Regional differences in waiting time to pregnancy: pregnancy-based surveys from Denmark, France, Germany, Italy and Sweden

S. Juul1,4, W. Karmaus2, J. Olsen3 and The European Infertility Subfecundity Study Group,*

1 Department of Epidemiology and Social Medicine, Vennelyst Boulevard 6, DK-8000 Aarhus C, Denmark, 2 Nordig Institute for Health Research and Prevention, Altc Kollaustrasse 32a, D-Z000 Hamburg 54, Germany and 3 Danish Epidemiology Science Centre at Department of Epidemiology and Social Medicine, Vennelyst Boulevard 6, DK-8000 Aarhaus, Denmark


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The objective of this study was examine geographical variation in couple fecundity in Europe. The study was based upon all recently pregnant (or still pregnant) women within well-defined geographical areas in Europe (Denmark, Germany, Italy, Sweden and France) at a given time period in 1992. Altogether, 4035 women responded to a highly structured questionnaire. Highest fecundity was found in Southern Italy and Northern Sweden; lowest fecundity was seen in data from the East German centre. Approximately 16% of the study population had a waiting time of more than 12 months to become pregnant. Most of the pregnancies were planned (64%) and approximately 14% were the result of contraceptive failures. The study shows that smoking, body mass index, age and parity did not explain the differences in fecundity found between the centres. Regional differences in fecundity exist and the causes may be genetic or due to variations in behavioural and environmental exposures.

Key words: environment/fecundity/infertility/pregnancy


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
There is evidence to suggest a decline in sperm quality over time (Carlsen et al., 1992Go), but this finding may be biased by geographical differences, differential selection, changes in the measuring or changes in other circumstances (Olsen, 1994Go). If the decline in sperm quality is real, it should have an impact on couple fecundity (Bonde et al., 1998Go). Unfortunately, there is surprisingly little evidence about trends in couple fecundity over time. There is also very little known about geographical differences and, if they exist, they should be taken into consideration when studying secular changes in semen quality or concentration.

In Europe, several investigations have provided a wide range of estimates of the prevalence of infertility, from 6% to 20% (Rachootin and Olsen, 1981Go, 1982Go, 1983Go; Hull et al., 1985Go; Rantala and Koskilles, 1986Go; Page, 1989Go; Templeton et al., 1990Go; Thonneau and Spira, 1990Go; Thonneau et al., 1991Go; Wagner and Stephenson, 1992Go). However, the applied methods vary and results are thus not comparable. In order to obtain comparable data, a European study group was established in 1990 (ESIS, European Studies of Infertility and Subfecundity) which implemented comparable studies on infertility and subfecundity in different European countries.

ESIS conducted both population-based and pregnancy-based surveys, using highly structured questionnaires. In this report we present findings on time to pregnancy (TTP) from the studies based upon pregnant women.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Concepts and definitions
Terms to describe fecundity are used differently by demographers, clinicians and epidemiologists. In the demographic context, the term fertility refers to the actual reproductive behaviour as expressed by the number of childbirths; clinicians often use the term to describe the biological ability to conceive, or the ability to produce live offspring. We use the term fecundity for the biological ability to give birth or to achieve a recognized pregnancy, subfecundity for evidence of decreased fecundity, and sterility for inability to conceive, e.g. due to hysterectomy or sterilization.

Fecundity is estimated for a population by the time to pregnancy (TTP): the time span, or the number of menstrual cycles exposed to unprotected intercourse, until conception (Baird et al., 1986Go). To determine TTP, the starting date of unprotected cohabitation and the date of conception are required. A TTP of one year or more is often used as a measure of subfecundity (Rowe and Farley, 1988Go).

Questionnaire and interview
A questionnaire was developed by a collective process during several cycles of pilot testing in most of the participating countries. The master questionnaire was in English, and ambiguity in the wording was checked by independent translations to other European languages and back-translation to English. The questionnaire was designed for self-administration, but in two centres in Italy it was administered by female interviewers.

The main contents of the questionnaire were: health and education, reproductive history, starting date and TTP for the current or recent pregnancy, exposures around the starting date, and intent to become pregnant (planning). A copy of the questionnaire is available on request (S.Juul).

The questionnaire was structured according to the past reproductive experience in order to better define the starting time (start of unprotected intercourse). The TTP question was phrased: `How long was it from that `starting time' until you became pregnant? (the date you became pregnant is the date you conceived)'. The women's response could be expressed in weeks, months or years.

Interviews/responding typically took 15–25 min. The unit of analysis was a pregnancy, and the mothers were used as informants.

Population
The surveys were conducted in five European countries between February 1992 and December 1992. The target group consisted of women who had just given birth to a live child or women who were pregnant. In Sweden and France, only part of the data were collected during pregnancy.

