In this potentially very important paper the investigators study the effects of various diets in the newborn period of premature infants versus the presence at ages 15-16 years of a plasma marker for the development of insulin resistance and non-insulin dependent diabetes in adults.  The marker is known as 32/33-SPI (split proinsulin).  Plasma concentrations were measured in 216 mid to late adolescents (13-16 years of age) who had been delivered prematurely (mean gestational age of 31 weeks and mean birth weight of 1.4 kg).  Of these preterm infants, 110 had received a low nutrient formula and 106 had received a high nutrient formula.

Not surprising, the preterm newborns fed the lower nutrient formula gained less weight prior to discharge compared to those receiving the higher nutrient formula.  The specific formulas were stopped when the infants were discharged from the neonatal unit or had reached a weight of 2000 gms.  At 16 years of age the children were re-evaluated and fasting serum concentrations of insulin, proinsulin, and 32/33-SPI were determined in specific assays.  As adolescents, the low nutrient group had lower levels of 32/33-SPI than the levels in the high nutrient group.  Levels of insulin, proinsulin and glucose were similar in the two groups.

After statistically adjusting for the effects of gender, gestational age, birth weight, neonatal morbidity, and other variables, the relationship between neonatal diet and concentrations of 32/33-SPI remained significant.  Further analysis revealed that high rates of weight gain in the neonatal period (basically a surrogate for higher caloric intake) - specifically within the first two weeks after birth - were most closely related to elevated levels of 32/33-SPI in adolescence which were independent of birth weight.  There was no association between values of 32/33-SPI and weight gain between two weeks of age and discharge from the nursery, discharge and 18 months, 18 months and 9 years, 9-12, and 13-16 years.  Preterm adolescents, fed a low nutrient diet at birth, did not differ in stature, weight, BMI, or sum of skinfold thickness compared with premature infants who were fed the high nutrient formula or from the control group of adolescents born at term. 

The investigators conclude that premature infants who were fed a low nutrient formula (albeit one that impaired neonatal weight gain) for several weeks after birth resulted in lower concentrations of 32/33-SP in adolescence, and by inference these subjects may be less likely to develop insulin resistance.  They hypothesize that the risk for developing insulin resistance in low birth weight neonates is not necessarily programmed by the intrauterine environment, but also by the immediate post partum extrauterine environment as exemplified by the high nutrient formula and more rapid weight gain that accompanies this diet.  They suggest that altering current feeding practices of preterm infants by lowering their caloric intake and decreasing their early rate of weight gain may prevent later development of insulin resistance, cardiovascular disease, and the dysmetabolic syndrome without adversely affecting their long-term growth.

Singhal A, et al.  Lancet 2003;361:1089-1097.

Editor’s Comment: Low birth weight infants are at risk for future development of the dysmetabolic syndrome (X) of dyslipidemia, insulin resistance, and type 2 diabetes mellitus.¹  It has been hypothesized that intrauterine factors that affect the fetal response to decreased blood flow or nutrient availability “program” the subsequent development of this syndrome - primarily by inhibiting tissue responsiveness to insulin.  However, there is no specific explanation that explains the cellular and molecular mechanisms by which low birth weight predisposes to insulin resistance.  The current work is of interest because it points to the possibility that post natal factors, in this instance rapid growth secondary to increased nutritional intake in very early life, contribute to the later development of insulin resistance.  Thus, this observation affords the possibility of an intervention that may prevent this long-term complication without negatively impacting the overall growth of the low birth weight subject.  Considered in the context of the findings is that partial nutrient restriction and growth hormone deficiency extend life in many species^2,3 including perhaps primates. Since the level of 32/33-SPI is only a marker of insulin resistance, it will be necessary for Singhal et al to continue to follow these subjects and to document the development of insulin resistance and other adverse events as, and if they occur.

Allen W. Root, MD

References

1. Goran MI, et al. J Clin Endocrinol Metab 2003;88:1417-1427.

2. Longo VD, Finch CE. Science 2003;299:1342-1346.

3. Tatar M, Bartke A, Antebi A. Science 2003;299:1346-1351.