| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research Communication |
1 From the Nutrition and Genomics Laboratory, the Lipid Metabolism Laboratory and the Epidemiology Program, Jean Mayer–US Department of Agriculture Human Nutrition Research Center on Aging at Tufts University, Boston; the Boston University School of Public Health, Boston; and the Framingham Heart Study, Boston University School of Medicine, Framingham, MA.
2 Supported by NIH/NHLBI grant no. HL54776, NIH/NHLBI contract no. 1-38038, and contracts 53-K06-5-10 and 58-1950-9-001 from the US Department of Agriculture Research Service. 3 Address reprint requests to JM Ordovas, Nutrition and Genomics Laboratory, JM USDA HNRCA at Tufts University, 711 Washington Street, Boston, MA 02111. E-mail: ordovas{at}hnrc.tufts.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: We examined whether dietary fat modulates the association between this polymorphism and HDL-cholesterol concentrations.
Design: We studied a population-based sample of 755 men and 822 women from the Framingham Offspring Study.
Results: The frequency of the A allele was 0.165. No significant differences were observed between G/G subjects and carriers of the A allele for any lipid variables. In multivariate linear regression models, HDL-cholesterol concentrations in women were associated with a significant interaction between polyunsaturated fatty acid (PUFA) intake as a continuous variable and APOA1 genotype (P = 0.005). By using 3 categories of PUFA intake, we found a significantly different effect of APOA1 genotype across PUFA categories in women. When PUFA intake was <4% of energy, G/G subjects had 
14% higher HDL-cholesterol concentrations than did carriers of the A allele (P < 0.05). Conversely, when PUFA intake was >8%, HDL-cholesterol concentrations in carriers of the A allele were 13% higher than those of G/G subjects (P < 0.05). No significant allelic difference was observed for subjects in the range of PUFA intake of 4–8% of energy. These interactions were not significant in men.
Conclusions: We found a significant gene-diet interaction associated with the APOA1 G-A polymorphism. In women carriers of the A allele, higher PUFA intakes were associated with higher HDL-cholesterol concentrations, whereas the opposite effect was observed in G/G women.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A common G-to-A transition located 75 base pairs (bp) upstream from the transcription start site of the APOA1 gene has been studied extensively. Initial reports showed that individuals carrying the A allele had higher concentrations of apo A-I and HDL cholesterol than did individuals with the G/G wild type (8, 9). Subsequently, studies that examined this association reported contradictory results. Although some were in partial agreement with the initial findings (10–17), others did not detect any significant association (18–23), and even an opposite association has been reported (24). A meta-analysis that included some of these studies showed that the rare A allele may be associated with mildly increased (by 0.05 g/L) apo A-I concentrations (25). It was suggested that the inconsistencies between studies could be the result of interactions with environmental factors that modulate the effect of this genetic polymorphism. Although some of these studies investigated the possible interaction with tobacco smoking (13–15, 17), none assessed the influence of long-term dietary habits in a large, population-based sample.
Dietary intervention studies including subjects with normal or slightly elevated cholesterol concentrations showed that the A allele is associated with increased LDL-cholesterol response to changes in dietary fat (26, 27). In one study (27), male and female subjects were first fed a diet high in saturated fatty acids (SFAs), followed by a diet rich in monounsaturated fatty acids (MUFAs) and a diet rich in polyunsaturated fatty acids (PUFAs). The hypocholesterolemic effect associated with a high-PUFA diet was significantly greater in G/A women than in G/G women. In men, however, the A allele was not a predictor of response. In another study (26), young male subjects were fed a low-fat diet followed by a diet rich in MUFAs. After consumption of the high-MUFA diet, plasma LDL-cholesterol concentrations increased significantly in G/A subjects but not in G/G subjects. Although the dietary interventions and subject characteristics differed between these 2 studies, in both studies carriers of the A allele were more responsive to dietary changes than were G/G subjects (27). Conversely, no significant gene-diet interactions were observed in subjects who were heterozygotes for familial hypercholesterolemia (28) or in subjects participating in a dietary intervention study in North Karelia (10).
