Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by persistent synovitis that can lead to joint destruction, disability, and extra-articular complications (1). Early diagnosis and accurate assessment of disease activity are critical to preventing irreversible damage and improving long-term outcomes (2). However, diagnostic challenges remain, particularly in early or seronegative RA, where patients lack rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA). Seronegative RA accounts for 20–30% of cases and often experiences delayed recognition and treatment initiation, resulting in worse prognosis (3).

Beyond conventional biomarkers such as C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) (4), several indices derived from routine complete blood counts (CBC) have emerged as potential markers of systemic inflammation. These include the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), the systemic immune-inflammation index (SII), and the systemic inflammation response index (SIRI). CBC-derived indices are inexpensive, widely available, and easily calculated, making them attractive for use in daily clinical practice (5, 6).

Growing evidence suggests that these indices reflect disease activity and outcomes in autoimmune diseases. In anti-neutrophil cytoplasmic antibody-associated vasculitis (AAV), NLR, PLR, MLR, SII, and SIRI are significantly elevated compared to healthy controls and correlate with disease activity and prognosis (7–9). Similar associations have been reported in systemic lupus erythematosus (SLE) and idiopathic inflammatory myopathies (10, 11). In RA, small studies have indicated that NLR and PLR are higher in patients than in controls and associated with disease activity scores, but the overall evidence remains limited (12–15).

Importantly, most prior RA studies did not stratify patients by serostatus, despite the clinical and immunological heterogeneity between seropositive and seronegative RA. Moreover, comparative evaluation of all five indices in a single cohort has been lacking, and their ability to predict active disease remains uncertain (16).

Therefore, the present study aimed to systematically evaluate the diagnostic performance and activity-related utility of NLR, PLR, MLR, SII, and SIRI in a large RA cohort, with a specific focus on differential performance between seropositive and seronegative patients.

Consistent with prior cohorts and meta-analyses, all five CBC-derived indices were elevated in RA versus controls, supporting their role as hematologic proxies of systemic inflammation. For example, Jin et al. showed that both NLR and PLR were markedly higher in RA than in non-RA groups (26), and meta-analysis likewise found higher NLR and PLR in RA versus controls (27, 28). These increases are biologically plausible given RA’s immune dysregulation, in which neutrophils and platelets are activated while lymphocyte counts decline, reflecting chronic systemic inflammation (12). In line with earlier work, we also observed positive correlations between NLR and disease activity markers (e.g., CRP, DAS28); similar associations have been reported in cohort studies and meta-analyses (29, 30). Moreover, several studies demonstrated higher NLR (and sometimes PLR) in active RA than in remission (31, 32).

To our knowledge, few studies have assessed all five indices in parallel within a single RA cohort; our analysis provides such head-to-head evidence and clarifies their comparative diagnostic and activity-related performance; most assessed one or two (e.g., NLR/PLR, or SIRI alone) (14, 33, 34). By examining them side-by-side, we showed that each index distinguished RA from controls (all P < 0.001 in our dataset).

Another contribution of this study is the explicit evaluation of seronegative RA (SN-RA). Prior literature often under-reports serostatus or focuses on seropositive disease (3). We found that SN-RA patients also exhibit elevated NLR, SII, and SIRI versus healthy controls, with levels broadly comparable to seropositive RA in our cohort. This aligns with data indicating that several CBC-derived indices do not differ materially by RF/ACPA status (33) and underscores their potential serology-independent utility. Given the diagnostic challenges of SN-RA, especially when traditional antibodies are absent, our findings support incorporating these indices as adjunct markers (3).

Despite overall agreement with prior work, two observations diverged from expectations. First, the PLR showed weaker diagnostic and activity-related performance in our cohort. Although higher in RA than controls, PLR contributed less to multivariable models and correlated only modestly with activity. Our results mirror a large cross-sectional analysis showing modest sensitivity (~57%) of PLR for active disease, underscoring its limited individual-level utility (35). Platelet counts are susceptible to confounding (iron deficiency, glucocorticoids, comorbidities), and antirheumatic therapy can dampen thrombocytosis, potentially lowering PLR’s signal-to-noise.

Second, the MLR displayed an attenuated association with RA activity. Although monocytes are central to synovial macrophage biology, we observed minimal differentiation of MLR across activity states and weak correlation with DAS28-CRP relative to NLR/SII/SIRI. This mirrors findings from Medicine (2021), where LMR (the inverse of MLR) did not differ significantly between active and remission groups and was not a useful independent activity marker (35). A plausible explanation is that peripheral monocyte counts fluctuate less with acute disease shifts (due to tissue trafficking or plateauing), yielding a narrower dynamic range for activity monitoring. By contrast, neutrophil-centric measures (NLR, SII, SIRI) may more promptly reflect cytokine-driven inflammatory surges.

