Real-world damage accrual in Sjögren’s disease: a 5-year multicentre GIRRCS cohort analysis Authors: Onorina Berardicurti, Annalisa Marino, Luca Navarini, Andrea Pilato, Marta Vomero, Letizia Pia Di Corcia, Irene Genovali, Damiano Currado, Lidia La Barbera, Rosaria Irace, Paola Conigliaro, Victoriya Pavlych, Valeria Guardo, Giorgio Carlino, Sara Ferrigno, Rosa Daniela Grembiale, Annamaria Iagnocco, Piero Ruscitti, Paola Cipriani, Maria Sole Chimenti, Francesco Ciccia, Giuliana Guggino, Roberto Giacomelli Published: 2026 ABSTRACT Abstract:IntroductionPrimary Sjögren’s disease (SD) is a systemic autoimmune disorder causing glandular dysfunction, sicca symptoms and extraglandular manifestations, which impair quality of life. Data on early irreversible damage and its predictors remain limited.MethodsWe conducted a multicentre retrospective cohort study of 429 adult patients with SD from the Gruppo Italiano di Ricerca in Reumatologia Clinica e Sperimentale network with ≥1 year of follow-up. Baseline assessments included disease activity (European League Against Rheumatism (EULAR), Sjögren’s Syndrome Disease Activity Index (ESSDAI)), patient-reported symptoms (EULAR Sjögren’s Syndrome Patient Reported Index (ESSPRI)), C reactive protein (CRP), Schirmer’s test (ST) and autoantibody status. Irreversible damage was evaluated using the Sjögren’s Syndrome Damage Index (SSDI) across ocular, oral and systemic domains. Predictors of damage accrual were analysed using Cox proportional hazards models and Kaplan-Meier curves. Exploratory Least Absolute Shrinkage and Selection Operator (LASSO) and principal component analysis (PCA) analyses assessed domain-specific contributions and relationships between subjective and objective measures.ResultsAt baseline, 48% of patients had SSDI≥1: ocular 25.4%, systemic 22.5% and oral 14.3%. After 5 years, 65% had cumulative damage: ocular 34.2%, oral 21.7% and systemic 40.8%. Higher baseline ESSDAI predicted damage across all domains. Elevated ESSPRI was associated with oral damage, while abnormal ST strongly predicted ocular damage. Anti-Ro/La antibodies, elevated CRP and recurrent oral infections were linked to higher domain-specific damage. LASSO analyses showed glandular, peripheral nervous system and pulmonary ESSDAI domains independently contributed to systemic damage. PCA confirmed that patient-reported symptoms and systemic activity represent complementary, independent dimensions.ConclusionIrreversible damage in SD occurs early, often present at diagnosis and affects multiple organ domains. Early identification of high-risk patients, based on clinical, serological and patient-reported measures, may enable timely interventions to prevent further damage, guide treatment and improve long-term outcomes. WHAT IS ALREADY KNOWN ON THIS TOPIC Sjögren’s disease (SD) causes glandular and systemic manifestations, impairing quality of life. Sjögren’s disease (SD) causes glandular and systemic manifestations, impairing quality of life. Irreversible damage occurs, but its early onset and predictors were unclear. Irreversible damage occurs, but its early onset and predictors were unclear. Few studies evaluated risk factors for early damage using baseline clinical and serological data. Few studies evaluated risk factors for early damage using baseline clinical and serological data. WHAT THIS STUDY ADDS Nearly 50% of patients had damage at diagnosis; 65% developed damage within 5 years. Nearly 50% of patients had damage at diagnosis; 65% developed damage within 5 years. Baseline European League Against Rheumatism (EULAR) Sjögren’s Syndrome Disease Activity Index, EULAR Sjögren’s Syndrome Patient Reported Index, abnormal Schirmer’s test and anti-Ro/La positivity predict domain-specific damage. Baseline European League Against Rheumatism (EULAR) Sjögren’s Syndrome Disease Activity Index, EULAR Sjögren’s Syndrome Patient Reported Index, abnormal Schirmer’s test and anti-Ro/La positivity predict domain-specific damage. Damage accrues early across ocular, oral and systemic domains, emphasising the need for early monitoring. Damage accrues early across ocular, oral and systemic domains, emphasising the need for early monitoring. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Early identification of high-risk patients enables timely interventions to prevent irreversible damage. Early identification of high-risk patients enables timely interventions to prevent irreversible damage. Comprehensive assessment should combine disease activity, serology and patient-reported symptoms. Comprehensive assessment should combine disease activity, serology and patient-reported symptoms. Routine damage evaluation should be incorporated in SD management and future trials to improve long-term outcomes. Routine damage evaluation should be incorporated in SD management and future trials to improve long-term outcomes. Introduction Primary Sjögren’s disease (SD) is a systemic autoimmune disorder characterised by lymphocytic infiltration of exocrine glands, leading to glandular dysfunction and sicca symptoms.1 2 Besides dryness, systemic manifestations and an increased risk of lymphoma have been reported.13 Furthermore, patients experience a significant symptom burden, resulting in a reduction in the quality of life (QOL) and workability.4 There is no specific treatment for SD, and existing strategies primarily focus on symptomatic management, with limited data regarding the efficacy of biological or synthetic Disease-Modifying Anti-Rheumatic Drugs on disease progression.57 Glandular and extraglandular manifestations are associated with both periepithelial immune infiltration and B-cell hyperactivation.8 While mechanisms of local glandular inflammation and damage are increasingly understood, data on systemic disease evolution, organ involvement and long-term damage accrual remain scarce.911 Although oral damage is widely recognised, end-organ damage is still considered rare.12 To date, two different damage indices have been proposed for SD: the Sjögren’s Syndrome Disease Damage Index (SSDDI) and the Sjögren’s Syndrome Damage Index (SSDI).13 14 The SSDDI is a global score including both exocrine and non-exocrine features. It was derived from a short-term study, and its sensitivity to change has not been evaluated.13 The SSDI, developed by the Systemic Lupus Collaborating Clinics, was reviewed by experts in rheumatology, oral medicine and ophthalmology. It includes three domains: ocular, oral and systemic and is sensitive to changes over 12 months.14 Damage assessment is considered a key point in the management of other autoimmune systemic diseases, that is, systemic lupus erythematosus (SLE): in fact, in patients with SLE, the evaluation of organ damage is routinely assessed and is considered an important measurement of disease burden. Damage denotes irreversible decline or loss of function because of the inflammatory activity of the disease, as well as treatment toxicity, comorbidities, ageing and/or incidental events.15 Thus, preventing damage accrual is a recognised therapeutic goal in SLE16 17 because the cumulative damage may have profound consequences, leading to functional impairment, chronic disability and reduced QOL. In SD, despite the availability of damage scores, the assessment of damage has received inexplicably less attention than the assessment of clinical activity, in both clinical and research settings.1821 However, few studies have suggested that irreversible systemic damage may accumulate over time in patients with SD, even in the absence of persistently high disease activity, highlighting that long-term organ damage may be under-recognised.22 23 Indeed, while recent longitudinal studies described damage evolution and its relationship with phenotypic heterogeneity in SD, these approaches were not specifically designed for identifying patients potentially at risk to develop systemic damage on the bases of clinical and serological variables at the disease onset. In this setting, both these approaches should be regarded as complementary rather than competing strategies, addressing different questions such as the weight of the clinical phenotype in damage occurrence and on the other hand the weight of specific clinical variables during the natural history of the disease. On these bases, this study aims to evaluate long-term