Search tips
Search criteria 


Logo of oenvmedOccupational and Environmental MedicineVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Occup Environ Med. 2007 December; 64(12): e9.
PMCID: PMC2095348

Psychosocial 2

031 Alcohol and occupation in Australia: results from the 2004–05 National Health Survey

D. R. Smith1, P. A. Leggat2. 1National Institute of Occupational Safety and Health; 2James Cook University

ObjectivesThe aim of this study was to provide the most recent estimates of Australia's national alcohol consumption rates stratified by occupation.

MethodsAlcohol consumption data were extracted from the 2004–05 National Health Survey, which achieved a response rate of around 90% and captured approximately 26 000 persons in all occupational categories. Participants were limited to those of working age (18–64 years), with data stratified by job category and gender during the analysis. Alcohol consumption was based on an individual's estimated daily consumption of alcohol during the previous week, with risk categories (high, medium and low‐risk) adapted from international substance use guidelines.

ResultsThe prevalence of high‐risk alcohol consumption among Australian workers in 2004–05 was estimated to be 6.1%, with very little differences reported between the employed (6.5%) and unemployed (6.3%). More than half of those who were currently employed were low‐risk drinkers (53.9%) however, compared to only 40.4% of the unemployed. By specific job category (both genders combined), high‐risk alcohol consumption was most commonly seen among electrical tradespersons (13.7%) and the least common among education professionals (1.6%). Associate professionals were the most likely to be low‐risk alcohol drinkers (67.0%), whereas elementary service workers were the least likely to be so (38.8%). Among Australian male workers, cleaners were the most likely to be high‐risk alcohol drinkers (15.2%), and education professionals (2.0%) the least likely. Among female workers, 11.2% of elementary service workers were high‐risk drinkers, compared to only 1.5% of education professionals.

ConclusionOverall, the current study suggests that Australian alcohol consumption patterns vary widely depending on occupation. Aside from its negative effect on general health, high‐risk alcohol consumption also has major implications in workplaces where safety may be compromised, particularly those where driving comprises a major part of daily work life. Effective harm reduction strategies and lifestyle support programs targeting vulnerable sections of the workforce are clearly needed.

Key wordsoccupation; high‐risk alcohol consumption; epidemiology

032 Disentangling depressive mood and tension in the SF‐36 scale

H. Burr, T. S. Kristensen, R. Rugulies, U. Bültmann, V. Borg. National Institute of Occupational Health, Denmark

ObjectivesThe SF‐36 Mental Health scale has been criticised for measuring an ambiguous construct. We analysed (1) if tension and depressive mood subscales have the same gender and age pattern and (2) if the risk factors for tension and depressive mood were the same.

MethodsThe Danish Work Environment Cohort Study (DWECS) contains: a cross‐section of 5636 employees (response rate 75%) for the year 2000; a 1995–2000 cohort of 4647 people being employees at baseline (response rate 67%). A tension scale was based on two SF‐36 items: “How much of the time during the past 4 weeks have you ‐ been a very nervous person?” and “‐ felt calm and peaceful?” A depressive mood scale was based on three SF‐36 items: “‐ felt so down in the dumps that nothing could cheer you up?”, “‐ felt downhearted and blue?”, “‐ been a happy person?” Responses were coded so that 100 indicated maximal tension/depressive mood and 0 minimal tension/depressive mood. In the cohort, linear regression analyses were conducted with tension and depressive mood in follow‐up as dependent and psychosocial work environment at baseline as independent variables controlled for age, family status, and tension and depressive mood, respectively, at baseline.

ResultsMean tension was highest in the younger age groups (18–29‐year‐old males: 14.7, females: 18.9) and lowest in the older (50–59‐year‐old males: 12.2, females: 14.1). Depressive mood showed a clear age pattern among females (18–29‐year‐old females: 15.1; 50–59‐year‐old females: 12.6). Among males, conflicts at work, low social support and job insecurity in 1995 predicted both tension and depressive mood in 2000. Low skill discretion predicted only depressive mood, young age predicted only tension. Among females, low social authority and conflicts at work in 1995 predicted both tension and depressive mood in 2000.

ConclusionIt seems that depressive mood and tension to a large degree occur in the same subjects and have mostly the same risk factors. However, the different age pattern and some few different risk factor patterns among men suggest that depressive mood and tension are different constructs.

