Article Text

PDF

Original article
Human service work, gender and antidepressant use: a nationwide register-based 19-year follow-up of 752 683 women and men
  1. André Buscariolli1,
  2. Anne Kouvonen1,2,3,
  3. Lauri Kokkinen4,5,
  4. Jaana I Halonen4,
  5. Aki Koskinen4,
  6. Ari Väänänen4,6
  1. 1Faculty of Social Sciences, University of Helsinki, Helsinki, Finland
  2. 2SWPS University of Social Sciences and Humanities, Wroclaw, Poland
  3. 3Administrative Data Research Centre Northern Ireland, Centre for Public Health, Queen’s University Belfast, Belfast, UK
  4. 4Finnish Institute of Occupational Health, Helsinki, Finland
  5. 5Faculty of Social Sciences, University of Tampere, Tampere, Finland
  6. 6School of Social Policy, Sociology and Social Research, University of Kent, Canterbury, UK
  1. Correspondence to Dr Anne Kouvonen, Faculty of Social Sciences, University of Helsinki, Helsinki 00014, Finland; anne.kouvonen{at}helsinki.fi

Abstract

Objectives To examine antidepressant use among male and female human service professionals.

Methods A random sample of individuals between 25 years and 54 years of age (n=752 683; 49.2% women; mean age 39.5 years). Information about each individual’s filled antidepressant prescriptions from 1995 to 2014 was provided by the Social Insurance Institution. First, antidepressant use in five broad human service categories was compared with that in all other occupations grouped together, separately for men and women. Then, each of the 15 human service professions were compared with all other occupations from the same skill/education level (excluding other human services professions). Cox models were applied and the results are presented as HRs for antidepressant use with 95% CIs.

Results The hazard of antidepressant use was higher among men working in human service versus all other occupations with the same skill/occupational level (1.22, 95% CI 1.18 to 1.27), but this was not the case for women (0.99, 95% CI 0.98 to 1.01). The risks differed between professions: male health and social care professionals (including medical doctors, nurses, practical nurses and home care assistants), social workers, childcare workers, teachers and psychologists had a higher risk of antidepressant use than men in non-human service occupations, whereas customer clerks had a lower risk.

Conclusions Male human service professionals had a higher risk of antidepressant use than men working in non-human service occupations. Gendered sociocultural norms and values related to specific occupations as well as occupational selection may be the cause of the elevated risk.

  • mental health
  • epidemiology
  • psychiatry
  • longitudinal studies

Statistics from Altmetric.com

Key messages

What is already known about this subject?

  • Human service professionals have high emotional demands at work and they are at a higher risk of antidepressant use.

What are the new findings?

  • Hazard of antidepressant use was higher among men working in human service versus men in non-human service occupations in the same skill/occupational level.

  • When looking at all human service professionals together, even if the overall prevalence of antidepressant use was higher among women than men, the corresponding risk for these women was not significantly different from the female population working in non-human service occupations.

How might this impact on policy or clinical practice in the foreseeable future?

  • The nature of face-to-face interaction, and the emotional labour associated with it, seems to influence mental health outcomes through gender-specific pathways.

Introduction

Human service professions are characterised by face-to-face interaction in which the employee provides customers with health, educational, social or commercial services.1 Because people in these occupations (including medical doctors, nurses, practical nurses, teachers and police officers) often work with particularly vulnerable populations, their performance is linked to experienced emotions.2 As such, these professionals engage frequently in emotional labour. They have to suppress their own feelings to ‘sustain the outward countenance that produces the proper state of mind in others’.3 This can easily create unbalanced relationships which can result in the depletion of one’s emotional resources and increase the risk of depressive symptoms.4 5 Indeed, high emotional demands associated with direct client interaction have been linked to affective and stress-related disorders.6 7 However, epidemiological studies on the relationship between emotionally demanding work and the leading mental disorder, depression, have mainly relied on cross-sectional data and self-reported measurements,8 9 which are subject to response bias and the risk of reverse causality.

