International Journal of Cardiology, 31(1992) 345-350 0 1992 Elsevier Science Publishers B.V. All rights reserved
CARD10
345 0167-5273/92/$05.00
01555
Microalbuminuria J. Woo a, C.S. Cockram
and other cardiovascular risk factors in nondiabetic subjects
a, R. Swaminathan
b, E. Lau ‘, A. Chan b and R. Cheung
b
Departments of a Medicine, b Chemical Pathology, ’ Community Medicine, The Chinese University of Hong Kong, Hong Kong (Received
30 January
1992; accepted
2 February
1992)
Woo, J, Cockram CS, Swaminathan R, Lau E, Chan A, Cheung R. Microalbuminuria cardiovascular risk factors in nondiabetic subjects. Int J Cardiol 1992;37:345-350.
and other
The association between urinary albumin : creatinine ratio and other cardiovascular risk factors such as age, blood pressure, obesity, glycemic indices, insulin and lipid profile was examined in a population in a Chinese community consisting of 795 men (mean age 35.8 f 8.8 yr) and 538 women (mean age 37.9 f 8.9 yr) with a normal glucose tolerance defined by WHO criteria. Men with a urinary albumin: creatinine ratio above the 90th percentile had higher systolic and diastolic blood pressures, fasting plasma glucose, 2-h glucose after a 75 g oral glucose load, and fasting serum insulin. Women with high urinary albumin : creatinine values had higher systolic and diastolic blood pressures, body mass index, waist-hip ratio, fasting insulin and triglycerides. Multivariate analysis showed that only systolic blood pressure and fasting glucose in men, and diastolic blood pressure and fasting insulin in women, independently contributed to urinary albumin: creatinine. When the effect of blood pressure was eliminated by excluding subjects with systolic blood pressure > 140 and diastolic > 90 mm Hg, only fasting insulin was associated with urinary albumin : creatinine in women. No associations were found for men. We conclude that microalbuminuria may be a marker for cardiovascular disease only because of its association with blood pressure in men, while in women, there is an additional independent association with fasting serum insulin. Key words: Microalbuminuria;
Insulin; Blood pressure; Obesity; Lipids; Cardiovascular
Introduction Microalbuminuria is an early indication of renal damage and in diabetic subjects it is associated with many other cardiovascular risk factors
Correspondence to: Dr. J. Woo, Dept. of Medicine, Prince of Wales Hospital. Shatin, N.T., Hong Kong. Tel. (852) 6363128. Fax (852) 6375396.
disease
[l-4] as well as being a predictor of mortality [5]. It has also been described as a marker for the presence of cardiovascular diseases even in nondiabetic subjects [6,7]. Since hypertension may give rise to renal damage, the association between microalbuminuria and cardiovascular disease may simply be mediated by hypertension. However, recent observations suggest that such an association is present independent of blood pressure [7,8]. Such studies have been confined to
346
minority racial groups and consisted of relatively few subjects. We examined the association between urinary albumin: creatinine ratio in a morning fasting sample and other cardiovascular risk factors such as age, blood pressure, obesity and glycemic indices, insulin, and lipid profile in a population in a Chinese community with normal glucose tolerance in order to confirm previous results and to determine the contribution of blood pressure to such associations.
Materials and Methods The study population consisted of employees from a public utility company and female staff of a general district hospital (excluding doctors, nurses, or paramedical staff) who participated in a community survey of the prevalence of diabetes mellitus and the lipid profile among Chinese subjects. Details of the population and survey method have been reported elsewhere [9]. Blood pressure was measured in the sitting position using an ordinary mercury sphygmomanometer. A mean of two readings was used. Height was measured with a tape with the subject standing upright against the wall. Weight was measured to the nearest 0.1 kg, with the subject in light clothing without shoes, using a bathroom scale. The body mass index was calculated as the weight (kg) divided by the square of the height cm). Waist circumference (cm> was taken as the minimum circumference between the umbilicus and xiphoid process, to the nearest 0.5 cm. The hip circumference was the maximum circumference around the buttocks posteriorly and the symphysis pubis anteriorly, to the nearest 0.5 cm. The mean of two measurements was recorded, and the waist-hip ratio was calculated by dividing the waist by the hip circumference. This parameter was used as an index of central obesity. A morning fasting urine specimen was collected for detection of microalbuminuria. This was estimated as the albumin:creatinine ratio. Urinary albumin was determined immunoturbidimetrically with use of a commercial antibody (Dak Patts a/s, Glostrup, Denmark) and a Cobas Bio Centrifugal analyzer [la], and creatinine concentration was measured by the Jaffe reaction on a
