Basal Metabolic Adaptation of the Evenki Reindeer Herders ofCentral Siberia
VICTORIA A. GALLOWAY,1,2* WILLIAM R. LEONARD,3 ANDEVGUENY IVAKINE4
1School of Medicine, University of Toronto, Toronto, Ontario, Canada2Department of Human Biology and Nutritional Sciences, University ofGuelph, Guelph, Ontario, Canada3Department of Anthropology, Northwestern University, Evanston, Illinois4Department of Immunology, University of Toronto, Toronoto,Ontario, Canada
Basal metabolic rate (BMR) is the energycost for the chemical processes essential forthe maintenance of life, including the pres-ervation of body integrity and homeostasis.Basal metabolism is comprised of the en-ergy costs of cardiac activity, thermoregula-tion, sympathetic nervous system activity,
Contract grant sponsor: the Natural Sciences and EngineeringResearch Council of Canada; Contract grant number: OGP-0116785.
*Correspondence to Victoria A. Galloway, 30 Charles St. W.Apt. 1607, Toronto, ON, M4Y 1R5, Canada. E-mail:[email protected]
Received 22 December 1997; Revision received 5 January1999; Accepted 26 January 1999
sodium-potassium ATPase activity, calciumpumping, and protein turnover (Ulijaszek,1992, 1996; Waterlow, 1988). The BMR isthe largest single component of total dailyexpenditure (TDEE), comprising between65–75% in sedentary humans (Jequier,1987; Soares and Shetty, 1991). Since it issuch a strong determinant of daily energyneeds, variation in basal metabolism is ofconsiderable interest to human biologistsand nutritionists.
Ethnic variation in BMR has been widelystudied, with BMRs of circumpolar popula-tions receiving particular attention (Itoh,1980; Roberts, 1978). Early studies reportedthat BMRs of traditional Alaskan Inuitwere elevated compared to Dubois (1927)standards (Rodahl, 1952) and to NorthAmerican Indians and Europeans (Milanand Evonuk, 1967). These researchers pro-posed that such metabolic elevations re-flected either an adaptation to high latitudestressors (e.g., extreme climate, a diet highin protein and fat), or was attributable toanxiety. More recently, Rode and Shephard(1995) reported that BMRs of the IgloolikNWT Inuit were significantly higher thanthose of reference standards (Consolazio etal., 1963; Schofield, 1985a,b) and Europeancontrols. These authors postulated a geneticmechanism for the high BMRs which maycontinue to be modified with increasing ac-culturation. Overall, then, the research todate suggests that indigenous high latitudepopulations may display a basal metabolicadaptation to severe, chronic, climaticstress.
Previous work among the Evenki reindeerherders of Siberia has indicated that theirresting metabolic rate (RMR) was higherthan predicted for their size (Katzmarzyk etal., 1994, 1996; Leonard et al., 1996). How-ever, it was unclear whether the elevationsin metabolic rate were attributable to short-term thermal stress or a longer-term devel-opmental and/or genetic adaptation. Thus,the purpose of this study is to examine evi-dence for a basal metabolic adaptation (i.e.,elevated BMR) in this indigenous Siberianpopulation. First, BMRs of the Evenki arecompared with those of Russian migrants(“controls”) living in the same communities.Next, the observed BMRs of both the Evenkiand Russians are compared to those pre-dicted from body weight and age using theregression equations developed by Schofield
(1985a,b), and body surface area (SA) usingthe norms of Consolazio et al. (1963).
METHODSSample
The Evenki are among the most numer-ous of the indigenous Siberian groups, todaynumbering about 30,000 (Hannigan, 1991).They were once reindeer hunters who, overtime, adopted a breeding and herding sub-sistence. Most of the Evenki live in collec-tivized villages, and during the spring sum-mer months they (especially younger males)leave for the forest in “brigades” to tend tothe reindeer.
Data for this study were collected fromJuly through September of 1995 amongEvenki and Russians residing in the vil-lages of Poligus, Surinda, and Baykit (con-secutively, in that order), Baykit district,Tunguska region of Central Siberia (63°N,97°E) (see Fig. 1). Mean monthly tempera-tures in this region range from −32.5°C inJanuary to +15°C in July.
Anthropometric, dietary, and energy ex-penditure data were collected from a sampleof 58 Evenki and 24 Russian adult volun-teers (ages 18–56 years). The Evenkisample included 39 females and 19 males;the Russian sample included 16 females and8 males. All of the Russian “controls” hadbeen regional residents for more than 10years. All data were collected in the villagehealth posts (three in total). The researchprotocol was approved by the Human Sub-jects Review committee of the University ofGuelph.
AnthropometryAnthropometric dimensions were taken
using standard techniques (Lohman et al.,1988; Frisancho, 1990) by a single re-searcher. Stature was measured to thenearest millimeter using a portable field an-thropometer. Body mass was assessed to thenearest 0.5 kg using a standing scale (SecaCorp., Columbia, MD). Skinfold thicknesses(triceps, biceps, subscapular, suprailiac,periumbilical, medial calf) were measuredto the nearest 0.5 mm using Lange callipers.Derived measurements included the sum ofskinfolds (SOS, mm), the body mass index(BMI, kg/m2), body surface area (SA, m2),percent body fat (%BF, %), and fat-freemass (FFM, kg). Body SA was estimatedbased on height and weight from the Baileyand Briars (1996) equation. Percent fat and
76 V.A. GALLOWAY ET AL.
Fig. 1. (A) Map of Russia showing location of study area; (B) study area showing location of Evenki villages.From Katzmarzyk (1993).
BASAL METABOLISM OF THE EVENKI 77
FFM were both determined with the equa-tions of Durnin and Womersely (1974) basedon four skinfolds (biceps, triceps, subscapu-lar, and suprailiac). Rode and Shephard(1994) compared body density and percentfat derived from these equations to directhydrostatic estimates in a sample of Inuitmales. They reported a 1–3% error in bodyfat attributable to the high muscle/bonemass ratio, and/or the high ratio of internalto superficial body fat of the Inuit.
