AR254-AN34-13 ARI 25 August 2005 15:1

Race and Ethnicity in PublicHealth Research: Models toExplain Health DisparitiesWilliam W. Dressler,1 Kathryn S. Oths,1

and Clarence C. Gravlee2

1Department of Anthropology, The University of Alabama, Tuscaloosa, Alabama35487; email: wdressle@tenhoor.as.ua.edu, koths@tenhoor.as.ua.edu2Department of Anthropology, Florida State University, Tallahassee, Florida 32306;email: cgravlee@fsu.edu

Annu. Rev. Anthropol.2005. 34:231–52

First published online as aReview in Advance onJune 14, 2005

The Annual Review ofAnthropology is online atanthro.annualreviews.org

doi: 10.1146/annurev.anthro.34.081804.120505

Copyright c© 2005 byAnnual Reviews. All rightsreserved

0084-6570/05/1021-0231$20.00

Key Words

cultural constructivism, psychosocial stress, racism, birth weight,blood pressure

AbstractThe description and explanation of racial and ethnic health dispari-ties are major initiatives of the public health research establishment.Black Americans suffer on nearly every measure of health in relationto white Americans. Five theoretical models have been proposed toexplain these disparities: a racial-genetic model, a health-behaviormodel, a socioeconomic status model, a psychosocial stress model,and a structural-constructivist model. We selectively review litera-ture on health disparities, emphasizing research on low birth weightand high blood pressure. The psychosocial stress model and thestructural-constructivist model offer greatest promise to explain dis-parities. In future research, theoretical elaboration and operationalspecificity are needed to distinguish among three distinct factors:(a) genetic variants contributing to disease risk; (b) ethnoracial orfolk racial categories masquerading as biology; and (c) ethnic groupmembership. Such elaboration is necessary to move beyond the con-flation of these three distinct constructs that characterizes much ofcurrent research.

231

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

Contents

INTRODUCTION. . . . . . . . . . . . . . . . . 232RACIAL AND ETHNIC

HEALTH DISPARITIES: THEEPIDEMIOLOGIC EVIDENCE 232

PATTERNS IN THELITERATURE ON RACIALAND ETHNIC HEALTHDISPARITIES . . . . . . . . . . . . . . . . . . . 234

EXPLAINING RACIAL ANDETHNIC HEALTHDISPARITIES . . . . . . . . . . . . . . . . . . . 234The Racial-Genetic Model . . . . . . . 235The Health-Behavior Model . . . . . 236The Socioeconomic Status Model 238The Psychosocial Stress Model . . . 238The Structural-Constructivist

Model . . . . . . . . . . . . . . . . . . . . . . . . 241RESEARCH NEEDS . . . . . . . . . . . . . . . 243CONCLUSION . . . . . . . . . . . . . . . . . . . . 245

Health disparities:differences inmorbidity, mortality,and access to healthcare amongpopulation groupsdefined by factorssuch associoeconomicstatus, gender,residence, and raceor ethnicity

NIH: NationalInstitutes of Health

INTRODUCTION

Understanding health disparities is a majorinitiative of the public health research estab-lishment in the United States (Woolf et al.2004) and around the world (Almeida-Filhoet al. 2003). Reducing such health disparitiesis one goal of the Healthy People 2010 ini-tiative, a program of the National Institutesof Health (NIH) intended to improve publichealth in the United States (U.S. Dep. HealthHum. Serv. 2000). “Health disparities” refersto differences in morbidity, mortality, and ac-cess to health care among population groupsdefined by factors such as socioeconomic sta-tus, gender, residence, and especially “race”or “ethnicity.” Research has examined differ-ences in health status among diverse racial andethnic groups; however, why the health dis-parities between black and white Americansare so large is the central question in thisresearch. On nearly every index measured,African Americans suffer in relation to Euro-pean Americans, and often in relation to other

racial and ethnic groups as well. Every field ofstudy concerned with public health, includ-ing epidemiology, health services research,health psychology, medical sociology, clinicalmedicine, nursing, and medical anthropology,contributes to this research.

The issues are immense. Approaching thisresearch in any sensible way demands thatsubjects from population genetics to the so-cial and cultural construction of the conceptsof race and ethnicity be addressed. And thereare deeply entrenched ideologies on all sidesof the issue. Separating ideology and theoryis always difficult, but it seems to be partic-ularly problematic in the study of the healtheffects of race and ethnicity. On the one hand,the biological reality of race continues to beobvious to many. On the other hand, it isequally clear to many researchers that race isa culturally constructed entity. Stating thesetwo perspectives so simply obscures the pro-found differences in epistemology that gen-erate them, differences that can overwhelmresearch progress on a topic that demands anintegration of realist and constructivist per-spectives. The study of racial and ethnic healthdisparities requires a biocultural perspectivein the fullest sense of the term.

The aim of this chapter is to examine howsome of these thorny questions have been ap-proached. The literature is vast, so our discus-sion will be selective, out of necessity. First,we briefly review the nature and breadth ofracial and ethnic health disparities. Second,we address the major alternative explanationsfor these disparities. Third, we turn to thequestion of the logical status of the constructsof race and ethnicity and consider how theymight be more effectively defined for cross-cultural research.

RACIAL AND ETHNICHEALTH DISPARITIES: THEEPIDEMIOLOGIC EVIDENCE

A research focus on health disparities can betraced to the Black Report examining socialclass and regional differences in health status

232 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

in Britain (Townsend & Davidson 1982), butrealization of the extent of ethnic and racialhealth disparities in the United States hasbeen a result of reports published by thepublic health research establishment, espe-cially the NIH and related agencies. Nickens(1986) drew attention to these disparitiesthrough a calculation of “excess deaths” inrelation to race or ethnicity. This is an epi-demiologic index in which predicted deathsare calculated for a population subgroup (e.g.,black men) using data from another pop-ulation subgroup (e.g., white men). Excessdeaths are the number of deaths observedthat exceed the predicted number. Nickensfound substantial disparities between ethnicand racial groups: Black Americans in partic-ular exhibited a large excess mortality com-pared with white Americans on virtually ev-ery cause of death. Other ethnic groups (e.g.,Asian and Pacific Islanders) exhibited fewerthan predicted deaths in some categories com-pared with white Americans, which indicatesthat health disparities can be observed inlower-than-expected rates for some groups aswell.

Wong & associates (2002) conducted asimilar analysis with more sophisticated tech-niques of adjustment. They also contrastedracial and ethnic disparities with educationaldisparities and found that different diseasescontribute to different disparities. In termsof disparities in mortality between less well-and more well-educated persons, coronaryartery disease, lung cancer, stroke, pneumo-nia, congestive heart failure, and lung dis-ease contributed most to lost years of life.With respect to racial or ethnic disparities inmortality, however, hypertension, HIV, dia-betes, and homicide were the greatest con-tributors. Therefore, there are a few dis-eases that contribute most to disparities inmortality.

Keppel et al. (2002) examined data be-tween 1990 and 1998 using a set of ten healthstatus indicators relevant to the Healthy People2010 goals. These indicators included a range

Ethnicity: adimension ofsociocultural systemsthat defines essentialgroup differencesand structures therelations amongpersons classified interms of thosedifferences. Thecriteria of essentialgroup difference willvary cross-culturally

HSIs: health statusindicators

of health conditions (coronary heart disease,cancer, tuberculosis) as well as indicators withmore direct social referents such as live birthsto teenagers ages 15–17 and suicide and homi-cide rates. Keppel et al. examined these indica-tors in relation to racial and ethnic categoriesemployed in the U.S. census: non-Hispanicwhite; non-Hispanic black; Hispanic; Amer-ican Indian or Alaska Native; and Asian orPacific Islander. For mortality related to dis-ease and traumatic injury, rates declined forall population subgroups between 1990 and1998. With respect to racial and ethnic dispar-ities, black Americans’ rates for six measures(total mortality, heart disease, lung cancer,breast cancer, stroke, and homicide) exceededother groups’ rates by a factor ranging from2.5 to almost 10 during both time periods.Other ethnic groups had higher rates for sui-cide (white Americans) and motor vehicleaccidents (American Indian and Alaskan Na-tives). Overall, Asians and Pacific Islanderstended to have the lowest mortality rates,although Hispanics were lowest for strokes.Keppel et al. concluded,

Based on this analysis relatively littleprogress was made toward the goal of elim-inating racial/ethnic disparities among theHSIs [health status indicators] during thelast 10 years. Progress toward the goal ofeliminating health disparities will requiremore concerted efforts during the next 10years. (2002, p. 12)

Clearly, understanding ethnic and racialhealth disparities demands a careful exami-nation of all groups in all societies in whichsuch disparities exist. However, these re-sults indicate that understanding the healthstatus differences between black and whiteAmericans is fundamental to understandinghealth disparities in general because this dif-ference contributes most to overall healthdisparities. In the remainder of this review,we place emphasis on understanding thesedisparities.

www.annualreviews.org • Race and Ethnicity in Health Research 233

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

Racial-geneticmodel: a model forthe explanation ofhealth disparitiesthat emphasizespopulationdifferences in thedistribution ofgenetic variants

Health-behaviormodel: a model forthe explanation ofhealth disparitiesthat emphasizesdifferences betweenracial and ethnicgroups in thedistribution ofindividual behaviorsrelated to health,such as diet, exercise,and tobacco use

Socioeconomicstatus model: amodel for theexplanation of healthdisparities that positstheover-representationof some racial andethnic groups withinlower socioeconomicstatuses

Psychosocial stressmodel: a model forthe explanation ofhealth disparitiesthat emphasizes thestresses associatedwith minority groupstatus, and especiallythe experience ofracism anddiscrimination

PATTERNS IN THELITERATURE ON RACIAL ANDETHNIC HEALTH DISPARITIES

A number of reviews have examined the extentto which race and ethnicity are used as vari-ables in public health research. Systematic re-views of several key journals show that (a) raceand ethnicity are among the most com-monly used variables in public health research;(b) their use is on the rise; and (c) they tend tobe used uncritically and without definition.

Jones and colleagues (1991) reviewed ar-ticles published between 1921 and 1990 inthe American Journal of Epidemiology and foundthat roughly two thirds of articles made ref-erence to race. By 1990 almost 80% of theresearch articles published included race as avariable. Comstock et al. (2004) updated theseresults and found that 77% of articles pub-lished from 1996 to 1999 in the American Jour-nal of Epidemiology and the American Journal ofPublic Health made reference to race or eth-nicity. Williams (1994) documents a similarpattern in the health services research liter-ature from 1966 to 1990, and Drevdahl andcolleagues (2001) show that more than 81%of reports published in 2000 in the journalNursing Research referred to race or ethnicity.

