Logo Lanfrica

valentineghanem-bit/alcohol-ncd-suicidality-ghana

Domain:

healthcare

Record type:

paper
Creator:
val
Host:
Alcohol Consumption Does Not Track Suicide or NCD-Mortality Trends in Ghana: A National Trend-Divergence Analysis and Two-Dimensional District Vulnerability Classification | Ghana | 2000-2022 # Alcohol Consumption Does Not Track Suicide or NCD-Mortality Trends in Ghana: A National Trend-Divergence Analysis and Two-Dimensional District Vulnerability Classification *Zenodo DOI will be minted on first tagged release; this badge will be updated at that time.* **Author:** Valentine Golden Ghanem | COCOBOD Cocoa Clinic, Ghana **ORCID:** 0009-0002-8332-0220 **Affiliation:** COCOBOD Cocoa Clinic, Ghana **Reporting standard:** STROBE (observational, ecological design) **Date:** 2026 **Status:** Manuscript in preparation ## 1. Abstract Alcohol use, noncommunicable disease (NCD) mortality, and suicide are frequently framed as a linked cluster of risks in sub-Saharan Africa, but whether national alcohol-consumption trends genuinely track these outcomes over time in Ghana has not been formally tested. This study analysed two independent data layers. A national panel (WHO GISAH/GHE/GHO compendia, 2000–2022) tracked alcohol consumption, heavy episodic drinking, crude suicide rate, and combined 30–70y NCD mortality probability via Mann-Kendall trend tests. Total NCD deaths were modelled via Poisson regression against alcohol consumption with mandatory overdispersion testing and Monte Carlo uncertainty propagation. A secondary, exploratory machine-learning model tested whether alcohol/expenditure covariates predicted year-over-year change in suicide rate. A district structural-vulnerability index (261 districts, 2021 census, six standardised determinants) was analysed via Getis-Ord Gi* and global Moran's I under two spatial-weights specifications. Ghana's crude suicide rate rose significantly, 2000–2021 (Mann-Kendall tau=0.69, p<0.0001), while heavy episodic drinking (tau=−0.55, p=0.0005) declined significantly and total alcohol per-capita consumption showed no trend (p=0.169); both significant trends were robust to Bonferroni/Holm correction under both the original and an autocorrelation-robust test. The Poisson model's apparent alcohol–NCD-mortality associat …

Licenses