Fann et al. (2004) used the Group Health Cooperative (GHC) of Puget Sound (450,000 members) in a prospective cohort study of TBI. All GHC members with a new diagnosis of TBI in 1993 who had been GHC members for at least a year were examined, evaluated for severity and matched with randomly selected GHC non-TBI members by sex, age, and enrollment date. Psychiatric illness was assessed for the year preceding TBI and for the 3-year period after TBI, as noted by the presence of a psychiatric diagnosis, filling of a prescription for a psychiatric medication, or use of psychiatric services. Psychiatric diagnoses were made according to ICD-9-CM by primary-care physicians. The authors collected 939 cases of TBI in 1993, for an overall annual TBI incidence of 475.2 per 100,000 person-years; 85.5% of the TBIs were mild. The risk of psychiatric illness was significantly increased after mild and moderate-severe TBI. Increased ORs were observed especially in patients with no prior history of psychiatric illness, within the first year after TBI: the OR was 2.1 (95% CI, 1.6–2.6) in those with mild TBI and 3.4 (95% CI, 1.9–5.8) in those with moderate to severe TBI. An approximate 1.5-fold increase in risk of psychiatric illness was also observed in the following 3 years in patients who had a diagnosis of a psychiatric illness before sustaining a mild TBI; no association was observed in such patients who sustained a moderate to severe TBI. Specifically, the new onset of a psychotic disorder was no greater after mild TBI in the following 3 years than in the year before the TBI, whereas after moderate to severe TBI, a diagnosis of psychosis were greater but not until the second and third years: 1–12 months after TBI, the OR was 2.8 (not significant); 13–24 months after TBI the OR was 5.9 (95% CI, 1.6–22.1); and 25–36 months after TBI the OR was 3.6 (95% CI, 1.0–12.3). The OR for any psychiatric diagnosis during the 3-year followup was increased considerably if there was a prior psychiatric illness.
A study by Achte et al. (1969) also found an association between TBI and psychosis. Data were collected in a Finnish hospital for brain injuries that housed all central nervous system–injured war veterans of the 1939–1945 Finnish Wars; 3,552 men comprise this cohort followed for 22–26 years. About 42% had a penetrating head wounds (shell splinters and gun shot wounds), and 58% had closed TBI. In this population, 317 (8.9%) out of the 3,552 veterans had had a diagnosis of psychosis (a rate that is 2–3 times the usual population-based rate of approximately 3–4%). In addition, 30.4% had epilepsy (44.2% of those with penetrating head injuries and 20.3% of those with closed TBIs), and 8.9% had aphasias (16.1% of those with penetrating head injuries and 3.7% of those with closed TBIs).
Godfrey et al. (1993), in a small but well-controlled study focusing on insight, found poor insight regarding behavioral impairment at 6 months after TBI. The defect appeared to attenuate with time. Increased insight regarding behavioral impairment was accompanied by emotional dysfunction. Henry et al. (2006) compared a group of TBI patients with their friends and close relativse for their ability to identify their own emotions (an aspect of insight) and found the TBI group impaired, less able to recognize emotion in others, externally oriented and less fluent on tests of semantic fluency.