Introduction

A medical condition that causes impaired mental function other than a psychiatric illness is known as a neurocognitive disorder1. Neurocognitive disorders (NCDs) are defined as deficits in neurological functions, motor functions, psychological functioning, and daily and workplace activities2. Even when effective antiretroviral therapy is widely used, NCD among people living with HIV/AIDS (PLWHA) continues to be one of the most common medical conditions3. HIV-associated neurocognitive disorder (HAND), which includes a variety of cognitive, motor, and/or emotional difficulties, can occur in PLWHA4.

Most HAND patients demonstrate asymptomatic neurocognitive impairment, particularly in cohorts with reduced blood viremia5,6. The neuronal injury that underlies the pathophysiology of HAND is either direct neurotoxicity by HIV and/or its viral proteins or, more critically, indirect neuronal damage through neural inflammation, the so-called “bystander effect”7. Individuals with suppressed HIV infection receiving the best possible antiretroviral medication perform poorly on neuropsychological tests, indicating that neurological impairments may persist despite treatment and expected long-term existence8. The subcortical regions are those that are most frequently affected by HIV; hence, changes in these structures result in a range of disorders known as HAND9.

A systematic review performed in the sub-Saharan region in 2020 estimated the pooled incidence of NCD in HIV to be 44.46%10. Worse neuropsychological function in this region’s cases are common11. Early on, PLWHA were predicted to have up to 17 years less life expectancy than that of HIV-negative individuals12. Ethiopia, as one of the regions in sub–Saharan Africa, has a prevalence that ranges from 34.07 to 53.20%9. The Amhara region has increased number of people with HIV infection13. The incidence rate per 1000 people was high in Dessie town (5.74), Bahir Dar city (4.27) and Gondar town (3.00)14,15. Patients with at least one cognitive illness had a higher death rate than those without (17.6% vs. 8.0%), particularly for AIDS-related deaths16. The most severe form, HIV-associated dementia (HAD), is rare, and milder varieties of cognitive impairment, such as asymptomatic neurocognitive impairment (ANI) and mild neurocognitive disorder (MND), made up more than 88% of all NCD types in HIV patients in a global study17. This study aimed to assess the current magnitude and factors associated with neurocognitive disorders among HIV patients by assessing several cognitive domains using tools sensitive for milder types of cognitive impairments which was lacking in previous studies.

Methods and materials

Study area

The study was conducted at selected public health facilities in Bahir Dar city that provide HIV/AIDS treatment and services. Bahir Dar is located 492.7 km from Addis Ababa, the capital city of Ethiopia. Currently, three public hospitals and nine health centers provide comprehensive HIV treatment. The data were collected from Felege Hiwot Referal Hospital and three health centers. According to the facility records, approximately 11,545 people received their HAART regimen in these facilities. Felege Hiwot Hospital had a monthly patient allocation of 1720 and was allocated (333), Abay HC has 340 and was allocated (66), Bahir Dar HC had 450 and was allocated (87), Han HC had 428 and was allocated (83) study participants.

Study design and period

A cross-sectional study was conducted to assess the proportion and associated factors of neurocognitive impairment among HIV patients from March 20 to April 30, 2023.

Source population and study population

The source population was all adults living with HIV and on HAART who attended the selected public health facilities in Bahir Dar city; the study population included all adults living with HIV and on HAART who attended the selected public health facilities in Bahir Dar city and who were available during the study period.

Inclusion and exclusion criteria

Individuals aged 18 years and older and who were receiving ART for more than a year and who were not diagnosed with CNS opportunistic infections such as toxoplasmosis, meningitis, or encephalitis were included, while those who were unable to respond and were on concomitant delirium; who were unable to hear or communicate; who were visually impaired; and who had a severe disability in their hands such as amputation were excluded.

Sample size determination

The sample size for this study was estimated by using the single population proportion formula by considering a population proportion (P) of 34.07%9, where zα/2 is the value of the standard normal distribution at the 95% confidence interval (α = 0.05) and 5% margin of error (d), and n is the estimated sample size, with a 1.5 design effect and 10% nonresponse. The final sample size was 569.

