Risk of cardiovascular disease among patients with sarcoidosis: a population-based retrospective cohort study, 1976–2013

Participants and study design

Through the resources of the Rochester Epidemiology Project, the population of Olmsted County is well suited for investigation of the epidemiology of sarcoidosis because comprehensive medical records for all residents seeking medical care for over six decades are available. The record linkage system allows ready access to the medical records, for inpatients and outpatients, from all healthcare providers for the local population, including the Mayo Clinic, the Olmsted Medical Center and their affiliated hospitals, local nursing homes, and the few private practitioners. The potential of this data system for use in population-based studies has previously been described [11]. This system ensures virtually complete ascertainment of all clinically recognised cases of sarcoidosis among the residents of Olmsted County. The system also allows complete ascertainment of the outcomes of interest (i.e. CVD) for cases and comparators.

The sarcoidosis cohort consists of 345 previously identified cases in Olmsted County, incident from 1976 to 2013 [12]. Potential cases were screened for sarcoidosis using diagnosis codes related to sarcoid, sarcoidosis and contextual noncaseating granuloma, and the medical records of all potential cases were individually reviewed. Inclusion required physician diagnosis supported by histopathology and radiological features of intrathoracic sarcoidosis, compatible clinical presentation, and exclusion of other granulomatous diseases such as tuberculosis. Tissue samples were considered positive if they demonstrated noncaseating granuloma without evidence of acid-fast bacilli or fungi. The only exception to the requirement of histopathological confirmation was stage I pulmonary sarcoidosis that required only radiographic evidence of symmetric bilateral hilar adenopathy in the absence of other identifiable causes. Isolated granulomatous disease of a specific organ except for the skin was also included if there was no better alternative diagnosis [13]. Cases with a diagnosis of sarcoidosis prior to residency in Olmsted County were not included.

For each sarcoidosis patient, one comparator without sarcoidosis at the time of the patient's sarcoidosis diagnosis was randomly selected and assigned an index date that corresponded to the sarcoidosis incidence date. Matching criteria were similar age (±3 years) and same sex.

The medical records of cases and comparators were then reviewed for cardiovascular events which included CAD, congestive heart failure (CHF), atrial fibrillation, cerebrovascular accident (CVA), transient ischaemic attack, peripheral arterial disease and abdominal aortic aneurysm. CAD included both nonfatal and fatal myocardial infarction. Classification was based on physician diagnosis from clinical notes supported by coronary angiography. Patients who underwent coronary revascularisation procedures (percutaneous transluminal coronary angioplasty and coronary artery bypass grafting) were also classified as CAD. CHF was classified based on the Framingham criteria for CHF [14]. Atrial fibrillation was classified based on an ECG. CVA was defined as ischaemic stroke, haemorrhagic stroke, subarachnoid haemorrhage or death from CVA. Classification required physician diagnosis verified by imaging studies or cerebrospinal fluid analysis. Transient ischaemic attack was classified based on physician diagnosis and negative brain imaging study. Peripheral arterial disease was classified based on a resting ankle–brachial systolic pressure index of ≤0.9 [15]. Abdominal aortic aneurysm was classified based on imaging showing the diameter of the abdominal aorta more than 1.5 times normal diameter [16]. Data on baseline cardiovascular risk factors, including smoking status, body mass index (BMI), diabetes mellitus, hypertension and dyslipidaemia, were also collected. The 10-year general Framingham risk score for CVD was calculated and the office-based 10-year Framingham risk score, which does not include laboratory values, was used when lipid values were unavailable [17]. Data on treatment with disease-modifying antirheumatic agents (DMARDs), biological agents and glucocorticoids were also collected. Follow-up was continued until death, migration or January 1, 2015. The study was approved by the Institutional Review Board of Mayo Clinic and Olmsted Medical Center. The need for informed consent was waived.

Statistical analysis

Descriptive statistics (percentages, mean, etc.) were used to summarise the characteristics of each cohort. Comparisons between cohorts were performed using Chi-squared and rank sum tests. Comparisons of prior CVD overall and by type were compared between cohorts using Fisher's exact tests to examine whether there were any increases in prevalence of CVD prior to development of sarcoidosis. The cumulative incidence of CVD overall and by type adjusted for the competing risk of death was estimated [18]. These methods are similar to the Kaplan–Meier method with censoring of patients who are still alive at last follow-up. However, patients who die before experiencing CVD events are appropriately accounted for to avoid overestimation of the rate of occurrence of CVD events, which can occur if such subjects are simply censored at death. For each type of CVD, patients whose first occurrence was prior to the diagnosis of sarcoidosis, or prior to the index date for subjects in the nonsarcoidosis comparison cohort, were excluded from the analysis of cumulative incidence. For the estimation of the cumulative incidence of CVD overall, the date of occurrence was the earliest date of occurrence of any of the individual CVD events.

Cox proportional hazards models were used to compare the rate of development of CVD, individually and in combination, between patients with sarcoidosis and the nonsarcoidosis comparison cohort. In addition, Cox proportional hazards models were used to assess the association of risk factors on the development of CVD events among patients with sarcoidosis. Risk factors of interest included those described above. Only events affecting an adequate number of patients (n≥20) were examined in these risk factor analyses. Additional analyses to investigate the association between treatment (DMARDs, biological agents or glucocorticoids) and CVD were also performed among patients with sarcoidosis. Treatment was modelled as a time-dependent covariate. A p-value of <0.05 was considered statistically significant for all analyses. Analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA) and R version 3.1.1 (R Foundation for Statistical Computing, Vienna, Austria).