Population based study of rates of multiple pregnancies in denmark, 1980-94

BMJ 1997; 314 doi: (Published 15 March 1997) Cite this as: BMJ 1997;314:775
  1. Tine Westergaard, epidemiologista,
  2. Jan Wohlfahrt, statisticiana,
  3. Peter Aaby, professora,
  4. Mads Melbye, professor and heada
  1. a Department of Epidemiology Research Danish Epidemiology Science Centre Statens Serum Institut Artillerivej 5 DK-2300 Copenhagen S Denmark
  1. Correspondence and requests for reprints to: Professor Melbye
  • Accepted 31 December 1996


Objective: To study trends in multiple pregnancies not explained by changes in maternal age and parity patterns.

Design: Trends in population based figures for multiple pregnancies in Denmark studied from complete national records on parity history and vital status.

Population: 497 979 Danish women and 803 019 pregnancies, 1980-94.

Main outcome measures: National rates of multiple pregnancies, infant mortality, and stillbirths controlled for maternal age and parity. Special emphasis on primiparous women ≥30 years of age, who are most likely to undergo fertility treatment.

Results: The national incidence of multiple pregnancies increased 1.7-fold during 1980-94, the increase primarily in1989-94 and almost exclusively in primiparous women aged ≥30 years, for whom the adjusted population based twinning rate increased 2.7-fold and the triplet rate 9.1-fold. During 1989-94, the adjusted yearly increase in multiple pregnancies for these women was 19% (95% confidence interval 16% to 21%) and in dizygotic twin pregnancies 25% (21% to 28%). The proportion of multiple births among infant deaths in primiparous women ≥30 years increased from 11.5% to 26.9% during the study period. The total infant mortality, however, did not increase for these women because of a simultaneous significant decrease in infant mortality among singletons.

Conclusions: A relatively small group of women has drastically changed the overall national rates of multiple pregnancies. The introduction of new treatments to enhance fertility has probably caused these changes and has also affected the otherwise decreasing trend in infant mortality. Consequently, the resources, both economical and otherwise, associated with these treatments go well beyond those invested in specific fertility enhancing treatments.

Key messages

  • National rates of multiple pregnancies have risen from 1.0% to 1.7% during 1980-94 in Denmark; this rise was most pronounced in recent years

  • Multiple pregnancy rates changed particularly among primiparous women ≥30 years of age, where the adjusted twinning rate increased threefold and triplet rate ninefold during1989-94

  • The dramatic increase in the twinning rate seems to be restricted to dizygotic twin pregnancies

  • These changes are believed to be associated with the increasing use of treatments to enhance fertility


The rates of multiple pregnancies have varied considerably during this century. These fluctuations have predominantly been explained by changes in maternal age and parity.1 New risk factors for multiple pregnancies have appeared with the introduction of hormonal induction of ovulation and advanced reproduction techniques.2 3 4 5 Such regimens may result in multiple pregnancies in about a quarter of the births.5 Despite the relatively small proportion of a population who undergo treatment to enhance fertility, their substantially increased risks of multiple pregnancies could have considerable impact on the national rates.

We took advantage of the population based national registers in Denmark to study trends in rates of multiple pregnancy that were not explained by changes in maternal age and parity patterns. Furthermore, we analysed to what extent changes in these rates influenced the national rates for stillbirths and infant mortality.

Subjects and methods

Data from the Danish Civil Registration System were used to obtain complete family histories. All liveborn children and new residents in Denmark are recorded in this register and ascribed a unique 10 digit personal identification number (the person number). Individual information is kept under the person number in all national registers, which enables high quality linkages between the different registers. The registration system was established on 1 April 1968, when all people who were alive and resident in Denmark were registered. It includes various data such as date of birth, sex, vital status, and information on parents. On the basis of this system we established a database that contains close to complete information on parity for all women born in Denmark who gave birth during the study period (January 1980 to September 1994). For this particular study, we added information from the Danish National Birth Registry on all stillborn children (born after 28 completed weeks of pregnancy) born during the period 1978-93. Data on stillbirths were not available for 1994. In this study a pregnancy was defined as a delivery. In order to identify multiple pregnancies we looked for children (live and stillborn) born to the same mother within two days (on each side of midnight). All children were assigned a number that indicated whether they were singleton, twin, triplet, quadruplet, or quintuplet. The mothers were assigned a parity number for each delivery.

