Research Article |
Corresponding author: Elena V. Selezneva ( evselezneva@hse.ru ) © 2024 Elena V. Selezneva, Oksana V. Sinyavskaya.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Selezneva EV, Sinyavskaya OV (2024) Self-rated health of the older population in Estonia and Russia: the impact of ethnic, cross-country differences and age at migration. Population and Economics 8(1): 52-76. https://doi.org/10.3897/popecon.8.e108855
|
Within the framework of the life trajectory paradigm, factors of poor health in older age may include ethnicity, as well as migration history of an individual. Estonia, with a large share of the Russian population, is a good example to analyze the impact of migration and changes in the ethnic environment on health throughout the life course. The purpose of the study is to assess differences in self–rated health of the older population (50+) living in Estonia and Russia, and identify reasons for these differences. The empirical basis of the study was data of the SHARE survey conducted in Estonia in 2010-2011 and the SAGE survey conducted in Russia in 2007-2010. The sample includes urban population aged 50+ in private households: 2,655 Estonians living in Estonia, 1,478 Russians living in Estonia, and 2,446 Russians living in Russia. The tested ordinal regression models show that the native-born in Estonia have a 39% higher chance of rating their health as good compared to Russians in Estonia, which is associated with differences in educational level in the population aged 50-64, while in the population aged 65+ it is associated with differences in living standards between the native-born and immigrants of the first and subsequent generations. At the same time, Russians from Russia are 70% (population aged 50-64) or 60% (population 65+) less likely to rate their health as good, which is related to the lifestyle and loneliness. Russians aged 65+ in Estonia who moved to the country at the age of 25+ have the same chances. The study negatives the healthy migrant effect identified in young immigrants, and also indicates health behavior and poor quality of social connections as possible reasons for poor health of the older Russian residents.
health, migration history, elderly, ordinal regression, self-rated health, Commonwealth of Independent States
The phenomenon of increased mortality (low life expectancy) in Russia against the background of countries with a high level of economic development is reflected in a number of scientific papers and has received various explanations related to both peculiarities of health behavior and individual characteristics of the population (for example, the level of education) (
In our opinion, studies involving Russian immigrants provide for bridging the gap in the analysis of differences between Russia and other countries in terms of health indicators, as well as the nature of these differences. Comparing the ethnic Russians in the country of origin and another state allows us to separate the reasons for health disparities related to institutions and individual characteristics of citizens, to assess impact of the national culture on health, which can determine both objective differences in health and biases in its subjective assessments. Estonia is one of the countries with a high share of Russians and is eligible for such comparisons. Sample surveys on ageing make it possible to conduct these cross-country and interethnic comparisons by a wider variety of health indicators, since questionnaires for this age group usually contain the widest range of health indicators. Therefore, older population living in Estonia and Russia is the focus of the research presented in the article.
Studies on older immigrants who have just relocated to the country explain the health advantage that the recently arrived usually have by the healthy immigrant effect (HIE): generally, the most active and healthy population of the donor country with possibly a superior health compared to the host country population decides to take a move (
About a third of the Estonian population at the time of data collection for the study were not ethnic Estonians
A long-term stay in the country (the highest migration flow from Russia to Estonia was registered after the end of World War II and until the 1980s and 1990s (
Individual researchers have attempted to analyze health characteristics of the Russian population living in the post-Soviet space outside Russia (
Studies conducted in Estonia also show that the Russian population has poorer health compared to the native-born population
Although Estonia, like Russia, is characterized by stagnation in life expectancy in the period from the 1960s to the end of the 1990s and is still lagging on this indicator from European countries, the local population of the country showed a less significant reduction in life expectancy compared to the Russian population of the country. By 2009, in non-titular ethnic groups residing in Estonia, life expectancy equalled to 67 years in males and 79.4 years in females, which was lower than similar indicators for ethnic Estonians, yet higher than in Russia (
Large-scale comparative studies reflecting differences in health of the population of Estonia and its ethnic groups, as well as Russia, have not been conducted until recently. Estonia, along with Russia, was among the countries studied by
A recent study showed that older Russians residing in Estonia are distinguished from the local population, as well as from the Russian population, by a high prevalence of cognitive impairments (based on the results of the immediate recall (
The purpose of this study is to assess differences in self–rated health of the older population (50+) residing in Estonia and Russia, and to identify reasons for these differences.
The study was conducted within the framework of the life trajectory paradigm, according to which human health is a subtotal of the accumulated throughout life impact of socio-economic situation and individual events that can positively or negatively affect health, including those related to immigration and incomplete adaptation to the conditions of the host society (
Hypothesis 1: the impact of discrimination. From the perspective of these paradigms, the most likely reason for disparities between residents of individual countries, ethnic groups or groups with different migration history (migration groups) is inequalities in education, employment characteristics and standard of living between the local population and migrants, which may also result from the social exclusion of the arrived population.
Faced with discrimination (if not during relocation from Russia to Estonia, then after the country gained independence), Russian immigrants could occupy a lower position in society compared to Estonians: they had difficulties finding employment, occupied less privileged positions in the labor market and as a result, in the 1990s and early 2000s they faced difficulties caused by a lower standard of living and less favorable housing conditions. At the same time, the nature of migration could hardly lead to any significant differences in the level of education between the Russian and Estonian older population, since during the period of active migration, especially the first waves, Estonia, the host country, acted as a territory of active development, rather than an already more developed host country.
