Because the data used in this study are publicly available on the Japanese government’s official website, ethical review of the data was deemed unnecessary6. The data from 1996 to 2020 was analyzed in this study.
The Comprehensive Survey of Living Conditions, Patient Survey, and Vital Statistics are conducted by the Ministry of Health, Labour, and Welfare, whose details and tabulated data are available on the website of the Statistics Bureau, Ministry of Internal Affairs and Communications, and were retrospectively analyzed in this study7,8,9.
The Comprehensive Survey of Living Conditions
In this cross-sectional national survey, the focus is on randomly stratified samples of households with their members. Beginning in 2001, a questionnaire about long-term care has been administered every 3 years to collect data about people who require long-term care (approximately 6000 individuals), according to the National Census, in 2500 districts. The results of this questionnaire reflect the proportion of individuals certified for long-term care under the Long-term Care Insurance System per 100,000 population (hereafter referred to as the “disability rate”), which is approximately equivalent to the prevalence of disability.
The Patient Survey
In this cross-sectional national survey, the focus is on randomly stratified samples of medical institutions, including hospitals with more than 500 beds and outpatient clinics. Because of the Tohoku earthquake and tsunami on March 11, 2011, the survey conducted in that year did not include data from medical institutions in Fukushima prefecture or in the Ishinomaki and Kesennuma medical areas of Miyagi prefecture.
At each medical institution, the surveys were performed during a designated 3-day period in October of each year. Physicians filled out the questionnaire and collected patients’ information. The International Classification of Diseases, Ninth Revision (ICD-9), was used to categorize diseases and injuries in the surveys until 1995, and the Tenth Revision (ICD-10) was used thereafter10.
For each disease or injury, the “treatment rate” was calculated as the estimated number of patients receiving treatment divided by the estimated total number of patients with the disease or injury and then multiplied by 100,000. The number of patients who continuously received medical care was estimated according to the following formula11:
Estimated number of patients receiving medical treatment = Estimated number of inpatients + Estimated number of outpatients at initial visit + (Estimated number of outpatients at return visit × Average interval since last visit × Adjustment factor [6/7])
The estimated total number of patients thus included those who were not receiving medical care at medical institutions on the date of the survey, and the treatment rate can be considered an approximation of the prevalence of medical conditions.
According to the Comprehensive Survey of Living Conditions in 2019, the most common cause of disability was dementia, followed by cerebrovascular disease and infirmity attributable to aging12. Other causes included bone fractures and falls. In this study, the following medical conditions were investigated because of their clinical significance, prevalence, potential to cause disability, and availability of data: malignant neoplasms, diabetes mellitus, hypertension, ischemic heart disease, cerebrovascular diseases, pneumonia, fractures, osteoporosis, and joint disorders (osteoarthritis and inflammatory polyarthropathies)3,4,5.
Vital statistics
Data on birth, marriage, and death registrations were collected from family registries. The data on total mortality rates and rates of mortality from cerebrovascular disease, heart disease, pneumonia, and malignant neoplasms were obtained. The cause of death was documented from death certificates issued by physicians.
Statistical analysis
We categorized the data by sex and evaluated the trends in disability rate, rates of treatment of medical conditions, and mortality rates in five age groups: 65‒69, 70‒74, 75‒79, 80‒84, and 85–89 years. The linear trend test was performed with R 4.2.0 (The R Foundation, Vienna, Austria) to evaluate the overall increasing or decreasing trends in the rates in each sex, with adjustment for the age groups. If the result of the initial trend test was statistically significant, linear regression was conducted to assess the trend in each age group. Two-sided p values of <0.05 indicated statistical significance.