It has been established that dietary fatty acids, mainly PUFAs, can modulate gene expression (29). Considering that the G-A polymorphism discussed above is located in a GC-rich DNA region, it was proposed that the presence of A or G may differentially influence the efficiency of APOA1 gene transcription. This hypothesis was tested in a series of in vitro experiments (30–34). However, these studies showed as much variability as did the population studies. Several reports found that the A allele was associated with lower transcriptional efficiency (30–32). Conversely, other investigators found increases in transcriptional efficiency associated with the A allele (34) or found no effect at all (33). Therefore, it was suggested that the observed effects could be a result of linkage disequilibrium with other regulatory sequences and that the combination of these elements may explain the contradictory results regarding APOA1 gene expression (16, 33).
We designed the present study on the basis of the gene-environment interaction hypothesis put forward to reconcile the discrepancies associated with this common genetic variant. The goal of our study was to examine whether dietary fat modulates the association between the APOA1 G-A polymorphism at -75 bp and HDL-cholesterol concentrations in a population-based sample of men and women in the Framingham Offspring Study.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Plasma lipid, lipoprotein, and apolipoprotein measurements
After subjects fasted for 12 h, venous blood samples were collected in tubes containing 0.1% EDTA. Plasma was separated from blood cells by centrifugation (1600 x g for 10 min at 4°C) and was analyzed immediately for lipid content. Plasma total cholesterol, HDL cholesterol, and triacylglycerol concentrations were measured as described previously (38). HDL cholesterol was measured after precipitation of apo B–containing lipoproteins with dextran sulfate and magnesium sulfate. LDL-cholesterol concentrations were estimated by using the equation of Friedewald et al (39). Plasma concentrations of apo A-I and apo B were measured with a noncompetitive enzyme-linked immunosorbent assay by using affinity-purified polyclonal antibodies (40, 41). CVs for total cholesterol, HDL-cholesterol, and triacylglycerol measurements were all <5%.
DNA isolation and genotyping
Leukocyte DNA was extracted from 5–10 mL whole blood by using the method described by Miller et al (42). Amplification of a 432-bp region of the APOA1 5' region was performed with the polymerase chain reaction in a DNA thermal cycler (PTC-100; MJ Research Inc, Watertown, MA) by using 250 ng genomic DNA and 0.2 µmol/L of each oligonucleotide primer (P1, 5'-AGGGACAGAGCTGATCCTTGAACTCTTAAG-3'; P2, 5'-TTAGGGGACACCTACCCGTCAGGAAGAGCA-3') in a 50-µL volume. Each reaction mixture was heated at 95°C for 5 min and followed by 30 cycles of amplification (95°C for 1 min, 58°C for 1.5 min, and 72°C for 2 min). The polymerase chain reaction products were digested with 10 units of the restriction endonuclease enzyme MspI (BRL, Rockville, MD) and the fragments were separated by using electrophoresis on a 3.5%-agarose gel. After electrophoresis, the gel was treated with ethidium bromide for 20 min and DNA fragments were visualized with ultraviolet illumination.
Dietary information
Dietary intake was estimated with the semiquantitative food-frequency questionnaire described by Rimm et al (43). This questionnaire includes 136 food items, with questions about intake of beer, wine, and spirits. Subjects are asked how often they consume each item per day, week, or month. Food-item intake frequencies are linked with nutrient data to estimate daily nutrient intakes.
Fat intake data were obtained in terms of absolute amounts (g/d). We then modeled the effect of fat in terms of nutrient density, ie, the ratio of energy from fat to total energy, expressed as a percentage. Intakes of total fat, SFAs, MUFAs, and PUFAs were calculated for each individual. These were included in the analyses as both continuous and categorical variables. To construct the categorical variables, intakes were classified into 2 groups according to the mean value of the population (ie, one group had intakes below the mean and one group had intakes above it). In addition, we defined 3 categories of PUFA intake (low, <4% of energy; middle, 4–8% of energy; and high, >8% of energy) on the basis of the frequency distribution and range of PUFA consumption in the population. PUFA intake ranged from 2.0% to 16.5% of total energy intake in men and from 1.2% to 13.7% of total energy intake in women.
