HIV-associated neurocognitive disorders in sub-Saharan Africa: a pilot study in Cameroon

Neurological complications affecting the central and peripheral nervous systems are common in HIV infection. The vast majority of studies on HIV-related neurocognitive disorders were performed in developed countries, on patients infected with HIV-1 subtype B [for recent reviews, see [1–4]]. Two-thirds of the estimated 33.2 million individuals currently living with the human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS) worldwide are in sub-Saharan Africa (SSA) and are infected with mostly non-B HIV subtypes. Thus, it is important to diagnose and categorize HIV effects on the central nervous system (CNS) in SSA [4, 5].

Autopsy studies of HIV-infected individuals have shown that the brain is the second organ most frequently affected by HIV infection [6]. Studies of neuroAIDS pathogenesis have established that following HIV infection, the virus rapidly crosses the blood-brain barrier and enters the brain, where it productively infects brain macrophages and microglia [for recent reviews, see [3, 7–9]]. Toxic HIV proteins and cytokines secreted by infected mononuclear phagocytes in the brain cause dysfunction of other CNS cells and neuronal death [10–12]. This HIV-induced brain pathology often results in behavioral, motor and cognitive abnormalities referred to as HIV-associated neurocognitive disorders (HAND) [13]. HIV-associated dementia (the most severe form of HAND) has been described as a sub-cortical dementia, yet clinical and pathological studies have also shown an association between cortical neurodegeneration and HIV-associated neurocognitive impairment in infected patients [4, 14, 15].

Like many countries in SSA, Cameroon still has a heavy burden of the HIV/AIDS epidemic. According to recent estimates, the prevalence of HIV infection in the overall adult population of Cameroon is 5.1%, with prevalence of 8.3% in large metropolitan areas like the capital city Yaoundé, 7.4% among women, and 26.4% in female sex workers [16–18]. Furthermore, Cameroon has a high genetic diversity of HIV strains including several group M HIV-1 subtypes, HIV-1 groups O and N, circulating recombinant forms and unique recombinant forms [19–23]. Therefore, it is important to understand the neurological and neurocognitive effects of HIV infection in this country.

A comprehensive NP battery was recently adapted and successfully used to produce normative data in healthy children in Cameroon [24]. Also, we have screened HIV-infected adults in Yaoundé for HAND and its risk factors using the International HIV Dementia Scale (IHDS) [25, 26]. Although the IHDS is simple and rapid, it is neither comprehensive nor able to assess specific cognitive functional domains. The goal of the current study was to adapt the HIV Neurobehavioral Research Center (HNRC, University of California at San Diego) international NP test battery in Cameroon, and to explore its validity for the detection of HAND in infected Cameroonians. This battery included 19 test measures designed to evaluate several ability domains known to be vulnerable to HIV effects on the brain, including abstraction/executive function, speed of information processing, attention and working memory, verbal fluency, motor ability, learning and delayed recall [27–29]. This preliminary report describes the first attempt to examine the effect of HIV infection on specific cognitive domain functions in infected individuals in Cameroon.

This is the first study of NP performance and impairment of specific cognitive domains using a comprehensive and validated NP assessment battery in HIV-infected individuals in Cameroon. After translation into French and when administered by trained examiners, this battery detected cognitive dysfunction in HIV-infected Cameroonians. Participants were able to understand, follow test instructions, and validly complete NP tests, including computerized tests such as the Category test, WCST-64, and PASAT-50. All HIV- and HIV+ subjects were recruited in the same settings/hospitals in Yaoundé, Cameroon, and shared similar demographic features. As in many SSA countries, it has been shown that HIV transmission in Cameroon is predominantly (> 90%) through unprotected heterosexual contact [50] which reduces the likelihood of comorbid conditions that may affect the CNS (e.g., substance use disorders, co-infection with Hepatitis-C due to injection drug use that is a common risk for HIV infection in other parts of the world). Cultural differences probably affect performances on NP tests in a similar way as education differences. However this will not necessarily affect the validity of the NP tests for measuring disease related CNS dysfunction because the basic abilities assessed (verbal fluency, learning, attention and working memory, cognitive processing speed, executive function) exist and have some relevance in any population. Moreover, these cultural differences are mitigated when controls and diseased groups are from the same population with similar culture, as in this study.

