Objectives: There are currently no military studies on non-benzodiazepine sedative hypnotic (n-BSH) use among combat injured active-duty service members. There were three objectives of this dissertation: (1) describe rates and patterns of n-BSH medications in this population; (2) examine how n-BSH use affects perceived Health Related Quality of Life (HRQOL); and (3) examine the role of traumatic brain injury (TBI) in n-BSH prescription patterns among these injured service members. Methods: Three population-based retrospective studies were conducted on all U.S. service members who sustained injuries during deployed operations between September 2001 and Aug 2019 and who received pharmacological treatment for insomnia. These studies utilized six data sources from the Department of Defense (DoD); the Expeditionary Medical Encounter Database (EMED), Military Health System Data Repository (MDR), Pharmacy Data Transaction System (PDTS), Defense Manpower Data Center (DMDC), Career History Archival Medical and Personnel System (CHAMPS), and the Wounded Warrior Recovery Project (WWRP). DMDC and EMED contains deployment and combat information for Operations Iraqi Freedom and Enduring Freedom. Classification of injury, TBI status, n-BSH prescription patterns, and clinical diagnoses (assessed via ICD-9 and ICD-10 codes) were abstracted from the MDR and PDTS. Mental and physical wellbeing questionnaires from the WWRP were utilized to assess health related quality of life (HRQOL). Results: n-BSHs were the most frequently prescribed medication class among service members who sustained combat injuries. Long-term prescriptions were largely driving the high prevalence rates of n-BSH, given that 42.6% of service members were prescribed n-BSHs for 6 months or longer. Additionally, they experienced high rates of post-traumatic stress disorder (50.6%), TBI (42.5%), depression (39.1%), anxiety (52.0%), and chronic pain (15.1%). After adjustment, there was no statistically significant relationship between n-BSH prescription characteristics with service members’ HRQOL, or between TBI and n-BSH prescription characteristics. Conclusions: Combat injured service members’ n-BSH prescription patterns exceed Food and Drug Administration and American College of Physician guidelines. While n-BSH patterns did not differ by TBI or HRQOL, the high prevalence of long-term n-BSHs is alarming. Given their high prevalence of comorbid conditions in this population, clinical management of these service members should include a multidisciplinary approach.

The COVID-19 epidemic has highlighted a number of important challenges for infectious disease Epidemiologic research: 1) scaling causal-inference efforts across the human disease phenome; 2) understanding the long-term consequences of a novel disease without robust longitudinal data; and, 3) leveraging non-traditional data types for infectious disease research. Our dissertation provides three examples of advanced Epidemiologic methods that illustrate how researchers may address one or more of these challenges.

Given the prevalence of multiple comorbidities, interrelated disease states may represent a more complete picture of COVID-19 infection severity risk compared to a disease-by-disease approach. We used a bias-adjusted, three-step Latent Class Analysis (LCA) method to identify patterns of comorbidities from 31 disease indicators; and, measured their relationship to severe COVID-19 infection among 176,894 participants in the UK Biobank cohort. We identified 5 distinct comorbidity patterns from 31 disease indicators, assessed using clinical diagnosis records from UK Biobank’s comprehensive EHR data linkage between 2015-2019. Our results identified significantly increased risk for severe COVID-19 infection, with substantial heterogeneity in effect sizes, for each of our 4 comorbidity latent classes compared to our ‘Healthy’ latent class.

We investigated the associations between genetic liability to severe COVID-19 infection, measured with Polygenic Risk Scores (PRS), and 31 comorbidity phenotypes derived from linked electronic health record (EHR) data in the past 20 years. PRS for very severe COVID-19 infection were associated with increased risk for uncomplicated diabetes, uncomplicated hypertension, obesity, and renal failure. Our research indicates that the same genetic composition that increases an individual’s risk for COVID-19 may also influence their risk for other important comorbid diseases.

Proximity to military bases may be an indicator of accessibility to military sexual partners; and, help identify important local HIV epidemics. We estimated the relationship between travel time to the nearest military base and HIV-status among 7,514 young women recruited at local venues. Our study found that adolescent girls and young women that meet or congregate near military bases were at a slightly elevated risk for HIV-infection in the combined sample, but only in 1 of our 4 military bases in stratified analysis.

