Is glycosylated haemoglobin a marker of fertility? A follow-up study of first-pregnancy planners

Niels Henrik I. Hjollund1,4, Tina Kold Jensen2, Jens Peter E. Bonde1, Tine Brink Henriksen3, Anna-Maria Andersson2, Niels Erik Skakkebæk2 and The Danish First Pregnancy Planner Study Team

1 Department of Occupational Medicine, Aarhus University Hospital, Norrebrogade 44, 8000 Aarhus, 2 Department of Growth and Reproduction, Rigshospitalet, Copenhagen and 3 Perinatal Epidemiological Research Unit, Department of Gynaecology and Obstetrics, Aarhus University Hospital, Aarhus, Denmark


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
We performed a follow-up study of time to pregnancy in a population of first-time pregnancy planners without previous reproductive experience. The objective of this paper is to report and discuss a finding of a strong relationship between glycosylated haemoglobin (HbA1C) and fertility. A total of 165 Danish couples without previous reproductive experience was followed from termination of birth control until pregnancy for a maximum of six menstrual cycles. HbA1C and reproductive hormones were measured at entry. Questionnaire data were collected at entry and once in each cycle during follow-up. The odds ratio (OR) for pregnancy per cycle decreased with increasing concentration of HbA1C (OR per percent HbA1C 0.4, 95% CI 0.2–0.9 for all six cycles and 0.2, 95% CI 0.1–0.5 in the first three cycles). A high concentration of HbA1C was associated with a high concentration of testosterone and a low concentration of inhibin A. No association was found between HbA1C and psychosocial distress. The reduced fertility among women with high HbA1C may be due to an association with subclinical polycystic ovaries as indicated by the hormonal profile.

Key words: HbA1C/polycystic ovary syndrome/reproduction/stress/testosterone


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Glycosylated haemoglobin (HbA1C) is formed by covalent binding of glucose to the N-terminal valine of the ß-chain in the haemoglobin molecule throughout the lifetime of the erythrocyte (Harrison and Isselbacher, 1994Go). In the clinical control of diabetic patients, HbA1C serves as an estimator of the time-averaged blood glucose concentration in the preceding 3–12 weeks. As a diagnostic tool, HbA1C may have advantages in relation to other methods, and a glycosylated proportion of 7.0% or higher has been proposed as indicative of the need for pharmacological intervention (Peters et al., 1996Go). We undertook a follow-up study of time to pregnancy which included HbA1C as a possible marker of biological response to stress. We report here on this finding and discuss possible physiological mechanisms.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
From 1992 to 1994, a total of 430 couples was recruited after nationwide mailing of a personal letter to 52 255 trade union members (metal workers, office and commercial workers, nurses and day-care workers) who were 20–35 years old, living with a partner and had no children. Only couples without previous reproductive experience who intended to discontinue contraception in order to become pregnant were eligible. A detailed description of the cohort and study design is provided elsewhere (Bonde et al., 1998Go).

The couples were enrolled into the study when they discontinued birth control and were followed during six menstrual cycles or until a pregnancy was recognized by their general practitioner. The couples were enrolled at two centres: the Department of Occupational Medicine in Aarhus (West Centre) and The Department of Growth and Reproduction in Copenhagen (East Centre). When the couples had notified the research team, they were invited to a consultation by a physician or a nurse who provided instructions and obtained a blood sample. The date of the consultation was not standardized according to the woman's menstrual cycle. At enrolment and on the 21st day of each menstrual cycle during follow-up both partners completed a questionnaire on demographic, medical, reproductive, occupational, psychological, and lifestyle factors. The woman recorded vaginal bleeding and sexual intercourse in a structured diary developed for the study.

HbA1C was determined in capillary whole blood or venous whole blood using an immunoturbidimetric assay (DCA 2000 HbA1C System, Bayer, Denmark). The validity of this assay has previously been tested (Mortensen et al., 1994Go). A serum sample was frozen for later analyses of reproductive hormones (Hjollund et al., 1998Go).

