1 Department of Biostatistics, University of North Carolina at Chapel Hill, Chapel Hill, NC, 2 Biostatistics Branch, National Institute of Environmental Health Sciences, Research Triangle Park, NC, 3 Department of Family and Preventive Medicine, University of Utah, USA, 4 Service de Biostatistiques, Centre Hospitalo-Universitaire, Lyon, France, 5 Department of Gynecological Endocrinology and Reproductive Medicine, Staedtische, Kliniken Duesseldorf gGmbH, Frauenklinik Benrath, Duesseldorf, Germany and 6 Department of Statistics, University of Padua, Padua, Italy
7 To whom correspondence should be addressed. e-mail: dunson1{at}niehs.nih.gov
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: Bayesian analysis/cervical mucus/day-specific pregnancy probabilities/menstrual cycle/ovulation
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Because properties of cervical mucus determine whether sperm will be capable of survival and transport to the ovum (Moghissi, 1973; Yudin et al., 1989
; Katz, 1991
; Kunz et al., 1997
), we hypothesize that mucus characteristics on the day of intercourse provide a clinically important predictor of the probability of conception independent of the timing relative to ovulation. In particular, consistent with the well known role of estrogenic mucus in enhancing progressive sperm motility (Eriksen et al., 1998
) and allowing for penetration, storage and transport of normal spermatozoa (Odeblad, 1968
, 1997; Menarguez et al., 2003
), we anticipate that the day-specific probabilities of conception will increase progressively with a ranking of the fertility of the mucus.
Using data from the European Study of Daily Fecundability (Colombo and Masarotto, 2000), we estimate the day-specific probabilities of conception according to both the timing of sexual intercourse relative to ovulation and a 14 ranking of the fertility of the mucus. Our data provide additional information not available in the World Heath Organization (1983
) study evaluating the use of vulvar mucus observations in estimating the fertile interval. Because the WHO study did not have a mucus-independent marker of ovulation day, the data cannot be used to address our hypotheses. In addition, the WHO study had problems with under-reporting of intercourse (Trussell and Grummer-Strawn, 1991
).
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Women kept daily records of basal body temperature (BBT), cervical mucus symptoms and intercourse. The daily mucus observations were classified according to Table I, ranging from a score of 1 (no discharge and dry) to 4 (transparent, stretchy, slippery). This 14 mucus scoring system is designed to summarize a wide variety of different mucus characteristics in a way that is predictive of the presence of fertile-type estrogenic mucus, which is characterized by a high score. If a discharge exhibited mixed characteristics, or if a woman observed multiple types of mucus through the course of the day, the highest matching category was chosen to assign the score. A primary goal of this study is to assess directly the extent to which the different levels of the mucus score predict a real difference in the conception probability.
|
In a previous analysis of these data, Dunson et al. (2002) found that nearly all pregnancies occurred from intercourse that took place in the 6-day window ending with the BBT-determined day of ovulation. This 6-day period was considered to be the fertile interval, and days outside this period were not taken into consideration. Cycles were excluded from the analysis if there were insufficient BBT data to determine the ovulation day, if there were no reported intercourse acts during the fertile interval, or if there was a day within the fertile interval on which intercourse occurred but mucus information was missing. Out of 6724 menstrual cycles of data with 487 pregnancies, 1473 cycles remained in the analysis, with 353 pregnancies. For the purposes of this study, pregnancy is defined as either an ongoing pregnancy of at least 60 days from the last menses or a clinically identified spontaneous abortion within 60 days of the last menses.
Bayesian statistical analysis approach
Modelling and estimation of pregnancy probabilities were carried out using a Bayesian hierarchical modelling approach (Dunson, 2001). This involves choosing prior distributions for unknown parameters in a statistical model based on previous information and updating this information with the data in the study to obtain posterior distributions, which represent the current state of knowledge about the unknown parameters. We base our inferences on summaries of the posterior distribution, including posterior means, 95% credible intervals and posterior probabilities.
