Behavioral approaches for reducing body weight in people with overweight or obesity often result in less weight or fat mass (FM) loss than anticipated from the intervention’s expected energy deficit (1), and weight regain is common (2). Several compensatory physiological and behavioral alterations may explain the observed modest weight loss and weight loss maintenance. Specifically, subjective hunger and orexigenic hormones may increase, whereas anorectic hormones, resting metabolic rate (RMR), and nonexercise physical activity (NEPA) may decrease in response to weight loss, which could ultimately oppose weight loss efforts (3–6). The success of behavioral weight loss interventions likely relates to changes in energy intake (EI; modulated by alterations in appetite), RMR, and NEPA or any combination of these factors. Although it is evident that weight loss induces several adaptations that ultimately support weight regain, it is unclear if these mechanisms differ according to behavioral weight loss modality. Some evidence suggests that acute and chronic exercise alters parameters of appetite in a manner favoring reduced EI (7); others have reported statistically significant but clinically negligible differences in EI after exercise interventions compared with no-exercise controls (8). Exercise may also preserve skeletal muscle—the largest component of fat-free mass (FFM)—and therefore prevent substantial declines in RMR. Conversely, EI restriction generally results in greater weight loss than exercise alone (9). Although lesser weight loss from exercise may be in part due to low doses of prescribed exercise energy expenditure (10), short-term energy deficits through EI restriction elicit greater compensatory responses in appetite than exercise (11). However, whether differences between diet and exercise strategies are apparent in interventions of longer durations is unknown.
Describing the unique metabolic and behavioral changes induced by different behavioral modalities may be useful in developing more effective and personalized weight loss strategies (e.g., modulating EI restriction and exercise amounts, program commencement, or goals in different populations). Therefore, the objective of this study was to characterize changes in appetite-related variables, RMR, and NEPA during an EI restriction (DIET) or aerobic exercise (AEX) weight loss intervention in adults with overweight or obesity. We hypothesized that appetite-related peptides and behaviors would change in a manner favoring lower EI in the AEX group, given the occurrence of greater short-term compensation with DIET and downstream effects of AEX on appetite parameters (7,11). Similarly, we anticipated that body composition–adjusted RMR changes would not differ between groups, whereas DIET would result in less effect on components of NEPA compared with AEX.
The primary objective of the randomized trial was to examine the effects of a 12-wk AEX intervention on brain responses to food cues as measured by functional magnetic resonance imaging between those who lose weight and those who do not; the DIET group was included to examine the overall effects of exercise itself compared with the associated weight/fat loss. Analyses presented here are secondary comparisons between weight loss modalities. The study was approved by the Colorado Multiple Institutional Review Board and registered on ClinicalTrials.gov (NCT02047721). Participants provided informed consent before undergoing any study procedures.
Participants were recruited from the University of Colorado Anschutz Medical Campus community and greater Aurora/Denver, CO, area using flyers and e-mail listservs. Inclusion criteria were body mass index (27–40 kg·m−2), age (21–55 yr), and currently participating in less than 2 h planned physical activity/week. Exclusion criteria are presented in Supplemental Content 1 (https://links.lww.com/TJACSM/A181). Only participants with subjective appetite parameters at both baseline and follow-up were included in the present analysis.
Participants were enrolled on a continuing, rolling basis from December 2014 until March 2019. After enrollment, participants were randomized to 12 wk of DIET or AEX using simple randomization, stratified by sex and using a random number generator in Excel (Microsoft; Redmond, WA). Only the study coordinator (A.H.) and K.T.L. had access to the randomization sequence; A.H. enrolled participants and assigned interventions according to the randomization sequence. Race and ethnicity were collected from self-reports. Participants completed evaluations of appetite, EI, body composition, RMR, and free-living physical activity at baseline and during the last week of the intervention. In the AEX group, the follow-up visit occurred at least 24 h after the last exercise bout.
Exercise and Diet Interventions
Details about the exercise and diet interventions as well as energetic compensation estimations are presented in Supplemental Content 1 (https://links.lww.com/TJACSM/A181). During the 12-wk program, AEX participants were instructed to perform aerobic exercise four times per week at 60%–80% of predicted heart rate maximum. The DIET intervention used several methods of in-person and self-monitoring strategies to support the goal of reducing EI by approximately 500 kcal·d−1. Participants in the DIET group were not counseled on physical activity or exercise nor encouraged to change behaviors aside from decreasing their EI; similarly, participants in the AEX group were not counseled on dietary intake nor encouraged to change behaviors aside from their exercise goals.