Investigators have observed similar pat-terns outside the United States. For example,Ellison & de Wet (1997) report that roughlyhalf of the articles published in the SouthAfrican Medical Journal in 1992–1996 men-tioned race, and the proportion was muchhigher (73.9%) in genetic studies. Britishhealth researchers also use racial and ethniccategories (Sheldon & Parker 1992), althoughthere is a decided preference for the conceptof ethnicity more than race in the UnitedKingdom.

Explicit definitions of race or ethnicity arerare. Williams’s (1994) review revealed thatnone of the 121 empirical studies publishedbetween 1966 and 1990 in the health servicesresearch literature defined race. Just 5 of the167 studies published in Nursing Research from1952 to 2000 defined the racial and ethnic

categories used, and these definitions oftenfailed to make clear the differences among cat-egories (Drevdahl et al. 2001). Similarly, in areview of the American Journal of Epidemiol-ogy and the American Journal of Public Health,Comstock et al. (2004) found that most re-cent articles neither specify why race or eth-nicity was included as a variable nor iden-tify the method by which either was assessed.These patterns appear to hold true else-where in the biomedical literature (Kaplan &Bennett 2003, Osborne & Feit 1992).

Social scientists are often quick to recog-nize these shortcomings in the public healthliterature, but it remains to be seen if they ap-proach race and ethnicity in any more sophis-ticated way. Preliminary results from an on-going content analysis of Medical Anthropologyand Medical Anthropology Quarterly (Gravleeet al. 2004) suggest that neither the conceptof race nor ethnicity is used as commonlyin medical anthropology as in public health.Medical anthropologists display a preferencefor ethnicity more than race, but an analyticdistinction between these concepts is seldommade. In addition, as in public health, raceand ethnicity typically are not defined, andthe methods by which groups and individu-als are assigned to racial or ethnic categoriesgenerally are not explicit.

EXPLAINING RACIAL ANDETHNIC HEALTH DISPARITIES

Dressler (1993) identified four general mod-els in the literature to explain health dispari-ties: a racial-genetic model; a health-behaviormodel; a socioeconomic status model; and asocial structural model. Changes in emphasisin the literature in the intervening ten yearsrequire both expansion and slight modifica-tion of these categories as follows: a racial-genetic model; a health-behavior model; asocioeconomic model; a psychosocial stressmodel; and a structural-constructivist model.

In the following sections, we sample fromthe health disparities literature, but most

234 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

of our examples will come from researchon low birth weight and essential hyperten-sion/chronic high blood pressure. These areuseful health indicators because they effec-tively bracket the life span, and they are amongthe health problems that contribute most tohealth disparities. Additionally, there is evi-dence that low birth weight and high bloodpressure are associated, and hence may belinked in the biocultural processes that gen-erate health disparities (Barker 2004).

The Racial-Genetic Model

There are large differences in rates of lowbirth weight (defined as birth weight less than2500 grams) and rates of hypertension (bloodpressure higher than 140 mm Hg systolicand/or 90 mm Hg diastolic) between blackand white Americans. Currently 13.3% ofblack women, versus 6.9% of white women,give birth to a low birth weight baby. Similarly,38.6% of black women and 34.8% of blackmen have high blood pressure, compared with22.6% of white women and 24.8% of whitemen (Cent. Dis. Control Prev. 2004a,b).

Especially with respect to high blood pres-sure, there have been appeals by some re-searchers to racial-genetic factors to accountfor these disparities (Boyle 1970). For bloodpressure, these appeals have been based onboth the differences in hypertension preva-lence between blacks and whites, and thegradient of blood pressure and hypertensionprevalence in relation to skin color withinthe African American community (Harburget al. 1978, Keil et al. 1977, Klag et al. 1991).Although poor birth outcomes are less of-ten explicitly attributed to racial-genetic dif-ferences, some researchers have suggested assuch (Wilcox & Russell 1990, and see critiqueby David 2001). In current literature it canbe difficult to find overt attribution of diseaserisk to a racial-genetic component, perhapsbecause of widespread knowledge of the cri-tique of race as a biological construct (Cooper1984, Montagu 1962, Kittles & Weiss 2003).More often than not, as noted above, differ-

ences associated with race or ethnicity are sim-ply reported.

Structural-constructivistmodel: a model forthe explanation ofhealth disparitiesthat emphasizes theintersection ofracially stratifiedsocial structures withthe culturalconstruction ofroutine goals andaspirations as thecause of differencesin morbidity andmortality. Explicit inthis model is thecultural constructionof the notion of“race” itself.

Prior to technological advances in molec-ular biology, evidence had accumulated that aracial-genetic explanation was untenable forbroad population differences in blood pres-sure. Literature reviews showed that preva-lence rates were extremely variable acrosspopulations in Africa and people of Africandescent in the New World. Later, focused re-search (eliminating alternative measurementhypotheses for these differences) demon-strated an east-west gradient in hyperten-sion prevalence: West African samples hadthe lowest prevalence (16%), West Indianpopulations had an intermediate prevalence(26%), and African American populations hadthe highest prevalence (33%) (Cooper et al.1997). With respect to birth weight, David &Collins (1997) found that the rate of low birthweight infants of African-born black womenin Chicago (3.6%) was closer to that of U.S.-born white women (2.4%) than to U.S.-bornblack women (7.5%). Similarly, Kleinman andassociates (1991), using a national data set,found that the risk of neonatal mortality was22% lower for foreign-born compared withU.S.-born black women, whereas there wasno difference in risk for whites on the ba-sis of birth country. More recently, Acevedo-Garcia et al. (2005) have shown that lower-educated foreign-born black women do notdiffer in birth outcomes from highly educatedU.S.-born black women (less well-educatedU.S.-born black women had the highest riskof low birth weight). These data at least sug-gest that birth weight and blood pressure aresubject to substantial environmental influenceand hence are not under strict racial-geneticcontrol, given the range of variability in preva-lence rates in genetically related populationsliving in different environments.

With the advent of technology for iden-tifying genetic variants, the importance ofa racial-genetic component in blood pres-sure has become even less tenable becausethe search for variant gene structures thatcontribute to blood pressure has not been

www.annualreviews.org • Race and Ethnicity in Health Research 235

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

NCHS: NationalCenter for HealthStatistics

particularly successful (Crews & Williams1999, Harrap 2003, Oparil et al. 2003); andthose candidate genes that do appear to beassociated with blood pressure are not differ-entially distributed across conventional racialgroups (Cooper et al. 1999, Daniel & Rotimi2003), nor do they differ between AfricanAmericans and first-generation African immi-grants (Bouzekri et al. 2004, Carlos Postonet al. 2001).

The racial-genetic model has not disap-peared, however. Belief in its importance inthe explanation of blood pressure disparitiescontinues in the form of Grim’s “slavery hy-pothesis” (Grim & Robinson 1996). This hy-pothesis posits that a salt-sparing genetic vari-ant was selected for in Africa (a kind of “thriftygenotype”) owing to chronic salt shortages.Then, enslaved Africans were subjected to ex-treme conditions of sodium deprivation in theMiddle Passage and under conditions of slav-ery, leading to high mortality rates, and thissalt-sparing genetic variant in New WorldAfrican-descent populations was further se-lected for. According to the slavery hypoth-esis, owing to the higher prevalence of thisracial-genetic trait, African Americans andother black populations in the western hemi-sphere retain more sodium when it is plentifulin the diet, resulting in high blood pressure(see below on the salt intake hypothesis forhypertension).

The slavery hypothesis is a controversialidea to account for racial and ethnic healthdisparities and has many critics (Curtin 1992,Jackson 1991). What is striking, however,is its wide acceptance based on virtually noempirical evidence. In a paper full of in-sight into how hypotheses diffuse, Kaufman& Hall (2002) demonstrate the level of credi-bility the hypothesis has received in both theprofessional and popular literature, even be-ing incorporated into recommendations onhow high blood pressure should be treated(Brownley et al. 1999), despite the lack of re-search. As Kaufman & Hall note, “The in-tellectual resilience of the Slavery Hypothesismay be attributable to several of its ideologic

components[. . . .]perhaps the most influentialof these in the modern era is the beguilingallure of a simplistic genetic determinism”(2002, p. 116). The slavery hypothesis mayowe its persistence to its reinforcement of folkmodels of race. As Kittles & Weiss (2003, p.34) point out, even specialists in genetics rou-tinely confuse technical and folk uses of theterm race. The slavery hypothesis may appearto be true simply because it is consistent with,and in turn reinforces, a Western Europeanand American cultural construction of race asa biologic entity (see also Braun 2002, Sankaret al. 2004).

In sum, this model for the explanationof racial and ethnic health disparities thatemphasizes genetic variants differentially dis-tributed across these groups appears to havelittle explanatory power.

The Health-Behavior Model

Here we subsume hypotheses that accountfor health disparities on the basis of discretebehaviors voluntarily adopted by individu-als. The health behaviors regarded as impor-tant usually include the combination of highcaloric intake and low physical activity, whichleads to obesity, smoking, excessive alcohol in-take, and the intake of specific dietary items,such as high salt intake or low potassiumintake.

Body composition is clearly associatedwith higher blood pressure (Sowers et al.2002), but differences in body composition donot explain health disparities. National Cen-ter for Health Statistics (NCHS) (Cent. Dis.Control Prev. 2004c) data show that in 1999–2000, 67.3% of white American men wouldbe considered overweight (a body mass in-dex >25.0), compared with 60.3% of AfricanAmerican men. This discrepancy thereforecould not account for the differences in hyper-tension prevalence between black and whitemen. Fifty-seven percent of white womenare overweight compared with 77.7% ofblack women. Although this seemingly couldaccount for prevalence differences between

236 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

black and white women, it apparently doesnot. Bell and associates (2004) recently an-alyzed data from the NHANES (NationalHealth and Nutrition Examination Survey)for women, and black women remained twiceas likely to have high blood pressure after con-trolling for obesity. Furthermore, virtually ev-ery study of blood pressure routinely uses ameasure of body composition (e.g., the bodymass index) as a control variable, with littleeffect on black-white disparities.

With respect to birth weight, heavierwomen tend to have heavier babies (Inst.Med. 1990), so the higher prevalence of over-weight among black women cannot explainthe higher prevalence of low birth weight.