Emotional demands associated with each occupation have been shown to be gendered.10 Related to gendered social roles and labour market structures, women are more often employed in the emotionally demanding human service professions. Tasks involving professional caring and empathising are often typical for female-dominated occupations and less typical for male occupations,11 and women are generally more engaged in emotional labour.3 12 13 Therefore, it can be expected that men working in emotionally demanding human service professions may have somewhat different social position and behavioural tendency when compared with their female counterparts. For instance, male nurses have been found to emphasise task-oriented behaviours, instead of people-oriented behaviours, to shape their work role to be more masculine.14 The burden of the emotionally taxing occupation may also have adverse gender-specific impacts on mental well-being. In fact, sparse studies on male nurses point to rather poor job satisfaction,15 and a tendency to view their job as a ‘dead end job’.16

To overcome the limitations associated with cross-sectional designs and self-reported measurements, recent studies have used register-based data to investigate mental health in human services professionals.1 17–19 Yet, from the perspective of gendered working life and occupational hazards these studies also have limitations. Their findings are either restricted to specific occupational groups,19 or have used short follow-ups.1 18 To fill these gaps, our long-term study uses register data for antidepressant prescriptions for a large representative sample of the Finnish working-age women and men with an exceptionally long follow-up period (1995–2014), and with a comprehensive set of occupational groups. Our hypothesis is that human service professionals have a higher risk of antidepressant use than those working in other occupations with the same skill/education level due to the emotional demands associated with their professions. These impacts might be more pronounced among male than female human service professionals.

Methods

Study population

We obtained a randomly selected 33% sample of 18–64-year-old permanent residents of Finland in 1994 from the Population Register maintained by Statistics Finland. For this study, we included individuals between 25 years and 54 years of age (n=752 683; 49.2% women; mean age 39.5 years). This age group was used as the younger professionals would just have entered the profession, whereas those older than 54 years old would retire at some point during the follow-up thereby reducing their exposure time to emotional demands related to their work.

Exposure

The Finnish Population Register maintained by Statistics Finland includes occupational class for each individual in our data at the end of 1994. The occupations are classified according to the 2001 International Standard Classification of Occupations (ISCO) codes. If the person was not employed at this time, he/she did not have an occupation and was excluded. To define emotionally demanding work, we identified 21 occupational groups related to human service. The criterion for emotional labour was that the job tasks involved face-to-face interaction in which the employee provides customers with health, educational, social or commercial services.1 Professions with small numbers of individuals were grouped together with broader occupational categories based on similar job tasks. For instance, ‘Midwifes’ was grouped together with the category ‘Nurses’. Occupations with low numbers that could not be grouped in any larger occupational categories were excluded (prison guards, firefighters, foreign exchange traders and auditors). The resulting 15 occupational groups were further grouped into five larger human service categories following the Wieclaw et al1 classification: health professionals; education professionals; social workers; customer services workers and miscellaneous (including police officers and psychologists).

Outcome

Information about each individual’s filled antidepressant prescriptions (ie, antidepressant use) (Anatomical Therapeutic Chemical code N06A20) from 1995 to 2014 was provided by the Social Insurance Institution that maintains the Finnish Prescription Register. This register includes dates of all drug purchases reimbursed to Finnish residents in non-institutional settings. Mean follow-up time until the first antidepressant use was 15.2 (minimum 0.1–maximum 19.0) years.

Covariates

Information about possible confounders was obtained from the Population Register. For each individual we included sex, age, marital status (married, single, divorced and widowed), income (higher or lower than €17 691 per year), level of education (a university degree vs not), place of residence (Uusimaa vs other counties) and unemployment status (being unemployed for more than 1 month at baseline or not).

