Atherosclerosis 188 (2006) 126–133
The effect of apolipoprotein E genotype on serum
lipoprotein particle response to exercise
Richard L. Seip a,∗ , James Otvos b , Cherie Bilbie a , Gregory J. Tsongalis a , Mary Miles c ,
Robert Zoeller d , Paul Visich e , Paul Gordon f , Theodore J. Angelopoulos g ,
Linda Pescatello h,i , Niall Moyna j , Paul D. Thompson a
a
Cardiology Division, Hartford Hospital, 80 Seymour Street, Hartford, CT 06102-5037, USA
b North Carolina State University, Raleigh, NC, USA
c Montana State University, Bozeman, MT, USA
d Florida Atlantic University, Davie, FL, USA
e Central Michigan University, Mt. Pleasant, MI, USA
f West Virginia University, Morgantown, WV, USA
g University of Central Florida, Orlando, FL, USA
h University of Connecticut, Storrs, CT, USA
i Lifespan Health System, Brown University, Providence, RI, USA
j Dublin City University, Dublin, Ireland
Received 16 June 2004; received in revised form 30 March 2005; accepted 1 June 2005
Available online 13 July 2006
Abstract
Exercise affects lipoprotein metabolism and apolipoprotein E (Apo E) genotype may alter changes in lipoprotein subclasses that occur
with exercise. The present study examined the effects of Apo E genotype (APOE) on the response of lipoprotein subclass concentrations to
long-term exercise. A prospective longitudinal study, conducted at seven centers, genetically screened 566 individuals to create three cohorts
of healthy adults, equal for gender and the most common APOE variants: E2/3 (n = 35), E3/3 (n = 40), and E3/4 (n = 31). Subjects with body
mass index (BMI) ≥31 or evidence of dyslipidemia or metabolic disease were excluded. All subjects exercised aerobically at 75% of maximal
heart rate for 40 min, four times weekly for 6 months. Fasting lipoprotein subpopulations were measured before and after exercise training
using proton nuclear magnetic resonance spectroscopy. Serum lipids for the entire cohort did not change with exercise training, but the LDL
subpopulation response varied by APOE. Small-sized LDL particles decreased only in the APOE3 homozygotes whereas medium-sized LDL
particles increased only in this group. These changes were directionally different from the responses in the E2/3 and E3/4 subjects (p < 0.05).
Neither exercise nor APOE variant affected overall LDL or HDL size or cholesterol concentration, but exercise decreased VLDL diameter
by 3.5 nm (p < 0.001) attributable to decreases in large VLDL in each APOE group. In conclusion, APOE variants influence the serum LDL
subpopulation response to exercise training in normolipidemic subjects. Subjects homozygous for APOE3 experienced the most beneficial
lipid effects from exercise training.
© 2005 Elsevier Ireland Ltd. All rights reserved.
Keywords: Apo E; Lipoproteins; Particle size; Aerobic exercise; Lipid metabolism
1. Introduction
∗
Corresponding author. Tel.: +1 860 545 5005; fax: +1 860 545 2882.
E-mail address: [email protected] (R.L. Seip).
0021-9150/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.atherosclerosis.2005.06.050
Lipoprotein subclass levels and particle size distributions
convey information about cardiovascular disease (CVD) risk
not provided by traditional blood lipid measures [1–4].
Nuclear magnetic resonance (NMR) analysis offers a rapid
R.L. Seip et al. / Atherosclerosis 188 (2006) 126–133
and efficient means of assessing the concentrations of 15
lipoprotein subclasses of varying size, as well as the overall average particle sizes of very low density lipoproteins
(VLDL), low density lipoproteins (LDL), and high density lipoproteins (HDL) [5]. Small LDL particles indicate
increased CVD risk [6].
Apolipoprotein E (Apo E) facilitates triglyceride clearance by mediating the binding of triglyceride-rich particles to
hepatic receptors [7], and common Apo E genotype (APOE)
variants (E2/3, E3/3, and E4/3) modulate LDL metabolism
[8]. APOE genotype is hypothesized to differentially affect
the formation of small, medium, and large LDL particles [9],
at least in part through the unique structural and biophysical properties that distinguish the three most common allelic
isoforms of Apo E (E2, E3, and E4). These characteristics
differentially affect the binding affinity of Apo E to hepatic
receptors [10], the activation of hepatic lipase by Apo E [11],
and the strength of Apo E interactions with heparan sulfate
proteoglycans [12], all of which are involved in the conversion of triglyceride-rich particles to LDL particles [9].
We have recently shown that common APOE variants
affect the lipid response to 6 months of aerobic exercise [13].
Exercise decreased the LDL-cholesterol to HDL-cholesterol
ratio in subjects with the APOE3/3 genotype (the most common APOE genotype [14]), but did not alter the LDL/HDLcholesterol ratio in APOE2/3 or 4/3 subjects [13]. Others
have demonstrated that exercise can decrease small LDLcholesterol and/or increase overall LDL size [15–19], but
genotype or phenotype analysis of Apo E was not performed
and these studies cannot determine whether the APOE variants affect the lipoprotein particle response to long-term
exercise training. We used NMR spectroscopy to obtain a
detailed analysis of lipoprotein particle subpopulations and
to examine further the effects of the most common APOE
variants on the lipoprotein particle cholesterol responses to
exercise training [13].
