1496

LDL Subclass Phenotypes and TriglycerideMetabolism in Non-Insulin-Dependent Diabetes

Kenneth R. Feingold, Carl Grunfeld, Miyin Pang, William Doerrler, and Ronald M. Krauss

Plasma low density lipoprotein (LDL) comprises multiple discrete subclasses differing in size, density,and chemical composition. A common, heritable phenotype characterized by the predominance of small,dense LDL particles (LDL subclass phenotype B) is associated with relatively increased concentrations ofplasma triglycerides, reduced levels of high density lipoprotein, and increased risk of coronary arterydisease in comparison with subjects with larger LDL (LDL subclass phenotype A). Population studieshave indicated that approximately 20-30% of adult men have phenotype B, and another 15-20% haveLDL of intermediate size. The lipid changes in phenotype B are similar to those that have been observedin patients with non-insulin-dependent diabetes mellitus (NIDDM). In the present study, we haveassessed LDL subclass phenotypes in normolipidemic men with NIDDM and in age-matched controlsubjects who had similar lipid levels. There was a greater than twofold increase in the percentage ofindividuals with the LDL B phenotype in the NIDDM subjects. The LDL B phenotype was associated withhigher plasma triglyceride levels and a trend toward lower high density lipoprotein cholesterol levelscompared with the LDL A phenotype in the NIDDM subjects, as has been previously observed in controlgroups. Indices of diabetic control, such as fasting and hemoglobin A, levels, were similar regardless ofLDL phenotype pattern, suggesting that glycemic control was not likely to account for the increase in theLDL B phenotype. In both control and NIDDM subjects, the clearance of trigryceride-rich lipoproteinswas slowed in the subjects with the LDL phenotype B compared with those with the A phenotype. Multipleregression analysis demonstrated that the diagnosis of NIDDM is an independent predictor of the LDLB phenotype. In summary, the present study demonstrates that NIDDM is associated with an increasedprevalence of the LDL subclass phenotype B, even in the absence of frank hyperlipldemia. Thus, geneticand metabolic factors leading to the predominance of small, dense LDL may contribute to the increasedrisk of vascular disease in patients with NIDDM. (Arteriosclerosis and Thrombosis 1992;12:1496-1502)

KEY WORDS • non-insulin-dependent diabetes mellitus • low density lipoprotein subclasses •triglyceride metabolism

Plasma low density lipoprotein (LDL) comprisesmultiple discrete subclasses differing in size,density, and chemical composition.1-6 Initial

studies demonstrated that approximately 65-75% of thegeneral population has a predominance of large LDLparticles, designated subclass pattern A; 20-30% of thepopulation has a predominance of small, dense LDLparticles, designated subclass pattern B; and a variableportion (15-20%) of the population has an intermedi-ate pattern.7'8 Complex segregation analysis has indi-

From the Department of Medicine (K.R.F., C.G., M.P., W.D.),University of California, San Francisco, and the MetabolismSection, Medical Service, Department of Veterans Affairs MedicalCenter, San Francisco; and the Donner Laboratory (R.M.K.),University of California, and the Molecular Medicine ResearchProgram, Research Medicine and Radiation Biophysics Division,Lawrence Berkeley Laboratory, Berkeley, Calif.

Supported by grants from the Bernard Bird Memorial Fund, theJewish Community Endowment Fund, Department of VeteransAffairs, and the National Institutes of Health (grants DK-40990and HL-18754). The work was conducted in part at the LawrenceBerkeley Laboratory through the US Department of Energy undercontract No. DE-AC03-76SF00098 to the University of California.

Address for correspondence: Kenneth R. Feingold, MD, Me-tabolism Section (11 IF), Department of Veterans Affairs MedicalCenter, 4150 Clement Street, San Francisco, CA 94121.

Received June 18, 1992; revision accepted September 4, 1992.

cated heritability of these phenotypic patterns, withevidence for a single major gene and an autosomaldominant pattern of inheritance of the LDL B pheno-type.78 However, the penetrance of the LDL B pheno-type is reduced in young males (<20 years of age) andpremenopausal females.7'8 Recently, the LDL subclassB phenotypes have been genetically linked to a locus onchromosome 19, near that of the LDL receptor genelocus.9

Increased plasma levels of triglycerides, very lowdensity lipoprotein (VLDL), and intermediate densitylipoprotein are associated with the LDL subclass Bphenotype.6-810 High density lipoprotein (HDL) cho-lesterol levels are reduced because of a decrease inHDL2 with no change in HDL3 levels.6-810 In parallelwith these changes, apoprotein B levels are increasedand apoprotein A-I levels are decreased in individualswith the LDL B phenotype.7 The lipid and lipoproteinvariations found in the LDL B phenotype are associatedwith an approximately threefold increased risk of myo-cardial infarction."

