Action to improve women's occupational health has been slowed by a notion that women's jobs are safe and that any health problems identified among women workers can be attributed to unfitness for the job or unnecessary complaining. With increasing numbers of women in the labor force, the effects of work on women's health have recently started to interest health care providers, health and safety representatives and researchers. We begin our summary of their discoveries with a discussion of women's place in the workplace and its implications for occupational health, followed by a brief review of some gender-insensitive data-gathering techniques. We have then chosen to concentrate on the following four areas: methods and data collection; directing attention to women's occupational health problems; musculoskeletal disease; mental and emotional stress. We conclude by pointing out some neglected occupational groups and health issues. [Article copies available for a fee from The Haworth Document Delivery Service: 1-800-342-9678. E-mail address: getinfo@haworth.com]

ABSTRACT. Action to improve women's occupational health has been slowed by a notion that women's jobs are safe and that any health problems identified among women workers can be attributed to unfitness for the job or unnecessary complaining. With increasing numbers of women in the labor force, the effects of work on women's health have recently started to interest health care providers, health and safety representatives and researchers. We begin our summary of their discoveries with a discussion of women's place in the workplace and its implications for occupational health, followed by a brief review of some gender-insensitive data-gathering techniques. We have then chosen to concentrate on the following four areas: methods and data collection; directing attention to women's occupational health problems; musculoskeletal disease; mental and emotional stress. We conclude by pointing out some neglected occupational groups and health issues. [Article copies available for a fee from The Haworth Document Delivery Service: 1-800-342-9678. E-mail address: getinfo@haworth.com]

KEYWORDS. Occupational health; women workers; methodology; musculoskeletal problems; stress

This paper focuses on four areas: (1) Methodological problems; (2) Directing attention to women's occupational health problems; (3) Musculoskeletal disease; and (4) Mental and emotional stress, including balancing home and professional responsibilities. Needless to say, many other topics could have been discussed. Risks for pregnant women are not given priority here because, in our view, these hazards have already been emphasized in research and prevention activities, to the exclusion of other important issues for women; however, we have integrated reproductive problems in other sections of the text and discuss work-related menstrual problems in our list of neglected research areas. We have not covered occupational cancer because of the recently published bibliography of studies of occupational cancer among women, listing them by profession and by cancer site (Zahm et al., 1994).(1)

The Growing Interest in Women's Occupational Health

The majority of North American women are employed full-time. Younger women have much higher rates of participation in the labor force than older women, whether or not they have children. Although women with children under three years of age participate less than others, this tendency is disappearing (Armstrong & Armstrong, 1993).

Although, overall, women's employment has a positive effect on their health (Walters, Lenton & McKeary 1994: p. 15) and employed women live longer than unemployed women or housewives (Silman, 1987; Waldron, 1991), risk factors present in some jobs may adversely affect women's health. Action to improve women's occupational health has been slowed by a notion that women's jobs are safe and that any health problems identified among women workers can be attributed to unfitness for the job or unnecessary complaining. However, the rise in the number of women in the labor force has sensitized some health care workers and scientists to the need to include women's concerns in their occupational health activities (Health Canada, 1993; Messing, Neis & Dumais, 1995).

Although it is too soon to speak of convergence of research directions, certain occupational groups have interested researchers: factory workers in repetitive tasks, hospital workers, solvent-exposed workers. Particular industries have received attention, because health problems have been identified (clothing, food-processing) and/or because they employ a large number of women (health care). Methodological issues have been discussed: How to analyse data by gender; how to take into account the different life patterns of women and men; how to identify health problems which arise in women's traditional work and for which strategies may not have been developed. Political questions are also asked: How to make sure that recognition of health hazards in women's work does not lead to denial of employment opportunities for women; should men and women be distributed in a more random way across employment categories or is each gender more "suited" to a specific type of work.

METHODOLOGICAL ISSUES

There are structural barriers to understanding the occupational health of women: (1) women and their work have been excluded from consideration by scientists in occupational health; (2) data collection has been gender-insensitive; (3) data analysis has been gender-insensitive; and (4) health outcomes analysed have not included effects from women's jobs.

Exclusion of Women and Their Work

In the past, women were excluded from studies of occupational health. Greenberg and Dement (1994) found a large excess of studies of occupational disease involving only males. A well-documented example is the field of occupational cancer. Zahm and colleagues found that of 1233 cancer studies published in 1971-1990 in the 8 major occupational health journals, only 14% presented analyses of data on white women and only 10% on non-white women (Zahm et al., 1994). Part of the reason for exclusion is the choice of jobs for study: miners, refinery workers and foundry workers are almost always men.

These exclusions create a circular situation where there is evidence of health problems only among men, leading to a reluctance to study women because of an impression that not many women get occupational disease. One consequence is that workers are studied as if they were all men. For example, male/female biological differences have not usually been taken into account in standard-setting and exposure assesssment. Greenberg and Dement describe the numerous differences that can affect their responses to toxic chemicals.

Measuring Instruments and Data Collection

Problems in government procedures can preclude women from being studied. For example, a Canadian study of the effects of agricultural exposures was forced to eliminate women since only the husband of a farm family was identified as a farmer in most provincial records (Semenciew et al., 1993). Also, many death certificates have not contained information on women's profession, in part because once a woman has retired she may be considered to be a housewife. Thus, a priority for research in women's (and men's) occupational health is establishment of appropriate data bases.

Research instruments and standards which have been derived with all-male populations are sometimes used without further validation on female populations. An example is the well-known Karasek questionnaire on job demands (Hall, 1989), or strength testing done with instruments validated on male populations (Stevenson et al., 1996; Messing & Stevenson, 1996).

Occupational prestige scales and social class scales are often used which use the husband's job to ascribe a score to the wife (Blishen, Carroll, & Moore, 1987; Heller et al., 1984). This causes problems when data on health are adjusted for social class, since some social class influences on health may be mediated through common family-revenue dependent factors such as nutrition and others may be specific to the individual situation such as job parameters.

The Nature of Women's Jobs

Women and men are not distributed at random over the labor force, but are segregated into specific industrial sectors and into female-majority jobs within these sectors (Armstrong & Armstrong, 1993: chapter 1). It has been calculated that for the sexes to be evenly distributed across job titles in North Carolina, three-quarter of women workers would have to change jobs, more than the change required for equitable distribution of black and white workers (Tomaskovic-Devey, 1993). The lists of principal jobs of women and men are very different, and women are more concentrated into a few professions. Women are over-represented in clerical and services jobs and underrepresented in manufacturing and resource generation. Also, women are more apt to work in small workplaces employing less than 20 workers (White, 1993: p. 168). About one-third of women are secretaries and clerks.

There is also vertical segregation in the job market. It is much more common for men to be in positions of authority, as witnessed by the fact that women in the top 20 jobs earned an average of 67% of the income of men in the same professional classification (Armstrong & Armstrong, 1993: p. 43). Women are 3 times as likely as men to work part-time and 3 times as likely to be temporary workers (Armstrong & Armstrong, 1993: p. 50-54).

Even more surprising than the extent of male-female job segregation is that of task segregation. Male cleaners mop and female cleaners dust (Messing, Chatigny & Courville, 1996). Women and men with the same job titles can have very different exposures, a fact which has implications for epidemiological research. For example, male and female gardeners may have different exposures to pesticides because women do more planting and weeding and men do more pruning (Messing, Dumais et al. 1994).

Women are shorter on average than men and are proportioned differently (Pheasant, 1986: chapter 1). Tools and equipment are not always available in the right dimensions for women. Therefore, even the same tasks do not necessarily interact the same ways with men's and women's bodies. Because of their physiology, women may develop ways to do specific tasks which are different from those of the average man. Lortie (1987) showed that, on average, female hospital orderlies lifted patients differently from men; they found ways to change lifting tasks into pushing and pulling tasks, in accordance with differential muscle strength. However, in a rigid, repetitive sorting task, where there was little control over task parameters and where certain dimensions of the work station caused problems for shorter workers, women had more work accidents than men (Courville, Vézina & Messing, 1992).

The fact that women and men are treated differently in the workplace makes it inappropriate to consider that women and men with the same job title have the same working conditions. In health research, using job title as a proxy for exposure may introduce inaccuracy and bias according to the gender of the worker. Hsairi et al. (1992) used expert estimates derived from job titles to classify workers in 13,568 jobs as exposed or not exposed to dust. Workers' reports of their own exposure and self-reports of symptoms of dust exposure (difficulty breathing, asthma) were correlated with experts' ratings. Self-reports of symptoms were better correlated with self-reports of exposure than with experts' estimates of exposure. Expert estimates were significantly closer to men's self-reports than women's. The authors interpret these results as implying a "better perception" of exposure by the men, but it is also possible that experts' estimates of exposure according to job title are based on experience with male job-holders.

