点击显示 收起
1 From the Institutes for Research in Extramural Medicine (GB, JMD, GN, CDAS, LMB, and RJH) and Cardiovascular Research (CDAS and RJH), VU University Medical Center, Amsterdam, Netherlands; the Center for Nutrition and Health, National Institute for Public Health and the Environment, Bilthoven, Netherlands (EJMF and MCO); and the Departments of Internal Medicine and Clinical Chemistry, Erasmus MC, Rotterdam, Netherlands (HJ).
2 Supported by the Dutch Diabetes Research Foundation (grant DFN 98901) and the Dutch Organization for Scientific Research, NWO (grant 940-35-034). 3 Address reprint requests to G Bos, Institute for Research in Extramural Medicine, VU University Medical Center, Van der Boechorststraat 7, 1081 BT Amsterdam, Netherlands. E-mail: g.bos.emgo{at}med.vu.nl.
ABSTRACT
Background: Gene-nutrient interactions affecting hepatic lipase (HL) activity may contribute to the interindividual variability of the cardiovascular disease risk associated with dietary fat intake.
Objective: We determined the associations of dietary fat intake with postheparin HL activity and the possible modifying effect of the HL 480CT polymorphism on these associations.
Design: Subjects were recruited from participants in the 20002001 follow-up examination of the Hoorn Study. HL activity was determined in postheparin plasma in a sample of 211 men and 218 women aged 6087 y. Information about dietary intake of the participants was obtained with a validated food-frequency questionnaire. Linear regression was performed, adjusted for age.
Results: Total dietary fat was positively associated with HL activity (standardized ß: 0.11; 95% CI: 0.02, 0.21), and this association was also seen for saturated fat (0.10; 0.01, 0.20) and monounsaturated fatty acid (0.10; 0.01, 0.19). We observed a significant interaction of the HL polymorphism with the relation between total fat intake and HL activity. The association of total fat with HL activity was stronger in subjects with CT (0.27; 0.11, 0.43) and TT (0.39; 0.22, 1.00) genotypes than in subjects with the CC genotype (0.06; 0.06, 0.18; P for interaction < 0.10). The interaction remained statistically significant in models that included age, sex, carbohydrate and protein intakes, and insulin or body mass index.
Conclusions: Higher intakes of total and saturated fat were positively associated with higher HL activity. In addition, the observed association of total fat with HL activity was modified by the HL480CT polymorphism, after adjustment for age, sex, carbohydrate and protein intakes, and insulin or body mass index.
Key Words: Hepatic lipase gene-nutrient interaction fat intake
INTRODUCTION
Dietary fat intake is an important predictor of serum lipids and lipoproteins (1-3). Reduction in saturated fat intake is associated with decreases in plasma LDL-cholesterol and HDL-cholesterol concentrations. Triacylglycerol and phospholipids in HDL and LDL particles are hydrolyzed by hepatic lipase (HL), a lipolytic enzyme, resulting in the formation of smaller and denser particles.
HL activity has been shown to be determined by genetic factors (4). A common 480CT substitution, also denominated LIPC 514CT (5), has been described in the promoter region of the HL gene (6). The T allele is associated with decreased plasma HL activity (6, 7). Dietary fat has recently been shown to modify the association between HL 480CT polymorphism and plasma HDL-cholesterol concentrations in such a way that the presence of the T allele was associated with higher HDL cholesterol, only in individuals who usually consume a low-fat diet. In contrast, the TT genotype was associated with lower HDL cholesterol in individuals who usually consume a high-fat diet (8). Information about the effects of dietary fat on HL activity is limited and contradictory. A high-fat diet (46% of energy) increased HL activity in 43 healthy men compared with a low-fat diet (24% of energy) (9). However, no differences in HL activity were observed after dietary fat restriction in 72 healthy women (10). In an intervention study with low-fat and high-fat diets in men, an increase in dietary saturated fat was associated with decreased HL activity (11).
In this cross-sectional observational study of 429 men and women, we investigated the association of both dietary fat intake and HL 480CT polymorphism with HL activity. In addition, we determined whether the HL 480CT polymorphism modifies the association of dietary fat intake with plasma postheparin HL activity.
SUBJECTS AND METHODS
Design and population
The Hoorn Study is a population-based cohort study of glucose metabolism and cardiovascular disease risk factors among the inhabitants of the municipality of Hoorn, Netherlands, which started in 1989 and consisted of 2484 subjects, as described before (12). In 20002001, a follow-up was conducted in selected subjects, then aged 6087 y. We invited all surviving subjects with type 2 diabetes (n = 176) to participate and also individuals chosen from random samples with normal glucose metabolism (n = 705) or impaired glucose metabolism (n = 193) on the basis of their glucose metabolism status at the previous examination in 19961998 (13). Of the 1074 individuals invited, 648 (60.3%) participated in the 20002001 follow-up examination. Among the reasons for not participating in the follow-up examinations were lack of interest (30%), comorbidity (23%), advanced age (7%), unwillingness to travel (6%), participation considered too time consuming (6%), and miscellaneous reasons (15%). Thirteen percent were complete nonresponders. For the present study, cross-sectional data of this 20002001 follow-up examination were analyzed. A sample of 585 participants was invited for the postheparin test, of which 566 participated. The Ethical Review Committee of the VU University Medical Center approved the study. Informed consent was obtained from all participants.
