DISPLACEMENT AND HEALTH STATUS IN LOW INCOME WOMEN: FINDINGS FROM A POPULATION-BASED STUDY IN GREATER BEIRUT

Abstract

In this paper we examined the relationship between internal displacement, social support and self reported health status of ever married women in three disadvantaged urban neighbourhoods in Lebanon.

Methods

This study was based on data from a cross sectional survey conducted in 2003 on 1869 ever married women residing in three urban disadvantaged communities in the outskirts of Beirut, Lebanon. The outcome variable was Self rated Health (good/bad) as assessed by the women. The independent variables included ever displaced status, social support, demographic, health behaviour, and socio-economic factors. Descriptive statistics and bivariate associations were provided using Pearson's chi-square tests. Unadjusted and adjusted odds ratios were then obtained from binary logistic regression models.

Results

Displacement was a significant risk factor for poor self reported health (OR=1.67; 95% CI= 1.35-2.07). Adjusting for demographic and socioeconomic factors decreased the association between displacement and self reported health but the relationship remained statistically significant. Women with poor support from the family, friends and neighbours were more likely to have poor health status. However, not exchanging support with the family members (OR= 1.87; 95% CI = 1.13 – 3.12) was significantly associated with poor self reported health only among displaced women but not among those who were not displaced.

Conclusion

Displacement and social support were negatively associated with women's health status but family support may play an important role in improving the health status of displaced women and not non-displaced women.

Introduction

Recently, there has been an interest in studying the relationship between internal displacement and health status. Forced population movements especially internally displacement is occurring at astounding levels around the globe. The number of internally displaced persons has exploded during the past few years; approximately 27 million people have fled their homes worldwide in order to escape war and persecution and to search for food and shelter often in unfavourable and remote locations without crossing international boundaries.^1,2

The link between forced migration and health has, to date, been addressed mainly in terms of infectious diseases risk during and immediately after mass population displacement ^3-7. Studies on refugee women's health have been few, and, for the most part, focused on the reproductive health of women.^{6, 8} Furthermore, previous studies rarely attempted to compare refugee women's health with those of their host country counterparts. The International Centre for Migration and Health (ICMH) with the cooperation and help of the Institute for Public Health in Sarajevo conducted a survey in 1996 to describe the current health situation of displaced people in Bosnia and Herzegovina. About 19% of displaced people as compared to 11% in the non-displaced group said they had suffered from major physical pain, and felt that this pain had interfered with their normal activities. ⁹ In another study, based on a survey of a clinical population conducted in 2000 to displaced people in East Timore, about ninety percent of subjects rated their present health as average to poor. ¹⁰

On the other hand, social support has been considered a major coping mechanism for the displaced, and an important mediating factor between displacement and health status. Social support has been associated with positive health outcome in Lebanon. ¹¹ Among individuals experiencing health problems, those with adequate social support tend to receive advice, services, material or financial aid from others and thus are likely to benefit from better medical and non medical care. ¹² The majority of the studies that examine the association between social support and health among displaced/refugees focused mainly on the effect of social support on mental health. ^{13, 14} It is widely known that family cohesion may help to reduce health problems. For instance, migrant children in Norway who had a close, supportive relationship with parents, tended to have a lower incidence of emotional disorder. ¹⁵ In addition, in an attempt to study the relationship between chronic stressor and health and social support among Namibian refugees it was found out that when social support is low, the relation between stress and poor health outcomes is high. ¹⁶

The effects of internal displacement and resettlement in Lebanon on women have not been previously investigated. Yet, the issue is of public health concern since many of Lebanese citizens faced internal displacement. This study seeks to explore the relationship between internal displacement and self reported health status in three disadvantaged urban neighbourhoods in Lebanon. Based on the results of previous studies, we hypothesize that displaced women have poorer self reported health compared to non-displaced women, after adjusting for relevant demographic and socioeconomic factors. Furthermore, we hypothesize that social support is associated with the health status of displaced women but not for the non-displaced.