These results would support caution in using job title to estimate exposure for both genders if the job exposure matrix has not previously been validated separately by gender. In addition, it may be unwise to adjust relationships between job title and disease incidence for gender, thus treating gender as a confounder when it may be a proxy for specific exposures.

Gender Insensitivity and Oversensitivity

Descriptors representing the place of people in society (gender, race, class) pose a special problem for epidemiological research. They may include higher probabilities of some biological characteristics (hormonal status, blood groups, nutritional status) but they also represent probabilities of different occupational exposures. If researchers simply adjust ("control") their analyses for gender, or if they include gender as a variable along with other exposure variables, the effects of gender-specific exposures may disappear from sight.

"Adjusting" for a variable while analyzing data means using a mathematical procedure to eliminate its effect. It is reasonable, for example, to adjust for smoking when examining the relationship of dust exposure to lung damage, because smoking is an independent determinant of lung damage and might confuse the issue if those exposed to dust smoked more or less than those not exposed. We may need to add a correction factor to the lung function of smokers before testing the relationship between dust exposure and lung damage. This procedure allows us to determine the effect of dust on the lungs while taking into account the well-established deleterious effects of smoking.

"Overadjusting" occurs when the variable adjusted for is a synonym for the exposure. This would happen if those more exposed to dust belonged to a particular ethnic group, perhaps because they were assigned more often to cleaning or maintenance. Adjusting for ethnicity would then diminish the possibility of finding an effect of the dust exposure on lung cancer.

Similarly, women and men have different exposures because of their different jobs. It is therefore appropriate to analyse exposure data for women and men separately before deciding whether they can be pooled. However, very many researchers adjust for gender without previous examination of the data. Studies which examine the health of workers often find that women workers report more symptoms of poor health or psychological distress than their male counterparts. The approach to these differences is often to "adjust" for gender, without any attempt at justification. Other studies treat gender as an independent determinant of health status. Adjusting or treating gender as an independent variable would be appropriate only if gender were an independent determinant of poor health reports, for example if women were weaker or complained more than men, or if the health effects had an important independent contribution from sex hormones or other biological differences (Mergler 1995).

An example is scientific studies in the effects of indoor air pollution. We searched the CISILO data base and found twelve epidemiological studies related working conditions to symptoms thought to be related to indoor air quality (Skov, Valbjorn & Pederson, 1989 & 1990; Norback, Michel & Widstrom, 1990; Hodgson et al., 1991; Harrison et al., 1992; Norback, Torgen & Edling, 1990; Kelland, 1992; Mendell & Smith, 1990; Menzies et al., 1993; Franck, Bach & Skov, 1993; Stenberg et al., 1995; Nelson et al., 1995). Women by and large suffered 2-4 times more symptoms than men, a fact that was mentioned in almost all of the articles. However, only three articles even discussed this difference in symptoms (Hodgson et al., 1991; Menzies et al., 1993; Stenberg et al., 1995), and only Stenberg and colleagues took into account the gendered division of labor in such a way as to allow us to understand why women had more symptoms. (Women in offices probably do more photocopying than men, with consequent increased exposure to ozone, toners and electrostatic effects; they share offices more often, resulting in more exposure to second-hand smoke and less ventilation per person.)

Three studies ignored gender entirely, treating all workers as undifferentiated. One of these noted that nurses had more symptoms than hospital administrators had different levels of symptoms but did not report the gender distribution or exposure specificities of either group (Kelland, 1992). Other studies adjusted the analysis for gender without reporting any details on symptoms or exposure by gender. The authors of one of them concluded their article by insisting on the necessity to control for gender (Stenberg et al., 1995). Seven others reported gender as if it were a separate cause of difference in symptom levels. For example, one reported without any other discussion that women were 2.6 times as likely to suffer central nervous system symptoms than men (Nelson et al., 1995). Another laconically reported "female" as one of a list of associations with symptoms that also included "handling carbonless paper," "hay fever," and "clerk." This kind of insensitive treatment leaves the reader with the impression that being female is a cause of the symptoms, perhaps because of weakness or psychological problems.

Pathology versus Positive Health

Occupational health researchers trained in medicine have often limited their interest to pathologies rather than indicators, signs or symptoms of deterioration in physical or mental states, reasoning that the presence of pathology guarantees that the problem examined is worthy of serious consideration. However, a requirement for diagnosed pathology may be premature when studying women's occupational health. Since the aggressors present in women's traditional work have been understudied, and the effects of even well-known conditions on women workers are often unknown, identification of occupational disease in women's work is embryonic. For example, women who handle money report unusual-looking and painful red streaks on their hands. A literature search revealed one article on nickel allergy among cashiers (Gilboa, Al-Tawil & Marcusson, 1988), but no other reference to skin disease among those handling money. It may be years before sufficient research enables us to decide whether to define it as an industrial disease.

The requirement for pathology has two further consequences. First, it forces the researcher to consider events that are rare among populations still at work. This requirement for populations of considerable size is a particular obstacle to identifying women's occupational health problems because women work in very small workplaces (White, 1993: p. 168).

Second and more important, the risks found in women's jobs are often undramatic and diffuse. In fact, obvious danger is a reason for excluding women from particular jobs. Thus, epidemiological studies which seek to link isolated, identifiable risk factors such as chemical exposures to well-defined pathologies are not well-adapted to discovering other types of problems. For example, restaurant workers are found among widely varying stressful conditions, carrying trays of various weights, with diverse scheduling problems and time pressures. Designing a study which isolates these factors is difficult.

DIRECTING ATTENTION TO WOMEN'S OCCUPATIONAL HEALTH PROBLEMS

Despite considerable progress in integrating women into the labor force, women are still found in specific jobs, where employment conditions are relatively unfavourable. This sexual division of labor affects women's health in 6 ways: (1) Women's jobs have specific characteristics (repetition, monotony, static effort, multiple simultaneous responsibilities) which may lead over time to physical and mental health; (2) Spaces, equipment and schedules designed in relation to the average male body and lifestyle may cause problems for women; (3) Segregation may cause health risks by causing task fragmentation and thus increasing repetition and monotony; (4) Sex-based job assignment may appear to protect the health of both sexes and thus distract from more effective occupational health promotion practices; (5) Discrimination against women is stressful in and of itself and may affect mental health; and (6) Part-time workers are excluded from many health-promoting benefits.

Unfortunately, prevention priorities are usually decided upon in ways that limit the attention paid to jobs usually done by women. In particular, in Canada and elsewhere, accident rates and (to a limited extent) chemical exposures have been primarily used to identify priority areas. These indicators do not work for most women's jobs. Women's health problems therefore have been attributed to their biology rather than to their jobs.

Accident Rates

When we examine statistics on compensation for occupational illness and injury, we notice that women and men show distinctive patterns. Men have from three to ten times more compensated industrial accidents and illnesses per worker (Pines, Lemesch & Grafstein, 1992; Robinson, 1989; Wagener & Winn, 1991; Laurin, 1991). When comparisons are made within the same industry, results of studies are variable: Sometimes women have more accidents than men, sometimes less (McCurdy, Schenker & Lassiter, 1989; Neuberger, Kammerdiener & Wood, 1988; Oleske et al., 1989; Tsai, Bernacki & Dowd, 1989; Wilkinson, 1987).

We did an approximate calculation of work accidents and industrial illnesses by industrial group in Québec for 1992. Women have more industrial disease than men (Boutin & Messing, in press). Many industrial diseases go unrecognized, since it is easier to recognize the occupational cause of a leg broken on the job than that of an allergy or inflammation that develops more slowly, away from the workplace (Kraut, 1994). Québec's Institut de Recherche en Santé et en Sécurité du Travail has compared official work accident and illness statistics with results from the Québec Health Survey and found several professions where they feel employment-related problems may be underestimated in official statistics (Gervais 1993). In addition, women's industrial illnesses are probably under-estimated to a greater degree than men's, for two reasons: (1) intoxications and other acute effects are more often found in jobs traditionally done by men in chemical factories, refineries and the like; and (2) getting recognition for accidents and occupational illnesses requires that the worker make a claim for compensation based on lost time and succeed in having that claim recognized. Research by Lippel and her colleagues has shown that women are much less successful than men at having their claims recognized at the Occupational Health and Safety Commission (Lippel, 1995).

Women's illness and injury rates may also be artificially lowered by a technical factor. Because women tend to work fewer hours than men at paid jobs, accident rates of women appear lower when, as is usual, the rates are calculated per worker rather than per hour worked. Of 14 studies comparing women and men, only two distinguished between occasional and permanent workers by reporting information on person-hours worked (Messing, Courville et al., 1994).

