Xiang-Yang Lou

Hidden Markov models (HMMs) are widely used to characterize latent state transition patterns in substance use. However, traditional HMM frameworks are incompetent when dealing with the complexities introduced by high-dimensional risk factors and varying time intervals, particularly in determining the number of hidden states and selecting variables for state transition parameters. To tackle the analytical challenges in the Population Assessment of Tobacco and Health (PATH) Study, a nationally representative longitudinal cohort study on tobacco use, we propose a continuous-time HMM framework with a regularization algorithm to identify multi-dimensional risk factors underlying complex poly-tobacco use transitions. We develop an elastic-net regularization on the transition covariates to identify informative covariates and improve model estimation accuracy. The inclusion of key covariates enables accurate determination of the number of hidden states. We incorporate survey weights and information on strata and clustering throughout the modeling framework. We demonstrate the validity of our approach in determining state numbers, identifying informative covariates, and estimating model parameters through a series of simulations. Application of the proposed approach to PATH data analysis revealed several demographic, behavioral, and psychosocial factors that contribute to the differential risks of transition between tobacco-use states among youth and young adults. The model’s capacity in identifying high-dimensional risk factors for underlying hidden variables substantiates its potential for enhancing public health research and informing interventions.

Background: Health-related stigma is widely acknowledged as a threat to public health and a barrier to managing chronic health conditions. Internalized stigma is a particularly strong predictor of poor health outcomes across health conditions, yet few evidence-based interventions are available. Peer support and counseling have been investigated as interventions for reducing internalized stigma. Typically, these interventions are developed and tested in disease-specific research, focusing on one health condition in isolation from others. This approach may limit knowledge and dissemination of support for health-related stigma across health conditions. Recent work has highlighted the need for research that breaks down traditional silos by using cross-cutting approaches to understand and reduce stigma. Objective: This study aimed to determine the feasibility and acceptability of a new group-based psychological intervention designed to reduce internalized health-related stigma among adults with different stigmatized chronic physical health conditions. Methods: A group intervention that was initially designed to address internalized weight stigma was adapted to be generalizable to other forms of internalized health-related stigma. This was done with input from Advisory Board members living with different stigmatized chronic health conditions and health professionals who specialized in these conditions. Adults with obesity, diabetes, HIV, skin diseases, chronic pain, or cancers were recruited to attend 12 weekly online group meetings. The average session attendance rate was computed with and without makeup sessions. A treatment acceptability questionnaire was completed at week 12. Primarily for feasibility testing, participants completed pre- and post-treatment questionnaires that assessed internalized health-related stigma and other relevant aspects of mental health and health-related quality of life. At baseline, participants were also asked to report reasons for perceived discrimination. Data collection occurred from December 2023 through April 2024. Results: In total, 10 adults were recruited within approximately 6 weeks, of whom 8 attended at least 1 treatment session and completed post-treatment questionnaires, with a 80% retention. The average session attendance rate was 95.8% with makeup sessions and 83.3% for those without makeup. Treatment acceptability ratings were high, with an overall acceptability rating of approximately 6.5 (SD 0.5) out of 7. Medium to large effect sizes were observed for changes in internalized stigma and some aspects of mental health. Almost all (n=7, 87.5%) of participants reported experiencing discrimination due to their health conditions, which accompanied a wide range of other reasons for perceived discrimination. Conclusions: Results showed high feasibility and acceptability of a transdiagnostic, online group psychological intervention for internalized health-related stigma delivered to adults with different types of stigmatized chronic physical health conditions. Given the small sample size and limited generalizability, testing in a large efficacy trial is needed to determine intervention benefits.

