mean estradiol

8/8/2019 Mean Estradiol

http://slidepdf.com/reader/full/mean-estradiol 1/9

AMERICAN JOURNAL OF EPIDEMIOLOGY

Copyright © 1987 by The Jo hns HopkiM University School of Hygiene and Public H ealthAll rights reserved

VoL 125, No. 5Printed in U SA .

SEX HORMONE LEVELS IN SERUM IN RELATION TO THE

DEVELOPMENT OF BREAST CANCER

DIANE K. WYSOWSKI,1-2 GEORGE W. COMSTOCK,1 KNUD J. HELSING,1AND

H. LORRIN LAU13

Wysowski, D. K., G. W. Comstock (Training Center for Public Health Research,Hagerstown, MD 21742-2067), K. J. Helsing, and H. L. Lau. Sex hormone levels

in serum in relation to the development of breast cancer. Am J Epidemiol

1987;125:791-9.

In 1974, approximately 13,000 fema le residents of Washington County, Mary-

land, donated 15 ml of blood as part of a project to determine if certain serologlc

factors were related to the development of site-specific cancer. Sera w ere storedat - 7 3 C. The present study reports the associations of serum levels of estrone,estradlol, estriol, androstenedione, progesterone, and testosterone with breast

cancer among 17 premenopausal cases diagnosed 8-132 months after Wood

was drawn, and 39 postmenopausal cases diagnosed 6-72 months after bloodwas drawn. Each case was matched to four controls selected from the other

serum bank donors. Matching factors were age, race, and time since last men-

strual period. Case s and controls who w ere taking estrogen-containing prepara-tions were excluded from this analysis. Sera were analyzed without knowledge

of case-control status. Differences in levels of serum hormones between cases

and controls were slight and not statistically significant

androstenedione; breast neoplasms; estradiol; estriol; estrogens; estrone; pro-

gesterone; testosterone

Evidence exists that suggests a role forendogenous sex hormones in female breastcancer etiology: Female sex is the leadingrisk factor for breast cancer (1-3); the ovar-ian hormones, estrogen and progesterone,

and the pituitary hormones, prolactin andsom atotropin, are primarily responsible formammary growth and development (4);mammary carcinoma can be induced in

male and female rodents by administrationof estrogens (5); transsexual males givenhigh doses of estrogens to induce breastdevelopment have developed breast cancer(5); several reproductive factors such as

number of children (6-8) and ages at men-arche (9, 10), at first full-term pregnancy(11, 12), and at natural or artifical meno-pause (13, 14) are associated with breast

Received for publication April 14,1986, and in finalform August 14, 1986.

1 Department of Epidemiology, School of Hygieneand Public Health, The Johns Hopkins University,Baltimore, MD .

2

Present address: Division of Epidemiology andSurveillance, Food and Drug Administration, Rock-ville, MD.

3Present address: Queen's Physicians Building,

Honolulu, H I.Reprint requests to Dr. George W. Comstock,

Training Center for Public Health Research, P.O. Box

2067, Hagerstown, MD 21742-2067.This research was supported by research grants CA

24758 and CA 36390 from the National Cancer Insti-tute and Research Career Award HL 21670 from theNational Heart, Lung, and Blood Institu te.

Dr. Abraham M. Lilienfeld played a major role inthe initiation of this study. Ann Brody maintained theWashington County Cancer Registry. Kathleen E.Trumpower helped with the manuscript. Th e authorsare grateful to them and to all the others who helpedwith the establishment of the serum bank.

791

b y g u e s t on S e p t em b er 2 5 ,2 0 1 0

a j e. ox f or d j o ur n al s . or g

D ownl o a d e d f r om

http://aje.oxfordjournals.org/



























792 WYSOWSKI ET AL.

cancer risk; and anti-estrogen treatmenthas been shown to be efficacious in reduc-ing certain malignant breast tumors (15).

Because of this evidence, investigators

for the past decade and a half have studiedabsolute and relative levels of various sexhormones (primarily estrogen, estrone, es-tradiol, estriol, progesterone, androgen andmetabolites, and prolactin) in the bodyfluids of breast cancer cases and controls(16-18) a nd of women a t high and low risksof breast cancer (19-34). Except for onestudy that reported urinary hormone levels(35), none are known to have compared

hormone levels in women who were subse-quently diagnosed with breast cancer andthose who were not. This report presentsresults obtained from such a prospectivestudy.