Key wordsmental health; psychosocial work environment; depressive symptoms

033 Characterising patterns due to mental health absences among workers on short‐term disability

I. Rivilis1, D. C. Cole1, L. Scott2. 1Institute for Work & Health; 2Organizational Solutions

ObjectivesThe growing number of mental health claims is a concern for Canadian employers and the working population in general. However, little research has been done to study the population of workers disabled by these conditions. In particular, little is known about claimants on short‐term disability (STD) due to mental health conditions. The purpose of this study was to use available disability management data to examine patterns in the occurrence and duration of short‐term disability due to mental health conditions, and to describe the characteristics of workers affected by these conditions.

MethodsWe collaborated with a work disability management consulting company to obtain administrative data on STD. We carried out a descriptive analysis to identify trends in mental health STD occurrence, as well as to elucidate factors that are associated with mental health disability in a population of STD claimants. Data were collected retrospectively over a 2‐year period on a purposeful sample of 10 client companies ranging in size (from 900 to 4000 employees), representing various sectors (eg, healthcare, manufacturing, communications, and utility sectors) and various private insurers. STD conditions were classified according to ICD‐10 codes. Reliability assessment was performed independently by two reviewers to assure consistent coding. Proportions and duration of mental health disorders were examined. Variations in STD frequency were examined by demographics factors (eg, job type, sex) and sector.

ResultsWe observed that STD due to mental health disorders were highly prevalent conditions in this population, and generally tended to have long durations. The proportion of mental health claims ranged from 8% to 28% across different companies. Some variations across demographic factors were also observed.

ConclusionVariability in the prevalence and duration of mental health STD demonstrates the need for further research to identify modifiable risk factors that could lead to effective work disability prevention and control strategies. Collaborating with a work disability management consultant provided the unique opportunity to access data from private insurers and across various organisations simultaneously, as well as the advantage of forming productive research–practitioner partnerships.

Key wordsmental health; short‐term disability; patterns

034 Do psychiatrists (PSYs) and occupational physicians (OPs) have different diagnostic preferences for cases of occupational psychological ill‐health?

E. O'Neill1, S. Turner2, R. McNamee3, L. Hussey2, S. Lines2, C. Roberts2, S. Taylor2, R. Parker2, R. Agius2. 1Department of Occupational Health and Safety, Manchester Royal Infirmary, UK; 2Centre for Occupational and Environmental Health, University of Manchester, UK; 3Biostatistics Group, University of Manchester, UK

ObjectivesThe Health and Occupation Reporting (THOR) network is a UK‐based voluntary surveillance scheme for occupational ill‐health, and involves more than 2200 physician reporters. Differences in reporting patterns to THOR by psychiatrists (PSYs) and occupational physicians (OPs) for psychological ill‐health have been noted. For THOR data 2002–2005, estimated annual incidence rates (per million) for anxiety and depression were 60 (95% CI 50 to 69) using PSYs' reports and 885 (95% CI 776 to 994) using OPs' reports. For “other work‐related stress” the rates were 6 (95% CI 3 to 9) and 624 (95% CI 534 to 714), respectively. This study investigated whether PSYs and OPs have different diagnostic preferences, and estimated the reliability of their diagnoses.

MethodsPhysicians reporting cases of psychological ill‐health to THOR were asked to produce case summaries, without a diagnosis, using a proforma. A collection of 50 summaries from PSYs and 50 from OPs was made. Each was sent to five PSYs and five OPs (raters), who independently assigned a diagnosis and judged whether the case was work related. Differences between PSYs and OPs in diagnostic proportions and judgments were assessed using two‐level multinomial/binomial models; the reliability of diagnostic categories within each group of physicians, and overall, was measured by inter‐rater κ coefficients.

ResultsDiagnostic proportions for PSYs and OPs demonstrated considerable overall agreement: for anxiety and depression they were PSYs 27%, OPs 27% and for “other work related stress” PSYs 14%, OPs 17%. κ Coefficients for the two groups were similar. Raters were able to differentiate well a psychotic diagnosis (κ 0.78, 95% CI 0.74 to 0.83) but less able to differentiate anxiety (κ 0.27, 95% CI 0.23 to 0.32), depression (κ 0.34, 95% CI 0.29 to 0.38) or “other work related stress” (κ 0.15, 95% CI 0.10 to 0.19). The OR of a case being classified as work related by PSYs compared to OPs was 2.39 (95% CI 1.68 to 3.38) (p<0.001).