We conclude that it is difficult to compare occupational accident rates among women and men when their working conditions are very different. It is, nevertheless, useful to use accident rates as an indicator for occupational health problems. But women's specific situation in the workplace must be taken into account.

Other Indicators

The frequency of occupational accidents can be used to identify priorities for intervention, but it is not the best indicator for women's jobs. Another approach would be to look at women's health problems and examine the relationship to work. According to Statistics Canada, women's most common problems are musculoskeletal problems, skin problems and hypertension (Statistics Canada 1995: p. 53). Women are more likely than men to be hospitalized for mental disorders (p. 37). How can we find out whether any of these problems can be attributed to occupation? One approach has been to examine health problems reported by women according to the sector of the economy where they work. Although employment sector is a poor indicator of job content, this type of gross analysis does suggest a need to study women's jobs. Musculoskeletal problems emerge as a specific risk associated with sales, restaurant and cleaning work. Psychological distress is found among those in sales, restaurant work and teaching. Allergies and skin conditions are common in white collar work, especially teaching and also in personal services such as hairdressing. Heart disease is found among cleaners, personal service workers, saleswomen and managers (Gervais, 1993).

Absence from Work

Since occupational accidents are relatively infrequent in the types of jobs usually held by women, Bourbonnais and colleagues have suggested that certified sick leaves might be a more useful indicator of occupational health problems for both sexes (Bourbonnais et al., 1992). Women are more often absent from work than men, and only part of the absence can be explained by family responsibilities (Akyeampong, 1992). Although they are not a specific indicator of working conditions, variations among workers in days lost or numbers of absences provide the possibility of identifying hazardous working conditions. Among nurses, sick leaves have been found to be related to various indicators of work load and to shift work (Bourbonnais et al., 1992). A recent study found that records of absence would be more useful if they kept more accurate accounts of maternity and parental leaves and if type of illness could be identified (Deguire & Messing, 1996).

Interviews

Not all illness leads to absence. Workers who are in pain can continue to work for years (Courville, Dumais & Vézina, 1994). It is therefore important to speak to women at work in order to hear first-hand about their occupational health problems. Instead, researchers have been relying more and more on interviews and "focus groups" to develop experiential data bases on women's occupational health problems. A study in progress at the Workplace Health and Safety Agency in Ontario is using this method to identify priority areas for research because the agency found that more traditional studies had not included women's problems (Carlan & Keil, 1995). Interviews with gardeners showed that women and men had differences in the pattern of pain, with women's pain concentrated in the upper limbs (Messing, Courville et al., 1994). Using these pain occurrences as indicators, researchers were able to recommend some changes in tools, equipment and work practices (Boucher, Messing & Courville, 1995).

MUSCULOSKELETAL DISORDERS

Although women live longer than men, women and men in many countries can expect to live a similar number of years in good health (Guyon, 1990; Silman, 1987). Put differently, women spend about twice as long as men being disabled. One cause of disability is muscle and joint problems. Several studies reveal that women are especially likely to have this type of problem, including arthritis, "rheumatism," carpal tunnel syndrome, and inflammations of various joints (tendinitis, bursitis, epicondylitis): One-fourth of women have arthritis and rheumatism (Statistics Canada, 1995: p. 45). Women are also twice as likely as men to have chronic backache (Jutras et al., 1989).

Repetitive Work

The major research area in women's occupational health is probably musculoskeletal problems, the majority of cases of compensated occupational diseases (Kraut, 1994). It is a very thorny area, because of the enormous difficulty researchers are having in defining conditions which pose a risk for health. A specific problem arises in the repetitive work so common in women's jobs in factories and offices. In many jobs assigned to women (as well as some assigned to men), the work cycle(2) is under ten seconds long and the same movements are repeated many thousands of times in a day (Vézina, Tierney & Messing, 1992). These movements can individually make trivial demands on the human body, but the enormous degree of repetition makes tiny details of the setup assume primary importance. A chair the wrong height or a counter the wrong width may cause constant oversolicitation of the same tendons or joints, yet the observer sees no problem.

Epidemiological studies require a disease to be consistently associated with a risk factor. However, the risk factors for musculo-skeletal problems depend on the interaction of the workers' specific dimensions with those of the work site. A counter the right width for a tall worker is too wide for a shorter worker to reach across (Courville, Vézina & Messing, 1992). Thus, it is impossible to arrive at reliable exposure estimates from job titles or even from cycle length. This problem has caused skepticism about the very existence of illness due to repetitive movements, with resulting difficulties in compensation and more suffering for workers (Reid, Ewan and Lowy 1991).

Several problems come up in regard to gender in the context of musculo-skeletal problems. First is the widespread tendency to adjust for gender. If gender is a proxy for exposure status, adjusting for gender would tend systematically to underestimate risks in jobs primarily held by women, for example those which are most highly repetitive. All the studies of carpal tunnel syndrome cited in a major review article (Hagberg et al., 1992: Table 2) adjusted for gender, even though it has been shown that gender is not related to carpal tunnel syndrome if anthropometric measurements related to wrist anatomy and physiology are taken into account (Stetson et al., 1992). Adjusting would be appropriate only if gender were an independent determinant of carpal tunnel syndrome, for example for hormonal reasons, rather than a determinant of job content or of inadequate job engineering. One study in the poultry-processing industry in France found that women reported much more often than men that their worksite was ill-adjusted to their size (Saurel-Cubizolles et al., 1991a). Women in factories studied by Silverstein et al. did more repetitive work than men (Silverstein, Fine, & Armstrong, 1987). If women in general experienced these conditions (as I think they do), they would provide an alternative explanation for excess repetitive strain injury among women.Office workers have recently been identified as a group susceptible to musculoskeletal disorders (On et al., 1995). Attention is being given to VDT keyboard design as well as to organizational factors. In Brazil, a government regulation limits secretaries to 8,000 keystrokes per hour. No such regulation is in effect in North America, although Billette et al. found workloads of up to 20,000 keystrokes per hour among Canadian data entry clerks (Billette & Piché, 1987).

In general, although there has been an attempt to introduce gender-fair guidelines for manual lifting tasks (Waters et al., 1993), no standards have been developed for the type of physical exertions in jobs traditionally assigned to women.

Static Muscular Effort

Static effort is exerted when muscles are contracted for long periods. This type of effort creates musculoskeletal and circulatory problems due to interference with circulation. Cleaning jobs (dusting high surfaces, bending over toilets) often require this type of posture. Many women's jobs in factories or services in North America (sales, hairdressing, tellers, cashiers) require standing for long periods of time, resulting in back and other musculoskeletal problems (Vézina, Chatigny & Messing, 1994). It should be noted that these workers usually work sitting down in Europe and Latin America.

Pregnant Workers

Pregnancy alters the shape of the body and thus the interaction with the work site (Paul, van Dijk & Frings-Dresen, 1994; Paul & Frings-Dresen, 1994). A study of precautionary leave given to pregnant workers in Québec showed that ergonomic considerations were the most common reason for giving such leave (Malenfant, 1993). In jurisdictions where no such leave is available, pregnant workers may risk physiological damage.

Psychological Distress and Musculoskeletal Problems

Several studies show a relationship between psychological stress and the occurrence of musculoskeletal problems (Bongers et al., 1993; Leino & Magni, 1993). Interpretation of this relationship is varied, extending from physiological explanations relating stress to muscle tension to accusations of "neurosis" or "hysteria." Whatever the cause, the relationship between work organization and musculoskeletal problems is well recognized. Since we know that women are more likely to be found in stressful jobs where there is insufficient job autonomy and high work overload (Hall 1989), they may be especially susceptible to stress-related injuries.

EMOTIONAL AND MENTAL STRESS

Any discussion with women (and often men) workers tends to identify "stress" as an important occupational health problem. Analysis of data from the 1987 Québec Health Survey (Vézina et al., 1992) showed that 18% of women workers and 9% of male workers suffered from psychological distress (Ilfield Scale). We can ask whether women "really" have more such problems, but it is undeniable that women consult more health practitioners and take more medication for mental problems than men (Statistics Canada 1995: p. 52). Women service workers were particularly likely to experience stress. Secretaries have also been identified by NIOSH as a group particularly prone to stress (Lippel, 1992: p. 167). Among female hospital workers, shift work was associated with psychiatric morbidity found during an annual medical examination (Estryn-Behar et al., 1990).

