INTRODUCTION
The goal of geriatrics is to promote a long, healthy life. In studying the lives and diseases of old population, it is crucial to examine the historical changes in lifespan (or life expectancy [LE]). The analysis of historical life expectancy has been reviewed by numerous scholars in the fields of geriatrics and gerontology.1) Demographic studies of the past are essential for understanding the lives and diseases of ancient populations,2) and average LE over time has been reported by several scholars in paleodemography.3-6)
For decades, bioanthropologists in Japan have also calculated LE using archaeological human remains.7-11) However, a significant limitation of studies based on skeletal specimens is the potential for severe bias, often due to the degradation of bones in the soil over time.12)
As a result, paleodemographic researchers have increasingly turned to historical resources. In Japan, studies on the average LE during the Tokugawa Shogunate period have been conducted by Hayami,13) Nakajima,14) Narimatsu,15) and others, using Edo-period documents such as the Shumon-Ninbetsu-Aratame-chō (宗門人別改帳; The Village People Register of Religious Faith and Relationship), etc. However, these documents mainly focus on commoners and exclude important data on samurai and other upper-class individuals. Additionally, they lack reliable information on infant mortality and other key demographic factors.
In this context, our study of documents related to the Tokugawa Shogun family—who were the highest ruling class during the Edo period—holds significant value. These documents are far more accurate than other sources and provide comprehensive data on the lifespan of Shogun family members, including detailed records of infant births and deaths.
Despite their invaluable demographic potential, these documents have yet to be properly analyzed. The objectives of this study are as follows: (1) to calculate the average LE of Tokugawa Shogun family members based on historical documents, (2) to estimate the paleodemographic patterns of the Edo period using survival curves and rates, and (3) to establish a foundation for applying the results of this study to other demographic data from the Edo period or contemporary Japan.
MATERIALS AND METHODS
Tokugawa shoguns were the de facto head of government and the highest authority in the Edo period, although the emperor's family nominally ranked higher. In this study, we used multiple documents about Tokugawa Shogun family: Tokugawa Jikki (徳川実紀), Shoku-Tokugawa Jikki (続徳川実紀), Tokugawa Bakufu Keifu (徳川幕府系譜), Tokugawa-Bakufu Shoke Keifu (徳川幕府諸家系譜).16) These are top-rated data of Edo period that recorded the birth and death dates of the Shogun Family members. Table 1 summarizes the data of Shogun family members numbered a total of 151 individuals (83 males and 68 females). In our study, the names of individuals are removed from the data set for ethical issues to personal information.
Since the date system used during the Tokugawa Shogunate period was based on the lunar calendar until the Meiji Restoration, we converted the dates in historical records to the solar calendar for a demographic comparison with modern populations. For this conversion, we utilized the Japanese calendar date database from the National Astronomical Observatory of Japan (https://eco.mtk.nao.ac.jp/cgi-bin/koyomi/caldb.cgi). Our study is pioneering in that it is the first to convert the birth and death dates from the lunar to the solar calendar, a step that has not been undertaken in previous paleodemographic studies of Edo period populations. LE was then calculated by sex, and survival curves were generated using the Kaplan-Meier method. Additionally, we evaluated the months of birth and death.
RESULTS
Table 1 summarizes the name, sex, date of birth, date of death, and age at death for members of the Tokugawa Shogun family. The mean LE at birth for males is 20.93±25.03 years, and for females, it is 20.53±25.16 years. The sex-pooled mean LE at birth is 20.75±25.00 years (Table 2). A t-test revealed no significant difference in LE at birth between males and females (p=0.9228).
Survival rates and survival curves, calculated using the Kaplan-Meier method, are presented in Table 3 and Fig. 1, respectively. These analyses reveal that approximately 30% of individuals died at age 0, and nearly 50% had died by age 3. This is visually represented by the age-at-death distributions in Fig. 2A and 2B. Our estimates of LE for Tokugawa Shogun children (n=102) at age 1 are 30.71±24.88 years (30.47±24.92 years for males, 31.02±25.12 years for females), and at age 2 (n=87), the LE is 35.83±23.38 years (34.6±23.8 years for males, 37.51±22.97 years for females). The LE at age 2 closely aligns with the findings of Hayami.17) Fig. 3A and 3B depict the months of birth and death for the Tokugawa Shogun family, respectively, showing a lower mortality rate in winter and a higher rate in summer.
DISCUSSION
Previous Reports on Edo People’s LE
To date, there have been several reports on the LE of Japanese populations prior to the 20th century, covering various groups. In summary, Kobayashi18) estimated the LE for males and females in Toraiwa Village (during the late Edo period) to be 36.8 years and 36.5 years, respectively. In his study, an individual's birth year was set to age 0 using a simple method that downgraded traditional age 1 to age 0. Hayami19) used Edo-period documents (Shumon-Ninbetsuaratame-cho) to estimate the LE of people at the time. While he was unable to calculate LE at birth, he reported the LE at age 2 as 32.0 years for males and 30.0 years for females. Narimatsu15) reported a LE of 37.7 years for males and 36.4 years for females in Niida Village, and a LE of 44.7 years for males and 43.3 years for females in Shimomoriya Village.20) Kalland and Pedersen21) reported an average LE of 44.7 years for males and 43.3 years for females in Senpukuji Village. Based on these reports, the average LE for Edo-period populations is generally estimated to have been in the late 30s to early 40s.
