Article Text

Download PDFPDF

The incidence of medically reported work-related ill health in the UK construction industry
  1. S J Stocks1,
  2. R McNamee2,
  3. M Carder1,
  4. R M Agius1
  1. 1Centre for Occupational and Environmental Health, Health Sciences Group, Faculty of Medical and Human Sciences, University of Manchester, Manchester, UK
  2. 2Health Sciences Methodology, Health Sciences Group, Faculty of Medical and Human Sciences, University of Manchester, Manchester, UK
  1. Correspondence to Dr S J Stocks, COEH, 4th Floor Ellen Wilkinson Building, University of Manchester, Oxford Rd, Manchester M13 9PT, UK; jill.stocks{at}


Objective Self-reported work-related ill health (SWI) data show a high incidence of occupational ill health and a high burden of cancer attributable to occupational factors in the UK construction industry. However, there is little information on the incidence of medically reported work-related ill health (WRI) within this industry. This study aims to examine the incidence of WRI within the UK construction industry.

Method Standardised incidence rate ratios (SRRs) were used to compare incidence rates of reports of medically certified work-related ill health returned to The Health and Occupation Reporting network (THOR) within the UK construction industry with all other UK industries combined.

Results Male UK construction industry workers aged under 65 years had significantly raised SRRs for respiratory (3.8, 95% CI 3.5 to 4.2), skin (1.6, 1.4 to 1.8) and musculoskeletal disorders (MSD; 1.9, 1.6 to 2.2). These SRRs were further raised for those working within a construction trade. The increased SRRs for skin disease within male construction industry workers were due to contact dermatitis (1.4, 1.2 to 1.6) and neoplasia (4.2, 3.3 to 5.3). For respiratory disease, the increased SRRs were due to non-malignant pleural disease (7.1, 6.3 to 8.1), mesothelioma (7.1, 6.0 to 8.3), lung cancer (5.4, 3.2 to 8.9) and pneumoconiosis (5.5, 3.7 to 8.0), but the SRRs for asthma (0.09, 0.06 to 0.11) and mental ill health (0.3, 0.1 to 0.4) were significantly reduced.

Conclusion The significantly raised SRRs for medically reported MSD and significantly reduced SRRs for mental ill health in construction workers confirm self-reported UK data. These SRRs provide a baseline of the incidence of WRI in the UK construction industry from which to monitor the effects of changes in policy or exposures.

  • Occupational health
  • occupational exposure
  • incidence
  • construction industry
  • epidemiology

Statistics from

Request Permissions

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

What this paper adds

  • The UK construction industry has high incidence rates of self-reported occupational ill health, especially musculoskeletal disorders, and a high burden of cancer attributed to occupational exposures.

  • There is a lack of information on the incidence of medically certified work-related ill health in the UK construction industry.

  • This study found significantly increased standardised incidence rate ratios (SRRs) for medically certified musculoskeletal disorders, skin neoplasia, contact dermatitis, non-malignant pleural disease, mesothelioma, lung cancer and pneumoconiosis and significantly reduced SRRs for asthma and mental ill health within UK construction industry workers compared to all other employment sectors combined.

  • Skin neoplasia is a significant work-related ill health problem among UK construction industry workers that may merit increased promotion of preventive measures.

  • This study confirms self-reported data for musculoskeletal disorders and mental ill health and provides a snapshot of the incidence of medically reported work-related ill health in the construction industry compared with all other industries and a baseline from which to monitor the effects of changes in policy or occupational exposures.


Construction workers worldwide are especially vulnerable to occupational ill health and injuries, even in prosperous countries.1 The UK construction industry employs over 2 million people, making it the UK's largest industrial sector.2 Self-reported work-related ill health (SWI) data compiled by the UK Health and Safety Executive (HSE) show an increased incidence of all work-related ill health and musculoskeletal disorders (MSD) and a reduced incidence of stress and depression or anxiety in construction industry workers.3 A self-reported survey of British construction workers also reported a high proportion of MSD (61%) and, contrary to the SWI data, high proportions of stress (14%).4 The self-reported data for SWI and MSD are consistent with medically certified occupational disability in German construction workers, but it is important to corroborate the self-reported data with medically certified evidence in the UK.5

Recently, it was estimated that exposures within the construction industry produce over half of the occupational attributable cancer deaths in Great Britain (comprising bladder, lung, mesothelioma, skin, sinonasal) as well as a high number of non-malignant skin cancer (NMSC) registrations.6 The most commonly implicated exposure was asbestos, followed by solar radiation.

