ABSTRACT
INTRODUCTION: the age reported by or on behalf of centenarians may be suspect unless proven correct. We report the validity of age reports in a population-based sample of centenarians living in New England and the prevalence of centenarians in an area within the North Eastern USA. METHODS: cohort study. All centenarians in a population-based sample detected by local censuses. Ages were confirmed by birth certificate. Type of residence and whether the subject was living independently were also recorded. RESULTS: from a population of about 450,000 people, 289 potential centenarians were reported by the censuses of the eight towns participating in the study. Of these, 186 (64%) had died at the time centenarian prevalence was determined. Of the 80 still alive, 13 (16%) had incorrect birth years recorded by the censuses. The specificity of the censuses for stating the number of centenarians alive and living in the sample was 28-31%. Using additional sources, only four more centenarians were located, indicating that the sensitivity of the censuses approached 100%. We had an 83% success rate in obtaining proof of age in those families we interviewed. In all instances, age and birth order of children were an important source of corroborative evidence and in no case did we detect inconsistencies with the families' reported ages of the centenarian subjects. Therefore, there were at least 46 centenarians or approximately 1 centenarian per 10,000 people. CONCLUSIONS: age validation can be performed for most centenarians in the North Eastern USA. Self or family reports of those between the ages of 100 and 107 years were dependable.
