Significance & Innovations
- In this study on time to pregnancy in women with axial spondyloarthritis, more than one-fifth of women were subfertile.
- Longer time to pregnancy was associated with longer disease duration, older age, and nulliparity.
- Findings suggest that young women with stable axial spondyloarthritis should be encouraged not to postpone pregnancy for too long.
Fertility is a person’s capacity to achieve pregnancy (3). Time from the start of actively trying to conceive to achieved pregnancy exceeding 12 months is often defined as subfertility (3). The prevalence of subfertility in the general population is estimated to be 9% (4). A study from 2015 demonstrated that 42% of women with rheumatoid arthritis (RA) were subfertile according to the above definition and that longer time to pregnancy (TTP) was associated with older age, nulliparity, disease activity, and use of prednisolone or nonsteroidal antiinflammatory drugs (NSAIDs) (5). Previous studies have found higher occurrence of subfertility in women with RA compared to healthy controls and women with systemic lupus erythematosus (6, 7).
Axial spondyloarthritis (SpA) is a chronic inflammatory rheumatic disease affecting the spine, as well as entheses and joints, with common onset in childbearing age (8). Ankylosing spondylitis (AS), now also called radiographic axial SpA because the diagnosis requires established sacroiliitis on radiographs, was traditionally seen as a disease affecting men. After the recognition of nonradiographic axial SpA, which is axial SpA without characteristic findings on radiographs, more women are diagnosed with axial SpA. Including nonradiographic axial SpA, the male:female ratio is 2–3:1 (9). A Norwegian study found that fertility rate and occurrence of childlessness were similar in women with SpA or unspecified arthritis compared to those with RA (2). To our knowledge, there are no studies on TTP in women with axial SpA.
Our aim was to study TTP and factors associated with TTP in women with axial SpA. Also, we wanted to compare women who conceived within 12 months to subfertile women with regard to preconception disease activity, health-related quality of life (HRQoL), medication, and factors that are known to affect fertility in the general population. Hypothesizing that fertility in axial SpA is similar to that in RA, we compared women with axial SpA to women with RA.
In this large registry-based study on fertility in axial SpA, we found a median TTP of 4 months in the subgroup of women enrolled when planning a pregnancy. Of these women, 32% were subfertile, defined as attempting to conceive longer than 12 months. In the total study population, where approximately one-half were already pregnant at enrollment, the median TTP was 2 months, and 21% were subfertile. Longer TTP was related to older age, nulliparity, and longer disease duration. There were no significant differences between women with axial SpA and women with RA in fertility-related outcomes.
To our knowledge, this is the first study on TTP in women with axial SpA. A Norwegian study demonstrated a lower fertility rate in women with RA and in a group of women with SpA or unspecified arthritis compared to healthy peers (2).
A study from 1999 showed a prevalence of subfertility of 16% in a general Western European population and a median TTP of 3 months (15). A later study based on population surveys from 25 countries found a mean prevalence of subfertility of 9% (4). In this setting, subfertility appears to be more common in women with axial SpA despite a median TTP similar to that of the general population. However, for direct comparison to the general population, we should have included a healthy control group.
The associations between subfertility and the factors older age, nulliparity, and smoking are well known (16–18). Lack of statistical power may explain why we did not find a significant association between nulliparity and longer TTP in the subgroup that was enrolled preconception. We have no reason to believe that the association between parity and fertility is different in women with axial SpA planning their pregnancy than in healthy women. Mean age and frequency of smoking in the study population were similar to mean maternal age at time of childbirth and the frequency of smoking in early pregnancy in the general Norwegian population (19). Thus, these factors do not seem to explain why subfertility is more common in women with axial SpA. Contrary to what is reported for the general population (20), we did not find an association between a high BMI and longer TTP. It is possible that some women struggling to conceive intentionally lose weight in order to improve fertility, diluting a negative association between a high BMI and fertility.
