Background

Infectious diseases pose a concerning threat to the health systems of both developed countries and countries with limited resources such as, for example, sub-Saharan countries [1, 2]. With 228 million malaria cases globally in 2018, future declines in the malaria burden caused by Plasmodium spp. infections are uncertain [3, 4]. Approximately 3.3 billion people are residing in malaria-endemic regions (parts of the Africa, Southeast Asia and Middle East) [5, 6].

The human immunodeficiency virus (HIV) is an emerging infectious disease agent defined by cellular immune system impairment [7]. HIV is a well-established global health burden, with > 36 million HIV-infected patients and > 1 million HIV-related deaths in 2017 [8]. While Plasmodium parasites causing human malaria are transmitted mainly by mosquitoes (Anopheles spp.) serving as biological vectors, malaria can also be transmitted directly via blood transfusion, needle sticks with contaminated needles and vertical transmission [9, 10]. The infection routes bypassing the biological vector are transmission routes shared by HIV and malaria [11]. Since HIV infection affects the immune system, the infected individuals are more susceptible to other infections [12,13,14,15]. Therefore, people living with HIV (including children, pregnant women and adults) are at risk for significant disease and may have fatal complications following infection [11, 16]. The vertical transmission option for both malaria and HIV facilitates co-transmission from infected pregnant women to their infants [17]. Since the co-infections of malaria and HIV can induce anemia, blood transfusion is often required, but blood transfusion can also contribute to the transmission of HIV and malaria [18, 19].

Although numerous studies have highlighted malaria prevalence in patients with HIV, there has been no comprehensive meta-analysis to demonstrate this prevalence in children, adults and pregnant women. Therefore, the aims of this systematic review and meta-analysis are to summarize malaria prevalence among HIV-positive children, pregnant women and adults, and to identify risk factors that increase the probability of HIV-positive patients being infected with malaria.

Methods

Search strategy

For inclusion in the present systematic review, the PubMed, Science Direct, Google Scholar, Scopus and Cochrane databases were searched for relevant English-language, full-text articles and abstracts published between 1 January 1983 (date of HIV discovery) and 1 March 2020. As the aim was to evaluate the prevalence of positive test results for malaria among HIV-positive and HIV-negative individuals, the following Medical Subject Headings (MeSH) terms were used: “Malaria” OR “Plasmodium” AND “prevalence” OR “epidemiology” OR “co-infection” AND “HIV” OR “AIDS” OR “acquired immune deficiency syndrome” OR “immunocompromised” OR “immunosuppressed” OR “immunodeficiency” AND “pregnancy women” OR “children” OR “adult” alone OR combined using “OR” and/or “AND”.

Study selection and data extraction

After an initial search of the databases, subject-related topics and their abstracts were double-checked, and then full texts of potentially eligible articles were selected for downloading. All potentially relevant full texts were reviewed by three independent reviewers (TM, HS, ASP). Discrepancies were resolved by discussion and consensus. The studies were assessed for quality using the Joanna Briggs Institute (JBI) checklist (Additional files 2, 3, 4, 5: Tables S1–S4). The required data were extracted by the reviewers and then re-checked. The criteria for inclusion in the review were: (i) peer-reviewed original research papers; (ii) cross-sectional and cohort studies that estimated the prevalence of malaria infection in HIV-positive and HIV-negative individuals; (iii) published papers in English; (iv) published online before 1 March 2020; and (v) sufficient sample size (n > 10). Any article that did not satisfy the above criteria were excluded. The reference lists of the eligible articles were also browsed manually to identify relevant papers that were not initially identified in the database search. Finally, details of each study were extracted using a data extraction form, including country, year of publication, first author, number of HIV+ and malaria-positive cases, education status of patients, alcohol consumption status, number of partners, marital status, level of CD4+ (cluster of differentiation 4) in HIV-positive patients, ART (antiretroviral therapy) status, sex protection status and diagnostic method (microscopy, serology or molecular). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were used to report the findings [20].