Statistical analysis

Cox proportional hazards regression models were used for data analysis (SAS V.9.4, SAS institute, Cary, North Carolina, USA). Results are presented as HRs for antidepressant use with 95% CIs. Individuals were followed-up from the 1 January 1995 until the first purchase of antidepressants, death or 31 December 2014, whichever occurred first.

Antidepressant use in the five broad human service categories was first compared with that in all other occupations grouped together. Then each of the 15 human service occupations were compared with all other occupations from the same skill/education level (excluding other human service professions from the reference group). The similarity of skills was determined by the first character of the four-digit ISCO code, which defines broader occupational classes. For instance, the ISCO code for medical doctors is 2221, the first character ‘2’ refers to ‘specialists’, which includes professions like architects, engineers, business specialists, among others. All non-human service professions whose first character was ‘2’, representing the same skill level, served as the reference group for medical doctors. The code for practical nurses is 5132; other human service professions that also have ‘5’ as the first character (referring to ‘service, sales, and nursing staff’) are sellers and retailers, protection and security employees, among others. All non-human service occupations whose first character was ‘5’ served as the reference group for these professions. This procedure was used to reduce the influence of other work stress factors related to different occupational status groups. The models were adjusted for age, marital status, income, education level, region and baseline unemployment.

Results

The distribution of human service professionals and the proportions of those with at least one antidepressant purchase during the follow-up by gender are shown in table 1. Around 11% of the entire sample (n=7 52 683) were employed in human service professions (n=85 849). Among human service professionals, health professionals, education professionals and customer services were the largest groups. Women were more predominantly featured in these professions than men; 85% of all human service professionals were women. Women had used antidepressants more often than men: in human service professionals, 34% of women had at least one antidepressant medication purchase during the follow-up compared with 27% of men. About 19% of the study participants who had purchased antidepressants had only one purchase during the follow-up while the majority had two or more antidepressant purchases.

Table 1

Distribution of human service professionals among male and female employees and the proportions of those with at least one antidepressant purchase during the follow-up (n=752 683)

Table 2 presents the sociodemographic characteristics of human service professionals and the hazards of antidepressant use by those characteristics. A higher proportion of women had high education compared with men (52% and 38%, respectively). However, men had higher incomes: 85% of the male population earned more than €17 691 per year, whereas the corresponding figure was only 60% for women. Men in human service occupations were on average 39.5 (SD=8.3) years old, while the mean age among all employed men in our data was 39.5 (SD=8.3) years. The proportion of men with a university degree among human service professionals was 62% and among all employed men 47%.

Table 2

Sociodemographic characteristics and antidepressant use in human service professionals by gender (n=85 849)

As table 2 further shows, being divorced and being widowed were associated with an increased hazard of antidepressant use among human service professionals (HR 1.39, 95% CI 1.36 to 1.41 for divorced men; and HR 1.47, 95% CI 1.44 to 1.49 for divorced women). Having high income and being unemployed at baseline were associated with an increased use of antidepressants in both genders, whereas living outside of Uusimaa and having a university degree were associated with a slightly decreased hazard. The results of each category were adjusted for other sociodemographic characteristics.

Table 3 presents the hazards of antidepressant use in the five broad categories of human service professions compared with all other occupations from the same skill/education level. The adjusted HR for antidepressant use in all human service professionals together was 1.22 for men (95% CI 1.18 to 1.27), and 1.00 for women (95% CI 0.98 to 1.01) when compared with non-human service employees. In both women and men, social workers had an increased risk (HR 1.62, 95% CI 1.46 to 1.81 for men and HR 1.19, 95% CI 1.13 to 1.24 for women). In education professionals, men had a slightly higher hazard for antidepressant use (HR 1.09, 95% CI 1.03 to 1.16) when compared with non-human service employees, while the opposite was observed for women (HR 0.92, 95% CI 0.90 to 0.95).

