Background: Previous studies have reported that the running speed of school-age children is influenced mainly by genetic factors. Children who are fast at short distances may possess dominant fast muscle fibers, while those who are fast at long distances possess dominant slow fibers. Fast muscle fibers are characterized as insulin-insensitive, while slow fibers are insulin sensitive. Therefore, a different relationship may potentially exist between running speed at short vs. long distances in school-age children and the prevalence of type 2 diabetes in adulthood.

Purpose: This study investigated the relationship between running speed in school-aged children and the prevalence of diabetes mellitus in adulthood.

Methods: 3,105 adults (aged 20−80 years) participated in a cross-sectional internet survey. In addition to the demographic and physical characteristics, we asked about relative running speeds (slow, somewhat slow, average, somewhat fast, fast) for short and long running distances during childhood. We also asked about the presence of diabetes at the time of the study in adulthood. Logistic regression models were used to evaluate the relationship between running speed during childhood and diabetes during adulthood, and multivariable-adjusted odds ratios (ORs) and 95% confidence intervals (95% CIs) after adjusting for sex, age, BMI, physical activity, smoking habit, and drinking habit.

Results: The number of participants with self-reported diabetes was 228. For short-distance running speeds, using the "slow" group as reference, the multivariable-adjusted ORs (95% CIs) of diabetes were 0.89 (0.56−1.40) for "somewhat slow", 0.73 (0.47−1.13) for "average", 0.62 (0.39−1.00) for "somewhat fast", and 0.51 (0.30−0.88) for "fast" group (P for trend = 0.004). For long-distance running speeds, using the "slow" group as reference, the multivariable-adjusted ORs (95% CIs) of diabetes were 0.67 (0.44−1.04), 0.62 (0.42−0.92), 0.54 (0.33−0.86), and 0.27 (0.12−0.59) in the corresponding groups (P for trend < 0.001).

Conclusions: Faster short-distance running speed in school-aged children was associated with a lower prevalence of diabetes in adulthood. In addition, we observed a similar, yet slightly stronger, association for long-distance running speed. However, prospective studies are needed.