damage accrual in a large, multicentre Italian cohort of patients with SD, as well as to explore the variables associated with the presence of damage at baseline and its development over time. Materials and methods Participants The Gruppo Italiano di Ricerca in Reumatologia Clinica e Sperimentale (GIRRCS) Sjögren’s syndrome cohort is an inception cohort including 448 consecutive adult patients (≥18 years) diagnosed with primary SD according to the American–European Consensus Criteria24 and followed in tertiary rheumatology centres experienced in SD management. For patients no longer in follow-up (due to loss or death), anonymised data were used in compliance with Italian privacy regulations (Garante per la protezione dei dati personali, Authorization n. 9/2016). Design Retrospective cohort study. Data collection Clinical records were reviewed to collect demographic, clinical and laboratory data. Baseline was defined as the first visit to the participating centres. Diagnostic delay was calculated as the time (in years) between symptom onset and diagnosis.25 Patients with missing data on key variables were excluded. Disease activity, damage and patient-reported symptoms were assessed using the European League Against Rheumatism (EULAR) Sjögren’s Syndrome Disease Activity Index (ESSDAI),24 26 SSDI14 and EULAR Sjogren’s Syndrome Patient Reported Index (ESSPRI),25 27 respectively. Outcomes The primary outcome was to describe systemic, ocular and oral damage accrual over 5 years. Secondary analyses explored baseline and disease-related variables associated with damage progression. Statistical analysis Continuous variables were summarised as mean±SD or median and IQR, according to data distribution assessed by the Shapiro-Wilk test. Categorical variables were reported as frequencies and percentages. Associations between clinical variables and damage accrual were first explored using univariate analysis. Damage-free survival was estimated using Kaplan-Meier curves and compared using the log-rank test. Predictors of damage accrual were then evaluated using Cox proportional hazards regression models. The proportional hazards assumption was assessed for all covariates. Missing data in the primary analysis were handled using multiple imputation by chained equations, applying logistic regression for categorical variables and predictive mean matching for continuous variables. To further explore the contribution of ESSDAI-specific domains to damage accrual, an exploratory Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression analysis was performed. In this model, each ESSDAI domain was included as a continuous variable reflecting its severity, together with sex and age at diagnosis. Missing values in covariates for this exploratory analysis were handled using median imputation. The optimal penalty parameter (λ) was selected through 10-fold cross-validation. Domains with coefficients shrunk to zero were considered non-predictive, whereas domains with non-zero coefficients were interpreted as contributing to the risk of damage accrual. HRs were obtained by exponentiating the estimated coefficients. Principal component analysis (PCA) was applied solely for exploratory purposes to provide an unsupervised description of the multivariate structure of clinical and laboratory data. All analyses were conducted using R software (V.3.0.3; R Foundation for Statistical Computing, Vienna, Austria), employing the survival, mice, survminer, glmnet, ggplot2 and ggradar packages. Results Patients’ characteristics Overall, 429 patients out of the total patients of the GIRRCS Sjögren’s syndrome inception cohort with at least 1 year of follow-up were included in the analysis. Among them, 140 patients had at least 2 years of follow-up, 118 had at least 3 years, 109 had at least 4 years and 96 had at least 5 years of follow-up. The patients’ median age was 52 years and 95% of them were female. The ESSDAI at the baseline was 4.52 (± 5.5) and 71% of participants had a low disease activity. The baseline characteristics of the cohort are summarised in online supplemental table 1. Rate of damage development and accrual At baseline, the mean diagnostic delay