Article Text

This article has a Reply. Please see:

Recent secular trends in multiple birth rates
  1. The Galton Laboratory
  2. University College London
  3. Wolfson House
  4. 4 Stephenson Way
  5. London NW1 2HE

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Editor—Dunn and Macfarlane1 reviewed the possible causes of the secular increase in multiple birth rates in England and Wales since 1980. They concluded: “Although the causes of the rise since 1980 in multiple births in general, and the dramatic rise in the triplet rate in particular, cannot be ascertained or quantified directly from routinely collected data, drugs used for subfertility and more recently, assisted conception, probably had a major role.”

    There can be no doubt that these authors are correct. But it is interesting to consider whether the rise is solely iatrogenic. There are reasons for suspecting that the “natural” twinning rate (which had been declining throughout the 1960s and early 1970s in England and Wales) had halted by 1980 and started to increase thereafter.

    Dunn and Macfarlane wrote: “Up to 1992, the increase was seen in all age groups except for women aged under 20.” This is not entirely correct: admittedly, the increase was small in that age group, but table 1 shows that an increase has occurred in that age group too. One may assume that few births to women of this age were preceded by treatment (of any sort) for subfertility. The inference is that the rise was “natural.”2

    Table 1

    Multiple maternity rates per 1000 inside and outside marriage, England and Wales 1976–95. Women under 20 only. (Data from OPCS Birth Statistics, Series FM 1)

    Derom et al 3 gave data for East Flanders, Belgium, for 1976–92, an area for which twin ascertainment was thought to be almost complete. The data included the numbers of “natural” twin pairs and the numbers preceded by all forms of medical assistance at conception. On the assumption that these data were typical of Belgium as a whole, I calculated2 that the “natural” twinning rate began increasing in Belgium after 1983.

    In short, there is some evidence that the substantial recent rise in multiple births in England and Wales (and Belgium) cannot be wholly attributed to medical treatments for subfertility. The point may be tested in England and Wales by checking the assumption noted above—namely, that most twin conceptions to women under 20 are not preceded by medical treatment for subfertility. Numbers may clarify the point. The rises in rates from 1976–80 to 1993–5 are about 12% within marriage and 22% outside marriage. If the percentages of twin pairs born 1993–5 (to women of this age) which had been preceded by medical treatment at the time of conception were as great as (or greater than) these values, then I am wrong. Otherwise, the inference is that the natural twinning rate has been increasing. The point is important because the natural dizygotic twinning rate may be regarded as a monitor of reproductive health (and the natural monozygotic twinning rate is roughly stable).

    Interestingly, the rate of twinning outside marriage at this age is now approaching the rate within marriage. For many years in England and Wales, twinning rates outside marriage were higher at all maternal ages except the lowest (less than 23). I have proposed that this curious feature may be explained hormonally.4 The suggestion is that in general, illicit sexual intercourse evokes high levels of erotic arousal and high female gonadotrophin concentrations (and hence increased risk of dizygotic twins). In contrast, in young women, illicit sexual intercourse may (or used to) produce considerable emotional turmoil and (perhaps because of anxiety) erotic constraint and low gonadotrophin concentrations. The fact that the rate outside marriage is approaching that within marriage might corroborate this suggestion. As the distinction in our society between licit and illicit sexual intercourse becomes blurred, one may expect differences between their emotional and hormonal concomitants (and their proposed effects on twinning) also to diminish.


    Drs Macfarlane and Dunn reply: We thank William James for his interesting comments. In trying to assess the extent to which the rise in multiple birth rates has been associated with the use of assisted conception and drugs for the medical management of subfertility, we did not assume that they have accounted for all the increase. On the other hand, we may not have looked sufficiently closely at the under 20 age group.

    The fact that the twinning rates inside and outside marriage are now very similar among women under 20 may be a reflection of the vast increase in the proportion of maternities outside marriage in this age group in England and Wales. This rose from just under a third in 1975 to over 80% in the 1990s. Both this increase and the overall percentage of maternities outside marriage have been similar for singleton and multiple births. Although the percentage of multiple births registered outside marriage was consistently slightly lower than that of singletons up to the mid 1980s, the differences between them were small.

    We are intrigued by William James’ theories about the past differences in multiple birth rates inside and outside marriage, but wonder what evidence he has to support them. Most births outside marriage in England and Wales are registered by both parents, but acknowledging paternity at birth registration does not necessarily constitute proof of paternity, irrespective of whether or not the couple are married.

    Linked Articles