Return to work and associations with psychosocial well-being and health-related quality of life in coronary heart disease patients: Results from EUROASPIRE IV

Abstract

Background

Coronary heart disease (CHD) can lead to loss of workability and early retirement. We aimed to investigate return to work (RTW) and its relationship towards psychosocial well-being and health-related quality of life (HRQoL).

Design

Secondary analyses were applied to cross-sectional data from the EUROASPIRE IV survey (European Action on Secondary and Primary prevention through Intervention to Reduce Events).

Methods

Participants were examined and interviewed at 6–36 months following the recruiting event. Psychosocial well-being and HRQoL were evaluated by completing the ‘Hospital Anxiety and Depression Scale’ and ‘HeartQoL’ questionnaire. Using generalised mixed models, we calculated the odds ratios for RTW. Depression, anxiety and adjusted means of HeartQoL were estimated accounting for RTW.

Results

Out of 3291 employed patients, the majority (76.0%) returned to work, of which 85.6% were men, but there was a general underrepresentation of women. Young (p < 0.001), high-educated (p < 0.001) patients without prior cardiovascular events (p < 0.05) were better off regarding RTW. No significant associations with CHD risk factors and cardiac rehabilitation were established. Those that rejoined the workforce were less susceptible to psychosocial distress (anxiety/depression, p < 0.001) and experienced a better quality of life (p < 0.001).

Conclusion

These findings provide evidence that non-modifiable factors (sociodemographic factors, cardiovascular history), more than classical risk factors, are associated with RTW, and that patients who resume work display better psychosocial well-being and HRQoL. Our results illustrate a need for tailored cardiac rehabilitation with a focus on work-related aspects, mental health and HRQoL indicators to reach sustainable RTW, especially in vulnerable groups like less educated and elderly patients.

Background

Cardiovascular diseases (CVDs) account for 45% of all deaths across Europe, with coronary heart disease (CHD) being the most prevalent cause (20%).¹ Mortality rates have declined due to evolutions in diagnosis, treatment and risk factor control.¹ However, the disease burden remains high, with CHD patients having substantial physical and mental problems² and displaying impaired health-related quality of life (HRQoL) scores.^2,3

Studies have highlighted that severe CVD may result in loss of workability, thereby leading to disability, early retirement and elevated risk of CVD mortality.^4,5

Return to work (RTW) after CHD has been linked to medical characteristics (e.g. low disease severity, no/few comorbidities, low-invasive intervention, short hospital stay, cardiac rehabilitation (CR) participation and duration), demographic distributions (e.g. young age, men, high education and socio-economic status), mental well-being (no depression/anxiety), individual characteristics (e.g. no type-A behaviour, no type-D personality, positive illness beliefs, low work stress, good work expectations, high job satisfaction), occupational characteristics (e.g. balanced job demands, autonomy, social support) and high HRQol.^6–15

Several studies^7,11,15 discussed the importance of RTW as an indicator of functional and mental rehabilitation after cardiac events. Patients that do return to work display better HRQoL scores and less anxiety and/or depression.^11,16

The heterogeneity of the RTW literature makes further exploration of this subject more than valid, especially in a large study population and by using disease-specific questionnaires to assess psychosocial well-being and HRQoL.

Based on the EUROASPIRE IV (European Action on Secondary and Primary prevention through Intervention to Reduce Events) survey in coronary patients, we aimed to investigate RTW and several characteristics at the time of the CHD event and, secondly, the relation of RTW to psychosocial well-being and HRQoL.

Methods

Study population and data collection

EUROASPIRE IV was conducted in 78 centres across 24 European countries from 2012 to 2013. Male and female patients were interviewed and examined 6–36 months after hospitalisation for a first or recurrent CHD event. Detailed information about the study design/methods can be found in the original publication.¹⁷ For the present study, only patients with a coronary artery bypass graft (CABG), percutaneous coronary intervention (PCI) or acute myocardial infarction (AMI) as recruiting event and employed at that time (full-time, part-time or self-employed) were eligible. An age restriction of <63 years was applied to include only patients with a fair chance at RTW considering age and the time between the recruiting event and the interview. Furthermore, when categorising our age variable, we took the possibility of an early retirement age into account: < 50 years; 50–<58 years; 58– < 63years. Education was divided into three levels: primary education was considered as primary school or less; secondary education as secondary school or high school completed; high education consisted of university/college degree.

Data were retrieved from the interview unless indicated otherwise, in accordance with the research questions.

RTW

Employment status, 6–36 months later, was dichotomised in RTW (full-time, part-time or self-employed) or no RTW (retired, unemployed, home maker, student). Self-employment was added as a dichotomous independent variable.

