Apolipoprotein E genotypes predict attendance rates at lipid clinic

Atherosclerosis 153 (2000) 461 – 468 www.elsevier.com/locate/atherosclerosis

Apolipoprotein E genotypes predict attendance rates at lipid clinic Ruth Frikke-Schmidt a,1, Hans H. Wittrup a,1, Anne Tybjærg-Hansen a,b, Hans Meinertz c, Peter Schnohr b, Børge G. Nordestgaard b,d,* a Department of Clinical Biochemistry, Herle6 Uni6ersity Hospital, DK-2730 Herle6, Denmark The Copenhagen City Heart Study, Bispebjerg Uni6ersity Hospital, DK-2400, Copenhagen NV, Denmark c Department of Medicine B, Di6ision of Cardiology, Copenhagen Uni6ersity Hospital, DK-2100 Copenhagen, Denmark d Department of Clinical Biochemistry, Glostrup Uni6ersity Hospital, DK-2600 Glostrup, Denmark b

Received 16 August 1999; received in revised form 24 December 1999; accepted 31 January 2000

Abstract Except for the rare o22 genotype it remains largely unsettled whether apolipoprotein E genotype influences an individual’s referral to lipid clinics. To test this hypothesis, we compared genotype distributions among 156 hypercholesterolemic and 83 hypertriglyceridemic patients attending a lipid clinic with that among 9241 individuals sampled from the Danish general population. The relative genotype frequencies of o22, o32, o42, o33, o43, and o44 were 0.005, 0.126, 0.026, 0.564, 0.251, and 0.027 in the general population, which differed from genotype frequencies in both hypercholesterolemic (x2: P= 0.01) and hypertriglyceridemic patients (x2: PB0.001). By comparison with o33, o44 predicted a 2-fold increase whereas o32 predicted a 2-fold decrease in the attendance rate at the lipid clinic for hypercholesterolemic patients (95% confidence intervals: 1.1 – 4.3 and 0.2 – 0.9). Among hypertriglyceridemic patients, o22, o42, o43, and o44 versus o33 predicted 13-, 3-, 112-, and 3-fold attendance rates at the lipid clinic, respectively (95% confidence intervals: 4.5–39.9, 1.2–8.4, 1.0 – 2.8, and 1.1 – 7.6). These findings are in accordance with the fact that o44 raises cholesterol levels, o32 reduces cholesterol levels, and o22, o42, o43, and o44 raise triglyceride levels in comparison with o33. These data suggest that hypercholesterolemic individuals carrying o44 and hypertriglyceridemic individuals carrying o22, o42, o43, or o44 are relatively more often referred to lipid clinics than carriers of o33. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Atherosclerosis; Apolipoprotein E; Genotype; Dyslipidemia; Genes

1. Introduction The apolipoprotein E (apoE) polymorphism consists of three alleles (o2, o3, and o4) on the long arm of chromosome 19 [1,2], coding for three protein isoforms, apoE-2, -3, and -4. The apoE-2 isoform differs from apoE-3 by a cysteine for arginine substitution at amino acid residue 158, while apoE-4 differs from apoE-3 by an arginine for cysteine substitution at residue 112 [3]. ApoE mediates interaction of mainly triglyceride-rich lipoproteins with lipoprotein receptors, the affinity of which depends on the apoE isoform [3].

* Corresponding author. Tel.: +45-4488-3297; fax: +45-44883311. E-mail address: [email protected] (B.G. Nordestgaard). 1 These authors contributed equally to this article.

Aside from gender the apoE polymorphism is the most important known genetic modulator of plasma lipids and lipoproteins in the population at large [3– 14]. It remains unsettled, however, whether apoE genotype influences referral of patients to lipid clinics. Patients are most often referred to lipid clinics as a result of clinically defined syndromes of familial hypercholesterolemia [15,16], combined hyperlipidemia [15], chylomicronemia syndrome [17], low high density lipoprotein (HDL) [18], or type III hyperlipoproteinemia [3]. The latter is precipitated by the rare apoE o22 genotype [3], while other lipid disorders are thought to be unaffected by apoE genotype. Nevertheless, it is still possible that the additional lipid raising effects of certain apoE genotypes cause an otherwise unnoticed lipid disturbance to become clinically recognizable, and thus causes the physician to refer the patient to a lipid clinic.

0021-9150/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 2 1 - 9 1 5 0 ( 0 0 ) 0 0 4 2 9 - 9

462

R. Frikke-Schmidt et al. / Atherosclerosis 153 (2000) 461–468

We tested the hypothesis that attendance rates at lipid clinics differ as a function of the six apoE genotypes. For this purpose 241 female and male patients attending a lipid clinic and 9241 individuals sampled from the general population, The Copenhagen City Heart Study, were genotyped.

