34. Differences in Medical Outcomes and Triggers for Disease Based on Pre-Morbid Weight Status in Adolescents With Anorexia Nervosa

Abstract

Youth with obesity who experience unsafe weight loss represent an under-diagnosed population in the eating disorder (ED) field. However, differences in medical outcomes and triggers for disease between those whose pre-morbid weights were classified as "overweight/obese" vs. “normal” are poorly understood, interfering with optimizing prevention and treatment. We aimed to examine the differences in medical complications and reported triggers for ED onset based on pre-morbid weight in adolescents with anorexia nervosa (AN). We performed a retrospective, electronic chart review of patients aged 9-19 admitted for complications of malnutrition due to AN or atypical AN (AAN) at a large, urban pediatric hospital from January 2015 to February 2020. Comparison groups were divided by their pre-morbid growth trajectories: “normal” [<85th percentile body mass index (BMI)] vs. “overweight/obese” (≥ 85th percentile). Demographic and outcome data (e.g., weight, vital signs, triggers for ED behaviors) were obtained. Reported triggers for disease were acquired, coded, and categorized into main themes via qualitative thematic analysis and then treated as binary variables. Binary logistic and linear regression analyses were conducted. 150 patients were identified. Mean (SD) age was 14.1 (2.3) years. 86% were female, 80% white, and 92% non-Hispanic/Latinx. 23% had pre-morbid weight in the overweight/obese category. We found no difference in age or length of stay between those with overweight/obesity vs. not. However, those with overweight/obesity had lost a significantly higher mean percentage of total body weight (%TBW) (22.3% vs. 17.9% [p=0.04]) and at a higher rate (11.4 vs. 5.7 pounds/month [p=0.18]) compared to those without overweight/obesity. Previous weight status was not a significant predictor for the presence of bradycardia, hypotension, or orthostasis. However, higher %TBW lost was significantly associated with the presence of bradycardia, when controlling for previous weight status and rate of weight loss (p<0.001). Furthermore, those with pre-morbid overweight/obesity had 7.6x the odds of reporting positive reinforcement about weight loss as a trigger for their ED (p=0.004) and 3.9x the odds of reporting weight-related teasing (p=0.003) compared to those without pre-morbid overweight/obesity when controlling for age and length of symptoms. Patients with AN/AAN who had pre-morbid overweight/obesity have significantly higher %TBW lost and rates of weight loss than those who were normal-weighted, which can place them at risk of dangerous complications, such as bradycardia. These patients are also more likely to report weight-related teasing and positive reinforcement for weight loss as triggers for disease. Understanding these data can help optimize prevention, assessment, and treatment of EDs in this under-diagnosed population.