Introduction

Rheumatoid arthritis (RA) is a systemic autoimmune disease characterized by chronic synovial inflammation and progressive joint destruction. Peripheral neuropathy is a frequent extra-articular manifestation of the disease. Electrophysiological studies report prevalence rates of approximately 30–40% in RA; however, substantial variability exists across cohorts, with some studies reporting lower prevalence (around 8%) and others demonstrating higher rates up to 75%, with many cases remaining subclinical and detected only on nerve conduction testing [1–5]. Clinically, neuropathy in RA most often presents with predominantly sensory symptoms such as numbness, tingling, or pain in the hands and feet and may also include sensorimotor involvement, mononeuritis multiplex, or focal entrapment neuropathies such as carpal tunnel or tarsal tunnel syndromes [4–6]. These manifestations can overlap with arthritis symptoms, significantly affecting the function and quality of life.

Chronic inflammation is central to the pathogenesis of RA. Conventional markers such as erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and composite indices like 28-joint Disease Activity Score (DAS28) are widely used to assess inflammatory burden and disease activity. However, these measures often fail to capture persistent, subclinical inflammation that contributes to long-term complications. Emerging evidence highlights the neutrophil-to-lymphocyte ratio (NLR) as a simple, cost-effective, and readily available biomarker that reflects systemic immune dysregulation by integrating both innate and adaptive immune responses. Elevated NLR typically indicates heightened inflammation, characterized by neutrophilia alongside relative lymphopenia. In RA, NLR is consistently higher than in healthy controls and shows strong correlation with disease activity [7–9]. Previous research suggests that NLR demonstrates a graded association with increasing disease activity and severity [10].

Despite growing interest in NLR as a marker of disease activity in RA, its role in identifying extra-articular complications such as peripheral neuropathy has not been systematically investigated. In other chronic diseases, including diabetic peripheral neuropathy, elevated NLR has been linked to the presence and severity of nerve dysfunction, supporting the concept that subclinical inflammation contributes to neural injury [11].

Serological markers such as rheumatoid factor (RF) and anti-cyclic citrullinated peptide (anti-CCP) are established indicators of autoimmune status in RA and have been associated with more severe disease and extra-articular manifestations. These autoantibodies may identify patients with higher inflammatory burden and therefore greater risk of complications [12, 13].

In this context, we proposed that elevated NLR would be associated with the presence of peripheral neuropathy in RA. The aim of this study was to examine the association between NLR and the presence of peripheral neuropathy in RA and to evaluate its potential as a simple biomarker for characterizing neural involvement. We also explored the relationship between NLR and serological status to determine whether autoimmune profiles influence the inflammatory environment linked to neuropathy.

Materials and methods

Results

Of the 230 patients with RA included in the study, 93 (40.4%) had peripheral neuropathy confirmed by clinical and electrophysiological assessments, while 137 (59.6%) had no neuropathy. Table 1 demonstrates the demographic, clinical, serological, and treatment characteristics of the study cohort. Patients with neuropathy were significantly older than those without neuropathy (mean age 56.2 ± 10.9 years vs. 47.4 ± 12.8 years, P < 0.001) and had a longer median disease duration (10.9 vs. 5.1 years, P < 0.001). No significant differences were observed in sex distribution between the two groups (P = 0.60). Disease activity, as measured by DAS28, was significantly greater among patients with peripheral neuropathy (5.7 ± 0.9 vs. 4.5 ± 1.0, P < 0.001). When stratified by serological status, 155 patients (67.4%) were seropositive, and 75 (32.6%) were seronegative. Seropositivity was more frequent in the neuropathy group than in patients without neuropathy (76.3% vs. 61.3%, P = 0.02). Traditional inflammatory markers were also elevated in patients with neuropathy. Mean ESR was significantly higher in the neuropathy group compared with those without neuropathy (59.8 ± 18.1 mm/h vs. 44.2 ± 17.3 mm/h, P < 0.001), and median CRP levels were likewise greater among patients with neuropathy (13.4 vs. 6.8 mg/L, P < 0.001). Patients with peripheral neuropathy exhibited significantly higher neutrophil counts and lower lymphocyte counts compared with those without neuropathy, which translated into a markedly elevated NLR (median 3.8 vs. 2.3, P < 0.001). With respect to treatment characteristics, a significantly greater proportion of patients with peripheral neuropathy were receiving glucocorticoids compared with those without neuropathy (62.4% vs. 44.5%, P = 0.006), and the median daily prednisolone equivalent dose was significantly higher in the neuropathy group (P < 0.001). There were no significant differences between the groups in the use of methotrexate, other non-biologic DMARDs (including sulfasalazine, leflunomide, hydroxychloroquine, etc.), biologic DMARDs, or JAK inhibitors. The majority of the patients in both groups were prescribed NSAIDs.

