European ancestry and resting metabolic rate in older African ... - Nature

European Journal of Clinical Nutrition (2011) 65, 663–667

& 2011 Macmillan Publishers Limited All rights reserved 0954-3007/11 www.nature.com/ejcn

ORIGINAL ARTICLE

European ancestry and resting metabolic rate in older African Americans TM Manini1, KV Patel2, DC Bauer6, E Ziv8, DA Schoeller3, DC Mackey4, R Li7, AB Newman5, M Nalls9, JM Zmuda5 and TB Harris2 for the Health, Aging and Body Composition Study 1

Department of Aging and Geriatric Research, University of Florida, Institute on Aging, Gainesville, FL, USA; 2National Institute on Aging, Laboratory of Epidemiology, Demography and Biometry, Bethesda, MD, USA; 3Department of Nutritional Sciences, University of Wisconsin, Madison, WI, USA; 4San Francisco Coordinating Center, California Pacific Medical Center, San Francisco, CA, USA; 5 Department of Epidemiology, University of Pittsburgh, Pittsburgh, PA, USA; 6Departments of Medicine and Epidemiology & Biostatistics, University of California, San Francisco, CA, USA; 7Office of Population Genomics, National Human Genome Research Institute, Bethesda, MD, USA; 8Departments of Medicine and Epidemiology and Biostatistics, Helen Diller Family Comprehensive Cancer Center and Institute for Human Genetics, University of California, San Francisco, CA, USA and 9National Institute on Aging, Laboratory of Neurogenetics, Bethesda, MD, USA

Background/Objectives: Resting metabolic rate (RMR) contributes 60–80% of total energy expenditure and is consistently lower in populations of African descent compared with populations of European populations. Determination of European ancestry (EA) through single nucleotide polymorphism (SNP) analysis would provide an initial step for identifying genetic associations that contribute to low RMR. We sought to evaluate the association between RMR and EA in African Americans. Subjects/Methods: RMR was measured by indirect calorimetry in 141 African American men and women (aged 74.7±3.0 years) enrolled in a substudy of the Health, Aging and Body Composition Study. Ancestry informative markers were used to estimate individual percent EA. Multivariate regression was used to assess the association between RMR and EA after adjustments for soft tissue fat-free mass (STFFM), fat mass, age, study site, physical activity level and sex. Results: Mean EA was 23.8±16% (range: 0.1–70.7%) and there were no differences by sex. Following adjustments, each percent EA was associated with a 1.6 kcal/day (95% Confidence interval: 0.42, 2.7 kcal/day) higher RMR (P ¼ 0.008). This equates to a 160 kcal/day lower RMR in a population of completely African ancestry, with one of completely European ancestry. Additional adjustment for trunk STFFM that partially accounts for high-metabolic rate organs did not affect this association. Conclusions: EA in African Americans is strongly associated with higher RMR. The data suggest that population differences in RMR may be due to genetic variants.

European Journal of Clinical Nutrition (2011) 65, 663–667; doi:10.1038/ejcn.2011.22; published online 6 April 2011 Keywords: admixture; energy metabolism; body composition; genetic mapping

Introduction Resting metabolic rate (RMR) is lower in African Americans when compared with individuals of European descent (Weyer et al., 1999; Gallagher et al., 2006). Adjustments for fat-free mass and environmental factors such as socioeconomic status do not completely explain this finding, and thus it is thought to result from genetic variants that are

Correspondence: Dr TM Manini, Department of Aging and Geriatric Research, University of Florida, Institute on Aging, 210 E. Mowry Road, 32611, Gainesville, FL, USA. E-mail: [email protected] Received 16 August 2010; revised 13 January 2011; accepted 13 January 2011; published online 6 April 2011

specific to the African American population (Fernandez et al., 2003). Considering that RMR contributes between 60 and 80% to total energy expenditure, it is important to know the contributions to RMR to increase our understanding of what drives differences in total energy expenditure. Additionally and although controversial, the low RMR found in African Americans may be a potential risk factor for the increased prevalence of obesity among this population (Weyer et al., 1999; Luke et al., 2006). New techniques have been developed to assess whether ethnically associated phenotypes have a genetic basis. One such technique takes advantage of the linkage disequilibrium found among admixed populations deriving their ancestry from separate ancestral populations. The admixed population contains regions of each chromosome from each

