The Incidence of Giant Cell Arteritis in Olmsted County, Minnesota: Apparent Fluctuations in a Cyclic Pattern

Methods

The details of our epidemiologic methods have been described previously [4]. Our unique medical records linkage system allows ready access to the medical records of all health care providers for the population of Olmsted County, Minnesota, including the Mayo Clinic and its affiliated hospitals; the Olmsted Medical Group and Olmsted Community Hospital; local nursing homes; and the few private practitioners. All medical records (inpatient and outpatient) of Olmsted County residents that showed a surgical index entry of temporal or occipital artery biopsy or a medical diagnosis of giant cell arteritis between 1986 and 1991 were reviewed, and information about clinical manifestations, laboratory findings, and disease course was collected. In cases where the diagnosis was questionable, three rheumatologists reviewed all of the medical information and reached a consensus. The medical records of all patients from the previous study (1950 to 1985) were also reviewed [3], and the two data sets were collated. The diagnoses of giant cell arteritis and polymyalgia rheumatica were made according to the 1990 American College of Rheumatology criteria [5] and the criteria of Chuang and colleagues [6], respectively. Only persons who had been residents of Olmsted County for at least 12 months before the diagnosis of giant cell arteritis were included as incidence cases.

Age- and sex-specific incidence rates were calculated using the number of incidence cases as the numerator and population estimates based on decennial census counts as the denominator; linear interpolation was used to estimate population counts for years between censuses. Overall rates were age- and sex-adjusted to the 1980 United States white population. Ninety-five percent CIs were computed for incidence rates. The annual incidence rates were graphically illustrated using a 3-year centered moving average to eliminate some of the random fluctuation over time. The month of onset of the first symptoms related to giant cell arteritis was used to calculate the season-specific incidence rates. A generalized linear model [7] with log-link function and Poisson error structure was used to evaluate the relation between log incidence (the dependent variable) and age, sex, and chronologic time (in 5-year intervals), two-way interaction terms, and higher-order polynomial terms for age and time. The models were fit using GLIM (Generalized Linear Interactive Modelling). The observations used for the regression analysis were the crude incidence rates for both sexes in the following age groups: 50 to 59 years, 60 to 69 years, 70 to 79 years, and 80 or more years; and in the following calendar time periods: 1950-1954, 1955-1959, 1960-1964, 1965-1969, 1970-1974, 1975-1979, 1980-1984, and 1985-1991. The midpoints of the age groups and the calendar time periods were used in the trend analysis.

Results

Between 1950 and 1991, 125 Olmsted County residents (103 women and 22 men) were diagnosed with giant cell arteritis. Of these, 81 (64.8%) had giant cell arteritis alone and 44 (35.2%) also had polymyalgia rheumatica. One hundred and fifteen (92%) had positive temporal artery biopsies, and all fulfilled the 1990 American College of Rheumatology criteria for giant cell arteritis [5].

Table 1 shows the age- and sex-specific incidence rates, the overall age-adjusted rates for each sex, and these same age-adjusted rates according to time periods. The best generalized linear model analysis of predictors of incidence showed a significant elevation of incidence in women compared with men (P < 0.0001), a significant increase with age (P < 0.0001), and a significant linear effect of calendar time (P = 0.002). This calendar-time effect predicted a 2.6% increase in incidence every 5 years (CI, 0.9% to 4.3%).

Figure 1 shows the annual incidence rates per 100 000 persons 50 years of age or more between 1950 and 1991. Cases appear to have clustered in five peak periods (1953-1955, 1959-1961, 1967-1969, 1974-1976, and 1982-1984), each of which lasted about 3 years. The peaks in this cyclic pattern occurred approximately every 7 years. At every peak, the incidence rate was more than twice the previous level. Plots of the deviance residuals, taken from the generalized linear model, over time did not show a deviation from the linear trend. However, the relatively small number of incidence cases provided us with limited statistical power to detect important nonlinear trends. The trends in incidence among men and women were similar except during the earlier time periods, 1950-1970. The irregular fluctuations seen in men between 1950 and 1970 cannot be interpreted due to the small number of cases (n = 4) diagnosed among men during this time.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1. Rates were calculated using a centered 3-year moving average. Annual incidence rates of giant cell arteritis in Olmsted County, Minnesota, per 100 000 persons 50 years of age or older.

Plots of the annual incidence rates per 100 000 persons 50 years of age or more showed that, among patients with giant cell arteritis but not polymyalgia rheumatica (n = 81), the five peaks remain clearly delineated. In fact, these peaks were somewhat less evident among patients with giant cell arteritis and polymyalgia rheumatica. Differences in season-specific incidence rates were not statistically significant.

Discussion

Our results, which show that giant cell arteritis is more common in women and in persons more than 70 years of age (Table 1), confirm the results of earlier studies [8, 9]. We show a statistically significant increase in the incidence of giant cell arteritis over the study period (Table 1), but heightened physician awareness and surveillance may account for this. However, our most remarkable finding was the apparent cyclic fluctuation in the incidence rate over the 42-year study period. This observation has not been previously reported. A regular repeating cyclic pattern in the development of giant cell arteritis over the 42-year period was apparent, with peaks appearing at intervals of about 7 years and lasting about 3 years. Although it seems unlikely that this regular cyclic pattern of incidence rates is random, we could not detect a statistically significant cyclic pattern. To suggest that this pattern is due to differences in disease ascertainment would imply alternating periods of high and low ascertainment; this is not plausible. The diagnosis in most cases (92%) was confirmed by temporal artery biopsy, making overascertainment unlikely. Furthermore, most physicians at our institution are very familiar with the presenting manifestations of giant cell arteritis, and rheumatologic consultation is readily available. Finally, the patients were identified using the medical records linkage system of the Mayo Clinic, which ensures almost complete ascertainment of all cases of any disease diagnosed in Olmsted County dating back 75 years [4]. The cyclic pattern is somewhat less evident in men than in women; this is probably due to the much smaller number of cases occurring among men. The cyclic pattern is also less evident among patients with giant cell arteritis and coexistent polymyalgia rheumatica than among those with giant cell arteritis alone. The exact nature of the relation between giant cell arteritis and polymyalgia rheumatica is uncertain. The differences we observed may reflect an unknown additional factor or factors that influence the development of polymyalgia rheumatica in patients with giant cell arteritis.

Illnesses with cyclic frequencies are most often infectious. An infectious cause has been previously postulated for giant cell arteritis and polymyalgia rheumatica [1, 10, 11]. Several reports of conjugal cases have been described and interpreted as evidence of a possible infectious cause [12-14]. Also, the presence of antibodies to intermediate filaments in patients with polymyalgia rheumatica or giant cell arteritis may indicate a viral cause [15] because these antibodies have also been found in serum specimens from patients with viral diseases [16]. Like giant cell arteritis [3, 17-23], other conditions (such as multiple sclerosis) that are suspected of having a viral cause have been noted to have an increased incidence at higher latitudes [24]. This implicates an environmental cause, in particular an infectious agent, that operates with increasing effectiveness or is more common at northern latitudes.

We have shown a statistically significant increase in the incidence of giant cell arteritis over the past four decades. Moreover, ours is the first study to report a regular cyclic pattern in giant cell arteritis incidence rates over time. We wish to emphasize, however, that although these results are intriguing, they are hypothesis-generating, and similar studies in other populations are needed to confirm our observations. Nonetheless, our findings support the hypothesis of an infectious cause for giant cell arteritis, perhaps in a genetically predisposed host. We hope that our data will serve as a stimulus for further research into the cause of this condition.