Stroke in Sierra Leone. the stroke risk factors for people with HIV: A prospective case-control study

Background: HIV infection rates are relatively low in Sierra Leone and in West Africa but the contribution of HIV to the risk factors for stroke and outcomes is unknown. In this study, we examined stroke types, presentation, risk factors and outcome in HIV stroke patients compared with controls. Methods: We used data from the Stroke in Sierra Leone Study at 2 tertiary hospitals in Freetown, Sierra Leone. A case control design was used to compare stroke type, presentation, risk factors and outcome in sero-positive HIV patients with HIV negative stroke controls. Controls were matched for age and gender and a 1:4 ratio cases to controls was used to optimize power. Analysis was performed using the Pearson x2 for categorical variable, Paired-T test and Mann-Whitney U test for continuous variables. A p-value of less than 0.05 was taken as the level of statistical significance. Results: Of 511 (51.8%) stroke patients tested for HIV, 36 (7.1%) were positive. Univariate unmatched analysis showed a stroke mean age of 49 years in HIV-positive versus 58 years in HIV-negative population (p = <0.001). In the case-control group, ischaemic stroke is the major type reported in both populations, HIV-negative population: 77 (53.5%) versus HIV-positive: 25 (69.4%) (p = 0.084). Hypertension is the most prevalent risk factor in both groups, HIV-positive: 23 (63.9%) versus HIV-negative: 409 (86.1%) (p = 0.001). Lower CD4+ count is associated in-hospital mortality (p = <0.001). Conclusion: These findings support the current call for timely management of stroke and HIV through integrated care.


Introduction
Stroke is a leading and rapidly increasing cause of adult deaths and disability in Sub-Saharan Africa (SSA) 1,2 .The human immunodeficiency virus (HIV) infection is a leading cause of death and disability in Africa particularly among the 25-49-year age group.The evidence for HIV as a risk factor for vascular disease has been demonstrated in a systematic review of over 300,000 HIV sero-positive individuals 3 .
With the advent of highly active antiretroviral therapy (HAART), HIV infection has become a chronic disease with a longer life expectancy in People Living with HIV (PLHIV).The increase in survival of PLHIV over the past few years has also contributed to a 10fold increase in the burden of non-communicable diseases (NCDs) in this population 4,5 .However, recent studies in Ghana and Malawi showed that in the first 6-months of initiation of HAART, PLHIV were associated with a higher risk of stroke occurrence compared to untreated HIV patients.This was attributed to immunosuppression and Immune Reconstitution Inflammatory Syndrome (IRIS) processes as a result of HAART use 6,7 .
Globally, the burden of non-communicable diseases such as stroke in PLHIV is expected to continue to rise to unprecedented levels, partly due to lifestyle changes, longer survival, and metabolic abnormalities related to antiretroviral therapy 5 .In a recent study in Malawi, PLHIV aged 45 years or younger were at high risk of stroke.Traditional risk factors for stroke, such as smoking, hypertension, and diabetes, can combine as a causal link with direct HIV infection, resulting in an almost 50% increased burden of stroke in PLHIV 6 .Metabolic syndrome among PLHIV was reported at 21.5% compared to 12% of uninfected patients in a recent systemic review of multiple cross-sectional studies comprising of about 500 HIV uninfected and 700 HIV infected participants in SSA 8 .Sierra Leone has a low HIV prevalence, with a national sero-prevalence of 1.7% 9 .However, despite these encouraging national statistics, less than 50% of PLHIV are receiving HAART 10 .As a result, many PLHIV in Sierra Leone are at greater risk of developing advanced HIV disease and therefore are also susceptible to stroke from untreated HIV infection compared to the risk attributed to early initiation of HAART 10,11,12 .The burden of both stroke and HIV in our study setting, Connaught Hospital, the adult tertiary referral hospital for Sierra Leone is high, reinforcing the need to know more about stroke in PLHIV to inform policy and practice [12][13][14] .
The Stroke in Sierra Leone (SISLE) register 14 was established in 2019 to determine the prevalence of stroke risk factors, stroke phenotypes, patient pathways to care, quality of hospital care and outcomes after stroke.In this case control study, we aim to examine the factors associated with stroke among PLHIV and compare traditional risk factors and clinical presentation with a gender and age-matched sample of stroke patients without HIV at Connaught Hospital in Freetown, Sierra Leone.

