates the true population mean. Since the certainty with which we can estimate the mean increases as the sample size increases, the standard error of the mean decreases as the sample size increases.] People are interested in knowing variability in ‘study population’ and not precision in estimate.

It is definitely appreciable that [line 294-297] ‘Quantile regression analysis’ [predicting changes in PmaxTotal is presented in Table 5: controlling for weekly exercise energy expenditure, exercise frequency per week (group), and changes in FM, REE and hunger] is used, however, a brief note (at least its non-parametric nature) on this ‘not so popular’ technique was desirable (my opinion may be little bias as a biostatistician). Surprisingly, median or other quartiles are not reported {though this non-parametric technique is used}. As you may be aware, there are disadvantages as well as limitations [like ‘monotonicity constraints’ as pointed out in ‘Journal of Machine Learning Research Non-parametric Quantile Estimation (2005)’ or computations are quite tedious compared to the least squares method] which is why technique, though very old, is seldom used. I believe [not sure] that “quantile regression” (because said in lines 270 that “All analyses were performed) is available in IBM SPSS Version 26.

Limitation mentioned in lines 462-64 [Liking of the test foods was not assessed as baseline liking scores did not influence food reinforcement, but if may have been interesting to see how or if liking of these tests foods changed as a result of the exercise intervention] is questionable. In the backdrop of the clear mention in lines 475-6 that this study is secondary of a larger study, [“the analysis, being a secondary aim of a larger study, was likely underpowered to detect significant differences in change scores”], editors may think about publication of such an article {though other things presented are alright}.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: renamed_5e1d4.docx

Click here for additional data file.

25 May 2020

Response to Reviewers

We would like to thank the reviewers and Dr. Buchowski for their time and effort into providing their important feedback. We have carefully addressed each comment and feel that this manuscript has improved significantly. Since we satisfied the previous reviewers’ comments, we are responding below only to Reviewer #3’s comments, with our response in italics below each comment.

COMMENTS – Manuscript PONE-D-20-01440R1

Title: “The consequences of exercise-induced weight loss on food reinforcement. A randomized controlled trial”

Important note: This review pertains only to ‘statistical aspects’ of the study and so ‘clinical aspects’ [like medical importance, relevance of the study, ‘clinical significance and implication(s)’ of the whole study, etc.] are to be evaluated [should be assessed] separately/independently.

My first question is when title says “A randomized controlled trial” then why this study is classified as [Article Type:] ‘Research Article’? If it is really ‘A randomized controlled trial’ then it is expected to estimate ‘required sample size’. For the study on this topic a small effect size is generally assumed. Even for medium effect size [type I error 5%, Power=80%] according to ‘table-2 on page 158 of an old but classical paper “A power primer” in Psychological Bulletin, 1992, vol.:112, pp 155-159 the required sample size is 52 per group. However, in this study total sample size is only 52. [For small effect size (type I error 5%, Power=80%) it is as large as 322]. Note that in lines 498-500 it is said that “it appears a large sample size is required to better elucidate food reinforcement changes after exercise, potentially taking sex effects into account”. But a priori estimation of the required sample size is not found.

Thank you for pointing this out, I initially selected ‘Research Article’ as this is what I believed a manuscript of this nature to fall under, although this study does fit NIH’s description of a clinical trial, so this may need to be updated. I have contacted the journal for advice in changing this designation.

As for the second part of this question, I’m not certain I am understanding this fully. I have provided a power analysis, starting on line 270, and have added detail regarding how and why the study is powered this way. In short, this study is powered to detect a significant difference in body fat loss between groups. Although not specific to the outcome measures of the present investigation, powering the study in this manner allows us the ability to assess the present outcomes in groups that differ in body fat loss, an important, clinically relevant, variable to consider.

What exactly you want to convey by ‘blocked on gender’ [a sedentary control group (no exercise) blocked on gender], the phrase used (in line 123) while describing group-III? {If that means ‘matched’ why the wide differences (68.4%, 85%, 78.8%, in Gr.I,II,III respectively)? It cannot be usual meaning of ‘block’ because group sizes are so different (19, 20, 14 in Gr.I,II,III respectively)}. Please explain.

You are correct, this wasn’t “blocked” in the traditional sense. Rather, men and women were randomized separately. This has been updated and explained on lines 124/125.

According to line 260 “Group differences in baseline study characteristics where tested via 1-way ANOVA” but P-values are not reported in table-2 displayed in lines 505- {which is in fact very good as it is often said that

To provide a description of baseline characteristics is entirely reasonable (since it is clearly important in assessing to whom the results of the trial can be applied), however, it does not require the division of baseline characteristics by treatment groups. Statistical comparison of baseline characteristics is not desirable at all [because even if P-value turns out to be significant (while comparing baseline characteristics despite random allocation), it is, by definition, a false positive] as you then are supposed to be testing ‘randomization’ then, which in any single trial may not balance all baseline characteristics because ‘randomization’ is a sort of ‘insurance’ and not a guarantee scheme}.

