The study was conducted during April and December 2014 in two different areas of the Tanzanian mainland. The first site was Bondo in the Tanga region, inhabited by 7970 people . Bondo lies at 309 meters above sea level, the coordinates of the area are 5°22'60" N and 38°34'60" E. The second study site was Hai in Kilimanjaro region, located at the foot of Mount Kilimanjaro, with 67225 residents and an elevation of 1000-1411 meters above sea level and the coordinates of 3° 12' 0" N and 37°13' 60" E. The selected study areas have diverse malaria transmissions and different ecologies and, malaria interventions have been differently deployed in these areas as well. Tanga used to be a high-endemic area in Tanzania and great effort has been done in controlling the disease. Kilimanjaro region is known to have very low transmission rates due to factors such as altitude and vector composition. Participant recruitment procedures and study design have been previously described , Figure 1.
Meetings were organized wherein the study staff explained the study and answered any questions in an open forum. Upon commencing participant recruitment, enrolment occurred in health facilities at each site. Participants were enrolled only after verification of potential subject eligibility, explaining the study in the Swahili language. Each member of the community had an equal chance of being selected as a participant. The list of all members in the village was listed and with a lottery method, each member of the population was assigned a number, after which participants were selected at random. The minimum age of the study participants was 2 years and above.
Sample size calculation
The sample size was calculated assuming the following parameters seroprevalence of anti-AMA/MSP 50% was used, P = 0.5 (proportion in the population), Power = 0.80, Alpha = 0.05 (two sided), Anticipated difference = 0.1, Alternative p = 0.4, Design effect = 2. Estimated required sample size in each site was, n = 194*2 = 388 per study, (Using STATA software). A minimum required sample was 776. In this study we enrolled 788 participants
A blood sample was obtained by finger prick. A portion of blood was used for malaria rapid test, which was performed by well-trained staff at each site. A blood spot was prepared for each participant, then dried and stored for further analysis. A 3.0 mm diameter circle of dried blood spot (equivalent to 2µl whole blood/1µl serum) was reconstituted in 200µl of sodium azide-phosphate buffered saline-tween (0.05%) (PBST/0.1% Azide). The solution is approximately 1:100 of whole blood which is equivalent to 1:200 dilution of antibodies concentration.
Enzyme-Linked Immuno-Sorbent Assay (ELISA)
Indirect immunosorbent Assay (ELISA)was performed using two Plasmodium falciparum surface antigens, Plasmodium falciparum MSP 119 (PfMSP 119) and Plasmodium falciparum AMA-1(PfAMA-1) . Briefly, the recombinant antigens (MSP_1 and AMA_1) were coated to each well of flat bottom high binding microtiter plates (Greiner bio-one, Germany) at a concentration of 0.5µg/ml in sodium carbonate-sodium bicarbonate buffer (pH 9.5) and incubated at 40C overnight. The plates were then washed three (3) times with phosphate-buffered saline and tween 20 (0.05%) and blocked with 1% (w/v) skimmed milk for 3 hours at room temperature. Samples and controls were then added in duplicate, positive controls were from pooled positive samples from highly endemic areas and negative controls were from European malaria naïve individuals. After three washes with Phosphate Buffered Saline (PBS), 50µl of horse-radish peroxidase-conjugated rabbit anti-human IgG diluted at 1:5000 in PBS was added to each well and incubated at room temperature for hours. O-phenylenediamine (Sigma-fast OPD) was used as the peroxidase substrate, the reaction was stopped by 2M sulphuric acid 15 minutes after adding the substrate. Optical density (OD) values were read using Microplate ELISA reader (Elx 808; USA), at the wavelength filter of 490nm.
Malaria parasite detection by polymerase chain reaction (PCR)
Parasite DNA was extracted using the simple chelex method, a dried blood spot of about 8mm in diameter was cut and placed in a 1.5 ml microcentrifuge tube containing 1ml of PBS with 10% saponin and incubated for 4 hours at a 4̊0C fridge. The spot was then soaked in PBS for15 minutes at 4̊0C, after the incubation, tubes were centrifuged, and the supernatant was discarded. 150µl of 6% chelex solution was then added and incubated at 95 ̊C for 10 minutes with vertexing periodically through the incubation time. The final centrifugation for 5 minutes at high speed was done and the supernatant was transferred to a clean microcentrifuge tube and stored at -200C freezer until used.
Plasmodium nucleic acid amplification was conducted using genus-specific reverse and forward primers (rPLU6-5`TTAAAATTGTTGCAGTTAAAACG3` and rPLU5-5`CCTGTTGTTGCCTTAAACTCC3`) targeting small sub-unit ribosomal RNA (ssurRNA) of the parasite. A reaction mix of 20µl per sample was used, 5µl of template DNA extracted from participants whole blood plus 15µl of nuclease-free water, dNTPs, Taq enzyme, buffers, and salts. Amplification conditions were, 95ºC for 5 minutes followed by 30 cycles of 940C for 1minute, 580C for 2 minutes and 720C for 5 minutes then one final extension cycle at 72ºC for 10 minutes [27,28]. Amplification products were run in Ethidium bromide agarose gel (2%) electrophoresis at 120 volts, 50 watts and 120mA. The amplified bands were visualized under an ultra-violet light trans-illuminator.
All data were analyzed using SPSS 20.0 software (SPSS Inc., Chicago, IL, USA) and GraphPad Prism8 software (San Diego, CA). Data cleaning was conducted before data analysis. Data were log-transformed before statistical analysis. After verifying that Optical density (OD) values were not normally distributed (p<0.0001; Anderson-Darling test), non-parametric tests were performed to compare the OD values. The Mann–Whitney test was used for the comparison of Antibody levels of two independent groups (i.e positive versus negative). The non-parametric Kruskal–Wallis test was used for the comparison of more than two groups (i.e age groups). Pearson’s Chi-squared (χ2) test was used to compare two proportions. All observed differences were considered significant at p<0.05. The cut-off for seropositivity among samples was determined as the mean OD of the negative control sera plus 3 standard deviations.