relationship between lipoprotein(a) phenotypes and albumin excretion rate in non-insulin-dependent...

Relationship between lipoprotein(a) phenotypes and albumin

excretion rate in non-insulin-dependent diabetes mellitus:

protective effect of ‘null’ phenotype?

C. HERNANDEZ, R. SIMO, P. CHACON,* A. SEGARRA,† D. LOPEZ* & J. MESADepartments of Endocrinology, *Biochemistry and †Nephrology, Hospital Universitari Vall d’Hebron,Barcelona, Spain

Received 15 October 1996; accepted 30 January 1997

Abstract. The possible association between lipo-protein(a) [Lp(a)] and albumin excretion rate (AER) isa topic that generates conflicting views. In addition,Lp(a) phenotypes have not previously been consideredas factors influencing AER. In order to clarify this issue,we studied 70 non-insulin-dependent diabetes mellitus(NIDDM) patients without clinically detectable macro-angiopathy, 27 with microalbuminuria and 43 without it.Both groups were matched for the known variables thatcould influence AER and serum Lp(a) levels. Lp(a) wasdetermined by enzyme-linked immunosorbent assay(ELISA), and Lp(a) phenotypes were assessed by elec-trophoresis followed by immunoblotting. Lp(a) pheno-types were grouped as follows: ‘small’ (F, S1 and S2),‘big’(S3 and S4) and ‘null’. The NIDDM patients withmicroalbuminuria presented higher serum Lp(a) concen-trations than the patients without it [15.7 mg dL–1 (95%CI 0.5–36.5) vs. 4.5 mg dL–1 (95% CI 0.1–18.5);P<0.001] and a direct correlation between Lp(a) andAER was observed (r ¼ 0.34;P<0.01). AER was signifi-cantly different when Lp(a) phenotypes were con-sidered [‘small’: median 19mg min–1 (range 1–195);‘big’: median 9.5mg min–1 (range 1–186); ‘null’:4mg min–1 (range 1–9); P¼ 0.04]. None of theNIDDM patients with a ‘null’ phenotype showed anAER of > 10mg min–1. In conclusion, this case–controlstudy provides evidence that microalbuminuria is asso-ciated with high serum Lp(a) in NIDDM without clini-cally detectable macroangiopathy. Furthermore, NIDDMpatients with a ‘null’ phenotype could be considered atlow risk for the development of microalbuminuria.

Keywords. Albumin excretion rate, diabeticnephropathy, lipoprotein(a), lipoprotein(a) phenotype,non-insulin-dependent diabetes mellitus.

Introduction

In non-diabetic subjects [1–3] and also in NIDDMpatients [4–6], microalbuminuria has been recognizedas a predictor of cardiovascular disease (CVD). Althoughthe well-established cardiovascular risk factors (hyper-tension, smoking habit, elevated plasma cholesterol andfibrinogen) are seen more frequently in patients withpersistent microalbuminuria than in normoalbuminuricdiabetic patients of similar age, sex and diabetes dura-tion, they cannot by themselves explain the highercardiovascular mortality in these patients. A number ofstudies have suggested that lipoprotein (a) [Lp(a)] is astrong independent predictor of CVD in non-diabeticsubjects [7–9]. Thus, it could be speculated that Lp(a)is involved in the excess of cardiovascular morbidityand mortality observed in diabetic patients with micro-albuminuria. However, in diabetic patients, the relation-ship between Lp(a) and CVD is less clear [10–13] andthe results of studies that attempt to find an associationbetween Lp(a) and microalbuminuria are conflicting[14–20]. It is possible that these discrepancies couldbe caused by different criteria used for the selection ofpatients, size of the population sample, study designand the statistical treatment of the results. In addition,overt macroangiopathy might act as a confoundingfactor because, according to Steno’s hypothesis, micro-albuminuria may result from widespread atherosclerosis[21]. In order to clarify further whether there is anyearly significant association between microalbuminuriaand Lp(a) concentration, and microalbuminuria and theLp(a) phenotypes, we performed a case–control studyof normo- and microalbuminuric NIDDM patients with-out clinically detectable macroangiopathy.