Basal energy expenditure measuresBMR (Kcal/min) was measured via open
circuit indirect calorimetry using the Aero-Sport TEEM 100 Metabolic Analyzer (Aero-Sport Inc., Ann Arbor, MI). This systemmonitors oxygen consumption (VO2, L/min)and carbon dioxide production (VCO2,L/min). The O2 and CO2 analyzers were cali-brated with external air and tanks of com-pressed gas containing 16.02% O2 and5.03% CO2 (Scott Specialty Gases, Troy,MI). Expired volumes were measured usingpneumotachometers that were calibratedwith a 3 liter syringe (Hans Rudolph, Kan-sas City, MO). Basal energy costs were mea-sured using low flow pneumotach heads.Basal VO2s were converted to kilocalories(Kcal) based on respiratory quotient (RQ)levels using the modified Weir formula(McArdle et al., 1991; Weir, 1949). All BMRmeasurements were done by the same tworesearchers using the same Aerosports andequipment.
The BMR protocol began with the sub-jects arriving at the community health poststhe night before testing to be briefed on theresearch protocol. To avoid potential anxi-ety during future measurement, the sub-jects familiarized themselves with theequipment, i.e., breathing with the mouth-pieces and nose clips (Soares et al., 1989).Information collected from subjects duringthe pretest meeting included: 1) standarddemographic data, 2) a 24-h dietary recall(including time of the subjects’ last meal), 3)smoking habits, and 4) general health his-tory. Blood pressures (BPs, mmHg) weremeasured with a standard sphygmometer tothe nearest 1 mmHg to prepare the volun-teers for the morning’s BP measurementsand to act as a reference value for anxiety.Subjects with BPs greater than 160 diastoleand/or 95 systole (classified as severe hyper-tensive according to WHO, 1978) were ex-cluded from data analysis. Heart rate (HR,
beats/min) data were collected concurrentlyusing Polar Vantage XL HR monitors (PolarElectro, Stanford, CT) to increase the powerof the anxiety referencing.
For the females, the menstrual cyclestage (follicular, luteinizing) and reproduc-tive status (pre-/peri-/postmenopausal/lactating/pregnant) were determined viaprotocol outlined by Curtis et al. (1996) andas outlined by WHO (1980), respectively.Each menstrual stage was representedequally by females from both ethnic groups.A total of eight Evenki females were lactat-ing, and two Evenki and four Russian fe-males were postmenopausal. The sampledid not contain pregnant or perimenopausalfemales.
Subjects were fitted with HR monitorsand they slept in the health posts overnighton cots with blankets. The next morning,the machines were calibrated and each sub-ject was awakened sequentially between6:30 and 7:00 AM for BMR monitoring. Atthe time of measurement, all subjects hadbeen in a fasted state for at least 10 h (mean4 11.6 ± 0.16 h). Prior to the testing, BPs,HRs, and forehead skin temperatures to thenearest 0.5°C were recorded with a stan-dard skin thermometer. Barometric pres-sure was also measured with a barometer,Model AP8110 (Geneq, Montreal, QUE), aswas air temperature with a standard ambi-ent thermometer by Bravo Sharp (HannaInstruments, Woonsocket, RI). When neces-sary, a portable space heater was used toheat the room to ensure the thermoneutralconditions. The mean ambient temperaturewas 23.1 ± 0.16°C, well within the rangespreviously found adequate by other re-searchers (Beall et al., 1996; Kashiwazaki,1990; Katzmarzyk et al., 1996; Rode andShephard, 1995). VO2, VCO2 and HR weremonitored continuously for 8 min, then av-eraged, after VO2, RQ, and HR readings sta-bilized. BMRs were later converted to Kcal.
Statistical methodsAll data were tested for normality using
the Kolmogrov-Smirnov goodness of fit test.For each variable, normality was tested atseveral levels: the entire sample, by ethnic-ity, by sex and ethnicity, by age and ethnic-ity, and by location and ethnicity. None ofthe measures showed significant deviationsfrom a normal distribution at the 0.05 sig-nificance level. One-way analyses of vari-ance and Tukey’s pairwise comparisons of
78 V.A. GALLOWAY ET AL.
means were used to determine group differ-ences in air temperature, fasting time,smoking status, menstrual/reproductivestatus, skin temperature, RQ, BP, and HR.If needed, analyses of covariance were usedto examine variation of these methodologi-cal parameters (as well as preliminaryanalyses of dietary protein) on BMR amonggroups.
Between group differences in anthropo-metric and metabolic parameters were as-sessed using Student’s t-tests (two-tailed).To test elevations in metabolic rate, pairedt-tests (two-tailed) were used to assess thedifferences between observed and predictedBMRs. Analyses of covariance were used tonormalize BMR data by unit body mass andFFM in addition to using the ratio method(Norgan, 1996; Poehlman and Toth, 1995).Predicted BMR were determined based ontwo standards, the Schofield (1985a,b)norms which estimate BMR from bodyweight based on age- and sex-specific equa-tions, and the Consolazio et al. (1963) normswhich estimate BMR per unit SA based onage and sex-specific equations (as presentedin Rode and Shephard, 1995). The Schofieldnorms have become the accepted standardfor estimating BMR in human nutrition re-search. According to the author, surfacearea did not make a significant difference inthe accuracy of the weight-derived predic-tive equations (FAO/WHO/UNU, 1985;Schofield, 1985a,b). In contrast, the Conso-lazio norms, and the convention of expres-sion BMRs relative to SA, are more commonamong physiologists and were the standardin earlier BMR research. Consequently,both comparisons are included to facilitatecomparisons with other studies of metabolicvariation. Age-related differences in meta-bolic rates were examined using correla-tional analysis.
All statistical analyses were tested to the
0.05 significance level and were done withthe Statistical Package for the Social Sci-ences (SPSS.PC) version 3.0 (Norusis,1988).
