Spondyloarthritis in the peripartum period: a systematic review of current knowledge and clinical implications

Abstract

Background:

Spondyloarthritis (SpA) commonly affects women of reproductive age, yet evidence on fertility, pregnancy outcomes and postpartum disease activity remains fragmented. The growing recognition of non-radiographic axial SpA (axSpA) and the wider use of biologics have increased the need for reproductive-health data in this population.

Objectives:

To systematically review the evidence on fertility, pregnancy course, therapeutic management, obstetric and neonatal outcomes and postpartum disease activity in women with SpA.

Design:

Systematic literature review.

Methods:

We performed a systematic search of PubMed/MEDLINE using predefined MeSH-based equations addressing fertility, pregnancy and the postpartum period in SpA. English and French studies were included without date restriction. Data extraction followed a PICO/PECO framework and findings were synthesized thematically according to PRISMA guidance.

Results:

Of 337 publications identified, 167 met the inclusion criteria. Evidence indicates a modest reduction in ovarian reserve and longer time-to-pregnancy in some women with SpA, partly influenced by non-steroidal anti-inflammatory drugs (NSAIDs) exposure. Disease activity during pregnancy is generally stable in PsA but more variable in axSpA; flares occur in up to 40% of pregnancies, especially after biologic discontinuation. Overall pregnancy outcomes are favourable, though risks of preterm birth, caesarean delivery and small-for-gestational-age infants are modestly increased. Biologic therapies, especially tumour necrosis factor inhibitors, appear safe and contribute to disease control. Postpartum flares are common, particularly in axSpA, and breastfeeding is generally compatible with most treatments.

Conclusion:

When disease activity is well controlled and treatment is optimized, reproductive outcomes in women with SpA are largely favourable. Avoiding unnecessary NSAID exposure, maintaining appropriate biologic therapy when indicated and close peripartum monitoring are key to improving outcomes. High-quality prospective studies remain needed to refine clinical guidance.

Introduction

Spondyloarthritis (SpA) is a chronic inflammatory rheumatic disease primarily affecting the axial skeleton, peripheral joints and entheses. ¹ Among its phenotypes – axial spondyloarthritis (axSpA), psoriatic arthritis (PsA), reactive arthritis and enteropathic arthritis – axSpA and PsA are the most frequently encountered in women of reproductive age. ² The onset of SpA typically occurs during the second or third decade of life, aligning with peak childbearing years. ³ In recent years, the clinical landscape of SpA has undergone a significant transformation. The recognition of non-radiographic axial SpA (nr-axSpA) has expanded diagnostic criteria and revealed a striking female predominance in this subgroup. Studies estimate that women represent up to 70% of patients with nr-axSpA, ⁴ contrasting with the historically male-dominated profile of radiographic axSpA (r-axSpA). This shift has profound implications for reproductive health, as a growing number of women with active SpA are navigating pregnancy, childbirth and postpartum recovery while managing chronic disease. Because most women with SpA are diagnosed during their fertile years and pregnancy is common in this population, understanding the interplay between disease activity, treatment safety and obstetric outcomes is timely and clinically essential. Pregnancy in women with SpA presents unique challenges. Although immunological adaptations of pregnancy may modulate disease activity, the peripartum period is frequently associated with disease flares, especially in axSpA. ² Moreover, SpA has been linked to increased risks of adverse obstetric outcomes, including preterm birth, small-for-gestational-age neonates, preeclampsia and caesarean delivery.^2,3

Despite the growing interest in reproductive health among patients with chronic inflammatory diseases, data on SpA in the peripartum period remain limited and heterogeneous. ⁵ Existing studies vary in design, population size and outcome measures, making it difficult to draw definitive clinical recommendations. Furthermore, the safety and management of biologic and conventional disease-modifying antirheumatic drugs (bDMARDs and csDMARDs) during pregnancy and lactation remain debated. ⁶

Given the increasing prevalence of SpA among women of childbearing age and the potential impact on maternal and neonatal outcomes, a comprehensive synthesis of current evidence is warranted. This systematic review aims to evaluate the existing literature on SpA in the peripartum period. By consolidating available data, we seek to inform clinical practice and identify gaps for future research.

Methods

We conducted a systematic review of the literature using the PubMed/MEDLINE database. The aim was to collect and synthesize publications addressing fertility, pregnancy and the postpartum period in women with SpA, to provide an up-to-date overview of existing data.

Three search equations were constructed using Medical Subject Headings (MeSH) terms, translated and validated via the CISMeF (Catalogue et Index des Sites Médicaux de la Langue Française) and the official MeSH database:

• Fertility: (‘Spondylarthropathies’[Mesh] OR spondyloarthritis[tiab] OR spondylarthritis[tiab] OR ‘Arthritis, Psoriatic’[Mesh]) AND (‘Fertility’[Mesh])

• Pregnancy: (‘Spondylarthropathies’[Mesh] OR spondyloarthritis[tiab] OR spondylarthritis[tiab] OR ‘Arthritis, Psoriatic’ [Mesh]) AND (‘Pregnancy’[Mesh])

• Postpartum: (‘Spondylarthropathies’[Mesh] OR spondyloarthritis[tiab] OR spondylarthritis[tiab] OR ‘Arthritis, Psoriatic’ [Mesh]) AND (‘Postpartum Period’ [Mesh]).

Because ‘Spondyloarthritis’ is not an official MeSH descriptor but is widely used in the contemporary literature, the search strategy combined the validated MeSH heading ‘Spondylarthropathies’ with the free-text terms ‘spondyloarthritis’ and ‘spondylarthritis’ to ensure comprehensive retrieval of both indexed and non-indexed publications.

The search was finalized on 11 June 2025.

Each search equation was structured according to the PICO framework to ensure clinical relevance and methodological consistency:

• Fertility question

• ○ Population (P): Women diagnosed with SpA (including PsA).

• ○ Exposure/intervention (I): Disease presence or activity or use of SpA treatments.

• ○ Comparator (C): Healthy women or those without SpA, when available or SpA women without treatment.

• ○ utcome (O): Fertility parameters such as time to pregnancy (TTP), ovarian reserve and use of assisted reproductive techniques.

• Pregnancy question

• ○ Population (P): Pregnant women with SpA.