was 3 years (±4.80) and the mean SSDI was 0.58 (±0.90); a total of 199 patients (47.6%) had an SSDI≥1 (figure 1a,b). The presence of damage was reported in 25.4% of participants in the ocular domain, in 14.3% of participants in the oral domain and in 22.5% in the systemic domain (figure 1c). Within the systemic domain, 6.5% of participants showed musculoskeletal damage, 5.5% showed damage in the pulmonary domain, 4.8% in the neurological domain and 3.8% in the renal domain. Cardiovascular and gastro-intestinal (GI) damage were reported in 1% of the participants. At baseline, 6.5% of participants had a history of malignancy (figure 1d). During follow-up, we observed an increase in the percentage of participants with SSDI≥1. After 5 years, 65% of the participants showed damage, specifically, 34.2% in the ocular domain, 21.7% in the oral domain and 40.8% in the systemic domain. At this time, 11% out of the participants showed damage in the musculoskeletal domain, 6% in the pulmonary domain, 9.1% in the neurological domain, 6.42% in the renal domain, 2.7% in the cardiovascular domain, 1.8% in the GI domain and 9.6% had a history of malignancy (figure 1d). We observed a significant increase in mean SSDI, when compared with the baseline after 5 years of follow-up (1.06±1.55, p<0.0001), figure 1b. (a) Histogram showing the percentage of Sjögren’s disease (SD) patients who developed damage at each timepoint during the study. T0: baseline; T1: 1 year; T2: 2 years; T3: 3 years; T4: 4 years and T5: 5 years. (b) Histogram showing the mean Sjögren’s Syndrome Damage Index (SSDI) at each timepoint during the study. T0: baseline; T1: 1 year; T2: 2 years; T3: 3 years; T4: 4 years and T5: 5 years. (c) Radar plot illustrating the percentage of patients with SD presenting ocular, oral and systemic damage over time. Orange line represents baseline (T0); blue, 1 year (T1); pink, 3 years (T3) and yellow, 5 years (T5). (d) Radar plot illustrating the percentage of patients with SD presenting damage in each systemic domain over time. Orange line represents baseline (T1); blue, 1 year (T1); pink, 3 years (T3) and yellow, 5 years (T5). Factors associated with higher SSDI at baseline At baseline, as reported in table 1, ocular damage was associated with positive Schirmer’s test (ST) (r=0.20, p<0.0001), higher C reactive protein (CRP) value (r=0.16, p=0.0017), higher ESSDAI value (r=0.14, p=0.0033) and the presence of anti-La antibodies (r=0.10, p=0.044). Oral damage was associated with higher ESSDAI value (r=0.26, p<0.0001), rheumatoid factor (RF) presence (r=0.20, p<0.0001), higher CRP value (r=0.18, p=0.0008), higher focus score (FS) (r=0.18, p=0.0125), recurrent oral infections (r=0.16, p=0.0011), higher ESSPRI value (r=0.15, p=0.011), presence of anti-Ro antibodies (r=0.14, p=0.0039), xerostomia (r=0.11, p=0.028) and xerophthalmia (r=0.11, p=0.022) (table 1). Systemic damage was associated with higher ESSDAI value (r=0.61, p<0.0001), recurrent oral infections (r=0.17, p=0.0006), higher ESSPRI value (r=0.15, p=0.0079), higher CRP value (r=0.14, p=0.011) and the presence of anti-Ro antibodies (r=0.10, p=0.036). (table 1). At baseline, PCA was performed to explore multivariate relationships between baseline features and the presence of damage (online supplemental figure 1). The first two principal components (PC1 and PC2) explained a substantial proportion of the variance in the dataset and were associated with the type of damage. PC1 was primarily influenced by patient-reported symptoms and glandular function measures, with sicca symptoms (26.7%), ESSPRI (25.1%), FS (19.6%) and ST (19.4%) contributing most strongly, representing the subjective symptom burden and glandular dysfunction domain. In contrast, PC2 was mainly driven by markers of systemic disease activity, including ESSDAI (38.9%), autoantibodies (25.8%), age at diagnosis (16.2%) and CRP (13.5%) (online supplemental table 3). Table 1Pearson’s correlations between ocular damage, systemic damage, oral damage and patients’ characteristics at baselineOcular damageSystemic damageOral damager95% CIP valuer95% CIP valuer95% CIP valueAge at diagnosis0.061−0.034 to 0.160.210.017−0.73 to 