Context

Lebanon had to face the challenges of internal displacement during its 16 years (1975-1990) civil war. The Lebanese civil war was unique in its experience in its nature and duration. ¹¹ Among the countries in Southeast Asia and North Africa and Middle East, Lebanon ranks second after Cyprus for the duration of displacement for both internal displaced people and refugees with 21 years. ¹⁷

Displacements were not continuous, but occurred in separate periods of the civil war due to internal strife, Israeli military invasions and fighting between Syrian forces and Lebanese militias. Up to one million people were uprooted from their homes during the civil war. ¹⁸ The war has caused fragmentation of a pluralistic society into fairly distinct sectarian areas. Today the displaced are among the poorest echelons of the Lebanese society. It was estimated that over 50% of the displaced were not able to meet their basic needs and 12.5% lived in absolute poverty ¹⁹.

This study is based on three urban impoverished communities housing displaced population in the suburbs of Beirut. The communities are: Nabaa, Hay El Sellom, and Burj el Brajneh. The populations of these communities share common socio-economic characteristics such as low income and education. Furthermore, most of them were displaced from their place of origin due to the regional and Lebanese wars. They differ, however, in their religious and ethnic makeup: residents of Hay El Sellom and Burj el Brajneh are Muslims, while the majority of those in Nabaa are Christians; and inhabitants of Hey El Sellom and Nabaa are mostly Lebanese while in Burj el Brajneh are all Palestinians. The three communities were founded, or otherwise developed, as a result of war-related displacements. Nabaa has grown mainly by an influx of Christians from Mount Lebanon who were displaced during the civil war (1975+). Hay el Sollum was developed in the aftermath of the 1982 Israeli invasion, and continued conflict in south Lebanon and the lack of work opportunities in the Beqaa region. Burj el Brajneh camp includes refugees who originally came from Palestinian villages in waves during or after the 1948 war. Although the camp is urban in character, its inhabitants still maintain a rural way of life similar to their ancestors.²⁰

Findings from our study showed that almost 50% of the sampled populations in the three communities were displaced. The largest region of origin of displaced people was Mount Lebanon with 78.3%, followed by the South with 6.8%.

Methods

Data source

The data used in this study were taken from the Urban Health Survey conducted by the Centre of Research on Population and Health (CRPH) at the American University of Beirut. The sample was selected using a probability proportional to size sampling design. The communities were first divided into blocks using area maps. A sample of households was then drawn randomly from each of the selected blocks. The survey was conducted in two phases. Phase I, which took place between May and July 2002, collected household information including demographics, education, general health and insurance coverage, migration, labour, and income. In this phase, 2797 households were successfully interviewed. Phase II consisted of three individual questionnaires on the health of adolescents, elderly, and women and was completed during the spring of 2003. The present paper used information from the reproductive health questionnaire. A total of 1869 ever married women from these communities were successfully interviewed. The data were collected using face to face interviews by interviewers recruited locally from three communities and trained on the questionnaire and interviewing techniques by CRPH staff and study investigators.

Dependent variable

The dependent variable was self rated health (SRH). It was measured by asking women to respond to a single question (“How would you rate your health?”) by selecting one of the five possible responses: “very good”, “good”, “fair”, “poor” and “very poor”. In this study, self reported health is dichotomized into good (the three first alternatives) and bad (the remaining alternatives) health status²¹.

Independent variables

The independent variables included ever-displaced status, social support, demographic and health variables. The main independent variable, displacement, was measured by asking a straightforward question, “Have you been ever displaced”?, with a dichotomous answer categories, yes (=1) and no (=0).

Social support was measured using three variables: exchanging favours with neighbour, exchanging help with family, exchanging help with friend or colleague. Each variable was based on two questions/items: “In the past month, did you help any of your friends or colleagues/family member or relative/neighbour?” “In the past month, did any of your friends or colleagues/ family member or relative/neighbour help you?”) Then, these answers were dichotomized by assigning a “Yes” (reference category) to the women who answered positively to the give and receive questions, and those who answered negatively to either question or to both were converted to a “No” answer.

The demographic variables included age, marital status and number of children. In this study age was classified into three 15 year categories “15-29” (reference category) and “30-44” and “45-59”. Marital status was defined as either currently married (reference category) or not (widowed, divorced or separated) ^{22, 23}. Number of children was categorized into four categories: childless women, women with one child (reference category), women with 2 to 6 children, and women with 7 or more children.