When stratified by serological status, NLR differed significantly according to the presence of neuropathy (Table 2). Among seropositive subjects without neuropathy, the median NLR was 2.4 (IQR: 1.7–3.2). In contrast, seropositive patients with neuropathy had a significantly higher median NLR of 4.1 (IQR: 3.0–5.4) (P < 0.001 vs. seropositive without neuropathy). Similarly, within the seronegative subgroup, individuals with neuropathy exhibited a higher median NLR of 3.3 (IQR: 2.5–4.5) compared with seronegative patients without neuropathy, who had a median NLR of 2.2 (IQR: 1.5–3.0) (P < 0.001).

Further analysis by neuropathy subtype showed a stepwise increase in NLR (Table 3). Patients with subclinical neuropathy (n = 39) had a significantly higher median NLR of 3.2 (IQR: 2.4–4.1) than those without neuropathy (P < 0.001). Median NLR was even greater in patients with clinical neuropathy (4.6, IQR: 3.3–5.8), and this difference was significant compared with the subclinical neuropathy group (P < 0.001).

A multivariable logistic regression analysis was performed to identify factors independently associated with peripheral neuropathy (Table 4). In the model adjusted for age, disease duration, DAS28, CRP, serological status, and glucocorticoid use, higher NLR was significantly associated with increased odds of peripheral neuropathy (adjusted OR = 1.92, 95% CI: 1.48–2.49; P < 0.001). Older age (adjusted OR = 1.02, 95% CI: 1.00–1.04; P = 0.02), longer disease duration (adjusted OR = 1.07, 95% CI: 1.02–1.12; P = 0.005), higher DAS28 (adjusted OR = 1.45, 95% CI: 1.11–1.89; P = 0.005), and seropositive status (adjusted OR = 1.51, 95% CI: 1.01–2.32; P = 0.04) were also significantly associated with neuropathy. CRP and glucocorticoid use were not significantly associated with peripheral neuropathy in the adjusted model.

To evaluate the incremental value of NLR for discriminating between patients with and without peripheral neuropathy, we compared models with and without NLR (Table 5). In the overall cohort (Figure 1A), the base model including DAS28-ESR, CRP, serological status, and disease duration yielded an AUC of 0.75 (95% CI: 0.69–0.81). The addition of NLR significantly improved the model performance, as reflected by a higher AUC (0.83, 95% CI: 0.77–0.89; P < 0.001) and better reclassification and discrimination metrics. When the cohort was analyzed by serological status, the pattern remained consistent. In both seropositive (Figure 1B) and seronegative (Figure 1C) subgroups, adding NLR to the base model led to significant increases in the AUC and better classification metrics, indicating that NLR enhanced the model’s ability to discriminate patients with peripheral neuropathy regardless of serological status.

Display full size

Figure 1. ROC curves of the base model and the base model augmented with NLR across the overall cohort and RA subgroups. ROC curves comparing the discriminatory performance of the base model and the model augmented with NLR for: (A) overall cohort, (B) seropositive RA, and (C) seronegative RA. For each cohort, the true positive rate (sensitivity) is plotted against the false positive rate (1 – specificity). The AUC for the base model and the base model plus NLR is indicated in the legend of each panel. The diagonal dashed line represents the reference line for random classifier performance. AUC: area under the receiver operating characteristic curve; NLR: neutrophil-to-lymphocyte ratio; RA: rheumatoid arthritis; ROC: receiver operating characteristic.

Discussion

RA is a chronic autoimmune disease characterized by persistent inflammation, joint damage, and increased morbidity and mortality. Inflammatory markers such as CRP and ESR are commonly used to assess disease activity, but they have significant limitations and may not reflect true inflammatory burden. Imaging studies frequently demonstrate active synovitis despite normal or low levels of these markers, indicating the need for additional, accessible indicators of systemic inflammation [15, 16].

In this cohort of patients with RA, we found that an elevated NLR was independently associated with the presence of peripheral neuropathy, including subclinical involvement, even after adjustment for key confounders such as age, disease duration, disease activity, conventional inflammatory markers, serological status, and glucocorticoid use. These findings support a role for systemic inflammatory burden in neural involvement and underscore the potential utility of NLR as a potential surrogate marker of systemic inflammation and immune dysregulation. This association remained consistent across serological subgroups, with both seropositive and seronegative patients with neuropathy demonstrating significantly higher NLR values compared with their respective counterparts without neuropathy.

