Risk Factors Assessed in Adolescence and the Later Risk of Stroke in Men: A 33-Year Follow-Up Study

Abstract

Background: Stroke is a common cause of death, and a leading contributor to long-term disability. The cost associated with the disease is great. Several modifiable risk factors for stroke have been found in older cohorts; however, no study to date has investigated the effects of these risk factors from late adolescence. Methods: The study cohort comprised 811,579 Swedish men (mean age, 18 years) that participated in the mandatory military conscription service in Sweden between 1969 and 1986. Some risk factors for stroke, such as body mass index, systolic and diastolic blood pressure, and cognitive function, were assessed at conscription. Aerobic fitness was also assessed at conscription, using a braked ergometer cycle test. Other risk factors for stroke, including stroke in subjects' parents, and socioeconomic factors including highest achieved level of education and annual income 15 years after conscription, were collected through national register linkage using the personal identification number. Stroke diagnosis among the study participants was tracked in the National Hospital Discharge Patient Register. Results: During a median follow-up period of 33 years 6,180 ischemic strokes and 2,104 hemorrhagic strokes were diagnosed in the cohort at a mean age of 47.9 years. Strong independent risk factors (all p <1.0 × 10−⁶) for ischemic stroke included low aerobic fitness (hazard ratio [HR], 0.84 per standard deviation [SD] increase), high BMI (HR, 1.15 per SD increase), diabetes (HR, 2.85), alcohol intoxication (HR, 1.93), low annual income (HR, 0.85 per SD decrease), and stroke in the mother (HR, 1.31). Similar risk factors were found for hemorrhagic stroke including low aerobic fitness (HR, 0.82 per SD increase), high BMI (HR, 1.18 per SD increase) alcohol intoxication (HR, 2.92), diabetes (HR, 2.06), and low annual income (HR, 0.75). The population attributable risks associated with all evaluated risk factors were 69% for ischemic stroke and 88% for hemorrhagic stroke (p < 0.001 for both). Conclusions: In the present study we have shown that several known risk factors for stroke are present already in late adolescence, and that they are independent of each other. The strongest risk factors were low physical fitness, high BMI, diabetes, low annual income and a maternal history of stroke. Several of the aforementioned risk factors are potentially modifiable.