“Parallel” autoanalyzer (American Monitor Corporation, IN, USA). Twenty ml of blood was taken from the antecubital vein after a 12-h fast, for measurement of complete blood picture, renal and liver functions, glucose, glycosylated hemoglobin (AlC), cholesterol (TC), triglycerides (TG), high density lipoprotein cholesterol (HDL), low density lipoprotein cholesterol (LDL), apolipoprotein AI (apo A-I), apolipoprotein B (Apo B), and lipoprotein (a) (Lp (a)>. Seventy-five g glucose monohydrate in 150 ml of water was given to each subject, and was consumed within 5 min. One ml of venous blood was drawn 2 h after glucose loading for estimation of plasma glucose. All specimens were transported in ice to the laboratory within 4 h of sampling. TC and TG were assayed enzymatically with commercial reagents (Baker Instruments Corporation, Allentown, PA 18103, USA). HDL was determined after fractional precipitation with dextran sulphate-MgCl, [ 111. LDL was calculated using the Friedewald’s formula [12]. Apolipoproteins AI and B were assayed by rate immunonephelometry (Array analyzer, Beckman Instrument Inc, Bera, CA 92621, USA). Lp (a> concentration was measured by immunoradiometric assay (reagents of Pharmacia Diagnostic AB, Uppsala, Sweden). Interassay coefficients of variation were: TC, 1.9% at 6.4 mmol/l; TG, 2.6% at 1.9 mmol/l; HDL, 5.4% at 0.86 mmol/l; Apo AI 2.2% at 136 mg/dl; Apo B, 2.8% at 85 mg/dl; and Lp (a), 5.3% at 52 mg/dl. Plasma glucose was assayed with the Trinder reagent (Diagnostic Chemicals Ltd., Charlottetown, Canada ClE 1BD) on a Cobas Mira analyzer (Hoffmann-La Roche & Co., Basle, Switzerland). AlC was measured by ion exchange high performance liquid chromatography, using a BioRad DIAMAT Analyzer (Bio-Rad Laboratories, CA, USA). Interassay coefficients of variation for glucose, and AlC were 2.0% at 6.6 mmol/l, and I 3.1% at values below 8.5%, respectively. Insulin assay was performed using a radioimmunoassay kit (Pharmacia Insulin RIA 100, Pharmacia Diagnostics AB, Uppsala, Sweden). The lower limit of detection was < 2 KU/ml. Crossreactivity with C-peptide was < 0.18% (by
1
One way ANOVA:
* p < 0.05, **
?
124.9
< 0.01, ***
20.6 (326)
24.7 (326)
0.74 (326)
0.9 (326)
0.3 (326)
1.0(326)
0.5 (63)
4.9 f
= waist-hip
ratio;
281.1 + 336.2 (266)
187.3 f 187.0 (63)
31.3 (52) 19.1 (52)
75.8 f
0.98 (52) *
1.05 f 151.5 f
0.9(52)
0.4 (52)
0.9 (52)
3.1 f
1.5 +
4.9 f
5.5 (52) ** 42.7 (52)
0.9 (52)
4.5 + 9.8 f 44.7 f
1.1 (50)
0.7 (52) 5.5 +
4.9 +
blood
pressure,
358.2 f 469.0 (52) and diastolic
+ 351.7 (220)
+ 20.3 (220)
31.6(220)
0.54 (220)
0.8 (220)
0.4 (220)
0.8 (219)
SBP. DBP = systolic
287.4
71.5
f
0.81+
2.9 k
1.5 f
4.7 +
71.0 + 22.0 (266)
0.59 (266)
0.9 (266)
0.4 (266)
0.8 (264)
18.2 (63)
80.9 +
0.81 k
3.0 k
1.5 k
4.7 +
1.2 (220)
1.1 (216)
0.6 (218)
8.9 f 5.2 (220) 48.9 + 37.7 (220)
4.3 f
5.4 +
4.9 +
9.7 (52) 3.9 (52) ** 0.07 (52) * 17(52)** 12 (52) **
39.2 f 24.3 f 0.82+ 116 + 74 f
37.7 + 9.2 (220) 22.8 + 3.3 (220) 0.79 k 0.06 (220) 109 & 14(220) 69 + 9 (220)
146.9
0.94 (63)
1.33 +
1.3 (266)
1.0 (263)
0.6 (266)
7.9 + 3.6 (266) 49.7 f 36.9 (263)
4.2 f
5.3 f
4.8 +
8.5 (266) 3.0 (266) 0.06 (266) 12(266) 9 (266)
2 90th percentile
50 to 90th percentile
145.2 k 35.5 (266)
0.9(63)
3.5 +
24.5 (63)
0.3 (63)
1.2 +
127.0 i
1.0(63)
5.3 +
8.7 + 5.4 (62) * 57.0 + 41.8 (62)
1.4(63)**
0.4 (63) **
5.4 f
4.9 f
< 0.001. BMI = body mass index; WHR
+ 267.8 (326)
k
k
1.22 f
222.5
0.86 (406)
f
3.3
f 293 (406)
0.9 (406)
f
k
k 4.0 (322) + 39Of321)
1.3 +
78.6
0.3 (406)
+
5.2
19.9(406)
0.9 (406)
f
7.3 45.9
+
4.2 (402) 39.3 (400)
f *
0.6 (326)
4.9 +
1.1 (325)
4.9 k
26.9 (406)
0.6 (406)
f
0.4 (326)
4.7 k
f
1.2 (405)
f
0.5 (406)
37.8 f 22.8 f 0.79 + 108 + 69 f
+
8.7 (63) 3.0 (63) 0.06 (63) 17(63)*** 11(63)*
4.7
f + f f +
38.3 23.7 0.88 128 80
+ 8.9 (326) & 3.0 (326) f 0.06 (326) + 12(326) I 9 (326)
35.6 22.9 0.86 123 79
8.8 (406) 2.8 (406) 0.06 (406) 11 (406) 8 (406)
f f + f k
fn = 538)
tolerance.