MATERIALS AND METHODS

In the autumn of 1974, 15 ml of bloodwere donated by each of 25,620 persons,representing approximately one-third of

the adult population of WashingtonCounty, Maryland. Another 182 specimenswere donated in July 1975. Slightly morethan half of the volunteers were women.Blood was collected in 15 ml Vacutainers(Becton-Dickenson and Co., Orangeburg,NY) and allowed to clot at room tempera-ture for approximately three hours. It wasthen kept at 5 C until the serum was re-moved, usually within several hours. Theserum , in two aliquots, was prom ptly frozenand stored with only trivial deviations from—73 C unt il it was thawed for the hormoneanalyses.

During the seven-year period from De-cember 1,1974 through November 30,1981,73 female participants were reported to theWashington County cancer registry as hav-ing bre ast cancer which h ad been diagnosedafter their blood had been drawn. Of thesecases, three said that they were taking oralcontraceptives or other female hormones,and hence were excluded from the presentanalysis. In addition, one case diagnosedless th an six mo nth s after blood was drawnwas also excluded. Of the remaining 69

cases, nine were premenopausal and 60were postm enopausal. Because of the smallnum ber of prem enopausal cases, eight moresuch cases who were reported to th e cancer

registry during th e nex t four years and w homet the study criteria were added. Becauselaboratory A, which performed the initialhormone assays, had been closed, the lasteight sets of sera from premenopausalwomen were examined by laboratory B.

Controls were selected from the otherserum bank donors and were matched tothei r respec tive cases by race, age, and tim esince last menstrual period. Age-matching

was done from a birth d ate listing of donorsto the serum bank, selecting the two olderand two younger women nearest in age tothe case who also met the other matchingcriteria. Matching on time since last men-strual period was done within the followingspecified intervals: 0-39 days, matched tothe same day; 40-365 days, matched within30 days; and one year, 2-4 y ears, 5-9 y ears,10-14 years, and 15 or more years, matche dto the same group. Before a woman wasfinally accepted as a control, the recordswere searched to be sure that 1) she wasnot listed in the cancer registry; and 2) if aparticipant in the 1975 county health cen-sus, she stated that she had not had adiagnosis of cancer. If a selected controlwas found to have said at the time of do-nating blood that she was taking oral con-traceptives or female sex hormones, a sub-stitute control was selected in the same

manner. Controls taking these hormonalpreparations have been excluded from theanalyses except when their inclusion isspecified.

Each set, consisting of serum from onecase and her matching controls, includingthose taking estrogens, was examined onthe same day with the same reagents. Eachserum specimen was identified only by acode number that was unrelated to case-

control status so that laboratory personnelknew only that a set contained sera from acase and controls. They could identify nei-ther the case nor which controls were tak-ing estrogens. Sera were examined for es-

b y g u e s t on S e p t em b er 2 5 ,

2 0 1 0

a j e. ox f or d j o ur n a

l s . or g





























SERUM HORMONES AND BREAST CANCER 793

trone, estradiol, estriol, androstenedione,progesterone, and testosterone. Both labo-ratories used radioimmunoassay methods;laboratory A followed p rocedures described

by Eminent et al. (36) with antisera ob-tained from Miles-Yeda, Rehovot, Israel,while laboratory B used their own antiseraand followed procedures typified by Abra-ham et al. (37). Duplicate assays were doneon each specimen, and the results reportedhere are the mean values of these two de-terminations. Quality control proceduresincluded the assay of known conc entrationsof hormones on each working day and of

occasional "b lind" replicates of serum spec-imens. Sera from one premenopausal andone postm enop ausal co ntrol were lost, leav-ing one case in each of these groups withonly three matched controls. In addition,among the eight premenopausal caseswhose sera were examined by laboratory B,there were insufficient sera for assays ofandrostenedione for three controls, of pro-gesterone for two controls, and of testoster-

one for one control. No set had less thanthree controls. In these sets, the averagevalue of the three examined controls wasused in place of the missing control value.

After sera from all but 21 postmeno-pausal cases and their controls had beenassayed, the results were examined. Be-cause addition of data from these 21 casesseemed unlikely to alter th e findings app re-ciably and because the hormone determi-

nations used up all the stored serum foreach subject, a decision was made to dis-continue the serum assays, thereby saving

sera from 105 women for possible futureuse.