Another manifestation of discrimination against immigrants, which in older ages can have a negative impact on health of the arrived, are problems with access to medical care (
Hypothesis 2: the impact of institutions for health maintenance. Access to health services can also cause loss of health of the native-born elderly population if the health system lacks necessary resources to provide assistance to citizens with increased demand for medical care and age-associated diseases or does not allow the older population to overcome barriers to receiving care associated with aging (
Hypothesis 3: the impact of national culture. Studies on the immigrant health confirm the importance of differences in the socio-economic status of the native-born and foreign-born populations, however, demonstrate that this group of factors does not fully explain peculiarities of the well-being of these groups (
Russians in Estonia may find unhealthy daily habits more typical than Russians in Russia, since in the former they may be provoked not only by culturally determined behaviors, but also by difficulties in acculturation or experiences associated with incomplete acculturation.
Official statistical data (Table
Healthy lifestyle indicators and access to medical care in Estonia and Russia in 2010
EU | Estonia | Russia | |
Health-related lifestyle | |||
Age-standardized prevalence of smoking among population aged 15 and over, % (WHO estimates) | |||
- males | 33.4 | 41.7 | 46.9 |
- females | 25.3 | 24.4 | 12.6 |
Consumption of strong alcoholic beverages by population aged 15+, liters of pure alcohol per capita | 2.3 | 5.6 | 5.4 |
Average annual intake of fruit and vegetables, kg per capita | 213 | 184 | 171 |
Age-standardized prevalence of obesity (BMI > 30 kg/m2), % (WHO estimates) | |||
- males | 20.4 | 17.9 | 15.6 |
- females | 20.4 | 20.8 | 26 |
Access to health care | |||
Prevalence of diabetes mellitus, % (2006) | 4.2 | 2.9 | 1.9 |
Total health expenditure, % of GDP | 9.6 | 6.3 | 5.8 |
Direct household expenditure on health care, % of total health expenditure | 16.2 | 18.6 | 43.3 |
Hypothesis 4: the impact of social connections. Social connections, having a significant impact on the older people health (
Russians in Estonia may have poorer health compared to Estonians, and probably compared to Russians in Russia, due to less developed social networks and low satisfaction with both close contacts and communication with people in general, which may be associated with incomplete acculturation in the host country.
Hypothesis 5: the impact of mental health. Another significant factor in self-rated health in older ages is such a component of mental health as a presence/absence of depression (Serby & Yu 2003;
Russians in Estonia may have poorer health compared to Estonians and probably compared to Russians in Russia, due to a greater prevalence of depression associated with acculturation stress in the host country.
Thus, the theoretical model of the study included both factors that can determine the increased vulnerability of an immigrant to health deterioration (related to residence in a donor country or integration into life in a new environment) and specific parameters that worsen health at an older age, regardless of migration history.
The empirical basis of the study was a combined set of data of two surveys on older people living in private households. Data of the first wave of the European Survey on Health, Ageing and Retirement in Europe (SHARE), conducted in Estonia in 2010-2011, cover the native-born population (Estonians) and the Russian population of the country of older age (50+). Together with the selected respondents aged 50+, their partners participated in the survey as well. The study is based on a random stratified
Since the respondent was free to choose the preferred language of the survey, Russian or Estonian, the survey probably reflects the distribution of the Estonian population of older ages by ethnicity as well – poor command of the Estonian language could hardly prevent the Russian-speaking population in Estonia to take part in the survey. The response rate at the household level equaled to about 60% (
Data of the Survey on Global Ageing and Adult Health (SAGE) conducted by the World Health Organization in 2007-2010 make it possible to assess health status of the elderly in Russia and compare health of Russians living in Estonia with it. The surveyed SAGE population is based on a sample from the World Health Survey (WHS) in 2003 and an additional set of respondents to ensure representativeness of the key socio-demographic indicators in line with the 2002 census data.
The SAGE sample is represented by both a subsample of the older population and a small subsample of the population aged18-49, which was excluded from the analysis. The response rate added up to 71.8% (WHOc).
In order to ensure comparability of ethnic groups, the rural population was excluded from the consolidated data set. Russians in Estonia are concentrated mainly in cities, and ethnic Russian dwellers of Estonian villages included in the sample are a small group, the data on which may be biased.
The final dataset for the analysis included 6,579 observations: 2,655 Estonians residing in Estonia, 1,478 Russians living in Estonia and 2,446 Russian respondents living in Russia.
Although more relevant surveys allow us to compare the three ethnic groups under study in terms of self-rated health and individual predictors of health, for example, the European Social Survey (ESS) in 2014 with the participation of two countries, the 8th wave of SHARE in Estonia (2020) and the National Study of the Older Generation conducted using a comparable methodology in Russia in 2021 (Human Capital…), in our opinion, the database used is preferable, since it provides a larger sample size and allows to track more Estonian citizens of Russian descent alive.
A regression analysis was used to test the study hypotheses on the origins of differences between the ethnic groups under study; ordinal regression models were constructed with health indicator being a dependent variable. At the first step, one independent variable was included in the model – the ethnic group. Further, other independent variables were introduced one after the other. A decrease in the significance level of the regression coefficient for an ethnic group below 0.05 was regarded as a situation where all predictors included in the model fully explain the differences between ethnic groups.
The leading indicator of health in the study is self–rated health - an indicator that reflects a general condition of a person. Despite the subjectivity of such assessments, the research results show that they still allow for a fairly accurate assessment of an individual’s well-being, including undiagnosed health problems, risk of death, possible deterioration of health in the future and demand for inpatient and nursing care (Idler & Benyamini 1997).
