Association of Apolipoprotein(a) Phenotypes in Children With Family

1609

Association of Apolipoprotein(a) Phenotypes inChildren With Family History of Premature

Coronary Artery Disease

Syed Islam, Bernard Gutin, Clayton Smith, Frank Treiber, M. Ilyas Kamboh

Abstract Although blacks have higher plasma levels oflipoprotein(a) [Lp(a)] than whites, the Lp(a) levels are notassociated with clinical coronary artery disease (CAD) orparental history of myocardial infarction in blacks. To explorewhether ethnic differences in the pathogenicity of Lp(a) arerelated to the thrombogenic component of Lp(a), this studyinvestigated in children the associations of apolipoprotein(a)[apo(a)] phenotypes and Lp(a) levels with family history ofpremature CAD. Subjects were 46 children aged 7 to 11 yearsdivided according to family history of premature CAD andassessed for Lp(a), apo(a) phenotypes, and other lipids andlipoproteins. The prevalence of small isofonns was higher inchildren with positive family history of premature CAD thanin children with negative family history of premature CAD

Numerous case-control and cross-sectional stud-ies have indicated a moderate to strong asso-ciation of lipoprotein(a) [Lp(a)] levels with

premature coronary artery disease (CAD),16 restenosisof previously dilated coronary arterial segments,7 pre-clinical atherosclerosis,8 and parental histories of myo-cardial infarction (MI),9-10 particularly among whites.Despite higher levels of Lp(a) in blacks compared withwhites, the Lp(a) levels were not associated with clinicalCAD or parental history of MI.10-11 These ethnic differ-ences in the pathogenicity of Lp(a) were unexplained.

The Lp(a) particle contains apolipoprotein (apo) (a),a highly glycosylated water-soluble protein that is linkedto water-insoluble apoB-100, the principal protein moi-ety of low-density lipoprotein (LDL).12 The apo(a)glycoprotein shows considerable homology with thekringle IV domain of plasminogen. The close structuralrelation between apo(a) and plasminogen is also re-flected by the immunologic cross-reactivity betweenthese two proteins.13 Because of this homology, Lp(a),through its apo(a) part, competes with plasminogen forbinding with fibrinogen or fibrin and thus attenuatesfibrinotysis.1413 Therefore, the apo(a) portion of Lp(a)plays a major role in the pathogenesis of CAD. It hasbeen speculated that the types of apo(a) isoforms ratherthan the absolute levels of Lp(a) may be important indetermining the risk of premature CAD and may par-

Received March 9, 1994; revision accepted July 19, 1994.From the Georgia Prevention Institute, Department of Pediat-

rics, Medical College of Georgia, Augusta, and the Department ofHuman Genetics, University of Pittsburgh (Pa) (I.K.).

Correspondence to Syed Islam, MBBS, DrPH, Georgia Preven-tion Institute, Department of Pediatrics, Medical College ofGeorgia, Augusta, GA 30912-3710.

© 1994 American Heart Association, Inc.

(32% versus 10%). Large isofonns were more prevalent inwhites (24% versus 6%), and medium-sized isoforms weremore prevalent in blacks (75% versus 52%). The black/whitedifference was smaller (19% versus 24%) in regard to smallisoforms. Lp(a) levels were inversely related to apo(a) size inboth blacks and whites (P=.O84 and P=M9, respectively).Single-banded small apo(a) isoforms predicted positive familyhistory of premature CAD, independent of ethnicity andLp(a) levels. Small apo(a) isofonns in children were indepen-dent predictors of family history of premature CAD. UnlikeLp(a), they appear to be equally pathogenic for blacks andwhites. (ArUrioscler Thromb. 1994;14:1609-1616.)

Key Words • lipoprotein(a) • family history of CAD •apo(a) phenotype • children

tially explain the ethnic differences in the pathogenicityof Lp(a). The size of apo(a) isoforms has been shown tobe inversely correlated with the levels of Lp(a) inwhites. Such an association has not been well estab-lished in blacks. Recent genetic studies have shown thatmost of the variability of Lp(a) can be explained byapo(a) gene polymorphism,16-17 and only certain formsof apo(a) (small) are related to high Lp(a) levels andclinical CAD.1819 Subsequently, it has been documentedthat other forms (intermediate size) are also related tohigh Lp(a) levels,20 and there is considerable variationin the Lp(a) levels within each category of the apo(a)isoforms.16'17'20 Initially, it was suggested that these vari-ations in the Lp(a) levels within each category of apo(a)isoform may be due to insensitive methods of measure-ment of apo(a) protein size.16 However, by means ofhighly sensitive techniques such as pulsed-field gel elec-trophoresis, a similar variability in the Lp(a) levels withineach DNA genotype has been observed.1617

Although DNA genotyping appears to be the "goldstandard," it is expensive. Recently it has been shownthat the modified Western blot (an immunoblottingtechnique) was highly sensitive and identified 34 differ-ent apo(a) size isoforms.21'22 In the homozygous formonly one band can be seen, and in the heterozygousforms two bands of different sizes have been identified.In the present study we used the Western blot techniqueof Kamboh et al22 to identify apo(a) size variationsamong black and white children aged 7 to 11 years andevaluated the relation of apo(a) isoforms to Lp(a) levelsand family history of premature CAD.

