| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Clinical Biochemistry, Herlev University Hospital (H.H.W., R.V.A., B.G.N.), DK-2730 Herlev, Denmark; Department of Clinical Biochemistry, Rigshospitalet, Copenhagen University Hospital (A.T.-H.), DK-2100 Copenhagen, Denmark; and Copenhagen City Heart Study, Bispebjerg University Hospital (A.T.-H., G.B.J., B.G.N.), DK-2400 Copenhagen, Denmark
Address all correspondence and requests for reprints to: Dr. Børge G. Nordestgaard, Department of Clinical Biochemistry, Herlev University Hospital, Herlev Ringvej 75, DK-2730 Herlev, Denmark. E-mail: brno{at}herlevhosp.kbhamt.dk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: The objective of this study was to investigate the influence of T(–93)G, G(–53)C, Asp9Asn, Gly188Glu, Asn291Ser, and Ser447Ter lipoprotein lipase genotypes on triglycerides, HDL, and IHD.
Design: The cross-sectional study involved 9004 adults. The prospective study consisted of 8817 adults developing 1001 IHD events over 23 yr. The case-control study involved 7818 non-IHD individuals vs. cohorts of 915 and 1062 IHD patients, respectively.
Setting: The study was performed in the Danish general population (the Copenhagen City Heart Study).
Participants: IHD was angina pectoris or myocardial infarction.
Main Outcome Measures: Triglycerides, HDL, and IHD were the main outcome measures.
Results: Cross-sectionally, triglycerides varied by genotype with 1.27 mmol/liter in women and 1.22 mmol/liter in men. HDL cholesterol varied by genotype with 0.49 mmol/liter in women and 0.60 mmol/liter in men. Prospectively, 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers had a hazard ratio for IHD of 1.6 [95% confidence interval (CI), 1.2–2.3]; 291Ser and 447Ter did not change IHD risk. In the case-control study, combining the cohorts of IHD patients, 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers had an odds ratio for IHD of 1.5 (CI, 1.2–2.1). 291Ser and 447Ter did not change IHD risk. Stratified for apolipoprotein E genotype, the odds ratios for IHD in 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers were 2.6 (CI, 1.2–5.5) among 
32 individuals and 2.4 (CI, 1.4–4.1) among 43 individuals.
Conclusions: Genetic variation in lipoprotein lipase is associated with differences in plasma triglycerides greater than 1 mmol/liter and differences in HDL cholesterol greater than 0.5 mmol/liter. A 1.6-fold risk of IHD in 9Asn (with –93G) heterozygotes and homozygotes combined is influenced by apolipoprotein E genotype.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
These six common genetic variants in LPL, therefore, may all affect levels of triglycerides and HDL (1, 2, 3, 4, 5, 6), and consequently, together could represent one of the most important genetic influences on the risk of ischemic heart disease (IHD). Because these variants have most often been studied one at a time, an understanding of the combined influence of LPL variation on triglyceride and HDL levels is sparse. Furthermore, because the risk of IHD mainly has been examined using family, cross-sectional, case-control, or case-referent designs (4, 5), the influence of LPL genetic variation on the risk of IHD prospectively is limited.
We first investigated cross-sectionally the combined influence of T(–93)G, G(–53)C, Asp9Asn, Gly188Glu, Asn291Ser, and Ser447Ter LPL variants on plasma lipoprotein levels in 9004 individuals from the adult Danish general population (12). Second, we studied the risk of IHD prospectively in 8817 general population participants with 1001 IHD events during 23 yr of follow-up. Previously, we examined these variants [except G(–53)C] with respect to the odds ratio for IHD in case-control studies one at a time (13, 14, 15, 16, 17); however, we never before used a prospective design on the general population with Kaplan-Meier curves and hazard ratios. Third, we performed two case-control studies in which 7818 non-IHD individuals were compared with IHD patient cohorts of 915 and 1062 patients, respectively. In prospective and case-control studies of the risk of IHD, we only examined the most common genotypes.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We used three different designs: 1) cross-sectional: the association between LPL genetic variation and lipid and lipoprotein levels was examined cross-sectionally in the entire general population cohort (n = 9004; Table 1
); 2) prospective: the risk of IHD was examined prospectively within this cohort (n = 8817), excluding those with IHD before study entry and those with rare genotypes (n = 187); and 3) case-control: the risk of IHD was also examined in two case-control studies separately (n = 915 and 1062 cases vs. 7818 controls) as well as combined (n = 1977 cases vs. 7818 controls), excluding participants with rare genotypes.