Weinberg's differential rule was used to estimate the number of dizygotic and monozygotic twin pairs–that is, the number of dizygotic twin pairs was calculated as twice the number of opposite sexed twin pairs and the number of monozygotic pairs was calculated as the total number of twin pairs minus the estimated number of dizygotic pairs.6

Adjustment for changes in maternal age and parity was done with a log-linea r binomial regression model7 with the SAS procedure proc genmod.8 Adjustment for parity was based on a categorisation into three groups (1, 2, and ≥3 para) and adjustment for maternal age on a categorisation into six age groups (<20, 20-24, 25-29, 30-34, 35-39, and ≥40 years) for multiple pregnancies and twin pregnancies overall, while for dizygotic and monozygotic twin pregnancies and for triplet pregnancies the adjustment was into four age groups (<25, 25-29, 30-34, and ≥35 years) because of smaller numbers.

Analyses of infant deaths (liveborn children dead within one year of birth) were performed only for children born during 1980-92 as data on deaths were not available for the whole year of 1994.


Multiple pregnancies

During January 1980 to September 1994 we recorded 803 019 pregnancies (deliveries) (table 1) among 497 979 women. Of these, 9904 (1.23%) were multiple pregnancies. There were 9689 (1.21 per 100 pregnancies) twin pregnancies, 206 (2.57 per 10 000 pregnancies) triplet pregnancies, eight quadruplet pregnancies, and one quintuplet pregnancy. During 1980-94, the mean age of the mother at first birth rose from 24.2 to 26.8 years and the mean age at second and later deliveries rose from 28.4 to 30.2 years. Overall, 38.7% of all deliveries occurred in primiparous women below 30 years of age and 7.4% in primiparous women ≥30 years. In comparison, 29.2% of all deliveries were recorded in multiparous women below 30 years of age and 24.7% in multiparous mothers ≥30 years. The proportion of women who had their first baby when they were aged ≥30 years rose from 4.9% in 1980 to 10.3% in 1994.

Table 1

 Distribution of all pregnancies (deliveries) in Danish women from January 1980 to September 1994

View this table:

The crude rate of multiple pregnancies increased from 1.03 per 100 pregnancies in 1980 to 1.18 in 1988 and increased further to 1.74 in 1994 (table 1). The crude twinning rate increased from 1.02 per 100 pregnancies in 1980 to 1.15 in 1988 and reached 1.67 in 1994 (table 1). Figure 1) illustrates the calendar effect on twinning rates by parity and maternal age. Table 2) shows changes in the twinning rates adjusted for maternal age and parity. For primiparous women ≥30 years of age there was a 2.7-fold (95% confidence interval 2.3 to 3.2) adjusted increase in the twinning rate from the period 1980-8 to 1994, corresponding to a yearly increase of 18% (15% to 20%) from 1989 to 1994. Primiparous women <30 years of age, multiparous women <30 years, and multiparous women ≥30 years experienced adjusted increases in the twinning rates of 1.3 (1.1 to 1.5), 1.2 (1.0 to 1.4), and 1.3 (1.2 to 1.5), respectively.

Fig 1

Triplet pregnancies showed a similar but more pronounced pattern (table 2). The crude triplet rate increased from 1.44 per 10 000 pregnancies during 1980-8 to 6.08 per 10 000 pregnancies in 1994 (table 1). For primiparous women ≥30 years of age the adjusted increase was 9.1-fold (3.2 to 25.5) from 1980-8 to 1994, while there was little change in the triplet rate for multiparous women <30 years (table 2), fig 2).