RESULTS
Menstrual/reproductive status (in women),fasting time, skin temperature, RQ, BP, andsmoking status are comparable between allgroups. HR is elevated in Evenki females vs.males, but are not significant predictors ofBMR variation. Mean air temperatures arenot comparable overall between the Poligusand the Surinda and Baykit locations (23.94vs. 22.37, 22.85°C, respectively); Evenki be-tween locations Poligus and Surinda (23.93vs. 22.37°C); Evenki females between loca-tions Poligus and Surinda (24.05 vs. 22.42°C);and Russian females between locations Poli-gus and Baykit (24.33 vs. 22.50°C). Of these,only air temperature between the regions ofPoligus and Surinda for the Evenki femalessignificantly account for the differences in ab-solute BMR (P 4 0.013) and BMR per unitFFM (P 4 0.004). However, location remainsa significant contributor to BMR variation (P4 0.013 and P 4 0.033, respectively) and,therefore, air temperature does not fully ac-count for the BMR differences between thetwo female location cohorts.
Table 1 compares the anthropometriccharacteristics between the Evenki andRussian subjects. On average, the Evenkiare significantly shorter (males by 14.7 cm,females by 7.7 cm) and lighter (males by20.5 kg, females by 19.6 kg) than their Rus-sian counterparts (P < 0.001). Even whenweight is adjusted for stature using theBMI, the Russians are significantly heavierthan the Evenki males (24.6 vs. 21.9 kg/m2;P < 0.001) and females (28.4 vs. 22.7 kg/m2;P < 0.01). The Evenki have significantly lessFFM (P < 0.001) and adipose tissue (P <0.01) than the Russians.
TABLE 1. Age and anthropometry (means and standard deviations) of Evenki and Russian men and womenfrom the Baykit district of Central Siberia*
Table 2 compares selected basal metabolicmeasures between Evenki and Russian menand women. For both sexes, mean BMRs(Kcal/day) are higher in the Russians thanin the Evenki. The differences are statisti-cally significant in males (1,896 [7.93 MJ]vs. 1,585 [6.63 MJ]; P < 0.05), but not fe-males (1,359 [5.69 MJ] vs. 1,274 [5.33 MJ]).To control for ethnic differences in body sizeand composition, BMR is also expressed perunit surface area, body weight, and FFM.This “ratio” method of expressing BMRs haslimitations compared to performing regres-sion/covariation analyses, since the formerassumes an isometric relationship betweenBMR and body mass and does not accountfor the Y-intercept (Norgan, 1996; Poehl-man and Toth, 1995). Average BMR per sur-face area is higher in the Evenki than in theRussians. The difference is significant inwomen (36.5 vs. 32.6 Kcal/m2/h; P 4 0.03),but not in men (41.8 vs. 41.1 Kcal/m2/h).Similar patterns are seen when BMR is ex-pressed per unit kg of body weight, 28.6 vs.25.3 Kcal/kg/d for males and 25.6 vs. 20.2Kcal/kg/d for females (P 4 0.001), and perunit kg fat-free body mass 33.3 vs. 32.3Kcal/kg/d for males and 36.6 vs. 31.6Kcal/kg/d for females (P 4 0.008). Overall,Evenki women show significantly highermetabolic rates than their Russian counter-parts, whereas the differences are lessmarked between Evenki and Russian men.These results are confirmed by covariationanalysis.
Table 2 also compares measured BMRsrelative to those predicted based on bodyweight and age and body SA. Relative to theSchofield predictions, both Evenki and Rus-sian men display modest, but not signifi-cant, elevations in BMR. As shown in Fig-
ure 2, Evenki and Russian men deviatefrom their predicted BMRs by +3.9% and+6.0%, respectively. Among the women,measured BMRs are slightly higher thanpredicted in the Evenki (+2.7%), but are be-low predicted levels in the Russian sample(−5.0%). Relative to the Consolazio SAnorms, Evenki and Russian men showmarked elevations in metabolic rate. Asshown in Table 2 and Figure 3, BMR/SA(Kcal/m2/h) among Evenki men is signifi-cantly higher than predicted (+11.8%; 41.8vs. 37.4 Kcal/m2/h; P < 0.05). For Russianmen, the difference between observed andpredicted BMR/SA is similar and but is notstatistically significant (+11.2%; 41.1 vs.37.0 Kcal/m2/h; P 4 0.07). Evenki womenare similar to their male counterparts inhaving metabolic rates significantly el-evated over predicted levels (+5.8%; 36.5 vs.34.5 Kcal/m2/h; P < 0.05). In contrast, Rus-sian women have metabolic rates that areslightly lower than predicted (−3.6%; 32.6vs. 33.8 Kcal/m2/h; n.s.), and significantlylower than their Evenki peers (P < 0.05).
Age differences in BMR
Table 3 presents correlations between se-lected metabolic indices and age for menand women of the Evenki and Russiansamples. Both Evenki and Russian menshow modest reductions in BMR with age;however, none of the correlations are sig-nificant.
In contrast, there is a stronger negativeassociation between metabolic rate and agein females. Among Evenki women, both ab-solute (i.e., Kcal/day) and relative to bodyweight and FFM, BMRs significantly de-cline with age. For example, Evenki femalesages 19–29 (n 4 19) have an average abso-lute BMR of 1,368 Kcal/d, BMR per unit sur-
TABLE 2. Basal metabolic rates (means and standard deviations) of Evenki and Russian menand women from the Baykit district of Central Siberia
Sex (group) nBMR (kcal/day) BMR/SA (kcal/m2/h) BMR/Wt
P2 0.170 <0.001 0.030 0.070 0.001 0.0081Measured vs. predicted values are significantly different at P < 0.05.2P values refer to differences between Evenki and Russians.
80 V.A. GALLOWAY ET AL.
face area of 39.84 Kcal/m2/h, BMR per unitbody weight of 28.5 Kcal/kg/d, and BMR perunit FFM of 38.9 Kcal/kg/d; whereas fe-males ages 30–56 (n 4 20) show 1,185Kcal/d (P < 0.05); 33.34 Kcal/m2/h (P <0.001), 22.9 Kcal/kg/d (P < 0.001); and 34.5Kcal/kg/d (P < 0.001), respectively. Evenkifemales also show deviations from predictedBMR that are negatively correlated with
age, such that subjects under 30 have BMRsthat are above predicted levels (mean 4+13.6%, Schofield norms; +12.1%, Consola-zio norms), whereas those 30 years andolder have BMRs below predicted levels(−7.8 and −0.4%, respectively). By age, theRussian females (n 4 5 ages 18–29; n 4 11ages 30–56) also show significant negativecorrelations between age and BMR per unitsurface area, BMR per unit weight, and per-cent deviation from the Schofield predic-tions, but not in absolute BMR, BMR perunit FFM, and percent deviation from theConsolazio norms.
