Cardiovascular disease (CVD) is the leading cause of mortality in developed countries.1 Atherosclerosis is the basis of CVD, being present from its early stages.2 Early intervention has been shown to improve outcome, but its cost-effectiveness is controversial.3 Initial treatment consists mainly of lifestyle changes, predominantly those related to diet and physical activity. Some authors suggest that such interventions require trained personnel, which significantly increases cost, without a notable benefit in terms of the number of cardiovascular events and deaths.3

The effect of different cardiovascular risk (CVR) factors when analyzed individually confirms their influence on the duration of sick leave episodes.4,5 In Spain, sickness absence is covered for both work-related and nonwork-related injuries and diseases, but with different regulations.6 Classification as an occupational disease is constrained by a specific list of conditions for defined occupations, developed according to the influence of definitive exposures.7 Occupational injuries, on the other hand, refer to those caused in the context of an accident at work or while commuting.7 The remainder of injuries and diseases are considered nonwork-related. In the case of nonwork-related sickness absence, sick pay extends from the fourth day of sickness absence to 12 months, with the possibility of an additional period of 6 months following an evaluation by the Social Security Institute.7 Sick leave from the beginning to the end of the episode must be certified by the patient's primary care physician and must be confirmed on a weekly basis.6 Occupational diseases and injuries generally involve additional benefits (eg, sick pay from the first day).7

In a previous study, we showed that asymptomatic workers at high CVR, with only a clustering of CVR factors (CVRF), and therefore with undiagnosed underlying early CVD, contributed to a significant increase in the cost of sick leave, and the occurrence of early cardiovascular events.8 According to our data, the estimated increase in the cost of sick leave for the whole Spanish working population was over €145 million per year, suggesting a huge potential for savings to be made.8 The aim of the present study, conducted in a population similar to that included in our previous study,8 was to investigate whether changes in CVR profile are associated with the length and cost of sickness absence.

This prospective cohort analysis was a part of the Ibermutuamur CArdiovascular RIsk Assessment (ICARIA) study, the methodology of which has been described elsewhere.9,10 Briefly, CVR factors and global CVR, as estimated using the SCORE (Systematic COronary Risk Evaluation) chart for European low-risk countries, were assessed in a broad and representative sample of the Spanish working population.9,10 All participants who underwent a routine medical examination were approached and included in the ICARIA cohort, provided they gave informed consent. Medical examinations were conducted, consisting of a structured interview, anthropometric and blood pressure measurements, and blood testing. For current analyses, all participants with 2 consecutive medical examinations approximately 1 year apart (365 ± 90 days), and therefore 2 subsequent global CVR estimates, were selected. Participants with coronary heart disease, cerebrovascular disease, peripheral artery disease, or diabetes diagnosed prior to the first medical assessment, were excluded. After the second CVR estimate was performed, data regarding all medically-certified sick leave episodes, and the total count of sick leave days, were obtained from the official register of the Ibermutuamur mutual insurance company during a 1-year follow-up (365 days).8 In Spain, mutual insurance companies provide health care for occupational injuries and diseases.7 They also collaborate with the National Social Security System in case management and distribution of economic support for both work-related and nonwork-related sickness absence episodes.7 The proportion of the working population covered by mutual insurance companies in Spain is 98% for work-related sickness absence and 83% for nonwork-related sickness absence.11 The official records held by these companies are fundamental to the conduction of epidemiological research into sickness absence (especially in the case of nonwork-related sickness absence) due to the lack of an official, centralized, nationwide registry in Spain.

All participants were informed about their CVR and were given recommendations regarding CVRFs control and lifestyle changes (diet and physical exercise). Furthermore, a clinical summary was sent to their primary care physician to encourage the implementation of lifestyle changes and to support any eventual introduction of drug therapy.

Figure 1 shows patient flow. The sample consisted of 179 186 participants, 72.1% of whom were men (Table 1). The mean age (± standard deviation) was 36.7 ± 10.4 years. When the workers were categorized into the 4 groups according to changes in CVR, there were significant differences in their distribution by sex and age (P < .001): 92.9% of participants had a SCORE that was stable at < 4% in the 2 estimates (70.5% men; mean age: 35.7 ± 9.72 years); 2.4% of them displayed a worsening, from an initial SCORE of < 4%, to ≥ 4% in the second estimate (90.1% men; mean age: 48.0 ± 9.89 years); 1.9% of participants displayed an improvement, from an initial SCORE of ≥ 4%, to < 4% at the second medical examination (90.2% men; mean age: 47.3 ± 9.78 years); finally, 2.7% of participants remained stable at ≥ 4% (97.0% men; mean age: 55.06 ± 8.12 years).

