HIV infection primarily affects individuals of working age (http://www.who.int/hiv/data/en/index.html). The introduction of HAART has led to life expectancies among HIV-infected individuals comparable to that of the background population . Access to HAART has in some studies from low-income countries been reported to have an immense impact on employment rates among HIV-infected individuals [2,3]. In some high-income countries, employment rates among HIV-infected individuals have been reported to be as low as 42–62% [4–7], and the risk of employment loss readily following an HIV diagnosis seems to be high . Risk factors of employment loss following an HIV diagnosis have been identified as comorbidities such as diabetes and hypertension, female sex and low socioeconomic status [6,8]. Even in countries with universal healthcare systems, socioeconomic status including employment status has been shown to be an independent prognostic risk factor [9–11].
In a Danish setting wherein HIV-related healthcare including medical treatment is provided free of charge, we aimed to estimate annual employment and disability retirement rates in the years following an HIV diagnosis among HIV-infected individuals compared with the background population.
Materials and methods
The study was performed as a cohort study including an HIV-infected population and a background population cohort matched on age, sex and country of origin.
We used the unique 10-digit civil registration number assigned to all Danish residents to link the following registries .
The Danish HIV cohort study
The Danish HIV cohort study (DHCS) is a population-based nationwide cohort study of all HIV-infected Danish residents receiving care at Danish HIV centres since 1 January 1995. The cohort includes prevalent cases with HIV diagnosis prior to 1 January 1995 and incident cases included from 1 January 1995. Individuals are consecutively enrolled. Data include demographics, date of HIV diagnosis, route of transmission, AIDS-defining events and antiretroviral treatment among other variables. CD4+ cell counts and HIV RNA measurements are extracted electronically from laboratory data files. Data are updated annually . The characteristics of the complete population in the DHCS have been described previously . In Denmark, the estimated HIV prevalence is approximately 0.09%.
The attainment register
The Attainment Register was established in 1981 and contains data on educational attainments on all Danish residents. Data are collected from educational institutions and the Danish Ministry of Education, and is administered by the government agency Statistics Denmark. Data are updated annually on 1 October .
Integrated database for labour market research
The Integrated Database for Labour Market Research was established in 1980 and contains data on occupational status, welfare services, social benefits and primary source of personal income on all Danish residents. Data are collected from the Danish tax authorities, educational institutions and employment services, and are administered by the government agency Statistics Denmark. Data are updated annually on 31 December .
Civil registration system
Civil registration system (CRS) contains data on vital status, country of origin and migration for all Danish residents. Data are updated continuously .
We included all HIV-1 infected individuals from the DHCS who were born in Denmark, between 18 and 60 years of age at study inclusion and were living in Denmark at date of study inclusion. Analyses were restricted to individuals younger than 60 years of age at study inclusion, as the general retirement age in Denmark was 65 years during the study period but with the possibility of early retirement from 60 years. The Danish welfare system including disability pension granting is described in details elsewhere . We excluded individuals who died or emigrated within the first year from date of study inclusion, reported IDU as a route of HIV transmission or were hepatitis C virus (HCV) coinfected. IDU and hepatitis C virus (HCV)-coinfected patients were excluded, as these subpopulations differ from the general HIV-infected population with respect to all outcome variables [17,18], and socioeconomic status among Danish HCV-infected patients has been described in other studies . Study inclusion was date of HIV diagnosis or 1 January 1996, whichever came last. The study period was 1 January 1996 until 31 December 2011.
Population controls were identified from CRS and extracted at random to constitute a background population comparison cohort. Eight population controls, individually matched on age, and sex, were identified per HIV-infected individual. All population controls were born in Denmark, were alive and living in Denmark at 1 year after date of study inclusion, and were assigned the same date of study inclusion as their corresponding HIV-infected individual.