We estimated the probability that intercourse would result in pregnancy on each of the days in the 6-day interval ending on the day of ovulation. In a cycle where intercourse occurred on more than one day during the fertile period, it is impossible to determine which act resulted in the pregnancy. Following Barrett and Marshall (1969), Wilcox et al. (1995
), Dunson et al. (2002
) and Stanford et al. (2003
) among others, we use a statistical model that allows for the incorporation of information from cycles where multiple intercourse acts occurred. Since most women contributed more than one cycle to the data, the model was also designed to account for within-woman dependency. The analyses presented in this article are based on the methods of Dunson and Stanford (2003
).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
This relationship is demonstrated in Figure 1, which shows the estimated day-specific pregnancy probabilities for the four mucus types. The day of lowest fertility was 5 days before ovulation, and the day of highest fertility was 3 days before ovulation. The difference in pregnancy probability between these two days ranged from 0.06 to 0.14, depending on mucus quality, while the difference in pregnancy probability attributable to increasing the mucus score from 1 to 4 ranged from 0.1 to 0.18. Thus the gain in pregnancy probability attributable to an increase from the lowest to highest mucus score is generally higher than the gain attributable to having intercourse 3 days before ovulation instead of 5 days before ovulation. Intercourse on any day in the 6-day window where the mucus is type 4 has a pregnancy probability that is 0.17, while the pregnancy probability does not exceed 0.13 on days with no secretions (mucus score = 1). Within the fertile window, the type of mucus observed on the day of intercourse is more predictive of conception than the timing relative to ovulation.
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Previous estimates of pregnancy probabilities on days relative to ovulation did not account for daily observations of the quality of mucus, though researchers have identified increased conception probabilities on days when secretions were observed compared with no secretions (Dunson et al., 2001) and in cycles with high mucus scores averaged over the fertile window (Stanford et al., 2003
). Our study demonstrates that the quality of mucus explains most of the relationship between the pregnancy probability and the timing of intercourse relative to ovulation. It is remarkable that even a rough categorization of mucus on a scale of 14, based on a womans own observations (Table I), explained more of the variability in the day-specific probabilities of pregnancy than could be attributed to timing of intercourse relative to ovulation.
Our results have important clinical implications. Because vulvar observations of cervical mucus predict not only the fertile days of the cycle but also the probabilities of conception within the fertile interval, monitoring of mucus provides additional information not provided by other methods for identifying the fertile interval. In particular, methods based on cycle monitoring by daily vaginal ultrasound and/or urinary LH detection are not informative about the probability of conception at a particularly time in the fertile interval within an ovulatory cycle. In addition, such monitoring is expensive and inconvenient and can miss the beginning of the fertile interval and even the most fertile days. Many women already rely on their own calculations to predict ovulation, often obtaining estimates different from results of ultrasound or LH detection (Gnoth et al., 2002). Hence, monitoring of mucus provides a useful clinical marker of days with high conception probabilities.
![]() |
Acknowledgements |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Billings EL, Billings JJ, Brown JB and Burger HG (1972) Symptoms and hormonal changes accompanying ovulation. Lancet 1,282284.[ISI][Medline]
Billings EL, Billings JJ and Catarinich M (1989) Billings Atlas of the Ovulation Method. Ovulation Method Research and Reference Centre of Australia, Melbourne, Australia.
Colombo B and Masarotto G (2000) Daily fecundability: first results from a new data base. Demogr Res 3,5.
Dorairaj K (1991) The modified mucus method in India. Am J Obstet Gynecol 165,20662067.[ISI][Medline]
Dunson DB (2001) Commentary: practical advantages of Bayesian analysis of epidemiologic data. Am J Epidemiol 153,12221226.
Dunson DB and Stanford JB (2003) Bayesian Order Restricted Inference for Count and Binary Data. Discussion Paper Series: Institute of Statistics and Decision Sciences, Duke University, Durham, NC.
Dunson DB, Baird DD, Wilcox AJ and Weinberg CR (1999) Day-specific probabilities of clinical pregnancy based on two studies with imperfect measures of ovulation. Hum Reprod 14,18351839.
