Article Text
Abstract
Objectives The authors conducted a population-based survey to examine gender differences in occupational exposure patterns and to investigate whether any observed differences are due to: (a) gender differences in occupational distribution; and/or (b) gender differences in tasks within occupations.
Methods Men and women aged 20–64 years were randomly selected from the Electoral Roll and invited to take part in a telephone interview, which collected information on self-reported occupational exposure to specific dusts and chemicals, physical exposures and organisational factors. The authors used logistic regression to calculate prevalence ORs and 95% CIs comparing the exposure prevalence of males (n=1431) and females (n=1572), adjusting for age. To investigate whether men and women in the same occupation were equally exposed, the authors also matched males to females on current occupation using the five-digit code (n=1208) and conducted conditional logistic regression adjusting for age.
Results Overall, male workers were two to four times more likely to report exposure to dust and chemical substances, loud noise, irregular hours, night shifts and vibrating tools. Women were 30% more likely to report repetitive tasks and working at high speed, and more likely to report exposure to disinfectants, hair dyes and textile dust. When men were compared with women with the same occupation, gender differences were attenuated. However, males remained significantly more likely to report exposure to welding fumes, herbicides, wood dust, solvents, tools that vibrate, irregular hours and night-shift work. Women remained more likely to report repetitive tasks and working at high speed, and in addition were more likely to report awkward or tiring positions compared with men with the same occupation.
Conclusion This population-based study showed substantial differences in occupational exposure patterns between men and women, even within the same occupation. Thus, the influence of gender should not be overlooked in occupational health research.
- Occupational exposures
- gender
- gender differences
- workforce survey
- epidemiology
- health and safety
Statistics from Altmetric.com
What this paper adds
The majority of occupational health and exposure assessment studies have traditionally been carried out in men. Therefore, very few studies have compared the distribution of occupational risk factors between women and men.
There are substantial differences in occupational exposure patterns between men and women, and these disparities were observed both between and within occupations.
The influence of gender should not be overlooked in occupational-health research.
Introduction
Women's work has traditionally been considered safe and less hazardous to health in comparison with men's work.1 This has resulted in a lack of information on occupational hazards for women workers,2 and our knowledge of occupational health has mainly been based on studies of men. However, clear gender differences in occupational morbidity have been observed, with males generally having higher rates of work-related injuries, cancer, hearing disorders and vibration-related diseases, whereas females are more likely to be affected by upper-musculoskeletal disorders, workplace bullying and sexual harassment.3 Gender differences in occupational distribution that is, men and women working in different jobs and therefore being exposed to different risk factors, play an important role in many of these differential outcomes.3 However, differences in occupational morbidity have also been observed for men and women with the same job title, suggesting that even in the same occupation, men and women are not equally exposed to particular risk factors for disease.
However, understanding gender differences in occupational exposure, both between and within occupations, is a necessary first step towards understanding gender differences in occupational morbidity. Very few studies have investigated the prevalence of occupational risk factors in women workers, or compared the distribution of risk factors between women and men. We conducted a population-based survey to examine gender differences in occupational exposure patterns and to investigate whether any observed differences can be explained by: (a) gender differences in occupational distribution; and/or (b) gender differences in tasks within occupations.
Methods
We conducted a nationwide telephone survey of a random sample of the New Zealand population aged 20–64 years over a 2-year period (2004–2006). The detailed study methodology is described elsewhere.4 Briefly, 10 000 potential participants were randomly selected from the Electoral Roll and sent a letter of invitation to take part in a telephone interview. The interview obtained information on lifetime work history (for jobs with a minimum of 6 months' duration), current exposures and workplace practices, and questions on selected health outcomes.
Participants were asked whether the following exposures were present (yes/no) in their current work environment: dust; smoke or fume; gas; oils and solvents; acids or alkalis; fungicides, insecticides, herbicides or timber preservatives; and other chemical products (including dyes, inks and adhesives). If a participant indicated exposure(s), they were also asked to state the name and source of the substance. Participants were also asked how often their current occupation involved exposure to physical and organisational factors, including awkward or tiring positions, awkward grip or hand movements, lifting, carrying out repetitive tasks, working at very high speed, working to tight deadlines, standing, using tools that vibrate and loud noise (all the time, three-quarters of the time, half of the time, one-quarter of the time or never). The questionnaire obtained information on whether participants worked for at least 3 h between midnight and 05:00 in the previous 4 weeks (night shift) and whether they regularly worked outside the hours of 08:00–17:00 (irregular hours). Participants were also asked to rate how stressful they found their current occupation on a five-point scale (not at all stressful, mildly stressful, moderately stressful, very stressful or extremely stressful). In addition, we asked individuals the ages of the members in their household and whether these people required looking after by the participant. The variable ‘household responsibility’ was created based on a positive report of looking after children (0–18 years) or older dependents (60+ years) in the household.
