1998 91: 2175-2179    

 Victor R. Gordeuk, Christine E. McLaren, Anne C. Looker, Victor Hasselblad and Gary M. Brittenham  Distribution of Transferrin Saturations in the African-American Population

http://bloodjournal.hematologylibrary.org/cgi/content/full/91/6/2175Updated information and services can be found at:

(1173 articles)Red Cells � collections: BloodArticles on similar topics may be found in the following

http://bloodjournal.hematologylibrary.org/misc/rights.dtl#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://bloodjournal.hematologylibrary.org/misc/rights.dtl#reprintsInformation about ordering reprints may be found online at:

http://bloodjournal.hematologylibrary.org/subscriptions/index.dtlInformation about subscriptions and ASH membership may be found online at:

. Hematology; all rights reservedCopyright 2011 by The American Society of 20036.the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by    

For personal use only. by on February 4, 2011. www.bloodjournal.orgFrom

Distribution of Transferrin Saturations in the African-American Population

By Victor R. Gordeuk, Christine E. McLaren, Anne C. Looker, Victor Hasselblad, and Gary M. Brittenham

To determine if transferrin saturations in African Americans

may reflect the presence of a gene that influences iron

metabolism, we analyzed the distribution of these values in

808 African Americans from the second National Health and

Nutrition Survey. We tested for a mixture of three normal

distributions consistent with population genetics for a major

locus effect in which the proportion of normal homozygotes

is p2; of heterozygotes, 2pq; of affected homozygotes, q2;

and in which p1q 5 1. Three subpopulations based on

transferrin saturation were present (P F .0001) and the fit to

a mixture of three normal distributions was good (P 5 .2). A

proportion of .009 was included in a subpopulation with a

mean 6 standard deviation transferrin saturation of 63.4% 6

5.7% (postulated homozygotes for a gene that influences

iron metabolism), while a proportion of .136 had an interme-

diate saturation of 38.0% 6 5.7% (postulated heterozygotes)

and .856 a saturation of 24.6% 6 5.7% (postulated normal

homozygotes). These proportions were consistent with popu-

lation genetics because the sum of the square roots of the

proportions with the lowest mean transferrin saturation (P 5.925) and the highest (q 5 0.093) was approximately 1

(1.018). The results are consistent with the presence in

African Americans of a common locus that influences iron

metabolism.

r 1998 by The American Society of Hematology.

I RON OVERLOAD IS COMMON in Africa,1 but thiscondition is not widely considered to be a problem among

African Americans whose ancestors originated in Africa. Recog-nized for more than 60 years,2 iron overload in Africa isetiologically related to increased dietary iron.3 A recent studysuggested that a non-HLA–linked iron-loading gene may alsobe implicated in the pathogenesis, with heterozygotes for theiron-loading locus developing iron overload only in the face ofhigh dietary iron, but with homozygotes becoming iron-loadedeven without increased dietary iron.4

In the United States, primary iron overload is regarded aspredominantly a problem among Caucasian Americans in theform of HLA-linked hemochromatosis. Based on screening thepopulation for homozygotes, the estimated gene frequency forthis recessive disorder is .067.5 Recently, we used a novelapplication of statistical mixture modeling to analyze thedistribution of transferrin saturations in white Americans and toestimate the prevalence of hemochromatosis heterozygotes inthat population. Our findings were consistent with a distinctdistribution of transferrin saturations in hemochromatosis heter-ozygotes and with a gene frequency of .07 to .08.6

The present study was prompted by a concern that a primaryiron overload condition may be present in the African-Americanpopulation, but be largely unrecognized and untreated. Wepostulated that, in an analogous situation to that of HLA-linkedhemochromatosis in whites, an iron-loading locus in AfricanAmericans might be manifested by distinct, statistically discern-ible subpopulations in the distribution of population transferrinsaturation data.

