Epidemiology of sarcoidosis in Japan

Sarcoidosis is a systemic granulomatous disorder of unknown aetiology, involving multiple organs. The disease occurs throughout the world, affecting both sexes and all ethnicities and ages 1. As the clinical phenotype varies in different areas and among ethnic groups, epidemiological surveys of sarcoidosis are important. However, few recent reports on the epidemiology of sarcoidosis are available, which remains a problematic issue for several reasons, including: 1) lack of a precise and consistent case definition; 2) variable methods of case ascertainment; 3) variability in disease presentation; 4) lack of sensitive and specific diagnostic tests, resulting in under-recognition and misdiagnosis of the disease; and 5) the paucity of systematic epidemiological investigations of cause 2.

Japanese sarcoidosis has a low incidence and prevalence, and is less severe than in western countries 3. Sarcoidosis in Japan is reported to have a much higher likelihood of ocular and cardiac involvement than in the West 4. Pulmonary sarcoidosis is recognised as a significant cause of death in western countries, but in Japan it is cardiac lesions that are found to be the main cause of death at autopsy 5, 6. In 1960, the Clinical Epidemiological Research Committee on Sarcoidosis conducted the first nationwide survey of the disease in Japan, followed by seven other surveys up to 1991 6. As there have been no further national epidemiological surveys in Japan since 1991, the current authors decided to undertake the present study.

MATERIALS AND METHODS

RESULTS

The present authors were able to obtain questionnaires from 34 out of the 47 prefectures of Japan (total population 101,404,000, which represents 79.4% of the Japanese population in 2004). The number of newly registered sarcoidosis patients accepted as having this intractable disease in 2004 was 1,679, of which 652 presented with insufficient histopathological findings and were excluded from the analysis. Therefore, a total of 1,027 patients with sarcoidosis were enrolled in the present study. Their characteristics are presented in table 1⇓. The study population was ethnically homogeneous (99.6% Japanese) and consisted of 364 males and 663 females.

Incidence rate

The average incidence rate was 1.01 per 100,000 inhabitants, 0.73 for males and 1.28 for females. The age-specific incidence rate is shown in figure 1⇓, displaying a biphasic pattern, with the first peak at 25–34 yrs and the second at 60–64 yrs. The first peak and the second peak were of almost the same frequency. The disease was found in all age groups (0–86 yrs), but only very rarely in individuals aged <20 or >80 yrs (0.9 and 0.4% of all cases, respectively). Age-specific incidence rates peaked in 20–34-yr-old males, and only 0.6% of cases were <20 yrs old. No cases were observed in males aged >80 yrs. Age-specific incidence rates in females showed a biphasic pattern with a first peak in 25–39-yr-olds and a second in females in their 50s and 60s. This second peak was approximately twice the size of the first peak. Only 1% of female patients were <20 yrs old and 0.7% were >80 yrs old. Familial sarcoidosis, defined as having a first or second degree relative afflicted by the disease, was found in 1.8% of cases.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1—

Incidence rate of sarcoidosis per 100,000 population stratified by a) age at diagnosis and b) sex. ▓: male; □: female. The total number of subjects was 938.

DISCUSSION

The clinical phenotype of sarcoidosis, such as which organs are affected, incidence and seriousness, varies according to geography and ethnicity. Therefore, epidemiological surveys are necessary, but there are few recent large scale surveys. National surveys have been carried out in Japan: the first took place in 1960 and the most recent, the eighth, in 1991 6. However, the present study is the first large scale investigation carried out in Japan with a detailed analysis of many different parameters. Institutional size, investigation period and eligibility criteria (e.g. only new onset or not, exclusively biopsy proven or not) were not identical for all the investigations previously performed. The present study examined newly biopsy-proven sarcoidosis patients in Japan. Even if noncaseous epithelioid cell granuloma is not proven, sarcoidosis patients tend to be diagnosed based on clinical symptoms, course and data from examination. In epidemiological surveys, as it is likely that the incidence rate would be underestimated when including only the biopsy-proven group, clinical diagnosis is important. However, no standard global criteria are available for the clinical diagnosis of sarcoidosis; therefore, only biopsy-proven cases were included for the present study. Data were obtained from an area including 79.4% of the entire Japanese population and, therefore, consider it likely that these data are representative and reflect the clinical features of Japanese sarcoidosis patients in general. The present study population was homogeneous in terms of ethnicity (99.6% Japanese), so it was not necessary to examine racial differences.

