JMIR Publications

ABSTRACT

Background:

Cardiorespiratory fitness (VO2max) is a key predictor of cardiovascular and other health-related diseases and is often impaired in individuals with obesity due to functional and structural limitations. Improving cardiorespiratory fitness enhances overall health and reduces mortality, making it an important indicator for preventing and treating obesity.

Objective:

The primary aim of this study is to develop an obesity-specific machine learning (ML) model that can accurately estimate VO2max, a key indicator of cardiovascular fitness, and make it accessible through a web-based application.

Methods:

A ML model will be trained to estimate VO2max in individuals with obesity using retrospective data combining assessments of VO2max tests and clinical parameters from adult patients with severe obesity body mass index (BMI ≥40.0 kg/m2, or 35.0-39.9 kg/m2 with at least one obesity-related comorbidity) from Vestfold Hospital Trust, Muritunet Rehabilitation Institution and Norwegian School of Sport Sciences. The ML-based VO2max estimation model will be presented as a web application, allowing easy access and interaction. The model’s estimations will be compared against direct VO2max measurements obtained from medical equipment across institutions as part of a prospective validation. Ethical approval has been obtained for the use of two databases in the initial model development; approval for the remaining data and prospective phase is pending.

Results:

Vestfold Hospital Trust, the Norwegian School of Sport Medicine, and Muritunet Rehabilitation have from 2013 to 2025 conducted > 2623 VO2max test and clinical parameters from 1,279 adults with severe obesity, both before, during and after lifestyle interventions. This comprehensive data set serves as the foundation for developing the ML model, which is presented through a user-friendly web application. A web application has been developed and will be tested with patients and healthcare personnel. The first scientific publication of the ML model is expected to be published in 2026. The results of the overall project are expected to be completed in 2028.

Conclusions:

This project aims to develop a machine learning model, which serves as a cost-effective tool for VO2max estimation in individuals with obesity, improving accessibility to this important health marker. This is the first known attempt in Norway to estimate VO2max in an obese population using machine learning, based on a unique clinical database. The project holds significant societal value with potential national and international relevance for healthcare and patient outcomes. Clinical Trial: ClinicalTrials.gov (Identifier: NCT07011108)