MATERIALS AND METHODS

Source of data. The second National Health and Nutrition Exami-nation Survey, which studied a representative sample of the noninstitu-tionalized United States population ages 6 months to 74 years from1976 to 1980, was the source of data for the present study. In this survey,20,322 persons were examined from 64 primary sampling units(counties or small groups of contiguous counties). Based on aninterviewer’s observation, each person was classified as white, black, orother. Of those surveyed, 2,763 (13.6%) were African Americans.Serum iron and total iron binding capacity were measured by amodification of the Automated Technicon AAII-25 method, which is acolorometric method using ferrozine; the transferrin saturation wascalculated from these values by dividing the serum iron by the total ironbinding capacity and multiplying by 100.7,8

Selection criteria. We selected transferrin saturations from African-American men and women aged 20 years to 74 years for whom themean corpuscular volume was between 80 fL and 100 fL and theerythrocyte protoporphyrin was, 70 mg/dL red blood cells. Additionalselection criteria included hemoglobin concentration$13.5 g/dL andhematocrit$40% for men and hemoglobin$11.5 g/dL and hematocrit$34% for women. We excluded subjects with abnormally low hemoglo-bin or hematocrit values because anemias of various causes areassociated with abnormally high9-11 or low12-14 transferrin saturations.We excluded subjects with abnormal values for mean corpuscularvolume because a low mean corpuscular volume can be associated withiron deficiency or inflammation and a high mean corpuscular volumecan be associated with megaloblastic conditions and drug effects,14 allof which can lead to altered transferrin saturations.12,13,15,16We ex-cluded subjects with elevated erythrocyte protoporphyrin levels becauseof the associations with iron deficiency and inflammation.9,17 Afterapplying the selection criteria, there were 836 individuals in the data set.As described later, an additional 28 subjects were excluded as possibleheterozygotes for HLA-linked hemochromatosis. Table 1 gives thenumbers of subjects excluded using specific criteria.

Adjustment of transferrin saturations for gender and diurnal varia-tion. Because transferrin saturation has a diurnal variation,15,18 theinclusion of samples obtained at different times of the day, withoutappropriate adjustment, in an analysis of distribution might alter the

From the Department of Medicine, The George Washington Univer-sity Medical Center, Washington D.C.; the Department of Mathematics,Moorhead State University, Moorhead MN; the Nutrition StatisticsBranch, National Center for Health Statistics, Hyattsville MD; theCenter for Health Policy Research and Education, Duke University,Durham, NC; and the Department of Medicine, MetroHealth MedicalCenter, Case Western Reserve University School of Medicine, ClevelandOH.

Submitted September 25, 1997; accepted October 28, 1997.Supported in part by a grant from the Offıce of Minority Health to the

Cell Biology and Metabolism Branch, National Institute of Child Healthand Human Development (Bethesda, MD) and by a grant from theNational Center for Health Statistics, Centers for Disease Control(Hyatsville, MD).

Address reprint requests to Victor R. Gordeuk, MD, Suite 3-428, 2150Pennsylvania Ave NW, Washington DC 20037.

The publication costs of this article were defrayed in part by pagecharge payment. This article must therefore be hereby marked ‘‘adver-tisement’’ in accordance with 18 U.S.C. section 1734 solely to indicatethis fact.

r 1998 by The American Society of Hematology.0006-4971/98/9106-0006$3.00/0

Blood, Vol 91, No 6 (March 15), 1998: pp 2175-2179 2175

For personal use only. by on February 4, 2011. www.bloodjournal.orgFrom

results. In addition, because of the relatively small size of the sampledata from men (368) and women (468), analyzing the individualdistributions could result in inappropriately large standard errors forparameter estimates. In this study, we adjusted transferrin saturationvalues for blood samples drawn in the afternoon or evening to reflectexpected values had the blood samples been drawn in the morning.These expected values were determined using linear regression analysisin the following manner. First, for each gender, the transferrin saturationvalues were stratified by time of blood collection (morning, afternoon,or evening) and each stratum was divided into deciles. Second, aregression equation was determined using the average value of transfer-rin saturations for each decile from blood samples drawn in the morningas the dependent variable. Predictors included an indicator variable forgender and the average value of transferrin saturations by decile fromblood samples drawn in the afternoon. Similarly, a separate regressionequation was formed for use with transferrin saturation values fromblood samples drawn in the evening. Third, the predicted averagemorning transferrin saturation value was then computed for samplesdrawn in the afternoon or evening using the appropriate regressionequation. To form the frequency distributions of transferrin saturationdescribed below, all of the values for samples drawn in the morning andthe predicted morning values for samples drawn in the afternoon orevening were used.