Alcohol consumption was calculated in g/d on the basis of the individual's reported frequency of consumption of alcoholic beverages during the previous year. Subjects were also classified as either nondrinkers (those who did not report consumption of alcohol) or as drinkers (those who reported drinking any amount of alcohol).
Statistical analyses
Triacylglycerol and apo B concentrations were log transformed and alcohol intake was square-root transformed to improve normality for statistical testing. Allele frequencies were estimated with the gene-counting method. Chi-square tests were conducted to examine whether the genotype frequencies were in Hardy-Weinberg equilibrium. To compare means between 2 independent groups, the Student's t test was used. For multiple comparisons of means, one-way analyses of variance were performed and P values for linear trends across categories were calculated by partitioning the between-groups sums of squares into trend components. Because of the marked sex differences in the variables of interest and the statistical significance of the interaction terms for sex in the regression models, statistical analyses were done separately for men and women. Pearson's product-moment correlation coefficients were calculated to describe unadjusted associations among continuous variables.
Multiple linear regression models with dummy variables for categorical terms were fitted to test the null hypotheses of no association between APOA1 polymorphism and HDL-cholesterol or apo A-I concentrations (dependent variables) after considering the effects of several predictors (body mass index, age, sex, smoking, alcohol consumption, fat intake, energy intake, and APOA1 polymorphism). Homogeneity of allelic effects according to sex or environmental factors was also tested by introducing the corresponding terms of interaction in a hierarchical way in the more parsimonious linear regression model. Regression coefficients and the proportion of variance attributable to each predictor were estimated from the models. For HDL-cholesterol and apo A-I variables, 6 regression models were fitted separately for men and women. In the core model (model 1), no interaction terms with fat intakes (as continuous variables) were considered. In models 2–6, interaction terms between APOA1 polymorphism (dichotomous) and intakes of total fat, SFAs, MUFAs, or PUFAs were tested. Additional multiple linear regression models with categorical variables for fat intake and for testing a gene-dosage effect for APOA1 polymorphism were fitted as described in the Results section. Regression diagnostics such as analysis of residuals and colinearity tests were used to check the assumptions and to assess the accuracy of the computations. All reported P values were two-sided, and 95% CIs for estimated coefficients were calculated. The SAS statistical package (version 6.12; SAS Institute Inc, Cary, NC) was used for the analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. No deviation from the Hardy-Weinberg equilibrium was observed (chi-square = 1.74; P > 0.05) for genotype distribution. The allele frequencies were 0.835 (95% CI: 0.822, 0.848) for the G allele and 0.165 (95% CI: 0.152, 0.178) for the A allele. There were statistically significant differences between men and women for all the variables examined, with the exceptions of age and genotype distribution.
|
. No significant differences between G/G homozygotes and carriers of the A allele were observed for age, body mass index, or any lipid, lipoprotein, or apolipoprotein variables. In women, mean HDL-cholesterol concentrations in the G/G and G/A + A/A groups were 1.46 ± 0.40 and 1.48 ± 0.41 mmol/L, respectively (NS). Likewise, male G/G homozygotes had HDL-cholesterol concentrations that were not significantly different from those of A-allele carriers (1.11 ± 0.29 and 1.12 ± 0.30 mmol/L, respectively; NS). Similar results were found for apo A-I concentrations. Mean dietary intakes of total fat, SFAs, MUFAs, and PUFAs did not differ significantly between the G/A + A/A and G/G groups.
|
. First, we fitted a core model (model 1) that included HDL-cholesterol concentration as the outcome variable and APOA1 genotype (G/G as reference), alcohol consumption (drinkers as reference), tobacco smoking (nonsmokers as reference), age, body mass index, and total energy intake as predictor variables. In model 1, APOA1 genotype adjusted for the other factors was not significantly related to HDL-cholesterol concentrations in women (B = 0.007 mmol/L for the carriers of the A allele, in comparison with the G/G homozygotes; NS). In model 2, we added total fat intake (expressed as a percentage of energy in continuous form) and its interaction term with the APOA1 genotype to model 1. This interaction term was not significant. For models 3 and 4, we started with model 1 and added SFA and MUFA intakes, respectively, along with the corresponding interaction terms with the APOA1 genotype. The P values for these interaction terms remained nonsignificant.
