Donate Help Contact The AHA Sign In Home
American Heart Association
Arteriosclerosis, Thrombosis, and Vascular Biology
Search: search_blue_button Advanced Search
« Previous Article | Table of Contents | Next Article »
Arteriosclerosis, Thrombosis, and Vascular Biology. 2002;22:133-140
doi: 10.1161/hq0102.101819
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bernstein, M. S.
Right arrow Articles by Morabia, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bernstein, M. S.
Right arrow Articles by Morabia, A.
Right arrowPubmed/NCBI databases
*Gene*GEO Profiles
*HomoloGene*UniGene
*Compound via MeSH
*Substance via MeSH
Medline Plus Health Information
*Exercise for Children
*Exercise for Seniors
*Exercise and Physical Fitness
Related Collections
Right arrow Acute coronary syndromes
Right arrow Clinical Studies
Right arrow Behavioral/psychosocial - stroke
Right arrow CT and MRI
(Arteriosclerosis, Thrombosis, and Vascular Biology. 2002;22:133.)
© 2002 American Heart Association, Inc.

Atherosclerosis and Lipoproteins

Physical Activity May Modulate Effects of ApoE Genotype on Lipid Profile

Martine S. Bernstein; Michael C. Costanza; Richard W. James; Michael A. Morris; François Cambien; Ségolène Raoux; Alfredo Morabia

From the Division d’Epidémiologie Clinique (M.S.B., M.C.C., A.M.), Hôpitaux Universitaires de Genève, the Lipid Laboratory (R.W.J.), Clinical Diabetes Unit, Geneva University Hospital, and the Division de Génétique Médicale (M.A.M.), Centre Médical Universitaire, Geneva, Switzerland, and INSERM U525 (F.C., S.R.), Faculté de Médecine Pitié-Salpétrière, Paris, France.

Reprint requests to Dr Martine S. Bernstein, Division d’Epidémiologie Clinique, Hôpitaux Universitaires de Genève, 25 rue Micheli-du-Crest, 1211 Geneva 14, Switzerland. E-mail martine.bernstein{at}hcuge.ch


*    Abstract
up arrowTop
 *Abstract
down arrowIntroduction
 down arrowMethods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Increased levels of physical activity may improve the lipid profile, but is this effect identical across apolipoprotein E (apoE) genotypes? A population-based cross-sectional survey conducted from 1999 to 2000 included 1708 randomly selected men and women aged 35 to 74 years. A validated physical activity questionnaire measured, for each participant, the total energy expenditure and its percentage used in high-intensity activities (%high-intensity activity), eg, brisk walking and sports. The effects of the apoEx%high-intensity activity interaction on the lipid profile were investigated by using multiple linear regression models. Among men, increased %high-intensity activity had greater protective effects in the apoE4 group regarding (1) high density lipoprotein (HDL) cholesterol (P<0.001), compared with either the apoE2 (interaction P=0.05) or apoE3 (interaction P<0.03) groups, and (2) triglycerides (P<0.03), compared with the apoE3 group (interaction P=0.07). A 10% increase of %high-intensity activity by an apoE4 man would correspond with a 0.07-mmol/L increase in HDL cholesterol and a -0.15-mmol/L decrease in triglycerides. Among women, only the protective effects of physical activity on HDL cholesterol in the apoE4 group versus the apoE2 group was statistically significant. Spending a larger fraction of the total energy expenditure in high-intensity activities may counteract the atherogenic effects of the {epsilon}4 allele on the lipid profile.


Key Words: apolipoprotein E • exercise • cholesterol • triglycerides • energy expenditure


*    Introduction
up arrowTop
 up arrowAbstract
 *Introduction
down arrowMethods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
The pathogenesis of atherosclerosis depends on the interacting effects of genetic and environmental factors.1–5 Increased levels of physical activity lower total and LDL cholesterol6–10 and increase HDL cholesterol.6,11 The plasma concentrations of lipoproteins are also modulated by apoE, which is on the surface of these lipid-rich particles. It binds with high affinity to receptors in the liver and extrahepatic cells, mediating the uptake of lipoproteins.12 The structural gene locus for plasma apoE is polymorphic,13 with 3 common alleles, designated {epsilon}2, {epsilon}3, and {epsilon}4, which determine 6 genotypes: 3 homozygous ({epsilon}2{epsilon}2, {epsilon}3{epsilon}3, and {epsilon}4{epsilon}4) and 3 heterozygous {epsilon}2{epsilon}3, {epsilon}2{epsilon}4, and {epsilon}3{epsilon}4), expressed in a codominant fashion. There are indications that the {epsilon}4 allele may be an independent risk factor for cardiovascular diseases that is associated with a more atherogenic profile, ie, higher total cholesterol, triglyceride, and LDL cholesterol levels than those observed in subjects harboring the {epsilon}2 or {epsilon}3 allele.12,14–16 A growing body of evidence suggests a protective antiatherogenic role for the {epsilon}2 allele in the general population, except under particular conditions.17

It is biologically plausible that the effects of physical activity and the apoE gene interact with respect to the levels of serum total cholesterol, HDL cholesterol, triglycerides, and LDL cholesterol. A beneficial effect of physical activity in apoE subgroups has been shown in children and young adults18,19 and in exercise training intervention.20 However, no population-based study has been reported that assessed whether, among adults, physical activity modified the association of apoE genotypes or phenotypes with lipid profile levels. Therefore, the purpose of the present study was to determine whether, in a general adult population, lipid profile levels observed across apoE genotype groups differed between more sedentary versus more active subjects. Energy expenditures (EEs), total and activity specific, were measured by using a self-administered questionnaire that was developed and validated in the target population.


