Objective: To examine sex differences in height-for-age z-scores and the percentage stunting among Bangladeshi children estimated using three growth references.
Design, setting and subjects: Data collected between 1990 and 1999 by Helen Keller International's nutritional surveillance system in rural Bangladesh were analyzed for 504 358 children aged 6-59 months. Height-for-age z-scores were estimated using the 1977 NCHS, 2000 CDC and 1990 British growth references.
Results: The shape of the growth curves for Bangladeshi boys and girls, and their positions relative to one another, depend on which of the three growth references is used. At 6 months of age the British reference showed no sex difference whereas the NCHS and CDC showed girls to have higher average z-scores than boys by 0.14 and 0.28 s.d., respectively. While all references showed a faster deterioration of girls' z-scores from 6 to 24 months, the magnitude and direction of the sex differences, and how they changed with age, were different. There was greater disagreement about girls' z-scores than boys. Discontinuities at 24 months in the NCHS and CDC produced jagged curves whereas the British curves were smooth.
Conclusions: The assessment of sex differences in linear growth depends on the growth reference used. Reasons for the different results need to be determined and may aid the final development of the new WHO international growth reference and the guidelines for its use. The findings suggest that anthropometry as a tool to explore the effects of societal gender inequality must be used with caution.