Temporal trends and geographic clusters of mortality from amyotrophic lateral sclerosis in Japan, 1995–2004

Abstract

The present study examined temporal trends and geographic clustering of amyotrophic lateral sclerosis (ALS) mortality in Japan, during 1995–2004, using vital statistics based on death certificates. ALS was usually diagnosed by neurologists according to clinical guidelines that complied with the El Escorial Criteria. The underlying cause of death for ALS was coded as G12.2A. Regression analysis was used to examine temporal trends. Spatial scan statistic was used to detect any area of elevated risk as a cluster. A total of 12,173 (6864 male and 5309 female) ALS deaths were reported. Annual crude mortality rate per 100,000 population was 1.07 (1.26 for males and 0.89 for females) in 2004. Although the overall temporal trend was stable, the trend increased in the 70+ years age group (p for trend, < 0.001 in males and < 0.05 in females), while it declined in the under 70 years age group (p for trend, < 0.01 for both sexes). Male preponderance and M/F ratio remained nearly constant over time. Three clusters were detected: two (p < 0.005 in males and p < 0.05 in females) in northeast and one (p < 0.05 in males) in west-central Japan. Further research is needed to clarify contributing factors for the observed trends and clusters in ALS mortality.

Introduction

Amyotrophic lateral sclerosis (ALS) is still a fatal neurodegenerative disease characterized by the selective loss of upper and lower motor neurons. Five to 10% of cases of ALS are familial; the others are believed to be sporadic [1]. While advances have been made in identifying disease-causing genes for familial ALS, very little is known about susceptibility genes or other risk factors for sporadic ALS. Many putative environmental risk factors (i.e., heavy metals, solvents, electrical and electromagnetic fields, poliovirus, mechanical trauma, heavy physical activity, cigarette smoking, and diet) have been previously reported; however, age and a family history are the only established risk factors for ALS [1], [2], [3], [4].

Variation in mortality over time and by geographic location, sex and ethnicity can often be a source of etiological clues [5]. The mortality rate from motor neuron disease (MND), of which ALS accounts for 85% or more [1], [5], was reported to have steadily increased from the 1950s to the 1990s in western countries [3], [5], [6], [7], [8], [9], [10]. In some European countries and the United States, a greater increase in ALS mortality was observed in females than in males in the past 30 years, causing a decrease in the male to female (M/F) ratio [5], [8], [9], [10].

Contrary to the trends noted in many other countries, Japan has shown an unusual pattern of mortality from MND for decades. The age-adjusted MND mortality rate rose from the mid 1950s, peaking in the early 1960s, and declined in the early 1970s [11], [12]. Thereafter, the rate slowly increased to that in the early 1950s for a period of 20 years between 1970 and 1990 [11]. A recent study reported that it decreased from 1995 through 2001, and the M/F ratio slightly increased [13].

The Western Pacific form of ALS, referred to as ALS and parkinsonism-dementia complex (ALS/PDC), was identified in the 1950s in three distinct geographic isolates: Guam, western New Guinea and the Kii Peninsula of Japan [14], [15]. Over the past four decades, the incidence of ALS/PDC has markedly declined in Guam [14]. On the other hand, a continuing high prevalence and incidence in Hohara in Mie prefecture [16] and Kozagawa in Wakayama prefecture [17] in the Kii peninsula are reported, although they temporarily declined in the 1980s.

In this study we examined the national ALS mortality data of Japan to reveal the recent temporal trends in ALS mortality and investigated whether or not any geographic clusters of ALS deaths exist in particular regions.