The Health and Occupation Reporting (THOR) network is a voluntary surveillance scheme for reporting cases of medically diagnosed work-related ill health (WRI).7

There is a clear need to address the lack of published information on the incidence of WRI in the UK construction industry.8 Therefore this analysis of THOR data aims to estimate incidence rate ratios in order to systematically compare medically certified WRI in the UK construction industry with all other employment sectors.


THOR has been described in detail elsewhere.7 Briefly, respiratory physicians, dermatologists, rheumatologists and psychiatrists return reports of WRI. Core reporters report every month, while sample reporters return a report for a randomly assigned month each year. Cases returned by sample reporters are multiplied by 12 in order to estimate the annual number of cases, hereafter referred to as ‘estimated reports’.

THOR occupational data are coded using the UK Standard Industrial Classification (SIC) and UK Standard Occupation Classification (SOC). A subset of THOR data for 2002–2008 representing the construction industry was selected using SIC (2007) code 45 and a subset representing the construction trades using UK SOC (2000) codes 5215–5216, 5223, 5241, 5311–5329, 8141–8149 and 9121–9129, corresponding to welders, pipe fitters, metal workers, electrical workers, bricklayers and masons, roofers, plumbing, heating and ventilating engineers, carpenters and joiners, glaziers and window fitters, construction trades n.e.c., plasterers, painters and decorators, scaffolders, road construction operatives, construction operatives n.e.c., labourers in building, woodworking and other construction trades n.e.c.

Standardised incidence rate ratios (SRRs) and rate ratios (RRs) based on the ‘estimated reports’ were calculated in order to compare reported incidence rates within the construction industry or construction trades with the incidence rates for all other employment sectors combined using the Labour Force Survey (LFS) 2005 data as denominators and for direct standardisation.2 9 Consideration was given to the fact that the LFS population includes working people only, whereas many reports to THOR originate from retired workers with long latency illness. Arguably, for those aged 65 and over, the sum of current and retired workers would be better denominators. (However for a rate ratio calculation, the ratio of these denominators is more important than absolute levels.) As the number of retired workers was not available, we used the LFS denominators (ie, working people) for those aged over 65: this would be valid if the ratio of people working aged over 65 to those retired is similar in the construction industry as in all other employment sectors. As this assumption may be questionable and there may in any case be differences in exposures in those over 65, SRRs and RRs were calculated separately for those under and over 65, and for males and females.9 Reports with missing age and/or gender data were not included, but these represented a small proportion of the data (MSD 4.5%, mental ill health 1.5%, respiratory disease 0.9%, skin disease 0.3%). The variance for the SRRs and RRs was estimated assuming a Poisson distribution of disease incidence, taking into account the increased contribution of the sample reports and the proportion of eligible physicians reporting to THOR (approximately 65% for dermatologists, 70% for respiratory specialists, 40% for rheumatologists and 10% for psychiatrists) by means of a finite population correction factor.9

Fisher's exact test was used to evaluate associations between occupation and diagnoses among reported actual (ie, not estimated) cases. These comparisons do not use population denominators.

Multicentre Research Ethics Committee approval (02/8/72) has been given for THOR.


The estimated numbers of UK construction workers based on LFS data (2005) show that male construction workers represent 13–14% of the UK workforce, whereas females represent only 1.7% (identified by SIC), or if identified by construction trades (SOC) 0.1%.

Between 2002 and 2008 over 2000 physicians reported to THOR and 28 068 actual case reports were returned; 3845 (14%) were within the construction industry (SIC) and 5538 (20%) were within construction trades (SOC). The majority of construction industry workers (88%) were employed within construction trades, the remainder being administrative, cleaning or managerial workers. As few reports originated from female construction workers (<1%), SRRs are only presented for males.

The SRRs for different types of WRI in males working within the construction industry (SIC) and construction trades (SOC) diagnosed by clinical specialists are shown in table 1. There were significantly raised SRRs for contact dermatitis, all types of skin neoplasia, non-malignant pleural disease (NMPD), mesothelioma, lung cancer, pneumoconiosis and MSD. In contrast, the SRRs for asthma and mental ill health were significantly reduced.