The measurement of HbA1C was included in the protocol from July 1994 and couples who were enrolled from that time onward were included (n = 203). Nineteen women (9.3%) refused to produce a blood sample. In some couples there was a delay from discontinuation of contraception use until the enrolment interview, and women who were later shown to be pregnant when the blood sample was obtained were not included if it was obtained later than 27 days after the first day of menstrual bleeding (n = 18). One single outlier of HbA1C was found (HbA1C = 7.9%); this woman had diagnosed diabetes mellitus and was excluded. No cycles with reported sexual abstinence from day 11 to 20 in the menstrual cycle resulted in a pregnancy and these cycles were excluded (n = 18 menstrual cycles). The analyses include 165 couples, contributing a total of 683 menstrual cycles (Tables I and IIGoGo).


View this table:
[in this window]
[in a new window]
 
Table I. Characteristics of 165 first-pregnancy planners dichotomized at the median value of glycosylated haemoglobin (HbA1C)
 

View this table:
[in this window]
[in a new window]
 
Table II. Characteristics of the menstrual cycles
 
The unit of observation was a menstrual cycle and each cycle was numbered according to number of menstrual cycles, since discontinuation of contraceptives. The association between HbA1C and fecundability was estimated by a logistic regression model while controlling for potential confounding factors (Hosmer and Lemeshow, 1989Go; SAS Institute Inc., 1989). This model is equivalent to the discrete version of the Cox proportional hazard model (Weinberg, 1994Go; SAS Institute Inc., 1996), and provides a fecundability odds ratio (OR), which is the estimated odds of conception in a menstrual cycle for the exposed group divided by the corresponding odds for the reference group. Potential confounding factors were selected by their biological relevance with respect to fertility, irrespective of their association with fecundability and covariates in the present study. Menstrual cycle numbers were entered as dummy variables to adjust for changes in baseline fecundability with increasing number of cycles.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
HbA1C was approximately normally distributed with a mean of 4.75% and a standard deviation of 0.33%. The 0, 33, 66, and 100 percentiles were 3.5, 4.7, 4.9, and 5.6% respectively. In bivariate analyses HbA1C had higher values among participants living in eastern Denmark and participants with a high caffeine consumption. HbA1C was not related to smoking or other lifestyle factors (including chocolate intake), psychological distress, body mass index, mean menstrual cycle length, frequency of sexual intercourse, or the timing of the blood sample with respect to the menstrual cycle (Table IGo).

The overall probability of conception per menstrual cycle (fecundability) was 13.3%. For women with HbA1C in the upper tertile the fecundability was 10.2% compared to 14.3% in the lower tertile. In the first three cycles of follow-up, the fecundability in the two groups was 7.3 and 15.3% respectively. Fecundability ORs for tertiles of HbA1C and for HbA1C as a continuous variable are listed in Table IIIGo. For women in the upper tertile compared with the lower tertile the odds for pregnancy per cycle were reduced by ~40% during up to six cycles of follow-up, and by 70% when including only the first three cycles of follow-up. The estimates changed only slightly after adjustment for possible confounding factors (Table IIIGo).


View this table:
[in this window]
[in a new window]
 
Table III. Fecundability odds ratios (OR) according to female percentage glycosylated haemoglobin (HbA1C)
 
All reproductive hormones varied according to the timing of the blood sample within the menstrual cycle in which the sample was obtained. To account for this, the menstrual cycle was categorized into 1 week periods. Within each period HbA1C and reproductive hormones were assigned as high or low by the 66 percentiles in the actual period. In all samples a high HbA1C was associated with a statistically significantly increased odds for a high value of testosterone and a low value of inhibin A, while no significant associations were seen with other reproductive hormones (Table IVGo). Analogous analyses based on the 80 percentiles provided similar results. Separate analyses of samples obtained within the first 10 days of the cycle demonstrated a borderline significant relationship between LH and HbA1C with wide confidence interval (OR for LH >80 percentile 5.0, 95% CI 0.9–26.9, n = 46). A high body mass index was weakly associated with high HbA1C, while, apart from a prolonged time to pregnancy, no relationship was found with the other clinical characteristics related to the presence of polycystic ovary (PCO) syndrome (Table IVGo).


