Abstract

<b>BACKGROUND</b> Two clinical formulae (CF conference formula and estimation based on 120% of average requirement for energy) have been recommended for the estimation of energy requirements in cystic fibrosis but their accuracy is unknown.<p></p> <b>AIM</b> To compare the accuracy of estimates of energy requirement derived from the two formulae.<p></p> <b>METHODS</b> Energy requirement, defined as total daily energy expenditure, was measured using the doubly labelled water method in 15 patients (six girls, nine boys; mean (SD) age, 10.0 (2.4) years) who were well and clinically stable. The accuracy of the formulae was assessed using calculation of biases and limits of agreement relative to measured energy requirement.<p></p> <b>RESULTS</b> Estimates from the CF conference formula were lower than measured values (mean paired difference, 0.52 MJ/day; 95% confidence interval (CI), −1.10 to 0.10), but this bias was not significant, and was smaller than that from the alternative formula (mean paired difference, 0.77 MJ/day; 95% CI, −0.20 to 1.74). Limits of agreement relative to measured total daily energy expenditure were narrower for the CF conference formula (−2.72 to 1.68 MJ/day) than for that based on 120% of estimated average requirement (−2.75 to 4.29 MJ/day), but with both formulae errors in estimation at the individual level were large.<p></p> <b>CONCLUSIONS</b> The CF conference formula offers improved prediction of energy requirements, but the accuracy of both formulae at the individual level is not sufficiently good for clinical purposes.