https://orcid.org/0000-0003-4613-2159
Figures
Abstract
Background
The number of people with dementia increases in an aging society; therefore, promoting policies for dementia throughout the community is crucial to creating a dementia-friendly society. Understanding the status of older adults with dementia in each region of Japan will be a helpful indicator. We calculated Dementia-free Life Expectancy and aimed to examine regional disparities and their associated factors.
Methods
We calculated Dementia-free Life Expectancy and Life Expectancy with Dementia for each secondary medical area in Japan based on the Degree of Independence in Daily Living for the Demented Elderly, using data extracted from the Japanese long-term care insurance claims database. We then conducted a partial least squares regression analysis, the objective variables being Dementia-free Life Expectancy and Life Expectancy with Dementia for both sexes at age 65, and explanatory regional-level variables included demographic, socioeconomic, and healthcare resources variables.
Results
The mean estimated regional-level Dementia-free Life Expectancy at age 65 was 17.33 years (95% confidence interval [CI] 17.27–17.38) for males and 20.05 years (95% CI 19.99–20.11) for females. Three latent components identified by partial least squares regression analysis represented urbanicity, socioeconomic conditions, and health services-related factors of the secondary medical areas. The second component explained the most variation in Dementia-free Life Expectancy of the three, indicating that higher socioeconomic status was associated with longer Dementia-free Life Expectancy.
Citation: Yoshikawa M, Goto E, Shin J-h, Imanaka Y (2023) Regional disparities in Dementia-free Life Expectancy in Japan: An ecological study, using the Japanese long-term care insurance claims database. PLoS ONE 18(5): e0280299. https://doi.org/10.1371/journal.pone.0280299
Editor: Sina Azadnajafabad, Tehran University of Medical Sciences, ISLAMIC REPUBLIC OF IRAN
Received: December 26, 2022; Accepted: May 7, 2023; Published: May 25, 2023
Copyright: © 2023 Yoshikawa et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data sets underlying the results described in this paper are available from the following web pages. Dementia-free Life Expectancy and Life Expectancy with Dementia used in this study are publicly available on the web page of the Department of Healthcare Economics and Quality Management Graduate School of Medicine Kyoto University. (http://med-econ.umin.ac.jp/DemFLE_SecondaryMA/) Data we used in aggregating regional variables are available on "e-Stat," a portal site for Japanese Government Statistics. (https://www.e-stat.go.jp/en).
Funding: This work was supported by Ministry of Health, Labour and Welfare Dementia Policy Research Program (Grant Number JPMH20GB1003), ISHIZUE 2022 of Kyoto University, and the JSPS KAKENHI (Grant Numbers JP19H01075, JP23H00448) from the Japan Society for the Promotion of Science (YI). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Today, the world’s population is aging, especially in developed countries, and the number of people with dementia is increasing [1]. Japan is one of the most aged countries, and it has the highest prevalence of dementia (23.3 per 1,000 people) in the world [2], which burdens people with dementia, their communities, and the nation as a whole [3, 4]. With the growing importance of addressing the social problems of dementia, the Japanese government released the "Framework for Promoting Dementia Care" in 2019 [5]. It aims to delay the onset of dementia and create a society where people can live their daily lives with hope, regardless of whether they have dementia. This framework also mentions the need to collect and analyze evidence of effective measures to promote at the community level.
Many studies have reported various healthy life expectancies to assess a population’s health status [6, 7]. For example, dementia-free life expectancy has been calculated in several countries via different methods, revealing that a healthy lifestyle is associated with a larger proportion of remaining years without dementia [8–11]. Those studies calculated dementia-free life expectancy by predicting the country’s prevalence from extracted data, but few have clarified differences among smaller units or regions within a country [12].
Under Japan’s Long-term care insurance (LTCI) system, people aged 65 years or older and persons aged 40 to 65 years with the specified conditions receive a "long-term care need certification" to utilize LTCI services such as home-visit care and daycare services. The level of need for long-term care is determined through a computerized primary judgment and a secondary judgment by a certification committee. Service applicants are then entitled to receive LTCI services according to their level of need. One of the criteria for assessing the level of need for LTCI services is the "Degree of Independence in Daily Living for the Demented Elderly," which evaluates the level of independence considering their dementia-related symptoms and behaviors. We can extract this information from the Japanese long-term care insurance claims database, which covers LTCI users all over Japan, therefore allowing us to make nationwide comparisons among regions in Japan.