Meta-analysis

The point estimate and corresponding confidence interval (CI) for the prevalence of malaria in HIV-positive individuals for each study were calculated. Incoherence and heterogeneity among studies were assessed using the I2 index and Cochran’s Q test, respectively, and the random-effects model (DerSimonian-Laird) was used for analysis. The heterogeneity among subgroups was tested by meta-regression analysis. The relationship between prevalence, year of publication and sample size was estimated by meta-regression. Additionally, a funnel plot relying on the Egger’s regression asymmetry test was used to assess the small effects of the study and the population bias. For the meta-analysis, the included studies were evaluated as a random sample of each study population, and the analyses were performed using StatsDirect (version 2.7.2) statistical software (StatsDirect Ltd., Altrincham, UK).

Results

The systematic search of the electronic databases identified 24,311 potentially relevant papers. The full-text of 212 articles was assessed, resulting in exclusion from the study of 106 papers due to their small sample size, the review or case report nature of the report, duplication and insufficient data. The remaining 106 papers fulfilled the inclusion criteria and were included in the present systematic review and meta-analysis. All of these 106 articles were published between 1983 and 2020 and present data from malaria-endemic regions in Africa (n = 103) and Asia (n = 3). The inclusion/exclusion criteria at each step of screening and eligibility and the number of selected papers are shown in Fig. 1.

Fig. 1
figure 1

Flowchart of study selection process

Full size image

All analyses were conducted in three subgroups: children (n = 17; Table 1; Fig. 2), adults (n = 57; Table 2; Fig. 3) and pregnant women (n = 32; Table 3; Fig. 4). The pooled malaria prevalence among HIV-positive children was 39.4% (95% CI = 26.6–52.9). The combined prevalence of malaria in HIV-positive adults was 27.3% (95% CI = 20.1–35.1), and the collective malaria prevalence among HIV-positive pregnant women was 32.3% (95% CI = 26.3–38.6) (Figs. 2, 3, 4). The funnel plot showing a statistically significant Egger’s regression suggests the possibility of publication bias (Additional file 1: Figure S1). The published risk factors associated with HIV and malaria, namely CD4+ level, ART consumption, sex, education, gravidity and age, were analyzed (Table 4). In adult patients with HIV, CD4+ count < 200 cells/µl predisposes the patient to malaria infection (odds ratio [OR] = 1.5, 95% CI = 1.2–1.7). In adult HIV-positive patients, age < 40 years old was found to be associated with a significant increase in the odds of being infected with malaria (OR = 1.1, 95% CI = 1–1.3). Also, for adult HIV-positive patients, being male and being treated with ART medication have been associated with a significant decrease in the odds of being infected with malaria (OR = 0.8, 95% CI = 0.7–0.9 and OR = 0.2, 95% CI = 0.2–0.3, respectively). CD4+ count < 200 cells/µl was found to predispose pregnant women with HIV to malaria infection (OR = 1.5, 95% CI = 1.1–1.9) (Table 4).

Table 1 Baseline characteristics of the included studies on malaria and human immunodeficiency virus co-infection in children
Full size table
Table 2 Baseline characteristics of the included studies on malaria and human immunodeficiency virus co-infection in adults
Full size table
Table 3 The baseline characteristics of the included studies on malaria and human immunodeficiency virus co-infection in pregnant women
Full size table
Fig. 2
figure 2

Forest plot diagram of malaria prevalence in human immunodeficiency virus-positive children (first author, year and country)

Full size image
Fig. 3
figure 3

Forest plot diagram of malaria prevalence in human immunodeficiency virus-positive adults (first author, year, and country)

Full size image
Fig. 4
figure 4

Forest plot diagram of malaria prevalence in human immunodeficiency virus-positive pregnant women (first author, year, and country)

Full size image
Table 4 Risk factors associated with malaria infection in human immunodeficiency virus-positive patients
Full size table

Discussion

Although extensive studies have been conducted on both HIV and Plasmodium spp. infections, a comprehensive meta-analysis aimed at precisely evaluating the prevalence of malaria infections among HIV-positive patients and related risk factors is lacking. Therefore, the aim of the present meta-analysis was to provide the pooled prevalence of malaria infection in HIV-positive children, pregnant women and adults and evaluate the related risk factors. The included studies represent African and Asian regions where both HIV and Plasmodium spp. are endemic. The pooled malaria prevalence in HIV-positive children, adults and pregnant women included in these studies was 39.4% (95% CI = 26.6–52.9), 27.3% (95% CI = 20.1–35.1) and 32.3% (95% CI = 26.3–38.6), respectively. In adult patients with HIV, receiving ART and having CD4+ count > 200 cells/µl were two factors significantly associated with malaria infection (P < 0.05).

