Relationship between insulin, insulin-like growth factor I, and dehydroepiandrosterone sulfate concentrations during childhood, puberty, and adult life.

The relationships between plasma insulin, insulin-like growth factor I (IGF-I) and dehydroepiandrosterone sulfate (DHEAS) concentrations in normal subjects have not been defined. We performed iv glucose tolerance tests on 102 normal subjects, aged 5-20 yr. The subjects were divided into 4 groups according to pubertal stage (Tanner): A, stage 1 (n = 22); B, stages 2 and 3 (n = 17); C, stages 4 and 5 (n = 20); and D, adult, greater than 17 yr (n = 43). The basal plasma IGF-I and insulin concentrations and incremental 0-60 min insulin areas in response to glucose rose significantly throughout puberty (P less than 0.001 for all parameters) and declined to prepubertal levels by the third decade of life. There was a strong positive correlation between log fasting plasma insulin vs. log plasma IGF-I (r = 0.625; P less than 0.001) and log incremental 0-60 min insulin areas vs. log plasma IGF-I (r = 0.572; P less than 0.001). Plasma DHEAS concentrations were measured in groups A-C (n = 59); these also rose throughout puberty. There was strong correlations between log plasma DHEAS and log basal or stimulated (incremental 0-60 min areas) insulin responses (P less than 0.001). To assess the relationship between plasma DHEAS and insulin before puberty, we analyzed the data from group A separately. Plasma DHEAS concentrations tended to be higher in children 9 yr of age or older than in those less than 9 yr old, whereas basal and stimulated plasma insulin levels were similar. We found no correlation between log plasma insulin (fasting or stimulated responses) and log plasma DHEAS concentrations in group A (P greater than 0.05). In conclusion, we found a strong relationship between plasma insulin and IGF-I throughout childhood and puberty and during adult life. This finding suggests that insulin may be important for normal growth during childhood. There was no correlation between plasma insulin and DHEAS concentrations in prepubertal children, which suggests that adrenarche does not influence insulin levels.