If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Medicine, Lillebaelt Hospital, Beriderbakken 4, 7100, Vejle, DenmarkDepartment of Regional Health Research, University of Southern Denmark, Denmark
Department of Medicine, Lillebaelt Hospital, Beriderbakken 4, 7100, Vejle, DenmarkDepartment of Regional Health Research, University of Southern Denmark, Denmark
Department of Medicine, Lillebaelt Hospital, Beriderbakken 4, 7100, Vejle, DenmarkDepartment of Regional Health Research, University of Southern Denmark, Denmark
Incidence and prevalence of sarcoidosis has increased from 2001 to 2015 in Denmark.
•
The contemporary sarcoidosis incidence is 14.5 per 100,000 citizens per year.
•
The contact prevalence of sarcoidosis is 77 per 100,000 citizens.
•
The age-associated incidence peaks for men and women between age 30–39 years.
•
Diagnostic work-up of sarcoidosis correlate well with incidence and disease severity.
Abstract
Introduction
The primary objective was to evaluate the prevalence and incidence of sarcoidosis, and secondly, to evaluate differences in incidence by age at diagnosis, gender, region, calendar year and treatment and to evaluate sarcoidosis-related diagnostic work-up.
Methods
Patients diagnosed with sarcoidosis from 2001 to 2015 and information on diagnostic procedures three months before and after initial diagnosis were identified in the Danish National Patient Register. Incidence proportion and prevalence proportion were calculated using the total population count of Danish citizens.
Results
We identified 8545 sarcoidosis cases. Mean age was 46.0 ± 15.0 years, male gender was overrepresented (56.2%) and systemic corticosteroid was initiated in 46% of cases. The prevalence was 77 per 100,000 citizens in 2015. From 2001 to 2015, the incidence varied from 11.3 to 14.8 per 100,000 per year. The age-associated incidence peaked at 30–39 years in both men (23.6 per 100,000 per year) and women (15.0 per 100,000 per year). Incidence varied from 10.4 to 15.7 per 100,000 per year among regions. In particular, the share of bronchoscopies and chest-computed tomography were high in the region with the highest incidence and low in the region with the lowest incidence. Invasive procedures were more frequently performed in patients treated with systemic corticosteroid.
Conclusion
We find an increasing incidence and prevalence of sarcoidosis, with a peak incidence for both men and women between 30 and 39 years of age. The share of procedures performed seems to correlate well with incidence and disease severity.
Sarcoidosis is a systemic disease characterized by the formation of non-necrotizing granulomas, primarily involving the lungs (90%) and frequently the lymphatic system, skin and eyes. The clinical presentation and prognosis of Sarcoidosis varies greatly. The onset can be acute or non-acute and the severity of sarcoidosis ranges widely from asymptomatic patients with accidental radiographic findings to patients with severe organ involvement. Spontaneous remission within the first two years is seen in approximately 50% of patients, whereas chronic or progressive disease is seen in 20% of patients [
]. The highest incidences have been observed in Europe, specifically in the Nordic countries, and in African American populations, whereas Asian countries have the lowest incidences [
]. The incidence of sarcoidosis in northern Europe varies from 7 to 19 in 100,000 per year; patients are predominantly young, with a peak incidence in the third to fifth decades [
Given the lack of specific diagnostic biomarker, the diagnosis of sarcoidosis is based on the clinical and radiological presentation, pathology with noncaseating granulomas, and exclusion of alternative diagnoses [
Joint statement of the American thoracic society (ATS), the European respiratory society (ERS) and the world association of sarcoidosis and other granulomatous disorders (WASOG) adopted by the ATS board of directors and by the ERS executive committee, February 1999.
]. However, applying these criteria in the clinical setting can be challenging because of the vast differences in not only clinical presentation, but also course of the disease and experience of the physician. Consequently, the diagnostic work-up will differ from case to case [
Joint statement of the American thoracic society (ATS), the European respiratory society (ERS) and the world association of sarcoidosis and other granulomatous disorders (WASOG) adopted by the ATS board of directors and by the ERS executive committee, February 1999.
