Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Subtyping patients with chronic disease using longitudinal BMI patterns

Md Mozaharul Mottalib, Jessica C Jones-Smith, Bethany Sheridan, Rahmatollah Beheshti

TL;DR

This study asks whether longitudinal BMI trajectories can reveal subtypes of risk for 18 major chronic diseases using a large US EHR dataset across six years. It introduces nine interpretable BMI-trajectory features and applies k-means clustering to create disease-specific subtypes, subsequently profiling clusters by demographic, socioeconomic, and physiological factors. The analysis identifies meaningful BMI-trajectory subtypes in 8 cohorts, with strong links between BMI patterns and diabetes, hypertension, and dementia, and notable patterns in cancer and other conditions. While focusing on interpretability, the work highlights potential for BMI-history-based risk stratification and suggests avenues for incorporating additional covariates and temporal models in future research.

Abstract

Obesity is a major health problem, increasing the risk of various major chronic diseases, such as diabetes, cancer, and stroke. While the role of obesity identified by cross-sectional BMI recordings has been heavily studied, the role of BMI trajectories is much less explored. In this study, we use a machine-learning approach to subtype individuals' risk of developing 18 major chronic diseases by using their BMI trajectories extracted from a large and geographically diverse EHR dataset capturing the health status of around two million individuals for a period of six years. We define nine new interpretable and evidence-based variables based on the BMI trajectories to cluster the patients into subgroups using the k-means clustering method. We thoroughly review each cluster's characteristics in terms of demographic, socioeconomic, and physiological measurement variables to specify the distinct properties of the patients in the clusters. In our experiments, the direct relationship of obesity with diabetes, hypertension, Alzheimer's, and dementia has been re-established and distinct clusters with specific characteristics for several of the chronic diseases have been found to be conforming or complementary to the existing body of knowledge.

Subtyping patients with chronic disease using longitudinal BMI patterns

TL;DR

This study asks whether longitudinal BMI trajectories can reveal subtypes of risk for 18 major chronic diseases using a large US EHR dataset across six years. It introduces nine interpretable BMI-trajectory features and applies k-means clustering to create disease-specific subtypes, subsequently profiling clusters by demographic, socioeconomic, and physiological factors. The analysis identifies meaningful BMI-trajectory subtypes in 8 cohorts, with strong links between BMI patterns and diabetes, hypertension, and dementia, and notable patterns in cancer and other conditions. While focusing on interpretability, the work highlights potential for BMI-history-based risk stratification and suggests avenues for incorporating additional covariates and temporal models in future research.

Abstract

Obesity is a major health problem, increasing the risk of various major chronic diseases, such as diabetes, cancer, and stroke. While the role of obesity identified by cross-sectional BMI recordings has been heavily studied, the role of BMI trajectories is much less explored. In this study, we use a machine-learning approach to subtype individuals' risk of developing 18 major chronic diseases by using their BMI trajectories extracted from a large and geographically diverse EHR dataset capturing the health status of around two million individuals for a period of six years. We define nine new interpretable and evidence-based variables based on the BMI trajectories to cluster the patients into subgroups using the k-means clustering method. We thoroughly review each cluster's characteristics in terms of demographic, socioeconomic, and physiological measurement variables to specify the distinct properties of the patients in the clusters. In our experiments, the direct relationship of obesity with diabetes, hypertension, Alzheimer's, and dementia has been re-established and distinct clusters with specific characteristics for several of the chronic diseases have been found to be conforming or complementary to the existing body of knowledge.

Paper Structure

This paper contains 16 sections, 7 equations, 1 figure, 5 tables.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Materials and Methods
  4. Data Processing
  5. Study Variables
  6. Weighted Mean
  7. Trend
  8. Normalized up-count and down-count
  9. Maximum BMI
  10. The maximum change in BMI
  11. Start and ending BMI categories
  12. Median
  13. Subtyping (clustering) process
  14. Results
  15. Discussion
  16. ...and 1 more sections

Figures (1)

  • Figure 1: Distribution of the clusters for the 8 chronic diseases and the overall BMI trends (the shapes of signals) for each cluster. The top chart (horizontal bars) in each sub-figure shows how the positive (those with the disease incidence) cases and negative cases (those without the disease incidence) are distributed within the identified subgroups (clusters). The bottom chart (vertical bars) shows the BMI trends corresponding to each cluster. The mean number of individuals ± (std) is shown for each cluster.