The T111I mutation in the EL gene modulates the impact of dietary fat on the HDL profile in women

The objective of the present study was to examine the impact of the T111I missense mutation in exon 3 of the endothelial lipase (EL) gene on HDL and its potential interaction effect with dietary fat. The study sample included 281 women and 216 men aged between 17 and 76 years from the Québec Family Study. Plasma HDL3-C levels of I111I homozygote women were higher compared with those of women carrying the wild-type allele (P = 0.03). These differences were not attenuated when adjusted for levels of obesity and were not observed among men. Dietary PUFA interacted with the T111I mutation to modulate apolipoprotein A-I (apoA-I) and HDL3-C levels among women. Specifically, a diet rich in PUFA was associated with increased apoA-I levels among women carriers of the I111 allele and with decreased apoA-I among women homozygotes for the wild-type allele (P = 0.002). A similar interaction was observed with plasma HDL3-C levels (P = 0.003). These interactions were not observed among men.In conclusion, the EL T111I mutation appears to have a modest effect on plasma HDL levels. The gene-diet interaction among women, however, suggests that the T111I missense mutation may confer protection against the lowering effect of a high dietary PUFA intake on plasma apoA-I and HDL3-C levels. The objective of the present study was to examine the impact of the T111I missense mutation in exon 3 of the endothelial lipase (EL) gene on HDL and its potential interaction effect with dietary fat. The study sample included 281 women and 216 men aged between 17 and 76 years from the Québec Family Study. Plasma HDL3-C levels of I111I homozygote women were higher compared with those of women carrying the wild-type allele (P = 0.03). These differences were not attenuated when adjusted for levels of obesity and were not observed among men. Dietary PUFA interacted with the T111I mutation to modulate apolipoprotein A-I (apoA-I) and HDL3-C levels among women. Specifically, a diet rich in PUFA was associated with increased apoA-I levels among women carriers of the I111 allele and with decreased apoA-I among women homozygotes for the wild-type allele (P = 0.002). A similar interaction was observed with plasma HDL3-C levels (P = 0.003). These interactions were not observed among men. In conclusion, the EL T111I mutation appears to have a modest effect on plasma HDL levels. The gene-diet interaction among women, however, suggests that the T111I missense mutation may confer protection against the lowering effect of a high dietary PUFA intake on plasma apoA-I and HDL3-C levels. Endothelial lipase (EL) has recently been cloned by two laboratories using different approaches (1Jaye M. Lynch K.J. Krawiec T. Marchadier D. Maugeais C. Doan K. South V. Amin D. Perrone M. Rader D.J. A novel endothelial-derived lipase that modulates HDL metabolism.Nat. Genet. 1999; 21: 424-428Google Scholar, 2Hirata K. Dichek H.L. Cioffi J.A. Choi S.Y. Leeper N.J. Quintana L. Kronmal G.S. Cooper A.D. Quertermous T. Cloning of a unique lipase from endothelial cells extends the lipase gene family.J. Biol. Chem. 1999; 274: 14170-14175Google Scholar). This lipase has been identified as a new member of the triglyceride (TG) lipase gene family and has been shown to be highly homologous to hepatic lipase (HL) and lipoprotein lipase (LPL), both playing key roles in lipoprotein metabolism. Despite strong similarities and conserved features among members of the lipase family, EL distinguishes itself from other lipases by being the only one expressed in endothelial cells and by its specific substrate preference (1Jaye M. Lynch K.J. Krawiec T. Marchadier D. Maugeais C. Doan K. South V. Amin D. Perrone M. Rader D.J. A novel endothelial-derived lipase that modulates HDL metabolism.Nat. Genet. 