Many real-world underlying processes, such as addiction status and its longitudinal change associated with substance use, which are influenced by multifaceted factors, are not directly observable but can be captured by competent statistical methods. Hidden Markov models (HMMs) are a powerful tool for characterizing temporal dependence and mixture of latent variables in longitudinal data analysis , whereas the current HMM approach is largely limited in high-dimensional settings due to lack of dimension reduction and variable selection techniques. This article presents a novel approach tailored to address these limitations within the framework of HMMs. To reduce bias in parameter estimation by effectively discerning and excluding insignificant covariates, we propose a two-step algorithm that integrates HMMs with regularized generalized linear models (GLMs). Initially , we combine the coordinate descent algorithm with the EM algorithm to identify significant covariates. Subsequently, we recalibrate the HMM using the selected covariates employing maximum likelihood estimation for precise parameter refinement. Simulations demonstrate our method can accurately identify significant covariates and provide unbiased estimates for both the Markov transition and GLM parameters. The proposed method has improved performance over regularized GLMs, with a false negative rate up to 38.2% lower in variable selection. In two real data analyses, our method yields meaningful results that estimate the level of persistence of the unobserved states and are consistent with the relevant studies. Additionally, it effectively selects the set of important 1 covariates. Taken together, the regularized algorithm-based HMM outperforms existing methods, offering a valuable addition to the analytical armamentarium.

Introduction: Certain subpopulations in the United States are highly vulnerable to tobacco initiation and addiction, and elimination of disparities among those groups is crucial to reducing the burden of tobacco use. Aims and Methods: This study evaluated the racial and ethnic differences in smoking initiation of menthol flavored cigarettes and cigars among never-users, and in subsequent tobacco use among new users of menthol-flavored products, using longitudinal data from waves 1–4 of the Population Assessment of Tobacco and Health Study. The outcomes of interest were new use of menthol-flavored products, and subsequent past 30-day and past 12-month cigarette and cigar smoking, irrespective of flavors, after initiation. Results: The percentages of new users of menthol-flavored cigarettes and cigars at waves 2–4 were disproportionately higher in non-Hispanic black and Hispanic than in non-Hispanic white people. Adjusting for age and sex, black people who first used any menthol cigars had higher risk of past 30-day use of the same cigar category at the subsequent wave (adjusted risk ratio, aRR 1.48; 95% confidence interval [CI] 1.11 to 1.96) and past 12 months (aRR 1.74; 95% CI 1.55 to 2.63) compared to non-Hispanic white smokers. Black people who first used menthol-flavored cigarettes had marginally higher risk of subsequent past 30-day cigarette use (aRR 1.44; 95% CI 0.99 to 2.10) compared with their non-Hispanic white counterparts. Conclusions: This study shows that racial and ethnic differences exist in both initiation of menthol-flavored tobacco products and product-specific subsequent use after first using menthol-flavored products; black and Hispanic people have higher rates of initiation; black people also have higher rates of subsequent use. Implications: Use of menthol flavors in tobacco products is confirmed to be a contributor to large disparities in tobacco use; black and Hispanic people are more likely to maintain smoking through use of mentholated products than non-Hispanic white people. The findings suggest educational and regulatory actions on menthol-flavored tobacco products including restricting the selective marketing to vulnerable communities and banning characterizing flavors in cigarettes and cigars may reduce tobacco-related disparities and inform the Food And Drug Administration’s evidence-based rulemaking process.

Genome-wide association study (GWAS) is the most popular approach to dissecting complex traits in plants, humans, and animals. Numerous methods and tools have been proposed to discover the causal variants for GWAS data analysis. Among them, linear mixed models (LMMs) are widely used statistical methods for regulating confounding factors, including population structure, resulting in increased computational proficiency and statistical power in GWAS studies. Recently more attention has been paid to pleiotropy, multi-trait, gene–gene interaction, gene–environment interaction, and multi-locus methods with the growing availability of large-scale GWAS data and relevant phenotype samples. In this review, we have demonstrated all possible LMMs-based methods available in the literature for GWAS. We briefly discuss the different LMM methods, software packages, and available open-source applications in GWAS. Then, we include the advantages and weaknesses of the LMMs in GWAS. Finally, we discuss the future perspective and conclusion. The present review paper would be helpful to the researchers for selecting appropriate LMM models and methods quickly for GWAS data analysis and would benefit the scientific society.