Residence location and BMRThe Evenki show marked differences in
BMR between residents of the two main in-digenous villages, Surinda and Poligus, theformer being more northerly, remote, andtraditional. The samples from the two com-munities are comparable in age for men(34.1 vs. 36.1 years) and women (30.6 vs.32.2 years). Of the anthropometric dimen-sions, only female stature differs signifi-cantly between the two communities (151.1cm in Surinda vs. 147.0 cm in Poligus; P <0.05). However, analyses show that statureis not a significant source of BMR variationbetween these two female location cohorts.
Table 4 compares metabolic measures be-tween the Evenki of Surinda and Poligus.Despite similarity in overall body size andcomposition, BMRs are significantly higherin the Surinda males (1,740 Kcal [7.28 MJ]vs. 1,333 Kcal [5.58 MJ], P < 0.05), and fe-males (1,394 Kcal [5.84 MJ] vs. 1,142 Kcal[4.78 MJ], P < 0.01). Significant differencesbetween the two groups are more pro-nounced after adjusting BMR for SA (males,46.6 vs. 34.6 Kcal/m2/h [P 4 0.006]; fe-males, 39.1 vs. 33.5 Kcal/m2/h [P 4 0.03]),body mass (32.0 vs. 23.5 Kcal/kg/d [P 40.005]; 27.2 vs. 23.8 Kcal/kg/d [P 4 0.030]),and FFM (37.4 vs. 27.6 Kcal/kg/d [P 40.010]; 39.9 vs. 32.9 Kcal/kg/d [P 4 0.002]).BMRs of individuals from the two villagesalso differ in predicted values (Table 4).BMRs of Evenki men from Surinda are, onaverage, 14.9% above the Schofield esti-mates (P 4 0.065) and 24.9% above theConsolazio values (P < 0.01). In contrast,Poligus men have BMRs that are 13.5% and6.8% below the Schofield and Consolazio es-timates, respectively. The percent devia-tions are also significantly lower than theirSurinda peers (P < 0.01 and P < 0.001, re-
Fig. 2. Percent deviations (±SD) from predicted lev-els based on body weight (Schofield, 1995a,b) for mea-sured BMR (Kcal/d) of Evenki and Russian males andfemales. Both Evenki and Russian men display modest,but not significant, deviations in BMR (+3.9% and+6.0%, respectively). Measured BMRs are slightlyhigher than predicted for the Evenki women (+2.7%),but are below predicted levels for the Russian sample(−5.0%).
Fig. 3. Percent deviations (±SD) from predicted lev-els based on body surface are a (Consolazio et al., 1963)for measured BMR (Kcal/d) of Evenki and Russianmales and females. Evenki and Russian men showmarked deviations in metabolic rate (+11.8% [P < 0.05]and +11.2% [P 4 0.07], respectively). Evenki womenhave metabolic rates significantly elevated over pre-dicted levels (+5.8%; P < 0.05), but the Russian womenshow BMRs that are slightly lower than predicted andsignificantly lower than their Evenki peers (−3.6%; P <0.05).
BASAL METABOLISM OF THE EVENKI 81
spectively). As with the men, Evenki womenfrom Surinda have BMRs that are signifi-cantly higher than predicted by both stan-dards (+10.6% [P < 0.05], to Schofield;+13.0% [P < 0.01], Consolazio). On the otherhand, BMRs of Evenki females from Poligusfall below predicted levels (−6.5% [P < 0.05],Schofield; −2.9%, Consolazio). The Evenkifemales from Poligus also show percent de-viations that are significantly lower thantheir Surinda peers (P < 0.01 and P < 0.05,respectively).
In contrast, Russian females from Poligus(n 4 6) and Baykit (n 4 8) show comparableBMRs, both in absolute terms (1,372 vs.1,325 Kcal/d) and expressed as units of bodymass, FFM, and SA. The Russian femalesample from Surinda was too small foranalysis, as were the Russian male samplesfrom Poligus and Surinda.
DISCUSSION
The purpose of this study is to examineevidence for a basal metabolic adaptation(i.e., elevated BMR) in the Evenki herders.BMRs of the Evenki were compared to thoseof Russian migrants (“controls”) living inthe same communities. The observed BMRsof both the Evenki and Russians were thencompared to those predicted from body
weight and age using the regression equa-tions developed by Schofield (1985a,b), andbody SA using the norms of Consolazio et al.(1963). It was expected that the Evenkiwould have measured BMRs that aregreater than predicted levels. And, it waspredicted that if such metabolic responseshave a genetic component, the Evenkiwould have greater elevations in BMR thanthe Russian “controls” who have been livingin the same communities for at least 10years.
The results indicated BMRs of Evenkimen and women that were absolutely lower(i.e., Kcal/day) than those of Russian coun-terparts living in the same villages. Thesedifferences, however, are attributable tovariation in body size and composition be-tween the two ethnic groups. Consequently,to account for the marked disparity in an-thropometric characteristics between thegroups, analyses of covariance were used tonormalize BMR for ethnic and sex differ-ences in body mass and composition, andthe standard approach of expressing BMRsper unit SA, body mass, and FFM (Butte etal., 1995; Ferraro et al., 1992; Jequier et al.,1987; Norgan, 1996; Poehlman and Toth,1995; Ravussin and Bogardus, 1989; Wein-sier et al., 1992) was followed. After stan-
TABLE 3. Correlations between selected measures of metabolic rate and age for Evenki andRussian men and women
Sex (group) BMR BMR/SA BMR/Wt BMR/FFM %Deviation14 %Deviation25
Differences between measured and predicted BMRs are different at 1P < 0.05, 2P < 0.01.P values refer to the differences between villages.
82 V.A. GALLOWAY ET AL.
dardizing BMR for unit SA or mass, Evenkiand Russian men showed similar metabolicrates, whereas Evenki women had higherBMRs than their Russian peers. The lack ofdifference between male ethnic groups mayin part be due to the small Russian malesample.
