Table 3. Ab-NK is associated with a reduced risk of clinical malaria episodes in children from Junju, Kenya.

The table shows the incidence risk ratios (IRRs), hazard ratios (HRs), and 95% confidence intervals (95% CI) comparing the outcomes of children with high versus low Ab-NK responses in modified Poisson and Cox regression models, respectively. Data are presented as grouped analysis based on outcomes of interest: children who experience at least one clinical episode of malaria (i.e., fever over 37.5°C plus parasitemia above a predefined threshold, n = 122). The “#” labeled panel shows the mean and 95% CI for the time to malaria episode analysis comparing children with high versus low Ab-NK responders who had their first malaria episode during 6 months of follow-up. All multivariate models were adjusted for the potential confounding effects of age and previous P. falciparum exposure.

Junju children n = 293)	Proportion of children with			Univariate analysis			Multivariate analysis
	High Ab-NK responses	Low Ab-NK responses		Ratio (95% CI)	P value		Ratio (95% CI)	P value
Acquired malaria episodes (n = 122)
Incidence risk ratio (cutoff values)
CD107a (6.32)	17% (10/56)	47% (112–237)		0.37 (0.11–0.67)	0.001		0.40 (0.22–0.71)	0.002
IFN-γ (0.93)	29% (15/51)	44% (107/242)		0.66 (0.42–0.97)	0.04		0.61 (0.39–0.97)	0.038
Time to treatment
Hazard ratio
CD107a	17% (10/56)	47% (112–237)		0.32 (0.17–0.61)	0.001		0.27 (0.13–0.52)	<0.000
IFN-γ	29% (15/51)	44% (107/242)		0.62 (0.36–0.99)	0.045		0.54 (0.31–0.97)	0.040
Time to first malaria episode (in days)#	Mean (95% CI)	Mean (95% CI)		P value
CD107a	178 (160–195)	137 (126–147)		0.001
IFN-γ	160 (139–182)	141 (131–151)		0.045