From the hospital discharge letter, risk factor information (yes/no) such as hypertension, dyslipidaemia, abnormal glucose metabolism, smoking status and obesity were retrospectively retrieved. Body mass index (BMI) at hospitalisation was calculated and combined with the discharge information to optimise our obesity data into a dichotomous variable (‘overweight or obese’). From self-reported medical history, we determined prior CHD as a CHD event which occurred before the recruiting event (including CABG, PCI, AMI, acute myocardial (AM) ischaemia, angina) and prior stroke/transient ischaemic attack (TIA) as an event which occurred before the recruiting event. CR was defined as attending ≥50% of the sessions.

Psychosocial well-being and HRQoL

Participants were asked to fill in the Hospital Anxiety and Depression Scale (HADS) and HeartQoL questionnaires. HADS depicts symptoms of anxiety (subscale HADS-A) and depression (subscale HADS-D) by way of 14 questions with a response scale of 0–3.¹⁸ HADS scores ≤ 7 were considered as normal, 8–10 as mild and ≥11 as moderate to severe symptoms.¹⁹ The HeartQoL questionnaire is a disease-specific tool for measuring HRQoL in CHD patients.^20,21 It consists of 14 items; 10 items focus on physical well-being and four items on emotional well-being, together providing a global (HeartQoL_global), physical (HeartQoL_physical) and emotional scale (HeartQoL_emotional), each ranging between 0 (worst) and 3 (best).

We retrieved medical history such as self-reported hypertension (HT), diabetes, dyslipidaemia and the occurrence of a new (since hospital discharge) CHD (including CABG, PTCA, AMI, AM ischaemia) from the interview after the recruiting event, which was defined as ‘recurrent CHD’. Because of an insufficient number of cases, other recurrent pathologies were not included. Regarding disease management, only CR and the use of antidepressants/anxiolytics were taken into account. Physical activity (PA) outside of work was divided into three categories: no PA, light PA (most weeks) and vigorous PA (1–≥3/week). Current smoking was defined as self-reported smoking and/or a breath carbon monoxide exceeding 10 ppm. BMI was calculated and was categorised as normal (<25 kg/m²), overweight (25–30 kg/m²) and obese (≥30 kg/m²). Patients were asked if and what actions they undertook to stop smoking, change their diet, lose weight and increase their PA. For each category the total information was dichotomised.

Statistical analyses

Descriptive analyses of the baseline-, RTW- and non-RTW groups were performed. This was reported separately for HADS-D/A categories and as HeartQoL_{global/physical/emotional} scores (mean, standard deviation) and lifestyle factors within the RTW and non-RTW groups.

All association models were assessed by way of generalised mixed models accounting for country-related clustering, and all baseline adjustments included age, sex, education and recruiting event.

Odds ratios (ORs) and 95% confidence intervals (CIs) by way of multilevel logistic regressions were calculated for RTW, with additional adjustments including clinical information at discharge, medical history, CR and self-employment.

HADS was dichotomised as an outcome (normal: < 8; symptomatic ≥ 8) for each subscale and ORs (95% CI) were calculated by logistic multilevel regressions. For HRQoL, the adjusted means (standard error) of HeartQoL scores were assessed by multilevel linear regressions. All HADS and HeartQoL models were additionally adjusted for RTW status, self-reported clinical risk factors, recurrent CHD, CR and lifestyle.

After analysing possible interaction terms (between age/sex and every independent variable, p < 0.01) only an age-stratification (<50 years, 50–<58 years, 58–<63 years) was performed for RTW, HADS-A, HeartQoL_global and HeartQoL_emotional. By including age as a continuous variable, the effect of age was analysed in each determined age group. A sensitivity analysis (HADS-D/A ≥ 11) was also performed to test the robustness of our results. All analyses were performed with IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, New York: IBM Corp and overall significance was defined as p < 0.05.

Results

From a total of 7998 CHD patients, 3291 (41.2%) were employed before the event, of which 2661 were aged < 63 years. Of those eligible, the employment status at the time of the interview was available for 2651 patients (99.6%). The median time (interquartile range) between index event and interview was 1.38 years (0.96–2.01 years). The majority of participants (n = 2014; 76.0%) returned to work and the baseline characteristics and results of the regression analysis can be found in Table 1. RTW was more prevalent in younger, high-educated patients without a foregoing cardiovascular event. Risk factors, as identified at discharge, and CR participation were not associated with a higher RTW rate when adjusted for sociodemographic and clinical factors and self-employment.

Age stratification (Appendix 1) showed lower RTW in hypertensive patients ≥58 years and younger patients treated with CABG. Higher RTW odds were found in young non-smokers.

As shown in Table 2, possible (HADS ≥ 8) or probable (HADS ≥ 11) depression (26.5%) and anxiety (35.5%) were more prevalent in the non-RTW group, but the majority in both groups were untreated. The non-RTW group was more obese and less physically active, but included more smokers undertaking actions to quit. HeartQoL scores were better, especially on the physical level, in the RTW group.

