Causes of death in patients with Down syndrome in 2014–2016: A population study in Japan

Abstract Despite the higher mortality rates in patients with Down syndrome compared with the general Japanese population, the life span has dramatically increased in Japan and other countries. We aimed to clarify recent causes of death in patients with Down syndrome in Japan. We calculated proportionate mortality and standardized mortality odds ratios (SMORs) among all deaths registered with Down syndrome as the cause of death (ICD‐10 code, Q90) in the Japanese National Death Registry Database in 2014–2016. In the study period, 762 in patients with Down syndrome died. The main causes of death were pneumonia/respiratory infections (20.5%), congenital malformations of the circulatory system (11.2%), other diseases of the circulatory system (9.2%), and aspiration pneumonia (8.4%). The SMORs (95% confidence intervals) were higher for natural death, defined as death of an elderly person with no other cause of death to be mentioned (55.73 [36.92–84.12]), early‐onset Alzheimer's disease, defined as Alzheimer's disease with onset <65 years of age (29.36 [16.44–52.44]), aspiration pneumonia (18.33 [14.03–23.96]), pneumonia/respiratory infections (8.11 [6.76–9.73]), congenital malformations of the circulatory system (8.07 [5.98–10.88]), and leukemia/lymphoma (2.16 [1.55–2.99]) but lower for malignant solid tumors (0.04 [0.02–0.06]) in patients with Down syndrome. Patients with Down syndrome had the greatest relative risk of dying from natural death, early‐onset Alzheimer's disease, and respiratory illnesses, highlighting the need for appropriate medical, health, and welfare services.

Studies from Europe, the United States, Sweden, and Denmark show that congenital heart defects, respiratory infections, dementia, pneumonia, and childhood leukemia are main causes of death in patients with Down syndrome, distinct from the leading causes of death observed in the general Japanese population according to recent statistics, which include malignant neoplasms, and heart disease. (Englund et al., 2013;Hasle et al., 2016;O'Leary et al., 2018;Santoro et al., 2021;Yang et al., 2002). However, several studies suggest that the life expectancy of patients with Down syndrome might have increased in many countries including Japan (Motegi et al., 2020;Presson et al., 2013;Yang et al., 2002), in parallel with advances in medical care and welfare. Despite this, the impact of changing demographics of this patient population including racial and ethnic differences (Santoro et al., 2016) on specific medical issues has not been extensively investigated.
We began this study to understand the recent changes in causes of death by age in patients with Down syndrome will allow the identification of the required role of medical, health, and welfare services and to determine the best approaches for treatment. The aim of this study was to clarify the causes of death in patients with Down syn- The data are then entered into the Vital Statistics Database and used for official statistics. Names, addresses, hospital names, and narrative text about multiple causes of death are not included in this fixed database. In 2007, an additional online vital statistics submission system was established for municipalities, facilitating the direct submission of death certificate data to the Ministry. This database holds all text data on submitted certificates, including name, address, hospital name, and narrative text about multiple causes of death. These two databases are stored separately.
In this study, our analyses included the data on primary causes of death based on ICD-10 codes for the 2014-2016 period, which were retrieved from the Vital Statistics Database, and the data on contributory causes of death based on narrative text for the 2014-2016 period, which were retrieved from the online vital statistics submission system database.
In our main analysis, we used data from the Vital Statistics Database between January 1, 2014 and December 31, 2016 and identified deaths in which Down syndrome was indicated as the primary cause of death (ICD-10 code, Q90). These data did not include those in which Down syndrome was registered as a contributory cause of death. Therefore, we complemented our analysis by extracting data from the online vital statistics submission system database for deaths in the 2014-2016 period. We retrieved all data with the word "Trisomy 21" or "Down" in the narrative text on multiple causes of death or contributory cause of death. We then linked these data to the corresponding vital statistics data. The linkage was performed based on the municipal code of the submission location and the certificate number. The distribution of causes of death in the general Japanese population were calculated from the Vital Statistics Database using e-Stat, an online source for Japanese Governmental Statistic ("e-Stat of government statistics").