Table 2

 Changes in twinning and triplet rates during 1989-94 by age of mother and parity. Values are rate ratios adjusted for age and parity (95% confidence intervals)

View this table:
Fig 2

The crude dizygotic twinning rate increased from 0.57 in 1980 to 1.29 per 100 pregnancies in 1994 and the monozygotic twinning rates remained stable at 0.45 and 0.38 per 100 pregnancies during the same period. Among primiparous women ≥30 years of age the adjusted dizygotic rate increased 4.0-fold (3.3 to 4.9) from 1980-8 to 1994 (table 3), with an adjusted yearly increase of 25% (21% to 28%) from 1989 to 1994. For primiparous women <30 years of age the adjusted increase in the dizygotic twinning rate in 1994 compared with 1980-8 was 1.8-fold (1.5 to 2.1). The corresponding increase for multiparous women was 1.4 (1.1 to 1.7) and 1.7 (1.5 to 2.0) for women <30 years and ≥30 years, respectively. None of the adjusted monozygotic twinning rates increased during 1989-94 for any of these four groups of women (data not shown).

Table 3

 Changes in rate of dizygotic twinning* during 1989-94 by age of mother and parity. Values are rate ratios adjusted for age and parity (95% confidence intervals)

View this table:

Death rates

The overall stillbirth rate was 0.46 and 0.47 per 100 children born during the periods 1980-8 and 1989-93, respectively. The rates among singletons were 0.42 and 0.43 per 100, among twins 2.24 and 2.00 per 100, and among triplets 6.67 and 5.04 per 100, respectively. None of the stillbirth rates differed significantly between the two periods.

In primiparous women ≥30 years of age the proportion of multiple births among stillborn children increased 1.4-fold (0.8 to 2.3) from 1980-8 to 1989-93 (table 4). The total rate of stillbirths in primiparous women ≥30 years, however, was the same (0.59 per 100 born children) in both periods.

Table 4

 Percentages (proportions) of multiple births among stillbirths and infant deaths by age of mother, parity, and period with adjusted relative increase in proportion of multiple births among stillbirths (from 1980-8 to 1989-93) and infant deaths (from 1980-8 to 1989-92)

View this table:

There was a 0.95-fold (0.89 to 1.00) adjusted decrease in the overall infant mortality from 1980-8 to 1989-92, which fell from 0.79 to 0.72 per 100 liveborn children. The infant mortality for singletons decreased 0.93-fold (0.88 to 0.99) from 0.72 to 0.65 per 100, while there was a non-significant decrease in the rate for twins from 3.82 to 3.34 per 100 and for triplets from 10.58 to 8.33 per 100.

In primiparous women ≥30 years of age the proportion of multiple births among infant deaths increased 2.3-fold (1.4 to 3.7) for children born during 1989-92 compared with 1980-8 (table 4). Among all livebirths in this group of women, however, there was a 0.90-fold (0.72 to 1.11) adjusted decrease in the total infant mortality over time. This was due to a 0.76-fold (0.59 to 0.97) adjusted decrease in the infant mortality among singletons from 1980-8 to 1989-92.


Maternal age and parity patterns fluctuate over time and significantly influence the rate of multiple pregnancies. The existence in Denmark of such complete national registers enabled us to study trends in rates of multiple pregnancy that were adjusted for such confounding effects. This is in contrast with previous studies that have either lacked this possibility or have only to some extent been able to distinguish between these factors.1 9 10 11 In the Netherlands a 1.3-fold increase was reported in the twinning rate and a 2.7-fold increase in the triplet rate from 1975 to 1989. “Natural” causes for the increase, such as increasing age of childbearing, could, however, not be ruled out in that study.9 In white Americans in the United States there was a 1.3-fold increase in the ratio of twins among liveborn children from 1980 to 1992.10 In the same population a 2.1-fold increase was reported in the rate of triplet and higher order multiple births from 1972-4 to 1985-9, which, adjusted for maternal age, amounted to a 1.8-fold increase, an increase that was seen particularly in women ≥30 years of age.11 The impact of parity on the increasing rates in the United States, however, could not be determined in either of these studies.10 11

We found a considerable increase in the national rates for multiple pregnancies in Denmark during 1980-94 that could not be explained by changes in maternal age and parity. This increase was primarily observed during the most recent period of 1989-94 and in particular for primiparous women ≥30 years of age. In this group of older women, the adjusted increase in the twinning rate was 2.7-fold and in the triplet rate as much as 9.1-fold during 1994 compared with the rates of 1980-8. The increase in the twinning rate was exclusively observed for dizygotic twin pregnancies and in particular among primiparous women ≥30 years of age who experienced a 4.0-fold adjusted increase during 1989-94.