Approximately 10% of the population have the APOE2/3
genotype whereas 20% have APOE4/3 [14]. The focused
study of APOE2/3 and 4/3 individuals through random sampling of the population yields few APOE2/3 or 4/3 subjects
thereby increasing the number of subjects required to obtain
sufficient sample sizes of the less frequent genotypes. This
is a major problem for exercise training studies, which are
labor intensive for both the subjects and investigators. Consequently, as described previously [13], we recruited subjects
based on APOE genotype to create a study cohort in which
the APOE2/3 and 4/3 genotypes were over-represented.
2. Methods
127
from 566 individuals screened for APOE genotype. Screening ceased when 50–60 subjects each with the less common
APOE2/3 and 4/3 genotypes – enough to statistically power
the study – were identified. Fifty-seven, 60, and 57 subjects
representing the APOE2/3, 3/3, and 3/4 genotypes, respectively, entered the study. Of these, 120 completed the 6-month
exercise program. In 14 subjects, samples were not obtained
either at the pre- or post-training timepoint, leaving complete
NMR-lipoprotein data for 106 subjects.
2.2. Subjects
Subjects were recruited if they were: healthy and without orthopedic problems, non-smokers, physically inactive,
between 18 and 70 years of age, had a body mass index (BMI)
≤31, and consumed less than two alcoholic beverages daily.
Subjects were considered physically inactive if they participated in vigorous activity less than four times/month for the
prior 6 months. Subjects underwent a medical history, physical exam, and a maximal exercise test to detect unreported
abnormalities and occult coronary artery disease. They were
reimbursed US$ 250 at the end of the study.
2.3. APOE genotype determination
DNA was extracted from leukocytes and APOE variants
determined using standard techniques [20].
2.4. Serum lipoprotein measurements
Serum was obtained after a 12 h fast before the start and
after 6 months of exercise training. Post-training samples
were obtained within 24 h of the penultimate and final exercise training session. Lipid levels in women before and after
training were obtained within 10 days of the onset of menses
to avoid variations in lipoprotein values [21]. Lipoprotein
subpopulation levels and mean particle size were determined
by NMR spectroscopy as previously described [1]. Concentrations of six VLDL subclasses (V1–V6), intermediate density lipoproteins (IDL), three LDL subclasses (L1–L3), and
five HDL subclasses (H1–H5) were determined. The L1–L3
subclasses were designated small, medium, and large LDL,
respectively. Large VLDL was the sum of the V5 + V6 fractions; medium VLDL, the sum of V3 + V4; small VLDL, the
sum of V1 + V2. Similarly, H1 + H2 were summed to give
small HDL-cholesterol and the sum of H3–H5 was taken as
large HDL-cholesterol. The lipid results in the present report
differ slightly from our prior report [13] which determined
lipids by enzymatic techniques. The present report presents
only NMR-derived lipid values. Total serum triglycerides are
determined enzymatically in both reports [13].
2.1. Study overview
2.5. Anthropometric measurements
This study was conducted by the Exercise and Genetics
Collaborative Research Group, a consortium of investigators at seven institutions* . Informed consent was obtained
Body weight and height were measured using balance
beam scales and wall mounted tape measures. Waist and hip
128
R.L. Seip et al. / Atherosclerosis 188 (2006) 126–133
girth measurements were taken using a cloth tape measure
[22].
cycles, cross-country ski machines, stair steppers, and rowing
machines.
2.6. Maximal exercise capacity
2.9. Exercise energy expenditure
Subjects underwent two pre- and one post-training maximal treadmill exercise tests. The first pre-training test was
designed to detect occult ischemia and to familiarize subjects with the measurement protocol, but was not used in
data analysis. The second pre-training test and the post-test
used the modified Astrand protocol [23]. Blood pressure and
12-lead electrocardiogram, as well as expired oxygen, carbon dioxide, and ventilatory volume were measured. Each
test site used its own metabolic measurement system for cardiorespiratory measurements and followed manufacturers’
calibration procedures. Maximal oxygen uptake was defined
as the average of the two highest consecutive 30-s values at
peak exercise.
Weekly exercise energy expenditure expressed as kilocalories per week was estimated from the average heart rates
recorded for exercise sessions of that week. From individual, pre-training maximal exercise test data plots of oxygen
uptake (VO2 ) versus heart rate, we estimated the VO2 corresponding to the exercise HR intensity and multiplied that
VO2 by bout duration (in minutes) to obtain total oxygen consumption for each bout. Each liter of oxygen was assumed to
represent 5 kcal of energy expenditure.
2.7. Dietary control procedures
Subjects were asked to maintain their usual dietary composition throughout the study. Dietary calories and composition
were assessed by random, 24-h dietary recall [24]. Trained
dieticians called the subjects by telephone on 1 weekday and
1 weekend day before the start and during the last month of
exercise training. Results from the two calls were averaged
to estimate dietary intake.
