Increasing energy flux to decrease the biological drive toward weight regain after weight loss – A proof-of-concept pilot study

Clinical Nutrition ESPEN xxx (2015) e1ee9

Contents lists available at ScienceDirect

Clinical Nutrition ESPEN journal homepage: http://www.clinicalnutritionespen.com

Original article

Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study* Hunter L. Paris b, 1, Rebecca M. Foright a, 1, Kelsey A. Werth a, Lauren C. Larson a, Joseph W. Beals b, Kimberly Cox-York a, Christopher Bell b, Christopher L. Melby a, * a b

Department of Food Science & Human Nutrition, Colorado State University, Fort Collins, CO 80523-1571, USA Department of Health and Exercise Science, Colorado State University, Fort Collins, CO 80523-1571, USA

a r t i c l e i n f o

s u m m a r y

Article history: Received 12 May 2015 Accepted 2 November 2015

Objective: Weight loss induces compensatory biological adjustments that increase hunger and decrease resting metabolic rate (RMR), which increase propensity for weight regain. In non-obese adults high levels of physical activity coupled with high energy intake (high energy flux) are associated with higher RMR and reduced hunger. We tested the possibility that a high flux state attenuates the increase in hunger and the decrease in RMR characteristic of diet-induced weight loss. Methods: Six obese adults [age (mean ± SE) ¼ 42 ± 12 y; body mass index (BMI) ¼ 35.7 ± 3.7 kg/m2] underwent measures of RMR, the thermic effect of a meal (TEM), and fasting and postprandial measures of hunger and fullness as well as plasma glucose and insulin. Following weight loss, subjects completed two 5-day conditions of energy balance in random orderdLow Flux (LF): sedentary with energy intake (EI) ¼ RMR (kcal/d)  1.35; and High Flux (HF): net exercise energy cost of ~500 kcal/d and EI ¼ RMR (kcal/d)  1.7. RMR was measured daily for each flux condition. The morning following each of the respective experimentally controlled HF and LF conditions (flux day 5), they underwent the same preweight loss tests and also reported their perceptions of hunger and fullness during the previous four days of HF and LF, respectively. Results: Average daily RMR was higher during HF (1926 ± 138 kcal/day) compared to LF (1847 ± 126 kcal/ day; P < 0.05). Perceived hunger at the end of day was lower (p < 0.03) and fullness throughout the day was higher (p < 0.02) in HF compared to LF conditions. On day 5 of each flux condition, the thermic effect of a meal and circulating glucose and insulin after the meal did not differ between HF and LF. Conclusion: Following weight loss, compared to a sedentary LF state of energy balance, a short-term HF energy balance state is associated with higher RMR, lower perceived hunger, and greater perceived fullness, all of which could help attenuate the biologic drive to regain weight. Given the pilot nature of this study and the relatively short period of time spent in the high and low flux states, future research is needed to address this research question in a larger sample over a longer time period. © 2015 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.

Keywords: Weight loss Metabolism Obesity Hunger Exercise Weight regain

1. Introduction Long-term weight loss results are usually modestdthe more common experience among individuals who diet to lose weight is ultimately, weight regain [1]. This recidivism is often attributed to

*

Presented at the Experimental Biology Meeting 2014, San Diego, CA, USA. * Corresponding author. Department of Food Science & Human Nutrition, Nutrition and Metabolic Fitness Laboratory, Colorado State University, 216 Gifford Building, Fort Collins, CO 80523-1571, USA. Tel.: þ1 970 491 6736. 1 Co-first authors.

volitional behavior and pressures from an obesogenic environment [2]. However, in addition to these factors, increasing emphasis is being given to the biological/metabolic regulators of energy balance. Two biological adjustments that are driving forces in weight regain are an increased appetite and reduced energy expenditure (EE). Weight loss from caloric restriction increases the internal drive to eat [3,4] and also leads to a reduction in total daily energy expenditure (TDEE) [5,6]. Dieting lowers resting metabolic rate (RMR), which contributes to an overall reduction in caloric expenditure and, coupled with an up-regulated internal drive to

http://dx.doi.org/10.1016/j.clnesp.2015.11.005 2405-4577/© 2015 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005

e2

H.L. Paris et al. / Clinical Nutrition ESPEN xxx (2015) e1ee9

eat, acts as a catalyst for regression back into positive energy balance and weight regain. Together the weight loss-induced homeostatic adjustments of increased hunger and reduced EE can promote a change toward positive energy balance, return to prediet levels of body weight, and can negate the health benefits of weight loss. Hill et al. [7] and MacLean et al. [8] have used the term ‘energy gap’ as a descriptor of the mismatch between desired and required calories following weight loss. The development of approaches to attenuate the biological drive to regain lost weight by attenuating this energy gap is central to enhancing the sustainability of weight loss and its health-related benefits. One possible approach to countering the tendency towards weight regain is by influencing energy flux. The concept of energy flux as referred to in this study pertains to the total throughput of energy (intake and expenditure) in the body. In order for a person in a low flux (LF) sedentary state to maintain energy balance, low energy intake is required to match the low expenditure. A high flux (HF) state is characterized by significantly greater energy throughput, with a higher expenditure coupled to the higher intake. In 1956 Mayer et al. [9], based on examination of Bengali workers, hypothesized that energy intake is more accurately regulated to match energy expenditure under conditions of higher rather than lower physical activity. This hypothesis is supported by more recent work by Stubbs et al. [10] in which decreased physical activity energy expenditure from 1.8 to 1.4 times RMR was not accompanied by a compensatory decrease in energy intake. There is also evidence that a high energy flux state resulting from exercise may increase resting energy expenditure. In three separate cross-sectional studies, we have shown that in contrast to a LF state, individuals in high energy flux owing to significant daily exercise, experience not only the additional energetic cost of the exercise itself, but also exhibit higher RMR values when measured the morning after each previous high flux day [11e13]. Given these findings, it is possible that maintenance of a HF state after weight loss, achieved by coupling a high level of physical activity to a higher dietary intake necessary to meet energy requirements could at least partially attenuate the two main problems that promote weight regaindincreased hunger and reduced energy expenditure. Therefore, in this pilot, proof-ofconcept study, we tested the hypothesis that, following clinicallyrelevant weight loss, individuals maintained in a HF state would exhibit higher RMR values and lower perceptions of hunger as compared to a LF state.