Occupations were coded using the New Zealand Standard Classification of Occupations (NZSCO) 1999.5 This classification scheme is based on the 1988 International Standard Classification of Occupations (ISCO-88). The NZSCO is a hierarchical skills-based classification which consists of nine major groups (one-digit code); 25 submajor groups (two-digit code); 99 minor groups (three-digit code); 260 unit groups (four-digit code); and 565 occupations (five-digit code). Table 1 shows an example of the different levels of classification.
The choice of the most applicable five-digit occupation code was based on the self-reported job title as well as the respondent's description of tasks.
Differences in current workplace exposure between men and women were assessed using prevalence ORs6 and 95% CIs using the unexposed as the reference group for each occupational exposure. In the case of job stress, individuals reporting no or mild work-related stress formed the reference group. We conducted unconditional logistic regression in STATA v10.0, adjusting for age (continuous variable).
To investigate whether gender differences in exposure were only due to gender differences in occupational distribution or could also be due to differences within occupations, we also conducted matched analyses where each male participant was matched (1:1) with a female participant on current occupation using the five-digit NZSCO code. ORs and 95% CIs were calculated using conditional logistic regression adjusting for age.
Categories of specific occupational exposures (yes/no), for example, acetone, caustic soda and timber treatment, were created using a word-search programme developed in SAS (version 9.1). The programme was designed to search keywords (including alternative spelling and trade names) in the ‘name of substance’ and ‘source of substance’ text fields. For each newly created exposure category, the original text was checked to ensure that the new category captured all of the exposed participants.
Results
A total of 3003 interviews were completed (the response rate was 37%). The characteristics of the total sample and the sample of males and females matched on occupation are described in table 2. Women comprised just over one-half of the total sample. In both samples, there was a higher proportion of females in the 35–44-year age group, a lower proportion in the oldest age group (55+ years) and a slightly higher proportion of Māori (the indigenous population of New Zealand) females than males. In the total sample, there were higher proportions of females in the professionals, technicians and associate professionals, clerks, and service and sales workers one-digit occupational group, whereas there were higher proportions of males in the legislators, administrators and managers, agricultural and fishery, trades, and plant and machine operators and assemblers one-digit occupational group. There were similar numbers of men and women in the elementary one-digit occupational group (eg, cleaners, labourers, packers and rubbish collectors).
Table 3 compares the prevalence of various occupational exposures between males and females in the total sample (n=3003), and between males and females in the same five-digit level occupation (referred to as the matched sample; n=1208). In the total sample, male workers were more than twice as likely to report exposure to dust and chemical factors. Males were also more likely to be exposed to loud noise (OR=2.70; 95% CI 2.29 to 3.18); use tools that vibrate (OR=3.80; 95% CI 2.94 to 4.90); work night shift in the previous month (OR=2.57; 95% CI 1.89 to 3.50); and work irregular hours (OR=1.76; 95% CI 1.44 to 2.15). On the other hand, women were 32% more likely to report carrying out repetitive tasks (OR=0.76; 95% CI 0.65 to 0.89) and 33% more likely to report working at very high speed (OR=0.75; 95% CI 0.65 to 0.87) a quarter of the time or more. Females were also 52% more likely to report looking after children (0–18 years) or older dependents (60+ years) in their household (OR=0.66; 95% CI 0.57 to 0.78).
Matched sample
There were 827 male participants with no female match for five-digit occupational code, and they were subsequently excluded from the matched analyses, which included 151 different occupations. For most exposures, matching on five-digit occupation attenuated gender differences in exposure (ie, for the dust and chemical factors, 67–87% of the excess risks observed for males were due to gender differences in occupational distribution). However, compared with women with the same occupation, men were still more likely to report exposure to smoke/fume/gas (OR=1.54; 95% CI 1.11 to 2.14) and oils and solvents (OR=1.62; 95% CI 1.16 to 2.27). Men were also twice as likely to use tools that vibrate (OR=2.06; 95% CI 1.29 to 3.29) and work irregular hours (OR=1.97; 95% CI 1.37 to 2.83), and three times more likely to work night shifts (OR=3.32; 95% CI 1.73 to 6.36) compared with women with the same occupation. Men were also about 50% more likely to report job stress compared with women in the same occupation.
Compared with men with the same occupation, female workers were 28% more likely to report carrying out repetitive tasks (OR=0.78; 95% CI 0.59 to 1.01), 43% more likely to report working at very high speed (OR=0.70; 95% CI 0.55 to 0.89), and 37% more likely to report awkward or tiring positions (OR=0.73; 95% CI 0.57 to 0.92). The elevated odds for women of household responsibility attenuated after matching on occupation but remained elevated by 32% (OR=0.76; 95% CI 0.59 to 0.98).
Specific occupational exposures
Table 4 compares the prevalence of specific occupational exposures between men and women for the total and matched samples.
