Background and Objective: In animal models, a rise in body temperature after cerebral ischemia consistently
produces more extensive brain damage. In humans, however, the relationship between body temperature and
stroke outcome has been far less extensively investigated, providing conflicting results. The objective of the present study is to determine whether body temperature recorded during the first 72 hours after admission is an independent predictor of mortality in acute cerebral infarction.
Material and Method: The medical records of patients admitted within 48 hours of onset of symptoms to Prasat Neurological Institute between 1 January 2002 and 31 December 2003, with a diagnosis of cerebral infarction, confirmed by CT or MRI of the brain were retrospectively studied. The relationship between the highest temperature recorded during the first 72 hours after admission and mortality during hospital stay was evaluated. Multiple logistic regression analysis included relevant confounders and potential predictors such as gender, age, hypertension, diabetes, smoking, previous stroke, ischemic heart disease, atrial fibrillation, consciousness and infections was performed.
Results: There were 332 patients included in the present study. During the first 72 hours after admission, 88 (26.5%) patients had fever (>37.5C). The highest temperature recorded during the first 72 hours after admission was a significant predictor of in hospital mortality in the final multivariate logistic regression model. For each 1C increase in body temperature the odd ratio of mortality rose by 3.95 (OR, 3.95; 95%CI, 1.84-8.45).
Conclusion: In the present study, the highest body temperature recorded during the first 72 hours after admission was shown to be a significant predictor of mortality in acute cerebral infarct patients.

Keywords: Body temperature, Cerebral infarction, Mortality, Prognostic factors