Abstract
Objective The disease course and burden of rheumatoid arthritis (RA) may differ between female and male individuals, but existing data on these differences are limited and often contradictory. Therefore, we investigated whether clinical outcomes and patient-reported outcomes (PROs) differ between female and male patients with RA over time.
Methods All female (n = 286) and male (n = 139) patients with RA according to 1987 and/or 2010 criteria from Treatment in the Rotterdam Early Arthritis Cohort (tREACH), a stratified single-blinded trial with a treat-to-target (T2T) approach and fixed medication protocol, were included. Clinical outcomes include disease activity, medication usage, sustained disease-modifying antirheumatic drug (DMARD)-free remission (SDFR), and radiographic progression. In addition, the following PROs were investigated: general health, pain, functional ability, health-related quality of life, fatigue, productivity loss, and a possible depression or anxiety disorder. For comparisons over time, a mixed model or Cox proportional hazard model was used. The mixed models were adjusted for age, initial treatment, and disease activity (Disease Activity Score in 44 joints [DAS44]).
Results Female patients had a higher DAS44 over time compared to male patients (β 0.36, 95% CI 0.25-0.47, P < 0.001), which also resulted in more treatment adjustments including use of biologic DMARDs (bDMARDs; 36% vs 24%, P < 0.001). Although not significant, first bDMARD survival seemed shorter in female patients (hazard ratio [HR] 1.4, 95% CI 0.8-2.6, P = 0.24). However, no differences were found in SDFR and radiographic progression. With regard to PROs, only functional ability differed significantly between sexes after adjusting for confounders, including disease activity (Health Assessment Questionnaire–Disability Index, β 0.10, 95% CI 0.04-0.17, P < 0.001).
Conclusion Clinical outcomes and PROs are intertwined, and both improve with a T2T management approach. Nevertheless, female patients with RA have higher disease activity, a greater need for bDMARDs—although these have lower efficacy—and more functional impairment over time, underscoring the need for sex-specific management recommendations. (Trial registration number: ISRCTN26791028)
Rheumatoid arthritis (RA) is more prevalent in female than male individuals.1,2 However, the pathophysiological mechanisms that cause these differences still have to be unraveled. We do know that the incidence of RA in female individuals decreases with age, suggesting that the disparity in sexes could be due to sex hormones.3 Moreover, male and female individuals may experience different disease courses and burdens, but data on these sex differences are falling behind.4 Moreover, previous literature showed contradictory results, which is possibly due to methodological issues in most of the aforementioned studies, including the lack of data from clinical trials.5
Sex differences in RA may be attributed to both biological (sex-related) and sociocultural (gender-related) factors, such as genetic predisposition, hormonal influence, and differing social roles.6 Hormones modulate immune responses, contributing to the observed differences in disease susceptibility and severity between male and female individuals.7 Additionally, variations in pain perception, muscle strength, and inflammatory markers further complicate the clinical picture in female individuals.8-10
Previous studies have already demonstrated that female patients with RA exhibit higher disease activity and lower remission rates compared to their male counterparts.11,12 Moreover, emerging evidence suggests that female patients respond less favorably to standard treatment regimens, including both conventional and biologic (b-) disease-modifying antirheumatic drugs (DMARDs).13-15 However, limited to no data exist on sex differences in bDMARD survival and sustained DMARD-free remission (SDFR). Currently, SDFR is an achievable outcome and the nearest proxy for cure, due to the recommended treat-to-target (T2T) management approach.16
Regardless of the T2T management approach and the improved outcomes, the burden of disease in patients with RA is still high. Therefore, at present, a dual T2T approach is recommended, in which the targets are both (1) control of inflammation, measured with composite disease activity indexes; and (2) control of disease impact, assessed using a range of patient-reported outcomes (PROs).17 The International Consortium for Health Outcomes Measurement (ICHOM) previously agreed on the most relevant PRO domains for patients with inflammatory arthritis. These domains are pain, activity limitation, overall emotional and physical health impact, fatigue, work/school/housework ability, and productivity.18
A few studies already demonstrated that PROs tend to be worse in female compared to male patients with RA. For example, Sokka et al showed that female individuals experienced more fatigue, pain, and functional impairment compared to male individuals, even after adjusting for the number of swollen joints.12 Shin et al, on the other hand, found that disease activity over time differed significantly between both sexes, but no differences were seen in PROs after correction for baseline disease activity.19 Although the aforementioned studies addressed the ICHOM domains to varying extents, none addressed all domains, and both studies had a short follow-up period.
