Background

Rheumatoid arthritis (RA) is a chronic autoimmune disease, which is characterized by synovial inflammation, cartilage damage, and bone erosion, and leads to severe physical disability [1]. The estimated prevalence of RA is 0.5-1.0% worldwide, and women are twice as likely to suffer from RA than men, with most cases occurring in women of childbearing age [1]. RA impairs the fertility, and compared to the general population, pregnancy outcomes are not satisfactory in women with RA, especially in those with high disease activity [2].

Several studies have reported the correlation between RA and adverse pregnancy outcomes [3,4,5,6]. A meta-analysis has reported that maternal RA increased the risk of autism spectrum disorders in offspring [7]. However, this meta-analysis has not reported the maternal outcomes and other fetal outcomes [7]. A meta-analysis performed by Huang et al. showed that maternal RA was significantly correlated with an increased risk of adverse maternal and fetal outcomes [8]; however, the association between disease activity and pregnancy outcomes was not explored in their meta-analysis.

Existing studies have shown that higher disease activity of RA was correlated with the higher risk of adverse pregnancy outcomes [9, 10]. A study by de Man et al. has reported that pregnancy outcomes of women with well-controlled RA was comparable with those of the general population [11]. A study by Langen et al. showed no association between disease activity and pregnancy outcomes in RA women, but they found that medication discontinuation increased the odds of adverse pregnancy outcomes at delivery [12]. Previous meta-analyses have reported pregnancy outcomes of women with and without RA without considering the disease severity [7, 8]. Given that it is important to assess pregnancy outcomes by disease status, there is a need to further examine the effect of disease severity on pregnancy outcomes within a population of women with RA.

We performed a systematic review and meta-analysis based on current available publications to systematically assess the association between the disease activity of RA and pregnancy outcomes in women with RA. We also examined the pregnancy outcomes in women with and without RA. The combination and analysis of data on this issue may provide useful clinical management and counselling for RA women.

Methods

The standard Cochrane methods were used in this meta-analysis, which performed according to Preferred Reporting Items for Systemic Reviews and Meta-Analyses (PRISMA) guideline [13]. The protocol of this meta-analysis was registered at PROSPERO (registration number: CRD42023402272).

Literature search strategy

Two researchers (JML and LX) searched the studies in four English databases (PubMed, Embase, Web of Science, and Cochrane Library) and three Chinese databases (China National Knowledge Infrastructure [CNKI], VIP, and Wan Fang) from inception to August 13, 2023. The search strategy included: “Arthritis, Rheumatoid” OR “Rheumatoid Arthritis” AND “Pregnancy” OR “Pregnancies” OR “Gestation” OR “Pregnancy Outcome” OR “Pregnancy Outcomes” OR “Outcome, Pregnancy” OR “Outcomes, Pregnancy” OR “Maternal outcomes” OR “Fetal outcomes”.

Inclusion and exclusion criteria

Studies were included if they met all the following criteria: (1) population: pregnant women with and without RA; (2) exposure and comparator: women with RA vs. women without RA, RA women with high disease activity vs. RA women with low disease activity; (3) outcome: adverse maternal and/or fetal outcomes; (4) study: observational studies; (5) language: published in English or Chinese.

Disease activity was assessed using Disease Activity Score-28 (DAS28), Health Assessment Questionnaire-Disability Index (HAQ-DI), pain score (PS), and patient’s global scale (PGS). DAS28 > 3.2 and HAQ-DI > 0.5 were defined as high disease activity [14].

Maternal outcomes included preeclampsia, gestational diabetes, hypertension, spontaneous abortion (pregnancy loss before 28 weeks of gestation), cesarean delivery, postpartum infection, postpartum hemorrhage, and maternal depression. Fetal outcomes included premature delivery (delivery at 28–37 weeks of gestation), stillbirth (delivery of a dead fetus at > 27 weeks of gestation), neonatal death within 30 days of birth, small for gestational age (SGA), birth weight, LBW (birth weight < 2500 g), low Apgar score (score at 5 min < 7), requiring intensive care, infantile autism (IA), congenital abnormalities, RA, Juvenile idiopathic arthritis (JIA), diabetes type 1, asthma, and epilepsy.

