8
Early Human Development, 11 (1985) 317-324 Elsevier 317 EHD 00650 Pattern of menstrual cycles and incidence of congenital malformations Alfred Spira *, Nadine Spira, Emile Papiernik-Berkauer and Daniel Schwartz Uniik de Recherches Statistiques and Unitt de Recherches sur la physiologic et la psychologie de la reproduction humaine, Institut National de la SantC et de la Recherche Mkdicale, France Accepted for publication 15 January 1985 An analysis has been made of the general reproductive characteristics of mothers, including the temperature curves of preconception and conception cycles, in a prospective, study which ended in the birth of 22 malformed infants and 894 normal infants. The differences observed have shown that the mothers of malformed infants had their first menstrual period at a later age (13.4 vs 12.8 years), their menstrual cycles were more often irregular (77% vs 40%) and lasted longer (32.9 vs 30.1 days). Moreover, they had a longer hypothermic phase during both their preconception cycle (21.7 vs 17.6 days) and during their conception cycle (24.0 vs 18.1 days), as well as a longer temperature rise during their conception cycle (3.7 vs 3.0 days). By taking into account the relation between these variables, we have been able to show that increased risk of malformation is associated with increase in the length of the hypothermic phase and the temperature rise of the conception cycle. congenital abnormalities; menstrual cycles Introductipn Numerous studies have shown that in various animal species aging of the gametes could be responsible for disturbances in fetal development, particularly hetero- ploidy, embryonic deaths and malformations [l-3,5,6]. However, it is impossible to measure the aging of the gametes directly in humans and its consequences for the * _To whom correspondence and reprint requests should be addressed: INSERM U 21, 16, avenue P.V. Couturier, 94807 Villejuif Cedex, France. 0378-3782/85/$03.30 0 1985 Elsevier Science Publishers B.V. (Biomedical Division)

Pattern of menstrual cycles and incidence of congenital malformations

Embed Size (px)

Citation preview

Early Human Development, 11 (1985) 317-324 Elsevier

317

EHD 00650

Pattern of menstrual cycles and incidence of congenital malformations

Alfred Spira *, Nadine Spira, Emile Papiernik-Berkauer and Daniel Schwartz

Uniik de Recherches Statistiques and Unitt de Recherches sur la physiologic et la psychologie de la reproduction humaine, Institut National de la SantC et de la Recherche Mkdicale, France

Accepted for publication 15 January 1985

An analysis has been made of the general reproductive characteristics of mothers, including the temperature curves of preconception and conception cycles, in a prospective, study which ended in the birth of 22 malformed infants and 894 normal infants. The differences observed have shown that the mothers of malformed infants had their first menstrual period at a later age (13.4 vs 12.8 years), their menstrual cycles were more often irregular (77% vs 40%) and lasted longer (32.9 vs 30.1 days). Moreover, they had a longer hypothermic phase during both their preconception cycle (21.7 vs 17.6 days) and during their conception cycle (24.0 vs 18.1 days), as well as a longer temperature rise during their conception cycle (3.7 vs 3.0 days). By taking into account the relation between these variables, we have been able to show that increased risk of malformation is associated with increase in the length of the hypothermic phase and the temperature rise of the conception cycle.

congenital abnormalities; menstrual cycles

Introductipn

Numerous studies have shown that in various animal species aging of the gametes could be responsible for disturbances in fetal development, particularly hetero- ploidy, embryonic deaths and malformations [l-3,5,6]. However, it is impossible to measure the aging of the gametes directly in humans and its consequences for the

* _To whom correspondence and reprint requests should be addressed: INSERM U 21, 16, avenue P.V. Couturier, 94807 Villejuif Cedex, France.

0378-3782/85/$03.30 0 1985 Elsevier Science Publishers B.V. (Biomedical Division)

318

quality and the development of the zygote. Indirect evidence suggests that delayed ovulation or overripeness in one or the other of the gametes might be responsible for the defective development of the ovum [9,10,15] or of the fetus [7,8,14].

Systematic collection of data supplied by a prospective study leading to 1075 conceptions enabled an analysis to be made of the relation between certain female characteristics, some elements of the temperature curves (especially those concerning the preconception and conception cycle) which may have a link with delayed ovulation and the risk of congenital malformation at birth.

