Lipids and Lp(a) Lipoprotein Levels and Coronary Artery Disease in Subjects With Non-Insulin-Dependent Diabetes Mellitus

mayo #

Lipids and Lp(a) Lipoprotein Levels and Coronary Artery Disease in Subjects With Non-Insulin-Dependent Diabetes Mellitus M.D., JAY M. HARRISON, M.S.,* M.D., AND BRUCE A. KOTTKE, M.D., PH.D.f

TIMOTHY O'BRIEN, M . B . , B . C H . , TU PETER JAMES DYCK,

T.

NGUYEN,

• Objective: To determine whether increased Lp(a) lipoprotein levels are associated with either non-insulin-dependent diabetes mellitus (NIDDM) or coronary artery disease (CAD) in patients with NIDDM and to examine the relationship between Lp(a) levels and glycémie control. • Design: We conducted a cross-sectional study of subjects with NIDDM who were participants in the Rochester Diabetic Neuropathy Study and healthy control subjects from the population of Rochester, Minnesota. • Material and Methods: Lipids and Lp(a) lipoprotein levels were compared in 227 subjects with NIDDM and 163 control subjects and, among the subjects with NIDDM, in those with (N = 96) and without (N = 131) CAD. The correlation between Lp(a) levels and glycosylated hemoglobin was investigated. • Results: Subjects with NIDDM had higher triglycéride and lower high-density lipoprotein cholesterol levels than did control subjects. Subjects with Coronary artery disease (CAD) is the most common cause of death in subjects with diabetes mellitus.1 The increased risk for cardiovascular disease in this group cannot totally be attributed to conventional risk factors.2 The nature of the diabetic factor responsible for this increased risk is unknown. Lp(a) is a lipoprotein that contains apolipoprotein B attached to apolipoprotein (a) by a disulfide bond.3 This parFrom the Division of Endocrinology/Metabolism and Internal Medicine (T.O., Τ.Τ.Ν.), Atherosclerosis Research Unit (T.O., T.T.N., B.A.K.), Section of Biostatistics (J.M.H., K.R.B.), Peripheral Nerve Center, Department of Neurology (P.J.D.), and Division of Cardiovascular Diseases and Internal Medicine (Β.Α.Κ.), Mayo Clinic Rochester, Rochester, Minnesota. *Current address: University of Florida, Gainesville, Florida. tCurrent address: The Watson Clinic, Lakeland, Florida. This study was supported in part by a grant from the American Diabetes Association, Minnesota Affiliate, and Grant NS 14304 from the National Institutes of Health, Public Health Service. Dr. O'Brien is supported in part by the W. L. Stephenson Fellowship in Clinical Nutrition. Address reprint requests to Dr. T. T. Nguyen, Division of Endocrinology/ Metabolism, Mayo Clinic Rochester, 200 First Street SW, Rochester, MN 55905. Mayo Clin Proc 1994;69:430-435

KENT

R.

BAILEY, P H . D . ,

NIDDM and CAD had higher total cholesterol, triglycéride, and low-density lipoprotein cholesterol levels and lower high-density lipoprotein cholesterol levels than did subjects with NIDDM without CAD. Subjects with NIDDM had significantly higher Lp(a) levels than did control subjects, but subjects with NIDDM and CAD did not have significantly higher Lp(a) levels than did those without CAD. Among subjects with NIDDM, the level of Lp(a) was not significantly correlated with glycosylated hemoglobin. • Conclusion: Although subjects with NIDDM have higher Lp(a) levels than do control subjects, Lp(a) does not seem to be associated with CAD in subjects with NIDDM. In this study, no association was found between Lp(a) level and glycémie control. (Mayo Clin Proc 1994; 69:430-435) ANCOVA = analysis of covariance; CAD = coronary artery disease; HDL = high-density lipoprotein; LDL = low-density lipoprotein; NIDDM = non-insulin-dependent diabetes mellitus