Anthropometrics and Body Composition
Body composition was measured at baseline (16 ± 9 d before study day visit) and follow-up using dual-energy x-ray absorptiometry (DXA; Discovery W, Hologic, Marlborough, MA; software version 13.4.1). Body weight was attained from the scan as the sum of all tissue mass. FFM was calculated as the sum of total lean tissue and bone mineral content. Adiposity was assessed as FM (body weight − FFM).
Participants were provided with a eucaloric (50% carbohydrate, 30% fat, and 20% protein) energy balanced run-in diet to consume for 3 d before each study day visit. Consumption of the food was confirmed verbally on each study day. Total daily EI requirements for both the run-in diet and study day visit meal were estimated as (23.9 × FFM from DXA) + 372 × 1.4 (12). Participants were asked to not consume any other food or beverages and to maintain usual activity during the run-in phase.
Study Day Visit
Study day visits occurred before and after the DIET or AEX intervention to assess appetite and EI. In premenopausal women, visits were scheduled in the follicular phase of the menstrual cycle (determined via self-report). After an overnight fast, participants reported to the University of Colorado Hospital Clinical and Translational Research Center (CTRC). An intravenous catheter was inserted into the antecubital vein for serial blood sampling of hormones. A fasting blood draw, visual analog scales (VAS) of appetite, and eating behavior questionnaires were completed. Participants then consumed a breakfast meal over ≤20 min that consisted of 30% of total daily EI requirements with a macronutrient composition identical to the run-in diet. EI requirements at follow-up were calculated with FFM measured at the end of the intervention. The food items provided differed according to participant EI needs and preferences. Blood sampling was performed in the fasting state and 30, 60, 90, 120, 150, and 180 min after the standard meal. After 180 min, participants consumed an ad libitum meal. Laboratory values and VAS were expressed as incremental area under the curve (AUC) using the trapezoid method (13).
Blood samples were collected from an intravenous catheter to assess total ghrelin, peptide-YY (PYY), and total glucagon-like peptide-1 (GLP-1) at each time point described above. A fasting value of leptin was also collected. All samples were collected in ethylenediamine tetraacetic acid tubes (plus added dipeptidyl peptidase-4 inhibitor for GLP-1) and stored at −800 C at the CTRC Core Laboratory until analysis. Samples were centrifuged at 3600 rpm for 10 min before analysis using radioimmunoassay for ghrelin (Millipore; 4.5% within-day precision, 93 pg·mL−1 sensitivity), PYY (Millipore; 5.3% within-day precision, 10 pg·mL−1 sensitivity), and leptin (Millipore; 5.8% within-day precision, 0.5 ng·mL−1 sensitivity) and enzyme-linked immunosorbent assay for GLP-1 (Mercodia, 6.7% within-day precision, 1.0 pmol·L−1 sensitivity).
Subjective Measures of Appetite States
Participants rated hunger, satiety, and prospective food consumption (PFC) using 100-mm VAS using pen and paper in the fasting state and 30, 90, 120, 150, and 180 min after the standard breakfast meal. VAS were not collected at 60 min because participants were completing functional magnetic resonance imaging testing at that time point (but had blood draws via an intravenous catheter).
Subjective Measures of Appetite Traits
During the study day visit, participants completed the Three-Factor Eating Questionnaire (TFEQ) in a fasted state to measure dietary restraint (conscious restriction of food intake to prevent weight gain or promote weight loss), disinhibition (tendency to eat opportunistically), and hunger (14). This questionnaire consists of 20 questions about restraint, 16 questions about disinhibition, and 15 questions about general levels of hunger. Each item is scored either 0 or 1; therefore, the maximum possible score is 20, 16, and 15 for the restraint, disinhibition, and hunger subscales, respectively.
Ad Libitum EI
Ad libitum EI was determined using two methods: at a lunch meal on the study day visit and over a 3-d, free-living period after the study day visit, as described in Supplemental Content 1 (https://links.lww.com/TJACSM/A181). All food was provided by the CTRC Nutrition Core, and uneaten food was weighed to attain an accurate estimate of energy and macronutrient intake. EI was expressed in absolute terms (kcal or kcal·d−1) and adjusted by body weight (kcal·kg−1).