Table 3

HRs, 95% CIs for antidepressant use in the five broad human service profession categories, by gender

Gender differences could be seen when we look at the 15 specific human service professions separately (see online supplementary table 1 for the unadjusted results, and table 4 for the adjusted results). Elevated hazards for antidepressant use for both men and women were found for the following professions: medical doctors, practical nurses, home care assistants, childcare workers, social workers, social care workers and special education teachers. In general, however, the associations were stronger in men and the strongest association was found for male social workers (HR 2.11, 95% CI 1.80 to 2.47). Among primary school teachers, the HR for antidepressant use was 1.15 (95% CI 1.03 to 1.27) in men, and 0.91 (95% CI 0.85 to 0.97) in women. The risk for antidepressant use was also increased among male nurses (HR 1.55, 95% CI 1.32 to 1.82), but not in female nurses, and in male (HR 1.36, 95% CI 1.08 to 1.73) but not female psychologists. In addition, women police officers had an increased risk of antidepressant use. On the other hand, male service clerks and women in travel services had a lower hazard of antidepressant use (0.85, 95% CI 0.75 to 0.98 and 0.83, 95% CI 0.70 to 0.98, respectively). We also performed sensitivity analyses and modelled the risk of having more than one purchase (a proxy for repeated antidepressant use). All main results remained highly similar, the HR became statistically non-significant only among men working as special education teachers (HR 1.20, 95% CI 0.94 to 1.53) and psychologists (HR 1.27, 95% CI 0.97 to 1.66).

Supplementary file 1

Table 4

Adjusted HRs, 95% CIs of antidepressant use in human service professions when compared with non-human service employees at the same skill/education level, by gender

Discussion

In this large register-based study, we found that particularly male human service professionals were at a higher risk of antidepressant use than their counterparts within the same occupational grade but working in non-human service occupations. Male health professionals and social workers had the most increased risks. However, it is important to note that the prevalence of psychological distress and depression in the overall population is generally higher among women than men.21 22 Thus, although the prevalence of antidepressant use among female nurses, and education professionals was higher than that of their male counterparts, the corresponding risk for these employees was not significantly different from the overall female population with a similar skill/educational background.

It has been argued that occupations in which one of the genders is in minority are associated with a higher risk of poor mental health;23 24 however, this distinction does not account for all the differences found. For instance, although male customer service employees were also a gender minority, they were found to have a lower risk of antidepressant use than their counterparts with similar skill/educational background. One reason for this can be that male customer services employees might have been subjected to work stress factors that are also common to non-human service occupations. What essentially differentiates these occupational groups where gender minorities are common is that health and social care professionals, education professionals and social workers are more likely to be exposed to compassion fatigue, that is, the emphatic exhaustion associated with the care of clients.25 Accordingly, the emphatic engagement of health and social care professionals with their patients and clients may make these employees more prone to develop vicarious traumatisation, burn-out syndrome, secondary traumatic stress, compassion fatigue and traumatic counter transference.26–28 Thus, our findings suggest the higher risk of antidepressant use in human service professionals is specifically related to the nature of face-to-face interaction and the emotional demands associated with it.

Because personality traits are important determinants of occupational choices29 as well susceptibility to depression,30 a competing explanation is that men selected into human services are more prone to develop mental illness to begin with,1 regardless of the emotional demands associated with their tasks. However, to the best of our knowledge, no previous studies have addressed the relationship between personality traits, occupational choices and susceptibility to depression. Neither had different studies addressing only one of these characteristics observed results that could suggest a converging point. For instance, high levels of openness/intellect are associated with counter gender-stereotypical occupational choices,29 whereas high neuroticism, low extraversion and conscientiousness are linked to depression.30 Although occupational selection may affect our results, it is not possible to evaluate its magnitude from register-based data.

