Maternal haemoglobin and birth weight in different ethnic groupsBMJ 1995; 310 doi: http://dx.doi.org/10.1136/bmj.310.6994.1601 (Published 17 June 1995) Cite this as: BMJ 1995;310:1601
Methods used do not support conclusions
- Sarah Vause,
- Michael Maresh,
- Khaled Khaled
- Research register Director Research registrar Medical Audit Unit, St Mary's Hospital, Manchester M13 0JH
EDITOR,—We have three comments on the methods used by Philip Steer and colleagues in their study of the relation between maternal haemoglobin concentration and birth weight.1
Firstly, Steer and colleagues repeatedly refer to the fall in haemoglobin concentration. This implies comparison of two values. They use only the lowest haemoglobin concentration and are therefore relating birth weight to minimum haemoglobin concentration and not to the fall in haemoglobin concentration.
Secondly, they base their analysis on the lowest haemoglobin concentration during the pregnancy without taking into account the gestation at which haemoglobin concentration was estimated. Surprisingly, they do this despite emphasising the importance of the relation between gestation, expansion of plasma volume, and haemoglobin concentration.
Thirdly, they take no account of multiple pregnancies, in which low birth weight and premature labour are often seen in conjunction with low haemoglobin concentrations. Multiple pregnancies should have been excluded from the analysis, and this should have been made clear in the paper.
We think that the conclusion in the abstract that “the magnitude of the fall in haemoglobin concentration in pregnancy is related to birth weight” is not supported by the methods used.