We studied 177 consecutive male patients aged 16-60 years with brain infarction verified by neuroradiology and analyzed the time of onset of stroke symptoms related to sleep and the role of possible or known risk factors for brain infarction. Brain infarction occurred relatively more often during the first 30 minutes after awakening than at any other time. In multiple stepwise logistic regression analyses, snoring was the only independent risk factor differentiating stroke occurring during sleep and stroke occurring either during sleep or during the first 30 minutes after awakening from stroke occuring at other times of the day. The risk ratios were 2.65 (95% confidence interval 1.32-5.29, p<0.005) and 3.16 (95% confidence interval 1.61-6.22, p<0.001), respectively. Other factors tested were age, arterial hypertension, diabetes mellitus, smoking, alcohol consumption, and body mass index. Arterial hypertension seemed to have an additive effect on the independent risk caused by snoring. {Stroke 1989;20:1311-1315) T here is a relation between the onset of stroke symptoms and the time of day, 1 -7 which may relate to the underlying pathophysiology of stroke type. Intracerebral hemorrhage, subarachnoid hemorrhage, and embolic infarction often occur during daily activities, whereas atherothrombotic brain infarctions often have their onset during sleep or the early morning hours.1 -2 Habitual snoring has been associated with arterial hypertension and angina pectoris.8 " 11 Snoring has also been suggested as a risk factor for brain infarction.1112 Heavy snoring is almost always present in obstructive sleep apnea, and obstructive sleep apnea syndrome has many harmful effects on the cardiovascular system. 13 -16 We studied the association between diurnal variations in ischemic brain infarction and potential risk factors for stroke, including the history of snoring and sleep apnea.
Subjects and MethodsWe studied 177 consecutive male patients aged 16-60 years admitted to the Meilahti University Hospital, Helsinki; we excluded from further study 10 patients whose sleeping habits could not be Received February 6, 1989; accepted May 16, 1989. obtained or in whom the time of stroke onset could not be determined. Brain infarction was diagnosed by a neurologist in the 167 patients and was verified in 165 (98.8%) by computed tomography, nuclear magnetic resonance imaging, single-photon emission computed tomography, or radioisotope planar scintigraphy. In many patients, more than one neuroradiologic method was used. When the clinical presentation indicated brainstem infarction, auditory evoked potentials were also recorded. We determined the occurrence of arterial hypertension (previous arterial hypertension with treatment or blood pressure > 150/100 mm Hg on admission and follow-up), coronary heart disease (previous myocardial infarction and/or clear signs of angina pectoris requiring medication), and diabetes mellitus (previous antidiabetic medication and/or special diet because of impaired glucose tolerance) from hospital ...