2. Methods

2.1. Subjects We collected 246 consecutive patients (45% women) attending the Lipid Clinic at Rigshospitalet (Copenhagen University Hospital) in Copenhagen; 241 of these were genotyped. The patients were referred from the Greater Copenhagen area, either from other hospitals, primary care physicians, or the Department of Cardiology at Rigshospitalet. Based on lipoprotein disturbance and family history, and prior to genotyping, patients were classified by the consultant in the lipid clinic (HM) as hypercholesterolemic [n =156; 97 as familial hypercholesterolemia (FH), and 59 as other low density lipoprotein (LDL) elevation], hypertriglyceridemic [n= 83; 51 as familial combined hyperlipidemia (FCHL), three as type III hyperlipidemia, five as chylomicronemia syndrome, and 24 as other types of hypertriglyceridemia], or as isolated low HDL cholesterol (n = 2). Most hypertriglyceridemic patients also had raised cholesterol levels, and thus had combined hyperlipidemia. This classification was not based on a single lipid measurement with specific cut-off points, but by evaluation of all diagnostic data available to the attending consultant (HM), who has run this lipid clinic for more than 25 years and followed most patients for several years. All patients classified as hypercholesterolemic had plasma LDL cholesterol \ 4 mmol/l, and all patients classified as hypertriglyceridemic had plasma triglycerides \ 2 mmol/l. Approximately 99% of the patients were Caucasians of Danish descent. The Copenhagen City Heart Study (third examination, 1991–1994) includes an almost equal number of women (55%) and men stratified into 10 year age groups from 20 to 80 years and above, drawn randomly from the Copenhagen Central Population Register with the aim to obtain a representative sample of the Danish general population [19 – 24]. A total of 9241 subjects were genotyped for the present study. Approximately 99% of participants were Caucasians of Danish descent. A total of 181 individuals from this sample either took lipid lowering drugs, or lacked information regarding this question. The study was approved by Danish ethical committees: no. 100.2039/91 Copenhagen and Frederiksberg committee, and no. KA 93273, Copenhagen County committee. All subjects gave informed consent.

2.2. DNA analyses DNA was amplified by PCR followed by restriction enzyme digestion with HhaI of the 244 basepair PCR product [25]. Due to the large sample size, we used a 5% agarose gel instead of a polyacrylamide gel to identify the apoE genotypes. On the agarose gel, however, we could not always detect the 48 basepair band which distinguished between the o22 and the o32 genotypes; we therefore retyped all o22 and o32 genotypes, using a second PCR (sense primer: 5%ACATGGAGGACGTGTGCGG%3; antisense primer: 5%ACGCGGCCCTGTTCCACCA%3) followed by digestion of the 250 basepair PCR product with HaeII (o22 homozygotes: 2× 187 bp; o32 heterozygotes: 1×187, 1× 152, 1×35 bp, and common bands of 32, 18, and 13 bp). The 187 and 152 bp band (o32) were clearly distinguishable on an agarose gel.

2.3. Other analyses Colorimetric assays were used to measure plasma levels of total cholesterol, HDL cholesterol, and triglycerides (all Boehringer Mannheim, Mannheim, Germany). LDL cholesterol was calculated from the Friedewald equation as: total cholesterol-HDL cholesterol-(triglycerides/2.2), all in mmol/l [26]. LDL cholesterol was only calculated in individuals with triglycerides B 4.5 mmol/l. Reported values for lipids and lipoproteins were all off lipid lowering treatment.

2.4. Statistical analyses A P-value B 0.05 on two-sided tests was considered significant. Data were analyzed using the SPSS software [27]. Results were similar when patients from the lipid clinic were compared with all individuals sampled from the general population or only with those not taking lipid lowering drugs (excluding 181 of 9241 subjects). We chose to present data from the former design. Chi-square tests and Student’s t-tests compared risk factor distribution and frequencies of apoE genotypes between patients and the general population. Lipids and lipoproteins were adjusted for age in 10 year groups in an analysis of covariance (ANCOVA). Kruskal–Wallis analysis of variance (ANOVA) examined differences in levels of adjusted lipids and lipoproteins as a function of apoE genotype; post-hoc two-genotype comparison was performed by the Mann–Whitney U-test. Logistic regression [28] allowing for 10 year age-groups explored the impact of apoE genotype on attendance rates at the lipid clinic. Interaction between apoE genotype and gender or age was explored in logistic regression models including apoE genotype, age, and in addition the age by apoE genotype interaction term, or apoE genotype, gender, age

R. Frikke-Schmidt et al. / Atherosclerosis 153 (2000) 461–468

and the gender by apoE genotype interaction term. The likelihood ratio test determined statistical significance.

