LDL particle size in mildly hypertriglyceridemic subjects: no relation to insulin resistance or diabetes

i

ATHEROSCLEROSIS Atherosclerosis 113 (1995)227-236

LDL particle size in mildly hypertriglyceridemic subjects: no relation to insulin resistance or diabetes S. Lahdenperga,

T. Sanea, H. Vuorinen-Markkolab,

P. Knudsen”,

M.-R. Taskinen*”

aThird Department of Medicine, University of Helsinki, Haartmaninkatu 4, FIN-00290 Helsinki, Finland ‘Second Department of Medicine, University of Helsinki, Helsinki, Finiand

Received 2 June 1994; revision received 3 October 1994; accepted IO October 1994

Abstract

We examined 18 Type 2 diabetic and 19 non-diabetic subjects in order to determine the association between insulin resistance and LDL particle size distribution in mildly hypertriglyceridemic and hyperinsulinemic subjects with and without Type 2 diabetes. Insulin sensitivity of the patients was characterized by their insulin-stimulated glucose uptake rate determined by euglycemic clamp technique. LDL particle size distribution was determined by nondenaturing polyacrylamide gradient gel electrophoresis. Type 2 diabetic and non-diabetic subjects had closely similar serum lipid and lipoprotein concentrations as well as the mean particle diameters of the major LDL peak (246 i 6 A and 244 f 6 A, respectively). To evaluate the effect of insulin resistance on LDL particle size the participants were categorized into two subgroups using the median of their insulin-stimulated glucose uptake rate (14.67 ~mol/kglmin) as a cut-off point. Neither lipid and lipoprotein concentrations nor the LDL particle size distributions differed between the more insulin resistant group (nine diabetic and nine non-diabetic subjects) and less insulin resistant group (nine diabetic and ten non-diabetic subjects). LDL particle size was not associated with the insulin-stimulated glucose uptake rate or with the mean 24-h concentration of serum insulin. Mean 24-h concentration of serum triglycerides was the strongest discriminator for LDL particle size (r = -0.44, P < 0.01). In conclusion, neither Type 2 diabetes nor insulin resistance seem to have any direct effect on LDL particle size in mildly hypertriglyceridemic subjects. The fact that LDL particle size was associated with serum triglycerides indicates that the effect of diabetes and insulin resistance on LDL particle size could be explained by the effects of insulin resistance and/or hyperinsulinism on VLDL metabolism. Keywordr:

Type 2 diabetes; Insulin resistance; Blood lipids; Hypertriglyceridemia;

1. I~ion Resistance

Small dense LDL

characteristics of syndrome X [ 11, more commonly referred to as metabolic syndrome or insulin resisto insulin-mediated

glucose uptake,

hyperinsulinemia, glucose intolerance, hypertension, dyslipidemia and abdominal obesity are the * Corresponding author.

tance syndrome [2]. New components of the metabolic syndrome are increased concentrations of plasminogen activator inhibitor-l and serum urate [1,3,4]. Insulin resistance is associated with elevations of serum total and VLDL

0021-9150/95/SO9.50 0 1995 Elsevier Science Ireland Ltd. All rights reserved SSDI 0021-9150(94)05450-W

triglycerides

and

228

S. Lahdenperii

et al. /Atherosclerosis

lowering of HDL cholesterol concentration, but data concerning LDL concentration is contradictory [5-lo]. Recently, LDL subclass phenotype B, characterized by predominance of small dense LDL particles, has emerged as a new component of the metabolic syndrome [ 111. Small dense LDL is a common feature in hypertriglyceridemic patients who have also been shown to be insulin resistant, glucose intolerant and hyperinsulinemic [ 121. The clustering of high triglycerides, low HDL and small dense LDL represents the atherogenic lipoprotein phenotype [ 131. Epidemiological studies have demonstrated an excess risk for cardiovascular disease in metabolic disorders clustered in the insulin resistance syndrome [ 14-161. Austin et al. first described two distinct LDL subclass phenotypes based on the LDL particle size distribution separated by gradient gel electrophoresis [ 151. Pattern A consists of the major LDL peak with LDL particle diameter more than 255 A, and of minor peak with smaller particle size [15]. In pattern B, small dense LDL is preponderant and the diameter of the major LDL peak is equal to or less than 255 A [ 151. Pattern B is associated with accelerated atherosclerosis and increased risk of myocardial infarction [ 151. The exact mechanism determining the properties of LDL particles is not well established. The fact that LDL particle size is genetically influenced is supported by data from complex segregation analyses in families, hereditability analyses in twins and linkage analyses [17-201. Substantial data suggesting that non-genetic factors modify the expression of the phenotype have recently emerged. Several studies have shown that prevalence of small dense LDL is associated with elevations of serum total and VLDL triglycerides and low HDL cholesterol concentration [2 I-241. Other environmental factors, such as diet, exercise, obesity, and use of /3-adrenergic blockers, have also been shown to influence on LDL particle size [25-281. Recent reports have demonstrated that small dense LDL particles (pattern B) are common in Type 2 diabetic patients 129,301. The aim of this study was to investigate the association between the degree of insulin resistance and LDL particle size distribution in mildly hypertriglyceridemic and hyperinsulinemic subjects with and without Type 2 diabetes mellitus.

