UC San Diego

Overweight and obesity impact two-thirds of adults in the United States. The food cue responsivity model posits that individuals respond both physiologically and psychologically to food cues, which leads to increased cravings, physiological changes, food consumption, and ultimately weight gain. Implicit approach behaviors occur in response to rewarding food cues. Thus, understanding the impact that automatic action tendencies have on biases to approach or avoid food cues may offer important mechanisms to target in eating behaviors.

Participants completed demographic and self-report questionnaires, a computerized approach-avoidance task (AAT-Food), a free-access Eating in the Absence of Hunger (EAH) paradigm, and anthropometric measurements. The AAT-Food consisted of three phases (baseline assessment, training, and post-training assessment) with high and low-calorie food picture stimuli. Participants were randomized to either a train towards high-calorie foods (90% towards, 10% away) or train away from high-calorie foods condition (90% away, 10% towards). Correlation analyses, paired samples t-tests, and linear regression models were used to assess the relationship of baseline approach bias to individual difference characteristics and EAH. Independent samples t-tests and linear regression analyses evaluated the impact of Body Mass Index (BMI) on approach biases. Mixed model ANOVAs and paired samples t-tests were used to assess differences in bias scores and EAH based on training condition.

Seventy-five participants (65.30% female; 28.00% Hispanic; 19.83  2.70 years; BMI: 23.91  3.97 kg/m2) completed all portions of this one-session laboratory study. Baseline approach-avoidance bias scores did not differ based on stimuli and were unrelated to both BMI and EAH (p’s > .05). None of the self-reported individual difference characteristics (reward-based eating drive, psychological responsivity to food, eating disorder cognitions, executive functioning, behavioral activation and inhibition, impulsivity) were related to approach bias to food stimuli (p’s > .05). Furthermore, training condition did not significantly impact approach bias or EAH (p’s > .05). Additional analyses revealed significant associations between individual difference characteristics and EAH, after adjusting for BMI and sex (p’s < .05).

These results represent an important step in understanding approach-avoidance biases to food stimuli. Additional research should be conducted prior to adapting computerized AAT-Food programs into interventions targeting clinical outcomes.