• ○ Exposure/Intervention (I): Disease activity, inflammatory state or therapeutic management during pregnancy.

• ○ Comparator (C): Healthy pregnant women, women without inflammatory rheumatic disease or women not exposed to treatments, when reported.

• ○ Outcome (O): Disease activity scores, pregnancy course, maternal complications and foetal outcomes.

• Postpartum question

• ○ Population (P): Women with SpA during the postpartum period.

• ○ Exposure/Intervention (I): Breastfeeding, postpartum immune and hormonal changes and treatment resumption.

• ○ Comparator (C): Women without SpA or with continued versus interrupted therapy.

• ○ utcome (O): Postpartum disease flares, breastfeeding safety and maternal or neonatal complications.

These PICO questions guided both the literature search strategy and the subsequent data extraction and synthesis.

Inclusion and exclusion criteria

Studies were selected based on predefined inclusion and exclusion criteria. Eligible publications had to be written in English or French and specifically address one or more of the following periods: fertility, pregnancy or postpartum. We included all types of studies – observational studies, interventional trials, case series, case reports and systematic reviews – without restriction on study design. Furthermore, they had to focus on women diagnosed with SpA, including PsA, and explore the relationship between the disease and these reproductive phases.

Articles were excluded if they did not pertain to SpA or did not address fertility, pregnancy or the postpartum period based on their title and abstract. Additional exclusion criteria included the absence of an abstract, publication in languages other than English or French and article types such as correspondence, illustrations or research protocols. Studies focusing exclusively on male fertility were also excluded. Notably, no restriction was applied regarding the year of publication, allowing for an analysis of the evolution of scientific interest over time.

Study selection process

The selection of studies was carried out in two distinct stages. First, titles and abstracts were screened to eliminate publications that were clearly irrelevant. Second, full-text articles were reviewed to confirm eligibility based on the inclusion and exclusion criteria. Duplicate entries were identified and removed to ensure a unique dataset.

Data extraction

To ensure a structured and clinically meaningful synthesis, the extracted information was organized according to the PICO/PECO framework – Population, Exposure/Intervention, Comparator and Outcomes.

Publications were categorized into five main thematic areas:

• The impact of SpA on fertility and preconception period

• The impact of pregnancy on SpA activity

• The use and safety of therapeutic interventions during pregnancy

• Pregnancy outcomes associated with maternal SpA, including obstetric and neonatal complications and delivery methods

• The postpartum period, notably focusing on the impact of the postpartum period and breastfeeding on SpA activity.

Synthesis of findings

Extracted data were synthesized and compared thematically. Key findings were organized and presented in summary tables to facilitate interpretation and highlight patterns across the literature.

Results

Flow diagram

The literature search yielded 337 publications across 3 research equations: fertility, pregnancy and postpartum. After removing duplicates and applying the predefined inclusion and exclusion criteria, 167 articles were retained for full-text analysis. Among these, 34 focused on fertility and preconception, 42 addressed SpA disease activity during pregnancy, 103 the therapeutic management, 80 the pregnancy outcomes and 36 explored the postpartum period (Figure 1). Several studies covered multiple phases or themes simultaneously.

Figure 1.

Flow diagram.

Temporal and proportional analysis of publications

To contextualize this research, we conducted a broader PubMed search for ‘Spondylarthropathies’[MeSH], which yielded over 30,000 publications during the same period. Thus, articles specifically addressing the peripartum period represent less than 1% of the total scientific output on SpA, underscoring the relative underrepresentation of reproductive health in this field.

However, a temporal analysis revealed a clear upward trend. Between 2000 and 2010, fewer than 20 articles addressed SpA in the context of pregnancy or fertility. From 2011 to 2020, this number more than doubled, with a notable acceleration after 2015. In the last 5 years (2020–2025), over 60 articles were published on this topic, reflecting growing clinical awareness and research interest (Supplemental Figure 1). This increase coincides with the broader recognition of nr-axSpA and its predominance in women, as well as the expansion of biologic therapies requiring reproductive safety data. This evolution of publication volume suggests a paradigm shift: reproductive health in SpA is no longer a marginal concern but an emerging priority in rheumatology research. Nevertheless, the literature remains fragmented, with few large-scale prospective studies and limited data on newer therapeutic agents.

Fertility and preconception

The impact of SpA on female fertility remains a subject of debate. Earlier studies, such as those by Ostensen et al.,^{7 –10} suggested that fertility in women with SpA was comparable to that of the general population. However, more recent research has identified a reduction in ovarian reserve among affected women, evidenced by lower anti-Müllerian hormone (AMH) levels and diminished antral follicle counts.^11,12 TTP was found to be prolonged in this population, with 21%—32% of women experiencing a TTP exceeding 12 months.^2,13 In some cohorts, subfertility rates reached as high as 45%. ²

Several factors were associated with reduced fertility, including advanced maternal age, nulliparity, longer disease duration, chronic pelvic pain, fatigue and psychological distress.^14,15 Interestingly, no direct correlation was found between disease activity and AMH levels. ¹⁶

Regarding pharmacological influences, non-steroidal anti-inflammatory drugs (NSAIDs), particularly COX-2 inhibitors, were linked to delayed conception.^17,18 Corticosteroids at doses above 7.5 mg/day were also associated with reduced fertility. ¹⁴ Sulphasalazine appeared to have no negative impact and was considered safe when supplemented with folic acid. Anti-tumour necrosis factor (anti-TNF) agents did not impair fertility and may even facilitate conception by reducing systemic inflammation.^16,19

Complementary examinations, such as AMH testing, were frequently used to assess ovarian reserve in this population.^12,16

The most recent fertility-related original studies and their main findings are summarized in Table 1. Some of the included studies evaluated broader populations of women with chronic inflammatory arthritis rather than exclusively patients with SpA. In these cases, SpA patients represented a subgroup of the cohort and, when available, their number is specified.

Table 1.

Fertility in women with SpA: summary of key studies.