The limitation of self-rated health is its dependence upon cultural context: representatives of certain age, social and ethnic groups may understand “good” or “poor” health in a different way, since they take into account different aspects of their well-being and use different standards for health rating (Krause & Jay 1994; Idler & Benyamini 1997;
Ultimately, according to researchers, self-rated health is a relatively reliable indicator of the respondent health status, even if it is used in studies involving interethnic comparisons (
Ethnicity in this study was determined based on data on the respondent’s country of residence and the respondent’s self-identification of the ethnic group affiliation, regardless of whether they have Estonian or Russian citizenship. The part of the consolidated dataset related to included only those who considered themselves ethnic Estonians or Russians, and respondents who identified themselves as Russians were included in the subsample of the consolidated array for Russia.
In the initial models (not presented in the article) with the three groups discussed above (Estonians, Russians in Estonia and Russians in Russia), using the available predictors it was possible to fully explain the differences only between Estonian and Russian males residing in Estonia. Differences in self-rated health between them disappeared after including age, family composition, financial situation, employment status and educational level, as well as socio-professional status in the current or previous employment in the model. Less significant differences between ethnic groups were obtained when a variable was included in the model in which Russians residing in Estonia were divided into subgroups by age at migration, as well as when regression models in different groups were tested by age rather than gender: respondents aged 50-64 and respondents 65 +. The division into these two age groups allows, first, to separate migration flows that differ in characteristics of the arrived, and second, the causes of health deterioration at different ages (for example, problems with access to medical care may have more significant consequences for people over 65).
Simultaneous separation of the sample by gender and age for the regression analysis was impossible due to a small number of observations in the subgroups of the Russian population in Estonia.
Gender was included in the regression models as a control variable that can determine both objective differences in health and differences in health rating by men and women (Idler & Benyamini 1997). It was added to the model in the first step, along with the “ethnic group” variable. Further, control variables such as age (the number of completed years within the age group under under), household type (whether an elderly person lives alone or with relatives) and marital status were included in the model. At the next stages of the regression analysis, variables were introduced into the model to control the impact of the key health factors such as:
Although this study is based on the life cycle paradigm, emphasizing the importance of characteristics and events that could affect health at different stages of life, regarding the empirical basis used, the influence of not all significant events can be controlled, even though indirect indicators. Thus, success of acculturation of the Russian population in Estonia during relocation to the country, in adulthood, or in the early 1990s, when independence of the state was declared, can be evaluated to the extent that the events taking place at that time influenced the level of education and success in the labor market, partly on social communication and mental health (presence of depression) at the time of the study. Outside of this work, there are events and, possibly, facts of discrimination based on ethnicity that have taken place long before the survey. It is also impossible to control selectivity of the Russian population in Estonia associated with the return of some part of this ethnic group to Russia.
The SHARE and SAGE data show (Fig.
Self-rated health of the older population in Estonia and Russia. Source: authors’ calculations based on SHARE (2010-2011) and SAGE (2007-2010) data
A possible impact of the socio-economic situation on health is indicated by the fact that the profile of ethnic groups in terms of self-rated health repeats the one observed when comparing these groups in terms of material security and employment status (Table
Differences by socio-demographic characteristic and health-affecting characteristic
