Investigation of the marked and long-standing spatial inhomogeneity of the Hungarian suicide rate: A spatial regression approach

BackgroundIn the last century Hungary had astonishingly high suicide rates characterized by marked regional within-country inequalities, a spatial pattern which has been quite stable over time.AimsTo explain the above phenomenon at the level of micro-regions (n=175) in the period between 2005 and 2011.MethodsOur dependent variable was the age and gender standardized mortality ratio (SMR) for suicide while explanatory variables were factors which are supposed to influence suicide risk, such as measures of religious and political integration, travel time accessibility of psychiatric services, alcohol consumption, unemployment and disability pensionery. When applying the ordinary least squared regression model, the residuals were found to be spatially autocorrelated, which indicates the violation of the assumption on the independence of error terms and - accordingly - the necessity of application of a spatial autoregressive (SAR) model to handle this problem. According to our calculations the SARlag model was a better way (versus the SARerr model) of addressing the problem of spatial autocorrelation, furthermore its substantive meaning is more convenient.ResultsSMR was significantly associated with the “political integration” variable in a negative and with “lack of religious integration” and “disability pensionery” variables in a positive manner. Associations were not significant for the remaining explanatory variables.LimitationsSeveral important psychiatric variables were not available at the level of micro-regions. We conducted our analysis on aggregate data.ConclusionOur results may draw attention to the relevance and abiding validity of the classic Durkheimian suicide risk factors - such as lack of social integration - apropos of the spatial pattern of Hungarian suicides.