Physical activity levels affect both weightand overall risk of disease, and there is ev-idence of differences among racial and eth-nic groups in levels of physical activity. Na-tionally, 34.4% of white men and 38.3% ofwhite women report being physically inac-tive compared with 45.1% of black men and55.1% of black women (Schoenborn et al.2004, p. 42). Bell et al. (2004) found that con-trolling for reported levels of physical ac-tivity made no difference in the differentialrisk of hypertension between black and whitewomen. With respect to low birth weight,strenuous occupational activity (such as stand-ing for long periods) does not alter differencesin low birth weight between black and whitewomen (Homer et al. 1990, Teitelman et al.1990).

Smoking is a risk factor that has beendirectly implicated in low birth weight(McCormick et al. 1990) but only tangen-tially, if at all, in association with blood pres-sure (Janzon et al. 2004). Again, smoking isnot a factor likely to account for health dis-parities because there are virtually no differ-ences in rates of smoking between black andwhite men (27.1% versus 25.2%) or betweenblack and white women (19.5% versus 22.2%)(Schoenborn et al. 2004, p. 21).

Alcohol intake is discouraged during preg-nancy because it contributes to low birthweight, whereas when considering blood

NCHS: NationalCenter for HealthStatistics

NHANES:National Health andNutritionExamination Survey

pressure there appears to be a J-shapedrelationship between alcohol consumptionand blood pressure; nondrinkers and milddrinkers (≤3 drinks/day) had comparablylow blood pressures, and heavy drinkers hadhigher blood pressures (Estruch et al. 2004).According to NCHS survey data, 70.8% ofwhite men versus 55.8% of black men re-port being drinkers; corresponding figures forwhite versus black women are 60.4% versus39.4%. Rates of heavy drinking, defined as14 or more drinks per week, are very simi-lar across groups, ranging from 2.2% amongAfrican American women to 5.6% amongwhite men (Schoenborn et al. 2004, p. 7).Again, there is little evidence that differencesin alcohol intake account for the large dispar-ities in health status; in studies that controlfor reported alcohol intake, racial and ethnicdifferences in blood pressure and birth weightremain unchanged, although there is some ev-idence that blood pressure may increase at alower level of alcohol intake for black men ver-sus other groups (Acevedo-Garcia et al. 2005,Bell et al. 2004, Fuchs et al. 2001).

With respect to specific factors and specificoutcomes, high salt intake has been hypoth-esized to account for black-white differencesin blood pressure (Sowers et al. 2002); how-ever, the link of salt intake with high bloodpressure continues to be controversial. As Mc-Carron (2000) notes, the salt intake hypoth-esis has been based on inappropriate animallaboratory models; the reporting of highly se-lective cross-cultural data; and questionableand unsubstantiated concepts like a geneticsalt sensitivity. Although reduction in dietarysodium can lead to small but sustained de-creases in blood pressure in some persons withhigh blood pressure, the evidence for salt in-take as an etiologic factor that accounts forracial and ethnic differences in hypertensionis slim to nonexistent (Chrysant et al. 1997).

Some investigators have attempted to eval-uate the combined effects of health behav-iors in reducing racial and ethnic healthdisparities. These studies enter physical ac-tivity, smoking, alcohol consumption, body

www.annualreviews.org • Race and Ethnicity in Health Research 237

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

SES: socioeconomicstatus

mass index, and various measures of nutrientintake simultaneously into an analysis, alongwith race or ethnicity. The results are mixed.In some studies, racial or ethnic differencesare left unchanged (Bassett et al. 2002, Bellet al. 2004). With these variables, plus de-pressive symptoms and anxiety, Jones-Webband coworkers (1996) were able to reducedifferences in mean blood pressure for blackand white men, but not for black and whitewomen. Finally, using prospective data, Liuet al. (1996) reported that black-white meanblood pressure differences at a 7-year follow-up could be reduced by 40%–50% throughthe inclusion of body mass index, alcohol andtobacco use, physical activity, and dietary in-take of calcium, potassium, and protein.

As observed ten years ago, these health be-haviors can be potent contributors to diseaserisk (Dressler 1993); there is little evidence,however, that alone or in combination healthbehaviors can explain racial and ethnic healthdisparities.

The Socioeconomic Status Model

The socioeconomic status (SES) model sub-sumes research that sees racial and ethnichealth disparities confounded with SES dis-parities in health. Race and SES are corre-lated (i.e., African Americans are overrepre-sented among lower SES groups), and someargue that controlling for SES will either re-veal the “true” effect of race or ethnicity or,if secondary to SES disparities, cause racialdisparities to disappear. Little or no consid-eration is given to why confounding occurs.This approach was encountered more oftenin earlier literature (e.g., Keil et al. 1977,Starfield et al. 1991) than recent studies (al-though see Dyer et al. 1999). In general, con-trolling for SES fails to account completelyfor racial and ethnic disparities, despite lead-ing to a reduction in the magnitude of groupdifferences. The failure of SES controls to ac-count for racial or ethnic differences has then,in turn, been used as “evidence” of some kindof residual racial-genetic effect. As Kaufman

and colleagues (1997) show, however, such aninference is almost never warranted becauseof the problems associated with trying to un-tangle race, ethnicity, and SES. Various formsof residual confounding occur, which in turnrender an inference regarding some kind ofracial-genetic effect unlikely, even after con-trolling for SES (see also Davey Smith 2000).

Some researchers have argued that theconfounding of SES and racial disparities isa function of the wider distribution of riskyhealth behaviors among lower class people,such as those behaviors reviewed in the pre-vious section (Liu et al. 1996, Stamler et al.2003). This again is an argument that race orethnicity is confounded with SES; however,as shown above, controlling for health behav-iors does not explain racial and ethnic healthdisparities.

More promising directions in research onSES examine the effect of residence in low-income communities on health (Williams &Collins 2001), as well as how SES may mod-erate racial or ethnic differences (Acevedo-Garcia et al. 2005).

The Psychosocial Stress Model

In earlier literature, after showing that con-trols for SES failed to account for racial andethnic disparities, reasoned speculation ledto a consideration of the stresses associatedwith institutional and interpersonal racism asa cause of these disparities (Clark et al. 1999,Williams & Collins 1995). The extension ofthe psychosocial stress model to studies ofracial and ethnic disparities was stimulatedby the logical difficulties and empirical weak-ness of alternative racial-genetic and health-behavior models. As usual, the serious con-sideration of the social production of diseasehad to await the exhaustion of alternative ap-proaches. Currently, this model looms large inthe literature, and it is possible to categorizeseveral approaches within this orientation.

The first approach can be best exemplifiedby the social epidemiologists Krieger (1999,2003) and Williams (Williams & Collins 1995,

238 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

Wyatt et al. 2003). In this approach, there isa clear distinction made between institutionalracism and perceived racism, the former re-ferring to the system of structured inequalitythat places black Americans lower on all indi-cators of economic well-being, and the latterreferring to the conscious perception of dis-criminatory acts and practices and the distressassociated with that perception. Institutionalracism results in the limited access of racialand ethnic minorities to resources, both in thesense of limited access to high-paying jobs oreducational opportunities and in the sense oflimited access to resources that would supportthe attainment of better health status (e.g., liv-ing in neighborhoods with markets that stockfresh fruits and vegetables, neighborhoods inwhich it is safe to walk for exercise). The con-cept of institutional racism has mainly offereda framework for the interpretation of racialand ethnic health disparities that is an alter-native to other (e.g., racial-genetic) models,providing what Krieger (1999, p. 310) calls an“indirect” approach to the study of discrimi-nation and health.

Perceived racism, by comparison, is mea-sured directly by self-reports of respondentsabout their experiences of discriminatory acts,both in institutional settings (e.g., on the job)and in mundane social interactions (Krieger1990, Krieger & Sidney 1996). The empiricalrecord for measures of perceived discrimina-tion is mixed. In a recent review, Williams andassociates (2003) report eleven studies that ex-amine the association of perceived discrimi-nation and blood pressure. Of these studies,three find a direct association, three find noassociation, and five find associations that per-tain only to particular subgroups (e.g., gen-der or occupational groups; see also Brondoloet al. 2003). A recent study reports a directassociation of perceived discrimination andblood pressure, although data were collectedfrom a convenience sample (Din-Dziethamet al. 2004). Two studies in the review byWilliams et al. (2003) examined perceived dis-crimination and low birth weight; one foundno association and one found a conditional as-

CARDIA:CardiovascularDisease in YoungAdults

sociation. A more recent study found that con-trolling for self-reported discrimination re-duced by half the risk for black women of re-porting having had a low birth weight baby(Mustillo et al. 2004; see also Collins et al.2004).

The second approach to the study of psy-chosocial stress and health disparities em-ploys a more general understanding of theterm stress as negative affect (depression, anx-iety) experienced by individuals, which in turncan be associated with deleterious health out-comes. This approach has been taken in theincorporation of psychosocial data into largenational studies, such as CARDIA (Cardiovas-cular Disease in Young Adults) and the variouswaves of the NHANES (see Williams 1999 fora useful discussion of national data sets). Jonaset al. (1997) and Jonas & Lando (2000) lookedat overall negative affect as a prospective pre-dictor of incident hypertension in two differ-ent follow-up waves of the NHANES, find-ing that those who report negative affect areat a higher risk for developing hypertensionand that this association is greater for AfricanAmericans. Davidson et al. (2000) found asimilar pattern of results using the CARDIAdata. Finally, using a subset of the CARDIAdata, Knox and colleagues (2002) found thatyoung African Americans who were more re-active to stressful stimuli in the laboratory inturn had higher ambulatory blood pressuresthree years later.

The third approach to the study of psy-chosocial stress and health outcomes is bestrepresented by the early work of Harburg andassociates (1973), and the subsequent workof James on the John Henryism hypothe-sis (James et al. 1983). These researchersadapted general models of the stress processto the specific ethnographic realities of theAfrican American community. For example,Harburg et al. (1973) argued that persons,black or white, living in high “socioecologicstress” areas (characterized by low SES andhigh rates of social instability as measuredby crime) were at a higher risk for stress-ful experiences on a daily basis, increasing

www.annualreviews.org • Race and Ethnicity in Health Research 239

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

the likelihood of high blood pressure. ForAfrican Americans, and especially darker-skinned black men, there was the added in-sult of racist interactions (with police or otherrepresentatives of the white power establish-ment). These racist interactions were in turnlikely to provoke hostility on the part of theblack participant in the interaction, who maythen suppress that hostility to avoid negativerepercussions. The model thus predicted thatdarker-skinned black men who lived in highstress areas and suppressed hostility wouldhave the highest blood pressures. Researchresults have been generally consistent withthese predictions, although the strength of theanger expression and suppression effect hasbeen found to be modest (Schum et al. 2003).

