An intergenerational and lifecourse study of health and mortality risk in parents of the 1958 birth cohort: (I) methods and tracing

Objectives. This paper describes the methods used to obtain information on mortality and cancer registrations for the parents of the British 1958 birth cohort, in order to create a dataset that can be used to examine intergenerational relationships on health and growth. Study design. Intergenerational cohort study. Methods. The 1958 cohort includes all births occurring during 1 week in March 1958 in England, Scotland and Wales. For more than four decades of follow-up, information has been collected on cohort members, their parents and children. Information on the National Health Service (NHS) numbers of the parents was not available, but other details were collated for the Office for National Statistics to trace and flag the biological parents of the cohort members. Results. Tracing was successful in 90.2% of fathers (n= 14,334) and 94.9% of mothers (n=15,076). The greatest success was achieved for parents in families where there was no indication for additional mother or father figures until the child was 16 years old (96.6% of the mothers traced, n=14,274; 94.3% of the fathers traced, n=13,256). Tracing rates were lower than average in unmarried mothers (59%) and for the small group who were separated, widowed or divorced in 1958 (81%); the rates were particularly poor for the corresponding fathers (24.4 and 54.7%, respectively). There were only small variations in tracing rates between different regions of Britain. Conclusions. The tracing rates achieved were generally very high despite the lack of NHS number, especially where there was family stability throughout the childhood of cohort members. Parental status will need to be considered in future studies. With the high tracing rates achieved, the dataset provides an important resource with which to evaluate multigenerational associations with health and development in parents, their offspring and grandchildren.