|
For cross-sectional and prospective studies, we studied, respectively, 9,004 and 8,817 individuals (55% women) participating in the Copenhagen City Heart Study (12, 13, 14, 15, 16, 17) (Table 1
). Participants gave written informed consent. This cohort was drawn randomly from the Copenhagen Central Population Register to obtain an age-stratified representative sample of the adult Danish general population. Of the 17,180 individuals invited, 58% participated, 54% gave blood, and 52% were genotyped for all six genetic variants for the present study.
Among the participants, 1083 have been recorded with IHD; of those, 82 were diagnosed before study entry and were thus excluded from the prospective study. Information on diagnoses of IHD (World Health Organization International Classification of Diseases, 8th edition, codes 410–414; 10th ed, codes I20–25) were gathered from 1976 through 1999 from the Danish National Hospital Discharge Register, the Danish National Register of Causes of Death, and medical records from general practitioners and hospitals.
For case-control studies, we included an additional 915 patients with IHD (26% women) ascertained through referral for coronary angiography (13, 14, 15, 17) (Table 1). Of the 992 patients referred, 96% had IHD, 94% were genotyped for all six genetic variants, and 92% with common genotypes were included in statistical analyses. IHD was diagnosed by experienced cardiologists based on characteristic symptoms of stable angina pectoris according to the guidelines of the European Society of Cardiology (18) on the basis of location, character, and duration of pain and the relation of pain to exercise plus at least one of the following: severe stenosis on coronary angiography (>70% stenosis of at least one coronary vessel or >50% stenosis of the left main coronary artery), previous myocardial infarction, or positive exercise electrocardiography test. Case-control studies were performed separately for this group of IHD patients (n = 915) and the IHD patients described above (ascertained from the general population; n = 1062) as well as for the two groups combined (all cases combined equal 1977); 7818 participants without IHD from the Copenhagen City Heart Study were used as controls (Table 1).
In all groups of participants, more than 99% were white, and roughly 99% were of Danish descent. The studies were approved by Danish ethical committees and were conducted in accordance with the Declaration of Helsinki. The number of individuals included in the various analyses presented in this paper differs slightly depending on the availability of data.
DNA analyses
Genotyping was described previously for five of the genetic variants: T(–93)G (15), Asp9Asn (15), Gly188Glu (14), Asn291Ser (13), and Ser447Ter (17). Using the principle of a previously designed method for analyzing DNA pooled from 10–20 individuals (14, 19), the promoter variant G(–53)C was detected in such pools by allele-specific PCR on genomic DNA using the sense primer 5'-GTTGGCAGGGTTGATCCTCAT-3', the antisense primer 5'-TTCCCTTGAGGAGGAGGAAGA-3', and the mutation-specific antisense primer 5'-GACTGGAAATATGCAAATAAAACTG-3' (mutation site in underlined italics and additional mismatch underlined only). DNA from individuals from pools testing positively for the G(–53)C variant were reanalyzed separately using identical end primers, digesting the PCR product with BclI, and separating on a 3% agarose gel. Apolipoprotein E genotyping was described previously (20).
Other measurements
Blood samples from participants of the Copenhagen City Heart Study were drawn in the nonfasting state, whereas blood samples from patients with IHD were drawn in the fasting state. Colorimetric and turbidimetric assays were used to measure levels of total cholesterol, HDL cholesterol, triglycerides, apolipoprotein AI, and apolipoprotein B (Roche, Mannheim, Germany). Hypertension was defined as the use of antihypertensive medication or blood pressure of 140/90 mm Hg or above. Diabetes mellitus was defined as self-reported disease treated with insulin, oral hypoglycemic agents, or diet. The body mass index was weight divided by height squared.