| Classification of causes of death
Two investigators (N.M. and Y.Y.) with medical knowledge categorized the ICD-coded causes of death among patients with Down syndrome into the following 27 groups: pneumonia/respiratory infections; congenital malformations of the circulatory system; other diseases of the circulatory system; aspiration pneumonia; natural death, defined as the death of an elderly person with no other cause of death to be mentioned; neoplasms; leukemia/lymphoma; early-onset Alzheimer disease, defined as Alzheimer disease with onset <65 years of age; cerebrovascular diseases; other diseases of the respiratory system; other diseases of the genitourinary system; other infectious diseases; endocrine, nutritional, and metabolic diseases; epilepsy; chronic bronchitis and pulmonary diseases; certain conditions originating in the perinatal period; sepsis; gastrointestinal disease; diseases of liver and biliary tract; other congenital malformations; other diseases including dementia, intellectual disabilities, and nervous system disorders; cerebral palsy, hydrocephalus, and anoxic brain damage; other diseases of the blood and immune system; sudden cardiac death; other respiratory disorders; diseases of the musculoskeletal system and connective tissue; and others. The ICD codes corresponding to these categories are shown in Appendix 1.

| Statistical analysis
First, we calculated the number of deaths for specific causes of death and age groups of 10 years (0, 1-9, 10-19, 20-29, 30-39, 40-49, 50-59, 60-69, 70-79, 80-89, and ≥90 years). Second, we selected five disease groups based on the most common causes of death in the present study, namely, aspiration pneumonia, pneumonia/ respiratory infection, congenital malformation of the circulation, leukemia/lymphoma, and natural death, as well as two causes of death which previous studies suggest occur in distinctly high and low frequency in patients with Down syndrome, namely early-onset Alzheimer disease, and malignant solid tumors (Antonarakis et al., 2020;Hasle et al., 2016;Rethore et al., 2020;Yang et al., 2002). We then compared the patients with Down syndrome with the general Japanese population and calculated standardized mortality odds ratios (SMORs) (K. R. Smith & Kliewer, 1995) to assess the age-specific relationship between causes of death and Down syndrome. SMORs were calculated by dividing the odds for death from a specific cause and the odds for death from all other causes among patients with Down syndrome with the corresponding odds in the general Japanese population (Miettinen & Wang, 1981). Third, we calculated proportionate mortality to determine the probability of death due to a specific cause, calculated as the number of deaths due to a cause divided by the number of living population (Proportionate Mortality, n.d.).
To calculate SMORs, we conducted logistic regression adjusting for age group and year of death (2014)(2015)(2016). To estimate 95% confidence intervals (CIs) for the SMORs, we assumed that the number of deaths and medical conditions followed a Poisson distribution (Rothman & Greenland, 1998).   Table 1 shows the distribution of causes of death among the patients with Down syndrome and in the general Japanese population, based on the major diagnostic categories according to ICD-10 codes. The most common cause of death in the patients with Down syndrome was diseases of the respiratory system (32.0%), followed by diseases of the circulatory system (13.4%), and congenital malformations, deformations, and chromosomal abnormalities (11.9%). On the other hand, the most common cause of death in the general Japanese population was neoplasms (29.6%), followed by diseases of the circulatory system (26.4%), and diseases of the respiratory system (16.0%).
3.2 | Cause of death among the patients with Down syndrome Table 2 shows the causes of death based on disease categories (Appendix 1) and age groups in the patients with Down syndrome.
The causes of death varied by age. In all age groups under 40 the most common cause of death was congenital malformations of the circulatory system (25.5-33.3%), while in the age groups from 40 to 79 the most common cause of death was pneumonia/respiratory infections (22.9-37.9%). In the 80-89-year-age group the most common cause of death was neoplasms (33.3%) and in the over-90-yearage group the most common cause of death was natural death (33.3%).

| SMORs and proportionate mortality of specific causes of death in the patients with Down syndrome
We calculated the SMORs for seven disease groups to compare the patients with Down syndrome with the general Japanese population ( with Down syndrome than in the general Japanese population. Figure 1 shows the proportionate mortality for these seven disease categories as causes of death in the patients with Down syndrome and in the general Japanese population. Compared to the general Japanese population, early-onset Alzheimer disease, aspiration pneumonia, pneumonia/respiratory infections, leukemia/lymphoma, and congenital malformations of the circulatory system were more likely to be a main cause of death whereas natural death and malignant solid tumors were less likely to be a main cause of death in the patients with Down syndrome. The proportionate mortality for natural death was lower in the patients with Down syndrome (4.7%) compared to the general Japanese population (6.5%). In addition, the patients with Down syndrome who died due to natural death were younger (youngest in the 50-59-year-age group) compared to the general Japanese population (youngest in the 60-69-year-age group). The proportionate mortality for natural death was consistently higher for all age groups except the 80-89-year-age group (50-59 years 5.4% vs. 0.0%, 60-69 years 7.1% vs. 0.1%, 70-79 years 20.7% vs. 0.9%, 80-89 years 0.0% vs. 5.2%, and ≥90 years 33.3% vs. 18.2%).
The proportionate mortality for early-onset Alzheimer's disease was higher in the patients with Down syndrome (1.6%) than in the general Japanese population (0%) and was concentrated within 50-69 years of age.
The proportionate mortality for aspiration pneumonia and pneumonia/respiratory infections in the patients with Down syndrome (8.4% and 20.5%, respectively) were at least double the proportionate T A B L E 1 Causes of death for patients with Down syndrome and general population in Japan in 2014-2016 General Japanese population a (n = 3,871,196) Patients with Down syndrome in Japan (n = 762)

