The Relationship between Diabetes and Retirement from Work: A Ten-Year Follow-Up in National Registers
Bryan Cleal* and Kjeld Poulsen
Steno Health Promotion Research, Denmark
*Corresponding author: Bryan Cleal, Steno Health Promotion Research, Gentofte, Denmark, E-mail: email@example.com
Int J Diabetes Clin Res, IJDCR-2-035, (Volume 2, Issue 4), Research Article; ISSN: 2377-3634
Received: June 26, 2015 | Accepted: July 18, 2015 | Published: July 22, 2015
Citation: Cleal B, Poulsen K (2015) The Relationship between Diabetes and Retirement from Work: A Ten-Year Follow-Up in National Registers. Int J Diabetes Clin Res 2:035. 10.23937/2377-3634/1410035
Copyright: © 2015 Cleal B, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Objective: To investigate the relationship between diabetes, mortality and exit from work using various Danish Pension schemes.
Methods: We linked the Danish National Diabetes register with socioeconomic and population-based registers and followed prevalent and incident cases of diabetes for ten years, adjusting for gender, age and socio-occupational group. We analysed the proportion of individuals with diagnosed diabetes before and after entering a normal retirement pen-sion, voluntary early retirement pension or disability pension. Risk ratios for entering pensions, emigration and mortal-ity were calculated using Poisson regression.
Results: Among 1,628,087 adults between the ages of 30 and 59, the relative risk for mortality was increased for those with incident and prevalent diabetes by 11-145%, compared to individuals without diabetes. Voluntary early retirement pension was the most common path out of the workforce, especially for individuals with diabetes. Compared to those without diabetes, the relative risk of disability retirement for individuals with prevalent and incident diabetes was 2.98 (95% CI, 2.89 - 3.07) and 2.40 (95% CI, 2.34 - 2.45), respectively. More than one fifth of participants were diagnosed with diabetes after retirement. There was a marked negative socio-occupational gradient for mortality, voluntary early retirement and disability pension.
Conclusion: Increasing the retirement age will increase the proportion of workers who develop diabetes while still in the workforce. Since the number and severity of complications are related to longer disease duration, it is likely that, under proposed retirement reforms, more employees will experience health problems managing their jobs, especially among the most vulnerable socioeconomic groups.
Actual and proposed increases to the retirement age in many Western countries, combined with the ongoing diabetes epidemic , raise the possibility that the prevalence of diabetes among the working population will increase consid-erably in the coming years. The most recent (2012) data from the Danish National Diabetes Register (DNDR)  indi-cate that the respective prevalence of diabetes for Danish females and males between 50 and 59 years of age is 6.7% and 8.5%; the respective prevalence of diabetes for females and males aged 60-69 years is 11.1% and 15.3%. Type 2 diabetes has generally been overlooked in the context of occupational safety and health, not least because it has tradi-tionally been perceived as primarily affecting people's health after retirement . Nevertheless, diabetes is influencing retirement choices today [4,5] and it is therefore expected to become more important as the retirement age increases. However, there is only scarce knowledge of how the incidence and prevalence of diabetes are distributed among differ-ent pension schemes.
In Denmark, pension benefits are obtained in three ways: normal retirement, voluntary early retirement and disability. All Danes are administratively assigned a normal retirement pension at the current qualifying age of 65 years, although some individuals may voluntarily continue to work. The Danish voluntary early retirement pension was implemented in 1979 and combines individual payment to an unemployment fund with a supplemental public welfare payment, cur-rently beginning at age 60. The original argument for voluntary early retirement was to incentivize employee attrition in crowded occupations and in general, to favour employment of younger people when jobs are scarce. To obtain this benefit, employees must be healthy and able to work full time until the first day of eligibility for a voluntary early pen-sion. A disability pension is granted when an employee's working capacity is permanently reduced to the extent that no work of any kind can be performed.
As part of a multi-partisan retirement agreement, the Danish Ministry of Finance estimated that normal retirement age will be 70.5 by the year 2045, compared with the current age of 65 . However, by 2022 both early voluntary and normal retirement ages will have already increased by two years. In addition, the duration of early voluntary pension gradually decreases from five to three years by 2023; beginning in 2045, this benefit will be available only when people are 67.5 years old.