RMR and NEPA
RMR was measured at baseline (15 ± 10 d before study day visit) and after the intervention (11 ± 8 d before the study day visit) using indirect calorimetry with a ventilated hood (TrueOne 2400 metabolic cart; Parvomedics, Sandy, UT). Gas analyzers and the flowmeter were calibrated before each test according to the manufacturer’s recommendations. Participants arrived for the test after fasting overnight and avoiding exercise for 24 h. Participants rested supine in a dimly lit, thermoneutral (20°C–23°C), quiet room for 25–30 min before testing. Respiratory gas exchange was collected for 20–25 min. The average of the last 15 min of O2 and CO2 data were used to calculate RMR using the Weir equation (15). RMR was expressed in absolute terms (kcal·d−1) and adjusted for body composition using a least squares best-fit linear regression equation for RMR as a function of FM and FFM, similar to several previous publications (16–18). This population-specific equation was then used to calculate adjusted RMR for each participant at baseline and follow-up, using their corresponding follow-up FM and FFM. The resulting equation was as follows:
where FM and FFM are expressed in kilograms. The difference between calculated and measured RMR was used as an index of RMR unexplained by body composition or RMR residuals. Negative residuals represent lower-than-predicted RMR and positive residuals represent higher-than-predicted RMR.
Physical activity and sedentary behaviors were characterized using activPAL3™ accelerometers (PAL Technologies, Glasgow, Scotland). activPALs were wrapped in a waterproof nitrile sleeve and attached to the midline of the anterior thigh, one-third of the way between the hip and the knee using waterproof Tegaderm™ (3M Company, St. Paul, MN). Participants were asked to wear the devices for seven consecutive days both at baseline and at the end of the intervention (within 2 wk of the intervention concluding, i.e., during weeks 10 or 11 of the 12-wk intervention). Data processing procedures are presented in Supplemental Content 1 (https://links.lww.com/TJACSM/A181).
Data were assessed using RStudio version 1.2.5033 (R Foundation for Statistical Computing, Vienna, Austria). Type I error rate was set at 0.05. No a priori power analysis was conducted for this secondary analysis. However, post hoc analysis indicated that this investigation had 80% power to detect changes of 10 mm fasting hunger, satiety, and PFC VAS within groups and 80% power to detect differences of 10 mm fasting hunger and satiety VAS between groups (19). Data are presented as mean ± SEM unless otherwise stated. Normality was determined using the Shapiro–Wilk test; nonparametric variables were expressed as median (interquartile range). Differences in parameters between intervention groups at baseline were determined via independent samples t-test (parametric variables) or Mann–Whitney U-test (nonparametric variables). Correlations among body composition and metabolic and appetite parameters were assessed via Pearson correlation coefficient (parametric variables) or Spearman’s rho (nonparametric variables). Linear mixed effects models with unstructured covariance compared outcome variables between DIET versus AEX. Each model consisted of time (baseline and follow-up), group (DIET or AEX), and their interaction term (group–time) as fixed effects and participants as random effects; post hoc comparisons between groups and within groups across time are presented due to the exploratory nature of these analyses. Quantile–quantile (Q-Q) plots for residuals in each model were visually inspected. Model residuals for body weight, fasting GLP-1, GLP-1 AUC, lunch EI (in kcal and kcal·kg−1), percent of time in light physical activity, average steps, number of sedentary bouts, and time spent in sedentary bouts >60 min were not normally distributed. Therefore, the log of these values was used in the revised mixed models, although the absolute and mean values are presented for ease of translation.
Of 69 individuals screened, 44 participants completed the study (n = 20 DIET, n = 24 AEX). One participant in the DIET group did not have body composition measured at follow-up, one in the AEX group did not have a blood draw at baseline or follow-up, one in the DIET group did not have a follow-up blood draw, and one in the DIET did not have a follow-up RMR. These individuals were not excluded from this analysis.
The average age of the entire sample was (mean ± SD) 37 ± 9 yr, body mass index was 30.6 ± 3.1 kg·m−2, and most participants were women (n = 35, 79.5%) (Table 1). At baseline, no differences in age, body weight, body composition, RMR, or subjective appetite parameters were observed between DIET and AEX groups (Table 1). As expected, men had higher FFM (67.9 ± 12.4 vs 47.5 ± 5.4 kg, P < 0.001) and RMR (1856 ± 291 vs 1524 ± 159 kcal·d−1, P < 0.001) and lower FM (28.9 ± 5.1 vs 37.4 ± 10.0 kg, P = 0.019) and body fat percent (30.0% ± 4.0% vs 42.4% ± 4.8%, P < 0.001). However, both sexes were analyzed together because the proportion of men and women were similar between intervention groups; changes in body composition, appetite, RMR, and NEPA were similar between sexes; and a small proportion of the sample were men. Average adherence to the duration of aerobic exercise was 83.5% ± 15.9%; the AEX group exercised 3.8 ± 0.5 d·wk−1 on average (95% of exercise frequency goal). The DIET group tracked food intake 6.6 ± 0.5 d·wk−1 (94% of intervention goal). Compensation was 29.8% ± 96.2% and 52.3% ± 37.9% in the AEX and DIET groups, respectively (P = 0.355).