1999; 21: 424-428Google Scholar, 2Hirata K. Dichek H.L. Cioffi J.A. Choi S.Y. Leeper N.J. Quintana L. Kronmal G.S. Cooper A.D. Quertermous T. Cloning of a unique lipase from endothelial cells extends the lipase gene family.J. Biol. Chem. 1999; 274: 14170-14175Google Scholar). There is indeed minimal sequence homology between the lid domain of EL and other intravascular lipases (1Jaye M. Lynch K.J. Krawiec T. Marchadier D. Maugeais C. Doan K. South V. Amin D. Perrone M. Rader D.J. A novel endothelial-derived lipase that modulates HDL metabolism.Nat. Genet. 1999; 21: 424-428Google Scholar, 2Hirata K. Dichek H.L. Cioffi J.A. Choi S.Y. Leeper N.J. Quintana L. Kronmal G.S. Cooper A.D. Quertermous T. Cloning of a unique lipase from endothelial cells extends the lipase gene family.J. Biol. Chem. 1999; 274: 14170-14175Google Scholar), a feature known to be critical in determining substrate specificity (3Dugi K.A. Dichek H.L. Santamarina-Fojo S. Human hepatic and lipoprotein lipase: the loop covering the catalytic site mediates lipase substrate specificity.J. Biol. Chem. 1995; 270: 25396-25401Google Scholar). Previous studies indicated that EL was primarily a phospholipase (1Jaye M. Lynch K.J. Krawiec T. Marchadier D. Maugeais C. Doan K. South V. Amin D. Perrone M. Rader D.J. A novel endothelial-derived lipase that modulates HDL metabolism.Nat. Genet. 1999; 21: 424-428Google Scholar, 2Hirata K. Dichek H.L. Cioffi J.A. Choi S.Y. Leeper N.J. Quintana L. Kronmal G.S. Cooper A.D. Quertermous T. Cloning of a unique lipase from endothelial cells extends the lipase gene family.J. Biol. Chem. 1999; 274: 14170-14175Google Scholar), with particular affinity for HDL compared with other lipoprotein subfractions (4McCoy M.G. Sun G.S. Marchadier D. Maugeais C. Glick J.M. Rader D.J. Characterization of the lipolytic activity of endothelial lipase.J. Lipid Res. 2002; 43: 921-929Google Scholar). In vivo studies in mice have shown that EL plays a major role in modulating HDL metabolism (1Jaye M. Lynch K.J. Krawiec T. Marchadier D. Maugeais C. Doan K. South V. Amin D. Perrone M. Rader D.J. A novel endothelial-derived lipase that modulates HDL metabolism.Nat. Genet. 1999; 21: 424-428Google Scholar, 5Ishida T. Choi S. Kundu R.K. Hirata K.I. Rubin E.M. Cooper A.D. Quertermous T. Endothelial lipase is a major determinant of HDL level.J. Clin. Invest. 2003; 111: 347-355Google Scholar, 6Jin W. Millar J.S. Broedl U. Glick J.M. Rader D.J. Inhibition of endothelial lipase causes increased HDL cholesterol levels in vivo.J. Clin. Invest. 2003; 111: 357-362Google Scholar, 7Ma K. Cilingiroglu M. Otvos J.D. Ballantyne C.M. Marian A.J. Chan L. Endothelial lipase is a major genetic determinant for high-density lipoprotein concentration, structure, and metabolism.Proc. Natl. Acad. Sci. USA. 2003; 100: 2748-2753Google Scholar). Because HDL-cholesterol (HDL-C) levels remain a key component of the lipid risk factors for coronary heart disease (CHD), the discovery of this new player in HDL metabolism warrants extensive investigation among humans. It has been suggested that more than 50% of the variation in HDL-C levels in humans may be genetically determined (8Cohen J.C. Vega G.L. Grundy S.M. Hepatic lipase: new insights from genetic and metabolic studies.Curr. Opin. Lipidol. 