The John Henryism hypothesis (Jameset al. 1983) is named for the mythic blacksteel driver who, in the face of seemingly in-surmountable odds, refused to be deterredin his efforts. In a series of studies, Jamesfound that persons in the black communitywho exhibit this tenacious and active copingstyle have higher blood pressure and a higherprevalence of hypertension if they also havefewer resources, such as higher educational at-tainment, for achieving their goals. While thefindings in research on John Henryism havegenerally been consistent, there have beennegative findings, and in some research the as-sociation has been obscured by other factors(James 1994) or was found to be contingenton variables such as gender (Dressler et al.1998a). This suggests that the John Henryismmodel may be in need of further specificationand elaboration.

Another example of this line of research isthe study of job-related stressors and incident-treated hypertension by Levenstein andcolleagues (2001). This study is relevantbecause institutional racism will result inindividual-level factors such as differential jobsecurity for blacks and whites, which in turnmay account for differences in disease rates.Levenstein et al. (2001) found that job inse-curity and unemployment were potent risksfor developing hypertension over a 20-year

follow-up, especially for men; however, as inother types of research, controlling for thesefactors did not substantially reduce the differ-ences in risk between blacks and whites.

Oths et al. (2001) used a similar line ofreasoning in the development of a model toaccount for ethnic differences in low birthweight. They adapted Karasek’s model of jobstrain, which, in its original form, was devel-oped to examine job stressors among white-collar workers. The central hypothesis of thismodel is that stress results from the imbal-ance of demands on the job and the re-sources a worker has to deal with those de-mands (see Markovitz et al. 2004). Oths et al.(2001) adapted this model to be appropriatefor poor and working-class black and whitewomen in Alabama who work in entry-levelservice jobs (e.g., fast-food restaurants andconvenience stores) and in factories such aspoultry processing, as well as for middle-class women in white-collar jobs. Prospec-tively, they found that women under high de-mands on the job who had little control (interms of being able to take a break, for exam-ple) had lower birth weight babies and thatthe effect was enhanced for black women ver-sus white women. The differences in birthweights between black and white women un-der low strain conditions were minimal; as jobstrain increased, the gap between black andwhite birth weights widened. Those who feltdiscriminated against on the job were nearlythree times more likely to suffer job strain asthose not discriminated against.

Finally, studies by Dressler (1990, 1991a)are relevant here. Like James’s studies ofJohn Henryism, Dressler eschewed the at-tempt to account for racial or ethnic dif-ferences in disease risk, focusing instead onfactors within the African American com-munity. On the basis of ethnographic ob-servations, he adapted the concept of statusincongruity, arguing that individuals’ strug-gles to achieve a middle-class lifestyle in theface of limited economic resources would bea potent stressor. At the same time, tradi-tional features of social organization in the

240 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

black community, especially reliance on theextended family for social support, wouldmoderate that stressor. He found that the in-teraction of status incongruence and socialsupport was associated with blood pressurewithin a Southern black community; how-ever, the interaction of kin support and statusincongruence was significant only for older(>45 years) respondents. For younger respon-dents, nonkin support buffered the pressor ef-fect of status incongruence.

Studies of psychosocial stress processesrepresent a conceptual advance over racial-genetic, health behavior, and SES modelsin one important sense: In the psychosocialstress model, there is an explicit attempt to in-tegrate, at least on some level, what is uniqueabout the experience of the African Ameri-can community in the United States and howthat singular experience generates a particularconfiguration of stressors that in turn is asso-ciated with health and disease. At the sametime, psychosocial stress models to explainracial and ethnic health disparities are subjectto all of the same criticisms of the conven-tional stress model. First, the measurementof many stressors, especially perceived racism,suffers from all the difficulties of distinguish-ing the accurate perception of a stressor fromcognitive and emotional efforts on the part ofan individual to cope with that stressor (Meyer2003), which can lead to complex interpreta-tions of results (Krieger 1990). Second, de-spite considerable refinement over the years,this emphasis on the individual and individ-ual perception tends to deflect attention awayfrom broader social and cultural fields thatgenerate stressors in the first place.

The Structural-ConstructivistModel

The next approach to the study of racialand ethnic disparities will be referred to asa structural-constructivist model, followingBourdieu (1990) and Dressler (2001). This ap-proach to research can be distinguished fromprevious approaches on several levels. First,

it takes into account explicitly the dual na-ture of human existence. On one hand, whatis taken to be the reality of life is in largepart a cognitive representation, constructedout of an amalgam of socially shared under-standings distributed within a society (this istermed a constructivist perspective). On theother hand, individuals are constrained by ex-ternal structures in which they are embed-ded, especially the ecology of social relation-ships created by the shared and distributedexpectations of others. Social, psychological,and biological processes occur within thisintersection of social structure and culturalconstruction.

The need for such a perspective can befound throughout the literature on racial andethnic health disparities in the repeated asser-tion that race is a socially or culturally con-structed concept (Krieger 2003); however, thestudy of how racial and ethnic categories areconstructed, and the implication for healthof those constructions, is rarely attempted.Furthermore, the importance of a construc-tivist perspective can be carried a step fur-ther. The logic underlying most research onhealth disparities is that finding and control-ling for the ways in which blacks and whitesdiffer will undo observed disparities. Thislogic seems unremarkable until examined fur-ther, as Kaufman & Cooper (1999) have done.These investigators argue that regarding raceor ethnicity per se as having true causal poten-tial is misplaced because the logical counter-factual argument cannot be made. The coun-terfactual argument asks what the effect ofrace or ethnicity would be if everyone whois “black” became “white,” and vice-versa.Kaufman & Cooper suggest that this argu-ment makes sense for some variables such asobesity because persons who are overweightcan logically be envisioned to lose weight andthose who are thin can be imagined to gainweight. In what sense can race or ethnicitybe imagined to fit this counterfactual argu-ment? Kaufman & Cooper argue that race orethnicity is such a dominant status category inthe United States that the counterfactual logic

www.annualreviews.org • Race and Ethnicity in Health Research 241

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

DWB: driving whileblack

fails because every aspect of life is dominatedso completely by racial or ethnic status, frombirth to death, that entering race or ethnic-ity into a statistical analysis cannot be readilycausally interpreted (see Berk 2004, pp. 81–103, for a similar but more general discussionof causal logic and categories such as race orethnicity).

Krieger (2003, p. 196) argues that the no-tion of “exchangeability” denied by Kaufman& Cooper does in fact exist because it is possi-ble to imagine a situation in which people, ir-respective of their skin color, are not subject tothe same racist interactions or, in keeping withthe true counterfactual, in which the racist in-teractions would be reversed. But perhaps theissue here is not the subtleties of counterfac-tual logical rigor, but rather the ethnographicrealities; that is, is it reasonable to assume thatthe same understanding or meaning of socialinteraction can be extended to persons in dif-ferent racial or ethnic groups? Probably not.Take for example the phenomenon of DWB(driving while black; see West 1993). The like-lihood and implication of being black and be-ing stopped by police while driving in a whiteneighborhood are different, for example, thanbeing white and being stopped by police whiledriving in a black neighborhood. In this exam-ple, and in many others we suspect, the coun-terfactual argument fails. These are simply in-commensurate phenomena.

Again, the issue is not so much the rigorof causal inference as it is how to examine thephenomenon at hand. It is, perhaps, more im-portant to understand, in an ethnographicallynuanced way, how the goals and aspirationsthat structure mundane social interaction areconstructed within racial and ethnic groups,and how these cultural constructions collidewith the social structure in which they areplayed out.

A number of studies have employed thisperspective. For example, Dressler and asso-ciates (Dressler 1991b, 1999; Dressler et al.1998b, 1999; Dressler & Bindon 2000) usedthis logic in studies of blood pressure in Braziland the United States. These investigators

examined the everyday goals and aspirationsfor a good life shared within communities. InBrazil, both Afro-Brazilians and white Brazil-ians were included, and a general culturalconsensus on a lifestyle (combining mate-rial goods and social behavior) representinga good life was found in two separate stud-ies. In the United States, research was carriedout exclusively within a Southern black com-munity, and a general cultural consensus onwhat constituted a good life was also found.In each setting, a lifestyle of domestic com-fort, not conspicuous consumption, was theconsensus model, and this notion of an incon-spicuous consumption was also emphasized innarrative data. Then, the degree to which in-dividuals were able in their own behaviors toapproximate this valued lifestyle was exam-ined, which is referred to as cultural conso-nance in lifestyle. Individuals who were ableto approximate better the valued lifestyle hadlower blood pressure. Furthermore, in Brazil,there was an interaction effect between cul-tural consonance and skin color, such that thepersons with darker skin color and higher cul-tural consonance had blood pressures lowerthan white Brazilians at any level of culturalconsonance. In the United States, these re-searchers found a similar interaction effectbetween cultural consonance and skin colorwithin the African American community, butthis interaction effect was specific to men aged25–44. These results suggest that the race ofan individual is malleable and subject to inter-pretation in a given social field. Specifically,where individuals can present themselves inmundane social interaction as having achievedwidely shared goals for socioeconomic attain-ment, in the way that these goals are encodedin culturally constructed lifestyles, the bioso-cial significance of skin color recedes.

Gravlee (2002, 2005; Gravlee & Dressler2005) extended this logic in a study of skincolor and blood pressure in Puerto Rico. Anearlier paper (Costas et al. 1981) showed thathigher blood pressure was associated withdarker skin color in Puerto Rico, after con-trolling for a variety of other factors. Gravlee

242 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

investigated the cultural construction of skincolor in Puerto Rico using cultural domainanalysis and cultural consensus analysis. Us-ing a standardized set of facial drawings de-rived from Harris’s (1970) work in Brazil, hefound that respondents agreed on the alloca-tion of phenotypic descriptors to standardizedfaces, and, like Harris in Brazil, he found thatthese attributions were a function of skin colorand hair type. He then investigated the asso-ciation of skin color and blood pressure usingdistinct measures of skin color that included(a) direct measurement of skin pigmentationby reflectance spectrophotometry, (b) self-rated skin color on a nine-point scale, and(c) an estimate of ascribed color derived fromlinking survey respondents to the culturalmodel of skin color, as determined by cul-tural consensus analysis. In bivariate and mul-tivariate analyses, skin reflectance was not as-sociated with blood pressure; however, bothself-rated and ascribed color were associatedwith blood pressure through interactions withSES. Individuals in lower SES groups hadsimilar blood pressures irrespective of theirattributed skin color category, whereas indi-viduals in higher SES groups had higher bloodpressures if they were also assigned to thecategory “black,” according to the culturalconsensus model (Gravlee 2002). Gravlee &Dressler (2005) also report that the discrep-ancy between self-rated color and skin pig-mentation is associated with blood pressurethrough an interaction with SES. They inter-pret this finding as a status incongruity effect.