Yassi et al. (1989) have pointed out that there is sometimes confusion between the notions of "unreal" problems and "mental" problems. In a study of telephone operators, they suggest that a collective stress reaction may be an appropriate response to some working conditions. However, these types of mental health problems are often dismissed as being due to innate characteristics of women rather than to organizational problems in the workplace. For example, a recent study of workplace determinants of depression scored women as depressed at a score of 23 and men at 17 on the same test (Goldberg et al., 1993). Thus, the fact that women had more symptoms of depression was "normalized" and could not be related to working conditions.

Refusing to examine the possibility that women workers' mental health symptoms are linked to jobs rather than gender has several consequences. First, there may be discrimination in compensation for stress (Lippel, 1992: p. 167). Second, there may be reluctance to investigate women's symptoms seriously. For example, women's symptoms of organic solvent exposure have been wrongly attributed to "hysteria" (Bowler et al., 1992; Brabant, Mergler & Messing, 1990). Similar uncertainty surrounds discussions of problems in women's traditional work such as indoor air quality and musculoskeletal disorders.

Stress and Cardiovascular Effects

The work of Karasek and others has related several workplace variables (degree of job control, level of demand) to effects on the cardiovascular system (Karasek et al., 1982; Johnson & Hall, 1988). Hall found that jobs assigned to women are characterized by a low level of decision latitude and more likely to be stressful (Hall, 1989). Unfortunately, most scientists who have studied heart disease by occupation have restricted their samples to men (Pickering et al., 1991).(3) Although coronary artery disease is the most common cause of death among women, and as many women as men report hypertension, heart disease is still thought of as a man's problem and studies have not been gender-sensitive (Stiengart et al, 1991; Doyal, 1995: p. 17). Several professions which are commonly held by women are among the ten professions with the highest diastolic blood pressure: Laundry and dry cleaning operatives, food service workers, private child care workers and telephone operators (Leigh, 1991). Women working on shifts (Kwachi et al., 1995), women with clerical or sales jobs (Haynes, 1991), and women reporting that their work is both hectic and monotonous have a higher incidence of coronary heart disease (Theorell, 1991). It has been found that during pregnancy, certain working conditions (noise, lifting weights) are associated with higher blood pressure (Saurel-Cubizolles et al., 1991b), this finding should be followed up to see whether such changes are permanent and to expand the determinants tested.

Multiple Causes and Effects

Stressors in a workplace can be diverse and they can affect multiple targets (Frankenhaeuser, 1991). Therefore a single factor, single effect model is not appropriate for stress investigations. Changing relationships between different job parameters and worker strain led Carayon et al. to propose a more global approach to occupational stress (Carayon, Yank & Lim, 1995).

A more global approach was tried in a study of primary school teachers, 82% of whom are women. Teachers have a high level of mental stress and are among the occupations with a significantly elevated risk of suicide (Boxer, Burnett & Swanson, 1995). An analysis of data from the Quebec Health Survey by job title showed that teachers, who had good mental health when under 29, were among the groups with the highest levels of mental problems between the ages of 45 and 64 (Gervais, 1993: p. 25). An investigation into the working conditions of primary school teachers found that no single stressor could be isolated; instead, low humidity and high temperatures in the classroom, constant standing, slightly flexed posture, mental concentration, needs of poor children, lack of recognition and other constraints combined to render the job difficult. However, the health and safety compensation system is ill-adapted to situations where the relationship between aggressors and occupational illness is complex and multifaceted (Messing, Seifert, & Escalona, 1997).

Balancing Home and Family

If women's and men's work differs in the workplace, their situations diverge even more outside the workplace. Studies done by Statistics Canada show that Canadian married men with children under five work 18.2 hours per week at domestic tasks and child care compared to 32.2 hours per week for married women with children under five and 23.8 for single mothers of children under five (Statistics Canada, 1994: Table 6.8). Walters and colleagues found that female registered nurses (with or without children) reported 24 hours a week spent on homemaking tasks (including car repairs and outdoor tasks) compared to 16 hours for male nurses (Walters et al., 1995). Women also assume a major role in eldercare, responsible for 70-80% of such care. Half of women now between 35 and 64 will have to care for an older relative at some time (Guberman, Maheu & Maillé, 1994).

Role overload is very stressful. Lee and colleagues found by an analysis of 300 telephone interviews with Ottawa-area women workers that over a quarter had thought of quitting because balancing work and family was too stressful. Over 85% felt that there were not enough hours in the day to accomplish everything they had to do and over half felt they did not have enough spare time for themselves. They felt they needed more support from the community and other family members, but workplace policies were also very important. Over one-third felt they would be better able to cope if they had more job security. Only one-third had flexible working hours, the most popular suggestion in response to the question "What could your employer do to help you balance work and family?" (Lee, Duxbury & Higgins, 1994).

But flexible working hours may carry their own difficulties, when needs of workers and employers do not coincide. Many workplaces (supermarkets, hospital workers, banks) now schedule workers according to just-in-time principles. Although workers have some liberty to change their hours, the cost may be high. Telephone operators' beginning time can vary as much as 10 hours from one day to the next and back again the following day, with 3-8 days' notice. Thirty operators filled out diaries on how they coped during a two-week period. Since none of the procedures undertaken to trade or change hours could be done during work time, breaks, lunchtime and family time were taken up by the 156 procedures used to change hours. To this should be added the 212 other procedures used to re-arrange existing day-care arrangements in order to meet new schedules (Prévost, Messing & Saint-Jacques, 1996).

This type of work-family conflict occupies a great deal of women's time and creativity. In addition, the conflict affects the health and serenity of mothers and children. Among telephone operators, 68.5% were above the Québec Health Survey's limit for psychological distress (Ilfeld test). On the other hand, increasing numbers of researchers show possible benefits from multiple roles. Barnett and Marshall find that employment can protect women from the effects of stressful family situations, but that results depend on the nature of employment (Barnett & Marshall, 1991).

Family or personal obligations are more often a cause of absence from work for women than for men (Statistics Canada: p. 82). This level of absence not only affects the perception of women's dedication to their jobs, it also results in their being unable to take sick days for their own illnesses.

NEGLECTED DISEASES AND OCCUPATIONS

This brief review does not enable us to do more than suggest a few ways of approaching problems which could have a serious effect on health or well-being or which might be easy to correct or which affect large numbers of women workers.

Guo and colleagues found that, particularly for women workers, professions where back pain had been most studied did not correspond to those reporting most back pain (Guo et al., 1995). Cleaners, waitresses and supermarket clerks were among the neglected groups. They suggest that research be oriented toward groups suffering health problems. This suggestion could also apply to other problems in occupational health.

Neglected Diseases and Health Problems

Occupational cancer: as mentioned, women's occupational cancers have been neglected, although attention is starting to be paid particularly to occupational breast cancer.

Sexual health: a recent review paper has pointed out that a variety of environmental agents and conditions may affect women's and men's sexual health (Bancroft, 1993). The author points out specific lacunae in understanding women's hormonal and sexual functioning. The section on industrial chemicals includes only male effects.

Reproductive health: in Canada, legislated maternity leave and provisions for precautionary leave for pregnant women have enabled pregnant workers to avoid some of the discriminatory practices found in the US, where a Supreme Court decision was necessary to prevent an employer from requiring that women of reproductive age exposed to conditions hazardous to fetuses (and male reproduction), be sterile. However, pregnant women have not yet been included in most standard-setting procedures. The fact that many working conditions are ill-adapted for pregnant workers is demonstrated dramatically by the fact that, in Québec, where fully-compensated precautionary leave is available for pregnant and nursing women exposed to conditions posing a danger for their health, one-third of pregnant women take such leave. Still, pregnant women often find themselves uncertain and anxious about which working conditions are dangerous; telephone lines should be set up to which pregnant women could refer for up-to-date information on risk factors.

Menstrual symptoms are among the most commonly-diagnosed disorders of women. During the mid-eighties, several researchers suggested that menstrual symptoms might be useful for the study of occupational effects on reproductive health, as well as indicative of health problems which should be addressed (Mergler & Vézina, 1985; Harlow, 1986). Parameters of the menstrual cycle which can be studied in relation to occupation include regularity and length of cycle, length and volume of flow, and symptoms of pain and discomfort associated with the periods. The latter symptoms are quite common and can be studied in normal populations. Some evidence has now accumulated, both on parameters of the cycle which vary with exposure (Messing et al., 1992; Harlow & Matoski, 1991), and on variations among working populations (Treloar et al., 1967; Shortridge, 1988).