Description of covariates
At study inclusion, we computed the following variables: Employment status prior to study inclusion was categorized as employed, if the individual was categorized as employed 1 and 2 years prior to study inclusion or unemployed, and if the individual was categorized as unemployed 1 and 2 years prior to study inclusion. Individuals with divergent employment status 1 and 2 years prior to study inclusion were not included in the analyses stratified by pre-HIV employment status. We included employment status 1 and 2 years prior to study inclusion in the definitions to minimize potential impact of HIV infection on employment before diagnosis.
Educational attainment was grouped into three categories: low educational attainment covering mandatory education of up to 9 years, medium educational attainment covering mandatory and vocational education of 9–12 years and high educational attainment covering formal education of more than 12 years .
Late presentation was defined as CD4+ cell counts less than 350 cells/μl. Very late presentation was defined as CD4+ cell count less than 200 cells/μl at HIV diagnosis and/or an AIDS-defining disease at/or within 14 days of HIV diagnosis regardless of CD4+ cell count .
Before, at and after study inclusion, study participants were categorized as employed in a calendar year if he or she was registered as an employee, a self-employed person or under training for a minimum of 6 months. A person was categorized as a disability pensioner in a calendar year if he or she received disability pension for a minimum of 6 months within that specific year.
Employment rates including 95% confidence intervals (CIs) were calculated on an annual basis as number of individuals employed in a specific calendar year divided by the total number of individuals under observation in the same year. Employment rates were calculated from year of study inclusion until 2011 or the year prior to the calendar year of death or emigration. Year of study inclusion was accounted as year 0, and subsequent included years were accounted values relative to year 0.
Disability retirement rates including 95% CI were calculated on an annual basis as number of individuals registered as disability pensioners in a specific year divided by the total number of individuals under observation in the same year. Disability retirement rates were calculated from the year of study inclusion until 2011 or the year preceding the year of death or emigration.
On the basis of year of study inclusion, HIV-infected individuals and population controls were grouped in pre-1996, 1996–1999 and 2000–2011 cohorts, and analyses were stratified according to these cohorts.
In the stratified analyses on employment rates and disability retirement rates (stratified by sex, educational attainment and employment status 2 years prior to study inclusion, route of transmission and by late/very late presentation) HIV-infected individuals and population controls included 1996–2011 were merged into one cohort.
The study was approved by the Danish data protection agency. Data analysis was performed using SPSS statistical software, version 19.0 (Norusis; SPSS Inc., Chicago, Illinois, USA).
We identified 2799 HIV-infected individuals, who fulfilled the inclusion criteria and were followed for a total of 27 343 person-years (PYR) (Table 1). Among the HIV-infected individuals, 412 (14.7%) died, 71 (2.5%) emigrated and two (0.1%) were registered as lost to follow-up. The comparison cohort included 22 369 individuals who were followed for a total of 237 137 PYR. Among the population controls, 1087 (4.9%) died, 312 (1.4%) emigrated and 11 (0.1%) were recorded as lost to follow-up (Table 1). One hundred and two (9%), 24 (5%) and 39 (3%) HIV-infected individuals were excluded from the pre-1996, 1996–1999 and 2000–2011 cohorts, respectively, because they died within the first year from study inclusion. In the pre-1996 cohort, fewer individuals received HAART within the first year after study inclusion and were more likely to present with AIDS prior to study inclusion. The 2000–2011 cohort had higher CD4+ cell count at study inclusion and fewer late presenters than the 1996–1999 cohort (Table 1).
