Is the Response of Serum Lipids and Lipoproteins to

Postmenopausal Hormone Replacement Therapy Modified

by ApoE Genotype?
Anna-Mari Heikkinen, Leo Niskanen, Markku Ryynänen, Marja H. Komulainen,
Marjo T. Tuppurainen, Markku Parviainen, Seppo Saarikoski

Abstract—Postmenopausal hormone replacement therapy (HRT) has favorable effects on the serum lipid profile, and it also
decreases the risk of cardiovascular diseases. The apolipoprotein E genotype has influence on serum levels of lipids and
lipoproteins; apoE allele e4 (apoE4) is associated with high total and LDL cholesterol levels. Genotype also influences the
lipid responses to treatment with diet and statins, but the effect of HRT in different apoE genotypes is unknown. We studied
the effects of HRT on the concentrations of serum lipids in apoE4-positive early postmenopausal women (genotypes 3/4 and
4/4) compared with apoE4-negative women (genotypes 2/3 and 3/3) in a population-based, prospective 5-year study. In all,
232 early postmenopausal women were randomized into 2 treatment groups: an HRT group (n5116), which received a
sequential combination of 2 mg estradiol valerate (E2Val) from day 1 to 21 and 1 mg cyproterone acetate (CPA) from day
12 to 21 (Climen), and a placebo group (n5116), which received 500 mg/d calcium lactate. Serum concentrations of total,
LDL, and HDL cholesterol and triglycerides were measured at baseline and after 2 and 5 years of treatment. A total of 154
women completed the final analysis. During the follow-up period, serum total cholesterol and LDL cholesterol concentrations
decreased in the HRT group in apoE4-negative women (8.1% and 17.1%, respectively; P,0.001) but did not change in the
HRT group in apoE4-positive women or in the placebo group. Serum HDL cholesterol concentrations decreased in the
placebo group (apoE4-negative, 3.9%, P50.015; apoE4-positive, 8.1%, P50.004) but did not change significantly in the HRT
group. Serum triglyceride levels tended to increase in both study groups and genotypes (15.1% to 36.2%, P,0.038 to 0.001),
but no differences were observed between the study groups or genotypes, respectively. Our finding was that in
postmenopausal Finnish women LDL cholesterol levels in apoE4-negative subjects respond more favorably to HRT than
those in apoE4-positive subjects. This finding has potential importance in postmenopausal women with hypercholesterolemia,
Downloaded from http://ahajournals.org by on April 13, 2019

if confirmed in other studies. (Arterioscler Thromb Vasc Biol. 1999;19:402-407.)

Key Words: ApoE genotype n postmenopausal hormone replacement therapy n LDL cholesterol n prospective study

A polipoprotein E (apoE) is a liver polypeptide that has an

important role in lipid metabolism by serving as a ligand
for the LDL receptor.1 In humans, there are 3 alleles of apoE
(e2, e3, and e4) and hence 6 different genotypes (2/2, 2/3, 2/4,
3/3, 3/4, and 4/4). ApoE genotype distribution, in particular
that of the apoE allele e4 (apoE4), is associated with total and
LDL cholesterol levels1–3 and also with cardiovascular mor-
bidity.4,5 The frequencies of apoE genotypes vary in different
age, sex, and race groups.6 – 8 ApoE polymorphism is esti-
mated to explain 4% to 15% of the variation in LDL
cholesterol concentrations.4,7 In postmenopausal women, this
variation has been reported to be greater than in premeno-
pausal women.9 The response to cholesterol-lowering diet
and statins has been shown to differ in subjects with different
apoE genotypes.7–12 It is well known that postmenopausal
estrogen therapy changes serum lipoprotein concentrations
favorably, which may explain about 25% to 50% of the
cardioprotective effect of estrogen,13 but the association
between apoE genotype and lipoprotein responses to post-
menopausal HRT is not known.
The aim of this placebo-controlled, prospective 5-year trial
was to investigate the response of HRT (sequential combina-
tion of 2 mg estradiol valerate [E2Val] and 1 mg cyproterone
acetate [CPA]) with respect to serum lipids in apoE geno-
types 3/4 and 4/4 (apoE4-positive subjects) compared with
genotypes 2/3 and 3/3 (apoE4-negative subjects) in a
population-based, randomized group of early postmenopausal
women.
Materials and Methods
Subjects
The population of the present study is a subgroup of the Kuopio
Osteoporosis Risk Factor and Prevention Study. In 1989 a postal
inquiry was sent to all 14 200 47- to 56-year-old women in Kuopio

Received November 7, 1997; revision accepted August 18, 1998.

