I suspect the sample was chosen to get two equal sets, as for an experiment like this you want to maximise the data from the two groups, as you are not trying to measure the incidence of the effect but rather the differences from the effect.