A machine-learned model for predicting weight loss success using weight change features early in treatment

Abstract Stepped-care obesity treatments aim to improve efficiency by early identification of non-responders and adjusting interventions but lack validated models. We trained a random forest classifier to improve the predictive utility of a clinical decision rule (>0.5 lb weight loss/week) that i...

Full description

Saved in:
Bibliographic Details
Main Authors: Farzad Shahabi, Samuel L. Battalio, Angela Fidler Pfammatter, Donald Hedeker, Bonnie Spring, Nabil Alshurafa
Format: Article
Language:English
Published: Nature Portfolio 2024-11-01
Series:npj Digital Medicine
Online Access:https://doi.org/10.1038/s41746-024-01299-y
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846147403510448128
author Farzad Shahabi
Samuel L. Battalio
Angela Fidler Pfammatter
Donald Hedeker
Bonnie Spring
Nabil Alshurafa
author_facet Farzad Shahabi
Samuel L. Battalio
Angela Fidler Pfammatter
Donald Hedeker
Bonnie Spring
Nabil Alshurafa
author_sort Farzad Shahabi
collection DOAJ
description Abstract Stepped-care obesity treatments aim to improve efficiency by early identification of non-responders and adjusting interventions but lack validated models. We trained a random forest classifier to improve the predictive utility of a clinical decision rule (>0.5 lb weight loss/week) that identifies non-responders in the first 2 weeks of a stepped-care weight loss trial (SMART). From 2009 to 2021, 1058 individuals with obesity participated in three studies: SMART, Opt-IN, and ENGAGED. The model was trained on 80% of the SMART data (224 participants), and its in-distribution generalizability was tested on the remaining 20% (remaining 57 participants). The out-of-distribution generalizability was tested on the ENGAGED and Opt-IN studies (472 participants). The model predicted weight loss at month 6 with an 84.5% AUROC and an 86.3% AUPRC. SHAP identified predictive features: weight loss at week 2, ranges/means and ranges of weight loss, slope, and age. The SMART-trained model showed generalizable performance with no substantial difference across studies.
format Article
id doaj-art-41e4ba0cb0474a50b6071a40354e6233
institution Kabale University
issn 2398-6352
language English
publishDate 2024-11-01
publisher Nature Portfolio
record_format Article
series npj Digital Medicine
spelling doaj-art-41e4ba0cb0474a50b6071a40354e62332024-12-01T12:46:08ZengNature Portfolionpj Digital Medicine2398-63522024-11-017111010.1038/s41746-024-01299-yA machine-learned model for predicting weight loss success using weight change features early in treatmentFarzad Shahabi0Samuel L. Battalio1Angela Fidler Pfammatter2Donald Hedeker3Bonnie Spring4Nabil Alshurafa5Department of Preventive Medicine, Feinberg School of Medicine, Northwestern UniversityDepartment of Preventive Medicine, Feinberg School of Medicine, Northwestern UniversityCollege of Education, Health, and Human Sciences, University of TennesseeDepartment of Public Health Sciences, The University of ChicagoDepartment of Preventive Medicine, Feinberg School of Medicine, Northwestern UniversityDepartment of Preventive Medicine, Feinberg School of Medicine, Northwestern UniversityAbstract Stepped-care obesity treatments aim to improve efficiency by early identification of non-responders and adjusting interventions but lack validated models. We trained a random forest classifier to improve the predictive utility of a clinical decision rule (>0.5 lb weight loss/week) that identifies non-responders in the first 2 weeks of a stepped-care weight loss trial (SMART). From 2009 to 2021, 1058 individuals with obesity participated in three studies: SMART, Opt-IN, and ENGAGED. The model was trained on 80% of the SMART data (224 participants), and its in-distribution generalizability was tested on the remaining 20% (remaining 57 participants). The out-of-distribution generalizability was tested on the ENGAGED and Opt-IN studies (472 participants). The model predicted weight loss at month 6 with an 84.5% AUROC and an 86.3% AUPRC. SHAP identified predictive features: weight loss at week 2, ranges/means and ranges of weight loss, slope, and age. The SMART-trained model showed generalizable performance with no substantial difference across studies.https://doi.org/10.1038/s41746-024-01299-y
spellingShingle Farzad Shahabi
Samuel L. Battalio
Angela Fidler Pfammatter
Donald Hedeker
Bonnie Spring
Nabil Alshurafa
A machine-learned model for predicting weight loss success using weight change features early in treatment
npj Digital Medicine
title A machine-learned model for predicting weight loss success using weight change features early in treatment
title_full A machine-learned model for predicting weight loss success using weight change features early in treatment
title_fullStr A machine-learned model for predicting weight loss success using weight change features early in treatment
title_full_unstemmed A machine-learned model for predicting weight loss success using weight change features early in treatment
title_short A machine-learned model for predicting weight loss success using weight change features early in treatment
title_sort machine learned model for predicting weight loss success using weight change features early in treatment
url https://doi.org/10.1038/s41746-024-01299-y
work_keys_str_mv AT farzadshahabi amachinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT samuellbattalio amachinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT angelafidlerpfammatter amachinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT donaldhedeker amachinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT bonniespring amachinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT nabilalshurafa amachinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT farzadshahabi machinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT samuellbattalio machinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT angelafidlerpfammatter machinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT donaldhedeker machinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT bonniespring machinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment
AT nabilalshurafa machinelearnedmodelforpredictingweightlosssuccessusingweightchangefeaturesearlyintreatment