Low parasitemia in submicroscopic infections significantly impacts malaria diagnostic sensitivity in the highlands of Western Kenya

Abstract

Asymptomatic malaria infections represent a major challenge in malaria control and elimination in Africa. They are reservoirs of malaria parasite that can contribute to disease transmission. Therefore, identification and control of asymptomatic infections are important to make malaria elimination feasible. In this study, we investigated the extent and distribution of asymptomatic malaria in Western Kenya and examined how varying parasitemia affects performance of diagnostic methods including microscopy, conventional PCR, and quantitative PCR. In addition, we compared parasite prevalence rates and parasitemia levels with respect to topography and age in order to explore factors that influence malaria infection. Over 11,000 asymptomatic blood samples from children and adolescents up to 18 years old representing broad areas of Western Kenya were included. Quantitative PCR revealed the highest parasite positive rate among all methods and malaria prevalence in western Kenya varied widely from less than 1% to over 50%. A significantly lower parasitemia was detected in highland than in lowland samples and this contrast was also observed primarily among submicroscopic samples. Although we found no correlation between parasitemia level and age, individuals of younger age group (aged <14) showed significantly higher parasite prevalence. In the lowlands, individuals of aged 5-14 showed significantly higher prevalence than those under age 5. Our findings highlight the need for a more sensitive and time-efficient assay for asymptomatic malaria detection particularly in areas of low-transmission. Combining QPCR with microscopy can enhance the capacity of detecting submicroscopic asymptomatic malaria infections.

Fig 1. Malaria prevalence of studied sites in western Kenya based on (A) microscopy and (B) conventional PCR.

Locality information can be referred to S1 Table. Areas of elevation below 1500 m were indicated by light gray and above 1500 m by dark gray. Black, blue, and red circles represent sites of low (<5%), moderate (5–25%), and high (>25%) malaria prevalence.

Fig 2. Boxplots comparing (A) prevalence rate detected by microscopy and conventional PCR methods for all sites as well as the lowland and highland sites separately.

Numbers above bars indicate number of sites included. Asterisks indicate level of significance; (B) parasitemia level and parasite DNA quantity obtained by microscopy and quantitative polymerase chain reaction (QPCR), respectively, between lowland and highland samples. Numbers above bars indicate number of individuals included. The central box represents the interquartile range and the whiskers represent the first quartile and the fourth quartile of the data. The median is shown as a lien through the center of the box and the ends of the whiskers correspond to the minimum and maximum in the data.

Fig 3. Boxplots showing the amount of parasite DNA detected by SYBR quantitative polymerase chain reaction (QPCR) analysis of subset samples that were diagnosed as positive by conventional PCR.

Comparison of parasite DNA quantity was made between (1) microscopy positive and negative samples; and (2) lowland and highland samples. Numbers above bars indicate number of individuals included. The central box represents the interquartile range and the whiskers represent the first quartile and the fourth quartile of the data. The median is shown as a lien through the center of the box and the ends of the whiskers correspond to the minimum and maximum in the data.

Fig 4. Scatter plot matrix showing correlation of estimates of parasite density obtained by microscopy and parasite DNA quantity by SYBR quantitative polymerase chain reaction (QPCR) analysis of blood samples.

Pearson’s product moment (r) correlation coefficients were indicated.

Fig 5. Histogram showing the mean malaria prevalence rate of the three age groups (under 5, aged 5–14, and over 14) in the lowlands and highlands.

Error bars indicated the standard deviation of the mean value. The level of significance was indicated.