abnormal hormone levels in men with coronary artery · abnormal hormone levels in men with coronary...

Abnormal Hormone Levels in Men withCoronary Artery

Barnett Ziimoff, Raymond G. Troxler, John O'Connor, Robert S. Rosenfeld,Jacob Kream, Joseph Levin, James R. Hickman, Alexander M. Sloan,

William Walker, Rosa L. Cook, and David K. Fukushima

Plasma concentrations and urinary excretions of various hormones and hormonemetabolites were measured In four groups. Group 1 was composed of 13 men withprior myocardial infarction; Group 2 contained 35 clinically normal men; Group 3consisted of 44 men with normal coronary arteriograms; and Group 4 was composedof 25 men with severe coronary artery disease shown on arteriogram but no Infarction.There were four major findings: Group 1 had significantly higher 24-hour meanplasma concentrations of estrone (E1), dehydrolsoandrosterone (DHA), and dehydro-isoandrosterone sulfate (DHAS) than Group 2, while Group 3 had the same levels asGroup 4; Group 4 had significantly lower urinary excretion of androsterone glu-curonlde (AG) than Group 3, while Group 1 excreted normal amounts. There are threepossible explanations for these findings: 1) myocardial Infarction occurring in menwith coronary artery disease may elevate the plasma levels of E1, DHA, and DHAS andeliminate the prelnfarctlon depression of urinary AG levels; 2) higher than averagelevels of E1, DHA, DHAS, and AG may favor the development of Infarction In men withcoronary artery disease; 3) higher than average levels of E1, DHA, DHAS, and AG mayfavor survival from any infarction that occurs In men with coronary artery disease.Experimental and epldemlologlcal evidence seems to favor the third possibility.(Arteriosclerosis 2:58-67, January/February 1982)

The classical, established risk factors for cor-onary artery disease are hyperlipidemia,

hypertension, and cigarette smoking. In addi-tion, the fact that women show a strikingly lowerdeath rate from coronary disease than men hasbeen recognized for over 40 years.1'2 This obser-vation has focused interest on hormonal factors

From the Departments of Medicine and Biochemistry,Montefiore Hospital and Medical Center, Albert EinsteinCollege of Medicine, Bronx, New York; the United States AirForce School of Aerospace Medicine, San Antonio, Texas;and Malcom Grow United States Air Force Medical Center,Andrews Air Force Base, Washington, D.C.

Supported in part by Grant RR-53 from the General Clin-ical Research Centers Branch, National Institutes of Health,and by Contracts F49620-78-C-0O76 and F49620-79-C-0136from the United States Air Force Office of ScientificResearch.

The opinions expressed in this paper are those of theauthors and do not necessarily represent the opinions of theUnited States Air Force.

Address for reprint requests: Dr. Barnett Zumoff, Depart-ment of Medicine, Beth Israel Medical Center, New York,New York 10003.

Manuscript received April 10, 1981; revision acceptedDecembers, 1981.

in the disease. A number of groups3"14 have in-vestigated the hormonal status of men withcoronary disease. Earlier studies3"7 measuredurinary hormone excretion and described abnor-malities of estrogen,4-5 androgen,367 and corti-coid6 metabolite excretion. More recent studieshave described plasma hormone levels, usingradioimmunoassay methods. Six studies810"14

have reported elevated plasma estrogen levels,one9 has reported elevated plasma cortisollevels, and two8-13 have reported elevations ofmany different hormones.

All these studies considered only patients whohad had one or more myocardial infarctions.Because any abnormalities found in such pa-tients might result from the infarction itself, wefelt it essential to study patients with proven cor-onary artery disease, as defined by the presenceor arteriographically demonstrated coronaryartery stenosis, who had never had a myocardialinfarct and to compare their data with those ofarteriogram-negative men and men who had hadmyocardial infarcts.

58

by guest on May 22, 2015http://atvb.ahajournals.org/Downloaded from

http://atvb.ahajournals.org/

HORMONES IN CORONARY DISEASE Zumoff et al. 59

Methods

Groups of Men Studied

Group 1: Postinfarction Patients

Group 1 consisted of 13 men aged 35 to 56years (mean ± so = 48 ± 5.9 years) who had hadone or more myocardial infarctions. Infarctionwas established in 11 cases by a typical history,typical electrocardiographic changes, andtypical enzyme changes; in one case, by a typicalhistory, an abnormal electrocardiogram compat-ible with prior infarction, and the presence of anadynamic segment of ventricular wall on ventric-ulography; in one case, by the presence of anadynamic segment and nonspecific electrocar-diographic abnormalities. In every case, theinfarction had taken place a minimum of 6months before our studies and the patient wasclinically well, ambulatory, working or capable ofworking, and taking no medication, exceptnitrates in a few cases. No patient had congestiveheart failure, hypertension, diabetes, or othermajor past or current disease; none had had acoronary bypass; none was significantly obese(more than 20% above desirable weight); and allhad normal thyroid, kidney, and liver function.