Perce ntage differences in horm one levelsbetween cases and controls were calculated

by subtracting the mean level for the con-trols from the level for the cases, dividingthe result by the control serum level, andmultiplying by 100. This was done sepa-rately by laboratory and by menopausalstatus. Tests of statistical significance ofthe m ean differences between case an d con-trol levels were performed by paired t tests.

R E S U L T S

Characteristics of the cases with respectto age, time since last menstrual period,and months from initial blood donation todiagnosis are summarized in table 1. Onecase and her four controls were black. Allother study participants were white.Matching of controls to cases by exact yearof age was achieved in 82 per cent of theattempts. Among the remainder, the meandeviation (ignoring sign) was only 2.0 years.

Premenopausal cases and controls werematched to the exact day since last men-strual period. All postmenopausal subjectswere at least two years past their last men-strual period; these cases and controls wereall matched within the following time pe-riods: 2-4, 5 -9,1 0-1 4, and 15 or more yearssince last menstrual period.

Mean serum levels of estrone, estradiol,estriol, androstenedione, progesterone, and

testosterone amon g cases and matched con-trols are shown in table 2. Among postmen-opausal women, serum levels of all six hor-

TABLE l

Characteristics of breast cancer cases, by menopausal status and laboratory, Washington County, MD,1975-1985

Menstrualstatus and

laboratory

PremenopausalLaboratory ALaboratory B

Postmenopausal*

No. ofsu ject*

98

39

Age

Mean

414361

(years)

Range

(25-49)(32-50)(36-90)

Time since lastmenstrual period

Mean

16 days8 days

14.4 years

Range

(2-28)(0-18)(2-15+)

Mon ths priorto diagnosis

Mean Range

34 (16-72)50 (8-132)28 (6-72)

* All assays were done in laboratory A.



l s . or g





























794 WYSOWSKI ET AL.

T A B L E 2

Mean serum levels of selected hormones among breast cancer cases diagnosed after having blood drawn in 1974

and their matched controls, by menopausal status and laboratory, Washington County, MD, 1975-1985

* * U l LU U 1 1 U

Laboratory A (9 cases, 36 controls)

Estrone

Estradiol

EstriolAndrostenedione

Progesterone

Testosterone

Laboratory B (8 cases, 31 controls)

Estrone

EstradiolEstriolAndrostenedione*

Progesteronet

Testosterone!

Laboratory A (39 cases, 156 controls)

Estrone

Estradiol

Estriol

Androstenedione

Progesterone

Testosterone

Cases

Premenopausal

77

100

37 6

2,191

1,810

30 9

66

11018

890

1,081

30 6

Postmenopausal

59

51

36 9

1,795

64 2

30 4

Mean level (pg/ml)

Control*

111

13 5

36 9

2,389

2,278

315

68

11 618

986

1,707

299

62

52

361

1,677

5 58

274

% difference

- 3 1

- 2 6

2

- 8

- 2 1

- 2

- 3

- 50

- 1 0

- 3 7

2

- 5

- 2

- 1

7

- 3

11

* Eight cases and 28 controls.

t Eight cases and 29 controls.

| Eight cases and 30 controls.

mones were virtually the same for casesand controls. Among premenopausalwomen, differences between cases and con-tro ls were similar in directio n for th e groups

examined by the two laboratories.Weighted average per cent differences be-

tween cases and controls for the 17 pre-

menopausal sets are as follows: estrone,—18; estradiol, —16; estriol, +1; androstene-dione, —9; prog estero ne, —29; and testos-teron e, 0. N one of the observed case-controldifferences achieved statistical significanceat the 0.05 level.

Dividing cases into those diagnosed early

and those diagnosed late in the 11-yearobservation period gave no consistent or

significant indications that case-controldifferences might be related to time of di-agnosis of the case and hence to the influ-

ence of unrecognized c ancers at the time of

donating blood (table 3).

