Phenotypic Variation in Patients Heterozygous for Familial Defective Apolipoprotein B (FDB) in Three European Countries
Abstract A glutamine-for-arginine substitution at amino acid position 3500 of apolipoprotein B (apo B) causes synthesis of LDL with reduced binding affinity to the LDL receptor (LDLR). The associated clinical syndrome has been named familial defective apolipoprotein B-100 (FDB). In 205 FDB patients from Germany (n=73), the Netherlands (n=87), and Denmark (n=45), we tried to assess determinants of variation in lipid concentrations. Besides age, sex, and geographic origin, variation in the LDLR gene was the most powerful determinant of variation in total cholesterol and LDL cholesterol levels. Polymorphic variation in the LDLR gene (SfaNI, exon 2; Nco I, exon 18) was associated with total cholesterol (TC) and LDL cholesterol (LDL-C) variation in women (SfaNI: P=.04 and .03 for TC and LDL-C, respectively; Nco I: P=.003 and .006, respectively), whereas the Ava II (exon 13) and the Pvu II (intron 15) polymorphisms were not. Combined information from all three LDLR exon polymorphisms showed that subjects with at least one S+A+N+ allele had 13% to 20% higher TC than non-S+A+N+ subjects (P=.02 [TC, men]; P=.01 [LDL-C, men]; P=.005 [TC, women]; and P=.004 [LDL-C, women]) and, together with age and geographic origin, accounted for 20% (women) and 19% (men) of the variation in LDL-C. The expected association of the apo E genotypes (e3e2, e3e3, and e3e4) with cholesterol concentrations was seen in S+A+N+ but not in non-S+A+N+ subjects and in P−P− but not in P+P+ or P+P− subjects. With regard to clinical expression, FDB patients had lower TC and LDL-C levels and a lower prevalence of cardiovascular disease than 101 Danish patients with familial hypercholesterolemia.
- familial defective apolipoprotein B
- familial hypercholesterolemia
- low-density lipoprotein receptor
- lipoprotein lipase
- apolipoprotein B
- apolipoprotein E
- Received November 21, 1995.
- Accepted July 18, 1996.
Familial defective apolipoprotein B-100 is an autosomal- dominant disorder due to a G-to-A substitution in the codon for amino acid 3500, resulting in substitution of glutamine for arginine.1 The mutation greatly reduces the affinity of apo B-100 for the LDLR.1 2 Its frequency has been estimated to be 1 in 500 to 1 in 700 in populations of Mid-European origin2 3 4 and 1 in 1250 in Denmark.5 In almost all patients, the mutation is on apo B genes carrying identical haplotypes, suggesting a common origin,6 but two other haplotypes have been described in FDB patients of German7 and Chinese-American origin.8
Patients heterozygous for the apo B-3500 mutation are usually clearly hypercholesterolemic, but cholesterol concentrations in plasma can vary from those found in FH to only modest hypercholesterolemia.1 9 10 11 12 In the 45 Danish patients involved in the present study, we have previously reported that one third had cholesterol levels below the 95th percentile for age and sex.13 Cholesterol concentrations can also vary significantly within the same family.13
März et al14 found that the impaired ligand function of apo B-3500 depended on the size of LDL particles: large LDL had normal or near-normal affinity to normal LDLR, and small LDL had greatly reduced affinity. A significant portion of apo B-3500–carrying lipoproteins was apparently catabolized through interaction of apo E with the LDLR or a specific apo E receptor.
The primary aim of the present study was to assess causes—other than the apo B-3500 mutation itself—of interindividual variation in TC and LDL-C concentrations in FDB heterozygotes. Proteins, other than the apo B-3500, likely to influence the conversion of VLDL to LDL and the catabolism of LDL precursors and LDL in FDB patients are LPL, apo E, the normal apo B protein (in heterozygous FDB), and the LDLR. To see whether genetic variation in any of these proteins adds to the interindividual variation in TC and LDL-C among FDB heterozygotes, we sought an association of polymorphic variation in these genes with lipid variation in 205 FDB heterozygotes.
An additional aim of this study was to assess whether the clinical phenotype of FDB differs from that of FH. We therefore compared plasma lipids and the frequency of CVD in FDB patients with data from a sample of 101 Danish FH index patients.
