Intrauterine growth retardation, insulin resistance, and nonalcoholic fatty liver disease in children

In their comment to our report (1), Fraser et al raise several issues that were not settled in our study. The first problem is the persistence into adulthood of the insulin resistance generated by intrauterine growth retardation (defined as small for gestational age) (SGA) and its potential for NAFLD. Very low birth weight confers an increased susceptibility to insulin resistance and metabolic diseases in adulthood (2), not attributable to body size composition or fat distribution. In adults born SGA, liver fat deposition and altered liver function tests may be part of a central axis of abdominal obesity. Dual-energy X-ray absorptiometry documented a moderately higher degree of abdominal obesity in young men born SGA, independently of BMI and waist-to-hip ratio, associated with an impairment of the insulin-stimulated glycolytic flux and a reduced forearm (muscle) glucose uptake in the face of normal whole-body glucose uptake (3). Ectopic fat accumulation in adults born SGA may thus derive from an abnormal function of the adipose tissue, suggested by hypoleptinemia, compared with individuals with normal weight at birth, even after correction for fat mass, gender and hyperinsulinemia (4). A second issue regards a possible gender difference in the risk of NAFLD associated with several variables. On the basis of indirect evidence, we have previously speculated that the circulating levels of estrogens within the physiologic range might be responsible for a “protective” effect on the development of steatosis (due to both NAFLD and, presumably virus-related steatosis (5). This would imply that, similar to what happens in cardiovascular disease, women in the fertile age are more spared than men from the development of steatosis and that this protection vanes after the menopause. Whether this also happens in SGA steatosis and NAFLD occurring in childhood remains to be determined. The relatively small size of our cohort does not allow a robust analysis in relation to gender. Whereas in males the association of SGA with NAFLD is maintained (OR, 9.59; 95% confidence interval, 2.52-36.52), the association is no longer significant in females (OR, 4.21; 0.58-30.66), but less than 25% of our NAFLD females were in the postpubertal age.