Spatial structure of depression in South Africa: A longitudinal panel survey of a nationally representative sample of households

Diego F Cuadros,Alain Vandormael,Rob Slotow,Jonathan K Burns,Andrew Tomita,Frank Tanser

doi:10.1038/s41598-018-37791-1

Abstract

Wider recognition of the mental health burden of disease has increased its importance as a global public health concern. However, the spatial heterogeneity of mental disorders at large geographical scales is still not well understood. Herein, we investigate the spatial distribution of incident depression in South Africa. We assess depressive symptomatology from a large longitudinal panel survey of a nationally representative sample of households, the South African National Income Dynamics Study. We identified spatial clusters of incident depression using spatial scan statistical analysis. Logistic regression was fitted to establish the relationship between clustering of depression and socio-economic, behavioral and disease risk factors, such as tuberculosis. There was substantial geographical clustering of depression in South Africa, with the excessive numbers of new cases concentrated in the eastern part of the country. These clusters overlapped with those of self-reported tuberculosis in the same region, as well as with poorer, less educated people living in traditional rural communities. Herein, we demonstrate, for the first time, spatial structuring of depression at a national scale, with clear geographical ‘hotspots’ of concentration of individuals reporting new depressive symptoms. Such geographical clustering could reflect differences in exposure to various risk factors, including socio-economic and epidemiological factors, driving or reinforcing the spatial structure of depression. Identification of the geographical location of clusters of depression should inform policy decisions.

Highlights

Mental health is a major global public health concern[1,2], with the Global Burden of Disease Study highlighting the full extent of depressive illness as a high impact disorder that is commonly comorbid with other mental and physical diseases, and that has major public health implications[3]
The study of the spatial structure of mental disorders at this large scale is important to match with and inform public health interventions or investments for improvement planned at a national or provincial level. Against this background, using a large ongoing national representative survey conducted in South Africa, we present the first assessment of the spatial structure of depression at national level
Using state of the art methodology for clustering detection, and data from a nationally representative survey conducted in different years in South Africa, we identified several geographical ‘hotspots’ of existing cases of depression in the northern and eastern parts of South Africa

Summary

Introduction

Mental health is a major global public health concern[1,2], with the Global Burden of Disease Study highlighting the full extent of depressive illness as a high impact disorder that is commonly comorbid with other mental and physical diseases, and that has major public health implications[3]. Spatial epidemiology has emerged as a promising approach for understanding the distribution and patterns underlying the processes and drivers of an epidemic This approach has been extensively used to study communicable and non-communicable diseases, and could provide important insights into mental health disorders, such as schizophrenia and depression[20,21,22]. These studies highlight the role of spatial epidemiology in understanding the contextual determinants of mental disorders Such approaches could support the development of strategic programs that are required to promote mental health, prevent mental illnesses, reduce treatment gaps, and develop and maintain effective and safe mental health services in those areas where they are needed most. The study of the spatial structure of mental disorders at this large scale is important to match with and inform public health interventions or investments for improvement planned at a national or provincial level

Methods

Results

Conclusion