Population-based study suggests an increased risk of Alzheimer’sdisease in Sjögren’s syndrome

Sjögren’s syndrome (SS) is a relatively common systemic autoimmune rheumatic disease, in which lymphocytic infiltration of salivary and lacrimal glands leads to immune-mediated secretory dysfunction. The resulting dryness of the mouth and eyes is termed “sicca syndrome.” SS is referred to as “primary” in patients who do not have an additional systemic rheumatic disease, and “secondary” when immune-mediated sicca syndrome coexists in patients with systemic lupus erythematosus (SLE), scleroderma, rheumatoid arthritis (RA), or other autoimmune rheumatic diseases [1]. Systemic involvement (extraglandular manifestations) is common and often results in potentially life-threatening complications, such as lymphoma [2, 3]. Neuropsychiatric complications, including headache, cognitive dysfunction, and mood disorders, are reported in approximately 20% of SS patients [4, 5].

Recently, an association was noted between neuropsychiatric symptoms and autoimmune diseases [4]. Previous studies indicate that inflammatory mechanisms may play an important role in increasing the risk of cognitive impairment and stroke [4‐7]. Most of those studies focused on rheumatoid arthritis (RA) or systemic lupus erythematosus (SLE) or described an increased risk of cardiovascular disease and autoimmune dementia [6‐13].

This nationwide, population-based study was designed to estimate and compare the risk of Alzheimer’s disease (AD) between primary SS patients and non-SS patients during a 10-year period that followed primary SS diagnosis. We were interested in determining whether medical comorbidities or primary SS might explain such an association.

Our study identified 4463 patients in the primary SS cohort and 22,315 patients in the non-SS cohort. The analysis of the demographic data is presented in Table 1. Men accounted for 13.5% of each cohort. After matching for gender, age, and the index year, we found no significant differences in age or sex between the primary SS and non-SS patients. The primary SS cohort, however, had a higher prevalence of hyperlipidemia (p < 0.0001), hypertension (p < 0.0001), and coronary artery disease (p < 0.0001). There were no significant differences in the distribution of comorbidities between the groups DM (p = 0.26), HF (p = 0.26), AF (p = 0.43), and stroke (p = 0.43).

Table 2 shows the crude and adjusted HRs for AD during the 10-year follow-up period, after the year index between patients and controls. We found that, of the 26,778 sampled subjects, 20 patients experienced AD during the 10-year follow-up period, including 7 patients (0.16%) from the study cohort and 13 patients (0.06%) from the comparison cohort. The incidence of AD was 2.68-fold (95% CI, 1.07–6.73; p < 0.05) greater in the primary SS cohort than in the non-SS cohort, with an adjusted HR of 2.69 (95% CI, 1.07–6.76; p < 0.05). Figure 1 shows the disease-free survival curves using the Kaplan–Meier method.

Table 3 presents the prevalence HR and adjusted HR for AD between cohorts. The HR for primary SS patients older than 70 years was 3.69 (95% CI 1.28–10.6; p < 0.05), and an adjusted HR was 3.69 (95% CI 1.27–10.7; p < 0.05).

There have long been concerns regarding the relationship between SS and neuropsychiatric syndromes, including headache, cognitive dysfunction, mood disorder, and polyneuropathy [4]. Nevertheless, previous studies that have examined this relationship have some shortcomings, which include a small sample size, or a definitive diagnosis of AD [12, 13]. The prevalence of AD in primary SS has been unknown until now [18, 19]. To our knowledge, this is the first study that investigates the risk of AD among primary SS patients in Asia. Our retrospective, population-based cohort study, which was conducted with data obtained from Taiwan’s NHIRD, demonstrated that a primary SS diagnosis was independently associated with a 2.69-fold higher risk of subsequent AD, even after adjusting for region of residence and selected comorbidities. Compared to the non-SS cohort, crude and adjusted HRs for AD were highest in older patients (≧ 70 years old).

