Methotrexate and glucocorticoid use in relation to cross-sectional and longitudinal associations of red blood cell distribution width in patients with rheumatoid arthritis Authors: Yoshifumi Ohashi, Mochihito Suzuki, Yasumori Sobue, Kenya Terabe, Shuji Asai, Shiro Imagama, Nobunori Takahashi Published: 2026 ABSTRACT ABSTRACTThis study investigated the clinical characteristics of red blood cell distribution width (RDW) in rheumatoid arthritis (RA) and the impact of treatment, including longitudinal dose changes. Among 691 RA patients enrolled in 2025, 591 with RDW recorded in both 2024 and 2025 were analyzed. Based on 2024 data, patients were classified into normal RDW (≤14.5%, n = 436) and high RDW (>14.5%, n = 155) groups. Logistic regression identified factors associated with high RDW. Associations with methotrexate (MTX) and glucocorticoid (GC) dose were examined, and one-year changes in RDW (ΔRDW, 2024–2025) were evaluated according to dose modifications. Patients with high RDW were older (70.6 vs 68.0 years), frequently used MTX (71.0% vs 57.6%) and GCs (41.3% vs 18.8%), had higher Disease Activity Score in 28 joints based on erythrocyte sedimentation rate (DAS28-ESR, 2.92 vs 2.68) and Health Assessment Questionnaire Disability Index (HAQ-DI, 0.64 vs 0.40), and lower hemoglobin (11.8 vs 12.9 g/dL). Multivariable analysis revealed that low hemoglobin (odds ratio [OR], 0.56), MTX use (OR, 2.47), and GC use (OR, 2.21) were independently associated with high RDW. The prevalence of high RDW increased with higher MTX (14.0%, 18.3%, 28.6%) and GC (14.0%, 20.7%, 43.2%) doses. Longitudinally, ΔRDW rose with MTX initiation or escalation (0.43 ± 0.91 vs −0.08 ± 1.06 or −0.26 ± 1.18), whereas GC changes were not significant. These findings indicate that RDW in RA is influenced more by treatment—particularly MTX use and dose escalation—than by inflammation. GC use also showed cross-sectional associations, though longitudinal effects were less evident. RDW should be interpreted in light of treatment rather than as a simple marker of disease activity.Key Words: rheumatoid arthritis, red cell distribution width, methotrexate, steroid, prednisolone INTRODUCTION Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by persistent synovial inflammation, which can lead to progressive joint destruction and functional impairment. The primary goal of RA treatment is to control disease activity early and prevent structural damage.1 Disease-modifying antirheumatic drugs (DMARDs) play a central role in treatment, with methotrexate (MTX) being the most widely used conventional synthetic DMARD.2 In recent years, the introduction of biological agents and Janus kinase (JAK) inhibitors, collectively known as biological and targeted synthetic DMARDs (b/tsDMARDs), has further advanced RA management.3 In addition, glucocorticoids (GCs) are still frequently used as adjunctive therapy for their rapid anti-inflammatory effects, particularly during the early stages of treatment or disease flares.4 Red blood cell distribution width (RDW), a parameter automatically calculated in routine complete blood counts, reflects variations in red blood cell size.5 Although traditionally used to help differentiate between types of anemia, RDW has recently garnered attention as a marker of systemic inflammation, oxidative stress, and poor nutritional status.6 In RA, chronic inflammation, nutritional deficiencies, and potentially the effects of medication may contribute to changes in RDW. Several studies have suggested a possible association between RDW and disease activity or physical function in patients with RA.7,8 However, the findings are inconsistent, and the utility of RDW as a reliable clinical biomarker in RA remains uncertain. Moreover, the effects of RA treatments on RDW have not been thoroughly investigated. Various medications, including DMARDs and GCs, may influence hematopoiesis, metabolic status, and inflammation, potentially affecting RDW, but evidence on these associations is limited.8,9 In this study, we aimed to investigate the clinical characteristics of RDW in patients with RA, including its relationship with commonly used treatments for RA. MATERIALS AND METHODS