Subjects and methods

Twenty-seven consecutive NIDDM patients ofCaucasian origin with persistent microalbuminuria andwithout clinically detectable macroangiopathy were

European Journal of Clinical Investigation(1997)27, 497–502

q 1997 Blackwell Science Ltd

Correspondence: Dr Rafael Simo´, Endocrinology Department,Hospital General Universitari Vall d’Hebron, Pso Vall d’Hebron,119–129, 08035 Barcelona, Spain.

recruited at the outpatient clinic of the Diabetic Unit of auniversity hospital (Vall d’Hebron, Barcelona, Spain)between February and May 1995. As a control group, 43NIDDM patients with normoalbuminuria and withoutclinically apparent CVD were individually matchedaccording to the following variables: age, gender, smok-ing habit, body mass index, blood pressure, diabetesduration, glycated haemoglobin (HbA1c), total serumcholesterol, low-density lipoprotein (LDL)-cholesterol,triglycerides and the presence of retinopathy (Table 1).Patients with renal failure, macroalbuminuria, acute orchronic infections, severe medical conditions (malig-nancy, liver cirrhosis, connective tissue disease, chroniccongestive heart failure), pregnancy or unstable diabeteswere excluded from the study. All diabetic patientstaking angiotensin-converting enzyme inhibitors werealso excluded. Informed written consent was obtainedfrom all the participants, and the study was approved bythe hospital’s human ethics committee.

The diagnosis of NIDDM was based on clinicalcharacteristics that included (a) no past history of keto-acidosis; (b) diagnosis of diabetes after 30 years of age;and (c) treatment by diet and/or oral agents or, in theinsulin-requiring patients, fasting serum C-peptidevalues> 0.30 nmol L–1.

To assess past and present evidence of CVD,a standardized cardiovascular questionnaire, a carefulphysical examination that included ankle–arm bloodpressure ratio and a 12-lead electrocardiogram (Minne-sota codes) were used [22]. Patients with a history of orpresent signs or symptoms suggesting CVD wererejected.

Presence of diabetic retinopathy was assessed byophthalmoscopy and fluorescein angiography. Thelatter was performed by injecting 5 mL of 20% sodiumfluorescein into an antecubital vein; rapid sequenceretinal photographs were taken with a fundus camera(Canon CF-60 UV). A retina specialist, unaware of

the clinical details of the patients, conducted theexaminations. Retinopathy was classified into threegroups: no retinopathy, non-proliferative retinopathyand proliferative retinopathy.

Body mass index was calculated as weight (kg)/heightsquared (m2). Weight and height were measured inlight clothing without shoes. Arterial blood pressurewas measured by one observer with a standard 12.5-cmcuff mercury sphygmomanometer as the mean of threereadings taken in the sitting position after 10 min ormore rest. Diastolic blood pressure was recorded atthe disappearance of the Korotkoff sounds (phase V).Hypertension was diagnosed according to a predeter-mined blood pressure level (systolic blood pressure>160 mmHg or diastolic blood pressure>95 mmHg) orif the patient was on antihypertensive therapy. Smokinghabits were also recorded, and smokers were definedas all subjects currently smoking more than one cigaretteper day.

Metabolic parameters were evaluated in venousblood drawn after overnight fasting. Blood was collectedinto EDTA for the HbA1c analysis, into sodium fluoridefor the glucose assay, and into plain tubes for lipidanalysis and renal biochemistry. Glucose was determinedusing a glucose oxidase method on a Hitachi 747 auto-analyser (Hitachi, Tokyo, Japan), and HbA1c wasmeasured by ion-exchange high-performance liquidchromatography (HPLC; Bio-Rad, CA, USA) (referencerange 4.7–6.6%). Total serum cholesterol and totaltriglycerides were tested on unfrozen samples by anautomated enzymatic method (Boehringer Mannheim,Mannheim, Germany) on a Hitachi 747 analyser. High-density lipoprotein (HDL)-cholesterol was determinedby assaying the cholesterol concentration in the super-natant obtained after precipitating lipoproteins withdensity lower than HDL by a mixture of phospho-tungstic acid and magnesium chloride. LDL-cholesterolwas calculated following Friedewald’s formula. Lp(a)was measured by ELISA in a separate aliquot of serumthat had been frozen before assay using a monoclonalLp(a) antibody technique (Macra Terumo, Newark,DE, USA). Our intra- and interassay coefficients ofvariation were 4.6% and 5.3% respectively. Lp(a)phenotypes were determined by sodium dodecylsulphate-polyacrylamide gradient gel electrophoresis(SDS-PAGE) under reducing conditions, followed byimmunoblotting. The method used in our study onlydiffers from the procedure described by Utermannet al.[23] in the electrophoretic gel (Excel Gel SDS Homo-geneous 7.5; Pharmacia). Sizes of apo(a) bands in sam-ples were estimated by comparison with mobilities ofapo(a) bands in a standard loaded gel in an immediatelyadjacent lane. According to their relative mobilitiescompared with apo B-100, apo(a) patterns were categor-ized into phenotypes F (faster than apo B-100), B(similar to apo B-100), S1, S2, S3 and S4 (all slowerthan apo B-100), and into the respective double-bandphenotypes. The ‘null’ phenotype was defined when theblots showed no bands at all.

