Adolescent and Adult Transitions From Major Depressive Disorder to Bipolar Disorder

Abstract

Importance: Bipolar disorder (BD) often first appears in adolescence after onset of major depressive disorder (MDD), but diagnosis and treatment are commonly delayed. This delay is a concern because untreated BD is associated with adverse long-term outcomes, a more recurrent disease course and difficult-to-treat illness, and suicide attempts and deaths.

Objective: To examine the association of age at MDD onset with early transition to BD and the subsequent use of psychiatric inpatient services as a severity indicator.

Design, setting, and participants: This retrospective cohort study analyzed comprehensive data sourced from the Stockholm MDD Cohort data from 1997 to 2018, which encompass both outpatient and inpatient care. Individuals with an initial MDD episode from January 1, 2010, to December 31, 2013, who transitioned to BD by December 31, 2018, were identified. Data were analyzed between September 5 and December 28, 2023.

Exposures: Post MDD assessments included a depression severity index, comorbidities, psychotherapy, psychotropic drugs, and electroconvulsive therapy.

Main outcomes and measures: The main outcome was the transition from MDD to BD, dichotomized as occurring early (within 3 years of MDD onset) or late (3 years after MDD onset). Secondary outcomes encompassed the use of psychiatric inpatient services post transition and patterns of medication usage. A robust propensity score matching framework was used to estimate outcomes.

Results: The final balanced cohort included 228 individuals, with an equal distribution between adults (n = 114; mean [SD] age, 24.5 [6.3] years; 96 female [84.2%]; 20 experiencing an early transition to BD [17.5%]) and youths (n = 114; mean [SD] age, 15.3 [1.6] years; 93 female [81.6%]; 8 experiencing an early transition to BD [7.0%]). Youths were substantially less likely to transition early (odds ratio, 0.42; 95% CI, 0.20-0.88; P = .02), despite having more outpatient visits (mean [SD] visits per month, 1.21 [1.07] vs 0.97 [0.98] for adults; P = .01). Both groups experienced substantially reduced inpatient care following a BD diagnosis, concurring with a marked decline in antidepressant use without increased lithium use.

Conclusions and relevance: These findings suggest that adolescents may experience delayed BD progression and that diagnosis substantially reduced inpatient care in all age groups, which coincided with a reduction in the use of antidepressants. These findings may inform pharmacologic strategies in patients with first-episode MDD at risk for BD.

Figure 1.. Comparison of Time Spent in Psychiatric Inpatient Treatment Before and After Diagnosis of Bipolar Disorder (BD) in Youths With a Prior First Episode of Major Depressive Disorder (MDD) in 2010-2013

Only individuals recorded as presenting to inpatient care either before or after their BD diagnosis are included in this illustration; however, individuals with 0 inpatient care days were still accounted for in the nonparametric tests, the results of which are also depicted here. The data points represent individual youths, with the lines emphasizing the within-individual change from before to after diagnosis This analysis of a matched and stratified sample of 114 youths and 114 adults showed a significant reduction in the percent duration of time for the entire period spent in inpatient care following a BD diagnosis, with the median of nonzero changes being reductions of 0.493% (IQR, 0.140%-2.119%; P = .002).

Figure 2.. Comparison of Time Spent in Psychiatric Inpatient Treatment Before and After Diagnosis of Bipolar Disorder (BD) in Adults With a Prior First Episode of Major Depressive Disorder (MDD) in 2010-2013

Only individuals recorded as presenting to inpatient care either before or after BD diagnosis are included in this illustration; however, individuals with 0 inpatient care days were still accounted for in the nonparametric tests, the results of which are also depicted here. The data points represent individual adults, with the lines emphasizing the within-individual change from before to after diagnosis. This analysis of a matched and stratified sample of 114 youths and 114 adults showed a significant reduction in the percent duration of time for the entire period spent in inpatient care following a BD diagnosis, with the median of nonzero changes being reductions of 0.444% (IQR, 0.081%-0.934%; P = .04).