Patients This study utilized data from the multicenter observational study (T-FLAG study), which was approved by the institutional review board. Initiated in 2020 by three rheumatologists from different institutions, the T-FLAG study enrolled patients with RA who fulfilled the 2010 ACR/EULAR classification criteria.10 The study protocol was approved by the institutional review board of Nagoya University Graduate School of Medicine (approval no. 2017-0271-6, October 25, 2017). Participants were informed about the study protocol and were provided the opportunity to opt out. All coauthors reviewed and agreed with the ethical procedures. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki, and all patient data were anonymized to protect confidentiality. In 2025, a total of 691 patients attending outpatient rheumatology clinics at three participating hospitals were assessed. After excluding 100 patients with missing RDW data in 2024 or 2025, 591 patients were included in the final analysis. Based on RDW values, patients were classified into two groups: normal RDW (RDW ≤14.5% in 2024, n = 436) and high RDW (RDW >14.5% in 2024, n = 155) (Figure 1). Because a clinically meaningful lower limit for RDW has not been established and low RDW has no recognized pathological significance, only the upper reference limit (14.5%) was used to define RDW elevation.11 Flow chart of participant selection RA: rheumatoid arthritis RDW: red blood cell distribution width Clinical assessments Clinical evaluations were conducted by three rheumatologists certified by the Japan College of Rheumatology. Evaluations included the assessment of swollen joint count (SJC), tender joint count (TJC), and documentation of current medications. Radiographic evaluation of joint damage was performed using bilateral hand X-rays, and joint destruction was classified according to the Steinbrocker staging system.12 This system categorizes radiographic damage into four stages: Stage I represents periarticular osteoporosis; Stage II includes joint space narrowing and early erosions; Stage III is defined by joint deformity caused by more advanced erosions; and Stage IV indicates severe joint destruction or bony ankylosis. Additionally, standard laboratory tests were performed to measure C-reactive protein (CRP), rheumatoid factor (RF), erythrocyte sedimentation rate (ESR), and matrix metalloproteinase-3 (MMP-3) levels. Assessment of disease activity and physical function Disease activity was assessed using the Disease Activity Score in 28 joints based on ESR (DAS28-ESR), which incorporates SJC, TJC, patient’s global health assessment, and ESR.13 Based on this score, clinical remission was defined as DAS28-ESR <2.6, low disease activity as 2.6≤ DAS28-ESR <3.2, moderate disease activity as 3.2≤ DAS28-ESR ≤5.1, and high disease activity as DAS28-ESR >5.1.14 Physical function was evaluated using the Japanese version of the Health Assessment Questionnaire Disability Index (HAQ-DI), with functional remission defined15,16 as a HAQ-DI score of ≤0.5. Comorbidities The presence of interstitial lung disease was confirmed through imaging studies, including chest radiographs and computed tomography, as documented in medical records. Renal function was evaluated using the estimated glomerular filtration rate (eGFR), with impaired renal function defined17 as an eGFR of <45 mL/min/1.73 m2. Comorbidities were also identified based on medical records. Assessment of high RDW across MTX dose categories To investigate the association between MTX dose and the prevalence of high RDW, we excluded patients receiving GCs from the analysis (n = 146). Since the median MTX dose among users was 8 mg/week, patients were categorized into three groups based on MTX dose: non-users (n = 164), those receiving 1 to <8 mg/week MTX (n = 120), and those receiving ≥8 mg/week MTX (n = 161). Assessment of high RDW across GC dose categories To assess the association between GC dose and the prevalence of high RDW, we excluded patients who were receiving MTX from this analysis (n = 361). GC dose was converted to prednisolone (PSL) equivalent for standardization. Among GC users, the median daily dose was 2.5 mg/day. Based on