From the Departments of Obstetrics and Gynecology (A.-M.H., M.H.K., M.T.T., S.S.), Medicine (L.N.), Clinical Genetics (M.R.), and Clinical
Chemistry (M.P.), University Hospital of Kuopio, Finland.
Correspondence to Dr. A.-M. Heikkinen, Department of Obstetrics and Gynecology, Kuopio University Hospital, PO Box 1777, FIN-70211 Kuopio,
Finland. E-mail anna-mari.heikkinen@pp.inet.fi
© 1999 American Heart Association, Inc.
Arterioscler Thromb Vasc Biol. is available at http://www.atvbaha.org

402
 Heikkinen et al February 1999 403

TABLE 1. Distribution of ApoE Genotypes and Gene

Frequencies in 2 Different Treatment Groups
HRT Placebo All
Allele/Genotype (n561) (n598) (n5159)
Gene frequency
e2 0.041 0.046 0.044
e3 0.779 0.781 0.780
e4 0.181 0.173 0.179
Genotype, N (%)
E2/3 1 (1.6) 8 (8.2) 9 (5.7)
E2/4 4 (6.6) 1 (1.0) 5 (3.1)
E3/3 40 (65.6) 57 (58.2) 97 (61.0)
E3/4 14 (23.0) 31 (31.6) 46 (28.9)
E4/4 2 (3.3) 1 (1.0) 3 (1.9)
E2/3 or E3/3 41 (67.2) 65 (66.4) 107 (67.3)
E3/4 or E4/4 16 (26.2) 32 (32.6) 48 (30.2)

(n57). There were 9 dropouts because of nonmedical reasons. The

Figure 1. Study design.
Province, Eastern Finland, to investigate osteoporosis risk factors
among perimenopausal women.14 The 464 voluntary postmeno-
pausal women who had their last menstrual period within 6 to 24
months before the study were included in the clinical osteoporosis
prevention trial. Exclusion criteria were restricted to general contra-
indications for HRT, including history of estrogen-dependent cancer,
Downloaded from http://ahajournals.org by on April 13, 2019
reasons for the remaining 17 dropouts in the HRT group were
miscellaneous, eg, abdominal pain (n53) or weight increase or
swelling (n52). In the HRT group, one of the following events
occurred: endometrial carcinoma, myocardial infarction, and tran-
sient cerebral ischemia. In the placebo group (n5116), 11 women
(9.5%) dropped out. The most common reasons reported were
climacteric symptoms or other medical reasons that required HRT
(n55). One hip fracture and one case of endometrial hyperplasia
occurred in the placebo group. Among the 53 dropouts 2 women
died; one in the HRT group because of rectal carcinoma and one in
the placebo group because of malignant melanoma. Additionally, 25

thromboembolic diseases, and medication-resistant hypertension.