This study aimed to calculate and explore the regional difference in Dementia-free Life Expectancy (De-FLE) based on the Degree of Independence in Daily Life for the Demented Elderly. These metrics will enable us to explore whether regional disparities exist in the dementia-related independence period and what factors are associated with these trends. We expect this study to help increase De-FLE and reduce regional disparities.
Method
Data source
We extracted cases from the Japanese long-term care insurance claims database of the Ministry of Health, Labour and Welfare (MHLW) by the Degree of Independence in Daily Living for the Demented Elderly. This database includes information on long-term care certifications and long-term care receipts collected nationwide. We also aggregated regional variables from publicly available data sources, shown in Table 1.
Geographical units
Secondary medical areas (SMAs), which consist of several municipalities, were used as the unit of this analysis. Japan’s Medical Care Act stipulates SMAs are the units that take comprehensive measures to foster cooperation between health care, medical care, and welfare. There are 335 SMAs in Japan, with a mean population of about 4 million in each SMA. We aggregated public statistical data by SMA and utilized it in the analysis. We used data in 2017, but data submission from long-term care insurers became mandatory after 2018. In 2017, some insurers did not submit data or had significantly fewer data. To omit those insurers, we checked the list of long-term care insurers against the specific extraction data and identified 334 SMAs. We excluded one SMA with missing data from the 2015 census owing to the Great East Japan Earthquake and included 333 SMAs in the analysis.
Dementia-free Life Expectancy
In our study, De-FLE refers to the period before dementia symptoms become severe enough to require support from others. We also defined Life Expectancy (LE) with Dementia as the period in which people receive long-term care due to cognitive difficulties, representing the period between life expectancy and De-FLE. To calculate these metrics, we utilized the "Degree of Independence in Daily Living for the Demented Elderly," which was created by the MHLW to facilitate objective and quick assessments by public health nurses, nurses, social workers, and care workers [13]. When the elderly in need apply for LTCI certification, investigators designated by the municipality visit their place of residence and judge their level of independence. Their family physicians also use the criteria to evaluate their independence level. The Degree of Independence in Daily Living for the Demented Elderly classifies applicants into five categories regarding their communication capacity, symptoms, and behaviors (S1 Table). In this study, we defined "unhealthy" as Independence Level II or higher to calculate De-FLE. Those assessed as Independence Level II usually live at home, but since it is challenging for them to live alone as they need support from the people around them, they often use home services in the daytime.
We extracted data from the Japanese long-term care insurance claims database in 2017 provided by the MHLW and calculated De-FLE using the number of independent degrees of daily living for the demented elderly. We defined LE with Dementia as the period between life expectancy and healthy life expectancy. The calculation followed the "Healthy Life Expectancy Calculation Program" published by the MHLW research group [14].
Statistical analysis
We conducted a partial least squares (PLS) regression with SMA-level De-FLE and LE with Dementia for both sexes at age 65 in 2017 as the objective variables and regional factors as explanatory variables. We selected explanatory variables based on previous studies and the World Health Organization Guidelines on Dementia: "Risk Factors for Dementia," and then aggregated them by SMAs, as shown in Table 1 [15].
PLS regression analysis is commonly used in econometrics, and in recent years, a few public health studies have adopted it [16, 17]. This model first extracts the latent components to have the largest covariance with the objective variables, then uses them to perform regression analysis. It can deal with multicollinearity among variables, making it useful for cases with many correlated explanatory variables [18]. We identified the number of latent factors as follows. First, we used leave-one-out cross-validation to identify the number of factors that minimizes the predicted residual sum of squares (PRESS) by this process [19]. The model with the number of latent factors identified here does not often show significant differences from a model with fewer factors. We performed Van der Voet’s test to test whether the difference was significant. Van der Voet’s test is a statistical test that compares the predicted residuals of different models [20]. We finally identified the least number of latent factors among the models that none were significantly different from the model with the smallest PRESS. All statistical analyses were conducted using JMP Pro 16.1 (SAS Institute Inc., Cary, NC).