this, patients were categorized into three groups according to GC (PSL) dose: non-users (n = 164), those receiving 1 to <2.5 mg/day PSL (n = 29), and those receiving ≥2.5 mg/day PSL (n = 37). Assessment of one-year change in RDW by MTX and GC dose adjustment Changes in RDW values associated with dose modifications of MTX or GCs between 2024 and 2025 were evaluated. The change in RDW was defined as ΔRDW = RDW in 2025 − RDW in 2024. For the MTX analysis, to eliminate the potential influence of GC dose modifications, only patients whose GC dose remained unchanged during this period were included (n = 529). Among these, patients who did not receive MTX in either 2024 or 2025 (n = 189) were excluded, resulting in a final cohort of 340 patients for the MTX analysis. Patients were classified into three groups: dose escalation group (patients who initiated or increased MTX dose; n = 36), dose unchanged group (patients with no change in MTX dose; n = 263), and dose reduction group (patients who reduced MTX dose; n = 41). ΔRDW was compared among these groups. In addition, MTX doses in 2024 and 2025 were compared within each group. For the GC analysis, to eliminate the potential influence of MTX dose modifications, only patients whose MTX dose remained unchanged during this period were included (n = 503). Among these, patients who did not receive GC in either 2024 or 2025 (n = 379) were excluded, resulting in a final cohort of 124 patients for the GC analysis. Patients were similarly classified into three groups: dose escalation group (patients who initiated or increased GC dose; n = 11), dose unchanged group (patients with no change in GC dose; n = 73), and dose reduction group (patients who reduced GC dose; n = 40). ΔRDW was compared among these groups, and GC doses in 2024 and 2025 were compared within each group. Supplementary analysis 1: one-year changes in DAS28-ESR, CRP, and MMP-3 according to MTX and GC dose adjustment To examine whether MTX or GC dose modifications were accompanied by changes in inflammatory status, we performed an additional longitudinal analysis using data from 2024 and 2025. The categorization of MTX and GC dose adjustment groups followed the same methodology described in the main Methods section (“Assessment of one-year change in RDW by MTX and GC dose adjustment”). One-year changes in DAS28-ESR, CRP, and MMP-3 were calculated as Δ values (2025 minus 2024) and compared across these dose–adjustment groups. Group comparisons were conducted using the Kruskal–Wallis test, followed by Bonferroni-adjusted post hoc tests when appropriate. Supplementary analysis 2: high RDW prevalence by MTX and GC dose in patients with normal hemoglobin levels Based on the World Health Organization (WHO) definition, hemoglobin levels ≥13.0 g/dL in men and ≥12.0 g/dL in women were considered normal, with values below these thresholds classified as anemia.18 To minimize the potential influence of anemia on RDW, we conducted a subgroup analysis including only patients with normal hemoglobin levels (n = 382). Supplementary analysis 3: impact of b/tsDMARDs on the proportion of high RDW Patients were stratified by the presence or absence of b/tsDMARD use in 2024, and the proportion of those with elevated RDW values was compared. To eliminate the confounding effects of MTX and GCs, RA patients receiving either MTX or GCs were excluded from this analysis. Missing data The breakdown of missing data is as follows: six for body mass index (BMI) (6/687, 0.9%), five for Steinbrocker stage (5/687, 0.7%), five for Steinbrocker class (5/687, 0.7%), 47 for anti-citrullinated protein antibody (ACPA; 47/687, 6.8%), 160 for RF (160/687, 23.3%), 12 for DAS28-ESR (12/687, 1.7%), 11 for ESR (11/687, 1.6%), five for MMP3 (5/687, 0.7%), 11 for HAQ-DI (11/687, 1.6%), and one for eGFR (1/687, 0.1%). Statistical analysis All statistical tests were two-sided, and a P-value of <.05 was considered statistically significant. Unless otherwise specified, data are presented as mean ± standard deviation. Patient characteristics were compared between the normal RDW and high RDW groups using univariate analysis (Mann–Whitney U test and Fisher’s exact test). Logistic regression