In total, 205 heterozygous FDB patients from 47 apparently unrelated families were studied. Most index patients were originally diagnosed as having FH. The Danish patients13 and a small number of German and Dutch patients10 12 have been described previously. Data from 101 Danish FH patients, diagnosed on clinical criteria including tendon xanthomata in the patient or in a first-degree relative as previously described,15 in whom the apo B-3500 mutation was excluded, were included for purposes of comparison of lipid concentrations and prevalence of CVD. None of the patients were treated with lipid-lowering diet or drugs at the time of blood sampling for determination of lipoprotein parameters.
Plasma concentrations of TC, HDL-C, and TG were measured by standard colorimetric methods (Boehringer). Six serum pools, selected to have TC concentrations covering the range from 4 to 11 mmol/L, were analyzed in the three participating laboratories (Amsterdam, Munich, and Aarhus), to assess differences in measurement levels. TC and TG levels differed slightly, corresponding to maximum deviations of about 0.2 mmol/L from grand means. The differences were constant across the concentration ranges covered, so that TC and TG data from the individual laboratories could be adjusted by simply adding or subtracting fixed values. LDL-C was calculated from the Friedewald formula.16 All blood samples were drawn after an overnight fast.
Genomic DNA was extracted from blood leukocytes by standard techniques.17 The apo B-3500 mutation was detected by the methods described by Defesche et al12 (Netherlands), Rauh et al10 (Germany), or Hansen et al18 (Denmark). All polymorphisms were detected after DNA amplification by restriction enzyme cleavage and/or size separation by gel electrophoresis. The common apo E genotypes were determined by the method of Hixon and Vernier.19 Apo B gene polymorphisms (SP24/27 polymorphism in exon 1,20 Xba I RFLP in exon 26,21 and 3′ VNTR length polymorphism22 ) were determined as previously described.23 LPL gene polymorphisms (HindIII and Pvu II) were determined as previously described.24 LDLR gene polymorphisms (SfaNI, Ava II, and Nco I) were determined essentially as described by Humphries et al.25 The LDLR Pvu II polymorphism was determined by long-distance polymerase chain reaction essentially as described by Plessis and Kotze.26
Distributions of plasma lipid concentrations were tested for normality, and triglycerides were loge transformed to obtain normal distribution. TC, LDL-C, and TG concentrations were significantly correlated with age for both men and women, and data were therefore adjusted by linear regression. Differences in age and lipid concentrations between men and women and between FDB and FH patients were assessed by Student's t test or Mann-Whitney test (TG), and differences between countries were assessed by ANOVA or the Kruskal-Wallis test (TG).
Apo E, apo B, LPL, and LDLR alleles were measured by gene counting. Differences in allele frequencies were evaluated by χ2 test. All tests of association between genotypes and cholesterol variation were performed separately for men and women. Associations between genotypes and cholesterol concentrations were assessed by ANOVA or Kruskal-Wallis test (TG). Multiple regression was used to estimate the linear effects of concomitants (age and geographic origin) and genotypes for each sex separately. The percentages of the total sum of squares attributable to concomitants and to genotype differences (after adjustment for the effects of concomitants) were estimated by the same regression analysis.27 All tests were performed on data from each country separately as well as on data from the whole sample. Significance was considered to be at the 5% level. Adjustment for multiple comparisons was not performed.28 To be considered biologically relevant, genetic associations with lipid variation should be consistent across countries.
Women had higher TC, LDL-C, and HDL-C than men in Denmark and the Netherlands but insignificantly so in Germany (Table 1⇓). Mean LDL-C concentrations differed slightly between countries. TC, plotted for each family (Fig 1⇓), indicates considerably greater within-family than between-family variation in TC in all three countries.
Together, age and geographic origin accounted for 12% (women) and 11% (men) of the interindividual variation in TC and 8% (both men and women) of the variation in LDL-C (Table 2⇓).
Genotype and allele frequencies for the polymorphisms examined here are given in Table 3⇓. There was no significant difference in the distribution of apo E, LPL, or LDLR alleles between countries. The distribution of all polymorphisms conformed to Hardy-Weinberg expectations, and the allele frequencies for all polymorphic markers were comparable to previously published frequencies in white populations.23 24 25 29 30 31
The SfaNI, the Ava II, and the Nco I LDLR RFLPs seemed to be associated with TC and LDL-C in a dose-dependent manner (Table 4⇓). The S+S+, A+A+, and N+N+ genotypes were associated with the highest concentrations and the S−S−, A−A−, and N−N− with the lowest concentrations. Heterozygous subjects had intermediate concentrations. These associations were statistically significant only in women, however, and only regarding the SfaNI and the Nco I RFLPs. On the other hand, the tendencies were consistent across countries.
