Introduction

Distal hand osteoarthritis (OA) is a common and debilitating joint condition that lacks disease-modifying treatments1. It primarily affects postmenopausal women and is characterized by deformity and structural changes in distal interphalangeal (DIP) and proximal interphalangeal (PIP) joints, leading to pain, reduced function, and impaired quality of life2. Established risk factors include female sex, advanced age, genetic predisposition, repetitive hand use3, and metabolic disturbances such as metabolic syndrome and obesity4,5,6.

Obesity is widely recognized as a risk factor for digital OA, particularly in the DIP and PIP joints7,8,9, with mechanisms thought to involve low-grade systemic inflammation and adipokine dysregulation7. Although some recent studies have reported weaker or null associations between body mass index (BMI) and hand OA10,11, earlier systematic reviews continue to support a positive link between obesity and distal hand OA, particularly at the DIP level9. In parallel, metabolic syndrome has been associated with radiographic hand OA and with distal joint pain, but not with involvement of metacarpophalangeal (MCP) or first carpometacarpal (CMC1) joints8.

In patients with rheumatoid arthritis (RA), obesity has been shown to increase the risk of RA development12,13 but is paradoxically associated with less structural joint damage despite contributing to higher Disease Activity Score 28 (DAS28) scores through elevated subjective symptoms (e.g., pain). Patients with RA and obesity often demonstrate slower radiographic progression, particularly in early RA treated with conventional synthetic Disease-Modifying Anti-Rheumatic Drugs (csDMARDs) or anti-tumor necrosis factor (anti-TNF) agents14, though they may exhibit poorer therapeutic response overall14,15. These observations have been attributed to lower adiponectin levels and other metabolic mediators influencing inflammation and bone remodelling14.

The relevance of obesity for OA development in RA remains poorly defined. While prior work has shown that DIP joint OA can progress independently of RA disease activity and serological markers such as rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA)16, it remains unclear whether obesity plays a meaningful role in this process. Notably, DIP joints are relatively spared from RA-related synovitis and erosive damage compared to MCP, PIP, or CMC joints, making them a more suitable site for assessing osteoarthritic changes in RA populations. Therefore, evaluating OA in DIP joints provides a unique opportunity to study non-inflammatory structural changes within the context of RA.

In this study, we aimed to determine whether baseline BMI, analyzed both continuously and categorically, was associated with the incidence and progression of DIP joint OA in patients with RA. We also examined associations between BMI and specific radiographic features of DIP OA, including joint space narrowing (JSN), osteophytes worsening, subchondral sclerosis, and central erosions.

Patients and methods

Data

We obtained data from the Swiss Clinical Quality Management in Rheumatic Diseases registry (SCQM)17. The SCQM is a national, multicenter registry in Switzerland that longitudinally collects clinical data and radiographic images from patients with RA to support disease monitoring and management. This study used anonymized data from the SCQM registry. All patients provided informed consent for their data to be used for research. The study was reviewed and approved for scientific integrity and appropriate use of SCQM data by the SCQM-affiliated coauthors. In accordance with Swiss human research regulations, no additional ethics approval was required for analyses using fully anonymized data.

Baseline and follow-up data

We analyzed radiographic data of DIP joints from baseline and the latest available follow-up for each patient. Radiographs were collected between November 1983 and October 2014. All available films, including those acquired in the 1980s, were digitized where necessary and rescored. Images that did not meet predefined quality criteria (exposure, positioning, or resolution) were excluded. Details of the scoring system and its reliability are provided in the Outcomes section.

Exposure

The primary exposure variable was baseline BMI (kg/m2), analyzed as a continuous variable in the primary analysis and categorized as Normal weight, Overweight, and Obesity in the sensitivity analysis (details provided below). Weight and height were measured and recorded by clinical staff (not self-reported).

