Keke, Chukwudi, authorElf, Jessica L., advisorLee, Debbie, committee memberKoslovsky, Matt, committee memberRojas-Rueda, David, committee member2025-09-012027-08-252025https://hdl.handle.net/10217/241862https://doi.org/10.25675/3.02182Tobacco smoking remains one of the leading causes of death and disability worldwide. Global estimates indicate that there are approximately 1.3 billion smokers worldwide; tobacco smoking contributed to an estimated 7.7 million deaths and 200 million disability-adjusted life years (DALYs) in 2019, primarily through ischemic heart disease, chronic obstructive pulmonary disease (COPD), cancer, and stroke. Low and middle-income countries (LMICs), including sub-Saharan Africa, shoulder a disproportionate burden of the global tobacco epidemic, where over 80% of the world's tobacco users live. The implementation of evidence-based programs and policies have led to a substantial decline in tobacco smoking and tobacco-related diseases in many high-resource countries over the past three decades; however, tobacco use continues to rise in sub-Saharan Africa due to its rapidly shifting demographic profile, increasing influence of tobacco industry, and competing public health challenges such as HIV. In South Africa, a leading economy in sub-Saharan Africa, there are approximately 7.6 million people with HIV (PWH). Advances in anti-retroviral therapy (ART) have significantly improved the life expectancy of PWH, however, PWH remains at a greater risk for cardiovascular disease and pulmonary comorbidities, including tuberculosis (TB) and chronic obstructive pulmonary disease (COPD). These co-morbid conditions are exacerbated by tobacco smoking. Alarmingly, smoking prevalence is much higher among PWH, with estimates ranging from 24%–74% compared to 25% in the general population. Despite the high burden of tobacco use among PWH in South Africa, evidence is limited on effective smoking cessation strategies in this setting. South Africa has ratified the Framework Convention on Tobacco Control (FCTC) and has since implemented several anti-tobacco legislations, including advertising bans, tobacco product taxes, and restrictions on public smoking; however, recent surveillance data suggests that almost half of young men initiate tobacco smoking by the age of 20. Given that tobacco-related diseases can be avoided with cessation at younger ages, intervention during this time would help prevent the nearly 1 billion deaths expected from tobacco use in the 21st century. The work in this dissertation addresses the burden of tobacco use among vulnerable populations that experience greater tobacco-related health disparities by informing effective strategies or combinations of strategies to improve cessation efforts among PWH and adolescents in low-resource settings. In Aim 1, we examine the association between urine nicotine metabolite ratio (NMR) and response to smoking cessation treatment among people with HIV in South Africa, using data from a large randomized controlled trial for smoking cessation among Black South African smokers with HIV. In the primary trial, participants were randomly assigned to receive either behavioral counseling alone (BC) or BC plus combination nicotine replacement therapy (nicotine patches and gum, c-NRT). Bio-banked urine samples collected at baseline were used to determine cotinine (COT) and 3-hydroxycotinine (3HC) levels; we then calculated the NMR as a ratio of 3HC to COT. We evaluated NMR as a categorical variable (high versus normal metabolizers), using a cut-off of the fourth quartile (≥0.3174 ng/mL). Self-reported smoking abstinence at 6 months was biochemically verified using exhaled breath carbon monoxide (CO) and urine cotinine test. Modified Poisson regression models examined the association between NMR (high vs. normal) and 6-month abstinence. Effect modification (on the additive and multiplicative scale) examined the role of treatment arm as a potential modifier of the relationship between NMR and smoking abstinence at 6 months. Linear regression models were used to examine the relationship between NMR (high vs normal) and secondary outcomes, including changes in exhaled carbon monoxide breath test, urine cotinine, and nicotine withdrawal at 6 months. NMR was not associated with smoking abstinence (adjusted risk ratio (aRR) = 0.82; 95% CI: 0.45, 1.49; p = 0.53). No evidence of effect modification by treatment conditions was observed on the multiplicative scale (aRR =1.17; 95% CI: 0.32, 4.30; p = 0.81) or additive scale (adjusted relative excess risk due to interaction (aRERI) = 0.10; 95% CI: -1.16, 1.36; p = 0.44). There was no association with changes in nicotine withdrawal (β = 1.01, 95% CI: (-1.91, 3.93; p = 0.49), exhaled breath CO (β = -1.38, 95% CI: (-4.52, 1.76, 1.01; p = 0.39) or urine cotinine (OR = 0.98, 95% CI : (-1.18, 3.15; p = 0.37). The results in AIM 1 showed no associations between NMR and response to smoking cessation treatment among Black South Africans with HIV in South Africa. Overall, our results do not provide further support for the potential use of the NMR as a tool in selecting smoking cessation treatment among PWH in this setting. In AIM 2, we assessed the efficacy of a repeat treatment of BC with or without cNRT for smoking cessation among PWH with prior treatment failure in South Africa. We used data from the second phase of the randomized controlled trial for smoking cessation among PWH described in Aim 1. Participants in each arm who were still smoking at the 6-month follow-up visit were offered a second round of their group-assigned intervention. The primary outcome was self-reported smoking abstinence biochemically verified using exhaled breath CO and urine cotinine test at 6 months post-treatment. Secondary outcomes include