Dodge H: A follow-up comparison of study participants and refusers within a rural elderly population. J Gerontol A Biol Sci Med Sci

Background. Survey rates are known to decline with age. Survey results can be affected by response bias if nonre-sponders are more, or less, likely than participants to suffer from the condition being studied. For instance, it is possi-ble that older adults with dementia would be less likely to participate in a study of dementing disorders. Methods. A random sample of a rural U.S. population aged 65+ years yielded 1,422 participants and 912 refusers in addition to others who were ineligible, inaccessible, or untestable. Participants and refusers were compared on age, sex, 5-year mortality, and causes of death suggestive of dementia as listed on death certificates. Results. Compared to participants, refusers were significantly older and more likely to be women, with mortality similar to that of participants at approximately 5-year follow-up. Death certificate data revealed no significant differ-ences in reported causes of death indicating or suggesting dementia. Conclusions. In this population, those who refused to participate in a dementia survey were not more likely to be ill or demented than those who did participate. F>R the optimal interpretation and comparison of theresults of research, potential sources of bias in the study populations should be known. Response rates in survey research decrease with age (1), potentially reducing general-izability and external validity of study results. Low respons