Socioeconomic variables included: household income, labour force participation, and educational level. Household income was initially classified into quartiles. In this study income was dichotomized into less than 3600 and more than 3601 (reference category) to test if any variations in the outcome exist between the 25% most deprived women (poor) and the rest, reflecting poverty status. Thus our measure is consistent with the widely used 50% of the median measure of relative poverty. ²⁴ Educational status was measured by asking the following question “what is the highest educational level that you have completed successfully?” and it was divided into four categories: “none”, “elementary”, “intermediate”, “secondary or more” (reference category). Labour force participation was defined according to the International Labour Organization (ILO) framework, distinguishing between those “in labour force” and “not in labour force” (reference category) based on a series of questions about work last week.²⁵

Health risk variables included: smoking cigarettes and visiting any health care centre during the past two months. The variable smoking cigarettes was measured by asking the following question “do you smoke cigarettes?” and included two groups “yes” or “no” (reference category). Visiting health care services at the time of the interview was divided into “yes to a health centre”, “yes, to a dispensary”, “yes, to a special clinic”, “yes, to a hospital”, “yes, to the emergency of a hospital”, “No”. In this study, it was dichotomized into “Yes” encompassing the six categories of “yes” answers and “No”. This is so because we are only interested in this study if the women were suffering from any physical health problem, indexed by visiting any health care centre. Finally, the community of residence was used to index the structural contexts associated with displaced women.

Data Analysis

All data analyses were conducted using SPSS for windows (version 13). Bivariate analysis was carried out to identify variables associated with SRH. All variables which showed significant bivariate associations with self rated health by chi-square test were entered into logistic regression. Unadjusted odds ratios with SRH were first calculated for each independent variable. Multivariate analysis was then performed using logistic regression models to investigate the association between SRH and displacement status and to determine whether social support variables had significant associations with SRH. The first model studied the main independent variable, isplacement, adjusting for all covariates. In the second model, we stratified the full model by displacement and studied each one individually adjusted for all covariates. The full model was tested for interactions. To study the possibility that the effect of number of children is modified by age, we tested the interaction between age and number of children and therefore number of children was excluded from the model. Also labour force participation was removed from the full model because it was not significant with SRH at the bivariate level and was highly related to marital status.

Results

Table 1 presents the characteristics of the women in the sample. Almost one out of every four women rated their health as bad (24%). Displaced women reported a higher proportion of poor SRH (28.8%) compared to non-displaced women (19.5%). The prevalence of social support (exchange help with their relatives, neighbours, and friends) was relatively low, ranging from 10% to 18.2%. Surprisingly, social support was fairly similar between the displaced and non-displaced women. Half of the interviewed women fell in the age category of 30-44 for both groups. However, the displaced women were older with 32.4% of them were 45-59 years old while only 18.3 % of non-displaced fell in that same category. The vast majority of women (90%) were currently married with a higher percentage (12.7%) of widowed or divorced among displaced women than the non-displaced (7%). Most of the participants (74.3%) have two to six children with a higher percentage (19.4%) of women having zero to one child among non-displaced than among the displaced (13.3%). The majority of displaced women lived in Burj el Brajneh while a larger number of the non-displaced lived in Nabaa. As for income, the percentage of poor displaced women was higher (28.9%) than the non-displaced poor women (23.6%). The majority of participants had an elementary level of education (42%) but the non-displaced had a lesser percentage of women with no education at all (25.6%) than the displaced women (33.8%). As for labour force participation, those who are in labour force were mostly displaced women (22.7%) than the non-displaced (16.3%). 63.2% of the participants did not smoke and only 36.8% smoked. The majority of the women (69.3%) did not visit any health care centre, with a slighter higher increase in the percentage of visitors (33.4%) among displaced than non-displaced women (28%).

Table 1.