smoking abstinence at 2 months post-retreatment and changes in exhaled breath CO at 2 months and 6 months post-retreatment. We assessed the proportions of participants who were abstinent at 6 months (and 2 months post-retreatment as a secondary outcome) based on the intent-to-treat analysis. Modified Poisson regression models were used to examine treatment effects on smoking abstinence and other secondary outcomes. Additionally, we assessed the association between treatment adherence and smoking abstinence at 6 months post-retreatment. At 6 months post-retreatment, a total of 35 participants (9%) were abstinent, including 15 (8%) participants in the BC group and 20 (11%) participants in the BC+cNRT group; p = 0.63. There was no significant difference in abstinence rates between treatment groups (RD = 3%; 95% CI: -9%, 3%; p = 0.40). This remained consistent after adjusting for potential confounders (aRR = 1.16, 95% CI: 0.39, 3.47; p = 0.79). We found no significant difference between the treatment groups in the median (IQR) changes in exhaled breath CO (BC: -1 (-6, 3) vs. BC+cNRT: -1 (-5, 4); p = 0.38; or urine cotinine (BC: -0.90 (-4, 0.4) vs. BC+cNRT: -1.4 (-4, 0.3); p = 0.42. At 2 months post-retreatment, 6 (3%) participants in the BC group were abstinent compared to 8 (4%) participants in the BC+cNRT group; p = 0.94. There was also no significant difference in abstinence rates between treatment groups (RD = 1%; 95% CI: -6%, 3%; p = 0.71) and after adjusting for potential confounders (aRR = 1.21, 95% CI: 0.42, 3.84; p = 0.79). We found no significant difference between the treatment groups in exhaled breath CO (BC: -2 (-7, 4) vs. BC+cNRT: -2 (-7, 2); p = 0.38) or urine cotinine (BC:-0.70 (-2.8, 0.9) vs. BC+cNRT: -0.9 (-4, 0.5); p = 0.52). Overall, the counseling sessions were moderately attended; there were no significant differences in median counseling sessions attended over the 6 months of retreatment between BC (4, IQR: 0, 5) and BC+cNRT groups (3, IQR: 0, 5) (p = 0.49). Self-reported patch use in the BC+cNRT group was low, with patch use reported by 64 (34%) participants at 2 weeks, 77 (41%) at 1 month, 70 (37%) at 2 months, and 68 (36%) at 3 months follow-up visit. We found no statistically significant associations between adherence and smoking abstinence at 6 months post-retreatment. The results in AIM 2 indicate that a moderate proportion of participants successfully quit following a second round of intervention. Abstinence rates in both treatment groups were low, with just 11% of participants in the BC+cNRT and 8% of participants abstinent at 6 months post-retreatment. Notably, adherence to the intervention was low. PWH in this setting who are unsuccessful after a first round of a smoking cessation intervention will likely benefit from repeat treatment, especially with attention to increasing adherence. In AIM 3, we used a qualitative and quantitative mixed-method approach to explore reasons for tobacco use maintenance, interest in quitting, barriers to cessation, and preferred quitting methods among adolescent smokers in South Africa. We conducted 6 focus groups (n = 38) to assess adolescents' opinions on tobacco maintenance and interest in quitting, as well as perceived barriers to and preferred methods for quitting tobacco. We then administered a structured questionnaire in a larger sample (n = 200) to collect information on socio-demographic characteristics, extent of tobacco use, previous quit attempts, interest in quitting, self-efficacy, and other substance use. Participants were adolescent smokers (aged 14-20) in South Africa. Focus group transcripts were analyzed using thematic analysis to identify major themes related to reasons for tobacco maintenance, interest in quitting, barriers, and preferred methods for quitting. Descriptive statistics were used to summarize participants' socio-demographic, tobacco, and substance use behavior. In the focus groups, we identified themes related to reasons for tobacco maintenance (i.e., dealing with stress, easy availability of cigarettes, household and peer smoking, personal wellbeing), interest in quitting (i.e., health concern, high cost of cigarettes, and social stigma), barriers to quitting (lack of social support from family and friends, co-use of substances and other tobacco products and social stigma) and preferred strategies to quitting (i.e., counseling, peer support). The survey data showed that most participants had low tobacco use dependence (n = 150; 75%) measured by the heaviness of smoking index (HSI). Participants reported moderate motivation to quit with a median motivation to quit score of 7 (IQR: 6, 8), and 122 (61%) participants reported a quit attempt in the past year, of which 92 (75%) tried quitting without any help. The result in AIM 3 suggests low levels of tobacco use dependence but moderate motivation to quit among adolescents in this population, indicating a critical window of opportunity to intervene before smoking behaviors become deeply entrenched in adulthood. Strategies to increase motivation to quit and peer counseling support should be considered when developing smoking cessation interventions for adolescents in this setting. Further, issues and needs that are specific to adolescents in low-resource settings, including mental health disorders and substance misuse, must be considered when developing smoking cessation programs for adolescents in this setting.born digitaldoctoral dissertationsengCopyright and other restrictions may apply. User is responsible for compliance with all applicable laws. For information about copyright law, please see https://libguides.colostate.edu/copyright.cessation interventionpeople living with HIVadolescentstobacco uselow resource settingsTextEmbargo expires: 08/25/2027.