Introduction In the context of delayed retirement age, we aimed to study the association between a score for global evaluation of perceived physical strain (PPS) at work and limitations after retirement in the GAZEL cohort.
Methods At baseline in 1989 and every year since then, the PPS question, ‘Do you find that your work is physically strenuous?’ was used as a proxy measure of general occupational physical exertion or load; it was coded into a score, and divided into four categories for men and three for women. The self-report question about limitations was asked in 1989 and 2012 (difficulties performing some daily life activities). Among men and women without limitations at baseline, relationships were studied between reported limitations in 2012 and a cumulative score based on PPS since 1989, allowing examination of the dose–effect relationship, and adjusted for age and perceived state of health at baseline.
Results From 1989 to 2012, 9326 participants without limitations at baseline were followed and filled out the 2012 questionnaire. In 2012, 12.1% of men (n=845) and 12.9% of women (n=302) reported limitations. Limitations in 2012 were associated with very high categorical PPS in men (OR 1.7 (1.4 to 2.2)) and high/very high categorical PPS in women (OR 1.6 (1.2 to 2.2)), with a significant trend.
Conclusions A positive association was found between preretirement physically arduous working conditions and limitations in daily activities after retirement. Findings offer a new insight for global evaluation of physical exposures during working life.
Statistics from Altmetric.com
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
In the context of delayed retirement age, global evaluation of physically arduous working conditions is needed for public health research and practice.
A global evaluation of physically arduous working conditions has been associated with limitations in daily activities after retirement in the GAZEL cohort.
A further step might be a job-exposure matrix approach in order to identify jobs for which early retirement could improve quality of life after retirement.
Many countries are trying to delay retirement age, in a context of economic constraints, ageing populations and public pension funds facing financial difficulties.1 One important barrier to raising the retirement age is premature ageing caused by arduous and difficult working conditions. Different studies have found that particular biomechanical exposures during the working life are associated with pain and other negative outcomes after retirement,2 ,3 as in the case of major chemical exposure related to cognitive decline.4 Furthermore, age and number of years at work are closely related to disability, and many scientific questions about the ability of workers to continue working into their later years remain.5 However, global evaluation of physically arduous working conditions is complicated and raises many questions from a public health perspective.6 ,7 A recent study has indicated that a single self-reported variable of perceived physical strain (PPS) at work is a reliable and valid proxy measure for physical load at work.8 We aim here to study the association between a cumulative score based on PPS during working life and limitations after retirement in the GAZEL cohort.
This analysis was conducted within GAZEL, a French occupational cohort composed of 15 011 men and 5614 women who were employees at Electricité de France-Gaz de France (EDF-GDF; GAZEL stands for GAZ and ELectricité) in 1989.9 Participants completed an annual self-report questionnaire, supplemented by EDF-GDF administrative records.
At baseline in 1989, the PPS question, ‘Do you find that your work is physically strenuous?’ was used as a proxy measure of general occupational physical exertion or load, and response options ranged from 1 (not at all strenuous) to 8 (very strenuous). Perceived state of health was used to measure overall health status. Responses to the question ‘How would you describe your general health?, with options from 1 (very good) to 8 (very poor)’ were recoded into three categories (1, 2; ‘good’), (3, 4, 5; ‘fair’) and (6, 7, 8; ‘poor’). The self-report question about limitations (‘do you have difficulties performing some daily life activities,’ with options ‘yes’ or ‘no’) was also included in the 1989 questionnaire.
In a follow-up, the PPS question was repeated every year until retirement, death or loss of contact, up to 2012. Our primary outcome of interest was the 2012 question about daily activity limitations.
At baseline, PPS was coded in four categories for men (1, 2, ‘low’), (3, ‘moderate’), (4 ‘high’) and (5, 6, 7, 8, ‘very high’), based on the distribution, and allowing at least 20% for both the ‘very high’ exposure group and the reference group (‘low’). Given the lower number of women exposed to physical constraints in the GAZEL cohort, PPS involved three categories for women, with merging of the ‘high’ and ‘very high’ options, that is, (1,2, ‘low’), (3, ‘moderate’) and (4, 5, 6, 7, 8, ‘high/very high’).
A score was computed in order to obtain a cumulative global evaluation of physically arduous work using PPS, by summing the answers provided yearly by each participant for the period from 1989 to 2012. To take into account the years where there was no response, the sum was divided by the number of years in which there were answers to the questionnaire and multiplied by 24 (the maximum number of questionnaires filled out in the period), and called sumPPS. To explore the dose–effect relationship, the sumPPS was divided into four categories for men (24–72; ‘low’), (72–96; ‘moderate’), (96–120; ‘high’), (>120 ‘very high’), and three categories for women (24–72; ‘low’), (72–96; ‘moderate’), (>96; ‘high/very high/’), noted as ‘categorical PPS’, so as to have a categorisation similar to that for the baseline (PPS multiplied by 24).
Among followed men and women without limitations at baseline, the relationships were studied between reported limitations in 2012 and categorical PPS since 1989, adjusted for age and perceived state of health at baseline.
