Infertility following trisomic pregnancies: A nationwide cohort study.

Satu Wedenoja, Mika Gissler, Seppo Heinonen, Juha Kere and Laura Tanner: Conceived, designed, and directed the study. Satu Wedenoja and Laura Tanner: Acquired funding. Satu Wedenoja, Mika Pihlajamäki, Mika Gissler, Juho Wedenoja and Hanna Öhman: Acquired and analyzed the data. Satu Wedenoja, Mika Pihlajamäki, Mika Gissler, Helena Kääriäinen and Laura Tanner: Wrote the draft of the article. All authors reviewed and revised the article. All authors read and approved the final article.

This study received funding from Päivikki and Sakari Sohlberg Foundation and Juhani Aho Foundation for Medical Research.

We detected a total of 6042 trisomic conceptions in Finland between 1987 and 2018. For building up the study series, we excluded 258 trisomic conceptions, of which 198 were recurrent trisomic pregnancies occurring in 99 women and 60 were conceptions for which no matched controls could be found from the Biobank data. The study series involved 5784 women with a laboratory‐confirmed diagnosis of T13, T18, or T21 in their fetus or newborn between 1987 and 2018. The control series involved 34 422 women matched on age, residence, and timing of pregnancy. Table  1 shows the characteristics at the time of the index pregnancy. Of these women, 2576 (44.6%) with a trisomic conception and 15 448 (44.8%) of controls were under 35 years at the time of the index pregnancy. Of trisomic conceptions, 4369 (75.6%) were associated with T21, 1065 (18.4%) with T18, and only 351 (6.1%) with T13. Characteristics of women with trisomic conceptions and matched controls in Finland between 1987 and 2018. Includes one individual with a combined trisomy 13 and 21. Includes six individuals, whose two different pregnancies were used as controls. We first studied all gynecology‐specific codes at any time before or after the index pregnancy (Table  S1 ). To this end, we included only the diagnoses that had been detected in at least five women with trisomic conceptions (either T13, T18 or T21). We observed an increased OR (1.19, 95% CI: 1.08–1.32) for female infertility (ICD‐10 diagnosis group of N97) among cases. In the subgroup analysis, N97 diagnosis showed association with T18 (OR: 1.29, 95% CI: 1.03–1.61) and T21 (OR: 1.17, 95% CI: 1.04–1.32) conceptions, but not for T13 (OR: 1.15, 95% CI: 0.75–1.72) (Table  2 ). Association of female infertility (N97) with trisomic conceptions among women who were pregnant in Finland between 1987 and 2018. Abbreviations: CI, confidence interval; OR, odds ratio. Includes one individual with a combined trisomy 13 and 21. Includes six individuals, whose two different pregnancies were used as controls in different subgroups. We also found inverse comorbidity for some disorders. Diagnoses of menstrual pain (N94) and female genital prolapse (N81) at any time yielded decreased odds in women with trisomic conceptions. However, results were null when assessed in the subgroups of T13, T18, and T21, and in the respective controls (Table  S1 ). We further studied the association between female infertility and trisomic conceptions regarding the time of the N97 diagnosis and the age of the women. We found no association between trisomic conceptions and female infertility before the index pregnancy (Table  S2 ). Instead, 245 (4.2%) individuals with a trisomic conception and 822 (2.4%) controls had a subsequent diagnosis of female infertility, corresponding to an adjusted OR of 1.81 (95% CI: 1.56–2.09) in the whole series (Table  3 ). In the subgroup analysis, the highest OR of 2.04 (95% CI: 1.48–2.77) arose after T18 conceptions, while OR of 1.77 (95% CI: 1.49–2.10) was observed after T21 conceptions, and results were null after T13 conceptions (OR: 1.57, 95% CI: 0.85–2.85) (Table  3 ). Association of female infertility (N97) with trisomic conceptions after the index pregnancy among women who were pregnant in Finland between 1987 and 2018. Abbreviations: CI, confidence interval; OR, odds ratio. Includes one individual with a combined trisomy 13 and 21. Includes six individuals, whose two different pregnancies were used as controls in different subgroups. We then studied the diagnoses of N97 in women according to the age groups. Subsequent diagnoses of female infertility (N97) were associated with trisomic conceptions both in women <35 years at the index pregnancy (OR: 1.71, 95% CI: 1.39–2.10) and in those ≥35 years at the index pregnancy (OR: 1.92, 95% CI: 1.57–2.36). These associations were observed both after T18 and T21 but not after T13 conceptions (Table  3 ). We did not study N97 subdiagnoses (N97.0 to N97.9) in detail because many women received several different N97 subdiagnoses and relative distribution of subdiagnoses was highly similar among cases and controls (Table  S3 ).

