@article{39afd3d340e04fb08cc4ec552f940fcf,
title = "Evaluating the impact of COVID-19 on severity at admission and response to inpatient treatment for adult and adolescent patients with eating disorders",
abstract = "Objective: The COVID-19 pandemic has been associated with increased hospitalization rates and worsened symptom severity in patients with eating disorders (ED), but most studies focused exclusively on adolescents. Further, research evaluating the impact of COVID-19 on response to inpatient treatment for ED is limited. This study aimed to compare demographic characteristics, symptom severity at admission, and discharge outcomes for adult and adolescent inpatients with EDs admitted before and after onset of the COVID-19 pandemic. We expected the post-COVID cohort would report elevated symptomatology and poorer response to treatment compared to the pre-COVID cohort and that this effect would be amplified for adolescents. Method: Patients were consecutively hospitalized adults and adolescents treated in a specialized behavioral integrated inpatient-partial hospitalization program for eating disorders between March 2018 and March 2022 (N = 261). Results: The effect of COVID-19 on symptomatology was moderated by age group such that adolescents in the post-COVID cohort, but not adults, reported higher levels of eating disorder and depressive symptoms compared to the pre-COVID cohort. No group differences were observed for discharge outcomes (rate of weight gain, length of stay, or percent target weight). Discussion: Findings with respect to elevated symptomatology in adolescents but not adults may reflect the particularly negative impact of social isolation on adolescents. Future research is needed to assess the impact of COVID-19 on long-term treatment outcomes including relapse at 1-year, as well as the potential impact of COVID-19 on treatment availability for chronically ill adults or those with public insurance. Public Significance: Patients with eating disorders (ED) admitted to a specialty inpatient program after the start of the COVID-19 pandemic were younger and more likely to be male than those admitted pre-pandemic. Adolescents admitted post-COVID, but not adults, reported elevated ED and depressive symptoms compared to the pre-COVID cohort. Group differences were not observed for treatment response. Future research should evaluate the impact of COVID-19 on relapse risk in EDs.",
keywords = "COVID-19, adolescent, adult, eating disorders, inpatient treatment, treatment response",
author = "Schreyer, {Colleen C.} and Vanzhula, {Irina A.} and Guarda, {Angela S.}",
note = "Funding Information: Demographic information is presented in Table 1. Independent t-tests indicated that patients in the post-COVID group were significantly younger than those in the pre-COVID group (t = 2.39, p =.018, d =.34). Chi-square tests indicated that the post-COVID group included significantly more adolescents (χ2 [1, 208] = 4.96, p =.026; OR = 1.91) and more men (χ2 [1, 208] = 4.06, p =.044; OR = 2.61). There were no differences in diagnoses between COVID groups (χ2 [3, 208] = 1.14, p =.768). Note: *p <.05, **p <.01. Abbreviations: Adol, adolescent; ARFID, avoidant restrictive food intake disorder; AN, R, Anorexia nervosa restricting subtype; AN-B/P, anorexia nervosa, binge/purge subtype; BMI, body mass index; BN, bulimia nervosa; OSFED, other specified feeding and eating disorder; PH, Partial Hospital; TW, Target weight. The reported F value represents the main effect of COVID (pre vs. post). The reported value represents the main effect of age group. The reported value represents the interaction effect of COVID and age group. Pre-COVID cohort total length of stay included inpatient + partial hospital stay. Post-COVID cohort total length of stay included only inpatient length of stay because partial hospital was closed. Only patients treated on the weight gain protocol were included. Participants not discharged for clinical improvement included those who were discharged against medical advice, were discharged due to financial or insurance limitations, and those transferred for other medical or psychiatric treatment. Patients in the post-COVID cohort could not step down to the partial hospital program as it was closed due to concerns about infection control. There were not sufficient ns in the post-COVID cohort to conduct a chi-square analyses. Age group moderated the relationship between COVID group and severity of symptomatology at admission on several measures (EDEQ, EDRSQ, BDI), such that adolescents in the post-COVID group reported higher levels of symptomatology than pre-COVID adolescents. This effect was not observed for adults (Table 2 provides means and standard deviations by group). EDEQ Global scores (F[1, 153] = 4.43, p =.037, η2 =.03) were higher post-COVID compared to pre-COVID for adolescents, but not adults (Figure 2). Body image self-efficacy and normative eating self-efficacy were lower at admission post-COVID compared to pre-COVID for adolescents, but not adults (F(1, 153) = 4.63, p =.033, η2 =.03 and F(1, 153) = 4.88, p =.029, η2 =.03 respectively; Figures 3 and 4). Additionally, depressive symptoms at admission were higher in the post-COVID compared to pre-COVID group in adolescents only (F(1, 153) = 5.58, p =.019, η2 =.04; Figure 5). No main effects of COVID