Due to widespread ART coverage, mortality due to HIV as the main cause of death has decreased drastically over the years [8]. Notwithstanding the extensive efforts to end the acquired immunodeficiency syndrome (AIDS) epidemic by 2030 (set down in the Joint United Nations Program on HIV/AIDS), a lot of the work remains to be done [122]. The troublesome high prevalence of HIV, the increased life expectancy of affected patients, the common co-transmission of HIV and malaria and a remarkable geographical overlap between malaria and HIV high prevalence areas have paved the way for higher rates of co-infections in HIV-positive individuals [123].

Although the incidence of malaria and mortality due to malaria declined significantly by 62% and 41%, respectively, between 2000 and 2015, WHO reported that malaria remained an endemic disease in 76 countries at the beginning of 2016 [124], with approximately 216 million malaria cases in that year. Fifteen countries of the sub-Saharan African region alone were reported to be responsible for 80% of the total malaria burden [125]. Therefore, it is believed that many challenges remain to be overcome in order to eliminate malaria [126]. Regarding the burden of HIV and malaria and the immunosuppressive nature of HIV, there is an urgent need to clarify malaria prevalence in HIV-infected patients and the related risk factors.

According to the results of this systematic review and meta-analysis, the majority of published HIV/malaria studies to date have been in African countries. Socioeconomic conditions and a desirable climate for the biological vector, both of which can facilitate malaria transmission, may be the main reasons underlying this result [127]. Based on our findings, more than one-third of pregnant and HIV-positive women have been infected by malaria, which is worrisome because of the vertical transmission nature of malaria and HIV, which predisposes neonates to other infectious diseases [128, 129]. Indeed, pregnant women are among the most susceptible and vulnerable groups infected with malaria due to the altered immune system during pregnancy [3, 130]. The weakened immune response and HIV infection can lead to even deeper attenuation of the immune system. It is well-recognized that a decline in CD4+ cell numbers is associated with attenuation of the cell immune system and an increased vulnerability to being infected with other infections [131]. Our finding that CD4+ cell count < 200 cells/µl is linked to increased susceptibility to malaria infection (OR = 1.5, 95% CI = 1.1–1.9) confirms this association. In essence, AIDS and malaria are each controlled by adaptive and innate immune mechanisms, and declining immunity caused by HIV infection will cause an increase in malaria severity. CD4+ cells are depleted by the HIV virus, which leads to an impaired immune response to many pathogens, including Plasmodium spp. [43]. This pattern was corroborated by Grimwade et al. [132] who observed that malaria incidence in persons with CD4+ T cell count ≥ 500/µl, between 200 and 499/μl and < 200/μl was 57, 93 and 140 per 1000 person-year, respectively, in Uganda. It has been postulated that HIV increases malaria incidence in adults based on CD4+ cell count categories [133].

This meta-analysis also revealed the worrying situation of malaria infection among HIV-positive children. Approximately 39.4% of HIV-positive children in the analyzed studies were infected with malaria. This is a much higher prevalence than that observed in several studies investigating general children populations in African countries, with the prevalence in these studies ranging from 1% in Kenya to 22% in Uganda, with 14.5% prevalence in Tanzania and 20% in the Democratic Republic of Congo [134, 135]. The observation of increased malaria prevalence in HIV-positive children supports our assumption that susceptibility to co-infection is high in HIV-positive individuals. It is interesting to note that much of the pathogenesis of malaria during pregnancy is mediated by the accumulation of Plasmodium-infected red blood cells in the placental intervillous space, termed ‘placental malaria.’ The placenta is also the key interface in mother-to-child transmission of HIV, especially that involving in utero transfer [136]. No remarkable association between receiving ART and HIV infection status has been noted in HIV-positive children (OR = 1.3, 95% CI = 0.2–6.6). Moreover, there has been no significant association between the CD4+ cell count and the probability of malaria infection (P > 0.05), possibly due to the small number of studies that have considered this factor.