1999; 10: 259-267Google Scholar). To date, only two studies have investigated how variants in the EL gene may contribute to variations in HDL concentrations (7Ma K. Cilingiroglu M. Otvos J.D. Ballantyne C.M. Marian A.J. Chan L. Endothelial lipase is a major genetic determinant for high-density lipoprotein concentration, structure, and metabolism.Proc. Natl. Acad. Sci. USA. 2003; 100: 2748-2753Google Scholar, 9deLemos A.S. Wolfe M.L. Long C.J. Sivapackianathan R. Rader D.J. Identification of genetic variants in endothelial lipase in persons with elevated high-density lipoprotein cholesterol.Circulation. 2002; 106: 1321-1326Google Scholar). Even though the allele frequency of several mutations in the EL gene identified by deLemos et al. (9deLemos A.S. Wolfe M.L. Long C.J. Sivapackianathan R. Rader D.J. Identification of genetic variants in endothelial lipase in persons with elevated high-density lipoprotein cholesterol.Circulation. 2002; 106: 1321-1326Google Scholar) was higher among individuals with high plasma HDL-C levels compared with a control group, none of the newly identified genetic variants showed significant association with plasma HDL-C levels. Among the 17 genetic variants identified in that study, one common mutation in exon 3 deserved greater scrutiny, as it was responsible for a significant amino acid change (T111I) that could potentially be associated with an altered intravascular EL activity. A recent report on a group of subjects from the Lipoprotein and Coronary Atherosclerosis Study (LCAS) observed higher plasma HDL-C and apolipoprotein C-III (apoC-III) levels among men carriers of the mutation (I111) as well as higher HDL-C/LDL-C and apoA-I/apoB ratios (7Ma K. Cilingiroglu M. Otvos J.D. Ballantyne C.M. Marian A.J. Chan L. Endothelial lipase is a major genetic determinant for high-density lipoprotein concentration, structure, and metabolism.Proc. Natl. Acad. Sci. USA. 2003; 100: 2748-2753Google Scholar). The aim of the present study was to characterize further the effect of this T111I missense mutation in exon 3 of the EL gene on the HDL profile of subjects from the Québec Family Study (QFS). As environmental factors such as dietary fat intake are also known to modify plasma HDL-C levels as well as CHD risk (10Mensink R.P. Katan M.B. Effect of dietary fatty acids on serum lipids and lipoproteins. A meta-analysis of 27 trials.Arterioscler. Thromb. 1992; 12: 911-919Google Scholar), we also sought to examine the potential impact of the interaction with the T111I variant and the diet on the HDL profile of these individuals. The QFS is a cohort of white men and women who were recruited to participate in studies designed to investigate factors involved in the etiology of obesity (11Bouchard C. Genetic epidemiology, association, and sib-pair linkage: results from the Québec Family Study.in: Bray G.A. Ryan D.H. Molecular and Genetic Aspects of Obesity. Louisiana State University Press, Baton Rouge, LA1996: 470-481Google Scholar). A total of 951 subjects from 223 families are currently enrolled in the QFS. Subjects gave their written consent to participate in this study, which received the approval of the Medical Ethics Committee of Laval University. Only subjects with a complete lipoprotein-lipid profile, with anthropometric measurements and genomic DNA available, and in whom dietary intake was assessed, were eligible to be included in the present analysis. None had diabetes or were treated for cardiovascular disease. The subsample used in the present study included 281 women and 216 men aged between 17 and 76 years. Body mass index (BMI) ranged between 16.8 and 64.9 kg/m2. The study sample comprised men and women from 172 two-generation families. There were 162 and 335 individuals from the parental and offspring generations, respectively. Participants