TABLE 1 -
Baseline Characteristics of DIET and AEX Groups.
||DIET, n = 20
||AEX, n = 24
|Sex, women, n (%)
|Race, n (%)
| Black/African American
| American Indian or Alaskan Native
|Ethnicity, n (%)
||38 ± 10
||35 ± 10
|Body weight, kg
||86.3 ± 11.5
||85.2 ± 13.8
|Body mass index, kg·m−2
||31.6 ± 3.5
||29.8 ± 2.6
|Fat-free mass, kg
||50.4 (41.2 to 59.6)
||49.1 (36.9 to 61.3)
|Fat mass, kg
||35.5 ± 7.6
||32.8 ± 7.3
|Body fat %
||41.1 ± 6.1
||38.8 ± 7.3
|Resting metabolic rate, kcal·d−1
||1567 (1317 to 1818)
||1554 (1270 to 1838)
|Fasting hunger, mm
||67.3 ± 17.3
||67.4 ± 19.3
|Hunger AUC, mm × 180 min
||5549 ± 2630
||5640 ± 2584
|Fasting satiety, mm
||11.5 (−12.5 to 24)
||4.5 (−11.3 to 22.0)
|Satiety AUC, mm × 180 min
||9242 ± 2512
||8790 ± 2881
|Fasting PFC, mm
||60.4 ± 19.4
||65.8 ± 14.8
|PFC AUC, mm × 180 min
||5726 ± 2425
||6331 ± 2856
||10.0 (1.5 to 18.5)
||9.0 (3.7 to 14.3)
||7.7 ± 3.8
||7.8 ± 3.8
||5.5 ± 3.3
||4.4 ± 3.0
P values attained from chi-square tests (sex, race, and ethnicity), independent samples t-tests (parametric continuous variables), or Mann–Whitney U-tests (nonparametric continuous variables). Parametric data are presented as mean ± SD and nonparametric data are presented as median (interquartile range).
Body Weight and Composition
Weight loss occurred in both the DIET (−3.4 ± 2.3 kg, P < 0.001) and the AEX groups (−1.7 ± 2.7 kg, P = 0.001), with no differences between groups (P = 0.058, group–time interaction). FM decreased in both groups (effect of time: P = 0.002; DIET = −2.1 ± 1.8 kg, P < 0.001; AEX = −1.3 ± 2.3 kg, P = 0.004) but was not different between groups (P = 0.215, group–time interaction). Conversely, FFM decreased in the DIET group (−1.2 ± 1.6 kg, P = 0.002) but was maintained in the AEX group (−0.4 ± 1.5 kg, P = 0.193), although this change did not differ between groups (P = 0.103, group–time interaction) (Fig. 1).
Changes in PYY, ghrelin, and GLP-1 are presented in Fig. 2 and Supplemental Content 2 (Table, https://links.lww.com/TJACSM/A182). No group, time, or group–time effects were observed for fasting or AUC values of appetite-related hormones. There were no correlations between baseline or changes in appetite-related hormones and changes in body weight, FM, or FFM (Supplemental Content 3, Table, https://links.lww.com/TJACSM/A183).
Subjective Measures of Appetite States and Traits
There were no group–time interactions for any appetite measures derived from the TFEQ. As shown in Fig. 3, dietary restraint increased over time in both groups (main effect of time: P = 0.001; DIET = 4.9 ± 1.2, P < 0.001; AEX = 2.8 ± 0.7, P = 0.003). Compared with baseline, disinhibition decreased in the AEX group (−1.5 ± 0.5, P < 0.001) and TFEQ-hunger decreased in the DIET group (−1.4 ± 0.5, P = 0.005). Dietary restraint was higher in the DIET than that in the AEX group at follow-up (6.6 ± 0.8 vs 6.3 ± 0.6, P = 0.021). There were also no differences in fasting or AUC hunger, satiety, or PFC collected by VAS within or between groups (Supplemental Content 2, Table, https://links.lww.com/TJACSM/A182). Changes in dietary restraint were negatively correlated with weight change (r = −0.433, P = 0.004) and FFM change (r = −0.383, P = 0.011), and AUC satiety at baseline was negatively correlated with FM change (r = −0.392, P = 0.009) in the entire sample (Supplemental Content 3, Table, https://links.lww.com/TJACSM/A183).