*    Methods
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 *Methods
down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Population and Sample Recruitment
Geneva (city and county, Switzerland) has a population of {approx}400 000 distributed over 242 km of land. Survey participants were asked to participate in a general population health survey. They were randomly selected uniformly from January 1999 to December 2000 to represent the 190 000 male and female noninstitutionalized residents aged 35 to 74 years. Approximately 0.5% of the target population was sampled each year from the official residents’ register. The intensive and standardized recruitment of a potential subject lasted from 2 weeks to 2 months.21

Data Collection
At home, survey respondents completed self-administered questionnaires about physical activity, diet, sociodemographic data, and smoking history. They returned the questionnaires to a mobile epidemiology clinic, where trained interviewers checked them for completion and measured their weight (lightly dressed without shoes) with a medical scale (precision 0.5 kg) and their height with a medical gauge (precision 1 cm). Overall, the clinic visit lasted {approx}20 minutes. Diet was recorded by using a semiquantitative food-frequency questionnaire previously developed and validated in the Geneva general population.22,23

Physical Activity Assessment
The physical activity–frequency questionnaire measures the total and the activity-specific EEs. It was developed in the Geneva general adult population and validated by using a heart rate monitor.24 It lists 70 physical activities grouped by general type (eg, occupational, housework, leisure time, and sports). The reference period was the past 7 days. Approximately 20 minutes is required for respondents to indicate the number of days (0 to 7) and the duration per day (0 to 10 hours, in 15-minute increments) they spent in performing each activity. Bedtime and waking hours are also reported. The duration for each physical activity is proportionally adjusted so that the total number of hours for sleep plus adjusted durations of physical activities equals 168 hours (7x24 hours).

The sex-, age-, and weight/height-specific basal metabolic rate (BMR), which is the energy expended by a fasting individual, resting supine, was calculated for each participant.25 Each physical activity was previously assigned a score indicating its intensity in terms of BMR multiples.25 For example, when sleeping, an individual expends 1 time the energy required by the BMR (1 BMR); the same individual performing an activity with a 3 BMR rating expends 3 times the BMR. The daily EE for the ith specific physical activity, EEi, for i, i=1, ..., 70, for a given study participant was calculated as follows: EEi=(BMR multiple for ith activity) x[(proportionally adjusted) duration of ith activity (min/d)] x[participant BMR (kcal/min)].

For sleep, EE71=1x[sleep duration (min/day)]x[participant BMR (kcal/min)]. The daily total EE (EETotal) for the study participant was then obtained as the sum of the study participant’s EEi, i=1,...,70, +EE71: Down


The physical activities were also grouped according to their intensity. For example, moderate intensity activities expending 3 to 3.9 times the BMR included most housekeeping chores, walking normally, and bicycling slowly ({approx}9 km/h). High-intensity and very high–intensity activities expending >=4 times the BMR included walking quickly or uphill (4.5 BMR), climbing stairs (6 BMR), and sports and heavy construction work (>6 BMR). Because the prevalence of performers of >6 BMR physical activities (eg, sports and heavy construction work) was very small, these very high-intensity activities were included with the >=4 BMR physical activities.

Lipid Profile Measurements
A fasting blood sample (10 mL) was collected between 7:00 and 10:00 AM in a hemoguard tube with EDTA, stored in a special suitcase at 4°C, and taken within 2 hours to the laboratory, where it was centrifuged (3000g, 30 minutes). Differential precipitation was performed on 2x0.4 mL aliquots. Total plasma cholesterol, triglycerides, and their concentrations in the lipoprotein subfractions were assayed immediately in an aliquot of the supernatant by using commercially available enzymatic kits (Bayer Technicon Diagnostics). Quality control of lipid measurements was ensured via the Centers for Disease Control and Prevention National Reference System/cholesterol protocol. HDL cholesterol was measured in fresh plasma after precipitation of lower density lipoproteins.26 LDL cholesterol was calculated as LDL=(total cholesterol-HDL-VLDL), where VLDL=triglycerides/2.2.27

DNA Analyses
For DNA extractions, 4 mL EDTA blood was collected from all consenting participants and stored at 4°C for up to 10 days. Total genomic DNA was extracted (Puregene Blood kit, Gentra Systems). The average yield was 30 µg/mL. DNA samples were stored in 10 mmol/L Tris/0.1 mmol/L EDTA at 4°C. The sequence of the apoE gene and of its protein are well known, and the complete DNA and mRNA nucleotide sequences are available on public databases (http://www.ncbi.nlm.nih.gov). The isoprotein apoE2 is a variant differing from the most common isoprotein, apoE3, by a single amino acid change (Arg158Cys), which leads to a much lower affinity for the cellular apolipoprotein receptor. The polymorphisms were assessed at the laboratory of INSERM-SC7/U525 in Paris. A description of the analytic techniques can be found online at http://genecanvas.idf.inserm.fr/.

Statistical Analyses
Except for estimating the population distribution of the apoE {epsilon}2, {epsilon}3, and {epsilon}4 allele frequencies, we excluded from further statistical analyses the following: (1) study participants with the genotype {epsilon}2{epsilon}4 (n=22) because of the potentially contradictory effects of these alleles on the lipid profile, which could not be thoroughly investigated with such a small sample, and (2) study participants taking cholesterol-lowering drugs (n=158). The apoE2 group included genotypes {epsilon}2{epsilon}2 (<1%) and {epsilon}2{epsilon}3; the apoE4 group included genotypes {epsilon}4{epsilon}4 ({approx}2%) and {epsilon}3{epsilon}4. The apoE3 group included only the homozygous genotype {epsilon}3{epsilon}3.

For descriptive purposes, we grouped age (35 to 44, 45 to 54, 55 to 64, and 65 to 74 years), education (primary [<9 years of school], secondary, and university [>=13 years and Swiss baccalaureate]), country of birth (Switzerland, Mediterranean [Italy, Spain, and Portugal], and others [mainly France and Germany]), and cigarette smoking (never [<100 cigarettes over lifetime], exsmoker [quit >=1 year before interview], and current [>=100 cigarettes over lifetime]). Body mass index (BMI) was computed as weight (in kilograms) divided by height (in square meters). Normal weight, overweight, and obesity were defined, respectively, as follows (kg/m2): BMI<25, 25<=BMI<30, and BMI>=30. Total daily EE, percentage of total daily EE expended in >=4 BMR physical activities (%high-intensity activity), and dietary fat and fiber were grouped into sex-specific tertile groups.

