Bonde, Jens Peter*; Joffe, Mike†; Sallmén, Markku‡; Kristensen, Petter§; Olsen, Jørn∥; Roeleveld, Nel¶; Wilcox, Allen**
From the *Department of Occupational Medicine, Åarhus University Hospital, Aarhus, Denmark; the †Department of Epidemiology & Public Health, Imperial College, London, U.K.; the ‡Department of Epidemiology and Biostatistics, Finnish Institute of Occupational Health, Helsinki, Finland; the §National Institute of Occupational Health, Oslo, Norway; the ∥Department of Epidemiology, School of Public Health, UCLA, Los Angeles, CA; the ¶Department of Epidemiology and Biostatistics, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands; and the **National Institutes of Environmental Health Sciences, Durham, NC.
This commentary is based on lectures given by Jørn Olsen, Mike Joffe, Markku Sallmén, Nel Roeleveld, and Allen Wilcox at a workshop held in relation to the 18th International Conference on Epidemiology in Occupational Health, September 11, 2005, Bergen, Norway.
Address correspondence to: Jens Peter Bonde, Department of Occupational Medicine, Aarhus University Hospital, DK 8000 Åarhus C, Denmark. E-mail: firstname.lastname@example.org.
Circumstantial evidence regarding the deterioration of male fecundity, together with the fact that the majority of infertility cases have unknown causes, make it important to identify risk factors that can be eliminated and thus pave the way for prevention of infertility. Rapidly increasing costs of artificial fertilization and cumbersome lengthy treatments of infertile couples with limited success rates add to the importance of such studies and contribute to a wider acceptance of epidemiologic studies of fertility. Worldwide fertility is now close to replacement level, and the previous fear of uncontrolled population growth should no longer stand in our way for studying environmental causes that impair fecundity.
More than 20 years ago, a seminal paper1 on subfecundity and infertility examined occupational risk factors in the Danish workplace. During the late 1980s, the use of time-to-pregnancy (TTP) was introduced and elaborated. TTP is the number of months or menstrual cycles it takes a couple to conceive and covers the entire distribution of waiting times from 0 (high-level fecundity) to several years (low-level fecundity).2,3 Valid information about TTP is easy and inexpensive to obtain by questionnaires and interview. With increasing interest in reproductive issues in public health, TTP soon became a widely used measure in epidemiologic studies. As time went on, it became clear that TTP studies needed to take several population characteristics into consideration, especially if the compared populations differed in such factors as desired family size, use of contraceptive methods, and sexual activity. Many specific pitfalls inherent in the TTP methodology were identified,4,5 and new types of biases have been added to the list in recent years.6 Lately, we have witnessed discussions on costly prospective designs versus inexpensive retrospective designs.7 Also, we have seen differing opinions as to whether it is possible to obtain valid retrospective information on secular trends in fecundity by the TTP methodology8,9 or on fecundity among different cultures.10,11 Is the TTP design option still to be considered a valid method for obtaining knowledge about fecundity in humans?
To obtain a direct measure of human fecundity, couples must attempt to use their reproductive capability. This is where the troubles start. First, not all persons get to try, even if they want children (for example, if they never find a partner). Second, some couples get pregnant while using contraception or by accident. Such pregnancies have no TTP value but provide important information. These couples may be more fecund than others, but how much more we do not know. As a result, those who provide a measurable TTP may underrepresent highly fecund couples. Third, not all couples succeed in conceiving, and TTP distributions that include only those who succeed will underrepresent less fertile couples, thus indicating the need for collecting data on unprotected intercourse that does not lead to a pregnancy.12 Fourth, other factors such as earlier reproductive experience and the intensity and persistence that couples invest in putting their fecundity to a test may also determine whether couples do or do not contribute TTP values.6
In addition to selective forces, causal inferences in TTP studies may be confounded by volitional factors related to sexual behavior. The probability of becoming pregnant in a menstrual cycle (ie, fecundability) depends on the timing and frequency of unprotected intercourse as well as on male and female biologic fecundity, which is the characteristic we seek to measure.
Strong assumptions of comparability among groups of couples are easily violated. This may be in part why several retrospective studies based solely on the TTP distribution have found higher fertility (shorter TTP) among women at the upper end of reproductive age.13 If older couples are less persistent in their efforts to obtain a pregnancy, the subfertile subgroup will be overrepresented among those who discontinue their pregnancy attempts.6 Likewise, the TTP distribution on its own may give biased estimates of fertility time trends because of shifts over time in social factors such as population access to safe contraception and induced abortion.9 With more effective and safe prevention, there are fewer contraceptive failures, and so the more highly fertile couples are kept in the population who has an eligible TTP value.
In occupational studies conducted in groups that are homogeneous with respect to socioeconomic and cultural factors, volitional factors such as reproductive behavior, use of contraception, and persistence of trying may not be a severe problem. However, these volitional factors are certainly relevant to comparisons over time, and comparisons of different regions and cultures.