Annual employment rates for the three cohorts of HIV-infected individuals and population controls are illustrated in Fig. 1. Compared with the population controls, the employment rates among the HIV-infected individuals were lower at study inclusion: 54.8 (95% CI 50.5–59.6) vs. 86.3% (95% CI 84.3–88.3), 74.6 (95% CI 66.9–83.2) vs. 85.9% (95% CI 82.8–89.0) and 77.4 (95% CI 72.8–82.2) vs. 87.2% (95% CI 85.5–89.0) for the pre-1996, 1996–1999 or 2000–2011 cohorts, respectively. Five years after study inclusion, employment rates were 56.1 (95% CI 51.4–61.1) vs. 85.6% (95% CI 83.6–87.6), 66.2 (95% CI 58.7–74.6) vs. 82.5% (95% CI 79.5–85.7) and 70.9 (95% CI 65.0–77.3) vs. 84.9% (95% CI 82.6–87.3) for HIV-infected individuals and population controls in the pre-1996, 1996–1999 or 2000–2011 cohorts, respectively. HIV-infected individuals in the pre-1996 cohort experienced a sharp drop in employment rates in the years preceding study inclusion, but this trend was not observed among HIV-infected individuals in the other cohorts (Fig. 1). Among HIV-infected individuals in the 1996–1999 and 2000–2011 cohorts, the employment rates dropped sharply in the first 2 years following HIV diagnosis, and thereafter the decrease in employment rates almost paralleled that of the population controls (Fig. 1).
Employment rates stratified by preinclusion employment status for HIV-infected individuals and population controls in the pre-1996, 1996–1999 and 2000–2011 cohorts are illustrated in Fig. 1. For the pre-1996, the 1996–1999 and the 2000–2011 cohorts, employment rates among HIV-infected individuals who were employed prior to study inclusion dropped sharply within the first 2 years following an HIV diagnosis compared with population controls. The increase in employment rates among individuals who were unemployed prior to study inclusion were higher among HIV-infected individuals than population controls in the 1996–1999 cohort, but for individuals in the pre-1996 and 2000–2011 cohorts, the increase in employment rates was almost the same for HIV-infected individuals and population controls (Fig. 1).
Employment rates at year 0 and 5 years from study inclusion for the 1996–2011 cohorts are summarized in Table 2. Compared with the population controls, employment rates at time of study inclusion were lower among HIV-infected individuals in all strata as defined by sex, educational attainment or employment status prior to study inclusion. Employment rates remained lower 5 years from study inclusion except for HIV-infected individuals who were unemployed prior to study inclusion whose employment rates at 5 years from study inclusion had increased more than among corresponding population controls (Table 2). HIV-infected individuals with late or very late presentation of HIV infection, of male sex or with medium educational attainment experienced the most pronounced declines in employment rates from study inclusion until 5 years after inclusion (Table 2). Further, employment rates among MSM were substantially higher at study inclusion than among heterosexually infected individuals, but MSM experienced a more pronounced decline in employment rates 5 years following an HIV diagnosis than heterosexually infected individuals (Table 2).
Disability retirement rates for the HIV-infected individuals and population controls in the pre-1996, 1996–1999 and 2000–2011 cohorts are illustrated in Fig. 2. Among HIV-infected individuals in the pre-1996 and 1996–1999 cohorts, disability retirement rates increased substantially in the first 2 years following study inclusion. A similar increase was not observed among HIV-infected individuals in the 2000–2011 cohort who had a more gradual increase in disability retirement rates in the years following study inclusion. Disability retirement rates 5 years from study inclusion dropped from 32.3% (95% CI 29.0–36.3) for HIV-infected individuals in the pre-1996 cohort, to 17.8% (95% CI 14.1–22.4) for HIV-infected individuals in the 1996–1999 cohort, and to 11.6% (95% CI 9.4–14.4) for HIV-infected individuals in the 2000–2011 cohort. For the corresponding population controls, the disability retirement rates were 5.8% (95% CI 5.3–6.3) in the pre-1996 cohort, 7.2% (95% CI 6.3–8.2) in the 1996–1999 cohort and 5.1% (95% CI 4.6–5.8) in 2000–2011 cohort.
Disability retirement rates stratified by sex, route of infection, educational attainment and preinclusion employment status at year 0 and 5 years from study inclusion are summarized for the 1996–2011 cohorts in Table 3. HIV-infected individuals, who were heterosexually infected, were men, had low or medium educational attainment, or with late or very late presentation had a larger increase in disability retirement rates 5 years from study inclusion compared with the rest of the HIV-infected population. HIV-infected individuals who were men or who had medium educational attainment had the highest increase in disability retirement rates compared with population controls.