I13 (I 995) 227-236

We specifically aimed to evaluate whether small dense LDL is linked to diabetes and/or insulin resistance, or related to hypertriglyceridemia, which is a common component in these two disorders. 2. Methods 2.1. Material Eighteen Type 2 diabetic and 19 non-diabetic subjects with hyperinsulinemia participated in the study. The subjects were recruited for an intervention study which addressed the effects of gemlibrozil treatment on lipoprotein and glucose metabolism [31-331. To be eligible for the study the participants had to have their serum triglyceride concentrations between 1.5 and 4.5 mmohl, fasting serum insulin ~60 pmol/l and BMI less than 30 kg/m* on screening. The mean serum triglyceride concentration at the 6-week pretreatment visit was 2.89 i 1.34 mmol/l in the diabetic and 3.01 f 1.32 mmol/l in the nondiabetic group. All the non-diabetic subjects had normal HbA,, and normal fasting blood glucose concentration, Subjects with abnormal hepatic, kidney or thyroid function tests, microalbuminuria > 30 &mm or unstable angina or MI within 6 months were excluded from the study. None of the participants received lipid-lowering drugs before the study. History of hypertension was received from one diabetic and two non-diabetic subjects. All Type 2 diabetic patients fullfilled the WHO criteria for diabetes [34]. Type 2 diabetic patients were treated with oral hypoglycemic drugs: eight with sulfonylurea alone, seven with sulfonylurea and metformin, one with sulfonylurea and guargum, and two with sulfonylurea, metformin and guargum. None of the Type 2 diabetic patients showed clinical signs of micro- or macrovascular diabetic complications. Four Type 2 diabetic and five non-diabetic subjects were smokers. During the run-in period of 6 weeks all study participants received counselling on diet and restriction of alcohol. All subjects gave their written informed consent to participate in the study. The study protocol was approved by the Ethics Committee of the Second Department of Medicine, Meilahti Hospital, University of Helsinki. Selected characteristics of the subjects are shown in Table 1.

S. Luhdenperii et al. /Atherosclerosis 113 (1995) 227-236 Table 1 Clinical characteristics (mean f SD.) of the subjects

n se% WV Age Clean BMI (kg/m2) fB-glucose (mmolll) HM,, (%)

c-peptide (Mloy1) S-Insulin (pmol/l) Glucose uptake rate ~mol/I@nin) Systolic BP (mmHg) Diastolic BP (mm&d Medication Beta blocker Cathannel blocker ACE-inhibitor

Type 2 diabetic subjects

Non-diabetic subjects

18 16/2

19 18/l

f it f f f f 16.61 f 55.9 21.4 9.2 7.6 0.88 68.4

10.3 2.3 1.7* 0.9* 0.31** 34.8

9.06

51.9 27.9

l

5.3 1.24 16.8

l

7.4

zt 1.8 5.0 * 0.6 0.3

f 0.46 f 33.6 19.22 zt 10.06 134 f 15 84 zt I

148 i 14” 91 f 12**

2

I

1

5

1

0

n, number of subjects; M, male; F, female; BMI, body mass index; HM,,, glycosylated hemoglobin; BP, blood pressure. *P < 0.001 vs. non-diabetic subjects. **P < 0.05 vs. non-diabetic subjects.

2.2. Metabolic tests

The subjects were admitted to the metabolic ward for the metabolic tests. Blood glucose, serum insulin and triglyceride concentrations were measured in blood samples withdrawn at 07:30 h, 11:30 h, 14:00 h, 16:00 h, 20:00 h, 2400 h. and at 04:OOh and 08:OO h in order to determine the mean 24-h concentrations of the variables. Insulin sensitivity of the patients was determined by using a two-step insulin clamp technique as previously described [31,35]. The insulin-stimulated glucose uptake rate expressed per kg of body weight, i.e. insulin sensitivity (normal value 34.2 f 4.2 ~mol/kg/min, mean f S.E.; Ref. [36]), was determined from the glucose infusion rate required to maintain normoglycemia during an insulin infusion rate of 1.0 mU/kgAnin (240-300 mm). The average glucose and insulin levels during the hyperinsulinemic period of the clamp (240-300 mm) were 6.7 i 1.5 mmol/l and 471.7 f 89.7 pmoVl in the diabetic group and 5.1 f 0.1 mmol/l and (P c 0.001 vs. diabetic group) and 529.2 f 92.3 pmol/l (P = NS vs. diabetic group)

229

in the non-diabetic group, respectively. The concentrations of blood glucose, glycosylated hemoglobin, serum C-peptide (normal value 0.50 f 0.03 nmol/l), insulin (normal value 36 * 6 pmol/l; Ref. [36]) and free insulin were determined as previously described [3 11. 2.3. Blood pressure

Blood pressure was measured after 5-10 min rest in the sitting position with a random zero spyghmomanometer (Random Zero Spyghmomanometer: Hawksley & Sons, Lancing, UK). 2.4. Lipids and lipoproteins

