Measuring energy needs during rehabilitation

Active Rehabilitation Training System for Motor and Cognitive Function Technologies and Systems for Assessing Human Energy Metabolism and Nutritional Rehabilitation

This project will try to use metabolic measurements and machine learning to predict calorie needs for adults recovering from stroke, traumatic brain injury, or major surgery.

Quick facts

• Study type: Observational
• Enrollment: 80 (estimated)
• Ages: 20 Years to 70 Years
• Sex: All
• Sponsor: Shenzhen Institutes of Advanced Technology ,Chinese Academy of Sciences (other)
• Locations: 1 site (Foshan, Guangdong)
• Trial ID: NCT07056504 on ClinicalTrials.gov

What this trial studies

Researchers will combine a doubly labeled water (DLW) database from healthy people with detailed, multimodal data collected from rehabilitation patients to build predictive models of energy requirements. Participants will undergo anthropometry, body composition, resting metabolic rate by indirect calorimetry, total energy expenditure by DLW and a metabolic chamber, activity monitoring, dietary intake and hunger ratings, sleep and cognitive testing, noninvasive brain monitoring, and gait analysis. Urine and other biological samples will be collected and DLW samples will be analyzed at a central laboratory, with machine learning used to develop injury-specific prediction algorithms. The goal is to characterize how energy metabolism changes across rehabilitation and to create tools to estimate individualized energy needs.

Who should consider this trial

Good fit: Adults aged 20–70 undergoing rehabilitation for stroke, traumatic brain injury, motor injuries, or major surgical recovery who can lie flat for testing, are not pregnant, and have no metallic implants are ideal candidates.

Not a fit: Patients with loss of autonomous mobility or consciousness, metallic implants (e.g., pacemakers), severe metabolic diseases (such as uncontrolled diabetes), claustrophobia, anorexia nervosa, uremia, or inability to lie flat for one hour are excluded and unlikely to benefit from participation.

Why it matters

Potential benefit: If successful, clinicians could use the models to personalize nutrition plans during rehabilitation, reducing under- or over-feeding and potentially improving recovery outcomes.

How similar studies have performed: Doubly labeled water and indirect calorimetry are well-established methods for measuring energy expenditure and have been successful in healthy and some patient populations, but applying machine-learning to create injury-specific energy-prediction models is a relatively novel approach.

Eligibility criteria

Inclusion Criteria:

* Male and female participants aged 20 to 70 years old;
* Rehabilitation patients with stroke or motor injuries;

Exclusion Criteria:

* Loss of autonomous mobility;
* Loss of consciousness;
* Individuals with metallic implants (e.g., pacemakers);
* Severe metabolic diseases (e.g., diabetes, other hereditary metabolic disorders);
* Patients with claustrophobia;
* Individuals with anorexia nervosa;
* Women in the preconception period, pregnancy, or lactation;
* Uremic patients;
* Inability to lie flat for one hour.

Where this trial is running

Foshan, Guangdong

• Guangdong Jianxiang Hospital Group — Foshan, Guangdong, China (RECRUITING)

Study contacts

• Principal investigator: Xueying Zhang, PhD — Shenzhen Institutes of Advanced Technology ,Chinese Academy of Sciences
• Study coordinator: Xueying Zhang, Doctor

Conditions: Energy Metabolism, Nutrition, Healthy