These studies take seriously the idea ofthe cultural construction of race or ethnicity,as well as how life goals are culturally con-structed within communities of color and theimplications for health when those goals arelimited by racial stratification.

RESEARCH NEEDS

To this point, we have used the terms race orethnicity in the same way they are generallyemployed in the literature, that is, without at-tempting to define them. The lack of explicit

definitions is widely recognized as a signifi-cant barrier to progress in research on healthdisparities. As Crews & Bindon (1991) pointout, biomedical researchers are expected todefine key concepts and to establish the va-lidity and reliability of measurement opera-tions, but the same researchers routinely as-sign participants to racial or ethnic categorieswithout further comment (see also Hahn &Stroup 1994). This pattern obscures healthresearchers’ responsibility to distinguish whatthey know about race as biological scientistsfrom what they know about race as encultur-ated members of society, and it illustrates howmuch the reality of race is taken for granted inthe United States. More important, it impedesefforts to understand the causes of health dis-parities among racially defined groups. Whenrace is treated as a proxy for some unspec-ified combination of environmental, behav-ioral, and genetic factors, rigorous tests ofthe precise causal mechanisms involved arethe exception, not the rule. For research toprogress, a conceptual model of race and eth-nicity is required; indeed, there is some move-ment in the biomedical literature to requirethe definition of race and ethnicity, and tospecify their relevance to the study, when theterms are used in publication (Davidoff 2000).We examine some of the ways in which sucha conceptual model might be developed andoperationalized.

Ethnoracialcategories:culturallyconstructed or folkcategories thatdenote essentialdifferencesconceived in terms ofbiological ancestry

We assume that the term race has no uni-versal biological referent when applied to thehuman species (Kittles & Weiss 2003); there-fore, as it is asserted frequently in the lit-erature, race is a cultural construct used bymembers of a society to explain perceived bi-ological differences among humans in spe-cific ethnographic settings. That being thecase, the most suitable terms for use in re-search would be “ethnoracial categories,” orperhaps “folk racial categories.” It is then in-cumbent on researchers to demonstrate, inany given setting, the cultural model that gen-erates ethnoracial categories and how thesecategories are employed in that ethnographiccontext. Research models exist, notably

www.annualreviews.org • Race and Ethnicity in Health Research 243

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

Harris (1970), Harris et al. (1993), Byrneet al. (1995), and Gravlee (2005). Using elic-itation techniques from cognitive anthropol-ogy, these researchers have systematically de-scribed ethnoracial descriptors where they arerelevant and demonstrated how those descrip-tors are or can be applied in specific ethno-graphic contexts. These techniques, especiallywhen combined with intensive ethnographicmethods, can provide a clear picture of thefolk racial model when the concept is em-ployed in a given setting (e.g., the link ofHarris’s work with Burdick’s (1998) in Brazil).Then, in epidemiologic surveys, when an in-dividual self-selects an ethnoracial category,or when such an ethnoracial category is ap-plied to an individual, the real socioculturalimport of that attribution can be understood(Gravlee 2005). Although it may seem like adaunting task to include such an ethnographi-cally detailed analysis in every study, it is prob-ably unnecessary to replicate these analyseseach time health research is conducted in thatsetting. It is more important that researchersare aware that folk racial categories are emic,ethnographically contingent constructs, andthat they make every effort to link their spe-cific operational procedures to what is under-stood about prevailing ethnoracial models inthat specific setting (e.g., Dressler et al. 1999).

In contrast, the term ethnicity can be ap-plied universally as an analytic construct. Theethnographic record suggests that there maybe a universal tendency to differentiate indi-viduals and social groups on the basis of fac-tors generally associated with attributions oftraditions and ancestry and the way in whichthose attributions are realized in the present(Gil-White 2001). Ultimately, ethnicity be-comes a fundamental way to define socialboundaries (Barth 1969). In a sense, ethnic-ity becomes a primary category in the analysisof any society. It is a category of socioculturalsystems analysis on par with the economic sys-tem, the kinship system, or the system of reli-gious belief and practice. Elaborated further,ethnicity is a social institution like kinship ormarriage, but a higher-level one in that, as an

organizing tool that assigns identity to mem-bers of a group, it actually incorporates, or ispartly constituted by, aspects of other institu-tions (kinship, economy). To be sure, ethnicgroup differences are culturally constructedwithin any given society. But what separatesethnic group differences and ethnoracial dif-ferences is that ethnic group differences mayor may not include a folk racial component.The use of the term race, in contrast, de-mands that an ethnobiological theory (or folkmodel of essential biological difference) bedemonstrated.

Elements of a definition of ethnicity havebeen offered by many, notably Montagu(1962), Barth (1969), Dominguez (1986),Crews & Bindon (1991), Gaines (2005), Oths(1999), and Gravlee (2005). The definition ofethnic groups within a society will incorpo-rate any of a number of dimensions that can beplaced into three broad categories—the cul-tural, the ancestral, and the referential—thesalient features of which will vary betweengroups. The cultural includes shared mod-els for both the mundane (e.g., language use,diet, dress, marriage rituals) and the more ab-stract (e.g., concepts of self, supernatural be-liefs) aspects of life. A sense of shared an-cestry includes territorial homeland, commonhistory (which may include ethnoracial dis-crimination), and kinship (whether construedbiologically), which may or may not incor-porate phenotypic or genotypic characteris-tics such as hair type, body build, or skintone. With respect to the referential, as eth-nic group labels fundamentally separate peo-ple into in- and out-groups (i.e., “we” ver-sus “they”), personal (or self-defined) andsocial (or other-defined) identity is an in-tegral component of ethnic definition. Folkracial categories may then be indexed here,when relevant, as an emic self-categorization,or as an eticly imposed descriptor used byothers.

In summary, rather than retaining the termrace with any kind of etic biological conno-tation, it can be seen as a part of a mean-ingful folk taxonomy that may (if present)

244 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

be incorporated into a more robust classifica-tion with worldwide applicability. Therefore,instead of finding that “current racial cat-egories capture ethnic status,” as Williams(1997, p. 325) holds, we prefer a model inwhich ethnic categories would subsume eth-noracial categories.

It should be clear that the scheme we sug-gest here invites the study of genetic factors inthe distribution of disease, in terms of the ex-plicit identification of genetic variants. Sucha scheme would distinguish the distributionof genetic variants, folk racial categories, andethnicity; each term would carve out a dis-tinctive phenomenon for analysis in a theo-retically and operationally explicit way. Wemight then be able, for example, to describethe distribution of disease separately in re-lation to genetic variants, ethnoracial cate-gories, and ethnicity, and in relation to thecombination of these factors. Furthermore,the associations of genetic variants, ethnora-cial categories, and ethnicity can themselvesbecome a focus of study. More to the point,however, our review of existing literature sug-gests that these three dimensions of humanbiology and social life have been routinelyoffered as risk factors for disease, while si-multaneously being routinely conflated. Theproposed scheme identifies factors believedimportant in existing literature and offers thepotential to distinguish among them in futureresearch.

CONCLUSION

As shown in quantitative analyses of the so-cial science and public health literature, eth-noracial categories and ethnicity have becomenearly standard variables included in researchon the distribution of disease. And, on nearlyevery indicator, but especially on several ofthe most important contributors to earlymortality, black Americans and white Amer-icans differ; black Americans suffer highermortality and morbidity. These comparisonshave, however, generated little understand-ing of these health disparities because they

SOCIAL STRUCTURE AND CULTURALCONSTRUCTIVISM

The dual nature of human existence has been prominent inthe history of social theory. Other terms could easily be sub-stituted for social structure and cultural construction, but thebasic issue has remained the same for more than 200 years: Ishuman behavior fundamentally a function of the external con-straints imposed on groups and individuals, or is it a functionof the way in which groups and individuals impose a mean-ingful structure on the world? Although most working socialscientists probably understand their subject matter as an inter-action of these aspects of human life, surprisingly little formaltheory has tried to integrate the two. An exception to this isthe earlier work of Pierre Bourdieu, who proposed a set ofconstructs (e.g., “habitus,” “social field,” “cultural space”) inwhich human behavior could be understood at the intersec-tion of social structure and cultural construction. This is aparticularly fruitful approach in the study of health disparitiesbecause it draws attention to the cultural construction of, forexample, racial categories, while being sensitive to the causalpotential of these arbitrary categories when there is a collec-tive acceptance of their “reality.”

are reported mostly without comment or, per-haps worse, with vague comments implicat-ing racial-genetic differences. A careful reviewof the literature indicates that such imputingof a racial-genetic basis for disease is with-out foundation, yet such suggestions cannotbut reinforce the general American culturalmodel of ethnoracial categories.

Other studies have sought the nonracial-genetic basis for these differences, and re-search on discrete health behaviors, althoughimportant in part to be sure, suggests thatthese cannot account for overall group differ-ences. What appears to offer a more potentexplanation for ethnoracial and ethnic healthdisparities is a model in which other formsof inequality, especially social and economicinequalities, generate life conditions that arechronically stressful over the life course ofblack Americans. What Geronimus (1992)refers to as “weathering,” or the chronic,allostatic load generated by this continuing

www.annualreviews.org • Race and Ethnicity in Health Research 245

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

adaptation to enduring structures of inequal-ity, then generates observed health disparities.It is a collision of the cultural construction ofmundane life goals with a social structure ofethnoracial stratification.

Understanding this process in terms of allits empirical complexity ultimately will re-quire a reconceptualization of key factors inthe process, notably the basic concepts of folkracial categories and ethnicity.

ACKNOWLEDGMENTS

The authors have benefited over the years from discussion of these issues with Russ Bernard,Jim Bindon, Woody Gaines, Sherman James, and the late Marvin Harris. The authors aloneare responsible for any errors.

SUMMARY POINTS

1. Differences in morbidity and mortality between conventionally defined racial andethnic groups have been widely documented, but these differences continue to bepoorly understood.