Figure 1.

Patient flow and baseline characteristics. CVR, cardiovascular risk.

(0.18MB).

Table 1 and Table 2 show incidence densities of new sickness absence episodes per 100 worker-years during a 1-year follow-up after the second CVR estimate, as well as the number of sickness absence days. The total incidence density of sickness absence episodes of all causes was 22.22 per 100 worker-years (95%CI, 22.02-22.43). With regard to the specific cause of sickness absence, incidence density was 15.75 per 100 worker-years (95%CI, 15.57-15.92) for nonwork-related sickness absence, 6.88 per 100 worker-years (95%CI, 6.76-7.00) for work-related sickness absence, and 0.12 per 100 worker-years (95%CI, 0.11-0.14) for CVD-related sickness absence.

As shown in Figure 2, after adjustment for covariates, the 1-year change in CVR profile remained significantly associated with the total count of sickness absence days at the end of the study (P < .001). Participants with a stable CVR of < 4% in both routine medical examinations had a lower count of sickness absence days than participants with a stable CVR of ≥ 4% (Figure 2). The group of participants who improved their CVR level from ≥ 4% to < 4% also showed a lower count of sickness absence days during follow-up in comparison with participants with a stable SCORE of ≥ 4%. This decrease was observed for the whole group of sickness absence episodes (RR, 0.91; 95%CI, 0.84-0.98), for nonwork-related episodes (RR, 0.89; 95%CI, 0.82-0.96), and for sickness absence due to CVD (RR, 0.66; 95%CI, 0.61-0.71), but not for work-related sickness absence (RR, 0.96; 95%CI, 0.87-1.05). In contrast, the group of participants that displayed a worsening of their CVR did not differ from participants with a stable SCORE of ≥ 4% in terms of nonwork-related and work-related sickness absence (P ≥ .05), but showed increased sickness absence due to CVD during follow-up (RR, 1.10; 95%CI, 1.04-1.17). Mean savings per participant in terms of sick pay associated with improvement in CVR were estimated at €40.03 per year (± €1766.37). When extrapolated to the whole Spanish working population at CVR ≥ 4%, the potential savings amounted to €52 026 686.80 per year (95%CI, €80 084 480.40-€1 503 558.30).

Figure 2.

Association of 1-year cardiovascular risk trend (365 ± 90 days) with the total count of sickness absence days during 1-year follow-up after the second evaluation of their CVR (SCORE charts). Poisson regression analyses (standard error correction). CVR, cardiovascular risk.

(0.25MB).

Table 3 shows the percentages of participants with differences in CVRFs and lifestyle, comparing those with improvement to those with no improvement in their CVR. These data show significantly higher percentages for those participants with an improvement in CVR, for all the parameters considered. The only exception was lifestyle (diet and physical exercise), which exhibited a positive trend that did not reach statistical significance.

When associations between CVRF progression and sickness absence were tested, tobacco consumption progression was consistently associated with sickness absence (Table 4). Workers who stopped smoking between the 2 medical examinations had a lower risk of sickness absence than those who continued to smoke (RR, 0.88; 95%CI, 0.84-0.92), although the risk was lower still for patients who were nonsmokers at both medical examinations (RR, 0.82; 95%CI, 0.81-0.84). The same trend was observed for nonwork-related sickness absence and sickness absence due to CVD, but not for work-related sickness absence.

Findings regarding the association between progression of hypertension and sickness absence were mixed (Table 4). Patients who were hypertensive at the first medical examination and under antihypertensive therapy at the second assessment had an increased risk of all-cause sickness absence, regardless of hypertensive status at the second visit (RR, 1.20; 95%CI, 1.11-1.30 for no hypertension; RR, 1.21; 95%CI, 1.14-1.28 for hypertension at the second visit). The same trend was also observed for nonwork-related sickness absence. Conversely, patients who were hypertensive at the first medical examination and under antihypertensive therapy at the second had a reduced risk of sickness absence due to CVD if blood pressure had been successfully controlled (RR, 0.59; 95%CI, 0.52-0.67), but not if blood pressure was still ≥ 140/90mmHg (RR, 1.41; 95%CI, 1.33-1.49). Findings regarding dyslipidemia progression were also mixed (Table 4).