Group 2: Clinically Normal ControlsThis group contained 35 men aged 21 to 85

years (mean ± SD = 44 ± 16 years) who were freeof any significant past or current disease, on thebasis of a complete history, physical examina-tion, and screening laboratory tests (SMA-6,SMA-12, complete blood count, chest x-ray, andelectrocardiogram), and who were taking nomedication. For plasma hormonal parametersthat show secular aging changes (testosterone,dehydroisoandrosterone, dehydroisoandros-terone sulfate, androsterone, androsterone sul-fate and follicle-stimulating hormone), onlythose values from the 16 age-matched controls(mean ± SD = 48 ± 5.5 years) were comparedwith those of the postinfarction patients; for allother plasma hormonal parameters, values fromthe total control group were compared withthose of the patients.

Group 3: Men with Normal CoronaryArteriograms

In this group there were 44 men, aged 35 to 55years (mean ± SD = 43 ± 5 years) who had beenreferred to the United States Air Force School ofAerospace Medicine (USAFSAM) for cardiacevaluation because of minor electrocardio-graphic abnormalities and/or arrhythmias ob-served in routine electrocardiograms. As part ofthis evaluation, they had negative coronary

arteriograms (defined as no more than minimalroughening in any artery). Men with significantpast or current disease or significant abnormali-ties on screening laboratory tests and mentaking medication of any kind were excludedfrom study.

Group 4: Men with Positive CoronaryArteriograms

This group contained 25 men, aged 35 to 55years (mean ± SD = 50 ± 5 years) similarlyreferred to USAFSAM and similarly studied, whowere found to have positive arteriograms (de-fined as more than 50% occlusion of the lumen ofone or more arteries). The exclusionary clinicaland laboratory criteria were the same as forGroup 3. Two of the men originally considered tobelong in this group were found to haveadynamic ventricular wall segments by ventricu-lography; in one, an earlier history of typicalinfarction symptoms was elicited retrospectively,and his electrocardiogram was considered to becompatible with prior infarction. Both men werereclassified into the postinfarction group (Group 1).

Design of the Study

All the men in Groups 1 and 2 had 24-hourplasma hormone studies, as previously de-scribed,15 in the clinical Research Center ofMontefiore Hospital and Medical Center. Ninemen in Group 1 and four men in Group 2 had24-hour urine collections for hormone analysiswhile in the Center. All the men in Groups 3 and4 had a 24-urine collection early during theirevaluation at USAFSAM. The urines were sent toMontefiore Hospital for "blind" analysis; thecode was broken after the analysis was com-plete. Sixteen men from Group 2 (mean age ± SD= 43 ± 4 years) and 13 men from Group 4 (meanage ± SD = 50 ± 6 years) traveled from theirhomes to Montefiore Hospital for a 24-hourplasma hormone study; before the study wasdone, they rested at least 1 day for each timezone traversed.

Parameters Measured

In the plasma studies, the 24-hour mean plas-ma concentration of 14 hormones or hormonemetabolites were measured: cortisol, testos-terone, dihydrotestosterone, dehydroisoandros-terone sulfate, androsterone, androsteronesulfate, estrone, estradiol, triiodothyronine,thyroxine, luteinizing hormone, follicle-stim-ulating hormone, and prolactin. In the urinestudies, 15 hormones or hormone metaboliteswere measured: testosterone glucuronide, dihy-drosterone glucuronide, androsterone glucuro-



60 ARTERIOSCLEROSIS VOL 2, No 1, JANUARY/FEBRUARY 1982

nide, androsterone sulfate, etiocholanoloneglucuronide, etiocholanolone sulfate, dehydro-isoandrosterone glucuronide, dehydroisoandros-terone sulfate, tetrahydrocortisol glucuronide,allotetrahydrocortisol glucuronide, tetrahydro-cortisone glucuronide, estrone glucuronide,estradiol glucuronide, estriol glucuronide, andfree cortisol. The 24-hour urinary excretion ofcreatinine was measured in all urinary studiesand all urinary excretion values were expressedas milligrams or micrograms per gram ofcreatinine, in order to "normalize" for differencesin body mass.

Analytical Methods

Plasma cortisol was determined by competi-tive protein-binding,16 testosterone and dihydro-testosterone were determined as described byBoyar et al.,17 dehydroisoandrosterone was de-termined as described by Rosenfeld et al.,18

dehydroisoandrosterone sulfate was determinedas described by Nieschlag et al.,19 androsteroneand androsterone sulfate were determined asdescribed by Kream et al.,20 triiodothyronine andthyroxine were determined as described byO'Connor et al.,21 estrone and estradiol weredetermined as described by Zumoff et al.,22

luteinizing hormone and follicle-stimulatinghormone were determined as described byMidgely,23-24 and prolactin was determined asdescribed by Sinha et al.25 Free cortisol wasextracted from urine and quantitated as de-scribed by Kream et al.26 Urines were thenadsorbed on columns of XAD-2,27 eluted, hydro-lyzed with /J-glucuronidase to yield the glucuron-ide fraction, and finally solvolyzed to yield thesulfate fraction. Tetrahydrocortisol, allotetra-hydrocortisol, and tetrahydrocortisone werequantitated in the glucuronide fraction by paperchromatographic separation and colorimetricmeasurement by reaction with blue tetrazolium,as previously described.28 All other steroids inthe respective urine fractions were quantitatedby radioimmunoassay, by the same methods asfor plasma hormones.

