Differential item functioning of PROMIS Sleep Disturbance-Short Form 8a across levels of health literacy in a community sample of adults.

Abstract

Limited health literacy may contribute to systematic bias in health assessments relying on self-report. The present study aimed to examine how varying levels of health literacy might affect scores on the Patient-Reported Outcomes Measurement Information System Sleep Disturbance-Short Form 8a, a commonly used tool to assess self-reported perceptions of sleep quality. A community-based sample of adults (≥18 years) from Boston and Chicago was included in the study. Health literacy was measured using the Health Literacy Assessment Using Talking Touchscreen Technology (Health LiTT). We conducted a differential item functioning analysis using an ordinal logistic regression and a graded response model. Monte Carlo simulations were employed to establish effect size thresholds. We analyzed data from 702 participants, with a mean age of 49.1 ± 15.6 years. The sample included 61.5% females, 41.6% Black participants, and 27.8% Hispanic participants, with 58% demonstrating low health literacy. Participants in the low health literacy group reported higher sleep disturbance levels compared to those in the adequate health literacy group. Four of the eight items on the Patient-Reported Outcomes Measurement Information System Sleep Disturbance-Short Form 8a scale exhibited statistically significant differential item functioning by health literacy, indicating that individuals with varying levels of health literacy interpret or respond to these items differently, even when they have the same underlying sleep disturbance level. The overall impact of differential item functioning on total scores, however, was negligible. The psychometric properties of the Patient-Reported Outcomes Measurement Information System Sleep Disturbance-Short Form 8a support its use across populations with varying levels of health literacy. (PsycInfo Database Record (c) 2026 APA, all rights reserved).