Atherosclerosis, 74 (1988) 5.5-63 Elsevier Scientific Publishers Ireland, Ltd. ATH 04214
Markku La&so * and Kaievi
r%i
Department of ll4edicine, Kuopio University Centsal Hospitad 70210 Kuopio (Finland) (Received 11 February, 1988) (Revised, received 20 June, 1988) (Accepted 27 June, 1988)
Coronary heart disease in insulin-dependent (IDDM) and in non-insulin-dependent diabetes (NIDD associated with lipid and lipoprotein changes favouring atherosclerosis. Whether lipid and lipoprotein abnormalities are associated also with peripheral vascular disease in both types of diabetes is largely unknown. Therefore, we studied lipid and lipoprotein levels and their association with claudication in a representative sample of diabetic and non-diabetic subjects in East Finland. Altogether 87 subjects had IDDM (43 men, 44 women), 264 subjects NIDDM (126 men, 138 women) and 120 subjects were non-diabetic controls (63 men, 57 women). Patients with IDD,M had an increased level of HDL and HDL,-cholesterol and patients with NIDDM a decreased level of HDL and HDL,-cholesterol and an increased level of total, LDL anA VLDL triglycerides than did non-diabetic subjects. Analyses in both types of diabetes by claudication status revealed that total and LDL-cholesterol and total and VLDL lower in those having claudication as triglycerides tended to be higher and HDL and HD L+holesterol compared to those without a da&cation symptom. Siarly, total choleste:ol/I-IDL-cholesterol ratio and LDL-cholesterol/HDL-cholesterol ratio were also more atherogenic in patients with claudication than in those without claudication. In conclusion, our results indicate that in both types of diabetes peripheral vascular disease is associz&d with lipid and lipoprotein abnormalities favouring atherosclerosis. is
Key words: Lipids; Lipoprotein; Insulin-dependent melhtus; Peripheral vascular disease
diabetes mellitus;
Non-insulin-dependent
diabetes
.L -Irktroduc~on * Present a&es Department of Medicine, Division of Endocrinology and Metabolism (V-IllG), Veterans Administration Medical Center, La Jolla, California 92161, U.S.A. Cowespondence ro: Markku Laakw. MD, Department of Medick, Division of Endocrinology and MetaboIism (VIIllG), Veterzxs Administration Medical Center, La Jolla, CA 92161, U.S.A. 0021-9150/88/$03.50
The risk of atherosclerotic complications is markedly increased both in insulin-dependent (IDDM) and non-insulin-dependent diabetes mellitus (NIDDM) [1,2]. According to several autopsy, clinical and epidemiological studies the preva-
0 1988 Ekevier Scientific Pubkhers Ireland, Ltd.
56
lence of coronary heart disease [3-61, cerebrovascular disease [7-81 and peripheral vascular disease (PVD) 19-121 occurs more frequently in diabetic patients than in corresponding control subjects. Although the etiology of the excess of atherosclerosis in diabetes has remained unsolved the lipid and lipoprotein abnormalities found particularly in NIDDM [13-151 favour the progression of atherosclerosis. Coronary heart disease in both types of diabetes is associated particularly with low levels of HDL and HDL,-cholesterol ]16,17]. It is not known whether abnormal lipid and lipoprotein changes seen in coronary heart disease are observed also in association with other atherosclerotic complications, like PVD. Previous studies on the relationship between PVD and lipids and lipoproteins have generally included only the measurement of total cholesterol and total triglycerides [4,11,12,18,19]. In only two studies [20,21] lipoprotein levels have been measured. in these studies [20,21] the classification of diabetic patients into the two main types was not based, however, on C-peptide levels or on other reliable indicators of the type of diabetes. In addition, these studies were not population-based studies but recruited their patients from diabetes clinics. Therefore, we have investigated lipids and lipoproteins in representative population samples of diabetic patients with and without PVD and compared the results with those of corresponding non-diabetic subjects free of PVD.