2.2. Experimental protocol As shown in Fig. 1, the study protocol involved four sequential phases: 1) pre-weight loss baseline testing; 2) diet-induced weight loss; 3) weight loss stabilization for three weeks; and 4) HF and LF conditions, which included 4 days of experimentally-induced high and low energy flux, respectively, followed by an experimental day in which the same tests performed at baseline were repeated. 2.3. Baseline testing Pre-weight loss baseline measurements occurred on two separate days. The first visit involved completion of a health history screening questionnaire and body composition analysis. For screening purposes, measurements of blood pressure via standard sphygmomanometry were also taken during this initial visit, as well as a 12-lead ECG at rest and during an incremental stationary cycle ergometry exercise test to exhaustion. The second visit was used to examine perceptions of hunger and satiety, blood glucose, and insulin in response to a small breakfast. Participants reported to the clinical laboratory between 0600 and 0800. Following measurement of RMR, a venous catheter was placed in an antecubital vein and connected to a saline IV drip for catheter patency. A fasting blood sample was obtained, which was then followed by subjects consuming a small liquid breakfast (Ensure, Ross Laboratories, Abbott Park, IL; 64% CHO, 22% fat, 14% protein) with a caloric value of 20 percent of measured RMR. Subjects were given 10 min to consume the liquid meal and then blood samples were collected at 30 min intervals for the next 3 h. For lunch, an ad libitum buffet of pre-weighed food was provided. Food items included a pre-made sandwich cut into small sections with the subject choosing the types and amounts of meat, greens and other vegetables, tomatoes, cheese, and dressing. They were also provided with a variety of other foods including chips, yogurt, apples, bananas, various candy bars, cookies, water, apple juice and milk. Subjects ate in a quiet room and were given 30 min to complete their meal. Remaining food was reweighed to determine the total amount of ingested food for each food item for later conversion to total kcal ingestion using diet analysis software (Nutritionist ProTM. Axxya Systems, LLC, Stafford, TX). Participants rated their perceptions of hunger and satiety at 30, 60, 90, 120, and 180 min following the lunch. This same experimental protocol was repeated following weight loss on day 5 of both HF and LF conditions, respectively. 2.4. Weight loss phase

2. Materials and methods 2.1. Study participants Eleven obese adults were initially recruited for the study. Criteria for inclusion were body mass index (BMI) between 30 and 43 kg/m2, age 18e55 years, normotensive, ability to exercise, and desire to lose weight. Individuals who were currently pregnant, breast feeding, smoking, using medications known to affect metabolism or appetite, had prior surgery for weight loss, and/or had dieted over the previous 12 months, were excluded from participation. Additionally, the presence of any contraindication at rest or during incremental exercise based on high blood pressure or an abnormal 12-lead electrocardiogram (ECG) were criteria for exclusion. The study protocol was approved by the Institutional Review Board at Colorado State University. The nature, purpose, and risks of the study were explained to each subject before written, informed voluntary consent was obtained. Of the 11 subjects initially enrolled, three discontinued participation in the study owing to inadequate treatment protocol compliance and two due to work conflicts.

Upon completion of pre-weight loss baseline measurements, participants were counseled on adherence to a moderate reduction in daily energy intake under free-living conditions in order to lose 7% of their baseline body mass over the course of 8e12 weeks. Dietary energy reduction was designed such that participants would lose approximately 0.5e1.0 kg per week, based on measured RMR and reported activity levels. Participants reported weekly to be weighed and to receive dietary counseling. Participants were asked to maintain their typical physical activity levels. None of the individuals were involved in a regular sustained formal exercise routine nor did they initiate an exercise program as part of the weight loss program. 2.5. Weight stabilization phase Following the targeted 7% reduction in body mass, participants entered a three-week period of weight stabilization, with counseling provided regarding caloric adjustments made to maintain energy balance and keep body mass constant. Subjects were

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005

H.L. Paris et al. / Clinical Nutrition ESPEN xxx (2015) e1ee9

e3

Fig. 1. Study Protocol with four phases: 1) Subjects completed baseline testing prior to weight loss; 2) subjects initiated a hypocaloric diet to produce 7% weight loss; 3) subjects entered a weight stabilization period for three weeks; and 4) HF and LF conditions, which included 4 days of experimentally-induced high and low energy flux, respectively, followed by an experimental day in which the same series of tests performed at baseline day 2 were repeated. The order of high and low flux was randomized.

weighed regularly during this period of time and directed to modify caloric intake upwards or downwards based on any small changes in body weight. A second incremental test to exhaustion was performed to determine VO2peak and the appropriate submaximal workload for the exercise performed during the HF condition.

2.6. Experimental flux conditions Three weeks after the onset of the stabilization phase, participants began a two-week protocol: five days of a LF condition, a three day washout period, and five days of a HF condition, or the reverse order of these conditions as randomly assigned.

