Open Access Highly Accessed Open Badges Research article

The mediating effect of social relationships on the association between socioeconomic status and subjective health – results from the Heinz Nixdorf Recall cohort study

Nico Vonneilich1*, Karl-Heinz Jöckel2, Raimund Erbel3, Jens Klein1, Nico Dragano2, Johannes Siegrist4, Olaf von dem Knesebeck1 and on behalf of the Heinz Nixdorf Investigator Group

Author Affiliations

1 Department of Medical Sociology and Health Economics, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

2 Institute of Medical Informatics, Biometry and Epidemiology, University Duisburg-Essen, Essen, Germany

3 Clinic for Cardiology, University Clinic Essen, Essen, Germany

4 Department of Medical Sociology, Heinrich Heine University Düsseldorf, Düsseldorf, Germany

For all author emails, please log on.

BMC Public Health 2012, 12:285  doi:10.1186/1471-2458-12-285

Published: 17 April 2012



Socioeconomic status (SES) is an important determinant of population health. Explanatory approaches on how SES determines health have so far included numerous factors, amongst them psychosocial factors such as social relationships. However, it is unclear whether social relationships can help explain socioeconomic differences in general subjective health. Do different aspects of social relationships contribute differently to the explanation? Based on a cohort study of middle and older aged residents (45 to 75 years) from the Ruhr Area in Germany our study tries to clarify the matter.


For the analyses data from the population-based prospective Heinz Nixdorf Recall (HNR) Study is used. As indicators of SES education, equivalent household income and occupational status were employed. Social relations were assessed by including structural as well as functional aspects. Structural aspects were estimated by the Social Integration Index (SII) and functional aspects were measured by availability of emotional and instrumental support. Data on general subjective health status was available for both baseline examination (2000–2003) and a 5-year follow-up (2006–2008). The sample consists of 4,146 men and women. Four logistic regression models were calculated: in the first model we controlled for age and subjective health at baseline, while in models 2 and 3, either functional or structural aspects of social relationships were introduced separately. Model 4 then included all variables. As former studies indicated different health effects of SES and social relations in men and women, analyses were conducted with the overall sample as well as for each gender alone.


Prospective associations of SES and subjective health were reduced after introducing social relationships into the regression models. Percentage reductions between 2% and 30% were observed in the overall sample when all aspects of social relations were included. The percentage reductions were strongest in the lowest SES group. Gender specific analyses revealed mediating effects of social relationships in women and men. The magnitude of mediating effects varied depending on the indicators of SES and social relations.


Social relationships substantially contribute to the explanation of SES differences in subjective health. Interventions for improving social relations which especially focus on socially deprived groups are likely to help reducing socioeconomic disparities in health.