This population‐based study demonstrates an association between trisomic conceptions and subsequent diagnoses of female infertility. The odds of infertility after trisomic conceptions were elevated by more than 80% in the whole series. Remarkably, this association appeared also in younger women, who were <35 years at the time of the trisomic pregnancy. Moreover, we found no similar associations for other gynecologic or reproductive disorders. Although these observational data cannot prove causality, this study suggests that similar mechanisms might predispose to trisomic conceptions and infertility, in both younger and older women. A main strength of our study is the large population‐based series of 5784 women and 34 422 matched controls. Unlike many other studies on human trisomic conceptions, we had a significant number of 2576 women <35 years at the time of the index pregnancy with trisomy. This allowed us to link diagnoses of female infertility also to younger women, for whom no major risk factors for trisomic pregnancies have been demonstrated so far. 21 Since the Finnish national healthcare system is responsible for all prenatal screening tests and maternal care, and coding of the disorders have been consistently performed during several decades, we are likely to have all laboratory‐confirmed trisomic conceptions and clinical diagnoses of infertility here. Because of the publicly funded maternity care, equal accessibility to the prenatal screening guaranteed by law, and high uptake of prenatal screening as well as strict uniform criteria for invasive procedures, our data is highly representative of the Finnish population. As an additional strength, control population was adjusted on maternal age, residence, and timing of pregnancy, potentially excluding major environmental or time‐dependent exposures, if any, to trisomy or infertility. The main limitation of our study was that this registry‐based analysis was unable to count miscarriages or medical abortions before the diagnosis of trisomy. Notably, more than 35% of early miscarriages are explained by an abnormality in the number of chromosomes. 22 Furthermore, most trisomic embryos are likely to be miscarried soon after conception, 23 and our series included only those trisomic pregnancies that received the genetic diagnosis. Moreover, data on gynecologic diagnoses were available only from 1996 onwards. Because we could not assess the time needed to get pregnant for the index pregnancy of our study, we cannot exclude the association between primary infertility and trisomic conceptions in our series. As for infertility diagnoses, we were able to find only those who sought help for infertility and received the diagnosis. The low number of women diagnosed with a trisomic conception and subsequent infertility, and highly similar infertility rates in women under 35 years and older, suggests that younger women were more likely to seek fertility testing than older women. An additional limitation is that our series included no data on the duration of infertility, number of conceptions before the infertility diagnosis, or conception method. Moreover, many of the pregnancies with trisomies were terminated, and we cannot rule out effects of abortion on subsequent pregnancies. However, no data supports the link between medical abortions and chances of getting pregnant in the future. 24 , 25 Although we had no data on parental balanced chromosomal translocations, causing 4% to 5% of trisomic conceptions and predisposing to infertility, 26 , 27 or confounders such as smoking, 10 their influence is likely to be minor at the level of the whole series. Infertility affects around 15% of all couples, and one of the major etiological factors is advanced maternal age. 28 In this study, diagnoses of female infertility at any time appeared in 8.2% of women with trisomic conceptions and were more common than the rate of 6.9% in the controls. These figures are lower than the self‐reported lifetime subfertility of 16% in Finland, 29 but in line with the findings that less than half of infertile women seek infertility treatments. 30 Although both T18 and T21 conceptions were associated with subsequent diagnoses of infertility, case counts were too low and confidence intervals too wide for T13 to examine this association independently. It is therefore possible that similar association could be observed for T13 with a larger patient cohort. Since no other human aneuploidies than T13, T18 and T21 were studied here, and all but laboratory‐confirmed diagnoses of trisomy were lacking, the magnitude of association between female infertility and previous trisomic conceptions remains obscure. Aneuploidy in oocytes is the leading cause of infertility. 31 On the other hand, previous studies found no association between non‐oocyte related etiologies of infertility and aneuploidies. 32 One of the major obstacles in studies of female infertility and the risk of trisomic conceptions is that they share advanced maternal age as a major risk factor. Moreover, couples seeking infertility treatments may not be representative of the general population and some nondisjunction events can be increased among them. 8 Our study setting and results are noteworthy for several reasons. First, this study was registry‐based analysis that involved national data on all trisomic conceptions and infertility diagnoses. Second, we had a significant number of women <35 years, who showed increased odds for future infertility after trisomic conceptions similarly to observations in older women. Third, no other pregnancy complications such as recurrent pregnancy losses, being associated with an increased aneuploidy rate, 32 were associated with trisomic conceptions in our series. Therefore, our results are novel and raise a possibility that some women have an age‐independent genetic predisposition to trisomic conceptions and infertility. This finding is supported by the observations that prior aneuploid pregnancies are, age‐independently, increasing the risk of future aneuploid conceptions. 32 , 33 In agreement with this, specific altered recombination patterns show association with maternal meiotic errors and some of them are age‐independent. 34 Moreover, several genetic loci have been linked to the risk of chromosomal non‐disjunction for T21, 35 and mitotic errors resulting in trisomic conceptions are rare but age‐independent. 36 That we found no association between primary infertility and trisomic conceptions may be related to the fact that information on the time elapsed to get pregnant for the first time were missing and infertility diagnoses were available only from 1996 onwards.