In the current study, the only disease-related factor associated with longer TTP was disease duration. This association has not been demonstrated in RA or systemic lupus erythematosus (5, 7). Studies have shown reduced levels of anti–müllerian hormone (an indicator of ovarian reserves) in women with AS and RA compared to healthy controls (21, 22). One study showed that HLA–B27 positivity was associated with lower anti–müllerian hormone (21). Whether some aspect of the pathophysiology of axial SpA over time affects fertility via reduced ovarian reserves or other biologic mechanisms is not known.
The current study confirmed previous findings of poor self-reported HRQoL with regard to vitality and pain in women with axial SpA and RA (7, 23). Studies have shown an association between HRQoL and sexuality in women with axial SpA (24, 25). We found no association between HRQoL and TTP that could explain an increased prevalence of subfertility.
In the current study, ~19% of women with axial SpA experienced pregnancy loss, compared to ~15% reported in the general population (26, 27). However, in order to study the risk of pregnancy loss in axial SpA, we need population-based case–control studies.
Regarding RA, we found that TTP exceeded 12 months in 24% of the total study population and 30% of the women enrolled preconception. In accordance with our study, a retrospective TTP study by Jawaheer et al found subfertility in 25% of women with RA (6). This study included a healthy control group and was able to demonstrate significantly longer TTP in women with RA. In the Dutch PARA cohort, Brouwer et al found a higher occurrence of subfertility; 42% of 245 women with RA had TTP exceeding 12 months (5). Brouwer et al demonstrated that TTP was associated with age, nulliparity, disease activity, and preconception use of NSAIDs or prednisolone. Similar to the current study, the PARA study had the combination of a prospective and retrospective design; 25% of women were pregnant at inclusion. There are several possible reasons for different findings regarding subfertility. The PARA study was conducted before 2008, while we included women until 2018. Only 15% of the women in the PARA cohort had ever used a biologic disease-modifying antirheumatic drug, while in our study 44% used TNFi in the last year before pregnancy. Additionally, our population had lower disease activity, with a mean DAS28-CRP score of 2.6 versus 3.6 in the PARA cohort.
The low disease activity in our study population, both in RA and axial SpA, may also explain why we found no association between disease activity and TTP. Our previous study of women with axial SpA in RevNatus revealed no significant difference between preconception BASDAI score and first trimester BASDAI score (28). Thus, we do not think that using the first trimester BASDAI score where the preconception score was missing relevantly affected our results.
One reason why we did not find associations between TTP and the use of NSAIDs or prednisolone may be that few women in our study population used prednisolone or NSAIDs continuously. Also, women struggling to conceive might have discontinued NSAIDs in order to facilitate pregnancy, diluting a possible adverse effect of NSAIDs on fertility. Reassuringly, we found no association between preconception TNFi and TTP.
Surprisingly, we found that MTX preconception was associated with shorter TTP in women with RA. Animal studies have suggested a harmful effect of MTX on ovarian reserve (29). In the PARA cohort, there was no association between MTX and TTP (30). We found no known variables explaining why women with RA using MTX preconception had shorter TTP in our study; age, disease duration, and disease activity did not differ significantly compared to those of women not using MTX. We can only speculate about reasons for the association. Women on MTX could have had stable disease activity for a longer period of time before deciding to get pregnant, which may have improved fertility. It is also possible that women who discontinue MTX because they desire to become pregnant have more concrete pregnancy plans and are therefore more aware of the importance of frequency and timing of intercourse.
Fertility is a complex biologic process involving gametogenesis, sperm transport, tubal patency, hormonal preparation of the endometrium, implantation, and the viability of the embryo. In addition, there are several psychosocial and cultural factors involved. Although they only tell part of the story, studies on TTP have proved useful in identifying factors with adverse effects on fertility (31, 32).
Retrospectively studying TTP in a group of women who have achieved pregnancy does not yield the same data as prospectively studying TTP in a group of women trying to conceive. In the former group, the sampling unit is the pregnancy, while in the latter the sampling unit is the attempt to become pregnant. In the current study, we used a combination of prospective and retrospective approaches. The main strength of the prospective approach is that it also includes infertile women. In addition, in prospective studies of TTP, it is possible to address underlying biologic processes. However, since RevNatus was not originally designed for studying fertility, we did not have information on ovulation, factors associated with male fertility, or frequency and timing of intercourse.