European ancestry and metabolism TM Manini et al

664 ancestral population that can be used to identify specific chromosomal loci. This approach is an efficient method to map complex health conditions that have a genetic origin (Patterson et al., 2004; Reich and Patterson, 2005). African Americans residing in the United States are a prime example of an admixed population, with ancestry from Africans and Europeans. As such, this admixed population can provide initial information on the possibility of having a genetic link with RMR. The purpose of this study was to determine the association between European ancestry (EA) and RMR within selfidentified African Americans. On the Basis of previous studies comparing European and African Americans (Weyer et al., 1999; Gallagher et al., 2006), we hypothesized that EA would be positively associated with RMR independent of adjustments for demographics and body composition in men and women.

likelihoods of individual ancestry from all of the markers was maximized as a function of that person’s ancestry, allowing ancestry to range between 0 and 1 (Chakraborty et al., 1986). Using STRUCTURE (Pritchard et al., 2000), we determined a 2-population model, where the proportions of African and European ancestry for each individual sum to 1, was the most appropriate. Tests of linkage disequilibrium among markers were carried out using EMLD, September 2003 Version (MD Anderson Cancer Center, Houston, TX, USA). As the markers are physically unlinked, then linkage disequilibrium between pairs of these markers should be due to heterogeneity in individual ancestry, which provides a complimentary analysis of evidence of genetic substructure and recent admixture (Choudhry et al., 2006). Individuals reporting European Descent were included for descriptive purposes (Table 1).

Subjects and methods

Resting metabolic rate and physical activity level RMR was measured by indirect calorimetry on a Deltatrac II respiratory gas analyzer (Datex Ohmeda Inc., Helsinki, Finland), detailed procedures are described elsewhere (Blanc et al., 2004). In a fasting state and after 30 min of rest, a respiratory gas exchange hood was placed over the participant’s head and RMR was measured minute-by-minute for 40 min. For calibration purposes, methanol burn tests were carried out in duplicate once or twice per month. Carbon dioxide recovery averaged 100.1±1.4% at the Pittsburgh site and 100.5±1.5% at the Memphis site. The gas-exchange ratios for methanol differed by 2.5% between sites (Memphis: 0.66±0.01, Pittsburgh: 0.68±0.01, Po0.001), and this difference did not demonstrate a trend over time. Therefore, a correction factor was used to equate the two study sites by dividing the respiratory ratios for participants enrolled at Pittsburgh by 1.025. The average day-to-day reliability of the Deltatrac II respiratory gas analyzer has been reported to be 4±74 kcal/day with a coefficient of variation of 3.6% (Cooper et al., 2009). Analyses were also performed to describe physical activity levels using the combination of RMR and total energy expenditure (TEE) (PAL ¼ TEE/RMR and Activity energy expenditure (AEE) ¼ (TEE*0.9)–RMR: thermic effect of food was assumed to be 10% of TEE in the equation). Using nonradioactive isotopes of H2 and O18 (i.e., the doubly-labeled water method), we determined TEE over a 14-day period before measurements of RMR. The methods for the doubly labeled water protocol have been described in detail in the following three publications by this group of authors (Blanc et al., 2004; Manini et al., 2006, 2009b).

Study sample In 1997–1998 investigators from the University of Pittsburgh and University of Tennessee, Memphis recruited 3075 participants aged 70–79 from a random sample of white Medicare beneficiaries and all age eligible self-identified black community residents to participate in the Health, Aging and Body Composition (Health ABC) study (Manini et al., 2009a). An energy expenditure substudy subsequently enrolled a random sample of 323 participants from the entire cohort (n ¼ 92 in 1998, n ¼ 125 in 1999 and n ¼ 85 in 2000). In all, 21 participants were excluded from this analysis because of failure to complete the protocol or failure of RMR data to meet a priori quality control criteria or poor quality genetic data, leaving an analytic sample of 279 participants (156 European Americans and 141 African Americans). Written informed consent was obtained from each participant. The authors declare no conflicts of interest with the research.