]. Proton emission tomography (PET) is not used routinely in diagnostic work-up. However, PET can be helpful in, e.g., detecting suitable locations for biopsy and identifying extra-pulmonary sarcoidosis, such as cardiac sarcoidosis [
]. Bronchoscopy with biopsy, bronchoalveolar lavage (BAL) and endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) are used to support the diagnosis, by detecting noncaseating granulomas in the biopsy material and to exclude other diseases (e.g., malignancy, tuberculosis or fungal infection) [
The diagnostic work-up in patients with sarcoidosis has improved in recent decades – mainly because of increased availability of procedures such as CT, HRCT, PET and endobronchial ultrasound (EBUS)/endoscopic ultrasound (EUS). Furthermore, the introduction of fast-track (cancer) pathways since 2008 also contributes to an increase in the use of conventional CT scans in Denmark. The Danish Health Authority reported a 250% increase in CT scans from 2003 to 2014 [
]. These ever-changing diagnostic realities may influence the incidence of sarcoidosis.
Thus, the aim of the present study was to evaluate the contemporary prevalence and incidence of sarcoidosis in adults at a national level. The secondary aim was to evaluate differences in incidence by age at diagnosis, gender, region, calendar year and SC treatment and thirdly, to evaluate sarcoidosis-related diagnostic work-up by SC treatment groups and region.
2. Materials and methods
2.1 Setting
The Danish National Health Service provides tax-funded universal health care, thus all Danish citizens have free of charge access to general practitioners and hospital treatment [
]. Since 2007, Denmark has been divided into five regions (Fig. 3), which are responsible for all treatment provided by the Danish healthcare system, including running public hospitals [
]. A unique ten-digit Civil Registration System (CPR) number is assigned to all Danish residents upon birth or immigration, enabling exact individual-level record linkage across all national registries and life-long follow-up [
Patients aged 18 years or older diagnosed with sarcoidosis from January 1, 2001 to December 31, 2015 were identified in the Danish National Patient Register (DNPR). Cases were followed until death, migration or end of follow-up in 2018.
The DNPR, which is an administrative registry, has achieved complete nationwide coverage on all non-psychiatric admissions since 1978 and out-patient clinic contacts since 1995 [
]. Registration, which is mandatory and submitted by the treating physician, is used for continuous monitoring of hospital and health service utilization and for billing purposes [
]. Registration includes diagnoses, administrative information and diagnostic procedures, but not results of procedures (e.g., spirometry values, bronchoscopy findings). Primary and secondary diagnoses are classified according to the International Classification of Diseases and Related Health problems 10th Revision (ICD-10). ICD-10 code D86, Sarcoidosis, was applied to identify patients with a primary or secondary diagnosis of pulmonary and/or extra-pulmonary sarcoidosis.
2.3 Incident sarcoidosis
Incidence proportion (IP) was assessed yearly from 2001 to 2015 and defined as cases with a first-time diagnosis of sarcoidosis in the DNPR per 100,000 citizens per year. To ensure cases were included on the date of their first-time diagnosis of sarcoidosis, a three-year washout period was applied, to exclude patients with a sarcoidosis diagnosis from 1998 to 2000.
To minimize the risk of misclassification, incident cases were restricted to individuals with >1 contact regarding sarcoidosis within the first year from the index diagnosis. Furthermore, because cancer and sarcoid-like reactions (often caused by cancer) are important differential diagnoses to sarcoidosis due to morphological and pathological similarities, individuals diagnosed with cancer (ICD-10 C00-C97) six months prior to or after the time of sarcoidosis diagnosis were excluded, to minimize the risk of misclassification.
2.4 Prevalent sarcoidosis
The prevalence proportion (PP) was defined as a contact prevalence and included the proportion of cases with a recorded contact regarding sarcoidosis within one calendar year per 100,000 citizens. To reduce the risk of misclassification, prevalent cases were restricted to cases with at least one more contact during a five-year span, that included the prevalent year and two years prior to/post the prevalent year. By applying this five-year span for a second visit we furthermore ensured the inclusion of cases with a last contact in 2001–2002 and cases with a first contact in 2014–2015 because a second visit was captured given the extended study period from 1998 to 2018.