Albumin excretion rate (AER) was evaluated by three

498 C. HERNANDEZ et al.

q 1997 Blackwell Science Ltd,European Journal of Clinical Investigation, 27, 497–502

Table 1. Clinical and metabolic characteristics of the patients

included in the study

With MA Without MA P

Patients (n) 27 43Age (years) 58.96 11.2 61.16 10.6 NSSex (M/F)% 44/56 39/61 NSSmokers (yes/no) 3/24 6/37 NSBMI (kg m–2) 29.86 4.6 28.76 4.9 NSHTA (yes/no)% 41/59 44/56 NSDiabetes duration (years) 10.76 9.5 11.76 8.6 NSHbA1c (%) 8.46 1.8 8.36 1.7 NSRetinopathy (%) NSAbsence 29.6 53.4Non-proliferative 51.8 41.9Proliferative 18.5 4.7Total cholesterol (mmol L–1) 6.246 1.52 5.576 1.21 NSLDL-cholesterol (mmol L–1) 3.896 1.10 3.596 1.07 NSTriglycerides (mmol L–1) 2.286 1.44 1.906 1.17 NS

Data are means6 SD.MA, microalbuminuria.

24-h urine samples collected within the 6 months beforethis cross-sectional study and expressed as the mean ofthe three determinations. Urine samples with patho-logical sediment were rejected. Normoalbuminuria wasdefined as<20mg min–1 and persistent microalbumi-nuria as 20–200mg min–1 in at least two out of threeurine collections. Urinary albumin concentration wasestimated by a double antibody radioimmunoassay(Diagnostic Products, Los Angeles, CA, USA). Noneof the patients followed a low-protein diet and nonehad historical, clinical or laboratory evidence of non-diabetic renal disease.

Statistical analysis

The chi-squared test was used for categorical variablesand the Student’st-test for continuous variables to verifythe absence of statistical differences. Lp(a) and AERresults were displayed as either median and confidenceinterval (CI) or median and range, in view of theirskewed distribution. Both Lp(a) and AER results werelogarithmically transformed when a parametric test wasused. Because of the low number of observations in manyof the statistical cells, the Lp(a) phenotypes were classi-fied by size into three groups: ‘small’ (phenotypes F, B,S1 and S2), ‘big’ (phenotypes S3 and S4) and ‘null’.When the patient had a double band, the smaller bandwas used to express the phenotype in the analysis. Thedifferences in Lp(a) concentrations among phenotypegroups were assessed by the Kruskal–Wallis test. Tocompare the concentrations of Lp(a) between patients

with and without microalbuminuria and to analysedifferences in phenotype distribution, the Mann–Whitney U-test and chi-squared test were used respec-tively. Correlation between AER and Lp(a) levelswas studied by linear regression analysis.ANOVA forAER to different Lp(a) phenotypes was covariated byLp(a) concentration (ANCOVA) in order to investigate theinfluence of Lp(a) phenotype itself on AER. AP-value(two-tailed) less than 0.05 was considered statisticallysignificant.

Results

The NIDDM patients with microalbuminuria presentedhigher serum Lp(a) concentrations than the patientswithout microalbuminuria [15.7 mg dL–1 (95% CI 0.5–36.5) vs. 4.5 mg dL–1 (95% CI 0.1–18.5); P<0.001]and the pattern of Lp(a) phenotype distribution wasalso different (P¼ 0.03) (Fig. 1). Furthermore, therewas a correlation between Lp(a) concentration andAER (r ¼ 0.34; P<0.01) (Fig. 2). The median of AERamong the three subsets of phenotypes was statisticallydifferent (P¼ 0.04. Table 2). This difference resultedfrom low AER detected in patients with ‘null’ phenotypeand remains at a significant level when the AER of

LP(A) PHENOTYPES AND AER IN NIDDM 499


Figure 1. Distribution of Lp(a) phenotypes in NIDDM patients withand without microalbuminuria (MA).