All blood samples for the lipid and lipoprotein analyses were collected in the morning after a 12-h fast. Serum was isolated immediately by centrifugation at 3000 rev./mm for 30 min at 4°C. Serum lipoprotein fractions (VLDL, LDL, HDL) were separated by sequential ultracentrifugation method [32,37]. LDL peak particle size was determined by nondenaturing polyacrylamide gradient gel electrophoresis using the method previously described [32,38]. Postheparin plasma lipoprotein and hepatic lipase activities were measured after an intravenous bolus of heparin (100 IU/kg body weight) as previously described [39]. The concentrations of serum lipids and lipoproteins were determined by using laboratory analyses previously depicted [32]. 2.5. Statistical analyses

The significances of differences between the groups were assessed with the Mann-Whitney non-parametric test (3s) using BMDP statistical software (University of California Press, 1988). The relationships between the variables were estimated with Pearson’s correlation coefficients (program 6D). In the case of non-normal distribution, log-transformation of the data was performed. Multiple stepwise regression analysis was performed using BMDP program 2R. 3. Results 3.1. Effect of Type 2 diabetes on metabolic parameters

Table 1 depicts the concentrations of fasting blood glucose and glycosylated hemoglobin levels

230

S. Ldtaknperti et al. /Atherosclerosis II3 (1995) 227-236

in the two groups. Fasting serum C-peptide concentration was decreased in Type 2 diabetic patients compared with the level in non-diabetic subjects (P c 0.05), indicating reduced insulin secretion in diabetic patients (Table 1). However, fasting serum insulin concentrations did not differ between the groups. Both diabetic and nondiabetic patients had also similar rates of insulinstimulated glucose uptake indicating that their degree of insulin resistance was comparable (Table 1). The ranges of insulin-stimulated glucose uptake rates were from 7.78 to 41.06 ~mol/kg/min in the diabetic and from 9.06 to 42.78 ~mol/kg/min in the

f ‘*.8

8

24 16 Time (hours)

8

Fig. 1. Twenty-four-hour profdes of blood glucose (upper panel), serum insulin (middle panel) and serum triglyceride (lower panel) concentrations in Type. 2 diabetic (0) and nondiabetic (0) subjects.

non-diabetic group. Diabetic patients treated with sulfonylurea only had lower HbAI, concentration than the diabetic patients receiving sulfonylurea combined with metformin and/or guargum (6.96 * 0.95 vs. 8.12 * 0.54%, P < 0.05). However, the insulin-stimulated glucose uptake rates were comparable in diabetic patients on sulfonylurea and on a combination therapy (14.84 i 6.50 vs 18.07 f 10.84 pmoll kg/mm, respectively, P = 0.79). Likewise, the serum lipid and lipoprotein concentrations did not differ between the diabetic patients treated with sulfonylurea only or with a combination of metformin and/or guarem (data not shown). Diurnal metabolic tests. Fig. 1 shows 24-h profiles of blood glucose, serum insulin, and serum triglyceride concentrations in the two groups. The mean 24-h concentration of the blood glucose was significantly higher in Type 2 diabetic patients than in non-diabetic subjects (10.3 i 2.2 vs. 5.0 f 0.3 mmolil, respectively, P < 0.001). In contrast, the mean 24-h concentrations of the serum insulin as well as triglycerides were similar in the Type 2 diabetic and non-diabetic groups (171.0 f 84.6 vs 183.6 f 80.4 pmol/l and 3.15 * 0.95 vs. 3.08 f 1.07 mmolil, respectively). A significant negative correlation existed between the insulin-stimulated glucose uptake and mean 24-h concentration of the serum insulin concentration (r = -0.48, P < 0.01). Mean 24-h concentration of serum triglycerides did not correlate with the insulin-stimulated glucose uptake nor with the mean 24-h concentration of serum insulin (t = -0.06, P= 0.74 and r = -0.03, P= 0.87, respectively). Serum lipids and lipoproteins. The concentrations of serum lipids and lipoproteins were closely similar in Type 2 diabetic and non-diabetic groups (Table 2). Both groups were mildly hypertriglyceridemic and they had lowering of HDLcholesterol concentration (Table 2). Hepatic lipase activity was slightly lower in Type 2 diabetic patients than in non-diabetic subjects (P = 0.05, Table 2). Lipoprotein lipase activity was comparable in the two groups. The particle diameters of the major LDL peak were 246 and 244 A in the two groups, indicating the preponderance of small dense LDL particles. Of note is that there were no

S. Luhaknpera”

et al. /Atherosclerosis

Table 2 Lipids and lipoproteins (mean f S.D.) of the study groups --

;ot-Chol (mmol/l) Tot-Tg (mmobl) HDL-Chol (mmohl) LDLChol (mmohl) VLDL-Chol (=oW VLDLTg (mmohl) Apo 3 (mg/dl) LDL particle diameter (A) Hepatic lipase @ol FFAMIh) Lipoprotein lipase bmol FFA/ml/h)