in middle-aged diabetic patients and non-diabetic subjects. The present study included 87 patients with IDDM (43 men, 44 women), 264 patients with NIDDM (126 men, 138 women) and 120 non-diabetic control subjects (63 men, 57 women). Of the patients with NIDDM, 77 (39 men, 38 women) were treated with diet only, 123 with oral drugs (60 men, 63 women) and 64 with insulin (27 men, 37 women). All diabetic patients included in the final study population fulfilled the WHO criteria for diabetes mellitus [22]. The 120 non-diabetic control subjects aged 45-64 years were drawn randomly from thespopulation register which included all persons living in the Kuopio University Central Hospital region. Therefore, our sample of non-diabetic subjects is fully representative of the non-diabetic population in the study area. In these subjects, diabetes was excluded or the basis of fasting plasma glucose measurements. None of the controls or diabetics received hypolipidemic drug therapy and none had a significant impairment in renal function assessed by serum creatinine level. None of the non-diabetic subjects included in this study had claudication on the basis of the Rose cardiovascular questionnaire [23]. All those diabetic and nondiabetic subjects were excluded who had suffered from definite or possible myocardial infarction verified at hospital according to the WHO criteria [24]. The participation rate in the study was 81% in diabetics and 79% in control subjects.
Subjects and methods
Methods Body mass index (BMI) was calculated according to the formula: BMI = weight (kg)fieight* (m). Alcohol intake was determined on the basis of subject’s estimate of the average number of drinks of beer, wine or spirits ingested per week (transformed to absolute alcohol in g/week). Exercise level was divided into two categories: (1) no or slight physical exercise during leisure time or work, and (2) regular physical exercise during leisure time (e.g. walking, bicycling, jogging, or swimming for at least 30 min at least twice a week) and/or heavy physical activity at work (e.g., heavy industrial work, farming, lumberjacking). The criterion for PVD was claudication as assessed by the Rose cardiovascular questionnaire [23]. Hypertension was defined as the use of anti-
Subjects All diabetic patients in Finland needing drug therapy are provided with it free of charge according to the Sickness Insurance Act. The Social Insurance Institution maintains a central register of diabetic subjects receiving drug reimbursement. This register includes practically all diabetic patients having or having had drug therapy for diabetes ( > 98%). On the basis of this register all diabetic patients, aged 45-64, were identified from the population living in the Kuopio University Central Hospital district (East Finland). The diabetic patients were originally identified for a study aiming to compare the prevalence of atherosclerotic vascular disease and its risk factors
57 hypertensive drugs or actual systolic blood pressure 2 160 mm Hg or diastolic blood pressure > 95 mm Hg. Neuropathy was defined as the absence of at least one Achilles tendon reflex. In all statistical analyses diabetics were classified as those having claudication (claud + ) and as those not having claudication (claud - ). Serum lipids and lipoproteins were determined from fresh serum samples drawn after a 12-h overnight fast. Lipoprotein fractionation was performed using ultracentrifugation and selectiveprecipitation with modification [25] of the method of Have1 et al. [26]. All spinnings were done at 10 o C using a Kontron TGA-65 ultracentrifuge (Kontron International, Switzerland). Serum samples were centrifuged at density (d) = 1.006 (105 000 x g, for 18 h), and very low density lipoprotein (VLDL) triglycerides (d < 1.006) were recovered as the top fraction. Total HDL-cholesterol was determined directly with dextrane sulphate and magnesium chloride precipitation, which correlates closely with the results obtained by ultracentrifugation [25]. Low density lipoprotein (LDL)-cholesterol (d = 1.006-l -063, including intermediate density lipoprotein) was calculated as a difference between the bottom fractions. On the average, the mean day-to-day variation in