Studies were excluded if they met one of the following criteria: (1) animal studies; (2) topic not meeting the requirements; (3) with incomplete data (data categories not meeting our requirements) or unable to extract data (contacting the authors for many times but no reply to obtain the original text); (4) duplicates of the same studies; (5) conferences, abstract, case reports, meta-analysis, and review.

Data extraction

Two of the authors (JML and LX) independently evaluated the data reported in the publications which were suitable for this research and cross-checked to ensure that no data were missed. The following data were extracted: the first author, publication year, region that the study performed, study design, total number of participants, separate number of women with RA and without RA, maternal age, maternal outcomes, and fetal outcomes. The third author (SHM) participated and resolved the disagreements by consensus.

Methodological quality appraisal

The quality of cohort studies and case-control studies was assessed using the Newcastle-Ottawa Scale (NOS), which was a nine-point scale and divided studies into poor quality (0–3 points), fair quality (4–6 points), and good quality (7–9 points) [15]. The quality of cross-sectional studies was evaluated using the Joanna Briggs Institute (JBI), which was a 20-point scale and divided studies into low quality (0–14 points) and high quality (15–20 points) [16].

Statistical analysis

The comparison results of categorical data were expressed as odds ratio (OR) and 95% confidence intervals (95%CIs), and the comparison results of continuous data were expressed as weighted mean difference (WMD) with 95%CIs. Cochran’s Q test and the I2 statistic were used to assess the heterogeneity between the studies. Random-effect model was used if heterogeneity was found (I2 values ≥ 50%) and fixed-effect model was used if I2 values < 50%. Subgroup analysis based on study design and regions was performed to identify the sources of heterogeneity. Sensitivity analysis was performed for all outcomes and publication bias was assessed using Begg’s test if more than nine studies were included [17]. Data were analyzed using STATA v15.1 (STATA Corporation, College Station, TX, USA). P < 0.05 was considered as statistical significance.

Results

Literature search and study characteristics

From the above-mentioned four English databases and three Chinese databases, 6,015 English articles and 137 Chinese articles were obtained. After removing the duplicates, 3,735 articles remained. By screening titles and abstracts, 3,637 articles were excluded because they were reviews or meta-analyses (n = 679), their topic did not meet the requirements (n = 1763), abstracts or case reports (n = 1068), or animal experiments (n = 127). Further, 57 articles were eliminated due to data unable to extract (n = 4) and topic not meeting the requirements after a careful assessment of full texts (n = 53). Finally, 41 eligible articles were included in the meta-analysis (Fig. 1) [3,4,5,6, 9,10,11, 14, 18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50]. Of the included articles, there were 32 cohort studies, 7 case-control studies, and 2 cross-sectional studies. For quality assessment of the studies, 14 studies, 25 studies, and 2 studies were assessed as good, fair, and poor quality, respectively (Table 1).

Fig. 1
figure 1

The flowchart of the studies selection

Full size image
Table 1 The characteristics of the included studies
Full size table

Systematic review and meta-analysis of the association between maternal RA and adverse maternal/fetal outcomes

Table 2 shows that RA was associated with an increased risk of preeclampsia (OR: 1.65, 95%CI: 1.53–1.78, I2 = 13.4%), gestational diabetes (OR: 1.61, 95%CI: 1.25–2.07, I2 = 72.0%), spontaneous abortion (OR: 1.32, 95% CI: 1.21–1.43, I2 = 25.8%), and cesarean delivery (OR: 1.62, 95%CI: 1.43–1.84, I2 = 87.9%). Study design or region was the source of heterogeneity for gestational diabetes and cesarean delivery. Table 3 displays that RA was associated with an increased risk of stillbirth (OR: 1.55, 95% CI: 1.17–2.06, I2 = 0.0%), SGA (OR: 1.48, 95%CI: 1.25–1.75, I2 = 85.4%), LBW (OR: 1.73, 95% CI: 1.46–2.06, I2 = 65.8%), congenital abnormalities (OR: 1.24, 95% CI: 1.13–1.37, I2 = 42.3%), diabetes type I (OR: 1.70, 95% CI: 1.41–2.06, I2 = 36.8%), and asthma (OR: 1.23, 95% CI: 1.16–1.30, I2 = 0%). Study design or region was the source of heterogeneity for birth weight and LBW.

