Ipek Ensari
Early Career Tenure-track Researcher, Icahn School of Medicine at Mount Sinai
2 active projects
Personal predictors of gestational diabetes and pregnancy outcomes
Scientific Questions Being Studied
This project aims to investigate modifiable lifestyle factors that are predictive of pregnancy outcomes in individuals with gestational diabetes. We will investigate physical activity as the primary predictor. We will also compare prevalence estimates among those with vs without gestational diabetes. The second aim of the project is to investigate these questions among those with co-morbid polycystic ovarian syndrome (PCOS) and gestational diabetes. The findings from the analyses will provide updated, novel findings on the relationship between objectively-estimated, longitudinal, and granular physical activity patterns and pregnancy outcomes among those with increased risk for adverse outcome. Third, we will assess access to care and insurance status as potential moderators.
Project Purpose(s)
- Disease Focused Research (gestational diabetes, PCOS)
- Population Health
- Social / Behavioral
Scientific Approaches
We will use the data from the EHRs, physical activity data from fitbits, and self-report health surveys (baseline health and medical history).
We will use the additive generalized models with unsupervised learning to identify potential latent profiles of activity patterns. We will use standard mixed effects regression models and depending on data availability causal inference modeling to assess the predictors of pregnancy outcomes in our defined cohort.
Anticipated Findings
We hypothesize that engaging in sufficient dose of regular physical activity will be a protective factor against adverse pregnancy outcomes in all patient groups, independent of disease and co-morbidity status. We furthermore expect to observe a dose-response in this relationship.
We anticipate that the overall physical activity volume will be higher among those without gestational diabetes (vs those with the diagnosis). We expect similar patterns for PCOS, though to a lesser magnitude.
Demographic Categories of Interest
- Access to Care
Data Set Used
Registered TierResearch Team
Owner:
- Ipek Ensari - Early Career Tenure-track Researcher, Icahn School of Medicine at Mount Sinai
Female reproductive disorders across diverse patient populations
Scientific Questions Being Studied
This project aims to investigate the following primary questions pertaining to disorders of the female reproductive anatomy:
1) Characterization of symptom clusters associated with endometriosis, and other women's pelvic pain disorders (e.g., endometriosis, dysmenorrhea) across diverse demographic and socioeconomic segments of the population
3) Modifiable lifestyle and health behaviors (e.g., physical activity) as predictors of phenotypes identified above, and socioeconomic factors that moderate this relationship
4) Feasibility and performance of different unsupervised learning methods for reliably characterizing disease sub-types and trajectories ("phenotypes")
These questions aim to address the knowledge gaps in the area of women's reproductive health across with respect to their diagnosis, treatment, and definition in clinical records, with a focus on patient populations at increased risk for health disparities and inequities.
Project Purpose(s)
- Disease Focused Research (female reproductive and related chronic pain disorders )
- Methods Development
- Control Set
Scientific Approaches
Data: We will use data from the electronic health records including structured data/codes on diagnosis, medications, treatments, patient-level factors (e.g., insurance, demographics), self-reported survey data, and wearable data to answer our research questions.
Methodology: We will primarily implement unsupervised learning approaches (e.g., hierarchical clustering) for mixed type data to analyze the survey and EHR variables. We will use functional data analytic methods for analyzing the longitudinal data from wearables (e.g., on physical activity from the FitBit devices). For any prediction and performance evaluation, we will rely on standard regression-based estimation and prediction modeling techniques.
Anticipated Findings
We expect to find sub-types based on symptom and QoL-related heterogeneity within the cohort of endometriosis patients, as well as differences in the documented prevalence and labels in EHRs based on socioeconomic factors and insurance status. We expect differences in reported treatment and self-management approaches, and habitual health behaviors across these groups. Finally, we anticipate that overall physical activity levels will vary across the sample; however, there will be distinct patterns of diurnal activity that are associated with disease status, self-management behaviors, and co-morbidities that might impair physical functioning. These findings can help identify potential targets of self-management interventions, as well as approaches for how to harmonize self-reported and passively-obtained patient-generated health data can augment the EHRs for conditions that are less frequently documented as standard practice to provide a comprehensive patient health profile.
Demographic Categories of Interest
- Race / Ethnicity
- Geography
- Access to Care
- Income Level
Data Set Used
Registered TierResearch Team
Owner:
- Ipek Ensari - Early Career Tenure-track Researcher, Icahn School of Medicine at Mount Sinai
Collaborators:
- Sahiti Kolli - Research Fellow, Icahn School of Medicine at Mount Sinai
Request a Review of this Research Project
You can request that the All of Us Resource Access Board (RAB) review a research purpose description if you have concerns that this research project may stigmatize All of Us participants or violate the Data User Code of Conduct in some other way. To request a review, you must fill in a form, which you can access by selecting ‘request a review’ below.
