Over 70% of Americans are either overweight or obese, putting them at risk for many chronic diseases including diabetes. Exercise is commonly used as a weight loss and weight loss maintenance strategy. However, exercise-induced weight loss is often much less than expected as individuals compensate for a large portion of the energy expended through exercise, resisting maintenance of the negative energy balance needed for weight loss. Our prior research, in agreement with others, point to increases in energy intake as the primary compensatory response when exercising for weight loss; however, mechanisms promoting this behavior have yet to be fully elucidated. With obesity and diabetes prevalence continually rising, innovative research is needed to identify novel mechanisms promoting energy compensation with exercise. The long-term goal of this proposal is to reduce the incidence and improve the outcomes of obesity-related diseases by developing interventions that will attenuate compensation for the energy expended through exercise and thus improve initial weight loss and weight loss maintenance. The present proposal will take the necessary first steps towards our long-term goal by identifying novel mechanisms promoting energy intake when exercising for weight loss.

One's reinforcing value of food, attentional bias and inhibitory control for food cues play an important role in feeding behaviors, independent of hunger. These behaviors are largely a product of the central dopamine reward system, which is also in play with exercise behavior. This provides mechanistic support for our central hypothesis, that exercise evokes increases in food reinforcement, attentional bias, and lowers inhibitory control for food cues to promote greater energy intake in effort to maintain energy homeostasis. The rationale for this project is by elucidating the mechanisms mediating energy compensation, future interventions can be designed that attenuate this response to improve the utility of exercise as a weight loss intervention to prevent and manage T2DM. The overall objective of the current proposal is to demonstrate an acute bout of exercise alters food reinforcement, attentional bias and inhibitory control for food cues. Upon completion, we will have a greater understanding of the mechanisms underpinning compensatory increases in energy intake when exercising. These findings will pave the way for future clinical trials testing this hypothesis in the context of a long-term exercise intervention. This contribution is significant, as the identification of novel mechanisms influencing energy compensation with exercise is needed to provide strong support for the development of novel, evidence-based interventions to attenuate this compensatory response to exercise, improving its efficacy for weight control and chronic disease management.

Over 70% of adults in the US are overweight or obese putting them at risk for a multitude of comorbidities including cardiovascular disease, hypertension, diabetes, dyslipidemia, metabolic syndrome, and certain cancers. Greater mortality rates with obesity have estimates of life expectancy reduced by 6-7 years with up to 300,000 deaths attributed to obesity annually in the US. Obesity treatment has therefore emerged as a prime focus of health care.

Increasing physical activity and energy expenditure is a universal recommendation for weight control, with exercise becoming the most prevalent, economical, and health-promoting treatment option for obesity.

Unfortunately, weight loss from an exercise program is often disappointing, and greater amounts of exercise may not always promote additional weight loss. An increasing body of evidence indicates individuals' concerted efforts to achieve weight loss through exercise are stymied by a coordinated set of compensatory responses working to maintain energy balance, conserving energy to ensure the energetic needs of vital organs are met. The primary compensatory response working to maintain energy homeostasis and thus opposing weight loss from an exercise-induced energy deficit is an increased drive for greater energy intake. This is often independent of hunger, as single bouts of exercise typically do not result in changes in appetite, food intake, or appetite- regulating hormones, while long-term exercise can actually improve the satiety response to a meal. The motivational aspects of eating, termed the reinforcing value of food, is a strong predictor of body weight and energy intake independent of hunger, although not previously studied as an important factor in energy compensation with exercise. Food reinforcement is developed by instigating a greater attentional processing towards food, termed attentional bias. These behavioral constructs are controlled by the central dopamine system, which mediates the rewarding aspects of reinforcing behaviors (such as eating) and is a more powerful driver of behavior than affective responses. Overeating occurs when the motivational wanting of food exceeds the capacity of inhibitory control for eating. As such, individuals with lower inhibitory control are more susceptible to the reinforcing aspects of food, and more likely to over-consume highly reinforcing, energy dense foods. With additional evidence eating and exercise behaviors share similar neuro-behavioral pathways possibly explaining why exercise can promote feeding in animal models, it would be of great interest to determine the influence exercise may have on food reinforcement, attentional bias, and inhibitory control for food cues in humans.