2.10. Data analysis
The main purpose of the study was to investigate whether
APOE affects LDL and HDL subpopulation responses to
exercise. Of secondary interest were gender interactive
effects. To test for these effects, 3 (genotype) × 2 (gender)
analyses of variance (ANOVA) were used for variables that
passed Levene’s test for equality of variances. In the absence
of a gender effect, the APOE genotype effect on change scores
was analyzed using analysis of covariance was employed
using baseline concentrations and change in triglycerides as
covariates (p ≤ 0.1). Post hoc tests (Fisher’s L.S.D.) were
employed when F ratios were significant. For hypothesis testing, significance levels were two-sided with alpha = 0.05.
2.8. Exercise program
Subjects underwent a 6-month, progressive, supervised
exercise program. Exercise session duration was increased
from 15 to 40 min during the first 4 weeks. Subjects exercised between 60 and 85% of their maximal exercise capacity
based on pre-training maximal heart rate. Once subjects could
perform 40 min of exercise, they continued this duration of
exercise 4 days a week for an additional 5 months. Subjects also participated in 5 min of warm-up and cool-down
so that each workout required 50 min. Subjects used a variety of exercise modalities including treadmills, stationary
3. Results
3.1. Baseline values
The numbers of men and women were similar across
genotypes (Table 1). There were no differences in baseline
anthropometric or exercise performance parameters among
the three APOE genotype groups [13]. Pre-training LDLcholesterol concentrations were higher in the APOE3/3 compared with the APOE2/3 subjects (Table 2). When adjusted
Table 1
Subject characteristics by APOE group
Variable
APOE genotype
2/3
3/3
4/3
N, male/female
Age (years)
BMI
Waist circumference
VO2 max (mL/(kg min))
14/21
38.9 ± 1.8
27.7 ± 0.8
35.2 ± 0.9
30.4 ± 1.3
21/19
40.0 ± 2.0
27.3 ± 0.8
34.9 ± 0.9
32.9 ± 1.2
16/15
37.6 ± 1.8
27.7 ± 0.9
35.7 ± 1.1
33.5 ± 1.4
Changes with training
BMI
Waist circumference
VO2 max (mL/(kg min))
−0.31 ± 0.18
−0.69 ± 0.45
4.3 ± 0.9
−0.36 ± 0.18
−0.36 ± 0.23
1.6 ± 0.5
−0.27 ± 0.14
−0.92 ± 0.31
4.2 ± 0.7
R.L. Seip et al. / Atherosclerosis 188 (2006) 126–133
129
Table 2
Baseline lipoprotein particle cholesterol and triglyceride concentrations (mg/dL) from NMR analyses, and change, before and after exercise training
(mean ± S.E.M.)
Baseline
Change with training
All
Total cholesterol
VLDL triglycerides
LDL-cholesterol
LDL total mass
HDL-cholesterol
212.1
111.5
131.0
1458
49.4
2/3
±
±
±
±
±
3.8
8.7
3.1
40
1.4
204.8
109.7
120.1aa
1344
50.9
3/3
±
±
±
±
±
6.4
13.0
4.6
70
2.3
220.7
118.7
137.8aa
1519
50.5
4/3
±
±
±
±
±
6.4
17.1
5.4
68
2.6
209.3
104.1
134.6
1506
46.5
All
±
±
±
±
±
6.6
13.7
5.4
68
2.2
6.0
−6.3
4.9
28
−0.5
2/3
±
±
±
±
±
3.1
5.3
2.5
30
0.8
2.4
−1.2
4.1
38
−0.4
3/3
±
±
±
±
±
5.0
9.3
4.1
40
1.3
7.7
−10.2
4.7
10
2.1
4/3
±
±
±
±
±
5.7
9.7
4.6
60
1.2
7.1
−7.2
6.0
70
−0.5
±
±
±
±
±
4.6
7.7
3.9
49
1.4
VLDL: very low density lipoprotein, IDL: intermediate density lipoprotein, LDL: low density lipoprotein, and HDL: high density lipoprotein. aa: Indicates
baseline LDL in APOE3/3 is different from in APOE2/3, p < 0.01, based on Fisher’s post hoc test.
Table 3
Baseline lipoprotein particle diameter (nm), and change, before and after exercise training (mean ± S.E.M.)
Particle category
Baseline
All
VLDL
LDL
HDL
‡
47.7 ± 1.2
20.92 ± 0.07
8.90 ± 0.04
Change with training
2/3
45.6 ± 2.1
20.98 ± 0.10
8.90 ± 0.08
3/3
49.7 ± 1.9
20.94 ± 0.11
8.94 ± 0.08
4/3
All
2/3
3/3
4/3
47.6 ± 2.2
20.84 ± 0.15
8.85 ± 0.08
−3.5‡
−1.7 ± 1.8
0.02 ± 0.08
−0.02 ± 0.04
−4.6 ± 1.6
0.14 ± 0.07
0.00 ± 0.04
−3.8 ± 1.9
0.07 ± 0.08
−0.04 ± 0.05
± 1.0
0.08 ± 0.04
−0.02 ± 0.02
Indicates significant pre- to post-exercise change for all subjects, based on F ratio for exercise effect, p < 0.001.
for the significant confounding effects of waist-to-hip ratio,
triglycerides, and percent body fat, baseline LDL-C levels
in Apo E2/3, E3/3, and E4/3 were 118.7 ± 5.0, 136.2 ± 4.8,
and 138.0 ± 5.2 mg/dL, respectively (p < 0.05, 2/3 versus 3/3
or 4/3). There were no other baseline differences in serum
lipids, lipoprotein classes, or particle size (Tables 2 and 3).