35.7 23.3 0.87 121 77
4.9 Glycosylated hemoglobin (%) 4.9 Fasting insulin (mU/I) 7.8 2-h insulin (mU/l) 48.0 Total cholesterol (mmol/l) 5.2 HDL-cholesterol (mmol/l) 1.2 LDL-cholesterol (mmol/l) 3.4 Triglycerides (mmol/l) 1.28 Apolipoprotein A-I fmg/dl) 123.4 Apolipoprotein B 79.3 fmg/dl) Lipoprotein (a) (mg/dl) 240.5
Age (yr) BMI ikg/m’) WHR (cm/cm) SBP (mmHg) DBP (mmHg) Fasting glucose (mmol/l) 2-h glucose (mmol/l)
glucose
< 50th percentile
with normal
Women
in subjects
2 90th percentile
excretion
50 to 90th percentile
(mg/mmol)
albumin
< 50th percentile
: creatinine
by urinary
Men (n = 795)
Albumin
Mean values f SD of variables
TABLE
348
weight). The coefficient of variation between kits was 5%. Subjects were included for analysis if the WHO criteria for normal glucose tolerance [13] were satisfied (fasting and 2-h plasma glucose < 7.8 mmol/l), and in the absence of a history of diabetes mellitus. None of the subjects had chronic liver, renal or biliary diseases. Results were analyzed using the Statistical Package for Social Sciences (Vers 3.0) for IBM PC. One-way analysis of variance was used to detect significant differences in mean values according to urinary albumin : creatinine groups. Multiple linear regression using the stepwise procedure was used to assess independent contributions of different variables on urinary albumin: creatinine.
transformation. Subjects were divided into three groups according to the urinary albumin : creatinine ratios: below the 50th percentile (0.33 mg/mmol for men and 0.63 mg/mmol for women), between the 50th and 90th percentile (0.33-1.29 mg/mmol for men and 0.63-2.10 mg/mmol for women), and above the 90th percentile (> 1.29 mg/mmol for men and > 2.10 mg/mmol for women). Table 1 shows the mean values of various cardiovascular risk factors according to urinary albumin : creatinine ratios. In men, subjects with urinary albumin: creatinine above the 90th percentile had higher systolic and diastolic blood pressures, fasting and 2-h glucose and fasting insulin. In women, those with a urinary albumin: creatinine ratio above the 90th percentile had higher systolic and diastolic blood pressures, body mass index, waist-hip ratio, fasting insulin and triglycerides. In order to examine the individual contributions of these parameters to urinary albumin : creatinine ratio, multiple stepwise regression using log,, (albumin : creatinine) as independent variable and parameters shown to be statistically significant in Table 1 as dependent variables was carried out (Table 2). Systolic blood pressure and fasting glucose in men, and diastolic blood pressure and fasting insulin in women, were identified as significant predicting factors. Systolic blood pressure in men accounted for the majority
Results Seven-hundred-and-ninety-five men (mean age 35.8 4~8.8 yr) and 538 women (mean age 37.9 f 8.9 yr) with normal glucose tolerance were studied. Since urinary creatinine excretion is lower in women, giving rise to higher urinary albumin : creatinine ratios, its association with various cardiovascular risk factors was studied separately in the two sexes. The distribution of urinary albumin:creatinine ratio was negatively skewed in both sexes, but became normal on logarithmic
TABLE 2 Multivariate analysis of parameters contributing to urinary albumin : creatinine ratio. Dependent variable
Independent
variable
Regression statistics B
SEB
%R2
F
0.005 0.08 - 1.50
0.001 0.04 0.23 Total
2 0.6
16.0 *** 10.2 ***
2.6
10.2 ***
0.004 0.002 0.15 Total
2 1
12.3 *** 9.7 ***
3
9.7 ***
Men
log,, urinary albumin: creatinine ratio
SBP Fasting glucose Constant
