Polycystic ovary syndrome (PCOS) is a common endocrine disorder1 characterized by ovulatory disorder, hyperandrogenism, and hyperinsulinemia.2–5 An accurate prevalence of the condition is difficult to determine as a result of differing diagnostic criteria used but is believed to be approximately 14% of women of reproductive age.6
Because menstrual irregularity, ovulatory disorder, hyperandrogenism, and hyperinsulinemia are frequently found in a woman with PCOS, she is at increased risk of infertility, miscarriage, and gestational diabetes in pregnancy.7–11 Furthermore, it has been proposed that early embryo development may be different in a woman with PCOS.12 It is known that the children born by in vitro fertilization (IVF) have a higher rate of congenital malformations13–15 and adverse health risks in childhood.16,17 Women with PCOS are more likely to seek treatment with IVF; however, any association of PCOS with congenital abnormalities has not been studied in detail.
We have compared the health outcomes between women hospitalized with PCOS and women who had hospitalizations without a PCOS diagnosis.18 We aimed to determine significant health associations for women with PCOS reported by others.10,19,20 This study extends the evaluation of a cohort of women with and without a history of PCOS and compare the perinatal and long-term hospitalization patterns and congenital anomalies of their offspring using data linkage of statewide hospital databases within Western Australia. Western Australia is remote with a migration rate of approximately 2.8% per annum, which is low by international standards.21
MATERIALS AND METHODS
Initial identification of women with a history of a PCOS diagnosis, selection of suitable women without a PCOS diagnosis, and extraction of long-term health outcomes, hospitalizations, pregnancies, and congenital birth defects was performed using several core health data sets (www.datalinkage-wa.org/data-linkage/data-collections) within the linked data from the Western Australia Data Linkage System that has been validated previously and used extensively for health research.22
The Western Australia Data Linkage System facilitates systematic record linkage from population-based administrative health data sets within Western Australia encompassing all hospitalizations at public and private hospitals recorded in the Hospital Morbidity Data System (since 1970), all pregnancies beyond 20 weeks of gestation recorded in the Midwives' Notifications System (since 1980), all congenital birth defects recorded in the Western Australia Register of Developmental Anomalies (since 1980), all Western Australia births and deaths recorded in the Western Australia Birth Registrations (since 1974), and Western Australia Death Registrations (since 1969). This system also includes the Western Australia Electoral Roll (since 1988) recording all Australian citizens who are residents aged 18 years and older eligible to vote to enable selection of controls for research studies. This study was approved by the Department of Health Human Research Ethics Committee.
Recording of a PCOS diagnosis was identified among all women who were hospitalized with a PCOS diagnosis (International Classification of Diseases, 10th Revision [ICD-10]: E28.2 or International Classification of Diseases, 9th Revision [ICD-9]: 256.4) using the Hospital Morbidity System hospitalization records between 1980 and 2011. Hospital admissions with a PCOS diagnosis as one of the recorded diagnoses on admission were recorded from 1983 initially with only one to two diagnoses per year and with increasing frequency over time until a plateau in 1997 with 250–300 admissions per year. Hence, study participants with a recorded PCOS diagnosis were defined as women who had reached 15 years of age in 1980, or later, and who were hospitalized with PCOS as one of the diagnoses noted on their admission between January 1997 and April 2011. Hospitalizations with a PCOS diagnosis in women younger than 15 years of age occurred from 1994 and were recorded in less than 1% of patients, hence restricting the maximal age of the patient to 15 years of age in 1980 ensured that no women with a PCOS diagnosis were missed and that all pregnancies that continued beyond 20 weeks of gestation were recorded in the Midwives Notification System active since 1980.
Women without a PCOS diagnosis were randomly selected from the Western Australia Electoral Roll or Western Australia Birth Registry to match women aged 18 years and older or younger than 18 years of age, respectively. Ten women without a PCOS diagnosis were assigned to each woman with a PCOS diagnosis while matching for age at first PCOS diagnosis from 1997, hence effectively also matching for year of birth. A high ratio of 10 women without PCOS to each woman with a PCOS diagnosis was chosen to also allow comparisons of hospitalizations with low prevalence between the women23 and to compare health outcomes of their offspring.
