PHYSICAL-ACTIVITY AND RISK OF SEVERE GASTROINTESTINAL HEMORRHAGE IN OLDER PERSONS

Objective.-To assess whether regular physical activity is associated w ith a decreased risk of severe gastrointestinal hemorrhage (GIH) in ol der persons. Design.-Cohort study with 3 years of follow-up. Setting.- Three communities of the Established Populations for Epidemiologic Stu dies of the Elderly. Participants.-A total of 8205 persons aged 68 yea rs or older, yielding 22 277 person-years of follow-up. Measurements.- The occurrence of severe GIH was defined as either a hospital discharg e diagnosis of gastrointestinal bleeding associated with blood transfu sion or death during the hospital stay or a nonhospital death with men tion of GIH on the death certificate. Physical activity was measured b y self-reported frequency of walking, gardening, or doing vigorous phy sical activity. Those participants doing the activity three times per week or more were compared with the remaining participants. Adjusted r elative risks (RRs) of GIH were controlled for age, gender, body mass index, blood pressure, chronic conditions, number of hospital admissio ns in the past year, and number and types of drugs taken. Results.-Sev ere GIH occurred in 241 participants (rate, 10.8 per 1000 person-years ). After adjusting for potential confounding variables, the RRs and 95 % confidence intervals (Cls) for severe GIH associated with walking, g ardening, and vigorous physical activity were 0.6 (0.4 to 0.8), 0.8 (0 .5 to 1.1), and 0.7 (0.4 to 1.2), respectively. The RR associated with a summary variable for the three activities was 0.7 (95% Cl, 0.5 to 0 .9). These results were consistent after stratifying on health status and disability or by excluding those who were not mobile, ie, not able to walk half a mile or climb a flight of stairs. Conclusions.-Regular physical activity is associated with a decreased risk for severe GIH in older persons.