Sampling technique and procedure

A multistage sampling technique was used. There were 13 public health facilities in the city. One health center did not provide comprehensive ART follow-up services. Considering the nine health centers and three public hospitals, there were a total of 12 public health facilities. Taking 30% of this total number or 1/3rd of the institutions, four facilities were chosen using the lottery method administered separately for the hospitals and health centers. Subsequently, using the facility HIV registry monthly patient appointment chart and proportional allocation, a computer-generated simple random sampling technique was administered to each of the four facilities to recruit the participants.

Study variables

Dependent

Neurocognitive disorders (yes, no).

Independent

Sociodemographic data

Age, sex, educational status, marital status, level of education, employment status and place of residence.

Clinical and HIV AIDS related

Clinical stage of the disease, viral load, CD4 count, drug adherence, opportunistic infections, type of ART drug regimen, impaired activity in daily living, previous history of neurologic disease and comorbidity.

Psychosocial related

Social support, depression, anxiety, communication about safe sexual intercourse.

Substance use-related variables

Alcohol consumption, cigarette smoking and khat chewing.

Operational definition

Neurocognitive disorder

A person with a score less than 10 on the International HIV Dementia Scale (IHDS) and less than 26 on the Montreal Cognitive Assessment (MoCA) was considered to have neurocognitive impairment, as recommended in a similar previous study18.

Asymptomatic neurocognitive impairment (ANI)

Patients who presented with the outcome and had impairments in at least two of the 6 cognitive domains, namely, language, attention, executive function, speed of memory recall, information processing speed and sensory and motor skills. Cognitive impairment has no impact on day-to-day functioning19,20.

Mild neurocognitive impairment (MND)

Patients who presented with the outcome and experienced cognitive impairment in at least two of the six listed cognitive domains. There are mild impairments in regular activities resulting in inefficiency in social interactions, domestic tasks, and at work19,20. Mild impairments were defined as a score of 6 or 7 on the Lawton scale.

Dementia

Those presenting with the outcome and having cognitive impairment in at least 2 of the cognitive domains mentioned above and causing marked interruption with daily activities19,20. A marked impairment is defined as a score of 5 or less on the Lawton scale. Classification was performed based on the results of the International HIV Dementia Scale (IHDS), Montreal Cognitive Assessment Scale (MOCA) and Lawton Instrumental Assessment of Daily Living (IADL) measurements.

Unclassified NCD

Patients who presented with the outcome and had cognitive impairment in only one of the domains mentioned above. There was no impairment in daily living.

Impairment in daily living

A person with a score of 8 was considered to be normal, a score of 6–7 was considered to indicate mild impairment, and a score of 5 or less was considered to indicate marked impairment in activities of daily living according to the LAWTON Instrumental Activities of Daily Living (IADL) tool.

Advanced stage of disease

Patients were classified as WHO-classified HIV AIDS clinical stage 3 or above.

Alcoholics

People who consumed 7 or more alcoholic units per week for men and 5 or more alcoholic units per week for women were considered alcoholics in this study. The number of alcoholic units was calculated by multiplying the volume of drinks consumed in a week by the respective alcohol amounts of the drinks divided by 1000 according to the National none-comunicable disease prevention and control (NCD) guideline 201621.

Cigarette smoking

Participants who smoke currently, regardless of the number or frequency of smoking days, were considered smokers.

Comorbidity

Participants had one or more long-term chronic conditions other than HIV, such as diabetes, hypertension, stroke, chronic kidney disease, or liver disease.

High viral load

Having more than 1000 copies of virus/ml of blood.

Khat chewer

Individuals who claimed to chew Khat currently, regardless of the frequency and amount of Khat, were considered Khat chewers.

Low CD4+ T-cell count

Patients with less than 500 cells/mm3 of blood.

Medication adherence

Drug adherence was calculated by dividing the number of pills taken by the patient by the number of monthly doses given and multiplying this value by one hundred22. Good adherence (> 95% intake), fair adherence (85–95% intake) and poor adherence (< 85% intake) according to the National HIV care comprehensive guideline.

Opportunistic infections

Patients with HIV infection caused by bacteria, viruses, fungi, or parasites after diagnosis.

Perceived social support

Participants who scored 3–8 on the Oslo 3 social support scale were considered to have poor social support, those who scored 9–11 were said to have moderate social support, and those with a score of 12–14 were said to have strong social support.

Depression and anxiety

Individuals who scored greater than or equal to 3 on the first two questions on the Patient Health Questionnaire (PHQ 4) were reported to have anxiety, and those who scored greater than or equal to 3 on the last two questions on the scale were said to have depression.

Data collection tools

The data were collected using a questionnaire consisting of sociodemographic, clinical and HIV-related, psychosocial and substance-related variables. Standard tools such as the IHDS, MoCA-B, Lawton instrument of daily living, PHQ-4 Depression Scale and Oslo-3 Social Support Scale were used. Interviews, document reviews and observations were the methods used to collect the data. A total of 59 questions were used.

The IHDS is a 3-item tool for assessing cognitive, motor, and psychomotor processes and was used to assess the presence of NCD. A cutoff point of less than 10 was used.

The 12-item MoCA, which can be scored out of 30, was used for evaluating cognitive performance. It has been recognized for its capacity to detect tiny variations in cognitive function. The MoCA was used to evaluate the following domains: visual perception, executive functioning, short-term memory, attention/concentration, language, and orientation23. In this study, a cutoff point of 26 was used. This study used the MoCA-B, which was developed to facilitate the detection of mild cognitive impairment in illiterate and less educated study subjects24.

Furthermore, the PHQ-4 is a four-item tool for assessing anxiety in the first two items and depression in the last two items25.

The IADL is an 8-item tool that is used to evaluate a patient's functional status; this tool is typically impaired in HIV patients26.

The Oslo 3 social support scale (OSSS 3) was used to assess social support. The OSSS-3 solely consists of three questions that inquire about the number of close friends, one's opinion of other people's concerns, and one’s relationship with neighbors, with an emphasis on the availability of practical help27.

The data were collected by five nurses from the ART department (one from each of the three health centers and two from the hospital). One supervisor was assigned to control the overall data collection procedure. Before being involved in the data collection, the data collectors received training and orientation on how to utilize the questionnaire and instruments, the ethical principles of confidentiality, and the data management. To ensure that all participants could understand the tools and questionnaire, the tools were first translated into Amharic and subsequently back-translated to English.

Data processing and analysis

Epi Data version 4.6 was used to code and enter the data, which were subsequently exported to STATA version 14 for further cleaning and analysis. To summarize the data, descriptive statistical analysis was performed. Pie charts, frequency/percentages, and tables were generated. The relationship between the dependent variable (NCD) and the independent variable (factor) was examined using binary logistic regression. First, simple binary logistic regression was performed. Variables with a cutoff p value < 0.25 were subsequently included in the multiple binary logistic regressions. The Hosmer and Lemeshow goodness of fit test was used to determine whether the model was adequate. The data were fitted with a chi-square of 7.43 and a p value of 0.49. For all the statistically significant tests, the cutoff value was a p value < 0.05 with a 95% confidence interval.

Data quality assurance

The questionnaires were standardized. In addition to the standard tools, some questions were adopted from the literature reviews. The data collection tool was pretested on 5% of the sample (28 participants) at a distinct private hospital. Based on the results of the pretest, language clarity, the appropriateness of the data collection methods, the expected completion time, and the necessary changes were considered. The data collection procedure was explained in a one-day training session for the data collectors. The lead investigator provided constant oversight and monitoring during the data collection period.

Ethics approval and consent to participate

The Helsinki Declaration for medical research involving human subjects was complied with. The College of Medicine and Health Science at Bahir Dar University granted ethical approval. A consent letter from the study settings was obtained from the administrators of each study facility. Written informed consent was obtained from each study participant. Every detail about the participants was kept private, and a strict code of ethics was enforced.

Results

Sociodemographic characteristics of the study participants.

