Article Text

Download PDFPDF

The recent economic recession and self-rated health in Estonia, Lithuania and Finland: a comparative cross-sectional study in 2004–2010
  1. Rainer Reile1,2,
  2. Satu Helakorpi3,
  3. Jurate Klumbiene4,
  4. Mare Tekkel5,
  5. Mall Leinsalu5,6
  1. 1Department of Public Health, University of Tartu, Tartu, Estonia
  2. 2Institute of Social Studies, University of Tartu, Tartu, Estonia
  3. 3Department of Lifestyle and Participation, National Institute for Health and Welfare (THL), Helsinki, Finland
  4. 4Institute for Health Research, Public Health Faculty, Lithuanian University of Health Sciences, Kaunas, Lithuania
  5. 5Department of Epidemiology and Biostatistics, National Institute for Health Development, Tallinn, Estonia
  6. 6Stockholm Centre on Health of Societies in Transition, Södertörn University, Huddinge, Sweden
  1. Correspondence to Rainer Reile, Department of Public Health, University of Tartu, Ravila 19, Tartu 50411, Estonia; rainer.reile{at}


Background The late-2000s financial crisis had a severe impact on the national economies on a global scale. In Europe, the Baltic countries were among those most affected with more than a 20% decrease in per capita gross domestic product in 2008–2009. In this study, we explored the effects of economic recession on self-rated health in Estonia and Lithuania using Finland, a neighbouring Nordic welfare state, as a point of reference.

Methods Nationally representative cross-sectional data for Estonia (n=10 966), Lithuania (n=7249) and Finland (n=11 602) for 2004–2010 were analysed for changes in age-standardised prevalence rates of less-than-good self-rated health and changes in health inequalities using logistic regression analysis.

Results The prevalence of less-than-good self-rated health increased slightly (albeit not statistically significantly) in all countries during 2008–2010. This was in sharp contrast to the statistically significant decline in the prevalence of less-than-good health in 2004–2008 in Estonia and Lithuania. Health disparities were larger in Estonia and Lithuania when compared to Finland, but decreased in 2008–2010 (in men only). In Finland, both the prevalence of less-than-good health and health disparities remained fairly stable throughout the period.

Conclusions Despite the rapid economic downturn, the short-term health effects in Estonia and Lithuania did not differ from those in Finland, although the recession years marked the end of the previous positive trend in self-rated health. The reduction in health disparities during the recession indicates that different socioeconomic groups were affected disproportionately; however, the reasons for this require further research.

  • Self-Rated Health
  • Eastern Europe
  • Health inequalities
  • Socio-Economic
View Full Text

Statistics from


The late-2000s financial crisis was in terms of its causes, magnitude and global reach similar to the 1929 Great Depression. Both had their origins in the excessive deregulation of financial markets1 ,2 and resulted in a persistent economic recession with consequences for health outcomes.3 ,4

Recent reviews on the health effects of economic recessions5–7 indicate that health is likely to be affected either directly via increased unemployment and reduced income or indirectly through the healthcare system. The policy responses to the current crisis varied across health systems in Europe and public spending on health tended to fall as the health sector is vulnerable to budget cuts.8 This is important as austerity measures are generally expected to have a negative effect on healthcare and health.6 ,9 Policy-related and income-related mechanisms7 may explain negative health outcomes such as increased levels of perceived stress and depression10 ,11 as a result of economic crisis. Economic stressors can affect individual health-related quality of life, mostly in mental health domains,12 but have also been associated with somatic symptoms,13 hospitalisations,14 increased sickness-absence leave15 and deteriorating self-rated health (SRH).16

At the same time, some authors have argued that health may even improve during recessions as a result of changing lifestyles and improvement in health behaviours.17 Although recessions do not always have major negative health effects at the aggregate population level,17 ,18 it is probable that unfavourable economic developments affect most of those on the lower levels of the social hierarchy.7 Rapid economic transformation may result in stress and negative life events, which is one of the many ways in which socioeconomic inequalities in health are (re)produced in society.19 An earlier research on the health effects of rapid societal change in Eastern and Central Europe in the 1990s showed that economic fluctuations may place additional strain on lower socioeconomic groups resulting in widening health disparities.20