The present meta-analysis reveals that, on average, 27.3% of HIV-positive adults are infected with malaria in endemic countries. One of the consequences of this alarmingly high figure can be manifested in blood transfusion. With the ever-increasing need for a blood transfusion due to environmental and heredity diseases such as sickle cell anemia [137], the prevalence of transfusion-transmitted HIV/malaria can be expected to be high. A study conducted in the sub-African region has demonstrated that about 10–15% of HIV transmission is related to blood transfusion [138]. Ahmadpour et al. [19] reported that transfusion-transmitted malaria is a significant challenge in sub-Saharan African regions. In terms of risk factors, CD4+ cell count of < 200 cells/µl predisposes HIV-positive adults to Plasmodium spp. infection (OR = 1.5, 95% CI = 1.2–1.7). However, the association between malaria and HIV is more complex than expected. Some studies have corroborated that CD4+ T cells, as the prime targets for reproduction by HIV-1, play a vital role in immune responses to malaria [131, 139]. Malaria infection leads to upregulation of proinflammatory cytokines and stimulates CD4+ cell activation, thus providing the ideal microenvironment for the spread of the HIV virus among the CD4+ cells. On the other hand, the selective infection of CD4+ cells by HIV leads to the loss of these cells [140]. It is assumed that the increased susceptibility of HIV-seropositive individuals to malaria is related to some immune system-modulating mechanisms, such as depletion of CD4+ cells [131, 141].

Age < 40 years has also been associated with a significant increase in the chance of HIV-positive adults becoming infected with malaria (OR = 1.1, 95% CI = 1–1.3). In HIV-positive adults, being male and receiving ART have been associated with a significant decrease in the risk of being infected with Plasmodium spp. (OR = 0.8, 95% CI = 0.7–0.9 and OR = 0.2, 95% CI = 0.2–0.3, respectively). This is an interesting finding when compared to individual studies that described a higher risk of malaria infections in males compared to females in the general population in north-east Tanzania, irrespective of their HIV status [134]. Thus, it appears that HIV status may potentially alter malaria susceptibility differently in male patients than in female patients. It is worth emphasizing that the reported figures may not reflect the current status of this co-infection because these endemic areas are limited in terms of healthcare resources, and testing may not be conducted on all people unless they show clinical symptoms. Furthermore, there is insufficient evidence to determine whether or not malaria-induced changes in CD4+ T cell counts or viral loads translate to accelerated HIV disease progression or death in areas of stable malaria transmission.

This is the first meta-analysis on malaria prevalence among HIV-positive patients. We broke down the data into three categories, namely infancy, pregnancy and adulthood, and identified the available risk factors for each group. Since there has been little research on the prevalence of malaria in HIV patients in malaria endemic areas, further studies are needed in this regard. Also, due to the incomplete data in the studies included in our meta-analysis, we were unable to evaluate some risk factors, including duration of illness, time of diagnosis and response to treatment. Unfortunately, no data on the health status of individuals having both malaria and HIV infection were provided in these studies. On the other hand, publication bias is one of the main concerns in systematic review studies. As expected, publication bias was observed in the analyzed studies. The main limitation of this systematic review and meta-analysis is related to the different study designs and varying laboratory methods used to determine infection status. Diagnostic methods have varying sensitivity and specificity and, therefore, the heterogeneous prevalence data reported may partially be caused by flaws in methodology. The use of an accurate, reliable and uniform diagnostic techniques would support the correct interpretation of results.

Conclusions

The current systematic review has revealed concerning prevalence data for malaria among HIV-positive persons, including children, adults and pregnant women. In view of the fact that malaria can quickly become a life-threatening condition in risk groups (e.g. people living with HIV), prevention, chemoprophylaxis, early diagnosis and treatment of clinical malaria are recommended. Recent information also indicates that malaria is associated with the availability of ART and CD4+ cell count numbers in adults. Therefore, the related risk factors should be given appropriate attention in HIV/malaria co-infected patients. As HIV infection affects the host immune response, future studies are needed to elucidate the pathogenesis aspects of this co-infection, as well as the severity of its complications, and to investigate possible drugs and drug effectiveness.