Statistical analyses
Statistical analyses were performed using SPSS and Stata software (SPSS, Inc., Chicago, IL). Two-sided P < 0.05 was considered significant. Correction for multiple comparisons was made by the Bonferroni method. A priori, we stratified by gender; however, to increase the statistical power when analyzing the risk of IHD, we also examined the two genders combined. Variation in levels of lipids and lipoproteins across LPL genotypes was evaluated using nonparametric tests, because neither triglyceride levels, HDL cholesterol levels, nor apolipoprotein AI levels completely follow a normal distribution. We tested for overall difference between genotypes using Kruskal-Wallis ANOVA, which showed that levels of triglycerides, HDL cholesterol, and apolipoprotein AI differed significantly between genotype groups. We therefore performed post hoc two-genotype comparisons using the Mann-Whitney U test. We tested for interaction between LPL genotype and a range of cardiovascular risk factors one at a time [age, total cholesterol, apolipoprotein B, lipoprotein(a), triglycerides, HDL cholesterol, apolipoprotein AI, fibrinogen, body mass index, glucose, alcohol consumption, smoking, physical activity at work, physical activity at leisure, hypertension, diabetes mellitus, menopause (women), use of diuretics, use of heart medication, use of antihypertensive medication, cholesterol-lowering therapy, and hormonal replacement therapy (women)], in the prediction of triglycerides, HDL cholesterol, and apolipoprotein AI levels, using two-factor (genotype by risk factor) interaction terms in an analysis of covariance. To obtain an approximately normal distribution, triglycerides and lipoprotein(a) were transformed logarithmically before these analyses.
Cumulative incidence of IHD as a function of age was plotted using Kaplan-Meier curves, with the log-rank test as a measure of statistical significance between genotypes. Cox regression analyses examined the role of genotype in time to first IHD event using the hazard ratios in the prospective study. Interaction of LPL genotype with gender and the cardiovascular risk factors mentioned above on IHD risk was also examined using two-factor interaction terms.
Unconditional logistic regression analysis examined the association between LPL genotype and risk of IHD in case-control studies. Analyses were made adjusting for 1) age only, 2) age and triglyceride levels, and 3) age and HDL cholesterol levels. Analyses made after stratification for apolipoprotein E genotype were adjusted for age. Interactions of LPL genotype with gender, apolipoprotein E genotype, and the cardiovascular risk factors mentioned above on IHD risk were also examined using two-factor interaction terms.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
). 9Asn was always found in combination with –93G [this combination is subsequently called 9Asn (with –93G) or *9]; 14 individuals carried –93G without 9Asn. Counting 9Asn (with –93G) as one variant, no individual carried more than two variants, compatible with a model in which each genetic variant is entirely separated from the others, forming an individual haplotype. The calculated frequencies of these haplotypes were 85.9% for noncarriers, 9.9% for 447Ter, 2.5% for 291Ser, 1.5% for 9Asn (with –93G), 0.12% for –53C, 0.08% for –93G, and 0.03% for 188Glu.
|
Plasma triglycerides varied as a function of genotype in both women and men, with the highest plasma levels in 188Glu heterozygotes and the lowest levels in 447Ter homozygotes (by ANOVA, P < 0.001 for both genders; Fig. 1); the absolute difference between these levels was 1.27 mmol/liter in women and 1.22 mmol/liter in men. Plasma HDL cholesterol and apolipoprotein AI levels also varied by genotype, but with the lowest levels in 188Glu heterozygotes and the highest in 447Ter heterozygotes and homozygotes (by ANOVA, P < 0.001 and P = 0.003; Fig. 1); the HDL cholesterol variation from the highest to the lowest level was 0.49 mmol/liter in women and 0.60 mmol/liter in men, whereas for apolipoprotein AI, the maximal variation was 16 mg/dl in women and 24 mg/dl in men. When we excluded individuals with diabetes mellitus, the results were similar (data not shown).
|
). The same applied to 291/N individuals compared with N/N individuals (except for triglycerides in men) and 447/447 individuals (except for apolipoprotein AI in women). Additionally, levels of triglycerides and HDL cholesterol were different in 447/447 men from *9/N and N/N men. After Bonferroni correction for multiple comparisons, only two-genotype comparisons with P < 0.001 remained significant. Thus, the borderline significance found for comparisons of the small genotype groups with other genotype groups disappeared after this correction. We observed no plausible multiplicative bivariate interaction between LPL genotypes and lipid and nonlipid cardiovascular risk factors in the prediction of plasma triglycerides, HDL cholesterol, and apolipoprotein AI levels.