3. Results Patients from the lipid clinic had higher levels of cholesterol, LDL cholesterol, and triglycerides, lower levels of HDL cholesterol, and a higher frequency of diabetes mellitus compared with individuals sampled from the general population (Table 1). The difference in triglycerides is probably underestimated because patients were measured in the fasting state while individuals sampled from the general population were measured in the nonfasting state.

3.1. All patients The distribution of apoE genotypes differed between

463

patients from the lipid clinic and individuals from the general population, in the two genders combined (x2: PB 0.001) as well as for each gender separately (women: x2: P= 0.009; men: x2: P= 0.001) (Table 2). Genotype frequencies in the general population were in accordance with those predicted by the Hardy–Weinberg equilibrium (x2: 0.20 BPB 0.30), whereas genotype frequencies in the patient sample deviated significantly from those predicted (x2: PB 0.02). On logistic regression analysis adjusting for age, carriers of the o22, o43, and o44 genotypes had 5-, 112-, and 2-fold attendance rates at the lipid clinic compared with the common o33 genotype [odds ratio (OR): 4.9, 95% confidence interval (CI) 1.9–12.7; OR: 1.4 (1.1–1.9); OR: 2.4 (1.3–4.2), respectively] (Fig. 1, all). These results were similar for each gender separately, although the effects of o43 and o44 in women and of o22 in men were not statistically significant (Fig. 1, women, men).

Table 1 Characteristics of patients attending a lipid clinic and of individuals sampled from the general populationa Women

Number of individuals Age (years) Cholesterol (mmol/l) HDL cholesterol (mmol/l) LDL cholesterol (mmol/l)b Triglycerides (mmol/l)c Body mass index (kg/m2) Diabetes mellitus (%)

Men

Patients

General population

Patients

General population

110 5891.5 10.590.5*** 1.490.1*** 7.09 0.3*** 3.29 0.5** 25.09 0.4 10***

5112 58 9 0.2 6.390.02 1.7 90.01 3.8 90.02 1.7 90.01 25.2 90.06 2

131 53 9 1.0** 8.7 90.2*** 1.1 90.05*** 6.0 9 0.2*** 4.0 90.5*** 26.8 9 0.3* 15***

4129 57 90.2 6.0 9 0.02 1.4 9 0.01 3.7 9 0.02 2.1 9 0.03 26.1 9 0.06 5

a Values are mean 9S.E.M. or percent. Lipids and lipoproteins are pretreatment values measured in the fasting state in patients, and in the nonfasting state in individuals sampled from the general population. x2-tests and t-tests compared female and male patients from the lipid clinic with women and men from the general population: *PB0.05, **PB0.01, ***PB0.001. b Excludes individuals with triglycerides \4.5 mmol/l. c To approach a normal distribution, plasma triglycerides were logarithmically transformed prior to t-tests; however, untransformed values are shown.

Table 2 ApoE genotype frequencies in patients attending a lipid clinic and in individuals sampled from the general populationa All

Women

Patients

Total o22 o32 o42 o33 o43 o44 x2b a b

General population

Patients

N

F

N

0.005 0.126 0.026 0.564 0.251 0.027

110 3 7 4 60 30 6

N

F

241 5 18 10 118 76 14

9241 0.021 48 0.075 1168 0.039 244 0.490 5211 0.315 2321 0.058 249 PB0.001

Men

F

0.027 0.064 0.036 0.545 0.273 0.055 P =0.009

General population

Patients

N

F

N

5112 28 655 127 2855 1298 149

0.005 0.128 0.025 0.558 0.254 0.029

131 2 11 6 58 46 8

N, number of individuals; and F, frequency. Comparison of frequencies in patient groups with those in the general population.

General population F

0.015 0.084 0.046 0.443 0.351 0.061 P = 0.001

N

F

4129 20 513 117 2356 1023 100

0.005 0.124 0.028 0.571 0.248 0.024

R. Frikke-Schmidt et al. / Atherosclerosis 153 (2000) 461–468

464

Fig. 2. Odds ratios and 95% confidence intervals (CI) for attendance of hypercholesterolemic patients at a lipid clinic as a function of apoE genotype relative to the common o33 genotype. The logistic regression model allowed for 10 year age-groups; indicates common genotypes in the population at large (o33, o43, o32); indicates rare genotypes in the population at large (o22, o42, o44). The number of subjects in the analyses is the same as in Table 3.