Figure 2. Relationship between serum Lp(a) and AER (r ¼ 0.34;P< 0.01).

Table 2.Serum Lp(a) and AER by phenotypes

Phenotype

Small Big Null(n¼ 34) (n¼ 28) (n¼ 8) P

Lp(a) (mg dL–1)Median 9 3 1.2 0.01Range 0.1–80 0.3–32 0.1–7.5

AER (mg min–1)Median 19 9.5 4.1 0.04Range 1–195 1–186 1–9

NIDDM patients with ‘null’ phenotype was comparedwith the other phenotypes as a whole, in a covariatedanalysis by Lp(a) concentration (ANCOVA; P¼ 0.02). Itshould be noted that, in all patients with null phenotype,the AER was lower than 10mg min–1. In addition, thestatistically significant difference in Lp(a) serum con-centration among the three subsets of Lp(a) phenotypesmakes the classification used in the present study suitable(P¼ 0.01; Table 2).

Discussion

Lp(a) is a cholesterol-rich, LDL-like particle, whoseprotein moiety contains apo(a) in addition to apo B-100[24]. Apo(a) shows a strong sequence homology withplasminogen [25] and exhibits a genetic size polymorph-ism that accounts for more than 90% of the variation inplasma Lp(a) levels and explains the skewed distributionof Lp(a) concentrations in the general population [26].Lp(a) is an independent risk factor for CVD [7–9] andseems to be related to the severity of this disease inangiographic studies [27,28]. However, the role of Lp(a)in the development of atherosclerosis in NIDDM patientsis less clear [10–13].

Microalbuminuria has been related to increased Lp(a)concentrations in subjects with NIDDM [15,17,18].However, other studies have not observed this associa-tion [14,16,19,20]. The divergent results concerningLp(a) levels in relation to AER could be partly causedby the small size of samples in most of these studies.Other contributing factors for these conflicting results arethat potential confounding variables on AER are notconsidered [29] and there is a lack of homogeneitywith respect to variables such as serum cholesterol [14]and triglycerides [30] that could influence serum Lp(a).Maioli et al. [31] have reported an association betweenserum Lp(a) and the severity of retinopathy in insulin-dependent diabetic (IDDM) patients, but this associationhas not been reported in NIDDM patients. In the presentstudy, we have considered all these potential biases in theselection of patients. In addition, although a link betweenmicroalbuminuria and increased Lp(a) levels may resultfrom widespread atherosclerosis [21], this possibilityhas not been considered previously. With regard tothis, we have excluded NIDDM patients with overtmacroangiopathy, although subclinical macroangiopathycould not be ruled out. With this limitation in mind,we have demonstrated in this case–control study thatNIDDM patients with microalbuminuria presentedhigher Lp(a) concentrations than normoalbuminuricpatients carefully matched by the mean factors thatcould influence serum Lp(a) and AER.

Increased Lp(a) levels have been reported in non-diabetic patients with heavy proteinuria of a differentorigin [32], and Lp(a) serum concentrations havebeen shown to rise and fall with exacerbations andremissions of proteinuria respectively [33]. This raisesthe possibility that albuminuria leads to increasedhepatic synthesis of Lp(a). The absence of NIDDM

with macroalbuminuria in this study makes the latterconsideration unlikely. On the other hand, recent experi-mental data suggest that lipid abnormalities may influ-ence the development and progression of renal injury[34,35], and it has been demonstrated in humans thatglomerular deposits of Lp(a) are linked to a less favour-able course of renal disease [36]. Studiesin vitro inhuman mesangial cells and matrix have shown that theseglomerular deposits are more likely to represent matrix-bound Lp(a) [37]. Matrix bound Lp(a) could enhancethe migration and adherence of leucocytes, which in turncould contribute to glomerular injury directly or throughthe oxidation of lipoproteins [38]. Alternatively, matrix-bound lipoproteins could interfere with cell–matrixinteractions, thereby inducing cell apoptosis [39]. There-fore, Lp(a) might play a role in the development ofmicroalbuminuria by direct renal damage.

