Do gonadal hormones confound between parental smoking and offspring sex ratio?

William H. James

The Galton Laboratory, University College London, Wolfson House, 4 Stephenson Way, London NW1 2HE, UK

Sirs—Obel et al.1 report that their data fail to confirm the suggestion of Fukuda et al.2 that periconceptional smoking of either parent tends to lower the associated offspring sex ratio (proportion male). There are at least a dozen other studies of this sort, and some confirm the propositions and others do not. It would be possible to perform a Mantel-Haenszel test across these data sets and thus derive an overall meta-analytic assessment. However, I suggest instead that such a procedure would be invalid because all of these data are bedevilled by the possibility of confounding.

A potential for confounding arises when some forms of behaviour (e.g. smoking) are tested as risk factors for outcomes which are dependent on gonadal hormone levels.3 The potential arises because both the behaviour and the outcome are independently associated with the same variables viz. parental hormone concentrations. There can be no reasonable doubt that when the smoking habit is initiated, smokers have higher testosterone and oestrogen levels than non-smokers4–6 on the average. So cross-sectional studies would be expected to yield the spurious suggestion that smoking protects against pathologies which are actually caused by low levels of gonadal hormones. (Epidemiologists should always be aware that such pathologies, e.g. Parkinson's disease, may in fact be at least partially caused by low levels of gonadal hormones.) These points are relevant for the following reasons.

  1. There is a great deal of evidence that mammalian (including human) offspring sex ratios are partially dependent on concentrations of the gonadal hormones (testosterone and oestrogens) of both parents around the time of conception; high hormone levels being associated with the subsequent birth of sons.7
  2. The only longitudinal study known to me suggests that smoking has the direct pharmacological effect of lowering men's testosterone levels.8 Moreover smoking is known to have anti-oestrogenic effects on women.9
If these propositions were true, then one would expect smokers (of both sexes) to produce high proportions of sons at the beginning of their reproductive histories, and lower proportions of sons later in these histories. Thus the problem of the relationship between parental smoking and offspring sex ratio is not well addressed by simple cross-sectional studies. I suggest that we ought to test whether (in contrast to that of non-smokers) the offspring sex ratio of smokers declines with birth order (viz. shows a negative birth order effect). If it does, the present ideas would be strongly supported.

References

1 Obel C, Henriksen TB, Hedegaard M, Bech BH, Wisborg K, Olsen J. Periconceptional smoking and the male to female ratio in the offspring —re-assessment of a recently proposed hypothesis. Int J Epidemiol 2003;32:470–71.[CrossRef][ISI][Medline]

2 Fukuda M, Fukuda K, Shimizu T, Andersen CY, Byskov AG. Parental periconceptional smoking reduces the male:female ratio of newborn infants. Lancet 2002;359:1407–08.[CrossRef][ISI][Medline]

3 James WH. Hypothesis: gonadal hormones act as confounders in epidemiological studies of the associations between some behavioural risk factors and some pathological conditions. J Theor Biol 2001;209: 97–102.[CrossRef][ISI][Medline]

4 Bauman KE, Foshee VA, Haley NJ. The interaction of sociological and biological factors in adolescent cigarette smoking. Addictive Behav 1992;17:459–67.[CrossRef][ISI][Medline]

5 Martin CA, Kelly TH, Rayens MK et al. Sensation seeking, puberty and nicotine, alcohol, and marijuana use in adolescence. J Am Acad Child Adolesc Psychiatry 2002;41:1495–502.[CrossRef][ISI][Medline]

6 Zuckerman M. Behavioral Expressions and Biosocial Bases of Sensation Seeking. New York: Cambridge University Press, 1994.

7 James WH. Evidence that mammalian sex ratios at birth are partially controlled by parental hormone levels at the time of conception. J Theor Biol 1996;180:271–86.[CrossRef][ISI][Medline]

8 Zmuda JM, Cauley JA, Kriska A, Glynn NW, Gutai JP, Kuller LH. Longitudinal relation between endogenous testosterone and cardiovascular risk factors in middle-aged men: a 13-year follow-up of former Multiple Risk Factor Intervention Trial participants. Am J Epidemiol 1997;146:609–17.[Abstract]

9 Michnovicz JJ, Naganuma H, Hershcopf RJ, Bradlow HL, Fishman J. Increased urinary catechol estrogen excretion in female smokers. Steroids 1988;52:69–83.[CrossRef][Medline]