3.2. Effects of exercise training
VO2 max increased 10.1% with exercise in the 106 subjects. As presented in detail elsewhere [13], the increase in
VO2 max was significantly less in the APOE3/3 subjects than
in the other two APOE groups. Exercise energy expenditure averaged 14.5 ± 1.0 kcal/kg per week during the last 2
weeks of training and was not different across groups. In
the 106 subjects studied presently, body weight decreased
1.0 ± 0.3 kg (p < 0.001), BMI fell 0.3 ± 0.1 units (p < 0.003),
and waist circumference fell 0.63 ± 0.19 in. (p < 0.055) overall as before [13]. There were no other significant changes in
anthropometric, exercise, or dietary parameters with exercise
training and no differences in anthropometric changes among
the Apo E genotype groups.
Serum lipids for the entire cohort were altered little by
exercise training, although triglycerides in the 106 studied here decreased 14% (136.0 ± 89.9–118.9 ± 7.9 mg/dL)
(p < 0.0003).
LDL particle subpopulations changed significantly with
exercise training and the changes were dependent on APOE
genotype (Fig. 1). Large VLDL decreased with exercise in
all subjects (F ratio = 9.22, p < 0.01) with no genotype effect
for any VLDL subclass response (Fig. 1, Panel A). Large
VLDL also decreased in men (p < 0.05, not shown), and
tended to decrease in women (p < 0.17, not shown). Mediumsized LDL increased whereas small-sized LDL-cholesterol
decreased in APOE3 homozygotes. These changes were
directionally different from responses in the APOE2/3 and
3/4 subjects (Fig. 1, Panel B) and statistically significantly
different from APOE2/3 (p < 0.01) and 3/4 for medium
LDL (p < 0.05), and from APOE2/3 and 3/4 for small LDL
(p < 0.05) (Fig. 1, Panel B). The training change in large HDL
was directionally different in APOE3/3 versus APOE2/3 and
3/4 but this was not statistically significant (Fig. 1, Panel C,
p < 0.18).
The change for each LDL subpopulation was significantly related to the respective baseline concentration (−0.41
to −0.48, p < 0.00001), but poorly related to changes in
BMI, waist circumference, and percent body fat (Table 4).
Baseline values for BMI, waist circumference, and percent
body fat were unrelated to LDL subclasses (not shown).
Fig. 1 presents estimated means statistically adjusted for
the baseline small, medium, or large LDL concentration
and the change in log triglycerides. The adjusted means
(±S.E.M.) for APOE2/3, 3/3, and 4/3 groups were 9.9 ± 6.4,
7.3 ± 6.0, and 8.4 ± 6.7 mg/dL, respectively, for large LDLcholesterol (APOE group effect, F ratio = 0.05, p NS);
−14.9 ± 4.8, 9.0 ± 4.5, and −6.8 ± 5.1 mg/dL for medium
LDL (F ratio = 6.91, p < 0.002); 5.0 ± 4.7, −8.4 ± 4.3, and
6.6 ± 4.9 mg/dL for small LDL (F ratio = 3.35, p < 0.039).
The APOE effect on LDL particle subclasses was generally
similar across gender (data not shown).
Neither training nor APOE affected overall LDL or HDL
diameter (Table 3). Training decreased VLDL diameter by
3.5 nm overall (p < 0.01) due entirely to decreases in large
VLDL in each group (Fig. 1, Panel A).
130
R.L. Seip et al. / Atherosclerosis 188 (2006) 126–133
cally, exercise training in APOE3/3 subjects decreased small
LDL particle cholesterol and raised medium LDL particle
cholesterol concentrations, whereas the opposite changes in
these subfractions were observed in the APOE2/3 and 4/3
subjects. This response was true for the entire cohort and the
pattern existed in both men and women.
Several studies have examined the effect of exercise
on LDL particle size but without regard to APOE variant. Williams et al. [15] used analytic ultracentrifugation to
demonstrate lower small LDL and higher large HDL mass in
runners compared to sedentary men. In another study, LDL
mean particle size decreased slightly in 46 previously sedentary men after 1 year of exercise training and the decrease
in small LDL was inversely related to running mileage [17].
Using NMR spectroscopy to measure LDL small, medium
and large subpopulations, Kraus et al. [18] and Kang et al.
[19] confirmed that exercise increased the mean diameter of
LDL by decreasing the small LDL subclass concentration
[18,19].
