Assessing the Prevalence of and Risk Factors for Disordered Eating Attitudes and Behaviors in Adolescents With Inflammatory Bowel Disease

pediatrics, body image dissatisfaction, disordered eating

Introduction

Adolescents living with inflammatory bowel disease (IBD) have a higher incidence of internalizing disorders, lower quality of life, and impaired social functioning relative to healthy peers.¹ Medical and psychosocial risk factors for eating pathology and body image concerns also affect youth with IBD, including medical hyperfocus on weight, medically related food restriction, body image uncertainty, and societal praise of thinness despite health issues.² Treatment-related risk factors may increase vulnerability to eating pathology and poor body image, including adverse effects of steroids and psychosocial sequelae related to surgery/medical devices (eg, body dissatisfaction after surgery).

Adults with IBD have reported high rates of body image dissatisfaction (BID)³; however, few studies have examined prevalence rates among adolescents with IBD. In a review of BID and disordered eating behaviors in pediatric chronic illnesses, 53% and 16% of patients with cystic fibrosis were found to have disordered eating attitudes and disordered eating behaviors, respectively, and patients with type 1 diabetes were more than twice as likely as healthy peers to engage in disordered eating behaviors.² A recent study of BID in newly diagnosed pediatric IBD found that older age, worse patient-reported disease status, and greater depressive symptoms were associated with higher levels of BID.⁴ The understanding of BID concerns in pediatric IBD beyond initial diagnosis is unknown and remains essential to care planning for medications and surgery. In addition to BID, young people with IBD may be at risk for disordered eating behaviors, though there is little known about this prevalence.

Adolescents with IBD who experience BID and disordered eating are vulnerable for poor outcomes given their need to maintain healthy nutritional status for growth and disease management. The current study assessed the prevalence of disordered eating attitudes and behaviors in adolescents with IBD and factors associated with disordered eating. Determining the prevalence in adolescents with IBD may help clinicians understand the extent of this problem. Discerning the contributions of the unique psychosocial and medical risk factors to this population may provide a framework for developing screening measures and interventions.

METHODS

This multisite, cross-sectional study collected data from sites in the ImproveCareNow (ICN) quality improvement collaborative. The ICN is a learning health care system dedicated to improving care in pediatric IBD.⁵ Four clinics across the country recruited participants from outpatient IBD clinics. The study was available to any adolescent within ICN via official ICN social media posts, which allowed for geographic diversity. Data were collected via the REDCap platform. Participants were recruited by study flyers that were posted or handed out in clinics and by ICN social media posts.

Inclusion criteria included (1) a formal diagnosis of IBD for ≥6 months; (2) the ability to speak, read, and understand English; (3) currently receiving IBD care at any ICN site; and (4) access to the internet (eg, through a computer). There were no exclusionary criteria.

Demographic data were collected, including age, ethnicity, and sex. Medical variables including IBD subtype, treatment history (ie, previous treatments), and surgical history were extracted from the ICN patient registry and matched to participants. The survey assessed self-reported disordered eating behaviors and attitudes, disease-specific health-related quality of life, and symptoms of anxiety and depression.

Disordered Eating Behaviors and Attitudes

The Eating Disorder Examination Questionnaire (EDE-Q) 16.0 is a 28-item self-report questionnaire that assesses disordered eating behaviors and attitudes on a 7-point Likert-type scale based on the past 28 days.⁶ The EDE-Q 16.0 is considered the gold standard for assessing eating pathology. The measure has 4 subscales (restraint, eating concern, shape concern, and weight concern) and generates subscale scores and a total score. Internal consistency for the current study was excellent at α = 0.93.

Health-Related Quality of Life

The IMPACT-III includes 35 questions on a 5-point Likert scale and is a health-related quality-of-life measure designed and validated for a pediatric IBD population.⁷ Internal consistency for the current study was excellent at α = 0.98.

Anxiety and Depression

The Revised Child Anxiety and Depression Scale (RCADS-25) is a 25-item measure of internalizing concerns in children and adolescents with strong psychometric properties, such as high internal consistency, test-retest reliability, and high convergent validity with validated internalizing measures for youth.⁸ The RCADS-25 includes 3 subscales: anxiety, depression, and total anxiety and depression. Internal consistency for the total score in the current study was excellent at α = 0.93.

RESULTS

Of the adolescents who completed the REDCap survey (n = 80), 62.5% (n = 50) participants met the inclusion criteria. Participants were 68% female (n = 34) and 28% male (n = 14), with 4% identifying as transgender or other. Participants reported a mean age of 17.42 years (SD = 2.91). Eighty-eight percent of participants were White (n = 44). Of the participants, 16% (n = 8) were in middle school, 34% (n = 17) were in high school, and 50% (n = 25) were in college. Medical characteristics and descriptives of the sample can be found in Table 1. Body mass index (BMI) data from the most recent visit were available for 25 participants. The BMI ranged from 17.27 to 31.2, with a mean BMI of 23.1 and a median BMI of 21.39. The BMI cutoffs revealed that 4% (n = 1) of these participants were categorized as underweight, 64% (n = 16) were categorized as normal, 24% (n = 6) were categorized as overweight, and 4% (n = 1) were categorized as obese.

TABLE 1.

Descriptive Data for the Study Cohort

	n	Mean	SD
Age (y)	50	17.42	2.91
Years diagnosed	49	5.6	4.0
Number of previous treatments	50	2.78	1.09
Number of current treatments	50	1.46	0.71
BMI (last visit)	25	23.1	3.93
EDE-Q total score	47	1.79	1.56
IMPACT total score	47	2.73	1.11
RCADS Anxiety subscale	47	9.81	6.08
RCADS Depression subscale	48	9.56	6.38
RCADS total score	46	19.63	12.08

Scores on the RCADS Anxiety scale indicated that 12.77% (n = 6) were borderline clinically significant or greater and that 4.26% (n = 2) were in the clinically significant range. Scores on the RCADS Depression scale indicated that 6.25% (n = 3) were borderline clinically significant or greater and that 2.08% (n = 1) were in the clinically significant range. When the scores were assessed for participants with comorbid anxiety and depression scores at the subclinical range or higher, 4% (n = 2) of participants were identified. Scores on the RCADS Total scale revealed that 8.7% (n = 4) were borderline clinically significant or greater and that 4.35% (n = 2) were clinically significant.

A stepwise regression analysis assessed the factors associated with disordered eating attitudes and behaviors, measured via EDE-Q total scores. Regression results can be found in Table 2. Findings revealed 2 factors explaining 50% of the variance in EDE-Q total scores (R² = 0.5; F[2, 41] = 19.49; P < 0.01). Results indicated that the RCADS Anxiety subscale (β = 0.68; P < 0.01) and years since diagnosis (β = 0.31; P = 0.01) were significantly associated with EDE-Q total scores.

TABLE 2.

Stepwise Regression Model Results of Factors Associated With EDE-Q Total Scores

Variable	B	SE B	β
Anxiety	0.18	0.031	0.68
Years since diagnosis	0.12	0.04	0.31
R2	—	0.50	—
Adjusted R2	—	0.47	—
F	—	19.49**	—

**P < 0.01.

Using the established guidelines for the clinical significance of scores for the EDE-Q to determine prevalence,⁹ 10% of participants reported clinically significant restraint, 6% reported clinically significant eating concerns, 30% reported clinically significant shape concerns, 22% reported clinically significant weight concerns, and 12% reported clinically significant total EDE-Q scores.

Discussion

The finding that one-third of participants reported body image concerns and more than 10% reported disordered eating behaviors suggests that adolescents with IBD may experience higher rates of disordered eating thoughts related to shape (ie, specific shape of a body part) and weight (ie, scale reading) as compared to disordered eating behaviors, consistent with the disordered eating literature.¹⁰ This finding suggests that mental health clinicians may consider initially targeting negative cognitions related to disordered eating and then tailor appropriate coping strategies to reduce or reframe thoughts and associated behaviors. For medical providers, important supportive strategies may include refraining from commenting on weight (eg, “You’re so skinny!”), discussing weight while showing visual growth curves, and being mindful about hospital roommate pairing during hospitalizations based on food intake status (eg, not rooming an adolescent who is NPO with an adolescent who can eat without restriction).² Although BMI data from the last medical visit were available for roughly half of the study cohort and identified the majority as having normal BMI, limited information on the weight and BMI of all participants limited the authors’ ability to interpret psychosocial findings in the context of weight status.

Findings revealed that greater anxiety and greater years since diagnosis were significantly associated with disordered eating behaviors. These findings differ from an emerging understanding of BID in newly diagnosed IBD, which has found that higher depressive symptoms are associated with BID; the current study findings are likely reflective of the distinct experiences of ongoing chronic illness care, such as increased anxiety about health and the ongoing impact of IBD on life, that patients with newly diagnosed IBD have not experienced. Anxiety and depression likely have differential impacts on an individual with IBD across care, and an initial increase in worries and changes in mood may be normative to adjusting to, and living with, a chronic disease. Currently, no screening measures exist for disordered eating behaviors in a pediatric IBD population, complicating the ability of medical providers to screen and triage adolescents. There also exists limited integration of psychological providers in pediatric gastrointestinal and IBD clinics, presenting an additional barrier to screening. Taken together, the development of brief screening tools, integrating psychological providers into gastrointestinal clinics, referring patients to pediatric psychology, and a workflow of screening including triaging to higher levels of care would aid in the timely and accurate identification of and intervention for adolescents with IBD and disordered eating behaviors.

Limitations for the current study include the cross-sectional design, small sample size, low statistical power, self-report of psychosocial and medical variables by adolescent participants, limited BMI data from the ICN registry, potential for anxiety and depression to be confounded by IBD symptoms, and lack of a control group. The cross-sectional methodology of the study provides an important, albeit limited, insight into body image and disordered eating concerns in adolescents with IBD. The current study does not address how the variables of interest may change over time along with the disease course. For example, this study did not assess whether BID was present before an IBD diagnosis and, if so, whether the symptoms or behaviors worsened, improved, or remained stable. Although a future longitudinal study would add to the clinical research question, baseline prevalence is important to establish cross-sectionally. The use of self-reported psychosocial and medical data is another limitation in that adolescents may respond according to social desirability biases and/or may inaccurately report medical information. Other medical variables, such as pubertal status, were unknown and would be important for future work to assess in the context of BID and disordered eating. Future research should gather the weight and BMI of all participants to better interpret psychosocial findings in the context of weight status. The anxiety and depression items on the RCADS-25 may be confounded by IBD symptomatology, such as fatigue or changes in appetite. Finally, the use of a convenience sample may have oversampled patients interested in body image or disordered eating and therefore limits the extent to which results are representative of adolescents with IBD.

Conclusions

In summary, the nontrivial rates of body image concerns and disordered eating behaviors among these adolescents with IBD suggest that clinicians and researchers should continue to evaluate for these concerns to best care for this population.