Background

The beneficial outcome of the highly active antiretroviral therapy (HAART) on people living with human immunodeficiency virus (PLHIV) is to increase quality of life and life expectancy [1]. The number of PLHIVs, aged over 50 years, has increased worldwide and is linked to an increased burden of noncommunicable diseases (NCDs). The double disease burden, HIV and NCDs in particular cardiovascular diseases (CVD), may cause a hectic burden on the economy of families and countries [2]. This situation requires global public health attention, particularly in low to middle income countries. From 1990 to 2015, the global burden of CVD attributable to HIV increased greater than three times from 0.74 million to 2.57 million disability-adjusted life-years [3]. A modeling study estimated that by the end of 2030 about 70% of PLHIV will be those aged older than 50 years, and 78% will have CVD [4].

A higher risk was observed of CVD among PLHIVs than that among HIV-negative individuals approximately 1.5 to 3 folds although the distribution of traditional risk factors for CVD among both groups did not differ [5,6,7,8,9]. Several recent studies suggested that underlying mechanisms of HIV-specific risk factors likely contributed to accelerated CVD in PLHIV, i.e., side effects of antiretroviral therapy (ART) and systemic chronic inflammation due to immune activation against HIV [5,6,7,8,9,10].

The Asia-Pacific region is the second region with a greater burden of HIV after sub-Saharan Africa, (5.8 and 25.7 million, respectively) [11]. The Asia-Pacific region has currently confronted the emerging challenge of CVD among PLHIV. A recent global burden of disease study revealed that the CVD population attributable to HIV was comparable with traditional risk factors [3]. Similarly, related studies suggested an increased incidence of CVD among PLHIV; however, most studies were conducted in high-income countries where epidemiologic evidence was unsuited for the Asia-Pacific region due to different socio-economic contexts [3, 7, 12–13]. To substantiate the situation of CVD and its risk factors among PLHIV remains indispensable for evidence-based public health in this low to middle-income region [3, 7, 9]. In this review, we addressed the knowable epidemiologic evidence of CVD among PLHIV in Asia-Pacific countries to provide existing scientific evidence to alert public health professionals in the region confronting the syndemic of HIV and CVD.

Methods

Search strategy

This systematic review of clinical and subclinical CVD among PLHIV in countries with the greatest impact of CVD attributable to HIV in the Asia-Pacific region, i.e., Thailand, Papua New Guinea, Bhutan, Cambodia, Myanmar, the Solomon Islands, Malaysia and Indonesia [3] was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [14] and reported following the Meta-analysis of Observational Research in Epidemiology guidelines [15]. This study has been registered in the INPLASY website with registration number INPLASY202290108 (https://inplasy.com/inplasy-2022-9-0108/).

The first author identified articles published in English through searching PubMed/MEDLINE, EMBASE, and the Cochrane Database of Systematic Reviews databases from any date to 31 December 2019 including Medical Subject Headings (MeSH) “Human immunodeficiency virus”, “People living with HIV”, “Cardiovascular”, “Cerebrovascular”, “Thailand”, “Cambodia”, “Myanmar”, “Bhutan”, “Papua New Guinea”, “the Solomon Islands”, “Malaysia”, or “Indonesia”. EndNote X8 (Clarivate Analytics, PA, USA) was used to collect, deduplicate, manage and review the searched articles. The detailed search strategy was provided in Additional file 1.

Study eligibility

To identify the eligibility of each study, the procedure was performed in a stepwise manner. First, the titles and abstracts of the identified articles were screened for appropriateness by the first author (WR) consulting with the senior author (MT). All identified articles through database searching under the search strategy were included. We excluded non-English articles, before HAART implementation articles, conference abstracts, case reports/case series and randomized controlled trials. We also excluded studies not using cardiovascular disease as the outcome, not involving PLHIV and conducted among nonadult populations (ages < 18 years). Next, the selected article full texts were independently reviewed by two authors (WR and MT) to collect pertinent data in greater detail. Study selection disagreements were resolved by authors’ discussion. Finally, we extracted data for qualitative synthesis.

Data extraction and quality assessment

Data were extracted on publication date, study area (country), study design, study period, sample size, age, CVD outcomes and measurements, incidence or prevalence of CVD outcomes among PLHIV and risk factors for CVD.