Outcomes

We investigated the following six radiographic outcomes: 1) incidence of DIP joint OA; 2) progression of DIP joint OA; 3) worsening of JSN of DIP joints; 4) worsening of osteophytes of DIP joints; 5) incidence of subchondral sclerosis on DIP joints; and 6) incidence of central erosions on DIP joints. All of our outcomes were binary (i.e., Yes or No). To investigate these radiographic outcomes, we created two cohorts: a ‘DIP joint OA incidence cohort’ and a ‘DIP joint OA progression cohort’.

We assessed the incidence and progression of DIP joint OA based on modified Kellgren-Lawrence (K/L) grades18: 0 = no DIP joint OA; 1 = doubtful DIP joint OA; 2 = mild DIP joint OA; 3 = moderate DIP joint OA; and 4 = severe DIP joint OA18. Our reader demonstrated good intra-reader agreement, evidenced by a weighted kappa exceeding 0.90. This agreement was tested using 200 randomly selected radiographs by rereading the images after two months interval. The reader examined images in a blinded fashion, without access to time or clinical information. A DIP joint was considered to have DIP joint OA if the DIP joint had a K/L grade of ≥ 2. For DIP joint OA, we investigated eight DIP joints per patient (bilateral 2nd to 5th digits). Our reader, using the Osteoarthritis Research Society International (OARSI) atlas19, provided semi-quantitative scoring for JSN and osteophytes (graded 0–3), and scored the presence of subchondral sclerosis, and central erosions. The reader had a good intra-reader agreement for JSN (kappa = 0.83), osteophytes (kappa = 0.77), subchondral sclerosis (kappa = 0.91), and central erosions (kappa = 0.92).

Our first outcome was the incidence of DIP joint OA, which we defined as a patient having it at the latest follow-up but not at baseline. We investigated this outcome in the 'DIP joint OA incidence cohort,' consisting of DIP joints without DIP joint OA at baseline. Our second outcome of progression of DIP joint OA was investigated in the ‘DIP joint OA progression cohort’, which consisted of DIP joint that had DIP joint OA at baseline. Progression of DIP joint OA was defined as an increase of ≥ 1 in the KL grade in at least one DIP joint between baseline and the latest follow-up. To test the robustness of our findings, we conducted a sensitivity analysis using a more stringent definition of progression, defined as worsening in at least two DIP joints (rather than one).

The outcomes of worsening of JSN, worsening of osteophytes of DIP joints, incidence of subchondral sclerosis, and incidence of central erosions were investigated in the ‘DIP joint OA incidence cohort’ and the ‘DIP joint OA progression cohort’. The worsening of JSN and worsening of osteophytes of DIP joints were defined as an increase ≥ 1 in the grade of JSN and osteophytes, respectively, in at least one DIP joint between baseline and the latest follow-up. The incidence of subchondral sclerosis, and incidence of central erosions were defined as the presence of subchondral sclerosis and central erosions, respectively, in at least one DIP joint at the latest follow-up while not present at baseline.

Selection of patients

This analysis was restricted to patients in the SCQM RA registry, which includes only individuals with a clinical diagnosis of RA. To establish the study cohorts, we first excluded patients with follow-up durations of less than two years and those with a baseline BMI in the underweight range (< 18.5 kg/m2) (Fig. 1). The two-year follow-up threshold was chosen to allow sufficient time for potential changes in DIP joint OA to become detectable. Additionally, patients with an underweight BMI were excluded as this may indicate underlying health conditions that could confound the association under investigation. Following these exclusions, we selected patients with at least one DIP joint evaluable on radiographs at both baseline and follow-up. From the remaining 2512 patients, we created the ‘DIP joint OA incidence cohort’ by including only the patients that did not have DIP joint OA at baseline, and the ‘DIP joint OA progression cohort’ by including only the patients with DIP joint OA at baseline (Fig. 1). Median radiographic follow-up was 5.0 years (IQR 3.2–7.3) in the incidence cohort and 4.9 years (IQR 3.2–7.7) in the progression cohort.

Fig. 1
figure 1

Selection of patients for our investigations. BMI: body mass index; DIP: distal interphalangeal; OA: osteoarthritis; RA: rheumatoid arthritis.