Delayed retirement is directly related to diabetes because disease duration is correlated with the risk of complications affecting fitness for work. Among complications potentially impacting the ability of individuals with diabetes to con-tinue working are ischaemic heart disease [7,8], stroke  and increased mortality . Due to a negative social gradi-ent in diabetes incidence [11,12], employees working in the lowest socioeconomic occupations may be more affected by these complications.
The aim of this study is to investigate how diabetes is related to exit from work, particularly through three Danish pen-sion schemes, and to examine the proportion of diabetes that is diagnosed before and after entering each pension scheme.
Diabetes incidence is derived from the DNDR, which provides incident data from 1995 onwards. The DNDR comprises linked records from the National Patient Register, National Insurance Service Register and Register of Medical Product Statistics [13-15]. To be registered as having diabetes, individuals must have a record of one or more of six criteria . Patients diagnosed with diabetes before 1995 are registered as prevalent cases, but there is no information on disease duration for these individuals. The DNDR does not distinguish between type 1 and type2 diabetes.
Study population and data sources
Using a unique person identification number, we linked DNDR with the Central Person Register and the Employment Classification Module to collect information on gender, age, date of birth, death, voluntary early retirement or migration [16,17]. Socio-occupational status at register entry was coded according to Statistics Denmark's classification system . Because of the small number of people employed in the military and agricultural sectors, these occupations were excluded from the study population.
Selection of the study population is shown in figure 1. In all, 1,628,087 Danish adults between the ages of 30 and 59 years who were in the labour market on January 1, 2001 were followed for ten years or until the occurrence of one of five exit options: death, first emigration or entry into normal retirement, voluntary early retirement or disability retire-ment.
Figure 1: Diabetes and career path
*DISCO Danish International Standard Classification of Occupations
§Events are death, emigration or pension View Figure 1
We calculated the proportion of individuals in the each of the five exit categories. Ten-year incidence of diabetes is calculated for the transitions between continue in work, normal retirement, voluntary early retirement and disability retirement. The population was stratified on sex and age (Figure 2). Poisson regression was used to estimate relative risk (RR) with 95% confidence limits for each of the five exit options during the ten-year follow-up period. Proc Genmod in SAS version 9.3 (SAS Institute; Cary NC, USA) was used to implement the analysis, stratified on gender, age group, diabetes status and socio-occupational class. Females, workers aged 30-49, people without diabetes and professionals were used as reference groups.
Figure 2: Labour market transitions among the male and female Danish population aged 30-49 and 50-59: Work; Incident Diabetes; Normal retirement; Voluntary early retirement; Disability retirement; Emigration; and Death View Figure 2
Table 1 shows the unadjusted distributions of ten-year incidence for each of the five exit categories stratified on diabe-tes status, sex and age. Mortality was higher among workers with incident diabetes, compared to people without diabe-tes, but only for younger members of the study population. If diabetes was present at entry into the study, then all groups had increased ten-year mortality. In contrast, emigration was lower for people with prevalent or incident diabe-tes. The normal retirement pension option was more common for males; females more often used the voluntary early retirement option, and the rate of early retirement was even higher for women with diabetes. Although the disability pension was used least frequently, the proportion of the incident population receiving a disability pension was 1.8-3.0 times greater than the comparable proportion of the population without diabetes. Similarly, the proportion of the prevalent population receiving a disability pension was 2.5-3.8 times greater than the comparable proportion of the population without diabetes.
Table 1: Relationship between diabetes status and work status after ten years View Table 1
To explore if diabetes was diagnosed before or after people entered a pension scheme, we calculated the proportion for each option (Figure 2). In the younger age group, it is only possible to receive a disability pension, which was more common for females (4.5% vs. 3.5% for males). Of females aged 30-49, 2.3% developed diabetes and 8.2% of these (0.2% of all females in this age group) subsequently entered a disability pension. Similarly, 0.2% of females in this age group with a disability pension subsequently developed diabetes. The total rate of ten-year diabetes prevalence among workers aged 30-49 was 3.7% for males and 2.7% for females.