The Western Australia Midwives Notification System records information on all pregnancies of at least 20 weeks of gestation (a detailed description of the stored data is included in the Appendix, available online at http://links.lww.com/AOG/A637).
The Western Australia Register of Developmental Anomalies records structural or functional anomalies present at conception or in pregnancy and diagnosed up to 6 years of age in all live births, stillbirths, and terminations of pregnancy in Western Australia.24 Up to 10 individual birth defects are recorded for each offspring. Each defect is categorized as major or minor. Most minor malformations that are nondisfiguring or not requiring treatment were excluded. Birth defects are coded according to the British Pediatric Association ICD-9 system and categorized into: cardiovascular, respiratory, gastrointestinal, urogenital, musculoskeletal, anomalies of integument, chromosomal defects, nervous system, anomalies of the eye, anomalies of the ear, face, or neck, and other anomalies not included in the preceding categories. Evaluation of congenital anomalies in the study was limited to the pregnancies of at least 20 weeks of gestation.
Hospitalizations that were recorded in the Hospital Morbidity System included one principal diagnosis, up to 20 secondary diagnoses, and up to 10 procedures per admission (detail of the diagnoses collected is included in the Appendix, available online at http://links.lww.com/AOG/A637).
Medians, interquartile ranges, and ranges were used to summarize continuous data. Mann–Whitney tests were used to compare continuous characteristics between the groups such as number of offspring. Hospitalization patterns, stratified by the PCOS exposure, were summarized with frequency distributions and using Kaplan–Meier cumulative diagnosis-free survival probabilities together with the estimates of hospitalizations probabilities by 10 years of age. Logistic regression and Cox proportional hazards regression were used to examine the effects of maternal PCOS diagnosis on adverse health outcomes. The PCOS effects were summarized using adjusted odds ratios (ORs) and hazard ratios with their 95% confidence intervals (CIs) after controlling for maternal age, offspring's age, or both and other relevant perinatal characteristics as appropriate for each outcome considered. All analyses were conducted using generalizing estimating equations that allowed modeling of potential correlation between births for the same mother (with “exchangeable” correlation structure) (a detailed description of the logistic regression performed to evaluate the influence of a PCOS diagnosis on perinatal outcomes, congenital abnormalities, and infant and child hospitalization is included in the Appendix, available online at http://links.lww.com/AOG/A637).
SAS 9.3 statistical software was used for data analysis. All hypothesis tests were two-sided and P values <.05 were considered statistically significant.
Overall 2,566 women with a history of PCOS diagnosis on their hospital admissions between January 1997 and April 2011 were matched by age and year of birth to 25,660 women without any prior recorded PCOS diagnoses and considered in the comparisons of their offsprings' perinatal and long-term hospitalizations (Table 1). Pregnancies were more frequent among women with a PCOS diagnosis (in 1,789 and 16,139 women without a PCOS diagnosis, respectively: 69.7% compared with 62.9%, P<.001), whereas the total number of pregnancies per woman was significantly lower compared with the women without a PCOS diagnosis, with a median number of two pregnancies in women with and without a PCOS diagnosis and their respective quartiles (interquartile range 1–3, maximum 7 and interquartile range 2–3, maximum 15) (P<.001) (ie, individually women with PCOS had fewer pregnancies each).
A summary of maternal demographics and pregnancy complications stratified by history of a PCOS diagnosis for all individual pregnancies is presented in Table 2, showing that women with PCOS were more likely to present in pregnancy with a pre-existing medical condition.
Women with a PCOS diagnosis were more likely to have had an IVF treatment (7.9% compared with 0.7%) and to have a multiple pregnancy (3.3% compared with 1.3%). Multiple pregnancies were more likely to occur after IVF treatment irrespective of a history of PCOS diagnosis on hospitalization (rates of multiple pregnancies of 6.8% and 9.3% after IVF treatment in women with and without PCOS, respectively, OR 0.71, 95% CI 0.38–1.33, P=.289). Women with a PCOS diagnosis were also more likely to develop pregnancy complications (Table 2).