Of the 61 specific exposures under study, 43 were more common among men (p<0.05) in the total sample. Specific exposures more than 10 times more common in the male working population compared with the female working population included: hydraulic oil; welding fumes; paint thinner; paint dust; kerosene; diesel fuel; printing; insulation material; sulfuric acid; timber treatment; fibreglass; and cutting fluid. Of the 61 specific exposures under study, six were more common among women (p<0.05): hair dye; textile dust; household dust; environmental tobacco smoke; bleach; and disinfectant. When comparing men and women with the same five-digit occupation, four exposures remained significantly more common in men: welding fumes, herbicides, wood dust and solvents.
Discussion
This study aimed to estimate the gender differences in occupational exposure in a representative sample of the working population of New Zealand.
The study has several limitations. First, the response rate was relatively low (37%), and the implications of this are discussed in more detail elsewhere.4 7 Briefly, differences between participants and non-participants were observed for age, ethnicity, deprivation and certain one-digit occupational groups. However, these differences were similar for males and females, and gender did not appear to be a significant determinant of refusal to participate or non-contact.7 The analyses presented here are also adjusted for possible determinants of non-response (eg, age), and it is therefore unlikely that the observed gender differences in exposure are due to gender differences in survey participation.
Second, the characterisation of exposure in this study was based on self-report, so we cannot rule out the possibility that gender differences in reporting or perception of exposure contributed to the results. However, a Swedish study reported no differences in the validity of reporting physical risk factors for musculoskeletal disorders between men and women.8 Third, the analyses only assessed exposure prevalence in the current occupation and did not take into account duration or intensity of exposure, which may also impact on gender differences in exposure and ultimately gender differences in occupational health. For example, female workers are more often employed part time and, therefore, more likely to experience shorter exposure duration.
On the other hand, this study has several important strengths. First, the men and women covered by this survey were representative of the total working population, as opposed to similar studies that were limited to selected occupation or industry groups.9 10 Second, gender differences were investigated for a wide range of occupational exposures, including specific chemical and physical exposures, as well as organisational factors and stress, thus not only focusing on exposures traditionally associated with men's work. Third, this is the first study that not only quantified the gender differences in occupational exposure at the population level, but also investigated whether any gender differences in occupational exposure exist for men and women working in the same occupation. Therefore, it was possible to investigate whether the observed gender differences in occupational exposure were entirely due to (a) the segregation of men and women into different occupations or could also be due to (b) men and women with the same occupation carrying out different tasks.
This study found that men were more likely to be exposed to many of the workplace substances under study. They were also more likely to be exposed to loud noise, vibrating tools, night-shift work and irregular hours. On the other hand, women were more likely to report repetitive tasks, working at high speed and exposure to certain workplace substances. These findings were not surprising, considering that throughout the industrialised world, men and women are concentrated in different occupations and industries; women are more likely to work as professionals (particularly in health and education), service and sales workers, and clerks (typically characterised by fast-paced and repetitive work tasks), while men are more likely to work in the agricultural, trades and manufacturing sectors (typically characterised by exposure to dusts and chemicals, and to physically demanding tasks such as heavy lifting). In this study, the actual gender differences in occupational exposure prevalence are quantified, showing that for general exposure categories, the gender difference in prevalence can be up to a factor of 4, while for more specific occupational exposures, the gender difference can be more than a factor of 10. The reported quantitative estimates for exposure prevalence in men and women can be useful in occupational-health studies focusing on both genders and where information on exposure is not available.
We also found that the different occupational distributions of men and women explained most of the observed differences in exposure prevalence; 67–87% of the excess in exposure to workplace substances, 88% of the increased odds of exposure to loud noise and 100% of the moderate excess in exposure to lifting were explained by differences in occupation between men and women. The European Working Conditions Survey also reported that occupational distribution played a major role in observed gender differences in chemical and physical exposures.11 In contrast, a cross-sectional study in the USA (36% women) reported that there were no major differences between men and women for exposure to dust, chemicals, noise and hand repetitions before and after adjusting for industry/job. However, this study examined only four industries in a limited socio-economic range.12
This study also found that, even after accounting for gender differences in occupation, gender differences in prevalence remained for several exposures; males were still more likely to report exposure to smoke/fume/gas, oils and solvents, vibrating tools, night-shift work and working irregular hours compared with females in the same occupation. Female workers were more likely to report repetitive tasks, working at very high speed, and awkward or tiring positions compared with their male counterparts. For the specific categories of workplace substances, in general, men were more likely to report exposure than women, even within the same occupation.