In the present study, the incidence rate was 1.01 per 100,000: 0.73 for males and 1.28 for females. The estimated incidence rate in Japan was reported as 1.2 for males and 1.4 for females in both the 1972 and 1984 national surveys 7. Since only 67.3% of patients had histological evidence in 1972 and 61.4% in 1984, it is proposed that the incidence rate for males has remained almost the same over this time period, but increased in females. However, rates cannot be compared directly in different populations as the data are different. For African-Americans, annual sarcoidosis incidence was reported to be as high as 36.5–81.8 per 100,000 8–10. In Sweden, the mean total incidence among individuals aged ≥15 yrs was 19 per 100,000 per year, 16.5 for males and 21.7 for females 11. In the USA, a population-based sarcoidosis incidence study revealed rates of 5.9 for males and 6.3 for females per 100,000 inhabitants per year 12. In the UK, the incidence rate was 5.0, 4.84 for males and 5.24 for females 13. In Japan, the incidence rate was not as high. In Korea, sarcoidosis is still a very rare disease, although it is increasing slowly, with an incidence rate of 0.125 per 100,000 in 1998 14. The reported incidence rate is similarly low in other regions of East Asia, i.e. China, Taiwan and Hong Kong 3, 15, 16. Consequently, it can be concluded that the incidence of sarcoidosis in the Asian population might be lower than in Caucasian and Black populations.

In previous Japanese studies 6, the female/male incidence ratio was: 0.97 in the period 1960–1964; 1.16 in 1965–1969; 1.13 in 1972; 1.14 in 1973–1977; 1.53 in 1984; and 1.70 in 1991. In the present study, carried out in 2004, the female/male incidence ratio was 1.82 (table 6⇓). Thus, it seems that the female/male ratio has been gradually increasing in Japan since 1960. In A Case Control Etiologic Study of Sarcoidosis (ACCESS) study 17, which enrolled only biopsy-proven patients in the USA, the female/male ratio was 1.74. Many other studies have also reported a slight female predominance 18. Because the data of this study were self-reported, and females were more likely to have subjective symptoms, the female/male incidence ratio may have been increased.

The disease shows a consistent predilection for adults aged <40 yrs, peaking in 20–29-yr-old subjects 17, 19. The peak age group in the ACCESS study was 35–44 yrs, in the UK it was 30–44 yrs 13, 17 and in Scandinavian countries there is a second peak incidence in females aged >50 yrs 11, 20, 21. The age-specific incidence rate displayed a biphasic pattern in the whole population and in females in the present study. The incidence rate for females has been biphasic for a long time in Japan 7, 11, 12. For females, the first peak was slightly higher in the 1972 national survey than currently. Conversely, the second peak was slightly higher in 1984 and was approximately twice the first peak in the present study. For females, the incidence rate showed a gradual increase when aged >50 yrs.

Abnormality at medical examination leading to diagnosis of sarcoidosis has shown a considerable decrease over the years, from 50.1% in 1960–1964 to 28.0% in 2004, whereas subjective symptoms have increased as the reason for visiting the hospital (table 6⇑). Eyesight disorder was the most frequent symptom (28.8%), with a high rate of ocular symptoms (30.9%). Previous studies in Japan had revealed that ocular symptoms detected at the first medical examination were frequent, at 24.8–38.9% 6. Ocular sarcoidosis is becoming more common in Japan 22, 23, and ocular involvement showed a large increase from 30.1% in 1960–1964 to 26.5% in 1965–1969 and 54.8% in 2004 (table 6⇑). Previous studies worldwide revealed that the lungs are affected in >90% of patients and that dyspnoea, dry cough and chest pain occur in one-third to one-half of all patients 22. The proportion of patients with bilateral hilar lymphadenopathy decreased from 90.7% in 1960–1964 to 95.3% in 1965–1969 and 75.8% in 2004. The proportion of patients with pulmonary parenchymal lesions increased slightly. Scadding stage I was the most common followed by stage II, the same pattern as in the ACCESS study 17. Skin lesions showed a decrease from 17.1% in 1960–1964 to 7.2% in 1965–1969 and an increase again to 35.4% in 2004. Erythema nodosum has been reported to have a high frequency in Europeans, but is uncommon in Black individuals and Japanese subjects 3, 22, 24.