View this table:
[in this window]
[in a new window]
 
Table IV. Odds ratio (OR) for having a high HbA1C (>4.9%) according to concentrations of different reproductive hormones and clinical characteristics thought to be related to polycystic ovary syndrome (PCO)
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The results indicate that female HbA1C may be a marker of fertility, especially in the first time period closest to the measurement, which is expected because HbA1C may vary over time. The weaker association in the later cycles could thus be due to increasing disparity between the value of HbA1C at enrolment and the level in the actual cycle. Little is known of intra-individual variation in HbA1C among non-diabetic women and in the present study HbA1C was measured only at enrolment. However, if our finding is true, high concentration of HbA1C is more likely to be a marker of a moderate increase in time to pregnancy than of infertility.

The concentration of HbA1C may change during pregnancy for physiological reasons, and bias may be introduced if the woman was pregnant when the blood sample was obtained. The date of onset of menstrual bleeding was established by questionnaire, diary, and interview. Hereby we obtained precise identification of the day of the blood sample in relation to the menstrual cycle. Eighteen women were excluded because they were actually pregnant when the blood sample was drawn. Their concentration of HbA1C was slightly lower (mean HbA1C 4.61% versus 4.77, P = 0.06) and inclusion of these women resulted in an even stronger relation with time to pregnancy. It is, however, impossible to know if these women became pregnant with a low concentration of HbA1C or developed it due to influence of an early pregnancy. Oral contraception is associated with reduced fecundity in the very first cycles after the discontinuation (Huggins and Cullins, 1990Go; Burkman, 1994). Separate analyses of 129 women who did not use oral contraceptives as the last method revealed similar results.

The level of HbA1C differed significantly between the East and West Centres (mean 4.83 versus 4.65, P < 0.0005). All couples at the East Centre were interviewed at the hospital and HbA1C was measured simultaneously from a capillary blood sample, while for the majority of couples at the West Centre (68%), a sample of venous blood was obtained for later analysis at the laboratory. Although the assay should be valid for both sample methods, this may explain the observed difference between the centres. Stratification by centre provided essentially identical results and in both strata the association reached the level of statistical significance [OR East Centre 0.2/% HbA1C (95% CI 0.1–0.7); West Centre 0.2/%HbA1C (95% CI 0.04–0.95)].

Possible physiological mechanisms
Polycystic ovary (PCO) syndrome is characterized by enlarged ovaries containing an increased number of follicles, and a hyperandrogenic hormonal profile involving increased plasma concentrations of testosterone and luteinizing hormone (LH), decreased serum hormone binding globulin (SHBG) and hyperinsulinism (Conway et al., 1989Go; Kaaks, 1996Go). The prevalence of PCO syndrome among premenopausal women has been estimated as 5–8%, but usually PCO syndrome is diagnosed only when women seek medical advice for hirsutism, oligomenorrhoea or infertility (Rich Edwards et al., 1994Go; Kaaks, 1996Go). Using ultrasound detection methods, the prevalence of an ovarian morphology typical of PCO was shown to be ~20% in population samples of adult women and a considerable proportion had a hyperandrogenic hormone profile (Polson et al., 1988Go; Franks, 1989Go; Clayton et al., 1992Go). These women generally have a normal menstrual pattern and normal body mass index. The frequency of hyperandrogenism without morphological changes is unknown, but may be even higher. Whether subclinical PCO confers subfertility is unclear.

