Historical cohort studies and the early origins of disease hypothesis: making sense of the evidence