This study was conducted in accordance with Ethical Guidelines for Medical and Biological Research Involving Human Subjects and the study was approved by the Ethics Committee, Kyoto University Graduate School and Faculty of Medicine (approval number: R0438) with a waiver of informed consent prior to data collection.
Results
Descriptive statistics for SMA-level LE, De-FLE, LE with Dementia, and regional variables
The mean SMA-level De-FLE at age 65 in 2017 was 17.33 years (95% confidence interval [CI] 17.27–17.38 years) for males and 20.05 years (95% CI 19.99–20.11 years) for females, a difference of 2.72 years, with standard deviations of 0.56 years and 0.54 years, respectively (Table 2).
PLS regression analysis
Cross-validation results showed that the PRESS was minimal when 15 factors were extracted. Then, three components were finally identified based on the outcome of Van der Vote’s Test. Factor loadings and the variable importance in projection (VIP) for each of the three factors are shown in Table 3.
The factor loadings in Table 3 show that Component 1 was positively loaded by the proportion of junior college or university graduates, the proportion of commuters traveling by railway or train, and the financial power index. In contrast, Component 1 was negatively loaded by the proportion of persons employed in primary industry and the proportion of elderly in the population. Thus, we named this component urbanicity. Component 2 was positively loaded by the proportion of persons employed in secondary industry and mean household size, while it was negatively loaded by the proportion of single households, tobacco taxes revenue per population, unemployment rate, and the proportion of single-mother households. We defined this as the socioeconomic conditions-related component. Component 3 was characterized by positive loadings on the number of general clinics per 100,000 people and the number of hospital beds per 1,000 people, and thereby named the healthcare resources component. The high score of each component represented the following variables: Component 1, high urbanization; Component 2, good socioeconomic conditions; and Component 3, rich healthcare resources.
Of the three components in Table 3, Component 1 had the largest coefficient of determination for the explanatory variable (31.2%), and Component 2 had the largest coefficient of determination for the objective variable (6.67%), which meant that Component 2 explained De-FLE the best of the three components.
Fig 1(A) shows the correlation and loading plot of variables for Components 1 and 2. De-FLE for males and LE with Dementia for both sexes were positively correlated with the score of Component 1. De-FLE for both sexes was positively correlated with the score of Component 2, especially for males. Some variables with high negative loadings with Component 2 had nearly zero loading on Component 1 (e.g., tobacco revenue per population and unemployment rate). Fig 1(B) and 1(C) show the correlation and loading plots of Components 2/3 and 1/3, respectively. All objective variables were positively correlated with Component 3. The variation in objective variables of Component 2 was much wider than that of Components 1 and 3.
(A) Component 1 and Component 2; (B) Component 2 and Component 3; (C) Component 1 and Component 3. Red triangles indicate the loadings of objective variables, and blue circles indicate the loadings of the explanatory variables. P1, Total population; P2, Proportion of child population; P3, Proportion of elderly population; P4, Mean household size; P5, Proportion of single-mother households; P6, Proportion of population aged 75 and over; P7, Proportion of single households; P8, Proportion of older single households; P9, Proportion of junior college or university graduates; S1, Number of dementia supporters per elderly population; S2, Tobacco taxes revenue per population; S3, Number of Suicides per population; E1, Unemployment rate; E2, Proportion of persons employed in secondary industry; E3, Proportion of persons working in the same municipalities; E4, Number of administrative officers; E5, Financial power index; E6, Proportion of people applied national pension contribution exemption; M1, Number of hospitals per 100,000 people; M2, Number of physicians; M3, Number of dental clinics per 100,000 people; M4, Capacity of elderly nursing facilities; M5, Proportion of welfare institutions services users; M6, Proportion of community-based services users; DFLE(F), De-FLE (Female); DLE (F), LE with Dementia (Female); DFLE(M), De-FLE (Male); DLE(M), LE with Dementia (Male).
Discussion
In this study, we calculated SMA-level De-FLE and LE with Dementia for both sexes and then conducted PLS regression analysis to examine regional disparities in Japan. De-FLE and LE with Dementia varied among SMAs, and the mean SMA-level De-FLE was longer for women than men. Three latent components were identified, which represented urbanicity, socioeconomic conditions, and healthcare resources. Socioeconomic conditions explained De-FLE the best.
