In 2017, unprecedented rainfall from Hurricane Harvey resulted in severe flooding in parts of Texas. While the association between the immediate effects of severe weather on mortality is well established, less is known about long-term mortality and how risk may differ across vulnerable populations, especially those reliant on health services. This study examines the association between severe weather exposure from Hurricane Harvey and mortality among vulnerable populations of older adults.

This retrospective cohort study used Medicare administrative claims data from beneficiaries aged 65+ who resided in Texas or Louisiana the year before and after Hurricane Harvey. Historic weather data was used to construct a 4-day measure of ZIP code tabulation area cumulative rainfall and identify locales exposed to high rainfall (above 75mm). Vulnerable older adult populations were identified. Attributable risk and relative risks (RRs) were estimated to quantify the impact of high rain exposure on mortality. Time-to-event analyses estimated associations between exposure to high rain and 12-month mortality.

The study included 1,730,693 beneficiaries; 535,794 (37.5%) were exposed to high rain in Texas, and 264,265 (87.5%) in Louisiana. Beneficiaries with Alzheimer’s disease and related dementias (ADRD) exposed to high rain had a 5% higher risk of mortality in fully adjusted models (HR=1.05, 95% CI: 1.02, 1.07) and 1,192 attributable deaths. High rain resulting from Hurricane Harvey was associated with a 13% higher risk of mortality among beneficiaries with end-stage renal disease (ESRD) in fully adjusted models (HR=1.13, 95% CI: 1.04, 1.23) and 85 attributable deaths.

The risk of mortality from exposure to severe weather differs based on the type of vulnerability. Those with an ADRD or ESRD diagnosis, in particular, had elevated mortality rates. High rain exposure had a significantly larger mortality impact on those of lower socioeconomic status. Emergency planning should pay particular attention to these highly vulnerable groups.

Determine how characteristics of deployment mild traumatic brain injury (TBI) and blast exposure influence the relationship between the functional brain connectome with cognitive outcomes and symptom severity.

N = 181 Iraq and Afghanistan combat veterans completed structured clinical interviews, cognitive testing, self-report questionnaires, and magnetoencephalography (MEG). MEG data were acquired in the resting-state with eyes open. MEG data were beamformed to identify brain regions active at rest. Functional brain connectomes representing the unique network present for a given individual were created using active brain regions identified for each participant. Network metrics describing these connectomes were calculated at the participant level. Cognitive tests included the WAIS-IV, Trail Making Test Parts A&B, and the Controlled Oral Word Association test. Due to differences in normative data across tests, raw scores were used in analyses. Symptom measures included the PTSD Checklist - 5 (PCL-5), Patient Health Questionnaire (PHQ-9), Neurobehavioral Symptoms Inventory (NSI), Quality of Life After Brain Injury (QOLIBRI), Pittsburgh Sleep Quality Index (PsQi), the Distress Tolerance Scale (DTS), and the PROMIS Pain Interference Scale (PROMIS-PI).

Hierarchical linear regression analyses revealed that several network metrics were significantly related to both cognitive outcomes and symptom severity after adjusting for demographic covariates and clinical characteristics.

The relationship between Global Efficiency (GE) and cognitive outcomes was moderated by deployment TBI on the WAIS-IV Full Scale Index (FSI), Perceptual Reasoning Index (PRI), and General Ability Index (GAI). In all cases, when deployment TBI was absent, greater GE was associated with poorer cognitive scores. The relationship between GE and symptom severity was moderated by the severity of blast exposure. Greater GE was associated with lower symptom severity at lower blast severities for the PHQ-9 and QOLIBRI A (thinking) and E (negative emotions). Moderation effects were also observed for the PSQI. In the absence of deployment TBI, greater GE was associated with better sleep quality; however, in the presence of deployment TBI, greater GE was associated with poorer sleep quality. Other connectome-outcome relationships were not consistently moderated by Deployment TBI or blast history

Results demonstrated relationships between several aspects the functional connectome of the brain with both cognitive outcomes and symptom severity beyond effects of common demographic and clinical variables. Moderation analyses revealed that the relationship between GE of the connectome and outcomes is frequently disrupted by deployment TBI and blast. GE is a measure of the ease of information transfer through the network. These results identified consistent relationships between GE and outcomes in the absence of deployment TBI or blast, but these relationships disappear when deployment TBI or blast are present. Participants in this study were on average 11 years post-TBI or blast exposure, suggesting these are chronic rather than acute effects. GE was significantly correlated with most symptom severity measures as well as the WAIS-IV PRI, GAI, VCI, and FSI. Future efforts to normalize the relationship between GE and outcomes following TBI may improve rehabilitation outcomes and directly affect areas of concern commonly reported by service members following TBI or blast exposure.