Our study demonstrates that during the period 1996–2011 in which HAART became generally available, employment rates among HIV-infected individuals have increased, but have remained lower than in the background population. Although the proportion of HIV-infected individuals who received disability pension declined substantially in the study period, the disability retirement rates were still higher among HIV-infected individuals than among population controls.
The decline in employment rates at time of and during the first years following an HIV diagnosis was not followed by a later increase in employment rates. Further, the decline in employment rates was to a large extend explained by a parallel increase in disability retirement rates, which indicates that initial loss of employment among HIV-infected individuals has long-term consequences, and that employment was not regained despite general access to HAART.
The increase in employment status among HIV-infected individuals in the period 1996–2011 may have several explanations. First, the general availability of HAART has led to less morbidity in the HIV population. Second, the HIV testing strategy has improved in the period why the proportion of late presenters has decreased (Table 1). Third, the new treatment options have transformed the perception of the HIV disease from an irreversibly to a reversible condition and thereby induced improved rehabilitation programmes for the HIV population.
HIV-infected individuals in the pre-1996 cohort were not included at time of HIV diagnosis but at 1 January 1996, and thereby constitute a prevalent cohort. In Denmark, HAART became generally available in 1996–1997 why individuals in the pre-1996 cohort have had very limited access to HAART prior to study inclusion. These individuals have therefore had an increased risk of HIV-related morbidity, which probably explains the decrease in employment rates and increase in disability pension rates seen in this cohort prior to study inclusion (Figs. 1 and 2). In line with the inclusion criteria, individuals in the pre-1996 cohort had to be alive at 1 January 1997(1 year from inclusion) and thereby had access to HAART in the study period. This could to some extent explain that employment rates and disability retirement rates have stabilized for the pre-1996 cohort after study inclusion. However, taking the prestudy inclusion loss of employment into account, the pre-1996 cohort has the greatest loss of employment over time. Furthermore, in Denmark, disability pension was until 1992 granted on the basis of an HIV diagnosis even without complications. In 1992, the criteria for granting disability retirement to individuals with an HIV diagnosis was changed to include an individual medical assessment of the actual loss of workability related to HIV/AIDS . Disability pension is granted on a lifelong basis in Denmark, which may partly explain the absence of regained affiliation with the labour market within the pre-1996 HIV-infected cohort.
A more gradual decline in employment rates following an HIV diagnosis was observed among HIV-infected individuals in the 1996–1999 and 2000–2011 cohorts. HIV infection seems to have a direct impact on employment status, as late or very late presentation of HIV infection was associated with lower employment rates and increased disability retirement rates. However, as employment rates and disability retirement rates diverged when stratified by educational attainment or sex, our findings might also suggest a more complex association between employment and HIV infection, which is unaffected by HAART. Schuring et al. found that among 2754 unemployed persons, conditions such as general health, physical functioning, mental health and social function were associated with the risk of remaining unemployed. In addition, Hultin et al.  found that within the Stockholm Public Health Cohort including 13 027 individuals, long-term sick leave increased the risk of disability pension fourfold regardless of health status.
Dray-Spira et al. found that the risk of employment loss among HIV-infected individuals increased with comorbidities such as diabetes, hypertension and depression, but not with characteristics of the HIV disease itself. HIV-infected individuals have a high incidence of non-AIDS comorbidity such as cardiovascular diseases, cancer, kidney diseases, liver diseases, osteopenia/osteoporosis and neurocognitive diseases , and some of these comorbidities may have been detected even before HIV diagnosis . In accordance with these findings, the increased disability retirement rates among HIV-infected individuals in the 1996–1999 and 2000–2011 cohorts could reflect an increased prevalence of non-AIDS comorbidity.