Women
Fasting insulin DBP Constant
0.015 0.005 - 0.72
* * * p < 0.001. SBP, DBP = systolic and diastolic blood pressure
349
of the variance to urinary albumin: creatinine compared with glucose, while fasting insulin was the major factor in women. These analyses were repeated after excluding subjects with systolic blood pressure > 140 and diastolic > 90 mmHg (70 males, 16 females). In women, body mass index and fasting insulin were significantly higher in subjects with increasingly high urinary albumin : creatinine values (body mass index: 22.8 * 3.0, 22.8 _t 3.3, 24.1 f 4.0 kg/m2, p < 0.05,one-way ANOVA; fasting insulin: 8.0 f 4.5, 8.8 + 5.1, 9.6 k 5.3, p 95 mmHg were excluded). Our present study of a population in a Chinese community consists of larger numbers than previous studies, and we have included only those with normal glucose tolerance based on WHO criteria (excluding those with impaired glucose tolerance). The effect of systolic blood pressure was also examined, since we used a criterion of systolic blood pressure > 140 and diastolic > 90 mmHg to exclude hypertensive subjects. As expected, before exclusion of hypertensive subjects, systolic and diastolic blood pressures were higher in the group with higher urinary albumin : creatinine. Other variables showing similar associations (fasting glucose, insulin, triglycerides, waist-hip ratio, body mass index) were all correlated with systolic and diastolic blood pressure. Thus in men, systolic blood pressure was positively associated with fasting glucose (Y = 0.20, p < 0.001) and fasting insulin tr = 0.19, p < O.OOl), and diastolic blood pressure had similar association. In women, systolic blood pressure was positively associated with body mass index (7 = 0.24, p < O.OOl>, waist-hip ratio (r = 0.25, p < 0.0011, fasting insulin (r = 0.18, p < 0.0011, and triglycerides (r = 0.23, p < 0.001). Similar values were observed for diastolic blood pressure. Therefore multivariate analysis was used to determine which factor(s) contributed independently to urinary albumin: creatinine. Since the distribution of the latter was skewed but was normalized on logarithmic transformation, log ,” urinary albumin: creatinine was used as the dependent variable. For men, blood pressure remained the primary contributory factor to urinary albumin : creatinine, but fasting glucose still had a small independent contribution. In women, both fasting insulin and blood pressure were important factors. Yeh et al. also noted a positive correlation in univariate analysis between log urinary albumin : creatinine, blood pressure, and serum insulin in 338 non-diabetic American Indians age 45-74 yr. However, after exclusion of hypertensive sub-
350
jects, only fasting insulin in women was associated with high urinary albumin : creatinine values. None of the factors in men were associated with urinary albumin : creatinine. Therefore, we can conclude that in subjects with normal glucose tolerance, microalbuminuria may be a marker for cardiovascular disease only because of its association with blood pressure. In women, although blood pressure contributed to microalbuminuria, fasting serum insulin has an equally, if not more important, role. This finding in women is compatible with the hypothesis that higher serum insulin levels may indicate insulin resistance which, at the level of the kidney, could contribute to raise intraglomerular pressure and hence increased microalbuminuria [7].
4
5
6
7
8
9
Acknowledgements We thank Merck, Sharp & Dohme Pharmaceuticals Co Ltd for financial assistance and Dr. C.W.K. Lam for arranging the lipid assays.
