Sickness absence among 299484 blue-collar workers in the trade industry during the Covid-19 pandemic

Background

Many blue-collar workers in the trade and retail industries have jobs that make it hard to avoid contacts with other people, which may have increased their risk for sickness absence (SA) during the Covid-19 pandemic. The aim was to investigate rates of SA and sociodemographic and occupational differences in risk of SA during the Covid-19 pandemic.

Methods

A prospective cohort study of all blue-collar workers in Sweden in the trade and retail industry aged 18-67 in 2018 (n = 299 484), followed 5 years (2016-2020) using linked microdata from nationwide registers. Descriptive statistics of rate of workers having had at least one SA-spell >14 days were calculated, and logistic regression was used to calculate odds ratios (OR) and 95% confidence intervals (CI) of having SA due to Covid-19 or related diagnoses (some infectious, respiratory, and symptom-based diagnoses).

Results

Their SA rates increased from fluctuating between 7.6%-8.2% in 2016-2019 to 10.0% in 2020. 0.05% had SA due to Covid-19 and 2.2% had SA due to Covid-19 or related diagnoses. Factors associated with having SA due to Covid-19 or related diagnoses were older age (OR age 55-64: 3.41, CI 3.04-3.82 compared to 18-25) and only elementary education (OR 1.50, CI 1.37-1.64 compared to university/college). Warehouse and terminal staff (reference category) was the occupational group with the highest risk of SA/DP due to Covid-19 or related diagnoses. Cashiers had the second highest risk, with CIs that overlapped 1 (OR 0.91, CI 0.77-1.06). All other occupational groups had significantly lower ORs (0.48-0.78).

Conclusions

The SA rates increased slightly during the Covid-19 pandemic. Within the trade and retail industry, the warehouse and terminal staff was the occupational group with the highest risk of SA due to Covid-19 or related diagnoses.

Key messages

• The rates of SA among blue-collar workers in the trade and retail industry increased slightly during the first year of the Covid-19 pandemic.

• Warehouse and terminal staff and cashiers were the occupational groups with the highest risk of SA due to Covid-19 or related diagnoses among blue-collar workers in the trade and retail industry.

Sickness absence and disability pension in white-collar workers in the trade and retail industry

Background

Very little is known about sickness absence among white-collars workers in the trade and retail industry, despite being a large and important group on the labour market. The aim was to investigate future sickness absence (SA) and disability pension (DP) in a cohort of privately employed white-collar employees in the trade and retail industry.

Methods

A prospective population-based cohort study of all 192,077 such white-collar workers (44% women) in Sweden in 2012, using linked microdata from three nationwide registers covering 2012-2016. Prevalence and mean number of SA and/or DP net days/year in general and by diagnoses categories were calculated for all and also stratified by sex. Logistic regression was used to calculate odds ratios (OR) and 95% confidence intervals (CI) between sociodemographic and work-related factors and future SA/DP.

Results

The proportion who had SA and/or DP was higher in women (10-13%, depending on year) than men (4-6%) each year. Each studied year, women had more mean SA/DP net days than men in the entire cohort, however, among those who had SA and/or DP, there were no gender differences regarding the mean number of net days. The mean number of SA/DP net days increased for both women and men each year, especially SA due to mental diagnoses. SA in 2012 was the strongest factor associated with SA/DP in 2016 (OR women 3.28, 95% CI 3.09-3.47; men 4.10, 95% CI 3.76-4.48). Work-related factors were only weakly or insignificantly associated with future SA/DP. The ORs for most factors were stronger for men than for women.

Conclusions

More knowledge about the mechanisms behind these results are needed. Some SA/DP measures showed large sex-differences, others not - it is important to use different measures to show the complexity of these phenomena. Several factors were more strongly associated with SA/DP among men than among women, indicating that there are other factors of importance for women.

Key messages

Prior sick leave with mental or somatic diagnoses and being in work in ages 66–71; a Swedish cohort

Background

As longevity and retirement ages are increasing, knowledge is needed on factors hindering extended working lives. The aim of this study was to explore how sickness absence (SA) and disability pension (DP) due to mental and/or somatic diagnoses before age 65 were associated with being in paid work when aged 66-71.

