Skip to main content
NCBI home page
The BMJ logoLink to The BMJ
. 1999 Jan 23;318(7178):259. doi: 10.1136/bmj.318.7178.259a

Association between birth weight and death from heart disease

Data do not support association

R J Jarrett 1
PMCID: PMC1114735  PMID: 9915745

Editor—Leon et al claim that their study is “the most persuasive evidence of a real association between size at birth and mortality from ischaemic heart disease.”1 However, the study is open to different interpretations and, in my view, inflicts a serious wound on the birth measurement and adult disease hypothesis.

Their table 3 shows that there was no significant association between birth weight and all cause mortality in either sex. Both sexes showed a positive association between birth weight and death from neoplasms and respiratory disease, though none of these were significant. The negative association between birth weight and deaths from circulatory disease was significant only in men. In women there were no significant associations between birth weight and any cause of death. Subsequently, most of the analysis concentrates on the association of death from ischaemic disease and birth measurements in men.

This study is claimed (probably correctly) to have unique features. It certainly has large numbers of deaths to analyse. Yet no significant association was found in women, and that in men, with ischaemic heart disease, was presumably compensated for by other causes of death, which made the association with all cause mortality non-significant. Thus, of all the possible associations with birth weight, and despite the large number of deaths, only the association between birth weight and death from ischaemic heart disease in men remains significant. I do not see how these data justify the key message “adult mortality from ischaemic disease increases as size at birth declines.”

There is also the question of socioeconomic confounding. Leon et al state that adjustment for socioeconomic circumstances produced only a small reduction in the strength of the association between birth weight and mortality from ischaemic heart disease. The problem of adjusting for socioeconomic factors in this context has been much debated. In this paper the adjustment brought the upper limit of the confidence intervals perilously near unity. If the precision of this measurement matched that of the others, even the single significant result might be in peril.

References

  • 1.Leon DA, O Lithell H, Vågerö D, Koupilová I, Mohsen R, Berglund L, et al. Reduced fetal growth rate and increased risk of death from ischaemic heart disease: cohort study of 15 000 Swedish men and women born 1915-29. BMJ. 1998;317:241–245. doi: 10.1136/bmj.317.7153.241. . (25 July.) [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 1999 Jan 23;318(7178):259.

Authors’ reply

David A Leon 1,2,3, Ilona Koupilová 1,2,3, Paul McKeigue 1,2,3, Hans O Lithell 1,2,3, Rawya Mohsen 1,2,3, Lars Berglund 1,2,3, Ulla-Britt Lithell 1,2,3, Denny Vågerö 1,2,3

Editor—Jarrett draws attention to the intriguing positive association between birth weight and neoplasms that we observed. Interest is growing in whether risk of certain cancers is increased in individuals who are large at birth.1-1 If true, this may indeed partly compensate for the negative effect of reduced size at birth on later mortality from circulatory disease. However, it is precisely the specificity1-2 of the negative association of size at birth with mortality from ischaemic heart disease that strengthens the case for this association having a causal element. Although not reaching significance, a similar association was observed in women, with risk of death from ischaemic heart disease falling with increased size at birth. We therefore stand by our assertion that risk of death from ischaemic heart disease increases as size at birth falls.

The question of socioeconomic confounding is an important one that we addressed. Jarrett is wrong to focus on confidence intervals. The main criterion for judging the extent to which there may be inadequate adjustment for a confounder is the magnitude of the change in the estimates of effect, rather than changes in the P values or width of confidence intervals. In our case the rate ratio for ischaemic heart disease associated with a 1 kg increase in birth weight was 0.77 in the crude data and 0.82 when adjusted for socioeconomic characteristics at the birth of the subject and at two points in adult life. We believe that this shows that although a small fraction of the crude association between size at birth and ischaemic heart disease is attributable to socioeconomic confounding, there remains a consistent and robust effect that is not.

The evidence for impaired fetal growth being associated with ischaemic heart disease in men is steadily accumulating. Unlike other studies, ours has largely been able to exclude the role of bias and confounding. Further epidemiological studies in other populations are desirable, particularly concerning the association in women and the potential modifying effects of factors such as obesity in adult life. However, the understandable scepticism about the reality of these associations that prevailed when the fetal origins hypothesis was first put forward some years ago should now give way to critical thinking and investigation of plausible underlying mechanisms. Among many other questions that need to be addressed is the extent to which maternal nutritional status drives this association.

References

  • 1-1.Trichopoulos D, Lipworth L. Is cancer causation simpler than we thought, but more intractable? Epidemiology. 1995;6:347–349. [PubMed] [Google Scholar]
  • 1-2.Hill AB. The environment and disease: association or causation ? Proc R Soc Med. 1965;58:295–300. doi: 10.1177/003591576505800503. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from BMJ : British Medical Journal are provided here courtesy of BMJ Publishing Group

RESOURCES