OBJECTIVE: To impute limited health literacy from commonly measured socio-demographic data and to compare it to the Short-Test of Functional Health Literacy in Adults (S-TOFHLA) for estimating the influence of limited health literacy on health status in the elderly.

METHODS: The Prudential Medicare Study assesses the S-TOFHLA score, leading to a "reference standard" classification of 25% of people with inadequate literacy; the National Health Interview Survey has no such assessment. We estimated a regression of S-TOFHLA on sex, age, years of schooling, and race/ethnicity in The Prudential Medicare Study data to derive a Demographic Assessment for Health Literacy (DAHL) score, and imputed inadequate literacy to the 25% with the lowest DAHL scores. Using regression, we then examined associations between several health status measures (including hypertension, diabetes, physical and mental SF-12) and inadequate literacy (imputed or test-based).

RESULTS: Estimates of association using imputed inadequate literacy closely approximate those obtained using S-TOFHLA-based inadequate literacy for most outcomes examined.

CONCLUSIONS: As few population surveys measure health literacy, the DAHL, a readily calculated health literacy proxy score, may be useful for expanding the scope of health literacy research in national survey data.