Attentional bias for food cues as a mechanism driving exercise-induced hyperphagia

Background: Exercise often prompts compensatory eating behaviors that serve to return the body to energy balance. Although maintaining energy homeostasis is an important, evolutionarily conserved trait, this may counteract purposeful weight loss efforts. Mechanisms independent of hunger are likely promoting post-exercise feeding behaviors. Such mechanisms may lie within central reward centers, where motivational drive to engage in reinforcing behaviors originate. An early process in developing behavioral reinforcement is fostering an attentional bias towards rewarding stimuli. We have demonstrated a large dose (500 kcal) of aerobic exercise can increase attentional processing for food cues (termed attentional bias), while a smaller dose (15-minute walk) can decrease attentional bias. Objective: Evaluate the effect of exercise mode (resistance vs. aerobic) and weight status on pre-post exercise changes in attentional bias towards food cues in healthy adults. Hypothesis: A low dose of resistance exercise (RE) will promote attentional bias towards food cues to a greater extent than a duration and intensity matched bout of aerobic exercise (AE), with individuals with obesity observing the greatest difference. Methods: 47 adults aged 30.33 + 12.3 years, with a BMI of 26.92 + 5.4 (mean + SD), completed a duration and intensity-matched bout of AE and RE in a crossover manner, at least one day apart. RE consisted of 7 machine exercises completed three times in a circuit manner. AE consisted of inclined walking on a treadmill to elicit the same heart rate and duration as the RE session. Attentional bias for food cues was assessed pre and post exercise using eye tracking technology (Tobbii Pro Fusion) and a validated dot-probe task. Data: Exercise intensity was measured via heart rate monitor (Polar), energy expenditure (kcal) and substrate oxidation (RER) were assessed via indirect calorimetry (AEI technologies and VOCO for AE and RE, respectively). Main outcome measure for attentional bias was the percentage of time fixated on food cues during the dot-probe task. Results: Duration was just over 20 minutes for both exercise sessions. Participants expended 151.4 + 7.8 kcal during RE and 168.4 +6.78 kcal during AE (mean + SE), which were not different. RER values averaged 1.03 + 0.02 for RE and 0.90 + 0.01 for AE (P< 0.01). This difference in RER was driven by Individuals without obesity, as those with obesity (BMI>30) did not demonstrate differences in RER between treatments (0.99 vs 0.92 for RE and AE, respectively). Pre to post-exercise changes in % of time fixated on food cues was not different between exercise conditions when considering the entire sample (+3.40% vs -1.56%, for RE vs AE, respectively, SE of difference: + 5.15, P=0.36), but different among participants with obesity: +6.97% vs -13.81% for RE vs. AE, respectively, + 10.21, P=0.05. Conclusions: As hypothesized, our 20 minute bout of AE did not promote attentional bias for food cues. Among all participants, RE did not differentially influence attentional bias. Among individuals with obesity, exercise mode significantly influenced changes in attentional bias towards food cues, with RE increasing it to a similar magnitude as the large dose of AE used in our prior work, while the low dose of AE promoted a greater than 2-fold decrease. These findings may indicate a short bout of AE may be less likely to instigate compensatory eating behaviors among individuals with obesity and thus be a preferred method for weight control. This abstract was presented at the American Physiology Summit 2026 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.