Anxiety, worry and cognitive risk estimate in relation to protective behaviors during the 2009 influenza A/H1N1 pandemic in Hong Kong: ten cross-sectional surveys

Background: Few studies have investigated associations between psychological and behavioral indices throughout a major epidemic. This study was aimed to compare the strength of associations between different cognitive and affective measures of risk and self-reported protective behaviors in a series of ten cross-sectional surveys conducted throughout the first wave of influenza A/H1N1 pandemic. Methods: All surveys were conducted using questionnaire-based telephone interviews, with random digit dialing to recruit adults from the general population. Measures of anxiety and worry (affective) and perceived risk (cognitive) regarding A/H1N1 were made in 10 serial surveys. Multivariate logistic regression models were used to estimate the cognitive/affective-behavioral associations in each survey while multilevel logistic models were conducted to estimate the average effects of each cognitive/affective measure on adoption of protective behaviors throughout the ten surveys. Results: Excepting state anxiety, other affective measures including anticipated worry, experienced worry and current worry specific to A/H1N1 risk were consistently and strongly associated with adoption of protective behaviors across different survey periods. However, the cognitive-behavioral associations were weaker and inconsistent across the ten surveys. Perceived A/H1N1 severity relative to SARS had stronger associations with adoption of protective behaviors in the late epidemic periods than in the early epidemic periods. Conclusion: Risk-specific worries appear to be significantly associated with the adoption of protective behaviors at different epidemic stages, whereas cognitive measures may become more important in understanding people's behavioral responses later in epidemics. Future epidemic-related psycho-behavioral research should include more affective-loaded measures of risk.