The women were approached either immediately after having given birth at a hospital or birth clinic or at a visit to an antenatal care centre after 20 weeks of gestation. The institutions were selected because they served geographically well-defined populations, and all pregnant women during the period of data collection were asked to participate. The sampling method excluded women who did not attend prenatal services or who gave birth at home, but <5% were excluded for this reason in any centre.

Statistical methods
TTP was recorded as weeks and/or months and/or years, e.g. a TTP of 3.5 months was recorded as 3 months + 2 weeks. For short TTP, up to three menstrual cycles to pregnancy were also recorded, as pilot experience showed that women recalled longer waiting times as months rather than as menstrual cycles. For this reason, the data in this paper were analysed according to passage of time. By using time as the outcome, TTP measures cover not only sexual activity and ability to conceive and sustain a conception, but also menstrual irregularities. For pregnancies occurring despite contraception, TTP was considered to be undefined and thus missing.

In this paper the TTP distribution was described as the cumulative proportion of pregnancies occurring up to a given waiting time.

Comparisons of TTP in different centres were also based upon Cox regression, using continuous TTP (not cycles) as the outcome. Observations longer than 12 months were censored. By design, there were no censorings within the first 12 months, and TTP was measured in continuous time, and it was therefore not considered appropriate to use a discrete time model. The term `pregnancy ratio' was used for the Cox regression coefficients, which should not be considered as fecundability ratios, as there were no data on TTP not leading to pregnancy. With this reservation, the pregnancy ratio was considered to be that between incidence rates of pregnancy, the incidence rate being the reciprocal of TTP.

Coding and data entry were performed locally, but combined to an anonymous dataset, using SPSS. Analyses were performed using SPSS for Windows.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Sample size, participation rates and administration of the questionnaire by centre are detailed in Table IGo. The overall participation rate was 83%, varying from 66% in East Germany to 98% in the Rome area.


View this table:
[in this window]
[in a new window]
 
Table I. Description of the samples from the different European countries
 
The reproductive experiences of the participants are described in Table IIGo. The variations in age and in reproductive experience reflect the actual age distribution and reproductive pattern in the regions included. The highest parity was seen in Umeå, Sweden, and the lowest in Rome and Milan, Italy. Table IIGo also details the distribution of potential confounders for the geographical comparisons.


View this table:
[in this window]
[in a new window]
 
Table II. Reproductive history and potential confounders
 
Overall, 14% of the pregnancies were the result of contraceptive failures, with a high value of 21% in Milan/Bergamo and a low of 9% in Paris (Table IIIGo). Most (64%) of the pregnancies were described as planned by the women, ranging from 81% in Paris to 41% in Halle and Rostock (East Germany). Similar figures were found for those who expected (or who had just had) their first baby. Information on TTP was obtained for almost all of the pregnancies not due to contraceptive failure.


View this table:
[in this window]
[in a new window]
 
Table III. Main information about time to pregnancy
 
TTP was recorded with some digit preference, with excess recordings of 12 and 24 months, compared with the neighbour categories. Thus, the cumulative TTP distribution for <12 months and TTP for <=12 months is presented in Table IVGo. The average proportion of TTP >12 months was 16%, with a low of 10% in Southern Italy and a high of 24% in East Germany. The TTP distribution did not differ significantly in the first pregnancy compared with other pregnancies. For planned pregnancies, the average proportion of TTP >12 months was 13%, with the same relative distribution between regions. For unplanned pregnancies, the proportion was 24%.


View this table:
[in this window]
[in a new window]
 
Table IV. Cumulated distribution of time to pregnancy (percentages) by region
 
To adjust for confounders, the TTP distributions were compared in Table VGo, using Cox regression analysis. The significant differences in the TTP distributions remained after adjustment for regional differences in body mass, smoking, frequency of intercourse and sexually transmitted diseases. After these adjustments, the French centre had the longest TTP and Southern Italy the shortest.


View this table:
[in this window]
[in a new window]
 
Table V. Pregnancy ratios (PR) according to geographical region. Cox regression on time to pregnancy in months
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The study shows substantial variations in fecundity between the centres, as estimated by the TTP distribution. To our knowledge this is the first international study on fecundity which is based upon an unselected population of pregnant women or mothers, a common protocol, and the same questionnaire. Results based upon a less comparable design have shown better couple fecundity in Finland than in the UK (Joffe, 1996Go). In spite of our use of a common protocol, however, the differences may be due to bias, e.g. related to comparability of populations and quality of information.

Comparability of populations
In all sites, the sampling included almost all pregnancies leading to a live birth or a pregnancy exceeding a duration of 20 gestational weeks within the stated time periods. Participation rates varied considerably, and this may have introduced a bias of unknown direction and magnitude in the comparison of fecundity.