The 24-hour mean plasma hormone concentra-tions were determined by sampling blood froman indwelling venous catheter every 20 minutes,pooling aliquots from each of the 72 samples,and measuring the hormone concentrations ofthe pool, as previously described.15

Statistical Methods

Most of the plasma hormone and hormone me-tabolite concentrations showed log-normalrather than normal distribution. Geometricmeans as well as arithmetic means were calcu-lated for each substance, and two comparisons

Table 1. Plasma Hormone Levels that Were Not Signif-icantly Different In Normal and Postlnfarctlon Men

Hormone

Cortisol (M9/CII)Androsterone sulfate

(M9/dl)Thyroxine <M9/dl)Testosterone (ng/dl)Dihydrotestosterone

(ng/dl)Androsterone (ng/dl)Triiodothyronine

(ng/dl)tEstradiol (pg/ml)Luteinizing hormone

(mlU/ml)Follicle-stimulating

hormone (mlU/ml)Prolactin (ng/ml)

Normal(n = 35)*

7.0 ±1.4

47 ±276.3 ±2.2421 ±104

102 ±3057 ±18

117 ± 1629 ±6.0

12 ±6.1

9.9 ±3.112 ±6.5

Post-infarction(n = 13)

7.3 ±1.3

65 ±485.0 ±0.9463 ±187

121 ±3371 ±36

116 ± 2029 ±8.0

14 ±5.3

9.2 ±6.512 ±4.3

All values are arithmetic means ± standard deviation.'Only the 16 age-matched controls were compared

with postinfarction patients for dehydroisoandros-terone sulfate, testosterone, androsterone, andros-terone sulfate and follicle-stimulating hormone, be-cause the concentrations of these hormones vary withage.

fThe mean value for the postinfarction group ex-cludes the three grossly subnormal values (see fig. 4).

Table 2. Plasma Hormone Levels In Men with Positiveor Negative Coronary Arterlograms

HormoneCortisol (jug/dl)Dehydroisoandros-

terone sulfate(Mg/di)*

Androsterone sulfate(Mg/dl)

Testosterone (ng/dl)Dihydroisoandros-

terone (ng/dl)Dehydroisoandros-

terone (ng/dl)*Androsterone (ng/dl)Triiodothyronine

(ng/dl)Estrone (pg/ml)*Estradiol (pg/ml)Luteinizing hormone

(mlU/ml)Follicle-stimulating

hormone (mlU/ml)Prolactin (ng/ml)

Positive(n = 13)6.7 ±1.4

105(39-285)

61 ±45478 ±184

117±44

361 (168-778)80 ±18

108 ±2262 (37-103)

26 ±10

11 ±2.3

13 ±5.810 ±5.6

Negative(n = 16)6.6 ±1.2

142 (51-396)

82 ±34533 ±178

114 ±41

422 (238-748)81 ±28

106 ±1957 (32-100)

32 ±14

13 ±5.5

13 ±8.812 ±5.3

*Values are geometric mean, with 95% confidencelimits in parentheses. All other values are arithmeticmean ± standard deviation.




were made for all of them: Group 1 vs Group 2,and Group 3 vs Group 4. The results for 11substances that showed no difference betweenGroup 1 and Group 2 (see Results section) arepresented in table 1 as arithmetic means withstandard deviations, for simplicity. Geometricmeans and 95% confidence limits are listed in thetable for the three substances that did differ(estrone, dehydroisoandrosterone, and dehydro-isoandrosterone sulfate); the individual datapoints were plotted logarithmically, and the sig-nificance of the Group 1 vs Group 2 differenceswere calculated by applying Student's t test, two-tailed and unpaired, to the logarithms of theplasma concentration.

Comparison of Group 3 vs Group 4 showed nodifference for any of the 13 plasma substancesmeasured. Table 2, which presents these data,contains arithmetic means and standard devia-tions for the 10 substances so presented in table1, and geometric means and 95% confidencelimits for the other three substances, to facilitatecomparison with the means shown for them intable 1.

In evaluating the statistical significance of thedifference in urinary androsterone glucuronideexcretion between Group 3 and Group 4, weused both Student's t test and the x2 t e s t withrespect to the Group 3 mean level as a discrim-inant, as further discussed in the Results section.

T3

z

oa:>-xt -oQ

oE

<

<_ l0 -

Z<UJ

s.orx

i

<M

I60-

I40-

I2O-

IOO-

80-

60

40

20-

I

NORMAL POST-INFARCTION

Figure 1. 24-hour mean plasma triiodothyronine Innormal men and in men who had a myocardial infarc-tion. In the postinfarction group, the values indicatedby x are clearly subnormal; the mean value for thisgroup (indicated by the horizontal dotted line) wascalculated omitting these three values, p = notsignificant.