Fig. 1. Odds ratios and 95% confidence intervals (CI) for attendance at a lipid clinic as a function of apoE genotype relative to the common o33 genotype. The logistic regression model allowed for 10 year age-groups; indicates common genotypes in the population at large (o33, o43, o32); indicates rare genotypes in the population at large (o22, o42, o44). The number of subjects in the analyses is the same as in Table 2.

and other LDL elevations (Table 3 and Fig. 2, familial hypercholesterolemia and other LDL elevation). These effects are in accordance with the effect of apoE genotype on levels of cholesterol in the general population: o32 carriers had lower cholesterol levels while o44 carriers had higher cholesterol levels compared with o33 carriers (Fig. 4, cholesterol).

3.2. Hypercholesterolemic patients The distribution of apoE genotypes differed between hypercholesterolemic patients and individuals sampled from the general population (Table 3; x2: P = 0.014). On logistic regression analysis adjusting for age, the o32 genotype decreased the attendance rate by 50% [OR: 0.5 (0.2–0.9)], whereas o44 carriers had a 2-fold attendance rate at the lipid clinic [OR: 2.1 (1.1 – 4.3)], when compared with the common o33 genotype (Fig. 2, all hypercholesterolemic patients). Similar trends were observed in both subgroups, familial hypercholesterolemia

3.3. Hypertriglyceridemic patients The distribution of apoE genotypes differed between hypertriglyceridemic patients and individuals sampled from the general population (Table 4, all patients; x2: PB 0.001). On logistic regression analysis adjusting for

Table 3 ApoE genotype frequencies in hypercholesterolemic patients attending a lipid clinic and in individuals sampled from the general populationa All patients

Total o22 o32 o42 o33 o43 o44 x2b a b

Familial hypercholesterolemia

Other LDL elevation

General population

N

F

N

F

N

F

N

F

156 1 9 4 83 50 9 P= 0.014

0.006 0.058 0.026 0.532 0.321 0.058

97 1 8 2 50 31 5 P= 0.270

0.010 0.082 0.021 0.515 0.320 0.052

59 0 1 2 33 19 4 P= 0.056

0.000 0.017 0.034 0.559 0.322 0.068

9241 48 1168 244 5211 2321 249

0.005 0.126 0.026 0.564 0.251 0.027

N, number of individuals; and F, frequency. Comparison of frequencies in patient groups with those in the general population.

R. Frikke-Schmidt et al. / Atherosclerosis 153 (2000) 461–468

465

Table 4 ApoE genotype frequencies in hypertriglyceridemic patients attending a lipid clinic and in individuals sampled from the general populationa

Total o22 o32 o42 o33 o43 o44 x2b a b

All patients

Familial combined hyperlipidemia

Type III hyperlipoproteinemia

Chylomicronemia

Other hypertriglyceridemia

General population

N

F

N

F

N

F

N

F

N

F

N

F

83 4 9 5 34 26 5 PB0.001

0.048 0.108 0.060 0.410 0.313 0.060

51 1 5 4 23 16 2 P=0.085

0.020 0.098 0.078 0.451 0.314 0.039

3 2 1 0 0 0 0 PB0.001

0.667 0.333 0.000 0.000 0.000 0.000

5 0 1 0 1 2 1 P= 0.185

0.000 0.200 0.000 0.200 0.400 0.200

24 1 2 1 10 8 2 P = 0.052

0.042 0.083 0.042 0.417 0.333 0.083

9241 48 1168 244 5211 2321 249

0.005 0.126 0.026 0.564 0.251 0.027

N, number of individuals; and F, frequency. Comparison of frequencies in patient groups with those in the general population.

age, o22, o42, o43, and o44 genotypes had 13-, 3-, 112-, and 3-fold attendance rates at the lipid clinic when compared with the common o33 genotype, respectively [OR: 13.4 (4.5–39.9); OR: 3.2 (1.2 – 8.4); OR: 1.7 (1.0– 2.8); OR: 3.0 (1.1–7.6), respectively] (Fig. 3, all hypertriglyceridemic patients). Similar trends were observed in the subgroups of patients with familial combined hyperlipidemia and other hypertriglyceridemia (Table 4 and Fig. 3, familial combined hyperlipidemia and other hypertriglyceridemia). These effects are in accordance with the effect of apoE genotype on levels of triglycerides in the general population: o22, o42, o43, and o44 carriers had increased levels of triglycerides compared with o33 carriers (Fig. 4, triglycerides). Among patients with type III hyperlipoproteinemia, two subjects carried o22 while one subject carried o32 (Table 4, type III hyperlipoproteinemia). Among patients characterized as chylomicronemic, one subject carried o32, one carried o33, two carried o43, and one subject carried o44 (Table 4, chylomicronemia).

patients, o22, o42, o43, and o44 all increased attendance rates when compared to o33. These findings reflected the effects of apoE genotype on cholesterol and triglyceride levels: o44 raised and o32 reduced cholesterol levels, and o22, o42, o43, and o44 all raised triglyceride levels.