Dray-Spira et al. further found that female sex, low occupational position and a nonpermanent affiliation with the labour market were associated with increased employment loss among HIV-infected individuals. However, in our study, HIV-infected men in contrast to HIV-infected women experienced a sharper decline in employment rates following HIV diagnosis than their corresponding population controls. In addition, individuals with medium educational attainment experienced a sharper decline in employment rates than individuals with low educational attainment. These findings could partly be explained by the fact that men with high or medium educational attainment under 50 years of age have the highest employment rates in the Danish population . Furthermore, we found that HIV-infected women and individuals with nonpermanent affiliation with the labour market, who had low employment rates at time of HIV diagnosis, seemed to regain employment following an HIV diagnosis compared with the population controls. These findings could reflect that some of the HIV-infected individuals have been unaware of their HIV infection, and that unattended HIV-related health conditions may have influenced affiliation with the labour market, emphasizing that early HIV diagnosis and timely initiation of HAART might improve workability among these individuals. Further, linkage to HIV care might have resulted in social initiatives, which can provide the basis for a more permanent affiliation with the labour market.
To our knowledge, this is the first nationwide study to estimate annual employment rates at time of and following an HIV diagnosis. The study includes a comparison cohort from the background population matched on age, sex and country of origin, which allows for estimation of relative employment and disability retirement rates. A major strength of this study is the high quality and coverage of the Danish registries, the population-based design with long and nearly complete follow-up, and access to valid data on employment status for all HIV-infected individuals and population controls. Data for the conditions that lead to granting of disability retirement were not included in the study, as we did not have access to these data.
Only individuals of Danish origin were included to minimize confounding from differences in migrations. However, the HIV-infected population in Denmark is multiform with respect to ethnicity , and the disparity in employment status identified in this study may be different in a multiethnic setting. In addition, we excluded HIV-infected individuals coinfected with HCV or who were infected with HIV through IDU, as these groups have been demonstrated to suffer from substantially increased mortality and morbidity [13,17,18,27,28]. Further, HCV coinfection and IDU among HIV-infected individuals are associated with an increased risk of employment loss following an HIV diagnosis , so to minimize confounding, we excluded HIV-infected IDUs or HCV coinfected from the study population.
Our study demonstrates that during 1996–2011, employment rates increased profoundly among HIV-infected individuals, but remained lower than in background population. As our results indicate that loss of employment among HIV-infected individuals following an HIV diagnosis might be permanent, prevention of employment loss at time of HIV diagnosis could result in significant individual and socioeconomic benefits.
Conflicts of interest
N.O. has received research funding from Bristol-Myers Squibb, GlaxoSmithKline, Boehringer Ingelheim and Gilead. C.P. has received research funding from Gilead. J.G. has received research funding from Abbott, Roche, Bristol-Myers Squibb, Merck Sharp & Dohme, ViiV, Swedish Orphan and Gilead. R.L., G.K., C.S.L., L.H.O. and U.B.D. report no conflicts of interest.
1. Obel N, Omland LH, Kronborg G, Larsen CS, Pedersen C, Pedersen G, et al. Impact of non-HIV and HIV risk factors on survival in HIV-infected patients on HAART: a population-based nationwide cohort study
. PLos One
2. Rosen S, Larson B, Rohr J, Sanne I, Mongwenyana C, Brennan AT, et al. Effect of antiretroviral therapy on patients’ economic well being: five-year follow-up
3. Larsen B, Fox MP, Rosen S, Bii M, Sigei C, Shaffer D, et al. Early effect of antiretroviral therapy on work performance: preliminary results from a cohort study of kenyan agricultural workers
4. Worthington C, O’Brien K, Zack E, McKee E, Oliver B. Enhancing labour force participation for people living with HIV: a multiperspective summary of the research evidence
. AIDS Behav
5. Dray-Spira R, Lert F. Living and working with HIV in France in 2003: results from the ANRS-EN12-VESPA Study
2007; 21 (Suppl 1):S29–S36.
6. Dray-Spira R, Legeai C, Le DM, Boue F, Lascoux-Combe C, Simon A, et al. Burden of HIV disease and comorbidities on the chances of maintaining employment in the era of sustained combined antiretoviral therapies use
7. Ezzy D, de Visser R, Grubb I, McConachy D. Employment, accommodation, finances and combination therapy: the social consequences of living with HIV/AIDS in Australia
. AIDS Care
1998; 10 (Suppl 2):S189–S199.
