Affect-Biased Health Perceptions: How Global and Daily Affect Influence the Correspondence Between Objective and Subjective Health in Older Adults

Research has revealed the presence of affective bias in self-reports and self-assessments of global health, resulting in a discrepancy between individuals? reports of their own health and those of more objective health assessments. This bias, which occurs with both negative and positive affect (NA/PA) on both trait and state levels, can be problematic in the contexts of medical treatment and academic research. The present study investigated the presence of affectively-biased health perceptions on both global and daily levels, and also explored cross-level affective bias effects. Participants were 152 older adults who completed: scales measuring global affect and perceived health; daily questionnaires for 56 days on affect, health events, and health satisfaction; and an in-person health assessment including a health interview, a physical, a blood assay, and the collection of medical records. On the global level, structural equation modeling was used to evaluate the presence of affective health bias, so that the relationship between an Objective Health factor and a Perceived Health factor is moderated by global NA and PA. On the daily level, multilevel modeling was used to test the extent to which the impact of a given day?s health events on that day?s health satisfaction is moderated by that day?s NA and PA. Additional models also tested the degree to which health bias is primarily a trait process (dependent on global affect) or a state process (dependent on day-level affect). Results revealed evidence for NA bias on the global level; PA did not significantly moderate the association between the Objective and Perceived Health factors, but results did suggest that PA serves to counteract NA bias to an extent. On the daily level, both NA and PA biased day-level health perceptions individually, but only NA maintained its moderating effect when both affective biases were tested simultaneously. None of the cross-level effects in the day- vs. global-level affect bias models were significant, indicating that health bias processes are largely confined within level of measurement/analysis. These results inform future procedures of medical professionals and academic researchers who rely on subjective health data for appropriate treatment decisions and accurate empirical conclusions.