The immunoglobulin G antibody response to malaria merozoite antigens in asymptomatic children co-infected with malaria and intestinal parasites

Background Co-infection with malaria and intestinal parasites is common in children in Africa and may affect their immune response to a malaria parasite infection. Prior studies suggest that co-infections may lead to increased susceptibility to malaria infection and disease severity; however, other studies have shown the reverse. Knowledge on how co-morbidities specifically affect the immune response to malaria antigens is limited. Therefore, this study sought to determine the prevalence of co-infection of malaria and intestinal parasites and its association with antibody levels to malaria merozoite antigens. Methods A cross sectional study was carried out in two villages with high transmission of malaria in Cameroon (Ngali II and Mfou) where mass drug administration (MDA) had been administered at ~6-month intervals (generally with albendazole or mebendazole). Children aged 1–15 years were enrolled after obtaining parental consent. A malaria rapid diagnostic test was used on site. Four (4) ml of peripheral blood was collected from each participant to determine Plasmodium falciparum infections by microscopy, haemoglobin levels and serology. Fresh stool samples were collected and examined by wet mount, Kato-Katz method and modified Ritchie concentration techniques. A Multiplex Analyte Platform assay was used to measure antibody levels. Results A total of 320 children were enrolled. The prevalence of malaria by blood smear was 76.3% (244/320) and prevalence of malaria and intestinal parasites was 16.9% (54/320). Malaria prevalence was highest in young children; whereas, intestinal parasites (IP+) were not present until after 3 years of age. All children positive for malaria had antibodies to MSP142, MSP2, MSP3 and EBA175. No difference in antibody levels in children with malaria-co infections compared to malaria alone were found, except for antibody levels to EBA-175 were higher in children co-infected with intestinal protozoa (p = 0.018), especially those with Entamoeba histolytica infections (p = 0.0026). Conclusion Antibody levels to EBA175 were significantly higher in children co-infected with malaria and E. histolytica compared to children infected with malaria alone. It is important to further investigate why and how the presence of these protozoans might modulate the immune response to malaria antigens.

