Faculty Bibliography

The complex nature of microbiome data has made the differential abundance analysis challenging. Microbiome abundance counts are often skewed to the right and heteroscedastic (also known as overdispersion), potentially leading to incorrect inferences if not properly addressed. In this paper, we propose a simple yet effective framework to tackle the challenges by integrating Poisson (log-linear) regression with standard error estimation through the Bootstrap method and Sandwich robust estimation. Such standard error estimates are accurate and yield satisfactory inference even if the distributional assumption or the variance structure is incorrect. Our approach is validated through extensive simulation studies, demonstrating its effectiveness in addressing overdispersion and improving inference accuracy. Additionally, we apply our approach to two real datasets collected from the human gut and vagina, respectively, demonstrating the wide applicability of our methods. The results highlight the efficacy of our covariance estimators in addressing the challenges of microbiome data analysis. The corresponding software implementation is publicly available at https://github.com/yimshi/robustestimates.

PURPOSE/UNASSIGNED:The purpose of this study is to describe diabetes distress and related factors among Chinese Americans with type 2 diabetes in New York City (NYC). METHODS/UNASSIGNED:We conducted a secondary data analysis of the baseline data from three research studies conducted among community-dwelling Chinese American adults with type 2 diabetes. Diabetes Distress Scale (DDS) was used to measure sources of diabetes distress including emotional-, regimen-, interpersonal-, and physician-related distress. A score of 2 or greater indicates moderate diabetes distress or higher. Patient Health Questionnaire-2 (PHQ-2) was used to measure depressive symptoms. Participants' sociodemographic information was also collected. Descriptive statistics were used to describe diabetes distress, and logistic least absolute shrinkage and selection operator (LASSO) regression was used to examine factors associated with diabetes distress level. RESULTS/UNASSIGNED:Data from 178 participants (mean age 63.55±13.56 years) were analyzed. Most participants were married (76.40%), had a high school degree or less (65.73%), had a household annual income < $25,000 (70.25%), and reported limited English proficiency (93.22%). About 25.84% reported moderate or higher overall distress. The most common sources of distress were emotional burden (29.78%), followed by regimen- (28.65%), interpersonal- (18.54%), and physician-related distress (14.04%). Participants who were younger, female, limited English proficient, and had elevated depressive symptoms were more likely to have higher diabetes distress. CONCLUSION/UNASSIGNED:Diabetes distress is prevalent among Chinese immigrants with type 2 diabetes, especially emotional- and regimen-related distress. Given the known link between diabetes distress and poor glycemic control, it is critical to screen for diabetes distress at primary care clinics and incorporate psychological counseling in diabetes care in this underserved population.

In this article, we present a flexible model for microbiome count data. We consider a quasi-likelihood framework, in which we do not make any assumptions on the distribution of the microbiome count except that its variance is an unknown but smooth function of the mean. By comparing our model to the negative binomial generalized linear model (GLM) and Poisson GLM in simulation studies, we show that our flexible quasi-likelihood method yields valid inferential results. Using a real microbiome study, we demonstrate the utility of our method by examining the relationship between adenomas and microbiota. We also provide an R package "fql" for the application of our method.

BACKGROUND:Recent studies have demonstrated considerable interindividual variability in postprandial glucose response (PPGR) to the same foods, suggesting the need for more precise methods for predicting and controlling PPGR. In the Personal Nutrition Project, the investigators tested a precision nutrition algorithm for predicting an individual's PPGR. OBJECTIVE:This study aimed to compare changes in glycemic variability (GV) and HbA1c in 2 calorie-restricted weight loss diets in adults with prediabetes or moderately controlled type 2 diabetes (T2D), which were tertiary outcomes of the Personal Diet Study. METHODS:The Personal Diet Study was a randomized clinical trial to compare a 1-size-fits-all low-fat diet (hereafter, standardized) with a personalized diet (hereafter, personalized). Both groups received behavioral weight loss counseling and were instructed to self-monitor diets using a smartphone application. The personalized arm received personalized feedback through the application to reduce their PPGR. Continuous glucose monitoring (CGM) data were collected at baseline, 3 mo and 6 mo. Changes in mean amplitude of glycemic excursions (MAGEs) and HbA1c at 6 mo were assessed. We performed an intention-to-treat analysis using linear mixed regressions. RESULTS:We included 156 participants [66.5% women, 55.7% White, 24.1% Black, mean age 59.1 y (standard deviation (SD) = 10.7 y)] in these analyses (standardized = 75, personalized = 81). MAGE decreased by 0.83 mg/dL per month for standardized (95% CI: 0.21, 1.46 mg/dL; P = 0.009) and 0.79 mg/dL per month for personalized (95% CI: 0.19, 1.39 mg/dL; P = 0.010) diet, with no between-group differences (P = 0.92). Trends were similar for HbA1c values. CONCLUSIONS:Personalized diet did not result in an increased reduction in GV or HbA1c in patients with prediabetes and moderately controlled T2D, compared with a standardized diet. Additional subgroup analyses may help to identify patients who are more likely to benefit from this personalized intervention. This trial was registered at clinicaltrials.gov as NCT03336411.