Cardiac sarcoid granulomas and death were significantly more frequent among Japanese sarcoidosis patients than Caucasians and African-Americans 5, 25. The course of cardiac sarcoidosis is variable and ranges from benign arrhythmias or high-degree heart block to sudden death 22. A previous study of 320 autopsied cases in Japan 5 revealed that mortality relating to cardiac sarcoidosis was found in 46.9% of the 194 (60.6%) sarcoidosis-related deaths, but an ante mortem clinical diagnosis of sarcoidosis had been made in only 26.7% cases. An abnormal electrocardiogram, heart wall motion or myocardial scintigraphy were categorised as abnormalities in cardiac laboratory findings. In the present study, the frequency of patients with abnormalities in cardiac laboratory findings was 23% (224 out of 979). However, it was not possible to establish the frequency of actual heart sarcoidosis in the present study, so patients with only minor electrocardiogram abnormalities were also included.

It is reported that hypercalcaemia occurs in 10–20% of sarcoidosis patients worldwide; previous studies in Japan revealed that the rate of hypercalcaemia is 5.1–8% in this population 26–28. Hypercalcaemia was significantly higher in males in the present study (10.8 versus 5.5%; p = 0.0051). In the ACCESS study, abnormalities of calcium metabolism were more frequent among Caucasian subjects, and significantly higher in males 17. Of 62 patients with hypercalcaemia, calcium in the urine was tested in only 17 (27.4%), of which eight (47.1%) showed hypercalciuria. Thus, in this small number of patients where calcium was assayed in both blood and urine, results suggest that there is a divergence between hypercalcaemia and hypercalciuria measurements.

In Japan, the annual tuberculosis incidence was reported as 23.3 per 100,000 in 2004, down from 354.9 per 100,000 in 1964 29. Despite this decrease, it is still high compared with other developed nations. Thus, many Japanese are already infected with tuberculosis and the majority of the uninfected have received bacille Calmette–Guerin vaccination. It is reported that the rate of tuberculin skin test negativity in Japan is low at 1.7–20% 30, 31. Therefore, the 73.1% of sarcoidosis patients with negative tuberculin skin tests is a remarkable finding.

Females were more likely than males to have eye (p<0.0001), skin (p<0.0001), muscle (p = 0. 0087) and neurological involvement (p = 0.0182), as well as hyper-γ-globulinaemia (p = 0.0169). In contrast, males were more likely to have diffuse lung shadow (p<0.0001), hypercalcaemia (p = 0.0051) and kidney involvement (p = 0.0119). In the ACCESS study, females were more likely to have eye and neurological involvement, and erythema nodosum, whereas males were more likely to have abnormalities in the calcium metabolism. Eye and neurological involvement, and abnormalities of calcium metabolism in the present study showed the same pattern as in the ACCESS study, but it is unknown why this should be so. Sharma 28 reported that there is no evidence of a relationship between sex and hypercalcaemia.

Familial sarcoidosis, defined as having an affected first or second degree relative, was identified in 1.8% of cases. There have been a number of reports of familial clustering in sarcoidosis, ranging from 1.7–17% of cases 32. The eighth national Japanese survey in 1991 identified familial sarcoidosis in 0.7% of all cases 33. Although there are thus far limited reports of familial clustering in Japan, there remains the possibility that familial sarcoidosis may gradually increase.

Statement of interest

None declared.