50-64 years old | 65 years old and older | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Estonians | Russians born in Estonia | Russians relocated at ages 0-17 | Russians relocated at ages 18-24 | Russians relocated at the age of 25 and over | Russians in Russia | Estonians | Russians born in Estonia | Russians relocated at ages 0-17 | Russians relocated at ages 18-24 | Russians relocated at the age of 25 and over | Russians in Russia | |
N | 1033 | 271 | 151 | 170 | 87 | 1174 | 1622 | 87 | 164 | 268 | 280 | 1272 |
Females, % | 58.7 | 59.8 | 58.3 | 60.0 | 59.8 | 62.9 | 62.0 | 70.1 | 64.6 | 61.6 | 62.1 | 71.5 |
Age, years | 57.8 | 57.1 | 58 | 58 | 59.2 | 56.6 | 74.6 | 74.9 | 73.1 | 75.1 | 75.5 | 74.3 |
(SD) | 4 | 4.1 | 4.2 | 4 | 3.5 | 4 | 6.5 | 7.1 | 4.8 | 6.2 | 5.7 | 6.2 |
Living alone, % | 16.7 | 18.8 | 17.9 | 17.1 | 8.0 | 17.6 | 31.9 | 33.3 | 26.8 | 27.6 | 31.1 | 41.3 |
Marital status, % | ||||||||||||
- married/ partnered | 64.9 | 67.2 | 71.5 | 72.9 | 82.8 | 67.1 | 57.5 | 56.3 | 60.4 | 62.7 | 60.4 | 40.1 |
- separated/divorced/never married | 27.9 | 25.5 | 23.2 | 17.1 | 13.8 | 18.7 | 16.7 | 18.4 | 14.0 | 10.8 | 10.0 | 6.2 |
- widowed | 7.3 | 7.4 | 5.3 | 10.0 | 3.4 | 14.1 | 25.8 | 25.3 | 25.6 | 26.5 | 29.6 | 53.7 |
Hardly making both ends meet, % | 9.7 | 21.8 | 23.2 | 24.7 | 19.5 | 29.9 | 5.7 | 21.8 | 17.1 | 16.4 | 14.6 | 28.2 |
Home owners, % | 60.2 | 64.2 | 58.9 | 56.5 | 55.2 | 90.4 | 58.9 | 65.5 | 56.7 | 57.8 | 61.8 | 91.2 |
Employment status, % | ||||||||||||
- employed | 76.7 | 72.3 | 60.3 | 65.9 | 62.1 | 59.5 | 15.2 | 13.8 | 6.1 | 7.1 | 6.1 | 8.7 |
- retired | 12.2 | 17.3 | 25.2 | 27.1 | 29.9 | 27.6 | 84.1 | 85.1 | 93.9 | 91.4 | 92.1 | 81.8 |
- other | 11.1 | 10.3 | 14.6 | 7.1 | 8.0 | 12.9 | 0.7 | 1.1 | 0 | 1.5 | 1.8 | 9.4 |
Years of education | 12.8 | 12 | 12.1 | 12.3 | 13 | 12.7 | 11.4 | 9.7 | 10.5 | 10.1 | 10.9 | 10 |
(SD) | 3 | 2.7 | 2.3 | 2.6 | 2.9 | 2.8 | 3.8 | 3.3 | 3.5 | 3.6 | 3.9 | 3.8 |
Socio-professional status, % | ||||||||||||
- law makers, top officials, top and mid-level managers, military servants | 17.6 | 8.1 | 13.2 | 8.8 | 13.8 | 29.6 | 15.2 | 9.2 | 12.2 | 9.0 | 17.5 | 22.0 |
- top-qualified specialists | 19.9 | 17.3 | 21.2 | 22.9 | 21.8 | 3.4 | 20.6 | 14.9 | 18.9 | 15.3 | 20.0 | 3.0 |
- mid-level specialists, officials, office and customer service specialists, sales and service workers | 20.7 | 22.1 | 11.9 | 12.9 | 16.1 | 17.4 | 18.1 | 23.0 | 9.8 | 13.4 | 10.7 | 15.0 |
- skilled manual workers | 17.1 | 28.4 | 33.1 | 30.0 | 27.6 | 27.9 | 17.8 | 24.1 | 36.0 | 28.0 | 22.9 | 36.0 |
- skilled workers using machines and machinery | 6.3 | 6.3 | 2.6 | 5.9 | 6.9 | 9.9 | 7.2 | 12.6 | 7.9 | 9.7 | 5.7 | 8.3 |
- unskilled workers | 18.3 | 17.7 | 17.9 | 19.4 | 13.8 | 11.9 | 21.1 | 16.1 | 15.2 | 24.6 | 23.2 | 15.7 |
Mother’s education level, % | ||||||||||||
- below secondary | 47.7 | 52.4 | 55.6 | 47.6 | 39.1 | 36.1 | 62.3 | 56.3 | 58.5 | 60.8 | 55.7 | 65.3 |
- secondary, specialized secondary and higher | 30.3 | 22.9 | 26.5 | 24.7 | 36.8 | 60.1 | 16.4 | 14.9 | 15.9 | 8.2 | 13.6 | 27.4 |
- no data | 22.0 | 24.7 | 17.9 | 27.6 | 24.1 | 3.7 | 21.3 | 28.7 | 25.6 | 31.0 | 30.7 | 7.4 |
Father’s education level, % | ||||||||||||
- below secondary | 41.3 | 39.5 | 39.1 | 40.6 | 40.2 | 29.5 | 56.1 | 42.5 | 43.3 | 50.0 | 42.9 | 49.1 |
- secondary, specialized secondary | 23.6 | 26.2 | 26.5 | 20.0 | 19.5 | 52.3 | 14.4 | 16.1 | 12.8 | 9.3 | 13.9 | 31.7 |
- higher | 6.9 | 4.8 | 7.3 | 7.6 | 13.8 | 8.9 | 4.0 | 3.4 | 6.1 | 4.5 | 3.2 | 5.2 |
- no data | 28.2 | 29.5 | 27.2 | 31.8 | 26.4 | 9.3 | 25.5 | 37.9 | 37.8 | 36.2 | 40.0 | 14.0 |
Frequency of alcohol consumption in recent 3 months, % | ||||||||||||
- never | 26.0 | 29.9 | 26.5 | 25.9 | 36.8 | 9.8 | 49.6 | 58.6 | 46.3 | 51.5 | 55.7 | 22.2 |
- sometimes/no data | 45.8 | 48.7 | 53.0 | 54.7 | 48.3 | 81.1 | 36.9 | 34.5 | 44.5 | 42.2 | 37.5 | 73.3 |
- often (once a week and more) | 28.2 | 21.4 | 20.5 | 19.4 | 14.9 | 9.1 | 13.5 | 6.9 | 9.1 | 6.3 | 6.8 | 4.4 |
History of smoking, years | 13.5 | 13.7 | 16.4 | 12 | 12.1 | 6.2 | 9.2 | 6.8 | 9.7 | 9.7 | 8.8 | 3.6 |
(SD) | 15.6 | 15.9 | 17.2 | 15.1 | 16 | 13.1 | 16.5 | 14.3 | 18.1 | 17.8 | 17.4 | 13 |
Intense physical activity rate % | ||||||||||||
- less than once a week/never | 37.5 | 48.7 | 53.6 | 44.7 | 51.7 | 64.8 | 58.6 | 79.3 | 73.8 | 76.5 | 78.9 | 80.2 |
- once a week | 19.1 | 12.9 | 14.6 | 13.5 | 17.2 | 3.2 | 11.3 | 5.7 | 11.6 | 9.3 | 12.1 | 3.1 |
- more often than once a week | 43.5 | 38.4 | 31.8 | 41.8 | 31.0 | 32.0 | 30.1 | 14.9 | 14.6 | 14.2 | 8.9 | 16.7 |
Moderate physical activity rate, % | ||||||||||||