Offspring of women with a PCOS diagnosis were at higher risk of preterm birth univariately (15.5% compared with 7.6%, P<.001) and after adjustment for confounding risk factors in pregnancy (OR 1.74, 95% CI 1.53–1.98) (Table 3). Low Apgar scores at 5 minutes were more frequent in offspring of women with a PCOS diagnosis (4.2% compared with 1.8%, P<.001) and remained significantly elevated after adjustments for pregnancy risk factors and being born preterm (OR 1.46, 95% CI 1.10–1.93). The rates of being born of low birth weight (11.4% compared with 6.2%) or born small for gestational age (8.7% compared with 7.2%) were significantly higher among the offspring of PCOS-exposed women univariately, but these increased risks were explained by the pregnancy risk factors and gestational age at delivery alone (Table 3). The likelihood of macrosomia (4,000 g) was not associated with maternal PCOS diagnosis (13.6% compared with 13.0%). Although stillbirths occurred more frequently in pregnancies of women with a PCOS diagnosis (1.8% compared with 0.6%, P<.001), this elevated risk was explained by maternal, pregnancy risk factors, and prematurity (OR 1.23, 95% CI 0.80–1.89). Neonatal deaths occurred in 20 (0.5%) and 44 (0.1%) of neonates born to mothers with and without a PCOS diagnosis, respectively (P<.001). Perinatal mortality was associated with maternal PCOS exposure univariately and remained significant after adjustments for pregnancy risk factors (2.3% compared with 0.7%, OR 1.49, 95% CI 1.02–2.18).
Offspring of women with a PCOS diagnosis were more likely to require admission to a special care nursery (14.1% compared with 7.9%), and their risk of an admission remained elevated after adjusting for maternal and pregnancy risk factors (OR 1.21, 95% CI 1.05–1.40) (Table 3).
Offspring of women with a PCOS diagnosis were younger with a lower proportion of individuals alive more than 6 years of age, which was the upper limit of age for recording of birth defects (60.8% compared with 68.3%; Table 1). Congenital anomalies among siblings occurred in 14.0% of women with a PCOS diagnosis (205/1,784) with 10.2% of women having more than one offspring with a congenital anomaly (21/205) and in 10.1% of women without PCOS (1,645/16,128) with 7.4% of women with multiple offspring with congenital anomalies (122/1,645).
The prevalence of congenital anomalies among all births and those limited to offspring who reached 6 years of age or died before 6 years of age is summarized in Table 4. Maternal PCOS diagnosis was associated with increased risk of any developmental anomaly (OR 1.20, 95% CI 1.03–1.40), any cardiovascular anomaly alone (OR 1.37, 95% CI 1.01–1.87), and urogenital anomaly alone (OR 1.36, 95% CI 1.03–1.81) (Table 5). The increased risk of any major anomaly present in the univariate analysis was no longer statistically significant after adjustment for other relevant covariates and sibship correlations (OR 1.14, 95% CI 0.96–1.35) (Table 5). No association between maternal PCOS diagnosis and other congenital anomalies combined, excluding cardiovascular and genitourinary, were found. The associations between maternal PCOS and less frequent congenital anomaly groups were only examined using univariate comparisons, as shown in Table 4.
The median age at the end of the hospitalization period for all 38,028 individuals was 8.6 years of age with 80% of these offspring followed until 15 years of age (maximum age 30 years). Hospitalizations for most ICD diagnoses were more frequent among offspring of women with a history of a PCOS diagnosis. Hospitalization rates categorized by ICD diagnosis codes are summarized in Table 6 and shown using the cumulative proportion of offspring hospitalized with each ICD diagnosis by 10 years of age (with select cumulative hospitalizations shown in Fig. 1).
In line with previous reports, our data suggest that PCOS is associated with a significantly increased risk of preeclampsia, gestational diabetes, preterm delivery, and cesarean delivery in pregnancy11,25,26 and we report a significant association with congenital abnormalities for offspring. Furthermore, a maternal PCOS diagnosis is associated with an increased rate of hospitalization for the child into adolescence.