Individuals with non-insulin-dependent diabetes melli-tus (NIDDM) have an increased incidence of atheroscle-rotic vascular disease that is not entirely accounted for bythe usual risk factors, such as elevated serum cholesterolconcentrations, decreased HDL cholesterol levels, hyper-

by guest on July 14, 2018http://atvb.ahajournals.org/

Dow

nloaded from

Feingold et al LDL Phenotypes in NIDDM 1497

TABLE 1.

Control

Diabetic

Demographic Data and

(«=87)

(n=29)

Age(years)

60±0.86

60±2.3

Plasma Lipld

BMI(kg/m2)

25+0.33

28±0.82*

Levels in Control and

Triglyceride

123±6.4

123±12.1

NIDDM Subjects

Cholesterol

207±3.0

190+5.5*

HDLcholesterol

46+1.3

42±2.0

LDLcholesterol

136±2.7

124±4.6t

NIDDM, non-insulin-dependent diabetes mellitus; BMI, body mass index; HDL, high density lipoprotein; LDL, lowdensity lipoprotein. All triglyceride and cholesterol measurements are in milligrams per deciliter.

•p<0.01 vs. control; tp<0.05 vs. control.

tension, cigarette smoking, etc.12-13 Furthermore, patientswith NIDDM commonly have a type of dyslipidemiacharacterized by elevations of plasma trigtycerides andreductions in HDL cholesterol similar to those seen inconjunction with the LDL subclass phenotype B.14'15

Therefore, in the present study, we determined LDLsubclass phenotypes in subjects with NIDDM but withoutsevere dyslipoproteinemia.

MethodsPatients

All subjects in the study were male. The subjects withdiabetes (n=29) had a history of two fasting serumglucose values greater than 140 mg/dl and were beingtreated with either oral hypoglycemic agents or insulin.None of the diabetic subjects had a history of ketoaci-dosis, and they were classified as NIDDM on the basisof clinical criteria. NIDDM subjects with fasting plasmatriglyceride levels greater than 325 mg/dl were excludedfrom this study. Control subjects (n=87) did not havediabetes, kidney failure, nephrotic syndrome, activeliver disease, or other disorders that would affect lipidmetabolism. None of the control subjects or diabeticswere receiving hypolipidemic therapy or were takingdrugs that might affect lipid metabolism. The controlsubjects were age matched to those with NIDDM.

Analytical MethodsPlasma triglyceride, cholesterol, HDL cholesterol,

glucose, and hemoglobin A, levels were determined bystandard laboratory procedures. LDL cholesterol con-tent was calculated by the method of Friedewald et al.16

Nondenaturing gradient gel electrophoresis of wholeplasma was performed on 2-16% polyacrylamide gra-dient gels and analyzed by using a lipid stain (oil red O)followed by scanning with a Transdyne RFT scanningdensitometer.7810 Particle diameters were calculatedfrom calibration curves by using protein standards ofknown size. Each subject's scan was classified into oneof three patterns (type A, type B, or intermediate) byone author (R.K.) using criteria previously described8"11

without knowledge of the clinical history or classifica-tion of the patient.

Triglyceride ClearanceAll of the NIDDM subjects and an equal number of

control subjects had the clearance of triglyceride-richparticles determined by the intravenous fat-tolerancetest as described previously.17 In brief, after fastingplasma samples were drawn, the subjects were given20% Liposyn II intravenous fat emulsion (Abbott Lab-oratories, North Chicago, 111.) at a dose of 0.1 g/kg bodywt. Plasma samples were then taken every 5 minutes for30 minutes and then every 10 minutes for the next 20

minutes. Clearance was measured by nephelometry.Regression of a semilog plot of clearance was linear foreach patient, and the half-time (t^) was determined byusing SIGMA-PLOT software. The t w for clearance by thismethod is not influenced by triglyceride levels in disor-ders with VLDL overproduction (reviewed in Refer-ence 17).

StatisticsDifferences between groups were analyzed by Stu-

dent's t test. Prevalence was compared by the x2 test.Multiple regression analysis and analysis of variancewere performed with the STATVIEW II statisticalpackage.