Therefore, our aim was to investigate whether clinical outcomes (disease activity, medication usage, SDFR, and radiographic progression) and all ICHOM-recommended PROs differ between female and male patients with RA who were enrolled in a randomized controlled trial (RCT) with a T2T management approach and fixed medication protocol.
METHODS
Patients. Data from the Treatment in the Rotterdam Early Arthritis Cohort (tREACH) were used. The tREACH trial was a multicenter, stratified, single-blinded RCT with a T2T management approach. Patients who were included in the original tREACH trial had to have at least 1 swollen joint and a symptom duration of < 1 year.20 Eligible patients were stratified into 3 groups according to their likelihood of progressing to persistent arthritis based on the prediction model of Visser et al.21 The 3 strata—low, intermediate, and high—correspond with probability tertiles of developing persistent arthritis. Further details can be found in the publications of Claessen et al and Visser et al.20,21
For the present study, we selected all female and male patients with RA (n = 425), according to 1987 and/or 2010 criteria22 (Supplementary Figure S1, available with the online version of this article). The inclusion and follow-up phases of the study have been completed.
Study design. The tREACH trial had a T2T management approach with a fixed medication protocol that aimed for low disease activity (LDA; Disease Activity Score in 44 joints [DAS44] < 2.4).23 Treatment alterations were made every 3 months, depending on the disease activity, and in the event of very active disease, an earlier visit could be planned based on the assessment of the rheumatologist.
Patients received either initial triple DMARD therapy (iTDT: methotrexate [MTX; 25 mg/week], sulfasalazine [SSZ; 1000 mg BID], and hydroxychloroquine [HCQ; 200 mg QD], with oral or intramuscular glucocorticoid [GC] bridging therapy); initial MTX with or without GC bridging therapy (iMTX); initial HCQ (iHCQ); or nonsteroidal antiinflammatory drugs (NSAIDs) or GCs (no DMARDs). Treatment was intensified in cases of active disease (DAS44 ≥ 2.4). Treatment intensifications occurred in the following order: triple DMARD therapy; MTX + etanercept (50 mg/week, subcutaneous [SC]); MTX + adalimumab (40 mg/2 weeks, SC) and MTX + abatacept (500-1000 mg/4 weeks, intravenous [IV], weight-dependent). Tapering of medications occurred if the DAS44 was < 1.6 at 2 consecutive visits. Medication was gradually discontinued, except for HCQ and NSAIDs, which were immediately stopped. In case of a flare (DAS44 ≥ 2.4) during tapering, full treatment was restarted, according to the stage in the protocol.
Data collection and outcomes. Patients visited the outpatient clinic every 3 months. At each visit, all clinical outcomes and PROs were collected, except for radiographs, which were obtained at baseline, 6 months, 1 year, and 2 years.
Disease activity, on which treatment decisions were based, was measured with the DAS44. The DAS44 is a pooled index that combines information from a 44-joint swollen joint count (SJC44), a 53-joint tender joint count (TJC53), erythrocyte sedimentation rate (ESR), and general health (GH) measured via a visual analog scale (VAS), into a formula to provide a numerical indicator of disease activity. Since female individuals tend to have higher ESR compared to male individuals, we also calculated DAS28 using C-reactive protein (DAS28-CRP), which uses a 28-joint count and CRP.
Medication data, including bDMARD usage and first bDMARD survival, were obtained from health records. SDFR was defined as being in remission (DAS44 < 1.6 and no swollen joints) without the use of any DMARDs for at least 6 months. Radiographic progression was assessed with the modified Total Sharp Score (mTSS), where a higher mTSS represents more radiographic progression.24 Radiographs were scored chronologically by 2 out of 3 qualified assessors, who were blinded for treatment allocation. The weighted overall κ was 0.67, with a 99% agreement rate. Median mTSS values and the percentages of patients with changes > 0.5 (radiographic progression) and > 1.0 (the smallest detectable change) in mTSS are reported.25
The following PROs were collected: GH, pain, functional ability, health-related quality of life (HRQOL), fatigue, productivity loss, and possible depression or anxiety disorder. GH was measured on a 0-100 mm VAS, where higher scores reflected greater severity. Pain was measured on an 11-point Likert scale, with higher scores indicating more pain experienced. The Health Assessment Questionnaire–Disability Index (HAQ-DI) was used to measure functional ability.26 The total scores range from 0 to 3, and higher scores indicate more functional impairment.