Multiple linear regression analyses of each lipid profile value were performed for men and women separately. For adjustment purposes, the potential confounders (age, BMI, percent dietary fat, and dietary fiber [in grams]) were treated as continuous, whereas each of the other grouped potential confounders (education, country of birth, and smoking) was recoded into 2 dummy (0/1) variables. The main independent variables were apoE group, also recoded into 2 dummy (0/1) variables (reference group, apoE4), and %high-intensity activity, all of which were treated as continuous. Two product terms reflecting the (apoEx%high-intensity activity) interaction effects were also included in the models. Two-dimensional plots of the multiple linear regressions of the lipid profile values on %high-intensity activity for each apoE group were adjusted for all the confounders and interaction effects. Main effects were tested via F tests for slopes, and interactions were tested via F tests for differences between slopes, with the apoE4 group being the reference.

The distribution of %high-intensity activity was right-skewed. However, because the linear regression results were quantitatively and qualitatively similar for the raw and transformed variable log(%high-intensity activity+1), only the results based on the raw variable are shown. The linear regressions were also visually compared with robust scatterplot smoothers.28 The regression lines captured sufficient trend details; hence, only these results are shown. Further investigations were made by analyzing the data for women after stratification by premenopausal versus postmenopausal status and by analyzing the data for women and men after stratification into the age groups <55 versus 55+ years, with age-adjustment within these 2 groups (see Discussion). Premenopausal, perimenopausal, and postmenopausal status was defined as 0 to 60, 61 to 364, and 365+ days since the last menstrual period. Women having had both ovaries surgically removed were considered postmenopausal. Women whose last period occurred <=60 days ago, who had ever had hormonal replacement therapy, and who were aged <55 or 55+ years were reclassified as perimenopausal or postmenopausal, respectively. All statistical analyses were performed with SAS (Statistical Analysis System, Inc) and S-PLUS.29


*    Results
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 *Results
down arrowDiscussion
 down arrowReferences
 
Overall, 57% each of the men and the women who were selected and contacted participated in the present study. Among the participants, 86% had a blood sample taken. There were 1888 participants (957 men and 931 women) in the "general population sample" before exclusions (see Methods). After the exclusions, 1708 participants (846 men and 862 women) constituted the "statistical analysis sample."

ApoE Allele Frequencies in Geneva
Table 1 shows the distribution of the apoE polymorphism in the Geneva general population sample. The genotype {epsilon}3{epsilon}3 was most frequent (67%), followed by {epsilon}3{epsilon}4 (18%) and {epsilon}2{epsilon}3 (12%). The overall frequencies of the apoE alleles ({epsilon}2 7%, {epsilon}3 82%, and {epsilon}4 11%) were identical to those observed in another Swiss sample.26


View this table:
[in this window]
[in a new window]
  		Table 1. Distribution of Apoe Genotypes Among the General Population Sample of Adults Aged 35–74 Years, Geneva, Switzerland, 1999–2000

Characteristics of the Statistical Analysis Sample
Table 2 shows the characteristics of the study participants by sex. The age distributions for men and women were identical to those of the target population (data not shown). The sex-specific tertiles of %high-intensity activity were different, with two thirds of the women (men) expending <9% (<14%) of their energy in such activities. Additional percentiles (not shown) indicated that {approx}30% of either sex expended <0.5% of their total daily EE in high-intensity activities; only 29% of the women expended >=10% of their energy in high-intensity activities compared with 44% of the men. The distributions of total cholesterol (in millimoles per liter, raw data) were similar for both sexes, with 10th percentiles at {approx}4.5 and 90th percentiles at {approx}7.1. The overall median of 5.6 was above the recommended concentration (5 mmol/L).


View this table:
[in this window]
[in a new window]
  		Table 2. Characteristics of the Statistical Analysis Sample, Geneva, Switzerland, 1999–2000

Effects of ApoE Polymorphism on the Lipid Profile
Men’s and women’s lipid profiles are presented together in Table 3. Individuals in the apoE3 group had the following concentrations for total cholesterol, HDL cholesterol, LDL cholesterol, and triglycerides (mmol/L): 5.69, 1.34, 3.81, and 1.21, respectively, with an HDL/total cholesterol ratio of 24.0%. Comparatively, subjects in the apoE4 group had respectively higher levels of total cholesterol (5.91 mmol/L) and LDL cholesterol (4.01 mmol/L), whereas those in the apoE2 group had lower total cholesterol (5.37 mmol/L) and LDL cholesterol (3.33 mmol/L) and higher HDL cholesterol (1.42 mmol/L) concentrations. All these differences were statistically significant.


View this table:
[in this window]
[in a new window]
  		Table 3. Adjusted Lipid Profile Means by ApoE Groups, Men and Women Together, Geneva, Switzerland, 1999–2000

Physical Activity Effects on the Lipid Profile
Before any stratification into apoE groups, the only statistically significant association was found in men, with increasing %high-intensity activity being associated with higher HDL cholesterol (1.25 mmol/L in the upper tertile of %high-intensity activity compared with 1.19 mmol/L in the lower tertile, P<0.0004). Triglycerides seemed to be lowered by increased physical activity in both sexes, but these apparent effects were not quite statistically significant. Triglyceride concentrations decreased from 1.47 mmol/L in the most sedentary men (first tertile of %high-intensity activity) to 1.31 mmol/L in the most active men (third tertile of %high-intensity activity, P=0.09); the corresponding concentrations in women were 1.10 mmol/L and 1.06 mmol/L (P=0.53). (Data are not otherwise shown.)

Effects of Interaction of Physical Activity and ApoE Genotype on the Lipid Profile
The adjusted multiple linear regression relationships between physical activity according to apoE groups and the lipid profile values are summarized in Figure 1 (men) and Figure 2 (women). Among men, the apoE4 group had significantly greater beneficial effects of increased %high-intensity activity on HDL cholesterol (slope, P<0.001) and triglycerides (slope, P<0.03) than either the apoE2 group (interaction probability value: HDL, P=0.05; triglycerides, P=0.41), or the apoE3 group (interaction probability value: HDL, P<0.03; triglycerides, P=0.07; Figure 1a and 1b). Translated into clinical terms, the estimated linear regression slopes imply, eg, that an increase of 10% in the %high-intensity activity by an apoE4 man would correspond with an increase in HDL cholesterol of {approx}0.07 mmol/L and with a decrease in triglycerides of {approx}-0.15 mmol/L.



