Letters

Authors defend methods used in their paper

BMJ 1999; 318 doi: http://dx.doi.org/10.1136/bmj.318.7185.737a (Published 13 March 1999) Cite this as: BMJ 1999;318:737
  1. Bertil Jacobson, Professor emeritus (bertil.jacobson{at}labtek.ki.se),
  2. Marc Bygdeman, Professor
  1. Department of Medical Engineering, F60 Novum, Huddinge University Hospital, SE-141 86 Huddinge, Sweden
  2. Department of Obstetrics and Gynaecology, Karolinska Hospital, SE-171 76 Stockholm, Sweden

    EDITOR—In his editorial1 Appleby discusses our paper on obstetric care and the proneness of infants in traumatic births to suicide as adults.2 He gives the incorrect impression that we did not consider the potential confounding effect of year of birth and fetal hypoxia (asphyxia). We did so by including these variables in the multivariate regression—as is evident from the text and table A in the additional information given on the BMJ's website. If year of birth is forced into the regression the estimated relative risk as well as the significance of the trauma score is not reduced.

    In the editorial Appleby claims that our findings are presented as supporting the imprinting hypothesis and that imprinting is the causal mechanism. This is not correct. The hypothesis served only in the design of the study and to predict the results. We would never have said that our data support the hypothesis. We could have said that the hypothesis cannot be rejected on the basis of the results of the study, which is an entirely different conclusion. But we did not even do this, and the hypothesis in not mentioned in the discussion. The essence of our paper is not the hypothesis but the statistical findings.

    Appleby suggests that any link between obstetric care and violent suicide occurs through mental illness, possibly caused by fetal hypoxia. As we stated in our paper, we cannot exclude the possibility that it may be the circumstances giving rise to the need for a traumatic intervention that cause the increased risk of suicide, rather than the intervention itself. Our results, however, do not support the notion that fetal hypoxia is the causal factor.

    References

    1. 1.
    2. 2.
    View Abstract