MethodsSubjects

The subjects were 46 children aged 7 to 11 years who wererandomly selected from among a sample participating in a

by guest on March 22, 2018

http://atvb.ahajournals.org/D

ownloaded from

http://atvb.ahajournals.org/

1610 Arteriosclerosis and Thrombosis Vol 14, No 10 October 1994

longitudinal study of the biobehavioral antecedents of CAD.23

They were divided according to reported family history ofpremature CAD. Ethnicity was determined by self-designationby parents. There were 23 children with positive family historyof premature CAD (13 whites, 10 blacks) and 23 with negativefamily history of premature CAD (16 whites, 7 blacks).Positive family history of premature CAD was defined asoccurrence of MI and/or CAD (angioplasty, coronary bypasssurgery, angina on medication) among any parent, grandpar-ent, uncle, or aunt before age 55 and was verified by contactingthe appropriate physicians. Twenty-three children were desig-nated as having positive family history of premature CAD onthe basis of history of at least one of the grandparents havinghad CAD before age 55; the remaining two children had atleast one uncle or aunt who suffered from CAD before age 55.No parents reported a history of premature CAD. All childrenand their parents signed informed consent forms in accor-dance with procedures of the Medical College of Georgia(MCG) Human Assurance Committee. In addition to demo-graphic information such as age, sex, race, and family history ofpremature CAD, all children were assessed for anthropometricfactors that may be related to lipids and lipoproteins. Restingsystolic and diastolic blood pressures were also measured.

Measurement of Lipids, Lipoproteins,and Apolipoproteins

Children came to the MCG phlebotomy laboratory after atleast 12 hours of fasting. Venous blood was collected by apediatric phlebotomist by venipuncture into Vacutainer tubescontaining EDTA. Except for apo(a) isoforms, all analyseswere performed in the lipid analytic laboratory, which hasbeen accredited by the College of American Pathologists andparticipates in the Centers for Disease Control and Preventionlipid standardization program. Total plasma cholesterol (TC)and triglycerides (TG) were measured with Olympus demandautoanaryzer methodology.24-25 Both assays were enzymaticend points. The coefficients of variation for TC and TG were1.23 and 2.75, respectively. High-density lipoprotein choles-terol (HDL-C) was measured on plasma supernatant afterprecipitation of very-low-density lipoprotein (VLDL) andLDL by dextran sulfate magnesium.26 VLDL cholesterol(VLDL-C) was isolated from plasma by ultracentrifugation(40 000 rpm, 18 hours, 15°C) and was measured directly asd<1.006 g/mL. LDL cholesterol (LDL-C) was determined byassaying the bottom fraction (d> 1.006 g/mL).27 We deter-mined apoA-1 and apoB-100 (apoB) from plasma using nephe-lometry (Behring Diagnostic). The reproducibility coefficientsfor apoA-1 and apoB with a subsample (n=19) were .83 and.92, respectively.

Measurement of Lp(a)Lp(a) was determined by an enzyme-linked immunosorbent

assay (Macro Lp; Terumo Corporation Diagnostic Division) inthe Clinical Nutrition Laboratory, MCG. The enzyme-linkedimmunosorbent assay was performed with the kit availablefrom Terumo Medical Corporation. This kit included stan-dards (six levels) containing Lp(a) in human plasma in buff-ered solution, which were used to plot a standard curve fromwhich controls and unknowns were calculated. An externalsource of Lp(a) controls (two levels) assayed by NorthwestLipid Research Center in Seattle, Wash, was also included inthe controls. All Lp(a) analyses were done in batches within 1month, and all standards, controls, and plasma samples wererun in duplicate. The reproducibility coefficient with a sub-sample (n = 19) was .93.

Measurement of Apo(a) IsoformsFrozen plasma samples were sent to the Department of

Human Genetics at the School of Public Health, University ofPittsburgh, by overnight mail in rigid polystyrene plastic boxescooled with dry ice. The apo(a) isoforms were measured by a

high-resolution sodium dodecyl sulfate-agarose gel electro-phoresis method, which resolved at least 30 apo(a) iso-forms.22-28 We mixed 10 to 15 nL of plasma with 30 fiL ofreducing buffer (1:2:10 ratio of 0-mercaptoethanol, 0.5%bromophenol blue in 5% glycerol, and 5% SDS), and themixture was heated for 5 minutes at 100°C. Electrophoresiswas performed on 1.5% agarose submarine gels (90 mmol/Ltris(hydroxymethyl)aminomethane (Tris), 90 mmol/L boricacid, 2 mmol/L EDTA, 0.1% SDS) in the Hoefer submarinegel unit with LKB power supply. Electrophoresis was carriedout in a tank buffer containing 45 mmol/L Tris, 45 mmol/Lboric acid, 2 mmol/L EDTA, and 0.1% SDS for 7 to 8 hours ata constant 25 W at 4°C. Each gel contained a mixture of fiveknown apo(a) isoforms as internal controls. After electropho-resis, proteins were transferred to a 0.45-/un nitrocellulosemembrane by electroblotting overnight by means of a HoeferTransphor cell at 90 V in 10 mmol/L Tris, 40 mmol/L glycine,and 5% methanol. After protein transfer, the membrane wasincubated with 5% powdered skim milk for 1 hour followed byincubation overnight with rabbit anti-human apo(a) antise-rum22 and finally with goat anti-rabbit IgG conjugated withalkaline phosphatase for 3 hours. Subsequently, apo(a) bandswere visualized by histochemical staining. For reliability, eachsample was run twice.

Anthropometric AssessmentBody weight and height (without shoes) were measured with

a Detecto scale and stadiometer. Waist and hip circumfer-ences were taken twice and averaged. Skin folds were mea-sured at triceps, biceps, subscapular, suprailiac, abdominal,thigh, and calf on the right side of the body with Langecalipers. Three sets of measurements were taken at each site tothe nearest millimeter. The mean of the three values was usedin all analyses.