European ancestry The methods used to determine EA in 141 individuals who self-identified as Black or African–American is detailed in two recently published articles by our group (Shaffer et al., 2007; Wassel Fyr et al., 2007). The subjects were genotyped, and 37 of these genes with ancestry-informative genetic markers were used to determine EA. These genes are known to differ in minor allele frequency by 0.6 or more (mean difference ¼ 0.79) between European and African population, and the specific analyses to determine EA are described previously (Shaffer et al., 2007). These markers occur at spacing of at least 20 centimorgans across chromosomes 1 to 22, and are not in linkage disequilibrium in either parent population. A maximum likelihood approach was used to estimate genetic EA where the likelihood of ancestry based on each individual’s genotype was assessed using the ancestral population allele frequencies. The sum of the log European Journal of Clinical Nutrition

Body mass and composition Fat mass (FM), soft tissue fat-free mass (STFFM) and trunk STFFM were measured using a Hologic 4500A Dual-Energy X-ray Absorptiometry Scanner (Hologic Inc., Watham, MA, USA). Body composition analysis was carried out using HOLOGIC

European ancestry and metabolism TM Manini et al 0.373 0.331 0.061 0.896 0.718 0.880 0.436 0.649 0.659 0.130 0.133 0.841 o0.001 (2.7) (42.5) (57.4) (5.2) (15.9) (0.09) (10.5) (10.2) (4.9) (0.74) (340) (215) (10.3) 75.1 20 27 27.8 77.3 1.66 47.9 26.6 23.4 1.6 660 1243 42.9 (2.9) (53.2) (40.4) (4.8) (14.2) (0.08) (9.2) (9.9) (4.1) (0.63) (316) (205) (3.7) 74.2 25 19 27.7 77 1.67 49.9 24.8 24.1 1.7 743 1264 20.3 (3.3) (57.4) (34.0) (5.4) (16.8) (0.09) (10.2) (10.0) (4.7) (0.61) (266) (201) (4.7) 74.7 27 16 28.2 79.5 1.67 50.5 26.2 24.2 1.4 616 1241 8.1 b

a

Values are expressed as number (percentage) or mean (s.d.) unless otherwise stated. Trunk soft tissue fat-free mass determined for 153 European American. c Physical activity level calculated as the ratio of total energy expenditure and resting metabolic rate. d Activity energy expenditure calculated as AEE ¼ (TEE*0.9)–RMR.

0.395 0.733 0.611 0.053 0.134 0.805 0.025 0.846 0.950 0.494 0.993 0.048 — (3.0) (48.9) (43.9) (5.1) (15.6) (0.09) (9.9) (10.0) (4.6) (0.67) (311) (206) (16.0) 74.7 69 62 27.9 77.9 1.66 49.5 25.8 23.9 1.5 673 1249 23.8 74.9 (2.9) 80 (51.3) 74 (47.4) 26.8 (4.7) 75.3 (15.3) 1.67 (0.09) 46.8 (10.2) 26.1 (8.2) 23.8 (4.9) 1.5 (0.67) 673 (267) 1300 (233) — Characteristics Age, years Women, no (%) From Pittsburgh, no (%) Body mass index, kg/m2 Body mass, kg Body height, m Soft tissue fat-free mass, kg Fat mass, kg Trunk soft tissue fat-free mass, kgb Physical activity levelc Activity energy expenditure, kcal/dayd Resting metabolic rate (kcal/day) European ancestry

Tertile 1 European ancestry (n ¼ 47) P-value African American (n ¼ 141) European American (n ¼ 156)

Table 1 Baseline characteristics of the participants stratified by race and by tertiles of European ancestry in African Americansa

Tertile 2 European ancestry (n ¼ 47)

Tertile 3 European ancestry (n ¼ 47)

P-value

665 software (version 8.21; Hologic Inc.). Calibration was performed three times per week using whole-body qualitycontrol phantoms outlined in the Hologic manual. Values of STFFM were calculated after removing mass due to bone mineral content (BMC) using the equation (FFM þ BMC)–BMC ¼ STFFM. Organ masses are known to contribute disproportionally to RMR and partially explain racial differences in RMR (Gallagher et al., 2006). To adjust for this potential confounder, trunk STFFM from DualEnergy X-ray Absorptiometry was used as a proxy for organ masses and added to the model. Justification for using this proxy originates from the strong associations seen with organ mass determined from magnetic resonance imaging (Bosy-Westphal et al., 2004).