To explicitly test for evidence of metabolicadaptation, the measured BMRs of theEvenki and Russian were compared to thosepredicted two sets of norms (Schofield,1985a,b; Consolazio et al., 1963), and toeach other. Relative to the Schofield (bodyweight) norms, mean BMRs of both Evenkiand Russian men were modestly (but notsignificantly) elevated relative to predictedlevels (+3.9 vs. 6.0%). Evenki women, on av-erage, also showed slightly elevated BMRs(+2.7%), whereas Russian women hadBMRs below predicted levels (−5.0%), andthe ethnic difference approached signifi-cance (P 4 0.06).
Compared to the Consolazio SA norms,Evenki and Russian men showed markedelevations in BMR, with those of the Evenkibeing significantly higher than predicted.Evenki women also displayed BMRs thatwere significantly higher than predicted,whereas the Russian women had slightlylower than expected values. Thus, in termsof percent deviation from predicted BMR,Evenki and Russian men showed more simi-lar metabolic responses (+11.8 vs. +11.2%),whereas Evenki women displayed greatermetabolic activity than Russian women(+5.8% vs. −3.6%; P < 0.05).
In addition to potential genetic variation,the observed difference in BMRs among theEvenki and Russian females may reflectlifestyle differences. Russian women areless active (Leonard et al., 1997b, 1998),have greater amounts of adipose tissue, andhave relatively lower energy intakes perunit mass than their Evenki counterparts(Galloway, 1996). In addition, Evenkiwomen consume a diet that is higher in pro-tein (Galloway, 1996; Leonard et al., 1998).Earlier BMR studies among circumpolarpopulations implicated traditional dietshigh in protein and fat to the elevatedBMRs observed in Inuit groups compared to“controls” (Hart et al., 1962; Rodahl, 1952),but the evidence was controversial. Thethermic effect of a meal is known to increasewith higher protein and fat intakes (Jameset al., 1990; Hill et al., 1995). However,Soares et al. (1988) showed that the effect of
the preceding day’s protein intake did noteffect BMRs in healthy adults. Preliminaryanalyses showed that protein intake as acovariant did not account for significant eth-nic differences in BMR per unit body weightand FFM, but partially accounted for theethnic differences in BMR per unit surfacearea. Clearly, the influence of diet on BMRsrequires further investigation.
Age-related differences in metabolic in-tensity were evident in the women of bothethnic groups. BMR per unit weight and SAwere strongly negatively correlated withage in Russian and Evenki women, suggest-ing marked declines in metabolic intensitywith aging. This is consistent with age-related changes described by Poehlman etal. (1996). In contrast, relatively little age-related decline in BMR was evident for theEvenki and Russian men.
This pattern of declining metabolic rateswith age seen in Evenki women is strikinglydifferent from that reported by Rode andShephard (1995) among the Inuit ofCanada. These authors found greater eleva-tions in BMR among their older subjects,relative to their younger and middle-agedindividuals. Rode and Shephard hypoth-esized that this difference was due to a moretraditional lifestyle, including higher con-sumption of protein-rich “country foods,”and more outdoor hunting activity amongthe older members of the community. Thereis considerable evidence to suggest that dif-ferences in physical activity can influencemetabolic rate (Bittel et al., 1988; Meijer etal., 1991; Poehlman et al., 1989; Zurlo et al.,1990). In contrast to the Inuit, the Evenki(especially women) show declines in physi-cal activity level as they age. It is theyounger Evenki who spend more time livingoutdoors in the herding brigades looking af-ter the reindeer herds, while the older mem-bers of the population have a more seden-tary life, with many living year-round aspensioners in the collective villages and in-doors.
Additionally, Evenki women show greaterage-related increases in body weight andfatness compared to men (Galloway, 1996;Leonard et al., 1994), thereby potentiallydecreasing the metabolic responses to coldair (Beall and Goldstein, 1992; Buskirk etal., 1963; Rode and Shephard, 1994). Au-thors of early BMR circumpolar studies (Mi-lan and Evonuk, 1967; Milan et al., 1963;Rennie et al., 1962) attributed elevated
BASAL METABOLISM OF THE EVENKI 83
BMRs of Inuit to their lower fat insulationand relatively increased activity patterns.Thus, changes in both activity patterns andbody composition are likely responsible forthe age-related declines in BMR seen inEvenki women. In contrast, Evenki maleage cohorts showed few age-related changesin anthropometric and BMR measures. Thismay be due to their adherence of more tra-ditional subsistence throughout life.
Residence location also had an importantinfluence on average BMRs for the Evenki.Subjects from the collective village ofSurinda had significantly higher BMRs (ab-solute, per unit surface area, weight, andFFM) than those of Poligus. Once more, theBMRs were significantly higher than refer-ence standards for the Evenki of Surinda,and considerably lower for the Evenki of Po-ligus. The higher BMRs in Surinda com-pared to Poligus are consistent with the factthat Surinda as a community is more north-erly and less acculturated, or more “tradi-tional.” The Evenki of Surinda appear to ad-here to a more nomadic lifestyle, involvinggreater cold exposure and physical activity.Dietary differences between the two com-munities may contribute to some of the dif-ferences in BMR. Surinda Evenki consumemore protein (reindeer meat) and fat-richfoods than Poligus Evenki (Galloway, 1996).However, preliminary analyses showed thatprotein intake as a covariant did not ac-count for the significant location cohort dif-ferences in BMRs (absolute, per unit surfacearea, weight, and FFM). Although furtherinvestigation is needed to examine any in-fluence of diet on BMR, these analyses sug-gest that other factors other than diet, suchas genetic variation, play important roles.
