Changes in lipoprotein composition during the menstrual cycle

Changes in Lipoprotein Composition the Menstrual Cycle

During

Hak-Joong Kim and Ronald K. Kalkhoff Composition of major plasma lipoproteins was studied in 14 normal women during different phases of the menstrual cycle for three consecutive months. The results were compared to measurements in ten normal age-matched men for a comparable period, to delineate possible sex differences in lipoprotein metabolism in young adults. Blood samples were obtained every 3-5 days after a 14-hr overnight fast and processed for determinations of total plasma cholesterol, LDL- and HDL-cholesterol, and apoproteins B and A-l. In premenopausal women, a significant, 10 % -25 % cyclical suppression of total plasma cholesterol, LDL-Chol, and LDL-apoB occurred during the luteal phase, which was significantly lower than unchanging concentrations found in men at any time interval. HDL-Chol remained in a significantly higher fixed concentration range in the female subjects as compared to the men. These sex differences in lipoprotein metabolism may have relevance to the reduced susceptibility of premenopausal women to atherosclerosis.

HIS ORIGINAL description of angina INpectoris in 1886, Heberden’ observed that the condition was much less frequent in women. Since that time, several epidemiologic studies have reported a sex difference in the incidence of cardiovascular disease in men versus premenopausal women ranging from 3/ 1 to 10/l .‘*’ These sex advantages are lost in oophorectomized or postmenopausal women4 suggesting that ovarian secretion of sex steroids may relate to this protective mechanism. Disturbances of lipid and lipoprotein metabolism have been implicated in atherogenesis.’ Moreover, normal, young women differ from age-matched men in composition of their plasma lipoproteins. Total cholesterol (Chol) is lower, cholesterol in the low density lipoprotein fraction (LDL) is reduced,’ and levels of highly density lipoprotein (HDL) and its cholesterol content are increased relative to men.* Many of these differences also disappear as women reach postmenopausal status.’ In the present study, plasma lipids, lipoproteins, and carrier apoproteins were measured throughout the menstrual cycle in healthy women. The data were compared to similar profiles obtained in age-matched men. The results demonstrate a pattern in menstruating

Mersfm/ism, Vol. 28, No. 6 (June), 1979

women that is cyclical and naturally protective against atherogenesis as compared to men. MATERIALS

AND

METHODS

The 14 women weighed 57.2 & 2.2 kg and were 33 + 2 yr old. The 10 men weighed 76 + 2.8 kg and were 30 + 2 yr old. All subjects were within 10% of their ideal body weight as defined by Metropolitan Life Insurance Tables (1959). The volunteers had negative family histories for diabetes mellitus and for unusual cardiovascular disease. None were diabetic, nor had they received any medications for 3 mo prior to the study. All women had regular menstrual periods for the three preceding months. The subjects were allowed their customary diets, but alcoholic beverages were withheld throughout the investigation. The I4 women had different menstrual cycle lengths, ranging from 21 to 37 days. Days within each cycle that bracketed the menstrual, follicular, ovulatory, and luteal phases were estimated, knowing total cycle length and basal body temperature changes. This is based on a large number of case studies of normal women previously reported by Adlercreutz and Tallqvist.9 Each woman was followed for three consecutive cycles. Blood samples were obtained every 3-5 days during this period, and results were grouped in one of the four phases of the cycle and averaged. For purposes of comparison, men were arbitrarily assigned 28 day cycles for two months. Blood obtained every 3-5 days were grouped in the first, second, third, and fourth week of these two hypothetic cycles. and the data for each week was also averaged. Blood samples were obtained from each subject after an overnight 14-hr fast and collected in chilled tubes containing 1 mg EDTA/ml blood. Plasma was separated immediately, and stored at 4OC until further fractionated into very low density lipoprotein (VLDL), LDL, and HDL within 7 days. After total plasma triglyceride (TG) and Chol concentrations were measured, cholesterol content of various lipopro-

Endocrine Metabolic Section, Department of Medicine, Clinical Research Center. Medical College of Wisconsin, and Milwaukee County Medical Complex, Milwaukee, Wis. Supported by grantsfrom TOPS Club, Inc.. Obesity and Metabolic Research Program, the Wisconsin Heart Association, and by USPHS General Clinical Research Grant 5-MOl-RROOOX Presented in part at the 50th Annual Meeting of the Central Society for Clinical Research, Chicago, Ill. November 4. 1977. Address reprint requests to Dr. Hak-Joong Kim, Medical College of Wisconsin. 8700 W. Wisconsin Avenue, Milwaukee, Wis. 53226. @ 1979 by Grune & Stratton. Inc. 0026-0495/79/2806~010$0I.00/0

663

KIM

teins (VLDL-Chol, LDL-Chol, HDL-Chol) was determined after isolation of these fractions by a combination of ultracentrifugation and precipitation. VLDL-Chol was obtained by measuring cholesterol content of the supernatant fraction after ultracentrifugation (d - 1.006 g/ml) for 18 hr at 100,000 g (Beckman L3-50) (LDL-Chol + HDLin a Spinco #sOTi rotor. ” Cholesterol Chol) and apolipoprotein B (apoB) were assayed from the infranatant. LDL was then precipitated from d = 1.006 infranatant fraction by heparin + MnCI,,” and cholesterol content of the supernatant fraction (HDL-Chol) was measured. LDL-Chol was calculated indirectly by subtracting LDL-Chol from the sum of cholesterol content of the d = 1.006 infranatant fraction (LDL-Chol + HDL-Chol). TriglycerideI and cholesterol” were determined by standard methods published elsewhere. Our method for measuring cholesterol” tends to run 4% or more higher than the Abell-Kendall method, as pointed out in Tonks’ review.14 ApoB content of total plasma and d = 1.006 g/ml infranatant fractions were measured according to the method of Sniderman et al.,” using antibody against either LDL or apoB and purified LDL as standard. There was a good correlation between the results obtained using antibody against LDL and against apoB (r = 0.99). Intra- and interassay coefficients of variations were 2% and 5%, respectively, in 26 assays. Apolipoprotein A- 1 (apoA- 1) of plasma and d = 1.006 g/ml infranate were measured by doubleantibody immunoassay of Schonfeld and PReger16 as modified by Fainaru, et al.,” using purified apolA-1 as standard. Intra- and interassay coefficient of variations were 5% and 15%. respectively, in 6 assays. Data obtained from women were compared to corresponding data from men using Student’s t test for unpaired data.‘* Changes from baseline values of the menstrual period in women and the first week in men were compared to other intervals within the same group by employing Student’s t test for paired data. In all instances, degrees of freedom were calculated on the basis of N = 14 for women and N for men. RESULTS

Total plasma triglyceride in women rose to highest concentrations at midcycle and fell during the luteal phase (Fig. 1). These changes, compared to values during the menstrual phase, were not significant. Peak TG levels in women were significantly higher than mean levels found in men at any time interval (p < 0.05). During the late luteal phase, women exhibited a significant fall in plasma cholesterol as compared to mean values obtained during the menstrual and early follicular phases (p < 0.01). Moreover, these two lowest points were also significantly lower than any mean values of the men during each week of the four week cycles 0, < 0.05). LDL-Chol in women, like total plasma cholesterol, was significantly depressed during the luteal phase as compared to the menstrual phase

Women (14)

130

130

AND

Men

KALKOFF

( 10)

1

i--t-i--1 .. l.

I

,

MENS FOLLIC' 0 ' LUTEAL

..

'

? MENS'FOLLIC' 0 ' LUTEAL

Phase of Cycle

Weeks

Fig. 1. Total plasma triglyceride and cholesterol concentrations during the menstrual cycle in 14 women and during a comparable period in 10 men. Each point for women represents en average of up to 52 determinations obtained during three separate cycles: for men, each point is an average of 20 determinations obteined during two arbitrary 4-wk cycles. Vertical bars denote SE of the meen. (‘1 Significance of the diierence between this mean and the highest values observed in the seme group, p < 0.06. (“1 Significance of the difference between this mean value in men end the lowest value in women, p < 0.050.01.

(p < 0.05, Fig. 2). All LDL-Chol values were also significantly below all mean values recorded for male subjects (p < 0.05-0.01). Although HDL-Chol tended to rise in women during the luteal phase, the increment above other phases was not significant (Fig. 2). However, the HDLChol in women was higher than in men at all times during the menstrual cycle (p < 0.050.01, Fig. 2). ApoB, the carrier protein in the LDL fraction, followed the same cyclical pattern that was demonstrated for LDL-Chol, i.e., a significant fall occurred in the luteal phase. Again, all concentrations were below those found in men (Fig. 3). There were no major fluctuations in apoA- 1, a major protein component of HDL, in either sex (Fig. 3). Although values were higher in women, the difference from men was not significant. Sex differences in lipoprotein composition were also expressed as the ratios of HDLChol/LDL-Chol and apoA- 1/apoB. In women, both ratios paralleled one another and reached a peak during the luteal phase, which significantly

LIPOPROTEINS

665

DURING

Women

(14)

Women

90

Men

(14)

(10)

60

1MENSI FOLLIC Phase

of Cycle

Phase

Weeks

Fig. 2. Plasma LOL-cholesterol and HOL-cholesterol during the menstrual cycle in women and during a comparable period in men. See legend for Fig. 1 for further explanations. (‘1 Significance of difference between these mean values and the highest mean in the menstrual phase, p < 0.05. (“1 Significance of difference between the mean values of men from each of the mean values in women, p < 0.06-0.01.

exceeded values during menstruation (p < 0.05, Fig. 4). These peak elevations were also greater than each of the four weekly ratios obtained in male subjects (p < 0.05).

I

0

I

I

LUTEAL

I

1

’

I

I

2

3

4

Weeks

of Cycle

Fig. 3. Plasma LDL-apog and total apoA-1 concentrations in women during their menstrual cycles and during comparable periods in men. (‘1 Significance of difference between mean values and the highest value during the menstrual phase in women, p -c 0.06. (“1 Significance of difference between mean values of men and each of the means of women, p < 0.05.

correlation between plasma LDL concentrations and total body cholesterol pool size exists.22 According to Goldstein and Brown2’ a major portion of LDL-Chol is catabolized in peripheral tissues following interaction with specific receptors. Thus, a heightened concentration of this

DISCUSSION

In the human female, total plasma cholesterol concentrations contrast with values in the male throughout a normal life span. However, the differences between sexes appear to pass through three different phases. Female cholesterol levels exceed those of the male between birth and puberty, although these differences are quite small.‘9*20During the adult reproductive period, the reverse is true.&* In the postmenopausal state, total plasma cholesterol concentrations in women are significantly higher than in men.’ It is of interest that LDL-Chol, like the total plasma cholesterol, follows the same pattern of difference between the sexes, whereas HDLChol is higher in the female, regardless of age.7*8*‘9 Recent studies suggest that apoB, the principal carrier protein of cholesterol in the LDL franction, functions primarily in the transport of cholesterol to peripheral tissues.2’ Thus, a direct

Women

3

UC j

Men (10)

*

2

Fm

(141

8 1 I@@--@-

OZ

o] ,

Moys FOLLIC’ 0 ‘LUTEAL Phase

of Cycle

’

0 I--

6-&--t+ ** 1

** 2

3

** 4

Weeks

Plasma HOL-cholesterol/LOL-cholesterol, and Fig. 4. total apoA-1 /LOL-epo6 ratios in women during their menstrual cycles and during a compareble period in men. (*) Significance of dfference between this mean value end the lowest mean during the men8trual phase in women, p < 0.06. (.*I Significance of difference between meen values of men and the highest mean of women, p < 0.06.

666

fraction in plasma may reflect diminished tissue catabolism, in part, and favor the acceleration of atherosclerosis. Consistent with this hypothesis is the finding of a higher incidence of atherosclerotic cardiovascular disease in subjects with elevated circulating levels of LDL-Chol, including familial type II (Fredrickson) hypercholesterolemia, nephrosis, and hypothyroidism.6~23~24 Conversely, HDL, with its major carrier protein apoA-1, may accentuate removal of cholesterol from tissues25 and promote its transport to the liver for ultimate catabolism. Moreover, plasma HDL-Chol levels inversely correlate with total cholesterol pool size.22 A predominance of this lipoprotein, then, could cause resistance to atherogenesis. Although the means by which this occurs is poorly understood, it may relate, in part, to the known activating effect of apoA-1 on lecithin-cholesterol acyl transferase (LCAT), an enzyme closely linked to tissue cholesterol removal from tissues.26 A recent study also suggest that HDL-Chol may be the principal source of biliary cholesterol, the excretion of which represents another key process for removal of this lipid.27 Decreased HDL concentrations in middle-aged and elderly persons are associated with a higher incidence of cardiovascular disease,28 and subjects with diabetes, hyperlipidemia, uremia, and obesity who exhibit low levels of circulating levels of HDL, develop atherosclerosis with greater frequency.29s30 Conversely, relatively higher plasma HDL concentrations observed in premenopausal women, certain ethnic groups, and those exhibiting the lonevity syndrome relate statistically to a lower incidence of atherosclerosis 31.33 Previous studies also have reported fluctuations of total serum cholesterol and lipoprotein subfractions during the menstrual cycle.9”4V35 However, the present study also shows a significant cyclical change in plasma LDL-Chol as well as apoB in the normal menstruating woman. During the luteal phase, these concentrations were significantly lower than levels observed in the menstrual phase of the same individuals and were depressed well below relatively fixed and higher concentrations of these three moieties in men. There are certain differences between results obtained in earlier work and the present investi-

KIM AND

KALKOFF

gation that deserve comment. Adlercreutz and Tallqvist’ observed a secondary rise of total plasma cholesterol in the late luteal phase. However, analysis of the raw data revealed that this change was not significant with respect to the levels measured at other intervals, including the postovulatory phase (period 6). In the study of Barclay et al.,3s total plasma LDL, as opposed to our more specific measurements of apoB and LDL-Chol, was shown to fall slightly between midcycle and the luteal phase, but the change was minimal and not significant. Oliver and Boyd34 observed depression of total plasma cholesterol and cholesterol esters at midcycle and early luteal phases with a subsequent secondary rise. However this study was conducted on women in the fed state, and comparison of the results to our own studies of fasted women is difficult. It is not certain which hormaonal changes that occur during the mentrual cycle are responsible for the fluctuations observed in the present investigation. The fall of total plasma cholesterol and plasma LDLChol as well as plasma apoB during the luteal phase follows the peak elevation of plasma 17P-estradiol, which occurs just prior to ovulation. That this estrogen has a substantial suppressive action on cholesterol is supported by results of previous studies of estrogen effects on this plasma lipid in ovariectomized or postmenopausal women and in middle-aged or elderly men. Since the major LDL-uptake site occurs primarily in a splanchnic bed (presumably liver)36 and estrogen administration increases this process,” it may be assumed that effects of this sex steroid are expressed to a great extent via actions on hepatic degradation of this lipoprotein. The delay between the achievement of peak plasma estradiol concentrations and the fall in cholesterol and its carrier protein may reflect a relatively long plasma half-life of LDL-apoB of approximately 4 days.38 However, we cannot exclude the possibility that other hormonal influences, such as cyclical secretion of androgens or progesterone-like steroids by the ovary,39 may alter cholesterol lipoprotein metabolism further. In contrast to the fluctuating levels of plasma cholesterol, LDL-Chol, and apoB in the menstruating women, higher, fixed concentrations of HDL-Chol and apoA-1 were observed in

