Asymptomatic only at first sight: malaria infection among schoolchildren in highland Rwanda

Plasmodium infection can cause a wide variety of illnesses, ranging from asymptomatic infection over uncomplicated malaria to severe and complicated disease [1‐4]. Owing to the enormous morbidity and mortality of clinical malaria in young children, the epidemiology and consequences of asymptomatic infection in different settings and age-groups has received comparatively little attention. Increasingly, however, the underrated impact of asymptomatic and/or submicroscopic infections is now being recognized [2, 3, 5‐10].

Partial immunity against malaria develops during childhood in endemic regions as a result of repeated exposure. Increasing age has consequently been associated with low and submicroscopic parasite density as well as with asymptomatic infection, particularly among school-age children [5, 11‐14]. Most of these infections remain undiagnosed and untreated [14, 15], and—in the case of schoolchildren—are not subject of targeted interventions. However, the significance of these asymptomatic Plasmodium infections is likely underestimated as they contribute to inflammatory responses, anaemia, cognitive impairment and delayed physical development [5, 9, 10, 15‐20]. Moreover, submicroscopic infections tend to persist for longer than symptomatic ones [13, 21], thereby constituting a sizeable transmission reservoir [2, 3, 14]. This has particular importance in areas of low endemicity considering the increase of submicroscopic (and frequently asymptomatic and untreated) infections with declining malaria transmission [2, 3, 11, 21].

In Rwanda, comprehensive malaria control measures since 2005 have been paralleled by 50% or higher declines in malaria morbidity and mortality [22]. In 2010, in the southern highland Huye district, PCR-assessed Plasmodium prevalence among rural children under 5 years of age was 16%, the majority of infections being asymptomatic [18]. This area represents a hypo- to meso-endemic setting as seen in many other parts of the country [12]. No data, however, is available as to the epidemiological characteristics and effects of seemingly asymptomatic and/or submicroscopic Plasmodium infections among school-age children living in the highlands of Rwanda. Such knowledge, nevertheless, forms the prerequisites for the design of targeted interventions, the reduction of transmission reservoirs, and, ultimately, elimination.

In this cross-sectional survey among largely afebrile children attending school in southern Rwanda, 22% were infected with Plasmodium spp., predominantly P. falciparum, and 41% of these infections were submicroscopic. Infection was predicted by older age, low altitude, previous treatment, and absence of electricity or bicycle. Despite its ostensibly asymptomatic presentation, infection was associated with anaemia, underweight, and a number of non-specific symptoms in the children´s histories.

Infection prevalence varied greatly between locations, with a tendency of decline closer to the city of Butare (located in Ngoma sector, Fig. 1). Such variation is considered to reflect multiple and geographically diverse factors including altitude, rainfall, topography, land use, SES and health-seeking behaviour [28]. Among these, altitude was the strongest predictor of infection in the present study. Declining infection prevalence with increasing altitude is a common pattern also in neighbouring countries [8, 28‐31]. For example, in Tanzanian children, infection prevalence declined by 5–21% for every 100-m increase in altitude [28, 30]. The extent of this effect even on a small range between 1560 and 1850 m in the present study is remarkable. The higher prevalence of Plasmodium infection at lower altitude, i.e., usually at the valley bottoms, is most likely explained by the closer proximity to vector breeding sites in stagnant waters and commonly practiced rice cultivation. In line with that, valley floors in neighbouring Burundi are considered sources of infection at higher altitudes [31]. Of note, the prevalence of Plasmodium infection did not show the usual decline with age, i.e., with cumulative exposure to malaria parasites. This suggests that the intensity of Plasmodium transmission in the study area is insufficient to build up immune responses during childhood, which prevent infection. Likewise, in neighbouring Tanzania, the age-dependent decline of infection prevalence has been found to vanish at an altitude similar to that of the present study area [28, 30]. In a previous survey in the study area [18], infection prevalence in young community children increased up to the age of 4–5 years and up to a level of 20–25%, i.e., a prevalence that was seen in most of the children in the present study. Children aged 10 years showed a significantly increased infection prevalence. This may reflect their comparatively lower use of bed nets as well as other behavioural characteristics of older children such as increased exposure due to longer staying outside in the evenings or traveling to areas of higher transmission. Similar to other studies from Rwanda [12, 18, 32], various indicators of low SES associated with Plasmodium infection but in multivariate analysis, only the protective parameters household availability of electricity and of a bicycle (each present in 16%) remained. While electricity reflects both, increased SES and proximity to rather urban settings [23], the latter may indicate increased access to malaria-related information, improved awareness and increased usage of curative services. Anti-malarial treatment, mostly with artemether-lumefantrine, within the preceding month was paradoxically associated with increased infection prevalence. This observation was also made among children under 5 years of age in the study area [18] and may stem from recurring parasitaemia following treatment. Drug resistance markers associated with reappearing parasitaemia following artemether-lumefantrine treatment are established in the study area [33, 34], but reports on actual treatment failure are absent. No information on the completeness of treatment was collected, i.e., on adherence to the 3-days regimen, but incomplete drug intake, e.g., in the community or after the first dose in a health facility, may have contributed to reappearing parasitaemia. Moreover, persisting gametocytes could have produced positive PCR results. Alternatively, this finding may reflect the aggregation of infections including rapid re-infections in a specifically vulnerable subset of the population [35].