Table 3 presents data on associations between depression/anxiety/HeartQoL and RTW.

RTW was associated with less depression, even after full adjustments. After age stratification, the odds for anxiety became higher in the case of RTW with rising age, and in the oldest group RTW appeared to be no longer significant (p = 0.08) for displaying symptoms of anxiety (see Appendices 2 and 3).

The sensitivity analysis (full model) showed a similar significant association between HADS-D ≥ 11 and RTW (OR 0.63; 95% CI (0.44, 0.90); p < 0.05) and between HADS-A ≥ 11 and RTW (OR 0.61; 95% CI (0.46, 0.82); p < 0.01).

Adjusted means for HeartQoL_{global/physical/emotional} (Table 3; Appendix 4), when CHD patients RTW, are higher than when they fail to RTW. After performing an interaction analysis for age, sex and HeartQoL, only age displayed a significant relationship for HeartQoL_{global/emotional}. After age stratification (Appendix 5), RTW remained significantly (p < 0.001) associated with HeartQoL_{global/emotional}. The highest HeartQoL_global scores were found in the youngest age group, whereas the highest HeartQoL_emotional scores were reported by the oldest age group. The difference between RTW and non-RTW participants, however, is the most prominent in the youngest age group for both HeartQoL domains.

Discussion

From the EUROASPIRE IV database, a large international cross-sectional study, we examined the proportion of CHD patients that RTW and the concomitant clinical information at discharge.

We found that 76% of our included CHD patients returned to work in the period of six months to three years after discharge. Our RTW rate is in line with most of the studies of the last 15 years displaying an average score (70–80%). Some studies^{6,11,13–15,22,23} report a higher RTW rate, but use a broader RTW definition (‘return to the workforce’, also including unemployed but workable participants) or are characterised by smaller populations, single treatment selection, combination of recruiting diagnosis, national settings, shorter follow-up period or alternative definitions of RTW outcome. Furthermore, apart from prior CHD and CR attendance, we focussed on the available clinical information at the time of the recruiting event. However, the design of our RTW model hinders accounting for clinical events during follow-up that may influence work resumption and social reintegration.

Although RTW is, in general, lower in women^{9,10,12,23–25} and female patients were found to display a worse CV risk factor profile in EUROASPIRE IV,²⁶ we could not find a significant association between RTW and gender. This is in line with the VIRGO study¹³ and the TRANSLATE-ACS study¹¹; however, the underrepresentation of women in our study needs to be considered.

As with previous studies,^{10,12,15,23,24,27} CHD patients who returned to work were younger and highly educated. Patients < 50 years were three times more likely to RTW, while for patients of 50–58 years the odds were about twice as high. With an increasing older working population, it remains important to enhance a continued participation of patients ≥58 years since they make up a large proportion of the CHD spectrum.

Multiple studies^10,13,15,27 have reported the beneficial effect of education on RTW. High education is often accompanied by better socio-economic status, more physically favourable work conditions but often high stress levels and sedentary conditions. Moreover, within educational group levels there are also variations of CVD prevalence between occupational sectors²⁸ due to distinctions in clinical health versus behavioural health in each profession. These educational/occupational variations emphasise the importance of multidisciplinary tailored reintegration beyond a focus on classical cardiovascular risk factors (CV RFs).

Similar to the VIRGO study¹³ and the TRANSLATE-ACS study,¹¹ a foregoing CHD event or TIA/stroke diminished the odds to RTW.

In contrast to previous studies,^8,13 CR was not associated with RTW, but the lack of detailed information on the different regional CR programs and the low attendance needs to be considered. Qualitative research by O'Hagan et al.⁷ has mentioned that some patients view CR as beneficial for lifestyle changes and building up capacity, but it lacks the occupational focus and barely contributes to RTW. This knowledge could prove an opportunity for more dialogue with occupational physicians as at such time European guidelines are being developed for CR involving work-related aspects (personal communication).

Regarding CV RFs, we only found HT and smoking, after age stratification, to be related to RTW, which is in contrast to previous research.^9,11,23 Lower RTW in hypertensive older patients could be due to a more challenging disease management, which may influence its compatibility with work demands. Younger patients displayed a negative influence of smoker status and CABG on RTW. Treatment options need to be carefully considered here, because although they may be better candidates for surgery, it seems to negatively impact RTW. Younger patients may have a worse CV profile or more physical demanding jobs and a more stressful family life.

Although previous studies^15,22 have shown conflicting results regarding the association between self-employment and RTW, we could not find a significant association between RTW and self-employment after adjustments. Of note, only a small percentage (9%, n = 239) of our population was self-employed.