It is believed that the cause is the deranged metabolic environment in women with PCOS leading to altered placentation and endovascular changes27 and potentially the increased markers of oxidative stress present in the circulatory system.28 In addition, women with PCOS are likely to have abnormalities of homocysteine metabolism related to the degree of hyperinsulinemia and hyperandrogenemia,29 although this is correctable with adequate folate intake. Further abnormalities identifiable in women with PCOS are an elevated serum plasminogen activator inhibitor-130 and an abnormal expression of some molecular markers with the endometrium,31 particularly relating to abnormalities in insulin-like growth factor binding protein-1, glycodelin, homeobox protein—HOXA 10, and endometrial progesterone resistance.31–34 Furthermore, evidence exists for impaired decidual trophoblast invasion in women with PCOS,32 the consequences of which may be a predisposition to growth restriction, preeclampsia, and prematurity.
After controlling for potential confounders, the neonate of a woman with PCOS is at an increased risk of admission to the neonatal nursery and at an increased risk of perinatal mortality, reported previously in univariate meta-analyses.11,26 Our study reports the association of an increased risk of congenital abnormalities among the offspring of mothers diagnosed with PCOS. Possible reasons for this increased risk include maternal diabetes, maternal obesity, and medication use. We controlled the data for the use of assisted reproduction; however, it is known that subfertility per se, even without the need to resort to fertility treatment, is associated with an increased predisposition to birth defects.35 In our study women with PCOS were at an increased risk of diabetes,18 and studies suggest that maternal pregestational diabetes increases the risk of cardiovascular, urogenital, musculoskeletal, and any anomalies in offspring.36 Furthermore, maternal obesity increases the risk of cardiovascular and other anomalies,37 and it is possible women with PCOS may have been prescribed more medication during pregnancy as a result of a pre-existing condition associated with an increase in malformations38 or the medication taken to assist conception.15
The strengths of the study include its population basis, the use of record linkage and hence avoidance of response bias, recall bias and loss to follow-up, and the ability to control for a large number of potential confounders. However, there exists some potential for misclassification bias within the data presented within this study. First, the diagnosis of PCOS was not standardized until 20042; hence, some women will have been labeled as having a PCOS diagnosis before the application of the new definition, but then subsequently did not meet the revised definition potentially reducing the strength of the association, because women with PCOS could potentially be in the comparison group. However, it would be expected that there may be a bias toward the coding of a diagnosis of PCOS for women with more clinically significant features of PCOS, that is, those women with a more extreme phenotype of PCOS (such as significant androgenic features or prolonged amenorrhea), than for those with less significant features, effectively strengthening any association. A potential weaknesses of the study is the fact that women with PCOS had to have had an admission to a hospital to be selected for the study; this may have resulted in the selection of a more “unhealthy population” of women diagnosed with PCOS.
In view of our findings and the purported association of periconception, metabolic derangements with poor implantation and placentation strategies to improve periconception health in women with PCOS may be expected to improve reproductive outcomes. There exists some limited evidence of an improved reproductive outcome for normal-weight women with PCOS in comparison with overweight women with PCOS with regard to the risk of miscarriage and gestational diabetes.39,40 Furthermore, evidence exists for the use of metformin for women with PCOS on reduced rates of miscarriage, gestational diabetes, and preeclampsia in pregnancy.41 Hence, preconception identification of women with PCOS may enable early intervention to improve long-term outcomes, although it is not clear if all women with PCOS have this increased risk of adverse outcomes or whether there are certain features of PCOS that are associated with particular outcomes for the mother and her offspring.
1. Hart R, Hickey M, Franks S. Definitions, prevalence and symptoms of polycystic ovaries and polycystic ovary syndrome. Best Pract Res Clin Obstet Gynaecol 2004;18:671–83.
2. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 2004;19:41–7.
3. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 2004;81:19–25.
4. Azziz R. PCOS: a diagnostic challenge. Reprod Biomed Online 2004;8:644–8.
5. Zawadzki J, Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach. In: Dunaif A, Givens J, Haseltine F, Merriam G, editors. Polycystic ovarian syndrome. Boston (MA): Blackwells; 1992:p. 377–84.