HRQOL was measured with the EQ-5D-3L questionnaire and 36-item Short Form Health Survey (SF-36).27,28 The EQ-5D-3L provides an assessment of the patient’s health status compared to the general Dutch population by evaluating 5 health dimensions. Scores range from below 0 to 1; 0 equals death and 1 equals perfect health. The SF-36, on the other hand, covers 8 domains and the summed scores per domain are transformed to a 0-100 scale.27 Physical component summary and mental component summary scores were calculated from these domains.
Fatigue was measured with a VAS, scores ranged from 0 to 100, and higher scores indicated more severe fatigue. Presenteeism, defined as working while sick, was used as a measure of productivity loss. Presenteeism was assessed using a numerical rating scale (0-10), which was then converted into a percentage of productivity loss, with 0% indicating no productivity loss and 100% indicating complete inability to work.
The presence of a possible depression and anxiety disorder was measured with the Hospital Anxiety and Depression Scale (HADS). The HADS consists of a depression subscale (HADS-D) and an anxiety subscale (HADS-A), with scores ranging from 0 to 21 for each subscale. A score > 7 is indicative of a possible depression or anxiety disorder.29
Statistical analysis. The t test, chi-squared test, or Wilcoxon rank-sum test was used to compare the differences in clinical outcomes as well as PROs between female and male patients with RA at diagnosis and after 2 years of follow-up.
Mixed models with an unstructured covariance matrix and random intercept and slope were used to compare all clinical outcomes (DAS44 and its components, DAS28-CRP, use of bDMARDs, and mTSS) and PROs (GH, pain, HAQ-DI, EQ-5D-3L, SF-36, fatigue, productivity loss, and possible depression [HADS-D > 7] or anxiety disorder [HADS-A > 7]) over the 2-year follow-up period. Sex, time (in months), age, initial treatment (iTDT, iMTX, iHCQ, and no DMARDs) and DAS44 were included as covariates in the mixed models. DAS44 was included as a time-varying covariate. Male patients were used as the reference catgeory.
On the other hand, Cox proportional hazard models were used to compare first bDMARD survival and SDFR over time in both sexes. No additional adjustments were made in the Cox proportional hazard models.
If a significant difference was found, then the difference was compared to the minimal clinically important difference (MCID). The MCID can be used as a criterion to determine if significant differences are also clinically relevant. An MCID was available only for DAS (DAS44 and DAS28-CRP)30,31 and all PROs.32-38
Since previous research already showed that female patients with RA have a higher disease activity compared to male patients with RA, a sensitivity analysis was performed in which female and male patients with RA were stratified by disease activity status (remission, LDA, and active disease) based on the DAS44 after 2 years of follow-up, after which all clinical outcomes and PROs at the same timepoint were compared between both sexes. Additionally, a sex-time interaction term was incorporated into our mixed models to further explore the differences between sexes over time.
To correct for multiple testing, a Bonferroni correction was applied to the mixed models by multiplying the P values with the 7 performed tests. A corrected P value ≤ 0.05 was considered statistically significant. Analyses were performed in Stata v.18.0 (StataCorp).
RESULTS
Patients. Baseline characteristics are given in Table 1. Of the 425 included patients with RA, 286 were female (67%). Symptom duration and autoantibody positivity (rheumatoid factor; 52% in female vs 49% in male patients) were divided equally between the sexes. Diagnosis on the other hand was made at a younger age (mean [SD]) in women: 52 (14) years compared to 58 (14) years in men.
Baseline characteristics for female and male patients with RA.
Clinical outcomes. At baseline, male patients had a higher SJC44 (median [IQR] 8 [5-14] vs 7 [4-11], P = 0.001), whereas female patients had a higher TJC53 (median [IQR] 11 [6-15] vs 7 [3-14], P < 0.001). Male patients also had significantly higher CRP levels (median [IQR] 11 [5-29] mg/dL vs 6 [3-16] mg/dL, P = 0.02), and reported better GH (median [IQR] 46 [25-65] vs 54 [37-70], P = 0.003). Despite these differences, disease activity scores (DAS44 and DAS28-CRP) at baseline were similar between both sexes (Table 1).
