Within‐subject weight variability in bulimia nervosa: Correlates and consequences

Abstract

Objective The oscillations between binge eating, purging, and dieting in bulimia nervosa (BN) may produce substantial within‐subject weight variability. Although weight variability has been predictive of eating‐ and weight‐related variables in community samples, it has not been empirically examined in eating disorders. The current study examined cross‐sectional and prospective associations between weight variability and BN pathology. Method Four weights were collected over an average of 42.02 days, and weight variability was calculated as the root mean square error around each individual's weight trajectory regression line. Linear regressions were performed to examine the association between weight variability and eating disorder psychopathology, cross‐sectionally at baseline and prospectively at 6‐month follow‐up, adjusting for baseline BMI. Results Weight variability was cross‐sectionally associated with eating pathology, but these relationships became non‐significant after adjusting for BMI. However, at 6‐month follow‐up, greater baseline weight variability predicted increases in body dissatisfaction, shape and weight concerns, and global eating pathology, even after adjusting for baseline BMI. Discussion These findings demonstrate, for the first time, that within‐subject weight variability predicts greater eating disorder pathology over time in BN. The results add to evidence that weight history variables contribute to BN psychopathology above and beyond well‐documented psychological dysfunction in BN.

Early within-person weight gain and variability during inpatient treatment for anorexia nervosa: Age-dependent effects on treatment outcome

“…It is notable that weight variability was not associated with any treatment outcome when simultaneously accounting for average daily weight gain, especially as recent research highlighted the predictive value of weight variability on core eating disorder pathology in individuals with bulimia nervosa (Chen et al, 2021). This is also in contrast to Hartmann et al (2007), where weight variability was associated with treatment failure.…”

“…For example, Hartmann et al (2007) found a statistically significant negative association of within-person weight variability during week 3-4 of inpatient treatment on treatment success at discharge. Recently, a similar analysis in a treatment-seeking sample of individuals with bulimia nervosa found significant associations of within-person weight variability with core eating disorder psychopathology at 6-months follow-up (Chen et al, 2021). Yet, it is unknown whether similar associations of weight variability with core eating disorder pathology can be found in AN.…”

“…RMSE and average daily weight gain were then winsorized at the 5% and 95% percentiles to manage the impact of extreme outliers and to meet assumptions for the level-2 models. A similar winsorization procedure was used to estimate weight variability in individuals with obesity and bulimia nervosa (Benson et al, 2020;Chen et al, 2021).…”

Early within‐person weight gain and variability during inpatient treatment for anorexia nervosa: Age‐dependent effects on treatment outcome

“…This might have reduced variance in weight variability in our sample compared to individuals with bulimia nervosa. Second, Chen et al (2021) used a slightly different level-1 model that included fixed effect estimates for weight change, and only included BMI and weight variability into their level-2 analyses, but not linear weight gain. Thus, their weight variability estimate might be confounded by weight gain for patients that increased weight during the study.…”

“…(2007) found a statistically significant negative association of within‐person weight variability during week 3–4 of inpatient treatment on treatment success at discharge. Recently, a similar analysis in a treatment‐seeking sample of individuals with bulimia nervosa found significant associations of within‐person weight variability with core ED psychopathology at 6‐month follow‐up (Chen et al., 2021). Yet, it is unknown whether similar associations of weight variability with core ED pathology can be found in AN.…”