ORIGINAL ARTICLES

Oral Contraceptive Use and Fasting Triglyceride,

Plasma Cholesterol and HDL Cholesterol
CHARLES H. HENNEKENS, M.D., DENIs A. EVANS, M.D., WILLIAM P. CASTELI-I, M.D.,
JAMES 0. TAYLOR, M.D., BERNARD ROSNER, PH.D., AND EDWARD H. KASS, M.D. PH.D.

SUMMARY Fasting plasma triglyceride, plasma cholesterol and high-density lipoprotein (HDL)
cholesterol levels were studied for 190 white women, ages 21-39 years, who were classified according to their
oral contraceptive (OC) usage patterns at two community surveys, 3 years apart. The mean level of fasting
triglyceride was higher among current OC users (95 mg/100 ml) than among nonusers (73 mg/100 ml)
(p = 0.002). After adjustment for the possible confounding effects of age, weight, current cigarette smoking
and fasting glucose level, current OC users still had a mean plasma triglyceride level 19 mg/100 ml higher than
that of nonusers (p = 0.007). Current OC users also appeared to have somewhat higher levels of total
cholesterol which were of borderline significance in crude and adjusted analyses. There was a nonsignificant in-
verse relationship of OC use with HDL cholesterol levels. Past use did not affect these results, indicating that
the OC-induced lipid changes were reversible.

A THREE- TO-FIVEFOLD INCREASE in risk of Subjects and Methods

myocardial infarction (MI) among current oral con-
traceptive (OC) users has been reported from two
British case-control studies, one of deaths,1 and the
other of survivors.2 This finding has also recently been
reported from a case-control study of women in the
United States.3 The results of these retrospective
studies are consistent with correlational data of OC
use and mortality trends among women from 21 coun-
tries.4
Initially, clinical5 and epidemiologic6 reports in-
dicated that OC use promoted venous and arterial
thromboembolism, suggesting an alteration in the
coagulation mechanism which might account for the
increased risk of MI among OC users. Recently,
however, it has been suggested that the effect of OC
use on risk of MI may be related to alterations in
levels of major coronary risk factors.7
In this report we examine the relations of current
and past OC use with fasting triglyceride, plasma
cholesterol and high-density lipoprotein (HDL)
cholesterol, as well as the confounding effects of ad-
ditional variables.
From the Channing Laboratory, Departments of Medicine and
Preventive and Social Medicine, Harvard Medical School, and
Peter Bent Brigham Hospital Division of Affiliated Hospitals
Center, Inc., Boston, Massachusetts, and the Framingham Heart
Study, Framingham, Massachusetts.
Supported by contract HD-5-2832 and research grant HD-0362-
11 from the National Institute of Child Health and Human
Development.
Dr. Hennekens is recipient of a Research Career Development
Award (HL 00286) from the NHLBI.
Address for reprints: Charles H. Hennekens, M.D., Channing
Laboratory, 180 Longwood Avenue, Boston, Massachusetts 02115.
Received July 6, 1978; revision accepted February 26, 1979.
Circulation 60, No. 3, 1979.
486
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2016
Study Population
The present investigation examined a subgroup of
middle-class white women aged 21-39 years, who had
been classified according to their OC use at two com-
munity surveys performed in East Boston,
Massachusetts in 1973 and 1976. Those eligible for in-
clusion were premenopausal, nonpregnant women
aged 21-39 years, whose OC use had been
characterized in 1973 and 1976. Of 286 women who
fulfilled these criteria, 190 (70%) agreed to attend the
study clinic after a 14-hour overnight fast.
Methods of Procedure
During each community survey, questionnaire in-
formation was obtained concerning cigarette smoking
(yes or no), age, and stated body weight. In the clinic,
fasting blood specimens were drawn by antecubital
venipuncture with the women sitting. The specimens
were collected between 8 and 10 a.m. in tubes that
contained sodium EDTA. The tubes were inverted
eight times and centrifuged, and the plasma was
decanted. In addition, a fasting blood glucose was
collected in a tube containing fluoride-oxalate, and
measured by the glucose oxidase method.8
Triglycerides were measured by a modification of
the Kessler-Lederer method.9 Cholesterol concen-
trations in the total plasma were measured by the
method of Abell-Kendall." HDL cholesterol fractions
were separated by precipitating the other lipoproteins
with heparin-manganese chloride, and measured by a
modification of the method of Burstein and
colleagues.1' These determinations were performed in
the laboratory of the Framingham Heart Study, which
participates in a quality-control program through a
cooperative arrangement with the Lipid Research
Clinics. 12
 LIPIDS AND ORAL CONTRACEPTIVES/Hennekens et al. 487