Our long-term goal is to reduce the incidence and improve the outcomes of obesity by developing interventions to attenuate compensation for the energy expended through exercise thereby improving exercise as a weight loss and weight loss maintenance strategy. The overall objectives of the current proposal are to 1) evaluate whether an acute bout of exercise alters food reinforcement, attentional bias and inhibitory control for food cues, and 2) assess a potential sex effect in this response. Our central hypothesis is that exercise training will evoke greater food reinforcement and attentional bias for food cues while reducing inhibitory control as a mechanism to promote greater energy intake in effort to maintain energy homeostasis. Our hypothesis has been formulated based on our preliminary data demonstrating a large (~1,000 kcal per week) compensatory response to a 12-week exercise intervention that could not be explained by changes in resting energy expenditure, physical activity outside the exercise intervention, or hormonal mediators of appetite. This compensatory response was observed across two different studies using various exercise doses, indicating a remarkable consistency in this compensatory response. Previous research in animal models have demonstrated exercise facilitates eating, with evidence of shared neuro-behavioral pathways between eating and exercise reward that may explain this connection. These behavioral constructs influencing energy intake have not been considered as mechanisms for energy compensation and likely contribute to the unexplained compensatory response we have previously observed. The rationale for this project is that with greater knowledge of the mechanisms behind energy compensation, future interventions may be designed that attenuate this response to improve exercise's utility as a weight loss strategy. To achieve the overall objectives, we will pursue the following specific aim and exploratory aim in obese, sedentary individuals, the population in greatest need for effective weight loss treatment: Specific Aim: Demonstrate changes in food reinforcement, attentional bias and inhibitory control for food cues after an acute bout of exercise. Based on our preliminary data and previous literature demonstrating exercise may increase food reward, our working hypothesis is an acute bout of exercise will cause food to be more reinforcing, increase attentional bias for food cues, and decrease inhibitory control for food cues.

Exploratory Aim: We will explore whether sex moderates these responses. Others have demonstrated sex differences in food reinforcement, although the role sex plays in this response to exercise is unknown.

At the completion of the proposed research, the expected outcomes are to have identified novel mechanisms that contribute to energy compensation during an exercise program. These results will have a sustained impact as they will provide a strong evidence-based proof of principle for development of novel interventions that attenuate the compensatory response to exercise, improving its efficacy for weight control.

Significance Importance of the Problem to Be Addressed. Greater than 70% of US adults are either obese (BMI > 30) or overweight (BMI 25-29.9). Poor dietary habits and physical inactivity are the top causes for this epidemic. Over 90% of US adults fail to meet physical activity recommendations when assessed objectively, while over 95% fail to meet recommendations for solid fat and added sugar intake [5]. The impact of obesity on society is realized in its associated comorbidities, including cardiovascular disease, hypertension, diabetes, dyslipidemia, metabolic syndrome, gallstones, osteoarthritis, sleep apnea, premature death, and many cancers such as breast, liver, and colon. The economic impact of obesity is startling, with annual health care expenditures between $92-117 billion, making obesity among the most expensive health problems in the US [8].

The continually escalating obesity epidemic suggests the majority of individuals are unable to meet their weight loss needs. Evolutionarily conserved mechanisms that defend against a negative energy balance are, in part, to blame [9]. These mechanisms include metabolic and behavioral responses to an energy deficit that function to conserve energy by either reducing energy expenditure or promoting energy intake. Metabolic adaptations reduce resting and non-resting energy expenditure to account for approximately 120 kcal/day through mechanisms related to thyroid function, skeletal muscle glucose oxidation, and sympathetic nervous system activity. Behavioral responses to an energy deficit, on the other hand, can account for far more than 120 kcal/day and include increases in energy intake and decreases in physical activity. Many agree the most impactful compensatory response working to resist an exercise-induced energy deficit is increases in energy intake, as the rate of energy intake far exceeds the rate of energy expenditure. The mechanisms underlying this compensatory response are poorly understood.

