Abstract
Objective
To investigate the risk of incident cognitive disorders in patients with primary Sjögren’s disease (pSjD) using nationwide Korean National Health Insurance Service data.
Methods
This retrospective cohort study included patients newly diagnosed with pSjD between 2010 and 2021 and propensity score-matched controls. The primary outcome was a composite of Alzheimer’s dementia (AD), vascular dementia (VaD), secondary dementia, unclassified dementia and mild cognitive impairment (MCI). Adjusted HRs (aHRs) and 95% CIs were estimated using multivariable Cox regression. Age-stratified analyses were also performed.
Results
A total of 12 731 patients with pSjD and 36 514 matched controls were analysed. pSjD was associated with increased risks of total cognitive disorders (aHR 1.48, 95% CI 1.36 to 1.63), AD (aHR 1.48, 95% CI 1.35 to 1.63), secondary dementia (aHR 2.24, 95% CI 1.45 to 3.48), unclassified dementia (aHR 1.43, 95% CI 1.21 to 1.70) and MCI (aHR 1.62, 95% CI 1.41 to 1.85). In age-stratified analyses, pSjD was associated with an increased risk of MCI among participants aged <65 years (aHR 1.46, 95% CI 1.06 to 2.01). Among those aged 65–74 years, pSjD was associated with increased risks of total cognitive disorders, AD and MCI.
Conclusions
pSjD was associated with increased risks of multiple cognitive disorders. The age-specific patterns—an increased risk of MCI among individuals aged <65 years and increased risks of total cognitive disorders, AD and MCI among those aged 65–74 years—underscore the need for clinical attention to cognitive symptoms across age groups in patients with pSjD.
Keywords: Sjogren's Syndrome; Epidemiology; Outcome Assessment, Health Care
WHAT IS ALREADY KNOWN ON THIS TOPIC
Primary Sjögren’s disease (pSjD) can involve the central nervous system, and cognitive dysfunction has been reported in some patients; however, epidemiological studies on the association between pSjD and dementia have shown inconsistent results, and the overall relationship remains controversial.
WHAT THIS STUDY ADDS
In this nationwide cohort study, pSjD was associated with an increased risk of overall cognitive disorders, Alzheimer’s dementia, secondary dementia, unclassified dementia and mild cognitive impairment, even after multivariable adjustment.
Age-stratified analyses demonstrated different patterns of association: pSjD was associated with an increased risk of mild cognitive impairment among individuals aged <65 years and with increased risks of total cognitive disorders, Alzheimer’s dementia and mild cognitive impairment among those aged 65–74 years.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
These findings raise awareness of cognitive morbidity across age groups in pSjD and support clinical attention to cognitive symptoms, including mild cognitive impairment before the age of 65 years and a broader spectrum of cognitive disorders among patients aged 65–74 years.
The age-specific patterns provide a basis for future studies to clarify the mechanisms underlying cognitive disorders in pSjD and to evaluate age-appropriate approaches to cognitive assessment and risk stratification.
Introduction
Primary Sjögren’s disease (pSjD) is a systemic autoimmune disease, and dysfunction of exocrine glands, particularly the lacrimal and salivary glands, is a hallmark manifestation.1 Lymphocyte infiltration into the exocrine glands and the presence of specific autoantibodies (anti-SSA/Ro) are key features distinguishing pSjD from other diseases that cause sicca symptoms.2 3 Treatment of pSjD is primarily based on symptomatic relief and attenuation of various systemic manifestations.4 pSjD can cause diverse manifestations depending on the organs involved, and major organ involvement often requires immunosuppressants and high-dose glucocorticoid therapy.4
Central nervous system (CNS) involvement in pSjD has a wide spectrum, ranging from headache and mood disorders to transverse myelitis and encephalitis.5 Cognitive impairment and ‘brain fog’ are also reported in a subset of patients with pSjD. Several epidemiological studies have reported an increased risk of dementia among patients with pSjD.6–9 In contrast, one study from South Korea reported that pSjD was not associated with dementia incidence,10 and a Mendelian randomisation study also found no association between pSjD and dementia.11 Therefore, the association between pSjD and cognitive disorders remains controversial.
The National Health Insurance Service (NHIS) is a single-payer, government-funded healthcare system in South Korea that covers almost the entire South Korean population.12 The NHIS database comprehensively collects healthcare information, including diagnoses, prescriptions and mortality data.12 Biannual health screening examinations are provided to individuals aged ≥20 years, and the results are integrated into the NHIS dataset. Thus, NHIS data constitute a longitudinal, nationwide, health-related dataset that can be used for diverse epidemiological and health policy studies.
In the present study, we evaluated the incidence and HRs of cognitive disorders—including Alzheimer’s dementia (AD), vascular dementia (VaD), secondary dementia, unclassified dementia and mild cognitive impairment (MCI)—in patients with pSjD using Korean NHIS data.
Materials and methods
Source of data
Korean NHIS data were used in the present study. The NHIS is a single-payer, government-funded medical insurance system established in 1963 that covers nearly the entire population of South Korea.12 Approximately 97% of South Korean residents are mandatorily enrolled in the NHIS and pay monthly health insurance premiums, while the remaining low-income population is covered through the Medical Aid programme without premium contributions. Healthcare reimbursement claims are collected and stored in the National Health Information Database (NHID). Since 2002, the NHID has compiled comprehensive nationwide data on healthcare utilisation, health screening examinations, socioeconomic and demographic factors and mortality information for the entire South Korean population.13 Claims data include diagnostic codes, procedures and prescription records, with diseases coded according to the International Classification of Diseases, 10th Revision (ICD-10) and medications categorised using Korean Drug and Anatomical Therapeutic Chemical codes. In addition, the South Korean government provides medical cost reduction for designated intractable rare diseases, which are strictly regulated; pSjD is included in this programme (code: V139). After approval by the NHIS review committee, anonymised data on patients with pSjD and healthy controls were provided from the NHIS database (approval no. KDCA-NHIS-2025-03-012).
Study population
The present study was designed as a retrospective cohort study. Patients with pSjD were defined as individuals who had an ICD-10 code for pSjD (M35.0) as their primary or secondary diagnosis, in combination with the intractable rare disease registration code (V139), during either inpatient or outpatient visits between January 2010 and December 2021. To identify newly diagnosed cases, individuals with a diagnosis code for pSjD (M35.0) during the washout period from 2008 to 2009 were excluded. Additional exclusion criteria were as follows:
Concomitant diagnosis of other autoimmune diseases, including rheumatoid arthritis, systemic lupus erythematosus, systemic sclerosis, mixed connective tissue disease, dermatomyositis, polymyositis and antiphospholipid syndrome.
Missing information on age or sex at the index date.
A history of cognitive disorders during 2008–2009.
The index date was defined as the date of the first recorded diagnosis of pSjD (M35.0). Healthy controls (HCs) were randomly selected from individuals without a diagnosis of pSjD at an initial ratio of 1:3. Subsequently, propensity score (PS) matching by age, sex, diabetes mellitus, hypertension, dyslipidaemia, ischaemic heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, chronic kidney disease and depression was performed, and controls were matched to patients with pSjD at a 1:3 ratio for the final analysis population. The ICD-10 codes used in the present study are summarised in online supplemental table 1 and a flow chart of patient selection is presented in online supplemental figure 1. Information on comorbidities, height, weight, smoking status, alcohol consumption and physical activity was extracted from the NHIS health screening database at the start of follow-up. All subjects were followed until 31 December 2023, until the first occurrence of an outcome event (AD, VaD, secondary dementia, unclassified dementia or MCI), or until death, whichever occurred first.