The participants were randomized by a computer to 4 treatment
groups: E2Val/CPA group, vitamin D3 group, E2Val/CPA1vitamin
D3 group, and calcium lactate group (placebo). Random allocation to
study groups was carried out by blocks using a computer, the block
size being 4, 8, or 12. The study was not blinded as to HRT. The data
analyses were done blindly. Those women who wanted to change
treatment groups were excluded from the analysis. The personnel
involved were unaware of the group allocations. The study design,
and in particular the adverse effect of vitamin D on the serum lipid
profile, has been described elsewhere in more detail.15 Therefore, for
this study, only the women using HRT or placebo without vitamin D3
were included: (1) HRT group (n5116): E2Val (2 mg) on cycle days
1 to 21, combined with CPA (1 mg; Climen, Schering AG) on cycle
days 12 to 21, with a treatment-free interval on cycle days 22 to 28;
and (2) placebo group (n5116): calcium lactate (Rohto Ltd), 500
mg/d, equivalent to 93 mg Ca21/d. Study design and formation of the
present study population are depicted by the flow diagram (Figure 1).
Written informed consent was obtained from the participants, and
the study design was approved by the ethics committee of Kuopio
University Hospital. The daily calcium intake was calculated as the
sum of calcium intake from milk, sour milk, yogurt (120 mg/dL), and
cheese (87 mg/slice). This also estimates indirectly the intake of
saturated fatty acids, as milk products are the most important source
of them in the Finnish diet.16 The weekly duration of physical
activity and the smoking and drinking habits were investigated. The
weekly consumption of alcohol was calculated as absolute ethanol
intake in grams. Each subject visited the outpatient clinic once a
year. Fasting blood samples were taken in the morning for determi-
nation of the lipid parameter values. If the serum total cholesterol
concentration was .6.0 mmol/L, dietary instructions were given. If
the serum total cholesterol concentration remained repeatedly above
7.0 mmol/L, the requirement for hypocholesterolemic medication
was considered on clinical grounds, and the subject was excluded
from the final analysis.
During the 5-year follow-up period 53 women of 232 (22.8%)
dropped out. Prospectively defined stopping rules were same as the
exclusion criteria. In the HRT group (n5116), there were 42 (36.2%)
dropouts, mostly with disorders of bleeding (n59) or headache
women (10.8%) were excluded from the final analysis. Eight women
were excluded as a result of interruptions in HRT ($1 month), 2
women in the HRT group and 4 women in the placebo group because
of hypocholesterolemic medication, 3 women from both groups
because of missing laboratory data, and 5 women because of apoE
genotype 2/4 (Table 1). The inclusion of these subjects did not alter
the conclusions of this study (data not shown). A total of 154 women
(66.4%) were included in the final analysis (Figure 1).
Analyses
The concentrations of total serum cholesterol, LDL cholesterol, HDL
cholesterol, total triglycerides, follicle-stimulating hormone (FSH),
and estradiol (E2) were measured at baseline and after 2 and 5 years
of treatment from fasting serum samples taken in the morning.
Serum E2 concentrations were measured by radioimmunoassays
(Sorin Biomedica; interassay coefficient of variation [CV] ,8.5%)
and those of FSH by luminescence immunoassay (Byk-Sangtec;
interassay CV,5.6%).
Concentrations of serum total cholesterol and triglycerides were
measured enzymatically (CHOD-PAP and GPO-PAP methods;
Boehringer) using a Hitachi 717 analyzer. The same cholesterol
method was also used for HDL cholesterol after removal of LDL and
VLDL through the use of dextran sulfate/MgCl2.17 The concentration
of serum LDL cholesterol was calculated by LDL cholester-
ol5total cholesterol2(HDL cholesterol10.453triglycerides).18 The
analytical variations of concentrations of total cholesterol, HDL
cholesterol, and total triglycerides within the series have been 1.5%,
4%, and 2%, respectively, at the levels in question using fully
automated methods for total cholesterol and triglycerides and semi-
automated measurement of HDL cholesterol. The long-term varia-
tions in total cholesterol, HDL cholesterol, and triglyceride analyses
were followed by assay of quality control fresh samples from
Labquality Ltd, and the between-series variations have been below
3%, 7%, and 4.5%, respectively. No drift in the analyte levels was
found during the present study.
The apoE genotype was determined from blood leukocytes. DNA
was extracted by standard phenol/chloroform extraction.19 ApoE
genotypes were analyzed by using the polymerase chain reaction
 404 Lipids, Lipoproteins, HRT, and ApoE Genotype

TABLE 2. Baseline Characteristics on 154 Postmenopausal Women According to Treatment

Groups and ApoE Genotypes
HRT, ApoE HRT, ApoE Placebo, ApoE Placebo, ApoE
2/3, 3/3 3/4, 4/4 2/3, 3/3 3/4, 4/4
(n541) (n516) (n565) (n532) P Value*
Age, y 52.560.3 52.160.6 52.460.3 52.660.4 0.607
Time since menopause, y 1.160.08 1.0060.16 1.1260.06 1.0760.09 0.709
Previous HRT use, y 0.6760.30 0.2860.14 0.3260.09 0.4760.18 0.989
Body mass index, kg/m2 26.260.6 27.261.0 26.060.5 26.660.6 0.584
Dietary Ca-intake, mg/d 726648 8556140 848655 776663 0.796
Smoking, pack-years† 8.3861.39 13.161.87 8.0161.92 6.7762.96 0.472
Current smokers, % 17.1 12.5 21.5 12.5 0.497
Physically active persons, %‡ 34.1 18.7 27.7 21.9 0.795
Alcohol, absolute ethanol g/week 18.364.9 26.2614.6 20.664.5 27.668.0 0.588
FSH, IU/L 65.764.5 47.866.3 62.863.3 61.665.4 0.163
E2, nmol/L 0.1460.02 0.1860.04 0.1360.02 0.1860.05 0.648
*Values are the mean6SEM by Kruskal-Wallis test, (in number of smokers and physically active persons, x2 test);
no statistically significant differences between the groups.
†Smoking is the life-time number of cigarettes/203365.
‡Physically active person is 3 or more hours of physical activity/week.