with plasma TG levels >4.5 mmol/l were excluded to allow the calculation of plasma LDL-C concentration with the Friedewald formula (N = 1 woman and 3 men). Inclusion of these individuals did not alter the results (data not shown). Postmenopausal status was defined as having been without menses for a year, and women were asked about the use of hormonal replacement therapy. BMI was derived from body weight divided by height squared (kg/m2). Waist circumference was measured according to procedures recommended by the Airlie Conference (12Lohman T.G. Roche A.F. Martorel R. The Arle (VA) consensus conference.in: Lohman T.G. Roche A.F. Martorel R. Anthropometric Standardization Reference Manual. Human Kinetics Publishers, Champagne, IL1988: 39-80Google Scholar). The amount of visceral adipose tissue was obtained by computed axial tomography as described previously (13Sjostrom L. Kvist H. Cederblad A. Tylen U. Determination of total adipose tissue and body fat in women by computed tomography, 40K, and tritium.Am. J. Physiol. 1986; 250: E736-E745Google Scholar). Plasma cholesterol and TG levels in the various lipoprotein subfractions were determined enzymatically with commercial kits as described elsewhere (14Leclerc S. Bouchard C. Talbot J. Gauvin R. Allard C. Association between serum high-density lipoprotein cholesterol and body composition in adult men.Int. J. Obes. 1983; 7: 555-561Google Scholar). Plasma LDL-C levels were calculated using the Friedewald formula (15Friedewald W.T. Levy R.I. Fredrickson D.S. Estimation of the concentration of low density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge.Clin. Chem. 1972; 18: 499-502Google Scholar). Plasma VLDL (d < 1.006 g/ml) was isolated by ultracentrifugation (50,000 rpm) using a Beckman 50.3 Ti rotor (Beckman, Palo Alto, CA) (16Moorjani S. Dupont A. Labrie F. Lupien P.J. Brun L.D. Gagné C. Giguère M. Bélanger A. Increase in plasma high density lipoprotein concentration following complete androgen blockage in men with prostatic carcinoma.Metabolism. 1987; 36: 244-250Google Scholar). HDL particles were isolated from the bottom fraction (>1.006 g/ml) after precipitation of apoB-containing lipoproteins with heparin and MnCl2 (17Albers J.J. Warnick G.R. Wiebe D. King P. Steiner P. Smith L. Breckenridge C. Chow A. Kuba K. Weidman S. Arnett H. Wood P. Shlagenhaft A. Multi-laboratory comparison of three heparin-Mn2+ precipitation procedures for estimating cholesterol in high-density lipoprotein.Clin. Chem. 1978; 24: 853-856Google Scholar). HDL2 was then precipitated from the d > 1.006 g/ml HDL fraction with a 4% solution of low-molecular-mass dextran sulfate (15–20 kDa) obtained from SOCHIBO (Boulogne, France) (18McManus R.M. Jumpson J. Finegood D.T. Clandinin M.T. Ryan E.A. A comparison of the effects of n-3 fatty acids from linseed oil and fish oil in well-controlled type II diabetes.Diabetes Care. 1996; 19: 463-467Google Scholar). The cholesterol concentration in the supernatant (HDL3) was measured and HDL2 was obtained by subtraction. The TG and cholesterol contents of the infranatant fraction were measured before and after the precipitation step for measurement of LDL and HDL compositions. The apoB and apoA-I concentrations were measured by the rocket immunoelectrophoretic method of Laurell (19Laurell C.B. Quantitative estimation of proteins by electrophoresis in agarose gel containing antibodies.Anal. Biochem. 1966; 15: 45-52Google Scholar) as previously described (16Moorjani S. Dupont A. Labrie F. Lupien P.J. Brun L.D. Gagné C. Giguère M. Bélanger A. Increase in plasma high density lipoprotein concentration following complete androgen blockage in men with prostatic carcinoma.Metabolism. 