Long-term sexual co-habitation offers no protection from hypertensive disease of pregnancy

Glen H. Hall1,4, Wendy L. Noble1, Stephen W. Lindow2 and Ewan A. Masson3

1 Hull Maternity Hospital and Departments of 2 Obstetrics and Gynaecology and 3 Medicine, University of Hull, Hull, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Long-term sexual co-habitation and previous pregnancies are thought to protect against the development of hypertensive disease of pregnancy. In order to test the hypothesis that pregnancies conceived after prolonged exposure to the partner's spermatozoa have reduced rates of hypertensive disease this study examined the outcomes of pregnancies of women who conceived by donor insemination as compared with women who conceived after IVF with partner's spermatozoa. This was a retrospective cohort study of 218 women attending an IVF clinic, 45 of whom conceived by donor insemination and 173 of whom conceived by partner's spermatozoa. Cases were identified from the IVF unit and data were extracted from patients' notes by blinded observers. Results showed no difference between the groups, with 22% of women who conceived with donor spermatozoa and 24% who conceived with partner spermatozoa developing some form of hypertensive disease of pregnancy. Insemination by partner's spermatozoa was not associated with a reduction of hypertensive disease and neither was donor spermatozoa associated with an increased incidence.

Key words: donor insemination/hypertension of pregnancy/IVF/pre-eclampsia


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Hypertensive disease of pregnancy is one of the most serious complications, affecting up to 30% of all pregnancies (Knutzen and Davey, 1997Go). It is the second most common cause of maternal mortality in the 1994–1996 triennial report on maternal mortality in the UK (Department of Health et al., 1998Go). The aetiology remains poorly understood, thus allowing little prospect of strategies for prevention. It has been known for some time that it is a disease which is commonest in first pregnancies and also in further pregnancies with a change of paternity (Trupin et al., 1996Go).

An inverse relationship has been suggested between the duration of sexual co-habitation and the incidence of pregnancy-induced hypertension (PIH) (Robillard et al., 1994Go), with rates of PIH in excess of 30% for under 4 months sexual co-habitation and rates of below 10% for over 12 months. Barrier contraception has also been linked to an increased rate of pre-eclampsia (Klonoff-Cohen et al., 1989Go). These observations have led to the development of various theories of immunological responses in the mother against the antigens in the partner's semen, which would seem to protect against the development of hypertensive disease of pregnancy.

This study aims to compare the incidence of hypertensive disease of pregnancy in women who conceived by assisted fertility techniques with their partner's spermatozoa to those who conceived with donor spermatozoa, to which they were completely immunologically naive. By selecting the population of women attending the local IVF clinic, most of the women presenting for treatment for subfertility would have been exposed to their partner's semen for a duration in excess of 3 years. They would also mostly be nulliparous and thus at greater risk of gestational hypertensive disease. The hypothesis was that those women conceiving by IVF with their partner's spermatozoa would have been exposed to its antigens for some time and should experience a protective effect, whilst those conceiving by donor spermatozoa would be immunologically naive to it and should experience increased rates of hypertensive disease.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A power calculation was carried out to calculate the sample size needed to achieve a decreased incidence from 30% to 10% in the rates of hypertensive disease, with an 80% power at a significance level of P = 0.05 [as based on the data from Robillard (Robillard et al., 1994Go)]. A total of 160 patients conceiving by partner and 40 conceiving by donor give a sufficiently large sample size. Consent was obtained from the Human Fertilisation and Embryology Authority (HFEA) to access patients notes at the infertility clinic at Princess Royal Hospital in Hull, and the study was approved by the local ethics committee.

Data were collected from the IVF unit on all the pregnancies achieved between 1991 and 1998 who booked for maternity care in one of the two local maternity units. Pregnancies from out of the region were excluded. The duration with partner (where documented), the method of conception and the origin of paternity (including previous number of attempts with each donor) were recorded and anonymized by the researcher (W.L.N.). The maternity notes were then traced at the two local maternity hospitals (Hull Maternity Hospital, Hull and Castle Hill Hospital, Cottingham, UK). When data were collected from the maternity notes, the researcher (G.H.H.) was blinded to the method of conception and origin of paternity. Details of obstetric outcome, antenatal, intrapartum and postnatal blood pressure readings, proteinuria and other complications were all recorded.

