Unemployment, Social Support and Health Problems

Results of the GEDA Study in Germany, 2009

Background: It is well documented that the unemployed have more health problems than the employed, and that social support facilitates coping with unemployment. The association of unemployment and social support with health was examined on the basis of representative data derived from a German study.

Methods: The GEDA study (Gesundheit in Deutschland Aktuell [Current Health in Germany]) was conducted in 2008/09 by the Robert Koch Institute, the nationwide governmental public health agency in Germany. For this paper, we analyzed data from the GEDA study that were obtained from 12022 persons aged 30 to 59. We used health indicators taken from the Four-Item Healthy Days Core Module of the Centers for Disease Control (CDC), and we measured social support on the Oslo Three-Item Social Support Scale. We report the findings on impairment in three distinct areas (physical, emotional, and functional) and the results of a multivariate statistical analysis.

Results: Unemployed persons aged 30 to 59 years suffer physical, emotional, and functional impairment more commonly than employed persons. Men and women with little social support are more likely to be impaired in these three areas whether they are employed or not. Regression analysis reveals that unemployment and social support have significant, independent effects on both the incidence of such impairments (OR_unemployed = 1.2–1.7, OR_supported = 0.4–0.9) and on their duration (IRR_unemployed = 1.3–1.8, IRR_supported=0.6–0.8) after age, income, and education have been controlled for.

Conclusion: Physicians should be mindful of the deleterious effect of unemployment on health and should encourage unemployed patients to participate in social networks, as the evidence shows that social support can have health benefits.

The health-related consequences of unemployment are a central subject of medical sociology, health psychology and social medicine and are regularly addressed by federal and regional health reporting bodies in Germany. Many studies have shown clearly that unemployed men and women suffer from more health problems and more frequently behave in ways that pose risks to their health than those who are employed (1–4). Unemployed men and women’s increased consumption of medical and psychotherapeutic resources are also relevant to health services research. Unemployed people use health services more often, are more affected by sleep disorders, depression, anxiety disorders, and addictions, and commit suicide more frequently (5, 6).

Health-related consequences of unemployment manifest as psychosocial health problems due to loss of income, job-related social contact, or social prestige (7–12). Social resources have a particularly marked effect on how well individuals cope with unemployment: the person affected copes with the loss of his/her job more successfully if he/she has a social support network and does not feel alone (2, 13, 14). This article describes the relationship between unemployment; social support; and physical, emotional, and functional complaints in the German population of working age (whether employed or unemployed), on the basis of representative data for 2009. As yet there are no representative analyses in Germany of the association between unemployment, social support, and health which examine both the probability of impairments of various kinds and their duration.

Methods

Data from the 2009 GEDA study (Gesundheit in Deutschland Aktuell [Current Health in Germany]) were used to analyze the relationship between unemployment and health complaints (Table 1 gif ppt). The GEDA study was a telephone survey conducted as part of the Robert Koch Institute’s health monitoring program between July 2008 and June 2009 (15, 16). Sampling was based on the Gabler-Häder design for random sampling of telephone numbers. Numbers were generated at random on the basis of entries in the telephone directory, and telephone numbers not in the directory could also be included in the sample. The reference population included all adults living in private households with landline telephones. The subjects of the survey were divided into a core subject area which remained constant in follow-up inquiries and a flexible subject area which addressed current problems and subjects. A total of 21 262 people were surveyed as part of the study. The response rate, i.e. the number of interviews conducted as a percentage of the total number of members of the population contacted (response rate 3 according to the American Association for Public Opinion Research, AAPOR), was 29.1% (e5). Subsequent analysis involved study participants’ information on their employment status, from the Four-Item Healthy Days Core Module of the United States Centers for Disease Control and Prevention (CDC) (17), their perceived social support (18), their education and training (19), and their net household income. Analysis included only study participants who were aged between 30 and 59 when surveyed (n = 12 022). It can be assumed that individuals of this age have already completed their education and training but have not yet retired (e1).

The data on employment status in this study are subjective assessments of individuals’ current situation. Participants were first asked whether they were currently employed (n = 9903). Those who answered no were then asked whether they were currently unemployed (n = 484). Thus no registration with an employment agency was needed for someone to be classified as “unemployed.” Those who were not part of the working population (n = 1635) were not included in these analyses. This definition is based on the concept of unemployment put forward by the International Labour Organization (ILO) (e1).

