Health consequences of gender-based harassment at work

Abstract

Background: Experiences at work with gender-based violence and harassment (GBVH) from superiors, co-workers or third parties are common. This includes a range of verbal or physical behaviors that can be of a sexual nature and can be one incidence or a pervasive part of the job. Decades of research show associations of sexual harassment with reduced mental and physical health, but many suffer severe methodological limitations, so that the magnitude of the consequences for workers’ health still is unclear. Also, non-sexual sexist experiences, here called gender harassment, have gained less attention despite being more common.

Aims: The studies aimed to further our understanding of different kinds of GBVH as occupational health hazards. Studies I-III investigated the health outcomes long-term sickness absence (study I), psychotropic medication use (study II) and alcohol-related morbidity and mortality (study III). Study IV, a systematic review, assessed the evidence from published results for the prospective association of GBVH with different health and occupational outcomes.

Methods: Studies I-III are prospective cohort studies. Information about sexual harassment from a superior or colleague, sexual harassment from a third party (not used in study I) and gender harassment from a superior or colleague was obtained from three survey items from the Swedish Work Environment Survey (SWES). SWES is a cross-sectional survey and conducted biannually on a representative sample of the Swedish working population. Data from several cohorts was pooled (Study I: cohorts 1999-2013, N: N=64 297, Study II: cohorts 2007-2013, N=23 449, study III: cohorts 1995-2013, N: 86 033). Information about the respective outcomes, and demographic and occupational information were linked from multiple registers. The continuous follow-up ranged from 1 year to 20 years in the respective studies.

In study I, relative rates (RR) of long-term sickness absence in the year after baseline were determined using modified Poisson regression. In study II, hazard ratios (HR) of incident use of psychotropic medication and in study III, the HR of alcohol-related morbidity or mortality were estimated with Cox proportional hazard models.

In study IV, bibliographic databases were systematically searched for prospective studies concerning exposure to GBVH in the work context and a health or manifest occupational outcome. After quality assessments, the results from the eligible studies with medium or high quality were grouped in clusters with similar exposures and outcomes and summarized in a narrative synthesis.

Findings: Study I showed weak statistically significant associations of reoccurring sexual and gender harassment with long-term sickness absence (RRs ranged from 1.04 to 1.06). Study II showed statistically significant associations of sexual and gender harassment with incident psychotropics use (HRs ranged from 1.20 to 1.31) and study III with alcohol-related morbidity and mortality (HRs ranged from 1.5 to 2.88).

In the systematic review, results from 29 eligible studies were summarized. They investigated mostly sexual violence and harassment (SVH). There were too few studies about physical health or occupational outcomes to synthesize the results and no consistent evidence for a prospective association of SVH with sickness absence. There was consistent evidence of an association of SVH with subsequent poor mental health and indications of an association with hazardous substance use.

There was no consistent evidence of gender differences in the association of the different kinds of GBVH with the investigated health outcomes in any of the four studies of the thesis. Furthermore, there was no consistent evidence of a difference in the association of sexual harassment from a member of the work organization compared to harassment from a third party with the investigated health outcomes.

Conclusions: Experiences with GBVH, both of a sexual and non-sexual nature are a risk factor for mental health and harm from alcohol use. If GBVH contributes to long-term sickness absence is unclear. Women and men appear to be similar in their susceptibility to the health effects of GBVH. Sexual harassment appears to be harmful regardless if it stems from a member of the organization or a third party.

More prospective and longitudinal studies with precise concepts and reliable assessments of different kinds of GBVH and objective health outcomes are needed to fully comprehend the health impact of work-related GBVH.