ResultsDemographic Data and Plasma Lipid Levels

Table 1 presents the demographic data and plasmalipid levels of control and NIDDM subjects. As ex-pected, body mass index is increased in the subjects withNIDDM. In these NIDDM subjects, plasma triglycerideand HDL cholesterol levels are similar to those of thecontrol group, whereas small but statistically significantdecreases in total plasma cholesterol and LDL choles-terol levels were seen. The absence of increases inplasma lipid levels in the NIDDM compared withcontrol subjects is most probably due to the fact that wespecifically excluded from this study NIDDM subjectswith elevated triglyceride levels (plasma triglyceridelevels >325 mg/dl) and those who were being treatedwith hypolipidemic drugs.

LDL Phenotype in Control and Diabetic SubjectsAs shown in Table 2, in these NIDDM subjects there

is an increase (116%) in the percentage of individualswith the LDL B phenotype and a decrease (41%) in theLDL A phenotype compared with control subjects. Thepercentage of individuals with the intermediate pheno-type is similar in NIDDM and control subjects. Thus,diabetes, a disorder that can cause alterations in triglyc-eride metabolism, is associated with an increased prev-alence of the LDL subclass phenotype B even whenhypertriglyceridemic subjects are excluded.

TABLE 2. LDL Phenotypes in

Control (n=87)

Diabetic (n=29)

A(%)

47

28

p< 0.025

Control and NIDDM

Intermediate (%)

29

21

NS

Subjects

B(%)24

52

p<0.025

LDL, low density lipoprotein; NIDDM, non-insulin-dependentdiabetes mellitus.

by guest on July 14, 2018http://atvb.ahajournals.org/

Dow

nloaded from

1498 Arteriosclerosis and Thrombosis Vol 12, No 12 December 1992

TABLE 3. Demographic Data and Plasma Llpid Levels in Controls Subjects According to LDL Phenotype

Phenotypc

A (n=41)Intermediate (n=25)B (n=2\)A vs. intermediateAvs. BIntermediate vs. B

Age(years)

60±1.358±1.661±1.6

NS

NS

NS

BMI(kg/m2)

24+0.4526±0.7026±0.48p<0.05p<0.01

NS

Triglyceride

87±4.6127±5.3187±16.3p<0.001/><0.001p<0.01

Cholesterol

202±4.1211 ±6.4214±4.9

NS

NS

NS

HDLcholesterol

52±1.843±2.337±1.6

/><0.01p<0.001p<0.05

LDLcholesterol

132+3.9139±5.2139±5.5

NS

NS

NS

LDL, low density lipoprotein; BMI, body mass index; HDL, high density lipoprotein. All triglyceride and cholesterolmeasurements are in milligrams per deciliter.

Demographic Data and Plasma Lipid LevelsAccording to LDL Phenotype

Table 3 presents the demographic data and plasmalipid levels in control subjects divided by LDL pheno-types. As was seen in previous studies, there is anincrease in plasma triglyceride levels and a decrease inHDL cholesterol levels in control subjects with the LDLB phenotype compared with subjects with the LDL Aphenotype. Control subjects with the intermediate LDLphenotype have plasma trigJyceride and HDL choles-terol levels that are between those for the A and Bphenotypes. Thus, the presence of the LDL B pheno-type is associated with an atherogenic lipid profile.

Table 4 presents the demographic data and plasmalipid levels in NIDDM subjects within each LDL phe-notype. As observed in control subjects, the LDL Bphenotype is associated with higher plasma triglyceridelevels. Additionally, there was a trend toward a decreasein HDL cholesterol levels in the NIDDM subjects withthe LDL B phenotype, but this difference did not quiteachieve statistical significance (/7<0.1). The plasmalipid pattern in the NIDDM subjects with the interme-diate LDL phenotype is similar to that of diabeticsubjects with the LDL A phenotype. However, plasmacholesterol and LDL cholesterol levels were lower insubjects with the intermediate phenotype comparedwith the LDL B phenotype. Importantly, indices of

diabetic control, such as fasting glucose and hemoglobinAi levels, are similar regardless of LDL phenotypicpattern, and thus glycemic control is not likely toaccount for the increase in the LDL B phenotype.

Multiple Regression Analysis of Factors Contributingto the LDL B Phenotype

A multiple regression model for the effect of keyvariables on the LDL B phenotype for the entire groupis shown in Table 5. Plasma triglyceride levels are thestrongest independent predictor of the LDL B pheno-type. Reduced plasma HDL levels also predict LDL Bphenotype. Most importantly, the diagnosis of NIDDMis also an independent predictor of the LDL B pheno-type. Similar regression coefficients for the lipid vari-ables were found for the control subjects when consid-ered separately, whereas none of the coefficients wassignificant for the NIDDM subjects considered sepa-rately (data not shown).

