Resting and total energy expenditure of patients with long-chain fatty acid oxidation disorders (LC-FAODs)

Abstract

The basis of medical nutrition therapy for patients with LC-FAODs is to provide adequate energy to maintain anabolism and prevent catabolism. In practice, energy needs are estimated based on formulas derived from normal populations but it is unknown if energy expenditure among patients with LC-FAODs is similar to the normal population. We measured resting energy expenditure (REE), total energy expenditure (TEE) and body composition in 31 subjects with LC-FAODs ranging in age from 7 to 64 years. Measured REE was lower than estimated REE by various prediction equations and measured TEE was lower than estimated TEE. It is possible that the lower energy expenditure based on prediction formulas from the normal population is due to differences in body composition; we compared body composition to normal data from the 2017-18 National Health and Nutrition Examination Survey (NHANES). Fat free mass and fat mass was similar between subjects with an LC-FAOD and NHANES normal data suggesting no difference in body composition. We then compared measured REE and TEE to normal published data from the Dietary Reference Intakes (DRI). Measured REE and TEE were significantly lower among subjects with LC-FAODs compared to normal published energy expenditure data. Our results suggests patients with a LC-FAOD exhibit a lower REE and therefore actually have a slightly lower TEE than estimated. Current prediction equations may overestimate energy expenditure of patients with a LC-FAOD.

Keywords: Energy expenditure; Fatty acid oxidation disorders; Resting energy expenditure; Total energy expenditure.

Figure 1. The mitochondrial long-chain fatty acid oxidation pathway:

Long-chain fatty acids are imported into the mitochondria via the carnitine palmitoyltransferase shuttle. Once inside the mitochondria, B-oxidation proceeds in a 4-step process catalyzed by very long-chain acyl-CoA dehydrogenase (VLCAD) and trifunctional protein (TFP). Participants in this study had inherited defects in CPT2, VLCAD and TFP, including long-chain 3-hydroxyacyl-CoA dehydrogenase (LCHAD). CPT1 = carnitine palmitoyltransferase 1, CACT = carnitine acylcarnitine translocase, CPT2 = carnitine palmitoyltransferase 2. Created in BioRender.com

Figure 2:. Resting and total energy expenditure of 31 subjects with a LC-FAOD by FFM.

Resting Energy Expenditure (REE) and Total Energy Expenditure (TEE) were measured at baseline (orange) and again after 4 months (blue). The corresponding shading represents the 95% confidence intervals of the measurement. A) The adjusted R² for the full model is 0.91. BMI category (normal, overweight/obese; p<0.001), age(p<0.001) and FFM(p<0.001) but not sex or timepoint (baseline or 4 months) were significantly related to measured REE. B) The adjusted R² for the full model is 0.68. Similarly, BMI category (p=0.002), age (p=0.006) and FFM(p<0.001) but not sex or timepoint were significantly related to measured TEE. kcal = kilocalories; kg= kilograms

Figure 3:. Measured REE (mREE) compared to predicted BMR and measured TEE (mTEE) compared to predicted EER in subjects with a LC-FAOD.

Spaghetti plots illustrate measured REE (mREE) compared to predicted BMR by various equations. Each line is one subject. Measured REE was significantly lower than predicted basal metabolic rate (BMR) by the Schofield, Harris-Benedict and Mifflin St Jeor equations. Measured TEE (mTEE) was lower than Estimated Energy Requirement (EER) by the DRI equations. Paired t-test * p< 0.05, ****p< 0.0001

Figure 4:. Similar Fat Free Mass (FFM) and Fat Mass (FM) between subjects with a LC-FAOD and matched controls.

(A) Fat Free Mass and (B) Fat Mass were similar between subjects with an LC-FAOD compared to age, sex and BMI matched NHANES 2017–18 data (n=252). Control subjects are in blue, subjects with a LC-FAOD in orange. The best fit line (solid line) and 95% confidence interval (dotted lines) are illustrated for FAOD (orange) and controls (blue). A) The adjusted R² for the full model is 0.99. Sex and BW (p<0.0001) but not age (p=0.67) or group (FAOD or control; p=0.08) are related to FFM. B) The adjusted R² for the full model is 0.99. Sex and BW (p<0.0001) but not age (p=0.85) or group (FAOD or control; p=0.11) are related to FM. kg=kilograms

Figure 5:. Dietary intake of subjects with a LC-FAOD compared to NHANES controls.

A) Violin plot of kilocalories per kg of body weight (kcal/kg) illustrates similar energy intake between subjects with FAOD (yellow 13.9±3.2) compared to NHANES controls (blue 15.0±5.7). The bold center dotted line is the median for each group, the smaller dotted lines indicate the lower and upper interquartile range. B) The macronutrient distribution is illustrated in the stacked bars. Percent protein was not different (FAOD 13.9±3.2, Control 15.0±5.7; p=0.14), % carbohydrate was higher (FAOD 53.4±10.0, Control 49.2±11.6; p=0.044) and % fat lower (FAOD 31.9±7.2, Control 35.5±8.5; p=0.02) among subjects with a FAOD compared to controls. Data in legend presented as mean±standard deviation of the mean. P-values from unpaired t-tests.

Figure 6:. Resting Energy Expenditure (REE), Total Energy Expenditure (TEE) and Activity Energy Expenditure (EE) of subjects with a LC-FAOD compared to controls.

REE (A) and TEE (B) are lower in subjects with a LC-FAOD compared to age, sex and BMI matched published control data (n=192). Control subjects are in blue, subjects with a LC-FAOD in orange. The best fit line (solid line) and 95% confidence interval (dotted lines) are illustrated for FAOD (orange) and controls (blue). A) The adjusted R² for the model is 0.97. Age, sex, body weight (BW) and group (FAOD or control) were significant effects in the model (p<0.001). There was a 61 ± 10 kcal difference between FAOD and controls. B) The adjusted R² for the model is 0.94. Sex, BW (p<0.001) and group (p=0.0015) were significant effects in the model. There was a 101 ± 33 kcal difference between FAOD and controls. C) Activity Energy Expenditure was not different between groups. The adjusted R² for the model is 0.81. BW (p<0.001) was but age (p=0.5) and group (FAOD or control p=0.14) were not significant effects in the model.