Metabolism Clinical and Experimental JULY 1994
VOL 43, NO 7 PRELIMINARY REPORT
Changes in the Low-Density Lipoprotein Profile During 17P-Estradiol-Dydrogesterone Therapy in Postmenopausal Women Marius J. van der Mooren, Jacqueline
de Graaf, Pierre N.M. Demacker, Anton F.J. de Haan, and Rune Rolland
Changes in the low-density Iipoprotein (LDL) subfractions, determined with gradient gel electrophoresis (GGE), were prospectively studied during hormone replacement therapy (HRT) in 23 healthy nonhysterectomized postmenopausal women. LDL subfractions were analyzed by densitometric scanning of the gels. We observed significant changes in the LDL subfraction profile toward a smaller particle size (P t: ,001). These changes could almost completely be attributed to the hormonal treatment regimen lP c ,011, and mav indicate an effect Partially opposite to the other reported changes in the lipid profile. Copyright $‘ 7994 by W.B. S&ders t’ompany
H
ORMONE REPLACEMENT therapy (HRT) reportedly reduces the risk of developing coronary heart dist-ase (CHD) in postmenopausal women, which may be in Parr. due to the beneficial changes in the lipid profile.1,2 Small low-density lipoprotein (LDL) particles are associated with both the male sex and increased age,3 and have also been identified as a risk factor for CHD.“-‘j We studied possible changes in the LDL subfraction profile, as found by gradient gel eiectrophoresis (GGE), in healthy postmenopausal women during 1 year of HRT. The changes were related to those in serum lipids, lipoproteins, ant: apolipoprotcins. SUBJECTS AND METHODS ‘I he study was approved by the ethical committee of our hospital, and before study entry the participants gave their written informed conlent. Recruited were healthy nonhysterectomized postmenopausal women (age. 54.3 2 3.5 years. mean rt SD) who were amt!norrhoeic for at least h months (56 t 40 months) and who were screened to have serum follicle-stimulating hormone concentralions within the range characteristic of the postmenopausal phase (84 2 26 IlJiL). Selection criteria and assays have been described elsewhere.? All women were treated with micronized 17p-estradiol (Zumenon: Solvay Duphar B.V.. Weesp. The Netherlands) 2 mg daily ancl with dydrogesterone (Duphaston; Solvay Duphar B.V.) 10 mg daily for the first half of each E-day treatment cycle, and both were oraily adminisl~re~~. t)et~rminati~~n of the LDL subfraction profile was performed in venous blood samples (evacuated collection tubes. Corvac, Becton Dickinson. Cedex. France) taken after a 12-hour fast, before and aftc.r 1 year of HRT in all 23 women who completed the l-year study period. Sera were stored at -80°C until the end of the study
Metabolism,
Vol43,
No 7
(July), 1994:
pp 799-802
and then assayed. Storage under these conditions has been shown not to influence the LDL subfraction profile.<-” LDL subfractions were separated by GGE as described previously.” using 2% to 16% polyacrylamide gradient gels (PAA ?/I6 gels, Pharmacia, Uppsala, Sweden). Serum aliquots were mixed with glycerol and bromphenol blue. Gels were pre-electrophoresed for 20 minutes at 70 V. and afterwards we added IO ~,LLof the prepared samples to each lane of the gel, and electrophorcsis was continued for another 30 minutes at 70 V, followed for 5.5 hours at 400 V and 8°C. After eiectrophoresis. the gels were fixated in ethanol and the lipoprotein bands were stained with Sudan Black B (Paragon Lipostain, Beckman Instruments, Palo Alto, CA) followed by destaining in ethanol. Samples of the same subject were assayed on the same gel and in the two nearest lanes. A serum standard was prepared by pooling sera to cover the complete spectrum of subfractions. In each gel. this serum was applied in two lanes to serve as a reference for the LDL bands in the samples. The pooled serum consisted of scra of six persons with a predominant LDL, (d = I.030 to 1.033 kg/L), LDL: (d = 1.033 to 1.04(1kg/L). or LDLi (d = I.040 to 1.045 kg/L) subfraction.& Gels were scanned in triplicate with the LKB 2207 Ultrascan XL enhanced laser densitometer ~Pharmacia~L~~. Uppsala, Swe-
From the Departments ncrl Medicine,
Sint Radboud,
Nijmegen,
Submitted
September
Supposed
by Sohq
Address Obstetrks
of Obstetrics
and Medical
Sk&tics,
& Gvnecology UtGerzitv
General
ffo.vpital
inter-
Nlimegen
The Netherlands. 7, I Y9.?; axepted
Jumrary
1. I Y&.
Dtcphar B. P:
re~~rlt requests to Rune Ro~lilnd, & ~~?~~~colo~* ~~~~rtrne~t
of
MD.
Phf).
Projkssor
in
Uh.~tet~c~ & ~~~~t,c~~lo~,
Uniwnity Hospital Nijmegert Sint Radbotrtl. PO Ik>w YlOl, 6500 HB Nijmegen, The Netherlunds. Copytight c 1994 by W. B. Saunders Cornpttfz) 00260495/9414307-0001$03.00!0 799
800
VAN DER MOOREN ET AL
0.35 mmol/L (P < .OOl), while serum triglyceride (TG) levels increased slightly but significantly (P < .05) from 1.27 t 0.69 to 1.45 t 0.73 mmol/L. Apolipoprotein (apo) A-I increased from 1,417 ? 191 to 1,672 -t 195 mg/L (P < .OOl), while apo B showed a slight nonsignificant decrease from 1,589 +- 434 to 1,554 ? 398 mg/L. Lipoprotein(a) decreased from 303 ? 467 to 239 ? 382 mg/L (P < ,001; all values are the mean ? SD).
den). The scans were analyzed by calculating the change in relative peak height after therapy to determine changes in the relative contribution of each LDL subfraction to total LDL. Furthermore, we evaluated the change in the migration distance and relative peak height of the predominant LDL band. StatisticalAnalysis All changes were tested using the Wilcoxon signed-rank test. Stepwise multiple regression analysis was used to examine significant contributions of serum lipids and (apo)lipoproteins to the variation in the LDL subfraction profile. By using the test on the intercept, we examined whether HRT had a significant additional influence on the LDL subfraction profile. Statistical analyses were performed with the Statistical Analysis System software package (SAS Institute. Gary. NC).
LDL Subfraction Profile
RESULTS
Lipids and Lipoproteins
During hormone therapy, mean LDL cholesterol decreased from 4.47 ? 1.52 to 3.84 ? 1.02 mmol/L (P < .OOl), parallel to the decrease in total cholesterol from 6.32 ? 1.46 to 5.91 2 1.00 mmol/L (P < .Ol). High-density lipoprotein (HDL) cholesterol increased from 1.49 ? 0.28 to 1.68 t
B
B
P
GGE of the individual sera revealed discrete LDL subfraction bands, which were identified as LDL,, LDL2, and LDL3, with particle sizes averaging 267,257, and 251 A, respectively (data originating from Austin et a15; Fig 1). This predominance of three LDL subfractions was confirmed by densitometric scanning of the gels. Each lane was scanned in triplicate; within all triplicates, the number of peaks and the migration distance of each peak agreed well, and the shapes of the densitometric curves were similar. Before treatment, the mean percentage contributions of each LDL band to total LDL, calculated as mean relative peak heights, were as follows: LDL,, 36% (SD 22%, median 34%, Q, 19%, Q3 55%); LDL2,44% (SD 18%. median 41%,
\\ \\ \
\
\ ILA
A
i #’
r’ \\
,’ \\
,
,’
‘> -# -1
-2
,, *’
,’
‘\
\\ ‘\
.
-3
Fig 1. Effect of HRT on the LDL subfraction profile of one subject as observed by GGE and densitometric scanning of the gel. During HAT, the size of the LDL shifts toward a predominance of smaller LDL particles. Top: Upper part (40%) of one gel showing the VLDL and LDL range, covering the molecular size between 180 and 800 A. Bottom: Midd/e: lanes of one subject (B) before and (A) after HRT, and of the pool serum (P), as shown in the frame (top). Leff: densitometric scan of the gel before HRT. Right: densitometric scan of the gel after HRT. 1, LDL, (267 A); 2, LDLz (257 A); 3, LDLI (251 A); lLp(a)-band. The values of molecular size distribution of the main subfractions originate from Austin et al.5
LDL SUBFRACTION
801
PROFILE AND HRT
18%, Q 55%); LDLX, 20% (SD 15%, median 15%, Qr it%, Q1 23%); and the predominant peak, 59% (SD 9%, median 58?h, Q, Sl%, Qj 63%). In 21 individuals (91%), the predominant peak was the LDLr or LDLl peak. Furthermore, analysis of data obtained by densitometric scanning revealed that HRT was associated with, on average, a small decrease in the mean relative heights of the LDL, (mean -4%, SD 16%, median -4%, Qr -14%, Qx + 1 Z; P = .1.54)and LDLz (mean -3%, SD 18%, median + 1 Z, Q, -22%, Qi + 10%; P = .930) peaks in combination with a considerable increase in the relative height of the I.DL3 peak (mean +7%, SD 12%, median +5%, Qr +l%, Q3 +7%; P = .O~)OS~.So the relative contribution of LDLi tc, total LDL increased, indicating a more dense subfraction profile. The mean relative height of the predominant peak decreased significantly (mean -50/c, SD 14%, median --5 %, Q, - 14% Qs 0%; P = .02.5), and its relative migration distance increased (mean t0.21 U. SD 0.58 U, median -0.13 U, Q, -0.20 U. Qs +0.70 U; P = .131), reflecting a change toward a smaller LDL particle size. Thus, during HRT the average size of LDL particles decreases. Stepwisc regression analysis on all lipids and (apo)lipoprc teins described here showed that only total serum apo B lcv~ls (as determined by ncphelometry) correlated significantly with the rrlative height of the LDLs peak (P < .05). An estimate of the additional effect of the given HRT on t hi: relative height of the LDL.1 peak, as found by the test on the intercept, is 0.08 (SE = 0.02; P < .Ol). So HRT had a significant additional influence on the relative height of the LDL; peak. Q,
DISCUSSION
Earlier, we reported on the changes in serum lipids and ( apo)lipoproteins during sequential 17@-estradiol-dydrogesterone therapy in postmenopausal women. With regard to the risk of CHD, these changes can be considered favorab e, with the exception of the small increase in serum TG.’ LDL cholesterol decreased while total serum apo B, which in these normotriglyceridemic women is largely present in LDL. remained almost similar. The observed changes in the LDL cholesterol to apo B ratio indicate a shift toward smaller LDL partic1es.j This could indeed be confirmed by GGE, a method that has proved to be valid for detecting LDL heterogeneity.~ Recently, Campos et al, who already demonstrated smaller LDL particles in postmenopausal wc,men,” also reported decreased LDL cholesterol and increased HDL cholesterol, apo A-I, and TG levels in postmenopausal women treated with conjugated estrogen monotherapy. In women with predominantly large LDL particles ( > 272 A), they found a significant decrease in the LDL peak area and in the particle diameter, whereas in women with predominantly small LDL particles (261 to 272 4’ these parameters did not change.“’
A possible explanation for the decreasing LDL particle size during HRT is given by Deckelbaum et al,” who hypothesized that the size and density of LDL are at least in part determined by exchange of TG from very-low-density lipoprotein (VLDL) for cholesteryl esters from LDL. Since TG levels increase during estrogen supplementation, due to increased VLDL synthesis, the hypothesized nlechanism explains how LDL particles become enriched by an excess amount of TG at the expense of cholesteryl esters, and then allow continued particle size reduction through lipase action, which may finaily result in lipid-poor and thus protein-enriched LDL particles of relatively high density. A predominance of smafl LDL particles has earlier been identified in a lipoprotein pattern with increased serum TG and decreased HDL cholesterol levels.” In previous studies comparing LDL particle sizes of patients with CHD and controls,jmh the LDL particle size difference between these two groups was reduced to nonsignificant after adjusting for TG and HDL cholesterol levels. This indicates that small LDL particles may not independently be associated with CHD. The reported association with low HDL cholesterol and increased TG suggests that the LDL. size reflects a series of alterations that is deviating in CHD. Indeed. LDL of different size, induced by lipid-exchange reactions, differ in their susceptibiI~ty to oxidative mc~di~catio1~in vitro.” a phenomenon considered to underlie the cxccssive accumulation of cholesterol in atherosclerotic lesions.‘” HRT, including the hormone treatment used in the present study. is associated with beneficial changes in the lipid and lipoprotein parameters predicting the risk for CHD.’ The favorable effect of HRT c(~ncerning the risk for atherosclerosis is supported by animal cxperimcnts in which it was shown that LDL cholesterol accumulation in the arterial wall is attenuated by estrogen monotherapy and combined estrogen-progestogcn therapy, even when plasma lipids, lipoproteins, and apoprotcins remain unchanged.‘” As a matter of fact, in these animal studies a decrease in LDL molecular weight was also obscrvcd.rq In summary, as a result of the diversity of intluenccs on women’s metabolism exerted by HRT, it remains difficult to make a proper CHD risk estimation based on the common lipid parameters. From earlier reports on changes in LDL density and LDL particle size during HRT in e.xperimcntai and clinical studies, which can be confirmed hy our own observations. it appears that a shift in LDLparticle distribution to a smaller size may not per se be positively correlated with cardiovascular disease. Clearly, thcrc is a need for more prospective research on the cffccts of HRT in the protection against cardio~scular discasc. ACKNOWLEDGMENT We thank Heidi Hak-Lemmers (Laboratory of General internal
Medicine) for excellent technical assistance.