Triglyceride Clearance and LDL PhenotypeIn the NIDDM subjects and in a subset of the control

subjects, the clearance of triglyceride-rich particles wasdetermined by the intravenous fat-tolerance test. Inthese control subjects, as shown in Figure 1A, meanplasma triglyceride levels are elevated in subjects withthe LDL B phenotype compared with the LDL A

TABLE 4. Demographic Data and Plasma Lipid Levels in NIDDM Subjects According to LDL Phenotype

Phenotype

A (n=8)Intermediate (n=6)D /»» 1 C\D [n—ij)A vs. intermediateAvs. BIntermediate vs. B

Age(years)

60±4.158±5.961±3.2

NS

NS

NS

A(n=8)Intermediate (r,

B (n = 15)

BMI(kg/m2)26±1.628±1.329±1.2

NS

NSNS

1=6)

A vs. intermediateAvs. BIntermediate vsi. B

Triglyceride

92±15.973±9.1

160+16.8NS

p<omp<0.001

Fasting glucose(mg/dl)

217±24.12O4±26.0201±21.0

NS

NS

NS

Cholesterol

190+7.0166+12.8200+7.6

NS

NSp<0.05

HDLcholesterol

48±6.0047±5.236±2.05

NSp<0.1

NS

HbA,*

13.5±1.5213.9±1.3912.5±0.98

NSNSNS

LDLcholesterol

124±6.0105 + 10.4132±6.48

NSNS

p<0.05

LDL, low density lipoprotein; BMI, body mass index; HDL, high density lipoprotein; HbA[, hemoglobin Ai. Alltriglyceride and cholesterol measurements are in milligrams per deciliter.

•Normal range, 4.3-7.7.

by guest on July 14, 2018http://atvb.ahajournals.org/

Dow

nloaded from

Feingold et al LDL Phenotypes in NIDDM 1499

TABLE 5. Multlvarlate Predictors of LDL Subclass Phenotype B 350

BMITriglycerideLDL cholesterolHDL cholesterolDiagnosis*

Regression coefficient

0.0140.0070.004

-0.018-0.399

SEM

0.0180.0010.0020.0060.148

PNS

<0.001NS

<0.01<0.01

LDL, low density lipoprotein; BMI, body mass index; HDL,high density lipoprotein.

•Presence (0) or absence (1) of non-insulin-dependent diabetesmellitus.

*~^*̂o>c

c

evel

ycerides

L

o>•cl__

300

250

200

150

100phenotype, but there is considerable overlap betweenthe two groups. As shown in Figure IB, the t1/2 oftriglyceride clearance is delayed in subjects with theLDL B phenotype compared with subjects with theLDL A phenotype (LDL B, 24.5±2.4 minutes versusLDL A, 12.3±0.6 minutes; p<0.001). Subjects with theintermediate LDL phenotype have a clearance midway

300

FIGURE 1. Bar graphs of plasma triglyceride levels (panelA) and triglyceride clearance (panel B) in control subjectsaccording to low density lipoprotein phenotype pattern.

50 -

INTERMEDIATE

c"E

oocai-oVo

50

40

30

20

10

n

B

-

-

-

-

8

8 1

A

A

A

i * .

INTERMEDIATE B

FIGURE 2. Bar graphs of plasma triglyceride levels (panelA) and triglyceride clearance (panel B) in diabetic subjectsaccording to low density lipoprotein phenotype pattern.

between these values (LDL intermediate, 17.4±1.6minutes; p<0.05 versus LDL B,/?<0.01 versus LDL A).Of note is that except for a single individual with LDLB, there is no overlap in clearance rates between theLDL B and LDL A groups, suggesting that the clear-ance rate of triglyceride-rich particles may be a strongmarker for the LDL B phenotype. After adjustment forfasting plasma triglyceride by analysis of covariance therelation of fat clearance and LDL phenotype remainedsignificant atp<0.05.

Figure 2A shows the plasma triglyceride distributionin the NIDDM subjects divided according to LDLsubclass phenotype. As expected, plasma triglyceridelevels are increased in subjects with the LDL B pheno-type, but again, there is considerable overlap betweengroups. Similar to our observations in control subjects,the clearance of triglyceride-rich particles (Figure IB) isdelayed in NIDDM subjects with the LDL B phenotypecompared with those with the LDL A phenotype (LDLB, 25.3±2.7 minutes versus LDL A, 13.8±1.8 minutes;

by guest on July 14, 2018http://atvb.ahajournals.org/

Dow

nloaded from

1500 Arteriosclerosis and Thrombosis Vol 12, No 12 December 1992

p<0.01). NIDDM subjects with the intermediate LDLphenotype have a clearance rate similar to LDL Asubjects (LDL intermediate, 12.3+0.4 minutes). In theNIDDM subjects, there is a greater overlap in clearancevalues between the LDL A and LDL B groups than thatobserved in control subjects. After adjustment forplasma triglyceride level by analysis of covariance, therelation of fat clearance to LDL phenotype in theNIDDM subjects was not significant.