Table 2 Summary results of the association between rheumatoid arthritis and maternal outcomes
Full size table
Table 3 Summary results of the association between rheumatoid arthritis and fetal outcomes
Full size table

Barnabe et al. reported no statistical difference in the risk of postpartum infection between pregnant women with RA and without RA (P = 0.875) [21]. Aljary et al. found no significant difference in the risk of postpartum hemorrhage between women with RA and without RA (P = 0.276) [49]. In addition, there was no significance in the risk of neonatal death within 30 days of birth (P = 0.477) [50] and epilepsy (P = 0.164) [6], while risks of RA [6] and JIA [5] were higher in infants born from mother with RA.

Systematic review and meta-analysis of the association between disease activity of RA and maternal/fetal outcomes

Table 4 summarizes the pooled results on the association between disease activity of RA and maternal or fetal outcomes. HAQ-DI > 0.5 was associated with an increased risk of premature delivery (OR: 1.82, 95% CI: 1.12–2.97, I2 = 0%), indicating that high disease activity of RA was significantly associated with the high risk of premature delivery. The forest plot regarding premature delivery by HAQ-DI was shown in Fig. 2.

Table 4 Summary results of the association between disease activity of rheumatoid arthritis and maternal/fetal outcomes
Full size table
Fig. 2
figure 2

Forest plot for the association between the disease activity of RA assessed by HAQ-DI and premature delivery

Full size image

We also found that there was a significant increase in the risk of cesarean delivery (OR: 2.29, 95% CI: 1.02–5.15, I2 = 0%) and premature delivery (OR: 5.61, 95% CI: 2.20–14.30, I2 = 17.8%) in RA women with DAS28 > 3.2, which reflected the unfavorable effect of high disease activity of RA on the cesarean delivery and premature delivery. Forest plots regarding cesarean delivery and premature delivery by DAS28 were shown in Fig. 3A and B.

Fig. 3
figure 3

Forest plots for the association between the disease activity of RA assessed by DAS28 and cesarean delivery (A) and premature delivery (B)

Full size image

De Man et al. reported that higher disease activity of RA showed a relationship with the lower birth weight of newborns [11]. Smith et al. reported that the increase of disease activity was associated with the increased risk of premature delivery [10]. In addition, Al Rayes found that the higher disease activity of RA was associated with the higher risk of spontaneous abortion, premature delivery, and neonatal intensive care unit (NICU) admission [9].

Sensitivity analysis and publication bias

The sensitivity analysis was carried out by sequentially excluding one of the studies each time, and the results were consistent, indicating every included study had equal sensitivity and did not impact the overall results (data not shown). Begg’s test revealed that there was no publication bias regarding to the risk of preeclampsia (Z = 0.66, P = 0.511), gestational diabetes (Z = 0.72, P = 0.474), cesarean delivery (Z = 0.05, P = 0.964), premature delivery (Z = 1.64, P = 0.102), SGA infant (Z = 1.04, P = 0.300), LBW (Z = 0.89, P = 0.373), and congenital abnormalities (Z = 0.34, P = 0.732) (Table 5).

Table 5 Publication bias of outcomes by Begg’s test
Full size table

Discussion

In this systematic review and meta-analysis, we systematically assessed the association between RA and pregnancy outcomes, and we also quantified the data on the association between the disease activity of RA and pregnancy outcomes. The results showed that pregnant women with RA had a higher risk of preeclampsia, gestational diabetes, spontaneous abortion, and cesarean delivery than those without RA. Infants born from RA mother had a higher risk of stillbirth, SGA, LBW, congenital abnormalities, type 1 diabetes, and asthma than those born from mother without RA. In addition, we found that high disease activity of RA was associated with the higher risk of premature delivery and cesarean delivery.