Disease activity over time, however, was significantly worse in female compared to male patients with RA (Table 2; Figures 1A,B). This applied to both the DAS44 (β 0.36, 95% CI 0.25-0.47, P < 0.001) and the DAS28-CRP (β 0.18, 95% CI 0.08-0.28, P = 0.001; Table 2). The difference in disease activity can be explained by the higher TJC53 (β 0.36, 95% CI 0.20-0.51, P < 0.001); however, female patients also had a higher ESR (β 3.43, 95% CI 1.76-5.11, P < 0.001; Table 2). After 2 years of follow-up, the disease activity (DAS44 and DAS28-CRP) was still higher in female compared to male patients with RA (Table 3).
Clinical outcomes over time stratified by sex. (A) Estimated mean DAS44, with 4 items (44-joint SJC, 53-joint TJC, erythrocyte sedimentation rate, general health on 0-100 VAS); (B) estimated mean DAS28, with 4 items (28-joint SJC, 28-joint TJC, CRP, general health on a 0-100 VAS); (C) proportion of patients using a biologic; (D) Kaplan-Meier curves for achievement of SDFR; (E) estimated mean mTSS; and (F) cumulative probability plot of 2-year radiographic progression. CRP: C-reactive protein; DAS44: Disease Activity Score in 44 joints; DAS28-CRP: Disease Activity Score 28 using CRP; mTSS: modified Total Sharp Score; RA: rheumatoid arthritis; SDFR: sustained disease-modifying antirheumatic–free remission; SJC: swollen joint count; TJC: tender joint count; VAS: visual analog scale.
Outcomes during the 2-year follow-up period, stratified by sex.
Outcomes per timepoint stratified by sex.
Although no differences in initial treatment were found, female patients had more treatment intensifications compared to male patients (Figure 2A). After 2 years, 36% of female patients were using a bDMARD compared to 24% of male patients. Of female and male patients with RA, 6% and 1% were already on their third or more bDMARD after 2 years of follow-up, respectively (Figure 2A). The odds (95% CI) of bDMARD use was 7.93 (3.18-19.77, P < 0.001) higher in female compared to male patients (Table 2, Figure 1C). Although the Kaplan-Meier curve indicated shorter first bDMARD survival times for female patients, the Cox regression analysis did not yield statistically significant results (hazard ratio 1.44, 95% CI 0.78-2.61, P = 0.24; Table 2, Figure 2B).
Medication usage and first biologic survival stratified by sex. (A) Medication usage over time; and (B) Kaplan-Meier curves for first bDMARD survival rate. bDMARD: biologic disease-modifying antirheumatic drug; bio: biologic.
On the other hand, no differences in SDFR and radiographic progression were found (Table 2, Figures 1D-E). Radiographic progression, defined as > 0.5-point increase in mTSS, occurred in 55 out of 268 female (21%) and 37 out of 134 male (28%) patients with RA (P = 0.11). If the smallest detectable change (> 1.0) is used, then 38 (14%) and 25 (19%) of female and male patients with RA had radiographic progression, respectively (P = 0.24). The cumulative probability plots for radiographic progression were superimposable (Figure 1F).
PROs over time. At baseline, female patients had a worse GH (VAS, median [IQR] 54 [37-70] vs 46 [25-65], P = 0.02), more functional impairment (HAQ-DI, median [IQR] 1.12 [0.48-1.62] vs 0.75 [0.32-1.12], P < 0.001), and experienced more fatigue (VAS, median [IQR] 59 [34-76] vs 37 [21-62], P < 0.001), compared to their male counterparts (Table 3). The MCID was exceeded for the HAQ-DI and VAS fatigue.
Irrespective of the sex, all PROs improved over time due to the T2T design (Figures 3A-H). After adjustment, which, among others, includes disease activity as a time-varying covariate, female patients still scored worse on the HAQ-DI over time (β 0.10, 95% CI 0.04 to 0.17, P < 0.001; Table 2). There was no significant effect of sex on the other PROs over time (Table 2, Figures 3A-H).