2.6.1. Low flux During the LF condition meticulous efforts were made to maintain tight energy balance. Body mass and RMR were measured each morning, and all daily meals and snacks were prepared in the nutrition research kitchen with food weights measured to the nearest gram and fluid volume measured to the nearest ml in order to provide subjects with their respective estimated energy intake requirement necessary to maintain energy balance, i.e. an energy intake target of RMR  1.35 [Physical activity level (PAL) ¼ 1.35 for a sedentary condition]. The macronutrient composition of the food provided each day as meals and snacks was kept constant at 50% energy carbohydrate, 35% fat, and 15% protein. Upon completion of the RMR test, subjects were provided with each day's meals in sealed containers with instructions for heating the food as necessary. They were told to consume all food provided and to ingest the food at regular meal intervals. They were further instructed to consume no other foods and beverages outside of those provided by the study investigators. They were asked to not ingest alcohol and coffee. They returned the previous day's food containers each morning of RMR measurement. If body mass began to rise or fall, small adjustments in the caloric content of given meals were made. Participants did not engage in any exercise during these four days and were asked to remain as sedentary as possible. The morning of day 5 included measurements of RMR, TEM, fat oxidation, ad libitum food intake, and fasting and postprandial hunger and fullness. The protocol for this day was the same as day 2 baseline pre-weight loss, which was described earlier.

2.6.2. High flux We estimated that for participants to remain in energy balance during the HF condition they would require an EI of approximately 1.7  RMR. Thus, EI was increased to compensate for the elevated level of physical activity. In a manner similar to that described for the LF condition, body mass and RMR were measured every morning, and food for the day was provided to the participant based on his or her RMR and body mass that morning. The macronutrient distribution of the meals and snacks consumed during HF was the same as LF. Based on our previous studies, we projected that participants would need to expend approximately 600e750 additional kcal/day through daily exercise plus the excess post-exercise energy expenditure and increased activities of daily living. Following morning measurements of body mass and RMR, participants used a cycle ergometer and/or treadmill to achieve a net increase in daily caloric expenditure of 500 kcal at an exercise intensity of 60% VO2peak. The net increase corresponds to a value exceeding by 500 kcal the energy expenditure during this same time period when in the LF condition. The duration of exercise sessions varied between individuals based on VO2peak values and ranged from 70 to 100 min. All exercise bouts took place either immediately following RMR measurements or later in the afternoon/early evening. Every exercise session was supervised by a member of the investigative team. Subjects completed a 5 min warm-up period and then the intensity of the exercise was increased to 60% VO2peak. Respiratory gas exchange measures were taken at regular intervals and any adjustments in intensity were made to maintain the workload at 60% VO2peak. Heart rate (beats/ min) was measured using a heart rate monitor (FT1, Polar Electro Inc., Lake Success, NY). Additionally, apart from the exercise, participants were asked to increase their usual number of daily steps by walking more in performing their usual tasks of daily living. Testing on day 5 of HF began in the morning immediately following the last exercise day of HF condition, and involved postintervention measurements according to the same experimental protocol as in the LF week. 2.7. Specific procedures 2.7.1. Body composition Height was recorded to the nearest millimeter and body mass to the nearest 100 g by use of a stadiometer and beam scale (Detecto,

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005

e4

H.L. Paris et al. / Clinical Nutrition ESPEN xxx (2015) e1ee9

Webb City, MO, USA). Fat mass (FM), fat free mass (FFM), and bone mineral content were measured using duel-energy X-ray absorptiometry (Hologic, Discovery W, QDR Series, Bedford, MA, USA). 2.7.2. Resting metabolic rate and resting fat oxidation Subjects arrived at the laboratory after a minimum 10-h fast prior to any significant physical activity on the day. Participants drove themselves to the laboratory, parked within 40 m, and walked to the clinical room where testing occurred. Calibration of the instruments (Mass Spectrometer (Perkin Elmer MGA 1100, MA Tech Services, Inc., St. Louis, MO) or Parvo (Parvo TrueOne 2400 Metabolic Measurement System, Parvo Medics, Sandy, UT)) with known gas concentrations took place prior to subject arrival. VO2 and VCO2 values were obtained using indirect calorimetry to estimate RMR while subjects lay quietly for 45 min in a dim lit room, with the first 15 min used for habituation and the last 30 min to determine RMR. The two different indirect calorimetry systems were used owing to other simultaneously conducted studies requiring the equipment. However, to maintain internal consistency, each participant was measured throughout the entire study with the same calorimetry system. The Weir equation was used to convert respiratory gas exchange measures to kcal [14]. The RMR measures for the indirect calorimetry systems have been previously shown to be reproducible with a typical margin of error of 2.5% [15]. Rates of fat oxidation (g/min) were calculated from VO2 and VCO2 values collected during minutes 15e45 of morning RMR measurements in the HF and LF conditions. The calculations were made according to a previously published equation [16]: Fat oxidation (g/ min) ¼ 1.695 VO2 e 1.701 VCO2. For each of the flux conditions, RMR was measured for 5 consecutive days. The RMR measured on day 1 was used only to calculate the required energy intake values for the first day of the flux condition. The last 4 RMR measurements reflecting the RMR measured the morning following each flux day. 2.7.3. VO2peak test Peak oxygen uptake was determined using an incremental exercise test to volitional exhaustion. Heart rate was recorded using a 12-lead ECG for baseline testing, and a short-range telemetry device (FT1, Polar Electro Inc., Lake Success, NY) for testing poststabilization. Participants were outfitted with a two-way nonrebreathing mouthpiece, valve, and headgear device (HansRudolph, St. Louis, MO). Exercise was performed on a cycle ergometer (Velotron Dynafit Pro, RacerMate, Inc., Seattle, WA, USA), and consisted of a 15e30 Watt/min continuous ramp protocol from 0 Watts. The test was terminated upon volitional exhaustion or once pedal cadence fell below 40 revolutions/min. Heart rate and blood pressure were recorded at rest, in the exercise position, and every two to 3 min during the protocol. The composition of expired gases, as well as ventilation, was measured continuously with a metabolic cart (Parvo TrueOne 2400 Metabolic Measurement System, Parvo Medics, Sandy, UT). The four highest consecutive 15-s average VO2 values were used to calculate VO2peak. 2.7.4. Plasma assays and analysis Fasting and postprandial blood glucose was measured using the glucose oxidase method on an automated glucose analyzer (2300 STAT Plus Glucose Lactate Analyzer, YSI Inc., Yellow Springs OH). Because overfeeding and underfeeding are associated with changes in plasma insulin concentrations, i.e. a change from zero energy balance to positive energy balance increases circulating insulin concentrations and a change to negative energy balance decreases circulating insulin [11]. Therefore, as a marker of energy balance during high and low flux conditions, we measured fasting insulin as well as postprandial insulin in response to the liquid breakfast meal used to determine TEM. A commercially available MILLIPLEX MAP