|
6% and >6% of energy), its interaction term with the APOA1 genotype remained statistically significant (P = 0.049). In women who consumed <6% of energy from PUFA, mean HDL-cholesterol concentrations for G/G homozygotes and carriers of the A allele were 1.49 ± 0.40 and 1.44 ± 0.40 mmol/L, respectively. In women who consumed >6% of energy from PUFA, these values were 1.44 ± 0.39 and 1.49 ± 0.39 mmol/L, respectively.
For men, the results of the same type of linear regression analysis are shown in Table 4. The magnitude of the regression coefficient for the APOA1 genotype was greater when dietary fat was added to the model, but none of the interactions between fat intake and the APOA1 genotype in men were statistically significant. When the continuous PUFA-intake variable was made into a dichotomous variable, the interaction term between this variable and the APOA1 genotype was also nonsignificant in men. Furthermore, no significant differences in HDL-cholesterol concentrations were found between the 2 groups with higher and lower PUFA intakes.
|
The correlation coefficients between PUFA intake and MUFA intake were r = 0.54 and r = 0.51 in women and men, respectively. Between PUFA and SFA intakes, the correlations were r = 0.22 in women and r = 0.14 in men. Because these correlations were not excessively high, we were able to retain these variables in the multivariate regression models without problems of colinearity and we also could obtain a more independent estimate of the regression coefficient for PUFA intake. Thus, when model 5 in Tables 3 and 4
was additionally adjusted for SFA intake and MUFA intake (model 6, not shown), the interaction term between APOA1 genotype and PUFA intake remained significant in women (B = 0.045; P = 0.007) but was not significant in men (B = 0.002; P = 0.86). In men, we observed a significant 4-way interaction (P = 0.023) between APOA1 genotype, PUFA intake, tobacco smoking, and alcohol consumption. However, this study did not have enough statistical power to analyze this effect within each of the 2 categories for PUFA intakes (ie, high and low), tobacco smoking (smokers and nonsmokers), or alcohol consumption (drinkers and nondrinkers).
Dietary fat intakes, APOA1 alleles, and apo A-I concentrations
Next we examined apo A-I concentrations as the dependent variable. Multivariate regression models 1–6 were fitted as previously described, for men and women separately (results not shown). In women, the P values for the interaction terms with APOA1 alleles were all not significant. Although the interaction term between the APOA1 genotype and PUFA intake was not significant for apo A-I concentrations in women, the effects were in the same direction as those observed for HDL-cholesterol concentrations. No significant interactions with smoking or alcohol consumption were observed in women. In men, the P values for the APOA1 x fat intake interactions were not significant for total fat, MUFA, SFA, and PUFA intakes. When alcohol and smoking interactions were also considered in model 6, a significant 3-way interaction (P = 0.035) between PUFA intake, alcohol consumption, and APOA1 genotype was found. The effect of this interaction was as follows: carriers of the A allele had lower concentrations of apo A-I than did G/G homozygotes, but apo A-I concentrations were higher with higher PUFA intakes in a linear pattern that differed between drinkers and nondrinkers. The slope of the line was higher in nondrinkers than in drinkers.
We also divided both male and female subjects into 3 categories of PUFA intake (<4% of energy, 4%–8% of energy, and >8% of energy) to obtain mean values that would allow direct comparison with other studies. Mean HDL-cholesterol and apo A-I concentrations by APOA1 genotype and sex across the 3 categories of PUFA intake are shown in Table 5. P values are shown for the unadjusted comparisons and for comparisons adjusted for age, BMI, alcohol consumption, tobacco smoking, and fat intake. In women, APOA1 genotype clearly had a different effect across the categories of PUFA intake. When PUFA intake was low, G/G homozygotes had 14% higher HDL-cholesterol concentrations than did carriers of the A allele (P < 0.05). When PUFA intake was high, HDL-cholesterol concentrations in carriers of the A allele were 13% higher on average than were concentrations in the G/G group. For apo A-I concentrations, the trend was not different from that observed for HDL cholesterol, but it was not significant. In men, an additional stratification by alcohol consumption or tobacco smoking was necessary to obtain significant differences between G/G homozygotes and A carriers (data not shown).
