Longitudinal Development of Portion Size Estimation in Patients With Anorexia Nervosa Receiving Inpatient Treatment—Evidence From Laboratory and Real‐Life Data

Background Food intake is highly aversive for patients with anorexia nervosa (AN), and fear of weight gain is closely linked to altered eating behaviours, including reduced portion sizes. This study investigated how portion size estimation develops longitudinally in patients with AN during inpatient treatment. We hypothesised that patients with AN would overestimate portion sizes compared to healthy controls, with overestimation being most pronounced at treatment onset and decreasing over the course of treatment. Methods 29 patients with AN were assessed three times during real‐life mealtimes in an inpatient setting and in a laboratory‐based task: shortly after inpatient admission, in the middle of treatment, and shortly before discharge. 19 healthy control participants were assessed cross‐sectionally. The laboratory task was performed twice, asking participants one time to select a portion they would be ready to eat completely and a second time selecting an adequate portion for a hypothetical other person of their age and gender. Results Patients with AN overestimated portion sizes as compared to control participants at baseline, both in the laboratory and under real‐life circumstances. Patients chose larger portions for a hypothetical healthy person. As treatment progressed, patients chose larger portions they would be ready to eat in the laboratory task, while they continued to perceive the real‐life lunch portions as large, indicating that they did not adapt to usual portion sizes and continued to overestimate. Discussion Portion size estimation in patients with AN differed depending on whether the task was implemented under laboratory or real‐life conditions, supporting the need for method combinations. A better understanding of mechanisms underlying portion size estimation is needed to advance mealtime support and other treatment approaches for AN.