Adjustment of the data set to account for a possible admixture ofCaucasian HLA-linked hemochromatosis genes.The data were modi-fied to take into account the possibility that the distribution oftransferrin saturations from African Americans is affected by individu-als who are heterozygotes or homozygotes for HLA-linked hemochro-matosis. The gene frequency for the HLA-linked hemochromatosislocus in the Caucasian population is estimated to be .067.5 Assuming a25% admixture of Caucasian genes in the African-American popula-tion,19 the gene frequency for the HLA-linked hemochromatosis locusin African Americans would be 25% of .067 or .017. Populationgenetics would then project for African Americans a proportion ofhomozygotes for the HLA-linked hemochromatosis gene of 3 per10,000 (.017 squared) and a proportion of heterozygotes of 33 per1,000. We assumed that the transferrin saturation values from theseprojected African American heterozygotes for the HLA-linked hemo-chromatosis locus would be normally distributed with the same meanand SD as found in our previous study of Caucasian Americans in thesecond National Health and Nutrition Examination Survey (NHANESII).6 Under this assumption, we removed 3.3% of the 836 values (n5

28) from the data set that were closest to 28 random values generatedfrom a normal distribution with a mean of 45.5% and a SD of 7%. In thecase when several transferrin saturation values were equidistant to arandomly generated value, one of these values was selected randomly.Because the number of transferrin saturations arising from homozy-gotes for HLA-linked hemochromatosis was projected to be less thanone, no further adjustment was made to the data.

Distribution of transferrin saturation values in African Americans.We examined the distributions of adjusted transferrin saturation valuesfor the remaining 808 men and women African Americans in theunweighted data using techniques developed for the analysis ofdistributions in grouped, truncated data.20 Transferrin saturation valueswere sorted into intervals and the frequency of values within eachinterval was computed. The physiologic models we considered were asingle normal distribution and a mixture of two or three normaldistributions. We have previously established that transferrin saturationsin a homogeneous population follow a normal distribution.6 Theexpectation-maximization algorithm was applied to the distributions oftransferrin saturation values for parameter estimation.20,21 Equal vari-ances were assumed for fit to mixtures of normal distributions becausemodels with unequal variances resulted in increased variances for thesubpopulation with the highest transferrin saturation that were biologi-cally implausible. The statistical test used to determine the best fittingmodel was based on the likelihood ratio statistic. For each observeddistribution, the maximized log-likelihood function for a mixture ofthree normal distributions was evaluated (Log L1) and compared withthe maximized log-likelihood function (Log L0) for either a singlenormal distribution or a mixture of two normal distributions. Signifi-cance of the likelihood ratio statistic,-2[Log(L0/L1)], was assessed byreferring to thex2 distribution with degrees of freedom based on thedifference between the number of parameters being estimated undereach model. A significance level of .05 was used. Similarly, themaximized log-likelihood function for a mixture of three normaldistributions was compared with that of a single normal distribution andto that of two normal distributions. Thex2 statistic was then used to testgoodness of fit of each observed distribution to the best fitting model.For the three-population model, the methods of Crump and Howe22

were used to compute confidence intervals for the proportion with thehighest mean.

Weighting the results to reflect the African-American population as awhole. The assumptions underlying our analysis were that transferrinsaturation values are independent and identically distributed, ie, eachobservation has an equal chance of being selected, and that allobservations come from the same distribution. However, becauseindividuals in the NHANES II sample did not have an equal probabilityof selection, sample weights must be used to calculate parameterestimates that reflect the United States population. For NHANES II, amultistage estimation procedure was used to calculate sample weightsso that point estimates would reflect the United States population.8,23

The methods described above were used to compute parameter esti-mates from the weighted transferrin saturation distribution to reflectresults for African-American men and women in the United Statespopulation fitting our exclusion criteria. It was not possible to adjust thevariance estimates to account for the complex design of NHANES II.