Postheparin plasma HL activity
HL activity was measured by using an immunochemical method as described previously in plasma collected 20 min after contralateral intravenous administration of 50 IU heparin/kg body weight (Leo Pharmaceutical Products, Weesp, Netherlands). One hundred seven samples (19%) were excluded from the analysis because of low activities for lipoprotein lipase (LPL) and HL in postheparin plasma. The cutoffs for low lipase activities were selected on the basis of visual inspection of a scatterplot. Activities were considered low if both LPL activity was <50 U/L and HL activity was <72 U/L. Inclusion of other subjects with incomplete heparin injections did not change the results (data not shown). Most excluded samples were among subjects with impaired glucose metabolism and type 2 diabetes. However, no statistically significant differences were observed between individuals who were excluded and the rest of the subjects with respect to lipoproteins, body mass index (BMI), and dietary fat intake (data not shown). Furthermore, HL activity showed the known associations with age, sex, HDL-cholesterol concentration, and triacylglycerol concentration (4, 14). The postheparin test was repeated in a sample of subjects with normal and impaired glucose metabolism and diabetes with both low LPL and HL activities (n = 10). Normal LPL and HL activities were measured in these subjects, indicating that insufficient heparin delivery was the cause of the low activities.
Genotyping HL polymorphism
DNA was extracted from blood. We used the polymerase chain reaction method as described by Berk-Planken et al (15) to assess the presence of the C>T variance in the HL gene promoter.
Glycemic control and lipids
All participants underwent a 75-g oral-glucose-tolerance test, except for participants with capillary fasting whole blood glucose concentrations 8 mmol/L or participants with a previous diagnosis of diabetes who were treated with oral glucose-lowering medication or insulin. Fasting glucose and 2-h postload glucose were measured in plasma with the hexokinase method (Roche Diagnostics GmbH, Mannheim, Germany). Glycated hemoglobin was analyzed by HPLC (reference range: 4.36.1%). Fasting plasma glucose, 2-h postload plasma glucose, total-cholesterol, HDL-cholesterol, and triacylglycerol concentrations were measured by enzymatic methods (Roche, Mannheim, Germany). LDL cholesterol was directly determined by the N-geneous assay (Genzyme, Cambridge, MA). Insulin was determined by using a two-site immunoradiometric test. Paired monoclonal antibodies were used (Medgenix, Diagnostics, Fleurus, Belgium).
Dietary intake
Information about the dietary intake of the participants was obtained by a validated food-frequency questionnaire (16, 17), which was linked to an extended version of the computerized Dutch Food Composition Table 1996 (18). Intake of macronutrients was expressed as the percentage of energy as fat [fat (in kJ)/total energy (in kJ) x 100].
Statistical analysis
Three glucose metabolism categories (normal, impaired glucose metabolism, and diabetes) were defined according to the World Health Organization criteria (19). We excluded people with missing data on primary variables of interest (glucose metabolism, n = 7; HL activity, n = 107; HL 480C
RESULTS
HL activity was significantly lower in carriers of the TT genotype than in subjects with the CC or CT genotype (P for trend < 0.001; Table 1). Furthermore, the carriers of the TT genotype had a more favorable lipid profile, ie, with higher HDL cholesterol (P for trend = 0.03). The SD for fasting glucose in the TT group is large, because there was one outlier with a fasting glucose concentration of 18.5 mmol/L. Nevertheless, we did not exclude this subject from the analysis; the extreme value was a true observation because the person was a diabetic subject. No significant differences in dietary fat intake across HL polymorphisms were detected. Genotype frequencies did not deviate from the Hardy-Weinberg equilibrium.
View this table:
TABLE 1. Baseline characteristics of the population according to hepatic lipase 480CT polymorphism1
The HL 480CT polymorphism was the strongest predictor of HL activity [ß (95% CI): 28 (54, 4) for the CT and 138 (210; 66) for the TT genotypes compared with the CC genotype]. Intakes of total fat, saturated fat, and monounsaturated fat were positively associated with HL activity, whereas intake of polyunsaturated fatty acid was not related to HL activity (Table 2). A test for interaction between total fat and HL polymorphism was significant (P < 0.10), indicating that the association of total dietary fat intake with HL activity differs across categories of HL 480CT polymorphism.