Author(s) and year	Study type	Population	N (patients)	Key findings	Impact on fertility
Wallenius et al., 2011 10	Population-based Norwegian National Population Registry	Chronic arthritides. Due to low numbers, the women with PsA, AS and UA were analysed as an entity labelled OCAs.	631 chronic arthritides, including 218 with OCA and 63,100 matching controls	Higher proportion of women with chronic inflammatory arthritides were nulliparous compared with the references, and relative fertility rates were reduced in all patient groups	Mild impairment overall
Henes et al., 2015 11	Case-control	RA, Behçet, SpA	33 RA, 32 SpA and 30 Behçet compared to healthy controls	Patients had significantly reduced AMH levels, with a median value for RA of 1.8 ng/mL (control 2.4 ng/mL; p = 0.009), for SpA of 1.5 ng/mL (control 2.3 ng/mL; p = 0.013) and for BD of 1.1 ng/mL (control 1.9 ng/mL; p = 0.007). HLA-B27 had a negative influence on ovarian reserve in SpA patients, whereas other serological parameters did not in the other diseases.	Reduced ovarian reserve
Yalçın Bahat et al., 2021 12	Case-control	r-axSpA	52 AS, 52 controls	Mean serum AMH levels were significantly lower in the study group when compared to the controls (2.203 ± 1.110 vs 1.188 ± 0.891, p < 0.001).	Reduced ovarian reserve
Ursin et al., 2021 13	Registry-based cohort (Norwegian)	axSpA	274 axSpA, 317 RA	TTP exceeded 12 months in 21% of women with axial SpA. Longer TTP was associated with older age, nulliparity and longer disease duration	Mild subfertility in axSpA
Alexander et al., 2021 20	Prospective cohort	Arthritis patients	129 arthritis patients (21 spA) and 198 controls without arthritis	Lower AMH than healthy controls. Long-term methotrexate use was not associated with an annual AMH decrease. No link between biologics and AMH decline	Reduced ovarian reserve. Relative neutral effect of biologics
Scrivo et al., 2022 16	Observational	SpA, PsA	20 SpA, 24 PsA and 44 healthy controls	AMH levels lower than healthy controls but not influenced by bDMARDs nor by disease activity	Reduced ovarian reserve. Relative neutral effect of biologics and disease activity
Hamroun et al., 2024 17	Prospective cohort (GR2)	SpA	88 SpA	Median TTP was 16.1 (95% CI (12.2–25.3)) months. The multivariate model found a significant association between TTP and age ((HR) (per year) 1.22 95% CI (1.08–1.40); p < 0.001) and the use of NSAIDs during preconception (HR 3.01 95% CI (2.15–3.85); p = 0.01).	NSAIDs negatively affect fertility
Micu et al., 2011 18	Observational	Inflammatory arthropathies	14 patients with inflammatory arthropathies (4 r-axSpA), 29 women with other musculoskeletal conditions and 449 women not exposed to NSAIDs	In 59 monitored cycles of patients with continuous NSAID exposure, 35.6% of LUF syndromes occurred, compared to 3.4% of LUF syndromes in untreated women (p < 0.001).	NSAIDs (particularly selective COX-2 inhibitors) exposed to LUF syndrome and may impair ovulation
Herzberger et al., 2025 19	Retrospective cohort	RA and PsA	4517 RA, 1415 PsA	Main outcome measure was positive pregnancy test rate. Both groups demonstrated lower rates of positive beta hCG after diagnosis, as compared to baseline rates before diagnosis. However, exposure to biological treatment did not negatively affect the likelihood of conception in either group. Beta hCG testing increased in both groups after initiation of biological treatments (RA, p < 0.01; PsA, p = 0.07).	Neutral effect of biologics

AMH, anti-Müllerian hormone; AS, ankylosing spondylitis; BD, Behçet disease; bDMARD, biological disease-modifying antirheumatic drug; CI, confidence interval; COX-2, cyclooxygenase-2; GR2, GR2 prospective cohort (study name); hCG, human chorionic gonadotropin (beta hCG); HLA-B27, human leukocyte antigen B27; HR, hazard ratio; LUF, luteinized unruptured follicle; NSAID, non-steroidal anti-inflammatory drug; OCAs, other chronic arthritides; p, p-value; PsA, psoriatic arthritis; RA, rheumatoid arthritis; r-axSpA, radiographic axial spondyloarthritis; SpA, spondyloarthritis; TTP, time to pregnancy; UA, unspecified arthritis.

Disease activity during pregnancy

Disease activity during pregnancy varied by SpA subtype. In women with PsA, 75%—80% experienced remission or low disease activity throughout gestation.^21,22 In axSpA, most studies report overall stable disease activity during pregnancy, although disease flares occur in a subset of patients (approximately 25%—30%), often peaking in the second trimester or following treatment discontinuation.^23,24 Several studies noted that disease activity tended to peak in the second trimester before declining in the third.^25,26 In approximately 21% of cases, pregnancy coincided with the initial onset of SpA symptoms. ⁹ The recent GR2 study ²⁷ is a French multicentre prospective cohort that included 71 pregnancies in women with SpA. At baseline (first trimester), disease activity was generally low, with a mean Ankylosing Spondylitis Disease Activity Score including C-reactive protein (ASDAS-CRP) of 1.9 ± 1.2 and a mean Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) of 3.0 ± 2.1. These indices remained stable throughout pregnancy: ASDAS-CRP: 1.9 ± 1.2 (T1), 2.0 ± 1.0 (T2), 1.7 ± 1.0 (T3), p = 0.60 and BASDAI: 3.0 ± 2.1 (T1), 2.8 ± 2.0 (T2), 2.9 ± 2.1 (T3), p = 0.92. Despite this apparent stability, disease relapse occurred in 28 out of 71 pregnancies (39.4%), defined by treatment intensification and/or a significant increase in disease activity scores. Of these, 20 patients required therapeutic escalation. In multivariable analysis, a history of biologic DMARD (bDMARD) use was significantly associated with relapse during pregnancy (odds ratio (OR) 5.4, 95% CI 1.1–27.3). These findings suggest that although most women with SpA maintain stable disease activity during pregnancy – particularly when treated with TNF inhibitor (TNFi) use (62% of patients received a TNFi during gestation) – nearly 40% remain at risk of flare. This underscores the importance of close monitoring and individualized therapeutic planning, especially for patients with prior bDMARD exposure. ²⁷

Key original studies on SpA disease activity during pregnancy are summarized in Table 2.

Table 2.

Disease activity during pregnancy in women with SpA.