- less than once a week/never | 12.0 | 17.3 | 19.2 | 11.2 | 14.9 | 17.0 | 23.9 | 41.4 | 24.4 | 33.6 | 30.7 | 31.7 |
- once a week | 12.6 | 13.7 | 9.9 | 14.1 | 14.9 | 2.6 | 13.7 | 6.9 | 20.7 | 13.8 | 12.5 | 3.4 |
- more often than once a week | 75.4 | 69.0 | 70.9 | 74.7 | 70.1 | 80.4 | 62.4 | 51.7 | 54.9 | 52.6 | 56.8 | 64.9 |
BMI, kg/м2 | 27.2 | 27.5 | 27.6 | 28.7 | 26.7 | 28.6 | 26.7 | 25.6 | 28.1 | 27.8 | 27.2 | 28.2 |
(SD) | 6.4 | 6.8 | 7.5 | 5.6 | 6.7 | 6.2 | 6.5 | 8.1 | 6.9 | 6.7 | 7.9 | 6.1 |
Signs of depression, % | 33.8 | 39.9 | 37.1 | 42.9 | 36.8 | 38.5 | 37.2 | 56.3 | 56.7 | 45.9 | 51.4 | 47.2 |
Does not show satisfaction with relations with loved ones, % | 9.4 | 10.7 | 10.6 | 11.2 | 12.6 | 16.4 | 10.7 | 16.1 | 7.9 | 9.7 | 10.7 | 24.7 |
Does not report receiving help from loved ones, % | 86.7 | 87.1 | 89.4 | 92.9 | 87.4 | 85.0 | 74.9 | 73.6 | 78.7 | 77.2 | 68.9 | 73.9 |
Thinks people can be trusted, % | 86.6 | 89.3 | 90.1 | 90.0 | 80.5 | 65.8 | 87.1 | 80.5 | 85.4 | 87.3 | 86.8 | 71.8 |
Number of hospitalizations for 12 months | 0.2 | 0.2 | 0.2 | 0.2 | 0.2 | 0.2 | 0.3 | 0.3 | 0.2 | 0.2 | 0.3 | 0.3 |
(SD) | 0.6 | 0.6 | 0.6 | 0.6 | 0.6 | 0.6 | 0.8 | 0.8 | 0.7 | 0.6 | 0.9 | 0.6 |
The profile of people with bad habits is the opposite of the one observed in people with poor health: groups with better (according to self-ratings) health rather than poorer health are more likely to drink alcohol and have a longer history of smoking. To a greater extent, the differences in self-rated health correspond to the profile of the groups according to BMI and intense physical activity.
Higher subjective health ratings of Estonians in older age may be associated with a lower prevalence of depression. These differences are especially pronounced in the age group over 65. On the contrary, Russians residing in Estonia, have the highest rates.
Contrary to expectations, the group with the lowest quality of social connections were Russians living in Russia rather than in Estonia. In the first group we observe the largest share of those who do not state they are satisfied with relations with loved ones, as well as the lowest level of depersonalized trust.
Differences in the frequency of hospitalizations in the groups under study turned out to be insignificant, suggesting an equal access to primary medical care across the groups.
The ordinal regression model for self-rated health among individuals aged 50-64 (Table
Odds ratios for ethnic groups in the ordinal regression models for self-rated health with a different set of predictors
M0 | M1 | M2 | M3 | M4 | M5 | M6 | M7 | M8 | M9 | M10 | M11 | M12 | M13 | M14 | M15 | M16 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
OR | OR | OR | OR | OR | OR | OR | OR | OR | OR | OR | OR | OR | OR | OR | OR | OR | |
50-64 years (control group – Russians born in Estonia) | |||||||||||||||||
- Estonians | 1.392** | 1.487** | 1.504** | 1.358* | 1.343* | 1.260 | 1.259 | 1.261 | 1.246 | 1.243 | 1.150 | 1.144 | 1.122 | 1.120 | 1.130 | 1.128 | 1.113 |
- Russians relocated at ages 0-17 | 0.711 | 0.759 | 0.748 | 0.771 | 0.867 | 0.846 | 0.828 | 0.819 | 0.798 | 0.808 | 0.817 | 0.814 | 0.783 | 0.784 | 0.772 | 0.773 | 0.763 |
- Russians relocated at ages 18-24 | 1.021 | 1.098 | 1.061 | 1.124 | 1.120 | 1.074 | 1.080 | 1.081 | 1.053 | 1.043 | 0.989 | 1.009 | 1.034 | 1.041 | 1.019 | 1.020 | 1.026 |
- Russians relocated at the age of 25 and over | 0.941 | 1.123 | 1.061 | 1.119 | 1.128 | 1.024 | 1.019 | 1.012 | 1.042 | 1.036 | 1.037 | 1.013 | 1.024 | 1.030 | 1.039 | 1.028 | 1.028 |
- Russians in Russia | 0.815 | 0.788 | 0.777 | 0.804 | 0.916 | 0.857 | 0.856 | 0.803 | 0.716* | 0.688* | 0.728* | 0.743* | 0.728* | 0.752 | 0.757 | 0.732* | 0.726* |
N | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 | 2881 |
Pseudo-R2 | 0.009 | 0.022 | 0.026 | 0.045 | 0.106 | 0.113 | 0.117 | 0.118 | 0.123 | 0.124 | 0.145 | 0.147 | 0.1717 | 0.173 | 0.177 | 0.177 | 0.193 |
Cut 1 | -1.624 | -6.027 | -6.172 | -5.663 | -4.668 | -3.563 | -3.614 | -3.560 | -3.119 | -3.254 | -2.243 | -2.740 | -3.196 | -3.245 | -2.694 | -2.781 | -2.922 |
Cut 2 | 0.838 | -3.515 | -3.645 | -3.050 | -1.762 | -0.626 | -0.660 | -0.602 | -0.138 | -0.270 | 0.837 | 0.352 | 0.025 | -0.017 | 0.552 | 0.466 | 0.406 |
65+ (control group – Russians born in Estonia) | |||||||||||||||||
- Estonians | 1.608* | 1.589* | 1.595* | 1.437 | 1.428 | 1.301 | 1.293 | 1.241 | 1.198 | 1.198 | 0.962 | 0.986 | 0.861 | 0.859 | 0.886 | 0.889 | 0.863 |
- Russians relocated at ages 0-17 | 1.080 | 0.934 | 0.938 | 0.931 | 1.030 | 0.986 | 0.991 | 0.979 | 0.922 | 0.922 | 0.844 | 0.879 | 0.913 | 0.890 | 0.885 | 0.888 | 0.829 |