(PCR) as described earlier,20 with slight modifications. The PCR
products were digested with HhaI (New England Biolabs). Digested
DNA fragments were analyzed using polyacrylamide gel electro-
phoresis and visualized by ethidium bromide staining.
Statistical Methods
The analyses were carried out on a treatment basis. Statistical
analyses of the longitudinal data were carried out using analysis of
Downloaded from http://ahajournals.org by on April 13, 2019
positive group, 3.362.0%; placebo–apoE4-negative group,
5.561.0%; and placebo–apoE4-positive group, 4.361.6%;
P50.708). Concentrations of serum E2 increased in the HRT
group but did not change significantly in the placebo group.
The 5-year concentrations in serum E2 were identical between
apoE4-negative and apoE4-positive subjects (HRT group,

0.1960.02 and 0.2360.06 nmol/L, respectively, P50.507;

variance for repeated measures (MANOVA). Log transformation of
the data were used if the values were not normally distributed. The
Student’s t test for paired data was used to test the significance of
differences within the 4 different treatment– genotype groups if there
was a time-related change within the group detected by MANOVA.
The significance between the groups was tested using MANOVA.
The 5-year LDL cholesterol levels adjusted for age, body mass
index, smoking, apoE4-allele, and the interaction for HRT and
apoE4-allele were examined by analysis of covariance.
One-way analysis of variance with the Newman-Keuls post hoc
test was used to test the significance of differences between lipid
concentrations at baseline and the relative lipoprotein changes in the
2 groups. Because the body weight, body mass index, and FSH levels
were not normally distributed after log transformation, one-way
analysis of variance with the Kruskall-Wallis test was used to test the
differences in these parameters between the groups. The Kruskall-
Wallis test was also used to test the differences between the study
groups with respect to other baseline data.
A P value of ,0.05 was considered statistically significant. The
results are reported as mean 6 standard error of mean (SEM). All the
data were analyzed using SPSS for UNIX (SPSS Inc).
Results
Frequencies of apoE alleles and genotypes are outlined
according to the treatment group in Table 1. The distribution
of alleles was similar to that reported earlier in Finnish and
Swedish populations2,6 but with a higher frequency of apoE4
compared with other white populations.21 There were no
statistically significant differences between the study groups
in the baseline characteristics or the laboratory parameters
(Table 2).
The relative body weight (difference between follow-up
and baseline, divided by baseline level, as a percentage)
increased similarly in all 4 treatment– genotype groups after 5
years (HRT–apoE4-negative group, 4.760.8%; HRT–apoE4-
placebo group, 0.0460.02 and 0.0460.02 nmol/L, respec-
tively, P50.861).
The concentrations of serum total cholesterol decreased in
the HRT group by 5.1% after 2 years (P50.014) and by 8.1%
after 5 years (P,0.001) in apoE4-negative subjects, but they
did not decrease significantly in apoE4-positive subjects. In
the placebo group, the concentrations of serum total choles-
terol did not change in apoE4-negative subjects, whereas they
had increased by 3.9% after 2 years in apoE4-positive
subjects (P50.015), but they had returned to the baseline
level after 5 years. The changes between the HRT and the
placebo groups were statistically significant (MANOVA,
P,0.001) (Table 3, Figure 2).
LDL cholesterol concentrations had decreased in the HRT
group by 10.3% after 2 years (P,0.001) and by 17.1% after
5 years in apoE4-negative subjects (P,0.001). There was a
slight but nonsignificant tendency for decreases in LDL
cholesterol concentrations in the HRT group in apoE4-
positive subjects and in the placebo group in apoE4-negative
subjects, whereas LDL cholesterol concentrations had in-
creased by 5.7% (P50.010) in the placebo group in apoE4-
positive subjects after 2 years but had returned to the baseline
level after 5 years. The time-related changes between the
treatment groups were statistically significant (MANOVA,
P,0.001) (Table 3, Figure 2).
Furthermore, in the multivariate model (analysis of covar-
iance), the LDL cholesterol levels at 5-year examination were
associated with apoE genotype, taking into account the
effects of age, body mass index, smoking, and interaction
between HRT and apoE4 allele (Table 4).
 Heikkinen et al February 1999 405