2. Although overall racial and ethnic group differences are complex, the largest dispari-ties are those between African Americans and European Americans. On virtually everyindicator of morbidity and mortality, blacks suffer in relation to whites.

3. A review of the literature on health disparities reveals five types of explanatory modelsthat have been employed to account for these differences; each explanatory modelemphasizes different sets of variables.

4. Models that emphasize both psychological and sociocultural factors in the causes ofhealth disparities appear to be most promising.

5. Future progress in this area will depend on the development of a satisfactory theoryof ethnic differences.

FUTURE DIRECTIONS/UNRESOLVED ISSUES

1. The conceptualization and measurement of the social and cultural dimensions of raceand ethnicity need to be improved.

2. More cross-cultural research is needed to accumulate empirical results outside thespecific sociocultural constraints of North American and Western European societies.

2. The degree to which ethnoracial and other traits construct ethnicity in a given groupis an empirical question that should be examined cross-culturally.

4. Future research needs to specify tests of hypotheses that include measures of bothgenetic and nongenetic differences between ethnic groups.

246 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

LITERATURE CITED

Acevedo-Garcia D, Soobader MJ, Berkman LF. 2005. The differential effect of foreign-bornstatus on low birth weight by race/ethnicity and education. Pediatrics 115:20–30

Almeida-Filho N, Kawachi I, Filho AP, Dachs JWD. 2003. Research on health inequalities inLatin America and the Caribbean. Am. J. Public Health 93:2037–43

Barker DJP. 2004. The developmental origins of well-being. Philos. Trans. R. Soc. London Ser.B 259:1359–66

Barth F. 1969. Ethnic Groups and Boundaries: The Social Organization of Culture Difference. Boston,MA: Little, Brown

Bassett DR, Futzhugh EC, Crespo CJ, King GA, McLaughlin JE. 2002. Physical activity andethnic differences in hypertension prevalence in the United States. Prev. Med. 34:179–86

Bell AC, Adair LS, Popkin BM. 2004. Understanding the role of mediating risk factors andproxy effects in the association between socio-economic status and untreated hypertension.Soc. Sci. Med. 59:275–83

Berk RA. 2004. Regression Analysis: A Constructive Critique. Thousand Oaks, CA: SageBourdieu P. 1990. In Other Words: Essays Towards a Reflexive Sociology. Transl. M Adamson.

Stanford, CA: Stanford Univ. Press (From French)Bouzekri N, Zhu X, Jiang Y, McKenzie CA, Luke A, et al. 2004. Angiotensin I-converting

enzyme polymorphisms, ACE level and blood pressure among Nigerians, Jamaicans andAfrican-Americans. Eur. J. Hum. Genet. 12:460–68

Boyle E. 1970. Biological patterns in hypertension by race, sex, body weight, and skin color.JAMA 213:1637–43

Braun L. 2002. Race, ethnicity, and health: Can genetics explain disparities? Perspect. Biol. Med.45:159–74

Brondolo E, Rieppi R, Kelly KP, Gerin W. 2003. Perceived racism and blood pressure: a reviewof the literature and conceptual and methodological critique. Ann. Behav. Med. 25:55–65

Brownley KA, Hurwitz BE, Schneiderman N. 1999. Ethnic variations in pharmacological andnonpharmacological treatment of hypertension. Hum. Biol. 71:607–40

Burdick J. 1998. Blessed Anastascia: Women, Race, and Popular Christianity in Brazil. New York:Routledge

Byrne B, Harris M, Consorte JG, Lang J. 1995. What’s in a name? The consequences of vio-lating Brazilian emic, color-race categories in estimates of social well-being. J. Anthropol.Res. 51:389–97

Carlos Poston WS, Pavlik VN, Hyman DJ, Ogbonnaya K, Hanis CL, et al. 2001. Geneticbottlenecks, perceived racism, and hypertension risk among African Americans and first-generation African immigrants. J. Hum. Hypertens. 15:341–51

Cent. Dis. Control. 2004a. National Center for Health Statistics. http://www.cdc.gov/nchs/data/hus/tables/2003/03hus066.pdf

Cent. Dis. Control. 2004b. National Center for Health Statistics. http://www.cdc.gov/nchs/fastats/pdf/nvsr52 10t43.pdf.

Cent. Dis. Control. 2004c. National Center for Health Statistics. http://www.cdc.gov/nchs/data/hus/hus04trend.pdf#069

Chrysant SG, Weir MR, Weder AB, McCarron DA, Canoss-Terris M, et al. 1997. There are noracial, age, sex, or weight differences in the effect of salt on blood pressure in salt-sensitivehypertensive patients. Arch. Int. Med. 157:2489–94

Clark R, Anderson NB, Clark VR, Williams DR. 1999. Racism as a stressor for African Amer-icans: a biopsychosocial model. Am. Psychol. 54:805–16

www.annualreviews.org • Race and Ethnicity in Health Research 247

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

Collins JW Jr, David RJ, Handler A, Wall S, Andes S. 2004. Very low birthweight in AfricanAmerican infants: the role of maternal exposure to interpersonal racial discrimination.Am. J. Public Health 94:2132–38

Comstock RD, Castillo EM, Lindsay SP. 2004. Four-year review of the use of race and ethnicityin epidemiologic and public health research. Am. J. Epidemiol. 159:611–19

A historicallyimportant paperbecause it was anearly challenge toconventional racialthinking in theprofessionalbiomedicalliterature, drawingespecially on theanthropologicalwork of AshleyMontagu.

Cooper RS. 1984. A note on the biologic concept of race and its application in epidemi-ologic research. Am. Heart J. 108:715–23

Researchersdemonstrateunambiguouslythat groups ofAfrican descentexhibit differentrates ofhypertension indifferentenvironmentalsettings, thuschallenging thenotion thathypertension isunder exclusiveracial-geneticcontrol.

Cooper RS, Rotimi C, Ataman S, McGee D, Osotimehin B, et al. 1997. The prevalenceof hypertension in seven populations of West African origin. Am. J. Public Health87:160–68

Cooper RS, Rotimi C, Ward R. 1999. The puzzle of hypertension in African-Americans. Sci.Am. 280:56–63

Costas R Jr, Garcia-Palmieri MR, Sorlie P, Hertzmark E. 1981. Coronary heart disease riskfactors in men with light and dark skin in Puerto Rico. Am. J. Public Health 71:614–19

Crews DE, Bindon JR. 1991. Ethnicity as a taxonomic tool in biomedical and biosocial research.Ethn. Dis. 1:42–49

Crews DE, Williams SR. 1999. Molecular aspects of blood pressure regulation. Hum. Biol.71:475–504

Curtin PD. 1992. The slavery hypothesis for hypertension among African Americans: thehistorical evidence. Am. J. Public Health 82:1681–86

Daniel HI, Rotimi CN. 2003. Genetic epidemiology of hypertension: an update on the Africandiaspora. Ethn. Dis. 13:S53–66

Davey Smith G. 2000. Learning to live with complexity: ethnicity, socioeconomic position,and health in Britain and the United States. Am. J. Public Health 90:1694–98

David RJ. 2001. Birthweights and bell curves. Int. J. Epidemiol. 30:1241–43David RJ, Collins JW Jr. 1997. Differing birth weight among infants of U.S.-born blacks,

African-born blacks, and U.S.-born whites. N. Engl. J. Med. 337:1209–14Davidoff F. 2000. News from the international committee of medical journal editors. Ann.

Intern. Med. 133:229–31Davidson K, Jonas BS, Dixon KE, Markovitz JH. 2000. Do depressive symptoms predict

early hypertension incidence in young adults in the CARDIA study? Arch. Intern. Med.160:1495–500

Din-Dzietham R, Nembhard WN, Collins R, Davis SK. 2004. Perceived stress following race-based discrimination at work is associated with hypertension in African-Americans. Soc.Sci. Med. 58:449–61

Dominguez VR. 1986. White by Definition: Social Classification in Creole Louisiana. NewBrunswick, NJ: Rutgers Univ. Press

Dressler WW. 1990. Lifestyle, stress, and blood pressure in a Southern black community.Psychosom. Med. 52:182–98

Dressler WW. 1991a. Social support, lifestyle incongruity, and arterial blood pressure in aSouthern black community. Psychosom. Med. 53:608–20

Dressler WW. 1991b. Social class, skin color, and arterial blood pressure in two societies. Ethn.Dis. 1:60–77

Dressler WW. 1993. Health in the African-American community: accounting for health in-equalities. Med. Anthropol. Q. 7:325–45

Dressler WW. 1999. Culture, skin color and arterial blood pressure. Presented at Annu. Meet. Am.Anthropol. Assoc., 98th, Chicago

Dressler WW. 2001. Medical anthropology: toward a third moment in social science? Med.Anthropol. Q. 15:455–65

248 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

Dressler WW, Balieiro MC, Santos JED. 1999. Culture, skin color, and arterial blood pressurein Brazil. Am. J. Hum. Biol. 11:49–59

Dressler WW, Bindon JR. 2000. The health consequences of cultural consonance: culturaldimensions of lifestyle, social support and arterial blood pressure in an African Americancommunity. Am. Anthropol. 102:244–60

Dressler WW, Bindon JR, Neggers YR. 1998a. John Henryism, gender, and arterial bloodpressure in an African American community. Psychosom. Med. 60:620–24

Dressler WW, Bindon JR, Neggers YR. 1998b. Culture, socioeconomic status, and coronaryheart disease risk factors in an African American community. J. Behav. Med. 21:527–44

Drevdahl D, Taylor JY, Phillips DA. 2001. Race and ethnicity as variables in Nursing Research,1952–2000. Nurs. Res. 50:305–13

Dyer AR, Liu K, Walsh M, Kiefe C, Jacombs DR Jr, Bild DE. 1999. Ten-year incidence ofelevated blood pressure and its predictors. J. Hum. Hyperten. 13:13–20

Ellison GTH, de Wet T. 1997. The use of ‘racial’ categories in contemporary South Africanhealth research. A survey of articles published in the South African Medical Journal between1992 and 1996. S. Afr. Med. J. 87:1671–79

Estruch R, Coca A, Rodicio JL. 2004. High blood pressure, alcohol, and cardiovascular risk.Eur. Soc. Hyperten. Sci. Newsl. 5:1–3

Fuchs FD, Chambless LE, Whelton PK, Nieto FJ, Heiss G. 2001. Alcohol consumption andthe incidence of hypertension. Hypertension 37:1242–50

Gaines AD. 2005. Race: Local biology and culture in mind. In Companion to Psychological An-thropology, ed. C Casey, RB Edgerton, pp. 255–78. Oxford: Blackwell

Geronimus AT. 1992. The weathering hypothesis and the health of African-American womenand infants. Ethn. Dis. 2:207–21

Gil-White FJ. 2001. Are ethnic groups “biological species” to the human brain? Essentialismin our cognition of some social categories. Curr. Anthropol. 42:515–54

Gravlee CC. 2002. Skin color, blood pressure, and the contextual effect of culture in southeastern PuertoRico. PhD thesis. Univ. Fla., Gainesville

Gravlee CC. 2005. Emic ethnic classification in southeastern Puerto Rico: cultural consensusand semantic structure. Soc. Forces 83:949–70

Gravlee CC, Dressler WW. 2005. Skin pigmentation, self-perceived color, and arterial bloodpressure in Puerto Rico. Am. J. Hum. Biol. 17:195–206

Gravlee CC, Sweet E, Abts M. 2004. Race, ethnicity, and racism in medical anthropology, 1977–2002. Presented at Soc. Appl. Anthropol., Dallas, TX

An articulation ofthe so-called“slaveryhypothesis” thatposits aracial-geneticpropensity tohypertensionamong AfricanAmericans.