In order to take into account failures to complete annual questionnaires between 1989 and 2012 (or retirement), multiple imputations of annual PPS were performed on the study population as sensitivity analyses. The annual PPS was built as a nominal variable by two generalised mixed models (for women and for men) adjusted for age, PPS, perceived health state in 1989, and limitations in 2012, with a quadratic relationship with time since 1989. An imputed categorical PPS was calculated on the basis of imputed annual responses to PPS between 1989 and 2012 (or retirement). A sensitivity analysis was also performed by not considering the last 5 years of the PPS and related categorical PPS, such as a trend test. Similar analyses using continuous variables were also performed. Authorisation from the appropriate Institutional Review Board (‘Commission nationale de l'informatique et des libertés’) was obtained.
In 1989, 15 011 men were included, aged 41–50 years, and 5614 women aged 36–50 years. At baseline, complete data on 18 930 participants were available (91.8%); 19.5% of men (n=2719) and 21.8% of women (n=1091) reported limitations; and 4.5% of men reported poor health (n=626) as did 4.8% of women (n=238). In 1989, 29.4% (n=4092) of the men reported having strenuous work in the group with the highest exposure, and so did 30.80% (n=1540) of the women. For 1989, a strong cross-sectional association was found between PPS and limitations (table 1).
In the longitudinal study, 11 485 participants were followed until 2012, and filled out the last questionnaire; mean age was 66.2 years (median 66 years, IQR=64–69); mean retirement age was 55.4 years (median 55, IQR=54–57), and the number of questionnaires filled out was 10.5 (median 11, IQR=8–13); 9326 (81%) had no limitation reported at baseline. Those who were followed were older, in better health, with a lower proportion of limitations, and lower exposure to PPS at baseline (p<0.0001). The sumPPS ranged from 24 to 193. In 2012, 12.1% of men (n=845) and 12.9% of women (n=302) reported limitations (participants with baseline limitations excluded). Limitations in 2012 were associated with the career-long summary score for categorical PPS in men and women (table 1), with a significant trend for men and women (p<0.05). A significant relationship between limitations in 2012 and continuous sumPPS was also found (p<0.05). Multiple imputation gave similar results, as did sensitivity results using a similar model and score without the last 5 years of exposure (data not shown).
We found a positive association between physically arduous working conditions and limitations in daily activities after retirement, when participants were aged 55–75 (women) and 65–75 years (men), in a cohort of now-retired French utility workers. Nonetheless there are issues that need to be considered.
First, attrition and non-response should be discussed. Attrition in the GAZEL cohort is very low (0.6% were lost to follow-up by December 2013)8 and is not likely to modify the findings. It is nevertheless plausible that workers with difficult working conditions are more likely to have medical problems, as it has been shown that participants having serious health conditions are less prone to answer the annual questionnaire;10 failure to answer may have led to an underestimation of the association between high PPS and limitation in 2012.
Possible reverse causation was taken into account by adjusting for self-reported health and exclusion of participants with limitations at baseline. However, impaired health in the follow-up period might have consequences for reported PPS and limitations. Further studies should also consider self-reported health every year. Future work will also explore the role of retirement age, years employed, medical conditions and other factors in the association between PPS and limitations in activities of daily living. Limitations were studied considering only activities in everyday life, without information on which daily life limitations were involved. Future studies should use a more complete assessment of disability.
PPS represents an emerging proxy for the global evaluation of physical exposure during working life. When exploring adverse long-term effects on health, global evaluation is indeed necessary.11 In a previous study using the GAZEL cohort, where physical working conditions during the working life were considered via a retrospective evaluation of the number of years of exposure to ergonomic strain, adverse physical working conditions showed long-term consequences for quality of life in older age.12 The present prospective analysis based on repeated evaluations strengthens these previous results by showing dose–response relationships.
In conclusion, this preliminary study offers new insights into the global evaluation of physical exposure during working life and its long-term impact on physical limitations. A further step might be a job-exposure matrix approach, in order to differentiate between those jobs for which early retirement could improve quality of life after retirement, and those for which later retirement may be appropriate.
The authors express their thanks to EDF-GDF, especially to the Service Général de Médecine de Contrôle and to the “Caisse centrale d'action sociale du personnel des industries électrique et gazière.” They also wish to acknowledge the “Population-Based Epidemiological Cohorts” unit responsible for the GAZEL database management. The GAZEL cohort Study was funded by EDF-GDF and INSERM and received grants from the “Cohortes Santé TGIR Program.” This study is supported by the French National Research Agency (ANR). Richard Carter helped us to correct the English.
Contributors AD, MG and ELS have initiated the work and designed the study. MZ, MC, AL and MG have participated to the conception and design of the GAZEL Cohort. AD, EH and MC have performed the analyses. AL, MG, MZ, ELS and AD have been involved in the interpretation of data. AD drafted the article. All of the other authors have reviewed it critically for intellectual content. All the authors gave final approval of the version submitted.
Competing interests AD has received fees from the national health insurance for work compensation and early retirement for his expertise in the field (indemnity independent of any decisions he has made about accepting or refusing cases), from Elsevier Masson as editor-in-chief of the journal “les archives des maladies professionnelles et de l'environnement.”
Patient consent Obtained.
Ethics approval CNIL.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement All Gazel data are available on request from Dr MZ, head of the Inserm UMS 011 (http://www.gazel.inserm.fr/en/).