Exercise is the most common weight loss strategy with a prevalence rate of 65% among those attempting to decrease body weight. Indeed, most people have the capacity to increase their energy expenditure to promote negative energy balance whereas, depending on an individual's aerobic fitness, exercise intensities can be maintained for prolonged periods at 2- 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. This has prompted the American College of Sports Medicine to issue separate recommendations to either maintain health or support weight loss through exercise. Exercise is also one of the primary components of the Diabetes Prevention Program (DPP) and a key aspect to the classic Look AHEAD trial primarily due to the role exercise is believed to play in weight loss and weight loss maintenance. However, reductions in body weight from exercise training are often less than expected due to the compensatory responses working to maintain energy balance noted above.

Our preliminary data, in agreement with others, point to increases in energy intake being the primary compensatory response to an energy deficit induced by exercise. Other purported compensatory responses include decreases in resting energy expenditure (REE), improved skeletal muscle work efficiency, and decreased habitual physical activity outside the exercise program [9]. We have twice demonstrated non-significant changes in REE, total energy expenditure, and habitual physical activity outside the exercise intervention despite a compensatory response of approximately 1,000 kcal/week. This supports the premise that increases in energy intake are largely responsible for this compensatory response to exercise, although the mechanisms behind this response remain unknown. Identifying the mechanisms for this compensatory response is the necessary first step in developing interventions that aim to attenuate the compensatory increases in energy intake that accompany an exercise program. Such interventions would vastly improve exercise as a weight control strategy.

An often-studied mechanism responsible for greater energy intake with exercise is increased hunger, assessed by hormonal mediators of appetite, lab-based food intake, and hunger/satiety scales [28-32]. In particular, elevated concentrations of the orexigenic hormone ghrelin and reductions in satiety-inducing signals such as peptide YY (PYY), glucagon-like peptide one (GLP-1), pancreatic polypeptide (PP), cholecystokinin (CCK), and leptin may all be altered to promote greater energy intake when participating in exercise training. However, single bouts of exercise typically do not result in the expected compensatory increases in appetite. Conversely, greater perceptions of fullness persist 24 hours after exercise, and long-term exercise can actually improve the satiety response to a meal. Thus, other mechanisms are likely involved, necessitating novel and innovative research to better understand the mechanisms behind this compensatory response to exercise.

Central dopamine (DA) systems are essential for experiencing reward (i.e. DA hypothesis of reward), where adaptations in this system were originally proposed to explain drug addiction. Interestingly, both exercise and eating behaviors are largely driven by this central DA system as genetic polymorphisms that control DA uptake and transport have been linked to exercise and eating reward. This offers the possibility for a neuro-behavioral cross-talk between these behaviors, explaining why individuals who exercise for weight loss often have distorted portion control, seek rewards from exercising in the form of food, and derive greater pleasure from high-fat, high-sugar, energy-dense food independent of hunger. There is evidence of such direct cross-talk between eating and exercise reward in rats, where wheel running that produces a low-level dopaminergic response facilitates eating. In humans, evidence points to exercise promoting a greater wanting to eat among individuals who overcompensate (eat more energy than they expend) due to exercise potentially having a sensitizing action that enhances food reward to increase energy intake. 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. The possibility for exercise to increase the rewarding aspects of food and thus energy intake is therefore a probable scenario, although this has not been fully evaluated in humans.

The rewarding aspects of eating are important volitional behavioral responses influencing energy intake, assessed by defining one's reinforcing value of food. 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. Food reinforcement is a more robust predictor of food intake than the hedonic value (i.e., liking) and is a strong predictor of body weight. Food deprivation resulting in energy deficits increases food reinforcement, and thus possible that an exercise-induced energy deficit will increase food reinforcement as a mechanism to promote feeding behavior and maintain energy balance. No study has been conducted characterizing and testing this behavioral response to exercise in humans.