ConclusionWhen presented with the same information, PSYs and OPs do not appear to have systematically different diagnostic preferences; reporting differences might instead be attributable to “case mix”. PSYs were more likely to consider a case to be work related than OPs, which may reflect differences in day‐to‐day workload, training or experience of the two groups of raters.

Key wordssurveillance; psychological ill‐health; work‐related

035 Chemical sensitivity in different occupational groups

B. E. Moen1, S. Torp2, T. Riise1. 1University of Bergen; 2Vestfold University College

ObjectivesThe diagnosis “multiple chemical sensitivity” was introduced several years ago. However, the existence of such a disorder is debated. Although there are doubts about the objectivity of this entity, these problems seem to occur, but it is unknown to what extent. Chemical exposure in working life has been suggested as causative agents for the diagnosis. The aim of the study was to examine the prevalence of chemical sensitivity in different occupational groups in a general working population, and to study their health, lifestyle and working environment.

MethodsAs a part of a cross‐sectional study of the population aged 40–44 years in the county of Hordaland, Norway, the participants answered a questionnaire on age, gender, occupation, educational level, life‐style factors and questions related to psychosocial work environment (demand, control, support questionnaire). The SF‐12 Health Survey was used for measuring health‐related quality of life. The participants also reported experienced gastrointestinal, central nervous, respiratory, musculoskeletal, genital or dermal symptoms on a five‐point scale. They were asked if any of these symptoms occurred when they experienced the smell of perfume, cooking, exhaust or similar odours. The study defined the presence of chemical sensitivity if at least one symptom from two organ groups was rated as 5 (almost always) and the person experienced the symptoms when exposed to the mentioned odours.

ResultsA total of 8598 men and 9983 women participated, participation rate: 63%. The prevalence for chemical sensitivity was 20% for women and 13% for men. Among women the prevalence was highest for the occupations drivers, receptionists, farmers, restaurant workers and cleaners. Among men, the prevalence was highest among painters, restaurant workers, mechanics, drivers and industrial assistants. Comparing these workers (n = 760) with the rest of the population revealed no differences in demographic factors, lifestyle factors or psychosocial work environment. However, the quality of life scores in this group were significant lower, both for the physical and mental summary scales.

ConclusionThe prevalence of chemical sensitivity is highest in occupations where the workers are exposed to specific smells. These persons have a reduced health‐related quality of life.

Key wordschemical sensitivity; occupation; quality of life

036 Unemployment and mortality during the national economic crisis in Korea

H. R. Kim, S. Y. Park. Department of Preventive Medicine, Catholic Industrial Medical Center, The Catholic University of Korea

ObjectivesThe aim of this study was to evaluate the effect of unemployment on mortality during the national economic crisis in Korea.

MethodsStudy subjects were 31 127 workers who underwent the worker's health examination in an occupational health clinic in Incheon, Korea in 1996. Workers who received benefits from the unemployment insurance system were defined as the unemployed group. The χ2 test was performed to compare causes of death between the unemployed and employed group, and logistic regression analysis was performed to evaluate the effect of unemployment on mortality. The unemployed group was divided into four subgroups according to the year of receiving unemployment benefits (1997–2000), and relative risk for mortality was calculated to compare each subgroup with the employed group by using Cox's proportional hazard model.

ResultsThe results of this study were as follows. First, mortality by cancer, cardiovascular and cerebrovascular disease in the unemployed group was higher than that in the employed group but was not statistically significant. Second, mortality during the first 4 years after being unemployed in the unemployed group was 1.80 times higher than that in the employed group even after the effect of age, sex and health status were adjusted. Third, female mortality in the 1998‐year subgroup was 2.28 times higher than that in employed group after the effect of age and health status were adjusted. Male mortality in the 1998‐ and 1999‐year subgroup were 1.71 times and 1.99 times higher, respectively, than that in employed group.

ConclusionThese results suggest that unemployment during the national economic crisis increased mortality in Korea.