|
During 23 yr of follow-up, there were 1001 IHD events among 8817 participants with the most common genotypes, who were free of IHD at study entry. Combining heterozygotes and homozygotes, Kaplan-Meier survival curves showed a cumulative incidence of IHD at age 75 yr of 32% for 9Asn (with –93G) heterozygotes and homozygotes combined compared with 18% for noncarriers (log-rank, P = 0.005; Fig. 2); equivalent cumulative incidences for 291Ser and 447Ter heterozygotes and homozygotes combined did not differ from that of noncarriers. Cox regression analyses adjusted for age showed a hazard ratio for IHD in 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers of 1.6 [95% confidence interval (CI), 1.2–2.3], whereas 291Ser and 447Ter heterozygotes and homozygotes combined vs. noncarriers did not confer altered IHD risk (Table 4). After stratification by gender, women and men with 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers had hazard ratios of 1.6 (CI, 1.0–2.8) and 1.7 (CI, 1.1–2.7), respectively. Analyzing only participants without diabetes mellitus gave similar results (Table 4). There was no statistical evidence for interaction between LPL genotype and gender (or other cardiovascular risk factors) in the prediction of IHD risk.
|
|
Age-adjusted logistic regression analyses using a cohort of 934 IHD patients as the case group, genders combined, showed an odds ratio of 1.6 (CI, 1.1–2.3) for 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers (Table 5); 291Ser and 447Ter did not confer altered IHD risk. After stratification by gender, 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers had odds ratios for IHD of 1.2 (CI, 0.6–2.4) in women and 2.3 (CI, 1.4–3.6) in men. Equivalent analyses using the 1083 participants with IHD from the Copenhagen City Heart Study as the case group showed odds ratios of 1.5 (CI, 1.0–2.2) for genders combined, 1.4 (CI, 0.8–2.5) for women only, and 1.7 (CI, 1.0–3.0) for men only. Because both case-control studies gave similar results, we combined the case groups in our other analyses. Analyses using the combined groups of 1977 IHD patients showed odds ratios for 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers of 1.6 (CI, 1.2–2.1) for genders combined, 1.3 (CI, 0.8–2.1) for women alone, and 2.1 (CI, 1.4–3.1) for men alone. Analyzing only participants without diabetes mellitus gave similar results (Table 5). Additional adjustment for levels of plasma triglycerides and HDL cholesterol also gave similar results (data not shown). There was no statistical evidence for interaction between LPL genotype and gender (or other cardiovascular risk factors) in the prediction of IHD risk.
|
32 individuals, 1.3 (CI, 0.9–2.0) for 33 individuals, and 2.4 (CI, 1.4–4.1) for 43 individuals (Table 6). There was no statistically significant interaction between LPL genotype and apolipoprotein E genotype in the prediction of IHD risk.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
32 or 43 apolipoprotein E genotypes conferred 2.5-fold IHD risk. Triglycerides and HDL
Numerous previous studies have documented that 188Glu, 9Asn (with –93G), and 291Ser alleles increase plasma triglycerides and reduce HDL cholesterol levels compared with the noncarrier allele (1, 2, 3, 4, 5, 6), in accordance with decreased plasma LPL activity in carriers of such alleles (1, 2, 3, 5). Likewise, it is well documented that the 447Ter allele vs. the noncarrier allele causes a reduction in plasma triglycerides and an increase in HDL cholesterol (2, 3, 4, 5), in accordance with increased LPL expression without changing specific activity in carriers of this allele (2, 9). A novel aspect of the current study is that we evaluated all these alleles together with the –93G and –53C promoter variants on triglycerides and HDL cholesterol levels in a large sample from the general population. This allowed us to demonstrate clinically important large differences in plasma triglycerides and HDL cholesterol levels caused by these LPL genetic variants. Because of this strong influence on levels of two important cardiovascular risk factors, genetic variation in LPL probably represents an important common genetic influence on risk of IHD.
IHD
The observation that 9Asn (with –93G) heterozygotes and homozygotes combined vs. noncarriers had increased risk of IHD is very plausible indeed, because 1) it is compatible with previous results obtained mainly in case-control studies (3, 4, 5, 6); 2) we could document this finding in both women and men in the prospective study; 3) because 9Asn leads to increased plasma triglycerides and reduced HDL cholesterol levels (1, 2, 3, 4, 5, 6), this variant mechanistically may lead to more severe atherosclerosis and consequently IHD; 4) 9Asn may also promote atherosclerosis via enhanced bridging function and augmented monocyte adhesion (11); 5) our case-control studies also documented increased risk of IHD associated with the 9Asn variant; and 6) even after correction for multiple comparisons using the Bonferroni method, 9Asn (with –93G) heterozygotes and homozygotes combined exhibited significantly increased risk of IHD in our prospective as well as case-control studies.