Subjects and Methods

Subsequently to the publication of articles in newspapers from 1979 to 1982, volunteer couples trying to conceive were observed in a prospective study according to the following protocol: All the couples supplied general information about the man, the woman and the couple, sent at the beginning of the study; the woman filled in a temperature curve chart for each menstrual cycle and sent it monthly; when pregnancy occurred, the couple supplied details of its development and outcome, after examination at birth by a midwife or a paediatrician. All the study subjects were those who responded to the notice in the newspapers and who were asked to collect data on a standardised form. The temperature curves were all analysed by the same person. Even if pregnancy occurred however, this person knew nothing of its outcome. An important parameter of the analysis was the duration of the hypother- mic phase beginning on the first day of the menses and ending on day 0, the last low point before the temperature rise [19]. The analysis of all the events occurring during the cycles was performed according to the data available quoted on the temperature sheets, especially the number of occasions when intercourse took place which, for the conception cycle, have been taken into account up to day 21 of the cycle. Other information was collected by means of questionnaires filled in by the couples themselves. The usual cycles of any woman have been classified as ‘irregular’ when she reported common variations of four days or more around the average length. The number of cases studied was made up of 1968 couples who agreed to participate in the study. At the time the analysis was undertaken, 1075 (54.6%) had established a pregnancy, with a known outcome in 1035 cases (96.3%). The latter resulted in 119 spontaneous abortions and 916 births, of which 22 infants were born with malforma- tions.

The statistical methods used were Student’s r-test for the comparison of two means on small samples and Kruskall and Wallis’ non-parametric test for the comparison of two means in the case of small samples with different variances [18]. The &i-square test was used for the comparison of proportions, and the comparison of correlation coefficients with zero to test the relationship between quantitative variables. Relative risk calculations and their confidence intervals were computed according to Fleiss [4]. A linear stepwise discriminant analysis was performed to compare the two groups while taking into account all possible confounding variables

WI.

319

TABLE I

Details of 22 cases of congenital malformations

Type of malformation Number

Major Malformation of the hands or feet Malformation of the heart Femoral agenesis Malformation of the kidney

Sub. total Minor Malformation of the hip Synodactyly Dystrophy of the thorax Angioma and bilateral hydrocele Naevus Dermic sinus

Sub. total

9 2 2 1

14

3 1 1 1 1 1

- 8

Results

A description of all the observed malformations is reported in Table I. Their total incidence is 2.4%, of which major malformations account for 1.5% and minor ones for 0.9%. All the subsequent analysis was performed first by including all the malformed infants and then by including only major malformations.

The main conception cycle characteristics according to the outcome of pregnancy can be found in Table II. The hypothermic phase is significantly longer in the case of congenital malformation. Moreover, in three cases of malformation, the hypother-

TABLE II

Main conception cycle characteristics according to the outcome of pregnancy (X + SD. or %)

Malformed Not malformed Signifi- (n=22) (n =894) cance

Duration of menstrual period (days) Duration of the hypothermic phase (days) Hypothermic phase 2 24 days (W) Temperature DO ’ (“C) Duration of temperature rise. (days) Temperature rise c 5 days (46) Coital frequency during cycle

’ Last day of the hypothermic phase. b By Student’s r-test.

5.3* 1.5 24.0 ic 10.0 53% 35.7* 1.6

3.7* 1.9 82% 14.3* 8.2

c By Kruskall and Wallis test. d By &i-square test.

4.9* 1.3 ns. b 18.1 k 5.7 P < 0.01 c 13% P<O.OOld 35.9*1.7 as. b

3.0* 1.3 P < 0.05 b 96% P < 0.05 d 12.7k7.5 n.s. b

320

TABLE III

Women’s age, habitual and preconception cycle characteristics according to the outcome of pregnancy (X f SD. or ‘%)

Malformed Not malformed Significance

Age at menarche (years) Age (years) Habitual duration of cycle (days) Women with irregular cycles (W) Duration of hypothermic phase in

preconception cycle (days) Hypothermic phase in preconception

cycle 2 24 days (W) Duration of temperature rise

in preconception cycle (days)