The classification of hypertensive diseases of pregnancy first proposed by Davey and MacGillivray (Davey and MacGillivray, 1988Go) was used. These criteria were applied strictly to classify the hypertensive diseases. The definition of proteinuria required either 300 mg per 24 h in a 24 h urine specimen or two mid-stream specimens more than 4 h apart with 2+ proteinuria on dipstick testing or 1+ by dipstick testing with measured specific gravity <1.03 and pH <8. Women were only classified as having proteinuria when full documentation allowed classification as described.

All results were analysed using Statistics Package for Social Sciences (SPSS Professional Statistics, SPSS Inc., Illinois, USA) and the {chi}2 test was used to show the level of significance of the rates of hypertensive disease in partner and donor sperm groups.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Both IVF unit and maternity unit data were available for 228 pregnancies. During the course of their pregnancies 10 of these women moved out of the area and were excluded from analysis because outcome data were not available.

There were no differences in the demographics between the partner and donor groups as outlined in Table IGo. Apart from one spontaneous conception the other 227 pregnancies were conceived using techniques that involved washed spermatozoa.


View this table:
[in this window]
[in a new window]
 
Table I. General characteristics of patients studied
 
In the 218 analysed pregnancies there were a total of 163 singletons, 49 twins and six triplets. The distribution of the multiple pregnancies was 45 twin pregnancies and six triplets in the partner group and five twins in the donor group. The mode of delivery was 51% spontaneous vaginal delivery, 9.5% operative vaginal delivery and 39.5% Caesarean section.

The frequency of hypertensive disorders is outlined in Table IIGo, and it is clear that there were no differences between partner and donor sperm groups. A subgroup analysis confined to nulliparous women excluded any effects of multiparity and again demonstrated no differences.


View this table:
[in this window]
[in a new window]
 
Table II. Classification of hypertensive disease by paternity
 
Table IIIGo shows results broken down in greater detail. Analysis with the {chi}2 test indicated no significant difference between either group.


View this table:
[in this window]
[in a new window]
 
Table III. Analysis of outcome
 
These data confirm there was no statistically significant association between these groups and the rates of hypertensive disease.

Multiple logistic regression analysis was performed, addressing all recorded variables which may have relevance, including: age, booking weight, booking height, booking systolic blood pressure, booking diastolic blood pressure, number of fetuses, sex of fetuses, parity, smoking, origin of paternity, early pregnancy loss rate, aspirin use, triple test result, use of antihypertensives and incidence of bleeding in pregnancy. The only variables which predicted a risk of hypertensive disease of pregnancy in the model were multiple pregnancy (P < 0.005) and raised booking systolic blood pressure (P < 0.0005).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
These results do not support the hypothesis that immunological exposure to semen is associated with a reduced risk of hypertensive disease. The study design, including power calculations, blinding of the researchers, patients managed in only two units, certainty of origin of paternity (avoiding patient reporting bias) and the very precise and rigorous definition of hypertensive disease of pregnancy (Davey and MacGillivray, 1988Go) make this methodologically sound.

In 1977 it was reported (Marti and Herrmann, 1977Go) that there was an increased incidence of mild to moderate hypertensive disease in a small group of primigravid women who had less preconceptional exposure to spermatozoa than did their control group. They postulated that sperm histocompatibility antigens induced an immunological tolerance that prevented an immune reaction. Some controversy exists over the expression of human leukocyte antigens (HLA) on spermatozoa, although it seems that HLA antigens are present in the seminal plasma fraction of the ejaculate (Singal and Berry, 1972Go) and that these may be adsorbed onto the surface of the sperm plasma membrane.

Since then much research has focused on the immunological theories. Findings have been discrepant in all areas of this subject and controversy prevails. The use of barrier contraception prevents maternal exposure to semen and may be associated with an increased risk of hypertensive disease. This is suggested by a case controlled study of 115 women with pre-eclampsia, which showed a 2.37-fold increased risk of pre-eclampsia among users of barrier contraception (Klonoff-Cohen et al., 1989Go). However combined data on 13 914 pregnancies from two prospective studies of pre-eclampsia rates showed no association between pre-eclampsia rates and barrier contraception (Mills et al., 1991Go).