Small dense LDL particles appear more atherogenic than
other LDL subpopulations [25]. Though the seven subclasses
of LDL determined by gradient gel electrophoresis (GGE) do
not directly correspond to the LDL subpopulations measured
by NMR spectroscopy, it is reasonable to assume both methods measure cholesterol native to atherogenic small LDL
particles. Both NMR spectroscopy and GGE detect clinically
relevant changes in LDL subpopulations in response to pharmacologic treatments [26].3
How APOE influences the LDL subpopulation response to
exercise training is unclear, but must result from differences
in clearance and/or production of the individual particles.
Apo E is found on both triglyceride-rich and HDL particles.
The Apo E2, E3 and E4 isoforms differentially affect the
rate of hepatic clearance of Apo E-containing chylomicrons,
VLDL, and IDL, through their interactions with the LDL
family of receptors [7]. Apo E has 299 amino acids with a
binding region between amino acids 140 and 160. Apo E3,
the most common or “wild-type” Apo E, contains a cysteine
at amino acid 112 and an arginine within the binding region
at amino acid 158 [10]. The E4 variant contains an arginine
at the 112 site and has normal or enhanced binding to the
LDL (B/E) receptor. The E2 variant contains a cysteine substitution within the binding region at amino acid 158 and has
only 1–2% of normal binding capacity reducing clearance of
Apo E2 containing particles [10]. Apo E3 and E4 protein iso-
4. Discussion
The present study is to our knowledge the first to demonstrate that the effect of aerobic exercise training on lipoprotein
subpopulations varies by common APOE variant. Specifi-
Fig. 1. Changes in VLDL (A), LDL (B), and HDL particle subpopulations (C) for all subjects. Data are means ± S.E.M. Large VLDL decreased
with exercise († p < 0.01, Panel A) with no differences among APOE groups
for changes in any VLDL subpopulation. For LDL, † indicates APOE3/3
different from APOE2/3 for medium-sized LDL (adjusted for baseline concentration and change in log triglycerides), p < 0.01; * indicates APOE3/3
different from APOE4/3 for both medium and small particles (small LDL
adjusted for baseline concentration and change in percent fat), p < 0.05. For
HDL, there were no significant differences among APOE genotypes for
changes in any category, but APOE3/3 tended to differ from APOE2/3 and
3/4 in the change in large HDL (p < 0.18).
R.L. Seip et al. / Atherosclerosis 188 (2006) 126–133
131
Table 4
Correlations between changes in LDL subpopulations and their respective baseline concentrations and with exercise-associated changes in factors known to
affect LDL size
Variable
Change in LDL size (nm)
Change in LDL subpopulation (mg/dL)
Large
Medium
Small
Baseline valuea
Age
Change, log TG
Change, weight
Change, %fat
Change, waist circumference
Change, BMI
−0.407‡‡
0.031
−0.420‡
0.001
−0.050
−0.111
−0.005
−0.408‡‡
−0.007
−0.367‡
0.160o
0.006
0.024
0.138
−0.476‡‡
−0.068
0.155
−0.130
−0.088
−0.006
−0.123
−0.431‡‡
0.026
0.155
0.119
0.112
0.118
0.117
−0.014
−0.312‡
0.196*
−0.005
0.036
0.182o
−0.012
0.149
−0.058
−0.085
−0.007
−0.069
Partial correlations (effect of baseline value removed)
Age
−0.021
Change, log TG
−0.352‡
Change, weight
0.060
Change, %fat
−0.053
Change, waist circumference
−0.100
Change, BMI
0.057
0.019
0.122
0.054
0.124
0.116
0.061
Note: The partial correlation coefficients demonstrate that correcting for baseline particle size does not importantly affect the results.
a Baseline LDL medium and LDL small concentrations were normalized using log[1/(1 + baseline value)] transformation.
o p < 0.07.
* p < 0.05.
‡ p < 0.01.
‡‡ p < 0.00001.
forms also bind with higher affinity to the lipoprotein receptor
like protein than does Apo E2 [27]. The higher affinity of E3
and E4 and their more rapid clearance increases the hepatic cholesterol pool and downregulates the hepatic receptor
population, producing higher fasting LDL-cholesterol concentrations in E3 and E4 compared to E2 [14]. In addition,
the in vivo clearance of E4 is greater than E3 [28] due to the
preferential association of E4 with VLDL over HDL [29].
The clearance of VLDL-cholesterol by this mechanism is
thought to further downregulate LDL receptors in E4 carriers [28] and explain the 5–10% higher LDL concentrations
in Apo E4/3 subjects compared to Apo E3/3 [14]. A similar pattern, though not statistically significant, could be seen
in the present study. Correction of baseline LDL-C for the
confounding effects of age and log-transformed triglycerides
lowered LDL-C to 136.2 ± 4.8 in APOE3/3 and raised LDLC to 138.0 ± 5.0 mg/dL in 4/3.
Exercise training reduces serum triglycerides in part by
accelerating the lipolysis of VLDL [30]. In the present study,
exercise significantly decreased large VLDL by 5–20 mg/dL
across genotypes, consistent with increased delipidation due
to lipoprotein lipase activity [31] and increasing lipoprotein
precursor for IDL and LDL formation.
We can only speculate as to how small LDL increased with
exercise training in both APOE4 and 2 allele variants in the
present study. In the absence of exercise, kinetic studies have
shown that the rate of VLDL to LDL conversion is increased
and the LDL catabolic rate is decreased in Apo E4/3 [8].
