Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Epidemiology and Population Health

Degree Program

Public Health Sciences with a specialization in Epidemiology, PhD

Committee Chair

DuPre, Natalie

Committee Co-Chair (if applicable)

Baumgartner, Rick

Committee Member

Baumgartner, Rick

Committee Member

Taylor, Kira

Committee Member

Little, Bert

Committee Member

Peiper, Nick

Author's Keywords

COVID-19; opioid epidemic; interrupted time series; mortality; overdose; urbanicity


Background: 585,000 people have died from an opioid overdose in the US between 1999 and 2020. The current opioid epidemic has been described as a quadruple wave of overdoses, due in part to the changes in prescription and illicit opioid supply as well as the underlying social and structural factors that led to a subsequent increase in demand. In order to mitigate the opioid epidemic through prevention and protection strategies, we must first understand the social and structural factors that are driving the increase in opioid misuse and abuse. To better understand the socioeconomic factors, we described socioeconomic profiles of US counties and examined their associations with rates of opioid overdose mortality in Aim 1. Since the beginning of the epidemic, rates of opioid-related overdose death have differed in rural and urban counties. We examined the association between urban residence and subsequent opioid overdose mortality in Kentucky, a state highly impacted by the opioid epidemic, and whether this association was modified by the COVID-19 pandemic (Aim 2). With the rise in opioid overdose deaths, people have sought out alternative substances that are advertised to have less side effects and lower abuse potential, such as kratom. Kratom is an herbal extract from evergreen tree leaves indigenous to Southeast Asia that has opiate-like properties. Six states have banned kratom over concerns about its potential for addiction; however, there is no scientific evidence regarding the impact of these laws on the opioid epidemic. Therefore, we examined this association between state-level kratom legislation and opioid overdose mortality across US states (Aim 3). Methods: In all analyses, opioid overdose mortality was classified using the International Statistical Classification of Diseases, 10th revision (ICD-10). Among deaths with drug overdose as the underlying cause, we captured those specifically involving an opioid analgesic including opium, heroin, prescription opioids (i.e., natural and semisynthetic opioids and methadone), synthetic opioids other than methadone, and other and unspecified narcotics. In a nationwide analysis (Aim 1), we identified patterns of demographic, socioeconomic and housing characteristics in US counties using principal components (PC) analysis and used Poisson regression to estimate adjusted relative risks (RRs) and 95% confidence intervals (CIs) of opioid overdose mortality for a one standard deviation increase in PC scores. We used data from all Kentucky inpatient and outpatient hospitalizations from 2016-2020 to estimate odds ratios (ORs) and 95% CIs of opioid overdose mortality for urban versus rural patients with multivariable logistic regression (Aim 2). Analyses were conducted separately for hospitalizations occurring before the pandemic and during the pandemic (April 2020 or later). Lastly, in Aim 3, we used an interrupted time series (ITS) analysis to estimate the association between state-level legislation banning kratom and opioid overdose mortality in the US. Results: We found that counties with an extremely disadvantaged socioeconomic profile had the highest rates of prescription opioid mortality (RR=1.17; 95% CI=1.13,1.21), but had inverse associations with illicit opioid mortality (RR=0.93 and 95% CI=0.88, 0.95). Counties with a high percentage of single individuals, high poverty rates and high education had the highest mortality rates from illicit opioids (RR=1.46; 95% CI=1.40, 1.53). Before the pandemic, Kentucky patients living in urban counties had 63% higher odds of opioid overdose death (OR=1.63; 95% CI=1.34, 1.97); however, during the COVID-19 pandemic, patients in urban and rural counties became more similar in regard to opioid overdose mortality (OR=0.72; 95% CI=0.45, 1.16; p-value for interaction =0.02). In regard to state-level legislation banning kratom and opioid overdose mortality, we found that a law banning kratom increased opioid overdose mortality in Indiana (RR=1.09; 95% CI=1.06, 1.12), Vermont (RR=1.11; 95% CI=1.05, 1.17), Wisconsin (RR=1.03; 95% CI=1.00, 1.05), Arkansas (RR=1.10; 95% CI=1.04, 1.15), and Alabama (RR=1.05; 95% CI=1.01, 1.09), but not in Rhode Island (RR=1.05; 95% CI=0.97, 1.13). Conclusions: In summary, we found that socioeconomic profiles were significantly associated with opioid overdose mortality rates in the US, and these profiles differed depending on the opioids involved. Though prescription and illicit opioid overdoses are closely intertwined, it is important to differentiate the deaths and examine them as distinct epidemics in order to craft more appropriate prevention and response efforts. In regard to urbanicity, we found that before the pandemic, living in urban counties was associated with higher opioid overdose mortality in Kentucky. However, urban and rural differences in opioid overdose mortality washed away during the pandemic, suggesting that the protective mechanisms that were preventing rural overdose deaths prior to the pandemic were no longer preventing rural deaths when the pandemic hit in 2020. Lastly, we observed that public health policy banning kratom that was put into place in an effort to thwart the addictive and dangerous properties of kratom, did not have a significant impact on the opioid epidemic.