Estimation of gene frequency for a possible locus that influences ironstatus in African Americans.Accepting the possibility that ourfindings might represent the presence of a locus that influences ironmetabolism among African Americans, we assumed that the threenormal distributions of transferrin saturation in our analyses wouldrepresent predominantly a subpopulation of normal homozygotes, asubpopulation of heterozygotes, and a subpopulation of affectedhomozygotes. We estimated the proportions of normal homozygotes,heterozygotes, and affected homozygotes as the proportions in thepopulations with the lowest, intermediate, and highest mean transferrinsaturations, respectively. According to the Hardy-Weinberg equilibriumequation,p2 (the proportion of normal homozygotes)1 2pq (theproportion of heterozygotes)1 q2 (the proportion of abnormal homozy-gotes)5 1. We estimated the gene frequency of the abnormal allele (q)as the square root of the proportion from the population with the highestmean transferrin saturation. We estimated the gene frequency of thenormal homozygotes (p) as the square root of the proportion from thepopulation with the lowest mean transferrin saturation. We then

Table 1. Exclusions From the Total Sample of African Americans

Aged 20 to 74 Years

Exclusion No.

Total sample 1,275

Excluded for

Abnormal mean corpuscular volume 188

Abnormal erythrocyte protoporphyin concentration 94

Abnormal hemoglobin concentration 141

Abnormal hematocrit 16

Possible heterozygote for HLA-linked hemochromatosis 28

Final analytic sample 808

The exclusions were applied sequentially, in the order shown.

2176 GORDEUK ET AL

For personal use only. by on February 4, 2011. www.bloodjournal.orgFrom

examined whether the modeled distributions were consistent withpopulation genetics for a major locus effect in whichp 1 q 5 1.

RESULTS

Analysis of unweighted data.The primary analysis wasperformed on transferrin saturations for 808 individuals. Thetransferrin saturations had been adjusted for sex and diurnalvariation and the data set had been adjusted for the possiblepresence of an HLA-linked hemochromatosis allele as de-scribed above. The unweighted data showed a significantlybetter fit to three normal populations than to two normalpopulations (likelihood ratio statistic, 32.1 with 3 degrees offreedom;P , .0001) or to a single normal population (likeli-hood ratio statistic 123.1 with 7 degrees of freedom,P ,.0001). The fit of the data to a mixture of three normalpopulations was good (P 5 .20; in a goodness-of-fit analysis, aP value above .1 indicates an acceptable fit). Figure 1A showsthe observed and fitted distributions for unweighted transferrinsaturation values. An estimated proportion of .856 of the

African Americans studied were included in a subpopulationwith a mean saturation of 24.6%, while .135 comprised asubpopulation with an intermediate mean saturation of 38.0%and .009 (95% confidence interval of .004 to .017) formed asubpopulation with a mean saturation of 63.4% (Table 2). Theseproportions are consistent with population genetics for a singlemajor locus affecting the distribution of transferrin saturations:the sum of the square roots of the proportion with the lowesttransferrin saturation (P5 .925) and of the proportion with thehighest saturation (q 5 .093) is approximately 1 (1.018).

Weighted results for the African American population.Theresults after the data (adjusted for sex, diurnal variation, and thepotential presence of HLA-linked hemochromatosis) wereweighted to reflect the United States population of AfricanAmericans as a whole are given in Table 2. The weighted resultsare similar to our primary results using unweighted data, whichsuggests that the unequal probability of selection in NHANESII did not have a major input on the transferrin saturationdistribution. An estimated proportion of .815 was included in asubpopulation with a mean saturation of 24.3%, while .172comprised a subpopulation with an intermediate mean satura-tion of 39.1% and .012 formed a subpopulation with a meansaturation of 61.0%. The weighted findings are also consistentwith population genetics for a single major locus affecting thedistribution of transferrin saturations: the sum of the squareroots of the proportion with the lowest transferrin saturation(P 5 .903) and of the proportion with the highest saturation(q 5 .112) is approximately 1 (1.015).