6. Agrawal R, Sharma S, Bekir J, Conway G, Bailey J, Balen AH, et al.. Prevalence of polycystic ovaries and polycystic ovary syndrome in lesbian women compared with heterosexual women. Fertil Steril 2004;82:1352–7.
7. Balen A, Rajkowha M. Polycystic ovary syndrome—a systemic disorder? Best Pract Res Clin Obstet Gynaecol 2003;17:263–74.
8. Wild RA. Long-term health consequences of PCOS. Hum Reprod Update 2002;8:231–41.
9. Hart R, Norman R. Polycystic ovarian syndrome—prognosis and outcomes. Best Pract Res Clin Obstet Gynaecol 2006;20:751–78.
10. Fauser BC, Tarlatzis BC, Rebar RW, Legro RS, Balen AH, Lobo R, et al.. Consensus on women's health aspects of polycystic ovary syndrome (PCOS): the Amsterdam ESHRE/ASRM-Sponsored 3rd PCOS Consensus Workshop Group. Fertil Steril 2012;97:28–38.e25.
11. Qin JZ, Pang LH, Li MJ, Fan XJ, Huang RD, Chen HY. Obstetric complications in women with polycystic ovary syndrome: a systematic review and meta-analysis. Reprod Biol Endocrinol 2013;11:56.
12. Wissing ML, Bjerge MR, Olesen AI, Hoest T, Mikkelsen AL. Impact of PCOS on early embryo cleavage kinetics. Reprod Biomed Online 2014;28:508–14.
13. Hansen M, Kurinczuk JJ, de Klerk N, Burton P, Bower C. Assisted reproductive technology and major birth defects in Western Australia. Obstet Gynecol 2012;120:852–63.
14. Hansen M, Kurinczuk JJ, Milne E, de Klerk N, Bower C. Assisted reproductive technology and birth defects: a systematic review and meta-analysis. Hum Reprod Update 2013;19:330–53.
15. Davies MJ, Moore VM, Willson KJ, Van Essen P, Priest K, Scott H, et al.. Reproductive technologies and the risk of birth defects. N Engl J Med 2012;366:1803–13.
16. Hart R, Norman RJ. The longer-term health outcomes for children born as a result of IVF treatment: Part I—General health outcomes. Hum Reprod Update 2013;19:232–43.
17. Hart R, Norman RJ. The longer-term health outcomes for children born as a result of IVF treatment. Part II—Mental health and development outcomes. Hum Reprod Update 2013;19:244–50.
18. Hart R, Doherty DA. The Potential Implications of a PCOS Diagnosis on a Woman's Long-Term Health Using Data Linkage. J Clin Endocrinol Metab 2015;100:911–9.
19. Balen A, Rajkowha M. Polycystic ovary syndrome—a systemic disorder? Best Pract Res Clin Obstet Gynaecol 2003;17:263–74.
20. Conway G, Dewailly D, Diamanti-Kandarakis E, Escobar-Morreale HF, Franks S, Gambineri A, et al.. The polycystic ovary syndrome: a position statement from the European Society of Endocrinology. Eur J Endocrinol 2014;171:P1–29.
21. Clark A, Preen DB, Ng JQ, Semmens JB, Holman CD. Is Western Australia representative of other Australian States and Territories in terms of key socio-demographic and health economic indicators? Aust Health Rev 2010;34:210–5.
22. Holman CD, Bass AJ, Rouse IL, Hobbs MS. Population-based linkage of health records in Western Australia: development of a health services research linked database. Aust N Z J Public Health 1999;23:453–9.
23. Hennessy S, Bilker WB, Berlin JA, Strom BL. Factors influencing the optimal control-to-case ratio in matched case-control studies. Am J Epidemiol 1999;149:195–7.
24. Bower C, Rudy E, Quick J, Rowley A, Watson L, Cosgrove P. Report of the Western Australian register of developmental anomalies. King Edward Memorial Hospital; 2014. Available at: http://www.kemh.health.wa.gov.au/services/register_developmental_anomalies/warda.htm
. Retrieved April 8, 2015.