Body fatness was measured with dual-energy x-ray absorp-tiometry (DEXA) (QDR 2000, Hologic Inc). This technique isa refinement of dual-photon absorptiometry, which has beenshown to be reliable and valid for determination of percentbody fat.29-30

Blood PressureBlood pressure was measured in a seated position with the

right arm at heart level with a Quinton 410 blood pressuremonitor after the child rested for at least 10 minutes. Theaccuracy of this instrument has been previously established.31

A total of five readings of systolic and diastolic blood pressurewere recorded with 1-minute intervals, and the last three wereaveraged to derive values for each child.

Statistical AnalysesInitial descriptive analyses for all variables were computed

by race, sex, and family history of premature CAD. Forvariables with standardized skewness and kurtosis exceedingthe range ±2, nonparametric analyses were used. The associ-ation of type of isoforms with Lp(a) levels was assessed by aKruskal-Wallis one-way analysis by rank. The association ofcategorized Lp(a) levels (<25 mg/dL versus ^25 mg/dL) andapo(a) isoforms with positive family history of premature CADwas evaluated by odds ratios, 95% confidence intervals (CIs)of odds ratios, and x2 f° r linear trend of proportions. Stratifiedanalyses were performed to evaluate the interaction or con-founding effect of race or sex in the association of categorizedLp(a) or apo(a) isoforms with family history of prematureCAD.32 Multiple linear logistic regression analyses were per-formed to evaluate the independent effect of apo(a) isoformson family history of premature CAD after controlling for race,sex, Lp(a) levels, and apoB.



ownloaded from


Islam et al Apo(a) Phenotypes in Children 1611

TABLE 1. Descriptive Characteristics of Sample According to Family History ofCoronary Artery Disease

Characteristics

Continuous variables

Age, y

Total cholesterol, mg/dL

LDL-C, mg/dL

VLDL-C, mg/dL

HDL-C, mg/dL

Trigtycerides, mg/dL

ApoA-1, mg/dL

ApoB, mg/dL

Lp(a), mg/dL

SBP

DBP

Percent body fat

Categorical variables

Ethnicity, % whites

Sex, % male

Apo(a), % small isoform

PositiveHistory

Mean

10.3

169.4

110.7

11.0

47.6

57.0

142.5

90.1

32.7

115.4

66.9

27.1

57

61

32+

Family(n=23)

(SD)

(0.6)

(26.7)

(23.7)

(3.8)

(14.6)

(22.9)

(24.1)

(18.8)

(23.9)

(8.0)

(9.1)

(10.9)

Negative FamilyHistory (n=23)

Mean

10.3

167.4

106.5

12.6

48.4

68.3

142.6

88.7

27.6*

113.4

64.9

22.0

70

43

10*

(SD)

(0.7)

(23.7)

(21.9)

(8.8)

(10.5)

(46.1)

(18.1)

(21.8)

(23.9)

(9.1)

(4.5)

(9.3)

LDL-C indicates low-density lipoprotein cholesterol; VLDL-C, very-low-density lipoprotein choles-terol; HDL-C, high-density lipoprotein cholesterol; Apo, apolipoproteln; Lp(a), lipoprotein(a); SBP,systolic blood pressure; and DBP, diastolic blood pressure. To convert mg/dL to mmol/L, multiply by0.02586.

*Based on 21 cases; tbased on 22 cases; Abased on 20 cases.

ResultsDescriptive statistics are provided in Table 1. There

were no significant differences between children withpositive family history of premature CAD and thosewith negative family history of premature CAD in lipids,lipoproteins, apolipoproteins, Lp(a), blood pressure,and body fatness. The estimates of body fatness ob-tained through skin folds and various circumferenceswere similar to DEXA-measured percent body fat.Therefore, we have only used DEXA-derived percentbody fat in the analyses.

As expected, Lp(a) values were not significantly cor-related with lipids, lipoproteins, apolipoproteins, bloodpressure, or anthropometric variables (Spearman rankcorrelations, all P>.Q5, data not shown). The apo(a)isoforms were classified according to the system ofKamboh et al,22-28 in which the apparent molecularweight ranged from 294 to 624 kD. To ensure compa-rability with other existing classification systems (eg,Utermann et al33 and Gaubatz et al34), the data werereanalyzed after reclassification according to other sys-tems. Craig et al35 reported excellent correlations be-tween 12 apo(a) isoforms determined by the SDS-polyacrylamide gel electrophoresis (PAGE) (Gaubatz)method and 30 apo(a) isoforms resolved by the SDS-agarose (Kamboh) method. Using the Kamboh classifi-cation scheme, the present study defined isoforms 1 to 8as large (approximate molecular weight of 531 to 624kD), which corresponded to 10 to 12 of Gaubatz and S5

of Utermann. Similarly, band sizes 9 to 13 of Kamboh(approximate molecular weight of 508 to 517 kD)corresponded to 7 to 8 of Gaubatz and S4 of Utermann.Band sizes 14 to 15 of Kamboh (approximate molecularweight of 458 to 464 kD) corresponded to 6 of Gaubatzand S3 of Utermann. Band size 16 of Kamboh (approx-imate molecular weight of 455 kD) corresponded to 5 ofGaubatz and S2 of Utermann. Finally, sizes 17 to 18 ofKamboh (approximate molecular weight of 391 to 414kD) corresponded to 2 to 3 of Gaubatz and S, ofUtermann.