tide was originally described by Berg4 in 1963. In 1972, an association was found between CAD and a pre-ß, migrating band,5"7 which was subsequently shown to be Lp(a).8 Lp(a) lipoprotein has been shown to be an independent risk factor for atherosclerotic vascular disease in subjects without diabetes.9 " The link between atherosclerosis and Lp(a) may be the striking homology found between apolipoprotein (a) and plasminogen;12·13 thus, a potential association may exist between the coagulation cascade and atherosclerosis. The presence of Lp(a) in atherosclerotic plaques is further evidence of a potential role for this particle in atherogenesis.14 Lp(a) levels have been reported to be normal" 17 or increased'8 " in subjects with non-insulin-dependent diabetes mellitus (NIDDM). The contribution of Lp(a) to the increased risk for atherosclerosis in subjects with diabetes is unknown. Velho and associates17 noted increased Lp(a) levels in subjects with diabetes who had a history of myocardial infarction. In contrast, Haffner and colleagues20 found no relationship between Lp(a) levels and CAD-associated

430

© 1994 Mayo Foundation for Medical Education and Research

Mayo Clin Proc, May 1994, Vol 69

mortality in subjects with diabetes mellitus. In the current report, we examine the following issues: the difference in Lp(a) levels between subjects with NIDDM and control subjects; the relationship between Lp(a) levels and diabetic control, as assessed by glycosylated hemoglobin levels in subjects with NIDDM; and the relationship between Lp(a) and CAD in subjects with NIDDM. MATERIAL AND METHODS Study Subjects.—Blood samples were obtained from 227 subjects (115 women and 112 men) with NIDDM who were participants in the Rochester Diabetic Neuropathy Study and 163 adult control subjects (85 women and 78 men) who were healthy persons from the general population of Rochester, Minnesota, without evidence of CAD on clinical examination and review of the medical history. All subjects were at least 35 years old. The diagnosis of diabetes was based on a fasting blood glucose level of more than 140 mg/dL on at least two occasions. Subjects were defined as having NIDDM on the basis of clinical criteria, including the age at onset of diabetes mellitus, absence of ketosis, and need for insulin therapy. The presence of nephropathy was defined as Albustix-positive proteinuria of more than trace amounts or a 24-hour urinary protein of more than 300 mg in men or 150 mg in women. The diagnosis of CAD was based on the following criteria: a history of myocardial infarction, electrocardiographic changes indicative of previous myocardial infarction (all subjects with diabetes in this study underwent electrocardiography), abnormal results on noninvasive cardiovascular testing, or abnormal findings on coronary arteriography. Laboratory Analyses.—Blood samples were withdrawn into tubes that contained ethylenediaminetetraacetic acid. Plasma was separated by centrifugation in a Beckman model TJ-6 tabletop centrifuge machine (Beckman Instruments, Inc., Fullerton, California) at 4°C and at l,500g for 20 minutes, and aliquots were frozen at -70°C for later analysis of Lp(a). Cholesterol and triglycérides were measured by standard enzymatic methods with the use of quality-control plasma pools.21-22 High-density lipoprotein (HDL) cholesterol was measured by using polyethylene glycol 6000 precipitation, as previously described.22·23 Low-density lipoprotein (LDL) cholesterol levels were calculated by using the Friedenwald equation when the triglycérides were less than 3.39 mmol/L. The Lp(a) lipoprotein level was determined by using a commercially available enzyme-linked immunosorbent assay, with use of a monoclonal antibody against apolipoprotein (a), which does not cross-react against plasminogen, and a second polyclonal antibody directed against the apolipoprotein (a) portion of Lp(a) (Terumo Medical Corp., Elkton, Maryland). Hemoglobin A, (glycosylated hemoglobin) was measured by affinity chromatography.24