HIV infection in Cameroon appears to be associated with impairment of cognitive function in domains similar to those most affected by HIV in geographic regions outside West Africa. Specifically, AIDS patients performed worse than non-AIDS HIV+ and HIV- groups on the global NP testing and NP tests assessing executive function, speed of information processing, working memory, and memory recall. In general, compared to HIV- control, the pattern of cognitive deficits was similar in non-AIDS HIV+ and AIDS groups, with deficits more pronounced in AIDS patients for all cognitive domains. Our results also indicate that neurocognitive impairment increases as the disease progress and is strongly associated with advanced immune deficiency and clinical disease, consistent with findings in the developed world. Effects from opportunistic infections are unlikely to have caused NP impairment because persons with a history of these diseases were excluded. All study participants were recruited at Yaoundé hospital out-patient clinics, appeared physically well and came to the clinics alone, without any support from a caregiver; thus, it is unlikely that patients in the AIDS group performed poorly because they were feeling too sick to perform at their best.

Other studies have shown increased rates of neurocognitive impairment with advanced stages of HIV infection [51–53]. Using this comprehensive NP battery, Heaton and colleagues [28, 29] showed that HAND is an important feature of HIV infection in China, and the impairment pattern, severity, and prevalence were similar to those reported in Western countries. Studies in India showed deficits in the domains of fluency, working memory, memory and learning in Clade C HIV-infected individuals, with AIDS patients showing greater impairment in visual working memory than non-AIDS and seronegative controls [32]. Robertson and colleagues [54] evaluated NP performance among HIV+ individuals in Uganda and found that compared to HIV- controls, the HIV+ group showed significant deficits on measures of learning and memory, speed of information processing, attention and executive function.

Clifford and colleagues [55] examined a group of workers in Ethiopia and found that HIV+ patients had slowed finger-tapping speed, but that the HIV+ and HIV- groups performed similarly in timed gait, Grooved Pegboard, and verbal fluency tests. Similarly, we found that performance of HIV- and HIV+ groups on the Grooved Pegboard, and on tests of verbal fluency, and on Trail making-A or Color Trails-I and II did not differ significantly (Table 1). In contrast, our more comprehensive NP battery found that HIV+ persons (mainly those with advanced disease) performed significantly worse than HIV- controls on tests of working memory (PASAT and WMS-III Spatial Span) and higher level executive functions (Category Test and WCST-64), with smaller (non significant with the current sample sizes) effect sizes in the same direction on tests of speed of information processing (WAIS-III Digit Symbol and Symbol Search, and Stroop Color Naming speed) and visual episodic memory (BVMT-R). Interestingly, in contrast to the latter result with visual episodic memory and findings in developed countries, our measures of verbal episodic memory (HVLT-R learning and recall) did not even show a small HIV effect. A smaller HIV effect for our verbal than visual episodic memory measures also was seen in rural China [29] and may suggest that the visual measures are more generalizable across cultures.

Several studies had demonstrated that demographic effects such as age, education, cultural and ethnic differences affect NP test performance [56–59]. A study of NP impairment among HIV+ individuals in South Africa showed that older age, lower education level, post-traumatic stress disorder, and alcohol use affected NP performance [60]. As expected, demographic effects were also present in our data as computation of Pearson's r showed that age and education significantly influenced NP performance, especially in the HIV- group, while gender had no major effect.

The cognitive impairments observed among HIV+/AIDS Cameroonians are likely due to HIV infection because our comparison groups were well matched for demographic factors and those with confounding conditions such as alcohol intoxication, systemic illness, CNS or psychiatric diseases were excluded. There are currently no normative adult NP test data for Cameroon or SSA, and large demographic effects noted in this study underscore need for more comprehensive local normative NP data to improve neurocognitive diagnosis [61]. Our subsequent studies will enroll more HIV- control subjects in order to establish normative data for NP testing in adult Cameroonians.

Our pilot study has a number of limitations including: 1) the small sample size which likely limits the power to detect differences in some individual NP test measures, precludes us from creating demographically adjusted test scores and from exploring possible effects of ART on cognitive performance; 2) the inability of the current study to estimate the population prevalence of cognitive impairment due to lack of local normative NP data; 3) lack of brain imaging and cerebrospinal fluid (CSF) data, due to limited resources and logistic difficulties, precludes us from exploring the relationship between HIV-induced neurologic abnormalities, HIV brain infection (CSF viral load) and cognitive performance; 4) this study was conducted in an urban setting and findings cannot be generalized to rural settings where inhabitants have much lower education levels.

A recent United Nations Programme on HIV and AIDS (UNAIDS) report estimates that over half million Cameroonians are currently living with HIV, of which only about 25% have access to antiretroviral drugs and ART coverage for prevention of mother-to-child transmission is less than 25% [17]. In our current study, only 15 of the 44 HIV+ individuals (34%) were on ART, including 6 of the 22 non-AIDS HIV+ and 9 of the 22 AIDS patients. In subsequent studies with larger sample size, we will evaluate the effect of ART on neurocognitive performance. Neurocognitive impairment predicts the presence of HIV encephalitis [62–64] and is an independent risk factor for death [65]. In resource limited countries like Cameroon, many HIV-infected individuals have no access to ART, and low CD4 cell counts and clinical symptoms remain the major criteria for selecting patients for ART treatment. Detection of neurocognitive impairment could target patients for ART, possibly reversing brain dysfunction in some cases, improving their productivity and quality of life, and perhaps even prolonging their survival.

Advanced immunosuppression from HIV infection is associated with HAND in Cameroon. AIDS patients showed impairment in executive function, speed of information processing, working memory, and memory recall and the pattern of deficits is similar to that found in many other areas of the world. Further longitudinal research with large HIV-infected population is needed to determine whether suppression of HIV replication through ART will reduce the prevalence of neurocognitive impairment in Cameroon as observed elsewhere.

United States investigators: GDK is Associate Professor and Vice-Chair in the Department of Pharmacology and Experimental Neurosciences, University of Nebraska Medical Center. RKH is Professor and Neuropsychologist at HNRC, University of California San Diego (UCSD). LAC was a Research Fellow at HNRC and recently moved to the University of New South Wales, Sydney, Australia. RJE is a Neurologist and Professor at the HNRC, UCSD. JAM is Professor, Epidemiologist, and infectious diseases specialist at the HNRC, UCSD. DRF is a Psychometrist at the HNRC, UCSD.

Cameroon investigators: AKN is a Neurologist, Head of the Neurology Department at the Yaoundé Central Hospital, and Associate Professor and Vice-Dean, Faculty of Medicine and Biomedical Sciences, University of Yaoundé I. KTC is a Neurologist at the Yaoundé Central Hospital. JYF is a clinician and research fellow at the Yaoundé Central Hospital. SE and RD are Psychologists and trained Psychometrists at the Yaoundé Central Hospital. DM is Head of the hematology lab at the Yaoundé University Hospital Center, and Professor in the Faculty of Medicine and Biomedical Sciences, University of Yaoundé I. EM is a technician in the hematology lab, University Hospital Center, Yaoundé. FB is Professor in the Faculty of Medicine and Biomedical Sciences, University of Yaoundé I. DMN is a counselor and social worker at the HIV Day-Care Hospital, Yaoundé Central Hospital.