analysis was performed to identify factors associated with high RDW, adjusting for age, sex, disease duration, BMI, albumin, hemoglobin, DAS28-ESR, b/tsDMARD use, MTX use, and GC use. To evaluate the effect of medication dose, the following analyses were conducted. For MTX dose, patients receiving GCs were excluded to avoid confounding. MTX users were divided into three groups based on the median dose of 8 mg/week: no use, 1 to <8 mg/week, and ≥8 mg/week. The Cochran–Armitage trend test was applied to examine trends in the proportion of high RDW across these categories. Similarly, to assess GC dose, patients receiving MTX were excluded. GC users were categorized into three groups based on the median dose of 2.5 mg/day PSL: no use, 1 to <2.5 mg/day, and ≥2.5 mg/day. The Cochran–Armitage trend test was used for this analysis as well. ΔRDW from 2024 to 2025 across three groups defined by MTX or PSL use status (dose escalation, dose unchanged, and dose reduction) were compared using the Kruskal–Wallis test with Bonferroni post hoc correction. Within-group comparisons of drug use over one year were assessed using the Mann–Whitney U test. To eliminate the influence of anemia, additional analyses were performed in patients with normal hemoglobin levels. In this subgroup, associations of high RDW with MTX dose, GC dose, and their combination were re-evaluated using the same methods. Finally, the proportion of patients with high RDW according to b/tsDMARD use was compared using Fisher’s exact test. All analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria), which is a modified version of R commander designed to incorporate statistical functions commonly used in biostatistics.19 RESULTS Patient background in normal and high RDW groups The mean age of patients was 66.7 years, and the mean disease duration was 13.4 years. GCs were used in 24.7% of patients, MTX in 61.1%, and b/tsDMARDs in 46.0%. The mean DAS28-ESR was 2.74, and the mean HAQ-DI was 0.46. Mean hemoglobin and serum albumin levels were 12.6 g/dL and 4.0 g/dL, respectively. Compared with the normal RDW group, patients in the high RDW group were older (68.0 vs 70.6 years, P = 0.025) and had higher rates of MTX use (57.6% vs 71.0%, P = 0.004) and GC use (18.8% vs 41.3%, P < 0.001). MTX dose was significantly higher in the high RDW group (7.5 vs 8.3 mg/week, P = 0.017), whereas GC dose showed no significant difference (3.0 vs 3.5 mg/day, P = 0.249). The high RDW group also had higher mean DAS28-ESR (2.68 vs 2.92, P = 0.004) and HAQ-DI scores (0.40 vs 0.64, P < 0.001), and lower mean hemoglobin (12.9 vs 11.8 g/dL, P < 0.001) and albumin (4.1 vs 3.9 g/dL, P < 0.001) levels (Table 1). Table 1 Patient background in normal and high RDW groupsOverallTotalNormal RDWHigh RDW P-valueAge, yearsMean (SD)66.7 (13.5)68.0 (13.4)70.6 (13.5)0.025Duration of disease, yearsMean (SD)13.4 (10.3)13.4 (10.1)13.5 (10.8)0.811Sex, female, %75.572.676.80.340BMI, kg/m2Mean (SD)22.1 (3.9)22.2 (4.0)21.7 (3.7)0.113Steinbrocker stage (1/2/3/4), %39.1/24.2/15.0/21.740.3/24.7/13.8/21.235.5/23.0/18.4/23.00.465Steinbrocker class (1/2/3/4), %42.3/46.6/9.9/1.246.1/45.6/7.4/0.931.6/49.3/17.1/2.0<0.001MTX use rate, %61.157.671.00.004MTX dose, mg/weekMean (SD)7.8 (2.9)7.5 (2.9)8.3 (3.0)0.017GCs use rate, %24.718.841.3<0.001GCs dose, mg/dayMean (SD)3.2 (1.9)3.0 (1.7)3.5 (2.1)0.249bDMARD/tsDMARD use rate, %46.049.237.40.014Anti-CCP antibody, U/mLMean (SD)164.7 (269.2)160.4 (253.8)176.9 (308.6)0.399Rheumatoid factor, IU/mLMean (SD)135.0 (314.0)125.8 (285.8)163.1 (387.4)0.152CRP, mg/dLMean (SD)0.40 (1.8)0.38 (2.07)0.44 (0.78)0.001ESR 1h, mmMean (SD)25.1 (21.7)23.6 (20.9)29.3 (23.4)0.005MMP-3, ng/mLMean (SD)103.2 (194.3)100.1 (217.3)111.8 (105.1)<0.001eGFR, mL/min/1.73 m2Mean (SD)64.7 (17.8)65.3 (17.8)63.0 (17.7)0.163Lung disease, %11.210.812.31DAS28-ESRMean (SD)2.74 (1.18)2.68 (1.23)2.92 (1.00)0.004HAQ-DIMean (SD)0.46 (0.70)0.40 (0.63)0.64 (0.83)<0.001Hb, g/dLMean (SD)12.6 (1.6)12.9 (1.4)11.8 (1.9)<0.001Alb, g/dLMean (SD)4.0 (0.4)4.1 (0.3)3.9 (0.3)<0.001Data are presented as mean ± standard deviation. GCs were converted to prednisolone-equivalent dose.BMI: body mass indexMTX: methotrexateGCs: glucocorticoidsb/tsDMARD: biological