Thermoneutral air temperatures duringBMR measurements were significantly dif-ferent between the Evenki females ofSurinda and Poligus. The fact that Surindais further north than Poligus (or Baykit)and, hence, has lower mean ambient tem-peratures is expected to affect short-termand long-term BMRs of the inhabitants ofthis community. However, evidence showedthat air temperature during testing hadlittle effect on BMR, since as a covariate, itdid not fully account for BMR (absolute, perunit surface area, weight, and FFM) differ-ences between the two female cohorts. Ad-ditionally, although BMRs were differentbetween the two Evenki male location co-horts, thermoneutral air temperatures were
comparable indicating no effect of air tem-perature on Evenki male BMRs. The factthat Poligus BMR measurements weretaken earlier in the summer than Surinda(end July to mid-August vs. mid to end ofAugust) with slightly lower daily environ-mental temperatures was also considered.The average air temperature between theRussian female sample of Poligus and Bay-kit was also significantly different; however,the BMRs between these two groups werecomparable. The fact that the Baykit BMRmeasures were taken the latest in the sum-mer (end of August to early September) withmore disparate and lower daily environ-mental temperatures is evidence to suggestthat the timing of BMR measures duringthe field season (i.e., consecutively betweencommunities) had little effect on observedBMRs. Alternatively, the possibility thatthe Evenki may be phenotypically more sen-sitive to environmental temperatures needsto be considered using endocrine along withmetabolic markers of cold adaptation.
Chronic and severe cold exposure has longbeen linked to altered levels of thyroidhormone (Itoh, 1980). In particular, tri-iodothyronine (T3) is dominant in metaboli-cally adapting to cold stress via nonshiver-ing thermogenesis (Freake and Oppenhei-mer, 1995). Altered thyroid status inhumans has been associated with prolongedresidence in circumpolar regions, and is re-ferred to as the “Polar T3 Syndrome” (Bojko,1997; Hackney et al., 1995; Harford et al.,1993; Reed et al., 1990, 1986). In addition, ithas been argued that variability in thyroidhormone levels seen in circumpolar nativesare one phenotype of populations geneti-cally adapted to northern climates(Saljukov et al., 1992; Tkachev et al., 1991).Preliminary analysis in this sample showsthat thyroid hormone levels are altered inEvenki compared to Russian females, andbetween the Evenki of Poligus and Surinda.
Genetic factors likely play an importantrole in determining BMR and in influencingthe degree to which it is mutable (Ravussinand Bogardus, 1989). Bogardus et al. (1986)showed a familial dependence of RMR inPima Indians. Twin studies of RMR(Bouchard et al., 1989) and BMR (Henry etal., 1990) also suggest a strong heritablecomponent. According to Rode and Shep-hard (1995), it remains conceivable that aninherited difference of basal metabolism has
84 V.A. GALLOWAY ET AL.
helped the survival of circumpolar popula-tions.
The population of Surinda is almost en-tirely Evenki. As of 1994, 95% o f the com-munity’s 613 residents were indigenous(Leonard et al., 1997a). In contrast, Poligusis a more heterogeneous community, with48% of the community’s population in 1994being indigenous. McComb et al.’s (1995)analyses of VNTR loci suggests greater ge-netic heterozygosity and relatively greatergene flow into Poligus than Surinda. Thefact that the Evenki of Surinda had elevatedBMRs during the summer months (i.e., withminimal thermal stress) in this more tradi-tional region with low genetic admixturemay indicate a long-term and genetic adap-tation of the Evenki that persists in theEvenki of Surinda. Alternatively, and per-haps more likely, the elevated BMRs of theEvenki of Surinda and the depressed BMRsof the Evenki of Poligus may reflect the factthat all Evenki possess a genetic potentialfor elevated BMRs and that BMR is tightlycoupled to the environment in which theyreside. The Evenki may thus possess a phe-notypic plasticity for BMR according to theenvironment. This dynamic form of cold ad-aptation would be inherently adaptive,since it would theoretically conserve energyfor the Evenki during warmer environmen-tal temperatures and less cold exposure.Lastly, long-term developmental acclimati-zation of the more traditional SurindaEvenki to more severe and chronic cold con-ditions is also a consideration.
The results point to a long-term and ge-netic potential for elevated BMR in theEvenki which continues to be modified byacculturation. BMR of the Evenki appearsto be multifactorial, and the genotype forbasal metabolic thermoregulation is pheno-typically expressed in a graded manner ac-cording to the environment. Higher BMRsare exhibited primarily in those Evenki liv-ing a traditional lifestyle with greater out-door cold exposure, i.e., the younger Evenkiand those inhabiting Surinda. Furtheranalyses will explore the extent to whichbody fat, diet, and thyroid hormone contrib-ute to differences in BMR among the Evenkiand Russians. Future field research will in-vestigate BMRs of Evenki and Russianyouth to explore the possibility of a develop-mental BMR adaptation, as well as BMRsduring the winter months.
In conclusion, this study has demon-
strated that after correcting for differencesin size and body composition, Evenki menhave similar BMRs compared to their Rus-sian peers, whereas Evenki women havehigher metabolic rates than their Russiancounterparts. When measured BMRs arecompared to predicted values based on theSchofield (1985a,b) norms, slight average el-evations are evident in Evenki men andwomen and in Russian men, while Russianwomen display BMRs that fall below pre-dicted levels. Relative to predicted valuesbased on body SA (from Consolazio et al.,1963), both Evenki men and women showsignificant (P < 0.05) elevations in BMR.Russian men also show high BMRs per unitSA, whereas Russian women have valuesthat fall slightly below predicted levels.
Within the Evenki, marked variation inBMRs was found in association with age(women) and residence location (bothsexes). The declines in BMR with age amongEvenki women are likely attributable to dif-ferences in activity, lifestyle, and body com-position. Differences associated with resi-dence location, on the other hand, may re-flect differences in the levels of geneticadmixture and/or gene expression, each inassociation with the levels of acculturation.As such, the patterns of variation in meta-bolic rate suggest the interaction of geneticand environmental factors.
ACKNOWLEDGMENTS
This study was conducted in collaborationwith Dr. Ludmilla Osipova (Institute of Cy-tology and Genetics, Russian Academy ofScience, Novosibirsk). We are grateful to allof the volunteers who participated in thisstudy. Additionally, we thank Mark Smith(AeroSport, Inc.) and Dr. Brian Wilson fortechnical assistance with the TEEM 100,and Anne Keenleyside and Marina Kazak-ovtseva for assistance with data collectionin the field.