Component 1 explained the region’s characteristics well, but its contribution to De-FLE can be described as small based on the coefficient of determination. The score of Component 2 had a positive correlation with De-FLE. Therefore, De-FLE will be longer for persons living in better socioeconomic conditions. Component 2 can be referred to as the most crucial component in explaining De-FLE because it had the largest coefficient of determination for the objective variables of all three components. This result was consistent with the well-known fact that maintaining good socioeconomic status is essential to staying healthy [21]. In addition, an Improvement in living conditions will prolong De-FLE. The score of Component 3 was positively correlated with De-FLE, suggesting that better access to healthcare resources was also positively associated with longer De-FLE. The development of a healthcare delivery system may result in longer De-FLE [22, 23].
The score of Component 1 was positively correlated with LE with Dementia for both sexes. Several studies in the US have reported higher mortality among people living in non-metropolitan than in metropolitan areas [24, 25]. Component 2 had a negative and weak correlation with LE with Dementia, contrary to De-FLE. Socioeconomic conditions may affect De-FLE more than LE with Dementia. The score of Component 3 was positively correlated with LE with Dementia, suggesting that access to healthcare resources was also positively associated with these metrics. A large number of people with dementia might bring about the need to develop a healthcare delivery system, thereby leading to an increase in physicians. However, a causal relationship remains unclear, and further research is warranted.
As discussed above, socioeconomic conditions best explain the variation in De-FLE, compared with urbanicity and healthcare resources. Most of the variables that illustrate Component 2 well had little loadings on Component 1, as shown in Fig 1. For example, variables with high negative loadings on Component 2 were the unemployment rate, the proportion of single households, and tobacco taxes revenue per population. These variables had little loadings on Component 1, meaning that these variables were negatively associated with De-FLE regardless of urbanicity. On the other hand, variables with high positive loadings on Component 2 were the mean household size and the number of dementia supporters per elderly person in the population, implying that they were also positively associated with De-FLE. These also had little loadings on Component 1, which means that they are positively associated with De-FLE in both urban and rural areas. We will further discuss socioeconomic conditions below.
The unemployment rate was a critical variable of Component 2. A previous study suggested a significant relationship between work and self-rated health among older adults [26]. A sense of fulfillment is also necessary because working only for financial gain diminishes the health benefits of working after retirement [27]. Measures to reduce the unemployment rate by sufficiently arranging rewarding jobs might work toward extending De-FLE.
Tobacco taxes revenue per population was also a crucial variable of socioeconomic conditions. The adverse effects of tobacco on overall health, not just dementia, have led to the enforcement of many tobacco control and policy measures [28, 29]. Social factors such as poor housing conditions, low wages, isolated parents, unemployment, and homelessness are closely linked to high rates of smoking and low rates of smoking cessation, and smoking is a significant cause of illness and premature death [30]. Given the above report and our study, smoking will not only directly but indirectly harm a person’s health status and De-FLE.
The number of dementia supporters per elderly person in the population also explains the characteristics of socioeconomic conditions. Previous studies revealed that extensive community resources contribute to longer healthy life expectancy [31, 32]. Improved initiatives, such as introducing dementia supporters, may play a critical role in encouraging government and community residents to better understand people with dementia, which will help prevent isolation among the elderly [33].
Variables representing family structure, such as the proportion of older single households and mean household size, were influential factors in explaining socioeconomic conditions. Previous research has found that household size was negatively associated with mortality from dementia [34]. Another previous study noted that people who live alone are more likely to move into a nursing home than those who live with a family member [35]. Our study suggests that household size may also be negatively associated with the number of people needing care due to dementia. However, it is estimated that the proportion of single households will continue to increase, accounting for 40% of all households by 2040 [36]. Given this social trend, promoting the development of community-based support systems will be imperative to avoid burdening family members and ensure that everyone receives appropriate care.
As for gender differences, the correlation between Component 1 and De-FLE was positive for males and nearly zero for females, which means urbanicity could affect males and females differently. Previous studies on the association between city size and health showed that areas with higher urbanicity had more people aged 100 years and over and a lower risk of physical disability [37–39]. In addition, Component 2, which represents socioeconomic conditions, correlated with De-FLE better for males than females. Several reports have found gender differences in the association between socioeconomic conditions and healthy life expectancy [40, 41]. This study found that socioeconomic conditions explain De-FLE particularly well for males, suggesting the need for community support considering gender.
