Aims We addressed the relevance of two usual beliefs about blood glucose response to exercise in insulin-treated diabetic patients. First, that a pre-exercise blood glucose value exceeding 250 mg/dL means that exercise should be avoided because it will result in worsening of hyperglycemia; second, the prediction of carbohydrate oxidation according to the level of exercise “glucose pulse” may have some value to predict how blood glucose levels will respond to exercise. Methods Twenty-four type-1 diabetics (12 men and 12 women, age 19–71 years) treated with basal-bolus or continuous insulin delivery by portable pump performed a steady-state exercise on ergocycle at 50% of their predicted P max (40–200 W) with measurement of blood glucose levels, blood lactate, and exercise calorimetry. This protocol lowers blood glucose ( p < 0.05), with a variable magnitude (−4 to −178 mg/dL [mean: −51 ± 9 mg/dL]), i.e., blood glucose decreased in 21 subjects (87.5%). Carbohydrate oxidation rates do not, however, predict the evolution of blood glucose. Baseline blood glucose G 0 (ranging between 89 and 270 mg/dL, i.e., on average 179 ± 14 mg/dL) was in fact negatively correlated (and not positively) to the variation of blood glucose during exercise ( r = −0.462; p < 0.05), so that the higher is G 0 the more blood glucose decreases. When G 0 is greater than 250 mg/dL blood glucose decreased on average by −97 ± 26 mg/dL ( p < 0.05). During these steady state bouts of exercise, which clearly decrease blood glucose, oxidation of carbohydrates does not predict the decline in blood glucose (minimizing the interest of the concept of “glucose pulse”), and a value of G 0 greater than 250 mg/dL does not predict exercise hyperglycemia so that it is no longer logic to set this value as a threshold for counter indicating exercise.
Read full abstract