Type 2 diabetic subjects

Nondiabetic subjects

5.65 18 2.82 1.12 3.31 0.92

6.36 19 f 1.33 3.14 Et 1.39 0.98 zt 0.17 3.63 f 0.88 1.26 f 0.79

f f * f f

1.11 0.96 0.21* 0.87 0.35

2.10 f 0.90 119 f 23 246 f 6

2.21 f 0.99 132 f 25 244 zt 6

36.1 f 20.9*

45.5 f 16.7

19.3

22.2 f 7.2

l

4.4

n, number of subjects; Tot-Chol, serum total cholesterol; TotTg, serum total triglyceride; HDLChol, HDL cholesterol; LDL-Chol, LDL cholesterol; VLDL-Chol, VLDL cholesterol; VLDL-Tg, VLDL triglyceride; Apo B, apolipoprotein B; FFA, free fatty acids. *P c 0.05 vs. non-diabetic subjects.

significant difference in the diameters of the major LDL peak between the two groups (Table 2). Overall the data indicates that diabetes per se had no influence on LDL size. 3.2. Effect of insulin resistanceon metabolic parameters and serum lipoproteins

To evaluate whether the LDL size is influenced by insulin resistance per se the participants were categorized into two subgroups according to their insulin-stimulated glucose uptake rates using the median of the values (14.67 ~mol/kg/min) as a cutoff point. More insulin resistant group 1 consisted of 18 subjects (nine Type 2 diabetic and nine nondiabetic subjects) with the mean value of the insulin-stimulated glucose uptake rate 11.17 f 2.06 ~ol/kg/min (range 7.78-14.56 pmovkg mm). The less insulin resistant group 2 included 19 subjects (nine Type 2 diabetic and 10 non-diabetic subjects) with a mean value of the insulinstimulated glucose uptake rate of 24.44 f 9.39 ~mol/kg/min (range 14.67-42.78 ~mol/kg/min,

113 (1995)

227-236

231

P < 0.001 vs. group 1). The groups did not differ

with respect to age, BMI and glycemic control. The mean 24-h concentration of blood glucose did not differ between the groups (7.5 f 2.8 vs. 7.7 f 3.4 mmoyl). The fasting and the mean 24-h serum insulin concentrations were higher in group 1 than in group 2 (90.6 f 33.6 vs. 54.0 it 24.0 pmolll, P < 0.001, and 208.8 f 82.2 vs. 147.0 f 69.6 pmol/l, P < 0.05, respectively). Serum lipid and lipoprotein concentrations of the more insulin resistant subjects did not differ from those of the less insulin resistant patients (data not shown). Also, the LDL particle sizes were closely similar in the two groups (243 f 6 vs. 246 f 5 A) despite the marked difference in the insulinstimulated glucose uptake rates. 3.3. Determinants of LDL particle size

The mean value of the major LDL peak (245 f 6 A) was smaller than 255 A in the whole study population, indicating preponderance of small dense LDL in our study subjects. To identify the determinants of LDL size we evaluated the interaction between different parameters, first using Pearson’s univariate correlation analysis. In the whole study group we found no correlation between the LDL particle diameter and the rate of insulin-stimulated glucose uptake (Fig. 2) nor between LDL particle diameter and mean 24-h concentration of serum insulin (r = -0.06, NS). However, an inverse correlation existed between the mean LDL particle diameter of the major peak and the logarithm of fasting VLDL triglycerides (r = -0.35, P < 0.05, Fig. 3) and the mean 24-h concentration of serum triglycerides (r = -0.44, P < 0.01). LDL particle size did not correlate with HDLcholesterol concentration in the whole study group (r = 0.20, NS). Neither did we find any correlation between LDL particle size and hepatic or lipoprotein lipase (r = -0.21, NS and r = 0.10, NS, respectively). LDL particle size was not correlated with systolic or diastolic blood pressure, either (r = 0.15, NS and r = 0.23, NS, respectively). We created three models of multiple regression analyses in order to discover the predictors of LDL particle size. The first model included body and lifestyle factors (sex, age, BMI, diagnosis of

232

S. Lahdenperii

et al. /Atherosclerosis

(3

9 0

10 insulin uptake

113

(1995)

I

g

20

30

40

50

227-236

0

stimulated glucose rate @moi/kg/min)

10 insulin uptake

20

I

1

30

40

stimulated glucose rate @moi/kg/min)

Fig. 2. Interrelations between LDL particle diameter of the major LDL peak and insulin-stimulated glucose uptake rate in the whole study population (r = 0.19, NS). Left panel: Type 2 diabetic patients (O), non-diabetic subjects (0). Right panel: more insulin resistant subjects 0, more insulin sensitive subjects (0).

diabetes, blood pressure, and beta blocker treatment). The second model included metabolic parameters (insulin-stimulated glucose uptake rate, HbAi, concentration, the mean 24-h concentrations of blood glucose, serum insulin and triglycerides). The third model consisted of parameters of lipoprotein metabolism (HDL-cholesterol and LDL-cholesterol concentrations, the mean 24-h concentration of serum triglycerides, hepatic and lipoprotein lipase activities). Finally, we incorporated the independent determinants of LDL

particle size into the same multivariate analysis (Table 3). In the whole study population these factors were diastolic blood pressure (model l), mean 24-h Tg and insulin-stimulated glucose uptake rate (model 2), and LDL-cholesterol concentration, LPL and HL activities (model 3). In the whole study population the mean 24-h concentration of serum triglycerides was the major determinant of LDL particle sire, although the hepatic and lipoprotein lipase activities were also independently associated with the LDL particle size (Table 3).