Ad Libitum EI
No significant group, time, or group–time interactions were detected for the ad libitum lunch or 3-d free-living EI in absolute terms (kcal) or adjusted for body weight (kcal·kg−1) (Supplemental Content 2, Table, https://links.lww.com/TJACSM/A182). There were also no differences in the proportion of macronutrients consumed during either condition within or between groups.
Absolute RMR change was not significantly different between groups (group–time interaction: P = 0.083). Within groups, absolute RMR decreased in the DIET group (−48 ± 21 kcal·d−1, P = 0.042) but was maintained in the AEX group (7 ± 27 kcal·d−1, P = 0.753) (Fig. 4). There were no group–time interactions or within- or between-group differences in RMR residuals (Fig. 4). Among all participants, absolute RMR change was correlated with FFM change (Spearman’s rho = 0.361, P = 0.018) and weight change (Spearman’s rho = 0.368, P = 0.015), but not with FM change (Spearman’s rho = 0.240, P = 0.121).
NEPA and Sedentary Behavior
Four participants did not have any accelerometer-derived data at baseline because of not wearing the device or device malfunction; at follow-up, 13 participants did not wear the accelerometer and 6 participants had <3 d of valid data. Therefore, valid accelerometer data were available in 39 participants at baseline (n = 17 DIET, n = 22 AEX) and 25 participants at follow-up (n = 15 DIET, n = 10 AEX). Weight or body composition parameters were not correlated with any NEPA or sedentary activity variable (baseline or change values) among all participants. No between- or within-group differences or group–time interactions were observed for sedentary, light physical activity, or moderate-to-vigorous physical activity parameters Supplemental Content 4 (Table, https://links.lww.com/TJACSM/A184).
This is the first investigation to delineate independent changes in appetite-related hormones, subjective appetite, ad libitum EI, RMR, and NEPA in adults with overweight or obesity randomized to a DIET versus AEX weight loss intervention. Although the DIET group lost 1.6 kg more weight on average than the AEX group, this difference and that of appetite hormones, subjective appetite, ad libitum EI, body composition–adjusted RMR, or NEPA changes did not differ between groups. However, increased dietary restraint was observed across time in both groups and was correlated with weight loss and FM loss, suggesting that this parameter was altered as a function of weight loss, regardless of intervention. We also found modest evidence to suggest that absolute RMR and TFEQ-hunger decreased during the DIET intervention and disinhibition decreased during the AEX intervention, although these changes were not different between groups.
Current understanding of the effect of diet versus exercise on appetite-related hormones originates exclusively from interventions of ≤4 d. For example, in one cross-over trial in healthy lean males, ghrelin AUC was higher and PYY AUC was lower after a single instance of EI restriction compared with a matched aerobic exercise energy deficit (~1150 kcal) and sedentary control (20). This finding was corroborated in healthy males with a smaller energy deficit (~480 kcal) (21) and in healthy females (~850 kcal deficit) (22). In contrast to these short-term interventions, the present study reported no significant within- or between-group changes in appetite-related hormones. Null differences in appetite hormone changes between groups may be partially explained by the lack of statistical significance in weight loss between groups. Because ghrelin increases in a compensatory manner to weight loss induced by combined diet and exercise (23), it is possible that the differences in weight loss between groups (i.e., 1.6 kg) were not large enough to induce significant differences in appetite regulation parameters. Our overall nonsignificant results are in line with that observed in a weight loss maintenance paradigm. In a study conducted by our group, 13 women with obesity were randomized to either a 12-wk diet or exercise training program for an average of 9.1% weight loss (24). Similar to the current results, there were no differences in weight loss maintenance, ghrelin, PYY, GLP-1, or leptin. These observations support the notion that the modality of weight loss or weight loss maintenance may not affect hormonal regulators of appetite. However, it is possible that studies of longer duration, larger sample sizes, or those that produce greater weight loss may reveal different appetitive hormone responses to a meal.