BACKGROUND:Emerging evidence suggests the potential mediating role of microbiome in health disparities. However, no analytic framework can be directly used to analyze microbiome as a mediator between health disparity and clinical outcome, due to the non-manipulable nature of the exposure and the unique structure of microbiome data, including high dimensionality, sparsity, and compositionality. METHODS:Considering the modifiable and quantitative features of the microbiome, we propose a microbial causal mediation model framework, SparseMCMM_HD, to uncover the mediating role of microbiome in health disparities, by depicting a plausible path from a non-manipulable exposure (e.g., ethnicity or region) to the outcome through the microbiome. The proposed SparseMCMM_HD rigorously defines and quantifies the manipulable disparity measure that would be eliminated by equalizing microbiome profiles between comparison and reference groups and innovatively and successfully extends the existing microbial mediation methods, which are originally proposed under potential outcome or counterfactual outcome study design, to address health disparities. RESULTS:Through three body mass index (BMI) studies selected from the curatedMetagenomicData 3.4.2 package and the American gut project: China vs. USA, China vs. UK, and Asian or Pacific Islander (API) vs. Caucasian, we exhibit the utility of the proposed SparseMCMM_HD framework for investigating the microbiome's contributions in health disparities. Specifically, BMI exhibits disparities and microbial community diversities are significantly distinctive between reference and comparison groups in all three applications. By employing SparseMCMM_HD, we illustrate that microbiome plays a crucial role in explaining the disparities in BMI between ethnicities or regions. 20.63%, 33.09%, and 25.71% of the overall disparity in BMI in China-USA, China-UK, and API-Caucasian comparisons, respectively, would be eliminated if the between-group microbiome profiles were equalized; and 15, 18, and 16 species are identified to play the mediating role respectively. CONCLUSIONS:The proposed SparseMCMM_HD is an effective and validated tool to elucidate the mediating role of microbiome in health disparity. Three BMI applications shed light on the utility of microbiome in reducing BMI disparity by manipulating microbial profiles. Video Abstract.

OBJECTIVES/OBJECTIVE:Although technology-supported interventions are effective for reducing chronic disease risk, little is known about the relative and combined efficacy of mobile health strategies aimed at multiple lifestyle factors. The purpose of this clinical trial is to evaluate the efficacy of technology-supported behavioral intervention strategies for managing multiple lifestyle-related health outcomes in overweight adults with type 2 diabetes (T2D) and chronic kidney disease (CKD). DESIGN AND METHODS/METHODS:, age ≥40 years), T2D, and CKD stages 2-4 were randomized to an advice control group, or remotely delivered programs consisting of synchronous group-based education (all groups), plus (1) Social Cognitive Theory-based behavioral counseling and/or (2) mobile self-monitoring of diet and physical activity. All programs targeted weight loss, greater physical activity, and lower intakes of sodium and phosphorus-containing food additives. RESULTS:Of 256 randomized participants, 186 (73%) completed 6-month assessments. Compared to the ADVICE group, mHealth interventions did not result in significant changes in weight loss, or urinary sodium and phosphorus excretion. In aggregate analyses, groups receiving mobile self-monitoring had greater weight loss at 3 months (P = .02), but between 3 and 6 months, weight losses plateaued, and by 6 months, the differences were no longer statistically significant. CONCLUSIONS:When engaging patients with T2D and CKD in multiple behavior changes, self-monitoring diet and physical activity demonstrated significantly larger short-term weight losses. Theory-based behavioral counseling alone was no better than baseline advice and demonstrated no interaction effect with self-monitoring.

Recent studies have suggested that the temporal dynamics of the human microbiome may have associations with human health and disease. An increasing number of longitudinal microbiome studies, which record time to disease onset, aim to identify candidate microbes as biomarkers for prognosis. Owing to the ultra-skewness and sparsity of microbiome proportion (relative abundance) data, directly applying traditional statistical methods may result in substantial power loss or spurious inferences. We propose a novel joint modeling framework [JointMM], which is comprised of two sub-models: a longitudinal sub-model called zero-inflated scaled-Beta generalized linear mixed-effects regression to depict the temporal structure of microbial proportions among subjects; and a survival sub-model to characterize the occurrence of an event and its relationship with the longitudinal microbiome proportions. JointMM is specifically designed to handle the zero-inflated and highly skewed longitudinal microbial proportion data and examine whether the temporal pattern of microbial presence and/or the non-zero microbial proportions are associated with differences in the time to an event. The longitudinal sub-model of JointMM also provides the capacity to investigate how the (time-varying) covariates are related to the temporal microbial presence/absence patterns and/or the changing trend in non-zero proportions. Comprehensive simulations and real data analyses are used to assess the statistical efficiency and interpretability of JointMM. This article is protected by copyright. All rights reserved.