4.1. o22 genotype The o22 genotype was overrepresented in this hyperlipidemic sample and, in particular, in the hypertriglyceridemic group, which yielded 5- and 13-fold attendance rates, respectively. Carriers of the o22 genotype have the propensity to develop type III hyperlipo-

3.4. Interactions ApoE genotype did not interact with gender or age in predicting attendance rate at the lipid clinic, all patients combined or in hypercholesterolemic or hypertriglyceridemic patients separately.

4. Discussion To compare whether apoE genotype influences an individual’s referral to a lipid clinic, we compared frequencies of each of the six apoE genotypes in patients attending a lipid clinic with apoE frequencies in individuals sampled from the general population. In hypercholesterolemic patients, o44 increased the attendance rate at the lipid clinic while o32 reduced the attendance rate when compared to o33. In hypertriglyceridemic

Fig. 3. Odds ratios and 95% confidence intervals (CI) for attendance of hypertriglyceridemic patients at a lipid clinic as a function of apoE genotype relative to the common o33 genotype. The logistic regression model allowed for 10 year age-groups; indicates common genotypes in the population at large (o33, o43, o32); indicates rare genotypes in the population at large (o22, o42, o44). The number of subjects in the analyses is the same as in Table 4.

466

R. Frikke-Schmidt et al. / Atherosclerosis 153 (2000) 461–468

holesterolemic patients compared with normolipidemic controls [29–32]. Furthermore, in a group of patients with familial hypercholesterolemia, the o2 allele seems to exert its LDL cholesterol lowering effect primarily in women [33]. In support of this, we found that the lower attendance rate associated with the o32 genotype was only present in hypercholesterolemic women (n= 84; OR: 0.3, 95% CI 0.1–0.8), and not in hypercholesterolemic men (n= 72; OR: 0.9, 95% CI 0.4–2.2).

4.3. o42 genotype

Fig. 4. Nonfasting plasma cholesterol and triglyceride levels as a function of apoE genotype in the general population, adjusted for age in 10 year groups, and average increase () for each genotype when compared to o33. Values are means 9 95% confidence intervals. Kruskal – Wallis analysis of variance: P-value B 0.001 for both cholesterol and triglycerides; on post-hoc Mann–Whitney U-test all two genotype comparisons (except o43 versus o44 for cholesterol, o43 versus o32 and o42 versus o44 for triglycerides) had P-values B 0.05.

proteinemia, characterized by a poor binding of the apoE2 isoform to lipoprotein receptors, thereby causing accumulation in plasma of remnants of both intestinal chylomicrons and hepatic VLDL, the b-VLDL. All o22 individuals have b-VLDL in plasma, however, some secondary factors such as age, obesity, glucose intolerance, alcohol consumption, postmenopausal status and hypothyroidism are necessary for the development of overt hyperlipidemia, and less than 10% of o22 individuals develop type III hyperlipoproteinemia [3]. In the present patient sample, two o22 carriers had type III hyperlipoproteinemia and the remaining three carriers were classified either as familial combined hyperlipidemia, other hypertriglyceridemia, or hypercholesterolemia by the attending consultant (HM). The interpretation of the observed 13-fold attendance rate should be that the o22 genotype is markedly over-represented in the overall hypertriglyceridemic sample compared with the general population sample.

4.2. o32 genotype The present finding in hypercholesterolemic patients that o32 reduces attendance rate at the lipid clinic to half that of o33 is supported by other studies reporting under-representation of the o32 genotype in hyperc-

Another novel finding is that carriers of the o42 genotype among hypertriglyceridemic patients exhibit a 3-fold attendance rate at the lipid clinic compared to o33. It is surprising that this genotype with a rather discrete effect on triglyceride levels in the general population exhibits a major increase in attendance rate at the lipid clinic. However, the finding is indirectly supported by previous studies which find frequencies of this genotype ranging from 0.042 to 0.069 in hypertriglyceridemic patients [30–32,34], larger than the frequency of 0.026 observed in our sample from the general population.