TABLE 3. Concentrations of Serum Lipids and Lipoproteins on 154

Postmenopausal Women According to Treatment Groups and ApoE Genotypes*
HRT, ApoE HRT, ApoE Placebo, ApoE Placebo, ApoE
2/3, 3/3 3/4, 4/4 2/3, 3/3 3/4, 4/4
(n541) (n516) (n565) (n532)
Total cholesterol, mmol/L
Baseline 6.5360.15 6.5760.14 6.2360.13 6.2360.16
2 years 6.2060.13† 6.4160.15 6.2660.12 6.4760.16†
5 years 6.0160.12§ 6.3060.15 6.1160.12 6.2160.15
LDL cholesterol, mmol/L
Baseline 4.3860.12 4.3760.16 4.2060.12 4.0460.16
2 years 3.9360.13§ 4.0960.15 4.1560.11 4.2760.12‡
5 years 3.6360.12§ 4.0360.13 4.0260.10 4.0560.14
HDL cholesterol, mmol/L
Baseline 1.6260.05 1.6660.14 1.5360.04 1.6160.07
2 years 1.6960.05† 1.6860.13 1.5460.04 1.6260.07
5 years 1.6760.05 1.5860.15 1.4760.04† 1.4860.06‡
Triglycerides, mmol/L
Baseline 1.1660.09 1.2060.15 1.1960.05 1.2860.11
2 years 1.2960.07† 1.4260.14† 1.2860.07‡ 1.2960.11
5 years 1.5860.15§ 1.5260.14† 1.3760.09§ 1.5060.10‡
*Values are mean6SEM; t test for paired data was used to test the significance of differences
within each treatment– genotype group; †P,0.05; ‡P,0.01; §P,0.001.

The concentrations of serum HDL cholesterol had in- years, the increase was 36.2% (P,0.001) in the apoE4-
creased by 4.3% (P50.042) and 3.1% (P5ns) after 2 and 5 negative HRT group, 26.7% (P50.030) in the apoE4-positive
Downloaded from http://ahajournals.org by on April 13, 2019

years, respectively, in the HRT group in apoE4-negative
subjects, but no significant changes in apoE4-positive sub-
jects were observed. In the placebo group, serum HDL
cholesterol concentrations had decreased after 5 years in
apoE4-negative subjects by 3.9% (P50.015) and in apoE4-
positive subjects by 8.1% (P50.004). However, no signifi-
cant changes between the treatment groups or genotypes were
observed (Table 3, Figure 2).
Serum triglyceride levels increased in all subjects irrespec-
tive of the genotype during the follow-up period. After 5
HRT group, 15.1% (P,0.001) in the apoE4-negative placebo
group, and 17.2% (P50.003) in the apoE4-positive placebo
group. No statistically significant differences between the
treatment groups or genotypes were observed (Table 3,
Figure 2).
In this study we observed a reduction in LDL cholesterol of
0.4 mmol/L between those with apoE4 allele and those
without it in subjects receiving HRT. When we set the a
(probability excluding type 1 error) to 5% and b (probability
of type 2 error) to 20% with SD of 0.4 mmol/L, the required
sample size will be 16.7 subjects per group. In the HRT group
there were 16 subjects with apoE4 allele and 41 without it;
therefore, the sample size here is adequate to show these
changes, given the changes are so prominent.
The lipid and lipoprotein concentrations were analyzed
yearly. The results of 1-, 3-, and 4-year examinations were in
line with the reported 2- and 5-year results (data not shown).

Discussion
The new finding in the present study was that the beneficial
response of total and, especially, LDL cholesterol to HRT in