Results

Plasma Hormone Concentrations

Group 1 vs Group 2

Eleven of the 14 hormones and hormone me-tabolites measured showed no significant differ-ence between these groups (table 1); in the caseof T3 (figure 1), three of the men in Group 1 hadsubnormal levels (56, 62, and 66 ng/dl) but nineothers had values whose range and mean wereessentially identical to those of Group 2; (T3 wasnot determined in the 13th member of thisgroup). The other three hormones, estrone (E1),dehydroisoandrosterone (DHA), and dehydroiso-androsterone sulfate (DHAS), showed signif-icant differences between Groups 1 and 2: thegeometric mean E1 in Group 1 (80 pg/ml; 95%confidence limits [CL] 49 to 131) was elevatedcompared with that of Group 2 (49 pg/ml; 95%CL 24 to 98) (p < 0.0001) (figure 2); the geo-metric mean DHA in Group 1 (444 ng/dl; 95% CL198 to 995) was elevated compared with that ofGroup 2 (298 ng/dl; 95% CL 143 to 619) (p <0.025) (figure 3), and the geometric mean DHASin Group 1 (112 /ug/dl; 95% CL 36 to 347) waselevated compared with that of Group 2 (72

I60-,

UJzoccV)UJ

<_ JQ .

<UJ

I

I

CM

80-

40 -

20-

95% CL.

I

95% CL.


Figure 2. 24-hour mean plasma estrone concentra-tion in normal men and In men who have had a myo-cardial infarction, p < 0.0001.




<

<_la.zUl

crI

I

CM

800-

400-

200-

100-

p< 0.025

95% CL. .

t r• 95% CL.

•


Figure 3. 24-hour mean plasma dehydroisoandros-terone concentration in normal men and in men whohave had a myocardial infarction, p < 0.025.

/xg/dl; 95% CL 26 to 202) (p = 0.05) (figure 4). It isnoteworthy that for each of these threehormones all but one of the postinfarctionpatients (Group 1) had values that were at orbelow the upper 95% confidence limit of the con-trol values (Group 2); the significant differencebetween groups was due in each case to cluster-ing of the patients' values in the upper half of thenormal range. This point bears on the interpreta-tion of the results, as will be amplified in theDiscussion.

Group 3 vs Group 4

None of the 13 hormones and hormone metab-olites measured (thyroxine was omitted in thiscase) showed a significant difference betweenthese two groups (table 2), including E1, DHA,and DHAS. Interestingly, the combined arterio-gram group (i.e., positives plus negatives) hadsignificantly higher mean plasma levels of DHA,DHAS, androsterone, and androsterone sulfate

p = 0.05

=5 200-

Q*r 100-

<_la.

55 50

CVJ25-

95%C.L.

t

95% CL.


Figure 4. 24-hour mean plasma dehydroisoandros-terone sulfate concentration in normal men and inmen who have had a myocardial infarction, p < 0.05.

a>c

o«O

6-,

x* = IO.9p<0.00l

t =2.16p<0.05

zocr

I 4oUlzocrUJ 2 .

COocra

cr<zcr3

t

NEGATIVE

ARTERIOGRAM

POSITIVE

ARTERIOGRAM

Figure 5. Urinary excretion of androsterone glu-curonide in men with positive or negative coronaryarteriograms. x2 = 10.9; p < 0.001; f = 2.16; p< 0.05.




Table 3. Plasma Hormone Levels that Were Significantly Higher In the Combined ArteriogramGroup than In Normal Men

HormoneDehydroisoandrosterone (ng/dl)Androsterone (ng/dl)Dehydroisoandrosterone sulfate (MQ/dl)Androsterone sulfate (/jg/dl)

Normalmen

(n = 16)298(143-619)

57 ±1883 ±4847 ±27

Arteriogramgroup

(n = 29)394(201-771)

80 ±23140 ±6573 ±40

Percentagedifference

32406955

(P)

< 0.025< 0.001< 0.001< 0.025

All values are arithmetic means ± standard deviation except dehydroisoandrosterone, whichfor comparability with table 4 is given as geometric mean with 95% confidence limits inparentheses.

than the normal controls (table 3). It is not clearwhy the military group showed higher levels ofadrenal androgens and androgen metabolitesthan did the civilian controls; the mean age of thetwo groups (47 ± 5 and 44 ± 16 years, respec-tively) and mean percentage deviation from idealweight (12 ± 8 and 14 ± 7, respectively) did notdiffer significantly.

Urinary Hormone Excretion

Group 3 vs Group 4

Of the 15 hormones and hormone metabolitesstudied, 14 showed no difference between thesegroups (table 4). Only androsterone glucuro-node showed a difference (3.2 ± 1.1 mg/g creati-nine in Group 3 vs 2.6 ± 0.9 mg/g creatinine inGroup 4; figure 5). Using Student's t test, two-tailed and unpaired, we found the difference tobe significant (p<0.05). An alternative statisticaltreatment, using the negative arteriogram

group's mean excretion of 3.2 mg/g creatinine asa "discriminant" and comparing the number ofpersons in each group with excretion above orbelow this value, yielded a x2 value of 10.9,equivalent to a statistical significance of p <0.001 for the difference between groups.