and targeted synthetic disease-modifying antirheumatic druganti-CCP antibody: anti-cyclic citrullinated peptide antibodyCRP: C-reactive proteinESR: erythrocyte sedimentation rateMMP3: matrix metalloproteinase 3DAS28-ESR: Disease Activity Score 28-erythrocyte sedimentation rateHAQ-DI: Health Assessment Questionnaire-Disability IndexHb: hemoglobinAlb: albuminRDW: red blood cell distribution width Patient background in normal and high RDW groups Data are presented as mean ± standard deviation. GCs were converted to prednisolone-equivalent dose. BMI: body mass index MTX: methotrexate GCs: glucocorticoids b/tsDMARD: biological and targeted synthetic disease-modifying antirheumatic drug anti-CCP antibody: anti-cyclic citrullinated peptide antibody CRP: C-reactive protein ESR: erythrocyte sedimentation rate MMP3: matrix metalloproteinase 3 DAS28-ESR: Disease Activity Score 28-erythrocyte sedimentation rate HAQ-DI: Health Assessment Questionnaire-Disability Index Hb: hemoglobin Alb: albumin RDW: red blood cell distribution width Factors associated with high RDW To identify factors associated with high RDW, logistic regression analysis was performed. The analysis revealed that lower hemoglobin levels (odds ratio [OR], 0.56; 95% confidence interval [CI], 0.48–0.66; P < 0.001), MTX use (OR, 2.47; 95% CI, 1.54–3.96; P < 0.001), and GC use (OR, 2.21; 95% CI, 1.41–3.46; P < 0.001) were independently associated with high RDW (Table 2). Table 2 Factors associated with high RDWFactorOdds ratio P-valueAge1.00 (0.99–1.02)0.759Sex, female0.72 (0.42–1.26)0.253Duration of disease1.00 (0.98–1.02)0.862BMI1.02 (0.97–1.09)0.428Alb0.70 (0.34–1.41)0.314Hb0.56 (0.47–0.66)<0.001DAS28-ESR0.95 (0.78–1.16)0.618b/tsDMARD use0.84 (0.54–1.31)0.446MTX use2.56 (1.56–4.21)<0.001GC use2.84 (1.78–4.52)<0.001BMI: body mass indexMTX: methotrexateGC: glucocorticoidb/tsDMARD: biological and targeted synthetic disease-modifying antirheumatic drugDAS28-ESR: Disease Activity Score 28-erythrocyte sedimentation rateHAQ-DI: Health Assessment Questionnaire-Disability IndexHb: hemoglobinAlb: albuminRDW: red blood cell distribution width Factors associated with high RDW BMI: body mass index MTX: methotrexate GC: glucocorticoid b/tsDMARD: biological and targeted synthetic disease-modifying antirheumatic drug DAS28-ESR: Disease Activity Score 28-erythrocyte sedimentation rate HAQ-DI: Health Assessment Questionnaire-Disability Index Hb: hemoglobin Alb: albumin RDW: red blood cell distribution width High RDW prevalence by MTX and GC dose The proportion of patients with high RDW increased with higher MTX dose. Specifically, the prevalence of high RDW was 14.0% in the non-MTX group, 18.3% in the group receiving 1 to <8 mg/week MTX, and 28.6% in the group receiving ≥8 mg/week MTX. A significant dose-dependent trend was observed (Cochran–Armitage trend test, P < 0.001). A significant difference was observed between the non-MTX group and the ≥8 mg MTX group (14.0% vs 28.6%, P < 0.001; Figure 2A). High RDW prevalence by MTX and GC dose RDW: red blood cell distribution width MTX: methotrexate GC: glucocorticoid PSL: prednisolone The proportion of patients with high RDW increased with higher GC (PSL) dose. Specifically, the prevalence of high RDW was 14.0% in the non-GC group, 20.7% in the group receiving 1 to <2.5 mg/day PSL, and 43.2% in the group receiving ≥2.5 mg/day PSL. A significant dose-dependent trend was observed (Cochran–Armitage trend test, P < 0.001). A significant difference was observed between the non-PSL group and the ≥2.5 mg PSL group (14.0% vs 43.2%, P < 0.001; Figure 2B). One-year change in RDW by MTX and GC dose adjustment Annual ΔRDW was evaluated with a focus on MTX dose. To eliminate the potential influence of PSL dose adjustments, the analysis was restricted to RA patients whose PSL dose remained unchanged over the one-year period. The ΔRDW in the dose escalation group was significantly greater than in both the dose unchanged group and the dose reduction group (dose escalation vs dose unchanged, 0.43 ± 0.91 vs −0.08 ± 1.06, P = 0.013; dose escalation vs dose reduction, 0.43 ± 0.91 vs −0.26 ± 1.18, P = 0.014). Although the ΔRDW in the dose reduction group was numerically