group or moderation effects of age group were found for levels of state and trait anxiety or percent target weight at admission (ps >.058). Note: *p <.05; **p <.01. Abbreviations: BDI, beck depression inventory; EDEQ, Eating Disorder Examination Questionnaire; EDRSQ, Eating Disorder Recovery Self-Efficacy Questionnaire; Inter., interaction; LOS, length of stay; STAI, State & Trait Anxiety Inventory; TW, target weight. F-value represents main effect of COVID (pre vs. post). F-value represents main effect of age group (adults vs. adolescents). F-value represents interaction effect of COVID and age group. Inpatient length of stay was longer post-COVID than pre-COVID, (F[1, 204] = 6.41, p =.012, η2 =.03) but there was no main effect by age group (p =.214) and moderation was not significant (p =.063). There were no main effects or moderation effects for total length of stay, rate of weight gain, or percent target weight at discharge (ps >.054, Table 2). Logistic regression showed that the likelihood of achieving target weight at discharge did not differ between pre- or post-COVID or by age group and moderation was not significant (Table 3). Note: R2 =.04. Demographic information is presented in Table 1. Independent t-tests indicated that patients in the post-COVID group were significantly younger than those in the pre-COVID group (t = 2.39, p =.018, d =.34). Chi-square tests indicated that the post-COVID group included significantly more adolescents (χ2 [1, 208] = 4.96, p =.026; OR = 1.91) and more men (χ2 [1, 208] = 4.06, p =.044; OR = 2.61). There were no differences in diagnoses between COVID groups (χ2 [3, 208] = 1.14, p =.768). Note: *p <.05, **p <.01. Abbreviations: Adol, adolescent; ARFID, avoidant restrictive food intake disorder; AN, R, Anorexia nervosa restricting subtype; AN-B/P, anorexia nervosa, binge/purge subtype; BMI, body mass index; BN, bulimia nervosa; OSFED, other specified feeding and eating disorder; PH, Partial Hospital; TW, Target weight. The reported F value represents the main effect of COVID (pre vs. post). The reported value represents the main effect of age group. The reported value represents the interaction effect of COVID and age group. Pre-COVID cohort total length of stay included inpatient + partial hospital stay. Post-COVID cohort total length of stay included only inpatient length of stay because partial hospital was closed. Only patients treated on the weight gain protocol were included. Participants not discharged for clinical improvement included those who were discharged against medical advice, were discharged due to financial or insurance limitations, and those transferred for other medical or psychiatric treatment. Patients in the post-COVID cohort could not step down to the partial hospital program as it was closed due to concerns about infection control. There were not sufficient ns in the post-COVID cohort to conduct a chi-square analyses. Age group moderated the relationship between COVID group and severity of symptomatology at admission on several measures (EDEQ, EDRSQ, BDI), such that adolescents in the post-COVID group reported higher levels of symptomatology than pre-COVID adolescents. This effect was not observed for adults (Table 2 provides means and standard deviations by group). EDEQ Global scores (F[1, 153] = 4.43, p =.037, η2 =.03) were higher post-COVID compared to pre-COVID for adolescents, but not adults (Figure 2). Body image self-efficacy and normative eating self-efficacy were lower at admission post-COVID compared to pre-COVID for adolescents, but not adults (F(1, 153) = 4.63, p =.033, η2 =.03 and F(1, 153) = 4.88, p =.029, η2 =.03 respectively; Figures 3 and 4). Additionally, depressive symptoms at admission were higher in the post-COVID compared to pre-COVID group in adolescents only (F(1, 153) = 5.58, p =.019, η2 =.04; Figure 5). No main effects of COVID group or moderation effects of age group were found for levels of state and trait anxiety or percent target weight at admission (ps >.058). Note: *p <.05; **p <.01. Abbreviations: BDI, beck depression inventory; EDEQ, Eating Disorder Examination Questionnaire; EDRSQ, Eating Disorder Recovery Self-Efficacy Questionnaire; Inter., interaction; LOS, length of stay; STAI, State & Trait Anxiety Inventory; TW, target weight. F-value represents main effect of COVID (pre vs. post). F-value represents main effect of age group (adults vs. adolescents). F-value represents interaction effect of COVID and age group. Inpatient length of stay was longer post-COVID than pre-COVID, (F[1, 204] = 6.41, p =.012, η2 =.03) but there was no main effect by age group (p =.214) and moderation was not significant (p =.063). There were no main effects or moderation effects for total length of stay, rate of weight gain, or percent target weight at discharge (ps >.054, Table 2). Logistic regression showed that the likelihood of achieving target weight at discharge did not differ between pre- or post-COVID or by age group and moderation was not significant (Table 3). Note: R2 =.04. Publisher Copyright: {\textcopyright} 2022 Wiley Periodicals LLC.",
year = "2023",
month = jan,
doi = "10.1002/eat.23855",
language = "English (US)",
volume = "56",
pages = "182--191",
journal = "International Journal of Eating Disorders",
issn = "0276-3478",
publisher = "John Wiley and Sons Inc.",
number = "1",
}