Patient-reported outcomes during the 2-year follow-up stratified for sex. (A) General health (0-100 VAS); (B) pain (10-point Likert); (C) functional ability (HAQ-DI); (D) quality of life (EQ-5D-3L); (E) fatigue (0-100 VAS); and (F) productivity loss (presenteeism), measured on a 0-100% productivity loss scale, where 0% indicates no impairment and 100% indicates complete impairment. Patients (%) with a possible (G) depression (HADS-D ≥ 8); and (H) anxiety disorder (HADS-A ≥ 8). All graphs represent the estimated mean with 95% CI, except for (G) and (H). HADS-A: Hospital Anxiety and Depression Scale, anxiety subscale; HADS-D: Hospital Anxiety and Depression Scale, depression subscale; HAQ-DI: Health Assessment Questionnaire–Disability Index; VAS: visual analog scale.
After 2 years of follow-up, female patients still had more functional impairment compared to male patients (HAQ-DI, median [IQR] 0.62 [0.12-1.32] vs 0.38 [0.0-0.78], P < 0.05), and again, the MCID was exceeded (Table 3). No other PRO differences were found after 2 years of follow-up (Table 3).
Sensitivity analysis. If female and male patients with RA were stratified by disease activity status (remission, LDA, and active disease), based on DAS44 after 2 years of follow-up, similar outcomes were found. Female patients who were in remission still experienced more functional impairment compared to their male counterparts (HAQ-DI, median [IQR] 0.75 [0.38-1.62] vs 0.50 [0.12-1.88], P = 0.01; Supplementary Table S1, available with the online version of this article).
After incorporation of a sex-time interaction term, female patients still had worse disease activity (DAS44: β 0.34, 95% CI 0.21-0.48, P < 0.001) and experienced more functional impairment (β 0.12, 95% CI 0.05-0.19, P = 0.02) over time compared to male patients and these differences did not change during follow-up (β 0.00, 95% CI −0.00 to 0.01, P = 0.67, and β 0.00, 95% CI −0.01 to 0.00, P = 0.24, respectively; Supplementary Table S2, available with the online version of this article). Although female patients had fewer swollen (β −1.0, 95% CI −1.5 to −0.5, P < 0.001) and more tender (β 2.2, 95% CI 1.4 to 3.0, P < 0.001) joints over time compared to male patients, these differences diminished during follow-up (β 0.1, 95% CI 0.0 to 0.1, P < 0.001 and β −0.1, 95% CI −0.1 to −0.1, P < 0.001, respectively).
DISCUSSION
The disease course and burden of RA may differ between female and male patients. Hence, we investigated whether clinical outcomes and PROs differed between the sexes. We found that female patients had a higher disease activity (DAS44 and DAS28) over time, which resulted in more treatment intensifications, including bDMARD usage. After 2 years, 6% and 1% of female and male patients with RA were already on their third or more bDMARD, respectively. Although not significant, first bDMARD survival seemed shorter in female patients. However, no differences were found in SDFR or radiographic progression. After correction for disease activity, among others, no differences were found in disease burden between the sexes, except for functional ability, which was worse in female patients with RA.
Previous literature has shown that RA diagnosis was made at a younger age in female patients and that the female-to-male ratio decreased after the age of 55, which suggests that sex hormones might affect the risk of disease.39 Our data are in accordance with previous literature. Aside from hormonal differences, other factors such as genetics or environment might contribute to the found differences between sexes in RA.40
Although disease activity at baseline was similar, female patients had more tender joints and worse GH. The higher TJC might be attributable to a different pain threshold or coping strategy. A large metaanalysis, for example, showed that the standardized mean difference in VAS pain was significantly higher in female compared to male patients with RA.41 Moreover, experimental pain studies using various stimuli showed that women are more sensitive to multiple pain modalities compared to men.42 Aforementioned reasoning could also explain why female patients have a higher DAS44 over time compared to male patients.12,43,44 Another reason might be the sex-related variability in ESR levels, which was also found in our study.10,45 For this reason, we also examined the DAS28-CRP and found results similar to those of the DAS44.