References 1 Christensen CK, Mogensen CE. The course of incipient diabetic nephropathy: studies of albumin excretion and blood pressure. Diabet Med 1985;2:97-102. 2 Wiseman M, Viberti G, Mackintosh D, Jarret RJ, Keen H. Glycemia, arterial pressure and microalbuminuria in type 1 (insulin-dependent) diabetes mellitus. Diabetologia 1984;26:401-405. 3 Jensen T. Stender S, Deckert T. Abnormalities in plasma
10
concentrations of lipoproteins and fibrinogen in type 1 (insulin-dependent) diabetic patients with increased urinary albumin excretion. Diabetologia 1988;31:142-145. Kapelrud H, Bangstad HJ, Dahl-Jorgensen K, Berg K, Hanssen KF. Serum Lpfa) lipoprotein concentrations in insulin dependent diabetic patients with microalbuminuria. Br Med J 1991;303:675-678. Mogensen CE. Microalbuminuria predicts clinical proteinuria in maturity onset diabetes. N Engl J Med 1984;310: 356-360. Yudkin JS, Forrest RD, Jackson CA. Microalbuminuria as a predictor of vascular disease in non-diabetic subjects. Ishington Diabetes Survey. Lancet 1988;2:530-533. Haffner SM, Stern MP, Gruber MKK, Hazuda HP, Mitchell BD, Patterson JK. Microalbuminuria. Potential marker for increased cardiovascular risk factors in nondiabetic subjects? Arteriosclerosis 1990;10:727-731. Yeh J, for the Strong Heart Study Investigators. Albuminuria and cardiovascular disease (CVD) risk factors in American Indians. Circulation 1990;82(suppl III):620. Woo J, Lau E, Chan A, Cockram C, Swaminathan R. Blood pressure and urinary cations in a Chinese population. J Human Hypertens 1992;6:299-304. Cheung CK, Swaminathan R. Rapid, economical immunoturbidimetric method for microalbuminuria. Clin Chem 1987;33:204-205. Talameh Y, Wei R, Naito H. Measurement of total HDL, HDL, and HDL, by dextran sulphate-MgClz precipitation technique in human serum. Clin Chim Acta 1986; 158:33-41. Friedewald WT, Levy RI, Frederickson DS. Estimation of plasma low density lipoprotein cholesterol concentration without the use of the preparative ultracentrifuge. Clin Chem 1972;18:499-512. World Heath Organization. Diabetes Mellitus. WHO Tech Rep Ser 1985;727:103-108. Modan M, Halkin H, Almog S et al. Hyperinsulinaemia. A link between hypertension, obesity and glucose intolerance. J Clin Invest 1985:75:809-817.
CARD10
345 0167-5273/92/$05.00
01555
Microalbuminuria J. Woo a, C.S. Cockram
and other cardiovascular risk factors in nondiabetic subjects
a, R. Swaminathan
b, E. Lau ‘, A. Chan b and R. Cheung
b
Departments of a Medicine, b Chemical Pathology, ’ Community Medicine, The Chinese University of Hong Kong, Hong Kong (Received
30 January
1992; accepted
2 February
1992)
Woo, J, Cockram CS, Swaminathan R, Lau E, Chan A, Cheung R. Microalbuminuria cardiovascular risk factors in nondiabetic subjects. Int J Cardiol 1992;37:345-350.
and other
The association between urinary albumin : creatinine ratio and other cardiovascular risk factors such as age, blood pressure, obesity, glycemic indices, insulin and lipid profile was examined in a population in a Chinese community consisting of 795 men (mean age 35.8 f 8.8 yr) and 538 women (mean age 37.9 f 8.9 yr) with a normal glucose tolerance defined by WHO criteria. Men with a urinary albumin: creatinine ratio above the 90th percentile had higher systolic and diastolic blood pressures, fasting plasma glucose, 2-h glucose after a 75 g oral glucose load, and fasting serum insulin. Women with high urinary albumin : creatinine values had higher systolic and diastolic blood pressures, body mass index, waist-hip ratio, fasting insulin and triglycerides. Multivariate analysis showed that only systolic blood pressure and fasting glucose in men, and diastolic blood pressure and fasting insulin in women, independently contributed to urinary albumin: creatinine. When the effect of blood pressure was eliminated by excluding subjects with systolic blood pressure > 140 and diastolic > 90 mm Hg, only fasting insulin was associated with urinary albumin : creatinine in women. No associations were found for men. We conclude that microalbuminuria may be a marker for cardiovascular disease only because of its association with blood pressure in men, while in women, there is an additional independent association with fasting serum insulin. Key words: Microalbuminuria;
Insulin; Blood pressure; Obesity; Lipids; Cardiovascular
Introduction Microalbuminuria is an early indication of renal damage and in diabetic subjects it is associated with many other cardiovascular risk factors
Correspondence to: Dr. J. Woo, Dept. of Medicine, Prince of Wales Hospital. Shatin, N.T., Hong Kong. Tel. (852) 6363128. Fax (852) 6375396.
disease
[l-4] as well as being a predictor of mortality [5]. It has also been described as a marker for the presence of cardiovascular diseases even in nondiabetic subjects [6,7]. Since hypertension may give rise to renal damage, the association between microalbuminuria and cardiovascular disease may simply be mediated by hypertension. However, recent observations suggest that such an association is present independent of blood pressure [7,8]. Such studies have been confined to
346
minority racial groups and consisted of relatively few subjects. We examined the association between urinary albumin: creatinine ratio in a morning fasting sample and other cardiovascular risk factors such as age, blood pressure, obesity and glycemic indices, insulin, and lipid profile in a population in a Chinese community with normal glucose tolerance in order to confirm previous results and to determine the contribution of blood pressure to such associations.