Outcome definition (Alzheimer’s dementia, vascular dementia, secondary dementia, unclassified dementia and mild cognitive impairment)
Cognitive disorders were operationally defined based on a combination of ICD-10 diagnostic codes and prescription of cognitive-enhancing medications, specifically acetylcholinesterase inhibitors (donepezil, rivastigmine or galantamine). The ICD-10 code for each cognitive disorder was as follows: AD (F00, G30), VaD (F01), secondary dementia (F02), unclassified dementia (F03) and MCI (F06.7). According to the Korean NHIS reimbursement criteria, prescription of acetylcholinesterase inhibitors requires documentation that patients meet predefined eligibility criteria based on standardised assessments of cognitive function and dementia severity. As a sensitivity analysis, cognitive disorders were alternatively defined using ICD-10 diagnostic codes alone, without requiring concomitant prescription of cognitive-enhancing medications. The associations between pSjD and cognitive disorders were reassessed using this alternative outcome definition to evaluate the robustness of the findings.
Covariates
Age was categorised into four groups: <65, 65–74, 75–84 and ≥85 years. Income level was classified into low, middle and high tiers based on health insurance premium levels, which reflect annual income. Body mass index (BMI) was classified as underweight (BMI <18.5 kg/m2), normal (BMI 18.5–22.9 kg/m2), overweight (BMI 23.0–24.9 kg/m2), obesity class I (BMI 25.0–29.9 kg/m2), obesity class II (BMI 30.0–34.9 kg/m2) and obesity class III (BMI ≥35.0 kg/m2) according to the Korean Society for the Study of Obesity.14 Smoking status was classified as people who have never smoked, people who previously smoked or people who smoke. Alcohol consumption was categorised by drinking frequency per week as: none, 1–2 days, 3–4 days or ≥5 days. Information on sufficient aerobic activity, hearing impairment and visual impairment was obtained from self-reported health screening questionnaires. Sufficient aerobic activity was defined as achieving ≥500 metabolic equivalent-minutes per week. Comorbidities were identified using ICD-10 codes, including: diabetes mellitus (E10–E14), hypertension (I10–I13, I15), dyslipidaemia (E78), ischaemic heart disease (I20–I25), cerebrovascular disease (I60–I69), chronic obstructive pulmonary disease (J43–J44), chronic kidney disease (N18–N19) and depression (F32–F33). The Charlson Comorbidity Index (CCI) was calculated from claims data and the number of healthcare visits was included as an indicator of healthcare utilisation frequency.
Statistical analysis and data management
Categorical variables were compared using the χ2 test and are presented as frequencies and percentages. Continuous variables are expressed as mean±SD. Cumulative incidence of cognitive disorders was assessed using Kaplan-Meier (KM) curves and compared using the log-rank test. Incidence rates (IRs) were calculated as the number of events per 100 000 person-years (PYs). The HC group served as the reference for calculating incidence rate ratios (IRRs). Cox proportional hazards regression models were used to estimate both unadjusted and adjusted HRs with 95% CIs for each cognitive outcome. Multivariable models adjusted for age, sex, income level, BMI, smoking and alcohol status, physical activity, hearing or visual impairment, diabetes mellitus, hypertension, dyslipidaemia, ischaemic heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, chronic kidney disease, depression, CCI and the number of healthcare visits. Interaction terms between pSjD and age, sex, income level, BMI, sufficient aerobic activity, hearing impairment and visual impairment were evaluated in multivariable Cox models. Given the possibility of competing risk bias arising from differential mortality, we conducted a sensitivity analysis using the Fine-Gray subdistribution hazards model to estimate subdistribution HRs, treating death as a competing event. A two-tailed p value <0.05 was considered statistically significant. All statistical analyses were performed using SAS Enterprise, V.7.3 (SAS Institute, Cary, North Carolina, USA).
Results
Baseline demographics of pSjD and healthy control groups
A total of 12 731 patients with pSjD and 36 514 PS-matched HCs were included in the present study. The mean age was 55.9 years in the pSjD group and 56.1 years in the HC group, and females accounted for 91.8% and 93.1% of each group, respectively. Although age, sex and comorbidities remained statistically different after PS matching due to the large sample size, the standardised mean differences were all below 0.1, indicating adequate balance between the matched groups. The CCI and number of healthcare visits were significantly higher in patients with pSjD. Detailed baseline demographic and clinical characteristics are summarised in table 1.
Table 1. Baseline characteristics of patients with primary Sjögren’s disease and healthy controls.
| Primary Sjögren’s disease | Control | P value | SMD | |
|---|---|---|---|---|
| (n=12 731) | (n=36 514) | |||
| Age (years, mean (SD)) | 55.88 (13.48) | 56.10 (13.12) | <0.0001 | |
| Age (years, median (IQR)) | 56 (48–65) | 56 (47–65) | <0.0001 | |
| Age (years) | 0.041 | −0.031 | ||
| <65 | 9463 (74.33%) | 27 149 (74.35%) | ||
| 65–74 | 2155 (16.93%) | 6336 (17.35%) | ||
| 75–84 | 997 (7.83%) | 2780 (7.61%) | ||
| ≥85 | 116 (0.91%) | 249 (0.68%) | ||
| Sex | <0.001 | 0.029 | ||
| Male | 1047 (8.22%) | 2510 (6.87) | ||
| Female | 11 684 (91.78%) | 34 004 (93.13) | ||
| Income | <0.001 | |||
| High | 5884 (46.22%) | 15 296 (41.89%) | ||
| Middle | 3121 (24.51%) | 10 312 (28.24%) | ||
| Low | 3726 (29.27%) | 10 906 (29.87%) | ||
| BMI* | <0.001 | |||
| Class III obesity | 43 (0.34%) | 224 (0.61%) | ||
| Class II obesity | 276 (2.17%) | 1494 (4.09%) | ||
| Class I obesity | 2487 (19.53%) | 9414 (25.78%) | ||
| Overweight | 2518 (19.78%) | 7666 (20.99%) | ||
| Normal | 5788 (45.46%) | 12 958 (35.49%) | ||
| Underweight | 730 (5.73%) | 1178 (3.23%) | ||
| Unknown | 889 (6.98%) | 3580 (9.80%) | ||
| Smoking status | <0.001 | |||
| People who smoke | 306 (2.40%) | 1397 (3.83%) | ||
| People who previously smoked | 749 (5.88%) | 2054 (5.63%) | ||
| People who have never smoked | 10 769 (84.59%) | 29 402 (80.52%) | ||
| Unknown | 907 (7.12%) | 3661 (10.03%) | ||
| Alcohol use (days/week) | <0.001 | |||
| ≥5 | 99 (0.78%) | 470 (1.29%) | ||
| 3~4 | 276 (2.17%) | 1187 (3.25%) | ||
| 1~2 | 1539 (12.09%) | 5499 (15.06%) | ||
| <1 | 9808 (77.04%) | 25 227 (69.09%) | ||
| Unknown | 1009 (7.93%) | 4131 (11.31%) | ||
| Sufficient aerobic activity | <0.001 | |||
| Yes | 1950 (15.32%) | 5656 (15.49%) | ||
| No | 9896 (77.73%) | 27 295 (74.75%) | ||
| Unknown | 885 (6.95%) | 3563 (9.76%) | ||
| Hearing impairment | <0.001 | |||
| Yes | 244 (1.92%) | 516 (1.41%) | ||
| No | 12 487 (98.08%) | 35 998 (98.59%) | ||
| Visual impairment | 0.183 | |||
| Yes | 136 (1.07%) | 341 (0.93%) | ||
| No | 12 595 (98.93%) | 36 173 (99.07%) | ||
| Comorbidities | ||||
| DM | 2531 (19.88%) | 6949 (19.03%) | 0.036 | −0.007 |
| HTN | 3658 (28.73%) | 10 541 (28.87%) | 0.772 | −0.017 |
| Dyslipidaemia | 5644 (44.33%) | 15 562 (42.62%) | 0.001 | −0.001 |
| IHD | 1161 (9.12%) | 3094 (8.47%) | 0.026 | −0.014 |
| CVD | 1207 (9.48%) | 3009 (8.24%) | <0.001 | 0.011 |
| COPD | 431 (3.39%) | 853 (2.34%) | <0.001 | 0.007 |
| CKD | 221 (1.74%) | 466 (1.28%) | <0.001 | −0.001 |
| Depression | 1906 (14.97%) | 5012 (13.73%) | 0.001 | −0.027 |
| Charlson Comorbidity Index | <0.001 | |||
| 0 | 2942 (23.11%) | 15 040 (41.19%) | ||
| 1~2 | 5917 (46.48%) | 13 980 (38.29%) | ||
| 3 over | 3872 (30.41%) | 7494 (20.52%) | ||
| Number of healthcare visits, mean (SD) | 227.24 (206.64) | 161.15 (182.31) | <0.001 |
BMI was categorised according to Korean criteria: underweight (<18.5 kg/m2), normal weight (18.5–22.9 kg/m2), overweight (23.0–24.9 kg/m2), class I obesity (25.0–29.9 kg/m2), class II obesity (30.0–34.9 kg/m2) and class III obesity (≥35.0 kg/m2).