0.110.73−0.056−0.15 to 0.0400.25Smoking habitus0.032−0.080 to 0.140.580.0068−0.10 to 0.120.90−0.10−0.21 to 0.0090.071Diagnostic delay−0.034−0.13 to 0.0630.49−0.056−0.15 to 0.0410.26−0.0053−0.10 to 0.0910.91Xerophthalmia−0.021−0.12 to 0.0740.66−0.065−0.16 to 0.0300.180.110.016 to 0.200.022Xerostomia−0.030−0.12 to 0.0650.54−0.048−0.14 to 0.0470.320.110.012 to 0.200.028Recurrent oral infections0.093−0.0039 to 0.190.0600.170.073 to 0.260.00060.160.065 to 0.250.0011ESSDAI0.140.047 to 0.230.00330.600.54 to 0.66< 0.00010.260.16 to 0.34<0.0001ESSPRI0.094−0.020 to 0.210.110.150.041 to 0.260.00790.150.034 to 0.260.011Focus score−0.038−0.18 to 0.100.590.0089−0.13 to 0.150.900.180.039 to 0.310.0125Schirmer’s test I0.200.10 to 0.30<0.0001−0.062−0.16 to 0.0410.240.038−0.065 to 0.140.47Gammaglobulins (mg/dL)0.015−0.10 to 0.130.800.026−0.093 to 0.140.660.044−0.076 to 0.160.47Cryoglobulins0.075−0.060 to 0.210.280.12−0.012 to 0.250.0740.12−0.012 to 0.250.075CRP0.160.062 to 0.260.00170.140.031 to 0.240.0110.180.074 to 0.280.0008RF0.040−0.062 to 0.140.440.057−0.045 to 0.160.270.200.10 to 0.30<0.0001ANA0.0054−0.091 to 0.100.910.031−0.066 to 0.130.530.079−0.017 to 0.170.11Anti-Ro0.070−0.028 to 0.160.160.100.0070 to 0.200.0360.140.046 to 0.240.0039Anti-La0.100.0024 to 0.200.0440.091−0.0076 to 0.180.0700.074−0.024 to 0.170.14ANA, anti-nuclear antibody; CRP, C reactive protein; CV, cardiovascular; ESSDAI, European Alliance of Associations for Rheumatology Sjögren’s Syndrome Disease Activity Index; ESSPRI, European Alliance of Associations for Rheumatology Sjogren’s Syndrome Patient Reported Index; GI, gastro-intestinal; MSK, musculoskeletal; RF, rheumatoid factor. Pearson’s correlations between ocular damage, systemic damage, oral damage and patients’ characteristics at baseline ANA, anti-nuclear antibody; CRP, C reactive protein; CV, cardiovascular; ESSDAI, European Alliance of Associations for Rheumatology Sjögren’s Syndrome Disease Activity Index; ESSPRI, European Alliance of Associations for Rheumatology Sjogren’s Syndrome Patient Reported Index; GI, gastro-intestinal; MSK, musculoskeletal; RF, rheumatoid factor. Damage accrual after 5 years of follow-up and damage-free survival After 5 years of follow-up, damage-free survival rates were 68.6% for ocular damage, 84.3% for oral damage and 60.9% for systemic damage. Overall, 118 participants developed ocular damage, 62 participants developed oral damage and 115 participants developed systemic damage during the observation period (figure 2a–c). Multivariable Cox regression analyses showed that positive ST results and higher ESSDAI scores at the baseline were significantly associated with an increased risk of ocular damage accrual, and in this setting ST showed the strongest association (HR 3.58; 95% CI 1.536 to 8.355; p=0.003), followed by ESSDAI (HR 1.029; 95% CI 1.00041 to 1.059; p=0.0468) (table 2). With regard to oral damage, higher baseline ESSPRI (HR: 1.229; 95% CI 1.078 to 1.401; p=0.00294) and ESSDAI (HR: 1.037; 95% CI 1.00080 to 1.074; p=0.0452) were both independently associated with an increased risk of damage accrual (table 2). As far as systemic damage is concerned, higher ESSSDAI values at baseline emerged as the only significant predictor of damage accrual (HR: 1.11; 95% CI 1.081 to 1.14; p<0.0001) (table 2). Higher ESSDAI scores were significantly associated with an increased risk of damage accrual. In the adjusted Cox models, patients with moderate and high disease activity showed a more than threefold and eightfold higher risk of damage, respectively, compared with those with low disease activity. When analysed as a continuous variable, ESSDAI remained an independent predictor of damage accrual, with a 10% increase in risk per 1-point increase (online supplemental figure 2 and online supplemental tables 3 and 4). Kaplan-Meier curves showing the estimated damage-free survival probability in patients with with Sjögren’s disease, referring to: (a) ocular damage, (b) oral damage and (c) systemic damage during the study period. Table 2Multivariate Cox proportional hazards model identifying predictors of ocular, systemic and oral damage in patients with Sjögren’s diseaseOcular