| DISCUSSION
In the present study encompassing the 2014-2016 period in Japan, we found that the main causes of death were diseases of the respiratory and circulatory systems in the patients with Down syndrome. We also observed that mortality due to pneumonia/ respiratory infections, aspiration pneumonia, leukemia/lymphoma, early-onset Alzheimer's disease, and natural death were higher Early-onset Alzheimer's disease = Alzheimer's disease with onset <65 years of age.
whereas mortality due to malignant solid tumors was lower in the patients with Down syndrome compared to the general Japanese population.

| Pneumonia/respiratory infections and aspiration pneumonia
We observed that one in four patients with Down syndrome died from pneumonia and that the proportionate mortality for aspiration Pneumonia has been reported to be the most common cause of death throughout life of patients with Down syndrome in developed countries such as Australia (Bittles et al., 2007) and Sweden (Englund et al., 2013), with a study from the United States reporting an SMOR of 7.61 (95% CI: 7.36-7.87) (Yang et al., 2002) compared with the general Japanese population. This estimate is very close to that estimated in the present study.
Patients with Down syndrome are known to have hypotonicity of orofacial muscles. Thus, early oral care and swallowing training (Hashimoto et al., 2014;Ronald et al., 2011) which have shown to be beneficial to patients with Down syndrome may be useful.

| Leukemia/lymphoma
We observed that one in five patients with Down syndrome in the 0-year-age group died from leukemia/lymphoma. The SMOR for leukemia/lymphoma was 2.16, consistent with previous studies from the United States (Yang et al., 2002) and Sweden (Englund et al., 2013).
The incidence of leukemia is higher in children with Down syndrome under the age of 5 years (Bittles et al., 2007). We observed that deaths due to leukemia occurred more often among those aged 0-29 years, especially among those under the age of 10 years, a finding consistent with the studies from the United States (Yang et al., 2002) and Sweden (Englund et al., 2013).
The high mortality due to leukemia may be explained by the higher incidence rate and the overall lower survival rate observed in children with Down syndrome compared to the general population. Children with Down syndrome are predisposed to acute myeloid leukemia (AML) by 150-folds (Murphy et al., 2019) and acute lymphoblastic leukemia (ALL) by 20-folds (Hitzler & Zipursky, 2005) compared to those without Down syndrome.
Meanwhile, ALL survival rate is very low (Pennella et al., 2018), and the AML survival rate is slightly high in patients with Down syndrome (Caldwell et al., 2014). In addition, treatment-related mortality and relapse rate in ALL are higher in children with Down syndrome than in children without Down syndrome (Izraeli et al., 2014).

| Early-onset Alzheimer disease
We observed that death due to early-onset Alzheimer's disease (onset <65 years of age) was more frequent in patients with Down syndrome compared to the general Japanese population, a finding consistent with the study from the United States (Yang et al., 2002). In our study the SMOR for early-onset Alzheimer was 29.36, higher from that estimated in a study about 20 years ago (Yang et al., 2002). Another recent study reported that 88% of patients with Down syndrome younger than 65 years of age suffered from dementia (McCarron et al., 2017). These findings suggest that susceptibility to dementia may be has increasing over the years. Given the continuing increase in deaths due to dementia and longer life expectancy in patients with Down syndrome (Englund et al., 2013;Motegi et al., 2020), it is equally probable that the proportionate mortality for Alzheimer's disease associated with advanced age will increase in patients with Down syndrome in future. (f) aspiration pneumonia *natural death = natural death in the elderly, which has no other cause of death to be described.
**early-onset Alzheimer's disease = Alzheimer's disease with onset < 65 years of age.  (Melville et al., 2005). Prevention of obesity and treatment of Alzheimer's disease from an early stage may lower the risk of dementia in these patients.
The increased SMOR for early-onset Alzheimer's disease was observed at a higher age range (ages 60-69) compared to the studies from the United States (ages 41-59 (Yang et al., 2002) and Sweden (ages 41-59) (Englund et al., 2013), which might be explained by several fac- there is a possibility that there are ethnic differences in polymorphisms in TFAM (mitochondrial transcription factor A), which encodes a key activator of mitochondrial transcription and is known to be related to the earlier onset of dementia among patients with Down syndrome (Ballard et al., 2016). Third, given that obesity is a known risk for early-onset Alzheimer's disease (Chuang et al., 2016), cultural differences in eating habits and obesity rates might have contributed to the observed lower disease risk (Ballard et al., 2016;Mazurek & Wyka, 2015).

