**5. Statistical analysis**

All results are presented as means ± standard deviation. The data were analyzed by two-way ANOVA followed by the Student-Newman-Keuls test when appropriate. Significant differ‐ ences in glucose concentrations and insulin doses and physical efficiency variables in relation to references ranges were determined using the Bonfferoni post-hock test. Pearson correlation coefficients were analyzed to determine the inter-variable relationships. All analyses were performed using the Statistica v. 9 statistical software package (StatSoft, Tulsa, OK, USA). Statistical significance was set at p < 0.05.