smaller than that in the dose unchanged group (−0.26 ± 1.18 vs −0.08 ± 1.06), this difference was not statistically significant (P = 0.633; Figure 3A). Regarding MTX dose over the one-year period, in the dose escalation group, the mean MTX dose increased from 4.2 ± 3.7 mg/week in 2024 to 8.7 ± 2.6 mg/week in 2025 (P < 0.001). In the dose unchanged group, the dose remained stable at 7.6 ± 2.7 mg/week in both 2024 and 2025. In contrast, the dose reduction group showed a decrease from 9.2 ± 3.0 mg/week in 2024 to 5.2 ± 3.4 mg/week in 2025 (P < 0.001; Figure 3B). One-year change in RDW by MTX and GCs dose adjustment RDW: red blood cell distribution width MTX: methotrexate GC: glucocorticoid Annual ΔRDW was also evaluated with a focus on GC dose. To eliminate the potential influence of MTX dose adjustments, the analysis was limited to RA patients whose MTX dose remained unchanged over the one-year period. No significant differences in ΔRDW were observed among the groups (dose escalation, 0.10 ± 2.18; dose unchanged, −0.43 ± 1.56; dose reduction, 0.02 ± 0.96) (Figure 3C). Regarding GC dose during the same period, the mean dose in the dose escalation group increased from 1.2 ± 1.1 mg/week in 2024 to 4.7 ± 2.9 mg/week in 2025 (P < 0.001). In the dose unchanged group, the dose remained stable at 3.3 ± 2.2 mg/week in both years. In contrast, the dose reduction group showed a decrease from 3.0 ± 1.6 mg/week in 2024 to 1.1 ± 1.1 mg/week in 2025 (P < 0.001) (Figure 3D). DISCUSSION In this study, we investigated the association between RDW and the use of MTX and GCs in RA patients. We first used the 2024 dataset to perform cross-sectional analyses, and then incorporated the 2025 data to conduct a one-year longitudinal analysis examining the initiation and dose of both MTX and PSL in relation to ΔRDW. We found that both MTX and GC use were significantly associated with elevated RDW. Notably, patients receiving ≥8 mg/week of MTX or ≥2.5 mg/day of PSL had a higher proportion of high RDW, demonstrating a dose-dependent trend. It should be noted, however, that the cutoff values for MTX and GCs were provisionally set based on median dose among users in this study. Therefore, they do not represent universally established thresholds above which RDW becomes elevated. Although RDW has been considered an indirect marker of inflammation in RA, its relationship with inflammatory markers such as DAS28-ESR and HAQ-DI has been inconsistent across studies.8,20 In our cohort, RDW showed only a weak correlation with DAS28, but a stronger association with the use of MTX and GCs. These findings suggest that RDW may not solely reflect disease activity, but may be strongly influenced by treatment status. In addition to these cross-sectional findings, we performed a one-year longitudinal analysis (2024–2025) focusing on changes in MTX and GC dose. In this analysis, patients who initiated or escalated MTX therapy showed a significant increase in ΔRDW compared with those whose dose remained unchanged or was reduced. Interestingly, MTX dose reduction was associated with a significant decrease in ΔRDW compared with the unchanged group. These observations indicate that RDW values fluctuate in parallel with changes in MTX dose. Although only a one-year follow-up was available, most previous studies evaluating the relationship between DMARDs and RDW have been cross-sectional,9,21 and thus our longitudinal data provide stronger evidence that MTX use directly influences RDW. This represents a key strength of the present study. However, for GCs, neither initiation nor changes in dose showed a clear association with ΔRDW. This likely reflects characteristics of our cohort: only 24.7% of patients were treated with PSL and the mean prescribed dose remained very low (approximately 3–4 mg/day). Under such conditions—low prevalence and a narrow, low-dose range—the analysis is inherently underpowered and any biological effect on RDW is expected to be small. In contrast to MTX, which is actively recommended and widely used as the anchor drug in RA,22 long-term PSL therapy is not recommended; rather, rapid tapering and discontinuation are encouraged.23 Consequently, the low frequency of use