While cognitive symptoms are commonly experienced, very little is understood concerning the mechanisms of cognitive dysfunction in primary SS. There are several reasons why primary SS patients may have an increased risk of developing AD. Primary SS is a relatively common systemic autoimmune rheumatic disease, in which lymphocytic infiltration of salivary and lacrimal glands leads to immune-mediated secretory dysfunction [1]. In the past, diagnosis has depended on the presence of typical clinical features and/or parotid gland swelling together with focal lymphocytic infiltration demonstrated on biopsy of minor salivary glands and lips. However, serologic findings recently have been recognized to have diagnostic values. Serologic findings include autoantibodies against SS-A/Ro and SS-B/La, antinuclear antibodies, anti-salivary gland antibodies, and rheumatoid factor [20, 21]. The presence of anti-Ro (SS-A) antibodies defines a subset of patients with Sjögren’s syndrome who have systemic clinical manifestations including vasculitis, hematologic abnormalities, and serologic hyperreactivity [22]. Recently, another antibody was found and a pathophysiological role for circulating anti-muscarinic acetylcholine receptor (mAChR) autoantibodies in patients with primary SS was proposed. These autoantibodies recognized and activated mAChRs in both salivary and lacrimal glands [23]. There is a strong regulatory action of parasympathetic stimulation on the secretion of lacrimal and salivary glands [23]. Lacrimal and salivary gland mAChRs are coupled to various signaling pathways, and the production of nitric oxide is of particular interest, because it is involved in several pathologic processes, including SS, where increased levels of nitrites are found in the saliva [23] mAChRs that, acting as an “agonist-like agent,” resulted in a primary, organ-specific dysfunction. Possibly, in primary SS and secondary SS, direct mAChR antibody-mediated tissue damage might occur through nitric oxide generation and accumulation, with an adverse effect on the lacrimal glands. Immunologic generation of nitric oxide could have cytotoxic effects on the cell, through the production of free radicals [23]. This could have a special pathologic role, particularly in SS, where the increased release of inflammatory mediators could induce uncommonly high levels of nitric oxide [23]. It is known that anti-Ro/SSA and anti-La/SSB antibodies are directed against ribonucleoprotein complexes. Their localization is mainly cytoplasmic, and the increased expression of Ro52 in SS and SLE patients is probably implicated positively or negatively in the alteration of cellular immune responses and in the increased apoptosis observed in primary SS patients. Ro52 in primary SS can act as a negative regulator of interferon production, ubiquitination, and degradation of the transcription factors IRF-3, IRF-7, and IRF-8 inducing major alterations in cytokine expression and production [24]. The inflammatory response is associated with an increased risk of cognitive decline [25]. Cevimeline (M1 muscarinic agonist, AF102B) is a drug used in patients with SS to improve sicca symptoms, and it is recognized in innovative therapies for Alzheimer’s disease [26]. The M1 mAChR appears to be linked with two other major hallmarks of this disease β-amyloid (Aβ) and hyperphosphorylated τ proteins. Activation of M1 mAChR could alter beneficially Aβ and τ proteins and apolipoprotein E (ApoE), as well as some processes involving certain G-proteins and neurotrophins [26, 27]. Muscarinic agonists induce differentiation in PC12 cells stably transfected with M1 mAChR (PC12M₁ cells), and these neurotrophic effects are synergistic with a variety of neurotrophins [26, 28]. Several studies showed that several neurotrophins (nerve growth factor [NGF], basic fibroblast growth factor [bFGF], and epidermal growth factor [EGF]) and muscarinic agonists induce neurite outgrowth via distinct intracellular pathways that cross-talk with each other. The agonist, via activation of M1 mAChR, restores learning and memory impairments with an excellent safety margin, mediate Aβ processing, and decrease τhyperphosphorylation, and thus may be useful in Alzheimer’s disease therapy [26]. In a recent animal study, a novel compound was reported, AF710B, a highly potent and selective allosteric M1 muscarinic and σ1 receptor agonist. AF710B exhibits an allosteric agonistic profile on the M1 muscarinic receptor; very low concentrations of AF710B significantly potentiated the binding and efficacy of carbachol on M1 receptors and their downstream effects (p-ERK1/2, p-CREB). The σ1 receptor (σ1R) is another potential target for drug development for AD, as it is considered to play a fundamental role in cognitive function. The compound could (i) mitigate cognitive impairments in the Morris water maze and (ii) decrease BACE1, SK3β activity, p25/CDK5, neuroinflammation, soluble and insoluble Aβ, Aβ, plaques and tau pathologies [29].

SS patients also were found to have significantly lower basal cortisol levels compared to healthy controls, which is thought to result from hypothalamic–pituitary–adrenal axis (HPA) deterioration [30, 31]. Le et al. [31] showed a significant correlation between cortical hypoperfusion and cognitive dysfunction. Another mechanism possibly linking SS to AD is the autoantibodies produced in SS, which may play a role in AD development. These autoantibodies can form an immune complex, and subsequently injure the endothelial walls of small blood vessels [19]. These studies further suggest that inflammation likely contributes to the pathogenesis of accelerating neurodegeneration among SS patients [19, 32].

We also found that, in comparison with patients without SS, primary SS patients were more likely to have hyperlipidemia, hypertension, and CAD. These factors also may explain the higher portion of AD in this cohort [9‐11]. Our findings match the prevalence community-based reports that indicate that primary SS patients have higher cardiovascular risk factors at primary SS diagnosis. Akyel et al. [11] reported that myocardial function is disturbed in SS patients and there is significant atrial electromechanical delay. Our findings support their observation, in that we found a higher prevalence of CAD. Therefore, cardiovascular risk factors should be taken into account when managing such patients.

Autoimmune-induced dementia is an acute or sub-acute disorder that affects memory, cognition, and behavior. More importantly, autoimmune dementia is reversible and may be misdiagnosed as an irreversible neurodegenerative disorder [12, 13]. However, the true incidence rate of autoimmune dementia is unknown; therefore, it is difficult to assign correct ICD-9-CM codes to autoimmune dementia. A review by Rossor et al. [33] showed that young onset dementia had a higher prevalence of genetic or metabolic diseases. Most autoimmune dementia occurs in patients younger than 45 years, and AD is far more common in the geriatric population [34]. Therefore, different age groups were used to check and prevent misclassification errors. In the present study, most patients with primary SS and AD were more than 70 years old. The diagnosis of AD is highly suitable.

One major strength of our study was that it was based on a nationwide, population-based, case cohort study, which minimized the problems of insufficient power. Furthermore, the potential for selection biases was minimized by the comprehensive coverage of the NHI system and a large sample size.

Nevertheless, there are several limitations in our study. First, the diagnosis of primary SS and AD was made using the ICD-9 code from the database. NHIRD does not provide personal information, such as smoking preference, education level, and body mass index. Intelligence and education level also may have some effect on the occurrence of AD. These co-variables could not be adjusted in the analysis. Second, NHIRD provides limited information on clinical characteristics, such as disease severity, laboratory data, and imaging information. Therefore, it is hard to know the relationship between clinical severity and brain region involvement in primary SS and AD. Finally, the incidence rate of AD may be underestimated because inpatient claims data were used.

In conclusion, our nationwide cohort study found that patients with primary SS have a higher prevalence of AD than patients without primary SS. Although the evidence suggests that primary SS is a risk factor for AD, very little is known about the duration or clinical severity of primary SS or how it may contribute to AD. Therefore, further research in this area is necessary to answer such questions, and to examine whether proper management of primary SS may help prevent the occurrence of AD.