Comparison of Androsterone GlucuronldeExcretion In All Groups

Figure 6 shows that the means and ranges ofGroups 1,2, and 3 were similar. Group 4, the menwith positive arteriograms, showed extremes ofrange that were also similar to those of the otherthree groups, but 85% of the values were clus-tered in the lower half of the range, so that themean of this group was significantly lower thanthose of Groups 1 (postinfarction) and 3 (nega-tive arteriograms). There were too few subjectsin Group 2 (clinically normal) to evaluate thedata statistically, but its mean was also higherthan that of Group 4.

Table 4. Urinary Hormone Excretion In Men with Positive or Negative Coronary Arteriograms

Hormone Positive (n = 25) Negative (n = 44)Androsterone sulfate (mg/g creatinine)Etiocholanolone glucuronide (mg/g creatinine)Etiocholanolone sulfate (mg/g creatinine)Dehydroisoandrosterone glucuronide (mg/g creatinine)Dehydroisoandrosterone sulfate (mg/g creatinine)Tetrahydrocortisol glucuronide (mg/g creatinine)Allotetrahydrocortisol glucuronide (mg/g creatinine)Tetrahydrocortisone glucuronide (mg/g creatinine)Testosterone glucuronide (M£)/9 creatinine)Dihydrotestosterone glucuronide (Mg/g creatinine)Free cortisol {ixg/g creatinine)Estrone glucuronide (Mg/g creatinine)Estradiol glucuronide (MQ/9 creatinine)Estriol glucuronide (txg/g creatinine)Creatinine (g/day)

0.5 ±0.42.4 ±1.40.2 ±0.20.1 ±0.10.6 ±0.61.6 ± 0.51.2 ±0.52.0 ±0.534±1821 ±1338 ±7.8

5.6 ±1.72.0 ±0.85.4 ±2.1

1.68 ±0.4

0.7 ±0.52.5 ±1.10.3 ±0.60.2 ±0.10.9 ±0.81.4 ±0.31.3 ±0.41.8 ±0.429 ±1522 ±1327 ± 95.3 ±1.21.8 ±0.75.4 ±1.9

1.70 ±0.28

All values are arithmetic mean ± standard deviation.




o

LL.

oO UJ

LJ Za: oo crX 3LJ O

> - I

<Z LU— z^ o

UJ

Ocro

6-,

if 4-

2-

•

r ftt

NEGATIVEARTERIOGRAM

POSITIVEARTERIOGRAM


Figure 6. Urinary excretion of androsterone glucuronide in the four groups studied: men withnegative arteriograms, men with positive arteriograms, normal men, and men who have had amyocardial infarction.

Discussion

Aside from the minor finding that a subset ofpostinfarction patients (3 out of 12) showed sub-normal plasma T3 levels (probably the nonspe-cific Iow-T3 syndrome), there were four majorobservations in this study: 1) postinfarction pa-tients showed significantly elevated plasma con-centrations of estrone (E1), dehydroisoandros-terone (DHA), and dehydroisoandrosteronesulfate (DHAS), compared with normal controls;2) men with positive coronary arteriograms butno history of infarction had values that did notdiffer from those of men with negative arterio-grams; 3) men with positive arteriograms but nohistory of infarction showed subnormal urinaryexcretion of androsterone glucuronide; 4) post-infarction patients excreted normal quantities ofthis steroid.

There have been no previous reports aboutplasma DHA or DHAS levels in coronary disease.Elevated plasma estrogen levels have been pre-viously reported by six groups, but the detailsvary. Wagner6 reported that postinfarction pa-tients showed elevations of estradiol, but he didnot measure estrone. Phillips10 and Entrican etal.11 reported that postinfarction patients showedelevations of both estrone and estradiol. Levinand Korenman12 reported that patients studiedwithin 2 days of an infarction showed elevated

levels of estradiol but not estrone. Pego et al.14

also reported elevated estradiol levels in such pa-tients, but they did not measure estrone.Pivovarov et al.13 reported elevated estradiollevels (estrone levels were not measured) in anincompletely characterized group of men withpositive arteriograms, some or possibly most ofwhom had had previous myocardial infarctionsbut none of whom were studied in the first fewweeks after infarction. We found elevated levelsof estrone but not estradiol. Despite these differ-ences, it can be said that all seven groups thathave studied postinfarction patients have re-ported elevated plasma estrogen levels. How-ever, until now no group had reported the plasmaestrogen levels of a group of well-characterizedmen with markedly abnormal coronary arterio-grams who had not had a myocardial infarction.Our finding that this group did not have elevatedplasma estrogen levels is therefore unique and,we think, of importance in formulating a hypoth-esis concerning the possible role of hormonalfactors in coronary disease.

Two groups of workers have described subnor-mal urinary excretion of androsterone after myo-cardial infarction. Rao7 reported subnormalexcretion of this steroid in men studied 1 to 2weeks after infarction, and Marmorston et al.6

reported decreased excretion in patients studied




at least 3 months after infarction. Neither groupstudied preinfarction patients like ours. Ourfindings of subnormal urinary excretion ofandrosterone glucuronide in preinfarction butnot in postinfarction patients is again unique.