Study site
The Stroke in Sierra Leone (SISLE) stroke register is a prospective hospital-based stroke register in Sierra Leone's two main adult referral hospitals in Freetown [methods described in 14.].These facilities act as tertiary hubs for both stroke care and HIV, serving over 1.2 million inhabitants in the Western Area and nationwide.Computed axial tomography scan (CT scan) for this study were done at a private facility and paid for by the SISLE project, as there are no functional CT-scanners or MRI in the government health system.At facility level, patients need to pay for all services, diagnostic investigations, and treatment, however, in an effort to reduce the cost barrier for patients to access care and reduce selection bias onto the register, all investigations in our study were funded by the SISLE project whilst patients covered the cost for treatment.

Study design and study population
The study is a case control study of PLHIV with stroke aged 18 years or older matched to HIV negative patients using the prospective observational hospital-based stroke register.We defined 'People living with HIV' as HIV positive patients, regardless of the time of diagnosis.From the point of being diagnosed with HIV, we considered them to fall into the category of "people living with HIV."The term was used inclusive of anyone who has the HIV virus, regardless of how long they have had the virus or what stage of the disease they are in.This includes people who have just been diagnosed and are in the earliest stages of the infection, as well as those who have been living with the virus for many years.Stroke patients with HIV were matched by gender and age (+/− 5 years) with the HIV negative group (144) of stroke patients in a 1:4 ratio.Due to the small number of HIV patients (cases), case-control ratio was increased to 1:4 to improve representation and enable large enough sample size to conduct analysis.The Stroke register included all stroke patients aged 18 years and older between May 1, 2019, and October 8, 2021.The SISLE stroke register methods have previously been described in detail 14 .

Clinical, laboratory and imaging data collections and definitions
Patients with suspected stroke were evaluated by a trained research team, using the National Institutes of Health Stroke Scale (NIHSS) 15 .Confirmed stroke patients were classified into sub-types by an experienced stroke physician, with reference to the clinical records and investigations; ischaemic, primary intracerebral haemorrhage, subarachnoid haemorrhage and undetermined stroke type.The Bamford classification 16 and TOAST criteria 17,18 were used to further sub-classify ischemic stroke.Brain neuroimaging using a CT scan was performed for 857 (87%) of patients.The Barthel ADL Index was used to measure dependency in self-care 19 HIV test was done sequentially dependent on the availability of a testing kit, using a fourth-generation rapid diagnostic kit by SD Bioline HIV-1/2 3.0 (Standard Diagnostics Inc).The HIV testing model is provider-initiated, applying to all patients seen in the hospital.However, access to HIV testing services in this hospital is dependent on the availability of test kits and human resources.CD4 cell count was determined using the Alere Pima ™ Analyzer (Abbott), a point-of-care testing platform with comparable performance to flow cytometry-based methods and validated in resource-limited settings 20 .CD4 laboratory test was done only for newly diagnosed HIV patients, however, this was again influenced by the availability of testing supplies.

Data collection and analysis
All the data were collected using a standardized paper-based report form and then double entered into REDCap ™ .Univariable analyses of patients tested for HIV (511) versus those not tested (475) and confirmed stroke patients with HIV positive status (36) compared to HIV negative (475) was done.Stroke patients with HIV were matched by gender and age (+/ − 5 years) with the HIV negative group (144) of stroke patients in a 1:4 ratio.The data was analyzed using IBM SPSS Statistics 27.Statistical analysis was done in contingency tables using the Pearson x 2 for categorical variable, Paired-Test and Mann-Whitney U test for continuous variable.Adjusted Odds ratios were calculated using multi-variable log binomial models.A p-value of less than 0.05 was taken as the level of statistical significance.

Ethical consideration
The study received ethical approval from the Sierra Leone Ethical and Scientific Review Committee on 18 th December 2018 and from the King's College London (HR-18/19-8467).Written informed consent was obtained from all participants and for those unable to give consent, assent was obtained from the next of kin.