Table 1

Directly standardised incidence rate ratios for males aged 65 years or under and incidence rate ratios for males aged over 65 years of reports made to The Health and Occupation Reporting Network (THOR) by clinical specialists to compare work-related ill health from 2002 to 2008 in the UK construction industry or construction trades with all other employment sectors combined

Within the construction industry, ultra violet light was specified as the suspected causal exposure in 215/216 (100%) actual reports of skin neoplasia, and asbestos in 1661/1739 (96%) reports of respiratory disease. Of the 400 contact dermatitis reports (54% allergic, 31% irritant, 13% mixed allergic/irritant, 2% not specified), the most common suspected causal agents were chromates (114/400, 29%), cement, plaster and masonry (41/400, 10%), resins (35/400, 9%), wet and dirty work (32/400, 8%), thiurams (30/400, 8%), cobalt salts (13/400, 3%), soaps and detergents (12/400, 3%), nickel salts (10/400, 3%) and flux (9/400, 2%). The more common tasks causing MSD were guiding or holding a tool (80/257, 33%), heavy lifting (47/257, 18%) and materials manipulation (40/257, 16%), and the most frequently reported movement was forceful upper limb or grip (118/257, 46%). Among mental ill health reports to THOR, the proportion of anxiety and stress was low compared with all other industries (26/66, 39% compared to 1691/2453, 69%, p=0.01) but for post-traumatic stress disorder (PTSD) the proportion was high (22/66, 33% compared to 243/2453, 10%, p=2×10−5). The most common precipitating events for mental ill health within construction industry workers were accidents and physical injury (24/66, 36% compared to 146/2453, 6%, p=5×10−10) and traumatic experiences of other people's injuries or fatalities (11/66, 17% compared to 58/2453, 2%, p=4×10−6), 10/11 of which involved fatal accidents.


These SRRs allow useful comparisons to be made between the incidence of medically certified WRI in individual employment sectors and the overall working population of the UK. The significantly raised SRRs for MSD and significantly reduced SRRs for mental ill health in construction workers are consistent with the self-reported data.2 The high rates of skin neoplasia in construction workers agree with reported high numbers of NMSC registrations.6 The low SRR for asthma is consistent with a Finnish study that reported increased risk of asthma among construction workers but noted that only 2% of cases had been classified as occupational by a chest physician, possibly because of the requirement for individual evidence of a causal association.10 Reporters to THOR usually report a suspected causal agent, but this is not required. The high SRRs for mesothelioma and lung cancer are consistent with previously published data from the UK and Germany.6 11 The increased SRR for pneumoconiosis is consistent with an increased prevalence in Dutch construction workers, probably related to exposure to quartz-containing dust.12

Although the low SRR for mental ill health is consistent with the SWI data, it does not agree with a survey of UK construction workers.4 This may be due in part to selection bias in the above survey or psychiatrists seeing only the more serious cases of mental ill health. Furthermore, construction workers may be reluctant to seek help for mental ill health or consult mental health professionals other than psychiatrists. However, the high proportion of reports where the precipitating event is an accident, especially a fatal accident, and PTSD suggests that reducing accidents within the construction industry might have beneficial effects beyond injury reduction.

Comparisons between the SRR and RRs for workers aged over 65 years may be made, but due consideration should be given to the aforementioned differences in the denominator. The possible reduction in the incidence of skin neoplasia in older workers might relate to susceptible workers being diagnosed at a younger age, a survival effect or reduced exposure to sunlight with increasing age or upon retirement. This reduction is particularly marked for melanoma, consistent with the hypothesis that intermittent exposure to intense sunlight plays a role.

The higher SRRs for WRI for workers within a construction trade (SOC) rather than the construction industry (SIC) imply that the actual task is the precipitating factor rather than other industry associated factors such as employee attitudes, for example reduced awareness of exposures or personnel protection equipment use in those not actually working in a trade.

Although the confidence intervals take in to account variance in the numerator as described in the Methods section, variance in the denominator could not be taken into account. This is mitigated by using rate ratios, but bias arising from differences in occupational group size estimates (LFS) or whether THOR reporters are representative of all eligible physicians was not considered.

In conclusion, SRRs allow comparisons of incidence rates in selected employment sectors with all other employment sectors using the THOR scheme and may be used as a baseline to monitor the effects of changes in policy or occupational exposures.


Thanks are due to all physicians reporting to THOR.



  • Funding The THOR project is partly funded by the UK Health and Safety Executive.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the North West NHS MREC 02/8/72.

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