Data Analysis TABLE 1. Fasting Plasma Triglyceride, Plasma Cholesterol

Initially, to evaluate the crude relationship between
current OC use and each of the lipid fractions, the
population was divided into those who did and those
who did not use OCs in the 1976 survey. A t test was
used to compare mean triglyceride, cholesterol and
HDL cholesterol values for current (1976) users and
nonusers.'3 In addition, current OC users were sub-
divided according to the composition of the OC
preparation being used, and the levels of the three lipid
fractions were compared for users of different
preparations.
Current OC users and nonusers were then further
classified according to OC use during the 1973 survey.
A weighted mean difference between current OC users
and nonusers was obtained by adjusting for past use,
using as weights the inverse of the variance for each
mean. 13
Finally, a multiple regression analysis13 was per-
formed, incorporating as additional variables
cigarette smoking, age, body weight and fasting
glucose. These variables were entered into the regres-
sion equation one by one in descending order of
strength of their association with the dependent
variable, in this case the value for the particular lipid
fraction (cholesterol, triglyceride or HDL
cholesterol). For each fraction, the residual associa-
tion with current OC use was evaluated by adding that
variable to the significant predictors.
Results
Table 1 shows the mean + SD for fasting
triglyceride, plasma cholesterol and HDL cholesterol
among current (1976) OC users and nonusers. Among
86 current OC users, the mean fasting triglyceride
level was significantly higher (p = 0.002) than that
found among 104 nonusers. Total plasma cholesterol
levels were also higher for current OC users than for
nonusers, but the difference did not attain statistical
significance (p = 0.08). HDL cholesterol values for
current users were somewhat lower than those for
nonusers (p = 0.08).
Past (1973) OC use had no apparent effect on
current (1976) levels of triglyceride, cholesterol or
HDL cholesterol. The difference in triglyceride levels
and HDL Cholesterol Levels Among Current (1976) Oral Contra-
ceptive (OC) Users and Nonusers
Current (1976) OC use
Lipid Yes No
fractions (n = 86) (n = 104) p
Triglyceride
(mg/100 ml) 95 42.4 73 53.9 0.002
Plasma cholesterol
(mg/100 ml) 198 35.4 189 36.4 0.08
HDL cholesterol
(mg/100 ml) 47 + 11.1 50 11.5 0.08
Values are mean = SD.
between current OC users and current nonusers was
significant even after adjustment for past OC use. The
nonsignificant elevation of total cholesterol in users as
compared with nonusers was virtually unchanged by
adjustment for past OC use, as was the small reduc-
tion in HDL cholesterol levels (table 2).
In the regression analysis for fasting triglycerides
(table 3), the significant variables were current smok-
ing (p = 0.001) and current OC use (p = 0.007), the
mean triglyceride levels for OC users being ap-
proximately 19 mg/100 ml higher than for nonusers.
Table 4 shows that the variables associated with
plasma cholesterol were age (p = 0.02), current smok-
ing (p = 0.07) and current OC use (p = 0.06). HDL
cholesterol levels (table 5) were inversely related to
cigarette smoking (p < 0.001), weight (p = 0.003) and
current OC use (p = 0.09).
Table 6 shows the composition of the various OC
preparations used by study participants. The mean
values of triglycerides, total cholesterol and HDL
cholesterol for users of the three most commonly
taken preparations are compared in table 7. The mean
triglyceride level was significantly higher for current
users of norethindrone 1.0 mg with mestranol 0.05 mg
(p = 0.01) and for users of norethindrone 1.0 mg with
mestranol 0.08 mg (p = 0.04) compared with the level
for nonusers. For users of norgestrel 0.5 mg with
ethinyl estradiol 0.05 mg, HDL cholesterol was
significantly lower (p < 0.001), while total plasma
cholesterol was significantly higher (p = 0.007) than
for nonusers.