Out of the 569 people who were approached, 501 (88.04%) were enrolled and participated in the study. Among them, 335 people (66.86%) were female, and 456 (91.02%) were urban dwellers. The mean age of the participants was 39.06 ± SD 9.23 years. Approximately two-thirds of the population (377; 75.25%) could read and write, and 359 (71.66%) had received formal education. A greater proportion of participants were daily laborers (106, 21.16%), followed by government employees (102, 20.36%) (Table 1).

Table 1 Sociodemographic characteristics of PLWHA attending public health facilities in Bahir Dar, Ethiopia, in July 2023 (n = 501).
Full size table

Clinical and HIV-related characteristics of the study participants

Approximately seventeen percent (83) of the participants were diagnosed with WHO clinical stage two disease, and 9 (1.80%) had advanced-stage disease. Among those with decreased CD4 T cells (< 500 cells/dl), 298 (59.48%) had decreased CD4 T cells/dl, and 26 (5.18%) had increased CD4 T cells/dl. A high viral load (> 1000 copies/ml of blood) was observed in 36 (7.19%) participants. Only eight percent (39) of the participants were receiving the second-line HAART regimen. Most of the participants claimed to strictly adhere to their medications (354 [70.66%], while 25 [4.99%] reported poor medication adherence). Approximately one-third of the participants 161 (32.14%) had a history of opportunistic infections. Nearly one-fourth of them (135; 26.95%) had a previous history of neurologic disease symptoms before their HIV diagnosis. A total of 80 (15.97%) people lived with chronic comorbidities other than HIV. Among these patients, 21 (26.25%) had diabetes mellitus, 25 (31.25%) had hypertension, 5 (6.25%) had cardiac disease, 3 (3.75%) had renal disease and 26 (32.50%) had other chronic comorbidities. Eighty-three (16.56%) of the participants reported impairment in activities of daily living (Table 2).

Table 2 Clinical and HIV-related characteristics of PLWHA attending public health facilities in Bahir Dar, Ethiopia, in July 2023 (n = 501).
Full size table

Psychosocial and substance use-related characteristics of the study participants

A total of 266 (53.09%) claimed to have sexual partners, and among them, 123 (46.24%) reported communicating about using condoms before having sexual intercourse. Regarding substance use, 38 (7.58%) of them reported consuming alcohol above seven alcoholic units (men) or five alcoholic units (women). The proportions of people who currently chewed khat or smoke cigarettes were 11 (12.20) and 6 (1.20%), respectively. Regarding social support, 265 (52.89) had poor social support, 152 (30.34%) had moderate social support, and 84 (16.77) had strong social support. More than one-third of the participants (28 [5.59] had anxiety, and 27 [5.39] had experienced depressive symptoms (Table 3).

Table 3 Psychosocial and substance-related characteristics of PLWHA attending public health facilities in Bahir Dar, Ethiopia, in July 2023 (n = 501).
Full size table

Proportion of patients with neurocognitive disorders among HIV patients

Of the study participants, 274 scored less than 10 on the IHDS and less than 26 on the MoCA. Thus, the proportion of NCDs was 54.7% (95% CI 50.62–58.77). The methods used to implement the IHDS and MoCA were as follows: timed finger tapping, psychomotor speed, and memory recall. The three assessments on the IHDS were individually scored from 1 to 4 and summed to a total score of 12. On the MoCA score, executive functioning (1), language fluency (2), orientation (6), calculation (3), abstraction (3), delayed recall (5), visual perception (3), naming (4), and attention (3) were the components, and each was summed up to 30. A total of 191 (18.56%) participants scored 10 or above on the IHDS, and 93 (18.56%) scored 26 or above on the MoCA.

Among the individuals with outcomes, 199 (72.63%) were asymptomatic, 50 (18.25%) had mild neurocognitive disorders, and 11 (4.01%) had severe neurocognitive disorders. The remaining 14 (5.11%) presented with the disorder but did not fall into the above categories (Fig. 1).

Figure 1
figure 1

Pie chart graphics on the classification of neurocognitive disorders among HIV patients.