Our study implicates chromosomal non‐disjunction resulting in human trisomies as a possible culprit for some cases of subsequent female infertility. These data suggest that not only the small risk of recurrent trisomy, 37 but also the possibility of infertility, are worth mentioning in the setting of genetic or clinical counseling of couples after trisomic pregnancies. Collectively, these data suggest that some mechanisms of trisomic conceptions are age‐independent and might decrease the chances of getting pregnant in the future. As this was an epidemiological study, a large multicenter prospective trial will be needed to confirm the findings of this study.

Human autosomal trisomies – characterized by a third chromosome instead of two – are among the most common causes of congenital birth defects and intellectual disability. 1 Their incidence and prevalence vary strongly depending on the maternal age. While around 20% of oocytes are aneuploid in women under 35 years of age, aneuploidy affects more than 50% of oocytes at advanced maternal age. 2 Up to 90% of fetuses with autosomal trisomies receive the extra chromosome from the mother. 3 Accordingly, the strong maternal age effect is seen for the prevalence of all survivable autosomal aneuploidies, namely trisomy 13 (T13), trisomy 18 (T18), and trisomy 21 (T21). 4 Why advanced maternal age, defined as 35 years or older, results in an exponential increase in the rate of aneuploidy remains poorly understood. 2 , 5 In older women, age‐dependent decay of meiotic mechanisms such as centromeric cohesion might trigger chromosomal non‐disjunction. 6 However, around 1% to 5% of all clinically recognized pregnancies are trisomic even in women younger than 35 years. 2 , 7 , 8 Among them, aberrant crossing over or recombination of chromosomes might play a major role. 9 Moreover, environmental exposures such as smoking or defective folic acid metabolism might modulate the pathogenesis. 6 , 8 , 10 , 11 , 12 Previous work has highlighted that some gynecologic or reproductive conditions such as low oocyte pool, infertility caused by anovulatory factor, recurrent pregnancy loss, and endometriosis increase the risk of trisomy. 13 , 14 However, evidence of an association between antral follicle pool or reproductive hormone levels with trisomic conceptions are lacking, 15 and chromosomal analysis of preimplantation embryos shows no association between anovulatory infertility and aneuploidy. 16 As for endometriosis, experimental models support increased formation of aneuploid oocytes, but clinical data are conflicting. 17 , 18 , 19 Whether diseases of female pelvic organs mediate the risk of trisomic conceptions remains largely unclarified. We conducted a cohort study using population‐based registry data collected over 30 years in Finland. We hypothesized that disorders of the female genital tract or pelvic organs, potentially having negative impact on fertility, might be associated with trisomic conceptions. Our secondary aim was to examine whether younger women, with a trisomic conception <35 years, differ from the older mothers in terms of their gynecologic and reproductive health.

The authors have no conflicts of interest.