Methods

A 6-year prospective population-based cohort study of all 98,551 people (48% women) in Sweden who turned 65 years in 2010 (baseline year) and had been in paid work at any point when aged 60-64. Microdata from nationwide registers were used. Exposure variables were SA (spells >14 days) and/or DP in 2005-2009, and the outcome variable was paid work at any point in 2011-2016. Logistic regression was used to calculate odds ratios (OR) with 95% confidence intervals (CI) for associations between exposures and outcome, controlling for sociodemographic factors in 2010, stratified by sex.

Results

Most women (56.0%) and men (66.3%) had no SA or DP when aged 60-64. Of the women, 42.7% and of the men 53.3% were in paid work after the age of 65. Those with SA due to mental diagnoses had lower OR of being in paid work (women 0.76; 95% CI: 0.69-0.84; men 0.74; 0.65-0.84). This association was weaker for SA due to somatic diagnoses (women 0.87; 0.84-0.91; men 0.92; 0.89-0.96). Having had SA due to both mental and somatic diagnoses was associated with lower OR for men (0.77; 0.65-0.91) but not women (0.98; 0.88-1.09). Full- or part-time DP had the strongest association with not being in paid work regardless of diagnosis group and sex (e.g., women mental DP 0.39; 0.34-0.45; women somatic DP 0.38; 0.35-0.41; men mental DP 0.36; 0.29-0.43; men somatic DP 0.35; 0.32-0.38).

Conclusions

SA due to mental diagnoses had a stronger association with not being in paid work after age 65 than SA due to somatic diagnoses. The results highlight the importance of identifying factors that hinder older workers with mental disorders to extend their working lives.

Key messages

Health and morbidity among those in paid work after age 64: a systematic review

Background

Despite the increase of labour market participation at older ages, very little is known about health and morbidity among those who remain in a paid work after age 64. The aim was to systematically review the scientific knowledge on health and morbidity among people aged above 64 years who are in paid work.

Methods

A systematic literature review of studies published in English in scientific journals in 2014-2020. We identified 18,972 unique publications, of which 66 were deemed relevant by at least two independent researchers. Quality judgements and data extraction were done by at least two independent researchers according to pre-specified templates.

Results

There was a great heterogeneity in the included studies regarding study design, included populations (both size and type), exposures, outcomes, covariates, measures, and analytical methods. Few were assessed as having high quality. Most studies (95%) were from OECD countries and results were about men to a greater extent than women. 42 of the 66 studies had results indicating that being in paid work after age 64 was associated with good health and less morbidity. Six studies presented at least one result showing the opposite; those in paid work had worse health than those not, while 21 studies presented at least one result showing that there were no health/morbidity differences between those in paid work and who were not. Only one study presented results regarding mortality. Many aspects had not been studied at all, or only in one or two studies.

Conclusions

Many studies had results indicating that those who were in paid work >64 had better health/less morbidity than those who were not, however, there was a great variety in the results. There are surprisingly few studies about health/morbidity among people in paid work after age 64, and those published are heterogeneous: it is thus not possible to draw conclusions regarding scientific evidence based on the currently existing studies.

Key messages

• Both study designs and results were very heterogenous in the 66 studies that presented results on health or morbidity among people in paid work after age 64.

• More and better studies are needed as well as greater clarity regarding study designs, populations, measures, analytical methods and definitions of central concepts such as work, health and morbidity.

Trends in associations between sickness absence, disability pension and paid work after age 65

Aim

The increasing workforce participation at higher ages may impact social insurance systems, however, this has hardly been studied at all. We investigated associations between sociodemographic factors, prior sickness absence, and prior disability pension, with being in paid work after age 65 and to what extent they then had any sickness absence, and if such associations changed over time.

Methods

We conducted three prospective cohort studies. We used longitudinal data from three nationwide registers linked at individual level covering all residents in Sweden who turned 65 years in 2000, 2005, or 2010 (N = 50,000, 68,000, and 99,000, respectively). Logistic regression was used to calculate odds ratios (OR) and 95% confidence intervals (CI) for the association between sickness absence and disability pension when aged 60-64 and a) being in paid work when aged 66-71, and b) having sickness absence when aged 66-71 among those in paid work, adjusted for sociodemographic variables in each of the three cohorts.