- Russians relocated at ages 18-24 | 0.771 | 0.769 | 0.777 | 0.767 | 0.823 | 0.803 | 0.811 | 0.805 | 0.772 | 0.772 | 0.726 | 0.759 | 0.681 | 0.662 | 0.651 | 0.654 | 0.610 |
- Russians relocated at the age of 25 and over | 0.666 | 0.678 | 0.688 | 0.656 | 0.711 | 0.651 | 0.647 | 0.655 | 0.643 | 0.643 | 0.574* | 0.593* | 0.561* | 0.553* | 0.570* | 0.572* | 0.549* |
- Russians in Russia | 0.793 | 0.734 | 0.755 | 0.732 | 0.825 | 0.804 | 0.817 | 0.731 | 0.592* | 0.592* | 0.574* | 0.598* | 0.527** | 0.560* | 0.576* | 0.571* | 0.552* |
N | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 | 3689 |
Pseudo-R2 | 0.020 | 0.047 | 0.026 | 0.061 | 0.084 | 0.113 | 0.094 | 0.118 | 0.112 | 0.112 | 0.151 | 0.152 | 0.1871 | 0.193 | 0.199 | 0.2 | 0.207 |
Cut 1 | -0.617 | -6.131 | -6.035 | -5.726 | -5.644 | -4.284 | -4.447 | -4.544 | -3.457 | -3.456 | -1.038 | -1.726 | -2.399 | -2.461 | -1.554 | -1.600 | -1.810 |
Cut 2 | 1.864 | -3.548 | -3.450 | -3.094 | -2.897 | -1.503 | -1.660 | -1.731 | -0.588 | -0.587 | 2.007 | 1.327 | 0.819 | 0.779 | 1.711 | 1.664 | 1.483 |
In the initial model for individuals aged 65+, with gender and ethnic group, Estonians also rate their health higher compared to Russians born in Estonia and other groups of the Russian population under study. However, at these ages, the gap between Estonians and Russians is more significant - the odds ratio adds up to about 60%. Statistical significance of these differences disappears after adding a subjective assessment of financial situation (M3) and housing ownership into the model.
In both age groups, 50-64 years and 65+, after including alcohol consumption (M8) and smoking (M9) in the model, differences in self-rated health between Russians born in Estonia and Russians in Russia become significant. Russians residing in Russia tend to rate health poorer than Russians born in Estonia, given similar parameters of unhealthy behavior.
With inclusion of the indicators of motor activity (M10) and BMI (M11) in the model, these differences increase in the age group of 50-64 years, remaining unchanged in the group 65+, however, the differences between Russians born in Estonia and Russians relocated to Estonia at the age over 25 become also significant. In the models for the 50-64-year group, these differences disappear if variables characterizing the quality of social contacts are included in the model (M13-M14), while in the models for the 65+ group they remain after adding variables on social contacts and accessibility of medical care (M16).
The study finds that according to the self-rated health in general, Russians residing in Estonia hold an intermediate position between Estonians who are more satisfied with their health at an older age and Russians in Russia who are less satisfied with their health. This result correlates with differences in life expectancy among representatives of the three ethnic groups under study and confirms that the subjective assessment of health, despite some influence of cultural characteristics, reflects key differences in the well-being of different ethnic groups and can be used in studies on the migration impact on population health and cross-country comparisons. However, the use of self-rated health as a health indicator could result in some underestimation of the extent of the differences in health status between the ethno–migration groups under study due to excessively high health ratings by categories of the population with a low health concern and therefore poorer health (men, Russian population in Estonia and Russia, citizens with low income and low level of education and socio-professional status).
The distributions of self-rated health for Estonian residents (Estonians and Russians) and residents of Russia obtained on the basis of SHARE and SAGE data indicate that in the age group of 65+, health disparities between ethnic groups and groups with different history of migration are more pronounced compared to the 50-64 age group.
If in the younger age group, only differences in the share of people with (very) good health are noticable, then in the older group the disproportion is also manifested in the distribution of those who consider their health to be (very) good or average. These differences are probably the result of the accumulated environmental influences or differences in life trajectories, including those related to migration. Research indicates that health in older age reflects the accumulated effects of factors a person has been exposed to during the life course (Smith & Kington 1997).