Human Metabolic Hormone Magnetic Bead Panel (Millipore, Billerica, MA, USA) was used to measure plasma insulin. 2.7.5. Measures of hunger and satiety Perceived hunger and satiety were determined using a 100 mm visual analog scale questionnaire at each 30 min interval following breakfast and lunch during baseline day 2 and day 5 of HF and LF, respectively. The questionnaire consisted of four questions: (1) How full do you feel? (2) How much could you eat right now? (3) How strong is your desire to eat? (4) How hungry are you? On this same day, participants were asked to complete a short questionnaire designed to address their overall feeling of hunger and satiety when reflecting retrospectively on the prior four days of LF or HF. Subjects answered the following questions by marking a vertical line on a 100 mm visual analog scale. (1) “On average, how hungry did you feel at the end of these 4 days?” with anchors with 0- Not hungry, plenty of food and 100- Very hungry, not enough food. (2) “On average, how hungry did you feel throughout each of these 4 days?” with the anchors 0- Not hungry, plenty of food and 100Very hungry, not enough food. (3) “On average, how full did you feel in the evening just prior to bedtime during these 4 days?” with the anchors 0- Never full, wanted more to eat and 100- Full, could not eat anymore. 2.8. Statistical analyses 2.8.1. Data analyses were performed using SPSS version 22 Comparisons among baseline, high flux, and low flux condition variables were made using repeated measures analysis of variance (ANOVA). To account for the possible effect of changes in body composition on RMR amongst paired observations, a mixed models approach was used. Least Significant Differences Tests were used for post-hoc comparisons among the three conditions. For single variables examined between low and high flux conditions only, data were analyzed using paired t-tests and for variables measured at multiple time points during the post-prandial period, a two-way repeated measures ANOVA (experimental condition  time) was used. Statistical significance was set at p < 0.05. Values are expressed as means ± standard errors. 3. Results 3.1. Subject characteristics The baseline physical characteristics of the study participants are shown in Table 1. All individuals (4 males, 2 females) exhibited BMI values greater than 30 kg/m2 and relatively low levels of cardiorespiratory fitness. In accord with our study design, the six participants succeeded in losing 6.9% of their body weight (preweight loss body mass: x þ SEM ¼ 110.7 þ 4.0 kg; post weight loss body mass x þ SEM ¼ 103.1 þ 4.0 kg) and weight loss was maintained during the scheduled three weeks of weight maintenance.

Table 1 Pre-weight loss baseline participant characteristics (n ¼ 4 males, 2 females). Variable

Mean ± SE

Age Height (cm) Body mass (kg) Body mass index (kg/m2) Body fat (%) Resting metabolic rate (kcal/day) Resting respiratory exchange ratio (VCO2/VO2)

42 176 110.7 35.7 42 1933 0.86

± ± ± ± ± ± ±

5 4 4.7 1.5 2 118 0.04

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005

H.L. Paris et al. / Clinical Nutrition ESPEN xxx (2015) e1ee9

3.2. Energy intake and body composition As per the experimental protocol, average daily energy intake for the HF condition (mean ¼ 3191 ± 262 kcal/day) was significantly greater than for the LF condition (2449 ± 182 kcal/day; p ¼ 0.001). Also consistent with the study protocol, the higher energy intake for HF was coupled with greater energy expenditure owing to daily monitored exercise, which resulted in no significant changes in body weight over the course of the HF and LF conditions. The magnitude of difference in dietary intake represents the additional energy expended in the HF condition resulting from increased physical activity (including exercise warm-up, the exercise bout itself, the excess post-exercise energy expenditure), the increased thermic effect of food owing to higher energy intake, and additional walking. Fig. 2 shows changes in body composition before and after weight loss. 3.3. Resting metabolic rate and resting substrate oxidation Daily RMR during the two flux conditions is presented in Fig. 3A. Average RMR during the four days of HF (1926 ± 138 kcal/day) was significantly higher (P < 0.05) than average RMR recorded over the four days of LF (1847 ± 126 kcal/day). Individual data showed five of six participants with a greater average RMR during the four days of HF compared to the four days of LF. A mixed model analysis showed the HF and LF RMR differences were not due to small differences in fat-free mass. The four-day average resting RER was lower in HF compared to the four days of LF, but the magnitude of the difference did not attain statistical significance (P < 0.09) (Fig. 3B). Owing to the lower RER value and the higher VO2 measured at rest, the calculated rate of fat oxidation (g/min) during HF (0.07 g/min) was significantly greater (P ¼ 0.015) than that of the LF condition (0.06 g/min) (Fig. 3C). 3.4. Thermic effect of feeding The TEF based on the liquid breakfast (relative energy intake based on 20% of measured RMR) prior to weight loss, as well as at