25 Treatment groups
The Danish Register of Medicinal Product Statistics (DRMPS) includes data on all prescription drugs dispensed in Danish community pharmacies since 1995 and consists of high quality data [
We did a subgroup analysis in relation to various treatment regimens as a proxy for disease severity. Cases, who redeemed prescriptions of systemic corticosteroid (SC; ATC H02AB), methotrexate (MTX; ATC L01BA01/L04AX03) or azathioprine (AZA; ATC L04AX01) three months prior to, and up to three year after, diagnosis, were identified in the DRMPS. The follow-up period continued until 2018, thus ensuring a minimum of three years follow-up after diagnosis for all incident cases.
The sarcoidosis case population was divided into a non-treatment group and a treatment group. The treatment group was then divided into four treatment groups:
1.
Short-term treatment group (redemption of ≥1 prescription of SC three months prior to and one year after diagnosis, but no redemption of prescriptions within the following 2 years)
2.
Late initiation treatment group (No redemption of prescribed SC three months prior to and one year after diagnosis, but redemption of ≥1 prescription within the following 2 years)
3.
Continuous treatment group (Redemption of ≥1 prescription of SC three months prior to and one year after diagnosis and redemption of ≥1 prescription within the following two years)
4.
+ MTX/AZA treatment group (Cases identified as in the continuous treatment group, but who also redeemed ≥ 1 prescription of MTX or AZA within the three-year follow-up period)
2.6 Diagnostic procedures
All diagnostic procedures in hospital settings are recorded in the DNPR. Information on diagnostic procedures both three months before and after the date of the index diagnosis included: spirometry, body-plethysmography, lung diffusion capacity, chest x-ray, chest CT, PET, bronchoscopy, EBUS/EUS, mediastinoscopy, thoracentesis as well as any procedure involving a biopsy (including skin, liver, lymph node and lung) (Supplementary 2).
2.7 Covariates
The Danish Civil Registration System (DCRS) has recorded information on date of birth, gender, civil status, residency and daily electronic update on vital status and migration since 1968 [
IP and PP were calculated using the total population count of Danish citizens ≥18 years of age at the beginning of each year as denominator. The annual population count was provided by Statistics Denmark [
]. Five-year IP were obtained by summing the denominator and numerator for each year. The denominator for computing regional incidence was based on the average population count from 2008 to 2015. Incidence was furthermore computed by age (18–29, 30–39, 40–49, 50–59, 60–69, 70+) and gender. Statistical analyses were performed using SAS 9.4 TS Level 1M5 (SAS, Inc., Cary, NC, USA).
3. Results
A total of 10,461 cases coded with sarcoidosis from 2001 to 2015 were identified in the DNPR. Cases with only one visit regarding sarcoidosis within the first year (1354 cases) and cases diagnosed with cancer −/+ 6 months within sarcoidosis diagnosis (562 cases) were excluded; consequently, 8545 cases were included in the sarcoidosis case population (Supplementary 1). Mean age at diagnosis was 46.0 years ± 15.0, male gender was overrepresented (56.2%) and one quarter of patients were aged 30–39 at time of diagnosis. More than half of the patients (54.4%) did not receive SC (Table 1 and Fig. 1).
Table 1Baseline characteristics of incident sarcoidosis cases.
Sarcoidosis case population
Total, n (%)
8545 (100.0)
Gender, n (%)
Male
4803 (56.2)
Age, years (SD)
Total
46.0 (15.0)
Male
45.0 (14.0)
Female
47.3 (16.2)
Regions, n (%)
Southern region
2203 (25.8)
Central region
2071 (24.2)
Northern region
1012 (11.8)
Capital region
2104 (24.6)
Region Zealand
1155 (13.5)
Civil status, n (%)
Living alone
2552 (29.9)
Married
4528 (53.0)
Cohabiting
1465 (17.1)
SC treatment groups, n (%)
Non-treatment
4651 (54.4)
Treatment group - total
3894 (45.6)
Short-term
1546 (18.1)
Late initiation
584 (6.8)
Continuous
1208 (14.1)
+MTX/AZA
556 (6.5)
Data are presented as n, n (SD), (%) or n (%). SD: standard deviation.
