Long-term trends in cardiorespiratory fitness and the incidence of type 2 diabetes.

Abstract

Whereas single assessments of cardiorespiratory fitness have been shown to predict lower incidence of type 2 diabetes, there are no data on long-term trends in fitness and risk. We investigated the relationship between long-term trends in fitness and the incidence of type 2 diabetes. A cohort of 4,187 Japanese men free of diabetes completed annual health checkups and fitness tests for estimated maximal oxygen uptake at least four times over 7 years (1979-1985). We modeled the trend in fitness over 7 years for each man using simple linear regression. Men were then divided into quartiles based on the regression coefficient (slope) from the model. During the follow-up period (1985-1999), 274 men developed diabetes. Hazard ratios (HRs) and 95% CIs for the incidence of diabetes were obtained using the Cox proportional hazards model. Men in the lowest quartile of the distribution decreased in fitness over the 7 years (median slope -1.25 ml/kg/min), whereas men in the highest quartile increased in fitness (median slope 1.33 ml/kg/min). With adjustment for age, initial fitness level, BMI, systolic blood pressure, cigarette smoking, alcohol intake, and a family history of diabetes and use of the lowest quartile, the HRs (95% CI) for the second through fourth quartiles were 0.64 (0.46-0.89), 0.40 (0.27-0.58), and 0.33 (0.21-0.50), respectively (P(trend) < 0.001). These results indicate that the long-term trend in fitness is a strong predictor of the incidence of type 2 diabetes in Japanese men.