Background: Data remain conflicting regarding the long-term metabolic consequences of prematurity and the impact of early nutrition and catch-up growth. Adiponectin and leptin are adipocyte derived proteins (adipokines) and are thought to be important regulators of insulin action.
Objectives: i) To investigate the influence of infant growth and contemporary body composition on adolescent adipokine secretion.
ii) To investigate the correlation between adipokine levels and insulin sensitivity.
Design, setting and participants: Participants were recruited from the Newcastle Preterm Birth Growth Study (NPBGS), a prospective, randomised study of infant growth. 102 underwent venepuncture and DEXA at a median of 11.5 years. Z-score weight changes between term and 12 weeks were compared with adolescent adipokine levels using multivariable linear regression to adjust for potential confounders (birthweight, gestation, fat mass index, sex and pubertal status). Insulin sensitivity was measured using HOMA 2.
Results: i) Overall, adipokine levels did not vary by sex. Stratification by pubertal status (Tanner stage 1 vs >1) showed significantly higher leptin levels in pubertal females than males and remained after adjustment for fat mass index. Infant growth patterns were not significantly associated with adolescent adipokine levels. Adiponectin levels were negatively correlated with adolescent fat mass index (Spearmans correlation =−0.25, P=0.01), while leptin was positively correlated (Spearmans correlation =0.90, P=0.00); these associations remained after multivariate adjustment.
ii) The correlations of leptin-adiponectin ratio (LAR) and leptin to insulin sensitivity (both Spearmans correlation =−0.58, P=0.000) were equal and stronger than that of adiponectin alone. (Spearmans correlation =0.20, P=0.05).
Conclusions: We have not shown an effect of early infant growth patterns on adipokine levels at adolescent follow-up and contemporary body composition appears more important. The sex difference in leptin levels in the pubertal cohort may reflect sex-based differences in body fat distribution, which only evolve at puberty. Our data indicate that the LAR is a useful surrogate marker of insulin sensitivity. The LAR could replace clamp methods and HOMA approximation when assessing insulin sensitivity in clinical research.
25 - 27 Nov 2015
British Society for Paediatric Endocrinology and Diabetes