View larger version (22K):
[in this window]
[in a new window]
  		Figure 1. ApoE genotype–specific multiple linear regressions of lipid profile values for %high-intensity activity, adjusted for age, BMI, education, country of birth, cigarette smoking, dietary fat, and dietary fiber and assuming 2-way (apoEx%high-intensity activity) interactions in the models. a, HDL cholesterol: apoE2 slope=0.0012, P=0.53, interaction P=0.05; apoE3 slope=0.0016, P=0.09, interaction P<0.03; apoE4 slope=0.0066, P<0.001 (Reference). b, Tri-glycerides: apoE2 slope=-0.0071, P=0.26, interaction P=0.41; apoE3 slope=-0.0016, P=0.60, interaction P=0.07; apoE4 slope=-0.0147, P<0.03 (Reference). c, Total cholesterol: apoE2 slope=0.0077, P=0.25, interaction P=0.94; apoE3 slope=0.0000, P=0.99, interaction P=0.26; apoE4 slope=0.0084, P=0.21 (Reference). d, LDL cholesterol: apoE2 slope=0.0097, P=0.10, interaction P=0.88; apoE3 slope=-0.0009, P=0.75, interaction P=0.16; apoE4 slope=0.0085, P=0.15 (Reference). Results are from male residents aged 35 to 74 years from Geneva, Switzerland, 1999 to 2000 (n=846).



View larger version (22K):
[in this window]
[in a new window]
  		Figure 2. ApoE genotype–specific multiple linear regressions of lipid profile values for %high-intensity activity, adjusted for age, BMI, education, country of birth, cigarette smoking, dietary fat, and dietary fiber and assuming 2-way (apoEx%high-intensity activity) interactions in the models. a, HDL cholesterol: apoE2 slope=-0.0056, P<0.05, interaction P<0.008; apoE3 slope=0.0012, P=0.48, interaction P=0.21; apoE4 slope=0.0058, P=0.07 (Reference). b, Tri-glycerides: apoE2 slope=0.0042, P=0.30, interaction P=0.37; apoE3 slope=-0.0039, P=0.13, interaction P=0.65; apoE4 slope=-0.0015, P=0.76 (Reference). c, Total cholesterol: apoE2 slope=-0.0078, P=0.31, interaction P=0.97; apoE3 slope=-0.0082, P=0.71, interaction P=0.53; apoE4 slope=-0.0133, P=0.36 (Reference). d, LDL cholesterol: apoE2 slope=-0.0040, P=0.56, interaction P=0.40; apoE3 slope=-0.0013, P=0.78, interaction P=0.19; apoE4 slope=-0.0133, P=0.11 (Reference). Results are from female residents aged 35 to 74 years from Geneva, Switzerland, 1999 to 2000 (n=862).

Among women, as among men, there was a statistically significant interaction effect for HDL cholesterol (apoE4 slope, P=0.07; interaction probability value: apoE2 versus apoE4 slopes, P<0.008; apoE3 versusapoE4 slopes, P=0.21; Figure 2a). There was, in addition, a significant decrease in HDL for the apoE2 group (P<0.05), which was not observed among apoE2 men. For triglycerides, there appeared to be a protective effect of increased physical activity for the apoE4 versus apoE2 (but not versus apoE3) groups (qualitative assessments, not statistically significant; Figure 2b) as opposed to the more protective effect for apoE4 versus apoE3 men (interaction, P=0.07) and an apparently more protective effect for apoE4 versus apoE2 men (not statistically significant) noted above (Figure 1b).

For LDL cholesterol, neither the slopes themselves nor any of the slope difference interactions were statistically significant (Figures 1d and 2d). The apparent sex difference is due solely to the equation used to calculate LDL cholesterol: LDL=(total cholesterol-HDL-triglycerides/2.2). Depending on the magnitudes and directions of the changes in any of total cholesterol, HDL, and/or triglycerides with increasing physical activity, it is biomathematically possible to have as a result an increase (eg, Figure 1d, for apoE4 men), a decrease (eg, Figure 2d, for apoE4 women), or even no change at all in LDL cholesterol.


*    Discussion
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 *Discussion
down arrowReferences
 
Our results indicate that the presence of the {epsilon}4 allele adversely affects plasma HDL and triglycerides but only among more sedentary persons. Indeed, among more active people, there were no differences in the lipid profile values across apoE genotypes. These results support previous findings suggesting that performing more physical activity increases HDL cholesterol and decreases triglycerides in those who need it the most.30–33

This is the first population-based study to assess the apoE–physical activity interaction in adults. Our results are not fully consistent with the 2 studies that have assessed this interaction in children and young adults. In Finnish males, aged 9 to 24 years,18 the effect of physical activity on LDL was absent in apoE phenotype E4/4 and moderately inverse in subjects with E4/3, E3/3, and E3/2 phenotypes. The Coronary Artery Risk Development on Young Adults (CARDIA) cohort of young US adults (aged 18 to 30 years)19 found no interaction of apoE phenotypes and self-reported physical activity and physical fitness with LDL and HDL levels at baseline. In an intervention study, 50 sedentary overweight men who were requested to follow an American Heart Association Step I diet had plasma lipoprotein lipids measured before and after 9 months of endurance exercise training.20 The results indicated that apoE2 men (n=6) had greater plasma HDL cholesterol increases with endurance exercise training.