LIPIDS, CORONARY ARTERY DISEASE, AND DIABETES

431

Statistical Analyses.—Comparisons of age, gender, and other clinical variables between groups (subjects with NIDDM versus control subjects; subjects who had NIDDM with versus without CAD) were done by t tests or by Pearson's χ2 tests. Distributions were summarized by means and standard deviations or by frequency percents. Group mean differences in Lp(a) and lipid levels between subjects with NIDDM and control subjects were analyzed with and without adjustment for age and sex. Unadjusted comparisons were done by two-sample t tests, whereas adjusted comparisons were done by analysis of covariance (ANCOVA) models in which age and sex were used as covariates. Because the distribution of Lp(a) was skewed, analysis was done on the transformed variable log[Lp(a) + 1]. Similarly, triglycérides were analyzed on the log scale. In order to understand the association of NIDDM with Lp(a) further, an ANCOVA model that included sex, age, NIDDM, and an interaction between age and NIDDM was estimated. Finally, t tests between NIDDM and control subjects were done separately for those younger than 60 years of age. The association of the various lipids with CAD was assessed both unadjusted (by / tests) and adjusted (by multiple logistic regression), with CAD as the dependent variable and with each lipid, in turn, plus age and sex as the independent variables. In addition, the possibility of a synergistic effect between Lp(a) and the other lipids was tested by including Lp(a) together with each lipid, as well as their interaction, in these logistic models. For investigation of the relationship between Lp(a) and glycosylated hemoglobin, Pearson correlation coefficients with the transformed Lp(a) variable were calculated. Fisher's exact test was used to analyze associations between pairs of dichotomous variables, such as nephropathy and presence of CAD. RESULTS Lipoprotein Profiles in Subjects With NIDDM Versus Control Subjects.—Lipid and Lp(a) lipoprotein profiles for subjects with NIDDM and control subjects are shown in Table 1. In comparison with control subjects, those with NIDDM had higher triglycéride levels (2.47 versus 1.34 mmol/L; P = 0.0001, ANCOVA) and lower HDL cholesterol levels (0.92 versus 1.21 mmol/L; P = 0.0001). Total and LDL cholesterol levels were approximately the same in NIDDM and control groups. Lp(a) levels were significantly higher in subjects with NIDDM than in control subjects without diabetes (17.5 versus 13.7 mg/dL; P = 0.0339, t test). When age and sex were adjusted by multiple regression, this difference was no longer significant (P = 0.17). When an interaction between diabetes and age (that is, different age effects in the two groups) was allowed, it seemed to indicate a significantly (P = 0.034) different age slope in the two groups— namely, a significant positive association in normal subjects

432

LIPIDS, CORONARY ARTERY DISEASE, AND DIABETES

Mayo Clin Proc, May 1994, Vol 69

Table 1.—Age, Gender, and Lipid Data for Subjects With Non-Insulin-Dependent Diabetes Mellitus and Control Subjects* Subjects Factor Age (yr) Gender (% male) Total cholesterol (N = 225 vs 163) Triglycerides§ (N = 225 vs 163) HDL cholesterol (N = 220vs 163) LDL cholesterol (N= 184 vs 160) Lp(a) (mg/dL)§

/ ' value ANCOVA

NIDDMt (N = 227)

Controlt (N=163)

itest

64.8 ±10.2 48.5

53.8 ± 13.7 47.9

0.0001 0.9191

5.14+1.02

5.04 ± 0.99

0.30

0.79

2.47 ±1.83

1.34 ±0.89

0.0001

0.0001

0.92 ± 0.28

1.21 ±0.39

0.0001

0.0001

3.24 ± 0.88 17.5 ± 19.4

3.21 ±0.93 13.7 ±14.5

0.77 0.0339

0.54 0.17

*ANCOVA = analysis of covariance; HDL = high-density lipoprotein; LDL = low-density lipoprotein; NIDDM = non-insulin-dependent diabetes mellitus; vs = versus. tValues are shown as mean ± SD. All cholesterol and triglycéride data are in mmol/L. +Fisher's exact test. §Data are summarized on natural scale, but analysis was performed on log scale. but no relationship and even a downward trend in those with NIDDM. Thus, age was positively associated with Lp(a) levels in those without but not in those with diabetes. A t test—on log[Lp(a) + 1]—restricted to the subjects younger than 60 years of age showed that the NIDDM-control difference was accentuated (P = 0.023). When the analysis was further restricted to those younger than 50 years of age, the t test P value was 0.017. The difference in Lp(a) levels between subjects with NIDDM and control subjects persisted after exclusion of those with diabetes who had nephropathy.

