Alteration of lipoprotein(a) concentration with glycemic control in non-insulin-dependent diabetic subjects without diabetic complications

Alteration of Lipoprotein(a) Concentration With Glycemic Control in Non-Insulin-Dependent Diabetic Subjects Without Diabetic Complications Hiroshi Nakata, Kazutoshi

Horita, and Masaaki Eto

Recently, a high plasma level of lipoprotein(a) [Lp(a)] has been considered an independent risk factor for atherosclerosis and its sequelae, particularly myocardial infarction. Patients with non-insulin-dependent diabetes mellitus (NIDDM) have an increased mortality rate from cardiovascular and cerebrovascular disease. Therefore, plasma concentrations of Lp(a) were determined and the relationship between fasting plasma Lp(a) level and diabetic control was investigated in NIDDM patients without any diabetic complications. Fasting plasma Lp(a) levels were measured using enzyme-linked immunosorbent assay kits (Terumo Medical Corp. Elkton, MD, Lp(a)] in 61 NIDDM subjects (30 men aged 56 +: 2.0 years, 31 women aged 53 + 2.1 years [mean + SEMI) who were without any diabetic macroangiopathy and microangiopathy such as retinopathy, nephropathy, and neuropathy and in 56 healthy age- and sex-matched controls. Plasma Lp(a) levels were significantly higher in the diabetic group than in the control group (23.5 f 2.5 v 11.7 f 1.4 mg/dL [mean f SEMI, P < .OOl). There was no significant correlation between log-transformed plasma Lp(a) levels and other factors such as age, sex, body mass index (BMI), blood pressure, duration of diabetes, fasting plasma glucose (FPG) level, glycosylated hemoglobin (HbA,c) level, and plasma lipid levels except for low-density lipoprotein cholesterol (LDL-C) levels in diabetic patients. A significant positive correlation was noted in diabetic patients between the changes of log Lp(a) and HbAlc levels after a 3-month follow-up period (P < .05). Our data indicate that NIDDM patients without any diabetic complications have high concentrations of plasma Lp(a) compared with healthy control subjects, and that improved glycemic control may reduce plasma levels of Lp(a). Copyright 0 1993 by W. B. Saunders Company

L

IPOPROTEIN(a) [Lp(a)], first described by Berg’ in 1963, is a low-density lipoprotein (LDL)-like substance with a specific apoprotein, apoprotein(a) [ape(a)], bound to apo B-100 by disulfide bridges (as recently reviewed2x3). The plasma level of Lp(a) varies from less than 1 mg/dL to more than 100 mg/dL in humans. From its properties including homology of partial amino acid sequence4 and cDNA sequence of ape(a) to plasminogen5 and its inhibition of the activation of plasminogen,6 Lp(a) is considered one of the risk factors of atherosclerosis. Recently, a high level of plasma Lp(a) has been reported to be an independent risk factor of ischemic heart disease, particularly myocardial infarction.‘-‘” In patients with diabetes mellitus, it is well known that lipoproteins and apoproteins have altered production and turnover rates and the mortality rate from cardiovascular and cerebrovascular disease is increased.” However, reports on Lp(a) levels in non-insulin-dependent diabetes mellitus (NIDDM) are few. In 1983, plasma Lp(a) levels in both insulin-dependent diabetes mellitus (IDDM) and NIDDM patients were first examined by Schernthaner et al;lz they noted the higher frequency of hyperlipoproteinemia of Lp(a) in diabetic patients. In 1987, Hughes et alI3 also determined plasma Lp(a) and lipoprotein levels and found that plasma Lp(a) levels were unaltered with insulin therapy. In 1990, significant reductions in plasma Lp(a) levels were noted following an improvement in plasma glucose levels in IDDM patients with high levels of serum Lp(a) by Bruckert et alI4 and in 1991 by Haffner et al.15 However, in 1992, Haffner et all6 reported a lack of effect of improved glycemic control on Lp(a) levels in NIDDM patients. Therefore, plasma concentrations of Lp(a) in patients with NIDDM without any diabetic complications were determined and their alterations with glycemic control were investigated in the present study. Metabolism,