e5

the end of both LF and HF conditions is presented in Fig. 3D. Data represent the caloric expenditure above resting metabolism for the 3 h following the standardized breakfast. When adjusted for caloric intake, group averages were not statistically significant between HF and LF, but demonstrated a trend for TEF to be greater in the HF condition (P ¼ 0.096). Individual results showed five of six participants with a higher TEF in the HF compared to the LF state. 3.5. Blood glucose and plasma insulin The fasting and post-breakfast blood glucose and plasma insulin determinations are shown in Fig. 4. Pre-weight loss fasting blood glucose concentrations were well within the normal range and there were not significant changes with weight loss (Fig. 4A). During the post-prandial period following breakfast, as expected there was a significant time effect for blood glucose, but not a time by condition interaction. Fasting and postprandial insulin decreased following weight loss. There was a significant time effect and a significant time by condition interaction with the lower insulin concentrations in the two flux conditions consistent with improved insulin sensitivity (Fig. 4B). The lack of a difference in fasting insulin concentrations for HF and LF provides some objective evidence of similar energy balance states between the flux conditions. 3.6. Perceived hunger and fullness during LF and HF days On day 5 of the LF and HF conditions, subjects were asked to rate their perceptions of overall hunger and fullness during the previous four-day periods (Fig. 5). In the face of energy intakes carefully matched to energy expenditure necessary to maintain energy balance and body weight in both experimental conditions, the participants indicated significantly less hunger (p ¼ 0.020) at the end of each of the days during HF compared to LF (Fig. 5A). They reported significantly greater fullness at the end of each of the days during high flux (p ¼ 0.015) (Fig. 5B). There was a trend for the subjects to exhibit less hunger throughout the day during HF compared to LF (p ¼ 0.09) (Fig. 5C). 3.7. Postprandial hunger and fullness and ad libitum food intake On the experimental day 2 and day 5 of HF and LF, i.e. the morning after experimentally controlled HF and LF states, there were no differences in subjects' reported perceptions of hunger and fullness during fasting or during postprandial periods following breakfast and lunch (data not shown). There were also no differences in ad libitum food consumption from the lunch buffet (Preweight loss: 898 ± 57 kcal, LF: 1047 ± 108 kcal, HF: 1046 ± 80 kcal). 4. Discussion The present study was designed to determine if, following weight loss, a weight-stable, HF state could minimize the energy gap by dampening the weight loss-induced effects on hunger, satiety and metabolic rate in obese individuals. The main findings of this study include lower subjective hunger, higher subjective fullness and higher average RMR in the weight-reduced participants during several days of a high flux state compared to a four-day low flux (sedentary) state.

Fig. 2. Body composition at baseline and the day following each of the low and high energy flux conditions. *P < 0.05 for body mass compared with baseline values. # P < 0.05; FFM was slightly but significantly higher and FM slightly but significantly lower in the HF condition compared to LF.

4.1. Resting and post-prandial energy expenditure and substrate oxidation The RMR data are consistent with previous work from our laboratories in which young male [11] and female adults [12] and older

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005

e6

H.L. Paris et al. / Clinical Nutrition ESPEN xxx (2015) e1ee9

A

C

*

0.14

1.4

1.2

1.0

0.10 0.08 0.06 0.04 0.02 0.00

0.8 Low Flux

High Flux

B

Low Flux

High Flux

Low Flux

High Flux

D 25

1.00

Thermic Effect of Feeding as Percent of Caloric Intake (%)

Respiratory Exchange Ratio (VCO 2 /VO2 )

*

0.12

1.6

Fat Oxidation (g/min)

Resting metabolic rate (kcal/min)

1.8

0.95 0.90 0.85 0.80 0.75 Low Flux

High Flux

20 15 10 5 0

Fig. 3. A) Participant data and condition means for resting metabolic rate; B), Resting respiratory exchange ratio; C) Resting fat oxidation for the LF days compared with the HF days; and D.) Individual values and condition means for the thermic effect of food (TEF) as a percent of the energetic value of the breakfast. Percent was calculated by dividing caloric expenditure above RMR for the 180 min following the meal, by the calories ingesting in the meal [(EE/EI)*100] (B). All values are for postprandial thermogenesis above RMR. *P < 0.05 compared to LF. Data are expressed as mean ± SE.

adults [13] were found to exhibit higher metabolic rates during high flux states compared to their own low flux states or compared to sedentary controls. To our knowledge, these data are the first to identify the experimental effect of HF on RMR and hunger/fullness in obese, weight-reduced individuals. We made no attempt in this pilot study to identify mechanisms for the elevation of RMR during the four days of HF compared to the four days of LF. However, our previous work supports the possibility that an elevated sympathetic support contributes to this increase in RMR. Previous literature has established the relationship between habitual exercise and b-adrenergic support of resting metabolism [17], as well as the role of HF in mediating this relationship and therefore elevating RMR [11,14]. b-Adrenergic thermogenic responsiveness has also been examined, and while past studies have examined endurancetrained, habitually exercising individuals, some research suggests that previously sedentary obese adults may have an elevated badrenergic thermogenic response to aerobic exercise training [18]. There is little evidence that exercise at 60% of VO2 max performed the previous day would acutely elevate recovery energy expenditure 12e18 h later when RMR was measured. In addition to sympathetic support, the key determinant of resting metabolic rate is FFM. Exercise in a HF state may conserve FFM to a greater degree than a LF state, such that reductions in RMR would be slightly offset [19]. In the present study, FFM was measured on the last day of each condition and was slightly higher in the HF state. Using a mixed model approach, our data showed no relationship between the change in FFM and the change in RMR (data not presented).