Data analysis Baseline comparisons were performed between European American and African American, and in African Americans across tertiles of EA (Table 1). Baseline participant characteristics were evaluated using analysis of variance for continuous variables and the w2 statistic for categorical variables. Multivariate linear regression was used to examine the adjusted associations of percent EA with RMR. We first tested whether sex and study site (Pittsburgh or Memphis) influenced the relationship between RMR and EA using a formal interaction test. We then examined EA in models adjusted for STFFM, FM, age, sex, AEE and study site. A separate model was created by adding trunk STFFM in an attempt to adjust for the effect of internal organ tissue mass.

Results Descriptive characteristics of European and African Americans enrolled in the energy expenditure substudy are listed in Table 1. At baseline, similar proportions of European Americans and African Americans originated from the study sites. The race groups had a similar age, sex composition, body mass, body height, FM, trunk STFFM and physical activity and AEE levels. STFFM was higher (P ¼ 0.025), body mass index was lower (P ¼ 0.053) and unadjusted RMR was higher (P ¼ 0.048) in European Americans. Mean levels of EA were 23% (median: 20.1%) and did not differ between sex. No demographic differences were noted across varying levels of EA categorized into tertiles listed in Table 1. We found no influence of sex (P ¼ 0.836) or study site (P ¼ 0.135) on the association between RMR and EA. Following multivariate adjustment, EA was significantly associated with RMR (Table 2). The model suggests that for every one percent EA there is a 1.6 kcal/day higher RMR and the adjustment for trunk STFFM had little influence on this relationship (Table 2). European Journal of Clinical Nutrition


666 Table 2 Association between resting metabolic rate (kcal per day) and percent European ancestry in African Americans (n ¼ 141) Model 1 (95% Confidence Interval) Intercept

Model predictors % European ancestry Soft tissue fat-free mass (kg) Fat mass (kg) Age (years) Sex (male ¼ 1, female ¼ 2) Study site (1 ¼ Pittsburgh, 2 ¼ Memphis) Activity energy expenditure (kcal/day) Trunk soft tissue fat-free mass (kg)

1128 (592, 1664)

1.6 18.6 0.08 11.8 31.3 8.5 0.02

(0.41, 2.7) (15.1, 21.1) (2.4, 2.6) (17.9, 5.7) (41.7, 104.3) (29.3, 46.4) (0.08, 0.04) NA

P-value

o0.001

0.008 o0.001 0.953 o0.001 0.398 0.656 0.529

Model 1 þ adjustment for trunk soft tissue fat-free mass b (95% Confidence Interval) 1137 (586, 1688)

1.6 17.9 0.08 12.0 30.7 8.0 0.02 1.3

(0.39, 2.7) (6.7, 30.5) (2.5, 2.6) (18.1, 5.7) (43.4, 104.8) (30.8, 46.7) (0.08, 0.04) (21.1, 23.6)

P-value

o0.001

0.009 0.002 0.948 o0.001 0.414 0.684 0.547 0.911

Note: Activity energy expenditure calculated as (TEE*0.9)–RMR, where TEE is the total energy expenditure determined using doubly-labeled water and RMR is the resting metabolic rate determined by indirect calorimetry.

Discussion Percent EA was estimated in 141 black men and women to determine whether RMR was associated with genetic variants. The data suggest a strong and consistent positive association between EA and RMR where higher RMR was associated with greater EA in African Americans. Although the estimates for EA differed between the two sites in the Health ABC Study, both estimates are consistent with previous reports where average EA values ranged from 17–30% (Parra et al., 1998; Smith et al., 2004; Reiner et al., 2005). This study supports and expands current knowledge on the association between EA and RMR. Ferna´ndaz and colleagues evaluated 145 African American women from three geographical locations in the United States and found a trend for an association between EA and RMR (Fernandez et al., 2003). Interestingly, the slope between RMR and EA was strikingly similar to the result reported here (slope ¼ 1.78), despite differences in study design by Fernandez et al.—for example being restricted to women only compared with women and men residing in different geographical regions. The results also support a wealth of information on reduced RMR in African Americans, as compared with European Americans. For example, the sleeping metabolic rate of African Americans has been reported at 85–100 kcal/day lower than European Americans (Weyer et al., 1999). We found a 1.6 kcal/day higher RMR for each percent EA, and thus a theoretical 160 kcal/day difference between a population of completely African ancestry and one with a completely European ancestry. The association between RMR and EA is thought to originate from selective environmental pressures on mitochondria (Wallace, 1994; Mishmar et al., 2003). Mitochondrial oxidative phosphorylation serves to generate ATP, reoxidize NADH and FADH2, and regulate temperature through heat production. These functions are partially dictated by the coupling state of the mitochondria European Journal of Clinical Nutrition