Author(s) and year	SpA subtype	Study type	N (patients)	Disease activity during pregnancy	Key findings
Ostensen, 1992 21	r-axSpA, PsA	Review of retrospective and prospective studies	77 r-axSpA, 15 PsA	Patients with PsA improved or even remitted in 80% of the pregnancies, whereas 80% of the AS patients had unaltered or aggravated disease symptoms	Pregnancy may induce remission in PsA. Aggravated or stable disease activity in r-axSpA
Østensen et al., 2004 28	r-axSpA	Prospective cohort	9 r-axSpA, 10 RA	Higher disease activity scores were found in the patients with r-axSpA, with a frequent increase of disease activity in the second trimester and mitigation of symptoms in the third trimester	Increased activity during the second trimester
Lui et al., 2011 29	r-axSpA	Case-control	19 r-axSpA, 33 controls	Improvement in pain was reported in 51% of r-axSpA patients, predominantly in the first trimester. Significantly more improvement in pain than in stiffness in all 3 trimesters	Pregnancy does not substantially aggravate disease activity or severity
Timur et al., 2016 30	r-axSpA	Retrospective cohort	60 pregnant women: 20 r-axSpA and 40 controls	Fourteen (70%) of 20 patients displayed decrement in ASDAS during pregnancy. Course of AS was unaltered in 6 (30%) of 20 patients during pregnancy	Pregnancy does not substantially aggravate disease activity or severity
Van den Brandt et al., 2017 31	axSpA	Prospective cohort	75 RA, 61 axSpA	Disease flares during pregnancy occurred in 29% of patients with RA and in 25% of patients with axSpA. In both diseases, active disease and TNFi discontinuation in early pregnancy were identified as risk factors for disease flares during pregnancy	To stabilize disease activity and to prevent a flare during pregnancy, tailored medication, including TNF inhibitors, should be considered beyond conception
Polachek et al., 2017 32	PsA	Prospective cohort	29 PsA	Arthritis improved or was stable in low activity in 24 (58.5%) of pregnancies	Arthritis activity trends toward a favourable course, while the skin disease shows a favourable course during pregnancy
Berman et al., 2018 33	PsA	Observational	25 PsA	Continuation of biologics therapy was associated with a low level of disease activity and a low probability of flare during pregnancy. Stopping treatment with biologics before pregnancy is associated with flare during pregnancy	No significant change in disease activity throughout pregnancy
Ursin et al., 2018 26	axSpA	RevNatus, Norwegian nationwide register	166 axSpA	Stable overall. Small increase in disease activity in the second trimester	Majority experienced stable, low disease activity
Polachek et al., 2019 34	PsA	Case-control questionnaire-based study	74 PsA, 77 controls	Most (58%) patients reported favourable joint activity during pregnancy	Stable disease in most cases
Ursin et al., 2019 22	PsA	RevNatus, Norwegian nationwide register	103 PsA	About 75% of the women were in remission or had low disease activity during and after pregnancy, according to DAS28. Although altogether stable, disease activity decreased in pregnancy. Women using TNFi in pregnancy had significantly lower disease activity	Disease activity remained low throughout pregnancy
Meissner et al., 2022 25	axSpA	EuNeP: EGR2 (France), RePreg (Switzerland), RevNatus (Norway) and Rhekiss (Germany)	304 axSpA	Pooled mean CRP was 4 mg/L before conception, peaking in the second trimester at 9.4 mg/L. BASDAI was below 4 at all time points.	Overall stable disease activity
Couderc et al., 2025 27	SpA	Prospective cohort (GR2)	71 SpA, 53 RA	Disease activity indexes remained stable in the 1st, 2nd and 3rd trimesters. Disease relapse during pregnancy occurred in 28 (39%) SpA patients, among whom 30 (24%) requiring a treatment intensification. In multivariable analysis, factors associated with disease relapse were nulliparity (OR: 6.5, 95% CI: 1.1–37.9) and a history of bDMARD use (OR: 5.4, 95% CI: 1.1–27.3) for SpA patients	Disease activity remained stable during pregnancy in women with SpA, but almost a quarter required major changes to their treatment

AS, ankylosing spondylitis; ASDAS, Ankylosing Spondylitis Disease Activity Score; axSpA, axial spondyloarthritis; BASDAI, Bath Ankylosing Spondylitis Disease Activity Index; bDMARD, biological disease-modifying antirheumatic drug; CRP, C-reactive protein; DAS28, 28-joint Disease Activity Score; EuNeP, European Network of Pregnancy Registries in Rheumatology; GR2, GR2 prospective cohort (study name); OR, odds ratio; PsA, psoriatic arthritis; RA, rheumatoid arthritis; r-axSpA, radiographic axial spondyloarthritis; RevNatus, Norwegian pregnancy register for rheumatic diseases (RevNatus); SpA, spondyloarthritis; TNFi, tumour necrosis factor inhibitor.

Pregnancy outcomes

Obstetric complications in women with SpA have been frequently reported. Maternal risks included preeclampsia, gestational diabetes, hypertension, placenta previa and oligohydramnios.^2,3,25 Foetal risks encompassed prematurity, intrauterine growth restriction (IUGR), low birth weight and increased rates of caesarean delivery.^23,35 Caesarean sections were more common and were often scheduled in advance. ³⁶ Nevertheless, several studies confirmed that uncomplicated pregnancies are achievable when disease activity is well controlled.^9,30,34

A large multicentre registry study published by Meissner et al. ²⁵ in 2022 investigated pregnancy outcomes in 304 women with axSpA, pooling data from 4 European rheumatology registries. A total of 332 pregnancies were analysed; the live birth rate was 98.8%, indicating highly favourable outcomes. Pre-eclampsia occurred in 2.2% of cases, preterm birth in 4.9%, low birth weight in 3.1% and small-for-gestational-age neonates in 9.5%. Caesarean delivery was reported in 27.7% of pregnancies, with nearly half of these being emergency procedures. Disease activity remained well controlled throughout gestation. The pooled mean CRP increased from 4 mg/L before conception to 9.4 mg/L in the second trimester, while BASDAI scores remained below 4 at all time points. Prior to pregnancy, 53% of patients were treated with TNFi, with 27.5% continuing treatment into the first trimester and 21.4% into the third. Most patients (78.4%) received pre-conception counselling and had planned pregnancies. These findings suggest that with close disease monitoring, individualized treatment strategies and multidisciplinary care in expert centres, pregnancy outcomes in women with axSpA can approach those of the general population.

