Influence of the age at diagnosis in the disease expression of primary Sjögren's syndrome. Analysis of 12,753 patients from the Sjögren Big Data Consortium

Objective. To analyse how the main components of the disease phenotype (sicca symptoms, diagnostic tests, immunological markers and systemic disease) can be driven by the age at diagnosis of primary Sjögren's syndrome (pSS). Methods. By January 2021, the participant centres had included 12,753 patients from 25 countries that fulfilled the 2002/2016 classification criteria for pSS. The age at diagnosis was defined as the time when the attending physician confirmed fulfilment of the criteria. Patients were clustered according to age at diagnosis. 50 clusters with more than 100 observations (from 27 to 76 years) were used to study the influence of the age at diagnosis in the disease expression. Results. There was a consistent increase in the frequency of oral dryness according to the age at diagnosis, with a frequency of <90% in patients diagnosed at the youngest ages and >95% in those diagnosed at the oldest ages. The smooth curves that best fitted a linear model were the frequency of dry mouth (adjusted R2 0.87) and the frequency of abnormal oral tests (adjusted R2 0.72). Therefore, for each 1-year increase in the age at diagnosis, the frequency of dry mouth increased by 0.13%, and the frequency of abnormal oral diagnostic tests by 0.11%. There was a consistent year-by-year decrease in the frequency of all autoantibodies and immunological markers except for cryoglobulins. According to the linear models, for each 1-year increase in the age at diagnosis, the frequency of a positive result decreased by 0.57% (for anti-Ro antibodies), 0.47% (for RF) and 0.42% (for anti-La antibodies). The ESSDAI domains which showed a more consistent decrease were glandular and lymph node involvement (for each 1-year increase in the age at diagnosis, the frequency of activity decreased by 0.18%), and constitutional, cutaneous, and haematological involvements (the frequency decreased by 0.09% for each 1-year increase). In contrast, other domains showed an ascending pattern, especially pulmonary involvement (for each 1-year increase in the age at diagnosis, the frequency of activity increased by 0.22%), and peripheral nerve involvement (the frequency increased by 0.09% for each 1-year increase). Conclusion. The influence of the age at diagnosis on the key phenotypic features of pSS is strong, and should be considered critical not only for designing a personalised diagnostic approach, but also to be carefully considered when analysing the results of diagnostic tests and immunological parameters, and when internal organ involvement is suspected at diagnosis.