Vol42, No 10 (October), 1993: pp 1323-1326

SUBJECTS AND METHODS Subjects Sixty-one NIDDM patients (30 men aged 56 lr 2.0 years, 31 women aged 53 ? 2.1 years [mean ? SEMI) without any diabetic complications such as retinopathy, nephropathy, neuropathy, and macroangiopathy and 54 healthy age- and sex-matched controls were studied. In diabetic patients, 31 were controlled with diet alone, 22 with an oral hypoglycemic agent, and eight with insulin. None of them had hepatic disease. Diabetic retinopathy was diagnosed according to changes of fundi as determined by an ophthalmologist’s examination. Diabetic nephropathy was diagnosed by the level of albuminuria, which was defined as the mean value of alubumin excretion rate (AER) in three timed overnight urine collections (normoalbuminuria, AER < 20 kg/min; microalbuminuria, AER = 20 to 200 kg/min; microalbuminuria, AER > 200 kg/min); only patients with normoalbuminuria were studied. Diabetic subjects with hypercreatininemia (serum creatinine > 1.4 mg/dL, renal insufficiency) were excluded from this study. Diabetic neuropathy was diagnosed when a diabetic subject was aware of neuropathic symptoms such as numbness, psychoesthesia, and pain in upper or lower extremities. Macroangiopathy was diagnosed when a diabetic subject had at least one of the following conditions: (1) a history of myocardial infarction characterized by a typical clinical picture, electrocardiogram (ECG) alterations, and enzymatic changes; (2) ischemic alterations in ECG (Minnesota code I, 1-3; IV, 1-4; V, l-3) with or without chest pain; (3) a history of cerebral infarction or transient ischemic attack characterized by a typical clinical picture; (4) brain computed tomographic changes; or (5) a history of arteriosclerosis obliterans characterized by a typical clinical picture. From the Second Department of Internal Medicine, Asahikawa Medical College, Asahikawa, Hokknido, Japan. Submitted September 1, 1992; accepted December 19, 1992. Address reprint requests to Hiroshi Nakata, MD, Second Depatiment of Internal Medicine, Asahikawa Medical College, Nishikagura 4-5-3, Asahikawa, Hokkaido, Japan 078. Copyright 0 1993 by W.B. Saunders Company 0026049519314210-0015$03.00/0 1323

NAKATA, HORITA, AND ET0

1324

WwW

Fasting plasma glucose (FPG) level was measured by the glucose oxidase method, and glycosylated hemoglobin (HbArc) level was determined by high-performance liquid chromatography (HPLC).” Serum creatinine level was measured by the routine method using the Jaffe reaction.tx Plasma total cholesterol (TC) and triglyceride (TG) levels were measured enzymatically. Plasma high-density lipoprotein cholesterol (HDL-C) level was determined enzymatitally following dextran sulfate/magnesium chloride precipitation. Plasma low-density lipoprotein cholesterol (LDL-C) level was calculated using the formula of Friedewald et al.lY Blood pressure was measured with a standard clinical sphygmomanometer (25 x J2-cm cuff); the mean of three measurements within a few months was used. Body mass index (BMI) was calculated as weight/height? (kg/m’).

1201

100

I

MeenS.E.Y.

* pco.oo1

Measurement of Plasma Lp(a) Blood samples were collected in EDTA from an antecubital vein after an approximate 12-hour overnight fast. Plasma samples were separated by centrifugation and stored at -80°C for up to 1 month until measurement. The plasma concentration of Lp(a) was measured by an enzyme-linked immunoassay technique using a monoclonal anti-Lp(a) antibody [Terumo Medical Corp, Elkton, MD Lp(a)]; the intraassay coefficient of variation (CV) was 7%. the interassay CV was 9%, and no cross-reactivity with plasminogen. LDL, very-low-density lipoprotein. or HDL was observed in this assay.?” Concentrations of plasma Lp(a), FPG. HbAlc. and lipids were remeasured after 3 months to examine the change from basal levels in all patients.

Statistics The nonparametric Mann-Whitney U test was used to compare the distribution of plasma Lp(a) levels in diabetic patients with those in control subjects. The Spcarman rank correlation coefficient was used to analyze the relationship between log-transformed plasma Lp(a) levels and other factors and the changes between them.

RESULTS Table 1 shows clinical characteristics and laboratory data of control subjects and diabetic patients without diabetic Table 1. Clinical Characteristics and Laboratory Data in Control and

Control wow

Diabetic grow

Fig 1. Plasma Lp(a) concentrations in control and diabetic subjects. Plasma Lp(a) levels in the diabetic group were significantly higher than in the control group (23.5 * 2.5 v 11.7 lr 1.4 mg/dL, P < ,001).

complications. No significant differences were found in age, BMI. blood pressure, and plasma lipid levels between both groups of subjects. Figure 1 shows plasma Lp(a) levels in control and diabetic subjects; Lp(a) concentrations were significantly higher in diabetic patients than in controls (23.5 + 2.5 v 11.7 % 1.4 mg/dL [mean & SEMI, P < .OOl). There was no significant difference in Lp(a) levels between diabetic patients in terms of sex and mode of therapy (Fig 2). Table 2 shows simple correlation coefficients between log-transformed plasma Lp(a) levels and other factors in diabetic patients. There was no correlation between log Lp(a) and age, BMI, duration of diabetes. blood pressure, and glycemic control as assessed by FPG and HbAic levels. Although a significant correlation was not found between log Lp(a) and TC. TG, and HDL-C levels, there was a slight positive correlation between log Lp(a) and LDL-C levels

Diabetic Subjects

Control n

(M/F)

Diabetic

38116

30131

51.0 & 1.2

54.5 + 0.7

22.5 2 0.4

23.5 + 0.5

Systolic

126 t- 1.9

129 2 2.0

Diastolic

77 * 1.4

80 +- 1.1

Age WI BMI (kg/m*) Blood pressure (mm Hgl

Plasma lipids (mg/dL) TC

196.2 t 4.2

195.7 + 4.5

TG

113.8 +- 7.7

116.5 -t 9.2

HDL-C

45.2 it 2.1

57.6 + 2.4

LDL-C

131.7 f 8.8

116.7 ? 4.2

FPG (mgidt)

94.5 * 1.1

153.6 2 5.8*

5.5 + 0.1

8.4 + 0.2*

HbA,, (%)

mgldl

‘= g g s :

30-

0” c ;._

20-

9 8 .g J

*: not significant

40f G--IA -r

lo-

Mode of therapy(n) Diet alone

31

Oral hypoglycemic agent

22

Insulin

8

NOTE. Values are means + SEM. FPG and HbA,, levels in the diabetic group were significantly higher than in the control group. lP

< .OOl “control group.

Fig 2. Plasma Lp(a) concentrations by gender and by different modes of therapy in diabetic subjects. There was no significant difference in Lp(a) levels.

1325

Lp(a) LEVEL WITH GLYCEMIC CONTROL IN NIDDM

Table 2. Simple Correlation Between Log-Transformed

Table 4. Simple Correlation Between the Changes of

Plasma Lp(a)

Log-Transformed

Levels and Other Factors in Diabetic Patients

iv

Plasma Lp(a) Levels and the Changes of Other

Factors in Diabetic Subjects After a 3-Month Follow-up Period

P

Age

.18

.116

R’

P

BMI

,139

29

AFPG

.04

.773

,178

AHbA,,

,283

,027

ATC

,052

,692

-.176

Duration of diabetes Blood pressure Systolic

,103

,432

ATG

-.112

,413

Diastolic

,079

,548

AHDL-C

-.071

,586

,497

ALDL-C

-.093

FPG

,134

HbA,, Lipids

.069

TC

,304 ,597

TG

-.161

,231

HDL-C

-.021

,873

,256

LDL-C

,054

(r,= .256, P = .054). Table 3 shows laboratory data before and after a 3-month follow-up period in the diabetic group; there was no significant change between these periods. Table 4 shows simple correlation coefficients between the changes of log-transformed plasma Lp(a) levels [Alog Lp(a)] and the changes of FPG, HbAic, and plasma levels . of hptds (AFPG, AHbAic, ATC, ATG, AHDL-C, and ALDL-C) after a 3-month follow-up period. Although Alog Lp(a) did not correlate with AFPG, ATC, ATG, AHDL-C, and ALDL-C, Alog Lp(a) did show a significant correlation with AHbAic (R2 = .283, P = .027; Fig 3). DISCUSSION

In 1983, plasma Lp(a) levels in diabetes mellitus were first examined by Schernthaner et al.12 Plasma Lp(a) levels were compared between 84 IDDM and NIDDM patients and 71 normal subjects; they found that the percentage of subjects with high Lp(a) levels (> 20 mg/dL) was 14% for the diabetic group but no more than 5% for the normal subject group, although there was no significant difference. They were the first to report the higher frequency of hyperlipoproteinemia of Lp(a) in diabetic patients. Furthermore, by continued evaluation of the diabetic patients, Schernthaner et alI2 reported that plasma Lp(a) levels did not correlate with HbAlc levels and that treatment with insulin did not produce any change in plasma Lp(a) levels. In 1987, Hughes et all3 also determined plasma Lp(a) and lipoprotein levels for 26 weeks after the start of insulin

.073

,242

NOTE. The change of log Lp(a) [Alog Lp(a)] correlated significantly with the change of HbA,, (AHbA,,).

