Psychosocial-Behavioral Phenotyping: A Novel Precision Health Approach to Modeling Behavioral, Psychological, and Social Determinants of Health Using Machine Learning

Abstract

Background The context in which a behavioral intervention is delivered is an important source of variability and systematic approaches are needed to identify and quantify contextual factors that may influence intervention efficacy. Machine learning-based phenotyping methods can contribute to a new precision health paradigm by informing personalized behavior interventions. Two primary goals of precision health, identifying population subgroups and highlighting behavioral intervention targets, can be addressed with psychosocial-behavioral phenotypes. We propose a method for psychosocial-behavioral phenotyping that models social determinants of health in addition to individual-level psychological and behavioral factors. Purpose To demonstrate a novel application of machine learning for psychosocial-behavioral phenotyping, the identification of subgroups with similar combinations of psychosocial characteristics. Methods In this secondary analysis of psychosocial and behavioral data from a community cohort (n = 5,883), we optimized a multichannel mixed membership model (MC3M) using Bayesian inference to identify psychosocial-behavioral phenotypes and used logistic regression to determine which phenotypes were associated with elevated weight status (BMI ≥ 25kg/m2). Results We identified 20 psychosocial-behavioral phenotypes. Phenotypes were conceptually consistent as well as discriminative; most participants had only one active phenotype. Two phenotypes were significantly positively associated with elevated weight status; four phenotypes were significantly negatively associated. Each phenotype suggested different contextual considerations for intervention design. Conclusions By depicting the complexity of psychological and social determinants of health while also providing actionable insight about similarities and differences among members of the same community, psychosocial-behavioral phenotypes can identify potential intervention targets in context.

Identifying High-Need Primary Care Patients Using Nursing Knowledge and Machine Learning Methods

“…Most similar to our work, Byrne and colleagues 27 have predicted housing instability and homelessness within the Veterans Health Administration using social history (e.g., branch of service, service use, and diagnosis) for personalized interventions, but their models are fit to a narrow population with a specific need, and do not take into account medical complexity. Burgermaster and Rodriguez 28 used sophisticated analytic approaches to define 20 different phenotypes predicting elevated weight (body mass index [BMI] ≥ 25 kg/m 2 ) and personalizing behavioral interventions based on psychosocial-behavioral characteristics. Their work yielded both positive and negative associations of elevated BMI, which has important clinical implications for personalizing care.…”

“…However, their models do not take into account the impact of the intersection of SDOH and chronic conditions on health outcomes. 28 Recently SDOH phenotypes predicting maternal health morbidity from income, stress, and immigration status show promise for improving maternal health outcomes, but the models were limited to a population of healthy individuals rather than those with chronic illness. 29 Thus, our study makes an important contribution to the literature given its pragmatic approach, consideration of the overlapping influences of SDOH and chronic conditions that can be easily replicated in community-based primary care settings, and reduction of bias through the intentional de-emphasis of costs of care as a predictor of risk.…”

A Strategic Research Framework for Defeating Diabetes in India: A 21st-Century Agenda

“… 55 There is also growing interest in studying psychosocial phenotypes, using cognitive and behavioral attributes of the patient, to tailor interventions for long-term behavioral change and adherence to interventions. 56 , 57

Figure 3: Research framework for precision medicine.

…”

“…55 There is also growing interest in studying psychosocial phenotypes, using cognitive and behavioral attributes of the patient, to tailor interventions for long-term behavioral change and adherence to interventions. 56,57 Collaborations between clinicians, epidemiologists, computer scientists, and statisticians to pursue approaches, such as multi-modal data integration and population segmentation, may be useful to characterize phenotypes for prevention, diagnosis, treatment, and prognosis. 58 Data fusion techniques to combine heterogeneous data from various sources, coupled with data analytics and machine learning algorithms, have significant potential to capture patterns and nuances in the phenotypes that may not be visible in the initial raw data extraction phase.…”

Hierarchy of demographic and social determinants of mental health: analysis of cross-sectional survey data from the Global Mind Project

“…Another challenge in mental health is that assessments are generally at the level of particular disorders and therefore do not provide an outcome of overall mental distress that aggregates across symptoms and disorders that tend to have high comorbidity 58–63. Thus, while these techniques have been used to understand the social determinants of health generally,51 64–66 they have not, to our knowledge, been used to predict mental health status from a large number of demographic and social determinants.…”