Indoor dust, immune activation, and household exposure risk after Hurricane Maria: a two-year pilot prospective cohort study in Puerto Rico

Abstract

Background: Hurricane Maria left many homes in Puerto Rico with prolonged flooding, structural damage, and indoor microbial growth, conditions that can drive respiratory illness and immune dysregulation. Objective: To assess the pro-inflammatory potential of post-flooding indoor dust and track two-year shifts in household risk profiles by integrating pro-inflammatory biomarkers, fungal exposure, and self-reported structural damage, mental health, and respiratory outcomes. Methods: Dust samples were collected from 50 households in a San Juan community one year and again from 35 of those same households two years after the hurricane. Dust sample extracts (DSEs) were prepared and incubated with pooled peripheral blood from volunteers residing outside Puerto Rico. Pro-inflammatory cytokines interleukin-1β (IL-1β), IL-6, and IL-8 were quantified using ELISA; airborne fungal abundance was assessed on malt extract and G25N agars. Survey data captured self-reported respiratory and mental-health scores using the SF-12v2 Health Survey, structural water damage, and remediation efforts. Principal component analysis (PCA) and k-means clustering were applied to identify household clusters, while Sankey plots were used to visualize the transitions of household risk. Results: IL-1β emerged as the most responsive cytokine, showing the greatest median reduction from Year 1 to Year 2 and across household water-damage categories. Three clusters emerged in Year 1, driven by structural damage, IL-1β, and fungal load. By Year 2, only two clusters remained, shaped more by self-reported respiratory and psychological distress than by structural damage. Sankey analysis revealed that some households transitioned into higher-risk profiles over time despite apparent household structural recovery. Impact: Indoor environmental hazards persist long after floodwaters in the aftermaths of hurricanes recede, yet few studies integrate biological, environmental, structural, psychosocial, and respiratory health data to track household risk over time. Using a human whole-blood pyrogen assay and multivariate clustering, this two-year study of Hurricane-María–affected homes in Puerto Rico shows that dust-induced IL-1β is a sensitive pro-inflammatory biomarker and that residual risk shifts from structural-microbial drivers to psychosocial-respiratory burdens during recovery. The multidomain “risk-profile” framework provides a reproducible template for post-disaster exposure assessment and can help public-health officials prioritize targeted remediation, health surveillance, and mental-health support as climate-related disasters increase.