Overfeeding increases energy expenditure and can also affect sensations of hunger and fullness. Thus it is of critical importance to address the question as to whether or not the elevation of RMR during HF compared to LF may have resulted from subjects being unintentionally overfed during HF or underfed during LF or both? Body weight was not different between HF and LF conditions, but weight is not sensitive to small energy imbalances over 5 days. However, we believe subjects were in energy balance and thus our results are not confounded based on the following reasoning. The mean energy intake provided the participants in the HF state (3191 kcal) was 742 kcal above the energy intake in the LF state (2449 kcal). This difference in mean energy intake was readily accounted for by the higher daily energy expenditure in the HF condition, which we estimated to be about 750 kcal based on the following: the exercise warm-up (z10 kcal), the net energy cost of exercise (measured 500 kcal), the excess post-exercise energy expenditure (z25 kcal) the higher thermic effect of food with the additional energy provided (estimated at 8% of the additional kcal z 60 kcal), the higher RMR (measured 79 kcal) and the additional steps taken (z75e100 kcal). Also, had subjects been overfed in HF and/or underfed in LF, one would expect that resting fat oxidation the following morning would have been higher after LF than HF. We found the opposite. Lastly, fasting insulin concentrations are sensitive to changes in energy balance, and there were not differences in insulin concentrations between the two flux conditions. Nevertheless, we recognize that an inpatient protocol using a room calorimeter to measure 24-h energy expenditure, rather than an our outpatient protocol using estimates of energy

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005

H.L. Paris et al. / Clinical Nutrition ESPEN xxx (2015) e1ee9

e7

A

A 80

110

90 80 70 Pre-Weight Loss Low Flux High Flux

60

Hunger Rating (mm)

Blood Glucose (mg/dl)

100

60

40

*

20

0

50 0

60

120

180

240

300

360

Low Flux

420

High Flux

Minutes

B

B

# 70

100

Fullness Rating (mm)

Insulin (uU/ml)

60 50 40 30 20 Pre-Weight Loss Low Flux High Flux

10

80 60 40 20 0

0 0

60

120

180

240

300

360

420

Low Flux

High Flux

Low Flux

High Flux

Minutes

C

expenditure (RMR  activity factor, pedometer counts, etc.) with their obvious limitations, is a next step in addressing the question as to whether or not increased energy flux can attenuate the weight loss-induced ‘energy gap’. As indicated above, resting fat oxidation was also higher in HF compared to LF. Several studies have observed a reduction in RER up to one day following exercise, indicating elevated fat oxidation [20e22]. Our findings coincide with this past research even in the face of adequate dietary compensation for the additional energy cost of exercise. Following vigorous exercise, elevated catecholamine concentrations may result in the increased release of fatty acids for up to 24 h post exercise [23], which could extend fat oxidation in the face of glucose being shuttled towards repletion of muscle glycogen [24]. A notable aspect of the energy gap is that following weight loss, the improvement in metabolic flexibility and heightened insulin sensitivity brought on by weight loss result in the preferential oxidation of carbohydrates, and the shuttling of dietary fats to adipose and hepatic stores [25]. Directing dietary fat

80

Hunger Rating (mm)

Fig. 4. Blood glucose and plasma insulin concentrations before weight loss and after 7% weight loss in response to a standardized breakfast and ad libitum consumed lunch buffet. A) Fasting (time 0) and 3-h postprandial blood glucose. No significant interaction or condition effects; B) Plasma insulin. For or the entire 180 min there was a significant time effect (p < 0.01) and a significant condition effect (p < 0.05), such that HF, LF < pre-weight loss. Data are expressed as mean ± SE. indicates time the meals were consumed.

60

40

20

0

Fig. 5. Overall perceived hunger/fullness during LF and HF phases completed at the end of each LF and HF condition (n ¼ 6). A) Individual and average ratings of perceived hunger based on the question “On average, how hungry did you feel at the end of these 4 days?” (*p ¼ 0.02). B) Individual and average ratings of perceived fullness based on the question “On average, how full did you feel in the evening just prior to bedtime during these 4 days?” (#p ¼ 0.015). C) Individual and average ratings of perceived hunger based on the question “On average, how hungry did you feel throughout each of these 4 days?” (p ¼ 0.09). Data are expressed as mean ± SE.

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005

e8

H.L. Paris et al. / Clinical Nutrition ESPEN xxx (2015) e1ee9

to triglyceride stores is energetically more efficient than converting carbohydrates and proteins into fats and storing them accordingly [26]. Therefore, less energy is expended, adipose triglyceride deposition increases. Thus, in the present study the higher fat oxidation under resting HF conditions may be important in attenuating the biologic drive to regain lost weight and body fat by helping to minimize positive fat balance and subsequent storage of lipids while still maintaining the metabolic health benefits associated with weight reduction. Weight maintenance via HF resulted in the thermic effect of feeding more closely resembling basal conditions than the lowered TEF observed in the LF condition, although this difference did not reach statistical significance. Because TEF contributes to approximately 5e10% of TDEE, increasing TEF exists as an additional avenue for increasing caloric expenditure and weight maintenance. While research regarding post exercise TEF is largely variable, some evidence shows that exercise may raise TEF through elevated glucose uptake, thermogenic hormone levels, and sympathetic activity [27].

acute weight loss on various outcome variables by maintaining subjects' weight loss for three weeks prior to their entry into high and low flux states. There are several limitations to the present study. 1.) The small sample size could obviously be viewed as a weakness. However, this experimental investigation was undertaken as a pilot, proof-ofconcept study, and despite the small sample, statistically significant differences in key outcome variables were found between the LF and HF conditions and the RMR and hunger/fullness responses were almost uniformly in the same direction among the study participants. 2.) To minimize the length of the study and burden for the subjects, we did not control for the menstrual cycles of the female participants. The extent to which this may have influenced our findings is unknown, but it is likely that the influence of energy flux on our dependent variables is far greater than circulating sex hormones associated with the menstrual cycle. 3.) Finally, the duration of the low and high flux conditions were short. Thus it is unclear if weeks and months of high flux would produce the same beneficial effects on resting metabolism and perceptions of hunger and satiety as were seen in this short-term study.