In order to study effects of the environment on human health, it is useful to have indicators of biological function which can be studied in normal exposed and unexposed populations, and which show sufficient variation to reflect exposure level (Stein & Hatch, 1987; Hatch & Friedman-Jimenez, 1991; Savitz & Harlow, 1991). The indicator of female reproductive health most often used is pregnancy outcome, which has been related to smoking, alcohol consumption, and many occupational exposures. However, a very small proportion of women are pregnant at any time, and studies of pregnancy outcome are often haunted by small numbers. Given the demographics of the North American working population, however, the vast majority of working women menstruate.

Disorders of the menstrual cycle have been explored in relation to occupational exposures to synthetic hormones (Mills, Jefferys & Stolley, 1984; Harrington et al., 1978), organic solvents (Panova, 1976), carbon disulfide (Zhou et al., 1988), chemical exposures of hairdressers (Blatter & Zielhuis, 1993) and night work (Uehata & Saskawan, 1982), although confounding factors were not examined in these studies. Amenorrhea has been associated with strenuous jobs such as athlete (Toriola & Mathur, 1986) or ballet dancer (Frisch, Wyshak & Vincent, 1980), but not with styrene exposure (Lemasters, Hagen & Samuels, 1985). No occupational variables have been considered in relation to cycle length, although exercise has been associated with long cycles among college students (Harlow & Matoski, 1991). Cycle anomalies were related to working conditions in poultry slaughterhouses; irregular cycles were associated with schedule variability and exposure to cold temperatures at work (Messing et al., 1992).

Dysmenorrhea or painful menstruation occurs when increased prostaglandin production and release by the endometrium during menstruation gives rise to increased abnormal uterine activity that produces ischemia and cramping pelvic pain. There may be other associated symptoms such as leg or backache or gastrointestinal upset. Menstrual pain can begin before or just after the onset of the menstrual flow. Premenstrual syndrome (PMS) is a less well-defined diagnostic category which refers to a group of symptoms thought to occur during the days preceding to the onset of menses. Since its diagnosis requires making an association with an event (menstruation) which has not yet occurred, reports of prevalence are not consistent (Gurevitch, 1995).

Prevalence estimates of perimenstrual symptoms vary greatly between studies, according to age, parity, contraceptive methods and other demographic characteristics. Severe dysmenorrhea was not associated with work in the reinforced plastics industry (Lemasters, Hagen & Samuels, 1985) or exposure to mercury in Italian lamp factories (Di Rosis, Anastatasio & Selvaggi, 1985), or to exposure to toluene (Ng, Foo & Yoong, 1992) but was associated with shift work and irregular shifts in Japanese hospitals (Uehata & Saskawa, 1982). Dysmenorrhea was found to be associated with several parameters expressing cold exposure and physical work load in poultry slaughterhouses in western France (Messing et al., 1993) and in Québec, where increased exposure to cold was associated with increased prevalence of dysmenorrhea and sick leave (Mergler & Vézina, 1985). It was found at a high level among hairdressers (Blatter & Zielhuis, 1993; Zita, 1989; Mortola et al., 1990).

In the past, research on dysmenorrhea was conditioned by the prevailing attitude that these symptoms had a primarily psychological base. For example, after reporting that beginning airline hostesses underwent unfavourable changes in the menstrual cycle 3.5 times as often as favorable changes, researchers commented, "There is not enough information to explain the pathophysiology of dysmenorrhea. The frequent association of dysmenorrhea with other (sic) neurotic symptoms is indicative of its psychological origin" (Iglesias, Terrés & Charravia, 1980).

Because of such attitudes, prudence is necessary when associating symptoms with the menstrual periods. Studies of the prevalence and etiology of back pain, a common occupational health problem among hospital workers, may be confused if perimenstrual back pain is not taken into account (Tissot & Messing, 1995).

Unfortunately, work-related dysmenorrhea does not fit conveniently into the occupational health and safety compensation system, which only provides for continuous absences rather than a day or two per month.

Little or no information is available on the relation between working conditions and age at menopause or menopausal symptoms (Sarrel, 1991). Age at menopause can be an indicator of exposure to environmental pollution, as shown by its relationship to smoking and a possible relationship to carbon disulphide exposure (Stanosz, Kuligowski & Pieleszek, 1995).

Violence

Violence in the workplace seems to be becoming more common; it is the major cause of women's fatal occupational accidents, now comprising 41% of women's occupational accident mortality in the US (National Institute for Occupational Safety and Health, 1993). Canada's figures are lower: an Ontario study found 25% of women's occupational fatalities to be due to violence (Liss, 1993; Liss & Craig, 1990), while in Québec in 1981-88 17% of women's occupational fatalities were due to violence, compared to 4% of men's (Rossignol & Pineault, 1992). (However, the absolute numbers of violent fatalities are higher in men; they have a fivefold excess in numbers of deaths due to workplace violence.)

Violence is a growing problem for some groups of women workers in contact with the public: bank tellers, convenience store cashiers, nurses, gas station attendants, teachers. A recent symposium sponsored by the American Psychological Association presented Canadian and American data on the specific risks to women in the helping professions, caught between government and angry clients in a time of cuts in social and health services. A 1991 survey of 800 Ontario nurses found that 59% reported physical assault, 17% sexual assault at some time during their professional lives. Ten percent had been physically assaulted during the previous month (Nurse Assault Project Team, 1992).

CONCLUSION

The system of recognition, compensation and prevention of industrial disease will have to adapt to the growing presence of women in the workplace and to the changing nature of work. Occupational health specialists will have to move beyond considering women only in the context of fetal protection. Women must receive appropriate compensation for their occupational health problems and prevention practices in the workplace must include women's jobs. Segregation of women and men into separate jobs should not be used as a way to avoid good health and safety practice. Discrimination against women in decisions regarding compensation of occupational illnesses must be eliminated. We must move toward recording accident statistics based on hours worked rather than on numbers of individual workers and use measures complementary to accident statistics in order to identify risky jobs, including qualitative interview methods. Labor standards and exposure limits will have to be reconsidered in the light of women's biological and social specificities.

In order to identify women's occupational health problems, gender-sensitive and women-centered occupational health research should be encouraged and supported. It should be recognized that such research may require innovative methods, and must involve the active participation of working women. Partnerships between workplaces and researchers should be actively encouraged through granting organizations. Assurances should be put in place so that women and their work are included in research supported by federal agencies. In bodies responsible for research grants, guidelines for gender sensitivity in health research must be developed, as has been done in some granting organizations (Eichler & Lapointe 1985). In this effort, inputs from both social sciences and biomedical sciences can complement each other effectively.

NOTES

(1.) The August 1994, November 1994 and March 1995 issues of the Journal of Occupational and Environmental Medicine report the results of the 1993 Conference on Women's Health: Occupation and Cancer sponsored by NCI and, among others, the National (US) Institutes of Health Office of Research on Women's Health.

(2.) The work cycle is the interval between repetitions of a single operation. For a university professor it would be the time between beginnings of successive introductory lectures to the same course (months), for an otorhinolaryngologist it could be the time between beginnings of successive tonsillectomies (hours or days), for a sewing machine operator it is the time between successive pant legs (6-10 seconds), for a data entry operator it is the time between characters entered, as little as 0.5 sec.

(3.) An article by the group of Robert A. Karasek, the pioneering researcher into stress and heart disease, mentions (Pickering et al., 1991, p. 179) that all the group's studies relating blood pressure to job strain had been done on men, although they intended to expand these studies.

REFERENCES

Akyeampong, E. B. (1992). L'absentéisme. Tendances sociales canadiennes, 25, 25-28.

Armstrong, P., & Armstrong, H. (1993). The Double Ghetto: Canadian Women and Their Segregated Work. McClelland and Stewart, Toronto. A detailed discussion of women's labor force participation is given in chapter 6 of this important study.

Bancroft, J. (1993). Impact of environment, stress, occupational and other hazards on sexuality and sexual behavior. Environmental Health Persp. 101 (Suppl. 2), 101-107.

Barnett, R., & Marshall, N. (1991). The relationship between women's work and family roles and the subjective well-being and psychological distress. In Frankenhaeuser, M., Lundberg, U., Chesney, M. Women, Work and Health. New York, Plenum Press. 111-136.

Billette, A., & Piché, J. (1987). Health problems of data entry clerks and related job stressors. Journal of Occupational Medicine. 29, 942-948.

Blatter, B. M., & Zielhuis, G. A. (1993). Menstrual disorders due to chemical exposure among hairdressers. Occupational Medicine. 43, 105-106.

Blishen, B. R., Carroll, W. K., & Moore, C. (1987). The 1981 socioeconomic index for occupations in Canada. Can. Rev. Soc & Anth. 24, 465-488.

Bongers, P.M., de Winter, C. R., Kompier, M. A. J., & Hildebrandt, V. H. (1993). Psychosocial factors at work and musculoskeletal disease. Scand. J. Work Environ. Health. 19, 297-312.