Elevations in NLR, PLR, SII, and SIRI reinforce a model in which innate immune activation (neutrophils/platelets/monocytes) coexists with relative lymphopenia in RA, driven by cytokine networks (e.g., TNF-α, IL-6) and sustained synovial inflammation (36–38). In particular, neutrophils and monocytes are implicated in synovial damage and systemic inflammation; thus, higher NLR and SIRI may index neutrophil/monocyte expansion and activation (36, 39). Conceptually, SII (platelets × neutrophils / lymphocytes) integrates pro-inflammatory myeloid responses with reduced lymphocyte-mediated regulation, offering a composite snapshot of systemic immune-inflammation (33).

In terms of clinical application, CBC-derived indices may serve as adjunctive tools to support early detection or triage, rather than as stand-alone diagnostic markers. In primary care settings, elevated NLR or SII values in patients presenting with inflammatory joint symptoms may raise suspicion for inflammatory arthritis while awaiting definitive serological testing (e.g., RF/ACPA) or imaging confirmation. Pooled evidence suggests acceptable diagnostic accuracy for NLR (AUC ≈ 0.70–0.76 for RA presence or active disease) and moderate accuracy for PLR in distinguishing RA from non-RA conditions, albeit with substantial heterogeneity across studies (28).

Beyond initial assessment, these indices may contribute to the monitoring of established RA as complements to clinical evaluation and conventional acute-phase reactants. Notably, imaging studies have shown that ultrasound can detect subclinical synovitis in patients classified as being in remission by DAS28, highlighting limitations of standard clinical and laboratory markers; in this context, modestly elevated NLR or SII values may reflect residual systemic inflammation and warrant closer follow-up rather than immediate treatment escalation (38, 40).

Finally, accumulating population-based evidence has linked NLR, SII, and SIRI to adverse long-term outcomes in RA, including all-cause and cardiovascular mortality as well as skeletal complications such as vertebral fractures (13, 39, 41, 42). These findings suggest potential prognostic relevance of CBC-derived indices, although their interpretation should remain cautious given their non-specific nature and susceptibility to treatment- and comorbidity-related confounding.

This study has three main limitations. First, CBC-derived inflammatory indices are non-specific markers of systemic inflammation and therefore cannot substitute for established diagnostic tools such as RF/ACPA testing, imaging modalities, or comprehensive clinical assessment; their potential value lies primarily in an adjunctive role. In addition, this was a single-center, retrospective, cross-sectional analysis, which limits causal inference and generalizability.

Second, the comparison was restricted to RA versus healthy controls, preventing assessment of test specificity against other inflammatory arthritides, osteoarthritis, or infection. CBC-derived indices may also be influenced by medications, comorbid conditions, smoking status, or iron deficiency. Moreover, certain RA therapies (e.g., IL-6 receptor blockade and JAK inhibitors) can alter leukocyte and platelet profiles as well as acute-phase reactants, potentially confounding CBC-derived indices and their correlations with disease activity.

Third, we lacked longitudinal sampling and outcome follow-up, precluding evaluation of dynamic changes, treatment responsiveness, flare prediction, or long-term prognostic performance. The cut-off values and AUC estimates reported here were derived within this dataset and have not yet been externally validated. Consequently, our findings should be considered exploratory and hypothesis-generating, pending confirmation in prospective, multicenter studies.

Future research should focus on prospective validation and on mechanistic work linking these hematologic indices to cytokine pathways and synovial immune phenotypes.

In conclusion, our study provides a comprehensive comparative evaluation of five readily available blood-derived inflammatory markers in RA. We demonstrated that NLR, PLR, MLR, SII, and SIRI are all elevated in RA patients relative to healthy individuals, reflecting the heightened innate immune activation characteristic of this disease. Among these indices, those incorporating neutrophils and platelets (NLR, SII, SIRI) showed the strongest associations with RA diagnosis and disease activity, whereas PLR and MLR appeared less responsive to disease fluctuations. Particularly MLR remained informative in seronegative RA, whereas SII showed limited diagnostic utility. Taken together, these inexpensive indices have potential clinical utility as adjuncts for early detection of RA and for monitoring inflammation and disease activity over time.