Increasing the reinforcing value of a behavior is accomplished through "incentive sensitization". Incentive sensitization is a behavioral process where repeated exposures increase the salience of a stimulus within the environment to produce an attentional bias. This produces neuroadaptations that increase the motivating or reinforcing value of the behavior. Attentional bias towards food cues is therefore an additionally relevant and important construct to consider when elucidating the motivational implications of eating. Increased attentional bias for food cues predicts the tendency to overeat and weight status, with obese individuals having greater attentional bias towards food cues, especially for energy-dense palatable foods. Overeating occurs when motivational wanting of food exceeds the capacity for inhibitory control for eating, as such, a heightened awareness/attentional bias of food or food reinforcement may lead to overeating in those with reduced inhibitory control. This puts individuals with poor inhibitory control at a greater susceptibility to food's rewarding effects, increasing energy intake and promoting weight gain. This association has not been assessed alongside exercise, and exercise-mediated increases in food reinforcement and attentional bias may follow different patterns than previously observed cross-sectional data, which has used primarily sedentary individuals. Based on our most recent preliminary data demonstrating 38/44 participants demonstrated some form of positive compensation after a 12-week exercise intervention, exercise may either lower inhibitory control as an additional mechanism to promote energy intake, or food reinforcement and attentional bias for food cues may be increased greater than what can be mitigated by one's inhibitory control.

Significance of the Proposed Research Contribution The expected contribution is an enhanced understanding of the mechanisms underpinning energy compensation to exercise. This contribution will be significant as this will provide the necessary information to develop future trials and treatments to facilitate more successful and sustained weight loss over time from an exercise program, reducing rates of obesity and related complications. These results are intended to both improve scientific knowledge and directly inform clinical practice as it pertains to obesity treatment and prevention. Currently, behavioral compensatory responses are not commonly considered a modifiable barrier to exercise-induced weight loss, possibly due to the unknown nature of the mechanisms behind them or how they may be modified. Novel mechanisms must be identified as it appears research on traditional mediators of energy intake (hunger/satiety) do not provide a definitive conclusion. Identifying the role food reinforcement, attentional bias and inhibitory control for food cues play in the exercise-induced increases in energy intake will lead to future research targeting these mechanisms to improve the effectiveness of exercise for weight control.

Innovation The status quo, as it pertains to obesity treatment, is ineffective with the continually escalating obesity epidemic serving as evidence. Exercise programs for obesity treatment often do not consider energy compensation as an integral aspect of weight loss. Our preliminary data and others have confirmed the compensatory response to exercise and that greater energy intake is the largest source of compensation. However, little is known of the mechanisms that mediate this compensatory response.

The proposed research is innovative because it challenges current obesity research paradigms by focusing on the underlying mechanisms of energy compensation with exercise. Specifically, food reinforcement, attentional bias and inhibitory control for food cues are volitional behavioral responses that will be evaluated as important mediators in the compensatory response to exercise. These constructs are strong predictors of weight status and energy intake but have yet to be evaluated in the context of exercise. With evidence that exercise can influence the neuro-behavioral processes underlying food reinforcement, attentional bias and inhibitory control for food cues, the present investigation's hypotheses are not only highly novel but also highly probable. The investigative team of Dr. Kyle Flack (PI, an experienced nutritionist, exercise and behavioral physiologist) coupled with Dr. Craig R. Rush (Co-I, an experienced behavioral scientist with expertise in attentional bias and inhibitory control) is an additionally novel aspect and uniquely suited to carry out this proposal. Successful completion will open new horizons for obesity treatment, including novel interventions that focus on attenuating exercise-induced alterations in food reinforcement, attentional bias and inhibitory control for food cues.

Approach Preliminary Data. Our preliminary studies on energy compensation with exercise have indicated non-significant changes in resting energy expenditure (kcal/24 hrs) and non-exercise physical activity despite an approximate 1,000 kcal per week compensatory response regardless of exercise energy expenditure (1500 vs. 3000 kcal/week) or exercise frequency (2 vs 6 sessions/week). This demonstrates the overall magnitude of energy compensation sedentary individuals exhibit when engaging in a 12-week exercise intervention that did not include dietary restriction. Importantly, the lack of pre-post changes in REE and non-exercise physical activity indicate changes in energy intake are the primarily responsible for this 1,000 kcal/week compensatory response. Additionally, changes in the hormonal mediators of appetite and satiety were not related to this compensatory response. For this reason, research focused on identifying novel mechanisms promoting increases in energy intake when exercising for weight loss is needed to develop interventions that work to mitigate this response and thus improve weight loss outcomes. The reinforcing value of food, attentional bias, and inhibitory control for food cues are behavioral constructs known to play a large role in eating behaviors and weight status. As we have noted, there is strong rational for investigating how these behavioral constructs are uniquely influenced by exercise, centered on the shared neuro-behavioral pathways within the DA reward system that may link exercise and eating behavior. This has been observed in rodents, where exercise that causes a dopaminergic response promotes eating, but has yet to be fully investigated in humans.