|
, the estimated mean HDL-cholesterol and apo A-I concentrations in G/G, G/A, and A/A women across the 3 categories of PUFA intake are shown. These means were adjusted for age, BMI, tobacco smoking, alcohol consumption, SFA intake, and MUFA intake. The P values shown are for the interaction between the APOA1 genotype and the 3 categories of PUFA intake. For HDL cholesterol, this interaction was significant and a clear gene-dosage effect was observed: subjects with the A/A genotype had the highest HDL-cholesterol concentrations in the category of PUFA intake >8% of energy and the lowest HDL-cholesterol concentrations in the category of PUFA intake <4% of energy. Moreover, when PUFA intake was analyzed as a continuous variable, the regression coefficient for A/A subjects adjusted for all covariates was 2.5 times higher than that of G/A heterozygotes (P < 0.05). In the regression model for apo A-I concentrations, we observed a gene-dosage effect similar to that for HDL cholesterol, but the effect was not significant. In men, the gene-dosage effect was not tested because of the higher-order interactions and limited sample size.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Some investigators reported that the G-A polymorphism is associated with variability in LDL-cholesterol or apo B concentrations rather than HDL-related variables (12, 27, 28), Akita et al (20), or Barre et al (19). Moreover, Matsunaga et al (24) observed that control subjects with the G/A genotype had significantly lower plasma concentrations of apo A-I. The inconsistency of these findings suggests that some of the observed associations could be the result of chance, especially considering the small sample sizes of several of the studies. Alternatively, the effects of the -75 bp G-A polymorphic site could be dependent on environmental factors that differ between study populations. This hypothesis gained support when Sigurdsson et al (15A allele was only associated with increased HDL-cholesterol or apo A-I concentrations in nonsmoking men. This gene-environment interaction was subsequently confirmed by a similar finding in Chinese subjects residing in Singapore (13). These findings are also consistent with the failure to detect significant associations in those studies with a high prevalence of smokers (18). Although the evidence suggests that smoking status may interact with the effect of genotype, many of the inconsistencies remain unexplained.
Some investigators reported that the G-A polymorphism is associated with variability in LDL-cholesterol or apo B concentrations rather than HDL-related variables (12, 27, 28). These findings provide some support for the alternative hypothesis that this mutation is not causative, but is in linkage disequilibrium with another functional mutation in one of the neighboring genes (16). Moreover, previous studies showed significant interactions between the G-A polymorphism and LDL-cholesterol response to different types of dietary intervention (26, 27), although other studies found no significant interactions (10, 28). Note that the gene-diet interactions for LDL-cholesterol response observed in the short-term dietary intervention studies (26, 27) were not significant when tested in the present population-based study examining long-term dietary habits. However, the trends in the present study (data not shown) were in the same direction as those previously reported (26, 27).
To address our primary goal, we examined whether fat intake is one of the factors that modify the association between the G-A polymorphism and HDL-cholesterol or apo A-I concentrations. Our study showed a significant interaction between dietary PUFA intake and the G-A polymorphism on plasma HDL-cholesterol and apo A-I concentrations. In general, our model showed that the A allele was associated with lower concentrations of HDL cholesterol and apo A-I. However, in subjects carrying the A allele, a higher PUFA intake may reverse the genetic effect. In women, this interaction was highly significant for HDL-cholesterol concentrations and approached significance for apo A-I concentrations. The complexity of this gene-diet interaction differed between men and women. In men, the PUFA-intake effect was only significant when interactions with alcohol consumption and tobacco smoking were also considered in the regression models (data not shown).