Effect of fertility treatment

Our ability to adjust for the confounding effect of maternal age and parity implies that the observed increase represents a realistic figure for the absolute increase attributable to other causes. The increases in the national multiple pregnancy rates seem closely related to the increasing use of ovulation induction and advanced reproduction techniques that may result in multiple pregnancy in about a quarter of births, ranging from a low of 7-9% for clomiphene citrate to a high of 25-40% for human menopausal gonadotropins and advanced reproduction techniques.2 3 4 5 12 A study of births after in vitro fertilisation reported 97% of the women to be primiparous with a mean maternal age of 32 years and a rate of multiple pregnancy of 22%.13 Our finding of a dramatic increase in the multiple pregnancy rate mainly among older primiparous women strongly supports the link to fertility enhancing treatment.

In Denmark the first child from in vitro fertilisation was born in the beginning of the 1980s. It was only after the mid-1980s, however, that this treatment became common practice. The number of clinics performing in vitro fertilisation and other advanced reproduction techniques went from one in the mid-1980s to six in the beginning of 199014 and at least 12 in 1993.15 In 1993 it was estimated that 800-1000 children had been born as a result of in vitro fertilisation in Denmark since the introduction of this treatment,15 and in 1994, 2929 women received treatment with advanced reproduction techniques.16 Although there are no exact figures available for Denmark concerning the use of induction of ovulation as a treatment to enhance fertility, there has been a definite increase in its use in recent years.17 Reduction procedures in multiple pregnancies have been applied in Denmark only in exceptional situations. To reduce the increase in multiple pregnancies caused by fertility enhancing treatment, however, in 1993 the Danish National Board of Health recommended that only two and never more than three oocytes or embryos should be transferred per treatment cycle, and when hormonal induction of ovulation is the only treatment a final ovulatory trigger should be given only if there are no more than three follicles ≥17 mm.17

It is generally accepted that most of the variation in the twinning rate worldwide is due to variation in the dizygotic rate and that monozygotic rates are fairly constant.18 Ovulation induction19 and in vitro fertilisation20 have, however, been reported to increase slightly the incidence of monozygotic twinning. Nevertheless, the results of our study would suggest that the effect on the monozygotic twinning rate is unimportant on a national scale and that the primary impact is an increasing dizygotic twinning rate. The validity of Weinberg's rule to estimate the number of monozygotic and dizygotic twins has been debated.21 22 23 It is, however, generally adopted by researchers all over the world, and it seems unlikely that the debated uncertainties should invalidate the overall interpretation of our findings.

Infant mortality

It is noteworthy that changed treatment regimens for a relatively small group of women have so drastically changed the overall national rates of multiple pregnancies. We also found a particular impact on infant deaths, in which the proportion of multiple births increased more than 2-fold from 11.5% to 26.9% for primiparous women ≥30 years of age. The total infant mortality, however, did not increase for this group of women because of a significant decrease in the infant mortality among singletons during the same period. Overall, this suggests that the introduction of new fertility enhancing treatments has retarded the otherwise decreasing national trends of infant mortality in Denmark. Multiple births are also known to be associated with higher risks of complications such as premature birth, low birth weight, and increased morbidity.11 24 Consequently, the resources, both economically25 26 and otherwise, associated with infertility treatment go well beyond those invested in the specific procedures.27

We may be only in the beginning of a new pattern for multiple pregnancies as the rate has increased each year in the 1990s. It should be noted that the mortality data presented here pertain only to the first years of this new trend. With longer follow up and an increasing proportion of multiple births, it may well turn out that the births related to fertility treatment have a greater impact on the national mortality rates. It seems essential that the trends in multiple pregnancies and infant mortality are monitored in future years to detect unwarranted consequences of these treatments.


Funding: Danish Medical Research Council and the Danish National Research Foundation.

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.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
View Abstract