| Natural death
We observed that 4.7% of patients with Down syndrome aged between 50 and 102 years died from natural death. This rate, which was higher than that reported in other countries, was however lower than the percentage of deaths from natural death in the general Japanese population in Japan. In the general Japanese population of Japan, the proportionate mortality of natural death in 2018 was 8.0% (Summary of Vital Statistics, 2018), which is higher than that in Australia (0.2%) and the United States (0.0%). This discrepancy may be due to two factors, long life expectancy of patients with Down syndrome in Japan (Motegi et al., 2020) and ambiguity of the medical definition of natural death.
The expected life span of patients with Down syndrome is 50-60 years.
In Japan, one in three people die after the age of 60 years; thus, it is possible that more patients with Down syndrome in Japan are dying from natural death compared to that observed in other countries. On the other hand, the rate of natural death is high even in the general Japanese population in Japan, and the high rate of natural death may be a reflection of the culture respecting the dead and the low postmortem examination rates (Maeda et al., 2013). However, there are no previous reports on the susceptibility of natural death in patients with Down syndrome compared to the general population, thus we do not know whether our findings are unique to Japan or the susceptibility has changed over the years.

| Malignant solid tumors
Our study showed that only 2.1% of patients with Down syndrome died from malignant solid tumors, which was considerably lower than the rate of almost 30% in the general Japanese population. Similarly, the low frequency of malignant solid tumors in patients with Down syndrome was reported in the United States (Yang et al., 2002). In our study the SMOR for malignant sold tumors was 0.04, close to the estimate in a study about 20 years ago conducted in the United States (Yang et al., 2002) suggesting the susceptibility has not changed over the recent years. Specifically, testicular cancer as a cause of death was lower in patients with Down syndrome in Japan compared to the United States (Yang et al., 2002) and Finland (Patja et al., 2006 (Hattori et al., 2000). Third, cell replication is slower in patients with Down syndrome compared to those without Down syndrome, leading to fewer replication errors in genes involved in tumorigenesis (Yang et al., 2002). Fourth, cells might be more prone to undergo apoptosis in patients with Down syndrome (Nizetic & Groet, 2012). Fifth, patients with Down syndrome generally have a shorter life span and undergo menopause earlier (Bittles & Glasson, 2004). Six, cancer screening rates might be lower in patients with Down syndrome, who are more likely to have intellectual disabilities, which are a known risk for low number and delayed screening (Rethore et al., 2020;Satge et al., 2014). For instance, several studies suggest that patients with Down syndrome might have difficulty in expressing cancer-related symptoms such as bloody stool, discomfort in diet, fatigue, and body ache due to lower social and communication skills (Mircher et al., 2017;Wiseman et al., 2015).
Provision of seamless transition from pediatric to adult health services is also important. For that purpose, updated guidelines on health care management of adult patients with Down syndrome should be implemented.
These guidelines should be understood not only by caregivers including parents but also the patients themselves so that they are aware of any changes in symptoms for early disease detection (Bull & Committee on Genetics, 2011;D. S. Smith, 2001). Future studies should focus on identifying the clinical features and associated social circumstances such as residence, family, and required care at a population-based level.
This is the first national report on the cause of death in patients with Down syndrome in Japan. The main strength of our study is the utility of national data. However, data obtained from death certificates have several limitations. First, the present study referred to the contents of death certificates and therefore did not include patients with death certificates which did not include Down syndrome as the diagnosis for various reasons. In addition, we cannot deny the possibility that the cause of death might have been misclassified where the true cause of death differed from that stated in the death certificate.
Second, we could not differentiate the types of Down syndrome (nondisjunction, translocation, or mosaicism) in the present study, due to the unavailability of relevant data.

| CONCLUSION
Using Vital Statistic Data, we found that the main cause of death were pneumonia/respiratory infections, congenital malformations of the circulatory system, other diseases of the circulatory system, and aspiration pneumonia in the patients with Down syndrome in Japan in the 2014-2016 period. In addition, the patients with Down syndrome were more likely to die from natural death, early-onset Alzheimer's disease, aspiration pneumonia, pneumonia/respiratory infections, congenital malformations of the circulatory system, and leukemia/lymphoma