The IP of sarcoidosis is displayed in Fig. 2. From 2001 to 2015, the incidence varied from 11.3 to 14.8 per 100,000 per year and the incidence tended to increase over time. The steepest increase was seen from 2007 (11.3 per 100,000 per year) to 2008 (14.8 per 100,000 per year.) The five-year IP from 2011 to 2015 was 14.5 per 100,000 per year (Supplementary 4). The non-treatment group had the most profound increase, from 6.7 to 8.1 per 100,000 per year (Supplementary 1).
Fig. 2Incidence proportions of sarcoidosis from 2001 to 2015.
Fig. 3Regional incidence proportions of sarcoidosis in Denmark from 2001 to 2015. IP: incidence proportion per 100,000 per year. Inhabitants >18 years.
Incidence varied between the five regions – from 10.4 to 15.7 per 100,000 per year (Fig. 3).
The age-associated incidence peaked from 30 to 39 years of age in both men (23.6 per 100,000 per year) and women (15.0 per 100,000 per year). Men had a steady decline in incidence as age increased, whereas for women it seemed to plateau from 40 to 69 years of age, after which it decreased (Fig. 4). Among women, the treatment groups displayed an incidence peak at 30–39 years of age except for the continuous treatment group, which had a steady increase in incidence, with a peak at 50–69 years of age (Supplementary 5).
Fig. 4Age-associated and gender-associated incidence from 2001 to 2015.
The prevalence of sarcoidosis increased continuously from 48 per 100,000 in 2001 to 80 per 100,000 in 2013, and plateaued at 77 per 100,000 in 2014-15 (Fig. 5).
Fig. 5The prevalence of sarcoidosis per 100,000 inhabitants. Prevalent cases have at least one visit in the prevalent year and, furthermore, at least one more visit in the prevalent year or two years prior/post to the index year.
The diagnostic work-up at both three months before and after the date of diagnosis is displayed in Fig. 6. Chest x-ray (84%), chest CT (59%) and spirometry (60%) were the most frequent examinations. Extended lung function was performed in 28% of patients and PET (3%) was rare. In particular, chest CT, PET, bronchoscopy, EBUS/EUS, biopsies and extended lung function varied between groups, with the majority of examinations performed in the MTX/AZA treatment group and the smallest share in the non-treatment group.
Fig. 6Share of diagnostic procedure among sarcoidosis cases according to: a) regions and b) treatment groups. Extended lung function test = body-plethysmography and/or lung diffusion capacity. PET = F-18-FDG positron emission tomography. EBUS = endobronchial ultrasound. EUS = endoscopic ultrasound. Biopsies = thoracentesis and any other biopsies.
The Southern Region had the highest share of chest CT (66%), bronchoscopy (57%) and spirometry (66%). The Central Region had the lowest share of chest x-ray (80%), chest CT (54%) and bronchoscopy (27%). EBUS/EUS and biopsy were most frequent in Region Zealand and most rare in the Northern Region.
4. Discussion
In this large combined cohort and cross-sectional study we find a contemporary sarcoidosis incidence of 14.5 per 100,000 citizens per year, corresponding to approximately 640 cases/year and a prevalence of 77 per 100,000 citizens in 2015. SC is initiated in 46% of cases within three years from initial diagnosis. Furthermore, diagnostic work-up seems to correlate well with incidence and disease severity.
4.1 Prevalence
The prevalence of sarcoidosis varies vastly between countries [
] and comparison is challenging because of diversity of study populations and definitions of prevalence. The prevalence reported in the present study reflects the proportion of the population affected by sarcoidosis each year. Sarcoidosis is primarily diagnosed in a hospital setting, where patients are usually followed for as long as they have symptoms, disease progression or complications, and 2–3 years after remission. Hence, cases who are discontinued are not included as prevalent cases in the subsequent years, because the disease has resolved in the majority of cases. As compared to a study on sarcoidosis from 2016, based on the Swedish National Patient Register, which reported a sarcoidosis prevalence of 64 per 100,000 citizens for cases with at least two visits, the prevalence in our study of 77 per 100,000 in 2015 seems high [
The observed increase in prevalence may reflect the increase in incidence, along with the ever-changing guidelines, which now recommend follow-up of all patients with sarcoidosis for at least two years, including mild cases [
]. Byg et al. studied Danish patients with sarcoidosis identified in the DNPR from 1980 to 1994 and found a lower overall incidence of 7.2 per 100,000 per year [
]. Differences in study design between Byg et al. and the present study are noteworthy when evaluating this two-fold increase in incidence. Firstly, Byg. et al. included individuals younger than 18 years and given the rarity of sarcoidosis in children, this results in a relatively lower IP. Secondly, out-patient are recorded in the DNPR since 1995, and thus only included in the present study, contributing to a higher IP in the present study. Noteworthy, two methodological differences will tend to overestimate the IP by Byg et al. as compared to the present study. Firstly, no wash-out period was applied by Byg et al. before 1980 and secondly, cases were included upon their first visit. The present study included a wash-out period of three years, making the index-date of sarcoidosis highly likely to be the first-time diagnosis and cases were required to have at least two visits to reduce the risk of misclassification. Although the latter two methodological differences are likely to have lesser impact on the IP they are important contributors to the correctness of the IP estimate.