and low dose of PSL observed in our study are unlikely to change in future practice. Therefore, clarifying the potential impact of PSL on RDW will require longer-term follow-up with sufficient variability in exposure. To further clarify whether changes in MTX or GC dose were accompanied by alterations in inflammatory status, we additionally analyzed longitudinal changes in ΔDAS28-ESR, ΔCRP, and ΔMMP-3 (Supplementary Material 1). These analyses showed no significant differences in any of these markers among the MTX dose adjustment groups, suggesting that factors unrelated to inflammatory activity—such as adverse effects, tolerability, or clinical symptoms not fully captured by composite indices—may have influenced MTX dose adjustments. For GC, ΔMMP-3 was lower in the dose-reduction group compared with the unchanged group, which may reflect pharmacological effects of GCs on MMP-3 production, secretion, or clearance. Serum MMP-3 values are known to be influenced by multiple therapies, including GC, as previously reported.24,25 Notably, the present study could not investigate the clinical reasoning behind MTX or GC dose escalation, reduction, discontinuation, or initiation. Evaluating these treatment decision factors will be an important direction for future research. Although our findings demonstrate a consistent association between MTX use and elevated RDW, the underlying biological mechanism remains unclear. MTX inhibits folate metabolism, a well-known effect that can impair erythroid maturation and lead to macrocytosis.26,27 Disrupted red blood cell maturation may increase cell size variability, providing a plausible explanation for higher RDW. However, confirming this hypothesis requires detailed red blood cell indices, which were beyond the scope of this study and should be addressed in future research. Given that RDW is affected by anemia,5 we performed a subgroup analysis limited to patients with normal hemoglobin levels. Consistent with the main results, higher MTX and GC dose were still associated with increased RDW in this subgroup (Supplementary Material 2). RDW is known to be influenced by various factors, including nutritional status, oxidative stress, bone marrow function, and drug exposure.6,28 Therefore, caution is warranted when interpreting RDW as a marker of disease activity in RA. Our findings highlight the need to consider treatment intensity and drug effects, particularly those of MTX and GCs, when evaluating RDW in clinical settings.29 We also conducted a sub-analysis to assess the effect of b/tsDMARDs on RDW. The results showed no significant difference in the proportion of high RDW between users and non-users of b/tsDMARDs (Supplementary Material 3), and multivariable analysis did not identify b/tsDMARD use as an independent factor associated with elevated RDW. Consistent with previous reports, we also did not find a significant association between b/tsDMARD use and increased RDW.29,30 Given the pharmacological characteristics of b/tsDMARDs, which do not directly interfere with folate metabolism or erythroid maturation—processes that may influence red blood cell size heterogeneity—it is plausible that these agents exert minimal direct effects on RDW.1 This supports the notion that MTX and GCs, rather than b/tsDMARDs, are the main treatment-related contributors to RDW elevation. Previous studies reporting an association between RDW and RA disease activity often included untreated or newly diagnosed patients.21,31 In contrast, our study population consisted mainly of patients in remission or with low disease activity, which may account for the differences in findings. As such, our results offer new insights into the interpretation of RDW in the context of well-controlled RA. Our findings suggest that elevated RDW in RA patients should not be interpreted solely as a sign of increased inflammation or nutritional deficiency. In our cohort, RDW was not associated with disease activity or with nutritional status as assessed by albumin levels, but was consistently associated with MTX and GC use, even among patients with normal hemoglobin levels. Recognizing that RDW can be influenced by treatment is important to avoid