Long-Term Risk for Mood and Anxiety Disorders After Pediatric Traumatic Brain Injury: A Population-Based, Birth Cohort Analysis

TL;DR

In this population-based birth cohort, isolated TBI before 10 years of age was not significantly associated with subsequent mood or anxiety disorder by 25 years of age overall. However, females with childhood TBI had a significantly increased risk of a subsequently diagnosed mood disorder compared with age-matched female referents.

Key findings

• Among 562 children with TBI before 10 years of age and 1124 matched referents, 115 persons with TBI and 215 referents had at least 1 mood or anxiety disorder; median age at first diagnosis was 18.1 years for TBI cases and 18.4 years for referents.
• Overall, childhood TBI was not significantly associated with anxiety disorder (adjusted hazard ratio [aHR], 1.01 [95% CI, 0.71-1.43]; P = .97) or mood disorder (aHR, 1.16 [95% CI, 0.92-1.47]; P = .21).
• Females with TBI had a significantly increased risk of mood disorder versus female referents (aHR, 1.40 [95% CI, 1.04-1.89]; P = .03), but males did not (aHR, 0.91 [95% CI, 0.61-1.36]; P = .65); the sex interaction was not significant (P = .12). Anxiety disorder was not significantly associated when stratified by sex (aHR, 1.12 for females and 0.84 for males; P = .55 for interaction).
• Injury severity analyses showed no significant associations for mood disorder with Definite or Probable TBI (aHR, 1.19 [95% CI, 0.75-1.89]; P = .44) or Possible TBI (aHR, 1.18 [95% CI, 0.89-1.55]; P = .25), and no significant anxiety associations by severity.

Methods

• Population-based birth cohort in Olmsted County, Minnesota: 5518 persons born January 1, 1976, through December 31, 1982, after exclusions. TBI before 10 years of age was confirmed by manual record review and classified with the Mayo Classification System: 8 Definite (1.4%), 140 Probable (24.9%), and 414 Possible (73.7%). Each of 562 TBI cases was matched to 2 age- and sex-matched referents from the same cohort without TBI at the index date (1124 referents).
• Outcomes were depressive, bipolar, and anxiety disorders through age 25, identified from clinical diagnostic codes and manual record review; bipolar disorder was combined with depressive disorders into mood disorder because of few events (5 persons with TBI, all males, and 8 referents). Reliability abstraction of 30 random records showed 87% agreement and κ = 73% (95% CI, 49%-98%).
• Cox proportional hazards models with age as the time scale compared risk between TBI cases and referents, adjusted for child’s race, mother’s age, and mother’s education level; analyses were also stratified by sex and injury severity, with Kaplan-Meier cumulative incidence estimates.

Limitations

• The paper states that Olmsted County residents’ age, sex, and ethnic characteristics are similar to Minnesota and the Upper Midwest, so results may not generalize to more ethnically or socioeconomically diverse US populations.
• The paper states that mood and anxiety disorders were identified by manual record review without direct clinical assessment, which could result in inaccurate classification and did not account for changes in diagnostic practices over time; however, the same blinded approach was used for all persons, and interrater reliability was adequate.
• The paper states that few Definite TBI cases limited power to assess moderate to severe TBI, that TBI after 10 years of age was not abstracted so the impact of repeat TBI is unknown, and that other outcomes such as substance abuse and behavioral/conduct disorders need further study. The text also makes evident that the sex-specific female mood finding should be interpreted cautiously because the formal sex interaction test was not significant and mood disorder sample sizes were small.