Research Design. This pilot study will use a counterbalanced cross-over design to assess the acute effects of a single bout of exercise and a single bout of sedentary activity on food reinforcement, attentional bias and inhibitory control for food cues among overweight to obese, sedentary adults (n=60, 30 males, 30 females). Differences in food reinforcement, attentional bias and inhibitory control for food cues will be determined between conditions (bout of exercise vs bout of sedentary activity), time (before exercise vs. after) and sex (male vs. female). This will allow us to determine how exercise influences the behavioral mediators of energy intake compared to both a non-exercise control condition and a pre-exercise condition.

Setting. All study activities will be conducted at the University of Kentucky (UK) Center for Clinical and Translational Science (CCTS), Laboratory of Human Behavioral Pharmacology (LHBP), and the Nutrition Assessment Laboratory (NAL), all located in Lexington Kentucky, on or surrounding the UK campus.

Participants, recruitment, and screening. Recruitment will include 72 community-dwelling overweight to obese (BMI 25-45 kg/m2) men (n=36) and women (n=36) aged 18-45 years who are not currently engaged in exercise or weight loss activities, nor have lost or gained over 5% of their current bodyweight in the previous 12 months. Eligible participants will also be (1) free of any cardiac, pulmonary, or metabolic health conditions, (2) able to safely engage in exercise, (3) not taking any medications or dietary supplements which may influence energy expenditure or intake and (4) have not been diagnosed with an eating disorder, clinical depression, or an anxiety disorder and (5) engage in less than 150 minutes of moderate to vigorous physical activity per week (assessed via accelerometry at baseline). Female participants must additionally be premenopausal and not pregnant or nursing. Ability to safely participate in exercise will be determined using a health history and Physical Activity Readiness Questionnaire (PAR-Q+) and reviewed by the study physician. Participants will be recruited from the greater Lexington, KY area using a combination of print and online media methods. Interested individuals will be screened for basic inclusion criteria (age, activity level, BMI) and, if deemed initially eligible, invited for a screening visit to further assess eligibility. Based on our previous 12-week exercise intervention performed at the University of Kentucky we will expect an overall retention rate of 85%, leaving an n=60 completing both assessment visits.

Procedures Visit 1 (week 1). The Study Coordinator will meet with each participant for a 30-minute orientation session. At this time, participants will have the study explained, questions answered, complete the necessary screening questionnaires (health history, PAR-Q+), complete an Accu-Check blood glucose test, and, if deemed eligible and interested, sign the informed consent. Upon enrollment, participants will also complete a food frequency questionnaire (DHQ-3), be assessed for body composition (BodPod), and have waist circumference measured. Participants will also be provided an ActiGraph accelerometer to objectively assess physical activity for the next seven days and scheduled for Visit 2 at least eight days later. Participants will be instructed not to purposely change dietary behaviors while enrolled in the study (i.e. start to diet).

Visits 2 and 3 (week 2 and 3). On the day of Visit 2, participants will record the specific foods, quantities, and time of day they consumed them prior to their testing visit. Participants will be instructed to eat the same foods at the same time of day prior to Visit 3 as they did prior to Visit 2. Visit 2 and 3 will be scheduled at the same time of day on the same day of the week where participants will be able to practice "usual" dietary behaviors, i.e. not on a holiday or a special occasion. Upon arriving to the LHBP exactly four hours post-prandial in the afternoon, participants will complete assessments for food reinforcement, attentional bias and inhibitory control for food cues before and twice after an acute bout of either exercise or sedentary activity, with one visit featuring exercise and the other visit featuring a sedentary activity. The order of the visits will be counterbalanced. Visual analog scales (VAS) will be performed for hunger and exercise enjoyment prior to each set of assessments. For female participants, these visits will occur while in the follicular phase, which will be assessed by a mensural cycle questionnaire. Participants will wear the accelerometer to record physical activity during the washout period between assessment visits and be instructed to remain sedentary.