PCO has no single organic aetiology, but chronic overstimulation of ovarian steroid synthesis caused by insulin or insulin-like growth factor secondary to decreased insulin sensitivity may be a central feature of the complex hormonal dysregulation leading to the syndrome (Kaaks, 1996Go). Plasma glucose concentration may still be within the normal range, but a small elevation in glucose concentration may be detectable in analyses of HbA1C. In one study of women with and without PCO syndrome, HbA1C was found to correlate well with the average glucose concentration throughout an oral glucose tolerance test, although the predictive value of diagnosing impaired glucose intolerance was low (Golland et al., 1989Go). In women with PCO syndrome including amenorrhoea, the cyclic pattern of reproductive hormone secretion is disturbed (Franks, 1989Go), but little is known of the variation across the menstrual cycle in asymptomatic women with hyperandrogenic hormonal status, but normal menstrual periods. In the present study only a single blood sample was obtained at enrolment, irrespective of the timing with respect to the menstrual cycle. On the other hand, the day in the menstrual cycle was precisely recorded and analyses of the relation between HbA1C and the reproductive hormones should provide unbiased estimates but with larger standard errors. A high concentration of HbA1C was associated with a high concentration of testosterone but a low concentration of inhibin A. While an increase in testosterone is a central feature of the PCO syndrome, the significance of inhibin is less clear (Lambert Messerlian et al., 1994Go; Anderson et al., 1998Go) and the role of inhibin in subclinical stages of PCO is not known. As inhibin A is believed to be produced mainly in large follicles and the corpus luteum (Lambert Messerlian et al., 1994Go), it may be speculated that a lower concentration of circulating inhibin A may be indicative of a less well functioning dominant follicle and/or corpus luteum.

We found an unexpected strong relationship between HbA1C and fertility and subsequent analyses revealed an association between HbA1C and testosterone. However, physiological mechanisms other than PCO related hormonal changes may explain the relationship between HbA1C and fecundability, but cannot be elucidated further using our own data. Although a relationship between subclinical hyperandrogenic hormonal status and HbA1C could be expected from previous studies, it is surprising that this is closely related to a temporary low fecundability. Even though the findings seem internally consistent and the associations are strong, the result was unexpected and could be due to chance or unconsidered sources of bias. This finding should be critically evaluated in other studies.


    Acknowledgments
 
We are indebted to the support of several trade union officials, in particular Ernst Bliesmann, Peter Olesen, Rigmor Laulund and Niels Nedergaard. Anders Nyboe Andersen, Department of Department of Gynaecology and Obstetrics, Rigshospitalet, Copenhagen is acknowledged for valuable comments. This study was mainly supported by a grant from the Aarhus University Research Foundation (J nr 1994–7430–1). Additional financial support was also given by the Danish Medical Research Council (J nr 12–2042–1), the Danish Health Insurance Foundation (J 11/243–91, J. 11/236–93) and the Institute of Experimental Clinical Research, Aarhus University. The activities of the Danish Epidemiology Science Centre are financed by a grant from the Danish National Research Foundation.


    Notes
 
4 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Anderson, R.A., Groome, N.P. and Baird, D.T. (1998) Inhibin A and inhibin B in women with polycystic ovarian syndrome during treatment with FSH to induce mono-ovulation. Clin. Endocrinol. Oxf., 48, 577–584.[ISI][Medline]

Bonde, J.P.E., Hjollund, N.H.I., Jensen, T.K. et al. (1998) A follow-up study of environmental and biologic determinants of fertility among 430 Danish first pregnancy planners: design and methods. Reprod. Toxicol., 12, 19–27.[ISI][Medline]

Burkman, R.T. Jr (1994) Noncontraceptive effects of hormonal contraceptives: bone mass, sexually transmitted disease and pelvic inflammatory disease, cardiovascular disease, menstrual function, and future fertility. Am. J. Obstet. Gynecol., 170, 1569–1575.[ISI][Medline]