where uncoupling causes heat production (Wijers et al., 2008). Lower RMR per unit STFFM has a theoretical advantage of maximizing the conversion of energy to ATP rather than heat. Mitochondrial mutations brought on by higher latitude environmental pressures may have initiated compensatory mechanisms for developing a higher RMR through higher rates of mitochondrial uncoupling in Europeans (Wallace, 1994). The higher RMR per unit STFFM also provides a method to compensate for additional energy intake in times of caloric abundance, a potential benefit in the obesity era not shared by African Americans. An alternative explanation was offered by recent studies, suggesting that racial differences in RMR could be partially or mostly explained by differences in fractional mass of highmetabolic rate organs (e.g., brain, liver, kidney, heart and spleen and so on) (Hunter et al., 2000; Gallagher et al., 2006). Gallagher and colleagues found that African Americans had a RMR that was 102 kcal/day lower than European Americans. However, after adjustment for the relative size of highmetabolic rate organs measured with MRI, they found that racial difference disappeared. In a follow-up study by the same group of investigators, Javed et al. found that the addition of brain volume explained B2% of the variance in RMR even after accounting for total STFFM (Javed et al., 2010). These data suggest that African Americans had a smaller proportion of STFFM as high-metabolic organs and thus previous reports that used total STFFM did not adequately account for this difference in the relative proportion between African American and European Americans. Trunk STFFM in our sample was similar across all comparison groups, which differs from the findings of Hunter et al. who found lower trunk STFFM in premenopausal African than European women (Hunter et al., 2000). With the known age-related decline in organ mass (He et al., 2009), the older age of our sample may have erased the ethnic differences in organ mass seen in previous studies. Nevertheless, statistical correction for trunk STFFM did not


667 reduce the association between RMR and EA in our sample. Unfortunately, organ-specific masses were not available to conduct more thorough statistical corrections to RMR. In conclusion, EA is strongly associated with RMR in African Americans after accounting for differences in body composition and demographics.

Conflict of interest The authors declare no conflict of interest.

Acknowledgements This study was supported by The NIA Claude D Pepper Center P30AG028740 and a grant from the Institute on Aging at the University of Florida. The Health, Aging and Body Composition Study was supported by the Intramural Research Program of the NIH, National Institute on Aging contracts N01-AG-6-2106, N01-AG-6-2101 and N01-AG-62103 with additional support from the National Institute of Diabetes and Digestive and Kidney Diseases. The National Institute on Aging Intramural Research Program designed the Health ABC study, supervised its conduct and participated in data collection. The coauthors, The Health, Aging, and Body Composition publications committee and representatives from The National Institute on Aging reviewed and approved the manuscript.

Disclaimer This article was prepared while Dr R Li was employed at the University of Tennessee. The opinions expressed in this article are the author’s own and do not reflect the views of the Department of Health and Human Services.

References Blanc S, Schoeller DA, Bauer D, Danielson ME, Tylavsky F, Simonsick EM et al. (2004). Energy requirements in the eighth decade of life. Am J Clin Nutr 79, 303–310. Bosy-Westphal A, Reinecke U, Schlorke T, Illner K, Kutzner D, Heller M et al. (2004). Effect of organ and tissue masses on resting energy expenditure in underweight, normal weight and obese adults. Int J Obes Relat Metab Disord 28, 72–79. Chakraborty R, Ferrell RE, Stern MP, Haffner SM, Hazuda HP, Rosenthal M (1986). Relationship of prevalence of non-insulindependent diabetes mellitus to Amerindian admixture in the Mexican Americans of San Antonio, Texas. Genet Epidemiol 3, 435–454. Choudhry S, Coyle NE, Tang H, Salari K, Lind D, Clark SL et al. (2006). Population stratification confounds genetic association studies among Latinos. Hum Genet 118, 652–664. Cooper JA, Watras AC, O’Brien MJ, Luke A, Dobratz JR, Earthman CP et al. (2009). Assessing validity and reliability of resting metabolic rate in six gas analysis systems. J Am Diet Assoc 109, 128–132.