Key wordsunemployment; mortality; national economic crisis

037 Longitudinal development of frequent and long‐term absenteeism

P. C. Koopmans1, C. A. M. Roelen2, J. W. Groothoff3. 1ArboNed Groningen; 2ArboNed Zwolle; 3University Medical Center Groningen, University of Groningen

ObjectivesFrequent and long‐term absentees were monitored over a period of 4 years after the base year with regard to the risk of recurrence.

MethodsBased on their absenteeism in the base year of 53 990 employees working at three large Dutch companies in the postal and telecommunications sector, the following cohorts were distinguished: 4126 frequent absentees (four times or more), 3585 long‐term absentees (duration more than 6 weeks), 979 “frequent and long‐term” absentees, and a reference population consisting of 45 300 employees. A log‐rate analysis was performed, with rates of frequent and long‐term absence rates as dependent variables and former absenteeism as explaining variable, corrected for employee and function characteristics.

ResultsThe odds ratio of developing a year with recurrent frequent absence in a period of 4 years amounts to 5.3 in men and 3.6 in women (as compared to the reference population). In former long‐term absentees the odds ratios are 1.7 in men and 1.3 in women. The odds ratio of developing a long‐term absence is higher for former frequent absentees (OR 2.1 in men and OR 1.7 in women) and former long‐term absentees (OR 2.1 in men and OR 1.6 in women) as compared to the reference population. “Frequent and long‐term” absenteeism gives the highest odds ratio of recurrent long‐term absenteeism (OR 2.6 in men and OR 1.8 in women). Risk groups for frequent absenteeism are females, younger employees, unmarried employees, employees in lower salary scales and employees in an urban workplace. Risk groups for long‐term absenteeism are females, older employees, employees in lower salary scales and employees in an urban workplace.

ConclusionEmployees with former frequent and/or long‐term absenteeism show a higher risk of relapse in the following years as compared to the reference population. Frequent absenteeism also serves as a signalling factor for future long‐term absenteeism. Attention can be given to specific risk groups.

Key wordslong‐term absenteeism; frequent absenteeism; longitudinal study

038 The prognostic value of injured workers' recovery expectations for timely return to work

M. C. Battié, D. Gross. University of Alberta

ObjectivesThere has long been an interest in identifying individuals reporting work‐related injuries who are at high risk for prolonged disability. However, previous research has demonstrated that predicting successful return to work is a complicated and challenging task involving multiple factors. We expanded on recent work that has attempted to broaden the scope of contextual factors influencing work disability by examining injured worker's expectations of recovery. In particular, we sought to determine whether perceptions of physical capacity to resume and sustain regular work activities, have prognostic value for timely, successful return to work.

MethodsWe conducted a prospective cohort study. Study participants were work injury claimants entering the primary rehabilitation centre of the Workers' Compensation Board of Alberta, Canada for rehabilitation or assessment between April 2001 and March 2002. All had open claims for back injury or clearly objective trauma‐related injuries and were on total temporary disability. A measure of patient expectations for recovery and resumption of work was added to data routinely collected at initial (baseline) assessment at the rehabilitation centre. Data on a variety of possible confounding factors were also gathered. Time in days from baseline measurements to suspension of total temporary disability status in the following year served as a proxy outcome measure of timely return to work and functional recovery. Cox proportional hazards regression analysis was used for univariate and multivariable analyses. Initially, analyses were performed on the entire dataset, then separately for back versus objective injury subgroups.

ResultsThe entire sample consisted of 536 claimants, including 317 back injury claimants and 219 objective injury claimants. Better recovery expectations were crudely associated with faster suspension of disability benefits (HRR 0.91 (95% CI 0.83 to 0.99), which diminished somewhat in multivariable analysis controlling for Pain Disability Index score, and other identified confounding factors (HRR 0.93 (95% CI 0.85 to 1.02)). The prognostic value of worker's expectations of recovery and return to work tended to be higher among back injury claimants than among those with clearly defined trauma and resultant pathology.

ConclusionRecovery expectations have a modest effect on timely return to work, which appears somewhat greater in claimants with back problems than clear trauma.

Key wordsrecovery expectations; prognostic value; claims outcomes

Articles from Occupational and Environmental Medicine are provided here courtesy of BMJ Group