Where we found double-banded isoforms, the lower-molecular-weight band was considered the predominantband. Gaubatz et al34 have shown that among heterozy-gote subjects, lower-molecular-weight isoforms weremore intense in 51.4%, while higher-molecular-weightisoforms were more intense among only 17.5%. Theremainder of the heterozygote subjects had bands ofequal intensity. In our 22 heterozygote children, alllower-molecular-weight bands stained more intenselythan the accompanying higher-molecular-weight bands.Nonetheless, we reanalyzed the data on heterozygotesusing the higher-molecular-band allele as the predomi-nant band. The percentages of medium isoforms in thepositive and negative family history groups were 73%and 91%, respectively. The percentages of mediumisoforms in the positive and negative family historygroups were 27% and 9%, respectively. None of thegroup differences were statistically significant. Thus,regardless of whether we used the lower- or higher-



ownloaded from



TABLE 2. Mean LJpoproteln(a) Levels by Isoform Types Among Whites and Blacks (n=41)*

Kamboh2*

Large/null

1-8, 0

Medium

9-13

14-15

Small

16

17-18

Isoform Classifications

.*» Gaubatz"

10-11

7-8

6

5

2-3

Utermann*3

s,

s4

S3

s,s,

n

6

7

6

2

4

Whites

Mean

5.9

32.4

35.4

19.0

36.0

(SD)

4.8

25.5

31.6

14.1

21.1

n

1

5

7

3

Blacks

Mean

21.8

20.3

54.4

41.8

(SD)

15.5

17.7

12.4

Total 25 16

P=.049t P=.O84

Values are milligrams per deciliter.•Isoform typing was available on 42 children, one of whom did not have lipoprotein(a) measured.t/Kruskal-Wallis one-way analysis by rank.

molecular-weight isoforms as predominant, we did notfind any significant association with family history.

We grouped the isoforms into several broad groupsfor three reasons. First, epidemiological studies18'19 re-ported a significant association of Si,S2 isofonns (lowmolecular weight, small size) with CAD in adults.Second, the isoforms that are not associated with CAD,if grouped together, allow one to perform meaningfulstatistical analysis without screening a huge population.Third, Kamboh et al28 reported that apo(a) allele fre-quencies of some apo(a) alleles were significantly dif-ferent in blacks and whites; the groupings in the presentstudy were similar to those of Kamboh et al, which wererequired for the evaluation of racial difference in typesof isoforms.

The prevalence of small apo(a) isoforms (size 16 andhigher) was higher in children with positive familyhistory of premature CAD (32%) than in children withnegative family history of premature CAD (10%) (Ta-ble 1). Simple linear regression of plasma Lp(a) levels

and original apo(a) phenotypes (before grouping)showed inverse correlations of Lp(a) levels and apo(a)isoform size with a significant slope (P=.OO16) and acorrelation coefficient of -.49. Table 2 shows that Lp(a)levels were inversely related to apo(a) isoform size inboth whites and blacks. However, marked differences inLp(a) levels were observed between isoform sizes 1 to 8versus others in whites and 1 to 13 versus others inblacks. In whites 24% had a large isoform, comparedwith only 6% of the blacks. In contrast, the medium-sized isoforms were more prevalent in blacks (75%versus 52%). For the small isoforms the black/whitedifference was smaller (19% versus 24%). Table 3 showsthe proportion of children with positive family history ofpremature CAD in each group of isoforms. There was afourfold increased risk of positive family history ofpremature CAD for the children with a small apo(a)isoform compared with those with a large apo(a) iso-form. The odds ratios for positive family history ofpremature CAD associated with size of apo(a) isofonns

TABLE 3. Association of Size of Apollpoproteln(a) Isoforms With Family History of CoronaryArtery Disease In Children (n=42)

Isoform Classification

Kamboh****

Large/null

1-8,0

Medium

9-13

14-15

Small

16

17-18

Gaubatz34

10-11

7-8

6

5

2-3

Utermann*»

s,

s«S3

s.s,

n

7

13

13

2

7

FH + ,

43

54

39

100

71

OR(MLE)

95% Cl(Exact)

1.00*

1.14t

4.19*

0.16-9.38

0.36-72.70

FH+ indicates positive family history of premature coronary artery disease; OR, odds ratio; MLE, maximumlikelihood estimation; Cl, confidence interval.

•Reference category, tcombined groups under medium isoform size, ^combined groups under smallIsoform size for trend in ORs (one-sided P=.11).



ownloaded from



TABLE 4. Stratified Analysis of Association ofCategorized LJpoproteln(a) Level (<25 mg/dL vs 225mg/dL) in Children With Family History of Premature

TABLE 5. Distribution of Single-BandedApolipoprotein(a) Isoforms by Size andLJpoproteln(a) Levels

coronary «n«ry UIBOOBO

n FH+,% OR 95% Cl

Overall

Low Lp(a) 22 40.9 1.0*

HighLp(a) 22 63.6 2.53 (0.63-10.43)

Whites

LowLp(a) 16 37.5 1.0*

HighLp(a) 11 63.6 2.92 (0.45-21.18)

Blacks

Low Lp(a) 6 50.0 1.0*

HighLp(a) 11 63.6 1.75 (0.15-21.73)

FH+ indicates positive family history of premature coronaryartery disease; OR, odds ratio; Cl, confidence interval; andLp(a), lipoprotein(a). To convert mg/dL to mmol/L, multiply by0.02586.

