Efforts for improving the measure of the burden of diseases are needed for setting priorities on resource allocation. Along with traditional indicators of health status (morbidity and mortality), disability has become an important indicator of burden.1 Disability is “the difficulty in functioning at the body, person, or societal levels, in one or more life domains, as experienced by an individual with a health condition in interaction with contextual factors”.2 Around 15% of the world's population experiences disability.3 One major component of the economic costs of disability is loss of labour productivity, as a result of work absences or reduced work.3 Data on disability-related productivity losses are scarce and difficult to obtain. In Canada,4 the total indirect cost attributable to short and long-term disability was around 42 billion dollars in 1998, the 6.7% of Canada's GDP.

Common conditions, including mental disorders, are important determinants of disability, but the interrelation between disorders and disability is complex. High prevalence of a health condition does not always mean high associated disability. In 2004, in Australia,5 arthritis was the most prevalent condition and the one with the highest proportion of disability, while autism was causing the highest impact on activity limitation, though it had the lowest prevalence rate. The number of co-occurring disorders, co-morbidities, is also important when considering the relationship between health conditions and disability. Recently, two reports6,7 from the World Mental Health Surveys showed that controlling by co-morbid conditions and their specific severity weights, mental and physical conditions yielded substantial full and partial limitation for performing activities of daily living worldwide. Marked differences by country level income were observed: while in high income countries, health conditions have a large effect on partial role limitation; in low and medium income countries a substantial impact on full role limitation was reported. The authors hypothesised on differences in prevalence rates of conditions as well as differences in welfare states as possible explanations of such cross-national differences.

Contextual factors (e.g., technology, attitudes, health services, legislation and policies) have a huge impact on the experience and in the extent of disability.3 In 2002, Spain reported one of the lowest prevalence percentages of disability in Europe, with 8.7% of the working-age population reporting disability.8 Recently, the burden of disease in Spain was measured.9 Neurological and mental disorders, malignant neoplasms, and cardiovascular disease, were the leading causes of DALYs (i.e., disability-adjusted life years). While comparative illness burden estimates are valuable for health planners, two important limitations can be underlined. First, valuation of the burdens of different conditions relies, basically, on experts’ judgments of vignettes10,11 rather than on the individual perception of s/he's own disability. Second, vignettes represents single conditions rather than realistic cases with a number of morbidities.12

Here we examine the effects of common conditions on number of days with either full or partial role limitation by using the Spanish data from the World Mental Health Survey (ESEMeD-Spain), which collected information on prevalence of mental disorders and correlates, along with information on physical conditions and disability.

Socio-demographic characteristics of the ESEMeD-Spain sample are shown in Table 1. The average age of respondents was 45.5 (SE=18.5), and women represented 51.4% (SE=1.7%). About one in two respondents (50.4%) were working at the time of the interview. Almost two-thirds (58.3%) of the sample has reached basic education. Nearly half of the sample (47.1%) had at least one common condition. Anxiety disorders were the most prevalent mental disorder (6.5%) while back/neck pain (17.0%) was the most prevalent physical condition. One fifth of the total sample had two or more conditions, with average of 1.7 conditions. Figure 1 shows the monthly distribution of days with full and partial role limitation. Overall, 7.7% of the sample had at least one day with full role limitation while double the respondents (14.2%) had at least one day with partial role limitation.

Figure 1.

Percentage of respondents with days with full and partial role limitation in the month previous to the interview (ESEMeD-Spain).

(0.08MB).

Table 2 shows the prevalence estimates of respondents with role limitations. About 42% of those respondents with mental disorders reported having role limitations. Among those with physical conditions, the figure was about 26%. Digestive, depression and anxiety were the conditions with the highest proportion of respondents with role limitation. Considerable differences in the number of days with full role limitation were observed across conditions. Respondents with any mental disorder reported a large number of days with full role limitation than those with physical conditions (34 vs 26, respectively). Digestive disorders, insomnia, depression and arthritis were the conditions with the highest number of days with full role limitation per year (93, 74, 47, and 46, respectively). Partial role limitation estimates were almost twice in respondents with mental disorders than with physical conditions (44 vs 28, respectively). Depression, arthritis, pain, and anxiety disorders were the conditions with the highest number of days with partial role limitation per year (60, 53, 46, and 44, respectively). Overall 20% of the sample reported to have had some day with role limitation.

Table 3 shows the additional number of days with full role limitation among respondents with common conditions, when compared to individual not having each of the individual conditions considered per year. Insomnia showed the highest mean number of additional days with full role limitation (59), followed by digestive disorders (57), and depression (25). On the other hand, depression (44), arthritis (43) and anxiety disorders (34) had the highest number of additional days with partial role limitation. Respondents with mental disorders had almost twice as much days with partial role limitation (42) than days with full role limitation (24). Those with physical conditions, on the other hand, had similar number of days for both types of role limitation (19 full and 21 partial role limitation).

In Table 4, PARPs of days with full and partial limitation for each condition are shown. Overall, all common conditions accounted for 73% of all days with full role limitation and 41% of all days with partial role limitation. Depression contributed to a similar proportion in both functioning outcomes (16% in full role limitation and 14% in partial role limitation). Arthritis (20%) yielded the highest attributable proportion in partial role limitation.

