Psychosocial Work Characteristics as Predictors for Burnout: Findings From 3-Year Follow Up of the PUMA Study Marianne Borritz, MD Ute Bu ¨ ltmann, PhD Reiner Rugulies, PhD Karl Bang Christensen, PhD Ebbe Villadsen Tage S. Kristensen, DrMedSci Learning Objectives • Relate baseline psychosocial work characteristics to three categories of burnout (personal, work-related, client-related) in a cross-sectional study of 1772 workers at human service organizations. • Identify any demographic or behavioral correlates of burnout. • Recall associations between burnout and psychosocial features of work in 952 workers in the human services sector who were followed up prospec- tively for 3 years, after adjusting for possible confounding factors, the degree of burnout at baseline, and psychosocial work characteristics themselves. Abstract Objective: The objective of this study was to investigate the impact of psychosocial work characteristics on burnout. Methods: A total of 1772 participants in different human service sector organizations were eligible for the cross-sectional analyses (baseline) and 952 for the prospective analyses. We measured 14 psychosocial work characteristics and three types of burnout. Linear regression models were used for analyzing associations between psychosocial work characteristics at baseline and burnout at baseline and at 3 years of follow up. Results: Low possibilities for development, high meaning of work, low predictability, high quality of leadership, low role clarity, and high role conflicts predicted burnout at 3 years of follow up after the psychosocial work characteristics were adjusted for each other, potential confound- ers, and burnout level at baseline. Conclusion: Psychosocial work characteristics were prospectively associated with burnout, suggesting that improving the psychos- ocial work environment may reduce future burnout in human service work. (J Occup Environ Med. 2005;47:1015–1025) B urnout is an increasing problem in highly developed countries. 1–3 The phenomenon burnout is described as a particular type of prolonged occu- pational stress that seems to occur most prominently among human ser- vices professionals and is character- ized by emotional exhaustion, loss of energy, and withdrawal from work. 3 The concept of burnout started as a “grassroots” description of pro- longed occupational stress among human service workers, in which former engaged employees gradually get overwhelmed of emotional ex- haustion, loss of energy, and with- drawal from work. Thus, the burnout concept was developed from field observations—not from theory. In a comprehensive review from 1998, Schaufeli and Enzmann conclude: “Burnout is not a new phenome- non—it has its root in the past. However, because of a unique con- stellation of several factors it was ‘discovered’ in the early 1970s as a particular type of prolonged occupa- tional stress that seemed to occur most prominently among human ser- vices professionals.” 3 Research on burnout was initiated in the mid- 1970s. Since then, more than 5500 studies on burnout have been pub- lished. 1,3 From the 1980s to the mid- 1990s, several cross-sectional studies were performed in addition to the former rather descriptive case sto- ries. A metaanalysis including these cross-sectional studies showed that emotional work combined with orga- nizational problems is associated with high levels of burnout. 4 Since From the National Institute of Occupational Health, Copenhagen, Denmark (Dr Borritz, Dr Bu ¨ltmann, Dr Rugulies, Dr Christensen, Mr Villadsen, Dr Kristensen); and the Department of Occupational Medicine, Bispebjerg Hospital, Copenhagen, Denmark (Dr Borritz). Marianne Borritz has no financial interest related to this article. Address correspondence to: Marianne Borritz, MD, National Institute of Occupational Health, Denmark, Lersoe Parkalle ´ 105, DK-2100 Copenhagen O, Denmark; E-mail: mb@ami.dk. Copyright © by American College of Occupational and Environmental Medicine DOI: 10.1097/01.jom.0000175155.50789.98 CME Available for this Article at ACOEM.org JOEM • Volume 47, Number 10, October 2005 1015