unnecessary diagnostic investigations or inappropriate treatment adjustments based on drug-related RDW changes. Previous studies have shown that a high RDW is associated with an increased risk of myocardial infarction, stroke, and all-cause mortality in RA,32 suggesting that accurate interpretation of RDW values—taking treatment status into account—may help improve cardiovascular risk stratification and support more individualized treatment monitoring strategies. In daily practice, rheumatologists often face the dilemma of whether to increase MTX dosage further or to switch early to b/ts DMARDs when disease activity persists despite MTX therapy.33 Although our current data cannot determine the predictive value of RDW for treatment response, the consistent association between RDW elevation and MTX or glucocorticoid exposure suggests that RDW may reflect cumulative treatment burden and residual therapeutic capacity.34 Patients with relatively low RDW might still have room for MTX escalation, whereas those with persistently high RDW could already be near their treatment limit and may benefit from an earlier transition to b/tsDMARDs. This interpretation remains exploratory and requires validation. To verify this hypothesis, interventional or pragmatic studies are needed. Specifically, prospective analyses of RA patients with inadequate response to MTX should compare RDW between those who improve and those who do not after MTX intensification, thereby clarifying whether RDW can guide individualized treatment intensity decisions. This study has several limitations. First, its cross-sectional and one-year longitudinal design precludes any conclusions regarding causality. Second, detailed laboratory data on nutritional status, iron, and vitamin deficiencies—which may influence RDW—were not available. Third, mean corpuscular volume (MCV), a key red blood cell index reflecting average erythrocyte size, was not consistently recorded in our observational study database. Because RDW reflects the variability in erythrocyte size rather than absolute cell size, RDW may increase even when MCV remains within the normal range. Therefore, we were unable to assess whether changes in red blood cell morphology, including potential MTX-related macrocytosis, contributed to elevated RDW.35 Fourth, treatment decisions, including MTX and GC use, were made at the discretion of attending physicians and may be confounded by unmeasured disease activity. In conclusion, RDW in RA appears to be influenced more by treatment—particularly MTX and GCs—than by inflammation. Clinicians should interpret RDW in the context of treatment status rather than as a simple marker of disease activity. Our analysis identified MTX and GC use as independent factors associated with elevated RDW, and the longitudinal data further demonstrated that MTX initiation or escalation is a decisive contributor to RDW elevation. Recognizing drug-related RDW changes may help avoid unnecessary investigations and support more personalized management of RA. AUTHORS’ CONTRIBUTIONS YO, NT, YS, and MS contributed to the study conception and design, data collection, analysis and interpretation, and drafting and critically revising the manuscript for important intellectual content. KT and SA contributed to data analysis and interpretation, and critical drafting/revision of the manuscript. KT and SI participated in the study design, data acquisition, and critical revision of the manuscript. All authors reviewed and approved the final manuscript. COMPETING INTERESTS The authors declare that they have no competing interests related to this study. AVAILABILITY OF DATA AND MATERIALS The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. ACKNOWLEDGEMENTS We sincerely thank the past and current members of Department of Orthopedic Surgery at Nagoya University Graduate School of Medicine, and Aichi Medical University in Nagakute, Japan, for their invaluable contributions to this study. Supplementary Material Funding Statement This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. REFERENCES Associated Data Supplementary Materials Data Availability Statement The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.