Effects of HIV status and linguistic medium on test performance of low literacy rural adults: Implications for neuro psychological test development in Zambia

Abstract

Studies in non-Western societies (China, India, Uganda, and Ethiopia) have reported differential performance on neuropsychological tests by HIV+ and HIV- groups. However, the impact of HIV on cognitive functioning has rarely been studied among individuals with low literacy, even though these represent the majority of individuals with HIV in most non-Western societies. One of the constraints is that most relevant tests are available only in English, which can be used only with persons with formal education. In addition, even when individuals are familiar with spoken English, it is possible that tests administered in a familiar language may be more sensitive to HIV-induced neurocognitive impairments than tests in English. This study followed an experimental design involving two versions of a set of tests – one in English and the other in Chichewa. The tests were Hopkins Verbal Learning Test for both immediate and delayed recall, Animal and Action Fluency tests of the Zambia Neurobehavioural Test Battery (ZNTB). The sample size comprised 28 HIV-infected (HIV+) (untreated with ART) and 22 uninfected (HIV-) rural residents (aged 40-65 years) with fewer than five years of education who tested as being illiterate in English, and who had Chichewa as their primary language but had some familiarity with English. These participants were recruited separately from the sample of the project as an extra sample to test their performance. Participants were tested twice, once in English and once in Chichewa. Counterbalancing of sequence was not planned but was done informally during data collection. The effects of HIV status, language of testing, and the interaction between HIV status and language, on individuals’ test performance were determined by two-way ANOVAS, with repeated measures and t tests. HIV positive respondents scored significantly lower than HIV negative respondents on all the Neuropsychological tests and the mean scores on the English medium version were consistently lower than scores on the L1 (Chichewa) version across all tests and all groups. HIV status and linguistic medium showed a significant interaction, with the difference between HIV positive and HIV negative groups being larger with the L1 (Chichewa) version, indicating that, this version discriminated more sharply between HIV positive and negative individuals than the English version. The study has demonstrated empirically the significance of language use in the assessment of HIV-associated neurocognitive functional disorders that remain highly prevalent and continue to represent a significant public health problem in this part of the world. Key Words: HIV status, linguistic medium, rural adults, neuropsychological test,Zambia.