Since the SfaNI, Ava II, and Nco I RFLPs are in strong linkage disequilibrium,25 30 we grouped FDB heterozygotes into those with at least one S+A+N+ LDLR allele and those without an S+A+N+ LDLR allele. ANOVA analysis of TC and LDL-C concentrations in these two groups suggested greater differences than when each RFLP was tested separately (Table 4⇑). Women with at least one S+A+N+ allele had 14% higher TC and 20% higher LDL-C than women without an S+A+N+ allele (P=.005 and .004, respectively). Men with at least one S+A+N+ allele had 13% higher TC and 17% higher LDL-C than men without an S+A+N+ allele (P=.02 and .01, respectively). Again, these associations were consistent across countries.
There was no association between the Pvu II RFLP and cholesterol concentration in either men or women (Table 4⇑).
Apo E, LPL, and the Normal Apo B Allele
There was no significant association between apo E genotypes and TC or LDL-C concentrations. We found an interaction between the apo E genotype–associated cholesterol variation and variation in the LDLR gene, however (see below).
Neither the SP24/27, the Xba I polymorphism, nor the diallelic haplotypes were associated with TC or LDL-C variation in either sex. Furthermore, there was no particular pattern of apo B haplotypes associated with higher or lower cholesterol concentrations in informative families.
There was no significant association between the LPL HindIII or Pvu II RFLPs and variation in TC or LDL-C. As in a previous study of normal Danish males,24 we found a significant association between the LPL HindIII RFLP and HDL-C variation (data not shown).
LDLR/Apo E Interaction
The association of higher cholesterol concentrations with the S+A+N+ allele was consistent across the common apo E genotypes but insignificantly so in e3e2 men. As shown in Fig 2⇓, the effect was greater in e3e4 subjects than in the other apo E genotype groups. Since the effect associated with the LDLR haplotypes was of similar magnitude and direction in both men and women, we analyzed the apo E interactions in both sexes together. Compared with non-S+A+N+ subjects, S+A+N+/e3e3 subjects had 11% higher TC (P=.04); S+A+N+/e3e4, 20% higher TC (P=.02); and S+A+N+/e3e2, 7% higher TC (P=.30; not significant). Thus, we saw the expected association of apo E genotypes with cholesterol variation (e3e2<e3e3<e3e4) in S+A+N+ subjects. This was also the case in subjects who were homozygous for the absence of the Pvu II restriction site (Fig 2⇓) but not in P+P+ or P+P− subjects.
FDB Compared With FH
Women as well as men with FDB had significantly lower concentrations of TC and LDL-C than the 101 Danish FH index patients (Table 1⇑). The frequency of tendon xanthomata and arcus cornea increased with age. Forty-five percent of FDB patients had tendon xanthomata by the age of 50 years (Fig 3⇓). The frequency of CVD reached ≈30% by the age of 60 years and further increased to ≈60% above 70 years of age. In all age groups, the frequency of CVD was lower in FDB patients than in FH (Fig 3⇓).
The present sample of 205 FDB heterozygotes from 47 families in three countries is the largest so far studied to ascertain causes of interindividual variation in plasma cholesterol concentrations. Since most of the patients had the apo B-3500 mutation on the same apo B gene haplotype background, they may have a common ancestor who for two reasons was probably separated from our patients by many generations. First, our population sample was in Hardy-Weinberg equilibrium for all polymorphisms except those in the apo B gene itself. Second, the allele frequencies of the polymorphisms studied were similar to those in other white population samples.23 24 25 29 30 31 The FDB patients in our study are therefore likely to represent the majority of FDB patients. A drawback of our study is that most of the index patients might have been selected because of a particularly unfortunate combination of genes leading to conditions such as very early atherosclerotic disease or very high cholesterol concentrations. Inclusion of FDB family members in the study sample does not exclude this potential bias, but there was no difference in any of the measured lipid concentrations between index patients and family members (data not shown). However, in the comparison with FH patients, such a selection bias would be expected to reduce a given difference between FDB and FH phenotypes.
Women had higher age-adjusted TC, HDL-C, and LDL-C than men. Since the sex difference in lipids applied to all 10-year age groups (data not shown), it is not likely to be due to estrogen effects on lipoprotein metabolism. A possible explanation is selective female survival: men with high LDL-C were possibly less likely than women to survive to be studied. The slightly younger mean age of the men is consistent with this idea, together with the fact that only 1 man of 87 was above 70 years of age, as opposed to 14 of 118 women. As in FDB, FH women had higher TC and LDL-C than FH men.
