Over the past two decades, important changes have taken place in the worldwide economy and consequently in labour markets [1, 2]. One of these changes has been the conversion of full-time to temporary or flexible contracts and reductions in social protections for workers [1–4]. Additionally, the demographic profiles of many European countries have changed as they are receptors of immigrant workers primarily from low-income countries [5–7]. There are an estimated 30 million immigrants living in the European Union (EU-27), of which 19.5 million were citizens of countries outside of the EU27. Germany, Spain, United Kingdom, France and Italy are the countries with the greatest proportion of foreign-born population, according to Eurostat data (2008) [6]. In 2008, 17% of EU-27 immigrants were located in Spain [6]. The proportion of immigrants in Spain that come from low-income countries has increased from 0.7% to 8.3% between 1996 and 2008 [8]. Nevertheless, it is important to consider the fact that this figure may be biased downward, since those immigrants in an irregular situation (undocumented) are less likely to be included in standard registries and therefore may not appear in the official statistics data. In Spain, it is estimated that 2% of all immigrant workers could be in this situation of invisibility [3, 9].

This new context in which immigration takes place makes it necessary to analyse the employment and working conditions of the immigrant population and the impact on health as a part of the research on determinants of health inequalities. Preceding studies have established that immigrant workers may face precarious work, employment difficulties, low salaries and lack of access to social benefits [10, 11], together with increased risk for occupational health and safety problems [12–15].

In spite of indicators that show worse employment conditions, previous studies have observed that immigrants are less likely to miss work for a health problem when compared with the autochthonous population after adjustment for age and the duration of sick leave [16]. Among the explanations suggested by the literature include: a) the healthy migrant effect, by which immigrants arrive in host countries with relatively good physical and mental health status due to selection as part of the migratory process [17], and b) presenteeism, whereby workers go to work despite being sick [18]. However, relevant information concerning the determinants of presenteeism is still relatively sparse [19–22]. The aim of this study is to assess the extent of sickness presenteeism in a sample of Spanish-born and foreign-born workers according to personal and employment characteristics and taking into account time living in Spain as a proxy for the healthy migrant effect.

Among those workers reporting good self-perceived health, immigrants living in Spain less than 2 years were more likely to report presenteeism in comparison to their Spanish counterparts (aOR 2.04 95%CI 1.38- 3.00). Among workers reporting good mental health, immigrants (living less than 2 years in Spain) were more likely to report presenteeism than Spanish workers (aOR 1.66 95%CI 1.10- 2.51); in this same subcategory of workers, immigrants reporting poor mental health and living in Spain more than 2 years were more likely to report presenteeism than Spanish workers (aOR 1.67 95%CI 1.05- 2.92).

A higher prevalence of sickness presenteeism was found in foreign-born workers living in Spain less than 2 years, after adjusting for potential confounders, and these differences are relevant in all the different subcategories of workers considered according to sociodemographic, employment and health status. To the best of our knowledge, this is a first approach to the assessment of sickness presenteeism in immigrant workers.

The higher sickness presenteeism in immigrant workers in comparison with the Spanish population may be explained by greater exposure to precarious employment conditions among immigrants [11, 15, 26, 27]. Interestingly, we observed more presenteeism in immigrants with temporary contracts than in immigrants with permanent contracts (in contrast to what was observed in Spanish workers). This situation highlights the relative importance of employment conditions (unfair or informal) in an already vulnerable population. This means that worse working and employment conditions are the pathway through which immigrants experience more health problems, especially after their second year in Spain. Therefore we would expect more sickness absence due to these health problems, but what we observe is more sickness presenteeism [28]. In a large proportion of cases, immigrant workers feel unprotected and vulnerable in the workplace, with less capacity to demand their rights [11, 13]. In other cases, their lack of knowledge of the social protections and difficulty navigating the health system may also be a contributor [29]. In addition, it is important to consider other cultural factors that vary according to the countries included in the sample. For instance, different perceptions of risk and self-perception of health and illness may exist between native and foreign-born workers [30]. Furthermore, we found in previous qualitative research that self-perception of health among immigrants is linked to life circumstances that prioritize economic sustainability of the self and family unit [27].

Other studies carried out in Spain have shown that sickness absence is higher in autochthonous workers than in immigrants [16, 31], and the same situation was found in our sample. A higher absenteeism was observed in Spanish workers in comparison with their foreign-born counterparts through additional analysis measured through the same questions from the survey.

