We read with interest the article on serum leptin concentrations by the Menon group(1). The authors have clearly reported the association of elevated serum leptin with obesity in a well designed study in Indian children.

In pre-pubertal normal weight children, serum leptin concentrations, as measured by radio-immunoassay, have been noted to be equal in girls and boys(2). In contrast, the authors(1) have observed higher leptin levels in girls than boys (23.5 ± 1.78 v 18.0 ± 7.6 ng/mL) with obesity. Assuming that the cohort of obese Indian children is almost entirely pre-pubertal (90% in Tanner stage 1), the hypothesis that gender dimorphism is likely to be due to a testosterone effect may not be correct. However, this assumption may be fallacious if there were greater proportions of children with Tanner stage 2 and 3 on the female side. One wonders whether higher leptin levels were due to greater adiposity in obese girls (as a cause, not an effect); if so, leptin levels adjusted for body mass index (BMI) and BMI standard deviation scores (BMI SDS) may have provided further information.

In this study(1), in spite of a relatively small sample size (10 girls vS 26 boys), a nominal significance of difference (P = 0.04), a wide standard deviation in boys (±7.6) and probable inclusion of 3 outliers with abnormal biochemical phenotypes, there remains the possibility that pre-pubertal Indian girls with obesity have greater leptin concentrations (adjusted for BMI SDS) than their male counterparts. This may indicate greater leptin resistance in Indian obese girls before the onset of puberty.

To improve our understanding of the origins and pathogenesis of leptin resistance and obesity in the Indian context, we suggest further study of leptin concentrations in boys and girls, with longitudinal follow up through puberty, supplemented by serial measurements of BMI, pubertal staging and measures of insulin resistance in comparison with healthy normal weight controls.

As for the 3 children with high BMI (>30 kg/m2) and unusually low serum leptin concentrations(1), it would be advisable to repeat the leptin measurement in these individuals, perhaps using a different assay(2). If serum leptin levels are truly low or undetectable, one must consider the possibility of leptin mutations in these children(3) and hence proceed to leptin gene sequencing.

Indraneel Banerjee,
Department of Pediatric Endocrinology,
Royal Manchester Children’s Hospital, UK
and