The main finding of the present study was the decrease in sickness absence in participants showing an improvement in their CVR profile from ≥ 4% to < 4%, according to the SCORE chart, during the 1-year follow-up period. Such a reduction in sickness absence was observed for nonwork-related and CVD absence. The association of CVR reduction with decreased sickness absence was still significant after adjustment for sex, age, occupation, tobacco consumption, and the incidence of prior sickness absence.

Our results suggest that positive changes in CVRFs are involved in sickness absence reduction during a 1-year follow-up. The improvement in CVR profile was the result of higher percentages of participants achieving controlled blood pressure and total cholesterol, low-density lipoprotein cholesterol, and triglycerides levels, and stopping smoking by the second medical examination. The percentages of participants with hypertension taking antihypertensive drugs, and participants with dyslipidemia receiving lipid-lowering therapy, also increased in this group 1 year after the initial assessment. A consistent association between tobacco consumption and sickness absence was observed, ranging from the lowest risk in nonsmokers at both examinations to the highest risk in those who were smokers at both. Blood pressure control among hypertensive participants appeared to be associated with a decrease in the risk of sickness absence caused by CVD. In contrast, the consistent association between antihypertensive treatment and increase in the risk of all-cause (and, more specifically, nonwork-related) sickness absence suggests that antihypertensive drug prescription could be interpreted as a severity marker (ie, antihypertensive drugs were only prescribed in the most serious of cases). If this were true, it could imply the need for a revision of current prescription practice, especially when patients are theoretically “young and healthy”. Our findings regarding progression of dyslipidemia are complex and could be related to the small number of participants receiving lipid-lowering therapy.

Our results also demonstrate that sickness absence of nonwork-related and work-related origin among participants with a worsening of their CVR from < 4% to ≥ 4% was similar to that observed in participants whose risk remained stable at ≥ 4%. Furthermore, sickness absence due to new-onset CV illnesses increased during the 1-year follow-up for this group. These findings should encourage occupational health care providers to focus health promotion programs not only on participants at high risk, but also on those at low risk who could potentially experience worsening of their risk.5

Several mechanisms have been proposed to explain the link between CVR and non-CVD sickness absence: the association of CVR with proinflammatory or prothrombotic states, which may contribute to a number of non-CVD diseases (eg, respiratory diseases, musculoskeletal pathology, or infectious diseases); involvement of health risk behaviors that are risk factors for other diseases; and a hypothetical underlying risky personality type.8

In a previous report, we showed that high CVR is associated with enormous cost in terms of sickness absence among the working population.8 The present results suggest that those costs could be significantly reduced in the short-term if CVR was successfully improved. Our findings are in line with a previous report on changes in health care, pharmacy, and short-term disability costs among manufacturing participants who improved their metabolic syndrome status.5 Prior research has not always demonstrated a reduction in sickness absence as a consequence of lifestyle modification.17 The reasons for this apparent discrepancy with the results reported here could be differences in the explanatory variables (estimated CVR or lifestyle parameters) or in the methodology of the studies. In some cases, the scientific evidence was obtained from selected populations, or was exclusively based on self-reported data.17 If our findings were to be confirmed, a reduction in sickness absence costs should be added to the decrease in the incidence of absences and in mortality associated with improvement in the control of CVRFs in most developed countries.18

Of importance, in our experience, 40% of participants with an elevated CVR profile at baseline showed an improvement 1 year later. In the remaining individuals, the CVR profile remained at ≥ 4%. In addition, the number of workers with a SCORE of ≥ 4% in the second medical examination increased as a consequence of the more than 4000 participants that moved from the < 4% to the ≥ 4% risk category. These results strongly suggest a need to improve the level of intervention used for our workers, with Health Promotion at Workplace Programs being potentially useful for this purpose.17 Changes in lifestyle are critical for CVR reduction,19 and the significant decrease in the cost of sick leave episodes that is associated with improvement in CVR profile is notable.

A stable or improved CVR level during a 1-year period, as estimated by SCORE charts for low-risk European countries, was significantly associated with shorter nonwork-related sickness absence, and shorter absence due to CVD, during a subsequent 1-year follow-up period. Further research will determine whether Health Promotion at Workplace Programs is cost-effective.

This study was funded by a research project grant (FIS PI12/02812) from the Health Institute Carlos III and the Spanish Ministry of Economy and Competitiveness.

None declared.