HDL-cholesterol measurement was 3.3% a.nd the daily variation 0.95%. HDL, and HDL, subfractions were separated running total HDL fraction at d= 1.125 (105050 x g, for 40 h), and the top and bottom fractions were isolated by the tube slicing technique. Cholesterol and triglycerides from the whole serum and from lipoprotein fractions were assayed by automated enzymatic methods (Boehringer-Mamrheim, West Germany). Glycosylated haemoglobin Al (normal range: 5.5-8.58; coefficient of variation: 6%) was determined by commercial column chromatography (Quick-&p Fast Haemoglobin Test System, Isolab. Inc., Akron, OH) after incubation in 0.9% saline solution for 12 h. Plasma glucose was measured by glucose dehydrogenase method (Merck Diagnostica, West Germany). The endogenous insulin secretion capacity was assessed in the post-prandial state by plasma C-peptide measurement after 1 mg of intravenous glucagon stimulation according to the method described by Faber and Binder [27]. Plasma C= peetide was assessed by radioimmunoassay (anti-
serum M 1230, NU:VQHradustri, Copenhagen, Denmark) [28]. The classification of diabetic patients into IDDM and NIDD was based on criteria suggested by the WHO [22]. In insulin-treated patients also C-peptide response to intravenous glucagon was used in the classification of different types of diabetes. If C-peptide level 6 min after 1 mg of glucagon exceeded 0.20 nmol/l and the patients had no history of ketoacidosis he/she was classified as having NIDDM. If C-peptide response was absent or I 0.20 nmol/l these patients were classified as having IDDM. In all patients having a history of ketoacidosis postglucagon C-peptide was I 0.20 nmol/l. Statistical methods Results for continuous variables are given as mean + SEM. Comparison between more than two groups was done by analysis of variance (ANOVA). Comparison between subjects with and without claudication or between diabetics (IDDM group or NIDDM group) and controls was made using Student’s t-test for independent samples and x2 test when appropriate. Adjustment for confounding factors was made by analysis of covariante (ANCOVA) when the groups of diabetics with (claud + ) and without (claud - ) claudication were compared. Logarithmic transformations of total and VLDL triglycerides were used in all statistical analyses including these variables.
Table 1 shows the clinical characteristics of the study groups. Patients with NIDDM were older and less physically active than non-diabetics. BMI was lower in IDDM and higher in NIDDM patients than in controls. No difference between controls and diabetics was seen in smoking or in alcohol intake. Both patients with IDDM and NIDDM had more neuropathy than controls and the prevalence of hypertension was higher among patients with NIDDM than among non-diabetic control subjects. As indicated in Table 2 patients with IDDM had higher HDL and HDL*cholesterol levels and lower LDL/HDL-cholester01 ratio than in corresponding controls in both sexes, In addition, female patients with IDDM had lower LDL and HDL,-cholesterol, lower total
58 TABLE 1 CLINICAL CHARACTERISTICS OF THE STUDY POPULATION Controls Men: No. of subjects
Age (yrs) BMI (kg/(m)‘) Duration of diabetes (yrs) Postglucagon C-peptide (nmol/l) Glucose (mmol/l) GH.bAl(%) Alcohol use (g/wk) Physically active (!%) Smokers ( W) Hypertensives (W) Neuropathy ( W) Women: No. of subjects
Age W) BMI (kg/(m)* ) Duration of diabetes (yrs) Postglucagon C-peptide (nmol/l) Glucose (mmol/l) GHbAl (I) Alcohol use (g/wk) Physically active (W) Smokers ( W) Hypertensives ( W) Neuropathy (%) *p
**Pco.ol,
63 53.7& 0.6 25.5f 0.3
57 53.8f 0.8 27.3& 0.6
5.3f 7.5*
20 f 51 5 41 9
0.1
0.1 9
Patients with IDDM
Patients with NIDDM
43 53.4 f 0.8 24.4 f 0.4 * 16.9 f 1.3 0.02f 0.01 10.5 + 0.9 *** 9.9 f 0.3 *** 69 k16 67 33 30 30 **
126 55.8 f0.5 27.4 f0.4 9.1 f0.4 0.93f0.19 10.9 *0.3 9.7 f0.2 56 f7 50 * 28 64 ***
44
138 58.0 f0.5 29.4 f0.5 8.8 kO.3 1.39fO.14 12.3 20.4 9.9 f0.2 4 *2 41 * 6 70 *** 20 *
55.1 f 0.7 24.3 f 0.5 *** 16.4 f 1.2 0.01* 0.01 12.4 f 0.7 *** 10.3 f 0.3 *** 6 k2 64 9 48 32 **
ANOVA P
** ***
*** ***
25 **
*** **
c 0.001 -z 0.001
*** ***
< 0.001 < 0.001 NS
* * * P < 0.001 (diabetics vs. controls); NS = not statistically significant.