BMI, body mass index; CKD, chronic kidney disease; COPD, chronicobstructive pulmonary disease; CVD, cerebrovascular disease; DM, diabetes mellitus; HTN, hypertension; IHD, ischaemic heart disease; SMD, standardised mean difference.
Incidence of cognitive disorders in pSjD and healthy controls
The IR of total cognitive disorders (a composite of AD, VaD, secondary dementia, unclassified dementia and MCI) was 799.7 per 100 000 PYs in the pSjD group and 603.6 per 100 000 PYs in the HC group. KM analysis demonstrated a significantly higher cumulative incidence of total cognitive disorders in patients with pSjD compared with HCs (p<0.0001; figure 1A). In subgroup analyses by dementia subtype, the cumulative incidence of AD, secondary dementia, unclassified dementia and MCI was significantly higher in the pSjD group than in the HC group (figure 1B, D, E and F), whereas the cumulative incidence of VaD did not differ significantly between the two groups (figure 1C, p=0.07). The detailed IRs and IRRs are presented in online supplemental table 2.
Figure 1. Cumulative incidence of (A) total cognitive disorders, (B) Alzheimer’s dementia, (C) vascular dementia, (D) secondary dementia, (E) unclassified dementia and (F) mild cognitive impairment between patients with primary SjD and HCs. HC, healthy control; pSjD, primary Sjögren’s disease. Red line: HC, blue line: pSjD.
Comparison of hazards for cognitive disorders between the pSjD and healthy control groups
In unadjusted Cox regression analysis, patients with pSjD had a significantly higher risk of total cognitive disorder than HCs (HR 1.30, 95% CI 1.19 to 1.42). The unadjusted risks of AD (HR 1.32, 95% CI 1.20 to 1.44), secondary dementia (HR 2.33, 95% CI 1.53 to 3.56), unclassified dementia (HR 1.52, 95% CI 1.27 to 1.82) and MCI (HR 1.56, 95% CI 1.37 to 1.78) were also significantly increased, whereas the risk of VaD was not significantly different (HR 1.25, 95% CI 0.98 to 1.58; p=0.067). After multivariable adjustment, pSjD remained associated with increased risks of total cognitive disorders (adjusted HR (aHR) 1.48, 95% CI 1.36 to 1.63), AD (aHR 1.48, 95% CI 1.35 to 1.63), secondary dementia (aHR 2.24, 95% CI 1.45 to 3.48), unclassified dementia (aHR 1.43, 95% CI 1.21 to 1.70) and MCI (aHR 1.62, 95% CI 1.41 to 1.85). The association with VaD remained non-significant (aHR 1.24, 95% CI 0.97 to 1.58; p=0.084; table 2). Across the full cohort, univariable analyses showed that older age, male sex, hearing impairment, diabetes mellitus, hypertension, dyslipidaemia, ischaemic heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, depression and a CCI ≥3 were each associated with increased hazards of all assessed cognitive outcomes (online supplemental table 3). In multivariable analyses, older age, low income, hypertension, cerebrovascular disease, depression and a higher CCI were associated with increased hazards of total cognitive disorders and AD (online supplemental table 4). In a competing risk analysis accounting for death using the Fine-Gray model, the results remained largely unchanged and were consistent with those from the primary Cox regression analysis (table 2). These findings suggest that competing mortality was unlikely to have materially affected the observed associations. Significant interactions were observed for sex with VaD and MCI; income level with total cognitive disorders, AD, unclassified dementia and MCI and hearing impairment with total cognitive disorders, AD, unclassified dementia and MCI (online supplemental table 5). In the sensitivity analysis defining cognitive disorders using ICD-10 diagnostic codes alone, the findings were largely consistent with the primary analysis. Patients with pSjD had increased risks of total cognitive disorders (aHR 1.53, 95% CI 1.44 to 1.63), AD (aHR 1.39, 95% CI 1.28 to 1.51), VaD (aHR 1.32, 95% CI 1.07 to 1.61), secondary dementia (aHR 1.84, 95% CI 1.27 to 2.66), unclassified dementia (aHR 1.54, 95% CI 1.34 to 1.77) and MCI (aHR 1.57, 95% CI 1.45 to 1.70). Fine-Gray competing-risk analyses yielded similar results (online supplemental table 6).
Table 2. Univariable and multivariable Cox regression analysis of cognitive disorders between pSjD and HCs.
| HR (95% CI) | ||||||
|---|---|---|---|---|---|---|
| Univariable | P value | Multivariable* | P value | Fine-Gray analysis† | P value | |
| Total cognitive disorders | ||||||
| HCs | Reference | Reference | Reference | |||
| pSjD | 1.30 (1.19 to 1.42) | <0.001 | 1.48 (1.36 to 1.63) | <0.001 | 1.48 (1.35 to 1.64) | <0.001 |
| Alzheimer’s dementia | ||||||
| HCs | Reference | Reference | Reference | |||
| pSjD | 1.32 (1.20 to 1.44) | <0.001 | 1.48 (1.35 to 1.63) | <0.001 | 1.48 (1.34 to 1.64) | <0.001 |
| Vascular dementia | ||||||
| HCs | Reference | Reference | Reference | |||
| pSjD | 1.25 (0.98 to 1.58) | 0.067 | 1.24 (0.97 to 1.58) | 0.084 | 1.24 (0.97 to 1.59) | 0.093 |
| Secondary dementia | ||||||
| HCs | Reference | Reference | Reference | |||
| pSjD | 2.33 (1.53 to 3.56) | <0.001 | 2.24 (1.45 to 3.48) | <0.001 | 2.24 (1.40 to 3.59) | 0.001 |
| Unclassified dementia | ||||||
| HCs | Reference | Reference | Reference | |||
| pSjD | 1.52 (1.27 to 1.82) | <0.001 | 1.43 (1.21 to 1.70) | <0.001 | 1.52 (1.26 to 1.83) | <0.001 |
| Mild cognitive impairment | ||||||
| HCs | Reference | Reference | Reference | |||
| pSjD | 1.56 (1.37 to 1.78) | <0.001 | 1.62 (1.41 to 1.85) | <0.001 | 1.62 (1.40 to 1.86) | <0.001 |
Adjusted with age (<65, 65–74, 75–84, ≥85 years), sex, level of income, body mass index (underweight, normal, overweight, class I/II/III obesity), smoking/alcohol status, sufficient aerobic activity, hearing and visual impairment, comorbidities, Charlson Comorbidity Index and number of healthcare visit.
Fine-Gray hazards model to estimate subdistribution HRs, treating death as a competing event.
HC, healthy control; pSjD, primary Sjögren’s disease.