Clayton, R.N., Ogden, V., Hodgkinson, J. et al. (1992) How common are polycystic ovaries in normal women and what is their significance for the fertility of the population? Clin. Endocrinol. Oxf., 37, 127–134.[ISI][Medline]

Conway, G.S., Honour, J.W. and Jacobs, H.S. (1989) Heterogeneity of the polycystic ovary syndrome: clinical, endocrine and ultrasound features in 556 patients. Clin. Endocrinol. Oxf., 30, 459–470.[ISI][Medline]

Franks, S. (1989) Polycystic ovary syndrome: a changing perspective. Clin. Endocrinol. Oxf., 31, 87–120.[ISI][Medline]

Golland, I.M., Vaughan Williams, C.A., Shalet, S.M. et al. (1989) Lack of predictive value of HbA1c for impaired glucose tolerance in polycystic ovary syndrome. Gynecol. Endocrinol., 3, 229–235.[ISI][Medline]

Harrison, T.R., and Isselbacher, K.T. (1994) Harrison's Principles of Internal Medicine. New York: McGraw-Hill.

Hjollund, N.H.I., Bonde, J.P.E., Jensen, T.K. et al. (1998) Semen quality and sex hormones with reference to metal welding. Reprod. Toxicol., 12, 91–95.[ISI][Medline]

Hjollund, N.H.I., Jensen, T.K. et al. (1999) Distress and reduced fertility. A follow-up study of first pregnancy planners. Fertil. Steril., in press.

Hosmer, D.W.J., and Lemeshow, S. (1989) Applied Logistic Regression. Wiley, New York.

Huggins, G.R., and Cullins, V.E. (1990) Fertility after contraception or abortion. Fertil. Steril., 54, 559–573.[ISI][Medline]

Kaaks, R. (1996) Nutrition, hormones, and breast cancer: is insulin the missing link? Cancer Causes Control, 7, 605–625.[ISI][Medline]

Lambert Messerlian, G.M., Hall, J.E., Sluss, P.M. et al. (1994) Relatively low levels of dimeric inhibin circulate in men and women with polycystic ovarian syndrome using a specific two-site enzyme-linked immunosorbent assay. J. Clin. Endocrinol. Metab., 79, 45–50.[Abstract]

Mortensen, H.B., Nielsen, M.R., Christensen, E. et al. (1994) Comparison of two new rapid assays for HbA1C determination in patients with diabetes mellitus. Ugeskr. Laeger, 156, 317–321.[Medline]

Peters, A.L., Davidson, M.B., Schriger, D.L. et al. (1996) A clinical approach for the diagnosis of diabetes mellitus: an analysis using glycosylated haemoglobin levels. JAMA, 276, 1246–1252.[Abstract]

Polson, D.W., Adams, J., Wadsworth, J. et al. (1988) Polycystic ovaries – a common finding in normal women. Lancet, i, 870–872.

Rich Edwards, J.W., Goldman, M.B., Willett, W.C. et al. (1994) Adolescent body mass index and infertility caused by ovulatory disorder. Am. J. Obstet. Gynecol., 171, 171–177.[ISI][Medline]

SAS Institute Inc. (1989) SAS/STAT User's Guide, Version 6, Volume 2. SAS Institute, Cary, NC.

SAS Institute Inc. (1996) SAS/STAT Software: Changes and Enhancements Through Release 6.11. SAS Institute, Cary, NC.

Weinberg, C.R. (1994) The influence of occupational activity on the menstrual cycle and fecundability. Epidemiology, 5, 563–564.[ISI][Medline]

Submitted on July 7, 1998; accepted on February 2, 1999.





This Article
Abstract
FREE Full Text (PDF )
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Request Permissions
Google Scholar
Articles by Hjollund, N. H. I.
Articles by Skakkebæk, N. E.
PubMed
PubMed Citation
Articles by Hjollund, N. H. I.
Articles by Skakkebæk, N. E.