cholesterol/HDL-cholesterol ratio and higher HDL triglyceride level than in non-diabetics. Patients with NIDDM had lower HDL and HDLr cholesterol, higher VLDL-cholesterol, higher total triglycerides, LDL triglycerides and VLDL triglycerides, and higher total cholesterol/HDLcholesterol ratio than in controls. In addition, female patients with NIDDM had lower LDLcholesterol, higher HDL triglycerides and higher LDL/HDL-cholesterol ratio than in corresponding non-diabetic controls. Table 3 shows the clinical characteristics of diabetic patients divided into 2 groups on the basis of claudication status. Clinical characteristics were quite similar in both diabetic groups. Fasting plasma glucose was lower in male IDDM patients with claudication compared to those
without claudication. Female NIDDM patients with claudication were less obese than those without claudication. Table 4 shows lipids and lipoproteins in diabetics by cl’audication status. In men, HDL-cholesterol was lower and total and VLDL triglycerides higher in IDDM pa.tients with claudication than in IDDM patients without claudication. These differecces persisted after adjustment for plasma glucose level (ANCOVA, P < 0.05) but the difference in HDL-cholesterol was no longer statistically significant after adjustment for plasma glucose level and total triglycerides. In female IDDM patients with claudication total and VLDL triglycerides were also higher than in those without claudication. In male NIDDM patients with claudication total and LDL-cholesterol levels, total
59 and VLDL triglyceride levels and total cholesterol/ HDL-cholesterol ratio and LDL-cholesterol/ HDL-cholesterol ratio were higher than in male NIDDM patients without claudication. In female NIDDM patients with claudication LDL-cholesterol and LDL-cholesterol/ HDL-cholesterol ratio were higher than in female NIDDM patients without claudication. These differences persisted after adjustment for BMI (ANCOVA, P c 0.05). Both in men and women, HDL and HDLr cholesterol tended to be lower in diabetic patients with elaudication than in diabetic patients without claudication in IDDM and in NIDDM but these differences were not statistically significant.
Although smoking and hypertension bave been identified as risk factors for PVD in diabetics [11,12,29] the association of lipids and lipoproteins with PVD has remained unclear [20,21]. Elevated total triglyceride level has been reported in some studies [X3-21] but in other studies [4,12] total cholesterol and triglyceride levels have been similar in diabetic patients with and without PVD. No reliable data based on unselected study populations exist on the relationship between lipoprotein levels and PVD where classification of diabetics had been based on objective criteria, like on
TABLE 2 LIPID AND LIPOPROTEIN CONCENTRATIONS (mmcQ1) IN CONTROLS AND IN DIABETIC SUBJECTS
Men: Cholesterol Total HDL HDL, HDLs LDL VLDL Total/HDL LDL/HDL Triglycetides Total HDL LDL VLDL Women: Cholesterol Total HDL HDL, HDL, LDL VLDL Total/HDL LDL/HDL Triglycerides Total HDL LDL VLDL
Controls
Patients with JDDM
Patients withNIDDM
ANOVA P
6.69kO.14 1.34*0.05 0.92 + 0.05 0.42 f 0.02 4.45 * 0.12 0.90 &0.05 5.29+0.20 3.55 f0.15
6.80 f 0.21 1.53 +0.07 * 1.11*&06 * 0.43 f 0.02 4.34+0.16 0.93 + 0.09 4.72 f 0.23 3.05 fC.18 *
6.65 f0.13 1.16f0.03 *** 0.75 * 0.03 * * 0.41 f 0.01 4.19f0.09 1.31*0.09 *** 5.29kO.19 ** 3.80 kO.10
NS -z 0.001 -z cm1 NS NS 0.002 c 0.001 0.002
1.48+0.08 0.12+0.01 0.30&0.02 1.04+0.07
1.56f0.14 0.13 + 0.01 0.36 + 0.02 1.08*0.12
2.58 kO.24 * * * 0.14+0.01 0.37 f 0.02 * * 2.07 f 0.22 * * *
c 0.001 NS NS < 0.001
7.28 f 0.20 1.45*&6x 1.01+0.05 0.44*0.02 4.94 f 0.16 0.89f0.09 5.28 + 0.20 3.59 kO.14
6.97kO.17 1.69kO.04 *** 1.3440.04 *** 0.35 f 0.01 * * * 4.49f0.15 * 0.79*0.06 4.22 + 0.14 * * * 2.73 f 0.12 * * *
7.30 kO.20 1.20f0.03 *** 0.79 j: 0.03 * * * 0.41 f 0.01 4.53 f 0.10 * 1.57kO.20 ** 6.77f0.38 *** 3.98 kO.10 *
NS -c 0.001
1.43 f 0.08 0.13 f 0.01 0.33 f 0.02 O-98&0.06
1.39+o.oil 0.18&0.01 *** 0.33 & 0.02 0.88 f 0.06
3.08 f 0.43 * * * 0.16 kO.01 * * 0.43 f 0.02 * * * 2.45 f 0.42 * * *
0.006 0.003 0.002 0.009
< 0.m 0.002 NS 0.011 -K0.001 < 0.001
* P < 0.05, * * P c 0.01, * * * P c 0,001 (Student’st-test; diabeticsvs. controls),NS = not statisticallysignificant.