Subgroup analysis of pSjD and HC groups
Next, we evaluated risk factors for cognitive disorders (total, AD and MCI) separately within the pSjD and HC groups. In the pSjD group, univariable analyses showed that older age, male sex, hearing impairment and visual impairment were associated with increased hazards of total cognitive disorders, AD and MCI. Compared with normal BMI, overweight and class I obesity were associated with increased hazards of all three outcomes; class II obesity was associated with total cognitive disorders and AD, whereas class III obesity was associated only with MCI. Insufficient aerobic activity was associated with total cognitive disorders and AD, but not with MCI. In multivariable analyses, older age, low household income, hearing impairment and depression were associated with increased risks of total cognitive disorders, AD and MCI. Hypertension and cerebrovascular disease were associated with total cognitive disorders and AD, whereas dyslipidaemia was associated with total cognitive disorders and MCI; a higher CCI was associated with total cognitive disorders and AD (table 3). In HCs, older age, hypertension, depression and a higher CCI were associated with all three outcomes. Low income and cerebrovascular disease were associated with total cognitive disorders and AD, while dyslipidaemia was associated with MCI (online supplemental table 7).
Table 3. Univariable and multivariable Cox regression analyses for total cognitive disorders, Alzheimer’s dementia and mild cognitive impairment in patients with primary Sjögren’s disease.
| Total cognitive disorder | Alzheimer’s dementia | Mild cognitive impairment | ||||
|---|---|---|---|---|---|---|
| Univariable | Multivariable | Univariable | Multivariable | Univariable | Multivariable | |
| HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | |
| Age (years) | ||||||
| <65 | Reference | Reference | Reference | Reference | Reference | Reference |
| 65–74 | 14.06 (11.23 to 17.60) | 11.63 (9.12 to 14.83) | 15.21 (12.05 to 19.20) | 12.57 (9.77 to 16.17) | 13.60 (10.00 to 18.50) | 10.30 (7.38 to 14.39) |
| 75–84 | 41.70 (33.26 to 52.29) | 21.88 (16.81 to 28.47) | 47.23 (37.36 to 59.73) | 24.78 (18.85 to 32.57) | 30.76 (22.38 to 42.27) | 16.10 (11.11 to 23.34) |
| ≥85 | 62.80 (43.18 to 91.34) | 20.15 (13.26 to 30.61) | 71.45 (48.51 to 105.25) | 23.18 (15.05 to 35.72) | 28.26 (14.38 to 55.53) | 11.48 (5.58 to 23.61) |
| Sex | ||||||
| Male | Reference | Reference | Reference | Reference | Reference | Reference |
| Female | 0.56 (0.45 to 0.69) | 0.83 (0.63 to 1.11) | 0.55 (0.44 to 0.69) | 0.79 (0.60 to 1.05) | 0.51 (0.38 to 0.70) | 0.72 (0.48 to 1.07) |
| Income | ||||||
| High | Reference | Reference | Reference | Reference | Reference | Reference |
| Middle | 0.67 (0.54 to 0.83) | 0.94 (0.76 to 1.17) | 0.66 (0.53 to 0.82) | 0.94 (0.75 to 1.17) | 0.75 (0.55 to 1.01) | 1.11 (0.82 to 1.52) |
| Low | 1.46 (1.25 to 1.72) | 1.47 (1.25 to 1.73) | 1.46 (1.24 to 1.72) | 1.44 (1.22 to 1.70) | 1.52 (1.21 to 1.93) | 1.53 (1.21 to 1.95) |
| BMI* | ||||||
| Class III obesity | 2.55 (0.95 to 6.86) | 1.26 (0.47 to 3.41) | 2.62 (0.98 to 7.06) | 1.09 (0.40 to 2.95) | 3.83 (1.22 to 12.04) | 2.70 (0.85 to 8.57) |
| Class II obesity | 1.99 (1.26 to 3.15) | 1.10 (0.69 to 1.75) | 2.11 (1.33 to 3.33) | 1.18 (0.74 to 1.88) | 1.57 (0.77 to 3.21) | 0.81 (0.39 to 1.66) |
| Class I obesity | 1.82 (1.50 to 2.22) | 1.10 (0.90 to 1.34) | 1.82 (1.50 to 2.22) | 1.03 (0.84 to 1.27) | 1.74 (1.32 to 2.29) | 1.00 (0.75 to 1.33) |
| Overweight | 1.51 (1.24 to 1.86) | 1.14 (0.93 to 1.40) | 1.51 (1.23 to 1.86) | 1.11 (0.90 to 1.37) | 1.52 (1.14 to 2.02) | 1.11 (0.84 to 1.49) |
| Normal | Reference | Reference | Reference | Reference | Reference | Reference |
| Underweight | 1.06 (0.73 to 1.54) | 1.20 (0.82 to 1.75) | 1.02 (0.69 to 1.50) | 1.15 (0.78 to 1.70) | 1.00 (0.59 to 1.72) | 1.03 (0.60 to 1.76) |
| Smoking status | ||||||
| People who smoke | 0.92 (0.55 to 1.53) | 1.03 (0.60 to 1.77) | 0.94 (0.56 to 1.57) | 1.08 (0.63 to 1.86) | 0.97 (0.48 to 1.96) | 1.21 (0.58 to 2.54) |
| People who previously smoked | 1.55 (1.17 to 2.06) | 1.17 (0.82 to 1.66) | 1.54 (1.15 to 2.07) | 1.10 (0.77 to 1.57) | 1.64 (1.11 to 2.44) | 1.20 (0.73 to 1.97) |
| People who have never smoked | Reference | Reference | Reference | Reference | Reference | Reference |
| Alcohol use (days/week) | ||||||
| ≥5 | 0.24 (0.16 to 0.37) | 0.50 (0.32 to 0.77) | 0.25 (0.17 to 0.38) | 0.52 (0.34 to 0.81) | 0.27 (0.16 to 0.48) | 0.51 (0.29 to 0.91) |
| 3~4 | 0.31 (0.13 to 0.76) | 0.42 (0.17 to 1.02) | 0.33 (0.14 to 0.79) | 0.44 (0.18 to 1.07) | 0.13 (0.02 to 0.90) | 0.17 (0.02 to 1.20) |
| 1~2 | 1.06 (0.48 to 2.38) | 0.71 (0.31 to 1.64) | 1.11 (0.49 to 2.47) | 0.78 (0.34 to 1.80) | 0.35 (0.05 to 2.49) | 0.21 (0.03 to 1.52) |
| <1 | Reference | Reference | Reference | Reference | Reference | Reference |
| Sufficient aerobic activity | ||||||
| Yes | Reference | Reference | Reference | Reference | Reference | Reference |
| No | 1.39 (1.10 to 1.75) | 0.84 (0.67 to 1.07) | 1.49 (1.17 to 1.89) | 1.01 (0.79 to 1.28) | 1.27 (0.93 to 1.74) | 0.83 (0.61 to 1.15) |
| Hearing impairment | ||||||
| Yes | 4.77 (3.62 to 6.28) | 1.53 (1.16 to 2.03) | 4.63 (3.49 to 6.15) | 1.47 (1.10 to 1.97) | 4.91 (3.33 to 7.23) | 1.53 (1.02 to 2.29) |
| No | Reference | Reference | Reference | Reference | Reference | Reference |
| Visual impairment | ||||||
| Yes | 2.18 (1.33 to 3.58) | 1.29 (0.78 to 2.13) | 2.25 (1.37 to 3.69) | 1.33 (0.80 to 2.19) | 2.52 (1.30 to 4.89) | 1.73 (0.88 to 3.38) |
| No | Reference | Reference | Reference | Reference | Reference | Reference |
| Comorbidities | ||||||
| DM | 3.39 (2.92 to 3.94) | 0.96 (0.81 to 1.14) | 3.45 (2.96 to 4.01) | 0.97 (0.82 to 1.16) | 2.99 (2.41 to 3.72) | 0.87 (0.68 to 1.12) |
| HTN | 5.50 (4.71 to 6.42) | 1.38 (1.15 to 1.67) | 5.63 (4.81 to 6.59) | 1.37 (1.13 to 1.65) | 4.58 (3.68 to 5.70) | 1.12 (0.86 to 1.46) |
| Dyslipidaemia | 3.07 (2.63 to 3.60) | 1.20 (1.00 to 1.43) | 3.17 (2.70 to 3.72) | 1.19 (1.00 to 1.43) | 3.73 (2.94 to 4.72) | 1.71 (1.31 to 2.23) |
