182 - Association of the social deprivation index to response to sertraline, but not citalopram or escitalopram, in pediatric patients

Poster Number: 182
Publication Number: 182.413

IDA AKA, Monroe Carell Jr. Children's Hospital at Vanderbilt, NASHVILLE, TN, United States; Sara Van Driest, Vanderbilt University Medical Center, Nashville, TN, United States; KATELYN ROSSOW, University of Louisville School of Medicine, Louisville, KY, United States

Background: Socioeconomic status (SES) has been associated with adverse health outcomes and response to some medications. However, the relationships between SES and outcomes of antidepressant use in children have not been assessed.

Objective: The purpose of this study was to determine if SES variables are associated with therapeutic response to commonly used antidepressants (sertraline, citalopram, and escitalopram) in children and adolescents.

Design/Methods: This retrospective cohort study used datasets previously derived from BioVU, a de-identified electronic health record and DNA repository. Inclusion criteria were exposure to the target drugs at age ≤18 years, adequate data to determine response, and available SES variables using US census tract data based on 3-digit zip codes [the composite social deprivation index (SDI) and 6 individual SES variables: median income, % assisted income, % high school education, % poverty level, % vacant housing, and % with no health insurance]. Clinical, demographic, exposure, and response data were manually extracted, blinded to SES. Associations between response, SES, genetic, and clinical variables were determined using Kruskal-Wallis test, Fisher’s exact test and multivariable logistic regression.

Results: For sertraline, 160 individuals (median age of 14.7 (IQR 11.3-16.4) years, 69% female) were included. 126 (79%) had a therapeutic response to sertraline. Responders had significantly higher SES for 4/6 variables (all but % vacant housing and % with no health insurance) as well as SDI [0.3 (IQR 0.2-0.4) vs 0.4 (0.3-0.4), p=0.004, for responders and non-responders, respectively). In multivariable analysis adjusting for age, race, sex, CYP2C19 metabolizer status, maximum dose, and concomitant medications, SDI was associated with response (OR 0.01, 95%CI 0.00-0.63). Among 318 individuals treated with (es)citalopram (median age 13.7, (10.4-16.8) years, 60% female), 241 had a therapeutic response. Older age (median age 13.8 vs. 13.1 years in responders vs. non-responders, p=0.01), but none of the SES variables were associated with response to (es)citalopram.Conclusion(s): SDI and other SES variables were associated to response to sertraline, but not (es)citalopram. Knowledge of differences in community risk could help improve drug outcomes for children and adolescents. Differences in the effects of SES on drug outcomes may be due to the different indications for these drugs, differences in out-of-pocket costs, or other factors still to be determined.