Since disease activity over time was higher in female patients with RA, treatment adjustments also occurred more often, leading to more bDMARD usage. Noteworthy is the fact that first bDMARD survival also seemed shorter in female patients with RA and after 2 years of follow-up, 6% and 1% of the female and male patients could already be classified as difficult-to-treat (D2T), respectively.46 However, no differences were found in SDFR rates, probably due to the follow-up period. Previous studies already noted differences in anti–tumor necrosis factor and rituximab response between the sexes, but these differences were also influenced by disease duration and prior treatment exposure.14,47 Although these studies provide insights into sex-based variations in treatment efficacy, they do not give insight into the proportion of patients who are D2T. Unlike the aforementioned studies, our present study sheds some more light on treatment patterns and efficacy, which underscores the need for a sex-specific management approach in order to optimize outcomes in female patients with RA. One approach might be the implementation of a dual T2T management approach that focuses on control of both inflammation and disease impact.
The disease impact can be measured with PROs, and female patients scored worse on all PROs (over time) compared to male patients. However, after adjustment, among other disease activities over time, female patients had more functional impairment only. This is in accordance with previous literature.48 A possible explanation might be that female individuals typically have lower muscle strength than male individuals and, consequently, the effect of RA on functional capacity tends to be more pronounced in this group.9 Other studies have also found differences in GH, pain, HRQOL, and fatigue.12,14,48,49 Aforementioned studies, however, did not adjust for disease activity over time and often had a short follow-up period. In addition to control of inflammation with treatment, there may be added value to patient education, physical therapy, and a healthy lifestyle, especially for reducing the disease impact in patients—particularly female patients—with RA.50
The strengths of our study are the study design and the broad clinical outcomes, as well as PROs that cover all ICHOM domains. An RCT with a fixed medication protocol is less susceptible to bias and confounding. In addition, adjusted models with DAS44 as a time-varying covariate were used to correct for the difference in DAS44 over time.
Although our broad outcomes have several advantages, they have also led to multiple testing. For this reason, we applied a Bonferroni correction and also compared the differences with known MCIDs from the literature. Second, PROs are susceptible to recall bias and nonresponse and are therefore difficult to interpret on a population level. Finally, dropout ratios were skewed during the first 2 years of follow-up. Reasons for dropout may differ between sexes and can vary from dissatisfaction with treatment during the first years of follow-up, to inactive disease in the last year of follow-up. Therefore, a sensitivity analysis was performed in which we compared female and male patients with similar disease activity status and found similar results.
In conclusion, clinical outcomes and PROs are intertwined, and both improve with a T2T management approach. Nevertheless, female patients with RA have a higher disease activity over time and subsequently use more bDMARDs, which tend to be less effective; further, female patients with RA also experience more functional limitations compared to their male counterparts. Therefore, a sex-specific management approach is needed to optimize outcomes in female patients with RA.
ACKNOWLEDGMENT
We would like to thank the participating patients in the tREACH trial for their willingness to contribute to the study and for their cooperation. Further, we would like to thank all rheumatologists from all participating centers.
Footnotes
CONTRIBUTIONS
GHK performed the statistical analysis and drafted the manuscript. PHPdJ contributed to the analysis. All authors contributed to the design, revised the manuscript, and read and approved the final manuscript. PHPdJ is the guarantor and takes responsibility for the integrity of the work as a whole, from inception to published article.
FUNDING
The tREACH trial was supported by an unrestricted grant from Pfizer BV (WI229707). Pfizer had no involvement in the study design; in the collection, analysis, or interpretation of data; in the writing of the report; or in the decision to submit for publication. The corresponding author had full access to all data and had final responsibility for the decision to submit the manuscript for publication. Data management was sponsored by the Dutch Arthritis Society (16–3-101).
COMPETING INTERESTS
The authors declare no conflicts of interest relevant to this article.
ETHICS AND PATIENT CONSENT
The study protocol was approved by the medical ethics committees at each participating center and this study was approved by the medical ethics committee of Erasmus Medical Center Rotterdam, the Netherlands (MEC-2006–252). All patients gave written informed consent before inclusion.
DATA AVAILABILITY
Data are available from the corresponding author upon reasonable request.
- Accepted for publication January 24, 2025.
- Copyright © 2025 by the Journal of Rheumatology
REFERENCES
SUPPLEMENTARY DATA
Supplementary material accompanies the online version of this article.