Pregnant women were selected late in pregnancy in order to avoid pregnancies leading to miscarriages, and women who aborted after recruitment were excluded from the sample according to the protocol. Since subfecundity correlates with abortion (Basso et al., 1997aGo), bias may have been introduced if not all abortions were excluded, though the bias would be small if present.

Of much more serious concern is the fact that the study relied upon a similar drive in pursuing a pregnancy attempt in all centres, since those who gave up are not included in the sample. We did, however, find similar results when restricting the study to pregnancy planners who are expected to provide better data on TTP and to have a more similar behaviour. Differences in the willingness to become pregnant hardly explain all the findings.

The locations included in the study are not representative of areas greater than regions, and the study provides no information on variation between countries.

Quality of information
The questionnaire was developed during several cycles of pilot testing, and the experience from the translation–back-translation process indicated that language differences give rise to few problems of comparability. However, cultural differences in pregnancy planning habits, use of contraceptive methods, etc. might lead to differences in the reporting of TTP, but this potential source of bias hardly explains the marked differences in fecundability.

The interviewers reported that the interviewees accepted being asked about reproductive and sexual experience, and they had the impression that the women provided honest answers. However, the tendency to digit preference indicates that the recall of TTP duration is not perfect at an individual level in spite of a short recall period. A previous study (Joffe et al., 1993Go) concluded that long-term recall of TTP duration is quite reliable, and we find no reason to believe that the quality of recall is different during pregnancy and shortly after delivery.

When the questionnaire was self-administered, some of the women had difficulties in finding the relevant questions.

The difference in the frequency of a TTP >12 months as a function of pregnancy planning (13% for planners and 24% for those who had unprotected sexual intercourse without planning a pregnancy) shows that TTP measures strongly depend upon how the key questions are phrased. A similar conclusion was reached in a study from USA (Marchbanks et al., 1989Go). These results warn against comparing data based upon different questionnaires, or questionnaires using different wordings of key questions.

In conclusion, interpretation of the TTP distribution in a pregnancy-based study is somewhat complex, since the pregnancy-based TTP distribution is conditional on a pregnancy actually occurring. On the other hand, geographical factors that potentially affect fecundity can be studied, since most determinants prolong TTP rather than cause sterility (perhaps with the exception of infections). The target population of a pregnancy-based study is easier to identify and to approach than the target population of a population-based study, and recall problems when reporting TTP are less since recall is short.

Fertility is a function of many factors, and fecundity plays a rather limited role in most countries with low fertility. We found low fecundity in areas with high fertility (East Germany) and high fecundity in areas with low fertility, such as Italy.

Many (36%) of the pregnancies (19% in France, 59% in East Germany) were not planned, and 14% of the pregnancies (from 9% in France to 21% in Northern Italy) occurred despite contraception. All these pregnancies had passed the gestational age for induced abortion which was available within the first 12 gestational weeks in all participating countries.

The study shows regional differences in couple fecundity which should be taken into consideration when studying how fecundity develops over time. In the search for environmental factors of importance for fecundity, such as hormonal disruptions acting in fetal life, geographical variation may give important clues and ecological studies related to environmental exposures are needed.

We have previously analysed data according to selected determinants (Basso et al., 1995Go, 1997bGo; Bisanti et al., 1996Go; Bolumar et al., 1996Go, 1997Go; Olsen et al., 1997Go), and the most important of these are mother's age, smoking habits and body mass index. Regional differences in these determinants did not explain the variation between the centres. Unfortunately, we probably have no good data on previous sexually transmitted infections or previous induced abortions. It is possible that these factors partly explain the geographical variation in TTP, but it is of course also possible that genetic factors play a role. The main aim of this study was to describe geographical variation, rather than to identify underlying reasons for the differences.

The study shows that differences in fecundity in Europe do occur, and resources should be devoted to finding out why. The use of simple survey methods may provide valuable and direct measures of couple fecundity, and monitoring TTP distribution over time for the same population may provide valuable data on major secular changes in couple fecundity. Great care should be taken when selecting proper data collection instruments, and uniform sampling criteria must be used.


    Notes
 
* European Studies of Infertility and Subfecundity is an EC/COST Converted Action Research Programme. Members of the Project management group were: S.Juul (project leader), W.Karmaus, J.Olsen, T.Fletcher, F.Bolumar, I.Figa-Talamanca, P.Thonneau and S.Pantelakis. W.Karmaus (Germany), L.Bisanti/A.Spinelli A (Italy ), I.Suhamburg (Denmark), P.Thonneau (France) and M.Wulff (Sweden) were responsible for the collection of data. Back

4 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Baird, D.D., Wilcox, A.J. and Weinberg, C.R. (1986) Use of time to pregnancy to study environmental exposures. Am. J. Epidemiol., 124, 470–480.[Abstract]