Common mental disorders and physical conditions accounted for a substantial proportion of days with full and with partial role limitation. While Spain showed a similar pattern of role limitation due to common conditions, the estimates are considerably higher than comparable data for high income countries.6,7,20 Mental and physical conditions contributed to a similar number of days with full role limitation (around 20 days per year), in contrast with partial role limitation, where mental disorders were responsible for twice the number of days than physical conditions (42 vs 21 days, respectively). This study extends the evidence that mental disorders contribute to substantial productivity losses, specifically through reduced daily performance.21–24 Overall, from a societal perspective, all nine conditions accounted for almost three-quarters of the total number of days with full role limitation, also much higher proportion than that reported for other developed countries.

Depression was associated with a remarkably high number of days with partial limitation. This is consistent with other studies where depression was strongly associated with worse performance at work.25,26 In a recent study with a similar methodology to ours20 carried out in the Netherlands, workers had nearly one month per year with a decreased performance due to major depression. Our finding is fairly consistent but still much higher since it is referring to all adult population and not only to employees. In fact, only half of the population was employed by the time of the study. So, it is worth pointing out that a large part of the burden associated to role limitation due to common conditions in general, and depression in particular, is being bear at a personal level in Spain. Although the scope of this study was not to consider functioning from an economic perspective, the enormous amount of days with partial role limitation reported here might have huge costs for society. It is also interesting to consider the healthy worker bias27 as a possible explanation of the large number of days with role limitation in a country with one of the lowest employment rate (62.7%) in Europe. While no causal relationships can be drawn from our cross-sectional data, it is widely known that there is a higher prevalence rate of health conditions, especially mental disorders in the unemployed population.28

A condition with a high prevalence rate, not always means higher disability. Depression and anxiety, despite of its relatively low prevalence rate,13 presented a higher impact on role functioning. This result is consistent with data reported by Gènova-Maleras et al.,9 where mental disorders ranked first as the most burdensome conditions in Spain. But, as mentioned above, burden of disease studies do not take into account co-morbidities. In our study, condition-specific ratings substantially change after the adjustment for co-morbidities. For instance, respondents with cardiovascular diseases had about 30 days with full role limitation and other 30 days with partial role limitation. But, after the adjustment for co-morbidity, both estimates dropped substantially and did not reach statistical significance. The same happened with other burdensome conditions such as pain, diabetes and respiratory diseases, which ranked in the top ten disease categories in the burden of disease study, and in ours, all three had almost none additional day with role limitation, after co-morbidity adjusment. We used interactions between the number of co-morbidities and each condition as the way of adjustment, allowing interactions to vary across conditions but not across particular pairs or higher number of disorders. While it is unlikely that our interaction model is optimal, we believe it is a useful approach to take into account co-morbidity.12

If the population were entirely unexposed to common physical and mental conditions, days with full limitation would be reduced by 73% and days with partial limitation by 41%. Health conditions explained a great proportion of days with full role limitation, but other non-health related factors come to play when referring to partial limitation.29,30 Work-related factors (e.g. shift work, physical work, employment position, among others) together with non work-related characteristics (e.g. family life, financial situation, adverse life events, among others)31 have shown to be relevant in explaining work performance. For instance, the ITSAL project (Immigration, Work and Health, for the Spanish acronym) found that the immigration status was one of the most important predictors of sickness presenteeism in Spain (OR=1.77; 95%CI: 1.24-2.53).32 We speculate that there are contextual factors that modulate the association between common health conditions and work functioning, which might be important to investigate in order to give a complete picture of determinants of lost productivity in Spain.

All authors have participated actively in the study, and have read and approved the submitted manuscript. J.M. Haro and C. García Forero were involved in the conception and design of the study and critically reviewed the manuscript; G. Barbaglia was involved in the conception and design of the study, the analysis and interpretation of data, and she was responsible for drafting the manuscript; N. Durán were involved in the analysis and interpretation of data and critically reviewed the manuscript; G. Vilagut and J. Alonso were involved in the conception and design of the study, the analysis and interpretation of data, and critically reviewed the manuscript. All authors provided final approval for the manuscript submitted.

This work was supported by the European Commission (QLG5-1999-01042, SANCO 2004123, EAHC 20081308), the Piedmont Region (Italy), Fondo de Investigación Sanitaria, Instituto de Salud Carlos III, Spain (FIS 00/0028-02), Ministerio de Ciencia y Tecnología, Spain (SAF 2000-158-CE), Departament de Salut, Generalitat de Catalunya, Spain, and other local agencies and by an unrestricted educational grant from GlaxoSmithKline. ESEMeD is carried out in conjunction with the World Health Organization World Mental Health (WMH) Survey. The EU Contribution to The World Mental Health (WMH) Surveys Initiative was funded by the Executive Agency for Health and Consumers (EAHC), European Comission (SANCO 2008-1308). G. Barbaglia is supported by Ministerio de Ciencia e Innovación Rio Hortega grant (CM 10-00099). C. García Forero is supported by Ministerio de Ciencia e Innovación FSE grant (JCI-2009-05486).

None.