This dissertation explores some of the ways that optimal control can be used to mitigate the effects of parametric uncertainty on the control's ability to achieve a desired endpoint condition on the state variables. Standard optimal control solutions rely on precise knowledge of parameter values which are difficult to measure in many practical applications. The result is that, when implementing the designed controls, there is likely to be some error between the desired state trajectories and the actual system outputs. Usually these errors are managed using feedback control using sensor measurements to correct the state deviations on the fly. However, recently there has been promising work done for generating controls which operate in the open-loop for generating state trajectories which are inherently robust to uncertainty in the parameters.

One avenue of applying optimal control for managing parameter uncertainty is unscented optimal control which borrows the low sample discretization of the Unscented Transform for approximating a Riemann-Stieltjes integral of an objective function. This method has proven effective for generating open-loop controls. In this dissertation we explore a family of sensitivity function based optimal control problems, which utilizes an approximation of the response to parameter deviations through a Taylor series, which can be used to likewise generate open-loop controls. We then use these sensitivity based problems as a lens for exploring the unscented control problem. Furthermore, we identify conditions for equivalence between the two optimal control problems. These inherently robust open-loop trajectories represent a single degree of freedom control.

In addition to generating the open-loop controls for a more robust state trajectory, this dissertation provides a problem formulation for generating a set of time-varying feedback gains. These gains use sensor measurement for both the sensitivity function based problem as well as the unscented problem frameworks. The unscented problem avoids linearizing the error dynamics which is shown to be advantageous for designing the feedback gains. The use of feedback gains represents a second degree of freedom when combined with a more robust open-loop control significantly improving the performance.

These concepts are demonstrated on a simulated nonlinear double gimbal mechanism with flexible effects where the unscented feedback control gives a nearly $100\%$ decrease in variance of the error of the trajectory endpoints and $82\%$ in the case of the purely open-loop. The mathematical model of the double gimbal as a second order set of differential equations represents a prevalent mechanical system in engineering. The controls generated by the neighboring unscented problem demonstrate a significant improvement in the robustness to the parameter uncertainty over the standard controls. The utility of these concepts is further illustrated experimentally using a nonlinear two link robotic arm with flexible joints.

BACKGROUND. The Global AIDS Strategy 2021-2026 called for more focus on the ‘missing’ men in areas with HIV to address the gender gap in testing and treatment. Men in the militaries of sub-Saharan Africa are a unique sub-population due to their culture and work environment and a priority for increasing HIV status awareness. METHODS. Data on active-duty men ages 18-49 years were combined from 20 Seroprevalence and Behavioral Epidemiology Risk Surveys conducted among militaries in sub-Saharan Africa from 2013-2020. Meta-analyses with sub-groups were used to compare HIV prevalence, HIV knowledge, condom use, and discriminatory attitudes with the general population. Associations of HIV knowledge, negative attitudes, perceived discrimination, and use of military HIV testing sites with HIV status awareness were determined using meta-regression analyses. Among men at higher-risk of acquiring HIV, imputation-based modeling was used to estimate the effect of changing identified associations on HIV status awareness at the population level. RESULTS. In Western/Central Africa, HIV prevalence and knowledge were higher in the military compared to the general population (2.7%vs1.3%). HIV status awareness was lower among men without correct HIV transmission knowledge, high perceived discrimination, and who didn’t use military testing facilities. Among high-risk men, the estimated difference in status awareness was +6.3% if 100% had correct HIV transmission knowledge, +3.8% if 100% had low perceived discrimination, and +2.7% if 100% used military testing. In Eastern/Southern Africa, HIV knowledge was higher in the military compared to the general population. Among high-risk men, only use of military testing was associated with status awareness with an estimated difference in awareness of +6.3% if 100% used military testing. CONCLUSIONS. The exposures related to military services may be putting men in the military in Western/Central Africa at increased risk of HIV. Identifying men living with HIV and getting them connected to treatment is vital for prevention of morbidity/mortality and transmission. Increasing HIV knowledge and use of military testing sites as well as decreasing perceived discrimination could increase overall status awareness among men at high-risk of acquiring HIV. Current military HIV programs should focus on proven interventions in these areas for HIV prevention and to increase status awareness.

