Back to gut feelings. Latent class analysis of daily perceived health variations identifies later responses to antidepressant treatment

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Major depressive disorder (MDD) is a prevalent and disabling condition, yet predicting individual response to antidepressants remains difficult. Traditional symptom-based assessments, typically conducted weeks apart, may miss early and dynamic changes. Daily self-reported measures of wellbeing dimensions, particularly via simple tools such as visual analog scales (VAS), could provide ecologically valid and patient-centered markers of early treatment response. This study aimed to identify distinct early trajectories of self-rated health during antidepressant initiation and examine their associations with baseline clinical characteristics and outcomes at 6–8 weeks. In this multicentre naturalistic study, 1536 outpatients with MDD initiating agomelatine monotherapy were recruited across 388 community psychiatry centers. The primary outcome was daily self-rated health during the first 14 days, assessed using the EuroQol VAS. Latent class analysis identified distinct trajectories of self-rated health evolution, with model selection based on Akaike and Bayesian information criteria (AIC/BIC), entropy, and clinical interpretability. A five-class solution was retained, including persistently high or low trajectories, stable intermediate patterns, and one subgroup showing marked early improvement despite low baseline values. Trajectories were associated with baseline clinical severity and different symptomatic and functional outcomes at 6–8 weeks. Patients with rapid early improvement had the highest remission rates, whereas those with severe baseline symptoms and limited early improvement had the poorest outcomes. These findings support the clinical utility of early wellbeing trajectories for personalized monitoring during antidepressant treatment.