Full size image

Factors associated with NCDs among HIV patients

A bivariable regression analysis was performed, and variables with p values less than 0.25, sex, education, marital status, residence, impairment in daily living, anxiety, history of opportunistic infection, type of drug regimen, recent CD4 count and viral load count, were identified.

Whereas the variables which were significant in the multiple logistic regression with < 0.05 significance were: Marital status (being widowed or divorced), rural residence, having a history of opportunistic infection, having a recent CD4 count of below 500, being in 1st line ART regimen consisting either Tenofovir, Lamivudine, Dolutegravir (TDF, 3TC, DTG) or Abacavir, Lamivudine, Dolutegravir (ABC,3TC, DTG) or Tenofovir, Lamivudine, Efavirenz (TDF, 3TC, EFV) complex and having impairment in activities of daily living.

Three widowed people were affected (AOR = 3.05 [95% CI 1.47–6.31]), and 1.9 divorced people were more likely to develop NCD (AOR = 1.95 [95% CI 1.16–3.28]) than married people were. In addition, people residing in rural areas were two times (AOR = 2.28 (95% CI 1.02–5.09)) more at risk for NCD than people living in urban areas were. Those with recent CD4+ T-cell counts less than 500 cells/dl were 1.6 times more likely to have NCD (AOR = 1.61 (95% CI 1.03–2.50)) than people with CD4+ T-cell counts between 500 and 1000. The data also revealed that opportunistic infections were 2.21 times more strongly linked to HAND (AOR = 2.21 (95% CI 1.42–3.41)).

People with impairments in daily living were 2.6 times more vulnerable to cognitive impairment than were those without impairments. Additionally, people who were on the first-line ART regimen were 2.9 times (AOR = 2.92 (95% CI 1.22–7.00) more likely to experience NCD than were those on other regimens. Finally, individuals with impairment in activities of daily living were 2.6 times (AOR = 2.64 (95% CI 1.39–4.99)) more likely to experience NCD than were those without impairment (Table 4).

Table 4 Multiple variable regression results for neurocognitive disorders among HIV patients and associated factors; Bahir Dar, Ethiopia; July, 2023.
Full size table

Discussion

The proportion of HIV patients with NCD in this study was 54.7%. These results are in line with those of studies conducted in East Africa (66.7%), West Africa (49.6%), southern Africa (48.4%)28, Brazil (53.2%)29, and Asia (52.03%)30.

However, this percentage was greater than that reported in studies conducted in different regions of Ethiopia (36.4% in South Wollo31, 33.3% in Mekelle2, and 41% in Northern Shoa32). These findings were also greater than those of studies from Tanzania (21.5%)33 and northern Nigeria34. The observed variations could be attributable to the various methodologies, variations in participant characteristics, sample sizes, and regional cultural differences.

However, the incidence of NCD in this study was lower than that in studies conducted in Brazil (73.6%)35, Kenya (81.1%)18 and southern Ethiopia (67.1%)36. The variation in the southern Ethiopian population could be attributed to differences in participant characteristics. The variance in prevalence from the Brazilian and Kenyan studies might have been caused by differences in viral subclades. It has been demonstrated that specific viral clades are more likely to be neuropathogenic and consequently to induce cognitive impairment. According to several studies, type D HIV is more neuropathogenic than the other subclades37. The most common subclade type in Ethiopia is C38, which is a less common cause of neurocognitive impairment than the A and D subtypes3.

This study showed that widowed and divorced people were more likely to be exposed to cognitive impairment than married people were. This finding was in line with a previous study in which being married tended to be protective36. This finding is also consistent with a study that identified widowhood as a risk factor for cognitive decline in elderly individuals39 and with a study reporting that HIV-positive people living alone are more susceptible to NCD33. This might be the case because people might be less motivated to lead healthy lives, seek out social support, and abstain from drug use after losing a loved one or becoming divorced36. People who live alone also experience stress, which over time affects brain nerve interactions and leads to neurocognitive deficits40.