TC, LDL-C, and TG increased with age in FDB patients, which is consistent with findings in normolipidemic populations. In contrast, LDL-C did not change with age in either men or women with FH. This could support the idea that the LDL-C elevation in FDB patients is more affected by lifestyle (eg, fat consumption and sedentary life) than it is in FH. The German FDB patients had the highest TC and LDL-C concentrations. This could be caused by selection criteria, but it could also be due to differences in fat consumption and/or different exercise habits. We have not performed detailed diet interviews, and we did not have information on body weight and height of a quality that allowed direct comparison between countries. In the Dutch patients, however, there was no correlation between body mass index and TC or LDL-C concentrations (R2=0 [men] and .04 [women], P=.98 and .22 for TC; R2=.008 [men] and .05 [women], P=.64 and .15 for LDL-C).
In addition to the well-known major contribution to decreased LDL catabolism of LDLR mutations in FH, several publications have indicated that minor, genetically determined alterations of LDLR function and/or expression may be associated with interindividual variation in cholesterol concentrations in the normal population.29 30 32 33 The most consistent findings have been that the P− variant of the Pvu II polymorphism in intron 15 is associated with higher cholesterol compared with the P+ variant.32 33 We found the same tendency, but it was not significant (Table 4⇑). A few studies have found that other polymorphisms in the LDLR gene are associated with cholesterol variation (eg, Nco I in exon 1830).
The higher TC and LDL-C concentrations associated with presence of the SfaNI (S+), Ava II (A+), and Nco I (N+) alleles and especially with presence of the haplotype S+A+N+ (Table 4⇑) suggest that one or more minor genetic alterations associated with the LDLR S+A+N+ variant significantly affect LDL catabolism in FDB patients. German FDB heterozygotes had a slightly higher frequency (0.60) of the S+A+N+ allele than Danish (0.5) and Dutch (0.51) patients, which may partly explain the higher TC and LDL-C in the German patients. Although the Pvu II RFLP was in significant linkage disequilibrium with both the Ava II, SfaNI, and Nco I RFLPs, the distribution of Pvu II alleles did not differ between S+A+N+ and non-S+A+N+ subjects (χ2 df=1=0.19; not significant). Further, combination of the Pvu II RFLP with the other three RFLPs did not add information to the association with cholesterol variation.
Together, age, geographic origin, and LDLR polymorphism categorized as presence or absence of at least one S+A+N+ allele accounted for 23% (women) and 21% (men) of the variation in TC and 20% (women) and 19% (men) of the variation in LDL-C (Table 2⇑).
LDLR/Apo E Interaction
Since the major part of LDL in FDB heterozygotes (≈70%) has very low affinity for normal LDLR,1 this considerable association between LDLR variation and cholesterol concentrations may be due to different affinities between the LDLR and the other ligand, apo E. Therefore, we analyzed whether the difference in cholesterol by apo E group differed between S+A+N+ and non-S+A+N+ carriers (Fig 2⇑). Since the direction and magnitude of association was essentially the same in men and women, these analyses were performed on the patient group as a whole to increase the statistical power. As can be seen from Fig 2⇑, the presence of an S+A+N+ allele greatly influences the between-group (apo E) variation in cholesterol concentration. Among non-S+A+N+ FDB heterozygotes there is hardly any difference in cholesterol levels, whereas in the S+A+N+ group, we saw the same association as in the general population: e3e4 subjects had higher and e3e2 subjects had lower cholesterol concentrations than e3e3 homozygotes. Probably due to the relatively low number of e3e2 subjects, there was no statistically significant difference in cholesterol concentrations between S+A+N+ and non-S+A+N+ carriers. The greatest difference in cholesterol concentrations is seen among e3e4 subjects. A possible explanation is a decreased affinity of the E4 protein variant for S+A+N+ LDLR variants. This would be expected to decrease remnant catabolism and therefore cause more VLDL to be converted to LDL in these patients.
A gene-gene interaction has previously been suggested for the apo E polymorphism and the LDLR polymorphism detected by the Pvu II RFLP in intron 15.33 In these studies, presence of the Pvu II cleavage site abolished the normal association of apo E polymorphism and cholesterol concentrations. We found the same pattern in the present study (Fig 2⇑). Further, it seems as if the presence of the S+A+N+ together with the P− variant potentiates the cholesterol-elevating effect of the e4 allele (Fig 2⇑). At present, however, we do not know the physiological basis for the S+A+N+ and the Pvu II associations with cholesterol variation.