DISCUSSION

Our analysis of transferrin saturations from African Ameri-cans studied in the second National Health and NutritionExamination Survey showed that three subpopulations ofindividuals could be detected. Using data that were adjusted forthe time of day of collection of the blood sample, for gender,and for the possible presence of an HLA-linked hemochromato-sis allele, we found that one subpopulation comprised an

Fig 1. Distribution of transferrin saturation values in African

Americans. The interval width is 3%. (A) Original sample of 808

values. (B) Weighted data for the African American population. The

dashed lines represent the fitted normal distributions representing

three subpopulations. The overall fitted mixture distribution is shown

with a solid line.

Table 2. Analysis of the Distribution of Transferrin Saturations

for African Americans

Postulated

Subpopulation

Including

Normal

Homozygotes

for a

Locus That

Influences

Iron

Metabolism

Postulated

Subpopulation

Including

Heterozygotes

for a

Locus That

Influences

Iron

Metabolism

Postulated

Subpopulation

Including

Abnormal

Homozygotes

for a

Locus That

Influences

Iron

Metabolism

Unweighted sample of 808 values

Proportion .856 .136 .009

Mean transferrin satu-

ration (%) 24.6 38.0 63.4

SD 5.7 5.7 5.7

Sample weighted to reflect the African-American population

as a whole

Proportion .815 .172 .012

Mean transferrin satu-

ration (%) 24.3 39.1 61.0

SD* 5.8 5.8 5.8

*Assumes a simple random sample.

TRANSFERRIN SATURATIONS IN AFRICAN AMERICANS 2177

For personal use only. by on February 4, 2011. www.bloodjournal.orgFrom

estimated proportion of .856 of the African Americans studiedand had a mean transferrin saturation of 24.6%. A secondsubpopulation was made up of an estimated proportion of .136and had a mean transferrin saturation of 38.0%. A thirdsubpopulation comprised a proportion of .009 and had a meantransferrin saturation of 63.4%. Limitations to this analysis arethat transferrin saturations were single rather than repeateddeterminations, that subjects were not screened for liver dis-eases that may be associated with elevations of transferrinsaturation, and that the sample size was small for a populationstudy. In addition, we were not able to fully account for thecomplex survey design when estimating variances.

The three subpopulations that we identified based on thismixture-modeling statistical analysis of transferrin saturationdata from African Americans are consistent with the presence ofa major locus that influences iron status. If we assume that themembers of the small subpopulation with the highest meantransferrin saturation may be abnormal homozygotes for a locusthat influences iron metabolism, then the abnormal genefrequency in this data set would be estimated to be .093.Similarly, if we assume that the members of the large subpopu-lation with the lowest mean transferrin saturation may benormal homozygotes for a locus that influences iron metabo-lism, then the normal gene frequency would be .925. Consistentwith population genetics for a single major locus, the sum of theestimated normal and abnormal gene frequencies is approxi-mately 1 (1.018).

If our present findings do reflect the presence in the African-American population of an abnormality of a major locus thatinfluences iron metabolism, then this locus may be either (1) theHLA-linked hemochromatosis gene that has heretofore beendescribed exclusively in people of European ancestry15,24or (2)some different gene that influences iron metabolism. The factthat HLA-linked hemochromatosis has only been reported inCaucasians and that a genetic iron-loading disorder that is notlinked to the HLA-locus may be common in Africa4 aresupportive of the possibility that we are observing the effects ofa genetic alteration that is unique to people of African ancestry.There is an admixture of Caucasian genes in the African-American population, estimated to be about 25%,19 but such anadmixture would not be sufficient to explain the findings of thepresent study. The gene frequency of HLA-linked hemochroma-tosis in the United States white population is estimated to be.0675 and an estimated 25% admixture of Caucasian genescould lead to an estimated gene frequency of only .017 forHLA-linked hemochromatosis in the African-American popula-tion. Furthermore, we made an adjustment to account for theprobable presence of HLA-linked hemochromatosis heterozy-gotes in the present study.

