1997 Articles

Alzheimer's disease after remote head injury: an incidence study.

Schofield, P W; Tang, M; Marder, K; Bell, K; Dooneief, G; Chun, M; Sano, M; Stern, Yaakov; Mayeux, R

Objective-To evaluate a history of recall bias is a potential problem with caseremote head injury as a risk factor for control studies, the apparent association subsequent dementia due to Alzheimer's between remote head injury and dementia disease. may be spurious. Two cohort studies,89 in Methods-271 participants of a commu- which the history of head injury was obtained nity based longitudinal study of aging in before the onset of dementia, failed to confirm north Manhattan without evidence of sig- an association with subsequent Alzheimer's nificant cognitive impairment were inter- disease. In this study we investigated the risk rogated for a history ofhead injury on two for subsequent incident Alzheimer's disease occasions at entry into the study. The associated with a history of head injury exminng physician sought a history of obtained from initially non-demented elderly head injury with loss of conciousness. subjects. Subjects were all participants in a Independently, a risk factor interviewer longitudinal, community-based study in the inquired about a history of head injury Washington Heights section of Manhattan in with loss of consiousness or amnesia, the New York City. duration of any loss of consiousness, and the date of the head injury. Patients were followed up with standardised annual Methods evaluations for up to five years to deter- Data were obtained from a community based mine the first occurrence of dementia. registry of conditions related to aging in north Results-Over the course of the study Manhattan.'0 To create this registry, nursing incident dementia due to probable or pos- homes, home healthcare agencies, private sible Alzheimer's disease was diagnosed practioners, and hospital admission and disin 39 patients. Cox proportional hazards charge lists were canvassed to identify service modelling showed that a history of head recipients aged 60 years or more who were injury with loss of consiousness reported invited to take a brief cognitive screening to the physician was associated with ear- examination'1 modified from the comprehenlier onset of dementia due to Alzheimer's sive assessment and referral interview.'2 disease (relative risk (RR) = 4*1, 95% Almost all subjects entered the study between confidence interval (95% CI) 1-3-12.7). December 1989 and November 1991. All subhead injury with loss of consiousness or jects who screened positive (score > 2) and a amnesia reported to the risk factor inter- randomly selected 26% of all subjects who viewer was not significantly associated screened negative (score < 2) were referred to with earlier onset of Alzheimer's disease a clinical evaluation team for comprehensive overall (RR 2*0, 95% CI 0.7-6.2), but those clinical assessments, which were repeated who reported loss of consiousness exceed- annually. Each annual assessment consisted of ing five minutes were at significantly a clinical evaluation by a physician (comprisincreased risk (RR 11*2, 95% CI 2*3-59 8). ing history and general medical and neuroIncident Alzheimer's disease was signifi- logical assessments), and a battery of cantly associated with head injury which neuropsychological tests administered by a occurred within the preceding 30 years trained tester. The standardised neuropsycho(RR 5.4, 95% CI 1-5-19.5). logical battery consisted of tests of memory Conclusion-The results of this cohort (the Buschke selective reminding test"3 and a study are consistent with the findings of multiple choice version of the Benton recogniseveral case-control studies suggesting tion test'4), orientation, abstract reasoning,'5 16 that head injury may be a risk factor for language,'7 "I and construction,'9 and usually Alzheimer's disease.