In addition to the 39 case-control sets of

sera from postmenopausal women that

were analyzed for the six hormones, therewere one case and 23 controls whose serawere also analyzed but who were excludedfrom the previous tables because they w eretaking estrogens when they gave blood.These 23 controls came from 17 matchedsets of controls. Including the set in whichonly the case was taking estrogens wouldhave made trivial changes in the results in

tables 2 and 3 since case-control differences

in this set were similar to those in the othersets. However, if the 23 controls takingestrogens had been included in the case-control com parisons, there would have beensome changes, as can be surmised from the


2 0 1 0































T A B L E 3

Differences in mean serum levels of selected hormones among cases diagnosed earfy and late after having blood

drawn in 1974 and their matched controls, by menopausal status and laboratory, Washington County, MD ,

1975-1985

Laboratoryand hormone

Laboratory A

No. of casesEstroneEstradiol

Estriol

AndrostenedioneProgesteroneTestosterone

Laboratory B


Estriol

AndrostenedioneProgesterone

Testosterone

Laboratory A


Estriol

AndrostenedioneProgesterone

Testosterone

6-23

3

- 2 0

- 2 8

- 1 5

- 2 5 6291

23

2

12

98

- 3

20

- 7 1

65

12

- 2

- 2

- 4 0

134

- 6 0

- 1 1

Months

Inpg/ml

24-72

from giving blood to diagnosis

24-132

Premenopausal

6

- 4 1

- 3 8

18

- 1 6 9- 8 4 8

- 2 0

6

- 7

- 4 0

0

- 1 9

- 8 1 3

- 1 4

Postmenopausal

27

- 3

- 1

15

11 1

4

48

6-23

- 1 8

- 2 1

- 4

- 1 1

13

7

17

89

- 1 4

24

- 4 1

21

- 3

- 4

- 1 1

8

- 1 1

- 4

As % of control value

24-72

- 3 7

- 2 8

5

- 7

- 3 7

- 6

- 5

- 2

4

7

1

18

24-132

- 1 0

- 3 4

0

- 1 9

- 3 7

- 5

findings in table 4. The two groups of con-

trols in the table are matched within eachset for the same characteristics used inselecting matched controls for cases,namely sex, race, age, and time since lastmenstrual period. For each hormone, themean values within each set for controlstaking estrogens and those not taking es-trogens were averaged, thus allowing forthe unequal numbers of controls on estro-gens in the 17 sets. Controls not taking

estrogens had considerably lower levels ofestrone, somew hat lower levels of estradiol,similar levels of estriol and androstenedi-one, and slightly higher levels of progester-one and testosterone than did those onestrogens. Had the controls taking estro-

gens been included in th e com parisons, the

case-control differences for estron e and es-tradiol would have become more stronglynegative, those for estriol, androstenedione,and progesterone would have remainedclose to zero, and the difference for testos-terone would have become somewhat largerin the positive direction.

D I S C U S S I O N

Th e study of normal wom en by Bulbrooket al. (35) most nearly resembles the pres-ent investigation. In 1971, they publishedthe results of a 10-year follow-up study of5,000 British w omen aged 15-55 years whoresided on the Isle of Guernsey. During the



l s . or g





























796 WYSOWSKI ET AL.

TABLE 4

Serum hormone levels among 67 postmenopausal controls who were not taking estrogens an d 23 postmenopausalcontrols who were taking estrogens, Washington County, MD , 1975-1985

No. of subjectsWeighted mean hormone levels (pg/ml)*

EstroneEstradiolEstriolAndrostenedione

ProgesteroneTestosterone

Postmf

No t taking

estrogens

67

6553

3391,773

285552

mopausol controls

Taking

estrogens

23

25591

3401,694

249474

Ratio (group

not tjilfingestrogens/group

taking

estrogens)

0.250.581.001.051.141.16

* See text for method of calculation.

follow-up period, 27 women developedbreast cancer. Each case was matched withbetwe en four and 13 contro ls on age, height,weight, parity, menopausal status, and dayof menstrual cycle. Their 24-hour urinespecimens, which had been collected at thebeginning of follow-up, were assayed for

various androgen metabolites. The inves-tigators found that the mean excretionlevels of androsterone and etiocholanolonein women subsequently diagnosed withbre ast c ancer w ere significantly lower thanthose in controls, indicating a general de-ficiency of androgen metabolite excretion.

Using historical information and urineassay results from the 1,485 women in thatstudy, a model was developed that enabledfive levels of risk of developing breast can-cer to be delineated (38). Between 1966 and1976, blood was collected from 386 of the1,485 women in the original study (30).Plasma estradiol levels showed no consist-ent relation am ong either pre- or postmen-opausal women with the risk of breast can-cer calculated from t ha t m odel. Plasm a pro-gesterone levels among premenopausalwomen during the luteal phase of the men-strual cycle showed a significant negative

association with breast can cer risk. No sim-ilar association was observed amongwomen whose blood was drawn during thefollicular phase nor among postmenopausalwomen.