Results

Although the percentage of individuals in paid work when aged 66-71 increased substantially between the cohorts (2000 cohort 28.7%, 2005 cohort 39.5%, 2010 cohort 48.2%), associations of sociodemographic factors with being in paid work or having sickness absence when aged 66-71 remained steady. Both sickness absence and disability pension when aged 60-64 were negatively associated with working past 65 (sickness absence 2000 cohort OR 0.86, 95% CI 0.82-0.90; 2010 cohort OR 0.89, CI 0.87-0.92 - disability pension 2000 cohort OR 0.41, 0.39-0.44; 2010 cohort OR 0.36, CI 0.35-0.38). Sickness absence when aged 60-64 was positively associated with sickness absence after 65 (2000 cohort OR 2.16, CI 1.86-2.51; 2010 cohort OR 2.15, CI 1.98-2.33) while disability pension was negatively associated with sickness absence after 65 (2000 cohort OR 0.69, CI 0.52-0.92; 2010 cohort OR 0.83, CI 0.68-1.01).

Key messages

Rates of people in paid work >65 increased much from 2000 to 2010; their sickness absence increased marginally, while associations of sociodemographic factors with paid work did not change over time. Prior sickness absence was associated with paid work and sickness absence after age 65, while disability pension before age 65 was associated with less sickness absence after 65.

Job demands/control among people in paid work age 55-64 and their labour market status 11 years later

Background

As discussions about extending working lives are ongoing, more knowledge is warranted on working conditions that are associated with labour market status in older age.

Aim

Among employees aged 55-64 years, associations between job demands/job control with their labour market status 11 years later were explored.

Methods

A population-based prospective cohort study using a job exposure matrix (JEM) and nationwide register data. All 616,818 individuals in Sweden who in 2001 were in paid work and aged 55-64, were assigned JEM values for job demands/control that were categorized into tertiles resulting in nine combinations of job demands/control (reference=medium control/medium demands). Follow-up was in 2012 regarding labour market status (paid work, old-age pension, low/no income, sickness absence emigrated, dead). Odds ratios (OR) and 95% confidence intervals (CI) were calculated using multinomial logistic regression (reference category=old-age pension) with adjustment for educational level, age, birth country, family situation and type of living area.

Results

Women in occupations with high control (OR low demands/high control 1.21, CI 1.06-1.39, medium demands/high control 1.32, CI 1.20-1.46, high demands/high control 1.29, CI 1.20-1.37); and men in occupations with high demands/high control (OR 1.11, CI 1.02-1.21) were more likely to be in paid work. Those in occupations with high demands at baseline were less likely to have low/no income at follow-up (OR women high demands/medium control 0.51, CI 0.37-0.68, high demands/high control 0.68, CI 0.50-0.92; men high demands/medium control 0.55, CI 0.31-0.96, high demands/high control 0.47, CI 0.32-0.74.

Conclusions

High job control combined with any level of demands for women and high demands for men aged 55-64 in 2001 were associated with higher OR of having main income from paid work in 2012, and high job demands combined with any level of job control were associated with lower OR of low/no income.

Key messages

Combinations of job demands and job control among workers aged 55-64 years in 2001 were associated with labour market situations in 2012. Using a job exposure matrix divided into tertiles allows for greater detail in measuring high, medium, and low demands and control than methods used in previous studies.

Psychosocial working conditions and future sick leave and disability pension trajectories in Sweden

Background

Psychosocial working conditions such as job demands and job control have been found to be associated with employee health and well-being, but studies on the associations with sickness absence (SA) and disability pension (DP) are scarce. We examined 11-year SA/DP trajectories and the association between psychosocial working conditions and subsequent SA/DP trajectories in the Swedish workforce.

Methods

Using a prospective cohort study with microdata we explored SA/DP trajectories among female and male employees, respectively, aged 30-53 years in 2001 in Sweden (1,076,042 women; 1,102,721 men). Group-based trajectory analysis was used to model annual mean SA/DP net days trajectories in 2002-2012. Based on a Swedish Job Exposure Matrix (JEM), individuals were assigned an age-, sex- and occupation-specific mean score for demands and control, respectively. Mean scores were categorized into tertiles and categorised into 3x3 combinations of exposure categories. Using multinomial regression we predicted trajectory group memberships for the JEM.