We propose that, in addition to the previously reported genotype-smoking interaction, the gene-diet interaction studied here could help explain some of the contradictory findings reported previously. According to our statistical model, a moderate consumption of PUFAs (4%–8% of total energy), common in most populations, would be consistent with the absence of significant genotype-phenotype associations as reported by several studies. An association between the G/G genotype and higher HDL-cholesterol concentrations would be expected in subjects consuming <4% of energy from PUFAs. However, such low PUFA intakes are uncommon in most populations, and only one study observed higher apo A-I concentrations in G/G homozygotes than in carriers of the A allele (24). Conversely, it would be expected that the A allele would be associated with significantly higher HDL-cholesterol concentrations in subjects consuming >8% of their daily energy intake as PUFAs than in subjects consuming lower amounts of PUFAs. Distinguishing the associations of various types of dietary fat with blood lipid concentrations in an observational study is difficult because of the multicolinearity of fat intakes. In the present study, separate multivariate regression models were computed for the different types of fat. Only PUFA intake showed significant results. When SFA and MUFA intakes were included in the PUFA regression model, the regression coefficient for PUFA intake remained significant.
The role of dietary fatty acids in regulating plasma lipoprotein concentrations is well documented (44). Compared with SFA intake, PUFA intake was shown to lower LDL- and HDL-cholesterol concentrations (45). However, these studies did not address genetic variability in response. When we considered the G-A polymorphism in our study, the lowering effect of PUFA intake on HDL-cholesterol concentrations was seen only in G/G subjects; the opposite effect was seen in carriers of the A allele. This observation, which requires further investigation in intervention studies, may be of particular relevance for dietary counseling designed to improve the lipid profile and thereby reduce the risk of atherosclerosis. However, we should keep in mind that other factors, such as BMI, alcohol intake, and physical activity, are also major modulators of HDL-cholesterol concentrations, response to diet, and genotype or phenotype associations (27, 46–49).
Dietary PUFAs can induce gene expression, as was shown in animal models and cultured cells (29). This polymorphic site, variably denoted as -75 bp, -76 bp, or -78 bp in different reports, is located in the 5' flanking region of the transcriptional start site of the apo A-I gene. Studies that used the more common G allele have not identified proteins binding at this specific site (33). However, a drug-responsive element was localized in the immediate neighborhood (50). In addition, the presence of the A substitution creates a 6-bp repeat that has homology to nuclear binding sites. Moreover, this repeat may allow the formation of a DNA secondary structure, which could interfere with protein interaction in the transcriptional apparatus. This agrees with results from some of the in vitro studies, which showed that the A allele was associated with lower in vitro transcriptional efficiency (30–32). Further evidence in support of this mutation and APOA1 gene transcription was provided by in vivo metabolic studies showing that subjects with the A allele had lower apo A-I production rates (31). These data are also consistent with our model, suggesting that under basal conditions the A allele is associated with lower HDL-cholesterol and apo A-I concentrations. Thus, PUFAs could interact differently with the transcriptional apparatus of the APOA1 gene, increasing transcription only in the presence of the A allele.
Another important finding of this research relates to the sex differences observed for genotype-PUFA interactions, which suggests a hormonal effect. Thyroid hormones, glucocorticoids, and estradiol enhance activity of the APOA1 gene, whereas retinoic acid and androgens decrease it (51, 52). Specifically, regulation of the APOA1 gene by estrogen may vary in terms of direction and magnitude depending not only on the presence of estrogen receptor-
and 17ß-estradiol, but also on the intracellular balance of estrogen receptor- and coactivators used by estrogen receptor- and the APOA1 enhancer (53). We hypothesized that the G-A polymorphism may be implicated in this mechanism. Alternatively, this mutation could be in linkage disequilibrium with a functional mutation in the promoter or enhancer regions of the APOA1 gene.
Although we cannot rule out the possibility of confounding in our PUFA hypothesis, it is unlikely that confounding would fully explain the strong interaction observed. The lowering effect of the A allele on HDL-cholesterol concentrations in women also showed a clear gene-dosage relation when the interaction with PUFAs was considered. A/A subjects had lower HDL-cholesterol concentrations than did heterozygous G/A individuals. Moreover, the elevating effect of PUFA intake on HDL-cholesterol concentrations was greater in A/A subjects than in G/A subjects.