There are several possible explanations for ourfinding that postinfarction patients showed threeplasma hormonal abnormalities that were notpresent in "preinfarction" patients (i.e., men withpositive coronary arteriograms but no history orevidence of prior infarction), and failed to showone urinary hormonal abnormality that was pres-ent in preinfarction patients. One explanation isthat infarction itself caused these changes. Thisis certainly possible, although it seems easier tobelieve that a physiological catastrophe like infarc-tion can produce new abnormalities than that itcan abolish old ones. In any case, if this is themechanism, it provides justification for the studyof preinfarction patients and suggests that theandrosterone glucuronide abnormality is theonly one of the four that is "basic" in coronaryartery disease, since it is present beforeinfarction.

A second possibility is that values for estrone,dehydroisoandrosterone, dehydroisoandroster-one sulfate, and androsterone glucuronide thatare in the upper half of the preinfarction rangepredispose to the development of infarction andare therefore over-represented in those whohave infarctions, thus raising their mean valuecompared with preinfarction patients. This is cer-tainly compatible with the distributions of hor-monal levels observed in the postinfarctionpatients.

A third possibility is that upper-half values forthese four hormones in preinfarction patients pre-dispose not to the development of infarction butto survival after an infarction occurs, and aretherefore higher than normal in postinfarction pa-tients who survive to be studied. This implies thatthe hormonal levels in the relatively well, highlyselected group of postinfarction survivors westudied are not typical of all postinfarctionpatients.

The best way to distinguish among these threepossibilities would be prospective studies. Un-fortunately, this is impractical. The relatively lowincidence of infarction even in a group of menwith positive arteriograms would mandate a verylarge number of man-years to achieve statisticalsignificance in such a study, and to do the studyin a reasonable time interval would thereforerequire a very large number of men. The logisticsand expense of such a study would beprohibitive.

An alternative, though less conclusive,approach would be to draw inferences from theknown pharmacological properties and epide-miological correlates of the four hormonesconcerned. On epidemiological grounds (i.e., the

protection that women have from myocardialinfarction and the loss of this protection inwomen castrated young), estrogens could beconsidered likely to favor a lower incidence ofinfarction. Indeed, an autopsy study29 has showna lower incidence of myocardial infarction inboth men and women treated with estrogens.The higher incidence of infarction in women whotake oral contraceptives30 and the higherincidence of reinfarction in postinfarction menwho take estrogens31 do not contradict thisfinding, since the estrogens used in both casesare all artificial synthetic compounds whoseingestion lowers the endogenous levels ofnatural estrogens; it has recently been reported32

that women who take noncontraceptive estro-gens do not have an increased incidence ofmyocardial infarction. Animal experiments3334

show that estrogens inhibit the formation of twotypes of experimental arteriosclerosis.

Additional light can be shed on this problemby considering the influence on myocardial in-farction of two disease processes in which af-fected men have chronically elevated plasmaestrogens, namely, obesity22-35 and cirrhosis ofthe liver.36 In obesity, it has been reported37 thatsurvival from an infarction is neither lower norhigher than in normal men; the influence ofobesity on the incidence of infarction has beencontroversial, but it now appears that if the con-founding effects of associated hypertension, dia-betes, and hyperlipidemia are factored out,obesity per se probably does not affect the inci-dence.38^*1 No information is available about theeffect of cirrhosis on survival from an infarction,but the weight of evidence is that the incidenceof infarction is markedly decreased in cirrhoticpatients.42"44

In summary, higher estrogen levels certainlydo not increase the incidence of myocardial in-farction and may indeed decrease it, and theavailable evidence does not permit us to evaluatethe possibility that higher estrogen levels mightincrease survival from an infarction.

The production rates of DHA and DHAS havebeen reported to be increased in obesity,45 sothat all three steroids whose blood levels we findelevated in postinfarction patients (DHA, DHAS,and estrone) are produced in increased amountsin obesity. Both DHA and DHAS have been re-ported to decrease angina,4647 an effect that hasbeen attributed to the ability of DHA to depresscardiac oxygen consumption,48 and which wouldcertainly be compatible with improved survivalfrom an infarction when higher DHA and DHASlevels are present.

Androsterone, which is a peripheral metaboliteof the testicular and adrenal androgens, waslong ago reported49 to produce a profound low-ering of cholesterol in all types of hypercholes-terolemia. This property and the present observa-




tion that urinary androsterone glucuronideexcretion is subnormal in men with positive cor-onary arteriograms make it unlikely that andros-terone would increase the incidence of myocar-dial infarction.

Finally, it would seem desirable to investigatethe effects of estrone, dehydroisoandrosterone,dehydroisoandrosterone sulfate, and andros-terone in one or more animal models of arterio-sclerosis.

Acknowledgments

We are grateful for the outstanding assistance of manytechnicians who helped carry out the arteriographic, blood-sampling, and radloimmunoassay tests, and without whosededicated efforts these studies could not have been done. Wethank Dr. Gerald Phillips for stimulating discussions thathelped us in the interpretation of the data, and Lt. ColonelGeorge Kush USAF, MSC (Ret.) whose Interorganizationalmediation made it possible to get these studies underway.