260~

3 -0.2

2t307

I 0

0

0.2

Log VLDL-Tg

0.4 (mmoih)

l

0.6

0.8

-0.2

0

0.2

Log VLDL-Tg

0.4

0.6

0.8

(mmoi/i)

Fig. 3. Interrelations between LDL particle diameter of the major LDL peak and logarithm of VLDL triglycerides in the whole study population (r = -0.35, P C 0.05) 0. Left panel: Type 2 diabetic patients (0, r = -0.48, P < 0.05), non-diabetic subjects (0, r = -0.20, NS) 0. Right panel: more insulin resistant subjects @, r = -0.21, NS), more insulin sensitive s&j&s (4 r = -0.55, p < 0.05).

S. Luhdenperii

et al. /Atherosclerosis

Table 3 Multivariate regression analyses using the particle diameter of the major LDL peak as a dependent variable

All subjects Mean 24-h Tg LPL HL Diastolic BP Glucose uptake rate LDL-Cholesterol NIDDM patients Mean 24-h Tg LPL HL LDL-Cholesterol Diastolic BP Glucose uptake rate Non-diabetic subjects Mean 24-h Tg LPL LDLCholesterol Glucose uptake rate HL Diastolic RP --

r2

F to enter

P

0.19 0.27 0.34 0.36 0.37

8.46 3.41 3.52 0.99 0.57


0.38

0.40

NS

0.15 0.31 0.42 0.49 0.51 0.51

2.78 3.57 2.63 1.87 0.50 0.00

co.05 co.01 <0.05 NS NS NS

0.27 0.38 0.52 o.s9

6.15 2.80 4.41 2.56

CO.001 co.05
0.60 0.60

0.19 0.07

NS NS

When the multiple regression analyses were performed separately in the Type 2 diabetic and nondiabetic subjects, the mean 24-h concentration of serum triglycerides, lipoprotein and hepatic lipase activities were independently associated with the LDL particle size in the Type 2 diabetic patients (Table 3). In the non-diabetic subjects the mean 24-h concentration of serum triglycerides was most closely associated with LDL particle size, but also LDLcholesterol concentration, LPL activity and insulin-stimulated glucose uptake rate were independent determinants of LDL particle size (Table 3). 4. Dlseuasion Two important observations emerged from this study. First, the present data suggest that neither Type 2 diabetes nor insulin resistance, measured as the insulin-stimulated glucose uptake rate, have

113 (1995)

227-236

233

any direct effect on LDL particle size in mildly hypertriglyceridemic subjects, even though the range of insulin-stimulated glucose uptake rate was wide (7.78-42.78 flohkglmin). Second, the observation that LDL particle size was related to serum triglyceride concentration indicates that the effect of diabetes and insulin resistance on LDL particle size could be explained by the effects of insulin resistance and/or hyperinsulinism on VLDL metabolism. According to the inclusion criteria, our study subjects had triglyceride concentrations between 1.5 and 4.5 mrnol/l. Austin et al. have shown that the prevalence of small dense LDL increases steeply if serum triglyceride values exceed 1.1 mmol/l, suggesting a threshold effect [ 131. In agreement we observed that our Type 2 diabetic subjects had preponderance of small dense LDL particles representing the atherogenic LDL pattern B. However, the diagnosis of diabetes was not an independent determinant of LDL particle size in our cohort of mildly hypertriglyceridemic patients. Likewise Stewart et al. [40] failed to find a correlation between LDL size and insulin resistance in Type 2 diabetic patients. The present data agrees with our previous observation that Type 2 diabetic and non-diabetic patients had similar LDL particle sizes when the groups were matched for serum triglyceride concentration [41]. The data confront the proposal that there is a direct association between small dense LDL and Type 2 diabetes [30]. This proposal is based on data from studies where insulin values are used as a measure of insulin resistance. Caution should be ensured because insulin values are only a moderate substitute for euglycemic clamp technique as a measure of insulin resistance [42,43]. In the present study the ambient LDL particle size was strongly correlated with serum triglyceride concentration. Thus our results support the view that the ambient LDL size within LDL pattern B may reflect abnormalities in the function of VLDL-IDL-LDL cascade which are commonly present in Type 2 diabetes. When the LDL particle size was specifically examined in relation to insulin resistance, we observed no difference in the LDL particle size between the more and less insulin resistant groups. Of note is the fact that the two groups had similar