low prevalence in the study area. A study conducted in this area (and other regions of Cameroon) by Louis-Albert Tchuem Tchuenté et al., (2012) reported a prevalence of S. haematobium of only 1.72%. Since a large sample size would be required to assess the impact of this pathogen on the Ab response to malaria, S. haematobium was not included in the study.
• Were the individuals asymptomatic to intestinal parasites infection too? No diarrhea, abdominal pain, etc.? Please clarify. Reply: Yes. To make the point clear, the Methods section has been revised and states that all children with clinical cases of malaria or intestinal parasites were not included in the study and referred to the local clinic/hospital by the attending physician for treatment. Thank you for the comment.
• (Page 6) It was mentioned that Plasmodium parasitemia was quantified. Did the authors observe any correlation between the Plasmodium parasite burden and the levels of IgG responses to the antigens? Reply: As expected, there was no correlation between parasitemia and malaria antibody levels.
• (End of Page 7) Please specify: If the cut-off is MFI+3*SD, how the standard deviation was calculated if the negative controls were pooled? Was this experiment repeated or used replicates? Traditionally, the negative controls are tested simultaneously in different wells of the plate, and the cut-off is calculated from those values. Reply: Pooled negative control plasma sample were run in triplicates on the same plates as the test samples in all experiments, as well as the positive controls. The cutoff was obtained by calculating MFI+3 SD of the triplicates on all plates in the experiment.
• Did the authors analyze the effect of helminth parasite burden (number of eggs/gram of stool) in those individuals with helminths? This valuable information was commented on but never included in the analysis. If not used,e I do not see the necessity of describing in the methods section Reply: The information has been deleted from the Methods section.
• For data analysis: • Before using ANOVA, did the authors checked for the normality of the variables? If yes, please specify, if not, calculate the normality of the variables and the other ANOVA assumptions. Reply: Yes, ANOVA was used to compare difference in age across the 4 groups (Table 2). However, comparisons of Ab MFI, which are not normally distributed, with age ( Fig. 1) were performed using the Kruskal-Wallis test. The Methods section (Data analysis) has been revised. Information in Fig. 1 legend was correct.
• If the authors have not-normal variables, they should use the Kruskal-Wallis nonparametric, and Dunn posthoc tests to verify differences between groups. Reply: Sorry for the mistake in the Methods section. The Kruskal-Wallis nonparametric test was performed in Fig 1 and 2. A posthoc test was not performed, as the goal was not to determine when peak Ab levels were obtained, but to determine if age had an influence on Ab levels. Since age was a variable, data for all age groups could not be combined, but rather age was taken into consideration during data analysis.
• Please check frequencies described in table 1 (MAL+IP-58.8%) vs. the values reported in the second line page 9. (59.4%). Reply: 59.4% is the correct value. The text has been revised.
• Sum of 58.8%+16.9% = 75.7% not 75.6%. Reply: Thank you for catching the error. The values in Table 1 and text have been revised and are now consistent.
• In table 1, please add a column with P-values to facilitate the interpretation of the differences between groups. Please report statistics of multiple comparisons between groups too. Reply: The comparisons requested by the reviewer were originally provided in the Table legend. To comply with the request, the p values have been moved to a column labeled "p values" and the method of analysis was retained in the Table legend. • What is the potential hypothesis to explain the increased values of parasitemia in the coinfected group? Reply: There is no significant difference in parasitemia between the two groups (p=0.1599). In fact, the higher parasitemia was found in young children who were intestinal parasite-negative (probably because very young children were in this group).
• Please comment in the text the presence of multi-parasitism in the studied individuals. Reply: We thank the reviewer for the comment. The following sentence has been added to the Results section. "Interestingly, all of the children had single parasite infections, and polyparasitism was not found." • (Page 11 table 3). Please include values of anemia and eosinophilia in individuals coinfected. In the current configuration is constructed is hard to determine the coinfection impact in anemia and eosinophilia values. Reply: Table 3 was designed to evaluate the influence of age on malaria, IP, anemia and eosinophilia. The number of co-infections are too small to be divided by age. In an attempt to address the Reviewer's comment, a separate Table was designed that compares the influence of no infections, malaria-positive only, and co-infections on percent with anemia and eosinophilia. The Table will be up-loaded as supplemental Table 1. It essentially showed that same results as expected, anemia was associated with malaria and eosinophils were associated with co-infections.
• (Page 11). In the sentence, "Thus, as children living in these villages increased with age, they developed partial immunity to malaria and anemia declined; whereas, the prevalence of IP and eosinophilia increased." In this sentence, it is necessary to specify that "protection" is protection against malaria symptoms. The table clearly shows that the frequency of malaria does not decrease with age, only the anemia. Reply: The sentence has been revised to read: "Thus, as children living in these villages increased with age, they began developing partial immunity to malaria symptoms and anemia declined; whereas, the prevalence of IP and eosinophilia increased.
• Please plot Age vs. Antibody levels for each protein to verify the correlation for each protein studied. Reply: The figure on the right confirms that Ab levels increase with age. The figure shows a linear regression analysis of Ab levels for MSP1, MSP2, MSP3 and EBA-175 using data from all 320 children, and includes the equation for the regression line, the R2 value (all positive), and p value (all significant). Thus, the figure confirms that Ab levels increase with age. We do NOT wish to include this figure in the MS since it is essentially identical to the one shown in Fig 1 B,C,D and E. In fact, we feel that the information in Fig 1B-E is easier for the reader to understand. Note: If the figure is not shown, it is provided in a separate document.
• As an exploratory analysis, I suggest joining all data and make a boxplot comparing MFI between MAl-PI-, MAL-PI+, MAL+PI-, and MAL+PI+. Mainly for MSP1, MPS2, and MSP3 group age 3-10 and 11-15 to check. Reply: We thank the Reviewer Thanks for the suggestion concerning exploratory analysis. A comparison of Ab levels in two of the above groups (MAL-,IP-, and MAL+,IP-) is shown in Fig 2. Unfortunately, the number of children in the MAL-,PI+ group is too small to provide valuable information. As stated above, children in the MAL-,PI+ group (n=54) are infected with a variety of intestinal helminths, cestodes and protozoa (see Table 2). With such a diverse range of pathogens, plotting the data as a boxplot will not provide useful information. In Fig. 2, the distribution of Ab levels in children co-infected with malaria and single intestinal pathogens is provided. We feel this approach is more informative than "dumping all pathogens together." • The sentence "E. histolytica is a gut amoeba that causes both intestinal and extraintestinal infections such as amebic colitis (dysentery) and liver or brain abscess. The protozoa cause a marked down-regulation of macrophage functions rendering the cells incapable of antigen presentation and unresponsive to cytokine stimulation (57)" does not explain the increase of antibody production in E. histolytica infected group. Why could a diminishing antigen presentation generate higher levels of anti-Plasmodium antigens? Reply: Very true! Not sure why that statement wasn't caught. The Discussion has been changed significantly. It now reads, "The decrease in macrophage function does not explain the increase in Ab to EBA-175. One possible explanation is that since malaria and E. histolytica…" Other observations/questions: • In the title, add "IgG" to Antibody response. Reply: IgG has been added to title (although not all of the co-authors agree this is necessary).
• Check all scientific names of parasite species for correct formatting in italics. (Example Entamoeba histolytica in the Results section in the abstract) Reply: The scientific name has been checked and are now in italics.
• Please, mention in the background the region where the study was performed. Reply: This information was included in the background section of the Abstract. It is also included in the Materials section.
• It is necessary to describe and discuss the role of MSP1, MPS2, MSP3, and EBA-175 as markers in serological studies. Reply: This information has been added to the Discussion.
• Considering that coinfection prevalence is relatively low, I consider that it is important to discriminate with colors or point shapes the individuals MAL-IP-, MAL+IP-, MAL-IP+, MAL+IP+ in Figure  • In page 6 subtitle "Laboratory detection, quantification and speciation of malaria parasites.", I will not use speciation here. I suggest "Diagnosis and quantification of Plasmodium sp. parasites. Reply: The header has been changed to read: "Laboratory detection of malaria parasites." • (Page 14-15) What type of parasite is "Amoeba"? What is the difference between "Amoeba" and E. histolytica? Traditionally, E. histolytica is considered an amoeba too. Reply: The figure has been revised to read Intestinal Protozoa. Thanks for pointing out the mis-classification.
• In table 1, to facilitate reading, please remove symbols % and /ul located in cells with data and add to the columns describing the variables. Reply: The symbols in the data cells have been removed.
• For consistency, unify parasitemia vs. parasitemia, anemia vs. anemia in the text and plots. Reply: The British spelling of parasitaemia, anaemia, and haemoglobin have been used through out the MS.
whereas, Ab are plasma proteins that bind specifically with an antigen. What was measured was IgG Ab. Since the serological assay measured IgG Ab that were recorded as MFI (median fluorescence intensity), we think the labels on the Y-Axis (Ab levels -MFI) reflect what was done. The Methods section makes it clear that the Ab were of the IgG class. [Note: Serum IgG levels (which implies mg/ml) were not measured.] • ( Figure 1E) Add, Change from EBA to EBA-175. Reply: Change has been made.
• Please verify all references formatting (For example, reference 42 is all in capital letters) Reply: References have been edited as requested by the reviewer.
Review #3: Comments were in the attachment. Reply: In revising the MS, all requested changes were made and additional information provided in the text, including information on the BLAST search. The only request we would not fully address is the prevalence of bednet use in the villages. The only information available is that very few children use bednets. Since the slide-positivity rate of 75.6% for P. falciparum, it is unlikely the bednets are having a major influence on the current study. The following information has been added to the MS in the Results section. "To determine if higher Ab levels in children co-infected with P. falciparum and E. histolytica might be due to cross-reactive epitopes, a BLAST search for sequence homology between EBA-175 and E. histolytica proteins was made. No similarities were found using Metablast, and only one hit was found using discontinuous metablast which had a span of only 38 nucleotides (~12 amino acids). Thus, there does not appear to be shared epitopes between these two pathogens that would explain the increase in Ab to EBA-175 in children with co-infections." Figure for Reviewer #2 confirming an increase in antibody levels with age.  (1). In 52 malaria endemic areas, individuals exposed to malaria infections gradually develop clinical 53 immunity (2) and commonly experience asymptomatic infections without fever or symptoms 54 and do not require antimalarial treatment. Asymptomatic infection results from partial 55 immunity that controls, but does not completely eliminate, malaria parasites, thus allowing 56 for constant presence of circulating parasites (2). 57 The prevalence of intestinal parasitic infections in children is fairly constant across sub-58 Saharan Africa with an average prevalence of 26% (3,4). In Cameroon, the prevalence in 59 children less than 18 years is 26.8% (5), while that for the general population is more than 60 28% The major intestinal parasites are Ascaris lumbricoides, Trichuria trichuria and 61 Entamoeba histolytica (6)(7)(8), but many cases of intestinal parasites go undetected. 62 Co-infections with malaria and intestinal parasites (IP) are common in malaria endemic 63 areas in sub-Saharan Africa (7,8) and infections with IP and Pf are both ranked among the 64 major cause of mortality and morbidity in sub-Saharan Africa. Several studies conducted on 65 IP (not including amoebas) and Pf have shown conflicting results. Some helminths suppress 66 different T-helper types and favor an increase in regulatory T (Treg) cell (9). Studies on 67 concomitant infections in humans suggest that A. lumbricoides infection may protect against 68 cerebral malaria (10,11), while other studies suggest that children infected by Schistosoma 69 mansoni may be more susceptible to P. falciparum infections and develop acute malaria 70 episodes (12,13). Also, it has been shown that the levels of TNF-α, IL-2, IL-10, IL-6 in 71 Plasmodium-helminth co-infected individuals were significantly higher than the malaria-72 positive (MP) group (14) dampening the immune response to malaria. However, little is 73 known regarding host immune responses to malaria in children co-infected with protozoan 74