1987; 36: 244-250Google Scholar). LDL and HDL size were determined by gradient gel electrophoresis as described earlier (20Lamarche B. St. Pierre A.C. Ruel I.L. Cantin B. Dagenais G.R. Despres J.P. A prospective, population-based study of low density lipoprotein particle size as a risk factor for ischemic heart disease in men.Can. J. Cardiol. 2001; 17: 859-865Google Scholar, 21Perusse M. Pascot A. Despres J.P. Couillard C. Lamarche B. A new method for HDL particle sizing by polyacrylamide gradient gel electrophoresis using whole plasma.J. Lipid Res. 2001; 42: 1331-1334Google Scholar). Total caloric intake, including carbohydrate, protein, and lipid intakes, was derived from a 3 day activity record, which was based on two weekdays and one weekend day (22Tremblay A. Després J.P. Leblanc C. Bouchard C. The reproductibility of a three-day dietary record.Nutr. Res. 1983; 3: 819-830Google Scholar). In order to design intronic primers for the amplification of exon 3, genomic sequences were sought for the intronic regions surrounding this EL gene exon. To do this, we compared the mRNA sequence of the EL gene with a contiguous genomic DNA region taken from sequences of an overlapping clone from GenBank (Accession Number AC091170). Intronic primers were then designed using the Primer 3.0 software available on the Whitehead Institute/MIT Center for Genome Research server (http://www-genome.wi.mit.edu/cgi-bin/primer/primer3.cgi). Exon 3 was amplified from genomic DNA using specific primers derived from the 5′ and 3′ ends of the intronic sequence (5′-ATTGGGAAGAAGGTCATATAGAAG-3′ and 5′-CTTAAGAAGATTGGGTTTGAGATCC-3′). PCR conditions were as follows: reaction volume was 50 μl, 1 unit AmpliTaq DNA polymerase with GeneAmp (Roche) in the buffer recommended by the manufacturer, 1.5 mM MgCl2, 0.2 mM deoxynucleoside triphosphate, primers at final concentration of 0.6 mM, and 2.5 ng of template genomic DNA. PCR products were purified with hydrophilic GF/C filter (Whatman). Sequencing reactions were performed using BigDye Terminators v3.0 (PE Applied BioSystems, Foster City, CA), and the products were analyzed on ABI 3700 automated sequencers (PE Applied BioSystems). The data files were collected using the Data Collection v1.1.1 and then processed using the Sequencing Analysis v3.7 software (PE Applied BioSystems). Sequence data were analyzed using the Staden Package software (Medical Research Council Laboratory of Molecular Biology, Cambridge, UK). Differences between genotype groups were assessed by ANOVA. The MIXED procedure in SAS (SAS Institute, v.8, Cary, NC) was used to adjust for nonindependence among family members. The prevalence of menopause among T111T, T111I, and I111I women was compared by the χ2 test. To explore the combined effect of the T111I mutation and dietary fat intake on the HDL profile, two genotype groups (T111T homozygotes vs. I allele carriers) were further divided according to the median of the distributions for the various dietary fats among women (total fat: 33.5% of daily calories; saturated fat: 10.3% of daily calories; monounsaturated fat: 10.9% of daily calories; and polyunsaturated fat: 4.2% of daily calories). Multivariate adjustment was performed to adjust for the potential confounding effect of other variables as indicated in the tables and the figure. Variables abnormally distributed were log transformed prior to analysis. All statistical analyses were performed using the SAS package. The frequency of the I111 mutated allele among subjects from the QFS was 32%, and the genotype frequencies distribution showed a Hardy-Weinberg equilibrium (χ2 = 0.01; P = 0.99). Table 1 shows anthropometric and dietary data in each of the genotypic groups. The three groups of women were similar with respect to age, BMI, energy intake, and dietary fat intake. Among men, differences between the T111T and the T111I groups were observed for the waist circumference (P = 0.03). Approximately 24% of women were menopausal (23% in T111T, 26% among T111I heterozygotes, and 14% among I111I homozygotes, P = 0.43). The T111I mutation was not associated with variations in plasma levels of apoB-containing lipoproteins or total plasma cholesterol and TG levels among women (Table 2). The HDL profile was also very similar among the three genotypes, with the exception of plasma HDL3-C levels. These levels, which were significantly higher among women homozygous for the mutated allele (I111I), compared with women heterozygous and homozygous for the wild-type allele (T111T and T111I), were independent of familial relationship and age. Further adjustment for menopausal status, the use of hormonal replacement therapy, and levels of visceral fat, BMI, or waist girth did not attenuate the impact of the T111I mutation on plasma HDL3-C levels (data not shown). Among men, no difference between the three genotypes for any of the lipid variables was observed (Table 3).TABLE 1Anthropometric characteristics of women and men from the QFS and energy and macronutrient intake according to the T111I missense mutation of EL geneT111/T111T111/I111I111/I111PWomenN = 153N = 100N = 28 Age (years)aP value after adjustment for the nonindependence among family members and age (except when the dependent variable is age) as well as for menopausal status and estrogen therapy (in women).40.2 ± 14.339.6 ± 14.435.8 ± 14.30.47 BMI (kg/m2)aP value after adjustment for the nonindependence among family members and age (except when the dependent variable is age) as well as for menopausal status and estrogen therapy (in women).29.0 ± 8.729.1 ± 8.427.7 ± 7.10.58 Waist circumference (cm)aP value after adjustment for the nonindependence among family members and age (except when the dependent variable is age) as well as for menopausal status and estrogen therapy (in women).86.6 ± 17.9cN = 151.86.9 ± 18.884.0 ± 17.30.63N = 119N = 73N = 18 Energy intake (MJ/d)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).8.32 ± 1.928.10 ± 2.048.02 ± 2.100.93 Total fat (% of energy)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).34.7 ± 5.933.5 ± 6.634.3 ± 5.40.84 SFA (% of energy)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).10.9 ± 3.210.2 ± 2.710.8 ± 2.90.62 MUFA (% of energy)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).11.6 ± 3.010.9 ± 2.611.0 ± 2.70.50 PUFA (% of energy)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).4.7 ± 1.74.4 ± 1.84.6 ± 1.90.67MenN = 77N = 117N = 22 Age (years)aP value after adjustment for the nonindependence among family members and age (except when the dependent variable is age) as well as for menopausal status and estrogen therapy (in women).40.5 ± 15.740.7 ± 15.443.6 ± 16.40.90 BMI (kg/m2)aP value after adjustment for the nonindependence among family members and age (except when the dependent variable is age) as well as for menopausal status and estrogen therapy (in women).29.7 ± 7.327.6 ± 6.628.6 ± 6.20.07 Waist circumference (cm)aP value after adjustment for the nonindependence among family members and age (except when the dependent variable is age) as well as for menopausal status and estrogen therapy (in women).99.8 ± 18.893.6 ± 16.998.5 ± 14.90.03dT111/I111 different from T111/T111.N = 67N = 95N = 14 Energy intake (MJ/d)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).11.46 ± 3.1011.22 ± 3.2011.88 ± 3.600.39 Total fat (% of energy)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).35.0 ± 5.534.7 ± 6.333.4 ± 5.80.75 SFA (% of energy)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).10.4 ± 3.210.4 ± 3.111.1 ± 3.90.68 MUFA (% of energy)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).11.6 ± 3.311.5 ± 3.411.7 ± 2.31.0 PUFA (% of energy)bMultivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).4.6 ± 1.84.3 ± 1.74.4 ± 1.90.75QFS, Québec Family Study; EL, endothelial lipase; SFA, saturated fatty acid; MUFA, monounsaturated fatty acid; BMI, body mass index. Data are presented as unadjusted values and are expressed as mean ± SD. BMI was log transformed prior to analysis in both men and women, as well as PUFA among men only. Energy intake is expressed in MJ/day. The conversion factor is 1 MJ = 239.23 kcal.a P value after adjustment for the nonindependence among family members and age (except when the dependent variable is age) as well as for menopausal status and estrogen therapy (in women).b Multivariate models further adjusted for BMI, carbohydrates (% of daily energy), and proteins (% of daily energy).c N = 151.d T111/I111 different from T111/T111. Open table in a new tab TABLE 2Mean plasma lipids among women from the QFS according to the T111I genotypeT111/T111 N = 153T111/I111 N = 100I111/I111 N = 28P aP value of the ANOVA comparing the three mutation groups after adjustment for the nonindependence among family members and age.P bMultivariate models were further adjusted for menopausal status, hormonal therapy, and BMI.mmol/lCholesterol Total4.91 ± 1.035.00 ± 1.194.66 ± 0.860.810.82 VLDL0.46 ± 0.260.50 ± 0.480.38 ± 0.200.280.36 LDL2.97 ± 0.893.06 ± 1.002.77 ± 0.770.970.94 HDL1.20 ± 0.291.19 ± 0.291.25 ± 0.250.090.14 HDL20.49 ± 0.220.48 ± 0.200.49 ± 0.160.510.66 HDL30.71 ± 0.150.71 ± 0.150.77 ± 0.170.030.03Triglycerides Total1.44 ± 0.621.44 ± 0.681.25 ± 0.600.410.49 VLDL0.86 ± 0.510.87 ± 0.580.74 ± 0.490.320.41 LDL0.30 ± 0.120.30 ± 0.110.25 ± 0.090.260.39 HDL0.28 ± 0.080.27 ± 0.080.28 ± 0.080.450.35mg/mlApoB Total0.96 ± 0.230.97 ± 0.230.90 ± 0.160.510.63 VLDLcN = 152 in the T111/T111 women.0.10 ± 0.050.10 ± 0.060.09 ± 0.040.430.50 LDL0.86 ± 0.210.87 ± 0.210.81 ± 0.160.660.74ApoA-I1.30 ± 0.181.29 ± 0.191.31 ± 0.200.290.21 LDL size (Å)dN = 151 in the T111/T111 women.264.8 ± 4.6264.0 ± 4.4264.1 ± 4.80.420.33 HDL size (Å)eN = 105, N = 70, and N = 17 among the T111/T111, T111/I111, and I111/I111 women, respectively.87.8 ± 3.687.3 ± 3.487.8 ± 3.10.820.70ApoB, apolipoprotein B; VLDL-C, VLDL cholesterol; TG, triglyceride. Data are presented as unadjusted values and are expressed as mean ± SD. Total cholesterol, VLDL-C, LDL-C, total triglycerides, VLDL-TG, LDL-TG, HDL-TG, and VLDL-apoB were log transformed prior to analyses.a P value of the ANOVA comparing the three mutation groups after adjustment for the nonindependence among family members and age.b Multivariate models were further adjusted for menopausal status, hormonal therapy, and BMI.c N = 152 in the T111/T111 women.d N = 151 in the T111/T111 women.e N = 105, N = 70, and N = 17 among the T111/T111, T111/I111, and I111/I111 women, respectively. Open table in a new tab TABLE 3Mean plasma lipids among men from the QFS according to the T111I genotypeT111/T111 N = 76T111/I111 N = 116I111/I111 N = 22P aP value of the ANOVA comparing the three mutation groups after adjustment for the nonindependence among family members and age.P bMultivariate models were further adjusted for BMI.mmol/lCholesterol TotalcN = 77 in the T111/T111 and N = 117 among the T111/I111 men.5.00 ± 0.954.96 ± 