Among study participants aged 50-59, exit from the workforce most frequently occurred through voluntary early re-tirement, which was chosen by 55.8% of females and 42.9% of males. The majority of workers aged 50-59 who devel-oped diabetes either remained in the workforce, exited the workforce via a normal retirement pension, or left through voluntary early retirement. The total ten-year diabetes prevalence for men and women aged 50-59 was 9.9% and 6.3%, respectively.
People with prevalent diabetes had the highest relative risk of mortality (RR 2.45; 95% confidence interval (CI), (2.30-2.60) and of receiving a disability pension (RR 2.98; 95% CI, 2.89 - 3.07), although incident diabetes was also associ-ated with increased retirement risk (Table 2). Compared to women, men were less likely to receive a disability pension (RR 0.71: 95% CI, 0.70 - 0.73) or voluntary early retirement pension (RR 0.77; 95% CI, 0.77 - 0.78) and more likely to die (RR 1.61; 95% CI, 1.57 - 1.66) and emigrate (RR 1.66: 95% CI, 1.62 - 1.70). After adjusting for sex, age, and diabetes, a clear negative socio-occupational gradient existed for mortality, voluntary early retirement and disability pension. Members of the highest socio-occupational groups, e.g., managers and professionals, remained in the labour market until reaching normal retirement age. The relative risk for emigration was highest for men, lower for all socio-occupational groups compared to professionals, and lower for those with diabetes and aged 50-59.
Table 2: Relative risk (RR) for labour market exit outcomes during ten-year follow up View Table 2
Our findings demonstrate that people with diabetes more frequently entered both voluntary early retirement and disability pensions than did people without diabetes. People with diabetes also had an increased relative risk of mortality compared to those without diabetes. The reason for the relative risk of normal retirement to be almost equal between incident and prevalent diabetes, might be explained by the much higher risk of dying for prevalent cases. An-other part of the explanation is the inverse directions of the socio-occupational gradients. In addition, the risk of entering both voluntary early retirement and a disability pension was highest for females and for those with the lowest socio-occupational positions.
Diabetes is not an "after retirement" disease
Our results also indicate that the greatest proportion of diabetes was diagnosed before normal retirement. Most people with type 2 diabetes are expected to live for many years without major problems after being diagnosed, and diabetes is also regarded as the condition with the least influence on workability compared to musculoskeletal, severe headaches, circulatory, respiratory, digestive and psychological chronic health problems, investigated in the Netherlands , which has a retirement age comparable to that of Denmark . However, in general, people with diabetes have in-creased absenteeism , especially around the time of diagnosis , and other authors have found that they retire earlier than do people without diabetes . Furthermore, people with diabetes report disabilities three times more often  than do people without diabetes; psychological and emotional problems like distress and depression are also asso-ciated with diabetes [25,26].
Unfortunately, even though previous reports may lack details about diabetes type, disease severity and complications, they are often generalised to all employees with diabetes, when a much more nuanced understanding is needed. To the best of our knowledge, complications with the most significant influence on attachment to the labour market are cardiovascular disease, renal disease, amputations , neuropathy and impaired vision . Although many complications are silent for years and may not be diagnosed until disease is advanced , they are nevertheless an important factor in remaining fit for work . Complications have the highest impact on the individual risk of job loss due to diabetes  and on the economic costs of diabetes . The implications of the new pension reforms being implemented in most OECD countries  will therefore need to be considered carefully.
The timing of diabetes diagnoses and entering pensions
The majority of participants who developed diabetes were diagnosed before entering one of the three pension schemes. However, 27% of females and 20% of males aged 50-59 were diagnosed after retirement; comparable figures among workers aged 30-49 were 7% and 6%. Since more than one fifth of the diabetes cases were diagnosed after retirement, substantially increasing retirement age over a relatively short period will increase the number of people with a longer duration of disease who is active in the workforce.