DiscussionThe factors that determine the size and composition

of LDL particles are uncertain. Studies have demon-strated that the phenotypic pattern of LDL is inherit-ed.7"9 In addition, environmental factors or other dis-ease states may also influence the size and compositionof LDL. First, the penetrance of the expression of theLDL B phenotype increases with age.7'8 Second, studieshave shown that weight loss induced by exercise de-creases the concentration of small, dense LDL.18 Lastly,we have recently found that in subjects with acquiredimmunodeficiency syndrome, an infection associatedwith hypertriglyceridemia,1719-20 the prevalence of theLDL B phenotype is increased 2.5-fold (K.R. Feingold,et al, unpublished observations). Thus, both genetic andenvironmental factors affect LDL phenotypic patternexpression.

The LDL B phenotype is associated with increasedplasma triglyceride levels.7'8'10 Whether hypertriglyceri-demia itself leads to the formation of small, dense LDLor whether those who inherit the LDL B phenotype aremore susceptible to the development of hypertriglycer-idemia is unresolved. The transfer of triglyceride fromtriglyceride-rich lipoproteins to LDL in exchange forcholesterol has been postulated to alter the metabolismof LDL and thereby ultimately lead to the formation ofsmall, dense LDL.21 However, small, dense LDL mayarise from distinct precursors whose levels are influ-enced by variations in triglyceride metabolism.22

Diabetes is associated with alterations in triglyceridemetabolism. In studies including many subjects withdiabetes, plasma triglyceride levels correlate directlywith indices of glycemic control, and improvements inglycemic control by oral hypoglycemic agents or insulintherapy lead to decreases in plasma triglyceride concen-trations.23-29 Studies have shown that the clearance oftriglyceride-rich lipoproteins is decreased and the pro-duction of triglyceride-rich lipoproteins is increased indiabetes.30-32 The decrease in the clearance of triglyc-eride-rich lipoproteins may be due to decreased activityof lipoprotein lipase, an enzyme whose synthesis isdependent on insulin.32-34 The increase in hepatic tri-glyceride production is thought to be secondary in partto the increased delivery of free fatty acids to the liverin individuals who are in poor metabolic control.30'31

In the present study, we demonstrate that the fre-quency of the LDL B phenotype is increased approxi-mately twofold in normolipidemic subjects withNIDDM, whereas the frequency of the LDL A pheno-type is decreased by 41%. The percentage of diabeticsubjects with the intermediate phenotype is similar tothat of control subjects. Studies by other investigatorsusing different techniques to examine LDL size andcomposition have demonstrated that diabetic LDLshave increased heterogeneity, with a tendency to

greater quantities of small, dense LDL.35-40 To the bestof our knowledge the present study is the first to classifya group of NIDDM subjects according to LDL subclassphenotype.

There were no differences in fasting plasma glucoseand hemoglobin A, levels (indices of glycemic control)in the diabetic subjects with either the LDL A, B, orintermediate phenotype. This suggests that poor glyce-mic control is not a major factor in determining LDLphenotypic pattern. As expected, plasma triglyceridelevels are increased and HDL cholesterol levels de-creased in NIDDM subjects with the LDL B phenotypecompared with NIDDM subjects with the LDL Aphenotype. Of note is that despite the exclusion ofdiabetics with increased plasma trigryceride levels (tri-glyceride values >325 mg/dl) and those on hypolipid-emic therapy, the present study still demonstrated anincreased prevalence of the LDL B phenotype. It islikely that a study of diabetic subjects with hyperlipid-emia, especially hypertriglyceridemia, would demon-strate an even higher prevalence of the LDL Bphenotype.