In this study, people residing in rural areas had greater exposure to NCD than people living in urban areas did. This result was supported by the findings of a previous study in which farming was associated with cognitive impairment41 since farming is most common in the countryside. The limited accessibility of early screening for cognitive deterioration may be one of the excuses. A study conducted among non-HIV cognitively impaired patients reported that agricultural work was related to cognitive decline for men residing in rural areas. The study also revealed that rural residents may have greater health issues, lower socioeconomic status, strained parent‒child relationships, and parental abuse, all of which are linked to neurocognitive impairment42; these findings could be similar for HIV-infected communities.

According to the present study, people with diminished CD4+ T cells were more prone to NCD than people with CD4+ T cells between 500 and 1000 cells/mm3 of blood. These findings could be supported by several previous studies17,30,41,43,44. This might be explained by the high vulnerability of people with low CD4+ T cells to neurological complications caused by infections45. A lower CD4+ T-cell count could be a biomarker for an increased viral titer in the central nervous system7.

The data suggest that having a history of opportunistic infection was directly associated with having a NCD. This could be supported by the findings of previous studies2,46, and the possible cause could be the occurrence of immunosuppression by the virus, possibly leading to infections that usually attack the central nervous system. These illnesses typically manifest when immunity levels decrease, a sign that one's health is at risk46, which most likely facilitates exposure to NCD.

People with impairments in daily living were also more vulnerable to cognitive impairment than were those without impairments. This finding aligns well with those of previous studies2,40,50. However, a study contrasted this finding by suggesting that an NCD has no association with activities of daily living47. The virus might have affected the subcortical regions of the brain that control motor function, working memory, executive function, information processing speed, learning, and recall of new information, which could account for the potential associations reported in this study48. Therefore, they experience difficulty completing their everyday tasks without help. Simultaneously, patients will also develop NCDs.

This study also suggested that people who were receiving the first-line HAART regimen had increased exposure to HIV-associated neurocognitive impairment compared to those who were receiving the second-line regimen. Among these people, those who had efavirenz in their regimen were at increased risk of developing NCD compared to those who were second-line agents. Previous studies that demonstrated a stronger association between NCD and efavirenz use, indicating the potential neurotoxicity of this drug, support this conclusion33,49. The second-line regimen was found to be protective against NCD, which could be justified by the drugs that are components of the second-line regimen, namely, zidovudine, nevirapine and lopinavir, which tend to prevent and reverse NCD due to their strong CNS penetration effectiveness score19,50.

Age and education were the two most strongly associated factors according to previous studies. However, there was no significant association in this study. The relatively young age of the participants in this cohort may be the cause of this difference. Although the majority of neuropsychological measurements are strongly correlated with education, the relationship between education and each test varies. This study used the Moca-B score, a tool that is suitable for illiterate individuals.

Limitations and strengths

One of the limitations of this study was that it was conducted among patients who were actively receiving HAART. This might have led to an underestimation of the current proportion of NCDs among HIV patients in this study because HAART reduces the prevalence of NCDs. Another limitation is the lack of normative data for NCD classification. Normative data for a specific population were obtained by performing standard neuropsychological tests on different cognitive domains on HIV-negative participants as controls. Additionally, confounders such as genetic disorders of cognition and behavior were not controlled. This study has several strengths. More cognitive domains were assessed using two standard tools than were assessed in previous studies conducted in the country. This approach helps better estimate the prevalence of NCD. Facilities with a larger proportion of HIV patients were used for the investigation. This approach will make it easier to generalize the findings to people attending public health facilities in the city. Its evaluation of ART regimens, which could have a considerable impact on NCDs, was also another strength.

Conclusions and recommendations

The overall proportion of NCDs among HIV patients in this study was greater than that in previous studies conducted in Ethiopia, and the factors associated with NCDs were being widowed or divorced, residing in a rural area, having a history of opportunistic infections, being on a first-line ART regimen with or without efavirenz and having impaired activities of daily living. The regional health bureau might consider integrating routine screening and timely intervention for NCDs as part of the existing comprehensive care provided by ARTs. Research institutions might consider conducting future follow-up studies and randomized control trials because they are essential for providing quality information regarding CNS penetration in ART regimens.