pathogens. 75
Studies suggest that children co-infected with malaria and intestinal helminths had 76 significantly decreased antibody levels to the malarial antigen apical merozoite antigen 1 77 (AMA-1) compared to those with P. falciparum or IP alone(15). Hence, infections with 78 intestinal helminths can stifle protective anti-plasmodial antibody responses (15). However, 79 increase in MSP3 IgG1-4 levels were significantly associated with children infected with 80 malaria alone compared to children co-infected with both parasites(15).

114
A cross sectional study was carried out in Ngali II and Mfou from January to May 2017, a 115 transitional period from the dry to wet season. Children who had lived in either of the villages 116 for at least six months and whose parents gave informed consent were included in the study. 117 All participants were systematically examined by a physician for clinical systems of malaria 118 and IP. Children who presented with symptoms of malaria, e.g., fever, headaches or 119 intestinal illnesses, e.g., diarrhea, vomiting were not enrolled. A total of 320 participants (140 120 from Ngali II and 180 from Mfou) aged 1-15 years participated in the study. Since both 121 villages have the same demographic features, data for the two villages were combined. 122 Blood collection and on-site testing for malaria 123 Venous peripheral blood (about 4mL) was collected by venipuncture using a butterfly needle 124 (G22) and a 5mL labeled EDTA tube from all 320 participants. Haemoglobin (Hb) was 125 measured using the HemoCue (AB Leo Diagnostics, Helsingborg, Sweden). On site, after 126 collecting the venous blood from the participants, a drop from the same collected blood was 127 placed on a CareStart TM Malaria pLDH/HRP-2 Combo Test (Access Bio Inc. USA) to detect 128 histidine-rich protein-2 (HRP-2) specific to Plasmodium falciparum and Plasmodium lactate 129 dehydrogenase (pLDH) pan-specific to Plasmodium spp (falciparum, P. vivax, P. malariae, 130 P. ovale). Results were read according to manufacturer instructions and recorded after 5 131

minutes. 132
Laboratory detection of malaria parasites 133 Ten microliters of whole blood were used to prepare thick and thin smears for malaria 134 parasite identification, speciation and quantification. The slides were air-dried overnight, and 135 the thin blood smears were fixed in absolute (100%) methanol. Both thick and thin smears 136 were stained using 10% Giemsa solution, washed with water and air-dried. Slides were then 137 microscopically examined (thin and thick smear) for the presence of malaria parasites by two experienced microscopists. The parasite density was determined by counting the number of 139 parasites against 200 leucocytes. The counts were expressed as the number of P. 140 falciparum-infected erythrocytes (IE) per microliter of blood (Pf IE/µl), assuming an average 141 leukocyte count of 8,000 cells/µl of blood (22). When the difference in parasitaemia between 142 the two readers was greater than 5 Pf IE/µl of blood, a third reader re-examined the slide 143 and the mean of the two closest values were considered. Also, a differential count for 144 eosinophil, lymphocytes, monocytes, neutrophils was obtained alongside parasitaemia and 145 different malaria species (by microscopy) 146

147
Plasma samples were tested for antibodies against the merozoite antigens MSP-142, MSP-2, 148 MSP-3 and EBA-175 using a multi-analyte platform assay with antigen-coupled magnetic 149 beads with different spectral addresses. Details of this assay used has been described 150 previously (23) (24). In brief, plasma samples were diluted 1:100 with PBS, 50µl of plasma 151 was incubated with 50µl antigen-coupled microspheres (2000 microspheres/test) for 60 152 minutes in the dark, washed with PBS, and then incubated at 500rpm for 60minutes at 25 °C 153 on a rotating shaker and using a magnet plate separator. Then, 100 µl of secondary Ab (R-154 phycoerythrin-conjugated, Affini Pure F(ab′)2 fragment, Goat anti-human IgG Fc fragment 155 specific, Jackson Immuno-research, West Grove, PA, USA, Cat no. 109-116-170) diluted to 156 2 µg/ml in PBS-1 % BSA was added to each well and incubated as above in the dark for 1 h. 157 The mixture is then washed and a minimum of 100 beads were read in a MAGPIX® reader. because of the small sample sizes of the groups. The one-way-ANOVA test was used to 183 compare all 4 groups after checking for normality (e.g., age). An unpaired t test was used to 184 compare the means of the MAL-,IP-vs. MAL+,IP-groups. Kruskal-Wallis test was used to 185 compare antibody levels, which are not normally distributed, among the groups or within the 186 MAL+IP+ groups. An individual was considered to have a co-infection if at least one IP 187 species and P. falciparum were present. Anaemia was considered when Hb values were < 188 11.5 g/dL and classified according to WHO (27,28