* Reference category.

increased as the size of apo(a) isoforms decreased. Thestratified analyses of the associations of categorizedLp(a) with family history of premature CAD revealedno interaction or confounding as a result of race or sex.The odds ratios for positive family history of prematureCAD associated with higher Lp(a) values in whites andblacks were 2.92 and 1.75, respectively. However, all theCIs for the odds ratios included the null value and were

PtNo.1

2

3

4

5

6

7

8

9

10

•)1

12

13

14

15

16

17

18

19

20

Band Size13

10

7

14

15

15

15

17

0

11

14

9

18

11

10

16

85

13

17

NomenclatureM

M

L

M

M

M

M

S

Null

M

M

M

S

M

M

sLL

M

S

Lp(a), mg/dL

17.4

27.8

4.1

48.7

3.4

33.7

72.8

27.6

0.4

67.1

22.8

10.4

12.5

19.0

10.8

29.0

12.51.1

70.2

25.3not statistically significant (Table 4). The odds ratiosand 95% CIs of odds ratios for positive family history ofpremature CAD associated with high Lp(a) levels were4.0 (0.46 to 40.34) and 1.67 (0.22 to 12.92) for girls andboys, respectively. However, Woolf s test for heteroge-neity of the odds ratios was not significant (P=A9).

Considerable variation in the Lp(a) levels was ob-served within each classified single- or double-bandedcategory. In general, large isoforms (molecular weightof 531 to 624 kD) were associated with low Lp(a) levels,while medium-sized (molecular weight of 458 to 517kD) and small (molecular weight of 391 to 455 kD)isoforms were associated with high Lp(a) levels (Tables5 and 6). However, none of the heterozygous children(who had double bands) had two small isoform bands,whereas three of the heterozygous children had twolarge isoform bands (patients 1, 5, and 12), and twochildren had two medium-sized isoform bands (patients7 and 16). This indicates that small isoforms are rare,and the chance that a heterozygote child will inheritsmall isoforms from both parents is very small. Al-though in the absence of family data the heterozygositycan be inferred from the presence of two distinct bandsin the children, the inference of homozygosity is lesstenable on the basis of observing one band only. Theobservation of a null allele makes it more complicated.In this sample one child had a null allele (ie, no visiblebands), which was associated with a very low level ofLp(a).

Table 7 shows that the size of apo(a) isofonns wassignificantly related to family history of premature CADamong the homozygotes (children with one band only),whereas no such association was seen in the heterozy-

Pt indicates patient; L, large (1-8); M, medium (9-15); S, small(2:16); and Null, no bands detected.

gotes. All four homozygotes with small apo(a) isoformshad a positive family history of premature CAD. TheLp(a) values did not correlate well with the isoformtypes in the homozygotes or in the heterozygotes,particularly in the children with positive family historyof premature CAD.

Table 8 shows linear logistic regression modeling ofthe risk of positive family history of premature CAD.The best model that predicted positive family history ofpremature CAD contained race and isoform. The re-gression coefficients indicate that after controlling forrace, children with small isoforms had a greater thanfourfold increased risk of having positive family historyof premature CAD. Sex, Lp(a) levels, and apoB levelsdid not add significantly to the model.

DiscussionIt has become increasingly clear that genetic suscep-

tibility plays a role in the pathogenesis of coronaryatherosclerosis, particularly when disease occurs at arelatively young age. One way of assessing the associa-tions of putative genetic markers in children with CADoccurring in the distant future is to use family history ofpremature CAD as a surrogate measure of futureCAD.10 This study evaluated the relation of apo(a)isoforms and Lp(a) levels with positive family history ofpremature CAD in black and white children. As ex-pected, Lp(a) levels were significantly higher in blacksthan in whites. Both higher levels of Lp(a) and smallapo(a) isoforms were associated with positive family



ownloaded from



TABLE 6. Size of Heterozygous Apollpoproteln(a)Isoforms and Upoproteln(a) Levels In Children

Pt No. Band 1 Band 2 Nomenclature Lp(a), mg/dL

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

5

7

8

14

6

4

9

5

8

6

7

6

6

11

7

10

7

6

5

6

5

Pt indicates patient;medium (9-15); and S,

8

17

15

18

8

18

14

14

13

16

14

8

15

18

10

14

15

11

11

15

12

Lp(a)small

L/L

L/S

L/M

M/S

L/L

L/S

M/M

L/M

L/M

L/S

L/M

L/L

L/M

M/S

L/M

M/M

L/M

L/M

L/M

L/M

L/M

, lipoprotein(a);(216).

21.8

47.6

58.3

46.6

8.2

59.5

43.2

71.3

16.4

9.0

21.8

9.0

74.4

50.3

43.2

63.7

6.1

12.7

3.9

73.0

29.4

L, large (1-8); M,

history of premature CAD. However, the relation ofsmall apo(a) isoforms with positive family history ofpremature CAD was far stronger than the relation ofLp(a) levels with positive family history of prematureCAD. Multiple linear logistic regression analyses re-

TABLE 8. Associations of Small Apolipoproteln(a)Isoforms With Family History of Premature CoronaryArtery Disease

Coefficient* OR 95% Cl

Saturated model

Race (1 =W, 2=B)

Sex(1=M, 2=F)

Isoform (2=small)

Lp(a) CAT (2=high)

ApoB

Zygosity (2=heterozygous)

Fitted model

Race(W, B)

Isoformt

.75

- .70

1.18

.28

.003

-.30

-.78

1.53

2.10

0.50

3.20

1.30

1.00

0.74

2.2

4.59

0.49-9.22

0.13-1.96

0.50-21.17

0.32-5.41

0.96-1.04

0.19-2.93

0.58-8.24

0.80-26.36

OR indicates odds ratio; Cl, confidence interval; W, white; B,black; Lp(a), lipoprotein(a); and Apo, apolipoproteln.

•Logistic regression coefficient.tUkelihood statistics comparing the model containing isoform

with that without it were marginally significant at P=.O65.

vealed that the associations of small apo(a) isoformswith positive family history of premature CAD wereindependent of race. In addition, the association ofsmall apo(a) isoforms with family history of prematureCAD was more clear in homozygotes (with one band)than in heterozygotes (one small band and anotherlarge or medium-sized band).