Demographic, socioeconomic and health behavior characteristics of the women

	Total		Displaced		Non-
Variable	N	%	N	%	N	%
Self rated health
Good	1451	75.9	675	71.2	769	80.5
Bad	461	24.1	273	28.8	186	19.5
Social support
Reciprocity (Family)
Exchange	280	14.7	145	15.3	135	14.2
No exchange	1625	85.3	801	84.7	815	85.8
Reciprocity (Neighbours)
Exchange	348	18.2	162	17.1	184	19.3
No exchange	1558	81.8	784	82.9	767	80.7
Reciprocity (Friends)
Exchange	190	10.0	98	10.3	91	9.6
No exchange	1715	90.0	848	89.7	859	90.4
Demographic
Age
15-29	432	22.6	153	16.2	276	28.9
30-44	989	51.7	485	51.4	502	52.7
45-59	485	25.4	306	32.4	175	18.3
Marital status
Married	1722	90.1	828	87.3	887	93.0
Widowed/divorced	189	9.9	120	12.7	67	7.0
Number of children
0-1	315	16.5	126	13.3	185	19.4
2-6	1420	74.3	707	74.6	709	74.3
7+	177	9.2	115	12.1	61	6.4
Community
Nabaa	651	34.0	260	27.5	389	40.8
Hey el Sellom	424	22.2	182	19.2	238	25.0
Burj El Barajneh	837	43.8	505	53.3	327	34.2
Socioeconomic
Income
3601+	1412	73.9	673	71.1	730	76.4
<3600	499	26.1	274	28.9	225	23.6
Education
None	564	29.5	318	33.8	242	25.6
Elementary	802	42.0	372	39.5	429	45.2
Intermediate	334	17.5	169	17.9	164	17.2
Secondary or more	198	10.3	82	8.7	114	12.0
Labour force participation
In labour force	372	19.5	215	22.7	155	16.3
Not in labour force	1538	80.4	732	77.3	799	83.7
Health Behavior
Smoking cigarettes
Yes	702	36.8	354	37.4	347	36.5
No	1204	63.2	592	62.6	604	63.5
Visit to health care centre
Yes	586	30.7	631	33.4	267	28.0
No	1325	69.3	316	66.6	687	72.0
Total	1912	100.0	948	100.0	955	100.0

^*Note: Weighted sample

Table 2 shows the unadjusted and adjusted estimated OR's with 95% CI for self reported health status. Displaced women were more likely to perceive their health as poor than the non-displaced (OR= 1.67; CI= 1.35-2.07). In terms of social support, reciprocity with families (OR=1.80; CI= 1.29-2.53) and friends (OR= 1.48; CI= 1.00-2.17) and neighbours (OR=1.80; CI= 1.33-2.46) were important risk factors associated with poor SRH. Consistent with the literature, older people in the sample perceived their health poorer (OR= 4.65; CI= 3.34-6.48) than younger women. Being widowed/divorced (OR= 2.03; CI= 1.480-2.781) was significantly associated with an increased risk of poor health. Women living in Hey el Sellom were more likely to report their health as poor (OR= 1.82; 95% CI= 1.36-2.44) followed by Burj el Brajneh (OR=1.80; 95% CI= 1.40-2.31) compared to Nabaa residents. Also, as expected being a woman with lower education (OR= 4.78; CI= 3.01-7.60) and lower income (OR= 1.53; CI= 1.21-1.93) was associated with an increased risk of poor health. Smokers perceived their health poorer (OR= 1.52; CI= 1.23-1.89) than non smokers and those who visited the health care centres (OR= 3.21; CI= 2.58-4.00) were more prone to report bad health status than those who didn't visit the centres.

Table 2.

Odds ratio for poor self rated health from logistic regression models

Variable	Unadjusted			Adjusted
	OR	P value	95%CI	OR	P value	95%CI
Displacement
No	1.00			1.00
Yes	1.67	0.00	1.35-2.07	1.30	0.04	1.02-1.65
Social support
Reciprocity (Family)
Exchange	1.00			1.00
No exchange	1.80	0.00	1.29-2.53	1.69	0.00	1.15-2.47
Reciprocity (Neighbours)
Exchange	1.00			1.00
No exchange	1.80	0.00	1.33-2.46	1.63	0.01	2.00-2.43
Reciprocity (Friends)
Exchange	1.00			1.00
No exchange	1.48	0.04	1.00-2.17	1.08	0.77	0.65-1.80
Demographic
Age
15-29	1.00			1.00
30-44	1.63	0.00	1.19-2.25	1.49	0.02	1.06-3.00
45-59	4.65	0.00	3.34-6.48	3.14	0.00	2.31-4.62
Marital status
Married	1.00			1.00
Widowed/Divorced	2.03	0.00	1.48-2.78	1.02	0.90	0.71-1.48
Community
Nabaa	1.00			1.00
Hey el Sellom	1.82	0.00	1.36-2.44	2.02	0.00	1.46-2.80
Burj el Brajneh	1.80	0.00	1.40-2.31	1.61	0.00	1.18-2.19
Socioeconomic
Income
3601+	1.00			1.00
< 3600	1.53	0.00	1.22-1.92	1.43	0.01	1.09-1.90
Educational level completed
None	4.78	0.00	3.01-7.60	2.24	0.00	1.36-3.68
Elementary	1.96	0.00	1.24-3.12	1.52	0.09	0.94-2.47
Intermediate	1.13	0.66	0.66-1.92	0.94	0.82	0.54-1.64
secondary or more	1.00			1.00
Health Behaviour
Smoking cigarettes
No	1.00			1.00
Yes	1.52	0.00	1.23-1.88	1.45	0.00	1.14-1.85
Visit to health care centres
No	1.00			1.00
Yes	3.21	0.00	2.58-4.00	2.87	0.00	2.26-3.64

