Influence Of Exercise And Non-Exercise Physical Activity On Mortality In Chinese Women

Charles E. Matthews,Adriana Jurg,Xiao-Ou Shu,Honglan Li,Gong Yang,Gi Li,Yu-Tang Gao,Wei Zheng
DOI: https://doi.org/10.1249/00005768-200605001-02703
2006-01-01
Abstract:PURPOSE: To describe the effect of both exercise and non-exercise physical activity on risk for mortality among Chinese women. METHODS: The Shanghai Women's Health Study (CA70867) is a prospective cohort study that enrolled 74,942 women between 1997 and 2000. The cohort has been followed for an average of 5.7yrs via in-person follow-up and linkage with vital disease registries (< 2% lost to follow-up). At baseline, in-person interviews were conducted to gather relevant exposure information (demographics, health behaviors, medical and weight history). Physical activity patterns were characterized in terms of exercise participation (adolescence [13-19 yrs], adulthood [last 5 yrs]), active transport (walking, cycling), household chores, stair climbing, and by type of occupation. An index of non-exercise activity (active transport, household, stair climbing) energy expenditure was calculated. Proportional hazards models were used to estimate hazard ratios and 95% confidence intervals (HR [95% CI]) and adjust for relevant covariates. Only women with no history of heart disease, stroke, or cancer at baseline were examined (N=67,480). RESULTS: Women aged 51.7 (SD=8.8)yrs at baseline contributed 384,889 person-years (PY) of observation and deaths from all-causes (n=1,108), cardiovascular disease (n=199), cancer (n=545), and other causes (n=364) were recorded during follow-up. Examination of exercise participation in each life-period, in mutually exclusive categories, revealed that exercise only in adolescence (HR=0.79 [0.62–0.99]), only in adulthood (HR=0.82 [0.70–0.97), and in both life-periods (HR=0.66 [0.52–0.83]) reduced risk for all-cause mortality, compared to women that reported no exercise. Increasing levels of non-exercise activities, by quartile (Q), were associated with reduced risk of all-cause mortality (Q1 =referent; HRQ2=0.80 [0.69–0.94]; HRQ3=0.64 [0.54–0.76]; HR =0.66 [0.56–0.79]). Occupational activity estimated from job codes was not associated with all-cause mortality. Exclusion of deaths occurring early in follow-up (first 2 yrs) did not substantively alter these results. The strongest effect of activity was noted for cancer and other causes of death, and there was no clear effect on cardiovascular mortality, which was predominantly derived from stroke in this analysis. CONCLUSIONS: Both exercise and common non-exercise activities (e.g., walking, cycling, and household chores) were associated with reduced risk for early mortality. The present findings support current health promotion efforts that encourage the public to increase their non-exercise activity levels.
What problem does this paper attempt to address?