The results among postmenopausalwomen in our study agree with those ofBulbrook et al. (30) in failing to find anassociation of risk with either serum estra-diol or progesterone. However, findingsamong our small series of premenopausalcases were not similar. Among premeno-

pausal women whose times from last men-strual period were 1-14 days (an approxi-mation of the follicular phase of Bulbrooket al.), cases in eight of the nine sets hadlevels of estradiol that were lower than theaverages of their ma tching con trols, and inseven of nine sets, case levels of progester-one were lower than the averages of theirmatching controls. Among premenopausalwomen with times from last menstrual pe-riod of 15 -28 days (an approxim ation of theluteal phase), there was no evidence of dif-ferences between cases and controls inlevels of either of these two hormones.

Lemon (39) hypothesized in 1969 thatthe relative levels of individual estrogenfractions (estrone, estradiol, and estriol)might be important breast cancer predic-tors. Specifically, he suggested that a lowurinary estriol excretion ratio might facili-tate cancer development. His hypothesiswas modified by Cole and MacMahon (40),who suggested th at a low estriol ratio in thefirst decade or so after puberty was animportant determinant of a woman's life-time risk of breast cancer. Since then,


2 0 1 0































doubts have been raised about some of theassumptions on which the estriol ratio hy-pothesis was based (2, 3, 41). Questionshave arisen about the use of urine as op-

posed to blood as an indicator of biologi-cally active estrogen (42) and about thepossibility that other hormones such asprolactin (27), progesterone (43), and an-drostenedione (33, 34, 44, 45) may play arole in breast cancer etiology. Investigatorshave focused more on absolute rather thanrelative levels of estrogen fractions in cross-sectional analyses of women at high andlow risks of breast cancer (24, 25, 27, 30-

32).Results of the present study do not sup-

port any of these hypotheses. Whether es-triol ratios are based on denominators us-ing the sum of estrone and estradiol or thesum of all three estrogens, they are slightlyhigher among cases than among controls.While the lower level of progesteroneamong prem enopau sal cases is in the direc-tion suggested by the findings of Cowan et

al. (43) that progesterone-deficient womenhad an increased risk of breast cancer, thedifference between levels for cases and con-trols could easily have occurred by chance.Un fortunately, we were able to look at onlysix horm ones. Im po rtant omissions in viewof current thinking were prolactin andlevels of bound and unbound estrogens.

The prospective nature of the presentstudy, its general population base, the use

of serum rather than urine specimens, andsimilar, blinded hormone assays for eachcase-control set are important strengths.The major weakness of the study is lack ofpower. The decision not to analyze the last21 postmenopausal case-control sets wasnot an easy one. While their inclusionwould have resulted in some added power,this would not have been great. To changethe present findings among postmeno-pausal cases to an interesting degree wouldhave required the results of the case-controlcomp arisons in the 21 sets whose sera wereno t assayed to be markedly different fromthose in the 39 sets whose hormonal anal-yses are reported here, an outcome that

seems highly unlikely. Another problem isthe markedly different levels of estriol re-ported by the two laboratories and the lessmarked differences in androstenedione

levels. While changes associated with in-creased storage time for the specimens ex-amined by laboratory B could be the expla-nation, this does not seem likely. For fiveof the six hormones, the levels reportedhere are close to or slightly above thoseeach laboratory considered normal forwomen. The exception is estriol which av-eraged 19 pg/m l in laboratory B, where thenormal level for nonpregnant females is

stated to be less than 10 pg/ml. However,there have been no indications of evapora-tive losses from th e stored sera. D iscussionswith laboratory personnel have failed toshed any light on reasons for the differentlevels. Regardless of the interlaboratorydifferences, the important feature is thatthe examination procedures were similarfor each set of a case and four controls andindependent of case-control status, so that

comparisons of cases and controls are validregardless of laboratory differences.