Strengths and limitations
We examined regional disparities in the time before the elderly needed substantial long-term care. Using LTCI data and official statistical data allowed us to make comparative observations not only for specific regions but also across all regions of Japan. On the other hand, there are some limitations to this study. First, an ecological fallacy, a phenomenon in which observed associations true at the population level do not apply at the individual level, might occur in this study. We considered variables reported to associate with individual-level health status in previous studies and found consistent results. In addition, this study aimed to identify factors associated with regional differences in De-FLE rather than to explore individual-level factors. Furthermore, regional variables, such as the proportion of unmarried people and educational background, may not fully apply to the population over 65. Therefore, we used such regional variables to measure a neighborhood environment rather than directly describing the nature of the older population. Finally, some official statistical data, such as educational background and commuting patterns, were collected from the 2010 Census results, and the situation may have changed since that time. Continuous monitoring with new data is warranted.
Conclusions
This study calculated Dementia-free Life Expectancy at the secondary medical area level as a new indicator for understanding the regional statuses of the elderly with dementia across the entire country of Japan and explored its association with regional variables. We found that socioeconomic conditions correlate with Dementia-free Life Expectancy more than urbanicity and healthcare resources factors. In order to achieve the extension of dementia-free life expectancy and the elimination of its regional disparities, it may be effective for the national and local governments to offer more comprehensive socioeconomic support and make efforts to gain the understanding and cooperation of local communities.
Supporting information
S1 Table. Degree of Independence in Daily Living for the Demented Elderly.
https://doi.org/10.1371/journal.pone.0280299.s001
(DOCX)
References
- 1. Nichols E, Steinmetz JD, Vollset SE, Fukutaki K, Chalek J, Abd-Allah F, et al. Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease 2019. Lancet Public Health. 2022;7(2):E105–E25. WOS:000763474800008. pmid:34998485
- 2. OECD. Care needed: improving the lives of people with dementia 2018. Available from: https://read.oecd-ilibrary.org/social-issues-migration-health/care-needed_9789264085107-en#page1.
- 3. Ikeda S, Mimura M, Ikeda M, Wada-Isoe K, Azuma M, Inoue S, et al. Economic Burden of Alzheimer’s Disease Dementia in Japan. Journal of Alzheimers Disease. 2021;81(1):309–19. WOS:000648474400026. pmid:33780371
- 4.
Ministry of Internal Affairs and Communications. Statistics Bureau Home Page/JAPAN STATISTICAL YEARBOOK 2018—Chapter 2 Population and Households [cited 2022 October 22nd]. Available from: https://www.stat.go.jp/english/data/nenkan/67nenkan/1431-02.html.
- 5.
Ministry of Health, Labour and Welfare. Framework for Promoting Dementia Care2019. Available from: https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000076236_00002.html.
- 6. Mathers CD, Sadana R, Salomon JA, Murray CJL, Lopez AD, World Health Organization, Global Programme on Evidence for Health Policy. Estimates of DALE for 191 countries: methods and results. 2000. https://apps.who.int/iris/handle/10665/67778.
- 7. Naito Y, Goto E, Lin H-R, Hara K, Sasaki N, Imanaka Y. Factors related to active life expectancy across secondary medical areas. Journal of the Japan Society for Healthcare administration. 2020;57(1):2–10.