LIPOPROTEINS DURING

MENSTRUAL

CYCLE

667

this group. Moreover, HDL-Chol values significantly exceeded those of the male subjects. However, the report from another laboratory utilizing the analytic ultracentrifuge suggests that certain subfractions of HDL (HDL,) may also increase significantly at midcycle. It is also possible that minor variations in HDL-Chol may remain undetected with our methods because of the long plasma half-life of apoA-I, which is approximately 4-6 days.40*4’Nevertheless, the present study documents the marked difference between premenopausal women and agematched men with respect to parameters measured in our laboratory. In addition, the differences between sexes are equally striking for the concentration ratios of HDL-Chol/LDLChol and apoA-1 /apoB. It is not known why HDL-Chol and the corresponding carrier protein apoA-1 are increased in young adult women. That estrogen may have an important role is suggested by previous studies in which administration of this hormone to both

pre- and postmenopausal women increased the concentration of these specific lipoproteins and apoproteins.33*42 During normal reproductive life, women appear to have at least two natural advantages over men with respect to cholesterol-related atherogenesis. The persistently high plasma HDL-Chol concentrations may augment the removal of this lipid; the significant suppression of LDL-Chol and apoB during the luteal phase of the menstrual cycle may represent decreased availability of an atherogenic lipoprotein to peripheral tissues. Whether this represents an estrogen action alone or in combination with other hormones or factors remains to be determined. ACKNOWLEDGMENT The authors wish to express their thanks to all volunteers; the nursing staff of the Clinical Research Center for their help; and to Patricia Ryan and Indira Kurup for their technical assistance in this study.

REFERENCES 1. Heberden W: Commentaries on the History and Cure of Diseases. New York, Hafner, 1962, pp 362-369 2. Levy H, Boas EP: Coronary artery disease in women. JAMA 107:97-102.1936 3. Kannel WB, Dawber TR, Kagan A, et al: Factors of risk in the development of coronary heart diseaseSix year follow-up experience. The Framingham study. Ann Intern Med 5533-50, 1961 4. Kannel WB, Hjortland MC, McNamara PM, et al: The Framingham study. Ann Intern Med 85:477-452,1976 5. Gofman JW, Young W, Tandy R: lschemic heart disease, atherosclerosis and longevity. Circulation 34:679697,1966 6. Adler&erg D, Schaefer LE, Steinberg AG, et al: Age, sex, serum lipids, and coronary atherosclerosis. JAMA 162:619622,1956 7. Carbon LA, Ericsson M: Quanitative and qualitative serum lipoprotein analysis. Part I. Studies in healthy men and women. Atherosclerosis 21:417-433, 1975 8. Russ EM, Eder HA, Barr DP: Protein-lipid relationships in human plasma. Part I. In normal individuals. Am J Med 1 I :468479,195 1 9. Adlercreutz H, Tallqvist G: Variations in the serum total cholesterol and hematocrit values in normal women during the menstrual cycle. Stand J Clin Lab Invest 1 l:l-9, 1959 10. Have1 RJ, Eder HA, Bragdon JH: The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest 34:1345-1353, 1955 11. Friedewalde WT, Levy RI, Fredrickson DS: Estimation of the concentration of low-density lipoprotein choles-

terol in plasma, without use of the preparative ultracentrifuge. Clin Chem 18:499-502, 1972 12. Van Handel E, Zilversmit DB: Micromethod for direct determination of serum triglycerides. J Lab Clin Med 50:152-157, 1957 13. Leffler HH, McDougald CH: Estimation of cholesterol in serum by means of improved technics. Tech Bull Regist Med Technol33:19-23, 1963 14. Tonks DB: The estimation of cholesterol in serum: A classification and critical review of methods. Clin B&hem 1:12-29, 1967 15. Sniderman A, Teng B, Jerry M: Determination of B protein of low density lipoprotein directly in plasma. J Lipid Res 16:465467, 1975 16. Schonfeld G, Pfleger B: The structure of human high density lipoprotein and the levels of apolipoprotein A-1 in plasma as determined by radioimmunoassy. J Clin Invest 54:236-246,1974 17. Fainaru M, Glengeaud MC, Eisenberg S: Radioimmunoassay of human high density lipoprotein apoprotein A-l. Biochim Biophys Acta 386:432443, 1975 18. Snedecor GW, Cochran WG: Statistical Methods (ed 6), chapter 4. Ames, Iowa State University Press, 1967, pp91-119 19. Carlson LA, Hardell LI: Sex differences in serum lipids and lipoproteins at birth. Eur J Clin Invest 7: 133-l 35, 1977 20. Morrison JA, DeGroot I, Edwards BK, et al: Plasma cholesterol and triglyceride levels in 6775 school children, ages 6-17. Metabolism 26:1199-1211,1977 21. Goldstein