In summary, PUFA intake modulates the effect of the G-A polymorphism in the APOA1 gene promoter, resulting in higher concentrations of HDL cholesterol in individuals carrying the A allele. According to these results, the known HDL-cholesterol-lowering effect attributed to the consumption of PUFAs does not apply to women carrying the A allele. Furthermore, this specific population group might benefit from a high-PUFA diet, which should increase HDL-cholesterol concentrations and thereby reduce CHD risk. Our results illustrate the complexity of polymorphism-phenotype associations and underscore the importance of accounting for interactions between genes and environmental factors in population genetic studies.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A substitution at position -75 of the apolipoprotein A-I gene promoter—evidence against a direct effect on HDL cholesterol levels. Arterioscler Thromb Vasc Biol 1995;15:1740–5.A transition) in the -78 position of the apolipoprotein A-I promoter increases transcription efficiency. J Biol Chem 1994;269: 17371–4.This article has been cited by other articles:
|
|
 |
|
  H. M Macdonald, F. E McGuigan, S. A Lanham-New, W. D Fraser, S. H Ralston, and D. M Reid Vitamin K1 intake is associated with higher bone mineral density and reduced bone resorption in early postmenopausal Scottish women: no evidence of gene-nutrient interaction with apolipoprotein E polymorphisms Am. J. Clinical Nutrition, May 1, 2008; 87(5): 1513 - 1520. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. A. Patsopoulos, A. Tatsioni, and J. P. A. Ioannidis Claims of Sex Differences: An Empirical Assessment in Genetic Associations JAMA, August 22, 2007; 298(8): 880 - 893. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M Ordovas Genetic interactions with diet influence the risk of cardiovascular disease Am. J. Clinical Nutrition, February 1, 2006; 83(2): 443S - 446S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. W Duff, P. Libby, J. M Ordovas, and P. R Reilly The future of living well to 100 Am. J. Clinical Nutrition, February 1, 2006; 83(2): 488S - 490S. [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. Mooradian, M. J. Haas, and N. C. W. Wong The Effect of Select Nutrients on Serum High-Density Lipoprotein Cholesterol and Apolipoprotein A-I Levels Endocr. Rev., February 1, 2006; 27(1): 2 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Mutch, W. Wahli, and G. Williamson Nutrigenomics and nutrigenetics: the emerging faces of nutrition FASEB J, October 1, 2005; 19(12): 1602 - 1616. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Deng, M. B. Elam, H. G. Wilcox, L. M. Cagen, E. A. Park, R. Raghow, D. Patel, P. Kumar, A. Sheybani, and J. C. Russell Dietary Olive Oil and Menhaden Oil Mitigate Induction of Lipogenesis in Hyperinsulinemic Corpulent JCR:LA-cp Rats: Microarray Analysis of Lipid-Related Gene Expression Endocrinology, December 1, 2004; 145(12): 5847 - 5861. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kaput and R. L. Rodriguez Nutritional genomics: the next frontier in the postgenomic era Physiol Genomics, January 15, 2004; 16(2): 166 - 177. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Martinez, M. S. Corbalan, A. Sanchez-Villegas, L. Forga, A. Marti, and M. A. Martinez-Gonzalez Obesity Risk Is Associated with Carbohydrate Intake in Women Carrying the Gln27Glu {beta}2-Adrenoceptor Polymorphism J. Nutr., August 1, 2003; 133(8): 2549 - 2554. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Kammerer, D. L. Rainwater, L. A. Cox, J. L. Schneider, M. C. Mahaney, J. Rogers, and J. L. VandeBerg Locus Controlling LDL Cholesterol Response to Dietary Cholesterol Is on Baboon Homologue of Human Chromosome 6 Arterioscler Thromb Vasc Biol, October 1, 2002; 22(10): 1720 - 1725. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, <4%; , 4%–8%; 
, >8% of energy) in women. Means were adjusted for age, body mass index, alcohol consumption, tobacco smoking, and intakes of energy, saturated fatty acids, monounsaturated fatty acids, and PUFAs. The P values shown were obtained for the interaction between APOA1 genotype and PUFA intake in the multivariate linear regression models.