Grim CE, Robinson M. 1996. Blood pressure variation in blacks: genetic factors. Semin.Nephrol. 16:83–93

Hahn R, Stroup DF. 1994. Race and ethnicity in public health surveillance: criteria for thescientific use of social categories. Public Health Rep. 109:7–15

Harburg EL, Erfurt JC, Hauenstein LS. 1973. Socioecologic stress, suppressed hostility, skincolor, and black-white male blood pressure: Detroit. Psychosom. Med. 35:276–96

Harburg E, Gleiberman L, Roeper P, Schork MA, Schull WJ. 1978. Skin color, ethnicity, andblood pressure I: Detroit blacks. Am. J. Public Health 68:1177–83

Harrap SB. 2003. Where are all the blood pressure genes? The Lancet 361:2149–51Harris M. 1970. Referential ambiguity in the calculus of Brazilian racial identity. Southwest. J.

Anthropol. 26:1–14Harris M, Consorte JG, Lang J, Byrne B. 1993. Who are the whites—imposed census categories

and the racial demography of Brazil. Soc. Forces 72:451–62Homer CJ, Beresford SAA, James SA, Siegel E, Wilcox S. 1990. Work-related physical exertion

and risk of preterm, low birthweight delivery. Paediatr. Perinat. Epidemiol. 4:161–74

www.annualreviews.org • Race and Ethnicity in Health Research 249

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

Inst. Med. Comm. Nutr. Status During Pregnancy Lactation. 1990. Nutrition During Pregnancy:Part I: Weight Gain. Part II: Nutrient Supplements. Washington, DC: Natl. Acad. Press

Jackson FL. 1991. An evolutionary perspective on salt, hypertension, and human genetic vari-ability. Hypertension 17:1129–32

James SA. 1994. John Henryism and the health of African Americans. Cult. Med. Psychiatry18:163–82

An early paperarticulating anongenetichypothesis toaccount forblack-whitedifferences in bloodpressure, notablefor its sensitivity tothe ethnographiccontext.

James SA, Hartnett SA, Kalsbeek WD. 1983. John Henryism and blood pressure differ-ences among black men. J. Behav. Med. 6:259–78

Janzon E, Hedblad B, Berglund G, Engstrom G. 2004. Changes in blood pressure and bodyweight following smoking cessation in women. J. Intern. Med. 255:266–72

Jonas BS, Franks P, Ingram DD. 1997. Are symptoms of anxiety and depression risk factorsfor hypertension? Arch. Fam. Med. 6:43–49

Jonas BS, Lando JF. 2000. Negative affect as a prospective risk factor for hypertension. Psycho-som. Med. 62:188–96

Jones CP, LaVeist TA, Lillie-Blanton M. 1991. “Race” in the epidemiologic literature: an exam-ination of the American Journal of Epidemiology, 1921–1990. Am. J. Epidemiol. 134:1079–84

Jones-Webb R, Jacombs DR, Flack JM, Liu K. 1996. Relationships between depressive symp-toms, anxiety, alcohol consumption, and blood pressure. Alcohol. Clin. Exp. Res 20:420–27

Kaplan JB, Bennett T. 2003. Use of race and ethnicity in biomedical publication. JAMA289:2709–16

Kaufman JS, Cooper RS. 1999. Seeking causal explanations in social epidemiology. Am. J.Epidemiol. 154:291–98

Kaufman JS, Cooper RS, McGee DL. 1997. Socioeconomic status and health in blacks andwhites: the problem of residual confounding and the resiliency of race. Epidemiology 8:621–28

A critique of theso-called “slaveryhypothesis” forhypertension,notable in the waythat it traces theintellectualdiffusion of ahypothesis that haslittle empiricalsupport.

Kaufman JS, Hall SA. 2002. The slavery hypertension hypothesis: dissemination andappeal of a modern race theory. Epidemiology 14:111–18

Keil JE, Tyroler HA, Sandifer SH, Boyle E Jr. 1977. Hypertension: effects of social class andracial admixture. Am. J. Public Health 67:634–39

Keppel KG, Pearch JN, Wagener DK. 2002. Trends in racial and ethnic-specific rates for thehealth status indicators: United States, 1990–98. Healthy People Statistical Notes No. 23.Hyattsville, MD: Natl. Cent. Health Stat.

Kittles RA, Weiss KM. 2003. Race, ancestry, and genes: implications for defining disease risk.Annu. Rev. Genomics Hum. Genet. 4:33–67

Klag MJ, Whelton PK, Coresh J, Grim C, Kuller LH. 1991. The association of skin color withblood pressure in U.S. blacks with low socioeconomic status. JAMA 265:599–602

Kleinman JC, Fingerhut LA, Prager K. 1991. Differences in infant mortality by race, nativitystatus, and other maternal characteristics. Am. J. Dis. Child. 145:194–99

Knox SS, Hausdorff J, Markovitz JH. 2002. Reactivity as a predictor of subsequent bloodpressure. Hypertension 40:914–19

Krieger N. 1990. Racial and gender discrimination: risk factors for high blood pressure? Soc.Sci. Med. 30:1273–81

Krieger N. 1999. Embodying inequality: a review of concepts, measures, and methods forstudying health consequences of discrimination. Int. J. Health Serv. 29:295–352

Krieger N. 2003. Does racism harm health? Did child abuse exist before 1982? On explicitquestions, critical science, and current controversies: an ecosocial perspective. Am. J. PublicHealth 93:194–99

Krieger N, Sidney S. 1996. Racial discrimination and blood pressure: the CARDIA study ofyoung black and white adults. Am. J. Public Health 86:1370–78

250 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

Levenstein S, Smith MW, Kaplan GA. 2001. Psychosocial predictors of hypertension in menand women. Arch. Intern. Med. 161:1341–46

Liu K, Ruth KJ, Flack JM, Jones-Webb R, Burke G, et al. 1996. Blood pressure in young blacksand whites. Circulation 93:60–66

Markovitz JH, Matthews KA, Whooley M, Lewis CE, Greenlund KJ. 2004. Increases in jobstrain are associated with incident hypertension in the CARDIA Study. Ann. Behav. Med.28:4–9

McCarron DA. 2000. The dietary guideline for sodium. Am. J. Clin. Nutr. 71:1013–19McCormick MC, Brooks-Gunn J, Shorter T, Holmes JH, Wallace CY, Heagarty MC. 1990.

Factors associated with smoking in low-income pregnant women: relationship to birthweight, stressful life events, social support, health behaviors and mental distress. J. Clin.Epidemiol. 3:441–48

Meyer IH. 2003. Predjudice as stress: conceptual and measurement problems. Am. J. PublicHealth 93:262–65

A classic paper thatchallenges thevalidity ofconventional racialcategories.

Montagu A. 1962. The concept of race. Am. Anthropol. 64:919–28Mustillo S, Krieger N, Gunderson EP, Sidney S, McCreath H, Kiefe CI. 2004. Self-reported

experiences of racial discrimination and black-white differences in preterm and low-birthweight deliveries: the CARDIA Study. Am. J. Public Health 94:2125–31

Nickens H. 1986. Health problems of minority groups: public health’s unfinished agenda. PublicHealth Rep. 101:230–31

Oparil S, Zaman MA, Calhoun DA. 2003. Pathogenesis of hypertension. Ann. Intern. Med.139:761–76

Osborne NG, Feit MD. 1992. The use of race in medical research. JAMA 267:275–79Oths KS. 1999. Who’s who in Alabama prenatal clinics. Presented at Annu. Meet. Am. Anthropol.

Assoc., 98th, ChicagoOths KS, Dunn LL, Palmer NS. 2001. A prospective study of psychosocial job strain and birth

outcomes. Epidemiology 12:744–46Sankar P, Cho MK, Condit CM, Hunt LM, Koenig B, et al. 2004. Genetic research and health

disparities. JAMA 291:2985–89Schoenborn CA, Adams PF, Barnes PM, Vickerie JL, Schiller JS. 2004. Health Behaviors of

Adults: United States, 1999–2001. Natl. Cent. Health Stat., Vital Health Stat. 10(219).Washington, DC: US GPO

Schum JL, Jorgensen RS, Verhaeghen P, Sauro M, Thibodeau R. 2003. Trait anger, angerexpression, and ambulatory blood pressure: a meta-analytic review. J. Behav. Med. 26:395–415

Sheldon TA, Parker H. 1992. Race and ethnicity in health research. J. Public Health Med.14:104–10

Sowers JR, Ferdinand KC, Bakris GL, Douglas JG. 2002. Hypertension-related disease inAfrican Americans. Postgrad. Med. 112:24–34

Stamler J, Elliott P, Appel L, Chan Q, Buzzard M, et al. 2003. Higher blood pressure inmiddle-aged American adults with less education—role of multiple dietary factors. J.Hum. Hyperten. 17:655–75

Starfield B, Shapiro S, Weiss J, Liang KY, Ra K, et al. 1991. Race, family income, and low birthweight. Am. J. Epidemiol. 134:1167–74

Teitelman AM, Welch LS, Hellenbrand KG, Bracken MB. 1990. Effect of maternal workactivity on preterm birth and low birth weight. Am. J. Epidemiol. 131:104–13

The analysis ofsocioeconomic andregional disparitiesin health in Britainthat was central inarticulating theissue of healthdisparities as animportant topic ofresearch.