References

1. Glendy RF, Levlne SA, White PD. Coronary disease inyouth. JAMA 1937;109:1775-1781

2. Matter AM, Dack J, Jaffe HL. Age, sex, and hypertensionin myocardial infarction due to coronary inclusion. ArchIntern Med 1939;64:676-786

3. Qertler MM, White PD. Coronary disease in young adults.Cambridge: Harvard University Press, 1954

4. Bauld WS, Glvner ML, Milne IG. Abnormality of estrogenmetabolism in human subjects with myocardial infarc-tion. Can J Blochem 1957;35:1277-1288

5. Bersohn I, OelofM PJ. Urinary estrogen levels in myocar-dial infarction. S Afr Med J 1958;32:979-983

6. Marmorston J, Geller PJ, Welner JM, Allaln CC, Para JH,Bush IE, Robert JB. An extension of the "coronary-proneprofile" based on an abnormal pattern of urinary steroidsIn myocardial infarction. Physiol Chem Phys 1970;2:337-350

7. Rao LGS. Urinary steroid excretion patterns after acutemyocardial Infarction. Lancet 1970:2:390-391

8. Wagner H. Endokrin-metabolische Storungen bei Arte-riosklerose. Stuttgart: Gustav Fisher Verlag, 1975

9. Troxler RG, Sprague EA, Albanese RA, Fuchs R, Thomp-son AJ. Association of elevated plasma cortisol and earlyatherosclerosis as demonstrated by coronary angiog-raphy. Atherosclerosis 1977;26:151-162

10. Phillips GB. Evidence for hyperoestrogenemia as a riskfactor for myocardial Infarction in men. Lancet 1976;2:14-18

11. Entrlcan JH, Beach C, Carroll D, Kenmure ACF, KlopperA, Mackie M, Douglas AS. Raised plasma oestradiol andoestrone levels in young survivors of myocardial infarc-tion. Lancet 19782:487^90

12. Levin LC, Korenman SG. Elevated estradlol levels Inacute Ml and coronary artery disease. Clin Res 1978;26:308A

13. Plvovarov VN, Rossels AN, Kasatklna LV, Kramer AA:Hormones in ischemic heart disease with coronaryatherosclerosis. Kardiologiia (Moscow) 1978;18:30-36

14. Pego R, Johnson M, Wltland M, Lurla M, Wleland RG.Increased estradlol levels in males during acute phase ofmyocardial infarction. Clin Res 1979;27:680A

15. Zumofl B, Rosenfeld RS, Strain QW, Levin J, Fukuthima

DK. Sex differences in the 24-hour mean plasma concen-trations of dehydroisoandrosterone (DHA) and dehy-droisoandrosterone sulfate (DHAS) and the DHA/DHASratio In normal adults. J Clin Endocrinol Metab 1980;51:330-333

16. Murphy BEP. Some studies of the protein-binding ofsteroids and their application to the routine micro andultramicro measurement of various steroids in bodyfluids by competitive protein-binding radloassay. J ClinEndocrinol Metab 1967:27:973-990

17. Boyar RM, Rosenfeld RS, Kapen S, Flnkelsteln JW, Rotf-warg H, WHtzman ED, Hellman L. Human puberty: simul-taneous augmented secretion of luteinizlng hormoneand testosterone during sleep. J Clin Inves 1974;54:609-618

18. Rosenteld RS, Rosenberg BJ, Fukuthima DK, Hellman L.24-hr secretory pattern of dehydroisoandrosterone anddehydroisoandrosterone sulfate. J Clin EndocrinolMetab 1975;40:850-855

19. Nleschlag E, Lorlaux DL, Lipsett MB. Radiollgand assayfor AS-S/S-hydroxysteroids. I. 30-hydroxy-5-androsten-17-oneand Its 3-sulfate. Steroids 1972;19:669-679

20. Kream J, Hellman L, Rosenfeld RS. Radiolmmunoassayof androsterone and androsterone-3-sulfate in plasma.Steroids 197627:727-739

21. O'Connor JF, Wu RY, Gallagher TF, Hellman L. Plasmathyroxine profile in normal man. J Clin Endocrinol Metab39:765-771, 1974

22. Zumoff B, Strain GW, Kream J, O'Connor J, Levin J, Fuku-thima DK. Obese young men have elevated plasma estro-gen levels but obese premenopausal women do not.Metabolism 30:1011, 1981

23. Mldgely AR Jr. Radiolmmunoassay: A method for humanchorlonic gonadotropln and human luteinizing hormone.Endocrinology 1966;79:1O-18

24. Mldgely AR Jr. Radiolmmunoassay for human follicle-stimulating hormone. J Clin Endocrinol Metab 1976:27:295-299

25. Slnha YN, Selby FW, Lewis U J, Vanderiaan WP. A homol-ogous radioimmunoassay for human prolactin. J ClinEndocrinol Metab 1973:36:509-516

26. Kream J, Fukuthima DK, ZumoH B. A rapid competitiveprotein-binding radioassay of urinary cortisol. ClinChem 1978:24:994