View this table:
TABLE 2. Standardized ß coefficients (95% CIs) for age-adjusted associations of dietary fat intake with hepatic lipase (HL) activity by HL 480CT polymorphism: univariate regression analyses1
In Table 3, stratified multivariate analyses of total fat intake and HL activity are shown. After adjustment for carbohydrate intake, protein intake, and insulin concentration (model 2), total fat intake tended to be positively associated with HL activity in subjects with CT and TT genotypes, but not in subjects with the CC genotype; however, none of these associations was statistically significant. These associations were seen after additional adjustment for insulin and not after adjustment for carbohydrate and protein intake only (model 1). After adjustment for glucose metabolism status or BMI (models 3 and 4), a positive association was observed between total fat intake and HL activity in the same direction as in models 1 and 2. These associations were statistically significant for the CT genotype in model 3. The overall F test for interactions between total fat and HL polymorphisms reached statistical significance after adjustment for carbohydrate and protein intakes, insulin (P = 0.07; model 2), and BMI (P = 0.04; model 4).
View this table:
TABLE 3. Standardized ß coefficients (95% CI) for associations of dietary total fat intake with hepatic lipase (HL) activity by HL 480CT polymorphism: multivariate regression analysis
DISCUSSION
In this population-based study, we describe the effects of dietary fat intake on HL activity. We found that the T allele was associated with lower HL activity, and this relation was independent of the amount and type of fat consumed. In addition, we found a stronger association of total dietary fat intake with HL activity in subjects with CT and TT genotypes than in subjects with the CC genotype, after adjustment for age, sex, carbohydrate and protein intakes, and insulin or BMI.
In the Framingham study, the T allele was associated with significantly greater HDL-cholesterol concentrations only in subjects consuming <30% of energy from fat (8). The observed gene-nutrient interaction with HDL cholesterol was in the same direction as our findings with HL activity. In line with our results, higher intake of saturated fat was associated with higher HL activity in 43 men in a study of Dreon et al (11). It has been described that HL activity is quite variable, indicating that other factors affect HL activity. HL activity is higher in men, in smokers, and in those with diabetes, and it is positively associated with intraabdominal fat and BMI (4, 14). Thus, HL activity is associated with several components of the insulin resistance syndrome (15, 20-23).
The design of our study did not allow us to investigate the mechanism by which dietary fat possibly interacts with the 480CT polymorphism. Investigations in rats have shown that the dietary lipid profile is an important predictor of tissue phospholipid composition (24), which may, in turn, influence insulin sensitivity by altering membrane fluidity and insulin signaling (25, 26). One possibility is a direct effect of fatty acids on HL expression. It has been shown that fatty acids enhance HL activity secreted from HepG2 cells (27). Therefore, HL activity could be modulated by a change in the supply of fatty acids to the liver either from the diet or from the omental fat stores. Additionally, it has been hypothesized that the underlying mechanism could be insulin resistance of the T allele at the level of gene expression (H Jansen, unpublished observations, 2004). Previous studies have reported that insulin up-regulates the activity of HL by insulin-responsive elements in the HL promoter, suggesting that variants in this promoter, as examined in this study, may affect the ability of insulin to stimulate HL activity (28). Jansen et al (6) have reported an association between variants of the HL promoter and insulin resistance. However, in our study, adjustment for insulin concentration did not change the observed association between total fat intake and HL activity in any of the genotypes.
Some limitations of our study should be noted. First, the number of subjects in the groups with the TT genotype was small (n = 12), thus reducing the statistical power for conclusions about this group. However, this study is the first to measure HL activity in such a large population and to investigate relations between HL activity and HL polymorphisms. Furthermore, we observed different associations in the CT genotype group. Second, our findings were obtained in the elderly. Therefore, we may have underestimated the association of dietary fat intake with HL activity because of selective morbidity and mortality of individuals with unhealthy lifestyle and food consumption and consequently unfavorable lipase activity or related lipoproteins.
In summary, total dietary fat and saturated fat were positively associated with HL activity. In addition, we observed a gene-nutrient interaction (in a model that included BMI), such that the association of the amount of total fat consumed with HL activity depended on the HL 480CT polymorphism. Gene-nutrient interactions may help to explain the different outcomes of plasma lipoprotein in response to dietary fat and the conflicting results about HL activity and cardiovascular disease risk.
ACKNOWLEDGMENTS
We thank the research assistants of the Diabetes Onderzoeks Centrum in Hoorn for their cooperation with this study. We also thank Annemarie Becker for carrying out the postheparin lipase test.
GB performed the data analyses and wrote the drafts and final paper. JMD, MCO, EJMF, GN, CDAS, LMB, RJH, and HJ provided advice concerning the presentation and interpretation of the results. JMD, EJMF, and HJ participated in the conception of the study. JMD, LMB, CDAS, RJH, and GN designed and collected data for the Hoorn Study. MCO provided the nutritional data. None of the authors had a conflict of interest.
REFERENCES