REFERENCES
Stampfer MJ, Cold&z GA, Willett WC, et al: Postmenopausal estrogen therapy and cardiovascular disease. Ten-year follow-up i.
from the Nurses’ Health Study. N Engl J Med 325:756-762, 1991 1. Bush ‘I’L. Fried LP, Barrett-Connor E: Cholesterol, lipoproteins. and coronary heart disease in women. Clin Chem 34:B60BSO. 1988
3. MeNamara JR. Campos H. Ordovas JM. et al: Effect of gender, age. and lipid status on low density lipoprotein subfraction distribution. Arteriosclerosis 7:483-490. 1987 3. Crouse JR, Parks JS, Schey HM, et al: Studies of low density lipoprotein molecular weight in human beings with coronary artery disease. J Lipid Res 26566-574. 1985
802
5. Austin MA, Breslow JL, Hennekens CH, et al: Low-density lipoprotein subclass pattern and risk of myocardial infarction. JAMA 260:1917-1921,1988 6. Campos H, Genest JJ Jr, Blijlevens E, et al: Low density lipoprotein particle size and coronary artery disease. Arterioscler Thromb 12:187-195,1992 7. Van der Mooren MJ, Demacker PNM, Thomas CMG, et al: A 2-year study on the beneficial effects of 17B-oestradioldydrogesterone therapy on serum lipoproteins and Lp(a) in postmenopausal women: No additional unfavourable effects of dydrogesterone. Em J Obstet Gynecol Reprod Biol52:117-123, 1993 8. Dormans TPJ, Swinkels DW, de Graaf J, et al: Single-spin density-gradient ultracentrifugation vs gradient gel electrophoresis: Two methods for detecting low-density-lipoprotein heterogeneity compared. Clin Chem 37853-858, 1991 9. Campos H, McNamara JR, Wilson PWF, et al: Differences in low density lipoprotein subfractions and apolipoproteins in premenopausal and postmenopausal women. J Clin Endocrinol Metab 67:30-35, 1988
VAN DER MOOREN ET AL
10. Campos H. Sacks FM, Walsh BW. et al: Differential effects of estrogen on low-density lipoprotein subclasses in healthy postmenopausal women. Metabolism 42:1153-l 158, 1993 11. Deckelbaum RJ, Granot E, Oschry Y. et al: Plasma triglyceride determines structure-composition in low and high density lipoproteins. Arteriosclerosis 4:225-231,1984 12. De Graaf J. Hak-Lcmmers HLM, Hectors MPC, et al: Enhanced susceptibility to in vitro oxidation of the dense low density lipoprotein subfraction in healthy subjects. Arterioscler Thromb 11:298-306,199l 13. Steinberg D, Parthasarathy S, Carew TE, et al: Beyond cholesterol. Modifications of low-density lipoprotein that increase its atherogenicity. N Engl J Med 320:915-924, 1989 14. Wagner JD, Clarkson TB, St Clair RW, et al: Estrogen and progesterone replacement therapy reduces low density lipoprotein accumulation in the coronary arteries of surgically postmenopausal cynomolgus monkeys. J Clin Invest 88:1995-2002, 1991