When vulnerability does not predict mortality: The uses and limits of heat vulnerability indices

Key Words: Heat vulnerability index • Heat-related excess mortality • Extreme heat • Heat waves • Urban climate risk • Spatiotemporal analysis

The study, published in Environmental Research and co-authored by faculty member Elizabeth Dzwonczyk, evaluates the long-term validity of Heat Vulnerability Indices (HVIs) by examining a 30-year, census-tract-level dataset of Denver, Colorado, from 1990 to 2019 to test whether neighborhoods identified as highly vulnerable match those experiencing the highest heat-related excess mortality. Utilizing multiple validation frameworks—including statistical association, classification performance, and spatial correspondence—researchers compared HVI scores against observed and modeled mortality across both moderate and extreme heat waves. The findings revealed that alignment between spatial vulnerability and actual mortality is weak, inconsistent, and highly dependent on heat intensity and the specific decade, with high-vulnerability neighborhoods rarely overlapping with peak mortality areas during extreme heat events. Rather than invalidating vulnerability indexing, the authors emphasize a crucial distinction: static, place-based HVIs are effective for mapping baseline susceptibility and systemic inequity, but they should not be relied upon to predict event-specific spatial mortality. Ultimately, this distinction provides critical guidance for environmental health researchers and urban planners, offering a clearer framework for operationalizing HVIs in climate adaptation strategies.

Want to know more? Click here to read the full paper!