Background and Purpose-Polymorphism of the apolipoprotein E gene (APOE
) may influence outcome after traumatic brain injury and intracerebral
hemorrhage, with the epsilon 4 allele being associated with poorer pr
ognosis. We investigated APOE allele distribution in acute stroke and
the effect of the epsilon 4 allele on outcome. Methods-APOE genotypes
were determined in 714 stroke patients: 640 ischemic stroke and 74 int
racerebral hemorrhage patients. The survival effect of the epsilon 4 a
llele was assessed with the use of a stratified log-rank test. A Cox p
roportional hazards regression model was used to estimate the independ
ent effect of epsilon 4 dose (0, 1, or 2) on survival, and logistic re
gression was used to determine the effect on 3-month outcome (good if
alive at home, poor if in care or dead). Results-Allele distribution m
atched the general population with no difference between the ischemic
and hemorrhagic groups. Survival in the entire cohort was unaffected b
y epsilon 4 dose. Improved survival with increasing epsilon 4 dose was
found in the ischemic group (relative hazard=0.76 per allele; P=0.04)
. If transient ischemic attacks were excluded, a trend for improved su
rvival persisted (P=0.06). With intracerebral hemorrhage, a trend was
seen toward reduced survival with epsilon 4 (P=0.07, log-rank test). T
hree-month outcome in the ischemic group was unaffected by epsilon 4 d
ose, and a trend toward poorer outcome with epsilon 4 was Seen for int
racerebral hemorrhage (P=0.10). Conclusions-The APOE epsilon 4 allele
had divergent effects on survival and outcome in ischemic and hemorrha
gic strokes in this population The reported adverse effect on patients
with intracerebral hemorrhage was supported. The favorable survival e
ffect on ischemic stroke patients requires further study.