The most important determinant of Lp(a) concentra-tions is the genotype at LPA, the locus encoding apo(a),which determines the size polymorphism of Lp(a) [26].The recent demonstration that distinct phenotypes mayhave different functional properties with regard to riskof thrombosis [40,41] points out the value of phenotypeassessment. To our knowledge, Lp(a) phenotypes havenot been considered previously in reports about therelationship between microalbuminuria and Lp(a) con-centrations. The improvement of electrophoretic tech-niques has indicated that Lp(a) is more variable thanoriginally suspected, and 34 different Lp(a) phenotypeshave been identified [42]. A variation in the ability ofapo(a) to move out of the endoplasmic reticulumexplains the inverse relationship between apo(a) pheno-type size and serum Lp(a) level [43]. At present, aconsensus for grouping the Lp(a) phenotypes does notexist, but it would be desirable in order to facilitatestatistical management of the data. Therefore, we havegrouped phenotypes by size as ‘small’ (F, B, S1 and S2),‘big’ (S3 and S4) and ‘null’, as reported previously byothers authors [11,44]. The clear statistically significantdifference in Lp(a) serum concentration demonstratedamong the three groups observed in our study reinforcedthis classification.

We feel that the most important result of this studyis the clear association between ‘null’ phenotype andlower AER. This result cannot only be attributed to thelower Lp(a) concentrations, characteristically associatedwith ‘null’ phenotype, because lower AER in thesepatients persists after considering Lp(a) concentrationin a covariated analysis. It should be emphasized thatmicroalbuminuria was absent in all NIDDM patientswith ‘null’ phenotype. To date, we cannot provide anyreliable explanation for this finding and, to our knowl-edge, there are no studies addressing this issue. However,it has been demonstratedin vitro that Lp(a) competi-tively inhibits the fibrin-dependent activation of plas-minogen to plasmin, and this action depends not onlyon the Lp(a) concentration but also on the Lp(a) pheno-type [40,41]. Moreover, plasmin plays a crucial role inthe catabolism of extracellular matrix proteins (ECM)[45,46]. Thus, it could be hypothesized that Lp(a) could



favour ECM accumulation and contributes to mesangiumexpansion, a capital event in early diabetic nephropathy.We could speculate from our observation that the ‘null’phenotype protects against ECM expansion, but futureinvitro studies focused on the differential effect of Lp(a)phenotypes on mesangium expansion will be needed toclarify this issue.

In view of our results, the meaning of ‘null’ phenotypedeserves an additional comment. ‘Null’ phenotype isassociated with low serum concentrations of Lp(a), andit is defined as no protein product being identified byimmunoblotting. Apo(a) alleles associated with ‘null’phenotype are not of a uniformly large size, but aredistributed over the entire size spectrum [47]. Recently,it has been demonstrated in baboons that two classesof ‘null’ apo(a) alleles exist, ‘transcript-negative null’alleles, which do not produce a detectable hepatic apo(a)mRNA transcript and ‘transcript-positive null’ alleles,which do not give rise to a detectable plasma protein,despite the presence of substantial levels of hepaticapo(a) mRNA. A reasonable hypothesis for this findingwould be the presence of a mutation in apo(a), whichprevents its correct folding and results in its retention inthe endoplasmic reticulum. Therefore, ‘null’ allelesrepresent extreme examples of both transcriptional andpost-transcriptional regulation of plasma Lp(a) levels[43]. The distribution of Lp(a) phenotypes is similar indiabetic patients to the general population, and theprevalence of the ‘null’ phenotype ranges between1.1% and 15.7% [44,47–49]. In the present study,when all patients were considered, the frequency of the‘null’ phenotype did not differ from the results obtainedby other authors. However, the percentage of ‘null’phenotype in the control group was higher (18.6%)than previously reported. This finding could be causedby the design of the study, because patients included inthe control group were without both cardiovasculardisease and microalbuminuria, two conditions thatcould be associated with low Lp(a) serum levels.

It has been suggested that susceptibility to developingmicroalbuminuria may be inherited in NIDDM [50]. Anumber of genetically conditioned mechanisms, such asthe sodium–lithium countertransport system [51], thepolymorphism of the angiotensin-converting enzymegene [52], insulin resistance [53] and the componentsof the glomerular vascular basement membrane [54],have been implicated. The link between Lp(a) pheno-types and AER shown in our study suggests that theLp(a) phenotype might be added as a new geneticcandidate for developing diabetic nephropathy, butfamilial aggregation studies are needed in order to con-firm this hypothesis.