Full size image

Statistical analyses

We used logistic regression to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) for the association between our exposure of BMI our six outcomes (i.e., incidence of DIP joint OA; progression of DIP joint OA; worsening of JSN; worsening of osteophytes; subchondral sclerosis, and incidence of central erosions) over the latest follow-up. Univariable (unadjusted) and multivariable (adjusted) analyses were performed. All of the multivariable analyses were adjusted for sex, and baseline values of the following variables: age, ACPA, RF, erythrocyte sedimentation rate (ESR), disease duration since first symptoms, number of tender joints (28 joints), number of swollen joints (28 joints), patient assessed general health visual analog scale (VAS) score, summed percentage of erosive joint surface destruction in 8 MCP joints (to adjust for RA related joint damage), and radiographic observation time. Additionally, adjusted for summed K/L grade in 8 joints at baseline for the outcomes for the progression cohort. These variables were selected because of their association with exposure and outcomes16,20. To evaluate whether the association between BMI and outcomes varied by radiographic observation time, we tested for effect modification by including an interaction term between BMI and radiographic observation time in the multivariable models.

We used R (version 4.3.1). for our analyses. We set our threshold for statistical significance as a two-tailed P value of less than 0.05 in all analyses. Baseline missing values were imputed using multiple imputations by chained equations (m = 50), assuming data were missing at random.

Sensitivity analysis

To complement the primary analysis, which treated BMI as a continuous variable, we conducted a sensitivity analysis using BMI categories. Specifically, BMI was categorized as Healthy Weight (18.5–24.9 kg/m2), Overweight (25.0–29.9 kg/m2), and Obesity (≥ 30.0 kg/m2). By comparing the findings from this categorical approach with those of the continuous analysis, we aimed to assess whether categorizing BMI provided distinct insights, thereby enhancing the clinical relevance of our results.

Results

In the DIP joint OA incidence cohort, which included 1,031 patients without DIP joint OA at baseline, 251 individuals (24.3%) developed incident DIP joint OA during follow-up (Table 1). Those who developed incident DIP joint OA were older on average (53.0 ± 9.6 years) compared to those who did not (50.6 ± 10.2 years, p < 0.001). Women were more likely to develop incident DIP joint OA (203 of 251, 80.9%) than men (48 of 251, 19.1%, p = 0.006). Baseline BMI (mean 25.4 ± 4.5 kg/m2 vs. 25.0 ± 4.6 kg/m2, p = 0.330) were similar in those with and without incident DIP joint OA. Patients with incident DIP joint OA had a longer disease duration since first symptoms (median 7.2 vs. 5.1 years, p = 0.012) at baseline, and longer radiographic observation time (median 6.1 vs. 4.6 years, p < 0.001). Other baseline characteristics, including serological markers (ACPA and RF) and tender or swollen joint counts, were comparable between patients with DIP joint OA incidence and no DIP joint OA incidence (Table 1).

Table 1 Baseline characteristics of patients in the DIP joint OA incidence cohort, stratified by incidence of DIP joint OA at follow-up.
Full size table

In the DIP joint OA progression cohort, which included 1,481 patients with DIP joint OA at baseline, 1006 individuals (67.9%) experienced progression of the disease (Table 2). Patients with progression were of similar age (60.0 ± 9.7 years) to those without progression (61.1 ± 10.6 years, p = 0.051). Women constituted a greater proportion of those with disease progression (786 of 1006, 78.1%) compared to men (220 of 1006, 21.9%, p = 0.046). Baseline BMI (mean 26.0 ± 4.6 kg/m2 vs. 25.8 ± 4.6 kg/m2, p = 0.473) were also similar. Notably, patients with progression had a shorter disease duration since the first symptoms (median 6.5 vs. 8.0 years, p = 0.010) at baseline but longer radiographic observation time (median 5.7 vs. 4.0 years, p < 0.001). Other clinical and radiographic characteristics, such as tender and swollen joint counts and summed erosive joint surface destruction, did not differ significantly (Table 2).

