The Authors Respond
Year: 2018
Type: letter
Abstract: To the Editor: Our previous letter1 regarding the paper by Dragano et al.2 on effort–reward imbalance and coronary heart disease (CHD) asked for more information on the association between effort–reward imbalance and CHD by several alternative operationalizations of effort–reward imbalance beyond the most popular operationalization, a ratio of effort–reward imbalance > 1, given various operationalizations in the literature and heterogeneous measures. Siegrist et al.3 replied that “comparison of alternative operationalizations of effort-reward imbalance and its subdimensions in relation to coronary heart disease is a relevant topic for future research.” Ingre et al.4 claimed in their recent letter regarding our previous letter that “none of the alternatives they suggest produce unbiased estimates in observation studies, because they are all susceptible to confounding by the main effects.” However, two different issues are mixed in their claim: (1) how to create a measure of effort–reward imbalance as a combination of high effort and low reward; and (2) whether to test it as an interaction term between effort and reward in addition to the main effects in statistical models. We addressed the former in our previous letter while Ingre et al.4 mainly addressed the latter. The latter is an important issue5 but should be discussed separately from the former for clarity. For example, a composite measure of effort–reward imbalance based on the medians of effort and reward scores, one of several alternative operationalizations that we suggested in our previous letter, can be used for testing interactions between effort and reward on health outcomes, such as CHD.6 We agree with Ingre et al.4 that the interaction effect between effort and reward as well as the main effects on health outcomes need to be examined and reported. Most importantly, we think that the nature of the interaction between effort and reward in effort–reward imbalance work stress theory needs to be better explicated because statistical tests for the interaction or any composite measure will depend on the theoretical assumptions of the nature of the interaction.6 Effort–reward imbalance theory could be enriched through empirical evidence, and precisely testing different operationalizations and the interaction between effort and reward could be useful and informative, mostly through analysis of large databases, which has high statistical power, such as those of the individual-participant data (IPD) work consortium group. Other pertinent issues in effort–reward imbalance theory need to be clarified, including the lack of clarity of the possible interactional role of overcommitment within the model, the different possible combinations of reward components in a multiple interaction term model (e.g., effort × salary, or effort × status, or effort × respect), or other methodologic issues that make the testing of effort–reward imbalance interactions harder, such as some reward items that are assessing effort–reward imbalance in itself and not rewards, as we have previously commented.7 BongKyoo ChoiCenter for Occupational and Environmental HealthSchool of Medicine and Program in Public HealthUniversity of CaliforniaIrvine, CA Arturo Juárez-GarciaCenter for Transdisciplinary Research in PsychologySchool of PsychologyUniversidad Autónoma del Estado de MorelosCuernavaca, México, [email protected] Paul LandsbergisState University of New York Downstate School of Public HealthBrooklyn, NY
Source: Epidemiology
Language: English
Subfield: General Health Professions
Field: Health Professions
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W4206982819