Twin births to mothers who are twins: a registry based study

BMJ 1996; 312 doi: (Published 06 April 1996) Cite this as: BMJ 1996;312:879
  1. Paul Lichtenstein, research associatea,
  2. Petra Otterblad Olausson, senior administrative officerb,
  3. A J Bengt Kallen, professor emeritusc
  1. a Karolinska Institute, Institute for Environmental Medicine, Box 210, 17177 Stockholm, Sweden
  2. b Centre for Epidemiology, National Board of Health and Welfare, 10630 Stockholm, Sweden
  3. c Tornblad Institute, University of Lund, Biskopsg 7, 22362 Lund, Sweden
  1. Correspondence to: Dr Lichtenstein.
  • Accepted 24 January 1996


Objectives: To estimate the risk of having twin infants for mothers who are twins; to investigate the genetic influence on twinning.

Design: Retrospective study of multiple births in two nationwide registries.

Setting: Sweden.

Subjects: Multiple births among 31586 deliveries between 1973 and 1991 to women who were twins.

Main outcome measures: Numbers of monozygotic and dizygotic twin births expected and estimated.

Results: Women who are dizygotic twins have a moderately increased risk of having twins (relative risk 1.30, 95% confidence interval 1.14 to 1.49) which seems to be completely the result of dizygotic twinning. When a mother is a monozygotic twin, her risk of having twins of the same sex is significantly increased (1.47; 1.10 to 1.97). This is the result of an excess of monozygotic twins (39 pairs estimated, 18 expected).

Conclusion: Women who are twins have an increased risk of giving birth to twins. Genetic components of monozygotic and dizygotic twinning seem to be independent.

Key messages

  • It is generally agreed that genetic factors are important for dizygotic but not for monozygotic twinning

  • Women who themselves are dizygotic twins have an increased risk of giving birth to dizygotic twins

  • Women who themselves are monozygotic twins have an increased risk of having monozygotic twins

  • The genetic components in monozygotic and dizygotic twinning are independent


It is generally agreed that genetic factors are important for dizygotic but not for monozygotic twinning. This has not yet been thoroughly investigated.

Among women who have already given birth to twins the risk for further twin births is increased even after age effects are controlled for.1 2 3 4 The risk seems to exist only for dizygotic twinning.1 4 5 A genetic influence is indicated by the increased number of twin births among female relatives of women who had given birth to dizygotic twins.1 2 3 4 6 7 A study of genealogical records in Salt Lake City, USA, found a twinning rate of 1.45% among mothers who had a twin brother (dizygotic twins) and 1.65% for their sisters, compared with the rate of 0.69% for fathers who had a twin sister.7

No firm evidence exists for a genetic influence on monozygotic twinning, although interview studies of mothers of monozygotic twins report an excess of monozygotic twins among the mothers' relatives.4 6 Increased monozygotic twinning rates have also been reported in selected pedigrees.8 9

Most evidence is based on either archival studies or self reports from twins or parents of twins. Both methods may give biased results. Archival studies may underreport twin maternities because stillbirths or early perinatal deaths of one twin in a pair (which occur more often in twin than in singleton deliveries) may be recorded as singleton births.4 Interview studies may overestimate twinning rates among first degree relatives in twin families, most likely owing to a heightened awareness of twinning, even if one of the twins was stillborn or died early. This may explain the relatively odd finding of a paternal role in dizygotic twinning as well as the purported common aetiology of monozygotic and dizygotic twinning.4


This study is based on a linkage of two nationwide Swedish registries: the Swedish twin registry,10 which encompasses data on all twins born in Sweden during the years 1926 to 1991, and the medical birth registry,11 covering 99% of all births in Sweden since 1973. Record linkage identified the female twins born 1926 to 1972 who gave birth in Sweden between 1973 and 1991.

The twin registry contains information on the sex of the twins and, for people born before 1959, zygosity determined on the basis of childhood resemblance.12 Zygosity was verified with serological data for 99% of the monozygotic and 92% of the dizygotic twins in a subsample.13 A high level of validity of this technique has been reported by others.14 15

The year of birth, maternal age, parity, and the number of infants in the birth were obtained from the medical birth registry. The observed numbers of multiple births was compared with expected numbers, which were calculated from the rate of multiple births in the medical birth registry stratified by maternal age and parity.

The estimated rate of monozygosity was based on Weinberg's assumption of equal numbers of same sex and different sex dizygotic twins.2 Thus, the number of monozygotic twins was estimated as the total number of twins minus twice the number of different sex twins. Although Weinberg's assumption has been questioned,16 the potential bias for this study is negligible because there is no reason to assume different biases among mothers who are twins and mothers who are singletons.


The odds ratio for having a same sex and different sex twin delivery for each maternal age or parity group is shown in figure 1. The maternal age effect on dizygotic twinning (clearer among the different sex twins because of inclusion of monozygotic deliveries among same sex twins), the decline in odds ratio among women above 40, and the parity effect have been described by Bulmer.1

Fig 1
Fig 1

Odds ratio in each maternal age group versus all other age groups (stratified by year of birth and parity) and in each parity group versus all other parity groups (stratified by maternal age) for having a twin delivery for all women in the medical birth registry, Sweden. Odds ratios are presented separately by type of delivery; 95% confidence intervals are marked with vertical bars

In table 1, the births have been divided into groups by maternal twin type. Overall, the twinning rate is 28% greater than in the general population.

Table 1

Number and type of births to Swedish mothers who were twins. Expected numbers were calculated after stratification by maternal age and parity

View this table:

Dizygotic mothers have a significantly increased risk of twinning. The risk increase is stronger for different sex twins than same sex twins. The estimated number of monozygotic twin births was not higher than the number expected, but the estimated number of dizygotic twin births was greater than the expected number (table 1). The ratio of estimated to expected number of dizygotic twins (1.63) was the same as for different sex twins.

For monozygotic mothers the number of same sex twin births was significantly increased (relative risk 1.47, 95% confidence interval 1.10 to 1.97). This is probably due to monozygotic twinning, as the ratio of estimated to expected number of monozygotic twins was 2.17 (39/18.0). With five births of different sex twins to monozygotic mothers there is a 95% probability that the total number of births of dizygotic twins will not exceed 18. The minimum number of monozygotic pairs is then 31—the total number of twin deliveries (49) minus the maximum number of dizygotic deliveries (18)—which clearly is in excess of the expected number, 18.0.


The risk of mothers who are dizygotic twins giving birth to dizygotic twins is significantly increased even after age and parity have been controlled for. These results are compatible with previously reported increased twinning in dizygotic twins and their relatives.12467 The underlying mechanism for dizygotic twinning is multiple ovulation,17 which has been shown to be affected by genetic factors.18 Segregation analyses have indicated that an X linked gene in sheep19 and an autosomal gene in humans20 affect multiple birth. Familial occurrences of dizygotic twinning found in this and other epidemiological studies are at least in part due to genetic factors.

The risk of twin births to monozygotic mothers was not significantly increased, but monozygotic mothers had significantly more same sex twin births. Comparison of estimated and expected numbers of monozygotic twin births suggests that this increase is a result of monozygotic twinning. Although other studies have suggested a familial component in monozygotic twinning, samples have been small and the methods susceptible to recall bias.46

This population based study shows that women who are twins have an increased risk of giving birth to twins. Contrary to results based on self reports, 4 this registry based study, free from recall bias, does not support the existence of a common aetiology for monozygotic and dizygotic twinning.


  • Funding The Swedish twin registry is supported by grants from the John D and Catherine T MacArthur Foundation and the Swedish Council for Planning and Coordination of Research (FRN).

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
View Abstract