TABLE 4. Analysis of Covariance of LDL Cholesterol Levels at

5-Year Examination
Figure 2. The mean 5-year changes (%) of serum total choles- Source of Variation F Value P Value
terol (total-C), LDL cholesterol (LDL-C), HDL cholesterol (HDL-
C), and triglycerides. †P,0.05, ‡P,0.01, §P,0.001 by t test for Age, y 0.415 0.524
paired data. HRT apoE4-neg indicates E2 (2 mg) 1 CPA (1 mg) Body mass index, kg/m2 2.985 0.094
cyclically, apoE genotype 2/3 or 3/3; HRT apoE4-pos, E2 (2 mg)
1 CPA (1 mg) cyclically, apoE genotype 3/4 or 4/4; Placebo ApoE4 allele 1.468 0.047
apoE4-neg, calcium lactate (500 mg/d), apoE genotype 2/3 or Current smoking 0.235 0.631
3/3; and Placebo apoE4-pos, calcium lactate (500 mg/d), apoE
Interaction for HRT and apoE4 4.298 0.290
genotype 3/4 or 4/4.
 406 Lipids, Lipoproteins, HRT, and ApoE Genotype
postmenopausal women was related to the apoE genotype.
Additionally, our results confirm the persistence of truly
long-term effects of sequential E2 (2 mg) and CPA (1 mg)
treatment on serum total and LDL cholesterol. This study was
placebo-controlled, population-based, successfully random-
ized, and well matched regarding baseline characteristics and
weight changes during the follow-up period, which all
strengthen our findings.
It is well established that postmenopausal estrogen therapy
has favorable effects on serum lipoprotein concentrations.22,23
However, a wide variation in the lipid responses has been
observed in previous reports, which have been attributed
either to the effect of the added progestin,23,24 and by larger
responses in hypercholesterolemic subjects.25 It has also been
suggested that positive lipid effects of HRT may diminish
with the duration of treatment.26
However, there are no studies in which the impact of
genetic factors on lipid responses to HRT have been investi-
gated. It is fairly well established that apoE allele 2 is
associated with lower and allele 4 with higher serum total and
LDL cholesterol concentrations compared with those associ-
ated with allele 3.1,3 Furthermore, there is evidence that
hormonal status modulates the lipoprotein variation related to
apoE genotype.7–9 In postmenopausal women, the association
between apoE phenotype and LDL cholesterol levels has
been stronger than in premenopausal women or in men.9 The
relatively high serum total cholesterol levels in subjects with
apoE4 have been suggested to respond more favorably to a
cholesterol-reducing diet, as these subjects show enhanced
Downloaded from http://ahajournals.org by on April 13, 2019
absorption of dietary cholesterol, especially in populations
consuming diets rich in saturated fat,10,11,27 although this is
not confirmed in all studies.28,29 Our findings are unlikely to
be caused by different dietary habits between the HRT and
placebo groups. During the 5-year follow-up period, LDL
cholesterol levels changed relatively little in the placebo
group, although menopause should result in an increase of
about 10%.30 This most likely reflects the decreased intake of
cholesterol and saturated fat during these years in the Finnish
population.31
The finding that apoE polymorphism determines the LDL
cholesterol response to HRT has not been previously recog-
nized. In the cross-sectional analysis of the Framingham
Offspring Study, no association between apoE genotype and
serum LDL cholesterol was found, but the number of post-
menopausal women using HRT was small.9 Although our
finding of this genetic determinant of LDL cholesterol to
HRT is a new one, there is evidence that apoE polymorphism
may influence the responses to statins in familial and nonfa-
milial hypercholesterolemia; apoE4-positive subjects have
more sluggish response than apoE4-negative ones.12,32 Taken
together, apoE polymorphism seems to be a clinically impor-
tant determinant of the various lipid-lowering interventions:
subjects with apoE4 allele have enhanced response to dietary
therapy, whereas the response to treatment with statins and
postmenopausal HRT is impaired.
ApoE plays a role in liver lipoprotein metabolism and
clearance, and LDL receptor affinity varies according to the
apoE isoform.1,4 ApoE4 downregulates hepatic LDL recep-