- 8. Dhana K, Franco OH, Ritz EM, Ford CN, Desai P, Krueger KR, et al. Healthy lifestyle and life expectancy with and without Alzheimer’s dementia: population based cohort study. Bmj-British Medical Journal. 2022;377. WOS:000789657500002. pmid:35418416
- 9. Grasset L, Peres K, Joly P, Sabathe C, Foubert-Samier A, Dartigues JF, et al. Secular trends of mortality and dementia-free life expectancy over a 10-year period in France. European Journal of Epidemiology. 2019;34(2):115–23. WOS:000458429900003. pmid:30649704
- 10. Jagger C, Matthews FE, Wohland P, Fouweather T, Stephan BCM, Robinson L, et al. A comparison of health expectancies over two decades in England: results of the Cognitive Function and Ageing Study I and II. Lancet. 2016;387(10020):779–86. WOS:000370418000037. pmid:26680218
- 11. Moreno X, Lera L, Moreno F, Albala C. Life expectancy with and without cognitive impairment among Chilean older adults: results of the National Survey of Health (2003, 2009 and 2016). Bmc Geriatrics. 2019;19(1). WOS:000511872500002. pmid:31878877
- 12. Farina MP, Hayward MD, Kim JK, Crimmins EM. Racial and Educational Disparities in Dementia and Dementia-Free Life Expectancy. Journals of Gerontology Series B-Psychological Sciences and Social Sciences. 2020;75(7):E105–E12. WOS:000593213100004. pmid:31111926
- 13. Ministry of Health Labour and Welfare. Outline of Survey of Institutions and Establishments for Long-term Care2010 22 Oct 2022. Available from: https://www.mhlw.go.jp/english/database/db-hss/dl/siel-2010-04.pdf.
- 14. Study group of Health Labour Sciences Research. Healthy life expectancy calculation table 2018 [cited 2022 October 22nd]. Available from: http://life.umin.jp.
- 15. World Health Organization. Risk reduction of cognitive decline and dementia: WHO guidelines2019 2022 October 22nd. Available from: https://www.ncbi.nlm.nih.gov/pubmed/.
- 16. Kabir A, Rahman MJ, Shamim A, Klemm RDW, Labrique AB, Rashid M, et al. Identifying maternal and infant factors associated with newborn size in rural Bangladesh by partial least squares (PLS) regression analysis. Plos One. 2017;12(12). WOS:000418564200056. pmid:29261760
- 17. Liu CY, Zhang XL, Nguyen TT, Liu JY, Wu TC, Lee E, et al. Partial least squares regression and principal component analysis: similarity and differences between two popular variable reduction approaches. General Psychiatry. 2022;35(1). WOS:000748914200002. pmid:35146334
- 18. Abdi H. Partial least squares regression and projection on latent structure regression (PLS Regression). Wiley Interdisciplinary Reviews-Computational Statistics. 2010;2(1):97–106. WOS:000214676800012.
- 19. SHAO J. Linear Model Selection by Cross Validation. Journal of the American Statistical Association. 1993;88(422):486–94. WOS:A1993LD68000012.
- 20. Vander Voet H. Comparing the predictive accuracy of models using a simple randomization test. Chemometrics and Intelligent Laboratory Systems. 1994;25(2):313–23. WOS:A1994PT15900016.
- 21. Marmot M, Friel S, Bell R, Houweling TA, Taylor S, Commission on Social Determinants of Health. Closing the gap in a generation: health equity through action on the social determinants of health. Lancet [Internet]. 2008 Nov 08; 372(9650):[1661–9 pp.]. Available from: https://www.ncbi.nlm.nih.gov/pubmed/18994664.
- 22. Kondo N, Mizutani T, Minai J, Kazama M, Imai H, Takeda Y, et al. Factors explaining disability-free life expectancy in Japan: the proportion of older workers, self-reported health status, and the number of public health nurses. Journal of Epidemiology. 2005;15(6):219–27. WOS:000232765300004. pmid:16276031
- 23. Laborde C, Crouzet M, Carrere A, Cambois E. Contextual factors underpinning geographical inequalities in disability-free life expectancy in 100 French departements. European Journal of Ageing. 2021;18(3):381–92. WOS:000582997100001. pmid:34483802
- 24. Cosby AG, Neaves TT, Cossman RE, Cossman JS, James WL, Feierabend N, et al. Preliminary evidence for an emerging nonmetropolitan mortality penalty in the United States. Am J Public Health. 2008;98(8):1470–2. Epub 20080612. pmid:18556611; PubMed Central PMCID: PMC2446448.
- 25. Walker BH, Brown DC. Trends in lifespan variation across the spectrum of rural and urban places in the United States, 1990–2017. SSM Popul Health. 2022;19:101213. Epub 20220822. pmid:36059373; PubMed Central PMCID: PMC9434220.
- 26. Watanabe S, Murayama H, Takase M, Sugiura K, Fujiwara Y. [Longitudinal association between work and self-rated health in older adults: a systematic review]. Nihon Koshu Eisei Zasshi. 2022;69(3):215–24. Epub 20211220. pmid:34924494.