Subjective ratings of dietary restraint, disinhibition, and hunger are key determinants of subsequent EI and may predict successful weight loss (25,26). As recently reviewed by Thivel et al. (11), several studies have shown that short-term energy deficit (≤4 d) through EI restriction in healthy adults induces increased hunger (20,27,28), food cravings (27), PFC (20), and general appetite (21,22) and lower satiety (20,28) compared with matched energy deficit through exercise or sedentary control conditions. Short-term EI restriction also leads to higher compensatory EI at a later meal compared with exercise or sedentary control conditions (20,22,27–29). Based on the TFEQ appetite trait measures, the DIET group in the present study exhibited decreased hunger after the intervention, the AEX group had decreased disinhibition, and both groups increased restraint compared with baseline, although there were no differences in the changes in these parameters between groups (i.e., no group–time interactions). A recent study found that adults with overweight or obesity who lose the most body weight during diet interventions have higher hunger at baseline and greater reductions in disinhibition and hunger during a dietary intervention (30). It is possible that those with higher trait hunger experience an uncoupling between hunger and EI during weight loss owing to the prescribed dietary counseling conducted as part of the intervention; this may not be reflected in laboratory settings of hunger, such as state hunger as assessed by VAS. Therefore, hunger may decrease in response to DIET as an ancillary function of the intervention itself (i.e., improved subjective appetite control). A similar phenomenon may occur with exercise; the current results mirror that of a recent publication showing decreased disinhibition in response to a 12-wk exercise training intervention in adults with overweight or obesity (31). These findings add to the body of literature, suggesting that specific aspects of eating behavior may be differentially altered according to weight loss modality. However, whether appetite traits are apparent over longer terms in people with obesity is unclear, although some evidence suggests that 2 yr of EI restriction in adults without obesity negligibly affects appetite (32). Future research should consider the predictive capacity of eating behavior traits for weight loss in longer-term trials.
Notably, the lack of significant differences in appetite, metabolic, and behavioral changes should not be viewed negatively, as previous research suggests that AEX may promote over the long term (33). Aerobic exercise increases activity energy expenditure, which may act as a tonic signal to increase EI (34). In other words, increased energy expenditure associated with exercise should theoretically promote increased EI in a coordinated effort to restore energy balance. However, our data suggest that 12 wk of AEX does not affect aspects of appetite, RMR, or NEPA in a manner that is substantially different than a DIET intervention. These data add credence to the notion that exercise may help support short-term weight loss by attenuating the potential compensatory increases in appetite in response to increased energy expenditure, although direct comparisons of these outcomes between DIET and AEX in more rigorous, longer-term interventions are lacking.
Although changes in appetite and EI were the primary outcomes of this analysis, we also investigated changes in RMR as an energy balance parameter that may differ according to weight loss modality. Whether lower-than-expected RMR (i.e., adaptive thermogenesis) occurs during weight loss is a subject of debate in the literature and may be dependent on the rate and extent of weight loss (16,35,36). The present investigation is the first to compare RMR adaptive thermogenesis between DIET and AEX interventions and demonstrates that lower-than-predicted RMR is not apparent after 12 wk of weight loss from either modality; however, this may be due in part to the modest weight loss and short duration of the study.
Contrary to our hypothesis, we found no persistent evidence that NEPA was negatively affected by participation in either intervention. These findings are consistent with research among individuals with overweight or obesity in which accelerometer-derived NEPA does not change with exercise or dietary restriction (18,37) but contradict several other findings in which NEPA decreases during weight loss interventions (38). A factor that may contribute to reduced NEPA is a high degree of weight loss. In a systematic review, Silva et al. (38) reported that intervention arms with decreased nonexercise activity thermogenesis (total energy expenditure−RMR) had higher weight loss compared with those without alterations in nonexercise activity thermogenesis. Reduced NEPA may change in a similar manner as nonexercise activity thermogenesis as a response to defend body weight after substantial weight reduction. If this speculation is indeed correct, weight loss in the present study was likely not enough to elicit compensatory responses in NEPA.
The present study has several strengths, including randomized design, individualized AEX prescriptions, and use of hormonal measures of appetite, objective EI, and accelerometer-derived NEPA (as opposed to participant self-recall). However, limitations should also be considered when interpreting our results. This was a secondary analysis not powered a priori for changes in appetite, body composition, RMR, NEPA, or sedentary behaviors between groups. A weight-stable control group was not included but should be considered in future studies to better interpret changes in energy balance parameters. Although we used a number of techniques for assessing adherence to both interventions, we are unable to attain exact metrics of energy deficits; however, estimates of compensation in relation to prescribed energy deficits did not differ between groups. Future studies may consider collecting more objective, controlled measures of adherence to EI restriction such as total energy expenditure via doubly labeled water in periods of energy balance to confirm these findings. Additionally, complete activPAL data were not available for the entire cohort, and exercise sessions were manually removed from the output. As a secondary analysis, the sample size may have been insufficient to detect between-group differences in outcomes, especially considering the moderate number of results that were not statistically significant but may be clinically relevant (e.g., body weight, fasting ghrelin, AUC ghrelin, RMR). Nevertheless, this investigation provides preliminary data necessary to support the generation of larger studies designed to compare metabolic and behavioral alterations in response to single- and combined-modality behavioral interventions.