Townsend P, Davidson N. 1982. Inequalities in Health: The Black Report. Har-mondsworth, UK: Penguin

www.annualreviews.org • Race and Ethnicity in Health Research 251

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

AR254-AN34-13 ARI 25 August 2005 15:1

US Dep. Health Hum. Serv. 2000. Healthy People 2010: Understanding and Improving Health.Washington, DC: US GPO. 2nd ed.

West C. 1993. Race Matters. Boston, MA: BeaconWilcox AJ, Russell I. 1990. Why small black infants have a lower mortality rate than small

white infants: the case for population-specific standards for birth weight. J. Pediatr. 116:7–10

Williams DR. 1994. The concept of race in Health Services Research: 1966 to 1990. Health Serv.Res. 29:261–74

Williams DR. 1997. Race and health: basic questions, emerging directions. Ann. Epidemiol.7:322–33

Williams DR. 1999. Monitoring racial/ethnic status in the U.S.: data quality issues. Ethn. Health4:121–37

Williams DR, Collins C. 1995. U.S. socioeconomic and racial differences in health: patternsand explanations. Annu. Rev. Sociol. 21:349–86

Williams DR, Collins C. 2001. Racial residential segregation: a fundamental cause of the racialdisparities in health. Public Health Rep. 116:404–16

Williams DR, Neighbors HW, Jackson JS. 2003. Racial/ethnic discrimination and health:findings from community studies. Am. J. Public Health 93:200–8

Wong MD, Shapiro MF, Boscardin WJ, Ettner SL. 2002. Contribution of major diseases todisparities in mortality. N. Engl. J. Med. 347:1585–92

Woolf SH, Johnson RE, Fryer GE Jr, Rust G, Satcher D. 2004. The health impact of re-solving racial disparities: an analysis of US mortality data. Am. J. Public Health 94:2078–81

Wyatt SB, Williams DR, Calvin R, Henderson FC, Walker ER, Winters K. 2003. Racism andcardiovascular disease in African Americans. Am. J. Med. Sci. 325:315–31

RELATED RESOURCES

The Problem of RaceF BoasAnthropology and Modern Life. 1960. Pages 18–62. New York: Norton

On the Race ConceptCL BraceCurrent Anthropology. 1964. Volume 5, pages 313–18

Dusk of Dawn: An Essay Toward an Autobiography of a Race ConceptWEB DuBois1940. New York: Harcourt, Brace and Co.

Inconsistencies in Coding of Race and Ethnicity Between Birth and Death in US Infants.A New Look at Infant Mortality, 1983 through 1985RA Hahn, J Mulinare, SM TeutschJournal of the American Medical Association. 1992. Volume 267, Issue 2, pages 259–63

Man’s Most Dangerous Myth: The Fallacy of RaceA Montagu1942. New York: Columbia University Press

252 Dressler • Oths • Gravlee

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

Contents ARI 12 August 2005 20:29

Annual Review ofAnthropology

Volume 34, 2005

Contents

FrontispieceSally Falk Moore � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � xvi

Prefatory Chapter

Comparisons: Possible and ImpossibleSally Falk Moore � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

Archaeology

Archaeology, Ecological History, and ConservationFrances M. Hayashida � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �43

Archaeology of the BodyRosemary A. Joyce � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 139

Looting and the World’s Archaeological Heritage: The InadequateResponseNeil Brodie and Colin Renfrew � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 343

Through Wary Eyes: Indigenous Perspectives on ArchaeologyJoe Watkins � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 429

The Archaeology of Black Americans in Recent TimesMark P. Leone, Cheryl Janifer LaRoche, and Jennifer J. Babiarz � � � � � � � � � � � � � � � � � � � � � � � 575

Biological Anthropology

Early Modern HumansErik Trinkaus � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 207

Metabolic Adaptation in Indigenous Siberian PopulationsWilliam R. Leonard, J. Josh Snodgrass, and Mark V. Sorensen � � � � � � � � � � � � � � � � � � � � � � � � � � 451

The Ecologies of Human Immune FunctionThomas W. McDade � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 495

vii

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

Contents ARI 12 August 2005 20:29

Linguistics and Communicative Practices

New Directions in Pidgin and Creole StudiesMarlyse Baptista � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �33

Pierre Bourdieu and the Practices of LanguageWilliam F. Hanks � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �67

Areal Linguistics and Mainland Southeast AsiaN.J. Enfield � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 181

Communicability, Racial Discourse, and DiseaseCharles L. Briggs � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 269

Will Indigenous Languages Survive?Michael Walsh � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 293

Linguistic, Cultural, and Biological DiversityLuisa Maffi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 599

International Anthropology and Regional Studies

Caste and Politics: Identity Over SystemDipankar Gupta � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 409

Indigenous Movements in AustraliaFrancesca Merlan � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 473

Indigenous Movements in Latin America, 1992–2004: Controversies,Ironies, New DirectionsJean E. Jackson and Kay B. Warren � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 549

Sociocultural Anthropology

The Cultural Politics of Body SizeHelen Gremillion � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �13

Too Much for Too Few: Problems of Indigenous Land Rights in LatinAmericaAnthony Stocks � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �85

Intellectuals and Nationalism: Anthropological EngagementsDominic Boyer and Claudio Lomnitz � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 105

The Effect of Market Economies on the Well-Being of IndigenousPeoples and on Their Use of Renewable Natural ResourcesRicardo Godoy, Victoria Reyes-Garcıa, Elizabeth Byron, William R. Leonard,

and Vincent Vadez � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 121

viii Contents

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

Contents ARI 12 August 2005 20:29

An Excess of Description: Ethnography, Race, and Visual TechnologiesDeborah Poole � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 159

Race and Ethnicity in Public Health Research: Models to ExplainHealth DisparitiesWilliam W. Dressler, Kathryn S. Oths, and Clarence C. Gravlee � � � � � � � � � � � � � � � � � � � � � � � � 231

Recent Ethnographic Research on North American IndigenousPeoplesPauline Turner Strong � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 253

The Anthropology of the Beginnings and Ends of LifeSharon R. Kaufman and Lynn M. Morgan � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 317

Immigrant Racialization and the New Savage Slot: Race, Migration,and Immigration in the New EuropePaul A. Silverstein � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 363

Autochthony: Local or Global? New Modes in the Struggle overCitizenship and Belonging in Africa and EuropeBambi Ceuppens and Peter Geschiere � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 385

Caste and Politics: Identity Over SystemDipankar Gupta � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 409

The Evolution of Human Physical AttractivenessSteven W. Gangestad and Glenn J. Scheyd � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 523

Mapping Indigenous LandsMac Chapin, Zachary Lamb, and Bill Threlkeld � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 619

Human Rights, Biomedical Science, and Infectious Diseases AmongSouth American Indigenous GroupsA. Magdalena Hurtado, Carol A. Lambourne, Paul James, Kim Hill,

Karen Cheman, and Keely Baca � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 639

Interrogating Racism: Toward an Antiracist AnthropologyLeith Mullings � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 667

Enhancement Technologies and the BodyLinda F. Hogle � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 695

Social and Cultural Policies Toward Indigenous Peoples: Perspectivesfrom Latin AmericaGuillermo de la Pena � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 717

Surfacing the Body InteriorJanelle S. Taylor � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 741

Contents ix

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

Contents ARI 12 August 2005 20:29

Theme 1: Race and Racism

Race and Ethnicity in Public Health Research: Models to ExplainHealth DisparitiesWilliam W. Dressler, Kathryn S. Oths, and Clarence C. Gravlee � � � � � � � � � � � � � � � � � � � � � � � � 231

Communicability, Racial Discourse, and DiseaseCharles L. Briggs � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 269

Immigrant Racialization and the New Savage Slot: Race, Migration,and Immigration in the New EuropePaul A. Silverstein � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 363

The Archaeology of Black Americans in Recent TimesMark P. Leone, Cheryl Janifer LaRoche, and Jennifer J. Babiarz � � � � � � � � � � � � � � � � � � � � � � � 575

Interrogating Racism: Toward an Antiracist AnthropologyLeith Mullings � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 667

Theme 2: Indigenous Peoples

The Effect of Market Economies on the Well-Being of IndigenousPeoples and on Their Use of Renewable Natural ResourcesRicardo Godoy, Victoria Reyes-Garcıa, Elizabeth Byron, William R. Leonard,

and Vincent Vadez � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 121

Recent Ethnographic Research on North American IndigenousPeoplesPauline Turner Strong � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 253

Will Indigenous Languages Survive?Michael Walsh � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 293

Autochthony: Local or Global? New Modes in the Struggle overCitizenship and Belonging in Africa and EuropeBambi Ceuppens and Peter Geschiere � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 385

Through Wary Eyes: Indigenous Perspectives on ArchaeologyJoe Watkins � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 429

Metabolic Adaptation in Indigenous Siberian PopulationsWilliam R. Leonard, J. Josh Snodgrass, and Mark V. Sorensen � � � � � � � � � � � � � � � � � � � � � � � � � � 451

Indigenous Movements in AustraliaFrancesca Merlan � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 473

Indigenous Movements in Latin America, 1992–2004: Controversies,Ironies, New DirectionsJean E. Jackson and Kay B. Warren � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 549

x Contents

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.

Contents ARI 12 August 2005 20:29

Linguistic, Cultural, and Biological DiversityLuisa Maffi � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 599

Human Rights, Biomedical Science, and Infectious Diseases AmongSouth American Indigenous GroupsA. Magdalena Hurtado, Carol A. Lambourne, Paul James, Kim Hill,

Karen Cheman, and Keely Baca � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 639

Social and Cultural Policies Toward Indigenous Peoples: Perspectivesfrom Latin AmericaGuillermo de la Pena � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 717

Indexes

Subject Index � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 757

Cumulative Index of Contributing Authors, Volumes 26–34 � � � � � � � � � � � � � � � � � � � � � � � � � � � 771

Cumulative Index of Chapter Titles, Volumes 26–34 � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 774

Errata

An online log of corrections to Annual Review of Anthropology chaptersmay be found at http://anthro.annualreviews.org/errata.shtml

Contents xi

Ann

u. R

ev. A

nthr

opol

. 200

5.34

:231

-252

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f W

isco

nsin

- M

adis

on o

n 04

/05/

10. F

or p

erso

nal u

se o

nly.