60 TABLE 3 CLINICAL CHAI’&CTERISTICS OF DIABETICS GROUPED ACCORDING TO CLAUDICATIGN S’i*ATUS AND SEX Patients with NIDDM
Patients with IDDM Claud -
Claud +
Claud -
Claud +
BMI (kg/(m)2) Duration of diabetes (yrs) Glucose (mmol/I) GHbAl (45) Alcohol use (g/wk) Physically active (W) Smokers (I)
33 52.8+ 0.8 24.3& 0.4 Klf 1.3 11.4* 1.1 9.8& 0.3 72 f17 67 33
10 554& 2.1 24.6&- 1.1 19.6* 3.7 7.3f 1.1 * lO.lf 0.5 55 f39 70 30
101 56.0*0.5 27.2f0.4 9.3 f 0.5 10.8f0.4 9.7 + 0.2 47 *7 52 26
25 55.1* 1.0 28.9&0.8 8.2f0.3 11.2+0.8 9.8 * 0.4 93 +22 40 36
Women: No. of subjects Age (yrs) BMI (kg/(m)’ ) Duration of diabetes (yrs) Glucose (mmol/l) GHbAl (%) Alcohol use (g/wk) Physically active (%) Smokers (%)
37 54.6+ 0.8 24.0+ 0.5 15.9* 1.3 11.7f 0.7 10.2f 0.3 7 _+3 68 5
7 57.4* 25.9+ 19.3* 16.2& 10.8f 2 f2 43 29
128 58.IkO.5 29.6 + 0.5 8.6f0.3 12.3 +0.4 lO.OfO.2 4 -+2 40 7
10 57.1 f 2.0 26.1 kO.7 * * 11.1* 1.3 12.3+ 1.2 9.0i0.6 0 50 0
Men: No. of subjects Age (yrs)
1.4 I.9 3.4 2.0 0.3
* P < 0.05, * * P < 0.01 (Student’s t-test; Claud C vs. Claud -); Claud - = no claudication, Claud + = claudication.
the measurement of C-peptide levels after glucagon stimulation. Therefore, we studied lipids and lipoproteins in representative samples of diabetics and non-diabetics with and without PVD assessed by claudication symptoms. Classification of diabetics into the two main types were done using clinical criteria suggested by the WHO [22] and using postglucagon C-peptide measurement which is a reliable indicator of insulin secretion capacity [30]. Therefore, patients with IDDM and NIDDM could be reliably classified into the two main types of the disease., Fatients with a previous history of myocardial infarction were excluded from our study because previous studies have shown that in both types of diabetes myocardial infarction is associated with abnormalities in lipid and lipoprotein metabolism [16,17]. The criterion for PVD was a history of claudication assessed by the Rose cardiovascular questionnaire [23]. According to Marinelli et al. [31] a history of claudication has a specificity of 96% and a sensitivity of 2%. Therefore, although this method cannot identify all patients with PVD almost all patients having a
history of claudication have PVD. Low sensitivity of this method is likely to underestimate in this study the true difference in lipid and lipoprotein abnormalities between those having and those not having claudication. Our study showed that patients with IDDM have an increased level of HDL and HDL,cholesterol and ::atients with NIDDM ha%-e a decreased level OF HDL and HDL,-cholesterol and an increased level of tot,al, LDL and VLDL triglycerides confirming the results of previous studies [13-E X-343. Therefrre, patients with NIDDM - in contrast to patieiiz with IDDM have in general En unfavourable lipid pattern with respect to the risk F!’arkerosclerosis. When we did analyses separately in both types of diabetes by claudication status in each sex atherogenic lipid and lipoprotein pattern was seen both in IDDM and NIDDM in those having a claudication symptom as compared to those without claudication. Total and LDL-cholesterol tended to be higher in patients with claudication than in those without claudication particularly in patients with NIDDM.