| IHD | 3.48 (2.93 to 4.15) | 1.00 (0.83 to 1.20) | 3.59 (3.00 to 4.28) | 0.97 (0.81 to 1.18) | 3.38 (2.63 to 4.36) | 0.98 (0.74 to 1.29) |
| CVD | 4.37 (3.71 to 5.15) | 1.30 (1.09 to 1.55) | 4.51 (3.82 to 5.33) | 1.31 (1.09 to 1.57) | 4.33 (3.42 to 5.50) | 1.22 (0.94 to 1.58) |
| COPD | 2.85 (2.17 to 3.74) | 1.31 (0.99 to 1.73) | 2.94 (2.24 to 3.88) | 1.42 (1.07 to 1.89) | 2.40 (1.57 to 3.66) | 1.07 (0.69 to 1.66) |
| CKD | 2.00 (1.28 to 3.12) | 0.85 (0.54 to 1.34) | 2.03 (1.29 to 3.21) | 0.90 (0.56 to 1.43) | 1.49 (0.70 to 3.14) | 0.73 (0.34 to 1.57) |
| Depression | 4.02 (3.46 to 4.67) | 2.54 (2.17 to 2.98) | 4.00 (3.43 to 4.67) | 2.40 (2.05 to 2.82) | 4.53 (3.65 to 5.62) | 2.58 (2.06 to 3.24) |
| CCI | ||||||
| 0 | Reference | Reference | Reference | Reference | Reference | Reference |
| 1~2 | 2.27 (1.67 to 3.08) | 1.40 (1.02 to 1.92) | 2.18 (1.61 to 2.96) | 1.33 (0.97 to 1.82) | 1.70 (1.15 to 2.53) | 0.94 (0.63 to 1.43) |
| 3 over | 7.88 (5.90 to 10.54) | 2.35 (1.69 to 3.28) | 7.82 (5.85 to 10.46) | 2.27 (1.63 to 3.18) | 5.73 (3.95 to 8.33) | 1.36 (0.88 to 2.11) |
| Number of healthcare visits | 1.00 (0.99 to 1.00) | 0.99 (0.99 to 0.99) | 0.99 (0.99 to 1.00) | 0.99 (0.99 to 0.99) | 0.99 (0.99 to 1.00) | 0.99 (0.99 to 1.00) |
BMI was categorised according to Korean criteria: underweight (<18.5 kg/m2), normal weight (18.5–22.9 kg/m2), overweight (23.0–24.9 kg/m2), class I obesity (25.0–29.9 kg/m2), class II obesity (30.0–34.9 kg/m2) and class III obesity (≥35.0 kg/m2).
BMI, body mass index; CCI, Charlson Comorbidity Index; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; CVD, cerebrovascular disease; DM, diabetes mellitus; HTN, hypertension; IHD, ischaemic heart disease; NA, not applicable.
Age-stratified subgroup analysis
We further evaluated the association between pSjD and cognitive disorders stratified by age (<65, 65–74, 75–84 and ≥85 years). Among participants aged <65 years, pSjD was associated with an increased risk of MCI (aHR 1.46, 95% CI 1.06 to 2.01), whereas the associations with total cognitive disorders (aHR 1.25, 95% CI 0.99 to 1.57; p=0.056) and AD (aHR 1.22, 95% CI 0.96 to 1.55; p=0.098) were not statistically significant. In the 65–74 years group, pSjD was associated with increased risks of total cognitive disorders (aHR 1.39, 95% CI 1.20 to 1.60), AD (aHR 1.41, 95% CI 1.22 to 1.63) and MCI (aHR 1.55, 95% CI 1.27 to 1.90). Among participants aged 75–84 years, pSjD was associated with increased risks of AD (aHR 1.16, 95% CI 1.01 to 1.33) and MCI (aHR 1.40, 95% CI 1.13 to 1.74), whereas the association with total cognitive disorders was not statistically significant (aHR 1.14, 95% CI 0.99 to 1.31; p=0.060). No association reached statistical significance in participants aged ≥85 years. Detailed results are presented in table 4.
Table 4. HRs for total cognitive disorders, Alzheimer’s dementia and mild cognitive impairment according to age stratification: a subgroup analysis.
| Under 65 years of age | HR (95% CI) | |||
|---|---|---|---|---|
| Univariable | P value | Multivariable* | P value | |
| Total cognitive disorders | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.18 (0.94 to 1.48) | 0.148 | 1.25 (0.99 to 1.57) | 0.056 |
| Alzheimer’s dementia | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.15 (0.91 to 1.46) | 0.230 | 1.22 (0.96 to 1.55) | 0.098 |
| Mild cognitive impairment | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.40 (1.02 to 1.92) | 0.037 | 1.46 (1.06 to 2.01) | 0.021 |
| 65~74 years of age | ||||
| Total cognitive disorders | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.37 (1.19 to 1.58) | <0.001 | 1.39 (1.20 to 1.60) | <0.001 |
| Alzheimer’s dementia | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.39 (1.21 to 1.61) | <0.001 | 1.41 (1.22 to 1.63) | <0.001 |
| Mild cognitive impairment | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.58 (1.29 to 1.93) | <0.001 | 1.55 (1.27 to 1.90) | <0.001 |
| 75~84 years of age | ||||
| Total cognitive disorders | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.16 (1.01 to 1.33) | 0.039 | 1.14 (0.99 to 1.31) | 0.060 |
| Alzheimer’s dementia | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.17 (1.02 to 1.34) | 0.028 | 1.16 (1.01 to 1.33) | 0.041 |
| Mild cognitive impairment | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.41 (1.14 to 1.75) | 0.002 | 1.40 (1.13 to 1.74) | 0.003 |
| ≥85 years of age | ||||
| Total cognitive disorders | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.16 (0.77 to 1.74) | 0.479 | 1.14 (0.75 to 1.74) | 0.533 |
| Alzheimer’s dementia | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.08 (0.71 to 1.64) | 0.711 | 1.07 (0.70 to 1.65) | 0.748 |
| Mild cognitive impairment | ||||
| HCs | Reference | Reference | ||
| pSjD | 1.54 (0.67 to 3.52) | 0.305 | 1.64 (0.69 to 3.90) | 0.267 |
Adjusted with sex, level of income, body mass index (underweight, normal, overweight, class I/II/III obesity), smoking/alcohol status, sufficient aerobic activity, hearing and visual impairment, comorbidities, Charlson Comorbidity Index and number of healthcare visit.
HC, healthy control; pSjD, primary Sjögren’s disease.
Among patients with pSjD aged <65 years, overweight status was associated with increased risks of total cognitive disorders, AD and MCI. In the 65–74 years pSjD group, low household income was associated with all three outcomes and class III obesity was associated with MCI. Among those aged 75–84 years, hearing impairment was associated with total cognitive disorders and AD. In the ≥85 years pSjD group, middle income and class II or III obesity were associated with some outcomes, although these estimates had wide CIs and should be interpreted cautiously (table 5). In the corresponding HC analyses, low income, several BMI categories and hearing impairment showed age-specific associations among participants aged <65 years, while underweight status was associated with all three outcomes in the 65–74 years group. Insufficient aerobic activity was not associated with cognitive outcomes in either of these HC age groups (online supplemental table 8).