tors, enhances liver lipoprotein uptake, and is associated with
increased serum LDL cholesterol concentrations.4 On the
other hand, oral estrogen therapy mediates an upregulation of
liver LDL receptors and therefore has a LDL-lowering
effect.22 Hence it is plausible, on the basis of our findings, to
suggest that the upregulation of LDL receptors induced by
estrogen is impaired in menopausal subjects with the apoE4
allele. The exact mechanisms remain to be demonstrated.
This study was not originally aimed at examining the
interaction between genotype and HRT. There were 42
(36.2%) dropouts in the HRT group. Although this number is
very low compared with long-term compliance of HRT, the
number of apoE4-positive subjects was relatively low in the
HRT group in the final analysis. The magnitude of the effect
of apoE4 was rather strong, and therefore the sample sizes are
adequate considering the homogeneity of the study popula-
tion and the large number of apoE4-negative subjects in the
present study. However, our findings are limited to the
Finnish population and need to be confirmed in other studies.
To conclude, the apoE genotype modulates the response of
serum LDL cholesterol to HRT in postmenopausal Finnish
women. This finding may be of potential importance, espe-
cially in the treatment of hypercholesterolemic postmeno-
pausal subjects, if confirmed in other studies.
Acknowledgments
We thank Schering AG and the Research Foundations of Leiras and
Orion for supporting this research. We also thank Sirkka Harle, Seija
Oinonen, and Liisi Saarela for technical help and Dr. Nick Bolton for
checking the language of the manuscript.
References
1. Mahley RW. Apoprotein E: cholesterol transport protein with expanding
role in cell biology. Science. 1988;240:622– 640.
2. Ehnholm C, Lukka M, Kuusi T, Nikkilä E, Uterman G. Apolipoprotein E
polymorphism in the Finnish population: gene frequencies and relation to
lipoprotein concentrations. J Lipid Res. 1986;27:227–235.
3. Dallongeville J, Lussier-Cacan S, Davignon J. Modulation of plasma
triglyceride levels by apoE phenotype: a meta-analysis. J Lipid Res.
1992;240:447– 454.
4. Davignon J, Gregg RE, Sing SF. Apolipoprotein E polymorphism and
atherosclerosis. Atherosclerosis. 1988;8:1–21.
5. Tiret L, Knijff P, Menzel H-J, Ehnholm C, Nicaud V, Havekes LM. ApoE
polymorphism and predisposition to coronary heart disease in youths of
different European populations: the EARS study: European Atheroscle-
rosis Research Study. Arterioscler Thromb. 1994;14:1617–1624.
6. Eggertsen G, Tegelman R, Ericsson S, Angelin B, Berglund L. Apoli-
poprotein E polymorphism in a healthy Swedish population: variation of
allele frequency with age and relation to serum lipid concentrations. Clin
Chem. 1993;39:2125–2129.
7. Kamboh MI, Aston CE, Ferrell RE, Hamman RF. Impact of apoli-
poprotein E polymorphism in determining interindividual variation in
total cholesterol and low density lipoprotein cholesterol in Hispanics and
non-Hispanic whites. Atherosclerosis. 1993;98:201–211.
8. Ferrières J, Sing CF, Roy M, Davignon J, Lussier-Cacan S. Apoli-
poprotein E polymorphism and heterozygous familial hypercholesterol-
emia: sex-specific effects. Arterioscler Thromb. 1994;14:1554 –1560.
9. Schaefer EJ, Lamon-Fava S, Johnson S, Ordovas JM, Schaefer MM,
Castelli WP, Wilson PWF. Effects of gender and menopausal status on
the association of apolipoprotein E phenotype with plasma lipoprotein
levels: results from the Framingham Offspring Study. Arterioscler
Thromb. 1994;14:1105–1113.
10. Miettinen TA, Gylling H, Vanhanen H, Ollus A. Cholesterol absorption,
elimination, and synthesis related to LDL kinetics during varying fat
intake in men with different apoprotein E phenotypes. Arterioscler
Thromb. 1992;12:1044 –1052.
11. Lopez-Miranda J, Ordovas JM, Mata P, Lichtenstein AH, Clevidence B,
Judd JT, Schaefer JE. Effect of apolipoprotein E phenotype on diet-
induced lowering of plasma low density lipoprotein cholesterol. J Lipid
Res. 1994;35:1965–1975.
12. Ordovas JM, Lopez-Miranda J, Perez-Jimenez F, Rodriguez C, Park J-S,
Cole T, Schaefer EJ. Effect of apolipoprotein E and A-IV phenotypes on

