To the best of our knowledge, this is the first study on falciparum malaria among schoolchildren from Hodeidah governorate in relation to haematological and nutritional indices. This study revealed that 8.0% of almost asymptomatic schoolchildren in Bajil district were infected with P. falciparum and over 96.0% of infections having low parasite densities. This prevalence is lower than the rates reported from Hodeidah based on community surveys (16.2%) in 2003  and among febrile patients (15.8%) in 2009 . It is also lower than the rates reported among febrile children seeking healthcare in Sana’a city between 1998 and 2000 (17.3%; 130/753) and children from Taiz governorate (18.6%; 83/447) in 2006 [30, 31]. It is also lower than the household-based malaria prevalence of 18.8% (136/735) reported from the southeastern governorate of Hadhramout , where over 99.0% of infections being caused by P. falciparum. However, it is substantially lower than those reported for schoolchildren from several African countries, including Malawi (60.0%), Kenya (42.0%), Uganda (27.6%), Cameroon (22.8–33.8%) and Tanzania (21.6–38.1%; 93/244) [33–39]. In contrast, it is higher than the rates reported among Kenyan children (4.3%) in a nationwide survey and asymptomatic schoolchildren from northwest Ethiopia (6.8%; 26/385) [40, 41]. The low prevalence of P. vivax (0.25%) among schoolchildren in the present study is in line with the low proportion of vivax malaria in Yemen [1, 32].
In Yemen, the prevalence of falciparum malaria has been reduced following the escalated control interventions of the NMCP since its launch in 2000; however, its burden among asymptomatic schoolchildren is high and can pose a threat to malaria control efforts. The asymptomatic nature of most infections in the present study could be attributed to the low-to-moderate parasitaemia levels and the absence of high-level parasitaemia. Asymptomatic cases usually go undiagnosed and untreated, potentially contributing to parasite transmission and the emergence and spread of drug resistance . Gametocyte carriage by more than one-third of asymptomatic children with uncomplicated falciparum malaria in the present study poses a threat to malaria control and elimination efforts. Such a large reservoir of asymptomatic gametocyte carriers contributes considerably to potential human-to-mosquito transmission . Therefore, identification and treatment of asymptomatic gametocyte carriers should be considered when tailoring malaria elimination strategies. This can be of public health significance in the context of the ongoing conflicts in the country, where the massive internal displacement of asymptomatic carriers from malaria-endemic to malaria-free areas can make the hosting communities prone to malaria epidemics. Apart from its impact on disease transmission, asymptomatic malaria can have consequences on the health and educational performance of children [34, 44, 45]. In Yemen, for example, an earlier study revealed that asymptomatic parasitaemia of P. falciparum can impair the cognitive function of Yemeni children . The findings of the present study underscore the need for complementing the household surveys conducted by the NMCP with school-based malaria surveys to help assess the impact of control interventions. School-age children are more preferred to adults for estimating parasite prevalence and density [46, 47], and school-based malaria surveys are reliable in estimating malaria epidemiology and assessing control interventions [48, 49].
The gender, age and residence of schoolchildren were not significant predictors of falciparum malaria in the present study. The lack of significant association between the age and malaria is consistent with that found among Tanzanian schoolchildren in a nationwide survey . In contrast, the male gender and age younger than 10 years were significantly associated with asymptomatic malaria among Ugandan and Kenyan schoolchildren [34, 38]. However, younger age was a significant predictor of malaria among primary schoolchildren from southern Malawi . Differences in age association with malaria among children could be attributed, among other factors, to differences in transmission intensity [47, 50], which is lower in Yemen compared to sub-Saharan African countries.
Independent of other factors, residence near water collections was significantly associated with a 2.6-fold higher risk of falciparum malaria among schoolchildren in Bajil. Similarly, people living near water collections in Hadhramout were at significantly higher risk of malaria . In the present study, only approximately one-third of schoolchildren reported sleeping under mosquito nets during the malaria transmission season, with no significant association with reduced malaria prevalence. This finding is consistent with those reported for Ugandan and Cameroonian children [34, 51] but inconsistent with that reported among Malawian schoolchildren . The low utilisation of mosquito nets among schoolchildren is in agreement with that (19.0%) reported at the community level in Hodeidah in 2016 . Low mosquito-net utilisation by children despite ownership is common in endemic countries. For instance, the utilisation of mosquito nets was reported among 19.0% and 32.4% of Kenyan and Malawian schoolchildren, respectively [36, 40]. Consequently, efforts should be made to identify the reasons for not utilising mosquito nets by schoolchildren to tailor appropriate educational interventions.