The observed increase in incidence is particularly high in the non-treatment group, supporting the assumption that the increase in incidence is mainly due to an overall increase in diagnostic procedures, leading to more accidental findings of sarcoidosis (Supplementary 4). This is furthermore supported by the temporal coincidence of the introduction of fast-track (cancer) pathways and the marked increase in incidence from 2008 (Fig. 2). Conversely, the +MTX/AZA groups almost double in incidence during the 15-year study period; this may indicate an increased awareness of the need to initiate steroid-sparing therapies, rather than reflecting an actual increase in severe/chronic sarcoidosis (Supplementary 4).
4.3 Age and gender
The average age at the time of diagnosis (46.0 years of age) and the fact that men are younger than women at the time of diagnosis are in line with several other reports [
]. However, in contrast to previous studies, the incidence among women does not display a second peak nor does it exceed the incidence of men at older age [
]. The vast majority of women diagnosed at a late age do not receive therapy, suggesting a milder course of disease. However, it is notable that the incidence of women receiving SC continuously in the three years following diagnosis shows a steady increase as age advances. This might indicate that the burden of severe/chronic sarcoidosis is more substantial in women diagnosed at an older age (Supplementary 4).
Distribution of gender among patients with sarcoidosis varies between studies and geographical regions. Two large studies from Asia found a predominance of female cases [
]. The present study finds a slightly higher share of male cases (56%), which is in line with a recent study that found 55% male cases based on nationwide data from the Swedish National Patient Registry [
]. On the contrary, several Nordic studies from 1950 to 1982 – before the establishment of the National Patient Registries and when citizens were offered routine chest x-ray screening – found a predominance of female cases [
Although these regional and age-associated differences between genders have been observed for decades, the causes are still unclear. Differences in occupational exposure, genetics and hormones have been suggested as possible contributors. However, no conclusion has been reached [
] found the highest prevalence of sarcoidosis in less density populated areas, which is in contrast to our findings of the lowest incidence in the Capital Region, which is by far the densest populated region in Denmark. Furthermore, the regional differences in IP is not reflected by differences in regional SC treatment since the share of non-treated cases show little variation between regions (51–56%), (Supplementary 3). A recent population-based Swedish study found vast regional differences in dispensed SC ranging from 30 to 50% of cases however, they did not find any consistency in the regional differences in treatment and prevalence [
], who also found the lowest incidence in the eastern part of Denmark (approximately corresponding to the Capital Region and Region Zealand). The explanation for these regional differences, that seem to have persisted for at least four decades, is unclear. Denmark has universal tax-funded healthcare with equal access for all citizens, and the socioeconomic and ethnic diversity is small. Differences in unmeasured factors, such as environmental, occupational and social exposures and genetic clustering between the western part of Denmark and the Capital Region might explain some of the regional differences, however this is yet to be explored.
4.5 Diagnostic work-up
This is the first study to investigate the diagnostic work-up of patients with sarcoidosis at a national level. As can be seen in Fig. 6, the diagnostic work-up differs vastly between regions, which might reflect local differences in diagnostic approach and accessibility of the various diagnostic procedures. The share of bronchoscopies and chest CT seems to correlate well with incidence, as the share of these two procedures are highest in the region with the highest incidence (Southern Region) and lowest in the region with the lowest incidence (Capital Region). The share of EBUS/EUS and biopsies is similar in these two regions. Use of PET is rare and does not seem to correlate with incidence; the Southern Region only performs PET in 0.3% of cases, as compared to 6.3% in the Capital Region. The impact of diagnostic approach on incidence is supported by Byg et al. [
], who found the highest incidence of sarcoidosis in the region with the most biopsies performed (48%), and in the region with the lowest incidence only 25% of cases had a biopsy.
