The study was conducted in children aged ≤ 12 years with acute, uncomplicated, P. falciparum malaria in Ibadan, a malaria endemic area in southwestern Nigeria 17. Fully informed consent was obtained from the parents/guardians of each child. Inclusion criteria were: fever or history of fever in the 24–48 h preceding presentation, pure P. falciparum parasitaemia ≥ 2,000 asexual forms/μL, absence of other concomitant illness, no history of antimalarial use in the 2 weeks preceding presentation, and negative urine tests for antimalarial drugs (Dill-Glazko and lignin). Patients with severe malaria 18, severe malnutrition, serious underlying diseases (renal, cardiac, or hepatic), and known allergy to study drug were excluded from the study. The study was approved by the Ethics Committee of the Ministry of Health, Ibadan, Nigeria.

After clinical assessment, blood was obtained for haematocrit determination and for quantification of asexual and sexual parasitaemia. Patients were treated with 25–30 mg/kg of amodiaquine base (Camoquine^®) given orally over 3 days. All patients waited for at least 3 h after to ensure the drug was not vomited. If it was, the patient was excluded form the study.

Oral paracetamol (acetaminophen) at 10–15 mg/kg 6–8 hourly was given for 12–24 h if body temperature was > 38°C. Patients were seen daily, at approximately the same time of the day for the first four days (days 0–3) and then daily on days 7, 14, 21, 28, 35 and 42 after treatment had begun. At each visit, patients were assessed clinically, and thick and thin blood smears were obtained for quantification of parasitaemia. The fever clearance time (FCT) was defined as the time taken for the body temperature to fall below 37.5°C and remain below this value for > 48 h.

Asexual parasite and gametocyte counts were measured daily for the first four days (days 0–3) and thereafter on days 7, 14, 21, 28, 35 and 42. Quantification in Giemsa-stained thick blood films was undertaken against 500 leukocytes in the case of asexual parasitaemia, and against 1000 leukocytes in the case of gametocytes, and from this figure, the parasite density was calculated assuming a leukocyte count of 6,000/μL of blood. Parasite clearance time (PCT) was the time interval from the start of antimalarial treatment until the asexual parasite count fell below the detectable levels in a peripheral blood smear. Capillary blood, collected before and during follow-up, was used to measure packed cell volume (PCV). PCVs were measured using a microhaematocrit tube and microcentrifuge (Hawksley, Lancing, UK). Routine haematocrit was undertaken on days 0, 3, 7, 14, 21, 28, 35 and 42. Blood was spotted on filter papers on days 0, 1, 3, 7, 14, 21, 28, 35 and 42 and at the time of re-appearance of peripheral parasitaemia after its initial clearance for parasite genotyping. Molecular genotyping was carried out as previously described 16.

Gametocyte sex determination was based on the following criteria 1920: males (microgametocytes) are smaller than females (macrogametocytes), the nucleus is larger in males than females, the ends of the cells are rounder in males and angular in females, with Giemsa the cytoplasm stains purple in males and deep blue in females, and the granules of malaria pigment are centrally located females and more widely scattered in males. The sex ratio was defined as the proportion of gametocytes in peripheral blood that were male 21. Gametocytes were sexed if the gametocyte density was ≥ 10/μL blood.

Data were analyzed using version 6 of the Epi-Info software 22 and the statistical programme SPSS for Windows version 10.01 23. Variables considered in the analysis were related to the densities of P. falciparum gametocytes and trophozoites. Proportions were compared by calculating χ² with Yates' correction or by Fisher exact or by Mantel Haenszel tests. Normally distributed, continuous data were compared by Student's t-tests and analysis of variance (ANOVA). Data not conforming to a normal distribution were compared by the Mann-Whitney U-tests and the Kruskal-Wallis tests (or by Wilcoxon ranked sum test). Kaplan-Meier plots are also presented to compare gametocyte carriage rates, and the duration of carriage of a male biased gametocyte sex ratio following treatment in those who were gametocytaemic at presentation. Differences in survival time were assessed by inspection of Kaplan-Meier curves and log-rank tests. The relationship between gametocyte sex ratio and gametocyte or asexual parasite density was assessed by linear regression. A multiple logistic regression model was used to test the association between a male biased sex ratio, that is, sex ratio ≥ 0.5 (yes or no at presentation or during follow up) and factors that were significant at univariate analysis: presence of fever, haematocrit < 25%, asexual parasitaemia > 20,000/μL, and gametocytaemia < 18/μL. Because the study was conducted over a period of six and a half year, time in years since the commencement of the study was included as a covariate in the model for pre-treatment male-biased sex ratio. All tests of significance were two-tailed. P-values of ≤ 0.05 were taken to indicate significant differences. Data were (double)-entered serially using the patients' codes and were only analyzed at the end of the study.

Between September 2000 and December 2006, 615 children (297 males, 318 females) with P. falciparum malaria, aged between 0.5–12 years (mean ± standard deviation [SD] = 6.5 ± 3.2 years) were enrolled. Of these, 612 (294 males, 318 females) completed at least 28 days of follow up and were analyzed (Table 1). The characteristics of the children were similar during the study periods except for the geometric mean parasite density that was significantly higher in 2006 than in 2000 and 2004 (P < 0.0001) (Table 1).

All children responded promptly to treatment, and none developed severe malaria. The overall mean (range) FCT was 1.1 (SEM 0.02) d, and was not significantly different between the years of enrolment (Table 1). None of the studied children had significant adverse effects as monitored by clinical symptoms, but overall, 46 children reported pruritus, which did not interfere with sleep.

The overall mean PCT (SEM) was 2.7 (0.005) d and was not significantly different during the study periods (Table 1). Recrudescent infections confirmed by polymerase chain reaction (PCR) occurred in 25 children: in 6, 8, and 11 children in 2000, 2004, and 2006, respectively. The characteristics of the recrudescent infections will be reported elsewhere.

Gametocytes were detected in peripheral blood in 122 (20%) children (in 66 children before treatment and in 56 children after initiation of treatment) (Table 2). Gametocyte detection rates before, during and after treatment were similar during the study periods (Table 3). Gametocyte densities during the study periods are summarized in Table 3. Overall, gametocytaemia increased significantly by days 3–7 during follow up (P < 0.0001) but did not differ between the study periods.

The probability of a mosquito infectivity following a blood meal is related to gametocyte density and the duration of carriage by the host. Figure 1 is a Kaplan-Meier plot (survival curve) of the cumulative probability of remaining gametocyte free following treatment with amodiaquine during the study periods. The probabilities were similar during the periods of the study (Log rank statistic = 0.8, P = 0.23).

In 122 children who were gametocytaemic at presentation or during follow up, a total of 2,286, 2,197, 3,108, 4,676, 2,880, 618, 54 and 576 gametocytes were counted on days 0, 1, 2, 3, 7, 14, 21, and 28, respectively. Of these, 2,050, 1,986, 2,843, 4,518, 2,699, 600, 54, and 440 gametocytes could be sexed on days 0, 1, 2, 3, 7, 14, 21, and 28, respectively. The corresponding number of patients in whom the gametocytes were counted was 66, 49, 47, 70, 50, 29, 5, and 4, respectively.

Following treatment with amodiaquine, gametocyte sex ratio increased significantly over the course of the infection. (Figure 2 and Table 4): 7.6% of the gametocytes were male at day 0, 25% at day 3, and 29% at day 7 (X² = 15.2, P = 0.0005). Gametocyte sex ratio at enrolment increased significantly during the study periods: the ratio was 0.4% in 2000, 4.8% in 2004, and 20% in 2006 (P < 0.0001), although it was still largely female-biased (Table 4). The variations in haematocrit, density of gametocyte and gametocyte sex ratio are shown in Figure 2. During the first week of follow up, gametocyte sex ratio increased as the packed cell volume decreased and the gametocyte density increased, suggesting that the latter may have been a direct contribution of the effect of amodiaquine on sex ratio, since, in general, low but not relatively high gametocytaemia is associated with male-biased sex ratio.

The possibility that gametocyte infectivity may be related to the duration of carriage of a less female biased sex ratio was examined. Figure 3 is a Kaplan-Meier plot of the cumulative probability of remaining a less female-biased sex ratio free following treatment with amodiaquine during the study periods. The probabilities were similar during the periods of the study (Log rank statistic = 2, P = 0.37).

The percentage of male gametocytes was negatively correlated with gametocyte density but not with asexual parasite density (Spearman's r = -0.46, P < 0.0001 and r = 0.009, P = 0.94, respectively with n = 66 in both cases.

The pre-treatment factors associated with a male-biased sex ratio are shown in Table 5. Age, gender or duration of illness before presentation appeared not to be associated with a male-biased sex ratio at enrolment. Absence of fever, a haematocrit < 25%, asexual parasitaemia > 20,000/μL, gametocytaemia < 18/μL, and enrolment in 2006 were significantly associated with a male-biased sex ratio at enrolment.

Following treatment, a haematocrit < 25% on days 0 and 7 and a parasitaemia greater > 20,000/μL of blood were independent predictors of a male biased sex ratio 7 days post initiation of treatment with amodiaquine (Table 6).