The apparent augmentation of the effect of 9Asn (with –93G) on IHD risk by apolipoprotein E 32 and 43 seems biologically plausible, because, like 9Asn (with –93G), 32 and 43 are associated with elevations in plasma triglycerides in our sample (20, 21). The combination of 9Asn (with –93G) heterozygosity and homozygosity with 32 or 43 apolipoprotein E genotypes could not be tested with sufficient statistical power in the prospective study because of the low number of individuals with these relatively rare genotypes.
Limitations
Despite our large prospective and case-control studies, due to limited statistical power we still may have been unable to document minor influences on the risk of IHD. For example, because both 291Ser and 447Ter heterozygosity and homozygosity influence plasma levels of triglycerides and HDL cholesterol, even larger studies than ours may be able to document small influences in IHD risk by these genetic variants. Likewise, due to small numbers, we excluded rare genotypes in the present analyses of IHD risk and, therefore, cannot exclude the possibility that 188Glu or the –93G (without 9Asn) and –53C functionally important promoter variants could influence the risk of IHD. A previous meta-analysis showed that 188Glu heterozygosity was associated with a 5-fold risk of IHD (4).
Conclusions
This study documents that LPL genetic variation is an important modulator of plasma triglycerides, plasma HDL cholesterol, and risk of IHD at the same scale or larger than the influence of the apolipoprotein E polymorphism (20, 22). 9Asn (with –93G) heterozygotes and homozygotes combined showed a 1.6-fold risk of IHD in prospective as well as case-control studies. Also important, we document that LPL 9Asn (with –93G) heterozygosity and homozygosity combined with 32 or 43 apolipoprotein E genotypes, found in one of 100 individuals, confers 2.5-fold IHD risk.
There are no potential conflict of interests for any of the authors.
|
Footnotes |
|---|
There are no potential conflicts of interest for any of the authors.
First Published Online January 17, 2006
1 H.H.W. and R.V.A. contributed equally to this study. 
Abbreviations: CI, 95% Confidence interval; HDL, high-density lipoprotein; IHD, ischemic heart disease; LPL, lipoprotein lipase.
Received August 1, 2005.
Accepted January 10, 2006.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
32 women are protected while 43 and 44 men are susceptible to ischemic heart disease. The Copenhagen City Heart Study. J Am Coll Cardiol 35:1192–1199This article has been cited by other articles:
 |
|
 |
|
  G. S. Sagoo, I. Tatt, G. Salanti, A. S. Butterworth, N. Sarwar, M. van Maarle, J. W. Jukema, B. Wiman, J. J. P. Kastelein, A. M. Bennet, et al. Seven Lipoprotein Lipase Gene Polymorphisms, Lipid Fractions, and Coronary Disease: A HuGE Association Review and Meta-Analysis Am. J. Epidemiol., December 1, 2008; 168(11): 1233 - 1246. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Frikke-Schmidt, B. G. Nordestgaard, M. C. A. Stene, A. A. Sethi, A. T. Remaley, P. Schnohr, P. Grande, and A. Tybjaerg-Hansen Association of Loss-of-Function Mutations in the ABCA1 Gene With High-Density Lipoprotein Cholesterol Levels and Risk of Ischemic Heart Disease JAMA, June 4, 2008; 299(21): 2524 - 2532. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Kathiresan, O. Melander, D. Anevski, C. Guiducci, N. P. Burtt, C. Roos, J. N. Hirschhorn, G. Berglund, B. Hedblad, L. Groop, et al. Polymorphisms Associated with Cholesterol and Risk of Cardiovascular Events N. Engl. J. Med., March 20, 2008; 358(12): 1240 - 1249. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) in levels of triglycerides, HDL cholesterol, and apolipoprotein AI in individuals with different LPL genotypes vs. noncarriers (N/N) in a cross-sectional study. Absolute values of differences from N/N levels are given below each panel. Mean levels for N/N individuals are shown at the baseline to the right. 447, 447Ter; 291, 291Ser; *9, 9Asn (with –93G); –53, –53C; –93, –93G; 188, 188Glu. The ANOVA used was the Kruskal-Wallis test.