Fig. 2

Relationship between parasite density and gametocyte carriage. a Estimated probability of gametocyte carriage as a function of parasite density. Gametocyte stages are more likely to occur in infections with low parasite density. b Parasite density (number of parasites / µl of blood) in gametocyte-negative (0) and –positive (1) patients. Each point represents a blood sample from a P. falciparum-infected patient (n = 676 gametocyte negative patients and 48 gametocyte carriers). The horizontal bold black line represents the median value of parasite density for each of the two groups, and the upper and lower boundaries of the box indicate the 75th and 25th percentile, respectively. c MOI level as a function of parasite density for both gametocyte-negative (0) and –positive (1) patients. Each colour line represents a linear relationship (± se) fitted to the MOI values for each group (red = non-carriers and blue = gametocyte carriers). The presence of gametocytes in patient blood was determined by microscopic observation and MOI was defined as the number of different parasite genotypes (based on msp1 and msp2 genotyping) co-existing within a particular infection. d Body temperature (°C) as a function of parasite density for both gametocyte-negative (in red) and –positive (in blue) patients. The lines represent a linear relationship (± se) fitted to the temperature values for each group. Because there were 52 missing values in the temperature dataset temperature, the sample size in d is different from that of (a) and (c) (n = 625 gametocyte negative patients and 47 gametocyte carriers)