Characteristics of NIDDM Groups With and Without CAD.—The characteristics of the subjects who had NIDDM with and without CAD are shown in Table 2. The subjects with NIDDM with (N = 96) and without (N = 131) CAD were similar relative to gender distribution, body mass index, glycémie control, and duration of diabetes mellitus. The subjects with NIDDM who had CAD were older (67.3 versus 62.9 years; P = 0.001) and had a slightly higher creatinine level (1.11 versus 0.97 mg/dL; P = 0.0009) than subjects with NIDDM without CAD.

Table 2.— Characteristics of Study Subjects With Non-Insulin-Dependent Diabetes Mellitus, Stratified by Absence or Presence of Coronary Artery Disease* Factor Age (yr) Gender (% male) BMI (kg/m2) Duration of diabetes (yr) Glycosylated hemoglobin (%) Creatinine (mg/dL)

P value ANCOVA

No CADt (N=131)

CADt (N = 96)

t test

62.9 ±10.8 45.0 30.4 ±6.1

67.3 ± 8.7 53.1 30.5 ± 5.6

0.001 0.23Φ 0.89

0.27

10.5 ±7.8

12.2 ±8.9

0.13

0.83

10.5 ±3.0 0.97 ± 0.22

10.5 ±2.7 1.11 ±0.32

0.87 0.0009

0.72 0.016

* ANCOVA = analysis of covariance; BMI = body mass index; CAD = coronary artery disease. tValues are shown as mean ± SD. tFisher's exact test.