Nevertheless, occupational studies of specific risk factors in homogeneous populations may still be complicated. For example, in cross-sectional studies, short TTP values will be overrepresented in the most recent pregnancies because sampling is biased on TTP length, leading to truncation bias.8 Then, if the exposure of interest is declining rapidly over calendar time, recent low exposure levels are associated with short TTP values and thus incorrectly indicate a harmful effect of the exposure. Several ways of addressing time-trend and truncation bias have been suggested.4,8,12 Furthermore, results may reflect reverse causality, whereby achieved fecundity (having a baby) affects occupational status or level of exposure. For example, farming couples may redistribute jobs when the family situation changes, or new mothers may not return to employment or only work part-time—the “reproductively-unhealthy worker effect.”14
In conclusion, studying fertility is difficult, but it is important. Time-to-pregnancy is not a measure of fecundability for an individual couple, but rather the best measure we have of couple fecundity for a population. Fertility studies should not be mistaken for simple survey studies; they are a method that is exceptionally susceptible to bias. Therefore, study design, questionnaire design, and statistical analysis need rigorous attention in every setting to minimize bias in the results. It has become clear that to address the various types of bias, TTP studies need to include information on at least 3 related conditions: the TTP distribution, unprotected intercourse not leading to pregnancy, and accidental pregnancies.
Prospective studies offer advantages with respect to collection of data on volitional factors such as timing and frequency of intercourse as well as accurate timing and assessment of exposure in relation to events during the menstrual cycle, fertilization, and implantation. However, prospective studies carry the burdens of low (and probably selective) participation rates and high costs.
Retrospective TTP studies on time trends and on crosscultural and regional differences are particularly at risk for biased findings. Nonetheless, they are a useful and cost-effective option for studies of occupational and similar risk factors if 2 conditions are met. First, care must be taken to ensure comparability of exposed and unexposed, in particular with respect to use of contraception and induced abortion, sexual behavior, and other volitional factors related to reproduction. Second, any systematic shift in exposure levels during the study period must be documented and addressed by measures to prevent truncation bias, and the study's conclusions must take into account the possibility that coincidental trends in exposure and in outcome may give rise to a spurious association (which can occur in any occupational or environmental study).
There is a need to further develop, standardize, and validate a TTP questionnaire that can be used in various settings and cultures. As a point of reference, existing widely used questionnaires can be used. The questionnaire should integrate not only the questions needed to define a TTP, but should also specify the additional information that is necessary to deal with the most important of the potential biases affecting TTP analyses and to perform standard sensitivity analyses.4,8
New methods of measuring couple fecundability are welcomed. For example, the ideas on current duration approach15 and case–cohort design16 may prove useful. It would also be helpful if biologic markers of fecundity, particularly female fecundity, could be developed and validated.
1. Rachootin P, Olsen J. The risk of infertility and delayed conception associated with exposures in the Danish workplace. J Occup Med
2. Baird DD, Wilcox AJ, Weinberg CR. Use of time to pregnancy to study environmental exposures. Am J Epidemiol
3. Joffe M. Male- and female-mediated reproductive effects of occupation: the use of questionnaire methods. J Occup Med
4. Weinberg CR, Baird DD, Wilcox AJ. Sources of bias in studies of time to pregnancy. Stat Med
5. Olsen J, Juul S, Basso O. Measuring time to pregnancy. Methodological issues to consider. Hum Reprod
6. Basso O, Juul S, Olsen J. Time to pregnancy as a correlate of fecundity: differential persistence in trying to become pregnant as a source of bias. Int J Epidemiol
7. Tingen C, Stanford JB, Dunson DB. Methodologic and statistical approaches to studying human fertility and environmental exposure. Environ Health Perspect
8. Joffe M, Key J, Best N, et al. Studying time to pregnancy by use of a retrospective design. Am J Epidemiol
9. Sallmén M, Weinberg CR, Baird DD, et al. Has human fertility declined over time? Why we may never know. Epidemiology
10. Juul S, Karmaus W, Olsen J. Regional differences in waiting time to pregnancy: pregnancy-based surveys from Denmark, France, Germany, Italy and Sweden. The European Infertility and Subfecundity Study Group. Hum Reprod
11. Tuntiseranee P, Olsen J, Chongsuvivatwong V, et al. Fecundity in Thai and European regions: results based on waiting time to pregnancy. Hum Reprod
12. Sallmén M, Lindbohm ML, Nurminen M. Paternal exposure to lead and infertility. Epidemiology
13. Jensen TK, Scheike T, Keiding N, et al. Selection bias in determining the age dependence of waiting time to pregnancy. Am J Epidemiol
14. Joffe M. Biases in research on reproduction and women's work. Int J Epidemiol
15. Keiding N, Kvist K, Hartvig H, et al. Estimating time to pregnancy from current durations in a cross-sectional sample. Biostatistics
16. Olsen J, Andersen PK. We should monitor human fecundity, but how? A suggestion for a new method that may also be used to identify determinants of low fecundity. Epidemiology
© 2006 Lippincott Williams & Wilkins, Inc.