Boucher, M., Messing, K., & Courville, J. (1995). Analyse du travail des jardinières et des jardiniers dans une municipalité. Report submitted to the municipality and to Local 301 of the Canadian Union of Public Employees.

Bourbonnais, R., Vinet, A., Vézina, M., & Gingras, S. (1992). Certified sick leave as non-specific morbidity indicator: a case-referent study among nurses. Brit. J. Ind. Med. 49, 673-678.

Bowler, R., Mergler, D., Rauch, S.S., & Bowler, R. P. (1992). Stability of psychological impairment: 2-year follow-up of former microelectronics workers' affective and personality disturbance. Women & Health. 18(3), 27-48.

Boxer, P. A., Burnett, C., & Swanson, N. (1995). Suicide and occupation: A review of the literature. Journal of Occupational and Environmental Medicine. 37 (4), 442-452.

Brabant, C., Mergler, D., & Messing, K. (1990). "Va te faire soigner, ton usine est malade: La place de l'hystérie de masse dans la problématique de la santé au travail," Santé mentale au Québec XV. pp 181-204.

Carayon, P., Yank, C.L., & Lim, S.Y. (1995). Examining the relationship between job design and worker strain over time in a sample of office workers. Ergonomics. 38 (6), 1199-1211.

Carlan, N., & Keil, M. F. (1995). Developing a proposal for a working women's health survey. Invisible: Occupational Health Problems of Women at Work. Messing, K., Neis, B., & Dumais, L., dirs. Charlottetown PEI: Gynergy Books, 292-306.

Courville, J., Dumais, L., & Vézina, N. (1994). Conditions de travail de femmes et d'hommes sur une chaîne de découpe de volaille et développement d'atteintes musculo-squelettiques. Travail et Santé. 10(3), S17-S23.

Courville, J., Vézina, N., & Messing, K. (1992). Analyse des facteurs ergonomiques pouvant entraîner l'exclusion des femmes du tri des colis postaux. Le travail humain. 55, 119-134.

Dawood, Y. (1990). Dysmenorrhea. Clinical Obstetric and Gynecology. 33, 168-178.

Deguire, S., & Messing, K. (1996). L'étude de l'absence au travail a-t-elle un sexe? Recherches féministes. 9(1), 1-20.

Di Rosis F., Anastatasio S. P., Selvaggi, L. et al. (1985). Female reproductive health in mercury lamp factories: Effects of exposure to inorganic mercury vapour and stress factors. Br J Ind Med. 42, 488-494.

Doyal, L. (1995). What Makes Women Sick: Gender and the Political Economy of Health. London: MacMillan Press Ltd.

Eichler, M., & Lapointe, J. (1985). On the Treatment of the Sexes in Research. Ottawa: Social Sciences and Humanities Research Council of Canada.

Estryn-Behar, M. et al. (1990). Stress at work and mental health status among female hospital workers. Brit. J. Ind. Med. 47, 20-28.

Franck C., Bach E., & Skov, P. (1993). Prevalence of objective eye manifestations in people working in office buildings with different prevalences of the sick building syndrome compared with the general population. Int Arch Occup Environ Health. 65, 65-69.

Frankenhaeuser, M. (1991). The psychophysiology of sex differences as related to occupational status. In Frankenhaeuser, M., Lundberg, U., & Chesney, M. Women, Work and Health. New York: Plenum Press. 39-61.

Frisch, R. E., Wyshak, G., & Vincent, L. (1980). Delayed menarche and amenorrhea in ballet dancers. N Eng J Med. 303, 17-19.

Gervais, M. (1993). Bilan de santé des travailleurs québécois. Montréal, Institut de recherche en santé et en sécurité du travail du Québec.

Gilboa, R., Al-Tawil, N. G., & Marcusson, J. A. (1988). Metal allergy in cashiers. Acta Dermatovenereologica. 68 (4), 317-324.

Goldberg, P., David, S., Landre, M-F., Fuhrer, R., Dassa, S., & Goldberg, M. (1993). An epidemiological study of depressive symptomatology and working conditions of prison staff. Proceedings of the 24th International Congress on Occupational Health. p. 110.

Greenberg, G. N., & Dement, J. M. (1994). Exposure assessment and gender differences. Journal of Occupational Medicine. 36(8), 907-912.

Guberman, N., Maheu, P., & Maillé, C. (1994). Travail et soins aux proches dépendants. Montréal: Editions Remue-mémage. p. 27-29.

Guo, H.-R., Tanaka, S., Cameron, L. L., Seligman, P. J., Behrens, V. J., Ger, J., Wild, D. K., & Putz-Anderson, V. (1995). Back pain among workers in the United States: National estimates and workers at high risk. Am. J. Ind. Med. 28, 591-602.

Gurevitch, M. (1995). Rethinking the label: Who benefits from the PMS construct? Women & Health. 23 (2), 67-98.Guyon, L. (1990). Quand les femmes parlent de leur santé. Québec: Ministère de la santé et des services sociaux. p. 20.

Hagberg, M., Morgenstern, H., & Kelsh, M. (1992). Impact of occupations and job tasks on the prevalence of carpal tunnel syndrome. Scandinavian Journal of Work Environment and Health. 18, 337-345.

Hall, E. M. (1989). Gender, work control and stress: A theoretical discussion and an empirical test. Int. J. Health Serv. 19, 725-745.

Harlow, S. D. (1986). Function and dysfunction: A historical critique of the literature on menstruation and work. Health Care Women Int. 7, 39-50.

Harlow, S. D., & Matoski, G. M. (1991). The association between weight, physical activity and stress and variation in the length of the menstrual cycle. Am J Epidemiol. 133, 38-49.

Harrington, J. M., Stein, G. F., Rivera, R. O., & Morales, A. V. (1978). The occupational hazards of formulating oral contraceptives. Int Arch Environ Health. 1, 12-15.

Harrison, J., Pickering, C. A. C., Faragher, E. B., Austwick, P. K. C., Little, S. A., & Lawton, L. (1992). An investigation of the relationship between microbial and particulate indoor air pollution and the sick building syndrome. Respiratory Med. 86, 225-235.

Hatch, M., Friedman-Jimenez, G. (1991). Using reproductive effect markers to observe subclinical event, reduce misclassification, and explore mechanism. Environmental Health Perspectives. 90, 255-259.

Haynes, S. (1991). The effect of job demands, job control and new technologies on the health of employed women. In Frankenhaeuser, Marianne; Lundbergh, Ulf; & Chesney, Margaret. Women, Work and Health: Stress and Opportunities. New York: Plenum Press. p. 157-169.

Health Canada. (1993). Proceedings of the Round Table on Gender and Occupational Health. Ottawa: Health Canada p. 155-161.

Heller, R. F., Williams, H., & Sittampalam, Y. (1984). Social class and ischaemic heart disease: Use of the male: female ratio to identify possible occupational hazards. J. Epid. and Comm. Health. 38, 198-202.

Hodgson, M. J., Frohliger, J., Permar, E., Tidwell, C., Traven, N. D., Olenchock, S. A., & Karpf, M. (1991). Symptoms and microenvironmental measures in nonproblem buildings. J Occup Med. 4, 527-533.

Hsairi, M., Kauffmann, F., Chavance, M, & Brochard, P. (1992). Personal factors related to the perception of occupational exposure: An application of a job exposure matrix. International Epidemiology Association Journal. 21, 972-980.

Iglesias, R. E., Terrés, A., & Chavarria, A. (1980). Disorders of the menstrual cycle in airline stewardesses. Aviat Space Environ Med. 1980, 51 (5): 518-520.

Johnson, J. V., & Hall, E. M. (1988). Job strain, workplace social support and cardiovascular disease. Am J Publ Health. 78, 1336-1342.

Jutras, S., Guyon, L., Renaud, M., Dandurand, F., & Bouchard, P. (1989). Comment les Québécois se tirent-ils d'affaire? Sciences sociales et santé. VII(4), 69-93.

Karasek, R., Baker, D., Marker, F., Ahlbom, A., & Theorell, T. (1982). Job decision latitude, job demands, and cardiovascular disease: A prospective study of Swedish men. Am J Publ Health. 71,694-705.

Kelland, P. (1992). Sick building syndrome, working environments and hospital staff. Indoor Environ. 1, 335-340.

Kraut, A. (1994). Estmates of the extent of morbidity and mortality due to occupational diseases in Canada. Am J Ind Med. 25,267-278.

Kwachi, I., Colditz, G. A., Stampfer, M. J., Willett, W. C., Manson, J. E., Seizer, F. E., & Hennekens, C. H. (1995). Prospective study of shift work and risk of coronary heart disease in women. Circulation. 92(11), 3178-3182.

