Results of this study expand our previous research on ethnic differences in insulin dynamics. Using mixed-model analysis, we were able to identify trends across puberty in the insulin secretory profile. We identified an association between first-phase insulin secretion and AfADM that began at Tanner stage 3, suggesting that the contribution of genetic make-up to β-cell function begins at reproductive maturity. Further, among peripubertal children, the relationship between AfADM and first-phase insulin secretion differed with adiposity. Among obese subjects, total body fat but not AfADM was a determinant of first-phase insulin secretion, whereas among lean subjects, AfADM was significant. AfADM tended to be associated with a lower percent, but not total, hepatic insulin clearance in the entire cohort. As such, genetic factors (ancestral genetic background) may play a role in determining postchallenge insulin responses in lean individuals, whereas the potential genetic effect influencing insulin dynamics is attenuated by a physiologic component among obese individuals. This is an important consideration to be taken into account in studies probing racial/ethnic and genetic contribution to diabetes risks.
The first phase of insulin secretion is essential for maintaining normoglycemia and achieving effective glucose homeostasis (36
). There are few studies and to our knowledge no longitudinal studies in children that evaluate the ethnic differences in insulin secretory profile among obese and lean children over the pubertal transition. The physiological relevance of a greater first-phase insulin response in AA is not known, but alterations in insulin secretory dynamics may be an early event in the etiology of impaired glucose tolerance or “prediabetes” (36
). It is possible that greater first-phase insulin secretion in response to a glucose challenge may lead to exhaustion of the pancreatic β-cells at a faster rate, and even to a loss of β-cell mass through increased susceptibility to apoptosis (38
). Such stress on the pancreas may be particularly detrimental during puberty, when there is a transient decline in insulin sensitivity and an associated increase in β-cell demand. We have previously demonstrated in this population that hyper-insulinemia during the pubertal transition may lead to greater fat mass deposition in AA relative to EA (13
). Whether the metabolic perturbations associated with puberty (i.e., insulin resistance, hyperinsulinemia) are more apparent among AA, and are related to increased disease risk in adulthood is not clear. Longitudinal data are needed to better understand the roles of insulin resistance and secretion in the development of chronic metabolic disease.
The determinants of greater first-phase insulin secretion among AA relative to EA have been unclear. Previous studies conducted by our group have suggested that AA children maintain a larger readily releasable pool of insulin in the pancreas, which could lead to increased first-phase β-cell sensitivity to glucose (3
). We have previously demonstrated greater first-phase insulin secretion among AA is independent of lower SI
among AA compared to EA (3
). Further, in this study, inclusion of SI
in all models controlled for any potential confounding due to differences in insulin sensitivity. Present data suggested that genetic factors may exert independent effects on insulin secretion during first-phase insulin secretion. Ethnic differences in factors regulating insulin secretion have been reported (40
). Higher glucagon-like peptide-1 concentrations have been documented in obese AA adults, when compared to obese EA adults, both in the fasting state and during an oral glucose tolerance test (41
). Glucagon-like peptide-1 is thought to increase β-cell proliferation (41
) and may account for the increased β-cell responsiveness among AA.
In accordance with the findings of Weiss et al.
) using an oral glucose tolerance test, our results indicated that total body fat had an independent effect on insulin secretion, such that greater total body fat was associated with greater insulin secretion. However, our study is unique in demonstrating that the effects of adiposity masked ethnic differences in insulin secretion. The mechanism through which adiposity affects insulin secretion is not known. Adipose tissue secretes free fatty acids and leptin, which have postulated effects on insulin secretion. Free fatty acids are believed to have an acute stimulatory effect on insulin secretion, but chronic exposure to elevated free fatty acids has been postulated to result in β-cell lipotoxicity, causing a decrease in insulin secretion (8
). Leptin has been shown to upregulate glucagon-like peptide-1 secretion in some studies (43
); glucagon-like peptide-1 in turn could stimulate insulin secretion. Leptin is also believed to directly affect pancreatic synthesis of insulin (44
In this study, the independent association of AfADM with first-phase insulin secretion was not observed when the analyses were conducted in a subgroup of obese subjects. Similarly, no ethnic difference in first-phase insulin secretion, assessed with the clamp technique, was observed in a study involving obese AA and EA children (6
). Furthermore, in this study, among obese but not lean children, total body fat was significantly and independently related to all phases of insulin secretion. These results suggested that adiposity may override the relative contributions of genetic factors in determining insulin secretion.
Higher AIRg previously documented among AA relative to EA also could be the result decreased hepatic insulin clearance. Lower percent hepatic insulin clearance, as determined by either the C-peptide to insulin molar ratio or mathematical modeling has been noted in both AA adults (15
) and children (3
), compared to EA. In addition, lower absolute insulin clearance, as assessed with the clamp technique, has been documented among AA relative to EA children (14
). In this study, the inverse association between AfADM and percent insulin extraction over the 180-min test period approached significance (P
= 0.055). Because the percent of insulin cleared by the liver is affected by the total amount of insulin secreted, we also calculated the absolute amount of insulin cleared by the liver. We postulated that AfADM would be associated with higher absolute insulin clearance, due to greater first-phase secretion observed among AA compared to EA. Although mean values for absolute insulin clearance were higher among AA than EA (), AfADM was not a significant determinant of absolute insulin clearance in the entire cohort. It is likely that the similar amount of insulin secreted over the 180-min test period between AA and EA resulted in a similar amount of absolute hormone cleared over this time.
Our results also indicated that total body fat was positively related to absolute insulin clearance. Previous studies suggested that obesity is associated with lower insulin extraction, as estimated by the C-peptide to insulin molar ratio (9
). This discrepancy is likely explained by the strong positive association between obesity and insulin secretion, as was observed in this study. Higher insulin secretion by individuals in the obese group was likely attributable to greater absolute insulin clearance. However, it also is possible that the molar ratio, which involves only basal, fasting concentrations of C-peptide and insulin, does not accurately reflect insulin clearance under dynamic conditions.
Because this study was conducted in a healthy population of children, it avoided the potential confounding effects (beyond those that exist during reproductive maturation) of factors that may affect insulin secretion and action in adults, such as smoking, alcohol intake, and chronic obesity. Such research using healthy children is useful for detecting potential inherent physiological factors contributing to ethnic differences in insulin secretion and action. This study used refined measures of both insulin secretion and clearance to avoid the pitfalls associated with surrogate estimates (45
). A limitation of our study was the absence of measures of physical activity and diet, which might have been be useful for identifying additional socioenvironmental contributors to the variance in insulin secretion.
In conclusion, greater AfADM was associated with greater first-phase insulin secretion among lean peripubertal children. This relationship became significant at pubertal stage 3, suggesting that the pubertal process affects the development of the β-cell response to glucose in a manner that differs with ethnic/genetic background. Among obese peripubertal children, adiposity appeared to mask the relation between genetic admixture and insulin secretion. Greater AfADM tended to be associated with lower percent, but not absolute, insulin clearance. An understanding of the differences in insulin dynamics based on body habitus and how these factors may differentially impact insulin dynamics are of particular significance to studies which probe a genetic contribution to diabetes risk. Longitudinal research is needed to determine the physiological ramifications, and potential association with disease risk, of relatively high first-phase insulin secretion among African Americans.