Study design and sample collection
The data presented here were generated from the malaria morbidity sentinel surveillance sites within the SMC program in Zinder, Dosso and Gaya districts located in western Niger, where malaria transmission is seasonal[15, 16]. Zinder and Gaya districts have implemented SMC with SPAQ respectively since 2014 and 2016; they were classified as mesoendemic and hyperendemic areas[15, 17]. SMC was not implemented in Dosso district, which was classified as hyperendemic and as a control district of the study[15, 17]. In 2016, SMC coverage in Zinder received once, at least 3 times and 4 times was 91%, 73% and 50%, respectively (Unpublished data). The coverage of Gaya district were 77.72%, 81.56%, 71.26%, and 69.47 respectively for round 1, 2, 3 and 4 (Unpublished data). All the 3 sites used ACTs as first line treatment for uncomplicated malaria and received universal coverage of bed nets. The seasonality of malaria transmission in these 3 sites is the same.
To assess the impact of SMC on the titer of antibodies to two asexual P. falciparum stage antigens, 6 health facilities in Zinder, Dosso and Gaya were selected. In this health facilities, malaria RDTs (SD Bioline) of randomly selected children aged 3-59 months were collected from symptomatic cases (fever + positive or negative RDTs) for serological analysis. For all RDTs collected, the date of consultation, the age, the gender, whether a test was performed and the result (positive or negative) were reported on the cassette. Samples were collected all three month at the same time in all sites between November 2015 to December 2016. The RDTs was stored at room temperature. The analyses were performed in April 2017. The average time of the RDT before testing across the sites was : 06 month for Zinder, 05 month for Dosso and 7 month for Gaya.
The total number of children to include was 249, and calculated based on 78% circumsporozoite protein antibody prevalence in children that received SMC for 1, 2 or 3 years  (95% CI) with a precision of 5%.
The malaria antigens used in this study included a recombinant antigen circumsporozoite protein (CSP) and glutamate-rich protein R2 (GLURP-R2). CSP antigen was a 44-aa NANP repeat-sequence peptide of the P. falciparum circumsporozite protein synthesised by Sygma Genosys, while GLURP-R2 (amino acids 706-1178, F32 strain) was produced at the Statens Serum Institutes of Copenhagen (Denmark) and was expressed in Escherichia coli.
Sera were extracted from filter paper inside RDTs cassette collected . RDTs have proximal, middle and distal parts according Cnops et al., description . The distal part of RDT contains a filter paper component that absorbed the residual blood solution. The cassettes was opened by sterile tweezers and distal part of each RDT was cut with sterile scissors in two or three pieces about 2 mm and eluted (all pieces obtained) into 300 µl of Phosphate Buffered Saline (PBS) from this fragment placed in 1.5ml Eppendorf tubes. The solution was stored at 4°C overnight. The solution is equivalent to 1/100 dilution of whole blood, with about half of the concentration of antibodies in plasma or serum resulting to a dilution of 1/200 . The elations CSP and GLURP-R2 total IgG antibody responses were quantified using ELISA .
The standard operating procedure developed by the African Malaria Network Trust was used to assess total IgG concentrations by Enzyme Linked Immuno Sorbent Assay (ELISA) to CSP and GLURP R2 as described previously . Briefly, recombinant proteins (0.1 µg/well) diluted in Phosphate Buffered Saline (PBS) were coated on MaxiSorp Nunc plates (Thermo Fisher Scientific, Denmark) and blocked with 3% powdered-milk + 0.1% of PBS-Tween 20. Sera samples were diluted at 1:200 for all recombinant proteins. Polyclonal goat anti-human IgG (Gamma) (Caltag) conjugated to HRPO diluted 1:3000 (Skybio, France) was used for revealing the reaction with 3,3',5,5'-tétraméthylbenzidine TMB as substrate and 0.2 M H2SO4 to stop the reaction. Standard curves were established using human IgG purified proteins (Binding Site, France) to determine the concentration of specific antibodies. Each point was tested in duplicate. Concentrations of standard IgG were: 500; 250; 125; 62.5; 31.3; 15.6; 7.8; and 3.9 µg/ml. The ADAMSEL FLP b039 software  was used to analyse the optical density (OD) of the plates at 450nm and interpolate the standard curve (µg/ml). Discordant duplicates (with a variation coefficient >15%) were dropped.
The Median test was used to analyze differences between IgG medians concentrations. The comparisons between IgG median concentrations in Zinder, Dosso and Gaya were performed to investigate the potential impact of SMC on Antibody responses. The comparison between IgG median concentrations were performed by Mann-Whitney test. Data were analyzed with SPSS software version 16.0. P-values ≤ 0.05 were considered were considered statistically significant.