Although the LDL B phenotype was related toplasma triglyceride values within both the control andNIDDM groups, plasma triglyceride levels in theseNIDDM subjects and the control subjects were identi-cal. Yet the NIDDM subjects had a higher prevalence ofthe LDL B phenotype, suggesting that rather than beingsimply related to increased plasma triglyceride levelsper se, the development of the LDL B phenotype maybe dependent on underlying changes in triglyceridemetabolism. As shown in Figures 1A and 2A, there isconsiderable overlap in plasma triglyceride levels be-tween individuals with the LDL A and B phenotypes. Incontrast, in control subjects the clearance of triglycer-ide-rich particles is delayed approximately twofold insubjects with the LDL B phenotype, and there is littleoverlap between the individuals with the LDL A and Bphenotypes. Only a single individual with the LDL Bphenotype has a clearance rate within two standarddeviations of the mean rate of clearance of the LDL Agroup (Figure IB). These results suggest that the slowedclearance rate of triglyceride-rich particles may be abetter marker for the LDL B phenotype in controls thanare plasma triglyceride levels. In NIDDM, there isgreater variability in the clearance of triglyceride-richlipoproteins, which may reflect the underlying metabolicdisorders. As a consequence there is a greater overlap inclearance rates between the LDL A and LDL B groups,although the degree of overlap of the clearance rates isstill somewhat less than that observed with plasmatriglyceride levels.

Interestingly, the diagnosis of diabetes is also anindependent determinant for the presence of the LDLB phenotype. Whether this indicates a direct relationbetween NIDDM and the LDL B phenotype or whetherthis is merely due to NIDDM causing disturbances inlipid metabolism that are not taken into account in ourmultivariate analysis is unclear. Both NIDDM and theLDL B phenotype are genetically determined, and it ispossible that they are linked. Insulin resistance is acharacteristic abnormality in NIDDM,41 and prelimi-nary studies have suggested that the prevalence of theLDL B phenotype is increased in nondiabetic subjectswho have insulin resistance.42 Thus, in NIDDM the

by guest on July 14, 2018http://atvb.ahajournals.org/

Dow

nloaded from

Feingold et al LDL Phenotypes in NIDDM 1501

situation is potentially very complex, with the increasedprevalence of the LDL B phenotype being secondary toa variety of factors, including hypertriglyceridemia, al-terations in lipid metabolism not reflected in plasmalipid levels, and perhaps to a linkage between twocommon genetic disorders.

Regardless of the cause of the increased prevalenceof the LDL B phenotype in NIDDM, these changeshave important clinical implications. The LDL B phe-notype is associated with an approximately threefoldincreased risk of myocardial infarction in nondiabeticsubjects." This risk may be due to associated lipidabnormalities, such as increased serum triglyceride lev-els and decreased HDL cholesterol concentrations.6"810

However, studies have also suggested that small, denseLDL (LDL B) per se may be more atherogenic. Small,dense LDL may bind to the arterial wall to a greaterdegree/3 Additionally, for any level of cholesterol thereare more apoprotein B-containing particles, which mayincrease the risk of vascular disease. Lastly, small, denseLDL have been shown to have increased susceptibilityto oxidation.4445

In summary, the present study demonstrates thatNIDDM is associated with an increased prevalence ofthe LDL B phenotype, even in the absence of frankhyperlipidemia. This may account for some of theincreased independent risk of vascular disease in pa-tients with NIDDM.

AcknowledgmentsWe thank P. Herranz for excellent editorial assistance. We

thank Laura Holl and Dulce Poyaoan for performing thegradient gel analysis and Bonnie Millar for assistance withstatistical analysis. The cooperation and support of our pa-tients and the staff of the Special Diagnostic and TreatmentUnit are greatly appreciated.

References1. Lindgren FT, Elliott HA, Gofman JW: The ultracentrtfugal char-

acterization and isolation of human blood lipids and lipoproteins,with applications to the study of atherosclerosis. / Phys ColloidChem 1951^5:80-93

2. Adams GH, Schumaker VN: Equilibrium banding of low-densitylipoproteins: III. Studies on normal individuals and the effects ofdiet and heparin-induced lipase. Biochim Biophys Acta 1970^10:462-472

3. Lee DM, Alaupovic P: Studies of the composition and structure ofplasma lipoproteins: Isolation, composition, and immunochemicalcharacterization of low density lipoprotein subtractions of humanplasma. Biochemistry 1970;9:2244-2252

4. AJbers JJ, Chi-Hong C, Aladjem G: Human serum lipoproteins:Evidence for three classes of lipoproteins in SF 0-2. Biochemistry1972;ll:57-63

5. Fisher WR, Hammond MG, Warmke GL: Measurements of themolecular weight variability of plasma low density lipoproteinsamong normals and subjects with hyper-beta-lipoproteinemia:Demonstration of macromolecular heterogeneity. Biochemistry1972;ll:519-529

6. Krauss RM, Burke DJ: Identification of multiple subclasses ofplasma low density lipoproteins in normal humans. J Lipid Res1982;23:97-104

