Epidemiological profile of Dupuytren’s disease in Taiwan (Ethnic Chinese): a nationwide population-based study

This is not only the first population-based epidemiological study on DD in Taiwan, but it is also is the largest epidemiological study that has ever been conducted on DD in an Asian country, and one of very few that uses as its source one nation’s entire stock of national health insurance claims data. Because of the low incidence, it is difficult and impractical to field community studies for an epidemiological approach. The NHIRD, however, is a prodigious resource for studying the epidemiology of rare diseases like DD. We found that both the incidence and prevalence rates of DD are much lower for ethnic Chinese in Taiwan than for Europeans [14,28,29] and US Americans [24,30]. However, the annual incidence rate of DD in Taiwan increased during the study period, especially for women. We also found that the most common comorbidities in ethnic Chinese DD patients were hypertension, diabetes mellitus, hyperlipidemia, ischemic heart disease, and COPD.

Other studies’ estimated prevalence rates of DD ranged from 0.2% [31] to 56% [32]; the large range is attributable to variations in age, population groups, and methods of data collection. Most of the epidemiological data were obtained by sampling specific populations [14,21,24,33] or by retrospectively reviewing charts [23,34]. For instance, one US study [24] estimated the prevalence rate of DD using an online survey from an Internet-based research panel representative of the US population, and one Belgian study [14] estimated it from a random visit by a physician to a market. Our estimate, however, is based not on a sampling, but on data from Taiwan’s NHIRD, into which data are entered without a selection bias. No similar information is available in the published literature in Taiwan or in any other Asian country. Because the diagnosis of DD was already recorded using ICD-9 codes, and because each diagnosis was made by a physician rather than being based on the patient’s memory on a health history questionnaire, a misclassification bias is unlikely. Because patients with only mild symptoms might not seek medical help, and therefore will not be entered in the database, the incidence and prevalence may be underestimated. However, DD is a proliferative and progressive disease, and most people who have it will ultimately need medical attention. Therefore, if the observation period is long enough, the underestimation will become smaller. Furthermore, because more than 98% of Taiwan’s population is ethnic Chinese [35], our data can represent the epidemiological features of DD in ethnic Chinese.

The two main foci of epidemiological studies are incidence rate and prevalence rate. The former is defined as the number of new cases of a disease over a specific time-period and the latter as the current number of cases of disease at a single point in time. Most previous epidemiological studies of DD that have quoted incidence rates were in fact referring to prevalence rate [15]. In the present study, we identified new cases every year from the NHIRD. The average annual incidence rate of DD in Taiwan during 2000–2011 was only 0.39/10⁵ (men 0.48/10⁵; women 0.29/10⁵), which was much lower than that of a UK study [29] (502,493 men; annual incidence: 34.3/10⁵ men), and of a US population-based study [24] (annual incidence: estimated at 30.0/10⁵ adults). We found that DD in Taiwan was much less prevalent than in Western countries [14,21,22,33]. Ethnic studies [14,17,21] show that DD is highly prevalent in Scandinavians. A US veterans-hospital-based study [30] reported that the prevalence of DD is highest in Whites and then Hispanic Whites, Blacks, and Asians.

The epidemiological reports on DD in Asia are only case series or else are limited to special groups. A Taiwanese study [19] reported 41 cases of DD in Taiwan during 1970–1988. Because it was case series from a single hospital, neither an incidence rate nor a prevalence rate could be reported. In addition, a Japanese cross-sectional study [36] of 1,154 patients reported prevalence rates of 19.7% for men and 9% for women. Because all the patients were more than 60 years old and lived in nursing homes, they did not represent a general population.

In the present study, we found a trend of rising incidence and prevalence rates from 2001 to 2011. We hypothesize that the trend is a reflection of the aging population because the prevalence and severity of DD increases with age [37]. Moreover, the incidence of DD increases with comorbidities like diabetes mellitus [9,31], hyperlipidemia [38,39], and hypertension [23,40], all of which are highly prevalent in the elderly [41-44]. Furthermore, the universal coverage and low copayment of Taiwan’s NHI system has increased the affordability of healthcare here, which increases the diagnostic rate for DD. In addition, increased medical knowledge from the Internet might also contribute to the higher contemporary rate of medical consultation [45].