Mayo Clin Proc, May 1994, Vol 69

LIPIDS, CORONARY ARTERY DISEASE, AND DIABETES 433

CAD Risk Profile in Subjects With NIDDM.—We found Although the number of subjects with nephropathy was no significant differences between the subjects who had small, Lp(a) levels did not seem to be associated with the NIDDM with and without CAD relative to either cigarette presence of this complication. Lp(a) Levels and Glycémie Control.—Levels of Lp(a) smoking (52% in the group with CAD versus 41% in those without; P - 0.12) or hypertension (53% in the group with were not significantly correlated with glycosylated hemoCAD versus 45% in those without; P = 0.20). With adjust- globin among the subjects with NIDDM (r = -0.12; P = ment for age, however, smoking became significantly asso- 0.0777). This finding prevailed for both subgroups of subciated with CAD (smoking was less frequent in older sub- jects with NIDDM (those with and those without CAD). jects than in younger subjects). Of the 227 subjects with NIDDM, 19 had nephropathy. Nephropathy was present DISCUSSION significantly more often in the group with CAD than in those Lp(a) lipoprotein is a risk factor for CAD in subjects without diabetes9 ' ' and seems to be related to the severity of CAD as without CAD (13.5% versus 4.6%; P = 0.027). Lipoprotein Profiles in Subjects With NIDDM With and assessed angiographically.10 The role of Lp(a) in atheroWithout CAD.—The lipid and Lp(a) values in subjects with genesis in diabetes is unknown. Researchers have postulated NIDDM with and without CAD are shown in Table 3. On that Lp(a) may, in part, explain the increased risk of vascular logistic regression, subjects with NIDDM who had CAD had disease associated with diabetes mellitus that exists after significantly higher levels of total cholesterol (5.26 versus controls have been imposed for conventional risk factors. 5.04 mmol/L; P = 0.0092), calculated LDL cholesterol (3.40 The role of Lp(a) in atherogenesis in NIDDM has not versus 3.13 mmol/L; P = 0.0097), and triglycérides (2.66 been addressed extensively. Joven and Vilella25 did not find versus 2.33 mmol/L; P = 0.0057) as well as lower HDL increased Lp(a) levels in 43 subjects with well-controlled cholesterol levels (0.82 versus 0.98 mmol/L; P = 0.0001). In NIDDM. Furthermore, administration of insulin to 30 subsubjects with NIDDM and CAD, however, Lp(a) levels did jects with NIDDM failed to decrease Lp(a) levels.26 Investinot differ significantly from those in subjects without CAD gators have reported that Lp(a) levels did not improve with (18.65 versus 16.67 mg/dL; P = 0.22). To determine better glycémie control in 12 subjects with NIDDM.27 In a whether the Lp(a) level acted synergistically with total cho- recent study, Haffner and associates15 found that Lp(a) levels lesterol, triglycérides, or HDL cholesterol level in the devel- were similar in subjects with NIDDM and control subjects. opment of CAD, we included Lp(a) and each of these vari- Corresponding findings were noted in two other studies.16·17 ables and an interaction term in an age- and gender-adjusted In contrast, our study of 227 subjects with NIDDM confirms logistic regression model. Neither the Lp(a) level nor the the findings of other studies18·19 in which Lp(a) levels were Lp(a) interaction with total cholesterol, triglycérides, or higher in subjects with NIDDM than in control subjects. HDL cholesterol level significantly affected the ability to Although in the current study the difference was no longer predict the presence of CAD among subjects with NIDDM. significant after adjustments were made for age and sex, this Table 3.— Lipid Data for Subjects With Non-Insulin-Dependent Diabetes Mellitus, Stratified by Absence or Presence of Coronary Artery Disease*

Factor Total cholesterol (N= 130vs95) Triglycérides:]: (N=130vs95) HDL cholesterol (N= 114vs70) LDL cholesterol (N = 114 vs 70) Lp(a) (mg/dL»

P value Logistic /test regression

No CADt (N= 131)

CADt (N = 96)

5.0411.00

5.26+1.03

0.085

0.0092

2.33+1.78

2.66+1.88

0.055

0.0057

0.98 ± 0.32

0.82 ±0.19

0.0001

0.0001

3.13 ±0.88 16.67+19.40

3.40 ± 0.87 18.65 ±19.36

0.039 0.36

0.0097 0.22

*CAD = coronary artery disease; HDL = high-density lipoprotein; LDL = lowdensity lipoprotein; vs = versus. tValues are shown as mean ± SD. All cholesterol and triglycéride data are in mmol/L. tData are summarized on natural scale, but analysis was performed on log scale.

