Metabolic adaptation fluctuates with different prediction equations: a secondary analysis based on a weight-loss clinical trial

Molian Tang¹Jialu Wang¹Yi Xiang²Renying Xu^1*

• ¹Department of Clinical Nutrition, Renji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
• ²Xinhua Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, Shanghai Municipality, China

Objective Metabolic adaption (MA) might be of clinical relevance in weight loss management. However, it is unclear whether resting metabolic rate (RMR) reduction in weight loss reflects true MA or fat free mass (FFM) loss. Methods We re-analyzed the data based on a weight loss trial of 131 patients (aged 33.3±6.7 years) with overweight/obesity. Anthropometric data, body composition, daily physical activity, sleep hour and dietary intake were collected for every four weeks (baseline, week 4, 8, 12, and 16). The linear mixed model was used to evaluate the absolute change in RMR and adjusted RMR [aRMR= RMR (kcal) divided by FFM (kg)] with two different equations (Katch-McArdle-determined vs. BIA-determined) for every four weeks after adjustment of age, sex, daily physical activity, sleep hours, dietary intake, and baseline FFM and fat mass (FM). Results Following the 16-week intervention, a significant reduction was observed in body weight [β: -5.6 kg; 95% Confidence Interval (CI): -6.3 kg, -5.0 kg], BMI (β: -2.3 kg/m2; 95%CI: -2.5 kg/m2, -2.0 kg/m2), FM (β: -4.7 kg; 95%CI: -5.2 kg, -4.1 kg), and FFM (β: -0.9 kg; 95%CI: -1.2 kg, -0.7 kg). Both Katch-McArdle-determined RMR and BIA-determined RMR presented a significant decrease between baseline and the end of the intervention (week 16). A small but statistically significant increase in Katch-McArdle-determined aRMR (β: 0.19 kcal/kg; 95%CI: 0.14 kcal/kg, 0.23 kcal/kg; adjusted p-value <0.0001) was confirmed by linear mixed models. While BIA-determined aRMR generally showed decreasing trends across the follow-up periods, only Week 12 demonstrated a statistically significant inverse association compared with the baseline (β: -0.21 kcal/kg; 95% CI: -0.29 kcal/kg, -0.13 kcal/kg, adjusted p-value =0.01). Conclusion The use of different prediction equations might account for variations in MA in this study. The results highlight the importance of preserving FFM during weight loss, thus to prevent reductions in RMR.