The Association of Changes in Cardiorespiratory Fitness and Body Mass Index with All-Cause Mortality

Abstract

The associations of changes in cardiorespiratory fitness (CRF) and body mass index (BMI) with mortality have not been fully investigated. PURPOSE: We examined the independent and combined associations of changes in CRF and BMI with mortality in 14,364 men aged ≥20 (mean age 44 ± 9 years at baseline) who were examined at least twice during 1974-2002. METHODS: All participants were free of CVD and cancer, and had a BMI ≥18.5 at both baseline and final examination. All achieved ≥85% of age-predicted maximal heart rate on the treadmill test at both examinations. Maximal metabolic equivalent tasks (METs) were estimated from the final treadmill speed and grade. We classified CRF change as loss (decline of at least 5%), stable (change less than 5%), and gain (increase of at least 5%) in maximal METs between baseline and final examination. BMI was calculated by measured weight and height and BMI change was classified as loss (decline of at 5%), stable (change less than 5%), and gain (increase of at least 5%). Mortality follow-up was through December 31, 2003 using the National Death Index. Cox regression analysis was used to quantify the associations of CRF and BMI changes with mortality. Baseline age, examination year, BMI, METs, interval between baseline and final examination, number of examinations during 1974-2002, and changes in lifestyle and disease conditions (smoking status, alcohol intake, physical activity, cardiac conditions, hypertension, diabetes), and BMI change or CRF change were included in the multivariate regression model. RESULTS: There were 917 deaths during 11.5 years of follow-up. The relative risks (RRs) and 95% confidence intervals (CIs) for mortality were 1.43 (1.20-1.70) for CRF loss, and 0.84 (0.71-0.99) for CRF gain compared with stable CRF. The RRs (95% CIs) for BMI change were 1.23 (1.02-1.48) for BMI loss, and 0.94 (0.77-1.14) for BMI gain compared with stable BMI. We also compared changes in CRF and BMI with both CRF and BMI stable as the reference. We observed higher RRs for mortality for CRF loss within all 3 BMI change categories. CONCLUSIONS: CRF reduction and BMI loss were associated with higher mortality risk, while CRF improvement was associated with lower risk compared with a stable CRF or stable BMI. In addition, CRF reduction was associated with higher risk regardless of the direction of BMI change.