| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Department of Medicine Obesity Research Center 1090 Amsterdam Avenue New York, NY 10025 E-mail: SBH2{at}COLUMBIA.EDU
Dear Sir:
Resting energy expenditure (REE) in humans is closely linked with body mass (1). A rise of 
10% or more above a predicted REE for body mass, age, and other predictor variables represents a hypermetabolic state (2). A corresponding decrease of 10% or more in resting metabolic rate represents a hypometabolic state.
Both absolute and relative REE elevations are well documented with hyperthyroidism, catabolic injury, overfeeding, fever, and many other conditions (3). Similarly, absolute and relative reductions in REE are recognized in hypothyroidism, semistarvation, and hypothermia. Periodic fluctuations in REE are also observed with variation in menstrual cycle activity (3). REE is thus a dynamic physiologic measure that varies in magnitude either up or down with many common physiologic states and pathologic conditions.
Although early investigators used body surface area to adjust REE for between-individual comparisons, the modern approach is to adjust REE for metabolically active tissue (1, 4). The usual compartment selected is fat-free body mass (FFM), although other approaches are recognized. Measured REE is first regressed against FFM for a group of fasting subjects and this establishes the basic relation between resting thermogenesis and body mass (4). Additional potential predictors of REE are then entered into the model, and usually others such as sex, fat mass, and age are found significant (4). However, the contribution of the variables other than FFM to REE prediction is very small and large subject samples (eg, several hundred subjects) are often required to achieve statistical significance. Hyper- or hypometabolism would presumably be established after accounting for REE variance secondary to these recognized physiologic predictors.
García Luna et al (5) in their letter examine the relation between REE and body mass in a heterogeneous group of 85 HIV-positive patients. The patients ranged from severely underweight to obese [body mass index (in kg/m2) range: 14–31] and hypometabolic to hypermetabolic (REE ranged from 85% to 143% of predicted based on height, weight, and age). The authors explore the possibility that REE is significantly correlated with plasma viral load, a finding that would suggest that energy expenditure is increased as part of the host response to viral infection. Studies examining this question have produced mixed results and García Luna et al did not detect a significant relation between REE and CD4 cell counts or plasma HIV RNA concentrations. However, the authors did not provide information on how REE prediction models were developed. Was FFM used as an independent variable representing metabolically active tissue? Were other covariates, such as sex and age, statistically significant in developed REE models? Was the magnitude of observed associations between REE and other predictor variables (eg, R2) similar to that reported by other investigators? Was the sample size adequate to detect the hypothesized difference?
We agree with García Luna et al that our understanding of the underlying mechanisms of widely observed hyper- and hypometabolism in HIV-positive patients remains an incompletely understood and fascinating problem. Because HIV is a prevalent infection, the opportunity remains to unravel the many and complex REE determinants in large scale, carefully executed, prospective studies. The information they could provide will give new insights into the many complex and interacting factors that determine REE and human energy requirements.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
