Latent Class Analysis of Adolescent Psychosocial Functioning and Course of Major Depression

Examining the influence of episode number and age of onset on individual depressive symptoms across episodes of major depression

Major depression is characterized by an episodic course with symptom manifestations differing across episodes. Previous work has found that symptom presentation differs across age. However, studies of symptom presentation have largely focused on symptoms in individual episodes, requiring further investigation of longitudinal symptom change. This study explored the impact of the initial age of onset, the number of episodes, and age of onset of each episode on individual depressive symptoms, while accounting for episode severity. We used data from the Oregon Adolescent Depression Project (N = 629) examining participants with at least one major depressive episode, assessed by diagnostic interview, across a 15-year follow-up. Multilevel logistic regression models revealed that approximately 20-25% of the main effects were significant and some were qualified by cross-level interactions. However, only a few associations remained robust after correcting for multiple comparisons. Specifically, older initial age of onset was associated with fatigue, younger initial age of onset for the first episode was associated with suicidal ideation, and a lower episode number was associated with weight loss. These findings highlight potential initial age of onset and scar effects influencing symptom manifestation, but require replication.

Data-driven subgrouping of youths with depression reveals that resilience is associated with higher physical functioning despite high symptom burden in the Texas Youth Depression and Suicide Research Network (TX-YDSRN)

BACKGROUND

The Patient-Reported Outcomes Measurement Information System (PROMIS) measure, which assesses past week status of seven domains (physical function mobility, anxiety, depressive symptoms, fatigue, peer relationships, pain interference, and pain intensity), represents a new paradigm using patient-reported outcomes. We used a data-driven approach with PROMIS to identify subgroups of youths receiving depression treatment.

METHODS

Youths (n = 721) enrolled in the Texas Youth Depression and Suicide Research Network who completed the PROMIS were analyzed. Latent class analyses (LCAs) identified subgroups and compared their baseline clinical/sociodemographic features.

RESULTS

Compared to population norms, our sample had worse than average physical function, anxiety, depression, fatigue, and pain interference. Using LCA, four subgroups were identified: 1) lower symptom severity and higher physical functioning (14.6 %); 2) higher symptom burden, higher pain interference/intensity, and lower physical functioning (52.7 %); 3) higher symptom burden, higher pain interference/intensity, but with higher physical functioning (9.2 %); and 4) higher symptom burden, but lower physical functioning and pain interference/intensity (23.6 %). Group 3 demonstrated higher resilience than Group 2. In contrast, Group 2 had higher anxiety than Group 4.

LIMITATIONS

Individuals may have different symptom profiles due to the observational nature of the study. Replication of these subgroups may be difficult, as future samples may differ in these characteristics. Further work may demonstrate the stability of these groups.

CONCLUSIONS

A data-driven analysis identified a small but significant subgroup with high physical functioning despite high symptom burden and pain, and this group reported higher resilience. Resilience-enhancing interventions may help improve functional outcomes in depressed youth.