Attrition in longitudinal population studies: Does it affect the generalizability of the findings?: An introduction to the series☆