Economic consequences of the recent crisis were especially strong in some Eastern European countries (along with Greece, Ireland and Spain).21 ,22 In Baltic countries, per capita gross domestic product (GDP) decreased by more than 20% within 1 year (2008–2009) and compared with 2007 the unemployment rates had more than tripled by 2010, reaching 17% and 18%, respectively, in Estonia and Lithuania. While per capita GDP also declined sharply in Finland (by 13%), the increase in the unemployment rate (reaching 8% in 2010) was not as dramatic as in the Baltic countries.23 Nevertheless, no comparative studies on the impact of crisis on health have been carried out so far in this region. This paper focuses on the health effects of the recent recession in Estonia and Lithuania, using data from a neighbouring Nordic welfare state, Finland, as a point of reference. To capture the possible recession-related health effects in 2008–2010, we analyse changes in the prevalence of less-than-good SRH in relation to the period of rapid economic growth in 2004–2008. As the health effects of economic downturns and their variation across the social ladder are still not fully understood, we compare changes in socioeconomic inequalities in SRH during the period of economic growth and the following recession.


Nationally representative data for the 2004–2010 period were drawn from the FinBalt Health Monitor project,24 a collaborative study of health-related behaviours conducted in Finland and the Baltic countries. Studies were conducted as a series of biennial cross-sectional postal surveys using a harmonised methodology and questionnaires. Random samples of individuals were selected from the countries’ national population registries with no replacement of non-respondents. The initial sample per year included 5000 persons in Estonia (in the 16–64 years age group) and in Finland (15–64 years). In Lithuania, the sample (in the 20–64 years old) consisted of 3000 persons in 2004–2008 and of 4000 persons in 2010. Response rates varied across countries ranging between 59 and 63% in Estonia (n=10 966), 54 and 62% in Lithuania (n=7249) and 57 and 68% in Finland (n=11 602) (table 1). Item non-response remained lower than 2% for all study variables. The analyses cover the 20–64 age group.

Table 1

Characteristics of the surveys and respondents, in the 20–64 age group

We used SRH as an indicator of subjective general health status. The question on SRH asked: ‘How would you assess your present state of health: (1) good; (2) reasonably good; (3) average; (4) rather poor; or (5) poor’. Answers were dichotomised into categories of less-than-good health (3–5) and good health (1–2).

The sociodemographic characteristics of age, ethnicity, education and employment status were used as independent variables. Age effects were analysed in three groups: (1) 20–34, (2) 35–49 and (3) 50–64 years old. For ethnicity, data were categorised into the (1) main ethnic group referring to the titular nation, and, (2) other ethnic groups. In Estonia and Lithuania ethnicity was self-reported; in Finland information on ethnicity was retrieved from registry data. In Estonia and Lithuania, education was measured by the highest level of completed education whereas in Finland years of full-time education were used. To achieve comparable educational categories across countries education was categorised into: (1) high, (2) intermediate and, (3) low levels. In Estonia and Lithuania high education refers to tertiary, university level education, intermediate education covers upper secondary or secondary vocational education and low education refers to lower secondary education or less. In Finland, high education refers to 15 or more years (full time) schooling, intermediate to 11–14 years, and low education refers to 10 years or less schooling. Economic activity was measured by self-reported activity status and dichotomised into (1) employed and (2) unemployed/non-active. The latter category consisted of persons who were studying, were homemakers, retired and not working or were unemployed. The ‘unemployed’ and ‘non-active’ categories were merged together due to the low number of cases in the ‘unemployed’ category in the selected years.

Trends in SRH were assessed by using age standardised prevalence ratios (using the direct method, 5-year age groups and the ‘old’ European standard population) with 95% CIs. The relative sociodemographic disparities in SRH by country and study year were measured by ORs with 95% CIs from multivariable logistic regression analysis. To assess the independent effect of each sociodemographic variable, all study variables were simultaneously added to the model. To assess whether the magnitude of sociodemographic disparities differed across study years (as compared to 2008), all sociodemographic variables in the fully adjusted model were separately tested for interaction with study year. An additional regression analysis was performed to assess the association between the study year and SRH, with all sociodemographic variables and study year simultaneously included to the model using 2008 as a reference year. All analyses were performed separately for men and women. Statistical analyses were conducted using SPSS V.18.0 (SPSS Inc).