The overall prevalence of peripheral neuropathy in the present study was approximately 40%, with subclinical involvement identified in 41.9% of patients. Previous studies have reported a wide range of prevalence estimates, largely attributable to variations in diagnostic criteria, study populations, and assessment methodologies. Despite this heterogeneity, electrophysiological evaluations consistently indicate a substantial burden of peripheral neuropathy in RA, encompassing both overt and subclinical forms [1–5]. Our findings are concordant with this literature and underscore the frequency of neural involvement even in the absence of overt neurological symptoms.

Seropositivity was observed more frequently among patients with neuropathy, consistent with prior literature indicating that extra-articular manifestations are more common in seropositive RA than in seronegative disease. In multiple cohorts, autoantibody positivity has been associated with a higher burden of systemic manifestations beyond the joints, including neurologic, pulmonary, and vasculitic complications, compared with patients lacking these antibodies [17–20]. The higher prevalence of neuropathy and elevated NLR observed among seropositive patients in our cohort suggests that autoimmune mechanisms may be linked to heightened systemic inflammatory activity, thereby increasing susceptibility to neural involvement.

We also observed that NLR was significantly elevated not only in patients with clinically overt neuropathy but also in those with subclinical involvement. The progressive increase in NLR across patients without neuropathy, those with subclinical involvement, and those with clinical neuropathy suggests a graded relationship between systemic inflammatory imbalance and the severity of neural involvement. This stepwise pattern strengthens the biological plausibility of the observed association and supports a potential causal pathway linking sustained inflammation to neural injury. Further, multivariable logistic regression analysis revealed that NLR was independently associated with the presence of peripheral neuropathy in RA. In the incremental analysis, inclusion of NLR substantially improved the discriminative performance of the model for distinguishing patients with peripheral neuropathy. Previous research has consistently demonstrated that NLR correlates with established disease activity indices such as DAS28, ESR, and CRP, and is elevated in patients with active RA compared with those in remission or healthy controls [16, 21–23]. However, the present findings extend this literature by demonstrating that NLR contributes independent diagnostic value for peripheral neuropathy in RA, even after adjustment for established correlates such as disease activity, serological status, and disease duration. This suggests that NLR not only reflects general systemic inflammation but may also be sensitive to inflammatory pathways linked with microvascular and neural injury in RA, offering incremental information beyond conventional markers.

The pathophysiological association between NLR and peripheral neuropathy in RA is supported by established immunological mechanisms underlying the disease. Neutrophils play a central role in RA pathogenesis through the release of pro-inflammatory cytokines (such as IL-1, TNF-α, and IL-6), reactive oxygen species, and proteolytic enzymes, which can contribute to endothelial dysfunction and microvascular injury [24, 25]. In parallel, relative lymphopenia reflects impaired immune regulation, particularly an imbalance between regulatory T cells (Tregs) and pro-inflammatory Th17 cells that sustains systemic inflammation [26–30]. Peripheral nerves are especially vulnerable to chronic ischemic and inflammatory damage, with microvascular injury promoting axonal degeneration and demyelination even in the absence of overt vasculitis. By capturing both innate immune activation and adaptive immune dysregulation, NLR may reflect this pathogenic milieu more effectively than ESR or CRP alone.

This study has certain limitations that merit consideration. The cross-sectional design precludes causal inference, and longitudinal studies are required to determine whether elevated NLR predicts the development or progression of neuropathy over time. While we adjusted for major confounders, including age, disease duration, disease activity (DAS28), conventional inflammatory markers (ESR, CRP), serological status, and glucocorticoid use in our multivariable analyses, residual confounding by unmeasured factors or treatment-related effects cannot be entirely excluded. ROC analysis was performed to evaluate the incremental discriminatory value of NLR; however, an optimal NLR cutoff was not derived. A specific threshold for NLR has not been established for peripheral neuropathy in RA, and cutoffs can vary across populations. Additionally, categorizing a continuous variable can reduce statistical power and obscure the underlying association. Future longitudinal and validation studies are warranted to establish clinically meaningful NLR thresholds for neuropathy assessment. Nevertheless, the relatively large sample size and systematic electrophysiological assessment strengthen the validity of the observed associations.

In conclusion, this study provides robust evidence that an elevated NLR is independently associated with the presence of peripheral neuropathy in RA, including subclinical involvement, after adjustment for key confounders such as age, disease duration, disease activity, conventional inflammatory markers, serological status, and glucocorticoid use. The observed association in both seropositive and seronegative subgroups highlights the relevance of systemic inflammatory balance in neural involvement in RA. The improvement in model discrimination observed with the inclusion of NLR beyond conventional clinical and laboratory parameters supports its potential utility as a simple and accessible biomarker for characterizing neural involvement in clinical assessments of RA patients. Integration of NLR into routine evaluation frameworks may enhance recognition of patients with concurrent neuropathy and contribute to more comprehensive clinical management.