As reported by others,34–36 we also found that compared with apoE3 individuals, individuals with the {epsilon}4 allele had a more atherogenic profile (higher total and LDL cholesterol) but also lower HDL cholesterol, whereas individuals with the {epsilon}2 allele had a protective lipid profile.12,34,35

We found little association between increased physical activity and the lipid profile before any stratification by apoE polymorphism. Only more active men seemed to have a small increase in HDL cholesterol. Previous studies have not been consistent regarding this finding. Some authors found a protective effect of exercise on the lipid profile,7–10,37–39 but others found little40 or no effect.41,42

However, 2 important issues may explain these apparently contradictory results. First, exercise may increase HDL cholesterol only in those who need it the most.30 Indeed, people with the highest triglycerides and lowest HDL cholesterol levels have the greatest possibilities for improving their lipid profile. Zmuda et al32 have already shown that the ability to increase HDL cholesterol levels through endurance exercise training was limited to subjects with low initial HDL cholesterol. These findings were supported by an intervention survey in men, aged 20 to 40 years, with type 1 diabetes mellitus. After 3 to 4 months of aerobic exercise (30 to 60 minutes of moderate intensity running 3 to 5 times per week), the largest improvement occurred in the subgroup with the lowest HDL/LDL ratio at baseline.33 The recent conclusions of the Health, Risk Factors, Exercise Training and Genetics (HERITAGE) study31 also were that regular endurance exercise appeared particularly helpful for improving the lipid profile of men with low HDL cholesterol levels along with abdominal obesity and elevated triglyceride concentrations. Failure to improve triglyceride metabolism in other subjects was proposed as a possible explanation.

Second, it is worth noting that the most important improvements in the lipid profile have been found to be associated with very high–intensity activities.43,44 A possible dose effect of physical activity on lipid profile has been discussed37 and is consistent with our results. The observed association of physical activity with the lipid profile might perhaps have been even stronger in Geneva if a larger part of the adult population had been exercising on a regular basis, therefore having a higher %high-intensity activity. This could also explain the observed lack of association between physical activity and the lipid profiles of women, who were more sedentary than men in our population. In Figures 1a and 2a, the association of HDL cholesterol and physical activity was similar in both sexes for apoE3 and apoE4. The only sex difference was observed for apoE2 and could be due to small numbers of apoE2 women and women spending larger fractions of their EE in high-intensity physical activities.

Neither age45 nor menopausal status46 was previously identified as a potential effect modifier. Among the 862 women in our sample, 418 (51%) were premenopausal, 62 (7%) were perimenopausal, and 346 (42%) were postmenopausal (of whom 48% were currently on hormone replacement therapy). For clarity, we excluded the perimenopausal women from the analyses of the data for women after stratification by premenopausal versus postmenopausal status. These stratified results were based on smaller sample sizes but nonetheless suggested that the overall interaction effects of physical activity and apoE were present in both menopausal subgroups, although the effects were somewhat clearer in postmenopausal women (data not otherwise shown). However, we do not believe that this finding was due to hormonal effects but rather to the effects of age, for the following reason: after stratification by the age groups <55 versus 55+ years for both sexes, we found that the interaction effects of physical activity and apoE were more pronounced among women and men who were aged 55+ years; the interaction effects were less pronounced among women and men aged <55 years (data not otherwise shown). The reason for this latter finding might be that older people (including postmenopausal women) were those with the most atherogenic lipid profile at baseline. Moreover, in a previous study, Couillard et al31 showed that men with relatively high triglyceride and low HDL cholesterol concentrations were also those who showed the most significant increase in HDL concentration and decrease in triglyceride concentration associated with exercise. In any case, the menopausal (hormonal) status per se does not seem to influence the physical activity response of serum lipids, as has been shown by others in the context of exercise training,46 or its interaction with the apoE genotype. Age may be a potential confounder, but this has been taken into account in the results, as all the analyses were adjusted for age.

The biological bases of the interactive effects of apoE and physical activity on HDL cholesterol remain elusive. Our results suggest that physical activity compensates for what appears to be reduced HDL cholesterol concentrations in sedentary subjects harboring the {epsilon}4 allele. It is very likely that physical activity has an impact on lipid levels through several pathways. The {epsilon}4 allele12 is thought to involve more efficient binding to receptors than other alleles, leading to downregulation of LDL receptor activity,47 thus slowing down the major catabolic pathway for lipoprotein particles; the effects are to favor high levels of triglycerides and low levels of HDL cholesterol. The aging process also reduces LDL receptor activity, which might accentuate the {epsilon}4 allele effect.48 Physical activity may benefit the lipid profile by increasing the lipoprotein lipase activity,49 which controls the release of free fatty acids from chylomicrons and VLDL to peripheral tissues,50 thereby decreasing the level of serum triglycerides.51 In our free-living, adult (aged 35 to 74 years), nonselected population, the effect of the {epsilon}4 allele slowing down the triglyceride metabolism in sedentary individuals appears to be counteracted by the effect of increased physical activity, mediated by a stimulation of the lipoprotein lipase. This hypothesis is supported by a recent study on a common variant of the lipoprotein lipase gene (D9N), which adversely affects plasma lipids.52 Physically inactive D9N carriers presented much higher levels of total cholesterol and triglycerides and lower HDL cholesterol compared with inactive noncarriers. However, this was not the case for physically active D9N carriers.

Another, and not exclusive, pathway may be through adrenocorticosteroid synthesis. High-intensity prolonged exercise has been shown to increase plasma adrenocorticotropin53,54 and, hence, cortisol. HDL cholesterol may be the preferred lipoprotein as a source for adrenocorticosteroid synthesis.55 Indeed, although glucocorticoid therapy often has adverse effects on the cardiovascular system, including dyslipidemia,56 corticotropin seems to contribute to the increase in HDL cholesterol level.57 This increase in HDL cholesterol might be more obvious in apoE4 individuals.

In conclusion, the present results indicate that increasing physical activity may compensate for the potentially deleterious effects of the apoE4 genotype on the lipid profile, at least in western (European) populations. It appears that this protection may be obtained by performing any high-intensity physical activity (ie, expending 4 times the BMR or more), such as brisk walking or sports. The required minimal sufficient duration of any such activity still needs to be determined. Individuals with the most atherogenic profile may be able to substantially reduce their cardiovascular risk by being more physically active. For this subgroup, physical activity changes may become a first choice because they would be easier to implement than dietary changes and would have less potential side effects than drugs. These findings, if confirmed, have high clinical and public health relevance because they may lead to more specific recommendations for the clinical prevention of cardiovascular diseases. However, our conclusions need to be confirmed by an intervention study.