7. Austin MA Krauss RM: Genetic control of low density lipopro-tein subclasses. Lancet 1986;2:592-595

8. Austin MA King MD, Vranizan KM, Newman B, Krauss RM:•Inheritance of low density lipoprotein subclass patterns: Results ofcomplex segregation analysis. Am J Hum Genet 1988;43:838-846

9. Nishina PM, Johnson JP, Nuggert JK, Krauss RM: Linkage ofatherogenic lipoprotein phenotype to the low density lipoproteinreceptor locus on the short arm of chromosome 19. Proc NatlAcadSciUSA 1992;89:708-712

10. Austin MA, King MC, Vranizan KM, Krauss RM: The athero-genic lipoprotein phenotype (ALP): A proposed genetic markerfor coronary heart disease risk. Circulation 1990;82:495-506

11. Austin MA, Breslow JL, Hennekens CH, Buring JE, Willet WC,Krauss RM: Low density lipoprotein subclass patterns and risk ofmyocardial infarction. JAMA 1988^60:1917-1921

12. Ruderman NB, Haudenschild C: Diabetes as an atherogenic fac-tor. Prog Cardiovasc Dis 1984;26J73-412

13. Feingold KR, Siperstein MD: Diabetic vascular disease, in Stoll-erman GH (ed): Advances in Internal Medicine. Chicago, YearBook Medical Publishers Inc, 1986, vol 31, pp 309-340

14. Howard BV: Lipoprotein metabolism in diabetes. J Lipid Res 1987;28:613-628

15. Kostner GM, Karadi I: Lipoprotein alterations in diabetes melli-tus. Diabetologia 1988^1:717-722

16. Friedewald WT, Levy RI, Fredrickson DS: Estimation of the con-centration of low density lipoprotein cholesterol in plasma withoutuse of the preparative ultracentrifuge. Clin Chem 1972;18:499-502

17. Grunfeld C, Pang M, Doerrler W, Shigenaga JK, Jensen P, Fein-gold KR: Lipids, lipoproteins, and trigtyceride clearance and cyto-kines in human immunodeficiency virus infection and the acquiredimmunodeficiency syndrome. J Clin Endocrinol Metab 1992;74:1045-1052

18. Williams PT, Krauss RM, Vranizan KM, Albers JJ, Terry RB,Wood PDS: Effects of exercise induced weight loss on low densitylipoprotein subtractions in healthy men. Arteriosclerosis 1989;9:623-632

19. Grunfeld C, Kotler DP, Hamadeh R, Tierney A, Wang J, PiersonRN Jr. Hypertriglyceridemia in the acquired immunodeficiencysyndrome. Am J Med 1989;86:27-31

20. Grunfeld C, Kotler DP, Shigenaga JK, Doerrler W, Tierney AWang J, Pierson RN, Feingold KR: Circulating interferon alphalevels and hypertriglyceridemia in the acquired immunodeficiencysyndrome. Am J Med 1991;90:154-162

21. Deckelbaum RJ, Granot E, Oschry Y, Rose L, Eisenberg S:Plasma triglycerides determines structure composition in low andhigh density lipoproteins. Arteriosclerosis 1984;4:225-231

22. Krauss RM: Relationship of intermediate and low density lipopro-tein subspecies to risk of coronary artery disease. Am Heart J1987;113:578-582

23. Glasgow AM, August GP, Hung W: Relationship between controland serum lipids in juvenile onset diabetes. Diabetes Care 1981;4:76-80

24. Lopes-Virella MF, Wohltmann HJ, Loadhold CB, Busc MG:Plasma lipids and lipoproteins in young insulin dependent diabeticpatients: Relationship with control. Diabetologia 1981;21:216-223

25. Sosenko JM, Breslow J, Miettinen OJ, Gabbay KH: Hyperglycemiaand plasma lipid levels: Co-variations in insulin dependent diabe-tes. Diabetes Care 1985^:40-43

26. Agardh CD, Nilsson-Ehle P, Scharsten B: Improvement of theplasma lipoprotein pattern after institution of insulin treatment indiabetes mellitus. Diabetes Care 1982^:322-325

27. Howard BV, Knowler WC, Vasquez B, Kennedy AL, Pettitt DJ,Bennett PH: Plasma and lipoprotein cholesterol and triglyceride inthe Pima Indian population. Arteriosclerosis 1984;4:462-471

28. Ostlund RE Jr, Samenkovich CF, Schectman KB: Quantitativerelationship between plasma lipids and grycohemoglobin in Type Ipatients. Diabetes Care 1989;12:332-336