LITERATURE CITED
Bailey BJR, Briars GL. 1996. Estimating the surfacearea of the human body. Stat Med 15:1325–1332.
Beall CM, Goldstein MC. 1992. High prevalence of ex-cess fat and central fat patterning among Mongolianpastorial nomads. Am J Hum Biol 4:747–756.
Beall CM, Henry J, Worthman C, Goldstein MC. 1996.Basal metabolic rate and dietary seasonality amongTibetan nomads. Am J Hum Biol 8:361–370.
Bogardus D, Lillioja S, Ravussin E, Abbott W, ZawadzkiJ, Young A, Knowler WC, Jacobowitz Rand Moll PP.1986. Familial dependence of the resting metabolicrate. N Engl J Med 315:96–100.
Bojko ER. 1997. Metabolic changes induced by adapta-tion to circumpolar conditions in Spitsbergen. Int JCircumpolar Health 56:134–141.
Bouchard C, Tremblay A, Nadeau A, Despres JP, The-riault G, Boulay MR, Lortie G, Leblanc E, Fournier G.1989. Genetic effect in resting and exercise metabolicrates. Metabolism 38:364–370.
Buskirk ER, Thompson RH, Whedon GD. 1963. Meta-bolic response to cold air in men and women in rela-tion to total body fat content. J Appl Physiol 18:603–612.
Butte NF, Moon JK, Wong WW, Hopkinson JM, SmithEO. 1995. Energy requirements from infancy to adult-hood. Am J Clin Nutr 62(Suppl):1047S–1052S.
Consolazio CF, Johnson RE, Pecora LJ. 1963. Physi-ological measurements of metabolic function in man.New York: McGraw-Hill.
Curtis VC, Birch E, Henry CJK, Ghusain-Choueire A.1996. Intraindividual variation in the basal metabolicrate of women: effect of menstrual cycle. Am J HumBiol 8:630–639.
Dubois EF. 1927. Basal metabolic rate in health anddisease. Philadelphia: Lea & Febiger.
Durnin JVGA, Womersley JA. 1974. Body fat measuredfrom total body density and its estimation from skin-fold thicknesses: measurements on 481 men andwomen aged from 16 to 72 years. Br J Clin Nutr 32:77–97.
FAO/WHO/UNO (Food and Agriculture Organiza-tion/World Health Organization/United Nations Uni-versity). 1985. Energy and protein requirements.WHO Technical Support Series No. 724. Geneva:WHO.
Ferraro R, Lillioja S, Fontvieille A, Rising R, BogardusC, Ravussin E. 1992. Lower sedentary metabolic ratein women compared to men. J Clin Inv 90:780–784.
Frisancho AR. 1990. Anthropometric standards for theassessment of growth and nutritional status. Ann Ar-bor: University of Michigan Press.
Galloway VA. 1996. Basal metabolic adaptation of theEvenki of Siberia. Master’s Thesis University ofGuelph, Ontario, Canada.
Hackney AC, Feith S, Pozos R, Seale J. 1995. Effects ofhigh altitude and cold exposure on resting thyroidhormone concentrations. Aviat Space Environ Med66:325–329.
Harford RR, Reed HL, Morris MT, Sapien IE, WardenR, D’Alesandro MM. 1993. Relationship betweenchanges in serum thyrotropin and total and lipopro-tein cholesterol with prolonged Antarctic residence.Metabolism 42:1159–1163.
Hannigan J. 1991. Statistics on the economic and cul-tural development of the northern aboriginal peopleof the USSR (for the period of 1980–1989). Ottawa:Indian and Northern Affairs.
Hart JS, Sabean HB, Hildes JA, Depogas F, HammelHT, Andersen KL, Irving L, Foy G. 1962. Thermaland metabolic responses of coastal Eskimos during acold night. J Appl Physiol 17:953–960.
Henry CJK, Piggott SM, Rees DG, Priestley L, Sykes B.1990. Basal metabolic rate in monozygotic and dizy-gotic twins. Eur J Clin Nutr 44:717–723.
Hill JO, Melby C, Johnson SL, Peters JC. 1995. Physicalactivity and energy requirements. Am J Clin Nutr62(Suppl):1059S–1066S.
Itoh S. 1980. Physiology of circumpolar populations. In:Milan FA, editor. The biology of circumpolar popula-tions. Cambridge: Cambridge University Press.
James PTW, McNeill G, Ralph A. 1990. Metabolism andnutritional adaptation to altered intakes of energysubstrates. Am J Clin Nutr 51:264–269.
Jequier E. 1987. Energy utilization in human obesity.Ann NY Acad Sci 499:73–83.
Jequier E, Acheson K, Schutz Y. 1987. Assessment ofenergy expenditure and fuel utilization in man. AnnuRev Nutr 7:187–208.
Kashiwazaki H. 1990. Seasonal fluctuation of BMR inpopulations not exposed to limitations in food avail-ability: reality or illusion? Eur J Clin Nutr 44(Suppl):85–93.
Katzmarzyk PT. 1993. The adaptive significance of en-ergetic expenditure among indigenous Siberian popu-lations. Master’s Thesis University of Guelph, On-tario, Canada.
Katzmarzyk PT, Leonard WR, Crawford MH, SukernikRI. 1994. Resting metabolic rate and daily energy ex-penditure among two Siberian populations. Am JHum Biol 6:719–730.
Katzmarzyk PT, Leonard WR, Stephen MA, Berti PR,Ross AGP. 1996. Differences between observed andpredicted energy costs at rest and during exercise inthree subsistence-level populations. Am J Hum Biol8:537–545.
Leonard WR, Katzmarzyk PT, Commuzie AG, CrawfordMH, Sukernick RI. 1994. Growth and nutritional sta-tus of the Evenki reindeer herders of Siberia. Am JHum Biol 6:339–350.
Leonard WR, Katzmarzyk PT, Crawford HM. 1996. En-ergetics and population ecology of Siberian herders.Am J Hum Biol 8:275–289.
Leonard WR, Keenleyside A, Ivakine EA. 1997a. Recentfertility and mortality trends among aboriginal andnonaboriginal populations of central Siberia. HumBiol 69:403–417.