Background: Transmission of avian mycobacteriosis is generally considered a contagious process, but is not well understood and environmental sources may be important. The large, dynamic population of birds at San Diego Zoo Global (SDZG) with complete population ascertainment over a 22-year period offered an opportunity to use social network analysis to understand disease epidemiology and test for patterns of contagion.

Objectives: Study one evaluated the social network structure of birds for evidence of a contagious process. Study two examined patterns of genetic similarities using whole genome sequencing (WGS) along pathways of network connectivity. Study three examined whether network connectivity predicts future disease.

Methods: Study one identified cases of mycobacteriosis and constructed a social network from enclosure histories (n=16,430) in the SDZG population. Stratification of network edges by spatial and temporal characteristics tested for contagion and other drivers of disease in directly- and indirectly-connected birds. Study two characterized mycobacteria isolated from 124/275 cases. For the subset with WGS (n=97), the probability of having similar genotypes given connectivity was estimated and significance determined from random permutation tests. Study three used longitudinal, mixed-effects logistic regression to evaluate the association between network exposure and mycobacteriosis development.

Results: Study one: Disease clustered significantly among directly- and indirectly-connected birds. The RR of disease given exposure to 2° contacts never housed in the same enclosure was 1.31 (p=0.004), providing strong evidence that a contagious process is present, because the association persisted when common environmental exposure was removed. Study two: Mycobacterium avium avium (MAA) and M. genavense were the most common species. The WGS showed genotypes of MAA were significantly related along paths of network connectivity; however, no significant patterns were identified for M. genavense. Study three: Results showed significant associations between direct (OR=2.15) and indirect (OR=1.56) exposure to positive birds (compared to no exposure) and mycobacteriosis. Risk-stratified models provided estimates with further characterization of exposure; not all findings were robust to model variation.

Conclusion: Social network analysis was a powerful method for evaluating complex contact patterns and mycobacteriosis. The data strongly support a contagious process, show low transmissibility, and provide new information on disease epidemiology.

Background: Posttraumatic stress disorder (PTSD) is a potential consequence of exposure to traumatic stress. PTSD has severe psychological, physical, interpersonal, and societal costs. The development of PTSD following traumatic stress is heritable. Studies of genetic risk factors have focused on common single nucleotide polymorphisms. The overwhelming majority of risk PTSD loci have yet to be identified. The integration of detailed PTSD phenotyping has shown promise to increase power to identify PTSD risk loci. Rare genetic variation has only been sparsely examined in the context of PTSD, yet evidence is emerging that rare variation such as copy number variation (CNV) is relevant to psychiatric disorders. Findings of large genetic studies of PTSD and co-morbid disorders make it possible to make causal inferences and provide mechanistic insights using Mendelian Randomization (MR).

Methods: This dissertation includes three studies, all of which were conducted among participants from the Psychiatric Genomics Consortium-PTSD data collection, a consortia made to investigate genomic risk factors for PTSD through meta-analysis of PTSD cohorts. Study 1 was a GWAS of PTSD symptom scores and lifetime trauma exposure phenotypes to identify common genetic risk variation for PTSD. Study 2 was an investigation of rare CNV and PTSD. Study 3 leverages GWAS summary statistics from PTSD and inflammatory phenotypes related to PTSD, using MR to make causal inferences about PTSD and inflammatory diseases.

Results: In study one, multiple genetic risk loci were identified for PTSD and for lifetime trauma exposure, with the two traits having a substantial degree of genetic overlap. In study two, PTSD risk was elevated in CNVs that crossed over known neurodevelopmental CNV regions and pathways related to the function of the nervous system and brain. In study three, PTSD had evidence of causal effect on asthma and psoriasis, as well as inflammatory biomarkers.

Conclusion: This dissertation enhances the general field of PTSD genetics, having identified novel common and rare risk variation, and supports hypotheses that PTSD has a causal relationship with certain comorbid diseases with an inflammatory component.

Background

Methods

Results

Conclusions