TABLE 2. Fasting Plasma Triglyceride, Plasma Cholesterol and HDL Cholesterol Among Four Subgroups
Classified by Oral Contraceptive (OC) Use in 1973 and 1976
OC Use in 1973 Yes Yes No No
OC Use in 1976 Yes No Yes No
(n =33) (n =44) (n =53) (n =60)
Lipid fraction
Triglyceride
(mg/100 ml) 96 40.5 66 i44.5 95 43.9 79 59.6
Plasma cholesterol
(mg/100 ml) 200 28.9 186 35.2 197 39.1 192 37.2
HDL cholesterol
(mg/100 ml) 46 - 12.0 50 = 11.9 48 10.5 50 11.2
Values are mean - SD.

Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2016

488 CI RCULATION VOL 60, No 3, SEPTEMBER 1979

T\13LE 3. Allltiple Regression Analysis of Current Oral Con- TARiLE 7. Triglyceride, Cholesterol and. HDL Cholesterol
traceptive (OC) Us5c and Othcr Variables on Fasting Plas-ma Levels Among Users of Various Oral Contraceptive (OC) Prepa-
7'rigljycerides rations
Regressionl Standard Total 1IDL
Variable coefficient error I test p OC Composition Triglyceride cholesterol cholesterol
CLurrent smokinig 26.68 6.91 :3.86 0.001 Norethindrone 1.0
mg with mestraniol
Current. OC tuse 19.1.) 6.94 2.76 0.007 0.t)5 mg (28) 100* 200 51
Norgestrel 0.3 mg
with ethiinyl
estradiol 0.05
TAI31: 4. M1ul.ltiple Regression Analtysis of Current Oral Con- mg(26) 93 2121 41+
traceptiv( Use and Other Variables on 7Total Plasmra Cholesterol Norethindrome 1.0
Rtegressioni Stantdard mg with mestranol
Variable coefficient error t test p 0.08 mg (12) 106 2053 31
Age 1.03 () 0.44 2.36 0.02 Other preparations
(20) 85 173 46
Current ()C Iuse 10.04 3.23 1.92 0.06
All current OC
Cturren1t sn-moking 9.41 3.13 1.83 0.07 isers (86) 95 1 9s 47
Currenit nonusers
(104) 73 18S9 30
T1ABL' 3. 31 u-lltiple Regression Analysis of Cutrrent Oral Con- Numbers in parentheses are the number of participants in
lruc(ptive (OC) Use and Other Variables on HDL Cholesterol each group.
*p 0.01 for differeince fromi rmieani value for currenit non-
I' egression Stanidard users.
Variable coefhicien1t error t test p tp 0.007 for difference from meanu value for current nion-
Current smokin1g -5.86 1.37 -3.74 <0.001 users.
lp < 0.001 for difference from mean value for current non-
Weight -0.09 0.03 -3.03 0.003 users.
Curren1t OC use -2.63 1.36 -1.70 0.09 p = 0.04 for difference from mean value for current non-
users.

Discussion reports of increased risk of coronary heart disease

These data indicate a strongly positive association
between current OC use and fasting triglyceride, a
positive relationship of borderline significance with
total plasma cholesterol, and a small and nonsignifi-
cant inverse effect on HDL cholesterol. These find-
ings are consistent with those of recently reported
cross-sectional studies;"4 15 the current study also
evaluates effects of additional variables, namely,
cigarette smoking, age, body weight and fasting blood
glucose.
These findings are of particular interest in view of
among current OC users.` The increase in fasting
triglyceride levels for current OC users suggests that a
probable effect of OC use is to increase the level of
very low density lipoproteins (VLDL). While
triglyceride level (as a measure of VLDL) has not been
shown to have an independent relationship with risk of
coronary disease,16 a recent report from the
Framingham study'7 suggests that in OC users, VLDL
may be more triglyceride-enriched as well as in-
creased. Whether this qualitative change in VLDL in-
creases or decreases the atherogenic potential is un-
known.