We used the Newcastle-Ottawa scale (N-O scale) to critically assess the quality of non-randomized studies [16]based on the study design, i.e., cohort, case-control and cross-sectional studies. The tool comprises three domains, i.e., participant selection, comparability and exposure/outcome assessment in the selected studies [17]. The N-O scale was then categorized in three levels following the standards of the Agency for Healthcare Research and Quality (AHRQ); good quality (> 3 stars in the selection domain, 1–2 stars in the comparability domain, 2–3 stars in the outcome/exposure domain); fair quality (2 stars in the selection domain and 1–2 stars in the comparability domain or 2–3 stars in the outcome/exposure domain); poor quality (< 1 star in selection domain or zero star in the comparability domain or < 1 star in the outcome/exposure domain) [17].

The data extraction and critical appraisal were conducted by the first author (WR). The second author (MT) independently checked and discussed all the results. In case of disagreement, the third author (NA) arbitrated. A detail of the quality assessment result is described in Additional File 2. We thematically analyzed and synthesized data from selected articles concerning the situation of CVD and risk factors which were classified as traditional risk factors and HIV-specific risk factors following Nou E et al. [18].

Results

Study selection and study characteristics

A total of 1,641 articles were identified through the literature search. After removing all duplicates, titles and abstracts of 1,467 records were screened, and 1,407 articles were excluded. Of the 1,407 excluded articles, 185 were published in the pre-HAART era, 75 constituted case reports and case series, 796 revealed irrelevant study objectives, 310 lacked cardiovascular outcomes and 41 lacked PLHIV. We further searched for 60 full texts and then excluded 14 abstract conferences, 6 systematic reviews, 5 studies among child populations, 19 studies in other regions and 5 studies not measuring CVD outcomes. Finally, 11 eligible studies were included in the summary and qualitative synthesis (Fig. 1). Among 11 eligible studies, the study quality of 9 studies was good, and 2 studies were poor (Additional File 2).

Fig. 1
figure 1

PRISMA Flow Diagram for Selection Eligible Study

Full size image

The 11 studies, comprised 7 cross-sectional studies, 2 retrospective cohort studies, 1 prospective cohort study and 1 case-control study. Among 8 counties with the greatest impact of CVD attributable to HIV in the Asia-Pacific region, we found eligible publications from only 3, i.e., Indonesia, Malaysia and Thailand. The sample size of each study ranged from 50 to 1,813 subjects (Table 1).

Table 1 Published studies on CVD outcomes among PLHIV in Asia-Pacific counties with the greatest impact of CVD attributable to HIV status
Full size table

CVD among PLHIV in the Asia-Pacific Region

Early detection of CVD was reported in 11 studies, 7 studies identified subclinical CVD diagnosed by CAVI [22], cIMT [19, 23,24,25,26], and ABI [19, 22, 27] and another 4 studies identified clinical CVD, i.e., stroke [20, 21, 28, 29] and atherosclerosis [21, 28].

Subclinical CVD, measured using ABI and cIMT reported in seven studies, ranged from 10 to 28% [19, 23–27]. A lower prevalence of subclinical CVD was reported in a study in a younger PLHIV population [25] than that in other studies [19, 20,21,22,24, 26, 27]. The prevalence of subclinical CVD such as atherosclerosis was higher in PLHIV than in nonHIV populations [19][23].

Clinical CVD, stroke and coronary artery diseases were reported in two cohort studies ranging from 0.37 to 1.17 /100 person-years [20–21] and the prevalence of 1.3% from a cross-sectional study [28] among PLHIV in Thailand. However, the incidence of ischemic stroke (2.35/100 person-years) was higher than hemorrhagic stroke (1.17/100 person-years) [20]. Clinical CVD was frequently observed in the early highly active ART era, and a higher incidence was observed in longer follow-up time.

Risk factors for CVD among PLHIV

The characteristics of CVD risk factor data among PLHIV collected by selected studies were classified in two groups: traditional risk factors and HIV-specific risk factors for CVD including adverse effects of antiretroviral therapy and factors related to systemic immune activation and HIV status [22].

Traditional risk factors for CVD among PLHIV

The traditional risk factors for CVD among PLHIV in the selected studies were frequently assessed. Moreover, these risk factors were prevalent in both PLHIV and nonHIV populations. Hypertension presented at 13 to 45%, diabetes mellitus was 5 to 24% [21, 24, 26, 27–28] and smoking was 13 to 45% [25–29]. Although the prevalence of traditional risk factors between the PLHIV and the nonHIV populations did not conclusively differ [21,22,23, 26], DM, dyslipidemia and family history presented synergistic effects on CVD risks reported in the PLHIV cohort.