Recent studies on malaria parasitaemia among school-age children have shown a prevalence range of <1 to 64% in East Africa and of 1 to 83% in West Africa [5]. Symptomatic malaria contributes to school absenteeism, impaired cognition and reduced school performance [5, 10, 16, 20]. As for asymptomatic parasitaemia, predominating in the present study, findings are less clear-cut: in Yemen [6] and DR Congo [36], no influence on children´s cognition was observed while this was the case in Uganda [10], Mali [16], and Kenya [9]. Overall, asymptomatic parasitaemia is abundant; it accounts for three in four infections detected in community settings [4], and many of those are submicroscopic [2, 3]. While some of these infections become symptomatic, most can remain asymptomatic for several months [37, 38]. Counterintuitively, the proportion of submicroscopic infections on a global scale is inversely correlated to transmission intensity, i.e., the proportion is highest at low endemicity [2, 3, 11, 21]. Asymptomatic carriers to a sizeable extent contribute to malaria transmission, also in low endemicity settings [2, 3, 7], even though the magnitude of this impact is not well established [39]. In the present study, more than 90% of Plasmodium infected children attending school were considered asymptomatic based on the absence of fever. Nevertheless, anaemia prevalence was doubled in infected children and mean Hb reduced by 1 g/dL. This degree of Hb reduction is approximately half of that observed among younger and also largely asymptomatic children below 5 years in the study area [18]. In addition, infected children more frequently had underweight and clinically assessed malnutrition than their non-infected peers. These findings argue against asymptomatic presentation but rather for a less obvious one. This is even more the case when considering the more frequently positive histories in infected as compared to non-infected children of fever, tiredness, weakness, poor appetite, abdominal pain, and vomiting. The influence of submicroscopic infections on the latter was similar to the one of microscopically visible parasitaemia but the effect on Hb levels was smaller and no association with underweight was observed. These findings add to the notion that “asymptomatic” Plasmodium infections do have subtle health consequences and should be regarded as potentially harmful [15]. In fact, symptomless children infected with malaria parasites have been shown to have low-grade inflammation as well as evidence of platelet and endothelial activation, in addition to reduced Hb levels [15].

The present study has limitations. The definition of asymptomatic was based on a single-time point assessment; potentially, some children were in a pre-symptomatic period of parasitaemia at the moment of examination. Likewise, PCR detection of malaria parasites may include gametocytes, e.g., persisting after treatment. Submicroscopic parasitaemia may reflect both, very early and thus asymptomatic infections, which later on could develop to clinical malaria, and low level infections persisting in individuals able to control parasitaemia. Of note, studies in murine models have demonstrated that plasmodial DNA is rapidly removed from circulation after treatment or inoculation of killed parasites [40]. Due to its design as a deworming study, questionnaire-based interviews focused on mainly general and gastrointestinal signs and symptoms whereas more intuitive malaria-related conditions, e.g., shivering or sweats, were not included. Although unfortunate, pronounced symptoms in this regard would likely have been detected upon clinical examination. Moreover, this circumstance enabled detecting significant associations of Plasmodium infection with less obvious “soft” symptoms. Plasmodium and Ascaris infections were not correlated and adjusting for the presence of the latter did not substantially affect the odds of symptoms due to the former. Infections with different Plasmodium species were combined since small numbers of non-falciparum malaria infections rendered separate analysis unsubstantial. In clinical terms, P. falciparum and P. ovale did not differ much, with the exception of a larger contribution to anaemia of the former.

The health and educational consequences of malaria and asymptomatic parasitaemia in school children as well as the role of this large group as a silent reservoir of infection are increasingly recognized [5, 9‐20]. School children are not specifically targeted in malaria control programmes so far, and due to the predominance of seemingly asymptomatic parasitaemia, these infections usually remain untreated. The use of insecticide-treated nets (ITNs) in school children is notoriously low [5, 41] but schools are ideally suited to provide education on malaria, prevention and ITNs. This may include information on the importance of adhering to the 3-days regimen of artemisinin-based combination treatment. Schools also can serve as hubs for ITN distribution, and boarding schools in endemic areas should equip their dormitories with ITNs. Chemoprophylaxis or intermittent preventive treatment (IPT) have shown to reduce malaria-associated morbidity in school children [42, 43]. IPT, also termed intermittent parasite clearance in schools (IPCs) or seasonal malaria chemoprevention (SMC), needs to account for the epidemiological characteristics of malaria in a given area, and can greatly reduce the proportion of Plasmodium infected children (reviewed in Ref. [5]), and also improve sustained attention [9]. School-based intermittent screening and treatment of malaria is another option but results have been disappointing so far [44]. It remains to be elucidated, which drug-based regimen fits best and is most cost-effective for the different transmission settings across SSA. Integration with other school-based health promotion activities, e.g., routine deworming, is also desirable.

Morbidity due to Plasmodium infection even if asymptomatic at first sight, associated impaired educational performance [5, 10, 16, 20] and contribution to transmission [2, 3, 7] argue for the inclusion of school children into improved surveillance and control activities.