Table 1.

The incidence of infectious and atopic diseases in childhood. Estimates of intervention effect sizes that can be detected in the cohort

	Infants to age one year							Children aged one to five years
		All infants		Infants of mothers with hookworm		Infants of mothers with S. mansoni			All children		Children of mothers with hookworm		Children of mothers with S. mansoni
	Expected rate, placebo group per year	Expected pyr	RR	Expected pyr	RR	Expected pyr	RR	Expected rate, placebo group per year	Expected pyr	RR	Expected pyr	RR	Expected pyr	RR
Pneumonia	0.25	1860	0.76	874	0.66	354	0.49	0.1	7948	0.81	3736	0.73	1510	0.59
Diarrhoea	1.9		0.91		0.87		0.80	1		0.94		0.91		0.86
Malaria	0.5		0.82/1.19		0.75/1.29		0.62/1.47	0.8		1.07		1.11		1.17
Eczema	0.25		1.28		0.70		2.19	0.05		1.30		1.76		2.31
Wheeze	0.08		1.51		2.20		3.14	0.04		1.34		1.76		2.31
TB infection	0.03		0.38		0.17		Not possible	0.03		0.67		0.54		0.32
Tuberculosis	0.005		Not possible		Not possible		Not possible	0.005		0.29		0.06		Not possible

The table shows expected disease incidence rate in the placebo group, and expected total (intervention + control) person years of follow-up for infants and children, for the whole study population and for subgroups whose mothers had hookworm or schistosomiasis. The smallest effects that the study has 80% power to detect, with a = 0.05 are shown as rate ratios. The direction of effect shown is that predicted by our hypothesis; both directions of effect are shown for malaria in infancy where either would be plausible. TB: Mycobacterium tuberculosis. Not possible: insufficient person years available in the study to detect an effect with 80% power and a = 0.05.