4.2. Hunger and fullness ratings Weight reduction, while beneficial to metabolic health, may actually contribute to the energy gap by increasing the rate of nutrient clearance, making transient the markers of satiety, and moving individuals more rapidly into a post absorptive state [8,28,29]. The latter is associated with increased hunger and reduced satiety. However, a HF state may attenuate the increased hunger based on our finding that weight-reduced obese participants retrospectively reported decreased subjective ratings of hunger and increased ratings of fullness during four days of HF compared to four days of the LF sedentary state. The study participants obviously consumed more food in the HF compared to the LF condition, but this was necessary to maintain energy balance. Thus, we believe that their greater perceived fullness during HF was not the result of overfeeding, but rather related to greater amounts of food consumed required to match their higher energy expenditure. A reduction in hunger could play a major role in countering weight loss recidivism, as higher hunger ratings predict greater weight regain [30]. In support of the benefits of a high flux state on appetite regulation, Rosenkilde et al. [31] found that only participants in a high exercise group (60 min/day at 67% VO2 max) experienced higher fasting and postprandial rating of fullness.

5. Conclusions

4.3. Experimental strengths and limitations

HLP and RMF are co-first authors and contributed to the study design, data acquisition, data analyses, subject monitoring, and manuscript writing; KAW contributed to data acquisition and supervision of participant weight loss, LCL contributed to the supervision of flux states and data analyses, JWB contributed to data acquisition, KCY supervised the plasma assays, CB contributed to study design, experimental oversight, and manuscript writing, and CLM contributed to the design of the study, experimental oversight, data acquisition, data analyses, and manuscript writing.

There are a number of strengths of this study. 1.) To our knowledge this is the first investigation designed to examine the possible role of a high flux state on key aspects of the energy gap following clinically relevant weight loss. 2.) While the study participants remained free-living, energy intake and expenditure were carefully controlled during the low and high flux states and there is every indication that energy balance was achieved during these conditions in the face of large differences in energy expenditure. All meals, snacks and food modules were provided to the participants based on measured rather than estimated RMR to minimize errors in estimating energy intake requirements. Daily physical activity was monitored by a pedometer and all exercise bouts performed during the high flux condition were supervised with measures of indirect calorimetry to achieve the target net energy cost of 500 kcal per exercise bout. Thus, the higher resting metabolic rates, greater reported satiety and less hunger for high compared to low flux are not likely attributable to positive energy balance during high flux and/or negative energy balance during low flux. 3.) Additionally, we controlled for the possible confounding effects of

We found a short-term HF energy balanced state resulted in lower self-reported hunger, higher fullness, and higher RMR and resting fat oxidation in weight-reduced, weight-stable, obese individuals when compared to a LF energy balanced state. Our findings support the possible role of a daily high energy flux state in attenuating the increase in hunger and the decrease in energy expenditure that accompany diet-induced weight loss. However, the greater satiety and lower hunger were not evident the morning following the last day of exercise in the HF state, suggesting that such benefits resulting from the HF state are likely to be more acute rather than long-lasting. Given the pilot nature of this study and the relatively short period of time spent in the high and low flux states, future research is needed to replicate these findings in a larger sample and determine the effects, if any, of a long-term, chronic state of high energy flux on the energy gap and maintenance of lost body weight.

Author's contributions

Funding This study was funded by a grant from the National Institute for Food and Agriculture (NIFA) of the USDA via the Colorado Agricultural Experiment Station to CM and CB.

Conflict of interest The authors report no personal financial interest in the work.