8. Dray-Spira R, Persoz A, Boufassa F, Gueguen A, Lert F, Allegre T, et al. Employment loss following HIV infection in the era of highly active antiretroviral therapies
. Eur J Pub Health
9. Delpierre C, Cuzin L, Lauwers-Cances V, Datta GD, Berkman L, Lang T. Unemployment as a risk factor for AIDS and death for HIV-infected patients in the era of highly active antiretroviral therapy
. Sex Transm Infect
10. Dray-Spira R, Gueguen A, Ravaud JF, Lert F. Socioeconomic differences in the impact of HIV infection on workforce participation in France in the era of highly active antiretroviral therapy
. Am J Pub Health
11. Legarth R, Omland LH, Kronborg G, Larsen CS, Pedersen C, Gerstoft J, et al. Educational attainment and risk of HIV-infection, response to antiretroviral treatment, and mortality in Danish HIV-infected patients
12. The Civil Registration System in Denmark, 2013.
The Central Office of Civil Registration. https://cpr.dk/in-english/
13. Obel N, Engsig FN, Rasmussen LD, Larsen MV, Omland LH, Sorensen HT. Cohort profile: the Danish HIV cohort study
. Int J Epidemiol
14. Quality declaration, The Attainment Register, Statistics Denmark
[Accessed on 26 February 2014]. http://www.dst.dk/en/Statistik/dokumentation/Declarations/the-attainment-register.aspx
16. Information on system on International Labour Standards, NORMLEX. http://www.ilo.org/dyn/normlex/en
. NORMLEX; [Accessed on 26 February 2014].
17. Larsen MV, Omland LH, Gerstoft J, Roge BT, Larsen CS, Pedersen G, et al. Impact of injecting drug use on response to highly active antiretroviral treatment in HIV-1-infected patients: a nationwide population-based cohort study
. Scand J Infect Dis
18. Larsen MV, Omland LH, Gerstoft J, Larsen CS, Jensen J, Obel N, et al. Impact of injecting drug use on mortality in Danish HIV-infected patients: a nation-wide population-based cohort study
19. Omland LH, Osler M, Krarup H, Weis N, Christensen PB, Roed C, et al. Socioeconomic status in HCV infected patients – risk and prognosis
. Clin Epidemiol
20. Antinori A, Coenen T, Costagiola D, Dedes N, Ellefson M, Gatell J, et al. Late presentation of HIV infection: a consensus definition
. HIV Med
21. Notice from the Ministry of Social affairs, Retsinfo. [Accessed on 26. February 2014]. www.retsinfo.dk
22. Schuring M, Mackenbach J, Voorham T, Burdorf A. The effect of re-employment on perceived health
. J Epidemiol Community Health
23. Hultin H, Lindholm C, Moller J. Is there an association between long-term sick leave and disability pension and unemployment beyond the effect of health status? A cohort study
. PLoS One
24. Deeks SG, Lewin SR, Havlir DV. The end of AIDS: HIV infection as a chronic disease
25. Lohse N, Gerstoft J, Kronborg G, Larsen CS, Pedersen C, Pedersen G, et al. Comorbidity acquired before HIV diagnosis and mortality in persons infected and uninfected with HIV: a Danish population-based cohort study
. J Acquir Immune Defic Syndr
26. LFS by sex and age –composition, stat.Extracts. http://stats.oecd.org/
. OECD; [Accessed on 26 February 2014].
27. Jarrin I, Lumbreras B, Ferreros I, Perez-Hoyos S, Hurtado I, Hernandez-Aguado I. Effect of education on overall and cause-specific mortality in injecting drug users, according to HIV and introduction of HAART
. Int J Epidemiol
28. Weis N, Lindhardt BO, Kronborg G, Hansen AB, Laursen AL, Christensen PB, et al. Impact of hepatitis C virus coinfection on response to highly active antiretroviral therapy and outcome in HIV-infected individuals: a nationwide cohort study
. Clin Infect Dis