Table 1 presents the basic characteristics of the study sample. The proportion of respondents by age varied slightly across countries with Estonia having the highest share of respondents in the youngest age group and Finland having the highest proportion in the oldest age group. Women had a higher average level of education than men in all countries. The proportion of respondents by self-reported employment status was comparable across countries. Estonia had the highest share of persons from other ethnic backgrounds. In nearly all countries, the relative proportion of older respondents, the highly educated and the main ethnic group increased over time.

The overall age-standardised and sex-standardised prevalence of less-than-good SRH (data not shown) decreased in Estonia from 59% in 2004 to 50% in 2008 and in Lithuania from 57% to 47%, respectively, with no remarkable changes observed in Finland. During the period of economic crisis in 2008–2010, the prevalence of poor health increased to 52% in Estonia and to 48% in Lithuania. Although the increase was not statistically significant, it marked the end of the previous positive trend of improving health status. A small and statistically insignificant increase occurred in 2008–2010 in Finland.

Among men (table 2) the prevalence of less-than-good health decreased in 2004–2008 in all countries considered and in nearly all sociodemographic groups (except among the not employed). For the 2008–2010 recession period, a slight and statistically non-significant increase was observed in the overall prevalence of less-than-good health among men in all countries. The prevalence of less-than-good health continued to decrease among the oldest age group in Estonia and Lithuania, and contrary to the previous trend, the prevalence of less-than-good health declined among not employed men in all countries. In Finland, a large increase in less-than-good health was observed in the non-titular ethnic group.

Table 2

Age-standardised prevalence (%) with 95% CIs for less-than-good self-rated health by sociodemographic variables among men, 2004–2010

Among women (table 3), a steep and statistically significant decline in overall less-than-good health was observed in Lithuania and Estonia in 2004–2008, whereas in Finland the prevalence remained nearly unchanged. In 2008–2010, a small and non-significant increase in the prevalence of less-than-good health was found among women in Estonia and Lithuania, whereas the prevalence slightly decreased among Finnish women. The patterns of change in 2008–2010 by sociodemographic indicator were less consistent than for men. The prevalence of less-than-good health continued to decline for women in the oldest age group in Estonia and Lithuania, and now also in Finland, though the trends reversed for younger age groups. Low educated women in Estonia and Finland experienced the largest improvement in health whereas in Lithuania, the prevalence of less-than-good health increased among low educated women. A slight improvement in health among not employed women was seen in Estonia and Lithuania, although the opposite was observed in Finland.

Table 3

Age-standardised prevalence (%) with 95% CI for less-than-good self-rated health by sociodemographic variables among women, 2004–2010

The results from multiple regression analyses are presented in table 4. Statistically significant differences between years 2004 and 2008 in SRH were found for men and women in Estonia and Lithuania but not in Finland. The differences between 2008 and 2010 in SRH were statistically insignificant in all countries. A similar sociodemographic gradient in SRH was observed in all countries in all study years with those being older, less educated or not employed having higher odds for less-than-good SRH. However, the sociodemographic disparities were much larger in the Baltic countries than in Finland. In 2004–2008, sociodemographic inequalities in SRH generally increased or remained the same among men in all countries. The increase in the magnitude of inequalities was particularly notable by employment status in all three countries (in Lithuania at statistically significant level) and by educational level in Estonia. Among women, however, there was a general tendency towards decreasing sociodemographic inequalities in 2004–2008, with the exception of Lithuanian women where the magnitude of inequalities grew by employment status and age. In 2008–2010, the magnitude of inequalities decreased among men in all three countries for nearly all sociodemographic variables, although statistically significant decrease was observed only in Lithuania for educational level and employment status. Among women, the patterns were less consistent in 2008–2010. In contrast to the previously observed decline in the magnitude of health inequalities, inequalities by educational level increased in Estonia and Lithuania and by employment status in Finland.

Table 4

Adjusted ORs for less-than-good self-rated health (SRH) by sociodemographic variables, 2004–2010


This study explored the impact of the 2000s economic crisis on the health status in the Baltic region-one of the regions hit hardest by the crisis. We found a slight (albeit statistically non-significant) decline in the health status of men and women in Estonia and Lithuania in 2008–2010 compared to the significant health improvement in 2004–2008. Socioeconomic health disparities tended to increase (or remained high) in 2004–2008, but decreased in 2008–2010 in Estonia and Lithuania, although changes among women were less consistent. In Finland, a country with a substantial welfare system—the prevalence of less-than-good health and health disparities remained fairly stable throughout the period.

In 2004–2007, the economic output of the Baltic countries grew faster than in most high-income countries in Europe.25 The magnitude of this economic change and its impact on well-being might help explain the noteworthy and almost universal improvements in health ratings in Estonia and Lithuania in 2004–2008. Between 2008 and 2010, a small deterioration in SRH occurred in Estonia and Lithuania. Although the increase in less-than-good SRH was statistically insignificant, it marked a sudden interruption of the previously observed positive trend. Our findings are consistent with the results from recent population-based studies from Greece, where slight (and statistically significant) worsening in SRH was observed during the recession.16 ,26 On the contrary, SRH improved slightly during the 1990s economic recession in Japan.27 Yet, Japan’s longer period of recession in the 1990s resulted in gradual macroeconomic changes compared to the rapid economic downturn that occurred after the 2000s financial crisis.

Economic recession can affect health through various mechanisms with loss of job and decreased income as a result of it being one of the most common explanations. In the context of recession, unemployment has generally been associated with negative health outcomes.5 ,7 For example, a recent study found that job loss due to enterprise closure increased the odds of experiencing fair or poor health by 54%, and among respondents with no pre-existing health problems, it increased the odds of a new health condition being reported by 83%.28 Our results showed that the prevalence of less-than-good health declined among the not employed in all countries (except among women in Finland) during the recession. At the same time, health deteriorated slightly among employed men (and women in Estonia) in all countries, thus reducing relative inequalities in less-than-good SRH. These findings might seem initially counter intuitive, yet several studies have found that during recessions, employed workers had an increased risk of developing hypertension and diabetes,29 and had more somatic and psychological symptoms, as the effects of unemployment in society may be mediated through pessimism about the future, higher work demands and financial problems.30 Also, reductions in health inequalities in times of high unemployment can be explained by reduced health selection. Indeed, several studies have shown, that when unemployment becomes widespread in society, health differences between the unemployed and employed are reduced, whereas during periods of low unemployment, unemployed individuals tend to be in poorer health than employed persons.31 ,32 This is in accordance with our results where an increase in health disparities among men by employment status was found during the economic boom in 2004–2008.

Kaplan5 noted that the negative health effects of economic change are undoubtedly greater among disadvantaged groups. Although the 1990s transition crisis in Eastern Europe affected those in a lower social position disproportionately and led to widening health disparities,20 a similar effect was not observed in our study. As the recent crisis in the Baltic countries was mostly driven by the mortgage bubble,21 ,25 it is likely that the crisis had a more profound effect on middle-socioeconomic to high-socioeconomic groups who had existing housing loans and who could thus be more vulnerable to the consequences of a shrinking job market.33 This is supported by our findings indicating improvement in the health status of lower educated and non-employed men whereas the opposite was observed among the higher educated and employed, thus reducing relative inequalities in SRH. For women, the variation in health inequalities was less consistent across countries, suggesting different response mechanisms of men and women to macroeconomic fluctuations. The gender-related aspects of health disparities during economic recessions would therefore require further research. Unexpectedly, ethnic differences in health presented a clear pattern of association only among women in Estonia with an overall trend towards declining disparities. This contrasts with previous studies from Estonia that found significant ethnic differences in SRH.34 Unhealthy behaviours35 and more frequent depression36 explained part of these inequalities but it is not known whether changes in these factors may underlie the positive health trend among the non-titular ethnic group seen in our study. Compared to Estonia and Lithuania, inequalities in less-than-good SRH were much smaller and remained unaltered during 2004–2010 in Finland which corresponds to the findings from previous studies31 ,37 reporting relatively stable socioeconomic disparities in health across time in the Nordic countries.

Despite the huge economic growth in the mid-2000s, the per capita GDP in Estonia and Lithuania still remained more than three times lower than in Finland. Differences in national wealth and public health expenditures, but also the differences in social spending and other policy-related aspects echoed in recent reviews5 ,7 could explain between-country variation in health status and its change during the study period. National health policy, which determines access to essential services for the population, is under considerable pressure during periods of economic crisis. A recent policy review8 reported an increase in user charges for some health services in Estonia and Finland, reductions in temporary sickness benefits paid by the health insurance fund in Estonia and Lithuania and cuts in the national health budget in Estonia in response to the financial crisis. Although these reductions in funding and coverage of health services were unfavourable, the effect of the financial crisis on health policy was similar in all three countries considered in this analysis8 and compared to cutbacks in Greece26 or in Spain,38 the scope of the reductions remained limited. In addition to healthcare, general social safety mechanisms may serve as a ‘cushion’ against rapid economic changes. Welfare state characteristics have been shown to account for about one-tenth of the variation in SRH between countries with stronger impact found for Scandinavian (and Anglo-Saxon) welfare regimes compared to East European (and Southern) welfare regimes.39 Although the benefits of a strong welfare state have also shown to offer better protection against unfavourable health effects in times of recession,40 our analysis found that the health effects of the recent recession were very similar in the (two) Baltic countries and in Finland, representing a Nordic welfare state.