In conclusion, a considerable part of the interindividual variation in TC and LDL-C concentrations in FDB heterozygotes can be accounted for by difference in age and sex and variations in the LDLR. Further, there may be an important interaction between one or more variants in the LDLR gene and the cholesterol variation associated with the apo E genotypes.
FDB Versus FH
Some studies have suggested that patients with FDB and FH have similar concentrations of plasma cholesterol,9 12 whereas others1 3 13 indicate that FDB patients in general are less severely hypercholesterolemic. Our results support the latter view, but an important source of bias in studies of this kind, including our own, is that FDB patients are diagnosed by molecular genetic analysis, whereas FH patients are identified by clinical criteria that typically include plasma cholesterol concentrations over the 95th percentile for the population. Since most of the FDB patients are ascertained through index patients originally diagnosed as FH heterozygotes, however, this bias would be expected to diminish the difference in the phenotypic appearance of FDB and FH patients. Thus, although our results suggest that hypercholesterolemia is less severe in FDB than in FH heterozygotes, the definitive answer to the degree of hypercholesterolemia in FDB and in FH has to await screening of an unbiased sample of the general population based on molecular-genetic diagnosis of both FDB and FH patients.
Tendon xanthomata were frequent in this sample of FDB heterozygotes (Fig 3⇑), but since the majority of the FDB patients were ascertained through index patients initially diagnosed as FH patients, the frequency of tendon xanthomata in an unbiased sample of FDB patients may be lower than in our sample. Compared with the 101 Danish FH index patients, the frequency of CVD is somewhat lower in FDB heterozygotes in all age groups (Fig 3⇑) although still well above the prevalence of CVD in the general Danish population in the same age groups.34 Thus, although presence of the apo B-3500 mutation can mimic the biochemical and clinical features of FH, our results support the concept that the FDB phenotype in regard to both LDL-C elevation and CVD is somewhat milder than the FH phenotype.
Selected Abbreviations and Acronyms
|FDB||=||familial defective apolipoprotein B-100|
|RFLP||=||restriction fragment length polymorphism|
This study was supported in part by a grant from the Danish Heart Foundation (to Dr Hansen). Drs Hans Meinertz (Rigshospitalet, Copenhagen) and Annebirthe Bo Hansen (University Hospital of Odense) are acknowledged for providing DNA and plasma samples from Danish FH patients. The authors wish to thank Pia Hornbek, Gitte Glistrup Nielsen, Anette Stenderup, Helmy Brink, and Alice van Zijl for excellent technical assistance.
Innerarity TL, Mahley RW, Weisgraber KH, Bersot TP, Krauss RM, Vega GL, Grundy SM, Friedl W, Davignon J, McCarthy BJ. Familial defective apolipoprotein B-100: a mutation of apolipoprotein B that causes hypercholesterolemia. J Lipid Res. 1990;31:1337-1350.
Soria LF, Ludwig EH, Clarke HRG, Vega GL, Grundy SM, McCarthy BJ. Association between a specific apolipoprotein B mutation and familial defective apolipoprotein B-100. Proc Natl Acad Sci U S A. 1989;86:587-591.
Miserez AR, Laager R, Chiodette N, Keller U. High prevalence of familial defective apolipoprotein B-100 in Switzerland. J Lipid Res. 1994;35:574-583.
Rauh G, Schuster H, Schewe CK, Stratmann G, Keller C, Wolfram G, Zöllner N. Independent mutation of arginine(3500)-glutamine associated with familial defective apolipoprotein B-100. J Lipid Res. 1993;34:799-805.
Bersot TP, Russell SJ, Thatcher SR, Pomernacki NK, Mahley RW, Weisgraber KH, Innerarity TL, Fox CS. A unique haplotype of the apolipoprotein B-100 allele associated with familial defective apolipoprotein B-100 in a Chinese man discovered during a study of the prevalence of this disorder. J Lipid Res. 1993;34:1149-1154.
Tybjaerg-Hansen A, Gallagher J, Vincent J, Houlston R, Talmud P, Dunning AM, Seed M, Hamsten A, Humphries SE, Myant NB. Familial defective apolipoprotein B-100: detection in the United Kingdom and Scandinavia, and clinical characteristics of ten cases. Atherosclerosis. 1990;80:235-242.