Consequently, with exercise VLDL lipolysis increases, and
we speculate that for APOE4, the increased production and
the extended residence time of LDL permits greater formation
of small, dense LDL from medium LDL [25,32].
In APOE2 subjects, the conversion rate of VLDL to LDL
is compromised [8] and another explanation for the increase
in small LDL is required. The rise in small LDL may be
a consequence of both the low affinity of Apo E2 for hepatic receptors [7] and the increased lipid transfer between
VLDL and LDL particles leading to the formation of small
dense LDL [33]. In the present study, APOE2/3 had slightly
higher baseline VLDL triglycerides (Table 2) and slightly
smaller VLDL particles (Table 3), consistent with an increase
in overall VLDL particle number. Increased VLDL favors formation of small dense LDL through the combined actions of
cholesteryl ester transfer protein (CETP) and hepatic lipase.
CETP facilitates the exchange of VLDL triglycerides to LDL
and LDL-cholesteryl esters to VLDL, and hepatic lipase
lipolyzes LDL-triglycerides, reducing LDL size [33]. The
increased plasma VLDL residence time in Apo E2 subjects
[18] provides more time for the modification of LDL particle
composition and the formation of small LDL. Although very
little of the total plasma Apo E is found in LDL [21], the
smallest, most dense LDL fraction is enriched in Apo E [34].
By increasing lipolysis, exercise might accelerate the production and subsequent accumulation of these poorly cleared
small LDL-Apo E2. Finally, some small LDL derive directly
from specific precursors [35] and direct conversion of small
VLDL to small LDL may be preferred in VLDL-Apo E2containing particles, but this is speculative.
We have previously reported increases in HDL-cholesterol
(determined enzymatically) with exercise training only in the
APOE3/3 subjects [13]. NMR determined changes in total
HDL-cholesterol and HDL particle distribution were not significantly different among the genotype groups in the present
analysis. Nevertheless, the present results suggest that these
132
R.L. Seip et al. / Atherosclerosis 188 (2006) 126–133
changes are produced by increases in large HDL particles
in the APOE3/3 and decreases in large HDL particles in
the APOE2/3 and 4/3 subjects. Changes in serum lipids for
the entire cohort in the present and prior report are indeed
small compared to those expected from exercise training.
A meta-analysis of 52 exercise training trials of >12 weeks
duration including 4700 subjects demonstrated an average
increase in HDL-cholesterol levels of 4.6% and reductions in
triglycerides and LDL-cholesterol concentrations of 3.7 and
5.0%, respectively [36]. In the present study NMR analysis indicated a non-significant change in HDL-cholesterol of
−1.0% for the entire cohort whereas triglycerides decreased
4.9% and LDL-cholesterol increased 3.6%. The paucity of
change in the present report is undoubtedly due in part to
the composition of the study population. APOE2/3, 3/3, and
4/3 subjects comprise 10, 62, and 22% of the general population, respectively [14]. The present study was designed
to include equal numbers of these three genotypes. Consequently, differences in the response among the APOE groups
magnified or diminished the response for the entire cohort.
For example, HDL-cholesterol increased 2.1 mg/dL or 4.2%
in APOE3/3, a response similar to the 4.6% observed in unselected subjects [37] but decreased 0.4 and 0.5 mg/dL, or
0.8 and 1.1%, respectively, in the APOE heterozygotes. The
heterozygous subjects exerted a larger effect on the average HDL-cholesterol response than normally observed in
unselected population-based studies because they were intentionally over-represented in the present sample. This may also
explain the increase in total cholesterol with exercise training
in the whole cohort.
There are several limitations to the present study. First, we
purposely sought carriers of the APOE2 or 4 alleles in order to
enlarge the APOE2/3 and 4/3 groups, following the premise
that functional differences in the protein products of the
APOE2, 3, and 4 alleles affect lipid metabolic responses. That
Apo E protein variants display many profoundly different
and physiologically meaningful biochemical behaviors justifies our approach. Nevertheless, we cannot exclude explanations related to possible underlying but unknown genetic
differences. Grouping subjects by APOE variants does not
guarantee genetic homogeneity within the 2/3, 3/3, and 4/3
groupings, because coding specificity does not generate identical APOE haplotypes [32,33,38]. In fact, within the classical
APOE2, 3, and 4 alleles, 18 additional sites of variation exist
at other exons or introns [34,39] that may obscure or override
some metabolic tendencies associated with the functionality
of the Apo E2, E3, and E4 proteins.
Perhaps even more importantly, there is no guarantee of
equivalent genetic heterogeneity elsewhere in the genome,
across the APOE groupings. There may be different degrees
of linkage disequilibrium between the individual APOE2,
3, and 4 alleles and other gene variants [40] to explain the
present findings. A more encompassing genomic analytic
strategy, based on genome-wide haplotyping that simultaneously accounts for genetic variation both within and without the APOE alleles studied here, is evolving [41] and
may demonstrate differences within the commonly accepted
APOE variants. As genes that determine variation in the
exercise response become known, genetic variability may be
found that outweighs the physiological consequences of the
polymorphisms studied here.