TTP studies that only include women who are already pregnant may be affected by right truncation bias so that women with longer TTP tend not to be included. Although excluding infertile women and underrepresenting subfertile women, the retrospective approach offers complimentary knowledge. This approach may include unplanned pregnancies not represented in a prospective study. Although telling less about fertility in a strictly biologic sense, studying the total population of women with axial SpA enrolled in RevNatus gives us useful knowledge on how long it takes for women with axial SpA to achieve pregnancy in a real-life setting. Generally, populations in retrospective TTP studies are considered more representative of the target population (32).
Despite being retrospective, recall bias is minimal in TTP studies where the study population is comprised of pregnant women (32). However, this design may cause other types of bias (32). We already mentioned 2 examples of possible behavior change bias: losing weight and discontinuing NSAIDs in order to achieve pregnancy in women struggling to conceive. It is not possible to examine whether our study was affected by planning bias because information on the planning of pregnancy was not recorded in RevNatus before 2016. In the current study, we included all women with TTP of 0 months regardless of pregnancy being planned or unplanned. We do not suspect pregnancy recognition bias. We included miscarriages, but the timing of recognition of pregnancy did not differ according to diagnoses or other relevant variables.
We do not suspect medical intervention bias. Fertility treatment was more common in women with RA than in women with axial SpA, but TTP in women referred to fertility treatment was comparable between diagnoses, showing no tendency for earlier referral of women with RA. When censoring TTP at 14 months in women receiving fertility treatment, as recommended by Joffe et al (32), the median TTP in both diagnoses was the same as in the original analyses. Left truncation bias, resulting in women with relatively shorter TTP not being included, is of particular importance when studying time trends. We do not suspect left truncation bias in the current study.
The main strength of this study is the large study population, where the majority fulfilled the ASAS criteria. In Norway, the majority of pregnant women with inflammatory rheumatic diseases are followed in the public specialist health care system and enrolled in RevNatus, which makes the registry representative of the population at large. However, there are limitations to generalizability. We cannot exclude that the healthiest women with axial SpA are followed in general practice and thus less likely to be enrolled in RevNatus. On the other hand, some women with axial SpA with high disease activity may never feel healthy enough for pregnancy and will therefore never be enrolled in RevNatus.
Our study would have been strengthened by considering the possible effects of SpA comorbidities. While little is known about fertility and psoriasis, subfertility has been related to disease activity in inflammatory bowel disease (33). However, in the current study, the subpopulations with psoriasis or inflammatory bowel disease were too small for analyses with sufficient statistical power.
In conclusion, women with axial SpA had an occurrence of subfertility surpassing the occurrence demonstrated in the general population. Longer TTP was associated with older age, nulliparity, and longer disease duration. Our findings suggest that women with stable axial SpA should be encouraged not to postpone pregnancy.
The authors thank the participating departments of rheumatology at the following hospitals for including patients in RevNatus: Betanien Hospital, Skien; Diakonhjemmet Hospital, Oslo; Haugesund Sanitetsforenings Rheumatism Hospital, Haugesund; Haukeland University Hospital, Bergen; Helse Førde, Førde Hospital, Førde; Helse Møre og Romsdal, Ålesund Hospital, Ålesund; Lillehammer Hospital for Rheumatic Diseases, Lillehammer; Nordland Hospital, Bodø; Oslo University Hospital Rikshospitalet, Oslo; private practice Anne Noraas, Kristiansand; Trondheim University Hospital, Trondheim; Sørlandet Hospital Kristiansand, Kristiansand; University Hospital of North Norway, Tromsø; Vestre Viken Hospital, Drammen; and Østfold Hospital, Moss.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Drs. Ursin and Wallenius had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design
Ursin, Skomsvoll, Salvesen, S. S. Koksvik, Jakobsen, Wallenius.
Acquisition of data
Ursin, S. S. Koksvik, Jakobsen, Wallenius.
Analysis and interpretation of data
Ursin, Lydersen, Wallenius.