Collectively, our results demonstrate that both DIET and AEX may reduce body weight and FM and suggest that FFM and RMR are preserved during AEX. Despite assumed higher energy expenditure, AEX did not lead to compensatory alterations in appetitive hormones, EI, RMR, or NEPA. However, modest evidence suggests that DIET and AEX may result in differential effects on disinhibition and hunger traits. Future interventions of longer durations, larger sample sizes, more diverse methods to induce energy balance change (e.g., resistance exercise, macronutrient modulation), and more robust methods of energy balance assessment are warranted and may help identify more targeted and personalized weight loss strategies.
The authors thank the core laboratory, dietary services, and metabolic kitchen of the University of Colorado Hospital Clinical Translational Research Center and the fitness center at the University of Colorado Anschutz Health and Wellness Center for their support in conducting this trial.
Drs. Legget, Tregellas, Melanson, and Cornier are supported by resources from the Rocky Mountain Regional Veterans Affairs Medical Center (Geriatric Research Education and Clinical Center; Research Service; I01CX00141 and Research Career Scientist Award [J.R.T.]; I01CX001949 [K.T.L.]). The contents do not represent the views of the United States Department of Veterans Affairs or the United States Government. This research was supported by the American Diabetes Association grant no. 1-14-TS-07 and NIH grant nos. UL1 TR000154, UL1 RR025780, T32 DK007658-29, and K01 DK100445. The funding bodies played no role in the design of the study; in the collection, analysis, and interpretation of data; or in the writing of the manuscript. The authors have no conflicts of interest.
The results of this study do no constitute endorsement by the American College of Sports Medicine.
1. Riou ME, Jomphe-Tremblay S, Lamothe G, et al. Predictors of energy compensation during exercise interventions: a systematic review. Nutrients
2. Anderson JW, Konz EC, Frederich RC, et al. Long-term weight-loss maintenance: a meta-analysis of US studies. Am J Clin Nutr
3. DeBenedictis JN, Nymo S, Ollestad KH, et al. Changes in the homeostatic appetite system after weight loss reflect a normalization toward a lower body weight. J Clin Endocrinol Metab
4. Pfluger PT, Kampe J, Castaneda TR, et al. Effect of human body weight changes on circulating levels of peptide YY and peptide YY3-36. J Clin Endocrinol Metab
5. Rosenbaum M, Hirsch J, Gallagher DA, et al. Long-term persistence of adaptive thermogenesis in subjects who have maintained a reduced body weight. Am J Clin Nutr
6. Schutz Y, Nguyen DMT, Byrne NM, et al. Effectiveness of three different walking prescription durations on total physical activity in normal- and overweight women. Obes Facts
7. Dorling J, Broom DR, Burns SF, et al. Acute and Chronic Effects of Exercise on Appetite, Energy Intake, and Appetite-Related Hormones: The Modulating Effect of Adiposity, Sex, and Habitual Physical Activity. Nutrients
8. Beaulieu K, Blundell JE, van Baak MA, et al. Effect of exercise training interventions on energy intake and appetite control in adults with overweight or obesity: a systematic review and meta-analysis. Obes Rev
9. Franz MJ, VanWormer JJ, Crain AL, et al. Weight-loss outcomes: a systematic review and meta-analysis of weight-loss clinical trials with a minimum 1-year follow-up. J Am Diet Assoc
10. Thomas DM, Bouchard C, Church T, et al. Why do individuals not lose more weight from an exercise intervention at a defined dose? An energy balance analysis. Obes Rev
11. Thivel D, Metz L, Julian V, et al. Diet- but not exercise-induced iso-energetic deficit induces compensatory appetitive responses. Eur J Clin Nutr
12. Grunwald GK, Melanson EL, Forster JE, et al. Comparison of methods for achieving 24-hour energy balance in a whole-room indirect calorimeter. Obes Res
13. Allison DB, Paultre F, Maggio C, et al. The use of areas under curves in diabetes research. Diabetes Care
14. Stunkard AJ, Messick S. The Three-Factor Eating Questionnaire to measure dietary restraint, disinhibition and hunger. J Psychosom Res