Voluntary early retirement has been very popular in Denmark. The Danish Welfare Commission asked people receiving a voluntary early retirement pension about their reasons for withdrawing from the workforce. One fifth of respondents reported they gave up work because of poor health, and one tenth reported being afraid of developing poor health if they continued working . However, it is unknown how much diabetes, in particular, influenced their decisions. We also found an elevated risk of receiving a disability pension among people with both incident and prevalent diabetes. However, diabetes is seldom the identified cause of disability retirement; common medical reasons are psychological or musculoskeletal in origin. Most of the senior workforce already has at least one chronic condition by approximately 60 years of age , and the additional development of diabetes may trigger entering a disability pension. However, we expect that, with impending retirement reforms requiring workers to remain in the workforce longer, workability will be affected for many more employees with diabetes.
Inequity factors in diabetes and work life
How would the diabetes epidemic influence a general pension scheme that does not differentiate according to socio-demographic factors? As shown in table 2, a significant difference existed in the patterns with which males and females applied to the three pension schemes. For instance, since females currently leave work earlier through voluntary early retirement, they will need to remain in the workforce even longer than will their male colleagues . We can expect a relative increase in workers with diabetes in jobs that traditionally employ women. Because the incidence of diabetes peaks later in women than in men , the largest increase will happen within the extended timeframe of the reform.
In addition, a significant negative socio-occupational gradient existed with respect to the options for exiting the work-force, even after adjusting for sex, age and diabetes. Diabetes is also distributed along a similar negative socio-occupational gradient [11,12,37,38], compounding the issue and highlighting the importance of effective strategies to address obesity, a major cause of diabetes  that is of great concern in occupational health .
Strengths and weaknesses
All data came from administrative registers covering the entire population, minimising selection bias . On the other hand, a diagnostic bias likely occurred because type 2 diabetes may be undiagnosed in a significant number of adults in the workforce. Similarly, diagnostic delays are common; many individuals have had diabetes for some time, perhaps several years, before they are diagnosed. However, the DNDR has been validated as a reliable source of information on diabetes in Denmark for epidemiological purposes .
A limitation of our study is that we cannot distinguish between type 1 and type 2 diabetes. However, since only roughly one in ten cases of diagnoses diabetes is of type 1, most of which are diagnosed during adolescence or in early adult-hood, we excluded people under the age of 30 years to minimise bias arising from this limitation. Another potential limitation is that legislation changed slightly during the ten-year follow up because of a global economic crisis. How-ever, we do not expect that this would substantially influence our conclusions. Finally, although it is possible that some members of the study population changed socio-occupational groups, we do not believe that this significantly affected our results , especially as it is a large register based survey following up for only ten years.
Implications for the labour market
Our results fuel the argument that the proportion of workers who will develop diabetes while still at work will increase significantly if demographic trends and diabetes incidence continue their current trajectories over the next three dec-ades. Furthermore, many more people will also have had diabetes of sufficiently long duration to be at increased risk of developing complications before retirement.
With regard to diabetes, applying a universal retirement scheme for all socio-occupational groups may ultimately be neither fair nor practical, since the lowest socio-occupational groups have a higher mortality rate, more frequently use disability benefits, and, as noted earlier, acquire diabetes more frequently. In addition to its direct effects, diabetes also amplifies the effects of many other physical and psychological conditions, forcing people out of work earlier than soci-ety requires.
As diabetes is predicted to continue as an epidemic, the extended retirement age will result in more people in the work-force having diabetes of longer duration. By 2050 most of the OECD-member countries will raise the retirement age to at least 67 years . Most of these countries have similar systems as the one we have described from Denmark and how to construct our future pension schemes is important to get on the international agenda. A fair approach will also need to take inequalities in health into consideration. In the future, most cases of diabetes will be diagnosed while peo-ple are still active in the labour market. A negative social gradient exists in which, compared to professionals, elemen-tary workers have more diabetes and a higher relative risk of mortality and more often use disability pensions and voluntary early retirement. The exact role of diabetes in relation to work fitness and retirement in the future labour market needs to be further investigated.