Differences in self-rated health between Russians residing in Russia and Estonia are a complex phenomenon that could not be fully explained by the regression analysis. However, the results indicating that these cross-country gaps in health status are determined by differences in health-related lifestyle, as well as low quality of social contacts in individuals aged 50-64. Overweight and insufficient physical activity, as well as bad habits, turn out to be significant lifestyle problems of the elderly in Russia, determining the lag in health indicators compared to the population of Estonia. A strong impact of physical activity and smoking on health of the elderly in Russia is identified by A. Selivanova and J. Cramm (Selivanova & Cramm 2014). The hypotheses of the block 3 are also confirmed: peculiar characteristics of health-related lifestyle are a significant reason for differences in self-rated health between ethno-migration groups.
Another reason for poorer health among Russians residing in Russia compared to Russians in Estonia, may be the fact that living in their ethnic homeland rather than a different cultural environment, they are less likely to be satisfied with their relations with loved ones, and also have difficulties trusting others. Researchers dealing with the problem of social connections in Russia believe that older people in Russia (primarily women) may more often experience loneliness as a result of early bereavement of a spouse (this is the result of a significant gender gap in life expectancy) (
Although linear distributions show that the share of people with depression is increased in certain groups of the Russian population in Estonia, this health predictor in the regression models explaining differences in health between ethno-migration groups, turned out to be insignificant – thus, the hypothesis 5 has been rejected. Based on the results of the analysis of the linear distributions and regression analysis, the hypothesis about the impact of differences in access to primary care on self-rated health in the ethno-migration groups has been rejected as well. The hypotheses 1 and 2 (regarding discrimination in receiving medical care) failed to be confirmed.
Following the work on cognitive abilities (
The results of both the linear distribution analysis and regression analysis suggest the impact of age and generation of immigration on health in old ages. If Russians born in Estonia (second-generation immigrants) or those relocated to the country as children are closer to native-born Estonians in terms of self-rated health, then those who immigrated to Estonia at the age over 24 resemble the population of Russia in a high level of dissatisfaction with health and increased risks of poor health. The similarity of the risks of poor health for those relocated to Estonia at the age over 24 and Russians in Russia is registered only in the regression model for the population aged 65+, rather than in the model for the population aged 50-64. There are two possible explanations for this. First, it can be assumed that the influence of individual events in life trajectory in Russia manifests only at the very advanced age, when a person starts losing the ability to self-care.
Second, we have very limited information about the older generations, which probably does not allow us to take into account differences in factors of the life trajectory of older and younger generations of immigrants: perhaps socialization of young people, primarily with regard to health standards, in Estonia and Russia before the end of the 1960s had more significant differences than in subsequent years.
Assessment of the impact of age at migration on health, according to the followers of this line of research, is an understudied phenomenon. However, it may be promising in studying the effects of immigration on health of the elderly population, since it reflects peculiar features of life trajectory that are important for understanding health in old ages (
The impact of the first mechanism in Estonia may be limited by a strong role of the State in stimulating migration. However, the second mechanism may explain good health of Russians born in Estonia or those relocated in childhood by higher integration into Estonian society (for example, due to the Estonian language acquisition in childhood) by the beginning of the transformations in the early 1990s and a lower level of stress associated with acculturation during this period.
The third mechanism may explain the increased risks of poor health in older ages among Russian-Estonians relocated in adulthood with unhealthy habits developed during their life in Russia.
Existing studies on health of the second generation of immigrants, as far as we know, are focused on populations with the healthy migrant effect, and their main result is the conclusion that the descendants of immigrants, unlike their parents, do not have any health advantages (
Studies show that on average, the native-born in Estonia have better access to material resources than the Russian population of the country, which is probably due to differences in the level of proficiency in the Estonian language, different educational trajectories and inequality in chances for obtaining citizenship (
The conducted research has both strengths and limitations. On the one hand, it is implemented on a unique empirical basis, making it possible to track a large number of Estonian residents immigrated to the country in the 20th century, and to conduct cross-country and interethnic comparisons of health status with their participation by a wide range of demographic and socio-economic indicators. The study contributes to understanding the impact of immigration on health of the elderly, as well as international comparisons to health-related issues among the elderly.
The study limitations include as follows: (1) the problem of self-selection and underestimation of individual health factors (individual risk factors, for example, alcohol abuse at a younger age can result in premature death in middle age and, as a consequence, selection of survivors; we cannot unambiguously determine whether this selection has an ethnic gradient), (2) the cross-sectional nature of the data, making it impossible to separate the ageing effect from the cohort effect, (3) the lack of comparable indicators in SHARE and SAGE for individual reasons for intergroup differences (scales for fixing restrictions in everyday life, nutrition characteristics, direct indicators of the primary care accessibility, comparable typology of urban settlement by size), and (4) the lack of data in SHARE on the problems of integration into the new Estonian society, which the Russians faced in the early 1990s.
The research has both theoretical and practical value. The analysis results show, first, a significant impact of material well-being, and therefore, social support programs, on health of the elderly. However, since this effect is cumulative, it is unlikely that such programs focused only on the elderly will be effective – a decent standard of living should be provided throughout the life course, as indicated by experts (WHOa). Second, the study confirms the importance of lifestyle - not so much bad habits, but a balanced diet and level of physical activity - for the state of health in older ages, as well as a rather tangible request for the development of measures on quality of nutrition and obesity control in the Russian population.