Available original studies on pregnancy outcomes and their main findings are summarized in Table 3.

Table 3.

Adverse pregnancy and neonatal outcomes in axial SpA and PsA.

Author(s) and year	Design/country	Population (inclusion criteria)	N (patients/pregnancies)	Control group (size/matching)	Maternal outcomes	Neonatal outcomes
Timur et al., 2016 30	Retrospective case-control/Turkey (single tertiary referral centre)	Pregnant women diagnosed with ankylosing spondylitis (modified New York criteria); most recent pregnancy per patient; ASDAS-CRP used	r-axSpA group: 20 pregnancies; Total: 60 (20 AS, 40 controls)	Hospital controls: 40 healthy pregnant women (2:1 matching by delivery sequence at same unit)	No significant increase in overall adverse maternal outcomes vs controls	No significant differences in birthweight, Apgar scores, NICU admissions or major neonatal outcomes; mean birthweight AS ≈ 3088 g vs controls ≈ 3273 g
Jakobsson et al., 2016 37	Nationwide register study/Sweden	Women with ankylosing spondylitis identified in national registers delivering during study period	Total included: 1470 singleton births; r-axSpA deliveries: 388 deliveries in 301 women	Population controls from general population; matched design; GEE used to account for clustering	Caesarean: Emergency CS 16.5% in AS vs 6.5% controls (OR 3.00, 95% CI: 2.01–4.46); Elective CS 9.8% vs 6.9% (OR 1.66, 95% CI: 1.09–2.54); use of epidural reported	Preterm <37 w: 9.0% in AS vs 4.9% controls (OR 1.92, 95% CI: 1.17–3.15); SGA: 3.1% AS vs 1.5% controls (OR 2.12, 95% CI: 1.00–4.50)
Remaeus et al., 2019 38	Nationwide cohort study/Sweden	All singleton pregnancies 1997–2014; exposure = PsA in national registers	Total: 41,485 pregnancies; PsA: 541; Unexposed: 40,944	Population controls: 40,944 non-PsA pregnancies; adjusted for maternal factors and year	Increased risks: preterm birth adjusted OR 1.63 (1.17–2.28); elective CS adjusted OR 1.47 (1.10–1.97); emergency CS adjusted OR 1.43 (1.08–1.88)	Neonatal: no increased risk for pre-eclampsia, stillbirth or other infant outcomes apart from preterm birth
Unal et al., 2020 39	Retrospective single-centre cohort/Turkey (2006–2019)	Pregnant patients with clinician-diagnosed ankylosing spondylitis; pregnancies ⩾24 weeks included	N  = 52 pregnancies	No external control group (single-centre descriptive analysis)	Preterm delivery: 9/52 (17.3%); Preeclampsia: 4/52 (7.7%); IUGR: 7/52 (13.4%); hospitalization during pregnancy 12/52 (23%); no significant differences by medication or disease duration	Median birthweight: 2950 g; Apgar <7: 3/52 (5.7%); NICU admissions: 11/52 (21.1%); IUGR 7/52 (13.4%)
Smith et al., 2020 40	Prospective cohort (OTIS Autoimmune Project)/USA and Canada (2004–2018)	Pregnant women enrolled before 20 w with physician-validated PsA or AS; disease activity measured (HAQ, RAPID3)	Total: 963 women; PsA: 117; r-axSpA: 129; Controls: 717	Healthy comparison cohort: n = 717; analyses adjusted for maternal age, BMI, parity, race, SES, smoking, comorbidity	PsA maternal: moderate preterm delivery (32–36 w) aRR 1.81 (1.01–3.26); preterm labour aRR 2.05 (1.21–3.48); oligohydramnios aRR 3.79 (1.34–10.74); Caesarean aRR 1.63 (1.26–2.12). AS: disease-activity signals (see text)	PsA neonatal: increased moderate preterm births and related outcomes; oligohydramnios noted. AS neonatal: increased NICU hospitalization aRR 1.67 (1.05–2.67); very preterm/VLBW rare
Mørk et al., 2021 41	Nationwide register-based cohort/Denmark (1997–2016)	All singleton births in Danish MBR; exposure = SpA diagnosis (AS, PsA, ReA, SpA-IBD, uSpA) prior to delivery; excluded pregnancies with missing gestational age	Total cohort: 1,199,610 pregnancies; SpA: 3728 pregnancies in 2479 women; No SpA: 1,195,882 pregnancies	General pregnant population from national registers (women without SpA; 1,195,882 pregnancies); comparisons adjusted for maternal age, smoking, BMI, year of birth	Higher rates/ORadj: preterm birth overall 1.55 (1.33–1.80); moderate preterm 1.56 (1.33–1.83); very preterm 1.47 (1.04–2.08); elective CS 1.44 (1.26–1.64); emergency CS 1.17 (1.04–1.33); epidural use 1.11 (1.02–1.20); preeclampsia not significantly increased in main analysis 1.22 (0.99–1.50)	Neonatal: higher proportion of preterm births; SGA not significantly increased ORadj 1.09 (0.97–1.23); Apgar < 7 and perinatal mortality not significantly different
Remaeus et al., 2022 42	Nationwide registry-based matched cohort/Sweden	Singleton pregnancies with physician-recorded PsA; analyses stratified by presence, timing and type of antirheumatic treatment (proxy for disease severity)	PsA: 921 pregnancies; Controls: 9210 (1:10 matched on maternal age, year, parity)	Matched population controls: 9210 non-PsA pregnancies (matched 1:10 on age, year, parity); adjusted for BMI, smoking, education, country of birth	Increased risks vs controls: preterm birth aOR 1.69 (95% CI: 1.27–2.24); elective CS aOR 1.77 (1.43–2.20); emergency CS aOR 1.42 (1.10–1.84). Risks varied by treatment: antirheumatic treatment during pregnancy associated with highest preterm risk aOR 2.30 (1.49–3.56); biologic exposure during pregnancy associated with markedly higher preterm risk aOR 4.49 (2.60–7.79). Increased preterm risk concentrated in first pregnancies	Neonatal: Preterm birth increased (especially with treatment exposure and in primiparas); other neonatal outcomes not consistently elevated in overall analyses
Meissner et al., 2022 25	Prospective pooled pregnancy registries/Europe	Women with physician-diagnosed axSpA enrolled in pregnancy registries	304 women; 332 pregnancies	No population control (pooled registry descriptive analysis)	Caesarean delivery: 27.7% (of pregnancies)	Neonatal: Live birth 98.8%; pre-eclampsia 2.2%; preterm birth (<37 w) 4.9%; Low birth weight 3.1%; SGA 9.5%
Redeker et al., 2022 43	Claims-based matched cohort/Germany (health insurance data 2006–2019)	Women with axSpA identified by diagnostic codes with deliveries during study period	AxSpA: 1021 pregnancy outcomes (928 deliveries, 80 abortions, 13 ectopic pregnancies)	Age and calendar year matched controls: 10,210 pregnancy outcomes (9488 deliveries, 615 abortions, 147 ectopic pregnancies)	Maternal: elective CS increased in axSpA vs controls aOR 1.52 (1.25–1.85). Treatment-stratified: preterm birth higher under NSAIDs vs no anti-inflammatory treatment aOR 2.22 (1.09–4.52); preterm under NSAIDs + csDMARDs/steroids/analgesics aOR 4.01 (1.93–8.34)	Neonatal: preterm birth and SGA increased under certain treatment combinations (SGA aOR 3.22 (1.34–7.73) under NSAIDs + csDMARDs/steroids/analgesics); bDMARDs: non-significant increases observed
Secher et al., 2022 44	Collaborative matched cohort (registry linkage)/Sweden and Denmark (2006–2018)	Singleton pregnancies with AxSpA or PsA identified in rheumatology registers and linked to national medical birth registers; data on antirheumatic treatment and disease activity	Pooled SpA: 1308 pregnancies (AxSpA 819; PsA 489)	Matched controls: ~13,080 pregnancies (1:10 matching on maternal age, parity, year of delivery)	Pre-eclampsia: AxSpA aOR 1.17 (0.76–1.78); PsA aOR 1.85 (1.10–3.12). PsA prepregnancy monotherapy associated with pre-eclampsia aOR 2.72 (1.44–5.13); axSpA – no significant treatment-related increase	Neonatal: not detailed for pooled SpA in the abstract; primary finding = increased maternal pre-eclampsia risk in PsA
Gangbe et al., 2022 45	Retrospective population-based analysis (HCUP-NIS)/USA	Hospital delivery admissions with ICD coding for PsA (ICD-9/10 code 696.0) compared with all other deliveries in the database	PsA: 419 deliveries; Controls: 9,096,369 deliveries	All non-PsA deliveries in HCUP-NIS (9,096,369); multivariable adjustment for demographic and clinical confounders	Maternal: higher rates of PIH aOR 1.58 (95% CI: 1.13–2.21). No significant differences in rates of preeclampsia, gestational diabetes, placenta previa, abruption, or modes of delivery (p > 0.05)	Neonatal: higher risk of SGA aOR 2.42 (95% CI: 1.49–3.93). No significant differences in preterm delivery, IUFD, congenital anomalies
Morin et al., 2023 46	Nationwide register-based cohort; temporal trend analysis/Sweden (births April 1, 2007–December 31, 2020)	Births to women with axSpA recorded in the Swedish MBR; linked to National Patient Register, Prescribed Drug Register and demographic registers; outcomes analysed over time and by antirheumatic treatment	axSpA births: 1580 singleton births (selected from 5666 recorded births); Comparators: 15,792 matched births	General population comparators matched 1:10 on year of delivery, maternal age and parity (15,792 comparators)	Increased risks vs comparators (modified Poisson regression): preterm birth RR 1.43 (95% CI: 1.13–1.80); pre-eclampsia RR 1.44 (1.08–1.92); elective caesarean delivery RR 1.59 (1.37–1.84). Temporal trends: annual decreases (~0.5 percentage points/year) in preterm birth, infant infection and caesarean delivery over the study period; concurrent increase in TNF inhibitor use during pregnancy	Neonatal: Serious infant infection increased RR 1.29 (95% CI: 1.05–1.59); SGA not increased; overall frequencies of preterm birth, infant infection and caesarean delivery declined over time.
Amikam et al., 2024 3	Retrospective population hospital discharge analysis (HCUP-NIS)/USA	Deliveries with ICD coding for AS 2004–2014	Total cohort: 9,096,788 deliveries; r-axSpA group: 383 women (3.8 per 100,000)	All non-AS delivery hospitalizations in HCUP (9,096,405 deliveries)	Maternal: Caesarean delivery increased in AS aOR 1.47 (1.14–1.91); gestational diabetes aOR 1.55 (1.02–2.33); placenta previa aOR 3.60 (1.60–8.12)	Neonatal: SGA aOR 2.19 (1.22–3.93); IUFD/stillbirth aOR 3.46 (1.11–10.83)