It is well known that externalising disorders, defined by antisocial behaviour and substance abuse are considerably at a higher level among men than women.30 Especially in the masculine occupational cultures (eg, metal industry) behavioural patterns are more inclined towards non-psychological coping strategies.30 This may also lead to the underuse of antidepressants among men working in non-human service occupations. Female professional cultures may also explain our findings (2; 15; 16; 17). Male doctors, nurses, teachers and social workers are subjected to a female type of occupational culture, which entails certain behavioural standards and social norms. Among male employees who work in the female dominated sector this may lead to the development of occupation/organisation-specific coping methods (ie, the use of antidepressants) closer to the average of the female rather than the average male population. Men who work in the female dominated professions of caring and empathy may therefore treat themselves in a more feminine way. This may partly explain the high risk of antidepressant use among men working in some human service occupations when compared with other men working in non-human service occupations. However, there may also be occupational selection effects.

Strengths and limitations

The main strengths of our study include the use of nationally representative register data for antidepressant prescriptions with an exceptionally long follow-up period (1995–2014), and a comprehensive set of occupational groups. The use of self-administered questionnaires is subject to misclassification bias,8 but the use of register data reduces this risk since purchase of antidepressant medication is only possible with a clinical diagnosis. Having a comprehensive set of occupational groups allowed us to assess the risk of antidepressant use in different human service professions compared with all other non-human service occupations with the same skill/education level together. This is an important addition to the literature as it allowed us to discuss how the emotional demands of different occupations are associated with antidepressant use. In addition, by including only occupations from the same skill/education level in the reference group we reduced the influence of work stress factors that are not directly associated with emotional demands.

Despite these strengths, our study also had some limitations. First, we used antidepressant use as a proxy for depression, however, even if the main indication of antidepressant use is depression, they are increasingly prescribed to treat other conditions such as anxiety, insomnia and chronic pain.15 Second, we cannot infer from the purchase of antidepressants that individuals are effectively adhering to the medication. Whereas our dependent variable was at least one antidepressant purchase during the follow-up, it has been found that up to 50% of the patients may discontinue the treatment prematurely.31 32 Third, the occupation of each individual was determined at the start of the follow-up period and possible changes could not be traced.

Conclusions

In this large-scale register study, we found that people working in occupations that contain human service are at increased risk of antidepressant use. The risks, however, differed between occupations and between women and men, the relative risks being the highest among male social workers when compared with the overall male population with the same skill/education level that does not work in human service. Although the incidence of antidepressant use was higher among women in human services professions, it was not significantly different from the overall female population. The nature of face-to-face interaction, and the emotional labour associated with it, seems to influence mental health outcomes through gender-specific pathways. For instance, while male social workers and health professionals had higher risks of antidepressant use, customer service employees had lower risks. In addition, being in a gender minority might also be associated with the observed higher risk of antidepressant use. Finally, the role of gender-specific occupational culture on antidepressant use provides an additional interpretation of our results.

References

View Abstract

Footnotes

  • Contributors AB, AKou, LK, JH, AKos and AKPV conceived and designed the experiments, AB and AKos analysed the data, AKos and AKPV contributed to data collection. AKPV is the guarantor of the study. All authors were involved in writing the paper and approved the submitted and published versions.

  • Funding This study was supported by the Academy of Finland (grant number 267172). AKou was supported by the Economic and Social Research Council (ESRC) (grant number ES/L007509/1) and the Medical Research Council (MRC) (Grant MR/K023241/1). The funders had no role in the data collection, analysis or writing of the paper.

  • Disclaimer The corresponding author (AKou) affirms that the manuscript is an honest, accurate and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

  • Competing interests None declared.

  • Patient consent Detail has been removed from this case description/these case descriptions to ensure anonymity. The editors and reviewers have seen the detailed information available and are satisfied that the information backs up the case the authors are making.

  • Ethics approval The Ethics Committee of the Finnish Institute of Occupational Health.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Because of legal reasons, we are unable to deposit the register data used in the present study with an open data repository. Interested researchers need to obtain these register data directly from the relevant data custodians: the Social Insurance Institution (SII) of Finland and Statistics Finland.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles