78. Estimating Goal Weights in Adolescents with Anorexia Nervosa and Atypical Anorexia Nervosa: Comparison of the Median BMI and Historical BMI Percentile

For adolescents with anorexia nervosa (AN), weight restoration is a key goal of treatment. Thus, determining an accurate estimated body weight (EBW) is critical. In the past, researchers have used the median BMI (mBMI) to determine patients’ EBW. However, the wide range of normal BMI growth trajectories means that individual patients’ growth charts should be considered when determining EBW. For patients with atypical anorexia nervosa (AAN), using historical growth records may be even more important. The primary objective of this study was to compare within-person differences between EBWs based on mBMI and historic BMI (hBMI) percentile in patients with AN and AAN. The secondary goal was to determine whether EBWs calculated from either method better predicted improvements in two physiologic measures of recovery: heart rate (HR) and menstrual status. Participants included 209 adolescents aged 9-20 years old (M = 15.47, SD = 2.21) with either AN (N = 163, 78%) or AAN (N = 46, 22%) completing treatment in a partial hospital program (PHP). Data were collected retrospectively at baseline and discharge from PHP. Paired sample t tests and Person’s correlation were used for analysis. For adolescents with AN, EBW derived from the hBMI was higher (M = 53.60kg, SD = 7.41) than the mBMI EBW (M = 52.91kg SD = 7.32), p = .02. In adolescents with AAN, hBMI EBW (M = 58.35kg, SD = 7.89) was also higher than mBMI EBW (M = 52.81kg, SD = 6.20), p <.001. The magnitude of the difference between hBMI EBW and mBMI EBW was larger for AAN (5.54kg) than AN (0.69kg). When looking at physiologic measures, upon entering PHP, HR was correlated with %hBMI EBW (r = .22) and %mBMI EBW (r = .16), meaning that HR increased as %EBW increased. At the end of PHP, HR remained correlated with %hBMI EBW (r = .17), but not %mBMI EBW. In AAN HR was not correlated with either method of EBW calculation at either time point. In patients with AN at baseline, %EBW (both definitions) was higher for people who were having regular periods (89% hBMI, 91% mBMI) than those who had amenorrhea (81% hBMI, 82% mBMI). People with AAN having regular periods at baseline also had higher %EBWs (95% hBMI, 104% mBMI) than those with amenorrhea (89% hBMI, 97% mBMI). Patients with AAN and amenorrhea did not meet the definition for malnutrition (<90% EBW) based on their mBMI EBW (97% EBW), but did per hBMI EBW (89% EBW). Findings suggest that EBWs derived from hBMI percentiles are higher compared to EBWs derived from the mBMI in both AN and AAN, though the magnitude of the difference is larger for AAN. The physiologic indicators HR and amenorrhea are predicted by both definitions of %EBW in AN. In AAN, hBMI EBW was an indicator of amenorrhea while mBMI EBW was not associated with physiological signs. An EBW based on historical growth records is likely a better marker of recovery than one based on mBMI, particularly for AAN.