aRR, adjusted risk ratio; AS, ankylosing spondylitis; ASDAS-CRP, Ankylosing Spondylitis Disease Activity Score including C-reactive protein; AxSpA, axial spondyloarthritis; bDMARD, biological disease-modifying antirheumatic drug; BMI, body mass index; CS, caesarean section; csDMARD, conventional synthetic disease-modifying antirheumatic drug; DMARD, disease-modifying antirheumatic drug; GEE, generalized estimating equations; HAQ, Health Assessment Questionnaire; HCUP-NIS, Healthcare Cost and Utilization Project – Nationwide Inpatient Sample; IUFD, intrauterine foetal death; IUGR, intrauterine growth restriction; MBR, Medical Birth Register; NICU, neonatal intensive care unit; NSAID, non-steroidal anti-inflammatory drug; OR, odds ratio; ORadj/aOR, adjusted odds ratio; PIH, pregnancy-induced hypertension; PsA, psoriatic arthritis; RAPID3, Routine Assessment of Patient Index Data 3; r-axSpA, radiographic axial spondyloarthritis; ReA, Reactive Arthritis; RR, risk ratio; SES, socioeconomic status; SGA, small for gestational age; SpA, spondyloarthritis; uSpA, Undifferentiated Spondyloarthritis; VLBW, Very Low Birth Weight (defined as birth weight <1500 g).