Table 2 Baseline characteristics of patients in the DIP joint OA progression cohort, stratified by progression of DIP joint OA at follow-up.
Full size table

BMI and the incidence of DIP joint OA (DIP joint OA incidence cohort)

There was no evidence of an association between BMI and the odds of incidence of DIP joint OA (Table 3). The odds ratio (OR) for the incidence of DIP joint OA was 1.02 (95% confidence interval [CI] 0.98–1.05) in the univariable analysis and was 1.02 (95% CI 0.98–1.06) in the multivariable analysis. Similarly, there was no evidence of an association between BMI and the odds of worsening of JSN, worsening in osteophytes, incidence of sclerosis, and incidence of erosion in either univariable or multivariable analysis (Table 3). None of the interactions between BMI and radiographic observation time were statistically significant (p > 0.05) across any outcomes investigated in the incidence cohort.

Table 3 Association of Body Mass Index (BMI) at baseline with the odds of outcomes investigated.
Full size table

BMI and the progression of DIP joint OA (DIP joint OA progress cohort)

There was no evidence of an association between BMI and the odds of progression of DIP joint OA (Table 3). The OR for the progression of DIP joint OA was 1.01 (95% CI 0.98–1.04) in the univariable analysis and was 1.012 (95% CI 0.98–1.04) in the multivariable analysis. Findings were consistent when applying the more stringent progression definition (≥ 2 joints), with no statistically significant associations observed. Similarly, there was no evidence of an association between BMI and the odds of worsening of JSN, worsening in osteophytes, incidence of sclerosis, and incidence of erosion in either univariable or multivariable analysis (Table 3). As in the incidence cohort, no statistically significant interactions were observed between BMI and radiographic observation time for any outcomes investigated in the progression cohort.

Sensitivity analysis

Our sensitivity analysis, in which we used BMI categories, showed no association of BMI categories with any of the outcomes either in univariable or multivariable analysis (Table 4).

Table 4 Sensitivity analysis: Association of Body Mass Index (BMI) categories at baseline with the odds of outcomes investigated.
Full size table

Discussion

Our study systematically explored the relationship between BMI and the incidence and progression of DIP joint OA in a large cohort of patients with RA, contributing novel insights to the understanding of OA pathogenesis in non-weight-bearing joints. Notably, we found no association between BMI, whether analyzed as a continuous variable or in categories, and the odds of incidence or progression of DIP joint OA. This includes outcomes related to radiographic structural changes including worsening of JSN, worsening of osteophytes, incidence of subchondral sclerosis, and incidence of central erosions in DIP joints. Furthermore, this lack of association was consistent across different durations of radiographic observation. These findings suggest that the biomechanical and metabolic factors linked to BMI may not substantially influence the risk of DIP joint OA in patients with RA.

In line with our findings, DIP joint OA seems less influenced by BMI, unlike weight-bearing joints such as the knee where elevated BMI associated with increased OA risk, likely through mechanical loading effects21. This distinction is consistent with earlier studies showing that BMI or weight change is associated with knee OA with inconsistent evidence for hip OA, while data specific to DIP OA remain scarce1.

Our findings contribute to the ongoing discussion on the role of inflammation and systemic factors in DIP joint OA pathogenesis in RA. Obesity is known to induce low-grade systemic inflammation that may influence OA in weight-bearing joints22,23, but our results suggest this mechanism may not substantially affect DIP joint OA. We examined ACPA and RF serostatus as RA-specific inflammatory markers. ACPA positivity was statistically less common among patients with incident DIP joint OA, suggesting a modest association; however, this was not observed in the progression cohort and should be interpreted with caution. This observation should not be interpreted as evidence that ACPA-negative status is protective or that ACPA positivity suppresses DIP OA, but rather as a modest, hypothesis-generating signal requiring confirmation in independent cohorts. No significant differences were observed for RF or other RA markers. These findings extend our prior work16, which reported that DIP joint OA progression in RA occurred independently of disease activity or serological status. Although that study16 did not evaluate BMI directly, our current results complement its conclusions by showing no association between BMI and DIP joint OA incidence or progression. Taken together, the absence of associations with BMI or inflammatory markers supports the hypothesis that DIP joint OA in RA may be driven primarily by local or joint-intrinsic mechanisms, rather than by systemic metabolic or inflammatory processes. However, local biomechanical consequences of RA, such as structural damage and deformities at the MCP or PIP joints, may alter hand biomechanics and shift loading patterns distally, potentially predisposing DIP joints to secondary osteoarthritic changes.