Table 5. HRs for cognitive disorders and associated risk factors in patients with primary Sjögren’s disease, stratified by age.
| Under 65 years of age | Total cognitive disorders | Alzheimer’s dementia | Mild cognitive impairment |
|---|---|---|---|
| HR (95% CI) | HR (95% CI) | HR (95% CI) | |
| Sex | |||
| Male | Reference | Reference | Reference |
| Female | 0.69 (0.34 to 1.38) | 0.84 (0.38 to 1.82) | 0.82 (0.29 to 2.31) |
| Income | |||
| High | Reference | Reference | Reference |
| Middle | 0.93 (0.56 to 1.53) | 0.96 (0.58 to 1.61) | 0.74 (0.35 to 1.56) |
| Low | 1.39 (0.90 to 2.15) | 1.29 (0.81 to 2.05) | 1.61 (0.90 to 2.87) |
| BMI* | |||
| Class III obesity | NA | NA | NA |
| Class II obesity | 0.64 (0.09 to 4.65) | 0.77 (0.10 to 5.61) | NA |
| Class I obesity | 1.18 (0.68 to 2.06) | 1.29 (0.72 to 2.31) | 1.18 (0.54 to 2.59) |
| Overweight | 1.96 (1.23 to 3.11) | 2.16 (1.34 to 3.50) | 2.16 (1.15 to 4.05) |
| Normal | Reference | Reference | Reference |
| Underweight | 1.33 (0.60 to 2.97) | 1.07 (0.42 to 2.72) | 1.94 (0.73 to 5.16) |
| Sufficient aerobic activity | |||
| Yes | Reference | Reference | Reference |
| No | 1.30 (0.75 to 2.26) | 1.40 (0.78 to 2.52) | 0.88 (0.45 to 1.72) |
| Hearing impairment | |||
| Yes | 0.93 (0.13 to 6.74) | 1.03 (0.14 to 7.46) | 1.72 (0.23 to 12.55) |
| No | Reference | Reference | Reference |
| Visual impairment | |||
| Yes | 2.11 (0.52 to 8.57) | 2.35 (0.58 to 9.58) | NA |
| No | Reference | Reference | Reference |
| 65~74 years of age | |||
| Sex | |||
| Male | Reference | Reference | Reference |
| Female | 1.02 (0.70 to 1.47) | 0.98 (0.67 to 1.43) | 0.79 (0.49 to 1.28) |
| Income | |||
| High | Reference | Reference | Reference |
| Middle | 1.16 (0.83 to 1.63) | 1.09 (0.77 to 1.55) | 1.53 (0.99 to 2.35) |
| Low | 1.97 (1.52 to 2.55) | 2.04 (1.57 to 2.64) | 1.85 (1.29 to 2.67) |
| BMI** | |||
| Class III obesity | 3.58 (0.88 to 14.59) | 3.67 (0.90 to 14.97) | 7.09 (1.71 to 29.40) |
| Class II obesity | 1.69 (0.85 to 3.35) | 1.80 (0.91 to 3.59) | 1.29 (0.47 to 3.58) |
| Class I obesity | 1.28 (0.95 to 1.73) | 1.22 (0.90 to 1.66) | 0.98 (0.65 to 1.49) |
| Overweight | 1.15 (0.83 to 1.59) | 1.12 (0.81 to 1.55) | 1.11 (0.72 to 1.70) |
| Normal | Reference | Reference | Reference |
| Underweight | 1.53 (0.85 to 2.74) | 1.56 (0.87 to 2.79) | 1.84 (0.91 to 3.75) |
| Sufficient aerobic activity | |||
| Yes | Reference | Reference | Reference |
| No | 1.28 (0.90 to 1.84) | 1.44 (0.99 to 2.09) | 1.07 (0.69 to 1.69) |
| Hearing impairment | |||
| Yes | 0.99 (0.54 to 1.81) | 0.90 (0.48 to 1.70) | 1.69 (0.86 to 3.33) |
| No | Reference | Reference | Reference |
| Visual impairment | |||
| Yes | 1.18 (0.53 to 2.67) | 1.16 (0.52 to 2.62) | 1.48 (0.54 to 4.02) |
| No | Reference | Reference | Reference |
| 75~84 years of age | |||
| Sex | |||
| Male | Reference | Reference | Reference |
| Female | 1.01 (0.72 to 1.42) | 0.95 (0.67 to 1.33) | 0.90 (0.54 to 1.50) |
| Income | |||
| High | Reference | Reference | Reference |
| Middle | 0.80 (0.55 to 1.16) | 0.83 (0.57 to 1.21) | 0.91 (0.51 to 1.60) |
| Low | 1.21 (0.94 to 1.57) | 1.23 (0.95 to 1.59) | 1.42 (0.96 to 2.10) |
| BMI* | |||
| Class III obesity | 0.94 (0.13 to 6.73) | 0.77 (0.11 to 5.55) | 2.43 (0.33 to 17.77) |
| Class II obesity | 1.33 (0.64 to 2.75) | 1.32 (0.64 to 2.75) | 1.27 (0.45 to 3.57) |
| Class I obesity | 0.93 (0.68 to 1.26) | 0.92 (0.68 to 1.26) | 1.03 (0.67 to 1.60) |
| Overweight | 0.77 (0.55 to 1.09) | 0.74 (0.53 to 1.05) | 0.75 (0.45 to 1.25) |
| Normal | Reference | Reference | Reference |
| Underweight | 0.70 (0.37 to 1.35) | 0.71 (0.37 to 1.37) | 0.15 (0.02 to 1.09) |
| Sufficient aerobic activity | |||
| Yes | Reference | Reference | Reference |
| No | 0.74 (0.51 to 1.08) | 0.79 (0.54 to 1.16) | 1.00 (0.54 to 1.86) |
| Hearing impairment | |||
| Yes | 1.69 (1.18 to 2.43) | 1.59 (1.09 to 2.30) | 1.62 (0.95 to 2.75) |
| No | Reference | Reference | Reference |
| Visual impairment | |||
| Yes | 1.37 (0.67 to 2.79) | 1.39 (0.68 to 2.83) | 2.34 (0.94 to 5.81) |
| No | Reference | Reference | Reference |
| ≥85 years of age | |||
| Sex | |||
| Male | Reference | Reference | Reference |
| Female | 0.70 (0.29 to 1.70) | 0.68 (0.28 to 1.66) | 0.29 (0.06 to 1.34) |
| Income | |||
| High | Reference | Reference | Reference |
| Middle | 5.09 (1.48 to 17.50) | 5.25 (1.51 to 18.29) | 0.97 (0.12 to 7.74) |
| Low | 3.24 (1.12 to 9.38) | 2.89 (0.98 to 8.47) | 1.71 (0.28 to 10.57) |
| BMI* | |||
| Class III obesity | 23.67 (2.12 to 264.44) | 22.72 (2.03 to 254.82) | NA |
| Class II obesity | 7.66 (1.25 to 46.94) | 7.16 (1.16 to 44.06) | NA |
| Class I obesity | 2.19 (0.56 to 8.44) | 2.19 (0.56 to 8.67) | 15.07 (1.00 to 227.37) |
| Overweight | 1.34 (0.34 to 5.29) | 1.25 (0.32 to 4.94) | 7.54 (0.52 to 109.84) |
| Normal | Reference | Reference | Reference |
| Underweight | 0.76 (0.09 to 6.62) | 0.76 (0.09 to 6.65) | NA |
| Sufficient aerobic activity | |||
| Yes | Reference | Reference | Reference |
| No | 1.85 (0.21 to 16.06) | 1.74 (0.20 to 15.12) | NA |
| Hearing impairment | |||
| Yes | 0.95 (0.39 to 2.34) | 0.99 (0.40 to 2.45) | 1.46 (0.28 to 7.80) |
| No | Reference | Reference | Reference |
| Visual impairment | |||
| Yes | NA | NA | NA |
| No | Reference | Reference | Reference |
BMI was categorised according to Korean criteria: underweight (<18.5 kg/m2), normal weight (18.5–22.9 kg/m2), overweight (23.0–24.9 kg/m2), class I obesity (25.0–29.9 kg/m2), class II obesity (30.0–34.9 kg/m2) and class III obesity (≥35.0 kg/m2).