Therapeutic management during pregnancy

Therapeutic management during pregnancy require careful consideration. In recent years, certolizumab pegol (CZP) has emerged as the preferred anti-TNF agent due to its minimal placental transfer, as demonstrated in the CRIB study, which measured CZP levels in maternal blood, cord blood and infant plasma and found negligible foetal exposure even during the third trimester. ⁴⁷ In contrast, other biologics were typically discontinued in late pregnancy and need more data. NSAIDs were considered safe during the first and second trimesters but were contraindicated in the third due to risks of foetal complications. Methotrexate was strictly contraindicated and recommended to be stopped before conception. Sulphasalazine was deemed compatible with pregnancy, and corticosteroids were used cautiously, as high doses above 10 mg/day were associated with increased risk of prematurity.

The 2024 update of the EULAR recommendations for the use of antirheumatic drugs in reproduction, pregnancy and lactation provides a comprehensive framework for managing patients with inflammatory rheumatic diseases, including axial and peripheral SpA. ⁶ These guidelines were developed through a systematic literature review and expert consensus, and they reflect the most current safety data available.

For patients with SpA, the recommendations emphasize the following:

• TNFi remain the cornerstone of treatment during pregnancy when disease control is needed. CZP is considered the safest option due to its minimal placental transfer. All TNFi may be used throughout pregnancy when needed to effectively control maternal disease, rather than being limited to the first trimester. For non-TNFi biologics, EULAR recommends they may be used if needed to effectively control maternal disease (lower level of evidence than TNFi) and specifically lists abatacept, ixekizumab, sécukinumab and ustekinumab; agents with very limited/no pregnancy safety data (guselkumab and risankizumab) should be reserved for situations where no other pregnancy-compatible medication can adequately control disease.

• NSAIDs are permitted during the first and second trimesters but should be avoided in the third trimester due to risks of premature closure of the ductus arteriosus and oligohydramnios. Regarding its use during preconception, consider discontinuation if prolonged time to conception.

• Corticosteroids may be used throughout pregnancy, with preference for the lowest effective dose. High-dose regimens are associated with increased risks of gestational diabetes and preterm birth.

• Methotrexate and léflunomide are contraindicated during pregnancy and require discontinuation in advance of conception.

• Sulphasalazine is considered safe and may be continued during pregnancy and lactation.

• During lactation, all bDMARDs are considered compatible with breastfeeding (including TNFi and non-TNFi such as IL-17i), based on minimal milk transfer and low infant oral bioavailability. Methotrexate and léflunomide remain contraindicated. EULAR states MTX ⩽25 mg weekly may be considered during breastfeeding only if no alternative compatible drug can be used (reflecting limited data).

The EULAR task force also stressed the importance of preconception counselling, shared decision-making and multidisciplinary care involving rheumatologists, obstetricians and midwives. These recommendations align with the trends observed in our review and provide a validated framework for clinical decision-making in women with SpA during the reproductive period.

CZP remains the biologic with the most reassuring pharmacokinetic data during pregnancy due to its minimal placental transfer. However, recent evidence and updated guidelines support a less restrictive approach to biologic therapy overall. Several biologics may be continued during pregnancy when disease control requires it, particularly during early and mid-pregnancy. Treatment decisions should therefore be individualized based on disease activity and maternal–foetal considerations.

Postpartum

The postpartum period was associated with a high risk of disease flares, particularly in women with PsA and r-axSpA.^24,33,48 However, patients who continued biologic therapy postpartum tended to maintain stable disease activity.^26,32 Breastfeeding appeared to have a protective effect, with lower disease activity observed among lactating women. ⁴⁹

The most recent original studies about disease activity and breastfeeding in women with SpA and their findings are summarized in Table 4.

Table 4.

Postpartum disease activity and breastfeeding in women with SpA.

Author(s) and year	SpA subtype	Study type	N (patients)	Postpartum disease activity	Key findings
Lui et al., 2011 29	r-axSpA	Case-control	19 r-axSpA, 33 controls	Postpartum pain scores in AS returned to prepartum levels in general.	Pregnancy does not substantially aggravate disease activity or severity in r-axSpA
Polachek et al., 2017 32	PsA	Prospective cohort	29 PsA	16 (40%) pregnancies had an unfavourable course during the 1-year postpartum (all of them manifesting as worsening) and 3 (7.5%) had a mixed course of improvement followed with worsening. A subgroup analysis of the 10 women with more than 1 pregnancy revealed that 80% (8 women) had a different course of arthritis activity during their different pregnancies and 60% had a different course during the different postpartum periods.	Most patients did not flare during postpartum
Ursin et al., 2018 26	axSpA	RevNatus, Norwegian nationwide register	166 axSpA	The lowest disease activity was found 1 year postpartum (mean BASDAI 3.35).	Stable, low disease activity
Berman et al., 2018 33	PsA	Retrospective case series	25	Disease activity of 16 (48%) worsened during the postpartum period. There was a trend toward aggravation of disease activity during the postpartum period compared to the pre-pregnancy level. A statistically significant aggravation postpartum was shown compared to the third trimester	Continuation of biologics therapy was associated with a low level of disease activity and a low probability of flare during pregnancy. Stopping treatment with biologics before pregnancy is associated with flare during the postpartum period
Ursin et al., 2019 22	PsA	RevNatus, Norwegian nationwide register	103 PsA	About 75% of the women were in remission or had low disease activity during and after pregnancy according to DAS28. Although altogether stable, disease activity increased within 6 months postpartum and was significantly higher than 6 weeks postpartum (mean DAS28 2.71 vs 2.45, p = 0.016). Women using TNFi in pregnancy had significantly lower disease activity.	Disease activity was highest 6 months postpartum, but altogether low and stable in the period from planning pregnancy to 1 year after delivery
Polachek et al., 2019 34	PsA	Case-control	74 PsA, 77 controls	50% reported worsening during the 1st postpartum year.	No major postpartum complications with relatively stable disease
Grøntvedt et al., 2025 49	axSpA	RevNatus, Norwegian nationwide register	363 axSpA	The proportion of patients breastfeeding at the postpartum follow-ups was 86% (347 women) at 6 weeks, 70% (221 women) at 6 months and 38% (104 women) at 12 months, respectively. ASDAS-CRP was higher for the non-breastfeeding group at 6 weeks (2.6 vs 2.2) and at 6 months (2.6 vs 2.1), and BASDAI was higher for the same group at 6 months (4.1 vs 3.2). CRP (mg/L) was significantly higher among the non-breastfeeding at 6 months (5.3 vs 3.3). VAS pain scores were higher for the non-breastfeeding group at 6 weeks (41.0 vs 31.6) and 6 months (43.9 vs 31.0). VAS fatigue was higher for the same group at 6 months (46.8 vs 37.8).	Breastfeeding associated with reduced disease-activity scores

