Background
In the past decades, intensive malaria interventions have resulted in a dramatic decline in global malaria morbidity and mortality. However, the global burden of malaria is still enormous: in 2016 there were about 216 million cases resulting in ~ 445,000 deaths [
1]. The Global Technical Strategy for Malaria 2016–2030 was endorsed by World Health Assembly, calling for a reduction in global incidence and mortality of malaria by at least 90% by 2030 [
2]. This ambitious plan should include all types of malaria infections, including severe and complicated, mild and uncomplicated, and asymptomatic infections [
3,
4]. Compared with asymptomatic malaria, acute malaria infections are often quoted as the “tip of an iceberg”, while the overwhelming majority of malaria infections are asymptomatic as identified by molecular methods with increasing sensitivities [
5‐
8]. Asymptomatic carriers of malaria are prevalent in both low- and high-endemicity regions and they are important reservoirs for sustaining malaria transmission because they persist for long time and harbour gametocytes that are infectious to
Anopheles mosquitoes [
9‐
15]. With regard to infectivity of sub-microscopic infections to mosquitoes, some studies have even reported that parasites from asymptomatic individuals were even more infectious to vectors than those from symptomatic cases [
16‐
18]. Since the prevalence, dynamics and transmissibility of asymptomatic malaria vary geographically and are influenced by complex factors involving parasites, hosts and environments [
19,
20], identification of these factors is particularly important for malaria control.
In the Greater Mekong Sub-region (GMS) of Southeast Asia, malaria distribution is heterogeneous and often concentrated along international borders [
21‐
24]. Border regions are not only difficult to access, but human migration across the porous borders also poses further danger of malaria re-introduction [
25]. Border malaria, a shared phenomenon of the GMS, is a major challenge for achieving the regional malaria elimination goal by 2030 [
26]. It requires coordinated control efforts from neighbouring countries. Recently, the asymptomatic
Plasmodium reservoir has been reported to be considerable, even in low-transmission settings along the international borders in GMS, including Thailand–Cambodia, Thailand–Myanmar and China–Myanmar borders [
8,
27,
28]. Within the GMS, Myanmar has the highest malaria burden. In addition, as a consequence of internal military conflict, refugees and internally displaced people (IDP) rushed to and settled down along the international borders. The overcrowded and poor living conditions in these camps increase the risk of infectious diseases [
29‐
31]. In the IDP camps and the nearby villages of Kachin State at the China–Myanmar border, previous surveys have detected continued malaria transmission and observed
Plasmodium vivax outbreaks in recent years [
32]. The potential spillover of disease outbreaks to neighbouring communities demands strengthened control efforts and close surveillance.
With continued transmission of malaria in the China–Myanmar border, this study wanted to address the potential of asymptomatic
Plasmodium infections as important reservoirs. The use of highly sensitive molecular detection methods allowed the detection of almost 20% of local residents as carriers of asymptomatic
Plasmodium infections [
8]. In the present study, further efforts were undertaken to study the seasonal dynamics and risk factors of asymptomatic malaria infections in the IDP camps and surrounding villages. Knowledge gained from this study may help identify a potential bottleneck transmission season and high-risk human populations to implement targeted radical elimination practices.
Discussion
Asymptomatic malaria was once thought the ‘forgotten’ malaria, but it is a major gametocyte reservoir to conduce mosquito infection and thus a major barrier for malaria elimination [
37]. In addition to the complicated features of border malaria observed elsewhere, settlements for refugees and IDP along the borders add another layer of complexity to the existing border malaria problem. This study provided further evidence about the significance of asymptomatic malaria as the reservoir for continuous malaria transmission, and also identified the individual- and household-level factors associated with asymptomatic malaria infection along the China–Myanmar border. This study found increased risks of asymptomatic infections in (1) males, (2) children (≤ 15 years), (3) those who failed to use bed net and IRS, (4) those who lived in poor houses, and (5) those lived farther away from a clinic. These findings provide the needed knowledge for guiding the malaria control and elimination program in border regions and in IDP camps.
Routine microscopy with a detection limit of 50–100 parasites/μL severely underestimates the burden of asymptomatic infections [
38]. In this study, nRT-PCR was confirmed to be a more sensitive technique for identification of asymptomatic
Plasmodium infections, revealing more than 15-fold higher prevalence of asymptomatic infections than microscopy. The results were consistent with those from other parts of GMS such as Thailand and Cambodia [
39‐
41], which showed that the majority of the asymptomatic infections were sub-microscopic. Despite that the prevalence was lower than in hyper-endemic regions in Africa [
42,
43], the > 20% prevalence of vivax infection is astounding and may explain the persistence and outbreaks of acute vivax malaria in this region. Like the acute malaria incidence detected from passive case surveillance [
24,
32], asymptomatic malaria along the border also exhibited seasonal variation. Compared with the dry season, the higher odds of infection during the rainy season might be related to increased density and species richness of the vectors [
44,
45]. Interestingly, seasonal variation was more obvious in the villages, which might be related to the different vector ecology and control efforts in villages and camps [
32].
