Circulating malignant cells in Burkitt's lymphoma: possible role in tumour dissemination.

Hence, distributions of height and (maturational) age for gender differ considerably and adjustments by linear regression cannot alter this: comparisons remain difficult. Secular trends further complicate such analyses.2 3 Because the height-lung function relationship during puberty is complex and varies with age and gender,4 5 it is desirable to study the sexes before and after attaining adult height. A contribution of body mass index to lung function may biologically relate to trunk size rather than fat distribution. Schwartz et al found correlations between body mass index and sitting height and concluded that sex related differences in lung function are partly explicable by the growth of the trunk.6 Indeed chest circumference and biacromial width correlate-independently of stature-with body mass index (r=080 and 055, in 995 girls, R Veeneklaas, personal communication). Explained variance is improved by body mass index but it is unclear if this holds equally for the whole age range studied, and it seems premature to infer that body mass has a positive effect on lung function in girls and in normal weight boys.