Materials and Methods The study population consisted of employees from a public utility company and female staff of a general district hospital (excluding doctors, nurses, or paramedical staff) who participated in a community survey of the prevalence of diabetes mellitus and the lipid profile among Chinese subjects. Details of the population and survey method have been reported elsewhere [9]. Blood pressure was measured in the sitting position using an ordinary mercury sphygmomanometer. A mean of two readings was used. Height was measured with a tape with the subject standing upright against the wall. Weight was measured to the nearest 0.1 kg, with the subject in light clothing without shoes, using a bathroom scale. The body mass index was calculated as the weight (kg) divided by the square of the height cm). Waist circumference (cm> was taken as the minimum circumference between the umbilicus and xiphoid process, to the nearest 0.5 cm. The hip circumference was the maximum circumference around the buttocks posteriorly and the symphysis pubis anteriorly, to the nearest 0.5 cm. The mean of two measurements was recorded, and the waist-hip ratio was calculated by dividing the waist by the hip circumference. This parameter was used as an index of central obesity. A morning fasting urine specimen was collected for detection of microalbuminuria. This was estimated as the albumin:creatinine ratio. Urinary albumin was determined immunoturbidimetrically with use of a commercial antibody (Dak Patts a/s, Glostrup, Denmark) and a Cobas Bio Centrifugal analyzer [la], and creatinine concentration was measured by the Jaffe reaction on a
“Parallel” autoanalyzer (American Monitor Corporation, IN, USA). Twenty ml of blood was taken from the antecubital vein after a 12-h fast, for measurement of complete blood picture, renal and liver functions, glucose, glycosylated hemoglobin (AlC), cholesterol (TC), triglycerides (TG), high density lipoprotein cholesterol (HDL), low density lipoprotein cholesterol (LDL), apolipoprotein AI (apo A-I), apolipoprotein B (Apo B), and lipoprotein (a) (Lp (a)>. Seventy-five g glucose monohydrate in 150 ml of water was given to each subject, and was consumed within 5 min. One ml of venous blood was drawn 2 h after glucose loading for estimation of plasma glucose. All specimens were transported in ice to the laboratory within 4 h of sampling. TC and TG were assayed enzymatically with commercial reagents (Baker Instruments Corporation, Allentown, PA 18103, USA). HDL was determined after fractional precipitation with dextran sulphate-MgCl, [ 111. LDL was calculated using the Friedewald’s formula [12]. Apolipoproteins AI and B were assayed by rate immunonephelometry (Array analyzer, Beckman Instrument Inc, Bera, CA 92621, USA). Lp (a> concentration was measured by immunoradiometric assay (reagents of Pharmacia Diagnostic AB, Uppsala, Sweden). Interassay coefficients of variation were: TC, 1.9% at 6.4 mmol/l; TG, 2.6% at 1.9 mmol/l; HDL, 5.4% at 0.86 mmol/l; Apo AI 2.2% at 136 mg/dl; Apo B, 2.8% at 85 mg/dl; and Lp (a), 5.3% at 52 mg/dl. Plasma glucose was assayed with the Trinder reagent (Diagnostic Chemicals Ltd., Charlottetown, Canada ClE 1BD) on a Cobas Mira analyzer (Hoffmann-La Roche & Co., Basle, Switzerland). AlC was measured by ion exchange high performance liquid chromatography, using a BioRad DIAMAT Analyzer (Bio-Rad Laboratories, CA, USA). Interassay coefficients of variation for glucose, and AlC were 2.0% at 6.6 mmol/l, and I 3.1% at values below 8.5%, respectively. Insulin assay was performed using a radioimmunoassay kit (Pharmacia Insulin RIA 100, Pharmacia Diagnostics AB, Uppsala, Sweden). The lower limit of detection was < 2 KU/ml. Crossreactivity with C-peptide was < 0.18% (by
1
One way ANOVA:
* p < 0.05, **
?
