Ill health - completed research

This page summarises the key findings from recent completed research projects relating to work-related ill health:

Analysis of the correlates of self-reported work-related illness in the Labour Force Survey


Key findings:

  • Males in work are 16% less likely to suffer from work-related ill-health than females in work
  • The risk of work-related ill-health increases steeply over the life course, peaking amongst those aged 50 to 59
  • Occupation is the most important risk factor for work-related ill health, particularly for musculoskeletal disorders
  • For stress, depression and anxiety, occupation and hours of work are jointly the most important risk factors, with longer hours being associated with higher rates of mental ill health
  • Job tenure also has a significant influence on rates of work-related ill health with those who have been in their jobs longer having a higher risk of work-related illness
  • A spouse or partner acting as a proxy respondent is associated with a 26% reduction in the likelihood that an individual is recorded as suffering from work related ill-health. This increases to 53% where the proxy respondent is not a spouse or partner
  • Respondents to the first wave of LFS interviews are most likely to report that they suffer a from a work-related ill-health condition, reflecting the better quality of data collected from face to face interviews compared with interviews in later waves that are conducted over the phone
  • Based upon both the differential quality of proxy respondents and responses provided during later waves of the survey, overall rates of under-reporting within the LFS would be expected to be in the order of 20-25%

A comparative analysis of self-reported and medically certified incidence data on work-related illness


Key findings:

  • Overall, the Labour Force Survey (LFS) reports 25% higher rates for work-related ill health compared with the THOR-GP scheme, possibly due to the LFS capturing less severe cases which are not referred to a GP
  • Stress and Musculoskeletal Disorders (MSDs) dominate both sources although rates of MSD are higher in THOR-GP and rates of stress, depression and anxiety are higher from the LFS. This could reflect the inclusion of some injuries in the THOR data and GPs possibly identifying other non-work factors as the underlying cause of stress;
  • Rates of work-related skin disease are substantially higher from THOR-GP than the LFS. This may reflect better awareness amongst the GP reporters who have been trained in occupational medicine of possible work-related causes compared to the awareness of the individual;
  • The LFS shows higher rates of "other" work-related illnesses;
  • The age and industry profile of work-related ill health sufferers is very similar from both sources;
  • The LFS shows higher ill health rates for skilled and professional workers compared with THOR-GP although this is confounded by the higher rates of stress that are found in the LFS.

Follow-up and assessment of self reports of work-related illness in the Labour Force Survey


Key findings:

  • An expert review panel accepted that in 77% of cases work was the main or contributory cause of the illness. In a further 10% of cases it may have exacerbated symptoms;
  • In 80% of cases, where information was available, the same decision was made by the respondent's doctor and the review panel;
  • Variables most strongly associated with apparent mistaken ascription were occupation and time since end of exposure to the work the respondent believed had caused or made their illness worse;
  • Psychological characteristics and illness beliefs of respondents do not appear to have a significant effect on the reliability of the reports, after accounting for occupation and time since end of exposure;
  • Work-related illness has an impact on the daily lives of individuals; the general physical and mental health of those affected tends not to be good;
  • There is no standard measure of work-related illness without problems; all available sources are subject to some kind of error. Self reports are broadly reliable; the level of mistaken reports will to some extent be counter-balanced by opposite biases. When sensibly interpreted, such surveys provide valid information not available from other sources.

Associated documents

Is this page useful?

Updated 2022-11-21