The non-treatment case group tend to have fewer diagnostic procedures performed than patients who are treated within the first three years. The + MTX/AZA treatment group has a particularly high share of diagnostic procedures performed, which might indicate more severe/complicated disease at the time of diagnosis in this subgroup of patients (Fig. 6).
4.6 Strengths, limitations and generalizability
Sarcoidosis is not validated in the DNPR, thus to reduce the risk of misclassification, cases were restricted to having a minimum of two sarcoidosis-related visits. However, we cannot rule out that a small share of individuals with only one visit did actually have sarcoidosis but were discontinued, which would tend to underestimate the incidence/prevalence of sarcoidosis. Furthermore, to address the potential misdiagnosis of cancer and sarcoidosis or sarcoid-like reaction, cases with a concurrent diagnosis of cancer and sarcoidosis were excluded. Cancer diagnoses have a high PPV in the DNPR thus, thus we expect misclassification of cancer to be minimal [
]. We cannot rule out, that cancer and sarcoidosis did occur concurrent in a small share of cases, which would tend to underestimate the incidence and prevalence of sarcoidosis.
Regional differences in coding of diagnostic procedures may tend to both over-estimate and under-estimate the share of diagnostic procedures. Furthermore, we cannot rule out that some diagnostic procedures were performed earlier than three months prior to diagnosis, which would tend to underestimate the share of diagnostic procedures.
The DRMPS provides information on collected medication, except for medication dispensed at hospitals, which could tend to underestimate the extent of treatment in some patients. On the contrary, we do not know if the collected drugs were ingested, which would lead to an overestimation of SC treatment in other patients. Furthermore, we have no information on the indications for treatment, nor if SC was prescribed but not collected, which could also lead to misclassification within the treatment groups.
The population-based study design, within a tax-funded, uniformly organised health care system and virtually complete follow-up on all cases, reduces the risk of selection bias [
]. Importantly, most cases of sarcoidosis are diagnosed in hospital settings thus, we expect our data to be largely representative of the entire Danish sarcoidosis population and generalizable to populations with similar demographics and standards of medical care as those in Northern Europe.
5. Conclusion and future implications
We find an increasing incidence and prevalence of sarcoidosis from 2001 to 2015. Our results suggest that the increase in incidence is widely seen among the milder cases (non-treatment group) and, put in a contemporary context, this is likely to be explained by an overall increase in diagnostic procedures. One quarter of cases are diagnosed at the age of 30–39 years. However, the age-associated incidence is markedly higher in men than women aged 30–69 years. Incidence seems to correlate well with the number of diagnostic procedures performed and with disease severity. Regional differences in incidence are observed and further research should focus on investigating risk factors, such as genetics and environmental factors, which potentially contribute to these regional differences.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Contributors
MGS, AL, RI and OH all contributed to the planning and designing of the study. AL and RI had full access to the data and take responsibility for the integrity of the data and the accuracy of the data analysis. The statistical analysis was primarily performed by RI. The first draft of the manuscript was written by MGS, with input from AL and OH. All authors revised the manuscript and accepted the final version. All authors take responsibility for the integrity of the work as a whole, including the data and analysis.
The authors have no competing risks. The study was approved by the Danish Data Protection Agency. The study involved only register-based data and therefore ethical approval was not required.
Data statement
Data are anonymized from the national databases and are not public available.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Joint statement of the American thoracic society (ATS), the European respiratory society (ERS) and the world association of sarcoidosis and other granulomatous disorders (WASOG) adopted by the ATS board of directors and by the ERS executive committee, February 1999.
00254-7&id=gr1.jpg){kind=link}
00254-7&id=gr2.jpg){kind=link}
00254-7&id=gr3.jpg){kind=link}
00254-7&id=gr4.jpg){kind=link}
00254-7&id=gr5.jpg){kind=link}
00254-7&id=gr6.jpg){kind=link}