Machine Learning for Reclassification of Obesity

Data-driven Clustering for Metabolic Classification of Obesity Using Machine Learning

The goal of this study is to employ or develop computational modeling techniques for the precise reclassification of obesity into subgroups. Clinical features, risks of noncommunicable diseases, as well as weight loss effects of bariatric surgery will also be studied and compared within the subgroups.

Study Overview

• Status: Completed
• Conditions: Obesity
• Intervention / Treatment: Diagnostic test: AI classification of patients with obesity

• Study Type: Observational
• Enrollment (Actual): 2495

Contacts and Locations

Study Locations

• China
  • Shanghai
    • Shanghai, Shanghai, China, 200072
      • Shanghai Tenth People's Hospital

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 10 years to 70 years (Child, Adult, Older Adult)
• Accepts Healthy Volunteers: Yes
• Genders Eligible for Study: All

• Sampling Method: Probability Sample

Study Population

Patients with overweight/obesity.

Description

Inclusion Criteria:

1. Patients with overweight/obesity
2. Patients with normal weight as controls

Exclusion Criteria:

1. had ever been performed with a bariatric surgery before the study's first visit is scheduled;
2. had taken exogenous insulin, medication that affects glucose metabolism, or uric acid drugs currently;
3. being diagnosed with type 1 diabetes, secondary diabetes, hereditary disease, or severe disease (e.g. malignant tumor, heart failure, liver failure, etc.);
4. in gestation of lactation;
5. did not have the complete data for model;
6. for normal-weight controls, patients with diabetes or hyperuricemia were excluded.

Study Plan

How is the study designed?

Number of groups / cohorts

5

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
NW; normal weight control
MHO; metabolic healthy obesity	Diagnostic test: AI classification of patients with obesity; Computational modeling techniques will be used for the precise reclassification of obesity into four subgroups, several variables according to the clinical experience and the modeling results will be selected for the cluster analysis.
LMO; hypometabolic obesity	Diagnostic test: AI classification of patients with obesity; Computational modeling techniques will be used for the precise reclassification of obesity into four subgroups, several variables according to the clinical experience and the modeling results will be selected for the cluster analysis.
HMO-U; hypermetabolic obesity with hyperuricemia	Diagnostic test: AI classification of patients with obesity; Computational modeling techniques will be used for the precise reclassification of obesity into four subgroups, several variables according to the clinical experience and the modeling results will be selected for the cluster analysis.
HMO-I; hypermetabolic obesity with hyperinsulinemia	Diagnostic test: AI classification of patients with obesity; Computational modeling techniques will be used for the precise reclassification of obesity into four subgroups, several variables according to the clinical experience and the modeling results will be selected for the cluster analysis.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Time Frame
Metabolic classification of patients with obesity using machine learning	baseline

Secondary Outcome Measures

Outcome Measure	Time Frame
Metabolic features in patients of different subgroups	baseline
Risks for noncommunicable disease in patients of different subgroups	baseline
Effect of bariatric surgery in patients of different subgroups	1 year after bariatric surgery

Collaborators and Investigators

• Sponsor: Shanghai 10th People's Hospital
• Collaborators: University of Pittsburgh; The Affiliated Nanjing Drum Tower Hospital of Nanjing University Medical School; Shanghai East Hospital; The Third People's Hospital of Chengdu

Publications and helpful links

General Publications

• Lin Z, Feng W, Liu Y, Ma C, Arefan D, Zhou D, Cheng X, Yu J, Gao L, Du L, You H, Zhu J, Zhu D, Wu S, Qu S. Machine Learning to Identify Metabolic Subtypes of Obesity: A Multi-Center Study. Front Endocrinol (Lausanne). 2021 Jul 14;12:713592. doi: 10.3389/fendo.2021.713592. eCollection 2021.

Study record dates

Study Major Dates

• Study Start (Actual): March 1, 2020
• Primary Completion (Actual): April 30, 2020
• Study Completion (Actual): June 20, 2020

Study Registration Dates

• First Submitted: February 21, 2020
• First Submitted That Met QC Criteria: February 21, 2020
• First Posted (Actual): February 25, 2020

Study Record Updates

• Last Update Posted (Actual): June 25, 2020
• Last Update Submitted That Met QC Criteria: June 23, 2020
• Last Verified: June 1, 2020

More Information

Terms related to this study

• Keywords: Obesity; Bariatric surgery; Machine learning; Clustering; Metabolic classification
• Additional Relevant MeSH Terms: Overnutrition; Nutrition Disorders; Overweight; Body Weight; Obesity
• Other Study ID Numbers: Obesity Reclassification

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No