The average age at the onset of DD is about 60 years old, and the incidence rises thereafter [13,37]. Similarly, in our study, the onset of DD age-dependently increased for men, and the incidence rate peaked at 2.94/10⁵ person-years at age 70–79 and dropped a little to 2.37/10⁵ person-years for men over 80. For women < 40 years old, however, the incidence rate jumped from 0.08/10⁵ to 0.81/10⁵ person-years for those 50–69 years old, and fell by 20% to 0.65/10⁵ person-years for those 70–79 and then by 55% to 0.36/10⁵ person-years for ≥ 80. These rates are different from those reported [46,47] for other ethnicities, in which the incidence continuously increased with age in women.

Another unique epidemiology profile for DD in ethnic Chinese is that there are no gender differences in incidence rates for patients < 60, but for those ≥ 60, the incidence rates increased more dramatically for men than for women. The male:female incidence rate ratio increases from 2.0 to 6.6 for those 60–69 and ≥ 80 years old. This is quite different from reports on other ethnicities in which the gender difference in the incidence rate diminishes with increasing age; specifically; the male:female incidence rate ratio approaches 1:1 for people ≥ 80 [17,25,46,47]. Additional studies are necessary to determine the underlying mechanisms.

Generally, DD is much less common in women than in men, with reported overall male:female incidence rate ratios ranging from 1.7:1 to 9.5:1[37,46,47]. Similar to what has been reported in the literature, our results show that both the incidence and prevalence rates are higher for men than for women in Taiwan. The male:female incidence and prevalence rate ratios, however, are 1.72 and 1.55, which are lower than those for Europeans and Australians [37,46], but close to that for US Americans [47]. Additionally, a US study [46] reported that women were older (mean age: 62.4 years) at their first operation (62.4 years) than were men (59.8 years). Other studies [14,21] also report that age at the onset of DD is about 10 years lower in men than in women. However, in the present study, the mean age at onset was about 7 years lower in women (53.72 years) than in men (60.69 years). This, however, might be attributable to genetic and environmental differences between populations. Moreover, the present study is an epidemiological profile report of DD in Taiwanese; thus, from our results, no direct evidence can be drawn determine the underlying mechanism for the early onset of DD in Taiwanese women.

A number of epidemiologic factors have been associated with the development of DD: e.g., diabetes mellitus, hyperlipidemia, hypertension, epilepsy, smoking, and alcoholism [13,15,37,48]. We found that hypertension, diabetes mellitus, and hyperlipidemia were the most common comorbidities both in men and in women. All of these diseases may impair microvascular circulation and cause localized ischemia, which increases free radicals and inflammatory cytokines [37], and concomitantly alters gene expressions of growth factors, adhesion molecules, and extracellular matrix components that, in turn, trigger the proliferation of DD tissue [49].

Our study has some limitations. First, our cohort was identified by ICD-9 codes coded by physicians. Patients with only mild symptoms might not seek medical care and thus were not included in our database. Therefore, the calculation of incidence and prevalence rates based on the NHIRD might underestimate the true rates. Second, data on several environmental factors believed to contribute to DD development, including alcohol intake, smoking, manual labor, or exposure to vibrations [37], were not available in the NHIRD. Only one man among our identified 1,078 patients had been diagnosed with alcoholism; this is undoubtedly an underestimated prevalence rate of habitual alcohol drinking for Taiwan. Third, although the administrative claims data allowed us to identify patients diagnosed with DD, they did not contain information about the severity of the disease. Fourth, there were no data about blood biochemistry, pathology reports, or finger function; thus, it was not possible to determine whether the disease was progressing or in remission. Fifth, although 98% of Taiwanese residents are ethnic Chinese, there are residents of other ethnicities, most of whom are aboriginal Austronesian Taiwanese [35]. Patients’ ethnicities were not recorded in the NHIRD; therefore, we did not know the distribution of DD in ethnic Chinese and aborigines. However, the highest population density of Taiwanese aborigines is in eastern Taiwan, and DD is lowest in that area. Sixth, although the NHI regularly investigates the quality of the ICD-9 coding in the claims database to ensure that diagnoses in particular are accurate, we still cannot exclude the possibility that some might be incorrect because of mistakes by the medical coders and physicians responsible for entering the codes. These limitations may compromise our findings.