- 27. Nemoto Y, Takahashi T, Nonaka K, Hasebe M, Koike T, Minami U, et al. Working for only financial reasons attenuates the health effects of working beyond retirement age: A 2-year longitudinal study. Geriatrics & Gerontology International. 2020;20(8):745–51. WOS:000544939100001. pmid:32618090
- 28. Llewellyn DJ, Lang IA, Langa KM, Naughton F, Matthews FE. Exposure to secondhand smoke and cognitive impairment in non-smokers: national cross sectional study with cotinine measurement. Bmj-British Medical Journal. 2009;338. WOS:000263976200006. pmid:19213767
- 29. Ohara T, Ninomiya T, Hata J, Ozawa M, Yoshida D, Mukai N, et al. Midlife and Late-Life Smoking and Risk of Dementia in the Community: The Hisayama Study. Journal of the American Geriatrics Society. 2015;63(11):2332–9. WOS:000365678500012. pmid:26503243
- 30. Wilkinson RG, Marmot MG. Social determinants of health: the solid facts, 2nd ed (en).2003. Available from: https://apps.who.int/iris/handle/10665/326568.
- 31. Blocker E, Fry A, Luebbers P, Burns J, Perales-Puchalt J, Hansen D, et al. Promoting Alzheimer’s Risk-Reduction through Community-Based Lifestyle Education and Exercise in Rural America: A Pilot Intervention. Kansas journal of medicine. 2020;13. pmid:32695261.
- 32. Kanamori S, Kai Y, Aida J, Kondo K, Kawachi I, Hirai H, et al. Social Participation and the Prevention of Functional Disability in Older Japanese: The JAGES Cohort Study. Plos One. 2014;9(6). WOS:000338701300084. pmid:24923270
- 33. Sari DW, Igarashi A, Takaoka M, Yamahana R, Noguchi-Watanabe M, Teramoto C, et al. Virtual reality program to develop dementia-friendly communities in Japan. Australas J Ageing. 2020;39(3):e352–e9. Epub 20200602. pmid:32483931.
- 34. You WP, Henneberg M. Large household reduces dementia mortality: A cross-sectional data analysis of 183 populations. Plos One. 2022;17(3). WOS:000776002200011. pmid:35239673
- 35. Mjorud M, Selbaek G, Bjertness E, Edwin TH, Engedal K, Knapskog AB, et al. Time from dementia diagnosis to nursing-home admission and death among persons with dementia: A multistate survival analysis. Plos One. 2020;15(12). WOS:000597150200027. pmid:33275638
- 36. National Institute of Population and Social Security Research. Household Projections for Japan: 2015–2040, Outline of Results and Methods 2018 [cited 2022 October 22nd]. Available from: https://www.ipss.go.jp/pp-ajsetai/e/hhprj2018/t-page_e.asp.
- 37. Fuentes P, Amador S, Lucas-Ochoa AM, Cuenca-Bermejo L, Fernandez-Villalba E, Raparelli V, et al. Sex, rurality and socioeconomical status in Spanish centennial population (2017). Aging-Us. 2021;13(18):22059–77. WOS:000704292100016. pmid:34565717
- 38. Bagavos C. Gender and regional differentials in health expectancy in Greece. Journal of Public Health Research. 2013;2(2):66–73. WOS:000214681100002. pmid:25170483
- 39. Tanaka K, Johnson NE. Social Integration and Healthy Aging in Japan: How Gender and Rurality Matter. Journal of Cross-Cultural Gerontology. 2010;25(2):199–216. WOS:000436655000006. pmid:20440547
- 40. Najar J, Aakre JA, Vassilaki M, Wetterberg H, Ryden L, Zettergren A, et al. Sex Difference in the Relation Between Marital Status and Dementia Risk in Two Population-Based Cohorts. Journal of Alzheimers Disease. 2021;83(3):1269–79. WOS:000703410000025. pmid:34420952
- 41. Okuzono SS, Shiba K, Kim ES, Shirai K, Kondo N, Fujiwara T, et al. Ikigai and subsequent health and wellbeing among Japanese older adults: Longitudinal outcome-wide analysis. Lancet Regional Health-Western Pacific. 2022;21. WOS:000793657100021. pmid:35141667