Other factors limit the generalizability of our results. The
study population consisted of relatively healthy adults. Subjects with BMI > 31 were excluded, as were individuals with
hyperlipidemia. It is possible that in metabolic disease states
such as metabolic syndrome, the ability of common APOE
variants to affect exercise-associated LDL size changes is
different.
In summary, we have demonstrated an APOE allelespecific LDL subpopulation response to exercise in healthy
subjects free of metabolic disease. In APOE3/3 homozygotes, medium LDL particles increased in concentration and
small LDL particles decreased. The opposite was observed
in APOE2/3 and 4/3 heterozygotes, suggesting that exercise
training produces the potentially most beneficial changes in
LDL particle size only in APOE3/3 individuals The NMR
findings extend our previous observation that the LDLto-HDL-cholesterol ratio improves most with exercise in
APOE3/3 subjects [13]. These findings are clinically relevant
since they help explain the variability in lipid and lipoprotein responses induced by exercise training [42]. The results
also demonstrate that the lipid response to exercise, like the
response to pharmacologic therapies, differs by genetic composition. Consequently, exercise genetic profiling or “exercisegenomics” may ultimately help predict population subgroups most likely to reduce their cardiovascular risk with
exercise training.
References
[1] Rosenson RS, Otvos JD, Freedman DS. Relations of lipoprotein
subclass levels and low-density lipoprotein size to progression of
coronary artery disease in the Pravastatin Limitation of Atherosclerosis in the Coronary Arteries (PLAC-I) trial. Am J Cardiol 2002;90:
89.
[2] Kuller L, Arnold A, Tracy R, et al. Nuclear magnetic resonance
spectroscopy of lipoproteins and risk of coronary heart disease
in the cardiovascular health study. Arterioscler Thromb Vasc Biol
2002;22:1175.
[3] Blake GJ, Otvos JD, Rifai N, Ridker PM. Low-density lipoprotein
particle concentration and size as determined by nuclear magnetic
resonance spectroscopy as predictors of cardiovascular disease in
women. CIRC 2002;106:1930.
[4] Soedamah-Muthu SS, Chang YF, Otvos J, Evans RW, Orchard TJ.
Lipoprotein subclass measurements by nuclear magnetic resonance
spectroscopy improve the prediction of coronary artery disease in
Type 1 diabetes. A prospective report from the Pittsburgh Epidemiology of Diabetes Complications Study. Diabetologia 2003;46:674.
[5] Otvos JD. Measurement of lipoprotein subclass profiles by nuclear
magnetic resonance spectroscopy. Clin Lab 2002;48:171.
[6] Williams PT, Superko HR, Haskell WL, et al. Smallest LDL particles
are most strongly related to coronary disease progression in men.
Arterioscler Thromb Vasc Biol 2003;23:314.
[7] Weisgraber KH, Innerarity TL, Rall Jr SC, Mahley RW. Atherogenic
lipoproteins resulting from genetic defects of apolipoproteins B and
E. Ann N Y Acad Sci 1990;598:37.
R.L. Seip et al. / Atherosclerosis 188 (2006) 126–133
[8] Demant T, Bedford D, Packard CJ, Shepherd J. Influence of
apolipoprotein E polymorphism on apolipoprotein B-100 metabolism
in normolipemic subjects. J Clin Invest 1991;88:1490.
[9] Berneis KK, Krauss RM. Metabolic origins and clinical significance
of LDL heterogeneity. J Lipid Res 2002;43:1363.
[10] Mahley RW, Huang Y. Apolipoprotein E: from atherosclerosis to
Alzheimer’s disease and beyond. Curr Opin Lipidol 1999;10:207.
[11] Thuren T, Weisgraber KH, Sisson P, Waite M. Role of apolipoprotein
E in hepatic lipase catalyzed hydrolysis of phospholipid in highdensity lipoproteins. Biochemistry 1992;31:2332.
[12] de Man FH, de Beer F, van de LA, Smelt AH, Leuven JA, Havekes
LM. Effect of apolipoprotein E variants on lipolysis of very low density lipoproteins by heparan sulphate proteoglycan-bound lipoprotein
lipase. Atherosclerosis 1998;136:255.
[13] Thompson PD, Tsongalis GJ, Seip RL, et al. Apolipoprotein E genotype and changes in serum lipids and maximal oxygen uptake with
exercise training. Metabolism 2004;53:193.
[14] Dallongeville J, Lussier-Cacan S, Davignon J. Modulation of plasma
triglyceride levels by apoE phenotype: a meta-analysis. J Lipid Res
1992;33:447.
[15] Williams PT, Krauss RM, Wood PD, Lindgren FT, Giotas C,
Vranizan KM. Lipoprotein subfractions of runners and sedentary
men. Metabolism 1986;35:45.
[16] Yu HH, Ginsburg GS, O’Toole ML, Otvos JD, Douglas PS, Rifai N.
Acute changes in serum lipids and lipoprotein subclasses in triathletes as assessed by proton nuclear magnetic resonance spectroscopy.