4.4. o43 and o44 genotypes The finding that o43 and o44 versus o33 overall have 112- and 2-fold attendance rates at the lipid clinic is indirectly supported by an over-representation of these genotypes in hyperlipidemic patients in previous studies [30–32,34]. It is plausible that o43 and o44 may influence attendance rates at lipid clinics of both hypercholesterolemic and hypertriglyceridemic patients, as both genotypes increase cholesterol as well as triglyceride levels, and as frequencies of o43 and o44 are over-represented in both hypercholesterolemic and hypertriglyceridemic patients in former studies [29– 32,34,35]. In the present setting, the over-representation of o43 and o44 genotypes did not seem to be isolated to specific subgroup phenotypes, as these two genotypes showed trends toward over-representation in patients with familial hypercholesterolemia, other LDL elevation, familial combined hyperlipidemia, as well as other hypertriglyceridemia.

4.5. Mechanism: susceptibility mutation for attending lipid clinic The apoE polymorphism has rather discrete effects on levels of lipids and lipoproteins in individuals in the general population [3], when compared with effects of LDL receptor and apoB mutations [16,22]. Despite the present findings, it is therefore questionable whether specific apoE genotypes alone will cause certain patients to be referred to lipid clinics with the exception of

R. Frikke-Schmidt et al. / Atherosclerosis 153 (2000) 461–468

the rare o22 genotype causing type III hyperlipidemia. It is more likely that the o44 genotype acts as a susceptibility mutation among hypercholesterolemic patients, and that o22, o42, o43, and o44 act as susceptibility mutations among hypertriglyceridemic patients for increased attendance at lipid clinics. In other words, the relatively small but additional lipid raising effect of these apoE genotypes causes an otherwise unnoticed lipid disturbance to become clinically recognizable, and thus causes the physician to refer the patient to a lipid clinic. In support of this idea, we have previously shown that the effect on plasma cholesterol levels of the apoB(Arg3500Gln) mutation depends on the background population with carriers identified in hyperlipidemic patients having levels of cholesterol approximately 2 mmol/l higher than carriers identified in the general population [22]. This additional increase in cholesterol levels in carriers among hyperlipidemic patients is not caused by the apoB mutations, but by susceptibility mutations like the apoE polymorphism and environmental or other factors present in these particular carriers. This is further supported by a study of several family members with the apoB(Arg3500Gln) mutation, where the o2 allele was associated with normal cholesterol levels or moderate hypercholesterolemia, whereas the o4 allele was associated with higher LDL cholesterol levels and the familial hypercholesterolemia phenotype with tendon xanthomas in the absence of an LDL-receptor defect [36]. Furthermore, apoE genotype also seems to modulate lipid levels in patients with LDL receptor mutations [29,33]. Thus, it could be that carriers of apoB mutations or LDL receptor mutations perhaps are more susceptible to be referred to a lipid clinic, if they in addition carry the cholesterol increasing o44 genotype, or that individuals with familial combined hyperlipidemia, diabetes mellitus or some other triglyceride raising condition are more susceptible to be referred to a lipid clinic if they in addition carry the triglyceride increasing o22, o42, o43, or o44 genotypes.

4.6. Limitations Since we genotyped about 9500 individuals, we cannot completely exclude misclassification of a few apoE genotypes; however, frequencies in the general population were in accordance with those predicted by the Hardy–Weinberg equilibrium and genotyping as well as data-base entry was scrutinized by two different researchers. Furthermore, attendance rates associated with the different apoE genotypes may not be identical to referral rates of new patients, as the patient population consisted of a mixed group of newly referred patients and of patients who had attended the lipid clinic previously. Referrals may also be influenced by the reputation of the lipid clinic, resulting for instance

467

in referral of more cases of familial hypercholesterolemia. Finally, selection bias naturally cannot totally be excluded; however, ‘controls’ and patients were recruited mainly from the same part of Copenhagen, the ‘controls’ were sampled to represent the general population, and patients attending the lipid clinic are most likely drawn from all classes of society, because hospital treatment is free of charge in Denmark.

5. Conclusion Approximately 30% of Caucasians carry apoE genotypes which increase the likelihood of referral to a lipid clinic, provided they have other genes or conditions also raising cholesterol or triglyceride levels.

Acknowledgements We thank Pia T. Petersen, Mette Refstrup and Hanne Damm for expert technical assistance. This study was supported by The Danish Heart Foundation, The Danish Medical Research Council, Chief Physician Johan Boserup and Lise Boserup’s Fund, Sven Hansen and Wife Ina Hansen’s Foundation, The Beckett Fund, European Organization for the Control of Circulatory Diseases, The Danish Stroke Association, and Director Jacob Madsen and Wife Olga Madsen’s Fund.