The dependent variables used were three overall indicators of health-related quality of life. All three were taken from the Four-Item Healthy Days Core Module (HRQOL-4) of the CDC (17). The questions had been translated into German. They addressed numbers of days in the last month with physical complaints (due to illnesses or injuries), mental problems (stress, depression, general mood), and functional limitations affecting respondents’ usual activities (self-care, work, recreation). In this study, the number of days respondents had suffered each of these impairments was analyzed.

The indicator social support is based on the Oslo Three-Item Social Support Scale, a scale used in Europe to measure perceived social support (18, 20). Study participants were asked three questions on how many close friends they had, how involved other people were in their lives, and the availability of help from their neighbors. According to the instructions for using the scale, the replies were used to calculate a cumulative total score (3 to 14 points), which was then allocated to one of the following categories: low (3 to 8), medium (9 to 11), high (12 to 14).

The control variables were participants’ age, level of education, and equivalized income. The authors classified participants’ level of education and training on the basis of the CASMIN model, which is used internationally. Information they themselves provided on their households’ net income was adjusted according to the new OECD-modified equivalence scale for household size. This corresponds to the German Federal Government’s report on poverty and wealth (e4). Unavailable information on participants’ income (n = 2797) was statistically estimated on the basis of regional statistical information from the database INKAR 2007 and information they themselves provided on household size, education, and age (n = 2654), using a multilevel model. In subsequent analyses, the authors first provided a descriptive representation of the distribution of the three dependent variables according to sex and next carried out a multivariate analysis of them on the basis of a regression model for count data (zero-inflated negative binomial regression) (21). The analyses were conducted using the program Stata, version 10.1 (22).

Results

Figure 1 (gif ppt) shows the average number of days with physical, emotional, or functional complaints according to age and sex. The number of days with physical and functional complaints increases with age in both men and women, while there is a smaller increase in emotional complaints. The number of days with impairments is lower in men than in women in all examined age groups. On average, 30 to 59-year-old men had had physical or emotional complaints on 6.6 days in the last 30 days, while the corresponding number for women in the same age group was 8.6 days. The surveyed men and women suffered health-related functional impairments affecting their daily activities on only 2.8 and 3.1 days respectively.

Figure 2 (gif ppt) shows the number of days with complaints according to employment status and sex. According to this information, the employed are significantly less frequently affected by physical, emotional, or functional complaints than the equivalent groups of unemployed men and women. There are bigger differences between the employed and the unemployed for women than for men. There were particularly marked differences in men’s number of days with functional impairments (difference employed vs unemployed = 2.8 days) and women’s number of days with emotional complaints (difference employed vs unemployed = 4.2 days).

The information in Figure 3 (gif ppt) is additionally differentiated according to the level of social support perceived by study participants in their surroundings. Overall, 71% and 65% of unemployed men and women respectively, but 86% of the employed, scored medium or high levels on the scale for social support. The group with social support shows significantly better results for all three types of impairment than the unemployed or the employed with only low levels of social support. Employed people with low levels of social support reported an average of 3.5 days with physical or emotional complaints, while those with medium or high support reported only 2 days. The corresponding figures were 6.2 and 4.4 days respectively for the unemployed. Physical and emotional complaints were particularly common in unemployed women with low levels of social support (9.6 and 11.6 days respectively).

Table 2 (gif ppt) shows the results of regression analysis for count data (zero-inflated negative binomial regression [e2]). The effect of unemployment and social support on the duration of impairments (incidence rate ratio, IRR) and the probability of their not occurring (odds ratio, OR) were examined. The effects were controlled for differences in participants’ ages, levels of education, and needs-adjusted household incomes.

The results show that for unemployed men and women the probability of not having suffered any complaints in the last 30 days is low, while it is significantly higher for respondents with social support than for those who perceive that they have little social support. When unemployed and employed men and women are compared, only differences in functional impairments are statistically significant, while the effect of social support is significant in all cases except for physical complaints in men.

The duration of impairments in the last month is correlated with employment status and perceived levels of social support, regardless of their overall probability of occurrence. In women the correlation between duration of complaints and employment status and social support is significant for all three types of complaint. In men, only the effect of social support on the duration of functional impairments is not statistically significant.