27. Bradlow HL. Extraction of steroid conjugates with neu-tral resin. Steroids 1968:11:265-272

28. Fukushlma DK, Zumoff B, Bulkln W, Hellman L. Effect of7/3, 17o-dimethyltestosterone (calusterone) on cortisolmetabolism in women with advanced breast cancer. JClin Endocrinol Metab 1976;43:38-45

29. Rlvln AU, Dlmitrorf SP. The Incidence and severity ofatherosclerosis in estrogen-treated males and in femaleswith a hypoestrogenlc or a hyperestrogenic state. Circu-lation 1954:9:533-539

30. Mann Jl, Vettey MP, Thorogood M, Doll R. Myocardialinfarction in young women with special reference to oralcontraceptive practice. Br Med J 1975:2:241-245

31. Coronary Drug Project Group. Findings leading to dis-continuation of 2.5 mg/day estrogen group. JAMA1973;226:652

32. Rosenberg L, Sloane D, Shapiro S, Kaufman D, StolleyFD, Mlettinen O. Noncontraceptive estrogens and myo-cardial infarction in young women. JAMA 1980:244:339-342

33. Rhee CY, Spaet TH, Stemerman MB, Lajam F, ShlangHH, Caruso E, Utwak RS. Estrogen suppression of sur-gically induced vascular Intimal hyperplasia in rabbits. JLab Clin Med 1977;90:77-84

34. Drouet L, Rhee CY, Clntron J, Spaet TH. Estrogeninhibition of experimental arteriosclerosis in the rabbit.Fed Proc 1978,37:474

35. Schneider G, Krlshner MA, Berkowltz R, Ertel NH.Increased estrogen production in obese men. J ClinEndocrinol Metab 1979:48:633-638




36. Kley HK, Keck E, Kriiskemper HL. Estrone and estradiolin patients with cirrhosis of the liver: effects of ACTH anddexamethasone. J Clin Endocrinol Metab 1976;43:557-560

37. Kannel WB, Sortie PD. Prognosis following change Inweight in persons with coronary disease. Paper pre-sented at USPHS Professional Association Symposium,Bethesda, MD, March 27-30,1978

38. Keys A, Taylor HL, Blackburn H, Brozek J, Anderson JT,Slmonson E. Mortality and coronary heart diseaseamong men studied for 23 years. Arch Intern Med 1971;128:201-214

39. Carlson LS, Bottlger LE. Ischaemic heart-disease in rela-tion to fasting values of plasma trlglycerldes and cho-lesterol. Lancet 1972;1:865-868

40. Gordon T, Castelll WP, HJortJand MC, Kannel WB,Dawber TR. Diabetes, blood lipids, and the role ofobesity in coronary heart disease risk for women. AnnIntern Med 1977;87:393-397

41. Pekonen I, Nlkklla EA, Kosklnen S, Penttlnen K, Sarna S.Association of serum lipids and obesity with cardiovascu-lar mortality. BrMed J 1977;2:1185-1187

42. MacDonald RA, Mallory DK. The natural history of post-

necrotlc cirrhosis. A study of 221 autopsy cases. Am JMed 1958;24:334-357

43. Hirst AE, Hadley GO, Gou I. The effect of chronic alco-holism and cirrhosis of the liver on atherosclerosis. Am JMed Sci 1965;249:143-149

44. Vandcek R. Atherosclerosis and cirrhosis of the liver. BullWorld Health Organ 1976:53:567-570

45. FeherT, Halmy L. Dehydroepiandrosterone and dehydro-epiandrosterone sulfate dynamics In obesity. Can J Bio-chem 1974:53:215-222

46. Fassatl P, Fassatl M, Sonka J, Lesensky K. New approachto the treatment of angina pectoris with the aid of DHEA.Cas Lek Cesk 1971; 110:606-609

47. Fassatl P, Fassatl M, Sonka J, Lesensky K. Dehydroepian-drosterone sulphate — a new approach to some cases ofangina pectoris therapy. Agressologie 1973:14:259-268

48. Sonka J, Hilgertova J, serclk J, Kratochvil O. In vitro andin vivo Inhibitory effects of dehydroepiandrosterone onrespiration. Endocrinol Exp 1987:12:165-169

49. Hellman L, Bradlow HL, Zumoff B, Fukushlma DK, Gal-lagher TF. Thyroid-androgen interrelations and thehypocholesteremic effect of androsterone. J Clin Endo-crinol Metab 1959:19:936-948

Index Terms: coronary artery disease • myocardial infarction • plasma hormone levelsurinary hormone excretion • estrogens • estrone • dehydroisoandrosterone •dehydroisoandrosterone sulfate • androsterone



W Walker, R L Cook and D K FukushimaB Zumoff, R G Troxler, J O'Connor, R S Rosenfeld, J Kream, J Levin, J R Hickman, A M Sloan,

Abnormal hormone levels in men with coronary artery disease.

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1982 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleArteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.2.1.581982;2:58-67Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/2/1/58World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at: Reprints:

document.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:


http://atvb.ahajournals.org/content/2/1/58

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://atvb.ahajournals.org//subscriptions/


abnormal hormone levels in men with coronary artery · abnormal hormone levels in men with coronary...

Documents