damageSystemic damageOral damageHR25–95% CIP valueHR25–95% CIP valueHR25–95% CIP valueGender0.9040.412 to 1.9840.800.2920.084 to 1.020.3970.9140.263 to 3.1720.885Age at diagnosis1.0100.995 to 1.0260.1841.00030.985 to 1.0160.9720.9940.972 to 1.020.55ESSDAI1.0291.00041 to 1.0590.04681.111.081 to 1.14p<0.00011.0371.00 to 1.0740.0452ESSPRI–––1.0550.928 to 1.20.3771.2291.078 to 1.4010.00294Xerophthalmia––––––3.2950.72 to 15.1390.122Xerostomia––––––1.1530.603 to 4.2050.826Recurrent oral infections––––––1.9150.973 to 3.7700.0596Focus score––––––1.0390.822 to 1.3140.731Schirmers’test3.5821.536 to 8.3550.00352––––––CRP1.00770.997 to 1.0180.1491.0081 to 1.0170.06151.00860.996 to 1.0210.185Anti-Ro–––1.0380.611 to 1.7640.92.2090.875 to 5.5740.0916Anti-La1.2370.833 to 1.8380.288––––––Rheumatoid factor––––––1.9990.996 to 1.0220.018CCS usage in the 1 year1.0340.693 to 1.5430.8691.3440.84 to 2.150.2151.1050.603 to 2.0240.741CCS, corticosteroid; CRP, C reactive protein; ESSDAI, European Alliance of Associations for Rheumatology Sjögren’s Syndrome Disease Activity Index; ESSPRI, European Alliance of Associations for Rheumatology Sjogren’s Syndrome Patient Reported Index. Multivariate Cox proportional hazards model identifying predictors of ocular, systemic and oral damage in patients with Sjögren’s disease CCS, corticosteroid; CRP, C reactive protein; ESSDAI, European Alliance of Associations for Rheumatology Sjögren’s Syndrome Disease Activity Index; ESSPRI, European Alliance of Associations for Rheumatology Sjogren’s Syndrome Patient Reported Index. To investigate the contribution of individual ESSDAI domains as predictors of systemic damage accrual, we performed an exploratory LASSO Cox regression. The analysis identified the glandular (HR 1.26), peripheral nervous system (HR 1.19) and pulmonary (HR 1.14) domains as the strongest independent predictors. Other domains, including articular, lymphadenopathy/lymphoma, constitutional, renal, muscular, haematological and biological domains, showed smaller non-zero effects (HR 1.02–1.11), whereas cutaneous and central nervous system domains, as well as gender and age at diagnosis, were shrunk to zero and did not contribute to the prediction of systemic damage (figure 3). Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression identifying European Alliance of Associations for Rheumatology Sjögren’s Syndrome Disease Activity Index (ESSDAI) domains associated with systemic damage accrual. (a) LASSO coefficient path plot showing how the coefficients of individual ESSDAI domains change across values of log(λ). Lines are coloured by domain. The vertical dashed line indicates lambda.min, where the optimal set of predictors is selected. Only domains with non-zero coefficients at lambda.min contribute to the prediction of systemic damage. (b) Table of HR for ESSDAI domains retained in the model at lambda.min. Glandular, peripheral nervous system and pulmonary domains were the strongest predictors, whereas cutaneous and central nervous system domains, as well as gender and age at diagnosis, were shrunk to zero. HR were calculated as exp(beta) per 1 SD increase in each domain score. (c) Cross-validation plot of the LASSO Cox model showing the mean cross-validated partial likelihood deviance versus log(λ). The vertical dotted line indicates the value of λ that minimises the cross-validated error (lambda.min). Discussion This study provides comprehensive data on the incidence and the characteristics of damage accrual in a large multicentre Italian cohort of patients with primary SD. Of note, at the time of diagnosis, nearly half of the participants (48%) had already accrued irreversible damage, before starting any treatment. This observation strongly highlights the clinical relevance of early diagnosis in SD, as a substantial proportion of patients already present disease-related damage at diagnosis, potentially limiting therapeutic effectiveness and influencing long-term outcomes. In our cohort, the diagnostic delay was approximately 3 years, confirming the growing concern that delayed diagnosis in rheumatic diseases may adversely affect prognosis and disease burden.28 29 At diagnosis, the most observed