234

S. Luha’enperii et al. /Atherosclerosis 113 (1995) 227-236

concentrations of VLDL triglycerides and blood glucose and differed only with respect to insulin sensitivity and serum insulin. Lack of any relation between LDL particle size and insulin-stimulated glucose uptake rate, although the range of insulinstimulated glucose uptake rate was wide (7.7842.78 ~molikg/min), also confronts a direct effect of insulin resistance on LDL size. The lack of association is especially evident in the more insulin resistant patients who had a narrower range of insulin-stimulated glucose uptake rate than less insulin resistant patients (Fig. 2), but despite this the range for the LDL peak particle diameters varied from 232 to 253 A (Fig. 2). Similarly to the large variation in LDL particle sizes there was more variability in VLDL-triglycerides (1.1-4.6 mmol/l) than in the glucose uptake rate. We also failed to find an association between LDL particle size and serum insulin concentration. This is not actually unexpected because in overt Type 2 diabetes serum insulin levels are not a reliable parameter of insulin resistance. Note that in the studies where an independent association has been found between serum insulin level and LDL subclass phenotypes, plasma triglycerides have, however, had much stronger association with pattern B than insulin

WI. Elevation of serum triglyceride concentration in insulin resistant states is mainly due to the increased production of VLDL-triglycerides in the liver [45,46]. It has been postulated that high levels of VLDL-triglycerides may increase CETP (cholesteryl ester transfer protein)-mediated transfer of cholesteryl esters and triglycerides between VLDL and LDL [21]. Thus, LDL becomes enriched with triglycerides and serves as a good substrate for hepatic lipase, which hydrolyses LDL-triglycerides resulting in the formation of small dense LDL particles [47]. In line we found that both hepatic and lipoprotein lipase activities were independent predictors of LDL particle size in this study population (Table 3). Recently Watson et al. identified hepatic lipase and VLDL-Tg as major determinants for LDL subclasses in young healthy subjects [48]. In conclusion, neither hyperinsulinemia nor insulin resistance were independent determinants of the ambient LDL particle size in mildly hypertriglyceridemic subjects with and without Type 2

diabetes, who had preponderance of LDL pattern B. On the other hand, serum triglyceride concentration was a discriminator for LDL particle size distribution in these cohorts. Our data support the view that small dense LDL is a feature of the insulin resistance syndrome as a consequence of abnormalities in VLDL metabolism. Since preponderance of small dense LDL becomes apparent over a range of triglyceride values which represent only mild hypertriglyceridemia (1.5-4.5 mmol/l), the practical implication is that therapeutic measures to treat the insulin resistance syndrome should aim to lower serum triglycerides as effectively as possible in order to prevent the deleterious metabolic consequences of hypertriglyceridemia. Acknowledgements We are grateful to Hannele Hilden, Leena Lehikoinen, Sirpa Rannikko and Kikka Runeberg for the excellent technical help. This study was supported by grants from the Finnish State Medical Research Council, the Sigrid Juselius Foundation, the Finnish Foundation for Cardiovascular Research, the Kyllikki and Uolevi Lehikoinen Foundation, Helsinki, Finland, and WamerLambert, Ann Arbor, Michigan. References 111 Reaven, G.M., Role of insulin resistance in human disease, Diabetes, 37 (1988) 1595.

121 Haffner, SM., Valdez, R.A., Hazuda, H.P., Mitchell, B.D., Morales, P.A. and Stem, M.P., Prospective analysis of the insulin-resistance syndrome (syndrome X), Diabetes, 41 (1992) 715. r31 Landin, K., Tengbom, L. and Smith, U., Elevated fibrinogen and plasminogen activator (PAI-1) in hypertension are related to metabolic risk factors for cardiovascular disease, J. Intern. Med., 227 (1990) 273. H. and Yki-JLrvinen, H., Hyperuri141 Vuorinen-Markkola, cemia and insulin resistance, J. Clin. Endocrinol. Metab., 78 (1994) 25. 151 Abbott, W.H.G., Lillioja, S. and Young, A.A. et al., Relationships between lipoprotein concentrations and insulin action in an obese hyperinsulinemic population, Diabetes 36 (1987) 897. 161 Garg, A., Helderman, J.H., KoBler, M., Ayuso, R., Rosenstock, J. and Raskin, P., Relationship between lipoprotein levels and in vivo insulin action in normal young white men, Metabolism, 37 (1988) 982.

S. Laha’enperd

et al. /Atherosclerosis

[7] Modan, M., Halkin, H., Lusky, A., Segal, P., Fuchs, Z. and Chetrit, A., Hyperinsuhnemia is characterized by jointly disturbed plasma VLDL, LDL, and HDL levels: a population-based study, Arterioscler. Thromb., 8 (1988)

227.