Our study has several limitations that need to be considered. First, this comparative study analyzed and interpreted trends in health status in the context of rapid macroeconomic change. As the repeated cross-sectional nature of the data does not allow us to determine causality, the occurrence of macroeconomic change during 2004–2010 is therefore only one possible interpretation for the observed variance in the microdata. Second, by having 2008 as the reference year we may have somewhat underestimated the magnitude of health effects during the period of economic growth and recession as economic growth was already slowing down by 2008 and may thus have had some early effect on SRH. Third, the survey response rates which ranged from 54% to 68% have been declining across the survey years (table 1). Low participation rate may bias the results, however, a previous study41 using data of late responders from the same FinBalt surveys showed that response bias had a minimal effect on prevalence indicators and no systematic bias between the countries was found. Finally, the operationalisation of the variables may have affected the results, especially in the case of education, where different formulations of educational attainment were combined resulting in a higher proportion of respondents with the highest and the lowest educational level in Finland. However, as our analysis did not focus on direct cross country comparisons by educational status, these differences have little relevance. Dichotomising SRH as good and less-than-good may have had some effect on the observed variation in SRH between the countries possibly related to the cultural differences of reporting the ‘average’ SRH.35 Nevertheless, the sensitivity analysis using different categorisation for SRH (data not shown) yielded fairly similar results regarding the magnitude of the changes. Therefore these factors are unlikely to have affected data quality significantly nor changed their effect substantially from other studies using a similar design.


The current study focused on the health effects of macroeconomic changes in Estonia and Lithuania, two Eastern European countries that experienced a significant economic boom and were then severely hit by economic recession. Although our comparative analysis did not find a statistically significant health effect of the recent economic crisis in 2008–2010, the end of the previous trend of improving health status during 2004–2008 highlights the negative consequences of the recession on health. The reduction in existing health disparities during the recession indicates that rapid economic fluctuations may affect different socioeconomic groups differently but not always in a predictable manner, neither is the impact similar in men and women. The causal explanations of such developments warrant further in depth research preferably utilising longitudinal study design, exploring different health outcomes and including a broader set of explanatory variables.

What is already known on this subject

  • Health status of the general population is likely to be affected by economic recessions.

  • Persons with lower socioeconomic status are more vulnerable to unfavourable economic developments and recessions are thus likely to affect socioeconomic inequalities in health.

  • Although the economic consequences of the recent recession were especially strong in some Eastern European countries, no comparative studies on the impact of crisis on health have been carried out in this region.

What this study adds

  • Recession in 2008–2010 interrupted the previous trend of improving self-rated health in Estonia and Lithuania and resulted in diminishing health inequalities in men.

  • Although the prevalence and inequalities in self-rated health were more stable in Finland in 2004–2010, the magnitude of changes in 2008–2010 was very similar in Estonia, Lithuania and Finland.


The authors would like to thank Dr Andrew Stickley for commenting on an earlier version of the manuscript.


View Abstract


  • Contributors RR and ML designed the study, ran the statistical analyses, and wrote the first and consecutive drafts of the paper together. SH, JK and MT were responsible for the collection of survey data used in this study and conducted the data cleansing. All authors revised the draft paper, provided critical comments and approved the final version of the paper.

  • Funding This work was supported by the Swedish Foundation for Baltic and East European Studies (grant number A052-10). ML's and MT's work was additionally supported by the Estonian Research Council (IUT5-1).

  • Competing interests None.

  • Ethics approval All surveys have been approved by respective national ethics review boards, that is, Tallinn Medical Research Ethics Committee in Estonia; the Lithuanian Bioethics Committee in Lithuania; and the Research Ethics Board of National Institute for Health and Welfare in Finland. In this study only anonymous secondary data were used.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.