Attrition in a longitudinal study of an aged population

Area of Science:

• Gerontology
• Epidemiology
• Sociology

Background:

• Longitudinal studies are crucial for understanding aging processes.
• Attrition, the loss of participants over time, poses a significant challenge in such research.
• Understanding attrition patterns is vital for ensuring the validity of findings in studies of older populations.

Purpose of the Study:

• To analyze attrition in a longitudinal study of older adults.
• To compare characteristics of reinterviewed participants with those lost to follow-up or deceased.
• To identify factors associated with attrition in a retirement community setting.

Main Methods:

• A 1959 probability sample (N=2544) of noninstitutionalized Florida residents aged 65+ was surveyed.
• Data included health status, practices, social factors, and activities.
• A follow-up in 1968 analyzed 2497 participants categorized as reinterviewed, lost to follow-up, or deceased.

Main Results:

• Approximately one-third of the sample was lost to follow-up, one-third deceased, and one-third reinterviewed.
• Reinterviewed individuals differed from dropouts in age, sex, housing, activity patterns, and chronic conditions.
• No significant differences were observed in education, household composition, occupation, or retirement status (when controlling for sex).

Conclusions:

• Attrition in longitudinal studies of older adults is substantial and selective.
• Participant characteristics influence the likelihood of remaining in or dropping out of a study.
• Findings highlight the need to account for attrition bias in gerontological research.