Although the associations of apo(a) isoforms withclinical CAD have been documented previously inadults,1819 it was thought to add little information tothat received from Lp(a) levels. However, an associationof apo(a) isoforms in children with positive familyhistory of premature CAD has not been documentedpreviously. Given that Lp(a) levels have autosomaldominant inheritance,33-36 it is not surprising that smallisoforms are associated with high Lp(a) levels. Although

TABLE 7. Association of Homozygoslty of Small Apollpoproteln(a) Isoforms With Family History of PrematureCoronary Artery Disease

Large/null

(1-8, 0)

Medium

9-13

14-15

Small

16

17-18

n

1

5

1

1

3

FH +

Lp(a),

Mean

4.1

37.3

72.8

20.0

21.8

Homozygotes

mg/dL

(SD)

(0)

(29.5)

(0)

(0)

(8.1)

P=.

n

3

2

4

0

0

037*

F H -

Lp(a),

Mean

4.7

18.2

27.1

mg/dL

(SD)

(6.8)

(1.1)

(19.1)

n

2

2

4

1

2

FH +

Lp(a), i

Mean

15.0

16.6

46.5

9.0

48.9

Heterozygotes

mg/dL

(SD)

(9.6)

(18.0)

(29.7)

(0)

(1.9)

n

1

3

4

0

2

P=0.973*

F H -

Lp(a),

Mean

9.0

24.1

56.4

53.0

mg/dL

(SD)

(0)

(16.6)

(24.1)

(9.1)

FH+ Indicates positive family history of premature coronary artery disease; F H - , negative family history of premature coronary arterydisease; and Lp(a), lipoprotein(a).

*By x2 test for linear trend of proportion for broad groups.



ownloaded from



based on a small sample size, our data suggest aninfluence of small apo(a) isoforms on family history ofpremature CAD independent of race and Lp(a) levels.

To our knowledge, this is the first study to show thatthe association of small apo(a) isoforms with positivefamily history of premature CAD was significant for thehomozygous small isoforms but was not significant forthe heterozygous small isoforms containing one smalland one medium-sized or large isoform.

Several issues need to be taken into account wheninterpreting the results of this study. All of the previousreports associating apo(a) isoforms with clinical CADused SDS-PAGE and immunoblotting, which isolated atleast six different common isoforms.33'37 However, thistechnique was not sensitive enough to detect low levelsof apo(a) protein and, as a result, the frequency distri-bution of apo(a) isoforms failed to fit the expectation ofthe Hardy-Weinberg equilibrium.38 The inconsistenciesin the prevalence of large isoforms in blacks and whitesdisappeared with further improvement of the apo(a)phenotyping.39-40 Kamboh et al22-28 reported a high-resolution SDS-agarose gel electrophoresis methodthat resolved at least 30 apo(a) isoforms in whites andblacks. Our classification of large, medium, and smallisoforms was similar to that of Kamboh et al,28 whoshowed significant black/white differences in the fre-quency distributions of large and medium-sized apo(a)isoforms. We also found a similar trend that indicated ahigher prevalence of medium-sized apo(a) isoforms inblacks than in whites and a higher prevalence of largeapo(a) isoforms in whites than in blacks. In this study thepresence of double-banded apo(a) isoforms (heterozy-gotes) was lower than in the previously reported studiesin adults.4142 However, these studies either did notinclude blacks or did not include subjects with clinicalCAD or positive family history of premature CAD.

In the present study blacks had a higher levels of Lp(a)than whites. However, the inverse association of Lp(a)levels with size of apo(a) isoforms was significant only forwhites. This observation further supports the ethnicdifferences in the distribution of apo(a) and associatedLp(a) levels. In the absence of family data, the evaluationof the effects of homozygosity and heterozygosity of theapo(a) phenotypes on levels of Lp(a) are less tenable.However, our results were congruent with previouslypublished reports that showed higher Lp(a) levels inheterozygotes than in homozygotes.344142 There weresignificant fluctuations of Lp(a) values within each iso-form type, and the variation in Lp(a) levels withinheterozygotes did not indicate codominant effects of thealleles, as previously suggested.33

Our data also showed that size of apo(a) isoforms wasmore important than Lp(a) levels in determining risk ofpositive family history of premature CAD. They furthershowed that the homozygous small apo(a) isoforms wereassociated with positive family history of prematureCAD. The proportions of single- and double-band iso-forms in our study were 48% and 50%, respectively, andthe remainder (2%) was null alleles. However, the pro-portions of single- and double-banded isoforms weresimilar in children with positive family history of prema-ture CAD and negative family history of prematureCAD.

In conclusion, size of apo(a) isoforms, particularlyhomozygous small (^16) isoforms, was a major deter-

minant of positive family history of premature CADindependent of race. The apo(a) genetic polymorphismclearly influences Lp(a) levels in both blacks and whites.However, the distributions of types of isoforms aredifferent. Further studies should be undertaken with alarge sample of parents with premature CAD and theirchildren to evaluate the influence of homozygous smallapo(a) isoforms in the pathogenesis of premature CAD.

AcknowledgmentsThis study was supported by grants from the National Heart,

Lung, and Blood Institute (HL-35073, HL-41781, HL-49074),the Medical College of Georgia Research Institute, and theAmerican Heart Association, Georgia Affiliate. We would liketo thank Cara Svitko for excellent technical assistance inapo(a) phenotyping and Angela Sheppard for help in prepa-ration of the manuscript.