Laurin, G. (1991). Féminisation de la main d'oeuvre: Impact sur la santé et la sécurité du travail. Commission de la santé et de la sécurité du Québec, Montreal, Quebec, Canada.

Lee, C., Duxbury, L., & Higgins, C. (1994). Employed Mothers: Balancing Work and Family Life. Ottawa: Canadian Centre for Management Development.

Leigh, J.P. (1991). A ranking of occupations based on the blood pressures of incumbents in the National Health and Nutrition Examination Survey I. JOM. 33,853-861.

Leino, P. & Magni, G. (1993). Depressive and distress symptoms as predictors of low back pain, neck-shoulder pain and other musculoskeletal morbidity. Pain. 53, 89-94.

Lemasters, G., Hagen, A., & Samuels, S. J. (1985). Reproductive outcomes in women exposed to solvents in 36 reinforced plastics companies. I Menstrual dysfunction. J Occup Med.27, 490-494.

Lippel, K. (1992). Le stress au travail (Cowansville, Québec: Les éditions Yvon Blais) 265p.

Lippel, K. (1995). Watching the watchers: How expert witnesses and decision-makers perceive men's and women's workplace stressors. In Messing, K., Neis, B. & Dumais, L. (Eds.) Invisible: Occupational Health Problems of Women at Work. (p. 265-291). Charlottetown PEI: Gynergy Books.

Liss, G. (1993). Examination of Workers' Compensation Claims Among Nurses in Ontario for Injuries Due to Violence. Toronto: Health and Safety Studies Unit, Ontario Ministry of Labor.

Liss, G. M., & Craig, C. A. (1990). Homicide in the workplace in Ontario: Occupations at risk and limitations of existing data sources. Can J Public Health. 81, 10-15.

Lortie, M., (1987). Analyse comparative des accidents déclarés par des préposés hommes et femmes d'un hôpital gériatrique. Journal of Occupational Accidents. 9, 59-81.

Malenfant, R. (1993). Le droit au retrait préventif de la travailleuse enceinte ou qui allaite: à la recherche d'un consensus. Sociologie et Sociétés XXV(1), 61-75.

McCurdy, S. A., Schenker, M. B., & Lassiter, D. V. (1989). Occupational injury and illness in the semiconductor manufacturing industry. Am J Industrial Med. 15,499-510.

Mendell, M. J., & Smith, A. H. (1990). Consistent pattern of elevated symptoms in air conditioned office buildings: A reanalysis of epidemiologic studies. Am J Public Health. 80, 1193-1199.

Menzies, R., Tamblyn, R., Farant, J-P., Hanley, J., Nunes, F., & Tamblyn, R. (1993). The effect of varying levels of outdoor-air supply on the symptoms of the sick building syndrome. New Engl J Med. 328, 821-827.

Mergler, D, & Vézina, N. (1985). Dysmenorrhea and cold exposure. J Reprod Med. 30, 106-111.

Mergler, D. (1995). Adjusting for gender differences in occupational health studies. In Messing, K., Neis, B., & Dumais, L. (eds). Invisible: Issues in Women Occupational Health and Safely/Invisible: La santé des travailleuses. Charlottetown, PEI: Gynergy Books. p. 236-251.

Messing, K., & Boutin, S. (1997). La reconnaissance des conditions difficiles dans les emplois des femmes et les instances gouvernemental es en santé et en sécurité du travail. Relations industrielles/Industrial Relations, 52(2): 261-291.

Messing, K., Chatigny, C., & Courville, J. (1996). L'invisibilité du trvail et la division léger/lourd: impact sur la santé et la sécurité du travail. Objectif Prévention. 19(2), 13-16.

Messing, K., Courville, J., Boucher, M., Dumais, L., & Seifert, A.M. (1994). Can safety risks of blue-collar jobs be compared by gender? Safety Science. 18, 95-112.

Messing, K., Dumais, L., Courvillle, J., Seifert, A.M., & Boucher M. (1994). Evaluation of exposure data from men and women with the same job title. Journal of Occupational Medicine. 36(8); 913-917.

Messing, K., Neis, B., & Dumais, L. (eds.) Invisible: Issues in Women's Occupational Health/La santé des travailleuses. Charlottetown, PEI: Gynergy Books.

Messing, K., Saurel-Cubizolles, M. J., Kaminski, M., & Bourgine, M. (1993). Factors associated with dysmenorrhea among workers in French poultry slaughterhouses and canneries. J Occup Med. 35,493-500.

Messing, K., Saurel-Cubizolles, M.-J., Bourgine, M. et al. (1992). Menstrual cycle characteristics and working conditions in poultry slaughterhouses and canneries. Scand J Work, Environment and Health, 18, 302-309.

Messing, K., Seifert, A.M., & Escalona, E. (1997). The 120-second minute: Using analysis of work activity to prevent psychological distress among elementary school teachers. Journal of Occupational Health Psychology, Vol. 2, No. 1.

Messing, K., & Stevenson, J. (1996). Women in procrustean beds: Strength testing and the workplace. Gender Work & Organization. 3(3), 156-167.

Mills, J. L., Jefferys, J. L., Stolley, P. D. (1984). Effects of occupational exposure to estrogen and progesteogens and how to detect them. J Occup Med. 26, 269-272.

Mortola, J. F., Girton, L., Beck, L., Yen, S.C. (1990). Diagnosis of premenstrual syndrome by a simple, prospective and reliable instrument: The calendar of premenstrual experiences. Obstetric and Gynecology. 76, 302-306.

National Institute for Occupational Safety and Health (NIOSH). Update October 25, 1993. Publication No. 94-101.

Nelson, N., Kaufman, J. D., Burt, J, & Karr, C. (1995). Health symptoms and the work environment in four nonproblem United States office buildings. Scand J Work Environ Health. 21, 51-59.

Neuberger, J. S., Kammerdiener, A.M., & Wood, C. (1988). Traumatic injuries among medical center employees. AAOHN Journal. 36, 318-325.

Ng, T. P., Foo, S.C., & Yoong, T. (1992). Menstrual function in workers exposed to toluene. Br J Ind Med. 49, 799-803.

Norback, D., Michel, I., & Widstrom J. (1990). Indoor air quality and personal factors related to the sick building syndrome. Scand J Work Environ Health. 16, 121-128.

Norback, D., Torgen, M., & Edling, C. (1990). Volatile organic compounds, respirable dust, and personal factors to prevalence and incidence of sick building syndrome in primary schools. Brit J Ind Med. 47, 733-741.

Nurse Assault Project Team. (1992). Nurse Assault Survey. Registered Nurses' Association of Ontario.

Oleske, D. M., Brewer, R. D., Doan, P., & Hahn, J. (1989). An epidemiologic evaluation of the injury experience of a cohort of automotive parts workers: A model for surveillance in small industries. J Occ Accidents. 10, 239-253.

On, C. N., Chia, S. E., Jeyaratnam, J., & Tan, K. C. (1995). Musculoskeletal disorders among operators of visual display terminals. Scand. J Work Environ Health. 21,60-64.

Panova, Z. (1976). Menstrual and reproductive problems and gynaecological morbidity in women occupationally exposed to petrol. Leptosi Na Hiegienno-Epidemiologic inata Sluzba. 20, 53-56.

Paul, J. A., & Frings-Dresen, M. H. W. (1994). Standing working posture compared in pregnant and non-pregnant conditions. Ergonomics. 37(9), 1563-1575.

Paul, J. A., van Dijk, F. J. H., & Frings-Dresen, M. H. W. (1994). Work load and musculoskeletal complaints during pregnancy. Scand J Work Environ Health. 20, 153-159.

Pheasant, S. (1986). Bodyspace. Taylor and Francis, London, England. Chapter 1.

Pickering, T. G., James, G. D., Schnall, P. L., Schlussel, Y. R., Pieper, C. F., Gerin, W., & Karasek, R. A. (1991). Occupational stress and blood pressure: Studies in working men and women. In Frankenhaeuser, Marianne; Lundbergh; Ulf; & Chesney, Margaret. Women, Work and Health: Stress and Opportunities. New York: Plenum Press. p. 171-186.

Pines, A., Lemesch, C., & Grafstein, O. (1992). Regression analysis of time trends in occupational accidents (Israel, 1970-1980). Safety Science. 15, 77-95.

Prévost, J., Messing, K., Saint-Jacques, Y. (1996). L'horaire de travail variable et la garde des enfants: Etude ergonomique sur un aspect de la conciliation des responsabilit és professionnel les et familiales. Actes du 3ième symposium québécois de recherche sur la famille. Trois-Rivières, Québec. le 27 octobre 1995.