treatment in 30 NIDDM patients, and reported that plasma Lp(a) level was unaltered with insulin therapy. In Japan, mean plasma Lp(a) levels in normal subjects were reported to range from 10 to 19 mg/dL, whereas those in diabetic patients were reported to range from 18 to 29 mg/dL.21 The present study using diabetic patients in our laboratory showed an approximate twofold higher plasma Lp(a) level in NIDDM patients than in normal subjects, although no difference was observed for sex or age. A significant reduction in plasma Lp(a) level was noted following an improvement in plasma glucose level in 10 poorly controlled IDDM patients with high levels of serum Lp(a) by Bruckert et all4 and in 12 IDDM patients by Haffner et a1.15However, in 1992 Haffner et alI6 reported a lack of decrease in Lp(a) levels with improved glycemic control in NIDDM patients. In our study, although no correlation was evident between log-transformed plasma Lp(a) levels and HbAic levels, a significant positive correlation was noted between the changes of log Lp(a) and HbAic levels after a 3-month follow-up period. Plasma Lp(a) level, which is hereditarily regulated, varies greatly among individuals. Log-transformed plasma Lp(a) levels did not show any correlation with HbAlc levels in the diabetic group as a whole. However, Lp(a) and HbAlc levels are speculated to increase from the respective basal ranges under the condition of high plasma glucose level, showing significant positive correlations when examined in W?aW =

0.4 -I

y=O.O3x+O.O5 , rs=0.283

(~~0.027)

Table 3. Laboratory Data Before and After a 3-Month Follow-up Period in the Diabetic Group Before

154.2 k 5.7

8.4 + 0.2

8.4 + 0.2

TC

195.7 2 4.5

201.1 * 4.5

TG

116.5 k 9.2

111.7 + 6.2

HbA,, (%)

153.6

After

k 5.8

FPG (mg/dL) Plasma lipids (mg/dL)

HDL-C

57.6 k 2.4

54.8 + 1.7

LDL-C

116.7 + 4.2

123.9 z!z4.1

23.5 + 2.5

26.8 k 2.7

Plasma Lp(a) (mg/dL)

NOTE. Values are means + SEM. There was no significant change in laboratory data.

’

-0.6 1 -10

-6

-6

-4

-2

0

2

Change of HbAlc Fig 3. Relationship between the changes of log-transformed Lp(a) levels and the changes of HbAlc levels in diabetic subjects after a 3-month follow-up period. The line represents the regression through data.

NAKATA,

1326

terms of their changes. Haffner et alih described several possibilities for the lack of decrease of Lp(a) level with improved glycemic control, including the small number of subjects and the short evaluation period in their study. In another study (unpublished), we found no correlation between the changes of log Lp(a) and HbAlc levels after a l-month follow-up period. Interestingly, it suggests the involvement of high plasma glucose levels in the occurrence of high plasma Lp(a) levels in diabetic patients. The mechanism by which plasma Lp(a) level increases in diabetes has not been well-clarified. In 1980, Krempler et alz2 reported in their study on the turnover of Lp(a) that the catabolic rate of Lp(a) was nearly constant, but that in patients with high levels of Lp(a) the synthetic rate was elevated in humans. It is assumed that in diabetes mellitus increased synthesis and/or decreased catabolism of Lp(a) leads to an increase in plasma Lp(a) level through an unknown mechanism. It is not yet known how a high plasma glucose level produces a high level of Lp(a). Hyperlipidemia is a complication frequently recognized in diabetes. The plasma Lp(a) level is not affected by the plasma lipid level and, unlike other lipoproteins, Lp(a) undergoes a different mechanism of metabolism, as was

HORITA, AND ET0

reported by Krempler et al.‘j Very recently, the study on dyslipidemia and Lp(a) in 1,065 cases by Boyer et alZJ showed no correlation between plasma Lp(a) level and plasma TC. TG, HDL-C, LDL-C, and apolipoprotein B levels. The results of our study were nearly comparable with the findings of Boyer et al.?“ The difference was a slight positive correlation observed between log-transformed plasma Lp(a) levels and LDL-C levels. Lp(a) is an LDL-like substance with an aspecific apoprotein, ape(a), bound to apo B-100 by disulfide bridges. Thus there is a possibility of LDL levels also being reIatively increased in diabetic patients with high levels of Lp(a), producing a correlation between them. Plasma Lp(a) levels are significantly higher in diabetic patients without any complications than in normal subjects. Our study suggests that a high plasma glucose level is probably one of the causes of a high plasma Lp(a) level. From

its properties,

Lp(a)

is considered

dent

risk factors

of atherosclerosis;

that

improvement

of glycemic

the development Lp(a)

one of the indepen-

therefore control

of atherosclerosis

it is presumed

may possibly by reducing

prevent

the plasma

level.

REFERENCES 1. Berg K: A new serum type system in man-The Lp system. Acta Pathol Microbial Stand 59:369-382, 1963 2. Utermann G: The mysteriesof lipoprotein(a). Science 246:904910, 1989 3. Scanu AM, Fress GM: Lipoprotein(a): Heterogeneity and biological relevance. J Clin Invest 85:1709-1715, 1990 4. Eaton DL, Fless GM, Scanu AM, et al: Partial amino acid sequence of apolipoprotein(a) shows that it is homologous to plasminogen. Proc Nat1 Acad Sci USA 84:3224-3228. 1987 5. McLean JW, Tomlinson JE. Lawn RM, et al: cDNA sequence of human apolipoprotein(a) is homologous to plasminogen. Nature 300:132-137, 1987 6. Edelberg JM, Gonzalez-Gronow M. Pizzo SV: Lipoprotein a inhibits streptokinase-mediated activation of human plasminogen. Biochemistry 28:2370-2374, 1989 7. Kostner GM, Avogaro P. Cazzolato G, et al: Lipoprotein Lp(a) and the risk for myocardial infarction. Atherosclerosis 3851-61, 1981 8. Rhoads GG, Dahlen G, Berg K. et al: Lp(a) lipoprotein as a risk factor for myocardial infarction. JAMA 256:2540-2544, 1986 9. Hoefler G. Harnoncourt F, Kostner GM: Lipoprotein(a)-A risk factor for myocardial infarction. Arteriosclerosis 8:398-401, 1988 10. Seed M, Hopplichler F. Reavely D, et al: Relation of serum lipoprotein(a) concentration and apoprotein(a) phenotype to coronary heart disease in patients with familial hypercholesterolemia. N Engl J Med 322:1494-1499, 1990 11. Kannel WB, McGee DL: Diabetes and cardiovascular risk factors: The Framingham Study. Circulation 59:8-13. 1979 12. Schernthaner G, Kostner GM, Dieplinger H, et al: Apolipoproteins (A-l. A-II, B), Lp(a) lipoprotein and lecithin: Cholesterol acyltransferase activity in diabetes mellitus. Atherosclerosis 49:277293.1983 13. Hughes TA, Clements RS, Fairclough PK. et al: Effect of

insulin therapy on lipoproteins in non-insulin-dependent diabetes mellitus (NIDDM). Atherosclerosis 67:105-l 14. 1987 14. Bruckert E. Davidoff P, Grimaldi A, et al: Increased serum levels of lipoprotein(a) in diabetes mellitus and their reduction with glycemic control. JAMA 26335-36, 1990 15. Haffner SM, Tuttle KR. Rainwater DL: Decrease of lipoprotein(a) with improved glycemic control in IDDM subjects. Diabetes Care 14:302-307. 1991 16. Haffner SM. Tuttle KR, Rainwater DL: Lack of change of lipoprotein(a) concentration with improved glycemic control in subjects with type II diabetes. Metabolism 41:116-120, 1992 17. Cole RA, Soeldner JS, Dunn PJ, et al: A rapid method for the determination of glycosylated hemoglobins using high-pressure liquid chromatography. Metabolism 27:289-301, 1978 18. Husdan H, Rapoport A: Estimation of creatinine by the Jaffe reaction. Clin Chem 14:222-238, 1968 19. Friedewald WT. Levy RI, Fredrickson DS: Estimation of the concentration of low-density lipoprotein cholesterol in plasma. without use of the preparative ultracentrifuge. Clin Chem 18499. SO?. 1972 20. Silberman SR, Armentrut MA, Saha AL: Macra Lp(a) for quantitation of human lipoprotein(a) by enzyme linked immunoassay. Clin Chem 36:951, 1990 (abstr) 71. Takegoshi T, Haba T, Mabuchi H. et al: Alterations of lipoprotein(a) in patients with diabetic nephropathy. Atherosclerosis X3:99-100, 1990 22. Krempler F. Kostner GM. Bolzano K. et al: Turnover of lipoprotein(a) in man. J Clin Invest 65:1483-1490, 1980 23. Krempler F, Kostner G. Bolzano K, et al: Lipoprotein(a) is not a metabolic product of other lipoproteins containing apolipoprotein B. Biochim Biophys Acta 575:63-70, 1979 24. Boyer H. De Gennes JL. Truffert J, et al: Lp(a) levels in different types of dyslipidemia in the French population. Atherosclerosis 85:61-69. 1990