National retrospective registry survey on the epidemiology of sarcoidosis in Finland 2002−2022

Abstract

BackgroundThe prevalence of sarcoidosis is known to be high in the Nordic countries. There are no recent research data on the incidence or prevalence of sarcoidosis in Finland. Our aim was to investigate the epidemiology of sarcoidosis in Finland through a retrospective registry-based study.MethodsWe made an information request to the Hilmo database on patients who had been treated in Finnish specialised care with a main diagnosis related to sarcoidosis. Data were requested for the period 1 January−31 December for the years 2002, 2012 and 2022. In addition, we examined the age and gender distribution and regional differences in these variables between the five university hospital districts covering the whole of Finland.ResultsThe incidence of sarcoidosis was 17‒19/100 000/year throughout the follow-up period. The prevalence of sarcoidosis in the ≥18-year-old population had risen from 85/100 000 in 2002–106/100 000 in 2022. There were considerable differences between university hospital districts: The highest prevalence rate was 170/100 000 in the Tampere University Hospital district in 2022, which was twice as high as in the Helsinki University Hospital district (84/100 000). The proportion of pulmonary sarcoidosis in all sarcoidosis cases decreased from 62% to 45% while the proportion of multiorgan sarcoidosis (D86.8) increased from 11% to 34%. The incidence of sarcoidosis was 15/100 000 and the prevalence was 82/100 000 in the age groups of ≥60 years in 2002. In 2022, the incidence in this same age group had risen to 20/100 000 and the prevalence to 109/100 000. In the ≥60-year-old population, the proportion of D86.8 increased from 11% to 35%.ConclusionsSarcoidosis was a more common disease in Finland than in previous studies. Multiorgan sarcoidosis among the elderly has increased over the past 20 years. This might be explained by changes in environmental factors associated with sarcoidosis. Significant regional differences in prevalence might be partly explained by familial clustering.