434

LIPIDS, CORONARY ARTERY DISEASE, AND DIABETES

Mayo Clin Proc, May 1994, Vol 69

result may be explained by the fact that Lp(a) levels in- a synergistic role in atherogenesis only in the presence of creased with advancing age in the control group but not in hypercholesterolemia.9 Neither Lp(a) nor the interaction of the subjects with NIDDM in this study. In fact, the differ- Lp(a) with total cholesterol, triglycéride, or HDL cholesterol ence in Lp(a) levels between those with and those without level, however, was a significant predictor of CAD in the diabetes was even more significant when subjects older than NIDDM group. Finally, subjects with diabetes and in60 years of age were excluded from the analysis. Why the creased Lp(a) levels may die prematurely and therefore relationship between age and Lp(a) levels differs in those would not be included in this cross-sectional study. with and those without diabetes is unclear. One possibility Another potential difficulty with the current study is the might be that subjects with diabetes and increased Lp(a) separation of subjects with and without CAD. The control levels die of vascular disease and therefore are not available group had no evidence of CAD on clinical assessment. The for inclusion in a cross-sectional study such as this. In- distinction in the subjects with NIDDM was based on the creased Lp(a) levels have also been found in subjects with criteria outlined in the Material and Methods section. Alimpaired glucose tolerance.28 In our study, Lp(a) levels though some subjects with NIDDM without clinical eviremained increased in subjects with NIDDM after exclusion dence of CAD may indeed have had underlying silent CAD, of those with diabetic nephropathy. One possible explana- conventional lipid risk factors were increased in the NIDDM tion for the conflicting results in studies of Lp(a) levels in group with clinically significant CAD. The subjects with NIDDM is the varied ethnic backgrounds of the populations diabetes and CAD had higher total cholesterol, triglycéride, studied. In a report by Haffner and colleagues, '5 71 % of men and LDL cholesterol levels and lower HDL cholesterol levand 77% of women with NIDDM were Mexican-American. els than did those without clinical CAD. Of interest, the In contrast, 38% of the male and 37% of the female control prevalence of hypertension and cigarette smoking was simisubjects were Mexican-American. Lp(a) distribution is lar in the subjects with NIDDM with and without CAD known to be influenced by ethnic origin.29 In our study, all (although after adjustment for age, a positive association was subjects were Caucasian. Histograms of Lp(a) distribution found between CAD and cigarette smoking in the subjects in our diabetic and control populations were the same as with NIDDM). This finding emphasizes the importance of those reported for Caucasians in the literature (data not lipid abnormalities in increasing the cardiovascular risk in shown).29 We found no relationship between Lp(a) levels subjects with NIDDM. and glycémie control, in agreement with other reports of Finally, the risk of CAD seems to be increased in subjects subjects with NIDDM,2527 although these have been cross- with insulin-dependent diabetes and nephropathy.30 In our sectional studies. study, subjects with NIDDM and CAD had a higher freWhether Lp(a) excess is a risk factor for CAD in diabetes quency of nephropathy than did those without CAD. Several mellitus is unclear. In a recent study, Lp(a) levels were not studies have reported increased Lp(a) levels in subjects with higher in subjects with NIDDM who died of CAD than in diabetes mellitus and microalbuminuria or macroalbumin3133 Two of these studies, however, dealt with insulincontrol survivors.20 Although that study was small, it points uria. 31 32 out that extrapolation of results from studies of subjects dependent diabetes mellitus, · and the other did not menwithout diabetes to the diabetic population may be inappro- tion the proportion of subjects with diabetes who were de33 priate. In contrast, Velho and coworkers17 found increased pendent on insulin. In an additional report, Lp(a) levels Lp(a) levels in subjects with diabetes and a history of myo- were not significantly higher in subjects with insulin-depencardial infarction. In our comparison of Lp(a) levels in a dent diabetes mellitus who had microalbuminuria than in 34 Lipid levels including large number of subjects who had NIDDM with (N = 96) and those without microalbuminuria. without (N = 131) clinically significant CAD, those with Lp(a) were unaffected by the presence of nephropathy in our study. CAD did not have higher Lp(a) levels. Several possible explanations for this finding can be proposed. First, although Lp(a) levels were higher in the overall CONCLUSION NIDDM group in comparison with the nondiabetic control Lp(a) lipoprotein levels are higher in subjects with NIDDM population, the difference was small. Indeed, Lp(a) seems to than in control subjects. Lp(a) levels in subjects with be a risk factor for CAD when levels exceed 30 mg/dL.'° NIDDM and CAD, however, do not differ significantly from Second, the NIDDM group without clinically significant those without CAD. Furthermore, no significant association CAD was younger than the group with such disease. We exists between Lp(a) levels and glycémie control or cannot exclude the possibility that clinically evident CAD nephropathy. These findings do not support the conclusion will eventually develop in subjects in the former group. that Lp(a) excess is independently atherogenic in NIDDM. Third, the Lp(a) level may be a risk factor for CAD only in The possibility exists that Lp(a) has a synergistic role in young subjects. Fourth, one theory is that the Lp(a) level has diabetes-related atherogenesis, but an interactive role of

Mayo Clin Proc, May 1994, Vol 69

Lp(a) with total cholesterol level in promoting CAD is not supported by our current study.

