Which Predictive Equation Of Resting Energy Expenditure Has The Best Agreement With Indirect Calorimetry In Elderly Patients On Hemodialysis?

Abstract

As the incidence of elderly patients initiating dialysis has been increasing, there is a need to develop specialized care to them. According to the European Best Practice Guideline, the energy needs of hemodialysis (HD) patients should be estimated by multiplying the resting energy expenditure (REE), obtained by predictive equations, for the physical activity factor. The predictive equation that yields better agreement to the indirect calorimetry in elderly patients on HD has not been rated yet. We aimed to evaluate the agreement between indirect calorimetry (IC) and the predictive equations of Harris&Benedict (HB), Schofield and WHO/FAO 1985. Fifty-seven elderly patients (38 males (67%); 69±6 years) on HD were included. The REE (kcal/day) estimated by the equations were higher than that measured by IC: IC (1246 ±288); HB (1443 ±279); Schofield (1357 ±232); WHO 1985 (1384 ±226) (P<0.05). The HB equation had the lowest intraclass correlation coefficient (ICC), the highest mean difference from the REE measured by IC and a high frequency of overestimation (Table).Empty CellIntraclass CC (r; 95% CI)Bland-Altman*REE overestimation (%[n])HB x REE0.53(0.31;0.69)188(-331; 70959.6 [34]Schofield x REE0.72(0.57;0.83)112(-248; 472)56.1 [32]WHO 1985 x REE0.74(0.60;0.84)139(-222; 499)59.6 [34]⁎Mean difference and interquartile range; CI: confidence interval Mean difference and interquartile range; CI: confidence interval In conclusion, all predicted equations overestimated the REE in elderly HD patients. Among them, the HB equation had the worse agreement with IC.