Total energy expenditure and its components in healthy young adults from Southwest China using doubly labeled water, indirect calorimetry, accelerometry, and energy intake.
Level 4 - case-series / case-control
Small single-cohort cross-sectional observational study comparing physiological measurement methods
PubMed 42537543 · doi:10.1016/j.nut.2026.113350
What was done
Twenty healthy adults aged 18–30 years from Southwest China completed a 7-day free-living assessment. Total energy expenditure (TEE) was measured using doubly labeled water (DLW), and resting metabolic rate (RMR) was measured by indirect calorimetry. Physical activity was monitored using triaxial accelerometry with activity records, and dietary intake was assessed by the duplicate-portion method with laboratory analysis. Physical activity level (PAL) was calculated as TEE/RMR, and thermic effect of food (TEF) was estimated by difference. DLW-measured TEE was compared with factorial TEE and equation-estimated TEE based on Chinese and FAO/WHO/UNU prediction equations.
What was found
DLW-measured TEE was 2039.63 ± 304.43 kcal/d and did not differ significantly from factorial TEE or equation-estimated TEE at the group level. Measured RMR was 1584.29 ± 232.97 kcal/d. Chinese equation-predicted RMR significantly underestimated measured RMR (p < 0.05), whereas the FAO/WHO/UNU equation was comparable. Measured PAL was low (1.29 ± 0.12). Component partitioning of TEE showed RMR accounted for 78.0%, physical activity energy expenditure accounted for 12.9%, and TEF accounted for 9.1%. Energy intake broadly agreed with DLW-measured TEE at the group level, though Bland-Altman analysis revealed individual-level variability.
Why it matters
This study shows that while standard equations may match group-level total energy expenditure in sedentary young Chinese adults, Chinese prediction equations can systematically underestimate resting metabolic rate.
Limits
The sample size was very small (n = 20) and recruited from a single region, limiting generalizability. The free-living assessment lasted only 7 days. Thermic effect of food was estimated by difference rather than measured directly.