Spatial clustering of codeine use and its association with depression: a geospatial analysis of nationally representative South African data

Abstract

Background: There is an alarming trend in sub-Saharan Africa in the use of codeine-containing pharmaceuticals, but its risk of common comorbidities, such as mental health, remains unquantified at a national-level. We investigated the relationship between codeine use and onset of depression in South Africa at a population-level. Methods: We used nationally representative panel data from the South Africa National Income Dynamic Study to investigate the relationship between exposure to codeine use in the community (i.e. residing in a codeine hotspot) and onset of depression. Geographical hotspots of codeine use were identified using Kulldorff spatial scan statistic in SaTScan. We quantified depression onset at Wave 5 (year 2017) between individuals residing inside and outside a codeine hotspot who were depression-free at enrolment (Wave 4: 2015) using generalised estimating equation (GEE) regression models. Results: We identified four statistically significant hotspots of codeine use, mostly located at the northern part of the country. Among 13,020 participants who were depression-free at enrolment, residing within a codeine use hotspot was significantly associated with higher subsequent onset of depression (aRR=1.21, 95% CI:1.08-1.44). Limitations: Data on diagnosis of depressive disorder were not available. Conclusion: South Africa, a resource scarce nation with chronically limited mental health services, is not spared from the global opioid epidemic and its impact on depression. Targeted scale-up access to agonist therapy to effectively treat (opioid) addiction in communities at risk for high exposure to codeine use could reduce the risk of subsequent mental health challenges.