61 TABLE 4 LIPID AND LIPOPRGTEIN CONCENTRATIONS TION STATUS AND SEX
(mmoI/l)
OF DIABETICS GROUPED
Patients with IDDM CIaud-
ACCORDING
TO CLAUDICA-
Patients with NIDDM CIaud +
CIaud -
Claud +
6.80&0.23 1.59*0.08 1.16f0.08 0.43 f 0.02 4.26i0.16 0.95 fO.ll 4.52f0.22 2.87f0.16
6.79*0.48 1.33*0.09 * 0.92*0.11 0.41 f0.03 4.59 f 0.45 0.87kO.14 5.41 f 0.65 3.67kO.53
6.49 f 0.14 1.18 f0.03 0.77*0.03 0.41 fO.O1 4.08 f 0.09 1.23 fQ.30 5.88&0.20 3.65 f 0.11
7.30f0.29 1.09*0.06 0.66*0.06 0.43 f 0.03 4.60 f 0.23 1.61 f0.18 7.06f0.40 4.37 f0.24
1.39f0.13 0.13*0.01 0.33iO.02 0.93f0.11
2.14kO.36 * 0.10+0.01 0.45 + 0.06 1.59*0.30 *
2.44i0.26 0.14f0.01 0.36+0.02 1.95 f0.24
3.14f0.56 * 0.14f0.01 0.44 f 0.04 2.57 f 0.48 *
LDL VLDL TotaI/HDL LDL/HDL
6.88&0.19 1.70*0.05 1.34*0.05 0.36*0.01 4.42kO.16 0.77f0.07 4.1s f0.15 2.67i0.12
7.43 f 0.29 1.66io.09 1.34io.09 0.33 * 0.02 4.88 f 0.35 0.88*0.18 4.61 f 0.45 3.05 f 0.41
7.25 f 0.21 1.2i f 0.03 0.80&0.03 0.41 f 0.01 4.49*0.11 1.55 f0.21 6.70*0&l 3.92f0.11
7.90 f 0.54 1.09kO.08 0.68f0.08 0.41 f 0.03 5.04*0.21 * 1.77 f 0.50 7.56f0.74 4.77*0.30 *
Triglycerides Total HDL LDL VLDL
1.30f0.08 0.18 fO.O1 0.32f0.02 0.80 +0.06
1.87i0.21 0.17*0.04 0.40f0.07 1.30f0.18
3.08 f 0.46 9.16 iO.01 0.43 f 0.02 2.48 + 0.45
3.06f0.77 0.17+0.03 0_46~0.05 2.43 + 0.72
choiesterol Total HDL HDLZ mL3
LDL VLDL TotaI/HDL LDL/HDL Triglycerides Total HDL LDL VLDL
.
*
* * *
Women: Cholesterol Total HDL mL2 H=3
* P c 0.05 (Student’s r-test; CIaud+ vs. CIaud-);
CIaud-
*
*
= no claudication, Claud+ = claudication.
HDL and HDL,cholesterol concentrations tended to be lower also in diabetics with claudication particularly in men but this difference was statistically significant only in IDDM patients Similarly, total cholesterol/HDL-cholesterol ratio and LDLcholesterol/HDL-cholesterol ratio were also more atherogenic in patients with claudication than in those without claudication. High total and VLDL triglycerides were a characteristic finding related to PVD particularly in TX& diabetic subjects confirming the results 01 previous studies [US-211. The abnormalities in total and VLDL triglycerides were more marked in IDDM patients with claudi-
cation than in NIDDM patients with claudication as compared to corresponding diabetics without claudication symptom. These results indicate that lipid and lipoprotein changes related to PVD seen in this study are similar to those of diabetic patients with coronary heart disease irrespective of the type of diabetes [16,17]. Therefore, high levels of total and LDLcholesterol, high levels of total and VLDL triglycerides and low levels of HDL and HDLr cholesterol favour atherosclerosis in different manifestations of atherosclerotic vascular disease. This is particularly striking in IDDM. Although
62 lipid and lipoprotein levels in IDDM are not generally atherogenic in patients with good metabolic control [34], those patients showing a manifestation of atherosclerotic vascular disease (coronary heart disease or PYD) have an atherogenic lipid/lipoprotein pattern. These observations need, however, a confirmation in prospective studies. Acknowledgements This study was supported by a grant from tke Medical Research Council of the Academy of Finland.