In conclusion, in NIDDM patients without clini-cally detectable macroangiopathy, microalbuminuria isassociated with high serum Lp(a) concentrations. Inaddition, the ‘null’ phenotype could be considered as aprotective factor in the development of microalbumi-nuria in NIDDM patients. An appropriately designedprospective study is warranted in order to confirmthis concept.

Acknowledgments

Part of this paper was presented at the 31st AnnualMeeting of the European Association for the Study ofDiabetes (Stockholm, 12–16 September 1995).

References

1 Yudkin JS, Forrest RD, Jackson CA. Microalbuminuria as predictorof vascular disease in non-diabetic subjects. Lancet 1988;ii:530–3.

2 Damsgaard EM, Froland A, Jorgensen OD, Mogensen CE. Micro-albuminuria as a predictor of increased mortality in elderly people.BMJ 1990;300:297–300.

3 Haffner SM, Stern MP, Gruber KK, Hazuda HP, Mitchell BD,Patterson JK. Microalbuminuria: potential marker for increasedcardiovascular risk factors in non-diabetic subjects? Arteriosclero-sis 1990;10:727–31.

4 Jarrett RJ, Viberti GC, Argyropoulos A, Hill RD, Mahmud U,Murrells TJ. Microalbuminuria predicts mortality in non-insulin-dependent diabetes. Diabetic Med 1984;1:17–19.

5 Mattock MB, Keen H, Viberti GCet al.Coronary heart disease andurinary albumin secretion rate in type 2 (non-insulin-dependent)diabetic patients. Diabetologia 1988;31:82–7.

6 Damsgaard EM, Froland A, Jorgensen OD, Mogensen CE. Eightto nine year mortality in known non-insulin dependent diabeticsand controls. Kidney Int 1992;41:731–5.

7 Rhoads GC, Dahle´n G, Berg K, Morton NE, Dannenberg AL. Lp (a)lipoprotein as a risk factor for myocardial infarction. JAMA1986;256:2540–4.

8 Rosegren A, Wilhelmsen L, Eriksson E, Risberg B, Wedel H.Lipoprotein (a) and coronary heart disease: a prospective case–control study in a general population sample of middle aged men.BMJ 1990;301:1248–51.

9 Sandholzer CH, Boerwinkle E, Saha N, Tong C, Utermann G.Apolipoprotein (a) phenotypes, Lp (a) concentration and plasmalipid levels in relation to coronary heart disease in a Chinesepopulation: evidence for the role of the apo (a) gene in coronaryheart disease. J Clin Invest 1992;89:1040–6.

10 Velho G, Erlich D, Turpin Eet al. Lipoprotein (a) in diabeticpatients and normoglycemic relatives in familial NIDDM. DiabetesCare 1993;16:742–7.

11 Ruiz J, Thillet J, Huby Tet al. Association of elevated lipoprotein(a) levels and coronary heart disease in NIDDM patients. Relation-ship with apolipoprotein (a) phenotypes. Diabetologia 1994;37:585–91.

12 O’Brien T, Chu B, Nguyen TTet al. Lipids and Lp (a) lipoproteinlevels and coronary artery disease in subjects with non-insulin-dependent diabetes mellitus. Mayo Clin Proc 1994;69:430–5.

13 Hiraga T, Kobayashi T, Okubo Met al. Prospective study oflipoprotein (a) as a risk factor for atherosclerotic cardiovasculardisease in patients with diabetes. Diabetes Care 1995;18:241–4.

14 Haffner SM, Gruber KK, Morales PAet al. Lipoprotein (a) con-centrations in Mexican Americans and non-hispanic whites: theSan Antonio heart study. Am J Epidemiol 1992;136:1060–8.

15 Jenkins AJ, Steele JS, Janus ED, Santamaria JD, Best JD. Plasmaapolipoprotein (a) is increased in type 2 (non-insulin-dependent)diabetic patients with microalbuminuria. Diabetologia 1992;35:1055–9.

16 Nielsen FS, Voldsgaard AI, Gall MA, Rossing P, Hommel E,Andersen P. Apolipoprotein (a) and cardiovascular disease in type2 (non-insulin-dependent) diabetic patients with and withoutdiabetic nephropathy. Diabetologia 1993;36:438–44.