References [1] Paik YK, Chang DJ, Reardon CA, Davies GE, Mahley RW, Taylor JM. Nucleotide sequence and structure of the human apolipoprotein E gene. Proc Natl Acad Sci USA 1985;82:3445– 9. [2] Das HK, McPherson J, Bruns GAP, Karathanasis SK, Breslow JL. Isolation, characterization, and mapping to chromosome 19 of the human apolipoprotein E gene. J Biol Chem 1985;260:6240– 7. [3] Mahley RW, Rall SC, Jr. Type III hyperlipoproteinemia (dysbetalipoproteinemia): the role of apolipoprotein E in normal and abnormal lipoprotein metabolism. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The Metabolic and Molecular Bases of Inherited Disease. New York: McGraw-Hill, 1995:1953–80. [4] de Knijff P, Havekes LM. Apolipoprotein E as a risk factor for coronary heart disease: a genetic and molecular biology approach. Curr Opin Lipidol 1996;7:59 – 63. [5] Boer JMA, Ehnholm C, Menzel HJ, Havekes LM, Rosseneu M, St J. O’Reilly D, Tiret L. Interactions between lifestyle-related factors and the apoE polymorphism on plasma lipids and apolipoproteins. The EARS study. Arterioscler Thromb Vasc Biol 1997;17:1675 – 81. [6] Utermann G, Albrecht G, Steinmetz A. Polymorphism of apolipoprotein E. I. Methodological aspects and diagnosis of hyperlipoproteinemia type III without ultracentrifugation. Clin Genet 1978;14:351 – 8. [7] Utermann G, Vogelberg KH, Steinmetz A, Schoenborn W, Pruin N, Jaeschke M, Hees M, Canzler H. Polymorphism of

468

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

R. Frikke-Schmidt et al. / Atherosclerosis 153 (2000) 461–468 apolipoprotein E. II. Genetics of hyperlipoproteinemia type III. Clin Genet 1979;15:37–62. Utermann G, Pruin N, Steinmetz A. Polymorphism of apolipoprotein E. III. Effect of a single polymorphic gene locus on plasma lipid levels in man. Clin Genet 1979;15:63–72. Reilly SL, Ferrell RE, Kottke BA, Kamboh MI, Sing CF. The gender-specific apolipoprotein E genotype influence on the distribution of lipids and apolipoproteins in the population of Rochester, MN. I. Pleiotropic effects on means and variances. Am J Hum Genet 1991;49:1155–66. Dallongeville J, Lussier-Cacan S, Davignon J. Modulation of plasma triglyceride levels by apoE phenotype: a meta-analysis. J Lipid Res 1992;33:447–54. Reilly SL, Ferrell RE, Kottke BA, Sing CF. The gender-specific apolipoprotein E genotype influence on the distribution of plasma lipids and apolipoproteins in the population of Rochester, MN. II. Regression relationships with concomitants. Am J Hum Genet 1992;51:1311–24. Wilson PWF, Myers RH, Larson MG, Ordovas JM, Wolf PA, Schaefer EJ. Apolipoprotein E alleles, dyslipidemia, and coronary heart disease. The Framingham Offspring Study. J Am Med Assoc 1994;272:1666–71. Reilly SL, Ferrell RE, Sing CF. The gender-specific apolipoprotein E genotype influence on the distribution of plasma lipids and apolipoproteins in the population of Rochester, MN. III. Correlations and covariances. Am J Hum Genet 1994;55:1001 – 18. Luc G, Bard JM, Arveiler D, Evans A, Cambou JP, Bingham A, Amouyel P, Schaffer P, Ruidavets JB, Cambien F, Fruchart JC, Ducimetiere P. Impact of apolipoprotein E polymorphism on lipoproteins and risk of myocardial infarction. The ECTIM Study. Arterioscler Thromb 1994;14:1412–9. Kane JP, Havel RJ. Disorders of the biogenesis and secretion of lipoproteins containing the B apolipoproteins. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The Metabolic and Molecular Bases for Inherited Diseases. New York: McGrawHill, 1995:1853 – 85. Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The Metabolic and Molecular Bases of Inherited Disease. New York: McGraw-Hill, 1995:1981–2030. Brunzell JD. Familial lipoprotein lipase deficiency and other causes of the chylomicronemia syndrome. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The Metabolic and Molecular Bases of Inherited disease. New York: McGraw-Hill, 1995:1913 – 32. Breslow JL. Familial disorders of high-density lipoprotein metabolism. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The Metabolic and Molecular Bases of Inherited Disease. New York: McGraw-Hill, 1995:2031–52. Appleyard M, Hansen AT, Jensen G, Schnohr P, Nyboe J. The Copenhagen City Heart Study. Østerbroundersøgelsen. A book of tables with data from the first examination (1976–1978) and a 5 year follow-up (1981–1983). The Copenhagen City Heart Study Group. Scand J Soc Med Suppl 1989;41:1–160. Wittrup HH, Tybjærg-Hansen A, Abildgaard S, Steffensen R, Schnohr P, Nordestgaard BG. A common substitution (asn291ser) in lipoprotein lipase is associated with increased risk of ischemic heart disease. J Clin Invest 1997;99:1606–13.