The study demonstrates that analysis of the older population, as well as the long-term consequences of immigration of the Russian population to a former republic of the USSR during the Soviet period, largely regulated by the state, fails to identify the healthy migrant effect. However, Russians can “benefit” from living outside their homeland: if they arrived in another country during the period of active socialization (in childhood or early youth) or were born in this country, they can adopt a higher health concern typical of the host country. According to self-rated health, as well as life expectancy, the Russian population aged 50+ in Estonia holds an intermediate position between the native-born population in Estonia and the population in Russia. At the same time, this group itself is heterogeneous, and the study has attempted to make a distribution by age at migration. Further studies can focus on analyzing the group composition, for example, by period of immigration to Estonia, and the impact on health.
Comparisons of Russians residing in Estonia and Russians in Russia should more correctly consider the type or size of an urban settlement in which individuals from those two countries reside; the population in Russia as a base for comparison is also very heterogeneous. The studies indicate significant and increasing inter-settlement differences in health of the residents in Russia (
ESS. European Social Survey. URL: https://www.europeansocialsurvey.org/ (accessed: 29.06.2023)
Human Capital Multidisciplinary Research Center. National Study of the Older Generation. URL: https://ncmu.hse.ru/elderly-study (accessed: 29.06.2023)
Reile R, Markina A (2010) Health inclusion. Perspectives of the providers on participation of migrants in health promotion in Estonia. URL: https://dspace.ut.ee/handle/10062/85291 (accessed: 29.06.2023)
Statistics Estonia (2011) Population and housing census 2011. URL: https://andmed.stat.ee/en/stat/rahvaloendus__rel2011 (accessed: 29.06.2023)
WHO (2008) Commission on Social Determinants of Health (CSDH) Closing the gap in a generation: health equity through action on the social determinants of health. Geneva: WHO. https://www.who.int/publications/i/item/WHO-IER-CSDH-08.1 (accessed: 01.09.2023).
WHOa Constitution. https://www.who.int/about/governance/constitution (accessed: 01.09.2023)
WHOb European Health for All database (HFA-DB). URL: https://gateway.euro.who.int/ru/datasets/european-health-for-all-database/ (accessed: 29.06.2023)
WHOc Study on Global Ageing and Adult Health (SAGE) https://www.who.int/data/data-collection-tools/study-on-global-ageing-and-adult-health#:~:text=WHO’s%20SAGE%20is%20a%20longitudinal,Russian%20Federation%20and%20South%20Africa (accessed: 01.09.2023)
The publication uses SHARE data 1, 2, 3, 4, 5, 6, 7 and 8 waves (DOI: 10.6103/SHARE.w1.710, 10.6103/SHARE.w2.710, 10.6103/SHARE.w3.710, 10.6103/SHARE.w4.710, 10.6103/SHARE.w5.710, 10.6103/SHARE.w6.710, 10.6103/SHARE.w7.711, 10.6103/SHARE.w8.100, 10.6103/SHARE.w8ca.100). SHARE data collection was funded by the European Commission’s Directorate General for Research and Innovation through grants FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3: RII-CT-2006-062193, COMPARISON: CIT5-CT-2005-028857, SHARELIFE: CIT4-CT-2006-028812), FP7 (SHARE-PREP: No. GA 211909, SHARE-LEAP: GA No. 227822, SHARE M4: GA No. 261982, DASISH: GA No. 283646), by Horizon 2020 (SHARE-DEV3: GA No. 676536, SHARE-COHESION: GA No. 870628, SERISS: GA No. 654221, SSH: GA No. 823782) and the Director General for Employment, Social Affairs and Social Integration through Grants VS 2015/0195, VS 2016/0135, SUN 2018/0285, SUN 2019/0332 and SUN 2020/0313. Additional funding was provided by the German Ministry of Education and Research, The Max Planck Society for the Advancement of Science and the US National Institute of Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06). -11, OGHA_04-064, HHSN271201300071C, RAG052527A). The data used for Estonia relate to the 4th wave of the survey in Europe.
The official website of the study: http://www.share-project.org/data-documentation/share-data-releases.html .
The second database used is WHO’s Global Ageing and Adult Health (SAGE). SAGE is conducted with the financial support of the U.S. National Institute of Aging under interagency agreements (OGHA 04034785; YA1323-08-CN-0020; Y1-AG-1005-0) and under research grants R01-AG034479 and R21-AG034263. The data used relate to the 1st wave of the survey in Russia.
The official website of the study: https://apps.who.int/healthinfo/systems/surveydata/index.php/catalog/68
The study was carried out under the auspices of the program “ERA.Net RUS plus” (grant RUS_ST2019-423 – LifeTraR) with the financial support of the Russian Foundation for Basic Research within the framework of the scientific project No. 20-511-76006.
The authors declare is no conflict of interest.
The authors express their gratitude to Luule Sakkeus, a research professor at the University of Tallinn, and Liili Abuladze, a researcher at the University of Tallinn, for constructive comments and support at all stages of the study.
Selezneva Elena Vladimirovna – Candidate of Sciences (PhD) in Economics and National Economy Management, Senior Research Fellow at the Centre for Comprehensive Social Policy Studies, Institute for Social Policy, National Research University “Higher School of Economics”, Moscow, 101000, Russia. E-mail: evselezneva@hse.ru.
Sinyavskaya Oksana Vyacheslavovna – Candidate of Sciences (PhD) in Economics and National Economy Management, Deputy Director at the Institute for Social Policy, Director at the Centre for Comprehensive Social Policy Studies, Institute for Social Policy, National Research University “Higher School of Economics”, Moscow, 101000, Russia. E-mail: osinyavskaya@hse.ru.
Odds ratios for ethnic groups in the ordinal regression models for self-rated health with a different set of predictors – with confidence intervals
Assessments of the ordinal regression models for self-rated health