Summary.Insulin was absorbed faster from the abdomen than from the thigh under resting conditions and during exercise. Exercise enhanced the rate of insulin absorption marginally. The fall of blood glucose during rest and exercise was not significantly different after insulin injection into either site. The faster absorption of insulin from the abdomen during rest and exercise was reflected in a sharper rise of serum growth hormone levels and urinary adrenaline excretion. Therefore exercise should not be taken immediately after injection of a large dose of soluble insulin, particularly into the abdomen.Key words: Insulin, insulin absorption, plasma insulin, blood glucose, exercise, growth hormone, urinary catecholamines.Physical exercise is one of the three corner stones of the treatment of the insulin-dependent diabetic. However, the fluctuations of blood glucose during exercise often exceed the desirable range. Previous investigations have shown that the rate of absorption of subcutaneously injected insulin varies during physical exercise depending on the site of injection [1][2][3][4] in man and rat. These findings might explain the large variations of blood glucose during exercise, being mediated by faster absorption of insulin and/or an increased glucose uptake by the working muscle. A relatively good correlation has been established between the rate of insulin absorption and the decline of blood glucose in diabetic subjects [5][6][7][8][9], although it is weaker in normal subjects [91.The aims of this study were to determine in normal subjects 1) the effect on blood glucose after injecting insulin into the thigh compared with the abdomen, 2) the concentrations of insulin in plasma and of some counter-regulatory hormones under these conditions and 3) the influence of physical exercise on these metabolic indices. Subjects and MethodsFour young healthy subjects (mean age 24.5 years, range: 21-31 years; mean weight 66.7 + 7.1 kg and mean height 174.9 _+_ 7.7 cm) gave their informed consent to the following study. Each of the subjects was tested on six occasions: NaC1 was injected into the arm for basal values. Actrapid insulin (Novo) was injected (0.12 U/kg body weight) 20 cm above the upper edge of the patella in the median line or 4 cm to the side of the umbilicus. The skin was folded and the depth of injection was 0.5-0.7 cm. Each test was carried out twice; once under resting conditions and a second time during exercise, consisting of 50% of the working capacity calculated according to the Pulse-Watt-Capacity 170 in the upright position (87.5 _+ 6 W; ergometical working capacity at a pulse rate of 170/rain). The subjects worked three 15-min periods each separated by 5 min of rest. Venous blood samples were drawn through an indwelling catheter in the antecubital vein of the contralateral arm to the injection site at 0, 15, 45, 65, 85, 120 and 150 min. The following determinations were carried out: blood glucose (glucose oxidase, YSI-Glucose Analyser 23A), plasma insulin by radioimmunoassay (detection limit:...