Response to: "Can Racial Differences in Resting Metabolic Rate be Explained by Body Composition?".

Abstract

Listen

Translate article

We are writing in response to the letter related to our original article “Resting Metabolic Rate Varies by Race and by Sleep Deprivation” (1). We appreciate that Avila et al. found our work to be of high interest. We would like to address the three concerns raised in their letter to the editor: (1) lack of adjustment of lean mass; (2) differences in baseline characteristics of the control and sleep-restricted groups; (3) smaller control group. We agree with Avila et al. that lean mass is an important contributor to resting metabolic rate. Indeed, for this reason, we entered fat-free mass as a covariate in our mixed-model ANOVA (Results section, Resting metabolic rate subheading, paragraph 2, page 2353). Fat-free mass was a significant covariate (P < 0.001), and we observed a significant main effect of race such that African Americans exhibited a lower resting metabolic rate than Caucasians. In response to Avila et al.'s concern regarding the role of lean mass, we conducted a one-way ANOVA with resting metabolic rate as a function of fat-free mass (mean resting metabolic rate/mean fat-free mass) as the dependent variable, gender and race as independent variables, and age as a covariate. We observed significant main effects of race (P = 0.02) and gender (P < 0.001) in this analysis of our sample (n = 47), further supporting our conclusion that resting metabolic rate varies by race. Avila et al. contend that our control and sleep-restricted subjects significantly differed in terms of resting metabolic rate during the baseline measurement. However, this was not the case; the difference between groups was not statistically reliable (P = 0.58). Furthermore, as indicated in our manuscript, we conducted a mixed-model ANOVA (Results section, Resting metabolic rate subheading, paragraph 1, page 2353) comparing the change in resting metabolic rate across days within subjects and assessed if this change differed between groups (this type of analysis does not compare groups on each day). As noted in our manuscript, we did not observe a change in resting metabolic rate across days within the control group but did observe a significant change in resting metabolic rate across days within the sleep-restricted group. Finally, Avila et al. questioned our randomization technique and wondered why our control group had fewer subjects than our sleep-restricted group. It is common practice for sleep restriction protocols to use weighted randomization (2, 3) whereby fewer subjects are randomized to the control group than the sleep-restricted group. Larger sample sizes are needed for the sleep-restriction condition in order to account for individual differences manifested in response to sleep restriction in neurobehavioral and energy balance measures (4, 5). Thank you again for your interest in our work. We hope this letter has addressed your concerns and clarified any misunderstandings arising from our original manuscript.