Demographic characteristics of patients with stroke
A univariable analysis of patients tested for HIV on stroke admission, including patients already known to have HIV versus those who were not tested is presented in Table I.Characteristics of patients tested vs those not tested were similar except for age (p = 0.005).
Of the 511 stroke patients who were tested for HIV, 36 were positive and 475 were negative.The mean age of PLHIV was 49 years compared with a mean age of 58 years for HIV negative patients (p = <0.001).There were more females in the HIV positive group (52.8%) compared to 49.1% in the HIV negative group (p = 0.666).Hypertension was the most prevalent risk factor in both groups but was significantly less prevalent in PLHIV 23 (63.9%) versus HIV negative 409 (86.1%) (p = 0.001).The use of any contraceptive method among females was more prevalent in PLHIV (3, 15.8%) compared HIV negative population (13, 5.6%) (p = 0.004), however, dyslipidemia was slightly more prevalent in HIV negative population (Table II).
There were no significant differences in stroke type, clinical manifestation (NIHSS), and outcome (hospital discharge and death) between PLHIV and HIV negative groups.Postdischarge follow-up using Barthel index to assess activities of daily living at 7days, 90 days, and 1 year after discharge from the hospital showed a significant improvement at 1 year in both groups (P = 0.004) compared to earlier periods (Table II Based on the demographic characteristics of PLHIV (36 confirmed stroke cases) matched with 4 HIV negative population (144 confirmed stroke cases).Due to the matching process, the mean age was similar for both PLHIV (49.1 years, standard deviation 15.1 years) and HIV negative population (49.94 years, standard deviation 14 years) (p = 0.748).The gender distribution was also similar, as slightly over 50% of patients in both groups were female (p = 0.823) (Table IV).

Risk factors for stroke, stroke severity and stroke outcomes
Hypertension was the most prevalent risk factor for stroke in both groups, however hypertension was less frequent in the HIV negative population; 123 (85.4%)HIV negative patients and 23 (63.9%)PLHIV (p = 0.002).Prior stroke history was slightly more predominant in the HIV negative group (20, 13.9%) compared to the PLHIV (4, 11.1%) (p = 0.702) (Table IV).
Furthermore, clinical severity of stroke presentation according to NIHSS shows nosignificant difference in stroke presentation among HIV negative patients (17.1, SD 9.1) compared to PLHIV (15.3, SD 7.8) (p = 0.287).The Barthel Index -7 days post stroke which measures functional dependence outcome shows that PLHIV (25.9, SD 23.8) were not significantly different compared to HIV negative patients (34.7,SD 29.8) (p = 0.198).These comparative analyses could be limited due to the small number of HIV cases.
A higher proportion of PLHIV patients (25, 69.4%) were discharged home from the hospital when compared to HIV negative patients (95, 65.9%), and this difference was not statistically significant (p = 0.855).On the other hand, the proportion of in-hospital mortality rate was slightly lower among PLHIV (11, 30.6 %) compared to HIV negative (49, 34%) patients) (p = 0.693).However, the in-hospital mortality rate showed a significant correlation with NIHSS (22.80,SD 7.69, p = <0.001),and the severely disabled (mRS 04-05) at discharge have higher mortality at 7 days post stroke in both groups (p = <0.008)(Table V).
Of the 36 PLHIV, only 14 (38.9%) had a documented CD4+ cell count with a low mean cell count of 202.9 cell/mm 3 (SD 342.93).CD4+ cell count testing was done only for newly diagnosed HIV patients and was influenced by the availability of testing kits.HIV-Stroke patients with CD4 + counts results (using CD4 count as a proxy for "newly diagnosed") showed no significant change in clinical presentation severity (NIHSS) (15.9, SD 7.4, p 0.598) and mortality (61, SD 46.3, p = 0.306) (Table VI).

Stroke subtypes and pre/post-stroke status
Ischaemic stroke is the major stroke type reported in both groups, however, ischaemic stroke was less frequent in the HIV negative population: 77 (53.5%) compared to PLHIV:25 (69.4%) (p = 0.084).Other details on stroke type are shown in Table VII.Using the OCSP (Oxfordshire Community Stroke Project) classification; 33.3% of PLHIV have partial anterior circulation infarct (PACI) compared with 42.9%% of HIV negative population.But the frequency of some forms of ischaemic stroke is almost similar, as 28.6% of HIV negative population have lacunar cerebral infarcts, compared with 29.9% of PLHIV (Fig. 2, 3 & 4)