*    Acknowledgments
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 up arrowDiscussion
 up arrowReferences
 
This project was supported by the Swiss National Science Foundation (No. 3200-054097). We thank I. Bouchardy, I. Moix, C. Monso-Hinard, and A. Schönbörner for technical assistance.

Received August 31, 2001; accepted October 26, 2001.


*    References
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 up arrowDiscussion
 *References
 

  1. Hagberg JM, Wilund KR, Ferrell RE. APO E gene and gene-environment effects on plasma lipoprotein-lipid levels. Physiol Genomics. 2000; 4: 101–108.[Abstract/Free Full Text]
  2. Lussier-Cacan S, Bolduc A, Xhignesse M, Niyonsenga T, Connelly PW, Sing CF. Impact of age and body size on inter-individual variation in measures of lipid metabolism: influence of gender and apolipoprotein E genotype. Clin Genet. 2000; 57: 35–47.[CrossRef][Medline] [Order article via Infotrieve]
  3. Manson JE, Tosteson H, Ridker PM, Satterfield S, Hebert P, O’Connor GT, Buring JE, Hennekens CH. The primary prevention of myocardial infarction. N Engl J Med. 1992; 326: 1406–1416.[Medline] [Order article via Infotrieve]
  4. Austin MA, King MC, Vranizan KM, Krauss RM. Atherogenic lipoprotein phenotype: a proposed genetic marker for coronary heart disease risk. Circulation. 1990; 82: 495–506.[Abstract/Free Full Text]
  5. Tornvall P, Karpe F, Carlson LA, Hamsten A. Relationships of low density lipoprotein subfractions to angiographically defined coronary artery disease in young survivors of myocardial infarction. Atherosclerosis. 1991; 90: 67–80.[CrossRef][Medline] [Order article via Infotrieve]
  6. 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: 1945–1949.[Abstract/Free Full Text]
  7. Marrugat J, Elosua R, Covas MI, Molina L, Rubies-Prat J. Amount and intensity of physical activity, physical fitness, and serum lipids in men: the marathon investigators [published erratum appears in Am J Epidemiol. 1996;144:317]. Am J Epidemiol. 1996; 143: 562–569.[Abstract/Free Full Text]
  8. Ready AE, Drinkwater DT, Ducas J, Fitzpatrick DW, Brereton DG, Oades SC. Walking program reduces elevated cholesterol in women postmenopause. Can J Cardiol. 1995; 11: 905–912.[Medline] [Order article via Infotrieve]
  9. Ponjee GA, Janssen EM, Hermans J, van Wersch JW. Effects of long-term exercise of moderate intensity on anthropometric values and serum lipids and lipoproteins. Eur J Clin Chem Clin Biochem. 1995; 33: 121–126.[Medline] [Order article via Infotrieve]
  10. Pronk NP, Crouse SF, O’Brien BC, Rohack JJ. Acute effects of walking on serum lipids and lipoproteins in women. J Sports Med Phys Fitness. 1995; 35: 50–58.[Medline] [Order article via Infotrieve]
  11. Leon AS, Connett J. Physical activity and 10.5 year mortality in the Multiple Risk Factor Intervention Trial (MRFIT). Int J Epidemiol. 1991; 20: 690–697.[Abstract/Free Full Text]
  12. Davignon J, Gregg RE, Sing CF. Apolipoprotein E polymorphism and atherosclerosis. Arteriosclerosis. 1988; 8: 1–21.[Abstract/Free Full Text]
  13. Utermann G, Hees M, Steinmetz A. Polymorphism of apolipoprotein E and occurrence of dysbetalipoproteinaemia in man. Nature. 1977; 269: 604–607.[CrossRef][Medline] [Order article via Infotrieve]
  14. Kuusi T, Nieminen MS, Ehnholm C, Yki-Järvinen H, Valle M, Esko A. Apoprotein E polymorphism and coronary artery disease: increased prevalence of apolipoprotein E-4 in angiographically verified coronary patients. Arteriosclerosis. 1989; 9: 237–241.[Abstract/Free Full Text]
  15. Eichner JE, Kuller LH, Orchard TJ, Grandits GA, McCallum LM, Ferrell RE, Neaton JD. Relation of apolipoprotein E phenotype to myocardial infarction and mortality from coronary artery disease. Am J Cardiol. 1993; 71: 160–165.[CrossRef][Medline] [Order article via Infotrieve]
  16. Nakata Y, Katsuya T, Rakugi H, Takami S, Ohishi M, Kamino K, Higaki J, Tabuchi Y, Kumahara Y, Miki T, et al. Polymorphism of the apolipoprotein E and angiotensin-converting enzyme genes in Japanese subjects with silent myocardial ischemia. Hypertension. 1996; 27: 1205–1209.[Abstract/Free Full Text]
  17. Mahley RW, Huang Y, Rall SC Jr. Pathogenesis of type III hyperlipoproteinemia (dysbetalipoproteinemia): questions, quandaries, and paradoxes. J Lipid Res. 1999; 40: 1933–1949.[Abstract/Free Full Text]
  18. Taimela S, Lehtimaki T, Porkka KV, Rasanen L, Viikari JS. The effect of physical activity on serum total and low-density lipoprotein cholesterol concentrations varies with apolipoprotein E phenotype in male children and young adults: the Cardiovascular Risk in Young Finns Study. Metabolism. 1996; 45: 797–803.[CrossRef][Medline] [Order article via Infotrieve]
  19. Schmitz KH, Schreiner PJ, Jacobs DR Jr, Leon AS, Liu K, Howard B, Sternfeld B. Independent and interactive effects of apolipoprotein E phenotype and cardiorespiratory fitness on plasma lipids. Ann Epidemiol. 2001; 11: 94–103.[CrossRef][Medline] [Order article via Infotrieve]
  20. Hagberg JM, Ferrell RE, Katzel LI, Dengel DR, Sorkin JD, Goldberg AP. Apolipoprotein E genotype and exercise training-induced increases in plasma high-density lipoprotein (HDL)- and HDL2-cholesterol levels in overweight men. Metabolism. 1999; 48: 943–945.[CrossRef][Medline] [Order article via Infotrieve]
  21. Morabia A, Bernstein M, Heritier S, Ylli A. Community-based surveillance of cardiovascular risk factors in Geneva: methods, resulting distributions, and comparisons with other populations. Prev Med. 1997; 26: 311–319.[CrossRef][Medline] [Order article via Infotrieve]
  22. Bernstein M, Morabia A, Costanza MC, Landis JR, Ross A, Flandre P, Luong BL, Kumanyika S, Sorenson A, Localio R. Nutritional balance of the diet of the adult residents of Geneva [in French]. Soz Praventivmed. 1994; 39: 333–344.[CrossRef][Medline] [Order article via Infotrieve]
  23. Morabia A, Bernstein M, Kumanyika S, Sorenson A, Mabiala I, Prodolliet B, Rolfo I, Luong BL. Development and validation of a semi-quantitative food questionnaire based on a population survey [in French]. Soz Praventivmed. 1994; 39: 345–369.[CrossRef][Medline] [Order article via Infotrieve]
  24. Bernstein M, Sloutskis D, Kumanyika S, Sparti A, Schutz Y, Morabia A. Data-based approach for developing a physical activity frequency questionnaire. Am J Epidemiol. 1998; 147: 147–154.[Abstract/Free Full Text]
  25. World Health Organization. WHO Consensus Report: Energy and Protein Requirements. Geneva, Switzerland: World Health Organisation; 1986.WHO Technical Report No 724.
  26. James RW, Valiotis C, Grab B, Pometta D. Phénotypes de l’apoprotéine E (apo E) et lipides sériques des diabétiques. Schweiz Med Wochenschr. 1987; 117: 2021–2024.[Medline] [Order article via Infotrieve]
  27. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972; 18: 499–502.[Abstract]
  28. Venables WN, Ripley BD. Modern Applied Statistics with S-PLUS. New York, NY: Springer; 1999.
