Dropout in supervised small-group exercise programs: a 7-year retrospective cohort study

Dropout to exercise programs remains a relevant challenge for physical education and public health professionals. Supervised small-group training is a promising strategy to enhance long-term adherence; however, evidence on its effectiveness in preventing dropout is still limited.

To investigate participation duration and predictors of dropout in a supervised small-group training program using retrospective data from a personalized studio to analyze the mean length of participation and factors associated with dropout among adults.

A retrospective cohort study was conducted with 587 participants enrolled between 2018 and 2024 at a training studio in São Paulo, Brazil. Variables analyzed included sex, physical activity level, body mass index (BMI), and reasons for initiating exercise. Kaplan-Meier survival analysis, Cox regression, and a decision-tree model (CHAID) were applied.

Mean participation was 13.8 ± 7.8 months, with a 12-month survival rate of 51.7%. Factors significantly associated with shorter participation were: male sex (HR = 0.614; p < 0.001), BMI ≥ 25 kg/m² (HR = 1.220; p < 0.05), weight-loss as the main goal (HR = 0.639; p < 0.001). The decision tree identified male sex, weight-loss goals, and insufficient physical activity as the strongest predictors of dropout (p < 0.05).

Dropout from supervised small-group training programs followed a progressive pattern, with nearly half of participants leaving within the first year. Male sex, overweight status, weight-loss goals complaints were the strongest predictors of early discontinuation, and the decision-tree analysis revealed that the combination of these factors further increased dropout risk.