Disaggregating depression into stable trait-like and fluctuating state-like components provides insights into potential mechanisms underlying depression; predictors of these components can point to potentially tractable risk factors for depression onset. Using Latent-Trait-Occasion modelling, we extracted state and trait depression components from the Short Mood and Feelings Questionnaire obtained in young people ages 11 to 26 from a population-based study in England (ALSPAC; N=7364). The proportion of state variance ranged from 91.9% (age 11) to 36.8% (age 26), decreasing during adolescence and stabilizing from age 22. The proportion of trait variance increased correspondingly. Female sex correlated with trait variance and with state variance only during adolescence (ages 12-19). Life events, school enjoyment, maternal depression, and parent-child relationship quality predicted state variance during adolescence, while anxiety was the main predictor during emerging adulthood. The greater role of state factors in early adolescence suggests that depression is more modifiable at younger ages.
Neufeld et al. (Tue,) studied this question.