EXCESSIVE WEIGHT GAIN IN PEDIATRIC INFLAMMATORY BOWEL DISEASE PATIENTS ON ANTI-TNF THERAPY

Abstract

Abstract Background Limited studies in patients with inflammatory bowel disease (IBD) suggest that anti-TNF therapy may be associated with excessive weight gain. The objective of this study was to compare the change in BMI z-score in children with Crohn Disease (CD) on anti-TNF therapy (CD exposed) to those not on anti-TNF therapy (CD unexposed) over time. We also sought to determine if anti-TNF exposure is a risk factor for excessive weight gain and obesity. Methods This was a retrospective cohort study utilizing the RISK database, a longitudinal inception cohort of pediatric patients with CD. Multivariable logistic, linear random mixed effects, and multilevel mixed effects generalized linear regression models were used to evaluate the association of anti-TNF therapy with excessive weight gain, change in BMI z-score, and obesity, respectively. Excessive weight gain was defined as patients with a normal baseline BMI z-score who became overweight/obese or had a delta BMI z-score >1 at the end of follow-up, underweight patients who became overweight/obese, and overweight patients who became obese or had a delta BMI z-score >0.5 at the end of follow-up. Results 625 (61%) exposed and 404 (39%) unexposed patients were included. There was no difference in age at diagnosis or sex between groups, however the CD exposed group had lower BMI z-score [-0.9 (IQR -1.8, 0.0) vs -0.5 (IQR -1.4, 0.3)] as well as higher disease activity index [52.5 (IQR 35.0, 72.5) vs 40.0 (IQR 25, 57.5)] and C-reactive protein (CRP) [4.5 (IQR 1.4, 13.0) vs 2.7 (IQR 0.7, 9.0)] at baseline. 28% of the group exposed to anti-TNF therapy met criteria for excessive weight gain. Exposed patients were at 1.43 (CI 1.06, 1.93) greater odds of excessive weight gain as compared to the unexposed group, controlling for baseline BMI z-score and sex. BMI z-score increased with anti-TNF exposure, particularly in the first year of follow-up and in children who were underweight at baseline, controlling for baseline BMI z-score, sex, CRP, and steroid exposure (Table 1). Male sex was associated with higher BMI z-score, while CRP and steroid exposure were inversely related to BMI z-score. Exposed patients had increased adjusted odds of obesity and morbid obesity: OR 1.72 (CI 1.34, 2.20) and OR 5.72 (CI 2.77, 11.33), respectively (Table 2). Male sex and baseline BMI z-score also increased the odds of obesity and morbid obesity. Conclusion CD patients with anti-TNF exposure have a significant increase in BMI z-score over time as compared to patients unexposed to anti-TNF therapy. While for most patients this increase in BMI z-score is restorative, a subset of patients gain excessive weight or become obese. Anti-TNF therapy may be a risk-factor for this phenomenon. BMI z-score at diagnosis and male sex may also be risk factors. Future studies evaluating changes in adiposity and energy expenditure should be pursued. Note: Interaction term between anti-TNF and baseline BMI z-score. Interaction term between anti-TNF and time. Interaction term between anti-TNF and quadratic time. Note: Table for risk factors for morbid obesity in CD patients not included due to text constraints but with similar findings.