Low validity of predictive equations for calculating resting energy expenditure in overweight and obese women with polycystic ovary syndrome.

BACKGROUND: Predictive equations are the main clinical tools for determining resting energy expenditure (REE). However, their adequate use in overweight and obese individuals is unclear. Thus, we investigated the best predictive equations for estimating REE in overweight and obese women with polycystic ovary syndrome (PCOS). METHODS: Eleven analyses were performed with prediction equations (pREE) based on anthropometric parameters in 30 overweight or obese women with PCOS without other chronic diseases. The measured REE (mREE) was calculated by indirect calorimetry. The validity of the equations was investigated by comparison, accuracy and agreement tests between pREE and mREE at both the individual and group level. RESULTS: Four analyses were similar to those of mREE, and smallest mean differences were observed for the World Health Organization/Food and Agriculture Organization of the United Nations/United Nations University (WHO/FAO/UNU) considering weight (W) [0.07 (1.13) MJ (16 [270] kcal)]. Individual accuracy was greater than 50% for Harris and Benedict, Müller and Lazzer equations. The percentage of REE underestimation ranged between 16.7% and 73.3%, whereas higher rates of overestimation were observed in the De Luis (66.7%) and Ireton-Jones (43.3%) equations. Mean bias at the group level was lowest in the WHO/FAO/UNU W and WHO/FAO/UNU considering weight and height (WH), Müller and Lazzer equations (-2.8 to 0.5). The WHO/FAO/UNU W and WHO/FAO/UNU WH formulas were optimal in individual agreement (33.3%). CONCLUSIONS: FAO/WHO/UNU W equations may estimate the REE in overweight and obese women with PCOS. However, the low individual accuracy and agreement in relation to mREE suggest caution regarding when to use the formula to perform an individual nutritional plan.