Dunson DB, Sinai I and Colombo B (2001) The relationship between cervical secretions and the daily probabilities of pregnancy: effectiveness of the TwoDay Algorithm. Hum Reprod 16,22782282.
Dunson DB, Colombo B and Baird DD (2002) Changes with age in the level and duration of fertility in the menstrual cycle. Hum Reprod 17,13991403.
Ecochard R, Boehringer H, Rabilloud M and Marret H (2001) Chronological aspects of ultrasonic, hormonal and other indirect indices of ovulation. Br J Obstet Gynecol 108,822829.[CrossRef][ISI]
Eriksen GV, Carlstedt I, Uldbjerg N and Ernst E (1998) Cervical mucins affect the motility of human spermatozoa in vitro. Fertil Steril 70,350354.[CrossRef][ISI][Medline]
Gnoth C, Frank HP, Bremme M, Freundl G and Godehardt E (1996) How do self-observed cycle symptoms correlate with ovulation? Zentralbl Gynakol 118,650654.[Medline]
Gnoth C, Frank-Herrmann P and Freundl G (2002) Opinion: natural family planning and the management of infertility. Arch Gynecol Obstet 267,6771.[CrossRef][Medline]
Hilgers TW and Stanford JB (1998) Creighton-Model NaProEducation Technology for avoiding pregnancy. J Reprod Med 43,495502.[ISI][Medline]
Insler V, Melmed H, Eichenbrenner I, Serr D and Lunenfeld B. (1972) The cervical score. Int J Gynaecol Obstet 10,223228.
Katz DF (1991) Human cervical-mucusresearch update. Am J Obstet Gynecol 165,19841986.[ISI][Medline]
Katz,DF, Slade DA and Nakajima ST (1997) Analysis of pre-ovulatory changes in cervical mucus hydration and sperm penetrability. Adv Contracept 13,143151.[CrossRef][ISI][Medline]
Kunz G, Beil D, Deiniger H, Einspanier A, Mall G and Leyendecker G (1997) The uterine peristaltic pump. Normal and impeded sperm transport within the female genital tract. Adv Exp Med Biol 424,267277.[ISI][Medline]
Marshall J (1968) A field-trial of the basal-body temperature method of regulating births. Lancet 2,810.[ISI][Medline]
Menarguez M, Pastor LM and Odeblad E (2003) Morphological characterization of different human cervical mucus types using light and scanning electron microscopy. Hum Reprod 18,17821789.
Moghissi KS (1973) Sperm migration through the human cervix. In Elstein M, Moghissi KS and Borth R (eds), Cervical Mucus in Human Reproduction. Scriptor, Copenhagen, pp. 128151.
Odeblad E (1968) The functional structure of human cervical mucus. Acta Obstet Gynecol Scand 47,5779.[Medline]
Odeblad E (1997) Cervical mucus and their functions. J Ir Coll Physicians Surg 26,2732.
Sinai I, Jennings V and Arevalo M (1999) The TwoDay algorithm: a new algorithm to identify the fertile time of the menstrual cycle. Contraception 60,6570.[CrossRef][ISI][Medline]
Stanford JB, Smith KR and Dunson DB (2003) Vulvar mucus observations and the probability of pregnancy. Obstet Gynecol 101,12851293.
Trussell J and Grummer-Strawn L (1991) Further analysis of contraceptive failure of the ovulation method. Am J Obstet Gynecol 165,20542059.[ISI][Medline]
Wilcox AJ, Weinberg CR and Baird DD (1995) Timing of sexual intercourse in relation to ovulationeffects on the probability of conception, survival of the pregnancy, and sex of the baby. N Engl J Med 333,15171521.
World Health Organization (1983) A prospective multicentre trial of the ovulation method of natural family planning. III. Characteristics of the menstrual cycle and of the fertile phase. Fertil Steril 40,773778.[ISI][Medline]
Yudin AI, Hanson FW and Katz DF (1989) Human cervical mucus and its interaction with sperm: a fine-structural view. Biol Reprod 40,661671.[Abstract]
Submitted on November 5, 2003; accepted on December 19, 2003.