Emotion work within eldercare and depressive symptoms: A cross-sectional multi-level study assessing the association between externally observed emotion work and self-reported depressive symptoms among Danish eldercare workers

Background: Danish professional caregivers have high rates of depressive symptoms. One proposed cause is exposure to emotion work. However, emotion work is usually measured by self-report which may bias results. Objectives: The objective of this study was to examine the association of emotion work, externally observed at the workplace, with self-reported depressive symptoms of professional caregivers. Design and data sources: The study was a cross-sectional observational study. Data was collected by 9 observers who assessed emotion work stressors and emotion work resources in 124 individual professional caregivers working in 56 work units across 10 eldercare homes. Emotion work stressors were defined as i) barriers for empathetic care, ii) taxing aggressive events, and iii) taxing non-aggressive events. Emotion work resources were defined as i) meaningful events, and ii) social interactions between professional caregivers and residents. Depressive symptoms were measured by a questionnaire sent to all professional caregivers at the 10 eldercare homes. We constructed two samples for analysis: a) a sample of 95 directly observed professional caregivers with full information on covariates, and b) a sample of 205 observed and non-observed professional caregivers with full information on covariates working in one of the 56 observed work units. Methods: Using multilevel regression models we analysed associations of individual and work unit averaged levels of emotion work with depressive symptoms among professional caregivers. Results: None of the three emotion work stressors were associated with depressive symptoms. Of the two emotion work resources, a high amount of social interactions between professional caregivers and residents were, contrary to expectations, related to higher levels of depressive symptoms at both the individual level and the work unit averaged level. Conclusions: The unexpected association between social interactions and depressive symptoms need to be replicated in future studies. These future studies should also investigate whether the association of social interactions and level of depressive symptoms depends on the content of the interactions between professional caregivers and residents. (C) 2016 Published by Elsevier Ltd.