An inverse relationship between the self-reported duration of sexual co-habitation before conception and the incidence of PIH has been reported in a population of Caribbean women (Robillard et al., 1994Go). In this study the incidence of PIH in pregnancies where the duration of co-habitation was under 4 months was 32%, and the incidence when the duration of co-habitation was over 12 months was 3%. This would seem to support the theory that immunological exposure to semen protects against PIH.

The issue of hypertensive disease in pregnancies from donor insemination has been addressed previously. The outcome of 584 pregnancies resulting from artificial insemination by donor was analysed retrospectively by an Australian group, who described an overall incidence of pre-eclampsia of 9.3% and claimed this was substantially increased over the expected (Need et al., 1983Go), although they went on to describe a background pre-eclampsia rate of 7.3% in their institution. A Canadian group looked at a total cohort of 81 patients, 44 with partner insemination and 37 with donor (Smith et al., 1997Go). They reported three cases of pre-eclampsia in the partner group and nine in the donor. A series of 44 pregnancies following donor insemination has been reported (Perkins, 1993Go) and no difference was found in hypertensive complications from expected rates in their institution.

A group from Finland (Laivuori et al., 1998Go) has studied a cohort of 123 pregnancies and found no increased rate of hypertensive disease with donor spermatozoa. However their report was confined to antenatal gestational proteinuric hypertension, which excluded a significant proportion of hypertensive disease of pregnancy. The current research adds to the data from these previous studies by including a larger cohort of patients than some and by methodologically addressing the question from another angle. Like with like was compared by using subfertility patients as controls, to exclude any confounding influence of subfertility on rates of hypertensive disease.

The immunological theories of hypertensive disease are also supported by the finding that oocyte donation may be associated with increased rates of the disease. A rate of 25.7% gestational hypertension in 35 recipients of oocyte donation versus 4.2% in a control group of 95 IVF patients has been reported in one series (Hendler et al., 1997Go). A group in the UK has studied a cohort of 33 women conceiving with donor spermatozoa, 27 with donor eggs and 12 women with embryo donation, and compared them with a control group of matched patients conceiving spontaneously (Salha et al., 1999Go). They found the incidence of both PIH and pre-eclampsia to be increased in the group with donated gametes (12.5% and 18.1% respectively), but the background rates of PIH of 2.8% and pre-eclampsia of 1.4% were much lower than that found in other studies. Another study of 1552 donor insemination pregnancies has reported an odds ratio of 1.4 for the development of pre-eclampsia when compared to normally conceived pregnancies, although the details of the classification of the data are not given clearly (Hoy et al., 1999Go).

The total of 218 completed pregnancies for which full maternity data could be examined and analysed gives the present study a cohort of patients greater than required by power calculations. The sample size needed was calculated to show a 20% reduction (from 30 to 10%) in the incidence of hypertensive disease as found in a cohort of Caribbean women (Robillard et al., 1994Go). In South Africa it was estimated that by using the same strict definitions of hypertensive disease the expected rate was 30% (Davey and MacGillivray, 1988Go). In the current study the total overall rate of hypertensive disease was 24%. A much larger sample size would be needed to exclude differences of a smaller magnitude in rates of hypertensive disease between donor and partner groups. The age distribution of patients, gestation at delivery, high percentage of nulliparous women and occurrence of multiple pregnancy was as would be expected for a cohort of women attending a fertility service. The mode of delivery was also in keeping with expectations, as the threshold to deliver by Caesarean section is recognized to be lower in women conceiving by assisted reproduction (Söderström-Anttila et al., 1998Go).

Hypertensive disease was observed in 23.9% of pregnancies in the current study, which is within keeping with that found in a cohort of South African women, with 40% of these cases being non-proteinuric antenatal gestational hypertension (Davey and MacGillivray, 1988Go). The distribution of cases through all categories of the classification of hypertensive disorders, from mild to severe, was uniform between donor and partner groups. Thus, within the confines of the power calculation, if any difference between donor and partner insemination existed it would have been detected.

This study does not show that partner insemination is associated with reduced rates of hypertensive disease of pregnancy. If pre-eclampsia has an immunological basis it is not reduced by previous exposure to the fertilizing spermatozoa. This study therefore challenges the evolving theory of sexual co-habitation.