Discussion

The presented results make it clear that in Germany unemployment is associated with worse physical, emotional, and functional health in both men and women. Unemployed men and women also suffer from their complaints for longer on average than the employed when their age, financial situation, and level of education are controlled for. When unemployed and employed people can rely on a supportive social network, this is associated with a lower risk and shorter duration of complaints.

It has already been documented many times that the health-related consequences of unemployment can be mitigated by social support (13, 14). These results are in line with the international literature in that they describe a correlation between unemployment, social support, and various aspects of health for Germany too—for the first time on the basis of representative data. The results also clearly indicate that the correlation with the duration of complaints is stronger than the correlation with the risk of their occurrence.

The limitations of this research are that it uses cross-sectional data which do not allow causal relationships to be inferred. The relationship between unemployment and health is reciprocal: health and health-related behavior may be either the cause or the result of unemployment (23, 24). The observed correlation between unemployment and health complaints must therefore be partly due to selection of unemployed individuals with health problems. However, in the 2009 GEDA study participants were also asked about their assessment of the reasons and consequences of their unemployment. 17% of unemployed individuals replied that their health had deteriorated as a result of their unemployment, while 18% stated health problems as the ultimate cause of their having lost their jobs. Further research should also address correlations between the duration of current unemployment, the availability of social support, and the health of those affected, in greater depth. This is not possible on the basis of current GEDA data. Interpretation of the results must also take into account that the health indicators used were self-reported. Nevertheless, previous cohort studies have successfully identified correlations between unstable employment and physiological parameters such as blood pressure, cholesterol levels, and body mass index (e3).

In summary, the current findings support the hypothesis that unemployment seems to be associated with various health complaints. It should therefore not be underestimated in medical practice. In view of the repeatedly documented health benefits of social support, doctors should encourage unemployed patients to engage more actively in non-professional networks instead of isolating themselves socially. However, the stress caused by loss of employment during a recession cannot be resolved by either the treating physicians or the social networks of those affected alone. Compensatory measures as part of social policy and employment policy must also buffer the effects of unemployment, so as to minimize negative effects on public health (25).

Acknowledgment
The authors would like to thank units FG25 and FG21 of the Robert Koch Institute for providing us with the data from the 2009 GEDA study.

Conflict of interest statement
The authors declare that no conflict of interest exists according to the guidelines of the International Committee of Medical Journal Editors.

Manuscript received on 9 July 2010, revised version accepted on 21 September 2010.

Translated from the original German by Caroline Devitt, MA.

Corresponding author
Dr. phil. Lars E. Kroll
Abteilung für Epidemiologie und Gesundheitsberichterstattung
Fachgebiet 24
Robert Koch-Institut
Postfach 65 02 61
13302 Berlin, Germany

@For eReferences please refer to:
www.aerzteblatt-international.de/ref0411

References

1. RKI: Arbeitslosigkeit und Gesundheit. Gesundheitsberichterstattung des Bundes. Heft 13. Berlin; Robert Koch-Institut 2003.

2. Hanisch KA: Job loss and unemployment research from 1994 to 1998: A review and recommendations for research and intervention. J Vocat Behav 1999; 55: 188–220.

3. Brenner H: Arbeitslosigkeit. In: Stoppe G, Bramesfeld A, Schwartz FW (eds.): Volkskrankheit Depression? Berlin: Springer 2006; 163–89.

4. Elkeles T: Arbeitslosigkeit, Langzeitarbeitslosigkeit und Gesundheit. Sozialer Fortschritt 1999; 6: 150–5.

5. Berth H, Förster P, Balck F, et al.: Arbeitslosigkeitserfahrungen, Arbeitsplatzunsicherheit und der Bedarf an psychosozialer Versorgung. Das Gesundheitswesen 2008; 70: 289–94.

6. Weber A, Hörmann G, Heipertz W: Arbeitslosigkeit und Gesund- heit aus sozialmedizinischer Sicht. Dtsch Arztebl 2007; 104(43): A 2957–62.