Until recently, only rare mention has been made in theliterature of African Americans with ‘‘hemochromatosis’’.25-29

Two recent reports underscore the fact that primary ironoverload does occur among African Americans.30,31 Further-more, one study raises that possibility that the condition maynot be rare.31 The present statistical study of the distribution oftransferrin saturation values is compatible with the possibilitythat an alteration in a gene that influences iron metabolism ispresent among African Americans. Clinicians should considerthe diagnosis of primary iron overload in African-American

patients and both treat the condition and screen family membersof affected subjects. Investigations to define the prevalence,clinical consequences, and causes of primary iron overload inAfrican Americans are needed.

REFERENCES

1. Gordeuk VR: Hereditary and nutritional iron overload. Balliere’sClin Haematol 5:169, 1992

2. Strachan AS: Haemosiderosis and Haemochromatosis in SouthAfrican Natives with a Comment on the Etiology of Haemochromato-sis. MD Thesis, University of Glasgow, Scotland, 1929

3. Bothwell TH, Seftel H, Jacobs P, Torrance JD, Baumslag N: Ironoverload in Bantu subjects. Studies on the availability of iron in Bantubeer. Am J Clin Nutr 14:47, 1964

4. Gordeuk V, Mukiibi J, Hasstedt SJ, Samowitz W, Edwards CQ,West G, Ndambire S, Emmanual J, Nkanza N, Chapanduka Z, RandallM, Boone P, Romano P, Martell RW, Yamashita T, Effler P, BrittenhamG: Iron overload in Africa. Interaction between a gene and dietary ironcontent. N Engl J Med 326:95, 1992

5. Edwards CQ, Griffen LM, Goldager D, Drummond C, SkolnickMH, Kushner JP: Prevalence of hemochrmatosis among 11,065 presum-ably healthy blood donors. N Engl J Med 318:1355, 1988

6. McLaren CE, Gordeuk VR, Looker AC, Hasselbad V, EdwardsCQ, Griffen LM, Kushner JP, Brittenham GM: Prevalence of hemochro-matosis heterozygotes in United States whites. Blood 96:2021, 1995

7. Gunter EW, Turner WE, Nease JW, Bayse DD: Laboratoryprocedures used by the Clinical Chemistry Division, Centers forDisease Control for the Second Health and Nutrition ExaminationSurvey (NHANES II) 1976-1980. U.S. Department of Health andHuman Services, Public Health Service, Centers for Disease Control,Center for Environmental Health, Nutritional Biochemistry Branch,Atlanta, GA, 1981

8. Fulwood R, Johnson CL, Bryner JD, Gunter EW, McGrath CR:Hematological and nutritional biochemistry reference data for persons 6months - 74 years of age: United States, 1976-1980. DHHS PublicationNo. (PHS) 83-1682. U.S. Department of Health and Human Services,Public Health Service, National Center for Health Statistics, Hyatts-ville, MD, December, 1982

9. Cartwright GE, Hughley CM, Ashenbrucker H, Faye J, WintrobeMM: Studies on free erythrocyte proto-porphyrin, plasma iron andplama copper in normal and anemic subjects. Blood 3:501, 1948

10. Finch CA, Deubelbeiss K, Cook JD, Eschbach JW, Harker LA,Funk DD, Marsaglia G, Hillman RS, Slichter S, Adamson JW, GanzoniA, Biblett ER: Ferrokinetics in man. Medicine 49:17, 1970

11. Hershko C, Graham G, Bates GW, Rachmilewitz EA: Non-specific serum iron in thalassaemia: An abnormal serum fraction ofpotential toxicity. Br J Haematol 40:255, 1978