    Acknowledgments
 
We would like to thank Anne Robertson and the staff of the IVF unit in Hull for their kind help.


    Notes
 
4 To whom correspondence should be addressed at: Centre for Magnetic Resonance Investigation, Hull Royal Infirmary, Anlaby Road, Hull, HU3 2JZ, UK. E-mail: g.h.hall{at}medschool.hull.ac.uk Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Davey, D.A. and MacGillivray, I. (1988) The classification and definition of the hypertensive disorders of pregnancy. Am. J. Obstet. Gynecol., 158,892–898.[ISI][Medline]

Department of Health, Welsh Office, Scottish Office Department of Health and Department of Health and Social Services, N.I. (1998) Why Mothers Die. Report on confidential enquiries into maternal deaths in the United Kingdom 1994–1996. HMSO, London.

Hendler, M., Dulitzky, D., Soriano, D. et al. (1997) Pregnancy outcome following oocyte donation. Am. J. Obstet. Gynecol., 176,S133.

Hoy, J., Venn, A., Halliday, J. et al. (1999) Perinatal and obstetric outcomes of donor insemination using cryopreserved semen in Victoria, Australia. Hum. Reprod., 14, 1760–1764.[Abstract/Free Full Text]

Klonoff-Cohen, H.S., Savitz, D.A., Cefalo, R.C. and McCann, M.F. (1989) An epidemiologic study of contraception and pre-eclampsia. J. Am. Med. Assoc., 262,3143–3147.[Abstract]

Knutzen, V.K. and Davey, D.A. (1997) Hypertension in pregnancy: perinatal mortality and causes of fetal death. S. Afr. Med. J., 51,675.

Laivuori, H.M., Hovatta, O.L.L. and Ylikorkala, R.O. (1998) Lack of previous exposure to paternal antigens does not predispose to hypertensive pregnancy complications. Hypertension in Pregnancy, 17,291–295.[ISI]

Marti, J.J. and Herrmann, U. (1977) Immunogestosis: a new etiologic concept of `essential' EPH gestosis, with special consideration of the primigravid patient. Am. J. Obstet. Gynecol., 128,489–493.[ISI][Medline]

Mills, J.L., Klebanoff, M.A., Graubard, B.I. et al. (1991) Barrier contraceptive methods and pre-eclampsia. J. Am. Med. Assoc., 265,70–73.[Abstract]

Need, J.A., Bell, B., Meffin, E. and Jones, R.W. (1983) Pre-eclampsia in pregnancies from donor inseminations. J. Reprod. Immunol., 5,329–338.[ISI][Medline]

Perkins, R.P. (1993) Pregnancy following donor insemination: implications for pre-eclampsia. J. Materno-Fetal Med., 2,52–54.

Robillard, P.Y., Hulsey, T.C., Perianin, J. et al. (1994) Association of pregnancy-induced hypertension with duration of sexual co-habitation before conception. Lancet, 344,973–975.[ISI][Medline]

Salha, O., Sharma, V., Dada, T. et al. (1999) The influence of donated gametes on the incidence of hypertensive disease of pregnancy. Hum. Reprod., 14, 2268–2273.[Abstract/Free Full Text]

Singal, D.P. and Berry, R. (1972) Soluble HLA antigens. Localization in the human seminal plasma fraction. Transplantation, 13,441–442.[ISI][Medline]

Smith, G.N., Walker, M., Tessier, J. and Millar, K.G. (1997) Increased incidence of pre-eclampsia in women conceiving by intrauterine insemination with donor versus partner sperm for treatment of primary infertility. Am. J. Obstet. Gynecol., 177,455–458.[ISI][Medline]

Söderström-Anttila, V., Tiitinen, A., Foudial, T. and Hovatta, O. (1998) Obstetric and perinatal outcome after oocyte donation: comparison with in-vitro fertilization pregnancies. Hum. Reprod., 13, 483–490.[ISI][Medline]

Trupin, L.S., Simon, L.P. and Eskenazi, B. (1996) Change in paternity: a risk factor for pre-eclampsia in multiparas. Epidemiology, 7,240–244.[ISI][Medline]

Submitted on May 15, 2000; accepted on October 23, 2000.