Rapid report on methodology: does loss to follow-up in a cohort study bias associations between early life factors and lifestyle-related health outcomes?