AS, ankylosing spondylitis; ASDAS-CRP, Ankylosing Spondylitis Disease-Activity Score including C-reactive protein; axSpA, axial spondyloarthritis; BASDAI, Bath Ankylosing Spondylitis Disease-Activity Index; CRP, C-reactive protein; DAS28, 28-joint Disease Activity Score; HAQ, Health Assessment Questionnaire; PsA, psoriatic arthritis; r-axSpA, radiographic axial spondyloarthritis; RevNatus, Norwegian pregnancy register for rheumatic diseases; TNFi, tumour necrosis factor inhibitor; VAS, Visual Analogue Scale.

Regarding treatment compatibility with breastfeeding, CZP and adalimumab were considered safe during breastfeeding.^50,51 Methotrexate remained contraindicated, and data on other biologics were limited. NSAIDs and corticosteroids were generally compatible with lactation, although timing precautions were advised for corticosteroid administration.

In the 2024 update of the EULAR recommendations for the use of antirheumatic drugs before conception, during pregnancy and lactation, special emphasis was placed on breastfeeding. ⁶ The task force reaffirmed that women should not be discouraged from breastfeeding if they are receiving compatible medications, and that postpartum therapeutic decisions should be made jointly with the patient, balancing the risks of drug exposure against those of uncontrolled disease. In this latest version, all bDMARDs are now considered compatible with lactation, based on their pharmacokinetic properties and minimal transfer into breast milk. Data remain more limited for non-TNFi biologics, but their structural and molecular similarities to TNF inhibitors support a low-risk assessment. Furthermore, the updated recommendations confirm that NSAIDs are compatible with breastfeeding and that methotrexate is contraindicated.

Discussion

This systematic review consolidates current knowledge on the impact of SpA during the peripartum period, encompassing fertility, pregnancy and postpartum outcomes. Our findings highlight both reassuring evidence and important gaps that warrant further investigation.

Several robust findings emerge from the literature. Fertility is not universally impaired in women with SpA, yet TTP is frequently prolonged, and ovarian reserve may be reduced in some cohorts.^{11 –14} During pregnancy, disease activity varies by subtype: PsA often improves,^21,22,32,33 whereas axSpA – particularly r-axSpA – tends to flare, most commonly in the second trimester.^{23 –25,27} When well-controlled, pregnancy outcomes are generally favourable, with live birth rates approaching those of the general population.^25,34 Nevertheless, active disease is consistently associated with increased risks of obstetric complications, including preeclampsia, IUGR, prematurity and caesarean delivery.^{2,3,35,36,40,46} Postpartum, flares are common, especially in women who discontinue biologic therapy,^26,32,33,48 whereas breastfeeding appears protective. ⁴⁹ CZP stands out as the safest biologic option during pregnancy and lactation, with reassuring pharmacokinetic data from the CRIB study.^47,51

Despite these advances, the evidence base remains limited and heterogeneous. Most studies are observational, with small sample sizes and variable outcome definitions, which limits comparability.^24,52 Data on newer biologics and small molecules are scarce, ⁵¹ and long-term child development outcomes remain largely unexplored. Postpartum disease activity is inconsistently reported, and interpretation of imaging findings – particularly sacroiliac MRI – can be confounded by physiological changes after delivery.^53,54 Finally, access to preconception counselling and multidisciplinary follow-up is not uniformly reported,^25,35 leaving gaps in understanding the broader determinants of outcomes.

These findings underscore several key messages for clinical practice. First, preconception counselling should be systematically offered, addressing fertility, medication safety and disease monitoring.^2,17,25 Second, maintaining low disease activity is crucial, as active disease rather than treatment per se appears to be the main driver of adverse outcomes.^36,40,46 Third, therapeutic planning should integrate evidence-based guidance: CZP, other TNFi and, in selected cases, other bDMARDs may be continued,^6,47 while methotrexate and leflunomide must be discontinued well in advance of conception. ⁶ Postpartum care requires vigilance, particularly regarding disease flares and breastfeeding support.^26,32,33,49 Importantly, our synthesis suggests a window of vulnerability in the postpartum period, during which disease flares are frequent and careful follow-up is essential. Standardized protocols for monitoring – combining clinical scores (BASDAI, ASDAS), inflammatory markers (CRP), and, where appropriate, imaging – could improve consistency of care and early detection of complications. Multidisciplinary coordination between rheumatologists, obstetricians, midwives and paediatricians is essential to optimize outcomes for both mother and child.^6,50,52

Future studies should aim to fill critical gaps. Large prospective registries with standardized outcome measures are needed to better define the risks of newer biologics and small molecules, as well as the long-term health of children exposed in utero or via breastfeeding.^25,51 More work is also required to clarify the biological mechanisms linking chronic inflammation to subfertility and adverse pregnancy outcomes.^{11 –14,17} Finally, interventional studies testing structured multidisciplinary care models could provide evidence on how best to deliver integrated reproductive health services for women with SpA.^6,52

Conclusion

This systematic review highlights the complex interplay between SpA and the peripartum period. While fertility is not universally impaired, women with SpA – particularly those with axial involvement – may experience delayed conception and reduced ovarian reserve. Pregnancy outcomes are influenced by disease activity and therapeutic management, with increased risks of preeclampsia, prematurity and caesarean delivery observed in several cohorts. The postpartum period remains a vulnerable phase, marked by frequent disease flares and diagnostic challenges.

Despite these risks, favourable maternal and neonatal outcomes are achievable when the disease is well controlled and care is coordinated across specialties. Anti-TNF agents such as CZP offer safe and effective options during pregnancy and lactation, and breastfeeding appears to confer protective benefits.

This review underscores the need for prospective, standardized studies to better define clinical pathways and optimize reproductive care in women with SpA. Rheumatologists, obstetricians and midwives must collaborate to ensure informed decision-making and tailored support throughout the reproductive journey.