13.4* 1.5 25.7k3.0 32.9 f 5.5 77% 21.7 f 8.3

18%

2.6 f 1.6

12.8 f 1.3 26.6 f 3.7 30.1+ 3.8 40% 17.6 f 5.2

10%

2.8f1.2

P < 0.05 a n.s. a P i 0.01 a P<O.OOlb P < 0.05 a

n.s. b

n.s. a

a By Student’s r-test. b By &i-square test.

mic phase lasted 41 days or more, which has never been observed in normal births. The duration of temperature rise is longer in the case of congenital malformation. The hypothermic phase in the cycle preceding the conception cycle is also longer in mothers who gave birth to malformed infants (Table III). Furthermore, these women had, on average, their first menstrual period at a later age and their menstrual cycles were often longer and more irregular. Yet there is no difference in age between the mothers of malformed infants and the mothers of infants without malformations.

Table IV shows the risks and relative risks of malformation according to different factors. The greatest risks were linked with the duration of the hypothermic phase in the conception cycle and habitual irregularity of the menstrual cycle. The risk of congenital malformation was 9.5% when the hypothermic phase lasted 24 days or more, versus 1.3% for shorter lengths of time (P -C 0.001). A detailed analysis of the data showed that the risk of malformation rose from 1.4% when the menstrual cycles were regular and the hypothermic phase of the conception cycle lasted less than 24 days, to 14.3% (relative risk = 10.2) when the menstrual cycles were irregular and the hypothermic phase of the conception cycle lasted 24 days or more. When the temperature rise in the conception cycle exceeded 5 days, the risk of malformation was multiplied by 4.7.

Except for the latter, the different factors under study were all positively linked together significantly (Table V). For these factors (except duration of temperature rise), a stepwise discriminant analysis was made between the malformed infants and those born without malformation in order to single out those factors which played a part concerning the risk of malformation. The only variable that differed signifi- cantly between the two groups was the duration of the hypothermic phase in the conception cycle. The coefficients of the other variables did not differ significantly from zero. When the hypothermic phase and the temperature rise were both abnormally long, the incidence of malformations was 20% (relative risk 23.4), but this observation was made on a group of only 5 infants. Although the file of each

321

TABLE IV

Risk of malformation and relative risk (95% confidence interval) according to the factors under study

Factor Risk of malformation (number at risk)

Relative risk (95% conf. int.)

Age at menarche 6 14 years

> 14 years

2.2% 1.0

(639) 5.0% 2.3 (0.8-7.2)

(272)

Regular cycles yes

no

1.0% 1.0

(529) 4.6% 4.8 (1.9-13.8)

(370)

Habitual duration of cycle < 35 days

> 35 days

1.4% 1.0

(714) 5.8% 4.1 (1.4-12.7)

(207)

Duration of hypothermic phase in preconception cycle

< 24 days

> 24 days

2.2% 1.0

(455) 4.1% 1.9 (0.4-9.2)

(51)

Duration of hypothermic phase in conception cycle

< 24 days

2 24 days

1.3% 1.0

(614) 9.5% 7.3 (3.0-21.1)

(95)

Temperature rise in conception cycle

Q 5 days

> 5 days

2.1% 1.0

(655) 10.0% 4.7 (1.4-18.8)

(30)

TABLE V

Correlation coefficient values between the different factors under study

Regularity Habitual du- Duration Duration Duration of cycle ration of hypothermic hypothermic temp. (regular = 0, cycle phase in phase in rise (irregular = 1) preconcep. concep.

Age at menarche 0.09 *

Regularity of cycle (regular = 0, irregular = 1) Habitual duration of cycle

Duration of hypothermic phase in preconception cycle Duration of hypothermic phase in conception cycle ‘P 005 **p <.; < 0.01; *** P < 0.001.

0.09

cycle

0.13

cycle

0.11 0.02 * ** 0.38 0.27 *** ***

0.34 **t

** n.s. 0.26 0.04 *** n.s. 0.44 0.06 *** n.s. 0.54 - 0.04 *** n.s.

- 0.07 n.s.

322

malformed infant was carefully studied, it was impossible to find any one particular type of malformation that could be linked to the duration of the hypothermic phase in the conception cycle.

