The psychological impact of deprivation in regions affected by conflict: a multilevel analysis of a cross-sectional survey in the occupied Palestinian territory

BACKGROUND Deprivation is an important determinant of poor health. Locality can be key in understanding variation in deprivation across a population. This study aimed to analyse how different forms of deprivation affect mental health among Palestinians, and how they account for locality effects in the occupied Palestinian territory. METHODS We used multilevel modelling to analyse data from the Socio-Economic & Food Security Survey 2014 conducted by the Palestinian Central Bureau of Statistics, which had a sample size of 7827 adults representing the same number of households. The main outcome is a General Health Questionnaire (GHQ) score, in which a higher score signifies worse mental health. Deprivation variables include subjective deprivation, material deprivation, food deprivation, and political deprivation (which was measured by use of the human insecurity scale). For the analysis, we included data on experience of different stressors (economic, political, health-related, and weather-related stress) reported at the household level in the 6 months preceding the survey, and we controlled for demographic characteristics, including age, gender, education, wealth, and region. We also conducted a two-level random effects multilevel regression, with locality as a proxy for neighbourhood. FINDINGS The model indicates significant variance at the locality (neighbourhood) level. There is a significant association between poor mental health and subjective, economic, political, and food deprivation; health, economic, and political stressors; age, and being a woman. Education beyond secondary school level and wealth have a significant inverse association with poor mental health. Individuals who indicated that they felt somewhat or very deprived have significantly higher GHQ scores than individuals who indicated that they did not feel deprived (β=1·73 and 4·33 for those who felt deprived and who did not feel deprived, respectively, p<0·0001). Food consumption was inversely associated with GHQ score (β=-0·01, p<0·0001) and food insecurity was positively associated with GHQ score (β=0·19, p<0·0001). Political deprivation, and health-related, political, and economic stressors were significantly positively associated with GHQ scores (β=0·043, 0·23, 0·35, and 0·19 respectively, p<0·0001). Age (β=0·079, p<0·0001) and being a woman were positively associated with GHQ score (β=0·26, p=0·0040), whereas education beyond secondary school level was inversely associated with GHQ score (β=-0·54, p<0·0001). INTERPRETATION The findings that the mental health of Palestinians is associated with various forms of deprivation and stressors, provide further evidence that political and social factors are determinants of health. Correlated factors include both subjective and objective measures, and suggest that although material conditions are important, people's subjective experiences are also important. Feeling deprived is an important correlate of mental health. The community effect suggests that services (or lack thereof), checkpoints and blockades, political situations, and other factors that vary across localities, may influence mental health issues at the neighbourhood level. FUNDING This project is part of the study Re-conceptualising health in wars and conflicts: a new focus on deprivation and suffering funded by the Middle East Centre at the London School of Economics and Political Science.