12. Cartwright GE, Wintrobe MM: Chemical, clinical, and immuno-logical studies on the products of human plasma fractionation. XXXIX.The anemia of infection. Studies on the iron-binding capacity of serum.J Clin Invest 28:86, 1949

13. Bainton DF, Finch CA: The diagnosis of iron deficiency anemia.Am J Med 37:62, 1964

14. Jandl JH: Blood. Boston, MA, Little, Brown, 198715. Edwards CQ, Griffen LM, Kushner JP: Twenty-four hour

variation of transferrin saturation in treated and untreated hemochroma-tosis homozygotes. J Intern Med 226:173, 1989

16. Gordeuk VR, Brittenham GM, McLaren GD, Spagnuolo PJ:Hyperferremia in immunosuppressed patients with acute nonlympho-cytic leukemia and the risk of infection. J Lab Clin Med 108:466, 1986

17. Devasthali SD, Gordeuk VR, Brittenham GM, Bravo JR, HughesMA, Keating LJ: Bioavailability of carbonyl iron: A randomized,double-blind study. Eur J Haematol 46:272, 1991

18. Hamilton LD, Gubler CJ, Cartwright GE, Wintrobe MM:

2178 GORDEUK ET AL

For personal use only. by on February 4, 2011. www.bloodjournal.orgFrom

Diurnal variation in the plasma iron level of man. Proc Soc Exp BiolMed 75:65, 1950

19. Chakraboty R, Kamboh MI, Nwankwo M, Ferrill RE: Caucasiangenes in American blacks: New data. Am J Hum Genet 50:145, 1992

20. Hassselblad V: Manual for the Grouped DISFIT Software.Version 1.1. Durham, NC, Duke University, 1992

21. Dempster AP, Laird NM, Rubin DB: Maximum likelihood fromincomplete data via the EM algorithm. J Roy Stat Soc Series B 39:1,1977

22. Crump, KS, Howe R: A preview of methods for calculatingconfidence limits in low dose extrapolation, in Krewski D (ed):Toxicological Risk Assessment. Boca Raton, FL, CRC, 1985, p 188

23. McDowell A, Engle A, Massey F, Maurer K: Plan and Operationof the Second National Health and Nutrition Examination Survey,1976-1980. Vital and Health Statistics. Series 1-No. 15. DHHSPublication No. (PHS) 81-1317. US Department of Health and HumanServices. Public Health Service. Hyattsville, MD, National Center forHealth Statistics, July 1981

24. Simon M, Le Mignon L, Fauchet R, et al: A study of 609 HLAhaplotypes marking for the hemochromatosis gene: (1) Mapping of thegene near the HLA-A locus and characters required to define a

heterozygous population and (2) hypothesis concerning the underlyingcause of hemochromatosis-HLA association. Am J Hum Genet 41:89,1987

25. Krainin P, Kahn BS: Hemochromatosis: Report of a case in aNegro; discussion of iron metabolism. Ann Intern Med 33:452, 1950

26. MacDonald RA, Mallory GK: Hemochromatosis and hemosid-erosis. Arch Intern Med 105:686, 1960

27. Rosner IA, Askari AD, McLaren GD, Muir A: Arthropathy,hypouricemia and normal serum iron studies in hereditary hemochroma-tosis. Am J Med 70:870, 1981

28. Muir WA, McLaren GD, Braun W, Askari A: Evidence forheterogeneity in hereditary hemochromatosis. Evaluation of 174 indi-viduals in nine families. Am J Med 76:806, 1984

29. Conrad ME: Sickle cell disease and hemochromatosis. Am JHematol 38:150, 1991

30. Barton JC, Edwards CQ, Bertoli LF, Shroyer TW, Hudson SL:Iron overload in African Americans. Am J Med 99:616, 1995

31. Wurapa RK, Gordeuk VR, Brittenham GM, Khiyami A, Schech-ter GP, Edwards CQ: Primary iron overload in African Americans. Am JMed 101:9, 1996

TRANSFERRIN SATURATIONS IN AFRICAN AMERICANS 2179

For personal use only. by on February 4, 2011. www.bloodjournal.orgFrom