All the results remained unchanged when studying only the major malformations group (first menstrual period at 13.7 years, usual cycle length 32.9 days, hypothermic phase during preconception cycle 23.3 days and during conception cycle 24.3 days, temperature rise during conception cycle 3.5 days).

Discussion

The number of cases of congenital malformations observed at birth in this study was more or less the same as that analysed in other studies undertaken in France during this period [16]. The number of malformed infants under study (only 22) was too small to enable a separate analysis of different types of malformations to be made. It is not known if either the parents or any sibs had the same malformations. The observed results were similar for malformations classified as major and for all malformations. The duration of the hypothermic phase in the conception cycle, calculated by analysing the temperature curves (18.4 days for the whole group of cycles, 17.5 days once the abnormally long cycles had been excluded) corresponded to other published results using the same methods [13,17,20,21].

Two completely independent factors seem to increase significantly the risk of congenital malformation at birth: the extended length of the hypothermic phase in the conception cycle and that of the temperature rise in the same cycle. The first of these factors, and the most important, is more often observed in women who had their first menstrual period at a later age and whose cycles are long and irregular. The extended length of the hypothermic phase in the conception cycle seems to be a usual characteristic of such women as suggested by the very high correlation coefficient (r = 0.54, P < 0.001) between the durations of the hypothermic phases in the conception and preconception cycles. The age of the women in the different groups does not seem to have any bearing on the results observed, as it was the same whether the infant was malformed or not.

Even though the sample studied is not representative of the general population, the fact that this is a prospective study and that the temperature curves have been analysed without any knowledge of the outcome of pregnancy, suggests that our findings have not been biassed. All the uncertainties associated with the analysis of temperature curves are comparable in both groups, their only effect being to reduce the power of the comparison done.

Since a relation exists between the length of the hypothermic phase plus the temperature rise in the conception cycle and the risk of malformation, it seems that the process responsible could be intrafollicular overripeness of the oocyte resulting from a longer hypothermic phase and a slow temperature rise. Such a process could act by itself, or in relation with paternal gametic factors. This result agrees with experimental studies made on animals, mostly rats and rabbits [l-3,5,6]. In particu- lar, it has been possible to show, in the latter, that a 60 hour delay in ovulation could

323

result not only in the early expulsion of an unimplanted ovum but also in the abnormal subsequent development of implanted ova. In humans the studies made by Guerrero [7,8] have brought to light a relation between intrafollicular aging of the oocyte and the risk of early spontaneous abortion. However in the present study, the only observed difference between spontaneous abortions and normal births is the higher incidence of women with irregular cycles in the former group (62% vs 40%, P < 0.001). Although chromosomal analysis of first trimester spontaneous abortions should be a more sensitive way to test our underlying hypothesis, this is not possible in such a prospective epidemiological study. The pathological studies undertaken by Hertig [9] and by Philippe [15] clearly indicate that ovular pathology depends in part on the ovulation date in the cycle. In Hertig’s series [9], there was one abnormal egg out of 13 (8%) when ovulation occurred on or before day 14 of the cycle and 12 abnormal eggs out of 21 (57%) when ovulation took place after day 15. In Philippe’s series [15] these figures are respectively 2 cases out of 8 (25%) and 15 cases out of 18 (83%). Jongbloet [ll] advanced similar suppositions indirectly when he analysed the seasonal variations in the incidence of the birth of malformed infants (the numbers in the present study are too small to allow confirmation of this result). More indirect evidence was supplied by Papiernik et al. [14] in a retrospective study of 289 conception cycles where a significant relation was found between the length of the hypothermic phase and the proportion of infants affected by intrauterine growth retardation.

The results of the present study suggest that there are some women whose general reproductive characteristics include a first menstrual period at a later age, long and irregular menstrual cycles, and an extended hypothermic phase, and that these women are much more likely to give birth to congenitally malformed infants. Confirmation of these results by a further study incorporating a much larger number of cases would determine if there was a possibility of pinpointing those women particularly at risk of giving birth to malformed infants, and for whom preventive measures could be envisaged.