124.9
< 0.01, ***
20.6 (326)
24.7 (326)
0.74 (326)
0.9 (326)
0.3 (326)
1.0(326)
0.5 (63)
4.9 f
= waist-hip
ratio;
281.1 + 336.2 (266)
187.3 f 187.0 (63)
31.3 (52) 19.1 (52)
75.8 f
0.98 (52) *
1.05 f 151.5 f
0.9(52)
0.4 (52)
0.9 (52)
3.1 f
1.5 +
4.9 f
5.5 (52) ** 42.7 (52)
0.9 (52)
4.5 + 9.8 f 44.7 f
1.1 (50)
0.7 (52) 5.5 +
4.9 +
blood
pressure,
358.2 f 469.0 (52) and diastolic
+ 351.7 (220)
+ 20.3 (220)
31.6(220)
0.54 (220)
0.8 (220)
0.4 (220)
0.8 (219)
SBP. DBP = systolic
287.4
71.5
f
0.81+
2.9 k
1.5 f
4.7 +
71.0 + 22.0 (266)
0.59 (266)
0.9 (266)
0.4 (266)
0.8 (264)
18.2 (63)
80.9 +
0.81 k
3.0 k
1.5 k
4.7 +
1.2 (220)
1.1 (216)
0.6 (218)
8.9 f 5.2 (220) 48.9 + 37.7 (220)
4.3 f
5.4 +
4.9 +
9.7 (52) 3.9 (52) ** 0.07 (52) * 17(52)** 12 (52) **
39.2 f 24.3 f 0.82+ 116 + 74 f
37.7 + 9.2 (220) 22.8 + 3.3 (220) 0.79 k 0.06 (220) 109 & 14(220) 69 + 9 (220)
146.9
0.94 (63)
1.33 +
1.3 (266)
1.0 (263)
0.6 (266)
7.9 + 3.6 (266) 49.7 f 36.9 (263)
4.2 f
5.3 f
4.8 +
8.5 (266) 3.0 (266) 0.06 (266) 12(266) 9 (266)
2 90th percentile
50 to 90th percentile
145.2 k 35.5 (266)
0.9(63)
3.5 +
24.5 (63)
0.3 (63)
1.2 +
127.0 i
1.0(63)
5.3 +
8.7 + 5.4 (62) * 57.0 + 41.8 (62)
1.4(63)**
0.4 (63) **
5.4 f
4.9 f
< 0.001. BMI = body mass index; WHR
+ 267.8 (326)
k
k
1.22 f
222.5
0.86 (406)
f
3.3
f 293 (406)
0.9 (406)
f
k
k 4.0 (322) + 39Of321)
1.3 +
78.6
0.3 (406)
+
5.2
19.9(406)
0.9 (406)
f
7.3 45.9
+
4.2 (402) 39.3 (400)
f *
0.6 (326)
4.9 +
1.1 (325)
4.9 k
26.9 (406)
0.6 (406)
f
0.4 (326)
4.7 k
f
1.2 (405)
f
0.5 (406)
37.8 f 22.8 f 0.79 + 108 + 69 f
+
8.7 (63) 3.0 (63) 0.06 (63) 17(63)*** 11(63)*
4.7
f + f f +
38.3 23.7 0.88 128 80
+ 8.9 (326) & 3.0 (326) f 0.06 (326) + 12(326) I 9 (326)
35.6 22.9 0.86 123 79
8.8 (406) 2.8 (406) 0.06 (406) 11 (406) 8 (406)
f f + f k
fn = 538)
tolerance.
35.7 23.3 0.87 121 77
4.9 Glycosylated hemoglobin (%) 4.9 Fasting insulin (mU/I) 7.8 2-h insulin (mU/l) 48.0 Total cholesterol (mmol/l) 5.2 HDL-cholesterol (mmol/l) 1.2 LDL-cholesterol (mmol/l) 3.4 Triglycerides (mmol/l) 1.28 Apolipoprotein A-I fmg/dl) 123.4 Apolipoprotein B 79.3 fmg/dl) Lipoprotein (a) (mg/dl) 240.5
Age (yr) BMI ikg/m’) WHR (cm/cm) SBP (mmHg) DBP (mmHg) Fasting glucose (mmol/l) 2-h glucose (mmol/l)
glucose
< 50th percentile
with normal
Women
in subjects
2 90th percentile
excretion
50 to 90th percentile
(mg/mmol)
albumin
< 50th percentile
: creatinine
by urinary
Men (n = 795)
Albumin
Mean values f SD of variables
TABLE
348
weight). The coefficient of variation between kits was 5%. Subjects were included for analysis if the WHO criteria for normal glucose tolerance [13] were satisfied (fasting and 2-h plasma glucose < 7.8 mmol/l), and in the absence of a history of diabetes mellitus. None of the subjects had chronic liver, renal or biliary diseases. Results were analyzed using the Statistical Package for Social Sciences (Vers 3.0) for IBM PC. One-way analysis of variance was used to detect significant differences in mean values according to urinary albumin : creatinine groups. Multiple linear regression using the stepwise procedure was used to assess independent contributions of different variables on urinary albumin: creatinine.