the low density lipoprotein response to HMG CoA reductase inhibitor
therapy. Atherosclerosis. 1995;13:157–166.
13. Barret-Connor E, Bush TL. Estrogen and coronary heart disease in
women. JAMA. 1991;265:1861–1867.
14. Tuppurainen M, Honkanen R, Kröger H, Saarikoski S, Alhava E. Osteo-
porosis risk factors, gynecological history and fractures in perimeno-
pausal women—the results of the baseline postal enquiry of the Kuopio
Osteoporosis Risk Factor and Prevention Study. Maturitas. 1993;17:
89 –100.
15. Heikkinen A-M, Tuppurainen M, Niskanen L, Komulainen M, Penttilä I,
Saarikoski S. Long-term vitamin D3 supplementation may have adverse
effects on postmenopausal hormone replacement therapy. Eur J Endo-
crinol. 1997;137:495–502.
16. Heinonen M, Lampi A-M, Hyvönen L, Homer D. The fatty acid and
cholesterol content of the average Finnish diet. J Food Comp Anal.
1992;5:198 –208.
17. Penttilä IM, Voutilainen E, Laitinen P, Juutilainen P. Comparison of
different analytical and precipitation methods for direct estimation of
serum high-density lipoprotein cholesterol. Scand J Clin Lab Invest.
1981;41:353–360.
18. Friedewald WT, Levy RI, Fredrickson D. Estimation of the concentration
of low-density lipoprotein cholesterol in plasma, without use of the
preparation ultracentrifuge. Clin Chem. 1972;18:499 –502.
19. Knowlton RG, Weaver EJ, Struyk AF, Knobloch WH, King RA, Norris
K, Shamban A, Uitto J, Jimenez SA, Prockop DJ. Genetic linkage anal-
ysis of hereditary arthro-ophthalmopathy (Stickler syndrome) and the
type II procollagen gene. Am J Hum Genet. 1989;45:681– 688.
20. Hixon JE, Vernier DT. Restriction isotyping of human apolipoprotein E
by gene amplification and cleavage with HhaI. J Lipid Res. 1990;31:
545–548.
21. Hallman DM, Boerwinkle E, Saha N, Sandholzer C, Menzel HJ, Csazar
A, Utermann G. The apolipoprotein E polymorphism: a comparison of
allele frequencies and effects in nine populations. Am J Hum Genet.
1991;49:338 –349.
22. Walsh BW, Sciff I, Rosner B, Greenberg L, Ravnikar V, Sacks FM. Effects
of postmenopausal estrogen replacement on the concentrations and metabo-
lism of plasma lipoproteins. N Engl J Med. 1991;325:1196–1204.
Downloaded from http://ahajournals.org by on April 13, 2019
Heikkinen et al February 1999 407
23. The Writing Group for the PEPI Trial. Effects of estrogen/progestin
regimens on heart disease risk factors in postmenopausal women: the
Postmenopausal Estrogen/Progestin Interventions (PEPI) trial. JAMA.
1995;273:199 –208.
24. Rijpkema AMH, Van der Sanden AA, Ruijs AHC. Effects of postmeno-
pausal oestrogen progestogen replacement therapy serum lipids and
lipoproteins: a review. Maturitas. 1990;12:259 –285.
25. Tikkanen MJ, Kuusi T, Vartiainen E, Nikkilä EA. Treatment of post-
menopausal hypercholesterolemia with oestradiol. Acta Obstet Gynecol.
1979;84:222–226.
26. Farish E, Fletcher CD, Dagen MM, Hart DM, Al-Azzawi F, Parkin DE,
Howie CA. Lipoprotein and apolipoprotein levels in postmenopausal
women on continuous oestrogen/progestogen therapy. Br J Obstet
Gynecol. 1989;96:358 –364.
27. Mänttäri M, Koskinen P, Enholm C, Huttunen JK, Manninen V. Apoli-
poprotein E polymorphism influences the serum cholesterol response to
dietary intervention. Metabolism. 1991;40:217–221.
28. Lefevre M, Ginsberg HN, Kris-Etherton PM, Elmer PJ, Stewart PW,
Ershow A, Pearson TA, Roheim PS, Ramakrishnan R, Derr J, Gordon DJ,
Reed R, for the DELTA Reasearch Group. ApoE genotype does not
predict lipid response to changes in dietary saturated fatty acids in a
heterogeneous normolipidemic population. Arterioscler Thromb Vasc
Biol. 1997;17:2914 –2923.
29. Pasagian-Macaulay A, Aston CE, Ferell RE, McAllister AE, Wing RR,
Kuller LH. A dietary and behavioral intervention design to lower coro-
nary heart disease: risk factors are unaffected by variation at the APOE
gene locus. Atherosclerosis. 1997;132:221–227.
30. Jensen J, Nilas L, Christiansen C. Influence on menopause on serum
lipids and lipoproteins. Maturitas. 1990;12:321–331.
31. Pietinen P, Vartiainen E, Seppäinen R, Aro A, Puska P. Changes in diet
in Finland from 1972 to 1992: impact on coronary heart disease risk. Prev
Med. 1996;25:243–250.
32. Carmena R, Roederer G, Mailloux H, Lussier-Cacan S, Davignon J. The
response to lovastatin treatment in patients with heterozygous familial
hypercholesterolemia is modulated by apolipoprotein E polymorphism.
Metabolism. 1993;42:895–901.