JL, Brown

MS; Atherosclerosis:

The low

KIM

density lipoprotein receptor hypothesis. Metabolism 26: 1257-1275.1977 22. Miller NE, Nestel PJ, Clifton-Bligh P: Relationships between plasma lipoprotein cholesterol concentrations and the pool size and metabolism of cholesterol in man. Atherosclerosis 23~535-547, 1976 23. Steinberg AD: Myxedema and coronary artery disease-A comparative autopsy study. Ann Intern Med 68:338-344, 1968 24. Berlyne GM, Mallick NP: lschaemic heart disease as a complication of nephrotic syndrome. Lancet 2399-400, 1969 25. Stein Y, Glangeaud MC, Fainaru M, et al: The removal of cholesterol from aortic smooth muscle cells in culture and Landschutz ascites cells by franctions of human high-density apolipoprotein. Biochim Biophys Acta 380:106118,1975 26. Glomset JA, Norum KR: The metabolic role of lecithin: Cholesterol acyltransferase: Perspectives from pathology. Adv Lipid Res 1l:l-65, 1973 27. Schwargz CC, Halloran LG, Vlahcevic ZR, et al: Preferential utilization of free cholesterol from high density lipoproteins for biliary cholesterol secretion in man. Science 200:62-64,1978 28. Gordon T, Castelli SP, Hjortland MC, et al: High density lipoprotein as a protective factor against coronary heart disease: The Framingham study. Am J Med 62:707714.1977 29. Bagdade JD, Albers JJ: Plasma high density lipoproteins in chronic hemodialysis and renal-transplant patients. N Engl J Med 296:14361439,1977 30. Gordon T, Castelli WP, Hjortland MC, et al: Diabetes, blood lipids and the role of obesity in coronary heart disease risk for women. Ann Intern Med 87:393-397, 1977 3 1. Rhoads GG, Gulbrandsen C, Kagan A: Serum lipo-

AND

KALKOFF

proteins and coronary heart disease in a population study of Hawaii-Japanese men. N Engl J Med 294:293-298,1976 32. Glucck CJ, Fallat RW, Millett F, et al: Familial hyper-alpha-lipoproteinaemia: Studies in eighteen kindreds. Metabolism 24: 1243-l 265, 1975 33. Barr DP, Russ EM, Eder HA: Influence of estrogens on lipoproteins in atherosclerosis. Trans Assoc Am Physicians 65:102-l 13, 1952 34. Oliver MF, Boyd GS: Changes in the plasma lipids during the menstrual cycle. Clin Sci l2:217-222, 1953 35. Barclay M, Barclay RK, Skipski VP, et al: Fluctuations in human lipoproteins during the normal menstrual cycle. Biochem J 96:205-209,1965 36. Sniderman A, Thomas D, Marpole D, et al: Low density lipoprotein: A metabolic pathway for return of cholesterol to the splanchnic bed. J Clin Invest 61:867-873, 1978 37. Hay RV, Pottenger LA, Reingold AL, et al: Degradation of ‘*‘I-labelled serum low density lipoprotein in normal and estrogen-treated male rats. Biochem Biophys Res Commun44:1471-1477.1971 38. Langer T, Strober W, Levy RI: The metabolism of low density lipoprotein in familial type. II hyperlipoproteinemia. J Clin Invest 51:1528-1536, 1972 39. Abraham GE, Chakmakjian ZH: Serum steroid levels during the menstrual cycle in a bilaterally adrenalectomized woman. J Clin Endocrinol Metab 37:581-587, 1973 40. Blum CB, Levy RI, Eisenbcrg S, et al: High density lipoprotein metabolism in man. J Clin Invest 60:795-807, 1977 41. Shepherd J, Patsch JR, Packard CJ, et al: Dynamic properties of human high density lipoprotein apoproteins. J Lipid Res 19:383-389, 1978 42. Albers JJ, Wahl PW, Cabana VG, et al: Quantitation of apolipoprotein A-l of human plasma high density lipoprotein. Metabolism 25:633-644, 1976