Using the Montreal Cognitive Assessment Test (MoCa) to screen for low cognitive performance in HIV positive heavy episodic alcohol users

Ori, Rasmita

Using the Montreal Cognitive Assessment Test (MoCa) to screen for low cognitive performance in HIV positive heavy episodic alcohol users

Thesis / Dissertation

2026

Publisher

University of Cape Town

Department

Department of Psychiatry and Mental Health

Faculty

Faculty of Health Sciences

Abstract

Background South Africa faces a dual public health crisis: concurrent HIV and high-risk alcohol consumption epidemics. Cognitive impairment in individuals with HIV (PLWH) is multifaceted, influenced by comorbidities and lifestyle factors. HIV infection correlates with increased alcohol use, potentially compromising treatment adherence and health outcomes. Globally, data on cognitive performance in PLWH with heavy episodic alcohol use are limited. Optimal assessment involves comprehensive neurocognitive testing with appropriate normative data; however, in resource-constrained settings, a concise, accessible screening tool administered by clinic staff is desirable. Early identification of at-risk individuals could enhance treatment outcomes and access to care. Given South Africa's resource limitations, this study investigates the utility of the Montreal Cognitive Assessment (MoCA) as a screening tool for detecting cognitive impairment in PLWH with heavy episodic drinking (HED) in Cape Town. Method: Sixty-six outpatients at Nolungile Clinic in Khayelitsha, Cape Town, underwent assessment encompassing sociodemographic data, MoCA and neuropsychological test battery scores, quantitative alcohol use, and clinical parameters. Participants were categorized into four groups based on HIV status and heavy alcohol consumption. Nonparametric Receiver Operating Characteristic (ROC) analysis determined the sensitivity and specificity of various MoCA cutoff scores against a gold-standard neurocognitive battery. Neurocognitive impairment was evaluated using the Global Deficit Score (GDS) and a novel multivariate method (NMM). Bivariate and chi-square tests identified associations between clinical variables and low cognitive performance. Results: The MoCA demonstrated high accuracy in differentiating participants with low cognitive performance, utilizing either the GDS (GDS; AUC = 0.901, 95% CI = 0.8090.992) or the NMM (NMM; AUC = 0.868, 95% CI = 0.766–0.969) to define low cognitive performance. Optimal MoCA cutoff scores were 19 (GDS) and 20 (NMM) with slightly lower specificity for the NMM (Sp = 0.661) compared to the GDS (Sp = 0.768), resulting in 19 false positives for the NMM and 13 for the GDS. At the standard MoCA cutoff of 26, 57 subjects (86.4%) were classified as having low cognitive performance. Using cutoff of 19 from ROC analysis, only 18 subjects (27.2%) were classified as impaired. Between group comparisons of low cognitive performance (HIV-/HED- = 5/19, HIV+/HED- = 6/18, HIV-/HED+ = 2/16, HIV+/HED+ = 5/13) revealed no significant difference (χ² = 2.923, df = 3, p = 0.404). MoCA cutoff scores showed no association with years of education or total socioeconomic status (SES) score. Similarly, no association was found between MoCA impairment and drinking score or HIV severity measures. Correlational analyses revealed a significant association between total MoCA score, SES, and NP domain scores (t = 2.436, df = 64, p = 0.01764). Significant associations were also observed for individual MoCA domains of naming and language with SES score. Low cognitive performance was significantly more prevalent in females (15/38 = 39.5%) than males (3/28 = 10.7%). Conclusion: This study demonstrates an association between lower cognitive performance and HIV, particularly among individuals with comorbid heavy alcohol use. This at-risk population would benefit from targeted screening to optimise treatment outcomes and access to dual diagnosis services. The MoCA maybe a valuable screening tool in resource constrained settings. Lowering the MoCA cutoff score further mitigates the overestimation of low cognitive performance. This research contributes to existing literature, showing a statistically significant association between SES score and both total MoCA score and specific cognitive domain impairment.

Keywords

HIV

alcohol consumption

Reference:

Collections

Masters

Full item page