Laakso, M., Sarlund, H. and Mykkgnen, L., Insulin resistance is associated with lipid and lipoprotein abnormalities in subjects with varying degrees of glucose tolerance, Arterioscler. Thromb., 10 (1990) 223. [91 Paolisso, G., Ferrannini, E., Sgambato, S., Varrichio, M. and D’Onofrio, F., Hyperinsuhnemia in patients with hypercholesterolemia, J. Clin. Endocrinol. Metab., 75 (1992) 1409. [lo] KarhapPii, P., Voutilainen, E., Kovanen, P.T. and Laakso, M., Insulin resistance in familial and nonfamilial hypercholesterolemia, Arterioscler. Thromb., 13 (1993) 41. [l 11 Reaven, G.M., Chen, Y.D.I., Jeppesen, J., Maheux, P. and Krauss, R.M., Insulin resistance and hyperinsulinemia in individuals with small, dense, low density lipoprotein particles, J. Clin. Invest., 92 (1993) 141. 1121 Sheu, W.H.H.,Shieh,S.M.,Fuh,M.M.T.,Shen, D.D.C., Jeng, C.Y., Chen, Y.D.I. and Reaven, G.M., Insulin resistance, glucose intolerance, and hyperinsulinemia: hypertriglyceridemia versus hypercholesterolemia, Arterioscler. Thromb., 13 (1993) 367. [13] Austin, M.A., King M-C, Vranizan, K.M. and Krauss, R.M., A&erogenic lipoprotein phenotype: a proposed genetic marker for coronary heart disease risk, Circulation, 82 (1990) 495. [I41 Fontbonne, A., Charles, M.A., Thibult, N. et al. Hyperinsulinemia as a predictor of coronary heart disease mortality in a healthy population the Paris prospective study, 15.year follow-up, Diabetologia, 34 [8]

(1991)

356.

[15] Austin. M.A., Breslow, J.L., Hennekens, C.H., Buring, J.E., Willeft, W.S. and Krauss, R.M., Low-density lipoprotein subclass patterns and risk of myocardial infarction, J. Am. Med. Assoc., 260 (1988) 1917. (161 Stamler, J., Vaccard, O., Neaton, J.D. and Wentworth, D., Diabetes, other risk factors, and 12-year cardiovascular mortality for men screened in the multiple risk factor intervention trial, Diabetes Care, 16 (1993) 434. (171 Austin, M.A., King, M.-C., Vranizan, K.M., Newman, B. and Krauss, R.M., Inheritance of low-density lipoprotein subclass patterns: results of complex segregation analysis, Am. J. Hum. Genet., 43 (1988) 838. [18] :LaBelle, M., Austin, M.A., Rubin, E. and Krauss, R.M., Linkage analysis of low-density subclass phenotypes and the apolipoprotein B gene, Genet. Epidemiol., 8 (1991) 269. (191

Nishina, P.M., Johnson, J.P., Naggert, K.J. and Krauss, R.M., Linkage of atherogenic lipoprotein phenotype to the low-density lipoprotein receptor locus on the short arm of chromosome 19, Proc. Natl. Acad. Sci. USA, 89 ( 1992) ‘708.

[20]

Austin, M.A., Newman, B., Selby, J.V., Edwards, K., Mayer, E.J. and Krauss, R.M., Genetics of LDL subclass

113

(1995)

227-236

235

phenotypes in women twins: concordance, heritability, and commingling analysis, Arterioscler. Thromb., 13 (1993) 687. [21] Deckelbaum, R.J., Granot, E., Oschry, Y., Rose, L. and Eisenberg, S., Plasma triglyceride determines structurecomposition in low and high density lipoproteins, Arterioscler. Thromb., 4 (1984) 225. [22] McNamara, J.R., Campos, H., Ordovas, J.M., Peterson, J., Wilson, P.W.F. and Schaefer, E.J., Effect of gender, age, and lipid status on low density lipoprotein subfraction distribution, Arterioscler. Thromb., 7 (1987) 483. [23] Cambert, P., Bouzerand-Gambert, C., Athias, A., Farnier, M. and Lallemant, C., Human low density lipoproseparated by gradient tein subfractions gel electrophoresis: composition, distribution and alterations induced by cholesterol ester transfer protein, J. Lip. Res., 31 (1990) 1199. [24] McNamara, J.R., Jenner, J.L., Li, Z., Wilson, P.W.F. and Schaefer, E.J., Change in LDL particle sire is asscciated with change in plasma triglyceride concentration, Arterioscler. Thromb., 12 (1992) 1284. [25] Lamon-Fava, S., Fischer, E.C., Nelson et al., Effect of exercise and menstrual cycle status on plasma lipids, low density lipoprotein particle size, and apolipoproteins, J. Clin. Endocrinol. Metab., 68 (1989) 17. [26] Campos, H., Willett, W.S., Peterson, R.M. et al., Nutrient intake comparisons between Framingltam and rural and urban Put&al, Costa Rica: associations with lipids, lipoproteins, apolipoproteins, and low density lipoprotein particle size, Arterioscler. Thromb., 11 (1991) 1089. [27] Campos, H., Bailey, S.M., Gussak, L.S., Siles, X., Ordovas, J.M. and Schaefer, E.J., Relations of body habitus, fitness level, and cardiovascular risk factors inchiding lipoproteins and apolipoproteins in a rural and urban Costa Rican population, Arterioscler. Thromb., II (1991) 1077. [28] Campos, H., Blijlevens, E., McNamara, J.R. et al., LDL particle size distribution Results from the Framingham offspring study, Arterioscler. Thromb., 12 (1992) 1410. [29] James, R.W. and Pometta, D., The distribution profiles of very low density and low density lipoproteins in poorlycontrolled male, type 2 (non-insulin-dependent) diabetic patients, Diabetologia, 34 (1991) 246. [30] Feingold, K.R., Gnmtield, C., Pang, M., Doerrler, W. and Krauss, R.M., LDL subclass phenotypes and triglyceride metabolism in non-insulin-dependent diabetes, Arterioscler. Thromb., 12 (1992) 1496. [31] Vuorinen-Markkoia, H., Yki-Jiirvinen, H. and Taskinen, M.-R., Lowering of triglycerides by gemfbrozil affects neither the glucoregulatory nor antilipolytic effect of insulin in Type 2 (non-insulin-dependent) diabetic patients, Diabetologia, 36 (1993) 161. [32] Lahdenperl, S, Tilly-Kiesi, M., Vuorinen-Markkola, H., Kuusi, T. and Taskinen, M.-R. Effects of gemftbrozil on low-density lipoprotein particle size, density distribution, and composition in patients with type II diabetes, Diabetes Care, 16 (1993) 584. [331 Sane, T., Knudsen, P., Vuorinen-Markkola, H., Yki-

236

[W] [35] [36]

[37]

[38]

[39]

[So]

S. L..ah&nperii et al. /Atherosclerosis 113 (1995) 227-236 Jirvinen, H. and Taskinen, M.-R., Lowering of triglycerides by gemfrbrozil affects neither the ghrcoregulatory nor antilipolytic effect of insulin in endogenous hypertriglyceridemia, Metabolism (in press). World Health Organization: WHO Expert Committee on Diabetes Mellitus, Second Report. WHO, Geneva, 1980, (Technical Report Series, no. 646) . DeFronzo, R.A., Tobin, J.D. and Andres, R., Glucose clamp technique: a method quantifying insulin secretion and resistance, Am. J. Physiol., 237 (1979) E214. Yki-J&men, H. and Taskinen, M.-R., Interrelationships among insulin’s antilipolytic and ghtcoregulatory effects and plasma triglycerides in nondiabetic and diabetic patients with endogenous hypertriglyceridemia, Diabetes, 37 (1988) 1271. Havel, J.R., Eder, H.A., Bragdon, J.H., Distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum, J. Clin. Invest., 34 (1955) 1345. Nichols, A.V., Krauss, R.M. and Musliner, T.A., Nondenaturing polyacrylamide gel electrophoresis. In: Segrest, J.P., Albers, J.J. (Eds.), Methods in Enzymology: Plasma Lipoproteins. vol. 128, Academic Press London, 1986, p. 417. Huttunen, J.K., Ehnhohn, C., Kinnunen, P.J. and Nikkilii, E.A., An immunochemical method for selective measurement of two triglyceride lipases in human postheparin plasma, Chn. Chii. Acta, 63 (1975) 335. Stewart, M.W., Laker, M.F., Dyer, R.G., Game, F., Mitcheson, J., Winocour, P.H. and Alberti, K.G.M.M., Lipoprotein compositional abnormalities and insulin resistance in type I1 diabetic patients with mild hypertriglyceridemia, Arterioscler. Thromb., 13 (1993) 1046.

[41] Tilly-Kiesi, M., Syviinne, M., Kuusi, T., Lahdenpera, S. and Taskinen, M.-R., Abnormalities in normolipidemic type II diabetic and nondiabetic patients with coronary artery disease, J. Lip. Res., 33 (1992) 333. [42] Laakso, M., How good a marker is insulin level for insulin resistance? Am. J. Epidemiol., 137 (1993) 959. [43] Scheen, A.J., Paquot, N., Castillo, M.J. and Lefebvre, P.J., How to measure insulin action in vivo, Diabetes Metab. Rev., IO (1994) 151. [44] Selby, J.V., Austin, M.A., Newman, B., Zhang, D., Quesenberry, C.P., Mayer, E.J. and Krauss, R.M., LDL subclass phenotypes and the insulin resistance syndrome in women, Circulation, 88 (1993) 381. [45] Tobey, T.A., Greenfield, M., Kraemer, F. and Reaven, G.M., Relationship between insulin resistance, insulin secretion, very low density lipoprotein kinetics, and plasma triglyceride levels in normotrigiyceridemic man, Metabolism, 30 (1981) 165. (461 Taskinen, M.-R., Factors influencing the altered lipoprotein systemin hypertriglyceridemia, In: Drugs Affecting Lipid Metabolism, Kluver Academic Publishers and Fondazioni Lorenzini, Netherlands, 1993, p. 467. I471 Zambon, A., Austin, M.A., Brown, B.G., Hokanson, J.E. and Brunzell, J.D., Effect of hepatic lipase on LDL in normal men and those with coronary artery disease, Arterioscler. Thromb., 13 (1993) 147. [48] Watson, T.D.G., Caslake, M.J., Freeman, D.J., Griffin, B.A., Hinnie, J., Packard, G.J. and Shepherd, J., Determinants of LDL subfraction distribution and concentrations in young normolipidemic subjects, Arterioscler. Thromb., 14 (1994) 902.