Repeated findings of weak association in a well- designed study can still lead to effective public health action. Strength of association supports a hypothesis of causality, but weak association supported by other evidences of causality are also important.