References1. Durrington P, Ishola M, Hunt L, Arrol S, Bhatnagar D. Apolipo-

proteins (a), Al, and B and parental history in men with early onsetischemic heart disease. Lancet. 1988;l:1070-1073.

2. Sandkamp M, Funke H, Schult H, Kohler E, Assraann G. Lipopro-tein(a) as independent risk factor for myocardial infarction at ayoung age. Clin Chem. 1990^6:20-23.

3. Armstrong V, Cremer P, Eberie E, et al. The association betweenserum Lp(a) concentration and angiographically assessed coronaryarteriosclerosis: dependence on serum LDL level. Atherosclerosis.1986;62:249-259.

4. Genest J, Jenner J, McNamara J, Ordovas J, Silberman S, WilsonP, Schaefer E. Prevalence of lipoprotein (a) excess in coronaryartery disease. Am J CardioL 1991;67:1039-1045.

5. Kostner G, Avogaro P, Cazzolato G, Marth E, Bittolo-Bon G,Qunici G. Lipoprotein Lp(a) and nsk of myocardial infarction.Atherosclerosis. 1981;38:51-61.

6. Rhoads G, Dahlen G, Berg K, Morton N, Dannenbert A. Lp{a)lipoprotein as a risk factor for myocardial infarction. JAMA. 1986;256:2540-2544.

7. Hoff HF, Beck GJ, Skibinski CI, Jurgens G, O'Neil J, Kramer J,Lytle B. Serum Lp(a) level as a predictor of vein graft stenosis aftercoronary artery bypass surgery in patients. Circulation. 1988;77:1238-1244.

8. Schreiner PJ, Morrisett JD, Sharrett AR, Patsch W, Tyroler HA,Wu K, Heiss G. Lipoprotein[a] as a risk factor for preclinicalatherosclerosis. Arterioscler Thromb. 1993;13:826-833.

9. Hoefler G, Harnoncourt D, Paschee, Mirtl W, Pfeiffer D, KostnerG. Lipoprotein(a), a risk factor for myocardial infarction. Arterio-sclerosis. 1988;8:398-4Ol.

10. Srinivasan S, Dahlen G, Jarpa R, Webber L, Berenson G. Racial(black-white) differences in serum lipoprotein(a) distribution andits relation to parental myocardial infarction in children.Circulation. 1991;84:160-167.

11. Sorrentino M, Vielhauer C, Eisenbart J, Fleis G, Scanu A,Feldman T. Plasma lipoprotein(a), protein concentration andcoronary artery disease in black patients compare with whitepatients. Am J Med. 1992^3:658-662.

12. Scanu AM, Fless GM. Lipoprotein(a): heterogeneity and bio-logical relevance. J Clin Invest. 1990;85:1709-1715.

13. Eaton DL, Fless GM, Kohr WJ, McLean JE, Xu Q-T, Miller CG,Lawn RM, Scanu AM. Partial amino acid sequence of apolipo-protein (a) shows that it is homologous to plasminogen. Proc NatlAcadSci USA. 1987;84:3224-3228.

14. Loscalzo J, Weinfeld M, Fless GM, Scanu AM. Lipoprotein (a),fibrin binding and plasminogen activation. Arteriosclerosis. 1990; 10:240-245.

15. Harpel PC, Gordon BR, Parker TS. Plasmin catalyzes binding oflipoprotein (a) to immobilized fibrinogen and fibrin. Proc NatlAcadSci USA. 1989;86:3847-3851.

16. Boerwinkle E, Leffert CC, Un J, Lackner C, Chliss G, Hobbs HH.Apolipoprotein (a) gene accounts for greater than 90% of thevariation in plasma lipoprotein (a) concentrations. J Clin Invest.1992;90:52-60.

17. Kraft HG, Kochl G, Menzel HJ, Sandholzer C, Uttermann G. Theapolipoprotein (a) gene: a transcribed hypervariable locus con-trolling plasma lipoprotein (a) concentration. Hum Genet. 1992;90:220-230.



ownloaded from



18. Sandholzer C, Saha N, Kark JD, Rees A, Jaross W, Dieplinger H,Hoppichler F, Boerwinkle E, Utermann G. Apo(a) isoformspredict risk for coronary heart disease: a study of six populations.Arterioscler Thromb. 1992;12:1214-1226.

19. Kark J, Sandholzer C, Friedlander Y, Utermann G. Plasma Lp(a),apolipoprotein (a) isoforms and acute myocardial infarction inmen and women: a case-control study in the Jerusalem population.Atherosclerosis. 1993;98:139-151.

20. Austin M, Sandholzer C, Selby J, Newman B, Krauss R, UtermannG. Lipoprotein (a) in women twins: heritability and relationship toapo(a) phenotypes. Am J Hum Genet. 1992^1:829-840.

21. Marcovina S, Zhang Z, Gaur V, Albers J. Identification for 34apolipoprotein(a) isoforms: differential expression of apolipopro-tein(a) alleles between American blacks and whites. BiochemBiophys Res Common. 1993;191:1192-1196.

22. Kamboh M, Ferrell R, Kottke B. Expressed hypervariable poly-morphism of apolipoprotein(a). Am J Hum Genet. 1991 ;49:1063-1084.

23. Treiber F, Musante L, Strong W, Levy M. Racial differences inyoung children's blood pressure responses to dynamic exercise. AmJDis Child. 1989;143:720-723.

24. Allain CC, Poon L, Chan C, Richmond W, Fu PC. Enzymaticdetermination of total serum cholesterol. Cbi Chem. 1974;20:470.Abstract.