1.095.07 ± 1.040.790.82 VLDL0.62 ± 0.420.54 ± 0.300.56 ± 0.340.570.99 LDL3.40 ± 0.793.45 ± 0.953.42 ± 0.740.890.93 HDL0.97 ± 0.211.00 ± 0.221.04 ± 0.240.820.96 HDL20.32 ± 0.160.33 ± 0.150.34 ± 0.140.830.84 HDL30.65 ± 0.120.67 ± 0.140.70 ± 0.160.680.86Triglycerides TotalcN = 77 in the T111/T111 and N = 117 among the T111/I111 men.1.72 ± 0.951.53 ± 0.731.75 ± 0.890.190.31 VLDL1.19 ± 0.801.05 ± 0.661.11 ± 0.680.500.84 LDL0.29 ± 0.120.28 ± 0.090.29 ± 0.140.300.56 HDL0.23 ± 0.060.22 ± 0.050.22 ± 0.050.330.66mg/mlApoB Total1.03 ± 0.241.03 ± 0.251.07 ± 0.250.930.97 VLDLdN = 115 among the T111/I111 men.0.11 ± 0.060.11 ± 0.060.12 ± 0.060.630.87 LDL0.91 ± 0.220.92 ± 0.230.95 ± 0.220.990.99ApoA-I1.20 ± 0.151.21 ± 0.151.22 ± 0.250.640.66 LDL size (Å)eN = 72, N = 114, and N = 21 among the T111/T111, T111/I111, and I111/I111 men, respectively.262.4 ± 5.1262.0 ± 5.0261.6 ± 5.20.870.76 HDL size (Å)fN = 51, N = 87, and N = 15 among the T111/T111, T111/I111, and I111/I111 men, respectively.84.4 ± 3.084.8 ± 3.382.7 ± 2.760.070.17Data are presented as unadjusted values and are expressed as mean ± SD. VLDL-C, HDL2-C, total triglycerides, VLDL-TG, and LDL-TG were log transformed prior to analyses.a P value of the ANOVA comparing the three mutation groups after adjustment for the nonindependence among family members and age.b Multivariate models were further adjusted for BMI.c N = 77 in the T111/T111 and N = 117 among the T111/I111 men.d N = 115 among the T111/I111 men.e N = 72, N = 114, and N = 21 among the T111/T111, T111/I111, and I111/I111 men, respectively.f N = 51, N = 87, and N = 15 among the T111/T111, T111/I111, and I111/I111 men, respectively. Open table in a new tab QFS, Québec Family Study; EL, endothelial lipase; SFA, saturated fatty acid; MUFA, monounsaturated fatty acid; BMI, body mass index. Data are presented as unadjusted values and are expressed as mean ± SD. BMI was log transformed prior to analysis in both men and women, as well as PUFA among men only. Energy intake is expressed in MJ/day. The conversion factor is 1 MJ = 239.23 kcal. ApoB, apolipoprotein B; VLDL-C, VLDL cholesterol; TG, triglyceride. Data are presented as unadjusted values and are expressed as mean ± SD. Total cholesterol, VLDL-C, LDL-C, total triglycerides, VLDL-TG, LDL-TG, HDL-TG, and VLDL-apoB were log transformed prior to analyses. Data are presented as unadjusted values and are expressed as mean ± SD. VLDL-C, HDL2-C, total triglycerides, VLDL-TG, and LDL-TG were log transformed prior to analyses. Among men, dietary intakes of PUFA showed an inverse relationship with plasma levels of HDL-C and HDL2-C (r = −0.18, P = 0.01, and r = −0.17, P = 0.02, respectively). Dietary intakes of monounsaturated fatty acids (MUFAs) were also negatively correlated with plasma HDL-C (r = −0.23, P = 0.002), HDL2-C (r = −0.17, P = 0.02), HDL3-C (r = −0.15, P = 0.05), and HDL size (r = −0.21, P = 0.01). We did not observe any significant relationship among dietary intakes of saturated fatty acids (SFAs) and total fat and the HDL profile in men. Among women, dietary intakes of total, PUFA, MUFA, and SFA showed no association with any HDL-related phenotypes (data not shown). In order to examine whether the T111I mutation influenced the relationship between dietary fat intake and the HDL profile, Spearman correlations were computed between T111T homozygotes and women carrying the I111 allele. Total fat intake (% of daily energy) was negatively correlated with plasma HDL3-C levels among women homozygous for the wild-type allele (r = −0.21, P = 0.02), but positively correlated among I111 carriers (r = 0.27, P = 0.01). PUFA intake (% of daily energy) show