Schuster H, Rauh G, Kormann B, Hepp T, Humphries S, Keller K, Wolfram G, Zöllner N. Familial defective apolipoprotein B-100: comparison with familial hypercholesterolemia in 18 cases detected in Munich. Arteriosclerosis. 1990;4:577-581.
Hansen PS, Meinertz H, Jensen HK, Fruergaard P, Launbjerg J, Klausen IC, Lemming L, Gerdes LU, Gregersen N, Faergeman O. Characteristics of 46 heterozygous carriers and 57 unaffected relatives in five Danish families with familial defective apolipoprotein B-100. Arterioscler Thromb. 1994;14:207-213.
März W, Baumstark MW, Scharnagl H, Ruzicka V, Buxbaum S, Herwig J, Pohl T, Russ A, Schaaf L, Berg A, Böhles H-J, Usadel KH, Gross W. Accumulation of ‘small dense’ low density lipoproteins (LDL) in a homozygous patient with familial defective apolipoprotein B-100 results from heterogenous interaction of LDL subfractions with the LDL receptor. J Clin Invest. 1993;92:2922-2933.
Hansen PS, Jensen HK, Meinertz H, Hansen ABB, Klausen IC, Gerdes LU, Hörder M, Gregersen N, Faergeman O. Apolipoprotein B and E gene polymorphisms and association with plasma lipids and atherosclerotic disease in familial hypercholesterolemia. Nutr Metab Cardiovasc Dis. 1994;4:204-208.
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.
Kunkel LM, Smith KD, Boyer SH, Borgaonkar DS, Wachtel SS, Miller OJ, Breg WR, Jones HW, Rary JM. Analysis of human Y-chromosome–specific reiterated DNA in chromosome variants. Proc Natl Acad Sci U S A. 1977;74:1245-1249.
Hansen PS, Rüdiger N, Tybjaerg-Hansen A, Faergeman O, Gregersen N. Detection of the apoB-3500 mutation by gene amplification and cleavage with Msp I. J Lipid Res. 1991;32:1229-1233.
Hixon JE, Vernier DT. Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with Hha I. J Lipid Res. 1990;31:545-548.
Boerwinkle E, Chan L. A three-codon insertion/deletion polymorphism in the signal peptide coding region of the human apolipoprotein B (apo B) gene directly typed by the polymerase chain reaction. Nucleic Acids Res. 1989;17:4003.
Priestley L, Knott T, Wallis S, Powell L, Pease R, Simon A, Scott J. RFLP for the human apolipoprotein B gene, I-V. Nucleic Acids Res. 1985;18:6789.
Huang LS, Breslow J. A unique AT-rich hypervariable minisatellite 3′ to the apo B gene defines a high information restriction fragment length polymorphism. J Biol Chem. 1987;262:8952-8955.
Gerdes C, Gerdes LU, Hansen PS, Faergeman O. Polymorphisms in the lipoprotein lipase gene and their associations with plasma lipid concentrations in 40-year-old Danish men. Circulation. 1995;92:1765-1769.
Humphries S, King-Underwood L, Gudnason V, Seed M, Delattre S, Clavey V, Fruchart J-C. Six DNA polymorphisms in the low density lipoprotein receptor gene: their genetic relationship and an example of their use for identifying affected relatives of patients with familial hypercholesterolemia. J Med Genet. 1993;30:273-279.
Plessis L, Kotze MJ. Detection of the Pvu II RFLP in intron 15 of the LDL receptor gene by long-distance PCR. Clin Genet. 1996. In press.
Ferrières J, Sing CF, Roy M, Davignon J, Lussier-Cacan S. Apolipoprotein E polymorphism and heterozygous familial hypercholesterolemia. Arterioscler Thromb. 1994;14:1553-1560.
Ahn YI, Kamboh I, Aston CE, Ferrell RE, Hamman RF. Role of common genetic polymorphisms in the LDL receptor gene in affecting plasma cholesterol levels in the general population. Arterioscler Thromb. 1994;14:663-670.
Humphries S, Coviello DA, Masturzo P, Balestreri R, Orecchini G, Bertolini S. Variation in the low-density lipoprotein receptor gene is associated with differences in plasma low-density lipoprotein cholesterol levels in young and old normal individuals from Italy. Arterioscler Thromb. 1991;11:509-516.
Appleyard M, ed. The Copenhagen City Heart Study: Østerbro-undersøgelsen: a book of tables with data from the first examination (1976-78) and a five year follow-up (1981-83). Scand J Soc Med. 1989;41(suppl 41):20-23 and 26-27.