17 Heller FR, Jamart J, Honore Pet al.Serum lipoprotein (a) in patientswith diabetes mellitus. Diabetes Care 1993;16:819–23.

18 Jerums G, Allen TJ, Tsalamandris Cet al.Relationship of progres-sively increasing albuminuria to apoprotein (a) and blood pressurein type 2 (non-insulin-dependent) and type 1 (inslin-dependent)diabetic patients. Diabetologia 1993;36:1037–44.

19 Gruden G, Veglio M, Cavallo-Perin Pet al. Lipoprotein (a) innon-insulin-dependent diabetic patients with normo- and micro-albuminuria. Horm Metab Res 1994;26:489–90.

20 Scnack C, Pietschmann P, Kno¨bl P, Schuller E, Prager R,

LP(A) PHENOTYPES AND AER IN NIDDM 501


Schernthaner G. Apolipoprotein (a) levels and atherogenic lipidfractions in relation to the degree of urinary albumin excretion intype 2 diabetes mellitus. Nephron 1994;66:273–7.

21 Deckert T, Feldt-Rasmussen B, Borch-Johnsen K, Jensen T,Kofoed-Envoldsen A. Albuminuria reflects widespread vasculardamage: the Steno hypothesis. Diabetologia 1989;32:219–26.

22 Blackburn H, Keys A, Simonson E, Rautaharju P, Punsar S. Theelectrocardiogram in population studies: a classification system.Circulation 1960;21:1160–75.

23 Utermann G, Menzel HJ, Kraft HG, Duba HC, Kemmler HG,Seitz C. Lp (a) glycoprotein phenotypes. Inheritance and relationto Lp (a)-lipoprotein concentrations in plasma. J Clin Invest 1987;80:458–65.

24 Utermann G, Weber W. Protein composition of Lp (a) lipoproteinfrom human plasma. FEBS Lett 1983;154:357–61.

25 McLean JW, Tomlison JE, Kuang W-Jet al. cDNA sequence ofhuman apolipoprotein (a) is homologous to plasminogen. Nature1987;300:132–7.

26 Boerwinkle E, Leffert CC, Lin J, Lackner C, Chiesa G, Hobbs HH.Apolipoprotein (a) gene accounts for greater than 90% of thevariation in plasma lipoprotein (a) concentrations. J Clin Invest1992;90:52–60.

27 Hoff HF, Beck GJ, Skibinski CI, Ju¨rgens G, O’Neil K, Kramer J.Serum Lp (a) levels as a predictor of vein graft stenosis aftercoronary artery by-pass surgery in patients. Circulation 1988;77:1238–44.

28 Rath M, Niendorf A, Reblin T, Dietel M, Krebber HJ, Beisiegel U.Detection and quantification of lipoprotein (a) in the arterial wallof 107 coronary bypass patients. Arteriosclerosis 1989;9:593–603.

29 Mogensen CA, Vestbo E, Poulsen PLet al. Microalbuminuriaand potential confounders. Diabetes Care 1995;18:572–81.

30 McConathy WJ, Trieu VN, Koren E, Wang Ch, Corder C. Trigly-ceride-rich lipoprotein interactions with Lp (a). Chem Phys Lipids1994;67/68:105–13.

31 Maioli M, Tonolo G, Paci´fico A et al.Raised serum apolipoprotein(a) in active diabetic retinopathy. Diabetologia 1993;36;88–90.

32 Karadi I, Romies L, Palos Get al. Lp (a) lipoprotein concentrationin serum of patients with heavy proteinuria of different origin. ClinChem 1989;35:2121–3.

33 Short CD, Durrington PN. Hyperlipidaemia and renal disease.Baillieres Clin Endocrinol Metab 1990;4:777–806.

34 Kasiski BL, O’Donnell MP, Cowardin W, Keane WF. Lipids andthe kidney. Hypertension 1990;15:443–50.

35 Coritsidis G, Rifici V, Gupta S, Rie J, Shan Z, Neugarten J.Preferential binding of oxidized LDL to rat glomeruliin vivo andcultured mesangial cellsin vitro. Kidney Int 1991;39:858–66.

36 Sato H, Suzuki S, Ueno Met al. Localization of apolipoprotein (a)and B-100 in various renal diseases. Kidney Int 1993;43:430–5.