25. Fredrickson DS, Goldstein JL, Brown MS. The familial hyperli-poproteinemias. In: Stansbury JB, Wyngaarden JB, FredricksonDS, eds. The Metabolic Basis of Inherited Disease. 4th ed. NewYork, NY: McGraw-Hill International Book Co; 1978:604-655.

26. Warnick G, Benederson J, Albers J. Quantitation of high-densitylipoprotein subclasses after separation by dextran sulfate and Mg+2

precipitation. Clin Chem. 1982;28:1574. Abstract.27. Lindgren FT, Jensen LC, Hatch FT. Blood lipids and lipoproteins:

quantitation, composition and metabolism. In: Nicholson GJ, ed.Blood Lipids and Lipoproteins. New York, NY: Wiley Interscience;1972:275-326.

28. Kamboh MI, Svitko CM, Williams ER, Ferrell RE, Pollitzer WS.Hypervariable polymorphism of apo(a) in blacks and whites asreflected by phenotyping. Chem Phys Ltpids. 1994;67/68:283-292.

29. Heymsfield S, Lichtman S, Baumgartner R, Wang J, Kamen Y,Aliprantis A, Pierson R. Body composition of humans: comparisonof two improved four-compartment models that differ in expense,technical complexity and radiation exposure. Am J Clin Nutr. 1990;52:52-58.

30. Pierson R, Wang J, Heymsfield S, Russell-Aulet, Manolo M,Tiemey M, Smith R, Thornton JC, Kehayias J, Weber DA, Dil-

manian FA. Measuring body fat: calibrating the rulers: inter-method comparisons in 389 normal Caucasian subjects. Am JPhysioL \99\\2f>\(Endocrinol Metab 24):E103-E108.

31. Treiber F, McCaffrey F, Strong W, Davis H, Baranowski T.Automated exercise blood pressure measurements in children: apreliminary study. Pediatr Exerc ScL 1991;3:290-299.

32. Morton NE. Outline of Genetic Epidemiology. Basel, Switzerland: SKarger, AG; 1982:94.

33. Utermann G, Menzel H, Kraft H, Duba H, Kemmler H, Seits C.Lp(a) glycoprotein phenotype: inheritance and relation to Lp(a)-lipoprotein concentration in plasma. J Clin Invest. 1987;80:458-465.

34. Gaubetz JW, Ghanem KI, Guevara J, Nava ML, Patsch W, Mor-nsett JD. Polymorphic forms of human apolipoprotein(a): inher-itance and relationship of their molecular weights to plasma levelsof lipoprotein(a). J Lipid Res. 1990;31:603-613.

35. Craig WY, Poulin SE, Ledue TB, Kamboh MI. Apolipoprotein (a):a comparison of isoforms identified by sodium dodecyl sulfate-poryacrylamide gel electrophoresis or by sodium dodecyl sulfate-agarose gel electrophoresis. Electrophoresis. 1993;14:1038-1041.

36. Sing CF, Schultz JS, Shreffler DC. The genetics of the Lp antigen,II: a family study and proposed models of genetic control. AnnHum Genet. 1974^8:47-56.

37. Kraft HG, Sandholzer C, Menzel HJ, Utermann G. Apolipopro-tein(a) alleles determine lipoprotein(a) particle density and con-centration in plasma. Arterioscler Thromb. 1992; 12:302-306.

38. Boerwinkle E, Menzel HG, Kraft HG, Utermann G. Genetics ofthe quantitative Lp(a) lipoprotein trait, III: contribution of Lp(a)glycoprotein phenotypes to normal lipid variation. Hum Genet.1989;2:73-78.

39. Sandholzer C, Hallman DM, Saha N, Sigurdsson G, Lackner C,Csaszar A, Boerwinkle E, Utermann G. Effects of the apolipopro-tein(a) size polymorphism on the lipoprotein(a) concentration in 7ethnic groups. Hum Genet. 1991 ;86:607-614.

40. Helmhold M, Bigge J, Muche R, et al. Contribution of the apo{a)phenotype to plasma Lp(a) concentrations shows considerableethnic variation. J Lipid Res. 1991^2:1919-1928.

41. Abe A, Noma A, Lee YJ, Hamaguchi H. Studies on apolipopro-tein(a) phenotypes, part 2: phenotype frequencies and Lp(a) con-centrations in different phenotypes in patients with angiograph-ically defined coronary artery disease. Atherosclerosis. 1992;96:9-15.

42. Pedro-Botet J, Senti M, Auguet T, Nogues X, Rubies-Prat J, AuboC, Vidal-Barraquer F. Apolipoprotein(a) genetic polymorphismand serum lipoprotein(a) concentration in patients with peripheralvascular disease. Atherosclerosis. 1993;104:87-94.



ownloaded from


S Islam, B Gutin, C Smith, F Treiber and M I Kambohcoronary artery disease.

Association of apolipoprotein(a) phenotypes in children with family history of premature

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1994 American Heart Association, Inc. All rights reserved.

Greenville Avenue, Dallas, TX 75231is published by the American Heart Association, 7272Arteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.14.10.16091994;14:1609-1616Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/14/10/1609World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at: Reprints:

document. Question and AnswerPermissions and Rightspage under Services. Further information about this process is available in the

which permission is being requested is located, click Request Permissions in the middle column of the WebCopyright Clearance Center, not the Editorial Office. Once the online version of the published article for

can be obtained via RightsLink, a service of theArteriosclerosis, Thrombosis, and Vascular Biologyin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:



ownloaded from

http://atvb.ahajournals.org/content/14/10/1609

http://www.ahajournals.org/site/rights/

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://atvb.ahajournals.org//subscriptions/


Documents

Association of Apolipoprotein(a) Phenotypes in Children With Family