15. Weir JB. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol
16. Fothergill E, Guo J, Howard L, et al. Persistent metabolic adaptation 6 years after “The Biggest Loser” competition. Obesity (Silver Spring)
17. Wolfe BM, Schoeller DA, McCrady-Spitzer SK, et al. Resting metabolic rate, total daily energy expenditure, and metabolic adaptation 6 months and 24 months after bariatric surgery. Obesity (Silver Spring)
18. Martin CK, Heilbronn LK, de Jonge L, et al. Effect of calorie restriction on resting metabolic rate and spontaneous physical activity. Obesity (Silver Spring)
19. Flint A, Raben A, Blundell JE, et al. Reproducibility, power and validity of visual analogue scales in assessment of appetite sensations in single test meal studies. Int J Obes (Lond)
20. King JA, Wasse LK, Ewens J, et al. Differential acylated ghrelin, peptide YY3-36, appetite, and food intake responses to equivalent energy deficits created by exercise and food restriction. J Clin Endocrinol Metab
21. Deighton K, Batterham RL, Stensel DJ. Appetite and gut peptide responses to exercise and calorie restriction. The effect of modest energy deficits. Appetite
22. Alajmi N, Deighton K, King JA, et al. Appetite and energy intake responses to acute energy deficits in females versus males. Med Sci Sports Exerc
23. Leidy HJ, Gardner JK, Frye BR, et al. Circulating ghrelin is sensitive to changes in body weight during a diet and exercise program in normal-weight young women. J Clin Endocrinol Metab
24. Foright R, Halliday TM, Melanson EL, et al. Effects of exercise during weight loss maintenance on appetite regulation in women. Transl J Am Coll Sports Med
25. Urbanek JK, Metzgar CJ, Hsiao PY, et al. Increase in cognitive eating restraint predicts weight loss and change in other anthropometric measurements in overweight/obese premenopausal women. Appetite
26. Thomas EA, Mcnair B, Bechtell JL, et al. Greater hunger and less restraint predict weight loss success with phentermine treatment. Obesity (Silver Spring)
27. Hubert P, King NA, Blundell JE. Uncoupling the effects of energy expenditure and energy intake: appetite response to short-term energy deficit induced by meal omission and physical activity. Appetite
28. Thivel D, Finlayson G, Miguet M, et al. Energy depletion by 24-h fast leads to compensatory appetite responses compared with matched energy depletion by exercise in healthy young males. Br J Nutr
29. Cameron JD, Goldfield GS, Riou ME, et al. Energy depletion by diet or aerobic exercise alone: impact of energy deficit modality on appetite parameters. Am J Clin Nutr
30. Hansen TT, Hjorth MF, Sandby K, et al. Predictors of successful weight loss with relative maintenance of fat-free mass in individuals with overweight and obesity on an 8-week low-energy diet. Br J Nutr
31. Beaulieu K, Hopkins M, Gibbons C, et al. Exercise training reduces reward for high-fat food in adults with overweight/obesity. Med Sci Sports Exerc
32. Dorling JL, Das SK, Racette SB, et al. Changes in body weight, adherence, and appetite during 2 years of calorie restriction: the CALERIE 2 randomized clinical trial. Eur J Clin Nutr
33. Doucet E, McInis K, Mahmoodianfard S. Compensation in response to energy deficits induced by exercise or diet. Obes Rev
. 2018;19(Suppl 1):36–46.
34. Hopkins M, Duarte C, Beaulieu K, et al. Activity energy expenditure is an independent predictor of energy intake in humans. Int J Obes (Lond)
35. Doucet E, St-Pierre S, Almeras N, et al. Evidence for the existence of adaptive thermogenesis during weight loss. Br J Nutr
36. Ostendorf DM, Melanson EL, Caldwell AE, et al. No consistent evidence of a disproportionately low resting energy expenditure in long-term successful weight-loss maintainers. Am J Clin Nutr
37. Church TS, Martin CK, Thompson AM, et al. Changes in weight, waist circumference and compensatory responses with different doses of exercise among sedentary, overweight postmenopausal women. PLoS One
38. Silva AM, Judice PB, Carraca EV, et al. What is the effect of diet and/or exercise interventions on behavioural compensation in non-exercise physical activity and related energy expenditure of free-living adults? A systematic review. Br J Nutr