With respect to prediagnostic serumlevels of hormones, this study leaves thesituation regarding premenopausal casesless clear th an one would like. Th e nu mbersof cases are not large. Matching of casesand controls by time since last menstrualperiod is only an approximation to match-ing on phase of the men strual cycle so tha t

control values m ay provide a less th an idealapproximation to expected values amongcases. In spite of these problems and theunexplainable laboratory differences, itseems unlikely that major case-control dif-ferences could be missed. The situation forpostmenopausal breast cancer is clearer.The differences between postmenopausalcases and controls are so slight that evenwith the modest number of cases in this

study it is not unreasonable to suggest thatprediagnostic serum levels of the six hor-mones do not play a meaningful role in thepathogenesis of breast cancer in this agegroup. Fu rtherm ore, th ere is no support forearlier fears that high levels of estrogens



D

ownl o a d e d f r om




























798 WYSOWSKI ET A L.

are associated with an increased risk of

breast cancer.

R E F E R E N C E S

1. Cutler SJ, Young JL Jr. Third National CancerSurvey; incidence data. Washin gton, D C: USG P O , 1975.

2. Thomas DB. Epidemiologic and related studies ofbreast cancer etiology. In: Lilienfeld AM, ed. Re-views in cancer epidemiology. Vol. 1. New York:Elsev ier /North Hol land, 1980.

3 . Kelsey JL. A review of the epidemiology of humanbreast cancer. Epidemiol Rev 1979;l:74-109.

4. Bloom W, Fawcett DW. A textbook of histology.10th ed. Philadelphia; WB Saunders Company,1975.

5. Nisker JA, Siiteri PK. Estrogens and breast can-cer. Clin Obstet Gynecol 1981;24:301-22.

6. W ynd er EL , Bross LJ, Hiraya ma T . A study of theepidemiology of cancer of the breast. Cancer1960;13:559-601.

7. Brinton LA, Wil l iams RR, Hoover RN, et al .Breast cancer risk factors among screening pro-gram participants . JNCI 1979;62:37-43.

8. He lmr ich S P, Shapiro S, Ros enbe rg L, et al. Riskfactors for breast cancer. Am J Epidemiol1983;117:35-45.

9. Valaoraa VG, MacMahon B, Trichopoulos D , etal. Lactation and reproductive histories of breastcancer patients in Greater Athen s, 1 965-67. Int JCancer 1969;4:350-63.

10 . Ravnihar B, MacMahon B, Ldndtner J. Epidemi-ologic features of breast cancer in Slovenia, 1965-1967. Eur J C ancer 1971;7:295-306.

11 . MacMahon B, Cole P, Lin TM, et al . Age at f irstbirth and breast cancer risk. Bull WHO1970;43:209-21.

12 . Bain C, Willett W, Rosner B, et al. Early age atfirst birth and d ecreased risk of brea st cancer. AmJ Epidemiol 1981;114:705-9.

13 . Feinleib M. Breast cancer and artificial meno-pause: a cohort study. JNC I 1968;41:315-29.

14 . Kelsey JL, Fischer DB, Holford TR, et al Exog-enous estrogens and other factors in the epide-miology of breast cancer. JNCI 1981;67:327-33.

15 . Rose C, Thorpe SM, Andersen KW, et al . Bene-ficial effect of adjuvant tamoxifen therapy in pri-mary breast cancer patients with high oestrogenreceptor values. Lancet 1985;1:16-19.

16. MacMahon B, Cole P, Brown JB, et al . Urineestrogens, frequency of ovulation, and breast can-cer risk: case-control study in premenopausalwomen. JNCI 1983;7(h247-50.

17 . Reed MJ, Cheng RW, Noel CT, et al . Plasmalevels of estrone, estrone sulfate, and estradioland the percentage of unbound estadiol in post-menopausal women with and without breast dis-ease. Cancer Res 1983;43:3940-3.

18 . England PC, Skinner LG, Cottrel l KM, et al .

Seru m o estradiol-17 /J in wom en w ith benign andmalignant breast disease. Br J Cancer 1974;30:571-6 .

19 . Wang DY, Bulbrook RD, Hayward JL. Urinaryand plasma androgens and their relation to famil-ial risk of breast cancer. Eur J Cancer 1975;11:873-7 .

20. Brennan MJ, Bulbrook RD, Deshpande N, et al .

Urinary and plasma androgens in benign breastdisease. Lancet 1973;l:1076-9.

21. BriggB M. Ethnic differences in urinary oestro-gens. Lancet 1972;1:324.

22 . MacMahon B, Cole P, Brown JB, et al . Urine

oestrogen profiles of Asian and North Americanwomen. Int J Cancer 1974;14:161-7.

