Secondary Logo

The Authors Respond

Dragano, Nico; Siegrist, Johannes; Kivimäki, Mika

doi: 10.1097/EDE.0000000000000826
Letters
Free
SDC

Institute of Medical Sociology, Medical Faculty, University of Düsseldorf, Düsseldorf, Germany, dragano@med.uni-duesseldorf.de

Institute of Medical Sociology, Medical Faculty, University of Düsseldorf, Düsseldorf, Germany

Clinicum, Faculty of Medicine, University of Helsinki, Helsinki, Finland, Department of Epidemiology and Public Health, University College London, London, United Kingdom

Back to Top | Article Outline

To the Editor:

To avoid bias in empirical science, a theoretical foundation, a sound operationalization of constructs, protocol-based testing, replication, and detailed reporting are important. The Individual Patient Data (IPD)–work consortium is devoted to those principles. Our analyses follow a strict protocol, including the definition and publication of the exposure measure before outcome analyses, the use of theory-based measures of exposure, operationalization according to common standards, and per-protocol analyses.

Peer feedback and comments are important to identify potential threats to validity not recognized by the researchers themselves. We thank Mikkelsen et al1 for pointing out two possible sources of bias in our IPD-Work study of the association between effort–reward imbalance at work and coronary heart disease.2 Yet, for the following reasons, we believe that bias in this particular case is unlikely.

First, Mikkelsen et al suspect that the analyses of rewards are confounded by socioeconomic status (SES). This appears to be a misunderstanding. The predefined exposure of our study was effort–reward ratio rather than its subcomponents.3 We adjusted the exposure–outcome associations for SES in the main analysis and, additionally, performed analyses stratified by SES. The results are available in the original publication, and neither adjustment nor stratification indicated substantial confounding.2

We published the operationalization of the effort–reward ratio concept, measured as a dichotomous variable, in 20143—3 years before the publication of our paper on effort–reward imbalance and coronary heart disease2—to provide the scientific community an opportunity to comment our protocol. The lead author of the paper was J. Siegrist, who developed the effort–reward imbalance theory.4 Mikkelsen et al did not comment on our operationalization at that time, but now, as part of their second point, they argue that the association between effort–reward imbalance and health outcomes should be examined using interaction tests on a multiplicative scale. This is not how Siegrist, the hypothesis generator, recommends the construct of effort–reward imbalance to be quantified. According to his ideas, the effort–reward imbalance score is a simple ratio measure to identify persons whose efforts exceed their perceived rewards, irrespective of the strength of the deviation. In this scenario, multiplicative interaction tests are not needed.5 Indeed, it is not uncommon in medical research to use simple composite variables to reflect the combined effect of two variables. Well-known examples include the body mass index and the hip–waist ratio.

In addition to these two points, the letter by Mikkelsen et al6 renews a previous debate about the measurement of job strain. The main exposure in our study was effort–reward imbalance, and it was not in our scope to perform in-depth analyses of the job strain measures. For those interested in this discussion, we would recommend reading our previous reply.7

Nico Dragano

Institute of Medical Sociology, Medical Faculty

University of Düsseldorf

Düsseldorf, Germany

dragano@med.uni-duesseldorf.de

Johannes Siegrist

Institute of Medical Sociology, Medical Faculty

University of Düsseldorf

Düsseldorf, Germany

Mika Kivimäki

Clinicum, Faculty of Medicine

University of Helsinki

Helsinki, Finland

Department of Epidemiology and Public Health

University College London

London, United Kingdom

Back to Top | Article Outline

REFERENCES

1. Mikkelsen S, Andersen JH, Ingre MRe: effort–reward imbalance at work and incident coronary heart disease. Epidemiology 2018;29:e34.
2. Dragano N, Siegrist J, Nyberg ST, et alIPD-Work consortium. Effort–reward imbalance at work and incident coronary heart disease: a multicohort study of 90,164 individuals. Epidemiology. 2017;28:619–626.
3. Siegrist J, Dragano N, Nyberg ST, et alValidating abbreviated measures of effort–reward imbalance at work in European cohort studies: the IPD-Work consortium. Int Arch Occup Environ Health. 2014;87:249–256.
4. Siegrist JAdverse health effects of high-effort/low-reward conditions. J Occup Health Psychol. 1996;1:27–41.
5. VanderWeele TJ, Knol MJA tutorial on interaction. Epidemiologic Methods. 2014;3:33–72.
6. Mikkelsen S, Andersen JH, Bonde JP, Hansen ÅM, Kolstad H, Thomsen JFJob strain and clinical depression. Letter to the editor. Psychol Med. 2017:1–2.
7. Madsen IEH, Rugulies R, Kivimäki MJob strain and clinical depression—authors’ reply. Psychol Med. 2017;2:1–2.
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.