|Home | About | Journals | Submit | Contact Us | Français|
Our study extends research on the social determinants of health by exploring the association between a new, potentially very significant dimension, cultural capital by type and self-rated health among low-income women living in outer Beirut, Lebanon.
Self-rated general health was assessed using household data from a cross-sectional survey of 1869 women, conducted in 2003. Three types of cultural capital were included: watching cultural TV programs, producing art (e.g., drawing, theatre performance) and consuming art or literature (e.g., attending exhibits, reading literary books). Associations between self-rated health status and types of cultural capital were assessed using odds ratios from binary logistic regression models.
With the exception of art production, lack of cultural capital increased the odds of self-perceived poor health status adjusting for socio-demographics and other risk factors. The adjusted odds ratios were 1.86 (95% CI: 1.07 to 3.22) for watching cultural TV programs and 1.52 (95% CI: 1.12 to 2.06) for consuming art. As expected, health risk factors, age, social support and community of residence were also associated with health status.
Two types of cultural capital were strong predictors of self perceived health status among women living in poor urban communities, regardless of social capital, income and other relevant risk factors.
The past few decades witnessed a heightened interest in the social determinants of health, culminating in the emergence of the sub-discipline of social epidemiology. Unlike traditional epidemiology, this new field takes a wider approach to health, explaining variations of disease and health status between individuals and groups by focusing on the social environment in which people live. Studies examining the impact of poverty, inequalities, gender, family structure, social networks and social capital on health and well-being have burgeoned in recent years. Our work seeks to extend this research by exploring an additional, potentially very significant, dimension of health in populations: cultural capital.
Coined by the late French sociologist, Pierre Bourdieu, cultural capital is a concept that has been utilized and debated mostly in the social science literature (1, 2). Following Bourdieu, we conceive of cultural capital as the ‘stock’ of artistic participation, aesthetic practices and knowledge individuals possess. Cultural goods have significant ‘symbolic values’ which serve to differentiate people into classes on the bases of aesthetic tastes and knowledge, reinforcing class boundaries and offering the elite an alternative set of goods for which to compete for ‘distinction’ (1). Much like economic capital, cultural capital contributes to the production and re-reproduction of class ‘distinction’ and hence inequality and social exclusion (3, 4). Building on this insight, a recent report using data from the urban health study found a significant association between mental distress and self-rated health and an overall index of cultural participation (5). However, some argue that the benefits of cultural participation are not primarily due to the nature of cultural activities themselves, but rather to the social networks that are accrued as a consequence of such activities (6, 7). This hypothesis was tested by Bygren et al (6) in a study that evaluated the impact of cultural practices on survival in a Swedish sample. They found a statistically significant relationship between cultural participation and survival, even when adjusting for social networks.
This study extends previous research by examining the association between types of cultural capital and self-rated health, adjusting for social capital, financial capital (income) and other relevant socio-demographic and health risk factors, among women in a context very different from that of most previous studies. Consistent with Bourdieu’s perspective, we hypothesized that cultural capital would be associated with self-assessed health regardless whether or not it involved social networking.
Our data included all ever-married women 15-59 years (n=1869) from the Urban Health Study (UHS) conducted in 2003 by the Center for Population and Health at the American University of Beirut. The UHS was based on an initial cross-sectional survey of 2,797 households selected randomly three poor communities in the greater Beirut: Naba’a, Hay El-Sollom and Burj El-Barajneh refugee camp. The communities located at the outskirts of Beirut were chosen purposefully by an expert group mainly on practical grounds including overall poverty conditions, lack of infrastructure, presence of rural-origin immigrants or displaced populations, ease of sampling, and proximity to Beirut proper. Although the three communities shared the same poor urban livings, they differed in their ethnic and religious make-up. While Naba’a and Hay El-Sollom housed a predominantly Lebanese population, Burj El-Barajneh consisted primarily of Palestinian refugees. As for religious affiliations, Naba’a was 80% Christian, and almost all of Hay el-Sollom and Burj El-Barajneh’s inhabitants were Muslims. Of the three communities, Burj El-Barajneh refugee camp, founded as a result of the 1948 Arab-Israeli war, was the most disadvantaged in terms of family income and other socio-economic indicators, owing to a long standing policy of exclusion from the public-sector services and the formal labour market. Until very recently, Palestinian refugees were not allowed to work outside the camps in many occupations including manual and clerical jobs. However, the United Nations Relief Works Agency for Palestinian Refugees (UNRWA) and other national and international non-governmental organizations (NGOs) provided aid and services to needy Palestinian refugees.