23 . Dickinson L E, MacM ahon B , Cole P, et al. Estro-gen profiles of Oriental and Caucasian women inHawaii . N Engl J Med 1974;291:1211-13.

24 . Bulbrook RD, Swain MC, Wang DY, et al . Breastcancer in Britain and Japan: plasma oestradiol-17/J, oestrone and p rogesteron e, and their urinarymetabolites in normal British and Japanesewomen. Eur J Cancer 1976;12:725-35.

25 . Gross J, Modan B, Bertini B, et al. Relationshipbetween steroid excretion patterns and breast can-cer incidence in Israeli women of various origins.JNCI 1977;59:7-11.

26. Cole P, Brown JB, MacMahon B. Oestrogen pro-files of parous and nulliparous women. Lancet1976;2:596-9.

27. Pike MC, Casagrande JT, Brown JB, et al . Com-parison of urinary and plasma hormone levels indaughters of breast cancer patients and controls .JNCI 1977;59:1351-5.

28 . Morgan RW, Vakil DV , Brown JB , et al . Estrogenprofiles in young women: effect of maternal his-tory of breast cancer. JNCI 1978;60:965-7.

29 . Wang DY, Moore JW, Thomas BS, et al . Plasmaand urinary androgens in women with varyingdegrees of risk of breast cancer. Eur J Cancer1979;15:1269-74.

30. Bulbrook RD, Moore JW, Clark GMG, et al.Plasma oestradiol and progesterone levels inwomen with varying degrees of risk of breastcancer. Eur J Cancer 1978;14:1369-75.

31. Boffard K, Clark GMG, Irvine JBD, et al. Serumprolactin, androgens, oestradiol, and progesteronein adolescent girls with or withou t a family historyof breast cancer. Eur J Cancer Clin Oncol 1981;17:1071-7.

32. Trichopoulos D, Brown JB, Garas J, et al. Ele-vated urine estrogen and pregnanediol levels indaughters of breast cancer patients . JNCI 1981;67:603-6.

33. Klinga W, von Hoist TH, Runnebaum B. Serumconcentrations of FSH, oestradiol , oestrone andandrostenedione in normal and obese women. M a-turitas 1982;4:9-17.

34. Poortman J, Thijssen JH, de Waard F. Plasmaoestrone, oestradiol and androstenedione levels inpost-menopausal women: relation to body weightand height. Maturitas 1981;3:65-71.

35 . Bulbrook RD, Hayward JL, Spicer CC. Relationbetween urinary androgen and corticoid excretionand subsequent breast cancer. Lancet 1971;2:395-8.

36. Em men t Y, Coll ins WT , Sommervil le IF. Ra-dioimmunoassay of oestrone and oestradiol in hu-

man plasma. Acta Endocrinol 1972 :69:567-82.37. Abraham GE, Buster JE, Lucas LA, et al . Chro-

matographic separation of steroid hormones foruse in radioimmunoassay. Anal Lett 1972;5:509-17 .

38. Farewell VT. The combined effect of breast cancerrisk factors. Cancer 1977;40:931-6.

39. Lemon HM. Endocrine influences on human


a j e. ox f or d j o ur n

al s . or g






























mammary cancer formation. Cancer 1969;23:781-90.

40. Cole P, MacMahon B. Oestrogen fractions duringearly reproductive life in the aetiology of breastcancer. Lancet 1969;l:604-6.

41 . Rudali G, Apiou F, Muel B. Mammary cancerproduced in mice with estriol. Eur J Cancer1975;11:39-41.

42. Longcope C, Pr att JH . Blood production rates ofestrogens in women with differing ratios of uri-nary estrogen conjugates. Steroids 1977^9:483-92 .

43. Cowan LD, Gordis L, Tonascia JA, et aL Breastcancer incidence in women with a history of pro-gesterone deficiency. Am J Epidemiol 1981;114:209-17.

44. Schindler AE, Ebert A, Friedrich E. Conversion

of androstenedione to estrone by hum an fa t tissue.J Clin Endocrinol Metab 1972;35:627-30.45. MacDonald PC, Edman CD, Hemsell DL, et aL

Effect of obesity on conversion of plasma andro-stenedione to estrone in postmenopausal womenwith and w ithout endometrial cancer. Am J ObstetGynecol 1978;130:448-55.

b y g u e s t on S e p t em b er 2 5

,2 0 1 0




























mean estradiol

Documents