damage was in the ocular domain, followed by systemic and oral domains. These data failed to confirm what was previously reported regarding the oral domain as the most important type of damage in SD,12 highlighting the importance of using SD specific damage scores. Our data suggest also that damage accrual in SD predominantly occurs within the first 3 years after diagnosis, across all domains. This early accumulation of irreversible damage may reflect higher inflammatory activity during the initial disease phases.30 Although the overall burden of damage in SD may appear lower than in other systemic autoimmune diseases such as SLE, this perception is likely influenced by the limitations of current SD-specific damage indices, which do not capture several treatment-related complications included in SLE damage scores, for example, osteoporosis and diabetes.31 32 The presence of systemic damage at diagnosis was significantly associated with high disease activity, as reflected by ESSDAI, together with high ESSPRI scores, increased CRP levels, recurrent oral infections and the presence of anti-Ro antibodies. Our data confirm the association between disease activity and damage, and mirror the data from Milic et al, who showed a significant association among pathologic salivary glands ultrasonography, higher ESSDAI values and damage in SD.33 The higher ESSDAI level and its association with damage confirm how, over time, the activity in some specific domain may induce damage and, consequently, any diagnostic delay allows damage to accrue prior to diagnosis. In fact, as reported, ESSDAI is higher in patients with early SD onset34 and correlates with disease duration.35 It is largely accepted that the ESSDAI Score and ESSPRI Score do not correlate36 and thus they provide complementary perspectives, both contributing, independently, to damage development. We observed a significant correlation between the presence of anti-Ro autoantibodies and systemic damage. Anti-Ro antibodies are associated with an earlier SD onset, salivary gland infiltration and dysfunction and a higher prevalence of extraglandular manifestations, such as lymphoma and interstitial lung disease, hypergammaglobulinemia as well as other features of B-cell activation.3639 In our opinion, more than a direct effect of autoantibodies in the pathogenesis of damage development, the specific clinical features of this subset of patients make them prone to a more severe disease and damage accrual. Ocular damage at baseline was associated with abnormal ST, higher ESSDAI, higher CRP levels and the anti-La antibodies positivity. Our findings confirm previously reported associations between dry eye severity, ST abnormalities and ocular damage.4042 Mild-to-moderate keratoconjunctivitis is associated with ocular irritation or visual disturbances. More severe diseases may result in corneal complications, including infectious keratitis, ulceration, scarring and subsequent loss of vision.4345 Anti-La antibodies have been linked to more severe glandular dysfunction and ocular involvement, potentially contributing to irreversible ocular damage.39 46 47 Taken together, all these results may explain the presence of ocular damage. Although CRP is generally considered an unreliable marker of disease activity in SD,48 elevated levels may reflect specific inflammatory manifestations, such as arthritis, which could indirectly contribute to damage development.49 50 As far as oral damage is concerned, at diagnosis, it was associated with higher ESSDAI scores, RF, increased CRP and higher FS. Notably, ESSDAI showed the strongest correlation, suggesting that systemic disease activity mirrors local glandular damage.9 50 51 ESSPRI does not influence damage development in the oral domain, and these data may reflect its low correlation with the salivary gland function. As reported, the ESSPRI dryness item correlates strongly with subjective xerostomia but showed a very weak or no correlation with objective measures, namely the stimulated and the unstimulated salivary flow.52 53 Surprisingly, no relationship between the presence of specific autoantibodies and the presence of oral damage was