The instrument consisted of two questionnaires: one for the household and one for ever-married women aged 15-59 at the time of the survey. The household questionnaire was completed by face-to-face interviews with a proxy respondent (i.e., any adult living in the household) in the spring of 2002, and the questionnaire for ever-married women was completed in 2003 through interviews with the women in the sampled households. The data were collected by female interviewers after undergoing a thorough 3 week training in interviewing techniques, data collection and quality control procedures and administering the instruments. The overall response rates were 88.3% and 77.8% for the household and women questionnaires, respectively. The study was approved by the Institutional Review Board at the American University of Beirut.
Self-rated health was measured by a question asking women to rate their level of health as either very good, good, fair, bad, or very bad. Responses were then dichotomized into “good” (very good, good, fair=0) and “poor” (bad and very bad=1) for analyses. Our primary independent variables were measures of cultural capital. We distinguished between three types of cultural capital: watching TV (drama, comedy, dance, cultural or news) programs, producing art (playing a musical instrument, dance or play performance, making sculpture, drawing or painting), or consuming art (attending the movies, exhibits, or public speeches, or volunteering for organizations involved in the arts) or literature (reading books, poetry, or plays). Responses to any item in the three types of cultural capital were dichotomized (0=yes, 1=otherwise). Although our instruments included detailed questions on each aspect of these broad cultural ‘practices’ as suggested by previous examples (8, 9), we decided to group responses into broad types because of skewed distributions and low prevalence of cultural involvement in this population. It should be noted that one of the cultural items may involve extra-familial social networking, namely consumption of art.
Other relevant predictors of self-rated health included: two social capital variables, namely, membership in clubs or other non-political groups (0=Yes, 1=No) and receiving help from friends or neighbors during the month before the survey (0=Yes, 1=No); household income adjusted for household size using the Organization for Economic Cooperation and Development (OECD) equivalent scale (10), measured in quartiles of Lebanese Lira (1500 LL=$1); self-reported health problem in the past two months (0=Yes, 1=No); currently smoking (0=Yes, No=1); age groups (coded in 15 years intervals, ranging from 15-29 to 45-59); educational level completed (less than elementary, elementary, intermediate level or higher); marital status (currently married, previously married); and community of residence (Naba’a, Hey El-Sollom, Burj El-Barajneh).
For the analysis, descriptive distributions of the variables used in the analysis were first examined. Unadjusted and adjusted odds ratios from binary logistic regression models were then computed using Stata for Windows (Version 8.2) statistical program (11). We applied sampling weights to adjust for unequal probabilities of selection into the sample and non-response at the PSU level. Failure to use sampling weights in the analysis of data from a population-based survey like this one may lead to serious biases in the estimates.
Table 1 shows the univariate distributions of all variables used in the analysis. Overall, 24.1% of women perceived their health as poor. The vast majority (96%) of them watched cultural TV programs, 28.4% consumed arts or literature, and only 7.2% produced any piece of visual or performing arts.