7. Jahoda M: Wieviel Arbeit braucht der Mensch. Weinheim: Beltz Verlag 1983.

8. Warr P: Work, unemployment and mental health. Oxford: Oxford University Press 1987.

9. Fryer JP: Employment deprivation and personal agency during unemployment: A critical discussion of Jahoda’s Explanation of the psychological effects. Social Behaviour 1986; 1: 3–23.

10. Brief AP, Konovsky MA, Goodwin R, et al.: Inferring the meaning of work from the effects of unemployment. J Appl Soc Psychol 1995; 25: 693–711.

11. Creed PA, Macintyre SR: The relative effects of deprivation of the latent and manifest benefits of employment on the well-being of unemployed people. J Occup Health Psychol 2001: 6: 324–31.

12. Janlert U, Hammarstrom A: Which theory is best? Explanatory models of the relationship between unemployment and health. BMC Public Health 2009; 9: 235.

13. Schwarzer R, Jerusalem M, Hahn A: Unemployment, social support and health complaints: A longitudinal study of stress in East German refugees. Journal of Community & Applied Social Psychology 1994; 4: 31–45.

14. Bjarnason T, Sigurdardottir TJ: Psychological distress during unemployment and beyond: social support and material deprivation among youth in six northern European countries. Soc Sci Med 2003; 56: 973–85.

15. Kurth BM, Lange C, Kamtsiuris P, et al.: Gesundheitsmonitoring am Robert Koch-Institut. Sachstand und Perspektiven. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2009; 52: 557–70.

16. Lampert T: Smoking, physical inactivity and obesity: association with social status. Dtsch Arztebl Int 2010; 107(1–2): 1–7.

17. Mielenz T, Jackson E, Currey S, et al.: Psychometric properties of the Centers for Disease Control and Prevention Health-Related Quality of Life (CDC HRQOL) items in adults with arthritis. Health Qual Life Outcomes 2006; 4: 66.

18. Dalgard OS, Dowrick C, Lehtinen V, et al.: Negative life events, social support and gender difference in depression: a multinational community survey with data from the ODIN study. Soc Psychiatry Psychiatr Epidemiol 2006; 41: 444–51.

19. Brauns H, Scherer S, Steinmann S: The CASMIN Educational Classification in International Comparative Research. In: Hoffmeyer- Zlotnik JHP, Wolf C, (eds.): Advances in cross-national comparison. New York: Kluwer 2003: 221–44.

20. Kilpeläinen K, Arpo A, ECHIM Core Group European Health Indicators: Development And Initial Implementation, National Public Health Institute (KTL) Helsinki 2008.

21. Long JS: Regression models for categorical and limited dependend variables. SAGE Publikations 1997.

22. StataCorp: Stata Statistical Software: Release 10.0. Stata Corporation, College Station, TX 2007.

23. Bartley M: Unemployment and ill health: understanding the relationship. J Epidemiol Community Health 1994; 48: 333–7.

24. Mathers CD, Schofield DJ: The health consequences of unemployment: the evidence. Med J Aust 1998; 168: 178–82.

25. Stuckler D, Basu S, Suhrcke M, et al.: The public health effect of economic crises and alternative policy responses in Europe: an empirical analysis. The Lancet 2009; 374: 315–23.

e1. Wingerter C: Arbeitsmarkt. In: Destatis, GESIS-ZUMA, WZB (eds.) Datenreport 2008. Ein Sozialbericht für die Bundesrepublik Deutschland, Bonn: Bundeszentrale für politische Bildung 2008: 109–21.

e2. Long S, Freese J: Regression models for categorial dependent variables using stata. Stata Press, College Station, TX 2001.

e3. Ferrie JE, Shipley MJ, Marmot MG, et al.: An uncertain future: the health effects of threats to employment security in white-collar men and women. Am J Public Health 1998; 88(7): 1030–6.

e4. BMAS: Lebenslagen in Deutschland. Der 3. Armuts- und Reichtumsbericht der Bundesregierung. Berlin: Bundesministerium für Arbeit und Soziales 2008.

e5. Müters S, Kamtsiuris P, von der Lippe E, Kroll LE, Lange C: Dokumentation zur Response in der Studie Gesundheit in Deutschland aktuell 2009. Internetveröffentlichung: http://www.rki.de/geda . Berlin: Robert Koch-Institut 2010.