Basso, O., Olsen, J., Bisanti, L. et al. (1995) Are seasonal preferences in pregnancy planning a source of bias in studies of seasonal variation in reproductive outcomes? Epidemiology, 6, 520–524.[ISI][Medline]

Basso, O., Olsen, J., Bisanti, L. et al. (1997a) Repeating episodes of low fecundability. A multicentre European study. Hum. Reprod., 12, 1448–1453.[Abstract]

Basso, O., Olsen, J., Bisanti, L. et al. (1997b) The performance of several indicators in detecting recall bias. Epidemiology, 8, 269–274.[ISI][Medline]

Bisanti, L., Olsen, J., Basso, O. et al. (1996) Shift work and subfecundity: a European multicenter study on infertility and subfecundity. J. Occup. Environ. Med., 38, 352–358.[ISI][Medline]

Bolumar, F., Olsen, J. and Boldsen, J. (1996) Smoking reduces fecundity: a European multicenter study on infertility and subfecundity. Am. J. Epidemiol., 143, 578–587.[Abstract]

Bolumar, F., Olsen, J., Rebagliato, M. et al. (1997) Caffeine intake and delayed conception: a European multicenter study on infertility and subfecundity. Am. J. Epidemiol., 145, 324–334.[Abstract]

Bonde, J.P., Ernst, E., Jensen, T.K. et al. (1998) Relation between semen quality and fertility: a population-based study of 430 first-pregnancy planners. Lancet, 352, 1172–1177.[ISI][Medline]

Carlsen, E., Giwercman, A., Keiding, N. et al. (1992) Evidence for decreasing quality of semen during past 50 years. Br. Med. J., 305, 609–613.[ISI][Medline]

Hull, M.G.R., Glazener, C.M.A., Kelly, N.J. et al. (1985) Population study of causes, treatment, and outcome of infertility. Br. Med. J., 291, 1693–1697.[ISI][Medline]

Joffe, M. (1996) Decreased fertility in Britain compared with Finland. Lancet, 347, 1519–1522.[ISI][Medline]

Joffe, M., Villard, L., Li, Z. et al. (1993) Long-term recall of time-to-pregnancy. Fertil. Steril., 60, 99–104.[ISI][Medline]

Marchbanks, P.A., Peterson, H.B., Rubin, G.L. et al. (1989) Research on infertility: definition makes a difference. The Cancer and Steroid Hormone Study Group. Am. J. Epidemiol., 130, 259–267.[Abstract]

Olsen, J. (1994) Is human fecundity declining – and do occupational exposures play a role in such a decline if it exists. Scand. J. Work Environ. Health, 20, 72–77.[ISI][Medline]

Olsen, J., Bolumar, F., Boldsen, J. et al. (1997) Does moderate alcohol intake reduce fecundability? A European multicenter study on infertility and subfecundity. Alcohol. Clin. Exp. Res., 21, 206–212.[ISI][Medline]

Page, H. (1989) Estimation of the prevalence and incidence of infertility in a population: a pilot study. Fertil. Steril., 51, 571–577.[ISI][Medline]

Rachootin, P. and Olsen, J. (1981) Social selection in seeking medical care for reduced fecundity among women in Denmark. J. Epidemiol. Community Health, 35, 262–264.[Abstract]

Rachootin, P. and Olsen, J. (1982) Prevalence and socioeconomic correlates of subfecundity and spontaneous abortion in Denmark. Int. J. Epidemiol., 11, 245–249.[Abstract]

Rachootin, P. and Olsen, J. (1983) The risk of infertility and delayed conception associated with exposures in the Danish workplace. J. Occup. Med., 25, 394–402.[ISI][Medline]

Rantala, M.L. and Koskilles, A.I. (1986) Infertility in women participating in a screening program for cervical cancer in Helsinki. Acta Obstet. Gynecol. Scand., 65, 823–825.[ISI][Medline]

Rowe, P.J. and Farley, T.M. (1988) WHO/HRP Biennial report 1986–87. WHO, Geneva.

Templeton, A., Fraser, C. and Thompson, B. (1990) The epidemiology of infertility in Aberdeen. Br. Med. J., 301, 148–152.[ISI][Medline]

Thonneau, P. and Spira, A. (1990) Methodological considerations on the estimation of the prevalence of infertility. Eur. J. Obstet. Gynecol. Reprod. Biol., 38, 43–52.[ISI]

Thonneau, P., Marchand, S., Tallec, A. et al. (1991) Incidence and main causes of infertility in a resident population (1 850 000) of three French regions (1988–1989). Hum. Reprod., 6, 811–816.[Abstract]

Wagner, M.G. and Stephenson, P.A. (1992) Infertility in industrialized countries: prevalence and prevention. Soz. Praventivmed., 37, 213–217.[ISI][Medline]

Submitted on July 31, 1998; accepted on January 29, 1999.