Acute Bout of Exercise. Participants will complete assessments of food reinforcement, attentional bias and inhibitory control before and after an acute bout of exercise during Visit 2 or visit 3 (depending on counterbalanced order). Participants will be provided a Polar A-300 heart rate monitor with smart-cal™ technology that factors the individual's sex, age, weight, activity level, and heart rate to produce an estimate of total energy in kilocalories (kcal) expended during the exercise session, with kcal expended displayed in real- time. Participants will engage in exercise on an elliptical ergometer (Octane Fitness ZR8) for the amount of time it takes them to expend 500 kcal under supervision of a researcher in the laboratory. Participants will exercise at a self-selected intensity of at least 65% heart rate reserve (HRR), an intensity commonly prescribed for sedentary individuals beginning an exercise program. Based on previous studies conducted in our lab, 500 kcal of energy expenditure is attainable for even the most sedentary individual with exercise sessions lasting between 30 and 70 minutes depending on body mass and intensity of exercise. Prescribing the exercise bout on energy expenditure rather than duration or intensity will standardize the amount of total work being done across all participants. While both the intensity and duration of the exercise could be controlled, due to differences in body mass of the participants, either the exercise time or total work would still vary. An energy expenditure of 500 kcal is specifically used to ensure the exercise bout is of sufficient volume (intensity * time) to produce weight loss if included in a long-term program and to elicit a dopaminergic response. Participants will be permitted to drink 0.5 l of water during and after the exercise bout. Participants will take a brief 15-minute rest after completing the exercise session before repeating assessments for food reinforcement, attentional bias and inhibitory control for food cues (post exercise assessment 1). Participants will repeat these assessments 60 minutes after cession of the exercise and sedentary activity bout (post exercise assessment 2).

Acute Bout of Sedentary Activity. Participants will complete assessments of food reinforcement, attentional bias and inhibitory control before and after an acute bout of sedentary activity (TV). Participants will watch TV for 60 minutes, choosing from a variety of television sitcoms (DVD) void of commercial advertisements that may feature food. Participants will have access to 0.5 l of drinking water during and after the test bout of TV viewing.

Statistical Analysis Linear mixed models will be fit with exercise and time as random effects to account for repeated measures and sex as a fixed effect. Response variables will be transformed if appropriate. With binary explanatory variables, an unstructured covariance matrix for the random effects will be assumed, but alternatives will be considered if needed. VAS scale ratings of hunger/satiety and exercise enjoyment will be assessed and included for adjustment as necessary. Power analysis: Because of our three response variables, we will use a Bonferroni corrected type I error rate of 0.05/3 = 0.017 for our regression models. In a linear regression model with n = 60, 30 males and 30 females, counterbalanced to start with and without exercise and then crossing over to the other treatment, we will have 80% power to detect models with R squared of 20.4% or more (nQuery 8.5.1). Our mixed models will improve model fit and decrease variance estimates compared with linear regression, thus increasing power or decreasing R squared type measures that can be detected.

Exploratory Aim: By including sex as a fixed effect in our linear mixed model approach, our study is designed and powered to determine the role of sex may play in the hypothesized exercise-induced changes to the behavioral mediators of energy intake assessed. Previous work in animals have identified a sex effect in eating reward/reinforcement where females have greater neurobehavioral drive to consume food. It is unknown, however, if sex may play a role in the response to exercise, warranting this exploration.

Inclusion Criteria:

• overweight to obese (BMI 25-45 kg/m2)
• not currently engaged in exercise or weight loss activities
• free of any cardiac, pulmonary, or metabolic health conditions
• able to safely engage in exercise
• female participants must be premenopausal and not pregnant or nursing.

Exclusion Criteria:

• Lost or gained over 5% of their current bodyweight in the previous 12 months.
• taking any medications or dietary supplements which may influence energy expenditure or intake
• have not been diagnosed with an eating disorder, clinical depression, or an anxiety disorder
• engage in less than 150 minutes of moderate to vigorous physical activity per week (assessed via accelerometry at baseline)

Collected one month after baseline assessments completed for each subject.

We originally desired to recruit 30 males and 30 females (60 total). However we were only able to recruit 32 individuals due to time constraints with sponsor. This limits out power in detecting sex differences, but does not the overall findings of the study.

Masking true propose of study- assessing eating behaviors.