^*Note: Weighted sample

After adjusting for socioeconomic and demographic indicators the odds ratios for displacement decreased 1.67 (95%CI: 1.35-2.07) to 1.30 (95% CI= 1.01-1.65) but it remained significant. Similarly the risks of low social support from the family (OR= 1.69; 95%CI= 1.15-2.47) and the neighbours (OR=1.63; CI= 2.00- 2.43) decreased but remained significant. In addition, controlling for other factors attenuated but did not eliminate the relationship between eldest age (OR=3.14; 95% CI=2.31-4.62), having no education (OR= 2.24; 95% CI=1.36-3.68), lower income (OR=1.43; 95%CI=1.09-1.90), smoking (OR=1.45; 95% CI=1.14-1.85) and visiting health care centres (OR= 2.87; 95%CI=2.26-3.64) and self rated health. However, when being a non working women or widowed/divorced were entered into the full model these associations became non- significant.

Table 3 reports the risks of poor self reported health among the displaced and non-displaced separately. We stratified by displacement status to examine if social support was associated with self rated health in displaced but not in the non-displaced groups. Results showed that for displaced women, not having social support from the family (OR= 1.87; CI= 1.13-3.12) and socioeconomic variables such as being poor (OR= 1.47; CI=1.02-2.12), being non educated (OR= 2.84; CI=5.76) emerged to be the strongest factors for SRH. As for non-displaced women, no form of reciprocity was significantly associated with self reported health status in these disadvantaged neighbourhoods. For them, demographic and lifestyle characteristics such as being middle aged (OR= 3.28; 95% CI= 1.84-5.87), visiting the health care centres (OR=3.85; 95% CI=2.68-5.54) and smoking seem to be of greater importance to health than the social relations within these communities. For both groups (displaced and non- displaced), age, smoking, visiting the health care centre and community were significantly associated with poor health. In our study, people living in Hey el Sellom were more prone to rate their health as poor. Finally, being widowed or divorced women was no longer significant after controlling for the other variables even though it was highly significant at the unadjusted level.

Table 3.

Adjusted odds ratio for poor self reported health among displaced and non-displaced women

Variables	Displaced			Non Displaced
	OR	P value	95 % CI	OR	P value	95% CI
Social support
Reciprocity (Family)
Exchange	1.00			1.00
No exchange	1.87	0.02	1.13-3.12	1.55	0.15	0.86-2.79
Reciprocity (Neighbours)
Exchange	1.00			1.00
No exchange	1.63	0.09	0.92-2.86	1.67	0.08	0.94- 2.96
Reciprocity (Friends)
Exchange	1.00			1.00
No exchange	1.20	0.60	0.59-2.47	0.95	0.89	0.45-1.98
Demographic characteristics
Age
15-29	1.00			1.00
30-44	1.39	0.20	0.84-	1.57	0.07	0.96-2.55
45-59	2.91	0.00	1.70-4.97	3.28	0.00	1.84-5.87
Marital status
Married	1.00			1.00
Widowed/Divorced	0.96	0.85	0.60-1.50	1.09	0.80	0.58- 2.04
Community
Nabaa	1.00			1.00
Hey el Sellom	2.30	0.00	1.43-	1.82	0.01	1.15-2.86
Burj el Brajneh	1.69	0.02	1.10-	1.60	0.04	1.01-2.54
Socioeconomic
Income
3601+	1.00			1.00
< 3600	1.47	0.04	1.02-2.12	1.35	0.20	0.89-2.04
Educational level completed
None	2.84	0.00	1.40-	1.62	0.20	0.79- 3.32
Elementary	1.70	0.14	0.85-	1.34	0.40	0.68- 2.67
Intermediate	0.67	0.33	0.29-	1.30	0.50	0.60- 2.80
Secondary or more	1.00			1.00
Health Behaviour
Smoking cigarettes
No	1.000			1.00
Yes	1.37	0.05	0.99-	1.68	0.00	1.16- 2.44
Visit to health care centres
No	1.00			1.00
Yes	2.31	0.00	1.68-	3.85	0.00	2.68- 5.54