Arterioscler Thromb Vasc Biol 1999;19:1945.
[17] Williams PT, Krauss RM, Vranizan KM, Albers JJ, Terry RB, Wood
PDS. Effects of exercise-induced weight loss on low density lipoprotein subfractions in healthy men. Arteriosclerosis 1989;9:623.
[18] Kraus WE, Houmard JA, Duscha BD, et al. Effects of the amount
and intensity of exercise on plasma lipoproteins. N Engl J Med
2002;347:1483.
[19] Kang HS, Gutin B, Barbeau P, et al. Physical training improves
insulin resistance syndrome markers in obese adolescents. Med Sci
Sports Exerc 2002;34:1920.
[20] Richard P, Thomas G, de Zulueta MP, et al. Common and rare genotypes of human apolipoprotein E determined by specific restriction
profiles of polymerase chain reaction-amplified DNA. Clin Chem
1994;40:24.
[21] Cullinane EM, Yurgalevitch SM, Saritelli AL, Herbert PN, Thompson PD. Variations in plasma volume affect total and low-density
lipoprotein cholesterol concentrations during the menstrual cycle.
Metabolism 1995;44:965.
[22] Callaway CW, Chumlea WC, Bouchard C, et al. Circumferences.
Anthropometric standardization reference manual. Champaign, IL.
Hum Kinet 1988:39–53.
[23] Pollock ML. Exercise in health and disease: evaluation and prescription for prevention and rehabilitation. Philadelphia: Saunders; 2003.
[24] Nutrient Data System. Minneapolis, MN: University of Minnesota;
2002.
[25] Lada AT, Rudel LL. Associations of low density lipoprotein particle
composition with atherogenicity. Curr Opin Lipidol 2004;15:19.
133
[26] Brown BG, Morse JS, Zambon A, Otvos JD, Lanman RB, Krauss
RM. Frequency, importance, and conversion of large/buoyant LDL
type during treatment with simvastatin plus niacin or their placebos:
comparison of four methods. CIRC 2003;108:IV-197 [abstract].
[27] Kowal RC, Herz J, Weisgraber KH, Mahley RW, Brown MS, Goldstein JL. Opposing effects of apolipoproteins E and C on lipoprotein
binding to low density lipoprotein receptor-related protein. J Biol
Chem 1990;265:10771.
[28] Utermann G. Apolipoprotein E polymorphism in health and disease.
Am Heart J 1987;113:433.
[29] Rubinsztein DC. Apolipoprotein E: a review of its roles in lipoprotein
metabolism, neuronal growth and repair and as a risk factor for
Alzheimer’s disease. Psychol Med 1995;25:223.
[30] Thompson P. What do muscles have to do with lipoproteins? CIRC
1990;81:1428.
[31] Seip RL, Angelopoulos TJ, Semenkovich CF. Exercise induces
human lipoprotein lipase gene expression in skeletal muscle but not
adipose tissue. Am J Physiol 1995;268:E229.
[32] Packard CJ, Demant T, Stewart JP, et al. Apolipoprotein B
metabolism and the distribution of VLDL and LDL subfractions.
J Lipid Res 2000;41:305.
[33] Lagrost L, Gambert P, Lallemant C. Combined effects of lipid transfers and lipolysis on gradient gel patterns of human plasma LDL.
Arterioscler Thromb 1994;14:1327.
[34] Chapman MJ, Laplaud PM, Luc G, et al. Further resolution of the
low density lipoprotein spectrum in normal human plasma: physicochemical characteristics of discrete subspecies separated by density
gradient ultracentrifugation. J Lipid Res 1988;29:442.
[35] Musliner TA, Krauss RM. Lipoprotein subspecies and risk of coronary disease. Clin Chem 1988;34:78.
[36] Tran ZV, Weltman A. Differential effects of exercise on serum lipid
and lipoprotein levels seen with changes in body weight. A metaanalysis. JAMA 1985;254:919.
[37] Leon AS, Gaskill SE, Rice T, et al. Variability in the response
of HDL cholesterol to exercise training in the HERITAGE Family
Study. Int J Sports Med 2002;23:1.
[38] Nickerson DA, Taylor SL, Fullerton SM, et al. Sequence diversity
and large-scale typing of SNPs in the human apolipoprotein E gene.
Genome Res 2000;10:1532.
[39] Stengard JH, Clark AG, Weiss KM, et al. Contributions of 18
additional DNA sequence variations in the gene encoding apolipoprotein E to explaining variation in quantitative measures of lipid
metabolism. Am J Hum Genet 2002;71:501.
[40] Hubacek JA, Pitha J, Adamkova V, Skodova Z, Lanska V, Poledne
R. Apolipoprotein E and apolipoprotein CI polymorphisms in the
Czech population: almost complete linkage disequilibrium of the
less frequent alleles of both polymorphisms. Physiol Res 2003;52:
195.
[41] Stephens JC, Schneider JA, Tanguay DA, et al. Haplotype variation and linkage disequilibrium in 313 human genes. Science
2001;293:489–93.
[42] Bouchard C, Rankinen T. Individual differences in response to regular physical activity. Med Sci Sports Exerc 2001;31:S446.