TABLE 6. Oral Contraceptive Preparations Used by Study Participants

Preparation Number of
Progestin Estrogen users
Norethindrone 1.0 mg Mestranol 0.05 mg 28
Norgestrel 0.5 mg Ethinyl estradiol 0.05 mg 26
Norethindrone 1.0 mg Mestranol 0.08 mg 12
Norgestrel 0.3 mg Ethinyl estradiol 0.03 mg 7
Ethynodiol diacetate 1.0 mg Mestranol 0.1 mg 4
Norethindrone acetate 1.0 mg Ethinyl estradiol 0.05 mg 3
Norethindrone acetate 1.0 mg Ethinyl estradiol 0.02 mg 2
Ethynodiol diacetate 1.0 mg Ethinyl estradiol 0.03 mg 2
Norethynodrel 2.3 mg Mestranol 0.1 mg I
Norethindrone 2.0 mg Mestranol 0.1 mg 1

Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2016

 LIPIDS AND ORAL CONTRACEPTIVES/Hennekens et al.
Total plasma cholesterol has been shown to be a
strong predictor of coronary risk, especially at
younger ages.'6 The consistency and strength of this
relationship suggest that OC-induced cholesterol
elevations, although small, may at least partially ex-
plain the increased risk of MI.
The present data also showed a small, nonsignifi-
cant inverse relationship between OC use and HDL
cholesterol. A recent report showed that different OC
preparations had varying effects on HDL cholesterol.
HDL cholesterol levels appeared to be directly related
to estrogen dose and inversely related to progestin
dose.'8 In the present study, the numbers were too
small to permit firm conclusions regarding the effects
of individual estrogens and progestins on the lipid
fractions.
The occurrence of MI in OC users without
angiographic evidence of atherosclerotic lesions19
raises the question of whether alterations in coagu-
lation also play a role. OC users have been noted to
have altered laboratory measurements of coagula-
tion,'4, 20. 21 and increased rates of thromboembolic
diseases." 6, 22, 23 Increased sensitivity of platelets to
epinephrine has been described both in OC users20 and
in familial hypertriglyceridemia.24
The qualitative changes within each of the lipopro-
tein fractions with OC use need further clarification,
as do effects of different OC preparations on various
lipoprotein fractions. Finally, further data are
necessary to quantitate possible interactions between
OC use, other coronary risk factors and MI.
References
1. Mann JI, Inman WH: Oral contraception and death from
myocardial infarction. Br Med J 2: 245, 1975
2. Mann JI, Vessey MP, Thorogood M, Doll R: Myocardial in-
farction in young women with special reference to oral con-
traceptive practice. Br Med J 2: 241, 1975
3. Rosenberg L, Hennekens CH, Rosner B, Belanger C, Rothman
KJ, Speizer FE: A case-control study of oral contraceptive use
and myocardial infarction in US women. Am J Epidemiol. In
press
4. Beral V: Cardiovascular disease mortality trends and oral con-
traceptive use in young women. Lancet 2: 1047, 1976
5. Stolley PD, Tonascia JA, Tockman MS, Sartwell PE, Rutledge
AH, Jacobs MP: Thrombosis and low-estrogen oral contracep-
tives. Am J Epidemiol 102: 197, 1975
6. Collaborative Group for the Study of Stroke in Young Women:
Oral contraception and the increased risk of cerebral ischemia
Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2016
489
or thrombosis. N Engl J Med 388: 871, 1973
7. Mann JI, Doll R, Thorogood M, Vessey MP, Waters WE: Risk
factors for myocardial infarction in young women. Br J Prev
Soc Med 30: 94, 1976
8. Trinder P: Detection of blood glucose using oxidase-peroxidase