This registry‐based nationwide cohort study was approved by the Finnish Institute for Health and Welfare (THL/2295/5.05.00/2019) and by the Biobank Borealis (BB_2021_5001). Research was performed according to the approved guidelines. No informed consents were required from the study subjects specifically for this study, but were available for biobanking the samples, as defined in the regulations of the national Biobank Act (688/2012). Trisomic conceptions were identified from the national Registry of Congenital Malformations. We included all women whose fetuses or newborns received a diagnosis of T13, T18, or T21 during pregnancy or after birth between 1987 and 2018. These trisomic pregnancies resulted in either miscarriage, abortion, or live birth but data on the pregnancy outcome were not included in the analyses. During the study period of 1987–2018, 34% of trisomic conceptions were diagnosed after miscarriage, abortion, or at birth, and 65% were diagnosed through the prenatal screening program offered as part of the routine antenatal care. During the study period, all pregnant women had been offered first‐trimester ultrasound and second trimester morphology ultrasound as well as screening programs of that time (serum screening or combined screening) free of charge. Uptake of these prenatal screening tests has traditionally been high. Invasive testing (chorionic villus sampling or amniocentesis) had been offered for those with a positive screening result or a known carrier status of a genomic rearrangement but, after 1990s, were no longer solely based on advanced maternal age. All congenital anomalies, including those diagnosed prenatally, have been registered to the national Registry of Congenital Malformations and, if resulting in birth, also to the Medical Birth Registry. We used data from the Finnish Maternity Cohort (FMC) held by the Biobank Borealis to assemble a control group of women without a diagnosis of trisomy in their pregnancy. The FMC cohort is a nationwide collection of serum samples taken from pregnant women for routine screening of congenital infections during the latter part of the first trimester or the early weeks of the second trimester. 20 We selected a maximum of six control subjects for each woman with a trisomic conception, and matched the controls on age, residence, and timing of the pregnancy. The age of the control subjects was identified on the index date. Residence was either the same town or county in Finland. Timing of the pregnancy was noted as the calculated conception date based on the first trimester screening date and was assigned to be not more than 2 months different from that of the women with trisomic pregnancies. Altogether, we found six matched controls for 5698 (99%) of all 5784 cases. Because we did not want to exclude the remaining cases in our patient series, we accepted five matched controls for 11 cases, four controls for 18, three controls for 15, two controls for 20, and only one control for 22 cases. To find matched controls, we accepted two different pregnancies from six women to serve as controls for different trisomic conceptions. These controls were utilized in subgroup analyses of T13, T18 or T21, but all individuals (cases and controls) were counted only once at the whole series level analyses. Initial case selection was based on the diagnosis of a trisomic conception. Matched controls were selected based on the absence of a diagnosis of trisomy in their pregnancy. Thereafter, the study women were linked based on their personal identification numbers to the national Care Registry for Health Care that involves all diagnoses related to the specialized healthcare, either inpatient admission or outpatient clinic contact. We utilized data spanning 1996 to 2019, when diagnoses were coded according to the International Statistical Classification of Diseases and Related Health Problems, 10th revision (ICD‐10) as described by the WHO. Data covered inpatients admissions, and from 1998 onwards, also outpatient visits. For the primary analysis, both inflammatory diseases of female pelvic organs (ICD‐10 diagnoses: N70–N77) and noninflammatory disorders of the genital tract (N80–N98) were studied for the whole study series. We calculated the number of women with each diagnosis at any time. Second, we studied the diagnoses that were associated with trisomic conceptions in subgroups of women with T13, T18, or T21 conceptions, and their respective controls. Third, we tested whether the associated diagnoses were coded before or after the index pregnancy, and whether they showed association with trisomic conception among younger (<35 years) or older (≥35 years) women of the study series. We characterized women with trisomic conceptions and control women by case–control groups (T13, T18 or T21), and by different age groups. The cases and controls were matched on the age, residence, and timing of pregnancy, while no other confounders were available. All comparisons were two‐valued (yes, no) and the outcome variables were two‐valued. We used chi‐square test with Yates correction to compare trisomic conceptions with controls regarding gynecologic diagnoses and calculated crude odds ratios (ORs) and 95% CIs. Each diagnosis was assessed independently among cases and controls. The statistical significance was set at P values less than 0.05, and two‐tailed P values were used. The analyses were conducted using the Graph Pad version 10.1.1 (GraphPad Software, LLC).

Table S1. Association of gynecology‐specific ICD‐10 codes with trisomic conceptions at any time before or after the index pregnancy. Table S2. Association of female infertility (N97) with trisomic conceptions before the index pregnancy. Table S3. N97 subdiagnoses among women with trisomic conceptions and in control women.