Introduction: Pancreatic cancer (PC) is currently the third-leading cause of cancer deaths in the United States. African Americans with PC have an increased incidence and worse survival outcome when compared to other racial groups. During the COVID-19 pandemic, there is evidence that hospital resources were allocated to treating immediate life-threatening conditions. Some of the daily highest case numbers were reported in the state of Florida with several peaks throughout 2020 and 2021. Additionally, the state of Florida has the second-highest rate of new cases of PC within the United States with an incidence of 4860/100,000. Our specific aim is to define the impact of COVID-19 between race, age, income, and gender on the survival time of newly diagnosed patients with pancreatic cancer in Florida. Materials and Methods: Patients with pancreatic adenocarcinoma diagnosed from January 1st, 2017 to October 31st, 2020 were identified through the statewide clinical research and network database called OneFlorida Clinical Consortium by using the ICD10 diagnosis code for pancreatic cancer. Patients were then placed into 3 cohorts based on date of pancreatic cancer diagnosis: pre-pandemic (01/01/2017- 09/30/2019), transition (10/01/2019-02/28/2020), and pandemic (03/1/2020-10/31/2020). Patients with a diagnosis of neuroendocrine carcinoma were excluded. Patients were followed for at least one year unless a death occurred. Summary statistics were reported for demographic variables (age, sex, income, gender). Kaplan-Meier analysis with log-rank test was performed to compare the difference in overall survival time among groups. Results: This retrospective study had a total of 934 unique patients available for analysis. Of the 934 patients, 81.3% were in the pre-pandemic cohort (n= 759), 8.2% transition cohort (n=77), and 10.5% pandemic cohort (n=98). There was a decrease in the rate of diagnosis from the pre-pandemic (23 per month) to pandemic cohort (12.2 per month). The demographic distribution of the sample was 23.4% Black, 68.7% White and 7.9% Other. The median age was 67 years (27–89). There were 49.8% women and 50.2% men. The median income was $52,915 ($23,704–$124,821). The differences in overall survival time were not significant for age and gender across the 3 cohorts. Income <$53,000 had significantly lower survival time across the 3 cohorts. African Americans had significantly lower survival time for pre-pandemic and transition cohort (p< .005), but Caucasians had the lowest survival time for the pandemic cohort (p <.005). When stratified for stage, the mean survival (in months) for White vs. Black populations was 37.8 vs. 26.1 for stage I, 37.6 vs. 27.3 for stage II, 28.5 vs.18.77 for stage III, and 20.7 vs. 21.7 for stage IV. Discussion: This study demonstrated a decrease in diagnosis & survival rate during the COVID-19 pandemic in Florida. Dissemination of resources should target these disparities in income and race. Citation Format: Guettchina Telisnor, Alexander S. Lim, Zhongyue Zhang, XiangYang Lou, Ibrahim Nassour, Bo Han, Edward Agyare, Sherise C. Rogers. Impact of COVID-19 on pancreatic cancer outcomes in Florida [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer; 2022 Sep 13-16; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2022;82(22 Suppl):Abstract nr A048.

Background Group format weight loss interventions have benefits over individual format, but privacy concerns may limit their uptake. Method In this study, adults with obesity and interest in losing weight were recruited nationally online and randomly assigned to view one of eight videos describing a hypothetical, group behavioral weight loss intervention. Based on three fully crossed factors, the videos varied on privacy features of intervention (present or not); matching participants to group based on weight loss barriers (matched or not); and intervention format (online or in-person). Participants rated their willingness to join, privacy concerns, and perceived effectiveness of these interventions. They further reported preference for individual or group format interventions and reason for preferences. Results Description of privacy features, matching of participants, and format did not affect willingness to join, privacy concerns, or perceived effectiveness of the intervention. Privacy concerns were associated with lower willingness to join and lower perceived intervention effectiveness, and greater social anxiety and weight stigma. More participants preferred individual over group format (40.1% vs 33.9%; 26% selected neither) and preference for individual format was associated with greater privacy concerns. Conclusion Strategies to address privacy concerns in group-based interventions warrant further attention.