Studies have reported the more prevalence of type 2 diabetes mellitus in RA patients than non-RA affected people [49, 51]. In our analysis, we found a higher risk of gestational diabetes in women with RA. This may be because that disease activity in pregnant RA women was usually controlled using glucocorticoids, which may decrease the sensitivity of peripheral insulin, increase the production of hepatic glucose, and inhibit the production and secretion of pancreatic insulin [52], thereby promoting the development of diabetes in patients with RA [53, 54]. Moreover, preeclampsia was commonly observed in RA women. The reason for the association between preeclampsia and RA remained unclear, but it was speculated that there was a common autoimmunologic factor between preeclampsia and RA [49]. In addition, RA women had an elevated risk of spontaneous abortion. Evidence has shown the higher rates of abortion after RA diagnosis [26, 55], especially in RA women with high disease activity during pregnancy [9].

Our analysis showed that RA was associated with the odds of SGA and LBW in infants. Strouse et al. have found that the odds of SGA were significantly increased in women with RA [3]. Some hypotheses explained how high disease activity led to LBW, such as high maternal cortisol level, vasculopathy, and high inflammatory cytokines levels that downregulate the activity of placental 11-hydroxysteroid dehydrogenase type 2 [56]. Moreover, we found that infants born from RA mother had a higher risk of congenital anomalies, and the heterogeneity of the pooled result was low (42.3%). Huang et al. also found the significant association between RA and congenital anomalies, and the heterogeneity of the pooled result was higher (78.4%) [8]. The reason for the difference in the heterogeneity for congenital anomalies may be that more studies included in our meta-analysis for this outcome, and the sample size was bigger, which might improve the statistical power. One study reported that children of a parent with RA had two-fold increased odds of type 1 diabetes [57]. We reported the similar finding that infants born from RA mother displayed the higher risk of type 1 diabetes. Asthma was a chronic inflammatory disease of the pulmonary system, and fetal allergic immune system was related to maternal inflammations [58]. Therefore, the status of maternal immune system was linked to childhood asthma. Yang et al. reported an elevated chance of asthma in children born to mothers with maternal RA [41], which was consistent with our analysis.

A significantly increased risk of cesarean delivery was found in women with RA in this meta-analysis. Huang et al. have reported that RA increased the risk of cesarean delivery, and speculated that differences in cesarean delivery rare may be associated with the disease activity [8]. A study by Zbinden et al. reported a significant association between high disease activity and the high odds of caesarean delivery in women with RA, implying that inadequate control of disease activity may lead to caesarean delivery [42]. In our meta-analysis, through quantitative analysis, we found that high disease activity of RA was associated with the higher risk of cesarean delivery. Our meta-analysis further provided the evidence for the association between active RA and cesarean delivery. We also found that the risk of premature delivery was increased with the higher disease activity of RA. Al Rayes et al. have reported the significant association between high disease activity and premature delivery [9]. The similar result was found in the study of Smith et al. [10]. Our findings suggested that a better disease control may be beneficial to improve the pregnancy outcomes of women with RA.

Our systematic review and meta-analysis further compare the pregnancy outcomes in women with RA and without RA, and also quantitively analyzes the association between disease activity and the risk of pregnancy outcomes in women with RA. There are some limitations in this study. First, heterogeneity may be caused by the different ages, onset time, and disease duration in included population, while data reported in the included studies are not enough to support us to conduct subgroup analysis for further exploration. Second, maternal smoking and drinking habits, family history, and rheumatoid treatment during pregnancy also affect maternal and infant outcomes. Since above information is not stated in all original studies, these data cannot be included in this meta-analysis. Third, disease activity is assessed using DAS28, HAQ-DI, PS, and PGS. Due to the limited number of studies, DAS28 and HAQ-DI are used for quantitative analysis. In the future, more studies are needed to further verify our findings.

Conclusion

Our meta-analysis showed that high disease activity of RA was associated with the increased risk of adverse pregnancy outcomes. RA women preparing for pregnancy should pay more attention to control the disease activity. It is essential to strengthen communications between patients, obstetricians, and rheumatologists to develop individualized treatment plans for women with high disease activity of RA who are preparing for pregnancy.