195
The study population 196 A total of 320 children were enrolled (Table 1). Among the children, 76.3% were slide-197 positive for malaria (MAL+), with 59.4% having malaria without intestinal parasites (MAL+,IP) 198 and 16.9% being coinfected with malaria and intestinal parasites (MAL+, IP+). All subjects 199 who tested positive for malaria using the rapid diagnostic field test were confirmed positive 200 by microscopy. Among children who were infected with malaria, 71.3% were infected with 201 only P. falciparum and 5% had P. falciparum and P. malariae. Interestingly, only 2.2% of the 202 children had IP without malaria and 21.6% were negative for both malaria and IP. 203 The mean age of the children changed with infection status among the 4 groups (p = 204 0.0001) with the lowest age found in uninfected children (6.4

years) and highest in children 205
with co-infections (9.3 years) ( Table 1). Malaria infections were found in all age groups; 206 whereas, none of the children under age 4 years had intestinal parasites. Mean 207 haemoglobin levels were lower in children infected with malaria, but the difference was of 208 marginal significance (p = 0.08; MAL-,IP-vs MAL+,IP-). The prevalence of anaemia was 209 higher in children who were infected with malaria (MAL+,IP-)(p=0.032), but not those with 210 co-infections (p >0.999) compared to children who were parasite-negative (MAL-,IP-). 211 Table 1 were also infected with malaria and 2.2% were IP+ but MAL-(  As expected, children aged 1 through 2 years did not have soil-transmitted IP and had 226 normal eosinophil levels; whereas, 63% of 1-2-year old children were infected with malaria 227 and had the highest prevalence of anaemia (Table 3). In contrast, in children 9-15 years of 228 age ~80% were slide-positive for malarial parasites, 24%-29% had intestinal parasites, and 229 10-38% had moderate eosinophilia. Thus, as children living in these villages increased with 230 age, they began developing partial immunity to malaria symptoms and anaemia declined; 231 whereas, the prevalence of IP and eosinophilia increased. 232 *Aneamia: Children with haemoglobin less than 11.5 g/dL. **Moderate eosinophilia: ≥1,500 eosinophils/mm 3 or ≥18.7% peripheral eosinophils A comparison of anaemia and eosinophilia among the 4 groups of children shown in Table 1  233 was made (S1 Table). Results showed that anaemia was associated with malaria infections 234 and eosinophilia was associated with IP. 235 236
Although different amounts of Ab were ultimately obtained to the different antigens, the 243 overall trend was for an increase in median Ab with age. Thus, it was important to take age 244 into consideration when making comparison between children infected with malaria 245 (MAL+,IP-), co-infected with malaria and IP (MAL+,IP+) and those who were not infected 246 (MAL-,IP-) at the time the study was conducted.  Table 3. 251 Since children below 3 years of age were not infected with IP, they were not included in the 259 comparative studies described below. Given that Ab prevalence and levels increased with 260 age, the study population was divided into 2 groups: children aged 3-10 years, a time period 261 when children were becoming infected with IP (Table 3) and those 11-15 years, mainly 262 children who had been infected repeatedly with malaria and may had lived with IP for a 263 period of time. As predicted, Ab levels were slightly higher in MAL+ children due to current 264 boosting compared to MAL-, but the differences were not significant (all p values >0.05) (Fig  265   2). 266 A comparison between Ab levels in children infected with malaria and co-infected 267 with IP was conducted. Children with helminths and cestodes were not included in the analysis because the sample sizes were too small. Ab levels were compared between 269 children aged 3-10 years infected with malaria (n=112) and co-infected with flagellate and 270 intestinal amoeba (n= 25 children), including G. lamblia (n= 15) and E. histolytica (n = 10 271 children) (Fig 2). Antibody levels did not differ between malaria-infected children with or 272 without intestinal amoeba for MSP1, MSP2 and MSP3; however, there were significantly 273 higher Ab levels to EBA-175 in children co-infected with malaria and intestinal amoeba (p = 274 0.018) ( Fig 2D). The higher Ab levels were due to E. histolytica infections (p=0.0026), and 275 not G. lamblia (p=0.3844). No differences were found between children aged 11 to 15 years 276 for any of the antigens between children with malaria (single infection) and co-infected with 277 any of the IP. 278 To determine if higher Ab levels in children co-infected with P. falciparum and E. 279 histolytica might be due to cross-reactive epitopes, a BLAST search for sequence homology 280 between EBA-175 and E. histolytica proteins. No similarities were found using Metablast, 281 and only one hit was found using discontinuous metablast which had a span of only 38 282 nucleotides (~13 amino acids) that had 82% similarity. Thus, there does not appear to be 283 shared epitopes between these two pathogens that would explain the increase in Ab to EBA-284 175 in children with co-infections. intestinal parasites, with 16.9% co-infections (Table 1 -2). This prevalence of malaria is 299 similar to those found in other highly endemic regions of the country (31), and the 300 prevalence of co-infections was 19.1%, which is similar to a prevalence of 18 -27% 301 reported in other regions of Cameroon (32,33). This high transmission is related to geo-302 ecological and climatic conditions at the time of the study which was the transition from the 303 dry to wet season, a period that favors vector breeding and distribution (34). 304 From Table 3, the prevalence of slide-positive malaria ranged from 61% to 90% in different 305 age groups implying that children in these villages were repeatedly exposed to malaria 306 throughout their lives. The current prevalence of malaria in 2017 is similar to that recorded 307 previously for Ngali II between 1998-2004, that ranged from 50% to 85% in 5-15 year old 308 children with an estimated entomological inoculation rate of 0.7 infectious bites/per/night 309 (~257 infectious bites annually)(35). Prior studies have established that repeated exposure 310 induces immunity to malaria, with development of anti-disease immunity followed by anti-311 parasite immunity (36-39). As a result, the highest prevalence of 56% anaemia was found 312 in young children (2,40,41) with a decline to 27% in 13 to 15-year-olds (  (44,45). The low prevalence of helminths is explained, in part by, the 320 fact that mass community de-worming is done biannually following the national infectious 321 disease guide-line for IP control program. The most prevalent intestinal parasites were the 322 protozoans, G. lamblia and E. histolytica (48). These protozoa are commonly found in damp 323 soil and contaminated water with a prevalence of 2-20% in Cameroon (50-53). These 324 results suggest children acquire their intestinal infections after learning to walk and interact 325 with the environment. Thus, children in the study population were exposed to malaria early 326 in life and began developing anti-malaria immunity prior to exposure to intestinal parasites. 327 Generally, both Ab prevalence and Ab levels increased with age in 1 to 15-year-olds living in 328 this high transmission area (Fig 1). Often, the presence of Ab is used as markers of 329 infection, including the merozoite antigens used in this study. This study compared antibody continued to mature as children developed into adolescents (Fig 1). Often individuals who 334 are MAL+ have higher Ab levels than MAL-individuals due to boosting of the Ab response 335 (36,38,39). In the current study, Ab levels did not differ significantly between MAL+ and 336 MAL-individuals, neither those who were 3-10 years nor 11-15 years-old. This result was 337 not surprising, since 75% of the children were slide-positive for malaria (Table 1) Prior studies have demonstrated that malaria-helminths co-infections can down regulate 344 malaria and orient the immune response via the Th2 response hence, making patients less 345 sick (20,36,57,58) whereas, others have demonstrated on the contrary that IP and malaria 346 co-infections increase malaria disease (13,56). Unfortunately, the current study could not 347 resolve the controversy because very few children had helminthic infections, due to frequent 348 treatment with albendazole via the mass drug administration program conducted by the 349 Ministry of Health and other random health campaigns. However, co-infections with malaria 350 and amoeba were relatively common. Ab levels to MSP1, MSP2 and MSP3 were similar in 351 children infected with P. falciparum alone and those with amoeba ( Fig 2); however, 352 significantly higher Ab levels to EBA-175 were found in children co-infected with malaria and 353 intestinal amoeba (p = 0.018). The higher Ab levels were due to E. histolytica infections (p = The prevalence of malaria was high in children 1-2 years old; whereas, intestinal parasite 376 infections occurred in children over 3 years old. Thus, immunity to P. falciparum began prior 377 to infection with soil-transmitted parasites. No differences were found in antibody prevalence 378 or levels in malaria-infected and co-infected children, except antibody levels to EBA175 were 379 significantly higher in children co-infected with malaria and E. histolytica. This is the first 380 report of an interaction between malaria and E. histolytica and antibodies to EBA-175 and 381 merits further evaluation. We heartily thank all the participants and their parents who let them participate in the study 400 in Ngali II and Mfou. We equally express our gratitude to the community health workers of 401 these areas for their assistance during sample collection.       specifically affect the immune response to malaria antigens is limited. Therefore, this study 7 sought to determine the prevalence of co-infection of malaria and intestinal parasites and its 8 association with antibody levels to malaria merozoite antigens. 9