^*Note: Weighted sample

Discussion

This study examined the associations between SRH and displacement and social support in a poor urban environment. Consistently with our expectations, displacement was a significant risk factor for poor self reported health in women after adjusting for demographic and socioeconomic factors. The second hypothesis was also supported as social support was a significant factor for health status among the displaced but not among the non-displaced.

Our findings that displaced women had worse self reported health status than the non-displaced counterparts are consistent with other studies. ^9,26,27 A study implemented in Jordan to assess the living conditions of Palestinian refugees and displaced as compared to non-displaced people showed that refugees and displaced are less satisfied with their health than the other group.

Our findings show that the factors that determine the self rated health of displaced women are different from those of the non-displaced. We found that for displaced women, not having social support from the family, the socioeconomic variables such as being poor, being non-educated emerged to be the strongest factors for self reported health. Generally speaking, the concept of being a displaced or refugee is associated with family disruption and for whom the traditional ties to relatives are weakened. ²⁶ This also to a certain extent the case among displaced and refugees in Lebanon. However the fact that displaced moved in large numbers probably explains why disintegration of families does not seem to be common here. In fact, 27.4% of displaced people have more than ten relatives (outside the household) living nearby whereas 22.4% of non-displaced have the same number of relatives.

Surprisingly, income was only significant for the displaced group. A possible explanation would be that in fact, the distribution of income by quartiles among displaced seem to be more homogenous and almost equal among the different categories (< 3600LB: 26.4%; 3600-6000LB: 24.7%; 6000-10200LB: 24.4%; >10200LB: 24.4%) whereas non- displaced people's were more dispersed across the different categories of income with the higher percentage for those earning between 6000 and 10200 L.B. (< 3600LB: 20.6%; 3600-6000LB: 22.5%; 6000-10200LB: 31.8%; >10200LB: 25.1%) Therefore it might be that non-displaced women are more able to use health care services when needed than others, whereas displaced people prefer to save their small revenues for other necessities of life.

Also interestingly, having no education was only significantly associated with health for displaced groups. The percentage of being in labour force among the two groups increased with the level of education ⁹ and reached its highest among women with secondary education, 42.2% among displaced and 27.8% among non displaced, whereas 19.2% among displaced and 13% among non displaced, of those without any educational background were in labour force. In addition, it has been found that 26.7% of displaced women and 20% of non displaced work in elementary occupations. This higher percentage of labour force participation among non-educated displaced women may have negative consequences on health.

As for non-displaced women, no form of reciprocity was significantly associated with self reported health status in these disadvantaged neighbourhoods. For them, demographic and lifestyle characteristics such as being middle aged, visiting the health care centres and smoking seem to be of greater importance to health than the social relations within these communities.

The study has some limitations. The main limitation is its cross-sectional nature makes it difficult to draw inferences about causal pathways between displacement, socio demographic and economic conditions and self reported health. Next, there is the possibility of bias from self reported data. However, self reported health status is a useful indicator of the health conditions of a population, and the stability over time of self reported health status was demonstrated, as well as the fact that this kind of subjective health assessment was a valid indicator of health in middle aged populations and could be used in cohort studies. ²⁸ Moreover, the test-retest reliability of self rated health status has been good. ²⁹

This study examined the health status of women only in three impoverished communities and cannot be generalized to the whole community. Some other limitations arise because the questionnaire didn't fully address certain topics such as biomedical risk factors.

Conclusions

This study has shown that displaced women were more likely to perceive their health status worse than non-displaced even after controlling for demographic, socioeconomic and lifestyle characteristics. In addition, reciprocity with families and neighbours were significantly associated with poor self reported health among women in disadvantaged populations. However, family support was the only form of social support was significantly associated with self rated health for displaced women. As for non-displaced people no form of social support was associated with health status.

In addition, variables that had significant associations with perceived health status varied between displaced and non-displaced women. Further in depth studies are needed to fully understand the link between displacement, social support and other outcomes.