29. Pfeifer MA, Brunzell JD, Best JD, Judzewitsch RG, Halter JB,Porte D Jr: The response of plasma triglyceride, cholesterol andlipoprotein lipase to treatment in non-insulin dependent diabeticsubjects without familial hypertriglyceridemia. Diabetes 1983;32:525-531

30. Reaven GM, Greenfield MS: Diabetic hypertriglyceridemia: Evi-dence for three clinical syndromes. Diabetes 1981;30(suppl2):66-75

31. Gibbons GF: Insulin, diabetes and hepatic very low density lipo-protein metabolism. Biochem Soc Trans 1989;17:49-51

32. Arbeeny CM, Eder HA: Effect of diabetes on the metabolism oftriacylgrycerol-rich lipoproteins. Biochem Soc Trans 1989;17t51-53

33. Pykalisto OJ, Smith PH, Brunzell JD: Determinants of humanadipose tissue lipoprotein lipase: Effects of diabetes and obesity onbasal and diet induced activity. J Clin Invest 1975;56:1108-1117

34. Brunzell JD, Porte D Jr, Bierman EL: Reversible abnormalities inpost heparin lipotytic activity during the late phase of release indiabetes mellitus. Metabolism 1975;24:1123-1138

35. Schonfeld G, Birge C, Miller JP, Kessler G, Santigo J: Apolipo-protein B levels and altered lipoprotein composition in diabetes.Diabetes 1979^23:827-834

by guest on July 14, 2018http://atvb.ahajournals.org/

Dow

nloaded from

1502 Arteriosclerosis and Thrombosis Vol 12, No 12 December 1992

36. Fisher W: Heterogeneity of plasma low density lipoproteins: Man-ifestations of the physiologic phenomenon in man. Metabolism1983;32:283-291

37. Suzuki N, Oikawa S, Hori S, Fujii Y, Sakuma E, Kotake H, NamaiK, Yoshie K, Goto Y: Appearance of multidisperse low densitylipoprotein and altered lipoprotein composition in non-insulindependent diabetes with type Ha hyperlipoproteinemia. Metabo-lism 1989;38:224-229

38. Iwai M, Yoshino G, Matsushita M, Morita M, Matsuba K, KazumiT, Baba S: Abnormal lipoprotein composition in normolipidemiadiabetic patients. Diabetes Care 1990;13:792-796

39. James RW, Pometta D: The distribution profiles of very low den-sity and low density lipoproteins in poorly controlled male Type II(non-insulin dependent) diabetic patients. Diabetologia 1991;34:246-252

40. Barakat HA, Carpenter JW, Malendon VD, Khazanic P, LeggettN, Heath J, Marks R: Influence of obesity, impaired glucose tol-erance and NIDDM on LDL structure and composition: Possible

link between hyperinsulinemia and atherosclerosis. Diabetes 199O,39:1527-1533

41. Moller DE, Flier JS: Insulin resistance —Mechanisms, syndromesand implications. N Engl J Med 1992;325:938-948

42. Katzel LI, Goldberg AP, Krauss RM: Altered relationships withinsulin lower HDL and raise trigrycerides in man with LDL sub-class B. (abstract) Artcrioscler Thromb 1991;11:1513a

43. Camejo G, Hurt-Camejo E, Bondjers G: Effect of proteoglycans onlipoprotein-cell interactions: Possible contribution to atherogene-sis. Curr Opin Upidol 1990;l:431-436

44. de Graaf J, HakLemmers HLM, Hectors MPC, Demacker PNM,Hendriks JCM, Stalenhoef AFH: Enhanced susceptibility to invitro oxidation of the dense low density lipoprotein subtraction inhealthy subjects. Arterioscler Thromb 1991;11:298-306

45. Tribble DL, Holl LG, Wood PD, Krauss RM: Variations in oxidativesusceptibility among six low density lipoprotein subtractions of dif-fering density and particle size. Atherosclerosis 1992^3:189-199

by guest on July 14, 2018http://atvb.ahajournals.org/

Dow

nloaded from

K R Feingold, C Grunfeld, M Pang, W Doerrler and R M KraussLDL subclass phenotypes and triglyceride metabolism in non-insulin-dependent diabetes.

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1992 American Heart Association, Inc. All rights reserved.

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

doi: 10.1161/01.ATV.12.12.14961992;12:1496-1502Arterioscler Thromb Vasc Biol. 

http://atvb.ahajournals.org/content/12/12/1496World 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.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on July 14, 2018http://atvb.ahajournals.org/

Dow

nloaded from