  29. S-Plus 2000 User’s Guide. Seattle, Wash: Data Analysis Products Division; 2001.
  30. Thompson PD, Rader DJ. Does exercise increase HDL cholesterol in those who need it the most? Arterioscler Thromb Vasc Biol. 2001; 21: 1097–1098.[Free Full Text]
  31. Couillard C, Despres JP, Lamarche B, Bergeron J, Gagnon J, Leon AS, Rao DC, Skinner JS, Wilmore JH, Bouchard C. Effects of endurance exercise training on plasma HDL cholesterol levels depend on levels of triglycerides: evidence from men of the Health, Risk Factors, Exercise Training and Genetics (HERITAGE) family study. Arterioscler Thromb Vasc Biol. 2001; 21: 1226–1232.[Abstract/Free Full Text]
  32. Zmuda JM, Yurgalevitch SM, Flynn MM, Bausserman LL, Saratelli A, Spannaus-Martin DJ, Herbert PN, Thompson PD. Exercise training has little effect on HDL levels and metabolism in men with initially low HDL cholesterol. Atherosclerosis. 1998; 137: 215–221.[CrossRef][Medline] [Order article via Infotrieve]
  33. Laaksonen DE, Atalay M, Niskanen LK, Mustonen J, Sen CK, Lakka TA, Uusitupa MI. Aerobic exercise and the lipid profile in type 1 diabetic men: a randomized controlled trial. Med Sci Sports Exerc. 2000; 32: 1541–1548.[Medline] [Order article via Infotrieve]
  34. Hallman DM, Boerwinkle E, Saha N, Sandholzer C, Menzel HJ, Csazar A, Utermann G. The apolipoprotein E polymorphism: a comparison of allele frequencies an effects in nine populations. Am J Hum Genet. 1991; 49: 338–349.[Medline] [Order article via Infotrieve]
  35. Lehtimäki T, Moilanen T, Aalto-Setälä Kea. Association of apolipoprotein E and B polymorphisms with serum lipids. Ann Med. 1991; 23: 657–662.[Medline] [Order article via Infotrieve]
  36. Corella D, Guillen M, Portoles O, Sabater A, Cortina S, Folch J, Saiz C. Apolipoprotein E gene polymorphism and risk of hypercholesterolemia: a case control study in a working population of Valencia [in Spanish]. Med Clin (Barc). 2000; 115: 170–175.
  37. Wei M, Macera CA, Hornung CA, Blair SN. Changes in lipids associated with change in regular exercise in free-living men. J Clin Epidemiol. 1997; 50: 1137–1142.[CrossRef][Medline] [Order article via Infotrieve]
  38. Halle M, Berg A, Konig D, Keul J, Baumstark MW. Differences in the concentration and composition of low-density lipoprotein subfraction particles between sedentary and trained hypercholesterolemic men. Metabolism. 1997; 46: 186–191.[CrossRef][Medline] [Order article via Infotrieve]
  39. Wood PD, Stefanick ML, Dreon DM, Frey-Hewitt B, Garay SC, Williams PT. Changes in plasma lipids an lipoproteins in overweight men during weight loss through dieting as compared with exercise. N Engl J Med. 1998; 319: 1173–1179.
  40. Crouse SF, O’Brien BC, Grandjean PW, Lowe RC, Rohack JJ, Green JS, Tolson H. Training intensity, blood lipids, and apolipoproteins in men with high cholesterol. J Appl Physiol. 1997; 82: 270–277.[Abstract/Free Full Text]
  41. Woolf-May K, Kearney EM, Jones DW, Davison RC, Coleman D, Bird SR. The effect of two different 18-week walking programmes on aerobic fitness, selected blood lipids and factor XIIa. J Sports Sci. 1998; 16: 701–710.[CrossRef][Medline] [Order article via Infotrieve]
  42. Leon AS, Rice T, Mandel S, Despres JP, Bergeron J, Gagnon J, Rao DC, Skinner JS, Wilmore JH, Bouchard C. Blood lipid response to 20 weeks of supervised exercise in a large biracial population: the HERITAGE Family Study. Metabolism. 2000; 49: 513–520.[CrossRef][Medline] [Order article via Infotrieve]
  43. Rubinstein A, Burstein R, Lubin F, Chetrit A, Dann EJ, Levtov O, Geter R, Deuster PA, Dolev E. Lipoprotein profile changes during intense training of Israeli military recruits. Med Sci Sports Exerc. 1995; 27: 480–484.[Medline] [Order article via Infotrieve]
  44. Ekstedt B, Jonsson E, Johnson O. Influence of dietary fat, cholesterol and energy on serum lipids at vigorous physical exercise. Scand J Clin Lab Invest. 1991; 51: 437–442.[Medline] [Order article via Infotrieve]
  45. Giada F, Vigna GB, Vitale E, Baldo-Enzi G, Bertaglia M, Crecca R, Fellin R. Effect of age on the response of blood lipids, body composition, and aerobic power to physical conditioning and deconditioning. Metabolism. 1995; 44: 161–165.[CrossRef][Medline] [Order article via Infotrieve]
  46. Grandjean PW, Crouse SF, O’Brien BC, Rohack JJ, Brown JA. The effects of menopausal status and exercise training on serum lipids and the activities of intravascular enzymes related to lipid transport. Metabolism. 1998; 47: 377–383.[CrossRef][Medline] [Order article via Infotrieve]
  47. Dong LM, Wilson C, Wardell MR, Simmons T, Mahley RW, Weisgraber KH, Agard DA. Human apolipoprotein E: role of arginine 61 in mediating the lipoprotein preferences of the E3 and E4 isoforms. J Biol Chem. 1994; 269: 22358–22365.[Abstract/Free Full Text]
  48. Pairitz Jarvik G. Genetic predictors of common disease: apolipoprotein E genotype as a paradigm. Ann Epidemiol. 1997; 7: 357–362.[CrossRef][Medline] [Order article via Infotrieve]
  49. Thompson PD, Yurgalevitch SM, Flynn MM, Zmuda JM, Spannaus-Martin D, Saritelli A, Bausserman L, Herbert PN. Effect of prolonged exercise training without weight loss on high-density lipoprotein metabolism in overweight men. Metabolism. 1997; 46: 217–223.[CrossRef][Medline] [Order article via Infotrieve]
  50. 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(pt 1): E229–E236.
  51. Wannamethee SG, Shaper AG, Alberti KG. Physical activity, metabolic factors, and the incidence of coronary heart disease and type 2 diabetes. Arch Intern Med. 2000; 160: 2108–2116.[Abstract/Free Full Text]
  52. Boer JM, Kuivenhoven JA, Feskens EJ, Schouten E, G, Havekes LM, Seidell JC, Kastelein JJ, Kromhout D. Physical activity modulates the effect of a lipoprotein lipase mutation (D9N) on plasma lipids and lipoproteins. Clin Genet. 1999; 56: 158–163.[CrossRef][Medline] [Order article via Infotrieve]
  53. Inder WJ, Hellemans J, Swanney MP, Prickett TC, Donald RA. Prolonged exercise increases peripheral plasma ACTH, CRH, and AVP in male athletes. J Appl Physiol. 1998; 85: 835–841.[Abstract/Free Full Text]
  54. Heitkamp HC, Schulz H, Rocker K, Dickhuth HH. Endurance training in females: changes in beta-endorphin and ACTH. Int J Sports Med. 1998; 19: 260–264.[Medline] [Order article via Infotrieve]
  55. Yaguchi H, Tsutsumi K, Shimono K, Omura M, Sasano H, Nishikawa T. Involvement of high density lipoprotein as substrate cholesterol for steroidogenesis by bovine adrenal fasciculo-reticularis cells. Life Sci. 1998; 62: 1387–1395.[CrossRef][Medline] [Order article via Infotrieve]
  56. Sholter DE, Armstrong PW. Adverse effects of corticosteroids on the cardiovascular system. Can J Cardiol. 2000; 16: 505–511.[Medline] [Order article via Infotrieve]
  57. Berg AL, Nilsson-Ehle P. Direct effects of corticotropin on plasma lipoprotein metabolism in man: studies in vivo and in vitro. Metabolism. 1994; 43: 90–97.[CrossRef][Medline] [Order article via Infotrieve]



