T h e ne w engl a nd jour na l o f medicine
Table 1. Characteristics of the Subjects.*
Lean Subjects (N = 10)
Overweight or Obese Subjects (N = 14)
was measured in the small intestine with the use of a telemetric pill (HQ). In three subjects (two lean and one overweight), the core-temperature measurement failed.
BMI Mean Range Body fat (%)‡ Mean Range
Body mass (kg)
Plus–minus values are means ±SD. BMI denotes bodymass index, calculated as the weight in kilograms divided by the square of the height in meters.
P values are based on unpaired ttests.
The percentage of body fat was determined with the use of dualenergy xray absorptiometry.
ware, version 2.85 (PMOD Technologies, May 2007). Brown-adipose-tissue activity (measured in kilobecquerels) and brown-adipose-tissue vol- ume (measured in cubic centimeters) were quan- tified in the region of interest by autocontour- ing the brown-adipose-tissue areas with a set threshold.
Energy expenditure was measured by means of a respiratory gas analyzer with the use of a ventilated hood system (Oxycon, Jaeger). The rest- ing metabolic rate was corrected for fat-free mass by means of linear regression, since 80% of in- dividual differences in energy expenditure can be explained by fat-free mass.18 The residuals of the regression between energy expenditure and fat- free mass were added to the average energy ex- penditure. Skin temperatures were measured con- tinually at 16 positions by means of iButtons (type DS1921H, Maxim).19 The mean skin tem- perature was calculated with a 14-point method (ISO-9886).20 Underarm–fingertip temperature gradients were calculated to obtain an estimate of blood flow in the fingers.21 Core temperature
The protocol for temperature (mildly cold, 16°C) was validated earlier.22 Every 15 minutes, each subject was asked whether he had experi- enced shivering. Body composition was measured with the use of a DEXA whole-body scanner (QDR 4500A, Hologic).
To ascertain whether brown adipose fat in the supraclavicular region could be identified with classic microscopy and an immunofluorescence assay for UCP1, samples were obtained during surgery for a multinodular goiter in a 46-year- old woman with normal thyroid function (body mass, 72 kg; height, 1.72 m; thyroid-stimulating hormone, 1.7 mU per milliliter; and free thyrox- ine, 12.8 pmol per liter [1.0 ng per deciliter]). Bi- opsy samples of both brown and white adipose tissues were obtained from the supraclavicular region on the right side. In this area, both fat types were observed in patches measuring approximate- ly 0.5 cm3. Sections (5 μm thick) were fixed for 15 minutes with 3.7% formaldehyde in phosphate- buffered saline, and UCP1 was detected with the use of rabbit polyclonal human UCP1 antiserum and with goat antirabbit fluorescein isothiocya- nate as the secondary antibody, together with the nuclei-staining dye 4,6-diamidino-2-phenylindole (0.5 mg per milliliter).
Data are reported as means ±SD. Data analysis was performed with the use of SPSS software, version 11, and Statview SE for MacIntosh (SAS Institute). Comparisons were made by means of a paired t-test (for findings under thermoneutral conditions as compared with mild cold exposure) and an unpaired t-test (for findings in lean men as compared with overweight or obese men). Lin- ear regression analyses were conducted to iden- tify correlations between variables. For the rela- tion between brown-adipose-tissue activity and body composition, exponential curve fitting was applied; therefore, log (brown-adipose-tissue ac- tivity) was plotted against body composition (BMI or percentage of body fat). Cook’s distance val- ues that were larger than 0.5 and Studentized deleted residuals larger than 4.0 were considered to be outliers. P values of less than 0.05 were considered to indicate statistical significance.
n engl j med 360;15
april 9, 2009
The New England Journal of Medicine Downloaded from nejm.org on February 3, 2015. For personal use only. No other uses without permission. Copyright © 2009 Massachusetts Medical Society. All rights reserved.