The observed gender differences within occupations could be due to gender differences in the assignment of tasks and/or men and women working in different industries. Men and women with the same occupation do not always carry out the same tasks. A few studies have examined men and women with the same job title and reported a gendered assignment of tasks.1 9 13 14 Several studies have reported that women are more likely to perform repetitive tasks compared with men in the same occupation.10 15 16 However, these studies have been based on small numbers or one specific occupation. Differences in task assignments by gender could be due to the different physical capabilities of men and women or socialised gender roles. The greater relative stature and muscular strength of men compared with women makes them more suited to physically demanding tasks such as heavy lifting. The same argument can be applied to the suitability of tasks involving dexterity and precision (and therefore repetitiveness) for women. However, Messing argues that social constructions of what is suitable work for men and women also play a role, and many physical tasks assigned to men can be adapted to women.1 For example, in an observational study of hospital cleaners, Messing et al concluded that there was no clear reason why women could not perform the ‘heavy’ work typically assigned to men.14
The observed differences within occupations could also be due to gender differences in industry distribution. In addition to gender differences in occupational distribution, men and women are typically concentrated in different industries. In the current study, men were more likely to work in the manufacturing and construction industries, and females were more likely to work in the health and community services and education industries (data not presented). Thus, a specific job title could entail different exposures depending on the industry context, although industry is less likely to be important for some occupations (eg, agricultural and trades workers) than others; for example, for the job title ‘sales worker,’ females are more likely to work in retail sales, whereas men are more likely to work as manufacturers' representatives (cited in Messing et al2). Therefore, while it is unclear whether gender differences within occupations are due to gender differences in task assignments and/or industry distribution, it is evident from our study that job titles, which are often used in studies of occupational health, may be poor surrogates for occupational exposure.
In the current study, women were not more likely to report high levels of job stress. Men were 52% more likely to report moderate stress and 43% more likely to report a very or extremely stressful job compared with females in the same occupation. A New Zealand study of 41 companies reported a higher risk of work-related stressors for men, even after adjusting for occupation.17 While women generally report more workplace stressors and poorer psychosocial working conditions than men, the current evidence for a gender difference in work-related stress as an outcome is inconsistent.3
The dual exposure of work demands and responsibility for the household may also have an impact on the relationship between exposure and health, particularly for women. The responsibilities for housework and childcare still largely fall on women. The current study found that women were 52% more likely to report looking after children (0–18 years) or older dependents in their household compared with men. This excess was attenuated after matching on occupation but remained elevated at 32%. The New Zealand Time Use Survey reported that women carry out more unpaid work than men regardless of employment status.18 The burden of household responsibilities may result in increased exposure to physically demanding activities as well as less time for recovery after work. Several studies have shown that women spend less time exercising or relaxing outside work hours compared with men.15 In addition, the responsibility for childcare may also restrict the type of work women can do—for example, part-time work, which is often associated with routine and repetitive work and lower job control. Unpaid work is an important aspect of women's occupational health and safety, and is rarely investigated in epidemiological studies.
In conclusion, this study has found that even in a country such as New Zealand, with relatively good gender equality in employment opportunities, men and women are far from equally exposed to occupational risk factors for disease. Male workers continue to experience a heavier burden of occupational exposure to most dusts and chemicals, and certain physically demanding tasks. However, it should be noted that for many of these occupational exposures, the prevalence in women workers was not negligible. For example, 7% of female workers reported exposure to solvents, and 4% reported exposure to engine emissions in the overall sample. Compared with men, women workers had a higher prevalence of repetitive tasks, working at very high speed, and certain exposures such as disinfectants, hair dyes and textile dust. Gender differences in exposure were in part explained by gender differences in occupational distribution, while for some exposures, differences in prevalence were even observed for men and women with the same occupation. Therefore, there is a need for caution when using job titles as a surrogate for exposure. This study illustrates that gender has a substantial impact on occupational exposure prevalence, even within the same occupation, and that the influence of gender should not be overlooked in occupational health research.
Acknowledgments
We thank T Whaanga, Z Harding, C Priest, P Whitson, M Skelly, P Taptiklis, E Drummond, A McCarty, N Holland, K Gray, A Hoskins, A Thomson, J Fargher, C Blackwell, E Turner, S Richards, K Crothall, A Harding, J Wilkie, J Dow and T McKenzie, who conducted the interviews, and R Jones, A Harding, Z Harding, A Thomson, C Tromop van Dalen, J Fargher, C Priest and M Hudson, who completed the data entry.
References
Footnotes
Funding This project was funded by a Joint Research Portfolio of the Health Research Council, the Accident Compensation Corporation and the Department of Labour, which issued a Request for Proposals (RFP) for a study of the burden of occupational ill-health in New Zealand (HRC 04/072). The Centre for Public Health Research is supported by a Programme Grant from the Health Research Council of New Zealand (HRC 02/159).
Competing interests None.
Ethics approval Ethics approval was provided by the Massey University Human Ethics Committee (WGTN 03/133).
Provenance and peer review Not commissioned; externally peer reviewed.