BMI, body mass index; NA, not applicable.
Discussion
In the present study, patients with pSjD had significantly increased risks of total cognitive disorders, AD, secondary dementia, unclassified dementia and MCI compared with PS-matched HCs, whereas the association with VaD was not statistically significant. These findings are noteworthy because they are derived from a large, nationwide cohort representative of the South Korean population, thereby reflecting real-world IRs of cognitive disorders in patients with pSjD. Given that pSjD is a systemic autoimmune disease with potential CNS involvement, our results support the hypothesis that patients with pSjD are at increased risk of cognitive dysfunction. These associations persisted after balancing major vascular, metabolic, pulmonary, renal and psychiatric comorbidities, adjusting for overall comorbidity burden and healthcare utilisation, and accounting for death as a competing event. Among participants aged <65 years, the risk of MCI was significantly increased, whereas the associations with total cognitive disorders and AD did not reach statistical significance. This age-specific finding supports clinical attention to cognitive symptoms even in younger patients with pSjD, but the observational nature of the study does not support routine cognitive screening of all patients or causal claims regarding prevention.
AD is the most common form of dementia, followed by VaD. AD is characterised by multifactorial pathogenesis, including neurodegeneration with amyloid-β plaque deposition and tau hyperphosphorylation as neuropathological hallmarks.15 16 In addition, secondary processes such as neuroinflammation, oxidative stress, mitochondrial dysfunction and blood-brain barrier disruption further contribute to AD pathogenesis.17 18 Genetic factors, particularly the apolipoprotein E ε4 (APOE ε4) allele, are also well-known risk factors for AD.16 19 In contrast, VaD is primarily driven by cerebrovascular pathology, including cerebral infarctions, white matter lesions, demyelination, small vessel disease, chronic hypoperfusion, ischaemia-reperfusion injury and atherosclerosis.20 In the present study, the incidence and adjusted hazard of AD were significantly higher in patients with pSjD, whereas the primary analysis showed no statistically significant association with VaD. The increased AD risk may be biologically plausible in light of potential shared immunopathological mechanisms between pSjD and AD. Failure of immune tolerance—potentially mediated by dysfunctional dendritic cells involved in central and peripheral tolerance—may contribute to autoimmunity in both diseases through enhanced T helper 17 responses, chronic neuroinflammation and autoantibody production.21 22 Moreover, the APOE ε4 allele, a major genetic risk factor for AD, has been reported to be associated with early-onset pSjD.23 Notably, most previous epidemiological studies evaluating dementia risk in pSjD did not differentiate dementia subtypes.6–10 24 25 By evaluating AD, VaD, secondary dementia, unclassified dementia and MCI separately, our study extends previous epidemiological evidence by providing subtype-specific risk estimates in pSjD.
In the present study, which excluded patients with antiphospholipid syndrome and accounted for major vascular and psychiatric comorbidities through PS matching, pSjD was not significantly associated with VaD (aHR 1.24, 95% CI 0.97 to 1.58; p=0.084). The Fine-Gray competing-risk analysis yielded a consistent result. However, when VaD was defined using ICD-10 codes alone, pSjD was associated with a significantly increased risk (aHR 1.32, 95% CI 1.07 to 1.61). These findings suggest that the estimated association between pSjD and VaD may be sensitive to the operational definition of the outcome. Requiring both a diagnostic code and prescription of cognitive-enhancing medication may capture a more restricted subset of treated patients and may fail to identify patients with VaD who do not receive such medications. Therefore, further studies using clinically adjudicated diagnoses and neuroimaging findings are needed to clarify the association between pSjD and VaD.
MCI is defined as a clinical condition characterised by measurable cognitive decline beyond what is expected for normal ageing, while functional independence in daily living is largely preserved.26 It is currently regarded as a heterogeneous clinical syndrome with diverse aetiologies, including neurodegenerative, vascular, psychiatric and metabolic causes, rather than simply a prodromal stage of dementia.26 Because a proportion of patients with MCI progress to dementia, longitudinal cognitive and functional follow-up remains clinically important.27 In the present study, pSjD was associated with an increased risk of MCI among participants aged <65 years (aHR 1.46, 95% CI 1.06 to 2.01), whereas total cognitive disorders and AD did not reach statistical significance in this age group (table 4). These findings support early clinical attention when subtle cognitive complaints or decline are suspected, including in relatively young patients with pSjD. Age-stratified analyses identified different correlates across age groups: overweight status was associated with total cognitive disorders, AD and MCI in patients aged <65 years; low household income was associated with all three outcomes and class III obesity with MCI in those aged 65–74 years and hearing impairment was associated with total cognitive disorders and AD in those aged 75–84 years (table 5). However, these subgroup findings, particularly those based on small numbers of events or wide CIs, should be considered exploratory. They may help identify patients who warrant closer assessment, but they do not establish that modifying these factors will prevent cognitive decline.
Several limitations should be acknowledged. First, inherent limitations of the NHIS database must be considered. Since clinical reconfirmation of diagnoses is not possible using administrative claims data, we relied on operational definitions. To enhance diagnostic validity of pSjD, we included only patients registered under the intractable rare disease code (V139), which requires fulfilment of the 2016 American College of Rheumatology/European Alliance of Associations for Rheumatology (EULAR) classification criteria. Cognitive disorders were defined using a combination of ICD-10 diagnostic codes and prescriptions for acetylcholinesterase inhibitors. This definition may have underascertained milder or untreated cases. Although analyses based on ICD-10 codes alone produced largely consistent findings for total cognitive disorders and most subtypes, the VaD estimate differed, indicating that outcome misclassification cannot be excluded. Second, information from health screening examinations (eg, BMI, smoking, alcohol consumption and physical activity) was incomplete because participation in government-funded health screenings, although freely available to individuals aged ≥20 years, is voluntary. This may have introduced selection bias and may limit the generalisability of findings involving health screening variables. Third, patients with pSjD had more healthcare visits than HCs, which could increase the likelihood of detecting cognitive disorders. Adjustment for the number of healthcare visits did not materially change the associations, but residual surveillance bias cannot be completely excluded. Fourth, the NHIS database does not contain detailed clinical information on pSjD, such as extraglandular manifestations, disease activity or serological profiles. In addition, prescription data for glucocorticoids and immunosuppressive agents were unavailable because they were not included in the original NHIS data request, and biologic therapies for pSjD are not captured because they are not reimbursed by the Korean National Health Insurance system. Therefore, we were unable to evaluate the associations between these disease-specific clinical characteristics or treatments and the risk of cognitive disorders. Fifth, despite PS matching and extensive multivariable adjustment for major comorbidities, overall comorbidity burden and healthcare utilisation, residual confounding by unmeasured clinical or socioeconomic factors may remain. Finally, as this was an observational epidemiological study, causal relationships and underlying biological mechanisms linking pSjD to cognitive disorders could not be established. Further basic and translational studies are required to elucidate the pathophysiological mechanisms underlying cognitive impairment in pSjD.
In conclusion, pSjD was associated with increased risks of total cognitive disorders, AD, secondary dementia, unclassified dementia and MCI in this nationwide PS-matched cohort. Therefore, when cognitive decline or memory impairment is suspected in patients with pSjD, prompt and appropriate cognitive evaluation should be considered, including in younger patients in whom the risk of MCI was elevated.
Supplementary material
The funder had no role in the analysis and interpretation of the data or in the writing of the manuscript.
Footnotes
Funding: This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (RS-2024-00339794).
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
Ethics approval: The study protocol was approved by the Institutional Review Board (IRB) of Konkuk University Medical Center (approval no. KUMC 2024-09-004). This study was conducted in accordance with the Declaration of Helsinki and its amendments. Given the use of anonymised secondary data, the requirement for written informed consent was waived by the IRB.
Data availability statement
Data are available on reasonable request.
References
- 1.Baldini C, Fulvio G, La Rocca G, et al. Update on the pathophysiology and treatment of primary Sjögren syndrome. Nat Rev Rheumatol. 2024;20:473–91. doi: 10.1038/s41584-024-01135-3. [DOI] [PubMed] [Google Scholar]
- 2.Kwon E-J, Park Y, Kwok S-K, et al. Clinical characteristics of chronic sclerosing sialadenitis as a distinctive entity from primary Sjögren’s syndrome. J Rheum Dis . 2024;31:230–43. doi: 10.4078/jrd.2024.0043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Shiboski CH, Shiboski SC, Seror R, et al. 2016 American College of Rheumatology/European League Against Rheumatism classification criteria for primary Sjögren’s syndrome: a consensus and data-driven methodology involving three international patient cohorts. Arthritis Rheumatol. 2017;69:35–45. doi: 10.1002/art.39859. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ramos-Casals M, Brito-Zerón P, Bombardieri S, et al. EULAR recommendations for the management of Sjögren’s syndrome with topical and systemic therapies. Ann Rheum Dis. 2020;79:3–18. doi: 10.1136/annrheumdis-2019-216114. [DOI] [PubMed] [Google Scholar]
- 5.Margaretten M. Neurologic manifestations of primary Sjögren syndrome. Rheum Dis Clin North Am. 2017;43:519–29. doi: 10.1016/j.rdc.2017.06.002. [DOI] [PubMed] [Google Scholar]
- 6.Liliang P-C, Liang C-L, Lu K, et al. Population-based study suggests an increased risk of Alzheimer’sdisease in Sjögren’s syndrome. Clin Rheumatol. 2018;37:935–41. doi: 10.1007/s10067-017-3940-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chen H-H, Perng W-T, Chiou J-Y, et al. Risk of dementia among patients with Sjögren’s syndrome: a nationwide population-based cohort study in Taiwan. Semin Arthritis Rheum. 2019;48:895–9. doi: 10.1016/j.semarthrit.2018.06.007. [DOI] [PubMed] [Google Scholar]
- 8.Hou T-Y, Hsu H-C, Lin T-M, et al. Higher risk of dementia in primary Sjogren’s syndrome. Ann Clin Transl Neurol. 2019;6:633–41. doi: 10.1002/acn3.737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lee K-A, Jeon H, Kim H-S, et al. Increased risk of dementia in patients with primary Sjogren’s syndrome: a nationwide population-based cohort study. Korean J Intern Med. 2025;40:330–8. doi: 10.3904/kjim.2023.536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Park H, Yim D, Ochirpurev B, et al. Association between dementia and systemic rheumatic disease: a nationwide population-based study. PLoS ONE. 2021;16:e0248395. doi: 10.1371/journal.pone.0248395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Yeung CHC, Au Yeung SL, Schooling CM. Association of autoimmune diseases with Alzheimer’s disease: a Mendelian randomization study. J Psychiatr Res. 2022;155:550–8. doi: 10.1016/j.jpsychires.2022.09.052. [DOI] [PubMed] [Google Scholar]
- 12.Lim SJ, Jang SI. Leveraging National Health Insurance Service data for public health research in Korea: structure, applications, and future directions. J Korean Med Sci. 2025;40:e111. doi: 10.3346/jkms.2025.40.e111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cheol Seong S, Kim Y-Y, Khang Y-H, et al. Data resource profile: the National Health Information Database of the National Health Insurance Service in South Korea. Int J Epidemiol. 2017;46:799–800. doi: 10.1093/ije/dyw253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kim K-K, Haam J-H, Kim BT, et al. Evaluation and treatment of obesity and its comorbidities: 2022 update of clinical practice guidelines for obesity by the Korean Society for the Study of Obesity. J Obes Metab Syndr. 2023;32:1–24. doi: 10.7570/jomes23016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mehta RI, Schneider JA. What is “Alzheimer’s disease”? The neuropathological heterogeneity of clinically defined Alzheimer’s dementia. Curr Opin Neurol. 2021;34:237–45. doi: 10.1097/WCO.0000000000000912. [DOI] [PubMed] [Google Scholar]
- 16.Nasb M, Tao W, Chen N. Alzheimer’s disease puzzle: delving into pathogenesis hypotheses. Aging Dis. 2024;15:43–73. doi: 10.14336/AD.2023.0608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lyu Y, Meng Z, Hu Y, et al. Mechanisms of mitophagy and oxidative stress in cerebral ischemia-reperfusion, vascular dementia, and Alzheimer’s disease. Front Mol Neurosci. 2024;17:1394932. doi: 10.3389/fnmol.2024.1394932. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hussain B, Fang C, Chang J. Blood-brain barrier breakdown: an emerging biomarker of cognitive impairment in normal aging and dementia. Front Neurosci. 2021;15:688090. doi: 10.3389/fnins.2021.688090. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Duong MT, Nasrallah IM, Wolk DA, et al. Cholesterol, atherosclerosis, and APOE in vascular contributions to cognitive impairment and dementia (VCID): potential mechanisms and therapy. Front Aging Neurosci. 2021;13:647990. doi: 10.3389/fnagi.2021.647990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mok VCT, Cai Y, Markus HS. Vascular cognitive impairment and dementia: mechanisms, treatment, and future directions. Int J Stroke. 2024;19:838–56. doi: 10.1177/17474930241279888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Manuel SL, Rahman S, Wigdahl B. Dendritic cells in autoimmune diseases and neuroinflammatory disorders. Front Biosci . 2007;12:4315. doi: 10.2741/2390. [DOI] [PubMed] [Google Scholar]
- 22.Oberstein TJ, Taha L, Spitzer P, et al. Imbalance of circulating Th17 and regulatory T cells in Alzheimer’s disease: a case control study. Front Immunol. 2018;9:1213. doi: 10.3389/fimmu.2018.01213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Pertovaara M, Lehtimäki T, Rontu R, et al. Presence of apolipoprotein E epsilon4 allele predisposes to early onset of primary Sjogren’s syndrome. Rheumatology (Oxford) 2004;43:1484–7. doi: 10.1093/rheumatology/keh383. [DOI] [PubMed] [Google Scholar]
- 24.Wang Z-Z, Liu M-S, Sun Z, et al. Risk of dementia or Parkinson’s disease in the presence of Sjögren’s syndrome: a systematic review and meta-analysis. Front Integr Neurosci. 2022;16:1027044. doi: 10.3389/fnint.2022.1027044. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lin T-M, Chen W-S, Sheu J-J, et al. Autoimmune rheumatic diseases increase dementia risk in middle-aged patients: a nationwide cohort study. PLoS One. 2018;13:e0186475. doi: 10.1371/journal.pone.0186475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Dunne RA, Aarsland D, O’Brien JT, et al. Mild cognitive impairment: the Manchester consensus. Age Ageing. 2021;50:72–80. doi: 10.1093/ageing/afaa228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Mitchell AJ, Shiri-Feshki M. Rate of progression of mild cognitive impairment to dementia--meta-analysis of 41 robust inception cohort studies. Acta Psychiatr Scand. 2009;119:252–65. doi: 10.1111/j.1600-0447.2008.01326.x. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available on reasonable request.