A cross sectional study was carried out in two villages with high transmission of malaria in 11
Cameroon (Ngali II and Mfou). Children aged 1-15 years were enrolled after obtaining 12 parental consent. A malaria rapid diagnostic test was used on site. Four (4)

19
A total of 320 children were enrolled. The prevalence of malaria by blood smear was 76.3% 20 (244199/230 (75. 6%) and prevalence of malaria and intestinal parasites was co-21 infections16.9% ( 54/320) (16.9%). Malaria prevalence was highest in young children; 22 whereas, intestinal parasites (IP+) were not present until after 3 years of age. All children 23 positive for malaria had antibodies to MSP142, MSP2, MSP3 and EBA175. No difference in 24 antibody levels in children with malaria-co infections compared to malaria alone were found, 25 except for antibody levels to EBA-175 were higher in children co-infected with intestinal 26

Conclusion:
3 Antibody levels to EBA175 were significantly higher in children co-infected with malaria and 4 E. histolytica compared to children infected with malaria alone. It is important to further 5 investigate why and how the presence of these protozoans can modulate the immune 6 response (Th1/Th2) to malaria antigens. (1,2). In malaria endemic areas, individuals exposed to malaria infections gradually develop 4 clinical immunity (2)(3) and commonly experience asymptomatic infections without fever or 5 symptoms and do not require antimalarial treatment. Asymptomatic infection results from 6 partial immunity that controls, but does not completely eliminate, malaria parasites, thus 7 allowing for constant presence of circulating parasites (2)(3). However, with most children 8 getting infected with several episodes of infections in a short period, this renders them more 9 prone to having clinical symptoms since the immune systems doesn't fully recover. 10

undetected. 16
Co-infections with malaria and intestinal parasites (IP) are common in malaria endemic 17 areas in sub-Saharan Africa (7,8) (8,9) and infections with IP and Pf are both ranked among 18 the major cause of mortality and morbidity in sub-Saharan Africa. Several studies conducted 19 on IP (not including amoebas) and Pf have shown conflicting results. Some helminths 20 suppress different T-helper types and favor an increase in regulatory T (Treg) cell (9)(10). 21

Studies on cConcomitant infections in humans have suggested suggest that A.scaris 22
lumbricoides infection may protect against cerebral malaria (10,11)(11,12), while other 23 studies, suggest that children infected by Schistosoma. mansoni may be were more 24 susceptible to P. falciparum infections and develop acute malaria episodes (12,13)(13, 14). 25 Also, it has been shown that the levels of TNF-α, IL-2, IL-10, IL-6 in Plasmodium-helminth 26 co-infected individuals were significantly higher than the malaria-positive (MP) group (14) Studies suggest that children co-infected with malaria and intestinal helminths had 1 significantly decreased antibody levels to the malarial antigen apical merozoite antigen 1 2 (AMA-1) compared to those with P. falciparum or IP alone(15) (16). Hence, infections with 3 intestinal helminths can stifle protective anti-plasmodial antibody responses (15) (16). 4 However, increase in MSP3 IgG1-4 levels were significantly associated with children 5 infected with malaria alone compared to children co-infected with both parasites(15) (16). 6 Malaria and other intestinal parasites overlap extensively in their epidemiological distributions 7 causing polyparasitism. Polyparasitism with intestinal parasites has been reported as one of 8 the contributing factors to hypo-responsiveness (16)(17), dampening of the immune 9 response by inducing a strong Treg response, which could in turn, blunt a strong response to 10 vaccines (17)(18). Equally, some studies have suggested an effect of IP on antibody 11 responses to P. falciparum gametocyte antigens that may have consequences on 12 transmission-blocking immunity (18)(19). 13 Effective elimination and future eradication of malaria will require not only vector control, but 14 also managing asymptomatic malaria patients and developing an effective vaccine. Given  have minimal access to portable water, with approximately one well per 500 inhabitants. 6 Currently, mass drug administration with albendazole is being performed twice a year by the 7 Ministry of Health, that is usually conduced in schools and symptomatic cases are sent to the 8 local clinic or hospital for follow up treatment. 9

Study pPopulation
10 A cross sectional study was carried out in Ngali II and Mfou from January to May 2017, a 11 transitional period from the dry to wet season. Children who had lived in either of these two 12 the villages for at least six months and whose parents gave informed consent were included 13 in the study. Since both villages (Ngali II & Mfou) were very similar in all features, data for 14 both village were combined. Vital parameters (temperature, pulse) and anthropometric 15 parameters (weight, height) were measured by assisting attendant nurses. These 16 parameters were used to calculate body mass index (BMI) and advice was given to the 17 parents of the participating children, as part of a service for participation. . All participants 18 were systematically examined by a physician for clinical systems of malaria and IP. Only 19 asymptomatic participants were included in the study. Children who presented with 20 symptoms of malaria, e.g., fever, headaches or intestinal illnesses, e.g., diarrhea, vomiting 21 were not enrolled. A total of 320 participants (140 from Ngali II and 180 from Mfou) aged  15 years participated in the study. Since both villages have the same demographic features, 23 data for the two villages were combined. 24