.

[21] Tybjærg-Hansen A, Agerholm-Larsen B, Humphries SE, Abildgaard S, Schnohr P, Nordestgaard BG. A common mutation (G-455“A) in the b-fibrinogen promoter is an independent predictor of plasma fibrinogen, but not of ischemic heart disease. J Clin Invest 1997;99:3034 – 9. [22] Tybjærg-Hansen A, Steffensen R, Meinertz H, Schnohr P, Nordestgaard BG. Association of mutations in the apolipoprotein B gene with hypercholesterolemia and the risk of ischemic heart disease. N Eng J Med 1998;338:1577– 84. [23] Agerholm-Larsen B, Tybjærg-Hansen A, Frikke-Schmidt R, Grønholdt MLM, Jensen G, Nordestgaard BG. ACE gene polymorphism as a risk factor for ischemic cerebrovascular disease. Ann Intern Med 1997;127:346 – 55. [24] Nordestgaard BG, Abildgaard S, Wittrup HH, Steffensen R, Jensen G, Tybjærg-Hansen A. Heterozygous lipoprotein lipase deficiency. Frequency in the general population, effect on plasma lipid levels, and risk of ischemic heart disease. Circulation 1997;96:1737 – 44. [25] Hixson JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HhaI. J Lipid Res 1990;31:545 – 8. [26] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499 – 502. [27] SPSS for Windows 1998. Base System User’s Guide and Professional Statistics, release 8.0. Chicago, IL: SPSS, Inc, 1998. [28] Snedecor GW, Cochran WG. The random effects model. In: Snedecor GW, Cochran WG, editors. Statistical Methods. Ames, IA: The Iowa State University Press, 1980:253 – 4. [29] Tonstad S, Leren TP, Sivertsen M, Ose L. Determinants of lipid levels among children with heterozygous familial hypercholesterolemia in Norway. Arterioscler Thromb Vasc Biol 1995;15:1009 – 14. [30] Ghiselli G, Gregg RE, Zech LA, Schaefer EJ, Brewer HB, Jr. Phenotype study of apolipoprotein E isoforms in hyperlipoproteinaemic patients. Lancet 1982;2:405 – 7. [31] Utermann G, Kindermann I, Kaffarnik H, Steinmetz A. Apolipoprotein E phenotypes and hyperlipidemia. Hum Genet 1984;65:232 – 6. [32] Assmann G, Schmitz G, Menzel HJ, Schultz H. Apolipoprotein E polymorphism and hyperlipidemia. Clin Chem 1984;30:641–3. [33] Ferrie`res J, Sing CF, Roy M, Davignon J, Lussier-Cacan S. Apolipoprotein E polymorphism and heterozygous familial hypercholesterolemia. Sex-specific effects. Arterioscler Thromb 1994;14:1553 – 60. [34] Lussier-Cacan S, Bouthillier D, Davignon J. Apo E allele frequency in primary endogenous hypertriglyceridemia (type IV) with and without hyperapobetalipoproteinemia. Arteriosclerosis 1985;5:639 – 43. [35] Leren TP, Børresen AL, Berg K, Hjermann I, Leren P. Increased frequency of the apolipoprotein E-4 isoform in male subjects with multifactorial hypercholesterolemia. Clin Chem 1985;27:458 – 62. [36] Davignon J, Dufour R, Roy M, Be´tard C, Ma Y, Ouellette S, Boulet L, Lussier-Cacan S. Phenotypic heterogeneity associated with defective apolipoprotein B-100 and occurrence of the familial hypercholesterolemia phenotype in the absence of an LDL receptor defect within a Canadian kindred. Eur J Epidemiol 1992;8(Suppl. I):10 – 7.