Population-based incidence and prevalence of systemic lupus erythematosus: the Michigan Lupus Epidemiology and Surveillance program

Abstract

Objective: To estimate the incidence and prevalence of systemic lupus erythematosus (SLE) in a sociodemographically diverse southeastern Michigan source population of 2.4 million people.

Methods: SLE cases fulfilling the American College of Rheumatology classification criteria (primary case definition) or meeting rheumatologist-judged SLE criteria (secondary definition) and residing in Wayne or Washtenaw Counties during 2002-2004 were included. Case finding was performed from 6 source types, including hospitals and private specialists. Age-standardized rates were computed, and capture-recapture was performed to estimate underascertainment of cases.

Results: The overall age-adjusted incidence and prevalence (ACR definition) per 100,000 persons were 5.5 (95% confidence interval [95% CI] 5.0-6.1) and 72.8 (95% CI 70.8-74.8). Among females, the incidence was 9.3 per 100,000 persons and the prevalence was 128.7 per 100,000 persons. Only 7 cases were estimated to have been missed by capture-recapture, adjustment for which did not materially affect the rates. SLE prevalence was 2.3-fold higher in black persons than in white persons, and 10-fold higher in females than in males. Among incident cases, the mean ± SD age at diagnosis was 39.3 ± 16.6 years. Black SLE patients had a higher proportion of renal disease and end-stage renal disease (ESRD) (40.5% and 15.3%, respectively) as compared to white SLE patients (18.8% and 4.5%, respectively). Black patients with renal disease were diagnosed as having SLE at younger age than white patients with renal disease (mean ± SD 34.4 ± 14.9 years versus 41.9 ± 21.3 years; P = 0.05).

Conclusion: SLE prevalence was higher than has been described in most other population-based studies and reached 1 in 537 among black female persons. There were substantial racial disparities in the burden of SLE, with black patients experiencing earlier age at diagnosis, >2-fold increases in SLE incidence and prevalence, and increased proportions of renal disease and progression to ESRD as compared to white patients.

Figure 1

Flow diagram of the ascertainment and verification of cases of systemic lupus erythematosus (SLE) in the Michigan Lupus Epidemiology and Surveillance program. For the category of not incident (inc) or prevalent (prev) during surveillance period, the prevalent cases included both newly diagnosed (incident) and existing cases (i.e., the incident cases are a subset of the prevalent cases). Dx = diagnosis; ACR = American College of Rheumatology; Rheum = rheumatologist.

Figure 2

Venn diagram depicting the overlap of cases identified by the 4 categories of case-finding sources used in the primary capture– recapture models. This diagram corresponds to unique prevalent cases classified as systemic lupus erythematosus (SLE) according to the American College of Rheumatology (ACR) definition. Capture– recapture modeling was performed using both 4-source and 5-source models; the models that used the 4 case-finding sources presented in this Venn diagram were determined to fit the data best and, thus, represent our primary capture–recapture modeling. The majority of cases in the registry (2,124 of 2,278 [93.2%]) meeting the ACR or rheumatologist definitions were ascertained from these 4 case-finding source categories. (The 2 case-finding sources that were not retained in the primary models were Medicaid and laboratory.) Based on the capture–recapture analysis of prevalent SLE cases meeting the ACR criteria, we estimated that an additional 7 prevalent SLE cases were in the source population (Wayne and Washtenaw Counties, Michigan, 2002–2004) but were not ascertained in the Registry. ESRD = end-stage renal disease (data from US Renal Data System).

Figure 3

A and B, Forest plots of the age-standardized incidence (A) and prevalence (B) rates of systemic lupus erythematosus (SLE), according to the American College of Rheumatology (ACR) and rheumatologist (Rheum) case definitions, categorized by sex and race/ethnicity. Values are the point estimates with 95% confidence intervals (95% CIs). Vertical dotted lines indicate the point estimates for the overall male and female rates according to the ACR definition. Numbers shown on the right y-axis are the average annual number of SLE cases during 2002–2004, with the age-standardized rate (ASR) per 100,000 persons and 95% CI. Pacif = Pacific.

Figure 4

A and B, Age-specific average annual incidence (A) and prevalence (B) rates (per 100,000 persons) of systemic lupus erythematosus, according to the American College of Rheumatology (ACR) case definitions, categorized by sex and race/ethnicity, in southeastern Michigan, 2002–2004. Values are the point estimates; shading in A and bars in B represent 95% confidence intervals.