25
Venous peripheral blood (about 4mL) was collected by venipuncture using a butterfly needle 26 (G22) and a 5mL labeled EDTA tube from all 320 participants. Haemoglobin (Hb) was 27 measured using the HemoCue (AB Leo Diagnostics, Helsingborg, Sweden). On site, after 28 collecting the venous blood from the participants, a drop from the same collected blood was 1 placed on a CareStart TM Malaria pLDH/HRP-2 Combo Test (Access Bio Inc. USA) to detect 2 histidine-rich protein-2 (HRP-2) specific to Plasmodium falciparum and Plasmodium lactate 3 dehydrogenase (pLDH) pan-specific to Plasmodium spp (falciparum, P. vivax, P. malariae, P. 4 ovale). Results were read according to manufacturer instructions and recorded after 5 5 minutes. 6 Laboratory detection, quantification and speciation of 7 malaria parasites. 8 Ten microliters of whole blood were used to prepare thick and thin smears for malaria 9 parasite identification, speciation and quantification. The slides were air-dried overnight, and 10 the thin blood smears were fixed in absolute (100%) methanol. Both thick and thin smears 11 were stained using 10% Giemsa solution, washed with water and air-dried. Slides were then 12 microscopically examined (thin and thick smear) for the presence of malaria parasites by two 13 experienced microscopists. The parasite density was determined by counting the number of 14 parasites against 200 leucocytes. The counts were expressed as the number of P. 15 falciparum-infected erythrocytes (IE) parasites per micro-liter of blood (Pf IE/µl), assuming an 16 average leukocyte count of 8,000 cells/µl of blood (22)(23). When the difference in 17 parasitaemia between the two readers was greater than 5 Pf IE/µl of blood, a third reader re-18 examined the slide and the mean of the two closest values were considered. Also, a 19 differential count for eosinophil, lymphocytes, monocytes, neutrophils was obtained 20 alongside parasitaemia and different malaria species (by microscopy) 21

Antibody Aanalysis
Plasma samples were tested for antibodies against the merozoite antigens MSP-142, MSP-2, 23 MSP-3 and EBA-175 using a multi-analyte platform (MAP) assay with antigen-coupled 24 magnetic beads with different spectral addresses. Details of this assay used has been 25 described previously (23)(24) (24)(25). In brief, plasma samples were diluted 1:100 with 26 PBS, 50µul of plasma was incubated with 50ulµl antigen-coupled microspheres (2000 27 microspheres/test) for 60 minutes in the dark, washed with PBS, and then incubated at 28   groups, because of the small sample sizes of the groups. The one-way-ANOVA test was 5 used to compare all 4 groups after checking for normality (e.g., age). An unpaired t test was 6 used to compare the means of the MAL-, IP-vs. MAL+, IP-groups. Kruskal-Wallis test was 7 used to compareisons antibodyies levels, which are not normally distributed, among the 8 groups or within the MAL+IP+ groups. An individual was considered to have a co-infection if 9 at least one IP species and P. falciparum were present. Anaemia was considered when Hb 10 values were < 11.5 g/dL and classified according to WHO (27,28)(28,29). To search DNA 11 sequences of P. falciparum EBA-175 and those of E. histolytica for possible cross-reactive 12 epitopes, PfEBA175 (ncbi.nlm.nih.gov/gene/2654998) was compared with E. histolytica 13 (ncbi.nlm.nih.gov/assembly/GCF_000208925.1) using Megablast for highly similar 14 sequences and discontinuous megablast for more dissimilar sequences. 15

17
The Sstudy Ppopulation 18 A total of 320 children were enrolled (Table 1). Among the children, 76.35.6% were slide-19 positive for malaria (MAL+), with 59.48.8% having malaria without intestinal parasites (MAL+, 20 IP-) and 16.9% being coinfected with malaria and intestinal parasites (MAL+, IP+). All 21 subjects who tested positive for malaria using the rapid diagnostic field test were confirmed 22 positive by microscopy. Among children who were infected with malaria, 71.3% were infected 23 with only P. falciparum and 5% had P. falciparum and P. malariae. Interestingly, only 2.2% of 24 the children had IP without malaria and 21.62.2% were negative for both malaria and IP. 25 The mean age of the children changed with infection status among the 4 groups (p = 26 0.0001) with the lowest age found in uninfected children (6.4

years) and highest in children 27
with co-infections (9.3 years) ( Table 1). Malaria infections were found in all age groups; 28 Field Code Changed Formatted: Font: 18 pt whereas, none of the children under age 4 years had intestinal parasites. Mean 1 haemoglobin levels were lower in children infected with malaria, but the difference was of 2 marginal significance (p = 0.08; MAL-,IP-vs MAL+,, IP-). The prevalence of anaemia was 3 higher in children who were infected with malaria (MAL+,IP-)(p=0.0324), but not those with 4 co-infections (p >0.999) compared to children who were parasite-negative (MAL-,IP-). Overall, 19.1% (61/320) of the children were positive for intestinal parasites, 16.9% of whom 1 were also infected with malaria and 2.2% were IP+ but MAL-( Table 2). The most frequent 2 major of helminthic parasites detected were A.scaris lumbrioides (2.8%) and single cases of 3 Trichura sp. and Strongyloides sp. Among the 320 children, 14.7% had detectable protozoan 4 infections, including 7.8% infected with Giardia lamblia, 5.9% with E.ntamoeba histolytica, 5 and 0.9% with Isospora sp. Very few children had intestinal cestodes ( Table 2). 6 Interestingly, all of the children had single parasite infections, and polyparasitism was not 7 found. 8 As expected, children aged 1 through 2 years did not have soil-transmitted IP and had 3 normal eosinophil levels; whereas, 63% of 1-2-year olds children were infected with malaria 4 and had the highest prevalence of anaemia (Table 3). In contrast, in children 9-15 years of 5 age ~80% were slide-positive for malarial parasites, 24%-29% had intestinal parasites, and 6 10-38% had moderate eosinophilia. Thus, as children living in these villages increased with 7 age, they began developinged partial immunity to malaria symptoms and anaemia declined; 8 whereas, the prevalence of IP and eosinophilia increased. 9 *Aneamia: Children with hHaemoglobin less than 11.5 g/dL. **Moderate eosinophilia: ≥1,500 eosinophils/mm 3 or ≥18.7% peripheral eosinophils A comparison of anaemia and eosinophilia among the 4 groups of children shown in Table 1  11 was made (S1 Table). Results showed that anaemia was associated with malaria infections 12 and eosinophilia was associated with IP. 13 With repeated exposure to malaria, Ab prevalence and levels increased with age to the four 1 merozoite antigens (Fig. 1). Among 1-to 2-year-olds, only 25% of the infants had Ab to 2 EBA-175 and MSP3, 30% had Ab to MSP2, but 80% had Ab to MSP1 (Figure 1). However, 3 by age 13-15 years, 60% had acquired Ab to MSP3 and >80% had Ab EBA-175, MSP2 and 4 MSP3 (Fig. 1A). Among Ab-positive childrenparticipants, Ab levels also increased with age 5 ( Fig. 1B-E). Although different amounts of Ab were ultimately obtained to the different 6 antigens, the overall trend was for an increase in median Ab with age. Thus, it was important 7 to take age into consideration when making comparison between children infected with 8 malaria (MAL+,IP-), co-infected with malaria and IP (MAL+,IP+) and those who were not 9 infected (MAL-,IP-) at the time the study was conducted.  Table  4 3. Fig1 B -E shows Ab levels (MFI) for children who were Ab-positive for each age group. Since children below 3 years of age were not infected with IP, they were not included in the 5 comparative studies described below. Given that Ab prevalence and levels increased with 6 age, the study population was divided into 2 groups: children aged 3-10 years, a time period 7 when children were becoming infected with IP (Table 3) and those 11-15 years, mainly 8 children who had been infected repeatedly with malaria and may had lived with IP for a 9 period of time. As predicted, Ab levels were slightly higher in MAL+ children due to current 10 boosting compared to MASL-, but the differences were not significant (all p values >0.05) 11