In a large number of studies, depression^10,13,24,27 and other mental comorbidities^6,10,24,29 have been suggested as a barrier for RTW, but to our knowledge we are the first to analyse depression/anxiety, using the HADS scale, in relation to RTW as an independent variable in a large CHD population.

In our employed group, about 16% (Table 2) were working while having depressive symptoms, which is similar to a study by Warraich et al.¹¹. Twenty three percent of our patients were able to work while experiencing anxiety. A possible deduction hereof could be that the presence of depression has more consequences regarding the ability to work than anxiety. The unfavourable combination of depression and work resumption was also emphasised by Smedegaard et al.²³ Our results from the regression analyses (Table 3), could also suggest that the relationship between RTW and depression/anxiety is unidirectional, since RTW has a protective effect on depression and anxiety. This is in line with the aforementioned studies,^11,16,23 although, due to the heterogeneity of measurements, comparisons should be carefully approached.

The sensitivity analysis also showed that RTW was less prominent in patients displaying probable depressive/anxiety symptoms. This emphasises the need for careful follow-up of psychological comorbidities, as shown in previous literature,^6,24 the integration of psychosocial guidance during rehabilitation and afterwards and a step-by-step RTW approach.

Although work has, in general, been linked to quality of life,³⁰ we could only retain two studies^{11,  16} where RTW was also included in the regression analysis of a HRQoL outcome in CHD patients. To the best of our knowledge, our study is the first to analyse HRQoL using the HeartQoL tool with RTW as a possible predictor, in employed CHD patients within a large multinational database.

RTW has been attributed as an important marker of the degree of functional rehabilitation, the patient's return to ‘normality’, which is reflected in our results. In all areas of HRQoL, RTW is associated with higher mean HRQoL scores. When adjusting for sociodemographic and clinical risk factors and lifestyle behaviour, the difference between the mean scores was the most prominent for HeartQoL_emotional, which suggests the importance of RTW for mental well-being. After age stratification, RTW was also more influential for HRQoL in patients <50 years. In the young, who were more plagued by common mental disorders and low HeartQoL_emotional, RTW could be a mediator.

While the cross-sectional design of EUROASPIRE IV limits the interpretation of results, the study population remains a major source of data.

Like most studies, our selected patients were employed before the CHD recruiting event, as this status has been shown to influence later RTW.¹⁵ This does discard, however, patients that, although unemployed, may have had the intention and motivation to regain employment after the event.

Unlike other studies, we applied a realistic age restriction, with patients who have a chance at RTW, since in this aging society a continued employment or participation of the elderly workforce is encouraged.

Contrary to previous RTW studies, which focus on a specific diagnostic event (CABG, PCI or AMI), we applied a wider definition of index event (PCI, CABG, AMI).

It is unclear whether patients who were still on sick leave at the interview were labelled as retired. Due to the single data collection in the interval time of 6–36 months, it is possible that some patients did RTW but then retired or were again on sick leave. As such, we have no information on the number of RTW attempts (full/partial), time to RTW or work sustainability. Information is also lacking about the patients' occupation, their social support, their work environment and their work (dis)ability. This hampers a thorough analysis of the RTW process in accordance with elaborations of the biopsychosocial health paradigm.³¹

EUROASPIRE IV provides broad information regarding psychosocial well-being (HADS, HeartQoL), but these were only measured at the time of the interview. As indicated in the study by Fiabane et al.,⁶ it is relevant to measure and follow-up HRQoL, depression/anxiety during the hospitalisation, during CR and at later time points.

In summary, to our knowledge this is the first study to investigate the relationship between HeartQoL and RTW, and to use RTW for analysing HADS-A/D in CHD patients.

CHD patients of EUROASPIRE IV returned to work at a reasonable rate (76%) compared to other studies. RTW is subject to non-modifiable factors like age, education and prior CAD/TIA/stroke. Contrary to our expectations, there was no significant typical gender gap, and CR along with CV RFs were not associated with RTW in this population.

Our results suggest that RTW is negatively associated with psychosocial well-being (depression, anxiety) and positively associated with HRQoL, whereby patients whom resumed their work reported less symptoms of anxiety/depression and better HeartQoL scores.

In consideration of a growing aging population accompanied by a longer working age, we suggest future research pay attention to a thorough follow-up of mental health and HRQoL at different stages of treatment and during RTW, and to CR programs including relevant work characteristics.

Author contributions

DDB, DDS and KK contributed to data acquisition. JVdC, DDB, DDS, EC and LB contributed to the design and conception of the work, and interpretation of the data. JVdC carried out the analyses and drafted the manuscript. JVdC, LB, DDB, EC and DDS critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.

Acknowledgements

The EUROASPIRE IV survey was carried out under the auspices of the European Society of Cardiology, EURObservational Research Programme. We thank the personnel of participating hospitals and all the patients who partook in the EUROASPIRE studies.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References