1 Research Center of CHU Sainte Justine, Université de Montréal 3175 chemin de la Côte-Sainte-Catherine, Montréal, Quebec H3T 1C5, Canada; Department of Mathematics and Statistics, Concordia University 1455 boulevard de Maisonneuve Ouest, Montréal, Quebec H3G 1M8, Canada.
2 Research Center of CHU Sainte Justine, Université de Montréal 3175 chemin de la Côte-Sainte-Catherine, Montréal, Quebec H3T 1C5, Canada.
3 Ingram School of Nursing, McGill University 680 rue Sherbrooke Ouest, Montréal, Quebec H3A 2M7, Canada.
4 Research Center of CHU Sainte Justine, Université de Montréal 3175 chemin de la Côte-Sainte-Catherine, Montréal, Quebec H3T 1C5, Canada; Department of Family Medicine, McGill University 5858 chemin de la Côte-des-Neiges, Montréal, Quebec H3S 1Z1, Canada.
5 Research Center of CHU Sainte Justine, Université de Montréal 3175 chemin de la Côte-Sainte-Catherine, Montréal, Quebec H3T 1C5, Canada; School of Kinesiology and Physical Activity Sciences, Université de Montréal 2100 boulevard Edouard Montpetit, Montréal, Quebec H3T 1J4, Canada.
6 Department of Mathematics and Statistics, Concordia University 1455 boulevard de Maisonneuve Ouest, Montréal, Quebec H3G 1M8, Canada; PERFORM Centre, Concordia University 7141 rue Sherbrooke Ouest, Montréal, Quebec H4B 1R6, Canada.
7 Research Center of CHU Sainte Justine, Université de Montréal 3175 chemin de la Côte-Sainte-Catherine, Montréal, Quebec H3T 1C5, Canada; Department of Pediatrics, Université de Montréal 3175 chemin de la Côte-Sainte-Catherine, Montréal, Quebec H3T 1C5, Canada.
8 Research Center of CHU Sainte Justine, Université de Montréal 3175 chemin de la Côte-Sainte-Catherine, Montréal, Quebec H3T 1C5, Canada; Department of Pediatrics, Université de Montréal 3175 chemin de la Côte-Sainte-Catherine, Montréal, Quebec H3T 1C5, Canada. Electronic address: melanie.henderson.hsj@gmail.com.

Introduction: Attrition in pediatric weight management programs is notoriously high. Greater understanding of its determinants is needed to inform retention strategies. We identified determinants of attrition in CIRCUIT, a healthy lifestyle intervention program for youth at risk of cardiovascular disease.

Methods: A one-arm intervention study of children aged 4-18 years who initiated the CIRCUIT program in the first five years of its existence (N = 403). We defined attrition as attending the baseline visit but ceasing attendance prior to the 1-year follow-up. Potential determinants of dropout included the child's age, sex, ethnicity, body mass index (BMI) z-score, family socio-demographic characteristics, and estimated driving time to the program, all measured at baseline. Associations were estimated bivariately, using chi-squared- and t-tests, and simultaneously in a multivariable logistic regression model.

Results: Of the 403 participants who started the program, 198 (49%) dropped out within 12 months of enrollment. Youth who dropped out were older (mean age 12.8y vs. 11.3y; p < 0.01), were less likely to live with both parents (62% vs. 71%; p = 0.05), and to have mothers who had completed high school (79% vs. 88%; p = 0.01). No group differences were observed for sex, ethnicity, baseline BMI z-score, fathers' education, or driving time to the program. In multivariate models, only older age at initiation of the intervention (OR: 1.2; CI: 1.1,1.3) and lower maternal education (OR: 2.0; CI: 1.0,3.8) were associated with dropout.

Conclusion: Improved tailoring of interventions to older pediatric participants and to families of lower maternal education may help reduce attrition in CIRCUIT and similar lifestyle intervention programs.