Antibody Levels to Malaria Merozoite Antigens
( Figure 2). 12 A comparison between Ab levels in children infected with malaria and co-infected with 13 IP was conducted. Children with helminths and cestodes were not included in the analysis 14 because the sample sizes were too small. Ab levels were compared between children aged 15 3-10 years infected with malaria (n=112) and co-infected with flagellate and intestinal 16 amoeba (n= 25 children), including G. lamblia (n= 15) and E. histolytica (n = 10 children) 17 (Fig. 2). Antibody levels did not differ between malaria-infected children with or without 18 intestinal amoeba for MSP1, MSP2 and MSP3; however, there were significantly higher Ab 19 levels to EBA-175 in children co-infected with malaria and intestinal amoeba (p = 0.018) (Fig.  20   2D). The higher Ab levels were due to E. histolytica infections (p=0.0026), and not G. lamblia 21 (p=0.3844). No differences were found between children aged 11 to 15 years for any of the 22 antigens between children with malaria (single infection) and co-infected with any of the IP. 23 To determine if higher Ab levels in children co-infected with P. falciparum and E. 24 histolytica might be due to cross-reactive epitopes, a BLAST search for sequence homology 25 between EBA-175 and E. histolytica proteins. No similarities were found using Metablast, 26 and only one hit was found using discontinuous metablast which had a span of only 38 27 nucleotides (~13 amino acids) that had 82% similarity. Thus, there does not appear to be 28 shared epitopes between these two pathogens that would explain the increase in Ab to EBA-1 175 in children with co-infections. 19.1% had intestinal parasites, with 16.9% co-infections (Table 1 -2). This prevalence of 4 malaria is similar to those found in other highly endemic regions of the country (31)(32), and 5 the . prevalence of co-infections was 19.1%, which is similar to a prevalence of 18 -27% 6 reported in other regions of Cameroon (32,33) (9,44). This high transmission is related to 7 geo-ecological and climatic conditions at the time of the study which was the transition from 8 the dry to wet season, a period that favors vector breeding and distribution (34)(33). 9 From Table 3 above, the prevalence of slide-positive malaria ranged from 61% to 90% in 10 different age groups implying that children in these villages were repeatedly exposed to 11 malaria throughout their lives. The current prevalence of malaria in 2017 is similar to that 12 recorded previously for Ngali II between 1998-2004, that ranged from 50% to 85% in 5-15 13 year olds, with an estimated entomological inoculation rate of 0.7 infectious bites/per/night 14

(~257 infectious bites annually)(35).[LEKE ET AL.] Prior studies have established that 15
repeated exposure induces immunity to malaria, with development of anti-disease immunity 16 followed by anti-parasite immunity (36-39)(34-37). As a result, the highest prevalence of 17 56% anaemia was found in young children (2,40,41) (3,38,39) with a decline to 27% in 13 to 18 15-year-olds (Table 3) suggest children acquire their intestinal infections after learning to walk and interact with the 1 environment. Thus, children in the study population were exposed to malaria early in life and 2 began developing anti-malaria immunity prior to exposure to intestinal parasites." 3 4 Generally, both Ab prevalence and Ab levels increased with age in 1 to 15-year-olds living in 5 this high transmission area (Fig. 1). Often, the presence of Ab is used as markers of 6 infection, including the merozoite antigens used in this study. This study compared antibody Since over 80% of 1-2-year-olds had Ab to MSP1, humoral immunity began to develop early 10 in life and continued to mature as children developed into adolescents (Fig. 1). Often 11 individuals who are MAL+ have higher Ab levels than MASL-individuals due to boosting of 12 the Ab response (36,38,39)(34,36,37). In the current study, Ab levels did not differ 13 significantly between MAL+ andthan MAL-individuals, neither those who were 3-10 years 14 nor 11-15 years-old. This result , however, was not surprising, since 75% of the children 15 were slide-positive for malaria ( Table 1). Because of high transmission, Therefore, it is likely 16 that children are becoming infected almost on a daily basis and either are in the process of 17 eliminating the new infection or reducing it to who were slide-negative had either been 18 recently infected or had submicroscopic levels. infections. Thus, most children living in 19 areas with high perennial transmission will test positive for malaria by PCR. In essence, the 20 immune response in individuals who are repeatedly infection would be similar to that produce 21 during chronic infections.Because of constant re-exposure, the resulting immune response 22 will be similar to that produced by a chronic infection. 23 Prior studies have demonstrated that malaria-helminths co-infections can down regulate 24 malaria and orient the immune response via the Th2 response hence, making patients less 25 sick (20,36,57,58)(21,34,55,56) whereas, others have demonstrated on the contrary that IP 26 and malaria co-infections increase malaria disease (13,56) (14,54). Unfortunately, the current 27 study could not resolve the controversy because very few children had helminthic infections, 28 due to frequent treatment with albendazole via the mass drug administration program 29 infections with malaria and amoeba were relatively common. Ab levels to MSP1, MSP2 and 2 MSP3 were similar in children infected with P. falciparum alone and those with amoeba ( Fig.  3 2); however, significantly higher Ab levels to EBA-175 were found in children co-infected with 4 malaria and intestinal amoeba (p = 0.018). The higher Ab levels were due to E. histolytica 5 infections (p = 0.0026), and not G. lamblia (p = 0.384). This result was unexpected. E. 6 histolytica is a gut amoeba that cause both intestinal and extraintestinal infections such as 7 amebic colitis (dysentery) and liver or brain abscess. Thise protozoa can cause a marked The prevalence of malaria was high in children 1-2 years old; whereas, intestinal parasite 4 infections occurred in children over 3 years old. Thus, immunity to P. falciparum began prior 5 to infection with soil-transmitted parasites. No differences were found in antibody prevalence 6 or levels in malaria-infected and co-infected children, except antibody levels to EBA175 were 7 significantly higher in children co-infected with malaria and E. histolytica. This is the first 8 report of an interaction between malaria and E. histolytica and antibodies to EBA-175 and 9 merits further evaluation. We heartily thank all the participants and their parents who let them participate in the study in 6 Ngali II and Mfou. We equally express our gratitude to the community health workers of 7 these areas for their assistance during sample collection. 8 Anna Babakhanyan, of Hawaii university provided the recombinant proteins (beads) used for 9 the assays and the entire staff of the Biotechnology Centre of University of Yaoundé I, 10 Cameroon for their tremendous work in the field during sample collection and sample 11 processing in the lab. 12 Competing iInterest:

20
Authors declare no competing interests. 21

Data Availability:
The database for the study can be found in the "Supporting Material File." The authors 23 approve of the availability of all data underlying the findings and without restriction upon 24 reasonable request from the corresponding authors. All-important data are within the paper.             (32), and the prevalence of co-infections was 19.1%, which is similar to the prevalence of co-infections of 18 -27% reported in other regions of Cameroon (9,44). The references have been added to the reference section.
• Do the authors have information about malaria and intestinal parasites last treatments? On page 17, it was commented that Albendazole treatment was frequent in these children. Deworming information will help the readers to understand why the prevalence of intestinal parasites was low compared with other studies in Cameroon. Additionally, reinforce in the discussion section that collecting/reporting that information is valuable for coinfection studies. Reply: In response to the Reviewer's suggestion, the following information has been added to the Methods section. "Currently, mass drug administration with albendazole is being performed twice a year by the Ministry of Health, that is usually conduced in schools and symptomatic cases are sent to the local clinic or hospital for follow up treatment." • (Figure 1 B, C, D, E) use the same scale limits for all plots. This is also useful to understand differences in levels of antigenicity between proteins. Reply: We understand the comment, but we do not wish to change the Y-axis on Fig 1, since it is risky to make a direct comparison of Ab levels between antigens in serological assays. A number of variables, including parasite strain, the system to produce recombinant proteins, protein purity, the amount of antigen used, number of exposed epitopes, dilution of plasma, etc., influence the overall results. Even when Luminex beads are covalently-coupled with saturating amounts of antigen, it is questionable if direct comparison of MFI can be made between antigens. Although our assays have been optimized and equivalence amounts of antigen used during bead-coupling, comparisons among the antigens may not provide accurate information about immunogenicity. In Figs1 B, C, D, E, the Y-Axis was selected to show the best distribution of the MFI results.
• (table 3) How could the authors explain increased eosinophilia with low levels of helminth infection? This mainly applies to the age group > 9 years-old. Reply: After age 2, children start becoming infected with helminths, resulting in an increase in eosinophil counts. During the biannual drug treatment campaign, helminthic infections are eliminated, but eosinophilia persists for a period of time. With increasing age, more children in the area become i) infected and ii) re-infected, resulting in an increase in prevalence of eosinophilia.
• Please comment in the text the presence of multi-parasitism in the studied individuals. Reply: We thank the reviewer for the comment. The following sentence has been added to the Results section. "Interestingly, all of the children had single parasite infections, and polyparasitism was not found." • (Page 11 table 3). Please include values of anemia and eosinophilia in individuals coinfected. In the current configuration is constructed is hard to determine the coinfection impact in anemia and eosinophilia values. Reply: Table 3 was designed to evaluate the influence of age on malaria, IP, anemia and eosinophilia. The number of co-infections are too small to be divided by age. In an attempt to address the Reviewer's comment, a separate Table was designed that compares the influence of no infections, malaria-positive only, and co-infections on percent with anemia and eosinophilia. The Table will be up-loaded as supplemental Table 1. It essentially showed that same results as expected, anemia was associated with malaria and eosinophils were associated with co-infections.
• (Page 11). In the sentence, "Thus, as children living in these villages increased with age, they developed partial immunity to malaria and anemia declined; whereas, the prevalence of IP and eosinophilia increased." In this sentence, it is necessary to specify that "protection" is protection against malaria symptoms. The table clearly shows that the frequency of malaria does not decrease with age, only the anemia. Reply: The sentence has been revised to read: "Thus, as children living in these villages increased with age, they began developing partial immunity to malaria symptoms and anemia declined; whereas, the prevalence of IP and eosinophilia increased.
• Please plot Age vs. Antibody levels for each protein to verify the correlation for each protein studied. Reply: The figure on the right confirms that Ab levels increase with age. The figure shows a linear regression analysis of Ab levels for MSP1, MSP2, MSP3 and EBA-175 using data from all 320 children, and includes the equation for the regression line, the R 2 value (all positive), and p value (all significant). Thus, the figure confirms that Ab levels increase with age. We do NOT wish to include this figure in the MS since it is essentially identical to the one shown in Fig 1 B, C, D and E. In fact, we feel that the information in Fig 1B-E is easier for the reader to understand. Note: If the figure is not shown, it is provided in a separate document.
• As an exploratory analysis, I suggest joining all data and make a boxplot comparing MFI between MAl-PI-, MAL-PI+, MAL+PI-, and MAL+PI+. Mainly for MSP1, MPS2, and MSP3 group age 3-10 and 11-15 to check. These data confirm that between the ages of 1 to 15 years, the amount of Ab increases with age, as the results of increasing Ab prevalence and Ab levels. Distribution of MFI for all 320 children by age. Figure show the regression line +/-95% CI. These data confirm that between the ages of 1 to 15 years, the amount of Ab increases with age, as the results of increasing Ab prevalence and Ab levels.