systems of non-carcinogenic chromogens. J Clin Path 22: 158,
1969
9. Kessler G, Lederer H: Fluorometric measurements of
triglycerides. In Technician Symposia: Automation in
Analytical Chemistry, edited by Skeggs LT Jr. New York, NY
Medical Inc, 1965, pp 341-344
10. Abell LL, Levy BB, Brodie BB, Kendall FE: A simplified
method for the estimation of total cholesterol in serum and
demonstration of its specificity. J Biol Chem 195: 357, 1952
11. Burstein M, Samaille J: Sur un dosage rapide du cholesterol lie
aux a aux ,B-lipoproteins du serum. Clin Chim Acta 5: 609,
1960
12. Manual of Operations: Lipid Research Clinics Program. Vol I:
Lipid and Lipoprotein Analysis. DHEW Publication no NIH
75-628, 1975
13. Snedecor GW, Cochran WG: Statistical Methods, 6th ed.
Ames, Iowa, Iowa State University Press, 1967
14. Meade TW, Chakrabarti R, Haines AP, Howarth DJ, North
WRS, Stirling Y: Haemostatic, lipid and blood pressure
profiles of women on oral contraceptives containing 50 ,ug or 30
,ug oestrogen. Lancet 2: 948, 1977
15. Wallace RB, Hoover J, Sandler D, Rifkind BM, Tyroler HA:
Altered plasma-lipids associated with oral-contraceptive or es-
trogen consumption. Lancet 2: 11, 1977
16. Kannel WB, Castelli WP, Gordon T: Cholesterol in the predic-
tion of atherosclerotic disease: new perspectives based on the
Framingham Study. Ann Intern Med 90: 85, 1979
17. Castelli WP, Kannel WB, Garrison RJ, Feinleib M,
McNamara PM: Blood lipoproteins in oral contraceptive users:
the Framingham Study. Circulation 56 (suppl III): III-44, 1977
18. Bradley DV, Wingerd J, Petitti DB, Krauss RM, Ramcharan
S: Serum high-density lipoprotein cholesterol in women using
oral contraceptives, estrogens and progestins. N Engl J Med
299: 17, 1978
19. Engel HJ, Handeshagen H, Lichtlen P: Transmural myocardial
infarction in young women taking oral contraceptives. Evidence
of reduced regional coronary flow in spite of normal coronary
arteries. Br Heart J 39: 477, 1977
20. Howie PW, Mallinson AC, Prentice CRM, Horne CHW,
McNicol GP: Effect of combined estrogen-progestagen oral
contraceptives, estrogen and progestagen on antiplasmin and
antithrombin activity. Lancet 2: 329, 1970
21. Carvalho ACA, Vaillancourt RA, Cabral RB, Lees RS,
Colman RW: Coagulation abnormalities in women taking oral
contraceptives. JAMA 237: 875, 1977
22. Vessey MP, Doll R: Investigation of relation between use of
oral contraceptives and thromboembolic disease. Br Med J 2:
199, 1968
23. Sartwell PE, Masi AT, Arthes FA, Greene GR, Smith HE:
Thrombo-embolism and oral contraceptives: an epide-
miological case-control study. Am J Epidemiol 90: 365, 1969
24. Carvalho ACA, Colman RW, Lees RS: Platelet function in
hyperlipoproteinemia. N Engl J Med 290: 434, 1974
Oral contraceptive use and fasting triglyceride, plasma cholesterol and HDL cholesterol.
C H Hennekens, D A Evans, W P Castelli, J O Taylor, B Rosner and E H Kass

Circulation. 1979;60:486-489
doi: 10.1161/01.CIR.60.3.486
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright 1979 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539

The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/60/3/486

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Circulation is online at:

http://circ.ahajournals.org//subscriptions/

Downloaded from http://circ.ahajournals.org/ by guest on June 16, 2016