In this study, we estimate that the LE at birth for the Tokugawa Shogun family was around 20 years (male, 20.93 years; female, 20.53 years). This finding is particularly intriguing because the Tokugawa Shogun family was at the height of political and economic power during the Edo period. Given their superior standard of living compared to ordinary people, it is unusual that they might have had a shorter life expectancy than commoners of the time.
However, there are important considerations when comparing our Tokugawa data to previous reports. One key factor is the different methods of age calculation. Traditionally, in Japan, a person’s age was counted as 1 at birth. In the Edo period, according to the lunar calendar, the most significant issue is that there was no concept of "age 0"; individuals were considered to be 1 year old at birth.22) As a result, the life expectancy of Edo-period people may have been overestimated in previous studies. This not only complicates comparisons with modern populations, but also introduces potential inaccuracies in paleodemographic analyses. In our study, we estimate LE by converting all birth dates to the solar calendar and treat the birth year as age 0. We believe this approach largely corrects the errors in previous studies, providing a more accurate representation of life expectancy.
Comparison to Modern Japanese old population
According to the Simple Life Tables published by the Ministry of Health, Labour and Welfare (2023), the LE of modern Japanese is 81.09 years for males and 87.14 years for females, among the highest in the world. The infant mortality rate in modern Japan is approximately 0.2%, meaning about 1.8 deaths per 1,000 live births occur before the age of 1 year.23)
In our study, the infant mortality rate for the Tokugawa Shogun family was much higher: 31.33% for males and 33.85% for females by the end of age 0 (Table 3). This translates to roughly 313 male deaths and 339 female deaths per 1,000 live births before the age of 1. The high infant mortality rate during the Tokugawa period suggests that such rates were likely consistent across pre-modern Japan, as clearly illustrated in the report by Kobayashi.18) In his study, the mortality rates for Toraiwa Village, Nagano Prefecture, during the late Edo period were 49.4% for males (at age 4) and 50% for females (at age 3), indicating that half of all children born in the Edo period did not survive past age 3 or 4.
Following this early period of very high infant mortality, individuals from the Tokugawa Shogun family seem to experience a slower decline in survival rates up to age 10, with the trend continuing into their 70s and 80s (Fig. 1). Since the survival curve after age 10 is nearly linear with a gradual downward slope, it suggests that the mortality rate remained relatively stable throughout this period. According to the classification of survival curves by Pearl and Miner,24) the Tokugawa Shogun family’s survival curve resembles a Type III curve from age 0 to 3 and a Type II curve after age 10. In summary, the death rates of the Tokugawa Shogun family after age 10 were significantly lower than during their infancy.
Causes of Death
In this study, the Tokugawa Shogun family exhibited a lower mortality rate in winter and a higher rate in summer. This pattern is likely due to the fact that, while they were able to stay warm in the elaborately constructed Edo Castle during the winter months, there were no measures to address the heat during the summer. According to Kito,25) in his study of farmers in the Edo and Meiji periods, digestive system diseases were the most common cause of death in summer, while cold weather-related illnesses were the leading cause of death in winter. The causes of death within the Tokugawa Shogun family may not have differed significantly from Kito’s findings. Additionally, it can be inferred that specific causes of death for the Tokugawa Shogun family likely included respiratory infections, bacterial and viral diarrhea, parasitic diseases, and other health issues prevalent throughout the year.2,12,26)
We also consider lead poisoning as a potential cause of death for the Tokugawa Shogun family. Anthropological studies suggest that high-ranking individuals in Edo-period Japan may have been exposed to elevated levels of lead.27-29) But where did this high level of lead come from? One likely source is the cosmetics used during the Edo period. White makeup powder (Oshiroi; 白粉), which is thought to have originated in China, began to be used in Japan as early as the Asuka-Nara period. During the Keicho period (1596–1615), the production of white makeup powder became widespread in Sakai City30) and continued to be used by the Japanese population until the 1970s.31)
We disagree with the notion that the lead in white makeup powder significantly contributed to the high infant mortality rate during the Edo period. This is because children of high-ranking individuals, such as Shoguns and Daimyos, were not breastfed by their noble mothers, who typically used white makeup powder, but rather by nannies from middle- and lower-class warrior families (or sometimes townspeople).32-34) However, for adults, the possibility of lead poisoning as a cause of death cannot be entirely ruled out, particularly for high-ranking females of the Edo period. We believe that further investigation into this issue is necessary in the future.
Conclusion
In this study, we focused on the records of the Tokugawa Shogun family, for which the dates of birth and death are clearly and accurately documented. While records of other high-ranking families, such as the Daimyo or Hatamoto, are often biased, the well-managed records of the Shogun family throughout the Edo period are particularly valuable for estimating the LE of Edo-period Japanese. To eliminate biases caused by the lunar calendar and the traditional concept of age 0, we converted the date records to the solar calendar and counted the birth year as age 0. Our findings reveal that the LE of the Tokugawa family was much lower than that of modern Japanese, despite their status as the supreme elite class during the Edo period. This low LE is primarily attributed to the high infant mortality rate, while the mortality rate after age 10 remained relatively stable.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
