The effect of nonresponse on prevalence estimates for a referent population: insights from a population-based cohort study. Atherosclerosis Risk in Communities (ARIC) Study Investigators

Ann Epidemiol. 1996 Nov;6(6):498-506. doi: 10.1016/s1047-2797(96)00104-4.


Characterization of nonrespondents, with the aim of detecting nonresponse bias, is a crucial component of prospective studies. This study was undertaken to investigate the demographic and health characteristics of nonrespondents to a population-based cohort study of cardiovascular disease, to determine whether early-stage nonrespondents differ from late-stage nonrespondents, and to estimate the bias in prevalence estimates for the source population. Sixty-seven percent of eligible subjects completed all phases of the cohort recruitment. Compared to respondents, nonrespondents were less likely to be married, less likely to be employed, and less likely to be well educated. Nonrespondents tended to describe their general health in less favorable terms and were more likely to be smokers. Their reported disease profile, however, was not dissimilar to that of respondents. For several demographic and health characteristics, including marital status, education, and smoking, early-stage nonrespondents differed from respondents more than did late-stage nonrespondents. For example, 42% of early nonrespondents were smokers compared to 37% of late nonrespondents and 22% of respondents. Overall, the bias in prevalence estimates related to nonresponse was small (< 5%) for most of the measured characteristics. Although nonresponse to health surveys is associated with typical attributes, early nonrespondents differ from respondents more than do late-stage nonrespondents. With few exceptions, however, a 33% nonresponse rate did not appear to introduce substantial bias into prevalence estimates for the source community.

Publication types

  • Comparative Study
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Age Distribution
  • Cardiovascular Diseases / epidemiology*
  • Cohort Studies
  • Epidemiologic Methods*
  • Female
  • Health Surveys
  • Humans
  • Male
  • Middle Aged
  • Minnesota / epidemiology
  • Multivariate Analysis
  • Patient Compliance
  • Prevalence
  • Reproducibility of Results
  • Selection Bias
  • Sex Distribution