Malnutrition represents a major public health problem in countries endemic for malaria. In Yemen, the ongoing complex emergency and humanitarian crisis besides the absence of school feeding programs aggravate this problem among schoolchildren. The present study revealed malnutrition in more than half of schoolchildren, with stunting and wasting being the most and least prevalent forms of malnutrition, respectively. This shows that chronic malnutrition is more prevalent than acute malnutrition among schoolchildren in the study area. Such predominance of stunting is consistent with those reported from Malawi, Laos and Cameroon [36, 37, 53] but inconsistent with that reported from Mount Cameroon . Of malnutrition forms, underweight was the independent predictor significantly associated with falciparum malaria among schoolchildren in the present study, where underweight children were approximately five times more likely to have falciparum malaria compared to their counterparts. This finding is in agreement with that reported among schoolchildren from high transmission settings of Uganda . In contrast, stunting was the significant predictor of falciparum malaria among children from Kenya, Malawi and Laos [36, 38, 53, 54]. Both stunting and wasting were significant predictors of malaria among children from southwest Cameroon . Although the specific interaction between falciparum malaria and underweight remains not fully understood, acute weight loss could be one of the nutritional consequences of falciparum malaria . Therefore, longitudinal studies are needed to assess the relationship between malaria and malnutrition and the impact of malnutrition on the treatment outcome with artemisinin-based combination therapies (ACTs) among schoolchildren in Hodeidah. It is noteworthy that the risk of treatment failure with ACTs can increase among malnourished children .
Regarding the haematological indices and according to the WHO’s criteria for the classification of malaria , mild anaemia was prevalent among more than half of P. falciparum-infected schoolchildren in the present study. It was an independent predictor of falciparum malaria, where anaemic schoolchildren were at approximately six times more likely to have malaria compared to non-anaemic ones. Such a significant association agrees with that among children from Kenya, Malawi and Laos [21, 36, 53], but disagrees with that among Ugandan schoolchildren . However, the establishment of a causal relationship is rather difficult in such a cross-sectional study, where malaria-associated anaemia is multifactorial. These factors include, among others, mechanical or autoimmune haemolysis, splenic sequestration of infected and non-infected RBCs and suppressed erythropoiesis [56–59].
The mean values of total WBCs and monocytes were significantly lower in infected than non-infected schoolchildren, but within the normal range for both groups. Consequently, it is difficult to establish any clinical implication from such differences in schoolchildren with uncomplicated malaria. Nonetheless, haematological changes are usually common in complicated or severe malaria. Low-to-normal WBC counts are usually observed in malaria patients, mainly due to their localization outside the peripheral circulation rather than actual depletion . In contrast, leucopenia was found to be significantly higher in schoolchildren with falciparum malaria compared to malaria-negative ones in Cameroon . On the other hand, monocytosis could be one of the frequent haematological changes and the most important leukocytic change characterizing malaria .
Thrombocytopenia can occur in P. falciparum-infected patients regardless of the exposure frequency or severity of the disease , which could be due to splenic sequestration, immune-mediated destruction and coagulation disturbances. Given that only one thrombocytopenic child was found in the present study, the association between thrombocytopenia and falciparum malaria was not tested statistically. Moreover, the significantly lower mean platelet count in infected than non-infected schoolchildren has no clinical implications because both were within the normal range. This could be attributed to the low-to-moderate levels of parasitaemia in uncomplicated malaria cases. In another context, thrombocytopenia was significantly associated with falciparum malaria in Hajjah governorate, northwest of Yemen . A significant association was also reported among Kenyan and Nigerian children [21, 63].
Overall, haematological indices are difficult to use in the prediction of malaria in the study district, even with significant differences between infected and non-infected schoolchildren. In this respect, mild anaemia was the only haematological abnormality observed. This could be partially attributed to the low transmission intensity in the study district compared to the studies in African countries. The impact of transmission intensity on the differences in haematological indices could not be ruled out . The utility of haematological indices as indicators of falciparum malaria should, however, be assessed in symptomatic and complicated infections and those with high parasite densities.
The present study is limited by its cross-sectional design that could not establish a causal relationship between malaria and haematological or nutritional abnormalities. Besides, its findings may not be generalizable to school-age children not enrolled in schools, who may represent a large proportion because of the ongoing complex emergency and humanitarian crisis. Nevertheless, this is the first study to provide essential information about falciparum malaria among schoolchildren in relation to haematological and nutritional indices in one of the largest districts of the most malaria-afflicted governorates in the country. Another limitation is the use of light microscopy for diagnosing malaria among schoolchildren, which overlooks submicroscopic infections and may underestimate malaria burden in the study setting. Therefore, there is a need for the molecular-based assessment of submicroscopic reservoir of falciparum malaria, preferably through school-based surveys, in pursuit of the efforts towards malaria elimination. Because the association between severe malaria and haematological and nutritional abnormalities could not be assessed in this school-based study, hospital-based studies are rather needed for this purpose. Longitudinal studies on the development of clinical malaria among schoolchildren with haematological and nutritional abnormalities compared to their normal counterparts are recommended.