LIPIDS, CORONARY ARTERY DISEASE, AND DIABETES

16. 17.

ACKNOWLEDGMENT We acknowledge the expert technical assistance of Nina D. Bren, who performed the Lp(a) assays, and the help of Douglas W. Mahoney with statistical analysis.

REFERENCES 1. Garcia MJ, McNamara PM, Gordon T, Kannel WB. Morbidity and mortality in diabetics in the Framingham population: sixteen year follow-up study. Diabetes 1974;23:105-111 2. Kannel WB, Gordon T, Castelli WP. Obesity, lipids, and glucose intolerance: the Framingham study. AmJClinNutr 1979; 32:12381245 3. Gaubatz JW, Heideman C, Gotto AM Jr, Morrisett JD, Dahlen GH. Human plasma lipoprotein(a): structural properties. J Biol Chem 1983:258:4582-4589 4. Berg K. A new serum type system in man—the Lp system. Acta Pathol Microbiol Scand 1963; 59:369-382 5. Dahlén G, Ericson C, Furberg C, Lundkvist L, Svärdsudd K. Angina of effort and an extra pre-beta lipoprotein fraction. Acta Med Scand Suppl 1972;531:11-15 6. Papadopoulos NM, Bedynek JL. Serum lipoprotein patterns in patients with coronary atherosclerosis. Clin Chim Acta 1973; 44:153157 7. Insult W, Najmi M, Vloedman DA. Plasma pre-beta lipoprotein subfractions in diagnosis of coronary artery disease [abstract]. Circulation 1972; 46(Suppl2):II-170 8. Dahlen G, Berg K, Gillnäs T, Ericson C. Lp(a) lipoprotein/pre-ß,lipoprotein in Swedish middle-aged males and in patients with coronary heart disease. Clin Genet 1975;7:334-341 9. Armstrong VW, Cremer P, Eberle E, Manke A, Schulze F, Wieland H, et al. The association between serum Lp(a) concentrations and angiographically assessed coronary atherosclerosis: dependence on serum LDL levels. Atherosclerosis 1986; 62:249-257 10. Dahlen GH, Guyton JR, Attar M, Farmer JA, Kautz JA, Gotto AM Jr. Association of levels of lipoprotein Lp(a), plasma lipids, and other lipoproteins with coronary artery disease documented by angiography. Circulation 1986;74:758-765 11. Rhoads GG, Dahlen G, Berg K, Morton NE, Dannenberg AL. Lp(a) lipoprotein as a risk factor for myocardial infarction. JAMA 1986; 256:2540-2544 12. Eaton DL, Fless GM, Kohr WJ, McLean JW, Xu Q-T, Miller GC, et al. Partial amino acid sequence of apolipoprotein(a) shows that it is homologous to plasminogen. Proc Natl Acad Sei U S A 1987; 84:3224-3228 13. McLean JW, Tomlinson JE, Kuang W-J, Eaton DL, Chen EY, Fless GM, et al. cDNA sequence of human apolipoprotein(a) is homologous to plasminogen. Nature 1987;330:132-137 14. Rath M, Niendorf A, Reblin T, Dietel M, Krebber H-J, Beisiegel U. Detection and quantification of lipoprotein(a) in the arterial wall of 107 coronary bypass patients. Arteriosclerosis 1989; 9:579592 15. Haffner SM, Morales PA, Stern MP, Gruber MK. Lp(a) concentrations in NIDDM. Diabetes 1992;41:1267-1272

18. 19. 20.

21. 22.

23.