The absence of a BMI–DIP joint OA association highlights the importance of distinguishing DIP joint OA pathogenesis from that of weight-bearing joints like the knee when making clinical recommendations. While weight loss in patients with elevated BMI remains a critical intervention for knee OA management due to its biomechanical benefits24,25, such recommendations may not be directly applicable to patients with isolated DIP joint OA. In cases of DIP joint OA, interventions targeting mechanical stress (e.g., ergonomic tools to reduce hand strain) may be more relevant than systemic strategies aimed at reducing adiposity.

Obesity has been paradoxically associated with reduced radiographic joint damage in RA despite higher subjective disease activity scores, a finding thought to involve adipokine-mediated effects on bone remodelling14. Our study did not find any association between BMI and DIP OA, which aligns with the notion that the effects of obesity on joint damage may differ across joint sites. However, our data do not suggest that elevated BMI has a beneficial effect on bone, but rather indicate an absence of association. Whether metabolic mediators such as adipokines contribute to DIP joint OA pathogenesis remains uncertain and warrants further investigation.

This study has several strengths, including a large sample size, detailed radiographic assessments, and multivariable analyses adjusting for potential confounders. Sensitivity analyses using BMI categories further strengthen the results. However, several limitations should be acknowledged. First, the cohort was predominantly composed of older White women, limiting the generalizability of our findings to younger, male, or more diverse populations. Second, although follow-up period was comparable to prior studies, it may not fully capture the slow progression of DIP joint OA in some individuals. Third, we lacked data on other metabolic or hormonal factors (e.g. lipid profiles, insulin resistance, post-menopausal changes), which could contribute to DIP joint OA pathogenesis and may partly explain the higher incidence observed in women. Fourth, information on hand dominance, occupational workload, and hand use was unavailable, preventing analyses of mechanical stress on DIP joint OA. Fifth, potential effects of biologic or targeted synthetic DMARDs (b/tsDMARDs) on bone remodelling, and thus on the BMI-DIP joint OA relationship, were not assessed. Sixth, while BMI is widely used as a proxy for body composition, it does not distinguish between adiposity and lean mass, which may have divergent effects on DIP joint OA risk. Seventh, because our analyses were restricted to DIP joints, the findings cannot be extrapolated to other hand sites (e.g., PIP or CMC), where RA damage complicates evaluation and the potential role of BMI may differ. Eighth, we modelled BMI and RA disease activity at baseline only; longitudinal changes in BMI or inflammation during follow-up could not be incorporated due to data structure. Lastly, while our study was powered to detect moderate effects (e.g., OR ≥ 1.15), it may have been underpowered to detect small associations, particularly for rare outcomes such as central erosions. As with all observational studies, causality cannot be inferred. Future research should examine the interplay between systemic metabolic and genetic factors in DIP joint OA, and advanced imaging modalities such as MRI could help clarify the role of local inflammation and microstructural changes.

In conclusion, our study found no association between BMI and either the incidence or progression of DIP joint OA in RA patients. These findings suggest that BMI-related biomechanical and metabolic factors may have a limited or non-detectable role in DIP joint OA pathogenesis in RA patients. This underscores the need for DIP joint-specific approaches in OA management and highlights the distinct nature of DIP joint OA compared to weight-bearing joint OA, such as the knee.