Heterogeneity in responses to a universal prevention program.

Because universal or primary prevention strategies often target heterogeneous populations, their effects may likewise be expected to be heterogeneous. We sought to explore the heterogeneity of outcomes of previously published results of a longitudinal Swedish study of a school-based socio-emotional learning program. By applying latent class regression analysis to two measures of well-being, we found three significantly different classes with different change trajectories that yielded different outcomes. We conclude that restricting outcome analyses of primary prevention programs to sample means may conceal important heterogeneity regarding individual outcomes.