Background
Depression is known to be common after traumatic brain injury (TBI) and associated with worse functional and psychosocial outcomes. However, there remains considerable uncertainty over the exact prevalence of the condition.
Aims
The aim of this study was to accurately assess the prevalence of post TBI depression and its changes over a period of one year. The associated demographic and injury features were also examined for possible association with risk of depression in the hope that those with higher susceptibility to depression may be identified.
Methods
The study population was a prospective cohort of TBI admissions to a teaching hospital emergency department over a two year period. Minimal exclusions were applied in order to recruit a representative TBI population who were then assessed in a specialist brain injury clinic at ten weeks and at one year post injury. Demographic and injury features were recorded to establish links with risk of depression which was recorded with a HADS (Hospital Anxiety and Depression Scale).
Results
Over a two year period, 774 individuals were recruited of whom 690 attended one year follow-up and 38 had died. Only 6% of the cohort was lost to follow-up after one year. The prevalence of depression at ten weeks was 56.3% [95% CI 52.8-59.8] and at one year 41.2% [95% CI 37.6-44.9]
A multivariable analysis identified the independent predictors of depression; at ten weeks these were TBI severity, abnormal CT scan, past psychiatric history, alcohol intoxication at the time of injury, female gender and non-white ethnicity. At one year the independent predictors were; abnormal CT scan, past psychiatric history, alcohol intoxication at the time of injury and female gender. TBI severity was no longer significant. Features such as injury aetiology, social isolation, age, length of stay and medical comorbidity were not associated with depression risk. All other outcome measures in the study, including psychosocial function, symptom severity and global overall outcome showed very high correlations with depression.
Discussion
The prevalence of depression is very high after TBI and associated with a number of injury features. While the prevalence drops over a year it still remains considerably elevated. There is also evidence that features related to the injury itself, such as TBI severity, become less significant in long term outcome compared to the initial period. It is possible that other psychosocial features such as personality and coping mechanisms are more important in determining long term outcome than injury features such as severity and aetiology. Some population features have been identified that may allow targeting of susceptible populations for intervention. The close correlations between all 4 outcome measures including depression suggest that they might be measuring a similar construct of emotional distress.
Future work will seek to reassess the prevalence of depression at three or five years as well as associated features, re-examining the relationship between various outcomes and use of interventions and treatments, especially in targeting at risk individuals.