25. Fauser BC, Devroey P, Diedrich K, Balaban B, Bonduelle M, Delemarre-van de Waal HA, et al.. Health outcomes of children born after IVF/ICSI: a review of current expert opinion and literature. Reprod Biomed Online 2014;28:162–82.
26. Boomsma CM, Eijkemans MJ, Hughes EG, Visser GH, Fauser BC, Macklon NS. A meta-analysis of pregnancy outcomes in women with polycystic ovary syndrome. Hum Reprod Update 2006;12:673–83.
27. Palomba S, Falbo A, Chiossi G, Muscogiuri G, Fornaciari E, Orio F, et al.. Lipid profile in nonobese pregnant women with polycystic ovary syndrome: a prospective controlled clinical study. Steroids 2014;88:36–43.
28. Murri M, Luque-Ramírez M, Insenser M, Ojeda-Ojeda M, Escobar-Morreale HF. Circulating markers of oxidative stress and polycystic ovary syndrome (PCOS): a systematic review and meta-analysis. Hum Reprod Update 2013;19:268–88.
29. Grodnitskaya EE, Kurtser MA. Homocysteine metabolism in polycystic ovary syndrome. Gynecol Endocrinol 2012;28:186–9.
30. Glueck CJ, Wang P, Fontaine RN, Sieve-Smith L, Tracy T, Moore SK. Plasminogen activator inhibitor activity: an independent risk factor for the high miscarriage rate during pregnancy in women with polycystic ovary syndrome. Metabolism 1999;48:1589–95.
31. Shang K, Jia X, Qiao J, Kang J, Guan Y. Endometrial abnormality in women with polycystic ovary syndrome. Reprod Sci 2012;19:674–83.
32. Palomba S, Russo T, Falbo A, Di Cello A, Amendola G, Mazza R, et al.. Decidual endovascular trophoblast invasion in women with polycystic ovary syndrome: an experimental case-control study. J Clin Endocrinol Metab 2012;97:2441–9.
33. Savaris RF, Groll JM, Young SL, DeMayo FJ, Jeong JW, Hamilton AE, et al.. Progesterone resistance in PCOS endometrium: a microarray analysis in clomiphene citrate-treated and artificial menstrual cycles. J Clin Endocrinol Metab 2011;96:1737–46.
34. Cermik D, Selam B, Taylor HS. Regulation of HOXA-10 expression by testosterone in vitro and in the endometrium of patients with polycystic ovary syndrome. J Clin Endocrinol Metab 2003;88:238–43.
35. Zhu JL, Basso O, Obel C, Bille C, Olsen J. Infertility, infertility treatment, and congenital malformations: Danish national birth cohort. BMJ 2006;333:679.
36. Vinceti M, Malagoli C, Rothman KJ, Rodolfi R, Astolfi G, Calzolari E, et al.. Risk of birth defects associated with maternal pregestational diabetes. Eur J Epidemiol 2014;29:411–8.
37. Stothard KJ, Tennant PW, Bell R, Rankin J. Maternal overweight and obesity and the risk of congenital anomalies: a systematic review and meta-analysis. JAMA 2009;301:636–50.
38. Ban L, Gibson J, West J, Fiaschi L, Sokal R, Smeeth L, et al.. Maternal depression, antidepressant prescriptions, and congenital anomaly risk in offspring: a population-based cohort study. BJOG 2014;121:1471–81.
39. Moran LJ, Hutchison SK, Norman RJ, Teede HJ. Lifestyle changes in women with polycystic ovary syndrome. The Cochrane Database of Systematic Reviews 2011, Issue 2. Art. No.: CD007506. DOI: 10.1002/14651858.CD007506.pub2.
40. De Frène V, Vansteelandt S, T'Sjoen G, Gerris J, Somers S, Vercruysse L, et al.. A retrospective study of the pregnancy, delivery and neonatal outcome in overweight versus normal weight women with polycystic ovary syndrome. Hum Reprod 2014;29:2333–8.
41. Zheng J, Shan PF, Gu W. The efficacy of metformin in pregnant women with polycystic ovary syndrome: a meta-analysis of clinical trials. J Endocrinol Invest 2013;36:797–802.