Social capital was rather low in an international perspective: 6% of women were members of a non-political club or association, and 22.1% received help during the month before the survey. The income distribution in this sub-sample of ever-married women showed a poorer background than that of all households, with 20.2% falling in the highest quartile and 29.4% in the lowest quartile. The majority (68.4%) of women had elementary education (6 years), about a fifth (19.6%) of them had at least intermediate education, and 12% had had less than elementary education. Half (50.1%) of the women were in the prime reproductive ages, 30-44 years, and about a quarter (24.8%) were in the older age group, 45-59 years. A small proportion of women (9.9%) in this population were widowed, divorced or separated. The health status indicators showed a disadvantaged population, with 40.4% of women reported smoking, and about half (50.8%) reported having health problems in the past two months. Finally, Hay El-Sellom had a smaller proportion (22.2%) of ever married women compared to either Nabaa (34%) or Burj El-Barajneh (43.8%).
Women in our sample were generally more likely to report bad health status with decreased cultural participation (Table 2). Women who did not watch TV cultural program, consume art and produce art were 2.65, 1.85, 1.25 more likely to report poor self rated health compared to women who did. With the exception of art production, unadjusted odds ratios for the cultural capital items were statistically significant. After adjusting for all other relevant factors, women who did not watch cultural programs, consume art and literature, and produced art were respectively, 1.86, 1.52, 1.12 times more likely to report poor health as compared to other women lacking involvements in theses cultural activities. Interestingly, women who produced art reported similar health status to non-producers of art.
Furthermore, social support, very low income, illness in past two months, smoking, age, and community were significantly associated with poor health status in these disadvantaged women. The adjusted odds ratios for health risk factors and age were relatively strong as would be expected, ranging from 1.41 for smoking to 6.16 for health problems in the past two months. On the other hand, group membership, educational level, and marital status were not associated with self-rated health. For income, only those in the lowest quartile were significantly more likely to report poor health status, but we found no socio-economic gradient (in terms of income or education) in this sample owing perhaps to the overall socio-economic homogeneity of these largely poor, urban communities.
Although a growing literature has emerged on the links between arts and health in recent years (12), few studies examined the association between various types of cultural capital and health status (13, 14) or survival (15, 16). Our findings demonstrated the importance of disaggregating types of cultural activity when investigating the association between cultural participation and health status. Consuming art or literature (i.e., reading), watching cultural TV programs were significantly associated with reported health status; producing art had no significant association with the outcome variable in our sample. Evidently, cultural activities were associated with self-rated health regardless of whether or not they involved social networking. Thus, our hypothesis was partially supported as actual involvement in the production of the arts e.g., painting, drawing or playing musical instrument was not associated with self-rated health. One possible explanation of this finding is that passive activities such as reading, watching TV programs and ‘consuming’ arts were ‘practiced’ more often, and that it was perhaps the frequency of cultural participation that matters to health status. Cultural capital may therefore be viewed as a ‘perishable commodity’ (13), requiring sustenance over time to have a real effect.
Although our study primarily focused on women’s cultural participation, the findings also demonstrated the importance of other demographic, socio-economic and risky behavioural variables to health status. Our findings are consistent with previous studies, showing significant associations between self-rated health and age (17, 18), low income (17-19), smoking (17), reported health problems (18), and community of residence (5). The findings pertaining to social capital measures were mixed: social support, but not group membership, was associated with self-assessed health. Our findings were somewhat different from those of several previous studies demonstrating associations between selected components of social capital and self-rated health in various settings (17, 19). The area of residence effect is also of interest because Nabaa, a predominantly Christian community, was advantaged in terms of self-rated health as compared to Hey El-Sollom or Bourj El-Barajneh camp, Muslim communities, reflecting perhaps religious affiliations of the populations. The area effect does not however reflect deprivation because the model adjusted for income and educational levels. There are perhaps other unobserved contextual (as opposed to compositional) variables, such as the quality of services and associations, behind the observed Nabaa advantage.