Wild S, Roglic G, Green A, Sicree R, King H (2004) Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 27: 1047-1053.
Carstensen B, Kristensen JK, Marcussen MM, Borch-Johnsen K (2011) The National Diabetes Register. Scand J Public Health 39: 58-61.
Shamshirgaran SM, Jorm L, Bambrick H, Hennessy A (2013) Independent roles of country of birth and socioeconomic status in the occurrence of type 2 diabetes. BMC Public Health 13: 1223.
Herquelot E, Gueguen A, Bonenfant S, Dray-Spira R (2011) Impact of diabetes on work cessation: data from the GAZEL cohort study. Diabetes Care 34: 1344-1349.
Moren-Hybbinette I, Moritz U, Schersten B (1989) Diabetes mellitus and disability pension. A descriptive study of all diabetic subjects granted disability pension in Sweden in 1980. Scand J Soc Med 17: 193-201.
Aftale om senere tilbagetrækning (2011) Agreement about extended withdrawal.
Natarajan S, Liao Y, Sinha D, Cao G, McGee DL, et al. (2005) Sex differences in the effect of diabetes duration on coronary heart disease mortality. Arch Intern Med 165: 430-435.
Almdal T, Scharling H, Jensen JS, Vestergaard H (2004) The independent effect of type 2 diabetes mellitus on ischemic heart disease, stroke, and death: a population-based study of 13,000 men and women with 20 years of follow-up. Arch Intern Med 164: 1422-1426.
Banerjee C, Moon YP, Paik MC, Rundek T, Mora-McLaughlin C, et al. (2012) Duration of diabetes and risk of ischemic stroke: the Northern Manhattan Study. Stroke 43: 1212-1217.
Brun E, Nelson RG, Bennett PH, Imperatore G, Zoppini G, et al. (2000) Diabetes duration and cause-specific mortality in the Verona Diabetes Study. Diabetes Care 23: 1119-1123.
Poulsen K, Cleal B, Willaing I (2014) Diabetes and work: 12-year national follow-up study of the association of diabetes incidence with socioeconomic group, age, gender and country of origin. Scand J Public Health 42: 728-733.
Agardh E, Allebeck P, Hallqvist J, Moradi T, Sidorchuk A (2011) Type 2 diabetes incidence and socio-economic position: a systematic review and meta-analysis. Int J Epidemiol 40: 804-818.
Andersen JS, Olivarius Nde F, Krasnik A (2011) The Danish National Health Service Register. Scand J Public Health 39: 34-37.
Kildemoes HW, Sorensen HT, Hallas J (2011) The Danish National Prescription Registry. Scand J Public Health 39: 38-41.
Lynge E, Sandegaard JL, Rebolj M (2011) The Danish National Patient Register. Scand J Public Health 39: 30-33.
Petersson F, Baadsgaard M, Thygesen LC (2011) Danish registers on personal labour market affiliation. Scand J Public Health 39: 95-98.
Sortso C, Thygesen LC, Bronnum-Hansen H (2011) Introduction and overview to Danish registers. Scand J Public Health 39:17-209.
DS Statistics Denmark (1997) SOCIO Danmarks Statistiks socioøkonomiske klassifikation: 1-18.
Leijten FR, Van den Heuvel SG, Ybema JF, van der Beek AJ, Robroek SJ, et al. (2014) The influence of chronic health problems on work ability and productivity at work: a longitudinal study among older employees. Scand J Work Environ Health 40: 473-482.
OECD (2013) Pensions at a glance 2013: OECD and G20 indicators: 1-364.
De Backer G, Leynen F, De Bacquer D, Clays E, Moreau M, et al. (2006) Diabetes mellitus in middle-aged people is associated with increased sick leave: the BELSTRESS study. Int J Occup Environ Health 12: 28-34.
Dray-Spira R, Herquelot E, Bonenfant S, Guéguen A, Melchior M (2013) Impact of diabetes mellitus onset on sickness absence from work--a 15-year follow-up of the GAZEL Occupational Cohort Study. Diabet Med 30: 549-556.
