Reliability of a submaximal treadmill test to determine critical speed in recreational marathon runners

The aim of the present study was to verify the reliability of T10 for marathon runners at recreational level.

Forty-three runners, with at least 18 months of experience in long distance running participated in the study. Six tests were conducted, including VO_2max assessment, T10-1 (test familiarization), T10-2 (retest) and three field tests on an athletics track at distances of 1.200 m, 2.400 m and 3.600 m. Critical speed (CS) was determined using a linear regression based on the distances and performance times. Reliability was evaluated using Pearson and Bland-Altman correlation.

The results revealed a significant difference in CS between the initial T10 test and T10 retest (15.1 ± 1.9 km h^-1 vs 14.3 ± 1.5 km h^-1) and between the T10 retest and field test (14.3 ± 1.5 km h^-1 vs 13.9 ± 1.5 km h^-1). The correlation between CS in the T10 retest and the field test was r = 0.83 (R² = 68%) indicating high reliability as confirmed by Bland-Altman analysis (0.40 km h^-1; 95% CI 0.13-0.67 km h^-1 - limits of agreement: 2.10-1.30 km h^-1).

The T10 test demonstrated acceptable reliability when applied to recreational marathon runners, do not replace the conventional procedure for CS assessment, but can be adopted as an alternative for predicting critical speed.