Reid, J., Ewan, C., & Lowy, E. (1991). Pilgrimage of pain: The illness experiences of women with repetition strain injury and the search for credibility. Soc Sci Med. 32(5), 601-612.

Robinson, J. C. (1989). Trends in racial inequality and exposure to work-related hazards, 1968-1986. AAOHN Journal. 37, 56-63

Rossignol, M., & Pineault, M. (1992). Mortality par accident de travail au Québec de 1981 à 1988. Département de santé communautaire, Hôpital du Sacré-Coeur de Montréeal.

Sarrel, P.M. (1991). Women, Work and Menopause. In Frankenhaeuser, M., Lundberg, U., & Chesney, M. Women, Work and Health. New York: Plenum Press. 225-237.

Saurel-Cubizolles, M-J., Bourgine, M., Touranchet, A., & Kaminski, M. (1991). Enquête dans les abattoirs et les conserveries des régions Bretagne et Pays-de-Loire. Conditions de travail et santé des salariés. Rapport à la Direction Régionale des Affaires Sanitaires et Sociales des Pays-de-Loire. Villejuif, France, INSERM Unité 149.

Saurel-Cubizolles, M. J., Kaminski, M., Du Mazaubrun, C., & Bréart, G. (1991). Les conditions de travail professionnel des femmes et l'hypertension artérielle en cours de grossesse. Revue Epid. et santé publ. 39, 37-43.

Savitz, J. A., & Harlow, S. D. (1991). Selection of reproductive health endpoints for environmental risk assessment. Environ Health. 90, 159-164.

Semenciew, R., Morrison, H., Riedel, D., Wilkins, K. Ritter, L., & Mao, Y. (1993). Multiple myeloma mortality aand agricultural practices in the prairie provinces of Canada. JOM. 35,557-561.

Shortridge, L. A. (1988). Assessment of menstrual variability in working populations. Rep Toxicol. 2, 171-176.

Silman, A. J. (1987). Why do women live longer and is it worth it? British Medical Journal. 293, 1211-1312

Silverstein, B. A, Fine, L. J, Armstrong, T. J. (1987). Occupational factors and carpal tunnel syndrome. Am J Ind Med. 11, 343-358.

Skov, P., Valbjorn, O., & Pedersen, B.V. (1990). The Danish Indoor Climate Study Group: Influence of indoor climate on the sick building syndrome in an office environment. Scand J Work Environ Health. 16, 363-371.

Skov, P., Valbjorn, O., Pederson, B.V. (1989). The Danish indoor climate study group: Influence of personal characteristics, job-related factors and psychosocial factors on the sick building syndrome. Scand J Work Environ Health, 15, 286-295.

Stanosz, S., Kuligowski, D., & Pieleszek, A. (1995). Concentration of dihydroepiandrosterone, dihydroepiandrosterone sulphate and testosterone during premature menopause in women chronically exposed to carbon disulphide. Meycyna Pracy. XLVI(4), 340.

Statistics Canada. (1994). Women in the Labor Force. Statistics Canada No. 75-507E. Table 6.8.

Statistics Canada. (1995). Women in Canada: A Statistical Report. Cat. No. 89-503E. Ottawa.

Stein, Z., & Hatch, M. (1987). Biological markers in reproductive epidemiology: Prospects and precautions. Environmental Health Perspectives. 74, 67-75.

Steingart, R. M. et al. (1991). Sex differences in the management of coronary artery disease. New England Journal of Medicine. 325,226-230.

Stenberg, B., & Wall, S. (1995). Why do women report "sick building symptoms" more often than men? Social Sciences and Medicine. 40(4), 491-502.

Stenberg, B., Eriksson, N, Mild, K. H. L., Höög, J., Sandström, M., Sundell, J., & Wall, S. (1995). Facial skin symptoms in visual display terminal workers. A case-referent study of personal, psychosocial, building and VDT-related risk indicators. Int J Epidemiology. 24(4), 796-803.

Stetson, D. S., Albers, J. W, Silverstein, B. A., Wolfe, R. A. (1992). Effects of age, sex, and anthropometric factors on nerve conduction measures. Muscle and Nerve. 15, 1095- 1104.

Stevenson, J., Greenhorn, D. R., Bryant, J. T., Deakin, J. M., & Smith, J. T. (1996). "Selection test fairness and the incremental lifting machine," Applied Ergonomics in press. Strength testing issues are further discussed in Messing, K. and Stevenson, J. 1996. A Procrustean bed: Strength testing and the workplace. Gender, Work and Organizations, in press.

Theorell, T. (1991). Psychosocial cardiovascular risks -- on the double loads in women. Psychother Psychosom. 55, 81-89.

Tissot, F., & Messing, K. (1995). Perimenstrual symptoms and working conditions among hospital workers in Quebec. American Journal of Industrial Medicine. 27, 511-522.

Tomaskovic-Devey, D. (1993). Gender and Racial Inequality at Work. Ithaca, New York, ILR Press.

Toriola, A. L, & Mathur, D. N. (1986). Menstrual dysfunction in Nigerian athletes. Brit J Obstet Gynaecol. 93, 979-985.

Treloar, A. E., Boynton, R. E., Behn, B. G. & Brown, B. W. (1967). Variation in the human menstrual cycle through reproductive life. Int J. Fertil. 12, 7-125.

Tsai, S. P., Bernacki, E. J., & Dowd, C. M. (1989). Incidence and cost of injury in an industrial population. Journal of Occupational Medicine. 31,781-784.

Uehata, T., & Saskawa, N. (1982). The fatigue and maternity disturbances of night workwomen. J Human Ergol. 11 suppl., 465-474.

Vézina, M., Cousineau, M., Mergler, D. et al. (1992). Pour donner un sens au travail: Bilan et orientations du Québec en santé mentale au travail. Boucherville: Gaëtan Morin éditeur.

Vézina, N., Chatigny, C., & Messing, K. (1994) A manual materials handling job: Symptoms and working conditions among supermarket cashiers. Chronic Diseases in Canada. 15(1), 17-22.

Vézina, N., Tierney, D., & Messing, K. (1992). When is light work heavy? Components of the physical workload of sewing machine operators which may lead to health problems. Applied Ergonomics. 23,268-276.

Wagener, D. K., & Winn, D. W. (1991). Injuries in working populations: Black-White differences. Am J Publ Health. 81, 1408-1414.

Waldron, I. (1991). Effects of labor force participation on sex differences in mortality and morbidity. In Frankenhaeuser, M., Lundberg, U., & Chesney, M. Women, Work and Health. New York: Plenum Press. 17-35.

Walters, V., Lenton, R., & McKeary, M. (1994). Women's Health in the Context of Women's Lives. Report submitted to the Health Promotion Directorate, Health Canada.

Walters, V., Beardwood, B., Eyles, J., & French, S. (1995). Paid and unpaid work roles of male and female nurses. In Messing, K., Neis, B. & Dumais, L. (eds.) 1995. Invisible: Issues in Women's Occupational Health and Safety/Invisible: La santé des travailleuses. Charlottetown, PEI: Gynergy Books. pp. 125-149.

Waters, T. R., Putz-Anderson, V., Garg, A., & Fine, L. J. (1993). Revised NIOSH equation for the design and evaluation of manual lifting tasks. Ergonomics. 36(7), 749-776.

White, J. (1993). Sisters and Solidarity. Toronto: Thompson Educational Publishing.

Wilkinson, W. E. (1987). Occupational injury at a midwestern health science center and teaching hospital. AAOHN Journal. 35, 367-376.

Yassi, A., Weeks, J. L., Samson, K., & Raber, M. B. (1989). Epidemic of "shocks" in telephone operators: Lessons for the medical community. Canadian Medical Association Journal. 140, 816-820.

Zahm, S. H., Pottern, L. M., Lewis, D. R., Ward, M. H., & White, D. W. (1994). Inclusion of women and minorities in occupational cancer epidemiological research, Journal of Occupational Medicine. 36(8), 842-847.

Zhou, S. Y., Liang, Y. X., Chen, Z. Q., & Wang, Y. L. (1988). Effects of occupational exposure to low-level carbon disulfide (CS(2)) on menstruation and pregnancy. Ind Health. 26, 203-214.

Zita, J. N. (1989). The pre-menstrual syndrome: "Dis-easing" the female cycle. In Tuana, N. (Ed). Feminism and Science. Bloomington, Indiana: Indiana University Press, pp 188-210.

For a complete reprint of this article contact Haworth Press by telephone (1-800-HAWORTH) OR EMail (getinfo@haworthpressinc.com). Article copyright The Haworth Press, Inc.