I do know many statisticians who scoff at the notion of testing for bassline group differences in a randomized trial. Although I’ve also received comments on papers I have submitted where reviewers ask me to test for baseline group differences despite a randomized design. I’m not a statistician, I have no preference, and have therefore removed the mention of testing for group differences (line 260) and simply state that baseline data is presented in table 2 (line 282).

If the study was to observe effect of ‘progressive ratio’, then why [as said in lines 148-9] ‘Participants in the two-day per week group were instructed to perform two long exercise sessions per week and encouraged to try to expend 1,000 kcal per session’ is not understood. Will you please thro light on purpose [of that action]. ‘Power Analysis’ account given in lines 271 onwards is not very convincing though ‘was based on our previous study’.

The progressive ration (or PR schedules) refers to the food reinforcement assessment. The two and six days of exercise per week were the exercise intervention. We were testing to see if the intervention exerted an effect on the food reinforcement measure (PR schedules). To make this clearer, we have added a couple lines at the end of the I intro, lines 87-89. As noted in the prior comment, we have added detail to the power analysis.

In Table 3 [Resulting data from the 12-week exercise intervention between groups that exercised either six or two days per week and a sedentary control group] data are mean ± SE whereas “It may please be noted that ‘data always should be presented as means ± standard deviation’ [and never as means ± standard errors]. The standard deviation (SD) and standard error (SE) of the mean measures two very different things. The standard error depends heavily upon the sample size. [Note: standard error of the mean tells not about variability in the original population, as the standard deviation does, but about the certainty with which a sample mean estimates the true population mean. Since the certainty with which we can estimate the mean increases as the sample size increases, the standard error of the mean decreases as the sample size increases.] People are interested in knowing variability in ‘study population’ and not precision in estimate.

We understand this notion of this perspective and the difference in SD and SE, which is why Table 1 included SD, as we are reporting strictly on our sample in this table. I have always been advised to report SE when presenting results from a clinical trial as we are trying to present our data not just specific to our data set but to convey it to the larger population. Although again, if SD is preferred in this scenario we are fine with making these changes. Table 3 has been updated to reflect SD.

It is definitely appreciable that [line 294-297] ‘Quantile regression analysis’ [predicting changes in PmaxTotal is presented in Table 5: controlling for weekly exercise energy expenditure, exercise frequency per week (group), and changes in FM, REE and hunger] is used, however, a brief note (at least its non-parametric nature) on this ‘not so popular’ technique was desirable (my opinion may be little bias as a biostatistician). Surprisingly, median or other quartiles are not reported {though this non-parametric technique is used}. As you may be aware, there are disadvantages as well as limitations [like ‘monotonicity constraints’ as pointed out in ‘Journal of Machine Learning Research Non-parametric Quantile Estimation (2005)’ or computations are quite tedious compared to the least squares method] which is why technique, though very old, is seldom used. I believe [not sure] that “quantile regression” (because said in lines 270 that “All analyses were performed) is available in IBM SPSS Version 26.

We appreciate the insight, and to better report all quartiles, we have included a histogram for delta PmaxTotal and have included this in the supplemental files and referenced it on line 292.

Limitation mentioned in lines 462-64 [Liking of the test foods was not assessed as baseline liking scores did not influence food reinforcement, but if may have been interesting to see how or if liking of these tests foods changed as a result of the exercise intervention] is questionable. In the backdrop of the clear mention in lines 475-6 that this study is secondary of a larger study, [“the analysis, being a secondary aim of a larger study, was likely underpowered to detect significant differences in change scores”], editors may think about publication of such an article {though other things presented are alright}.

It is true that this was a secondary aim of a larger study, but the analysis was centered on the regression analysis, identifying predictors in food reinforcement. Just because there were no statistically significant changes in Pmax, the regression analysis was still able to demonstrate significant findings. With changes in PmaxTotal over +20 in the 6-d group and -39 and -14 in the 2-d and control groups, respectively, this seems like a fair speculation and worthy of publication, albeit probably not in Nature or N Engl J Med. This notion that liking may have changed during the intervention to influence food reinforcement at post-testing was brought up by another reviewer before, questioning if liking was assessed at 12-weeks and if there were changes in this measure. Since we did not assess liking at 12 weeks, we considered this a limitation.

Submitted filename: response to reviewers2.docx

Click here for additional data file.

2 Jun 2020

The consequences of exercise-induced weight loss on food reinforcement. A randomized controlled trial

PONE-D-20-01440R2

Dear Dr. Flack,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Maciej S. Buchowski

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #3: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #3: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #3: (No Response)

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #3: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: No

8 Jun 2020

PONE-D-20-01440R2

The consequences of exercise-induced weight loss on food reinforcement. A randomized controlled trial

Dear Dr. Flack:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Maciej S. Buchowski

Academic Editor

PLOS ONE