Fernandez JR, Shriver MD, Beasley TM, Rafla-Demetrious N, Parra E, Albu J et al. (2003). Association of African genetic admixture with resting metabolic rate and obesity among women. Obes Res 11, 904–911. Gallagher D, Albu J, He Q, Heshka S, Boxt L, Krasnow N et al. (2006). Small organs with a high metabolic rate explain lower resting energy expenditure in African American than in white adults. Am J Clin Nutr 83, 1062–1067. He Q, Heshka S, Albu J, Boxt L, Krasnow N, Elia M et al. (2009). Smaller organ mass with greater age, except for heart. J Appl Physiol 106, 1780–1784. Hunter GR, Weinsier RL, Darnell BE, Zuckerman PA, Goran MI (2000). Racial differences in energy expenditure and aerobic fitness in premenopausal women. Am J Clin Nutr 71, 500–506. Javed F, He Q, Davidson LE, Thornton JC, Albu J, Boxt L et al. (2010). Brain and high metabolic rate organ mass: contributions to resting energy expenditure beyond fat-free mass. Am J Clin Nutr 91, 907–912. Luke A, Durazo-Arvizu R, Cao G, Adeyemo A, Tayo B, Cooper R (2006). Positive association between resting energy expenditure and weight gain in a lean adult population. Am J Clin Nutr 83, 1076–1081. Manini TM, Everhart JE, Anton SD, Schoeller DA, Cummings SR, Mackey DC et al. (2009a). Activity energy expenditure and change in body composition in late life. Am J Clin Nutr 90, 1336–1342. Manini TM, Everhart JE, Patel KV, Schoeller DA, Colbert LH, Visser M et al. (2006). Daily activity energy expenditure and mortality among older adults. JAMA 296, 171–179. Manini TM, Everhart JE, Patel KV, Schoeller DA, Cummings S, Mackey DC et al. (2009b). Activity energy expenditure and mobility limitation in older adults: differential associations by sex. Am J Epidemiol 169, 1507–1516. Mishmar D, Ruiz-Pesini E, Golik P, Macaulay V, Clark AG, Hosseini S et al. (2003). Natural selection shaped regional mtDNA variation in humans. Proc Natl Acad Sci USA 100, 171–176. Parra EJ, Marcini A, Akey J, Martinson J, Batzer MA, Cooper R et al. (1998). Estimating African American admixture proportions by use of population-specific alleles. Am J Hum Genet 63, 1839–1851. Patterson N, Hattangadi N, Lane B, Lohmueller KE, Hafler DA, Oksenberg JR et al. (2004). Methods for high-density admixture mapping of disease genes. Am J Hum Genet 74, 979–1000. Pritchard JK, Stephens M, Donnelly P (2000). Inference of population structure using multilocus genotype data. Genetics 155, 945–959. Reich D, Patterson N (2005). Will admixture mapping work to find disease genes? Philos Trans R Soc Lond B Biol Sci 360, 1605–1607. Reiner AP, Ziv E, Lind DL, Nievergelt CM, Schork NJ, Cummings SR et al. (2005). Population structure, admixture, and aging-related phenotypes in African American adults: the Cardiovascular Health Study. Am J Hum Genet 76, 463–477. Shaffer JR, Kammerer CM, Reich D, McDonald G, Patterson N, Goodpaster B et al. (2007). Genetic markers for ancestry are correlated with body composition traits in older African Americans. Osteoporos Int 18, 733–741. Smith MW, Patterson N, Lautenberger JA, Truelove AL, McDonald GJ, Waliszewska A et al. (2004). A high-density admixture map for disease gene discovery in african americans. Am J Hum Genet 74, 1001–1013. Wallace DC (1994). Mitochondrial DNA sequence variation in human evolution and disease. Proc Natl Acad Sci USA 91, 8739–8746. Wassel Fyr CL, Kanaya AM, Cummings SR, Reich D, Hsueh WC, Reiner AP et al. (2007). Genetic admixture, adipocytokines, and adiposity in Black Americans: the Health, Aging, and Body Composition study. Hum Genet 121, 615–624. Weyer C, Snitker S, Bogardus C, Ravussin E (1999). Energy metabolism in African Americans: potential risk factors for obesity. Am J Clin Nutr 70, 13–20. Wijers SL, Schrauwen P, Saris WH, van Marken Lichtenbelt WD (2008). Human skeletal muscle mitochondrial uncoupling is associated with cold induced adaptive thermogenesis. PloS One 3, e1777.

European Journal of Clinical Nutrition