This article has been cited by other articles:


Home page
 Am J EpidemiolHome page
M. C. Costanza, E. Cayanis, B. M. Ross, M. S. Flaherty, G. B. Alvin, K. Das, and A. Morabia
Relative Contributions of Genes, Environment, and Interactions to Blood Lipid Concentrations in a General Adult Population
Am. J. Epidemiol., April 15, 2005; 161(8): 714 - 724.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
A. Rundle
Molecular Epidemiology of Physical Activity and Cancer
Cancer Epidemiol. Biomarkers Prev., January 1, 2005; 14(1): 227 - 236.
[Abstract] [Full Text] [PDF]

Home page
 J. Lipid Res.Home page
R. Elosua, J. M. Ordovas, L. A. Cupples, C. S. Fox, J. F. Polak, P. A. Wolf, R. A. D'Agostino Sr., and C. J. O'Donnell
Association of APOE genotype with carotid atherosclerosis in men and women: the Framingham Heart Study
J. Lipid Res., October 1, 2004; 45(10): 1868 - 1875.
[Abstract] [Full Text] [PDF]

Home page
 Hum Mol GenetHome page
A. Morabia, E. Cayanis, M. C. Costanza, B. M. Ross, M. S. Flaherty, G. B. Alvin, K. Das, and T. C. Gilliam
Association of extreme blood lipid profile phenotypic variation with 11 reverse cholesterol transport genes and 10 non-genetic cardiovascular disease risk factors
Hum. Mol. Genet., November 1, 2003; 12(21): 2733 - 2743.
[Abstract] [Full Text] [PDF]

Home page
 ANGIOLOGYHome page
G. Kolovou, D. Daskalova, and D. P. Mikhailidis
Apolipoprotein E Polymorphism and Atherosclerosis
Angiology, January 1, 2003; 54(1): 59 - 71.
[Abstract] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted