Exposure to tropical storms and hurricanes during pregnancy can influence neonatal and birth outcomes such as low birth weight and preterm birth. These outcomes are impacted by the disruption of healthcare and infrastructure, as well as stress, injury, and changes in nutrition. However, little is known about how far from the storm center maternal and neonatal impacts occur, nor how storms affect spatial patterns of maternal health. This study aims to assess spatial patterns and distance metrics for four birth outcomes (low birth weight, very low birth weight, preterm birth, and very preterm birth). The geospatial analysis included multiple buffers of 30, 60, and 100 kilometers and local spatial autocorrelation statistics. The results were predominately insignificant, with some key exceptions. The difference-in-difference analysis found a statistical association between hurricane exposure and preterm birth, with reductions post-storm. Across all three models, we found exposure to Hurricane Isabel and low birth weight was statistically significant at the 30 and 100-km spatial buffers. We also found preterm birth was also associated with exposure at the 30km buffer. We had also found significant differences in clustering before and after Isabel made landfall, with new clusters forming along the storm track. Our findings will provide a framework for emergency preparedness during tropical cyclones for mothers and their children.
Despite affecting up to 20% of women and being the leading cause of preventable deaths during the perinatal and postpartum period, maternal mental health conditions are chronically understudied. This study is the first to identify spatial trends of perinatal mental health conditions and relate these patterns to place-based social and environmental factors that drive cluster development. Among recorded hospital births from 2016 to 2019 in North Carolina, we analyze perinatal mood and anxiety disorders (PMAD), severe mental illness (SMI), and maternal mental disorders of pregnancy (MDP). Significant spatial clustering for all three outcomes was concentrated in smaller urban areas in the western, central piedmont, and coastal plains regions of the state. Mixed-effect linear regression modeling was used to examine the association of patient and community-level factors with elevated cluster risk. Results indicate that age, race, ethnicity, racial and socioeconomic segregation, urbanicity, access to healthcare services, insurance, food security, and access to greenspace have significant influences on cluster risk. These results provide important contextual and spatial information concerning at-risk populations for maternal mental health conditions within the state of North Carolina and can better inform targeted public health interventions and drive future research analysis
In the last few decades, wildfires have increased in frequency. Such increased frequencies can be attributed to climate change, which has been a major contributor to several natural disasters in the recent past. Increased wildfires potentially lead to significant health hazards in humans, particularly in vulnerable groups including pregnant women and the elderly. It is important to study the human health burden of wildfires primarily because of two reasons. Firstly, climate change will continue to have its impacts in the coming future, and hence the frequency of wildfires is expected to increase. Secondly, particulate matter from wildfires is more toxic compared to ambient PM2.5. Hence, we aim to forecast the long-term and short-term effects of wildfire on human health. In that direction, it is necessary to be able to forecast particulate matter emissions (E), which in turn depends on the forecasted burned area (BA). We have adopted a two-dimensional cellular automata model for modeling wildfire burned area, where a region with a wildfire is considered as a square lattice with regular grids, each of which is in one of the following three states: Unburned, Ablaze and Burned. The parameters of the model take into account the effects of covariates, including weather conditions, wind speed and direction, topology, etc. The model is trained on historical wildfire data to estimate parameters, which can then be used for forecasting future burned area. We demonstrate the empirical performance of our model via a simulation study, and show its applications to historical wildfire data. We plan to combine the described model with a suitable statistical regression model of the form E= f(BA) for forecasting future emissions.
Human health is intimately linked to the state of the environment. Air pollution continues to rise at an alarming rate, wreaking havoc on human health. Currently, for individuals to obtain life-saving insights, they must travel to locations of specialized equipment like hospitals or buy expensive health equipment. There is a dire unmet social need for individuals to have the ability to easily, accurately, and holistically measure, monitor, and detect changes in health conditions in a given environment. “EnviroSenseHealthMonitor” is a one-stop, cost-effective, easy-to-use solution that includes a device and mobile application for people to get healthier in their own living environment.
This project lies at the intersection of agricultural waste management, sustainability, and medicine. It delves into innovation in medicine and biotechnology through sustainability by integrating bioprocessing engineering focused on creating a magnetic core-shell nanoprobe derived from biochar used for novel applications in various medical and biological systems, including targeted drug delivery and tumor ablation by stepping into sustainability in oncology. The nanoprobe consists of a highly magnetic core material made from biochar, coated with a thin shell of tissue-compatible material.
Particulate matter 2.5 is associated with increased morbidity and mortality through its impacts on the cardiovascular and respiratory systems. Understanding air quality data allows communities to make informed decisions and advocate for change. This study provides a review of different approaches to air quality analysis and risk communication that are well-suited for low-cost monitoring networks. This work was done in partnership with Robeson County, a community with a monitoring network that has been sharing raw data with its members. Raw data can be overwhelming and difficult to interpret for average citizens. The goal of this work is to provide communities with a starting point for interpreting air quality data by determining methods of risk communication that are informative, effective, and feasible for low-cost air quality monitoring networks. The review identified methods from university community partnerships, government organizations in several countries, and public health and environmental organizations. All methods were evaluated using a decision matrix with the following 7 categories which were ranked by descending importance: 1) comprehensibility, 2) applicability, 3) data accessibility, 4) connection to health outcomes, 5) communication method, 6) data time period, and 7) frequency. For methods that were deemed a good fit draft versions were created. Line graphs and pie charts with corresponding air quality index colors were deemed to be the most accessible and effective method for preliminary communication. Certain programs, such as AirQ+, were found to be feasible and effective to estimate health impacts of PM2.5 for communication. Moving forward we will collect and share real time air quality data in Robeson County using these methods.
Malaria is a life-threatening vector-borne disease responsible for over 200 million cases and 600,000 deaths in 2020. While malaria-related deaths have steadily decreased by over 40 percent from 2000 to 2020, there are still regions where malaria-related deaths and burdens continue to pose a significant health threat. The Amazon River Basin is the largest source of malaria in the Americas. In the northwest region of Loreto, in the Peruvian Amazon, malaria has the highest incidence and mortality rate in Peru. Despite concerted efforts from the Project for Malaria Control in Andean Border Areas (PAMAFRO), malaria cases have increased in Loreto since the program ended in 2011. Local malaria transmission and control is determined and complicated by various factors, including human migration, limited healthcare access, remote surveillance challenges, and land-use change. Increased malaria incidence and mosquito breeding site proliferation has been connected to deforestation and human migration from several extractive industries, including logging, gold mining, and hydrocarbons. There is still limited literature on how concessionary land activities relate to malaria incidence. This study aims to understand the impact of extractive land concessions from mining, logging, and hydrocarbons on the malaria incidence rate in Loreto, Peru, from 2015 to 2020. Secondary data on land concessions and malaria was collected from the Peruvian Ministry of Energy and Mines (MINEM); the Geological, Mining, and Metallurgical Institute (INGEMMET); the Global Forest Watch (GFW); the Agency for Supervision of Forest Resources and Wildlife (OSINFOR); Peru’s National Institute of Statistics and Informatics (INEI); and Loreto Directorate of Health for longitudinal analysis in ArcGIS and R. The main objectives of this research are to: (1) investigate outbreak patterns across active, inactive, and non-concessionary sites in Loreto, Peru, and (2) determine the populations and communities at highest risk of malaria exposure in concessionary areas.
Flood events do not impact communities equally and often expose the legacy of public policies, practices, and plans that result in certain communities experiencing more damage after flood events. At the same time, extensive research has shown environmental disamenities (also referred to as Locally Unwanted Land Uses) are often located in Black, Indigenous, People of Color (BIPOC), and low-income communities compared to white and wealthier communities. Outside these areas of affluence, BIPOC communities are often exposed to a variety of hazards, whether that be flooding, polluting industries, or a mixture of natural and human-made hazards. Differences in impacts and recovery from disasters can widen societal inequalities. Communities with high social vulnerability (as defined by income, race, age, and other characteristics) typically have a slower recovery trajectory than those with low social vulnerability. Little is known about how communities are impacted by the interaction between environmental hazards (flooding) and human-made hazards (living near polluting industries). In this project, I examine the impact of flooding from Hurricane Matthew (2016) and Hurricane Florence (2018) on four different types of environmental disamenities (hog concentrated feeding operations, solid waste facilities, hazardous waste facilities, and toxic release inventory sites) in 22 southeastern North Carolina counties and compare the level of social vulnerability of communities with flooded disamenities. The study area is ideal for this analysis because of its extensive flood risk and numerous unwanted land uses, with hog-concentrated animal feeding operations (CAFOs) as the most prominent. Through geospatial analysis of disamenities and using social vulnerability and environmental justice as a guiding theoretical framework, I find that 61% (n=1,455) of environmental disamenities outside of designated floodplains, experienced flooding post-Matthew and post-Florence, and most are located in census block groups with high social vulnerability. These findings show how historically marginalized communities are not only disproportionately impacted by natural hazard events, but also by polluting industries which pose significant health and environmental risks after flooding occurs. Unpacking and addressing these disparities is essential to ensure everyone can live, work, and play in safe and healthy spaces regardless of demographic characteristics or geography.