Publication Date



Physical activity and overweight are associated with myocardial infarction (MI). However, their joint association with MI remains unclear. Our objective was to examine the independent and joint association between leisuretime physical activity (LTPA), body mass index (BMI) and MI. This prospective cohort study included 16,572 men and women (47.5% women) aged 20–54 years who took part in the second Tromsø Study. At baseline in 1979–80 LTPA was assessed by questionnaire. Data on MI was collected and adjudicated through hospital and causes of death registries between 1979 and 2013. Cox proportional hazards models were used to examine the independent and joint associations between LTPA, BMI and MI. The final sample included 16,104 individuals. During a median follow up of 34 years, 1613 incident cases of MI were recorded. Physical inactivity and elevated BMI were both independently associated with MI (p for trend 0.02 and<0.001). In joint analyses, normal weight, inactive individuals had a 20% higher risk of MI compared to their active counterparts (hazard ratio (HR) 1.20 (1.02–1.41)). The highest risk of MI was seen in obese, inactive individuals when compared to normal weight, active individuals (HR 3.20 (2.30–4.44)). The risk of MI increased with increasing BMI regardless of the activity level. HRs were lower for active compared to inactive individuals within the same BMI category. The findings suggest that LTPA and BMI are independently associated with risk of MI. LTPA seems to attenuate but not eliminate the risk of MI associated with excess bodyweight.


Mary MacKillop Institute for Health Research

Document Type

Journal Article

Access Rights

ERA Access

Access may be restricted.