The results of our study are not easily comparable with other research, due to the lack of studies of presenteeism in immigrants or that compare native-born and foreign-born workers. In the case of the few studies of presenteeism in the working population, the general prevalence found in our study- considering both immigrant and Spanish workers- is similar to that of one European study [22] in which 48% of men and 51% of women had to attend work at least once, despite being sick. In another study [19] higher rates of presenteeism were shown: 73% for women and 67% for men. In any case, because of the complexity of the topic and the indicators used in those studies, the comparison and the derived conclusions must be interpreted carefully, especially in terms of international comparisons. It is important to consider the key points suggested by the authors, for instance, the relationship between presenteeism and low monthly income and occupational social class and employment.

Despite these limitations, it is very relevant to study sickness presenteeism due to its social, economic and health implications [19, 32, 33], which include its association with loss of productivity, its association with poor organizational climate in the workplace and its possible relationship with future sickness absenteeism [20, 32, 34–36]. One of the advances made by this research is that we have evaluated presenteeism from a perspective of the social determinants of health, analysing how the inequalities faced by immigrants originate give rise to situations of vulnerability in host countries [15, 37]. Presenteeism is an emerging concept that can be viewed as an inequalities index.

Interpretation of these results should take into account the study's limitations. The data for this study was collected under quota sampling methodology, and the potential effects on the ability to extrapolate findings to a larger population should be noted. The hidden nature of immigrant populations, especially the undocumented subgroup, has prevented a completely accurate description of the population of interest, prohibiting random probability sampling. We do not have information concerning non-response across the target groups. Response rates in other research in similar populations vary. For example, in the Angeles County Health Survey [38], the use of random digital dialing and computer-assisted telephone interviews in different languages obtained a response rate of 18%, and the Vietnamese American Community Issue Survey, a bilingual, computer-assisted phone interview of 500 Vietnamese American adults (18 and older) examining participation in political protests, obtained a response rate of 25.5% [39]. The slightly lower response rate of Spanish workers in this case could be due to their being targeted more often for surveys. The higher response rate among immigrants could be due to the fact that the immigrants might feel a sense of solidarity with the workers conducting the survey, who are immigrants themselves, and their desire to voice their opinions.

It should be mentioned that there is the possibility of misclassification in the variable presenteeism. The variable assumes that sick workers that did not miss work during the time period in question is due to presenteeism and not a possible lack of employment. In other cases, if health problems are mild, workers do not need to miss work. More research is needed to standardized items to measure sickness presenteeism. The cross-sectional design of the study avoids assessing causality in the observed relationships. Prospective studies following a cohort of foreign-born and Spanish-born workers are needed to overcome these limitations. It is possible that there are some unobservable individual characteristics that have an influence on both decision to migrate and the subsequent prevalence of sickness presenteeism. Nevertheless, considering the characteristics of the sample, we think that the interest variables would be distributed randomly in Spanish and foreign-born populations and this situation would confirm our main findings. Likewise, people with an acceptable level of Spanish language sufficient to answer the questions included in the questionnaire were selected, and for that reason this study also leaves out other immigrant groups. There are immigrant groups in Spain proceeding from sub-Saharan countries, Asia and Oceania, which were not included in this study due to technical difficulties and economic resource limitations. Also, it is important to include in successive analyses workers in other age groups than those included in this study.

The ITSAL survey was conducted at the end of an economic cycle characterized by high rates of employment. In fact, according to the Active Population Survey -EPA-, the employment activity rates of the immigrant population are higher the rates for the native-born in the four community regions and the cities covered by the survey [40]. For that reason, it will be interesting to see how the situation changes as economic dynamics impact a job market characterized by higher unemployment and lower job stability. As such, we suppose that the rate of presenteeism could be higher now than when the original study took place.

It is important to consider that the models do not control for other labor and working conditions in the cities where the study was carried out. For that reason it is important to conduct further research about the impact of specific labour market conditions on health conditions in Spanish and immigrant populations by means of a multilevel approach. Also, complementary analyses are currently being carried out by the research team focused of the impact of the economic crisis in European Countries and particularly in Spain. Our focus is on how economic changes have impacted the employment, working and health conditions in immigrants compared with the Spanish-born population.

The higher presenteeism found in the immigrant population when compared with the Spanish-born population, and the demographic and working conditions that may be related to this finding, suggest inequalities that require discussion and debate in the field of public health. The implementation of specific policies to reduce social inequalities in health is clearly linked to developing strategies related to protecting the health of immigrant workers. Immigrant workers should be entitled to the same standards of social security, and of health and safety in the workplace, that Spanish workers enjoy. In addition, further study should explore ways to extend sickness leave and effective social protection to all workers in Spain as a potential strategy for preventing sickness presenteeism.