Unlike previous studies however, our findings did not demonstrate a socio-economic gradient in self-rated health, and neither low education nor low/medium income was associated with health status. We argued that the lack of significant associations between education and income (save the lowest quartile) and health status in these three overwhelmingly poor neighbourhoods was perhaps due to the overall homogeneous economic standing of the three communities. In fact, the three communities fell in the lowest income bracket when measured on a national scale. Also surprising was the lack of significant association between marital status and self-assessed health owing perhaps to a small number of divorced or widowed women in our sample.
The significance of two important dimension of cultural capital in predicting self-rated health after adjusting for relevant conventional demographic, socio-economic and health risk factors, calls for explanation. Given the cross-sectional design of our study, reverse causation cannot be ruled out. Nevertheless, we consider possible pathways linking cultural participation to health that can be investigated further using alternative study designs. We suggest three possible pathways linking cultural participation to health status: (1) intermediary physiological stimulation and related psychosomatic responses to cultural activities; (2) cultural capital may serve as a marker of social status much like income or other material resources; and (3) cultural participation may be a proxy for women’s autonomy, particularly mobility, and may indirectly be associated with good health status. Although we briefly discuss each possible pathway separately, it is likely that two or more factors have a cumulative effect on perceived health status.
Recent research in the area of art and health demonstrated the possibility of certain intermediary emotive states during culturally-oriented activities (21, 22), neuroimmunological explanations of brain signals that change hormone levels (23), or other positive psychosomatic responses to the experience of cultural immersion (24). Although research in this area is nascent, a growing body of literature showed the healing potential of creative, arousing, emotionally, or intellectually engaging activities (25). It may be that cultural participation produce a more stimulating environment for women than social activities alone, impacting their psychological and physical health.
Cultural capital could also be conceived of as a dimension of social stratification, much like income, and hence contributes to the production and maintenance of social hierarchies (1, 26). Following Bourdieu’s study of the links between taste and class in France (1), social scientists investigated the links between cultural capital and a host of socio-economic outcomes, particularly educational achievement. Cultural capital, much like financial and social capital, is fungible (3) and some can invest in cultural ‘goods’ for accessing and accumulating economic gains. In other words, cultural capital is an important feature of socioeconomic status and social hierarchy. Wilkinson (27) argued that equality and better health are associated because the former leads to improvement in social cohesion and high self-esteem. Experimental evidence shows that social hierarchy can be associated with chronic stress, aggression, and coronary artery atherosclerosis (28). Here, we argue that cultural capital was associated with health status because it reflected social hierarchy in this impoverished context.
Finally, a plausible explanation for the association between cultural capital and health is that cultural participation is a proxy for women’s autonomy. Since there were no movie theatres or exhibition centres in the urban neighbourhoods of our study, women going to movies or exhibitions must have been able to leave their neighbourhood either accompanied by their husbands or a lone. It is widely reported that women’s autonomy, particularly mobility and decision-making power, is associated with women’s well being and with a range of reproductive behaviours, and may be particularly important in our contexts characterized by patriarchy and gender inequities.
Interventions involving cultural programs in disadvantaged urban communities may be an efficient way to improve health status, especially among women. However, further studies preferably based on longitudinal data should be undertaken before interventions are considered. Our study included a range of different items on cultural participation in addition to conventional risk factors from a large representative sample of women in a developing country, and this was perhaps its main strength. However, our study suffered from several shortcomings including the cross-sectional design, lack of some common risk factors such as obesity, physical inactivity, and the inclusion of women only from three disadvantaged communities in the sample. Although self-reporting of general health is known to be associated with mortality and morbidity (29), physical health status has many specific dimensions which were excluded in this study due to the lack of requisite data. Furthermore, while our social capital measures focused on social networking, they did not cover a whole range of other dimensions of the concept (30). Studies using longitudinal or experimental research designs as well in-depth case studies may be needed to confirm the associations found in this report.
This study was part of a larger multi-disciplinary research project on urban health sponsored by the Center for Research on Population and Health at the American University of Beirut, and supported by grants from the Wellcome Trust, the Mellon Foundation and the Ford Foundation.
Competing interests: None declared.