Rumball-Smith J, Barthold D, Nandi A, Heymann J (2014) Diabetes associated with early labor-force exit: a comparison of sixteen high-income countries. Health Aff (Millwood) 33: 110-115.
Mayfield JA, Deb P, Whitecotton L (1999) Work disability and diabetes. Diabetes Care 22: 1105-1109.
Gois C, Akiskal H, Akiskal K, Figueira ML (2012) Depressive temperament, distress, psychological adjustment and depressive symptoms in type 2 diabetes. J Affect Disord 143: 1-4.
Fisher L, Skaff MM, Mullan JT, Arean P, Mohr D, et al. (2007) Clinical depression versus distress among patients with type 2 diabetes: not just a question of semantics. Diabetes Care 30: 542-548.
Sorensen J, Ploug UJ (2013) The costs of diabetes-related complications: registry-based analysis of days abscent from work. Economics Research International: 1-8.
Fu AZ, Qiu Y, Radican L, Wells BJ (2009) Health care and productivity costs associated with diabetic patients with macrovascular comorbid conditions. Diabetes Care 32: 2187-2192.
Kanavos P, Aardweg Svd, Schurer W (2012) Diabetes expenditure, burden of disease and management in 5 EUcountries: 1-113.
Waclawski ER, Gill G (2013) Diabetes mellitus and other endocrine disorders. In: Palmer KT, Brown I, Hobson J. Fitness for work - The medical aspects. Oxford: Oxford University Press.
Kraut A, Walld R, Tate R, Mustard C (2001) Impact of diabetes on employment and income in Manitoba, Canada. Diabetes Care 24: 64-68.
Dall T, Mann SE, Zhang Y, Martin J, Chen Y, et al. (2008) Economic costs of diabetes in the U.S. in 2007. Diabetes Care 31: 596-615.
Velfærdskommissionen Kapitel 7 (2006) Senere tilbagetrækning fra arbejdsmarkedet. Extended withdrawal from the labour market: 315-96.
Virtanen M, Oksanen T, Batty GD, Ala-Mursula L, Salo P, et al. (2014) Extending employment beyond the pensionable age: a cohort study of the influence of chronic diseases, health risk factors, and working conditions. PLoS One 9: e88695.
Agardh EE, Ahlbom A, Andersson T, Efendic S, Grill V, et al. (2007) Socio-economic position at three points in life in association with type 2 diabetes and impaired glucose tolerance in middle-aged Swedish men and women. Int J Epidemiol 36: 84-92.
Carstensen B, Kristensen JK, Ottosen P, Borch-Johnsen K; Steering Group of the National Diabetes Register (2008) The Danish National Diabetes Register: trends in incidence, prevalence and mortality. Diabetologia 51: 2187-2196.
Espelt A, Borrell C, Roskam AJ, Rodríguez-Sanz M, Stirbu I, et al. (2008) Socioeconomic inequalities in diabetes mellitus across Europe at the beginning of the 21st century. Diabetologia 51: 1971-1979.
Cleal B, Hannerz H, Poulsen K, Andersen LL (2014) Socio-economic status and incident diabetes mellitus among employees in Denmark: a prospective analysis with 10-year follow-up. Diabet Med 31: 1559-1562.
Waly MI, Essa MM, Ali A, Al-Shuaibi YM, Al-Farsi YM (2010) The global burden of type 2 diabetes: a review. Int J Biol Med Res 1: 326-9.
Bonde JP, Viikari-Juntura E (2013) The obesity epidemic in the occupational health context. Scand J Work Environ Health 39: 217-219.
Thygesen LC, Ersboll AK (2014) When the entire population is the sample: strengths and limitations in register-based epidemiology. Eur J Epidemiol 29: 551-558.
Hansen MB, Jensen ML, Carstensen B (2012) Causes of death among diabetic patients in Denmark. Diabetologia 55: 294-302.
Van Raalte AA, Martikainen P, Myrskyla M (2014) Lifespan variation by occupational class: compression or stagnation over time? Demography 51: 73-95.