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005

H.L. Paris et al. / Clinical Nutrition ESPEN xxx (2015) e1ee9

Acknowledgments The authors thank the participants for the time and effort devoted to this study and to Janelle Davis, Kyle Sevits, and Jessie Wilburn for their technical assistance. References [1] Kraschnewski JL, Boan J, Esposito J, Sherwood NE, Lehman EB, Kephart DK, et al. Long-term weight loss maintenance in the United States. Int J Obes (Lond) 2010;34:1644e54. [2] Crawford D, Jeffery RW, French SA. Can anyone successfully control their weight? Findings of a three year community-based study of men and women. Int J Obes Relat Metab Disord 2000;24:1107e10. [3] Sumithran P, Prendergast LA, Delbridge E, Purcell K, Shulkes A, Kriketos A, et al. Long-term persistence of hormonal adaptations to weight loss. N Engl J Med 2011;365:1597e604. [4] Gilbert JA, Drapeau V, Astrup A, Tremblay A. Relationship between dietinduced changes in body fat and appetite sensations in women. Appetite 2009;52:809e12. [5] Heilbronn LK, de Jonge L, Frisard MI, DeLany JP, Larson-Meyer DE, Rood J, et al. Pennington CALERIE Team: effect of 6-month calorie restriction on biomarkers of longevity, metabolic adaptation, and oxidative stress in overweight individuals: a randomized controlled trial. J Am Med Assoc 2006;295: 1539e48. [6] Melby CL, Schmidt WD, Corrigan D. Resting metabolic rate in weight-cycling collegiate wrestlers compared with physically active, noncycling control subjects. Am J Clin Nutr 1990;52:409e14. [7] Hill JO, Wyatt HR, Peters JC. Energy balance and obesity. Ciruclation 2012;126: 126e32. [8] Maclean PS, Bergouignan A, Cornier MA, Jackman MR. Biology's response to dieting: the impetus for weight regain. Am J Physiol Regul Integr Comp Physiol 2011;301:R581e600. [9] Mayer J, Purnima R, Mitra KP. Relation between caloric intake, body weight and physical work: studies in an industrial male population in West Bengal. Am J Clin Nutr 1956;4:169e75. [10] Stubbs RJ, Hughes DA, Johnstone AM, Horgan GW, King N, Blundell JE. A decrease in physical activity affects appetite, energy, and nutrient balance in lean men feeding ad libitum. Am J Clin Nutr 2004;79:62e9. [11] Bullough RC, Gillette CA, Harris MA, Melby CL. Interaction of acute changes in exercise energy expenditure and energy intake on resting metabolic rate. Am J Clin Nutr 1995;61:473e81. [12] Burke CM, Bullough RC, Melby CL. Resting metabolic rate and postprandial thermogenesis by level of aerobic fitness in young women. Eur J Clin Nutr 1993;47:575e85. [13] Bell C, Day DS, Jones PP, Christou DD, Petitt DS, Osterberg K, et al. High energy flux mediates the tonically augmented beta-adrenergic support of resting metabolic rate in habitually exercising older adults. J Clin Endocrinol Metab 2004;89:3573e8. [14] Weir JB. New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol 1949;109:1e9.

e9

[15] Sevits KJ, Melanson EL, Swibas T, Binns SE, Klochak AL, Lonac MC, et al. Total daily energy expenditure is increased following a single bout of sprint interval training. Physiol Rep 2013;1:e00131. [16] Peronnet F, Massicotte D. Table of nonprotein respiratory quotient: an update. Can J Sport Sci 1991;16:23e9. [17] Bell C, Seals DR, Monroe MB, Day DS, Shapiro LF, Johnson DG, et al. Tonic sympathetic support of metabolic rate is attenuated with age, sedentary lifestyle, and female sex in healthy adults. J Clin Endocrinol Metab 2001;86: 4440e4. [18] De Glisezinski I, Crampes F, Harant I, Berlan M, Hejnova J, Langin D, et al. Endurance training changes in lipolytic responsiveness of obese adipose tissue. Am J Physiol 1998;275:E951e6. [19] Ryan AS, Pratley RE, Elahi D, Goldberg AP. Resistive training increases fat-free mass and maintains RMR despite weight loss in postmenopausal women. J Appl Physiol 1995;79:818e23. [20] Bahr R, Ingnes I, Vaage O, Sejersted OM, Newsholme EA. Effect of duration of exercise on excess postexercise O2 consumption. J Appl Physiol 1987;62: 485e90. [21] Bielinski R, Schutz Y, Jequier E. Energy metabolism during the postexercise recovery in man. Am J Clin Nutr 1985;42:69e82. [22] Weststrate JA, Hautvast JG. The effects of short-term carbohydrate overfeeding and prior exercise on resting metabolic rate and diet-induced thermogenesis. Metabolism 1990;39(12):1232e9. [23] Maehlum S, Grandmontagne M, Newsholme EA, Sejersted OM. Magnitude and duration of excess postexercise oxygen consumption in healthy young subjects. Metabolism 1986;35:425e9. [24] Gore CJ, Withers RT. Effect of exercise intensity and duration on postexercise metabolism. J Appl Physiol 1990;68:2362e8. [25] Jackman MR, Steig A, Higgins JA, Johnson GC, Fleming-Elder BK, Bessesen DH, et al. Weight regain after sustained weight reduction is accompanied by suppressed oxidation of dietary fat and adipocyte hyperplasia. Am J Physiol Regul Integr Comp Physiol 2008;294:R1117e29. [26] Schutz Y. Concept of fat balance in human obesity revisited with particular reference to de novo lipogenesis. Int J Obes Relat Metab Disord 2004;28(Suppl. 4):S3e11. [27] Segal KR, Gutin B, Albu J, Pi-Sunyer FX. Thermic effects of food and exercise in lean and obese men of similar lean body mass. Am J Physiol 1987;252: E110e7. [28] Corpeleijn E, Mensink M, Kooi ME, Roekaerts PM, Saris WH, Blaak EE. Impaired skeletal muscle substrate oxidation in glucose-intolerant men improves after weight loss. Obesity 2008;16:1025e32. [29] McLaughlin T, Schweitzer P, Carter S, Yen CG, Lamendola C, Abbasi F, et al. Persistence of improvement in insulin sensitivity following a dietary weight loss programme. Diabetes Obes Metab 2008;10:1186e94. [30] Elfhag K, Rossner S. Who succeeds in maintaining weight loss? A conceptual review of factors associated with weight loss maintenance and weight regain. Obes Rev 2005;6:67e85. [31] Rosenkilde M, Reichkendler MH, Auerbach P, Torang S, Gram AS, Ploug T, et al. Appetite regulation in overweight, sedentary men after different amounts of endurance exercise: a randomized controlled trial. J Appl Physiol 2012;115: 1599e609.

Please cite this article in press as: Paris HL, et al., Increasing energy flux to decrease the biological drive toward weight regain after weight loss e A proof-of-concept pilot study, Clinical Nutrition ESPEN (2015), http://dx.doi.org/10.1016/j.clnesp.2015.11.005