1 Pyi%ki, K. and Laakso, M., Macrovascular disease in diabetes mellitus. In: J.I. Mann, K. Py&%l%and A. Teuscher (Eds.), Diabetes in Epidemiological Perspective, Churchill Livingstone, Edinburgh-London-Melbourne-New York, 1983, pp. 183-247. 2 Pyiir& K., Laakso, M. and Uusitupa. M., Diabetes and atherosclerosis: an epidemiological view, Diabetes/Metabolism Rev., 3 (1987) 463. 3 Robertson, W.B. and Strong, J.P., Atherosclerosis in persons with hypertension and diabetes mellitus, Lab. Invest., 18 (1968) 538. 4 West, K.M., Ahuja, M.M.S., Bennett, P.H., Czyzyk, A., De Acosta, 0-M.. Fuller, J.H., Grab, B., Grabauskas, V., Jarrett, R.J., Kosaka, K., Keen, H., Krolewski, A.S., Miki, E., Schliak, V.. Teuscher, A., Watkins, P.J. and Strober, J.A., The role of circulating glucose and triglyceride concentrations and their interactions w:th other “risk factors” as determinants of arterial disease in nine diabetic population samples from the WHO multinational Study, Diabetes Care, 6 (1983) 361. S Gordon, T. and Kannel, W.B., Predisposition to atherosclerosis in the head, heart, and legs. The Framingham study, J. Am. Med. Assoc., 221 (1972) 661. 6 Cruz-Vidal, M., Gracia-Palmieri, M-R., Costas, R., Sorlie, P.D. and Havlik, R.J., Abnormal blood glucose and coronary heart disease: the Puerto Rico Heart Health Program, Diabetes Care, 6’(1983) 556. 7 Grunnet, M.L., Cerebrovascular disease: diabetes and cerebral atherosclerosis, Neurology, 13 (1963) 486. 8 Kannel. W.B. and McGee, D.L., Diabetes and cardiovascular disease. The Framingham Study, J. Am. Med. Assoc., 241 (1979) 2035. 9 Conrad, M.C., Large and small artery occlusion in diabetics and non-diabetics with severe vascular disease, Circulation, 36 (1967) 83. 10 Melton, L.J., Macken, K.M., Palumbo, P.J. and Elveback, L.R., incidence and prevalence of clinical peripheral vascular disease in a population-based cohort of diabciic patients, Diabetes Care, 3 (1980) 650.
11 Janka, H.U., Stand], E. and Mehnert, H., Peripheral vascular disease in diabetes mellitus and its relation to cardiovascular risk factors: screening with the Doppler ultrasonic technique, Diabetes Care, 3 (1980) 207. 12 Paisey, R.B., Arredondo, G., Villalobos, A., Lozano, O., Guevara, L. and Kelly, S., Association of differing dietary, metabolic, and clinical risk factors with macrovascular complications of diabetes: a prevalence study of SO? Mexican type II diabetic subjects, I, Diabetes Care, 7 (1984j 421. 13 Laakso, M,, Voutilainen, E., Sarlund, H., Are, A., Py&%ll, K. and PenttilSi, I., Serum lipids and lipoproteins hi middle-aged non-insulin-dependent diabetics, Atherosclerosis, 56 (1985) 271. 14 Barrett-Connor, E., Grundy, S.M. and Holdbrook, M.J., Plasma lipids and diabetes mellitus in an adult community, Am. J. Epidemiol.; 115 (1982) 657. 1s Walden, C.E., Knapp, R.H., Wahl, P-W., Beach, K.W. and Strandness, E., Sex differences in the effect of diabetes mellitus on lipoprotein triglyceride and cholesterol concentrations, N. Engl. J. Med., 311 (1984) 953. 16 Laakso, M., Voutilainen, E., Py&ti& K. and Sarlund, H., Association of low HDL and HDLl cholesterol with coronary heart disease in noninsulin-dependent diabetics, Arteriosclerosis, S (1985) 6S3. 17 Laakso, M., Pyiir%, K., Sarlund, H. and Voutilainen, E., Lipid and lipoprotein abnormalities associated with coronary heart disease in patients with insulin-dependent diabetes mellitus, Arteriosclerosis, 6 (1986) 679. 