observed. During the 5-year follow-up, higher baseline ESSDAI values were consistently associated with the development of ocular, oral and systemic damage, supporting a role of disease activity in damage accrual over time, in line with previous studies.23 An association was observed between baseline disease activity and subsequent damage risk. In the adjusted Cox models, patients with moderate and high disease activity showed a more than threefold and eightfold higher risk of damage accrual, respectively, compared with those with low disease activity. Consistently, ESSDAI remained an independent predictor of damage accrual when analysed as a continuous variable, with a 10% increase in risk for each additional point. Overall, these findings suggest that higher disease activity at baseline is associated with an increased risk of irreversible damage over time, highlighting the potential importance of early disease control in patients with SD. In our study, exploratory PCA highlighted a clear separation between subjective symptoms and objective systemic disease activity measures, with ESSPRI and sicca symptoms loading independently from ESSDAI and serological markers, using stratification by type of damage. This finding, consistent with a previous report,54 reinforces the concept that patient-reported outcomes and systemic inflammatory activity represent distinct but complementary dimensions of SD and both should be considered when assessing disease activity and predicting damage accrual. We are aware that our study has some weaknesses, such as the retrospective design, the similar genetic background of the cohorts (only Caucasian participants) and the limited number of participants, thus suggesting the need to design similar studies enrolling more patients, possibly performing the evaluation in participants with different genetic backgrounds, as demonstrated by the results of the Big Data Sjogren Consortium, which showed the importance of genetic differences for some specific SD features.55 Notably, some authors of this paper are already members of the Big Data Sjogren Consortium and have already applied for a research question to assess the progression of damage in the largest cohort worldwide. However, our study also has some strengths. First, this is the largest multicentre cohort of well-characterised SD participants evaluating damage accrual, so far, a largely underestimated issue in this disease. We also identified variables that were significantly associated with the development of damage. All the participants were evaluated using standardised methodologies and protocols. Of note, all the centres in this study have a very long experience in SD treatment, resulting in an important homogeneity for the management of the patients, reducing the bias in patients’ selection, strengthening the quality of our results and making the data more reliable. Finally, unlike previous long-term studies exploring heterogeneous damage trajectories over extended follow-up in primary SD, our analysis was primarily focused on damage present at diagnosis and during the early disease course; however, when interpreted in the context of the existing literature, our data complement these observations by showing that irreversible damage is frequently already present at diagnosis and predominantly accrues within the first years after disease onset. This different methodological approach, already discussed in the Introduction, reflects a distinct study aim rather than a limitation, as longitudinal and early disease-focused analyses address complementary aspects of damage accrual in SD.23 In conclusion, this study highlights and confirms that damage accrual in SD is not rare, but rather a frequent and early phenomenon, often present at diagnosis. Preventing the evolution of irreversible damage requires prompt diagnosis and targeted strategies aimed at controlling disease activity, particularly in patients with higher ESSDAI scores. To do that, a very early identification of patients at greater risk, based on their own clinical and serological markers, may enable earlier interventions and ultimately improve long-term outcomes. Supplementary material