HIV-specific risk factors for CVD

Although several selected studies intended to identify associations between HIV-specific risk factors and CVD among PLHIV [21,22,23,24,25,26, 28, 30], only 3 of 11 studies demonstrated the association of HIV-specific risk factors and subclinical or clinical CVD status among PLHIV in this region [24]. Poor immune system (CD4 cell count < 200 or CD8/CD4 ratio < 1) increased the risk of subclinical CVD [24, 25]. A cross-sectional study in Thailand reported the association between low nadir CD4 counts (< 200 cells/mm3) and carotid artery stenosis (abnormal cIMT > 0.9 mm) and/or presence of carotid plaques (adj OR 1.80; 95%CI 1.02–3.18) [24] similar to a study in Indonesia (β= -0.791) [25]. One cohort study reported a relationship between the duration of ART exposure with abnormal cIMT [19]. However, other HIV-specific risk factors such as antiretroviral therapies, duration of antiretroviral therapies exposure, statin use, fibrosis-4 index and high sensitivity c-reactive protein did not present any association [24]. Although clinical CVD prevalence between nonHIV and PLHIV did not significantly differ in both populations aged above 55 years, the onset of subclinical CVD such as abnormal cIMT and carotid plaques was earlier among the PLHIV than among nonHIV [19, 22,23,24].

Discussion

This systematic review revealed a higher prevalence of subclinical and clinical CVD among PLHIV than that in the nonHIV population, a higher risk of subclinical CVD among poorer immune PLHIV and synergistic interaction between diabetes mellitus, dyslipidemia and family history on CVD risk among PLHIV. Although this systematic review underscored the significant risk of CVD among PLHIV in the Asia-Pacific region, the limitation in existing data remains. First, published data included small sample sizes which were less likely to detect the statistical association. Additionally, cohort studies did not have a longer follow-up time and were less likely to capture clinical CVD risks. NonHIV populations in three cross-sectional studies were not tested for HIV so the nonHIV populations likely mixed with populations with and without HIV. Those limitations can lead to underestimating the CVD risk; therefore, the CVD risk among PLHIV in this study can be used to raise attention from public health professionals in this region. Additionally, most studies used a cross-sectional design which did not capture the risk of CVD over time, and our study includes only English publications to prevent the challenges of accurate translation. This study did not meta-analyze the pooled effect of CVD risk factors because HIV/CVD studies are very few and likely inadequate for statistical analysis.

Obviously, the prevalence of subclinical atherosclerosis among PLHIV are the same as that in nonHIV populations in those study populations aged above 50 years [22–23, 26]. This situation reflected the double burdens of diseases in this region. The PLHIV regularly visit hospitals and attend health education sessions to improve healthy lifestyles to reduce the risk of NCDs [22, 33]. However, a prospective cohort study reported a higher prevalence of subclinical CVD compared with nonHIV populations which was consistent with data from other parts of the world [3, 5, 8–9, 35]. The higher CVD risk among PLHIV is related to multifactorial factors, both traditional and HIV-related risk factors for CVD [5,6,7,8,9].

The associations between the major traditional risk factors and CVD among PLHIV found in our review were consistent with results from other related reviews [6–7, 36]. These factors including diabetes, hypertension, dyslipidemia and smoking served major roles to increase oxidative stress in the cardiovascular system and led to chronic systemic inflammation, endothelial dysfunction, atherosclerosis progression and direct effects on cardiac performance through abnormal hormones or cytokines secretions. [36–42] Identifying extremely high risks of CVD among PLHIVs with DM, dyslipidemia and family members with CVD remains crucial.

The existing data were only from cross-sectional studies which cannot assess all potential exposures and CVD events over time. the prospective ascertainment for CVD among PLHIV and potential exposures will provide crucial information to identify future optimal interventions in this region. Although the association between HIV-specific risk factors and CVD has been reported in other studies [7, 42], HIV management may differ across regions especially between high and low to middle income countries. The study of HIV-specific risk factors for CVD and synergistic effects between traditional and HIV-specific risk factors should be conducted in this region.

Conclusion

The limited existing data suggested the risk of early CVD among PLHIV. Extreme CVD risk among PLHIV with DM, dyslipidemia, and family history should highlight the need of NCDs intensive prevention program. We identified the crucial gaps in HIV/CVD work from the Asia-Pacific Region and recommended a further prospective study with large sample size and longer follow-up time or conducting a meta-analysis to better capture CVD risk, and interaction between HIV-related and traditional risk factors in this vulnerable population in this region.