Leonard WR, Galloway VA, Ivakine EA. 1997b. Under-estimation of daily energy expenditure with the fac-torial method: implications for anthropological re-search. Am J Phys Anthropol 103:443–454.
Leonard WR, Galloway VA, Ivakine EA, Osipova L.1998. The influence of acculturation and lifestylechange on the health of indigenous Siberians. Am JPhys Anthropol Suppl 28:150 (Abstract).
Lohman TG, Roche AF, Martorell R. 1988. Anthropo-metric standardization reference manual (abridgededition). Champaign, IL: Human Kinetics.
McArdle WD, Katch FI, Katch VL. 1991. Exercise physi-ology: energy, nutrition, and human performance, 3rded. Philadelphia: Lea & Febiger.
McComb J, Blagitko N, Cumuzzie AG, Schansield SM,Sukernik RI, Leonard WR, Crawford MH. 1995.VNTR DNA variation in Siberian indigenous popula-tions. Hum Biol 67:217–229.
McCormack PD, Reed HL, Thomas JR, Malik MJ. 1996.Increase in rT3 serum levels observed during ex-tended Alaskan field operations of naval personnel.Alaska Med 38:89–97.
Meijer GAL, Janssen GME, Westerterp KR, VerhoeverF, Saris WHM, en Hoor F. 1991. The effect of a5-month endurance-training programme on physicalactivity: evidence for a sex-difference in the metabolicresponse to exercise. Eur J Appl Physiol 62:11–17.
Milan FA, Evonuk E. 1966. Oxygen consumption andbody temperatures of Eskimos during sleep. J ApplPhysiol 22:565–567.
Milan FA, Hannon JP, Evonuk E. 1963. Temperature
86 V.A. GALLOWAY ET AL.
regulation of Eskimos, Indians, and Caucasians in abath calorimeter. J Appl Physiol 18:378–382.
Norgan NG. 1996. Measurement and interpretation is-sues in laboratory and field studies of energy expen-diture. Am J Hum Biol 8:143–158.
Poehlman ET. 1996. Energy intake and energy expen-diture in the elderly. Am J Hum Biol 8:199–206.
Poehlman ET, Toth MJ. 1995. Mathematical ratios leadto spurious conclusions regarding age- and sex-related differences in resting metabolic rate. Am JClin Nutr 61:482–485.
Poehlman ET, Melby CL, Badylak SF, Calles J. 1989.Aerobic fitness and resting energy expenditure inyoung adult males. Metabolism 38:85–90.
Ravussin E, Bogardus C. 1989. Relationship of genetics,age and physical fitness to daily energy expenditureand fuel utilization. Am J Clin Nutr 49:968–975.
Reed HL, Burman KD, Shakir KMM, O’Brian JT. 1986.Alterations in the hypothalamic-pituitary-thyroidaxis after prolonged residence in Antarctica. Clin En-docrinol 25:55–65.
Reed HL, Brice D, Shakir KMM, Burman KD,D’Alesandro MM, O’Brian JT. 1990. Decreased freefraction of thyroid hormones after prolonged Antarc-tic residence. J Appl Physiol 69:1467–1472.
Rennie DW, Covino BG, Blair MR, Rodahl K. 1962.Physical regulation of temperature in Eskimos. JAppl Physiol 17:326–332.
Roberts DF. 1978. Climate and human variability, 2nded. Menlo Park, CA: Cummings.
Rodahl K. 1952. Basal metabolism of the Eskimos. JNutr 48:359–368.
Rode A, Shephard. 1994. Predication of body fat contentin an Inuit community. Am J Hum Biol 6:249–254.
Rode A, Shephard RJ. 1995. Basal metabolic rate ofInuit. Am J Hum Biol 7:723–729.
Saljukov VB, Lemza SV, Kucher AN, Puzyrev VP. 1992.The role of heredity factors in phenotypic variabilityof hormone in the population genetically adapted tocircumpolar environment. Arct Med Res 51:143–149.
Schofield WN. 1985a. Predicting basal metabolic rate:new standards and a review of previous work. HumNutr 39C(Suppl 1):5–41.
Schofield WN. 1985b. An annotated bibliography ofsource material for basal metabolic rate data. HumNutr 39C(Suppl):42–91.
Soares MJ, Shetty PS. 1991. Basal metabolic rates andmetabolic economy in chronic undernutrition. Eur JClin Nutr 45:363–373.
Soares MJ, Sequeirs J, Shetty PS. 1988. The effect ofthe preceding day’s protein intake on basal metabolicrates in young adults. Br J Nutr 60:425–431.
Soares MJ, Sheela ML, Kurpad AV, Kukarni RN,Shetty PS. 1989. The influence of different methodson basal metabolic rate measurements in human sub-jects. Am J Clin Nutr 50:731–736.
Tkachev AV, Jevgeny RB, Ramenskaya EB. 1991. En-docrine status and plasma lipids in inhabitants of thenorthern European part of the USSR. Arct Med Res50:148–151.
Ulijaszek SJ. 1992. Human energetics methods in bio-logical anthropology. Yrbk. Phys Anthropol 35:215–242.
Ulijaszek SJ. 1996. Energetics, adaptation, and adapt-ability. Am J Hum Biol 8:169–182.
Waterloc JC. 1988. The variability of energy metabo-lism in man. In: K Baxter and I MacDonald, editors:Comparative nutrition. London: John Libbey, pp 133–147.
WHO (World Health Organization). 1978. Arterial hy-pertension. Technical support series no. 628. Geneva:WHO.
WHO (World Health Organization). 1980. Research onthe menopause. Technical report series No. 720.Geneva: WHO.
Weinsier R, Schutz Y, Bracco D. 1992. Reexamination ofthe relationship of resting metabolic rate to fat-freemass and the metabolically active components of fat-free mass in humans. Am J Clin Nutr 55:790–794.
Weir JB. 1949. New methods for calculating metabolicrate with special reference to protein metabolism. JPhysiol 109:1–9.
Zurlo F, Larson K, Bogardus C, Ravussin E. 1990. Skel-etal muscle metabolism is a major determinant ofresting energy expenditure. J Clin Inv 86:1423–1427.