BACKGROUND:Type 2 diabetes (T2D) in Chinese Americans is a rising public health concern for the US health care system. The majority of Chinese Americans with T2D are foreign-born older immigrants and report limited English proficiency and health literacy. Multiple social determinants of health limit access to evidence-based diabetes interventions for underserved Chinese immigrants. A social media-based diabetes intervention may be feasible to reach this community. OBJECTIVE:The purpose of the Chinese American Research and Education (CARE) study was to examine the potential efficacy of a social media-based intervention on glycemic control in Chinese Americans with T2D. Additionally, the study aimed to explore the potential effects of the intervention on psychosocial and behavioral factors involved in successful T2D management. In this report, we describe the design and protocol of the CARE trial. METHODS:and psychosocial and behavioral outcomes. RESULTS:This pilot RCT study was approved by the Institutional Review Board at NYU Grossman School of Medicine in March 2021. The first participant was enrolled in March 2021, and the recruitment goal (n=60) was met in March 2022. All data collection is expected to conclude by November 2022, with data analysis and study results ready for reporting by December 2023. Findings from this pilot RCT will further guide the team in planning a future large-scale study. CONCLUSIONS:This study will serve as an important first step in exploring scalable interventions to increase access to evidence-based diabetes interventions among underserved, low-income, immigrant populations. This has significant implications for chronic care in other high-risk immigrant groups, such as low-income Hispanic immigrants, who also bear a high T2D burden, face similar barriers to accessing diabetes programs, and report frequent social media use (eg, WhatsApp). TRIAL REGISTRATION/BACKGROUND:ClinicalTrials.gov NCT03557697; https://clinicaltrials.gov/ct2/show/NCT03557697. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)/UNASSIGNED:DERR1-10.2196/42554.

Importance:Interindividual variability in postprandial glycemic response (PPGR) to the same foods may explain why low glycemic index or load and low-carbohydrate diet interventions have mixed weight loss outcomes. A precision nutrition approach that estimates personalized PPGR to specific foods may be more efficacious for weight loss. Objective:To compare a standardized low-fat vs a personalized diet regarding percentage of weight loss in adults with abnormal glucose metabolism and obesity. Design, Setting, and Participants:The Personal Diet Study was a single-center, population-based, 6-month randomized clinical trial with measurements at baseline (0 months) and 3 and 6 months conducted from February 12, 2018, to October 28, 2021. A total of 269 adults aged 18 to 80 years with a body mass index (calculated as weight in kilograms divided by height in meters squared) ranging from 27 to 50 and a hemoglobin A1c level ranging from 5.7% to 8.0% were recruited. Individuals were excluded if receiving medications other than metformin or with evidence of kidney disease, assessed as an estimated glomerular filtration rate of less than 60 mL/min/1.73 m2 using the Chronic Kidney Disease Epidemiology Collaboration equation, to avoid recruiting patients with advanced type 2 diabetes. Interventions:Participants were randomized to either a low-fat diet (<25% of energy intake; standardized group) or a personalized diet that estimates PPGR to foods using a machine learning algorithm (personalized group). Participants in both groups received a total of 14 behavioral counseling sessions and self-monitored dietary intake. In addition, the participants in the personalized group received color-coded meal scores on estimated PPGR delivered via a mobile app. Main Outcomes and Measures:The primary outcome was the percentage of weight loss from baseline to 6 months. Secondary outcomes included changes in body composition (fat mass, fat-free mass, and percentage of body weight), resting energy expenditure, and adaptive thermogenesis. Data were collected at baseline and 3 and 6 months. Analysis was based on intention to treat using linear mixed modeling. Results:Of a total of 204 adults randomized, 199 (102 in the personalized group vs 97 in the standardized group) contributed data (mean [SD] age, 58 [11] years; 133 women [66.8%]; mean [SD] body mass index, 33.9 [4.8]). Weight change at 6 months was -4.31% (95% CI, -5.37% to -3.24%) for the standardized group and -3.26% (95% CI, -4.25% to -2.26%) for the personalized group, which was not significantly different (difference between groups, 1.05% [95% CI, -0.40% to 2.50%]; P = .16). There were no between-group differences in body composition and adaptive thermogenesis; however, the change in resting energy expenditure was significantly greater in the standardized group from 0 to 6 months (difference between groups, 92.3 [95% CI, 0.9-183.8] kcal/d; P = .05). Conclusions and Relevance:A personalized diet targeting a reduction in PPGR did not result in greater weight loss compared with a low-fat diet at 6 months. Future studies should assess methods of increasing dietary self-monitoring adherence and intervention exposure. Trial Registration:ClinicalTrials.gov Identifier: NCT03336411.