Discussion
This study is the first to examine the demographics, risk factors, stroke severity (NIHSS), stroke types, and outcomes for HIV-related strokes at the main adult referral hospital in Sierra Leone.The stroke register shows that stroke occurs 10 years earlier on average in PLHIV than HIV negative population, this is similar to studies done in West Africa 7,21 .HIV testing in these facilities is not age-biased, but the availability of test kits and human resources determines access to these services.The 10-year age gap observed in stroke onset among PLHIV could be associated with HIV-related factors such as endothelial dysfunction mechanisms, which result in vasculopathy, atherosclerosis, and minor vessel diseases.These factors may be interconnected to trigger stroke onset in PLHIV 6 .
Similar to the non-HIV population, hypertension is the most prevalent risk factor for stroke in PLHIV, however significantly different between PLHIV and HIV negative population.As recent studies in Sierra Leone have shown a high burden of hypertension in PLHIV and HIV negative population, these findings, similar to a study in Cameroon, support the current calls to address comorbid NCDs in PLHIV through integrated care in low-income countries [22][23][24] .
Although, the evidence of HIV as a risk factor for stroke in Africa is less clear, with conflicting study results [25][26][27][28] but in small studies of populations differing in HIV prevalence and medication take up.Ischaemic stroke is the most dominant stroke sub-type in both HIV and non-HIV stroke groups.The prevalence of ischemic stroke is similar to the SSA rate for PLHIV and HIV negative populations 29,30 .Previous studies reporting common subtypes of ischaemic strokes in both South Africa and Uganda support our findings and reinforces the fact that the management pathway for stroke and HIV should address practical challenges such as the HAART -pill access and burden, drug-drug interactions and side effects 29,30 .Furthermore, ischaemic stroke subtypes showed a different distribution in PLHIV and HIV negative population, presumably linked to the lower proportion of TACI.As this finding is linked to multiple factors, including HIV-induced endothelial dysfunction and prevailing opportunistic infections such as varicella zoster virus, especially in patients with low CD4 cell counts, it reflects on the call for early initiation of antiretroviral therapy and retention in HIV Care 31 .Furthermore, while hypertension was a most prevalent risk factor for stroke in PLHIV and HIV-negative populations, challenges with early HIV diagnosis and treatment is clearly noted in this study (Table VII), thereby underscoring our call to strengthen HIV care and reduce the risk of opportunistic infections and baseline metabolic changes in the endothelium and other tissues.
In both PLHIV and HIV-negative stroke patients, more than 50% of ischemic strokes sub-type were classified as unknown.This fact is attributable to the lack of important investigations such as carotid imaging, prolonged ECG monitoring and echocardiography in our setting.Owing to limited resources in the management of PLHIV in the country, CD4 cell count was reported for only 38.9% of PLHIV, 85% of whom have advanced HIV disease (CD4 cell count less than 200 cells/mm 3 ).We add our voices for sustainable plans and funding for CD4 cell counts as it provides strategic clinical and public health information in the management of HIV in Africa 32 .Although stroke severity (NIHSS) at admission did not show significant correlation to HIV-status of patients (p = 0.147), stroke severity (NIHSS) correlated with mortality (p = <0.001).Furthermore, the 90-day mortality rate in PLHIV (41.7%) and HIV-negative (47.2%) populations reflect the health system challenges faced by these hospitals and the high mortality burden of both diseases 12,14 .Nonetheless, in PLHIV and HIV-negative stroke survivors, Barthel indices at discharge, at 90 days and at 1 year showed significant improvement.
Our study has limitations.Challenges in obtaining other investigations to confirm stroke sub-type and inability to perform HIV testing and CD4+ cell counts for a large proportion of patients significantly impact the chances of comprehensively analyzing the relationship between stroke and HIV.Also, the study was limited due to the relatively small sample size of PLHIV and HIV-related data.The use of a hospital-based register means only admitted stroke patients could be studied.Nonetheless, this study is the first to provide information on HIV-associated stroke using a prospective stroke registry from two national referral hospitals in Sierra Leone

Conclusion
We reported stroke in HIV positive patients at the main tertiary hospital in Sierra Leone.This study shows that stroke I 10 years earlier in PLHIV than in the HIV negative population and hypertension was the most prevalent risk factor for stroke in both groups.These findings support the current call for timely diagnosis of HIV, HIV screening of stroke patients and addressing risk factors for stroke in PLHIV through integrated care.

Figure 2 .
Figure 2.Stroke types and location of confirmed HIV positive +ve vs matched HIV negative−ve in case-control study.

Figure 3 .
Figure 3.Location distribution of ischaemic stroke in confirmed HIV +ve vs matched HIV −ve in the case-control study.

Table I
Univariate analysis of tested for HIV and not tested for HIV in the stroke register, Connaught Hospital.Univariate analysis of confirmed HIV positive and HIV negative stroke patients in the stroke register at Connaught Hospital.

Table III
Stroke types and OCSP (Oxfordshire Community Stroke Project) classification of patients with known HIV status in the stroke register.
J Stroke Cerebrovasc Dis.Author manuscript; available in PMC 2024 September 01.

Table IV
Comparison of baseline demographic and clinical features between HIV positive and matched HIV negative stroke patients in case-control study.

Table VI
HIV-related factors on People Living HIV with stroke.Mann-Whitney U test.
J Stroke Cerebrovasc Dis.Author manuscript; available in PMC 2024 September 01.

Table VII
Stroke types and OCSP classification of confirmed HIV positive vs matched HIV negative in case-control study.