Exposure to smog, floods,1 or earthquakes2 has been associated with a lower sex ratio (proportion of males) in offspring conceived at the time of the exposure episode. We hypothesised that exposure to severe life events before and during the periconceptional period might be associated with a decline in the sex ratio.
Methods and results
All Danish women who gave birth from 1 January 1980 to 31 December 1992 were identified by the population based medical birth registry. By linking personal identification numbers to the Danish statistical office’s fertility database, we identified all the women’s partners and older children. A partner was defined as the father of the child or the partner with whom the woman was cohabiting on 1 January in the year of birth. Severe life events were defined as death or admission to hospital for cancer (codes 140 to 209 (excluding 208 polycythemia vera) of the international classification of diseases, eighth revision) or acute myocardial infarction (code 410) in the partner or older children. By linking the personal identification numbers to the mortality registry and to the national registry of patients, we identified all deaths and all admissions to hospital of partners and children between 1978 and 1992.
We identified all women exposed to severe life events in the year of birth and the previous year but included only the women exposed before the second trimester. Women with diabetic partners were excluded, as it has been reported that the sex ratio in their offspring is higher. The time of exposure was recorded as the date of death or first admission to hospital. Information on the possible confounding variables was obtained from the registries.
The exposed cohort consisted of 3072 singleton pregnancies exposed to severe life events and a control cohort of 20 337 singleton pregnancies randomly selected among all pregnancies without such exposure in the observation period. Of all the offspring, 945 had congenital malformations, chromosomal abnormalities, or hereditary diseases.
Using logistic regression analysis, we found that the proportion of boys was 49.0%in the exposed group and 51.2% the control group, giving a crude overall lower ratio between the two sex odds (sex odds ratio) of 0.92 and an adjusted ratio of 0.91 (95% confidence interval 0.84 to 0.99) (table). Exclusion of offspring whose mothers were exposed during the first trimester and all offspring with congenital malformations gave similar results (data not shown).
The nearer the exposure was to conception, the lower the overall sex odds ratio (test for trend P=0.013) and the lower the sex ratio for exposure by an older sibling only (P=0.035) and for exposure by death of a partner only (P=0.016).
Comment
The results show that severe life events may reduce the sex ratio, especially for exposures around the time of conception, because of differential conception or differential abortion of male embryos. The study cohort was large, and the registration of death, cancer, and the offsprings’ sex is known to be accurate and complete in the registers used. The results are consistent with results for natural disasters.1,2
The study has some limitations. The exact dates of the actual stress exposure may in most cases have started long before our date of exposure. For this reason, we also included women exposed during the first trimester. More factors may have contributed to a dilution of the effect—for example, other stressors may have affected both the women in the exposed group and those in the control group, and the partners may not have been the father of the offspring in all cases.
Psychological stress related to severe life events may alter the sex ratio through changes in sexual activity, changes in hormones around the time of conception,3 reduced semen quality,2,4 or an increased rate of early male abortion.5
Table.
Adjusted sex odds ratios and percentage of boys as function of severe life events before and in periconceptional period and stratified for timing of exposure
| Type of cohort | Numbers | Boys (%) | Sex odds ratio (95% confidence interval)* |
|---|---|---|---|
| Controls | 20 337 | 51.2 | 1.00 |
| Exposed women: | 3 072 | 49.0 | 0.91 (0.84 to 0.99) |
| Exposed 13-16 months before† | 71 | 52.1 | 1.04 (0.65 to 1.69) |
| Exposed 7-12 months before† | 789 | 49.6 | 0.94 (0.81 to 1.08) |
| Exposed 0-6 months before† | 1 922 | 48.9 | 0.91 (0.83 to 1.01) |
| Exposed first trimester | 290 | 46.0 | 0.81 (0.64 to 1.02) |
| Source of exposure: | |||
| Older sibling: | 2 661 | 49.4 | 0.93 (0.85 to 1.01) |
| Exposed 13-16 months before† | 56 | 55.4 | 1.18 (0.70 to 2.00) |
| Exposed 7-12 months before† | 693 | 50.8 | 0.98 (0.84 to 1.15) |
| Exposed 0-6 months before† | 1 771 | 49.1 | 0.92 (0.83 to 1.02) |
| Exposed first trimester | 141 | 43.6 | 0.74 (0.53 to 1.03) |
| Partners | 411 | 46.5 | 0.83 (0.68 to 1.01) |
| Type of life event: | |||
| Death | 2 575 | 49.0 | 0.92 (0.84 to 1.00) |
| Exposed 13-16 months before† | 44 | 54.7 | 1.15 (0.63 to 2.08) |
| Exposed 7-12 months before† | 650 | 50.5 | 0.97 (0.83 to 1.14) |
| Exposed 0-6 months before† | 1 718 | 49.2 | 0.92 (0.83 to 1.02) |
| Exposed first trimester | 163 | 40.6 | 0.65 (0.47 to 0.89) |
| Cancer | 426 | 48.1 | 0.88 (0.73 to 1.07) |
| Acute myocardial infarction | 71 | 49.6 | 0.94 (0.59 to 1.59) |
Sex ratio of exposed/sex ratio of non-exposed; and adjusted for maternal and partners’ ages as well as for school education, maternal parity (nulliparous v multiparous), previous spontaneous abortions (yes/no), residence (capital, major city, or rural), and congenital malformation in offspring (yes/no).
Before first day of last menstruation.
Acknowledgments
DH and JO also work, as paediatrician and professor respectively, at the Danish Epidemiology Science Centre, Statens Serum Institut, in Copenhagen.
Footnotes
Funding: Danish Medical Research Council, Health Insurance Foundation, Dagmar Marshall Foundation, and Danish Medical Association Research Foundation. The activities of the Danish Epidemiology Science Centre are financed by a grant from the Danish National Research Foundation.
Competing interests: None declared.
References
- 1.Lyster WR. Altered sex ratio after London smog of 1952 and the Brisbane flood of 1965. J Obstet Gynecol Br Commonwealth. 1974;81:626–631. doi: 10.1111/j.1471-0528.1974.tb00529.x. [DOI] [PubMed] [Google Scholar]
- 2.Fukuda M, Fukuda K, Shimizu T, Møller H. Decline in sex ratio at birth after Kobe earthquake. Human Reprod. 1998;13:2321–2322. doi: 10.1093/humrep/13.8.2321. [DOI] [PubMed] [Google Scholar]
- 3.James WH. Evidence that mammalian sex ratios at birth are partially controlled by parental hormones levels at the time of conception. J Theor Biol. 1996;180:271–286. doi: 10.1006/jtbi.1996.0102. [DOI] [PubMed] [Google Scholar]
- 4.Fenster L, Katz DF, Wyrobek AJ, Pieper C, Rempel DM, Oman D, et al. Effects of psychological stress on human semen quality. J Androl. 1997;18:194–202. [PubMed] [Google Scholar]
- 5.Pratt NC, Lisk RD. Effects of social stress during early pregnancy on litter size and sex ratio in the golden hamster (Mesocricetus auratus) J Reprod Fert. 1989;87:736–769. doi: 10.1530/jrf.0.0870763. [DOI] [PubMed] [Google Scholar]