18 Santen, R.J., Willis, P.W. and Fajans, S.S.. Atherosclerosis in diabetes mellitus. Correlations with serum lipid levels, adiposity and serum insulin level, Arch. Intern. Med., 130 (1972) 833. 19 Vyden, J.K., Thomer, J., Nagasawa, K., Takano, T., Groseth-Ditrich, M.F., Perlow, R and Swan, H.J.C., Metabolic and cardiovascular abnormalities in patients with peripheral arterial disease, Am. Heart J., 90 (1975) 7C3. 20 Zimmerman, B.R., Palumbo, P.J., O’Fallon, W.M., Ellefson, R.D., Osmundson. P.J. and Kazmier, F.J., A prospective study of peripheral occlusive arterial disease in diabetes. III. Initial lipid and lipoprotein findings, Mayo Clii. Proc., 56 (1981) 233. 21 Beach, K.W., Brunzell, J-D., Conquest, L.L. ‘and Strandness, E., The correlation of arteriosclerosis obliterans with lipoproteins in insulin-dependent and non-insulin-dependent diabetes, Diabetes, 28 (1979) 836. 22 World Health Organization Expert Committee on Diabetes Mellitus, 2nd Report, Technical Report Series 646, World Health Organization, Geneva, 1982. 23 Rose, G., Blackbum, H., Gillum, R.F. and Prineas, R.J., Car’diovascular survey methods, 2nd edn., World Health Organization, Geneva, 1982. 24 World Health Organization. Proposal for the multinational monitoring of trends and determinants in cardiovascular disease and protocol (MONICA project), WHO/MNC/IZ. 1, Rev. 1, World Health Organization, Geneva, 1983. 2s Penttilti, I., Voutilainen, E., Laitinen, P. and Juutilainen. P., Comparison of different analytical and precipitation methods for the direct estimation of serum high-density lipoprotein cholesterol, Stand. J. Clin. Lab. Invest., 41 (1981) 353.
63 26 Have& R.J., Eder, H.A. and Bragdon, H.J., The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum, J. Clin. fnvest., 34 (1955) 1345. 27 Faber, O.K. and Binder, C., C-peptide response to glucagon. A test for the residual B-cell function in diabetes mellitus, Diabetes. 26 (1977) 605. 28 Hedmg, L.G., R.adioimmu~ological determination of human C-peptide in serum, Diabetologia, II (1975) 541. 29 Beach, K.W. and Strandness, D-E., Arteriosclerosis obliterarts and associated risk factors in insuhn-dependent and non-insulin-dependent diabetes, Diabetes. 29 (1980) 882. 30 Madsbad, S., Krarup, T., MeNair, P., Christiansen, C., Faber, O.K. and Binder, C., Practical clinical value of the C-peptide response to ghrcagon stimulation in the choice of
31
32
33
34
treatment in diabetes mellitus, Acta Med. !&and., 210 (1981) 153. Marinehi, M.R., Beach, K-W., Glass, M.J,. Primozich, J.F. and Strandness, DE., Non-invasive testing vs. chmcal evaluation of arterial disease, J. Am. Med. Assoc., 241 (1979) 2031. Briones. E.R., Mao, S.J.T., Palumbo, P.J., G’FaUon, W.M., Chenoweth, W. and Kottke, B.A., Analysis of plasma lipids and lipoproteins in insulin-dependent and non-insulin-dependent diabetics, Metabolism, 33 (1984) 42. Nikkil8, E.A. and Hormila, P., Serum lipids and Iipoproteins in insulin-treated diabetes. Demonstration of increased high density lipoprotein concentrations. Diabetes, 27 (1978) 1078. N&kit& E.A., Plasma lipid and lipoprotein abnormalities in diabetes. In: RJ. Jarrett (Ed.), Diabetes and Heart Disease, Elsevier Science Publishers, Amsterdam, 1984, pp. 134-167.