Results

The highest rate of women were in occupations with low demands and control (24.8%), while the highest rate of men (22.9%) was in occupations with high demands and control. We found three SA/DP trajectories for women (low, medium, high increasing) and two for men (low, high increasing). In fully adjusted models, those in occupations with low demands and low control were at higher risk of belonging to the high increasing SA/DP trajectory compared to those in occupations with high job demands and control in both women (OR 3.86; 95% CI: 3.75-3.97) and men (OR 3.0; 2.99-3.16).

Conclusions

Low job demands and low job control were associated with more high increasing future SA/DP trajectories compared to high job demands and job control in both women and men.

Key messages

In Sweden, women are more often in occupations characterized by low job demands and low job control and men are more often in occupations with high job demands and high job control. Occupations with low job demands and low job control were associated with more adverse SA/SP trajectories compared to occupations with high job demands and high job control in both women and men.

Job demands & control among employees in Sweden aged 55-64 and labour market status 11 years later

Background

As discussions about extending working lives are ongoing, more knowledge is warranted on how psychosocial working conditions are associated with labour market status in older age.

Aim

Among employees aged 55-64 years, explore associations between job demands/control with their labour market status 11 years later, using a job exposure matrix (JEM).

Methods

A population-based prospective cohort study using nationwide register data. All 616,818 individuals in Sweden who in 2001 were in paid work and aged 55-64, were categorized using JEM into 9 groups, based on tertiles. They were followed up in 2012 regarding their labour market status (main income from: paid work, old-age pension, marginalised (no income/social assistance), sickness absence >183 net days, emigrated, dead) using multinomial logistic regression for odds ratios (OR) and 95% confidence intervals (CI), controlling for labour market status and sociodemographics in 2001. Analyses were stratified by sex.

Results

The majority (women: 84.9%, men: 80.3%) had main income from old-age pension at the 11-year follow-up; 4.7% from paid work (women: 3.9%, men: 5.6%). Those initially in jobs with high demands were less likely to be marginalised at follow-up (OR women high demands/medium control 0.51, CI 0.38-0.68, high demands/high control 0.68, CI 0.50-0.92; OR men high demands/medium control 0.55, CI 0.31-0.96, high demands/high control 0.47, CI 0.30-0.73). Those in occupations with low demands were less likely to be in paid work (OR women low demands/low control 0.56, CI 0.51-0.62, low demands/medium control 0.63, CI 0.58-0.69; OR men low demands/low control 0.56, CI 0.51-0.63, low demands/medium control 0.63, CI 0.58-0.69).

Conclusions

High job demands with high job control among people aged 55-64 in 2001 were associated with higher rate and OR of having main income from paid work in 2012, and high job demands regardless of level of job control were associated with less marginalisation.

Key messages

Levels of job demands when aged 55-64 were associated with labour market status 11 years later for women and men, while levels of job control were less so. High job demands were associated with a higher likelihood of being in paid work and lower likelihood of being marginalised at the end of follow-up.

Sick leave among people in paid work after age 65: A Swedish population-based study covering 1995, 2000, 2005 and 2010

Aims: Extending working life into older age groups is discussed in many countries. However, there is no knowledge about how this affects rates of sick leave. The aim of this work was to investigate rates of sick leave among people in paid work after retirement age and if such rates have changed over time. Methods: Swedish nationwide register data on people aged >65 years and living in Sweden in 1995, 2000, 2005 and 2010 were analysed. All people with a sufficiently high work income to be eligible for public sick leave benefits were included. The proportions in paid work and compensated rates of sick leave for people aged 66–70 and ≥71 were analysed by sex, educational level, country of birth, living area, and employment type and sector. Results: The percentage of people in paid work at ages 66–70 years increased from <10% in 1995 to 24% in 2010 and among those aged ≥71 years from 2.7% in 1995 to 3.5% in 2010. The rates of sick leave among working people aged 66–70 years were 3.3% in 1995 and 2.4% in 2010 and for people aged ≥71 years the rates of sick leave were 2.2% in 1995 and 0.2% in 2010. Women had higher rates of sick leave than men in 2005 and 2010, but lower in 1995 and 2000. In 2010, the rates of sick leave were similar between employees and the self-employed, and higher among employees in the public sector than among employees in the private sector. Conclusions: Rates of sick leave among workers aged >65 years were lower in 2010 than in 1995, despite much higher rates of labour market participation in 2010.