Houston hurricane Harvey health (Houston-3H) study: assessment of allergic symptoms and stress after hurricane Harvey flooding

Abstract

Background: In August 2017, Hurricane Harvey caused unprecedented flooding across the greater Houston area. Given the potential for widespread flood-related exposures, including mold and sewage, and the emotional and mental toll caused by the flooding, we sought to evaluate the short- and long-term impact of flood-related exposures on the health of Houstonians. Our objectives were to assess the association of flood-related exposures with allergic symptoms and stress among Houston-area residents at two time points: within approximately 30 days (T1) and 12 months (T2) after Hurricane Harvey's landfall.

Methods: The Houston Hurricane Harvey Health (Houston-3H) Study enrolled a total of 347 unique participants from four sites across Harris County at two times: within approximately 1-month of Harvey (T1, n = 206) and approximately 12-months after Harvey (T2, n = 266), including 125 individuals who participated at both time points. Using a self-administered questionnaire, participants reported details on demographics, flood-related exposures, and health outcomes, including allergic symptoms and stress.

Results: The majority of participants reported hurricane-related flooding in their homes at T1 (79.1%) and T2 (87.2%) and experienced at least one allergic symptom after the hurricane (79.4% at T1 and 68.4% at T2). In general, flood-exposed individuals were at increased risk of upper respiratory tract allergic symptoms, reported at both the T1 and T2 time points, with exposures to dirty water and mold associated with increased risk of multiple allergic symptoms. The mean stress score of study participants at T1 was 8.0 ± 2.1 and at T2, 5.1 ± 3.2, on a 0-10 scale. Participants who experienced specific flood-related exposures reported higher stress scores when compared with their counterparts, especially 1 year after Harvey. Also, a supplementary paired-samples analysis showed that reports of wheezing, shortness of breath, and skin rash did not change between T1 and T2, though other conditions were less commonly reported at T2.

Conclusion: These initial Houston-3H findings demonstrate that flooding experiences that occurred as a consequence of Hurricane Harvey had lasting impacts on the health of Houstonians up to 1 year after the hurricane.

Keywords: Disaster epidemiology; Environmental exposure assessment; Extreme weather events; Flooding; Geographic information system; Hurricanes; Post-disaster rapid response research; Post-flooding environmental stressors; Post-flooding respiratory outcomes.

Fig. 1

Original study neighborhoods based on Federal Emergency Management Agency preliminary damage assessment, Harris County, Texas

Fig. 2

Location and concentration of Houston-3H Study participants in the Houston metro area, Harris County, Texas. ** We used the “Near” tool in ArcGIS Pro 2.1 to measure the Euclidean distance from each participant’s address to the nearest “original neighborhood” centroid (center). After visual inspection of address clustering around the centroid, a 7.5-mile radius was used as the cutoff for the “customized neighborhood boundary.” Most participants (n = 303) lived inside the customized neighborhood boundaries. Those who lived outside the customized neighborhoods were assembled into an “Others” group (n = 44)

Fig. 3

Logistic regression (Odds Ratio) of allergic symptoms on flooding exposures; time point 1 (N = 206) and time point 2 (N = 266). Note (i): Odds ratios are indicated by the size of the circle, and colors indicate level of significance after false discovery rate (FDR) correction. ¹ No exposure refers to preceding six exposures; subjects indicating “No Exposure” were not exposed to visible mold, dirty water, sewage, debris, fumes, or fuel/oil. ² No Clean-up Work refers to four previous work experiences; subjects indicating “No Clean-up Work” did not remove water, remove mud, perform tear out work, or perform a major repair

Fig. 4

Linear regression of stress level (0 to 10) on flooding exposures; time point 1 (N = 206) and time point 2 (N = 266). Note (i): Area Deprivation Index (ADI) is a composite measure of “neighborhood” socioeconomic disadvantage. Higher score represents increasing socioeconomic disadvantage. Note (ii): Linear regression results for all study participants (left), high ADI participants (middle), and low ADI participants (right). Note (iii): Colors indicate level of significance after false discovery rate (FDR) correction. ¹ No exposure refers to preceding six exposures; subjects indicating “No Exposure” were not exposed to visible mold, dirty water, sewage, debris, fumes, or fuel/oil. ² No Clean-up Work refers to four previous work experiences; subjects indicating “No Clean-up Work” did not remove water, remove mud, perform tear out work, or perform a major repair