37 Kramer-Guth A, Greiber S, Pavensta¨dt H et al. Interaction of nativeand oxidized lipoprotein (a) with human mesangial cells and matrix.Kidney Int 1996;49:1250–61.

38 Ramaswamy C, Staedman R, Jordan N, Westwick J, Williams JD.Lipoprotein (a) stimulates leukocyte adherence to endothelial cellsin vitro (abstract). J Am Soc Nephrol 1995;6:907.

39 Sugiyama H, Kashihara T, Sekikawa Yet al. Cell–matrix interac-tions modulate cell apoptosis (abstract). J Am Soc Nephrol1995;6:912.

40 Hervio L, Chapman M, Thillet J, Loyau S, Angle´s-Cano E. Doesapolipoprotein (a) heterogeneity influence lipoprotein (a) effects onfibrinolysis?. Blood 1993;82:392–7.

41 Leerink CB, Duif PFCCM, Verhoeven Net al. Apolipoprotein (a)isoform size influences binding of lipoprotein (a) to plasmin-modified des-AA-fibrinogen. Fibrinolysis 1994;8:214–30.

42 Marcovina SM, Zhang ZH, Gaur VP, Albers JJ. Identification of 34apolipoprotein (a) isoforms: differential expression of apolipopro-tein (a) alleles between American blacks and whites. BiochemBiophys Res Commun 1993;191:1192–6.

43 White AL, Hixson JE, Rainwater DL, Lanford RE. Molecular basisfor ‘null’ lipoprotein (a) phenotypes and the influence of apolipo-protein (a) size on plasma lipoprotein (a) level in the baboon. J BiolChem 1994;25:9060–6.

44 Klausen IC, Berg Schmidt E, Lervang HH, Gerdes LU, Ditzel J,Faergeman O. Normal lipoprotein (a) concentration and apolipo-protein (a) isoforms in patients with insulin- dependent diabetesmellitus. Eur J Clin Invest 1992;22:538–41.

45 Vassalli JD, Pepper MS. Membrane proteases in focus. Nature1994;370:14–15.

46 Baricos WH, Cortez SL, el-Dahr SS, Schnaper HW. ECM degrada-tion by cultured human mesangial cells is mediated by a PA/plasmin/MMP-2 cascade. Kidney Int 1995;47:1039–47.

47 Gaw A, Boerwinkle E, Cohen JC, Hobbs HH. Comparativeanalysis of the apo (a) gene, apo (a) glycoprotein, and plasmaconcentrations of Lp (a) in three ethnic groups. Evidence for nocommon ‘null’ allele in the apo (a) locus. J Clin Invest 1994;93:2526–34.

48 Geroldi D, Bellotti V, Buscaglia Pet al.Characterization of apo (a)polymorphism by a modified immunoblotting technique in anItalian population sample. Clin Chim Acta 1993;221:159–69.

49 Csaszar A, Dieplinger H, Sandholzer Cet al.Plasma lipoprotein (a)concentration and phenotypes in diabetes mellitus. Diabetologia1993;36:47–51.

50 Gruden G, Cavallo-Perin P, Olivetti Cet al. Albumin excretionrate levels in non-diabetic offspring of NIDDM patients with andwithout nephropathy. Diabetologia 1995;38:1218–22.

51 Krolewski AS, Canessa M, Warram JHet al. Predisposition tohypertension and susceptibility to renal disease in insulin-dependentdiabetes mellitus. N Engl J Med 1988;18:140–5.

52 Cambien F, Poirier O, Lecerg Let al. Deletion polymorphism inthe gene for angiotensin-converting enzyme is a potent risk factorfor myocardial infarction. Nature 1992;359:641–4.

53 Forsblom CM, Eriksson JG, Ekstrand AV, Teppo AM, TaskinenMR, Groop LC. Insulin resistance and abnormal albumin excretionin non-diabetic first-degree relatives of patients with NIDDM.Diabetologia 1995;38:363–9.

54 Deckert T, Kofoed-Enevoldsen A, Norgaard K, Borch-Johnsen K,Feldt-Rasmussen B, Jensen T. Microalbuminuria. Implicationsfor micro- and macrovascular disease. Diabetes Care 1992;15:1181–91.



relationship between lipoprotein(a) phenotypes and albumin excretion rate in non-insulin-dependent...

Documents