RESUMEN
BACKGROUND: Low-density asymptomatic Plasmodium infections are prevalent in endemic areas, but little is known about their natural history. The trajectories of these infections and their propensity to fluctuate to undetectable densities can affect detection in clinical trials and field studies. We aimed to classify the natural history of these infections in a high transmission area over 29 days. METHODS: In this longitudinal cohort study, we enrolled healthy, malaria-asymptomatic, afebrile, adults (age 18-59 years) and older children (age 8-17 years) in Katakwi District, Uganda, who were negative for Plasmodium infection on rapid diagnostic tests. Participants were instructed to self-collect one dried blood spot (DBS) per day for a maximum of 29 days. We excluded people if they were pregnant or taking antimalarials. During weekly clinic visits, staff collected a DBS and a 4 mL sample of venous blood. We analysed DBSs by Plasmodium 18S rRNA quantitative RT-PCR (qRT-PCR). We classified DBS by infection type as negative, P falciparum, non-P falciparum, or mixed. We plotted infection type over time for each participant and categorised trajectories as negative, new, cleared, chronic, or indeterminate infections. To estimate the effect of single timepoint sampling, we calculated the daily prevalence for each study day and estimated the number of infections that would have been detected in our population if sampling frequency was reduced. FINDINGS: Between April 9 and May 20, 2021, 3577 DBSs were collected by 128 (40 male adults, 60 female adults, 12 male children, and 16 female children) study participants. 2287 (64%) DBSs were categorised as negative, 751 (21%) as positive for P falciparum, 507 (14%) as positive for non-P falciparum, and 32 (1%) as mixed infections. Daily Plasmodium prevalence in the population ranged from 45·3% (95% CI 36·6-54·1) at baseline to 30·3% (21·9-38·6) on day 24. 37 (95%) of 39 P falciparum and 35 (85%) of 41 non-P falciparum infections would have been detected with every other day sampling, whereas, with weekly sampling, 35 (90%) P falciparum infections and 31 (76%) non-P falciparum infections would have been detected. INTERPRETATION: Parasite dynamics and species are highly variable among low-density asymptomatic Plasmodium infections. Sampling every other day or every 3 days detected a similar proportion of infections as daily sampling, whereas testing once per week or even less frequently could misclassify up to a third of the infections. Even using highly sensitive diagnostics, single timepoint testing might misclassify the true infection status of an individual. FUNDING: US National Institutes of Health and Bill and Melinda Gates Foundation.