This study has demonstrated that immigrant population attended in the north-eastern area of primary health care, compared with the natives, perceives a lower overall social support, which is also reflected in the four dimensions of social support (emotional, instrumental, social interaction, affective).
In a recent a study with the purpose to describe and compare immigrant (n = 46) and native Swedish patients (n = 46) in physical limitation, emotional state, social support and self-care , the authors observed that only the dimension of emotional social support was significantly worse (p = 0.048) in immigrants than Swedes, therefore immigrants had a greater need for emotional support that the natives. Another cross sectional study conducted in Spain compared quality of life of native Spanish's (n = 1009) and immigrants (n = 226) in the entire school population (12-18 year olds), excluded of the study those students who did not have enough knowledge of Spanish language to answer the questionnaire . The results show that Spaniards had significantly high social support (42.2%) than immigrants (33.5%) (p = 0.02).
Moreover, these differences in social support are also shown in the size of the support network, where natives have averaged more than 9 members providing support and immigrants do not reach the average of 6 persons. These figures are lower than those found in other Spanish studies conducted with immigrant populations (9-10 members)  and even conducted with the Spanish population (6 members) .
Adittionaly, studies in the field of psychosocial sciences have described the social support characteristics of the immigrant population  but there are fewer studies that compare the social reality of the immigrant population with the natives in the area of health sciences . To the best of our knowledge, in this respect our study would be an important contribution.
Regarding the possible variables associated with lack of perceived social support for immigrants, our data show that socioeconomic status, marital status, stress and self-reported health status are risk factors significantly associated with lack of social support. In a study conducted with people living in Canada, a strong correlation between the different dimensions of social support was identified; a positive correlation between physical health and perception of social support and a negative correlation between stress and low socioeconomic status were also found . These results are consistent with the findings of our study, since the worst monthly income levels and stress were associated with lower global social support and its four dimensions.
In general, a better self-reported health status in the Spanish population than in immigrants has been found. These results are consistent with findings in other studies in our country, showing that the immigrant population perceives that their health is worse than the Spanish population . As suggested in a previous research , one's perception of adequate social support is associated with better self-reported health status (excellent/very good health status reduce the prevalence of lack of social support by 54% [1-0.46]). So, the key factor to understanding social health in immigrants is social support.
Social networks have a direct effect on health, by the interaction with others and by social participation, which promotes healthier life behaviours and greater self-esteem and social competence . This may explain the association between health perception status and social support of our population as observed in a previous study .
Moreover, our study has found that women have better perception of social support than men, except in the instrumental support. These results are coherent with other studies that have described higher perceived social support in women, especially in emotional and affective dimensions [34, 35]. In several studies, the difference in perception of social support among both genders is due to various factors such as marital status, education level, age and socioeconomic conditions [36, 37]. In this sense, our data demonstrate a relationship between perceived social support and marital status, so that being married is positively associated with social support and being divorced negatively, confirming the findings of other studies [11, 38].
One of the sociodemographic characteristics with greater strength of association with the lack of social support is monthly income. A low level of income, below the minimum salary, is a stressful situation that favours the loss of social skills and increases stress. Both conditions reduce the availability of resources giving poor people less social support. This is also confirmed by Palomar and Cienfuegos results , in which after an analysis of variance and multiple regression, they found three socioeconomic levels (extreme poor, moderate poor and no poor) to explain the relationship between perception of social support and social and personal characteristics. They found that the extreme poor, compared with the other two groups, perceived little social support from family and friends, but like the moderate poor, perceived high levels of support from the people of the community in which they lived. For its part, the moderate poor, compared to extremely poor, reported a better support of children, and compared with non-poor, a better support of children and higher levels of perceived social support by neighbors.
Another study carried out by the Aragonese government (Spain) in 1993, which analyzed the interaction and social support in family units, eligible for economic assistance, only 47% of participants had social support, provided mainly by family and 6.9% lacked any type of social support. Women had more social interaction than men, as in our study, the relationship between socioeconomic status and social support was demonstrated .
The lack of social support is a variable with a known association with stress level, both are closely linked. Research has indicated that social support serves as a resource to minimize the negative effects of facing a stressful situation , such as migration to a foreign country. The results of this study are consistent with this line and suggest that people with a good perception of social support have lower level of stress in coping with stressful life situations.
Another variable of interest that can explain the degree of social support in the immigrant population is the time spent in the host country, a phenomenon already studied by Lin and Hung  that support the findings of our study, that shows the longer one lives in the country of migration the more social support it has. This relationship can be explained by several circumstances: a better understanding of the host language, a better sociocultural adaptation to the new society and by the immigrants' origin.
This study has some limitations. The main limitation is the exclusion of immigrants who did not have sufficient understanding of the Spanish language. To try to control it, it should have been conducted a transcultural adaptation to the different possible languages of immigrant assessment interview, although this was not done because the questionnaires used are validated for the Spanish population but not available in other languages. Therefore, not all immigrant groups in Spain are represented in the study population, as immigrants with poor language comprehension were not included in the study. Additionaly, it's possible that the desire (conscious or not) to please his physician with his participation in the study had included some patients with a poor understanding of the questions that have answers providing input and an unreliable, introducing an information bias in the study.
However, in general, the inmigrant population included for the study is representative of Spanish immigrants (according to statistics released by the Spanish Home Office and other public and private institutions), coming from latinoamerica, and less from Europe and North of Africa (Morocco). Another important limitation is the small size of some of the subpopulations (Asian, African and European). This forces a cautious interpretation and limits the ability to generalize the results to the examined groups of immigrant population.
Other limitation is the high number of women participants. This situation is due to the fact that women use primary health care services more often than men, as seen in the study of Esteban Peña and Health Survey of Castilla la Mancha (Spain), where the population of women studied was 67% and the use of primary health care services for women was 82%, respectively [42, 43].
In addition, the cross-sectional design of this study limits the possibility of establishing causal relationships between variables.
Despite the limitations, this research offers an insight into personal and social factors that the immigrant population expressed as an important perceived lack of social support.