Among the demographic factors, male gender and young children (≤ 15 years) were significantly associated with asymptomatic infections. Such a gender bias of malaria infection has been reported in other studies [
19,
46‐
49]. In this case, males were mostly farmers and soldiers, who are engaged in more agricultural and forest-related activities, which might increase the chances of infection [
50]. Age is considered one of the most important factors associated with protective immune response in malaria endemic areas. Many previous studies revealed adults are asymptomatic parasite carriers because they have acquired strong immunity from repeated exposures to malaria parasites, whereas infections in young children are symptomatic because their anti-malarial immunity is still developing [
51‐
53]. Thus, the finding that school children (≤ 15 years) had higher odds of asymptomatic malaria infections seems contradictory to findings from Africa. Yet, this finding corroborated our earlier studies, which showed that school children had increased odds of having acute vivax malaria [
32,
33]. Consistent results were also obtained from a series of recent studies revealing that school-age children were an under-recognized reservoir of malaria infection showing either higher levels of asymptomatic infection or gametocyte carriage [
54‐
56]. Such higher prevalence of asymptomatic infections in school-age children might be resulted from increased
Plasmodium exposure due to neglected bed net use and drug treatment. A study conducted in Papua New Guinea showed that immunity to
P. vivax may be acquired at a younger age than that seen with
P. falciparum [
57], which might be another reason for what was observed in this study. This highlights that malaria control initiatives should specifically target school children for malaria elimination along the border.
Malaria transmission is mainly influenced by two key factors: gametocyte infectivity and vector density. Thus, effective vector-based malaria control measures such as the use of IRS and bed nets, especially long-lasting insecticide-treated nets (LLINs) are particularly important in reducing malaria incidence [
58]. This study found that non-users of bed net or IRS had much higher risks of asymptomatic malaria infection in whichever seasons or sites. Moreover, the risk was much higher in rainy season than in dry season. Additionally, odds in the surrounding villages were 2.34 times higher than those in the IDP camps, which may be largely due to higher vector densities in villages [
32,
44]. This result was in line with the higher asymptomatic malaria infection rates in local villages than in IDP camps in the peak transmission season [
31,
32,
59]. Vector control interventions were insufficient in local villages as compared to those in the IDP camps, as the latter were helped by multiple non-governmental agencies. Bed nets in local villages were mostly conventional (> 99%), whereas > 60% of the nets in the IDP camps were LLINs, and more frequent IRS were performed in the IDP camps during the rainy season [
32]. Interestingly, such a trend of asymptomatic infections was reversed in the early dry season, when mosquito density is low. Although reasons for this are not clear, it is possible that the higher asymptomatic infection rate in the IDP camps might be due to migration of human populations from more interior areas with higher malaria endemicity as the camp population grew [
29,
30]. This study emphasized the importance of vector control strategies in reducing asymptomatic infections.
This study also found that household and the surrounding environmental factors affected the prevalence of asymptomatic malaria. Firstly, houses constructed with wood or bamboo increased the risks of infection, corroborating findings from other studies [
60‐
62]. Since the house structures reflect the socio-economic status of the family, it may correlate with the use of personal protection measures [
63,
64]. Interestingly, longer distance to the clinic is another factor associated with increased asymptomatic infections. The distance from residents in JHK and MSY villages to the nearest clinic is almost 1500 m, and half of the recruited populations from these 2 villages were infected. One reason might be due to potential delays for patients from these 2 villages to seek treatment, especially during the rainy season when short travel becomes prohibitive due to poor road conditions [
65]. Other factors associated with the locations of the houses likely reflect the surrounding environment. For example, the higher prevalence of infections in these 2 villages might be due to their closer proximity to forest edges, where mosquito vectors associated with forested areas are more abundant as shown from the earlier study [
32].
The findings of these 3 seasonal cross-sectional surveys have provided insights about persistent malaria transmission along the China–Myanmar border, which is needed for malaria elimination on both sides of the border. A sensitive molecular method verified the presence of a large submicroscopic reservoir of malaria parasites in both village and IDP populations. This study also identified risk factors associated with increased prevalence of asymptomatic infections in both populations. On the individual level, males, school children, and those who did not use bed net were the as high-risk populations. On the household level, worse living conditions, failure to use the mosquito control-based interventions and farther away from the clinic were also associated with increased risks of parasite infections. While increasing the coverage of effective malaria interventions will be the key to control border malaria, additional measures such as mass drug administration (MDA) may be considered in order to reduce the infected population. MDA with chloroquine and primaquine has been found to be successful in eliminating vivax malaria in central China [
66]. Furthermore, better management of malaria cases such as the recently described ‘1-3-7’ strategy is needed in the final phase of malaria elimination [
67].
This study has several limitations. First, the data collection was confined to the available residents in the households, which led to the female-biased study population and over-representation (54.6%) of office/in-house workers, which undermines the generalization of the findings. Given that males, farmers and soldiers might be the high-risk populations, future studies should ensure a better representation of the study population. Second, the use of convenience samples inevitably adds additional biases, and future study designs need to be more systematic. Third, the study did not take into account the effect of human migration on malaria transmission. It is important to find out whether and how much migration of people from more malaria-endemic interior areas to the camps contributes to the malaria transmission. Furthermore, the significance of cross-border migration for malaria importation to the neighbouring Yunnan province of China also deserves special attention. Finally, the comparison between the local villages and camps needs to be more quantitative since their accesses to malaria treatment and prevention services were different. Because the whole region is targeted for malaria elimination, provision of quality malaria control services to the entire populations will ensure achieving the goal.
Authors’ contributions
YZ planned and performed the field and experimental work and analyzed data and engaged in drafting manuscript. JZ, YHZ and JQZ participated in field and experimental work. QL, YH and ZY participated in data analysis. YC, LC, QW and QF conceived the study and commented on the manuscript. All authors read and approved the final manuscript.