24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

435

Csâszâr A, Dieplinger H, Sandholzer C, Karâdi I, Juhâsz E, Drexel H, et al. Plasma lipoprotein(a) concentration and phenotypes in diabetes mellitus. Diabetologia 1993;36:47-51 Velho G, Erlich D, Turpin E, Neel D, Cohen D, Froguel P, et al. Lipoprotein(a) in diabetic patients and normoglycemic relatives in familial NIDDM. Diabetes Care 1993;16:742-747 Guillausseau PJ, Peynet J, Chanson P, Legrand A, Altman JJ, Poupon J, et al. Lipoprotein(a) in diabetic patients with and without chronic renal failure. Diabetes Care 1992;15:976-979 Heller FR, Jamart J, Honore P, Derue G, Novik V, Galanti L, et al. Serum lipoprotein(a) in patients with diabetes mellitus. Diabetes Care 1993; 16:819-823 Haffner SM, Moss SE, Klein BE, Klein R. Lack of association between lipoprotein (a) concentrations and coronary heart disease mortality in diabetes: the Wisconsin Epidemiologie Study of Diabetic Retinopathy. Metabolism 1992;41:194-197 Barr SI, Kottke BA, Mao SJT. Improved method for determination of triglycérides in plasma lipoproteins by an enzymic kit method [letter]. Clin Chem 1981;27:1142-1144 Grillo F, Izzo C, Mazzotti G, Murador E. Improved method for determination of high-density-lipoprotein cholesterol. II. Enzymic determination of cholesterol and high-density lipoprotein fractions with a sensitive reagent. Clin Chem 1981;27:375-379 Izzo C, Grillo F, Murador E. Improved method for determination of high-density-lipoprotein cholesterol. I. Isolation of high-density lipoproteins by use of polyethylene glycol 6000. Clin Chem 1981; 27:371-374 Willey DG, Rosenthal MA, Caldwell S. Glycosylated haemoglobin and plasma glycoprotein assays by affinity chromatography. Diabetologia 1984;27:56-58 Joven J, Vilella E. Serum levels of lipoprotein(a) in patients with well-controlled non-insulin-dependent diabetes mellitus [letter]. JAMA 1991;265:1113-1114 Hughes TA, Clements RS, Fairclough PK, Bell DSH, Segrest JP. Effect of insulin therapy on lipoproteins in non-insulin dependent diabetes mellitus (NIDDM). Atherosclerosis 1987;67:105-114 Haffner SM, Tuttle KR, Rainwater DL. Lack of change of lipoprotein (a) concentration with improved glycémie control in subjects with type II diabetes. Metabolism 1992;41:116-120 Davies M, Rayman G, Day J. Increased incidence of coronary disease in people with impaired glucose tolerance: link with increased lipoprotein(a) concentrations? BMJ 1992;304:1610-1611 Sandholzer C, Sana N, Kark JD, Rees A, Jaross W, Dieplinger H, et al. Apo(a) isoforms predict risk for coronary heart disease: a study in six populations. Arterioscler Thromb 1992;12:1214-1226 Borch-Johnsen K, Kreiner S. Proteinuria: value as predictor of cardiovascular mortality in insulin dependent diabetes mellitus. BMJ 1987;294:1651-1654 Kapelrud H, Bangstad H-J, Dahl-J0rgensen K, Berg K, Hanssen KF. Serum Lp(a) lipoprotein concentrations in insulin dependent diabetic patients with microalbuminuria. BMJ 1991;303:675-678 Jenkins AJ, Steele JS, Janus ED, Best JD. Increased plasma apolipoprotein(a) levels in IDDM patients with microalbuminuria. Diabetes 1991;40:787-790 Takegoshi T, Haba T, Hirai J, Kitoh C, Saga T, Yamazaki Y, et al. Alterations of lipoprotein(a) in patients with diabetic nephropathy [letter]. Atherosclerosis 1990;83:99-100 Jay RH, Jones SL, Hill CE, Richmond W, Viberti GC, Rampling MW, et al. Blood rheology and cardiovascular risk factors in type 1 diabetes: relationship with microalbuminuria. Diabet Med 1991; 8:662-667