Association of respiratory diseases with humidifier disinfectants exposure and its latency: A study of health damage reporter cohort in South Korea

Following widespread use of humidifier disinfectants (HDs) in South Korea from the late 1990s until their recall in 2011, clusters of respiratory illnesses emerged. While causality has been established for severe cases, evidence linking non-specific or chronic respiratory diseases to HD exposure remains insufficient for recognition or compensation. We aimed to assess the relationship between HD exposure and a broad spectrum of respiratory diseases and determine the latency of disease onset. We linked national health insurance data to self-reported exposure records for 6061 individuals who filed compensation claims for HD-related health damage. We conducted individual-level interrupted time series (ITS) analysis over a 15-year period (60 months before and 120 months after first HD use). We evaluated eight respiratory diseases and calculated excess rate ratios (ERRs) and latency patterns by disease type, severity, and HD chemical ingredients. Significantly elevated risks were observed for all eight diseases post-HD exposure. The strongest associations were observed for pneumonia (ERR=14.04), asthma (ERR=8.54), and interstitial lung disease (ERR=8.26). Most disease onset was typically concentrated within the first few years post-exposure but extended up to a decade for chronic conditions. ERRs stratified by chemical type revealed distinct ingredient-specific risk profiles. Despite potential collider bias, comparative analyses using national population data showed consistency in relative risk estimates. This study provides robust epidemiologic evidence for the long-term health consequences of HD exposure. The findings support revisions in compensation policies and underscore the need for extended health surveillance of affected individuals. However, the results should be interpreted with caution due to potential collider bias and possible overrepresentation of affected individuals.