Acknowledgements

The authors wish to give special thanks to Jeanine Wattiaux and Marie-Christine Le Hene for their participation in all phases in this study. This work was supported by a grant from the Mutuelle G&n&ale de 1’Education Nationale (M.G.E.N.), Paris, France.

References

1 Bomsel-Helmreich, 0. (1976): The aging of gametes, heteroploidy, and embryonic death. Int. J. Gynaecol. Obstet. 14, 98-104.

2 Butcher, R.L. and Fugo, N.W. (1967): Overripeness and the mammalian ova. II. Delayed ovulation and chromosome anomalies. Fertil. Steril. 18, 297-302.

324

3 Butcher, R.L. and Page, R.D. (1981): Role of the aging ovary in cessation of reproduction. In: Dynamics of Ovarian Function, pp. 253-271. Editors: N.B. Schwartz and M. Hunzicker. Raven Press, New York.

4 FIeiss, J.L. (1981): Statistical Methods for Rates and Proportions, pp. 57-82. John Wiley and Sons, New York.

5 Fugo, N.W. and Butcher, R.L. (1966): Overripeness and the mammalian ova. I. Overripeness and early embryonic development. Fertil. Steril. 17, 804-814.

6 Fugo, N.W. and Butcher, R.L. (1971): Effects of prolonged estrus cycles on reproduction in aged rats. Fertil. Steril. 22, 98-101.

7 Guerrero, R.V. and Larotot, C.A. (1970): Aging of fertilizing gametes and spontaneous abortions. Am. J. Obstet. Gynecol. 107, 263-267.

8 Guerrero, R.V. and Rojas, 0.1. (1975): Spontaneous abortion and aging of human ova and sperma- tozoa. N. Et-&. J. Med. 293, 573-575.

9 Hertig, A.T. (1967): The overall problem in man. In: Comparative Aspects of Reproductive Failure, pp. 11-41. Editor: K. Benirschke. Springer-Verlag, New York.

10 Iffy, L. and Wingate, M.B. (1970): Risks of rhythm method of birth control. J. Reprod. Med. 5, 96-102.

11 Jongbloet, P. (1975): The effects of preovulatory overripeness of human eggs on development. In: Aging Gametes, pp. 300-329. Editor: R.J. Blandau. Karger, Base].

12 Kleinbaum, D.G. and Kupper, L.L. (1978): Applied Regression Analysis and Other Multivariable Methods, pp. 431-433. Duxbury Press, North Scituate, MA.

13 Matsumoto, S., Nogami, Y. and Ohkuri, S. (1962): Statistical studies on menstruation; a criticism on the definition of normal menstruation. Gunma J. Med. Sci. 11, 294-318.

14 Papiernik, E., Spira, A., Bomsel-Helmreich, 0. and Lebel, S. (1979): Ovarian overripeness and intrauterine growth retardation. Lancet 2, 1025-1026.

15 Philippe, E. (1980): Les nidations precocement abortives. A propos de 18 oeufs previlleux et de 24 oeufs villeux jeunes. J. Gynecol. Obstet. Biol. Reprod. 9, 513-521.

16 Roux, C., Migne, G., Mulliez, N. and Youssef, S. (1982): Frequence des malformations a la naissance. Etude dune matemite parisienne pendant cinq ans. J. Gyndcol. Obsttt. Biol. Reprod. 11, 215-226.

17 Saito, M., Yazawa, K., Hashiguchi, A., Kumasaka, T., Nishi, N. and Kato, K. (1972): Time of ovulation and prolonged pregnancy. Am. J. Obstet. Gynecol. 112, 31-38.

18 Schwartz, D. (1963): Methodes Statistiques a I’Usage des Medicins et des Biologistes, pp. 255-256. Flammarion, Paris.

19 Vincent, B. (1964): Atlas de courbes thermiques, 110 pp. Centre de Documentation et dInformation Conjugale, Names.

20 Vollmann, R. (1977): The menstrual cycle. In: Major Problems in Obstetrics and Gynecology, Vol. 7, pp. 73-188. W.B. Saunders, Philadelphia.

21 W.H.O. (1967): Biology of fertility control by periodic abstinence. Report of a WHO Scientific Group. WHO Technical Report Series, No. 360, 20 pp.