transformation. Subjects were divided into three groups according to the urinary albumin : creatinine ratios: below the 50th percentile (0.33 mg/mmol for men and 0.63 mg/mmol for women), between the 50th and 90th percentile (0.33-1.29 mg/mmol for men and 0.63-2.10 mg/mmol for women), and above the 90th percentile (> 1.29 mg/mmol for men and > 2.10 mg/mmol for women). Table 1 shows the mean values of various cardiovascular risk factors according to urinary albumin : creatinine ratios. In men, subjects with urinary albumin: creatinine above the 90th percentile had higher systolic and diastolic blood pressures, fasting and 2-h glucose and fasting insulin. In women, those with a urinary albumin: creatinine ratio above the 90th percentile had higher systolic and diastolic blood pressures, body mass index, waist-hip ratio, fasting insulin and triglycerides. In order to examine the individual contributions of these parameters to urinary albumin : creatinine ratio, multiple stepwise regression using log,, (albumin : creatinine) as independent variable and parameters shown to be statistically significant in Table 1 as dependent variables was carried out (Table 2). Systolic blood pressure and fasting glucose in men, and diastolic blood pressure and fasting insulin in women, were identified as significant predicting factors. Systolic blood pressure in men accounted for the majority
Results Seven-hundred-and-ninety-five men (mean age 35.8 4~8.8 yr) and 538 women (mean age 37.9 f 8.9 yr) with normal glucose tolerance were studied. Since urinary creatinine excretion is lower in women, giving rise to higher urinary albumin : creatinine ratios, its association with various cardiovascular risk factors was studied separately in the two sexes. The distribution of urinary albumin:creatinine ratio was negatively skewed in both sexes, but became normal on logarithmic
TABLE 2 Multivariate analysis of parameters contributing to urinary albumin : creatinine ratio. Dependent variable
Independent
variable
Regression statistics B
SEB
%R2
F
0.005 0.08 - 1.50
0.001 0.04 0.23 Total
2 0.6
16.0 *** 10.2 ***
2.6
10.2 ***
0.004 0.002 0.15 Total
2 1
12.3 *** 9.7 ***
3
9.7 ***
Men
log,, urinary albumin: creatinine ratio
SBP Fasting glucose Constant
Women
Fasting insulin DBP Constant
0.015 0.005 - 0.72
* * * p < 0.001. SBP, DBP = systolic and diastolic blood pressure
349
of the variance to urinary albumin: creatinine compared with glucose, while fasting insulin was the major factor in women. These analyses were repeated after excluding subjects with systolic blood pressure > 140 and diastolic > 90 mmHg (70 males, 16 females). In women, body mass index and fasting insulin were significantly higher in subjects with increasingly high urinary albumin : creatinine values (body mass index: 22.8 * 3.0, 22.8 _t 3.3, 24.1 f 4.0 kg/m2, p < 0.05,one-way ANOVA; fasting insulin: 8.0 f 4.5, 8.8 + 5.1, 9.6 k 5.3, p 95 mmHg were excluded). Our present study of a population in a Chinese community consists of larger numbers than previous studies, and we have included only those with normal glucose tolerance based on WHO criteria (excluding those with impaired glucose tolerance). The effect of systolic blood pressure was also examined, since we used a criterion of systolic blood pressure > 140 and diastolic > 90 mmHg to exclude hypertensive subjects. As expected, before exclusion of hypertensive subjects, systolic and diastolic blood pressures were higher in the group with higher urinary albumin : creatinine. Other variables showing similar associations (fasting glucose, insulin, triglycerides, waist-hip ratio, body mass index) were all correlated with systolic and diastolic blood pressure. Thus in men, systolic blood pressure was positively associated with fasting glucose (Y = 0.20, p < 0.001) and fasting insulin tr = 0.19, p < O.OOl), and diastolic blood pressure had similar association. In women, systolic blood pressure was positively associated with body mass index (7 = 0.24, p < O.OOl>, waist-hip ratio (r = 0.25, p < 0.0011, fasting insulin (r = 0.18, p < 0.0011, and triglycerides (r = 0.23, p < 0.001). Similar values were observed for diastolic blood pressure. Therefore multivariate analysis was used to determine which factor(s) contributed independently to urinary albumin: creatinine. Since the distribution of the latter was skewed but was normalized on logarithmic transformation, log ,” urinary albumin: creatinine was used as the dependent variable. For men, blood pressure remained the primary contributory factor to urinary albumin : creatinine, but fasting glucose still had a small independent contribution. In women, both fasting insulin and blood pressure were important factors. Yeh et al. also noted a positive correlation in univariate analysis between log urinary albumin : creatinine, blood pressure, and serum insulin in 338 non-diabetic American Indians age 45-74 yr. However, after exclusion of hypertensive sub-
350
jects, only fasting insulin in women was associated with high urinary albumin : creatinine values. None of the factors in men were associated with urinary albumin : creatinine. Therefore, we can conclude that in subjects with normal glucose tolerance, microalbuminuria may be a marker for cardiovascular disease only because of its association with blood pressure. In women, although blood pressure contributed to microalbuminuria, fasting serum insulin has an equally, if not more important, role. This finding in women is compatible with the hypothesis that higher serum insulin levels may indicate insulin resistance which, at the level of the kidney, could contribute to raise intraglomerular pressure and hence increased microalbuminuria [7].
4
5
6
7
8
9
Acknowledgements We thank Merck, Sharp & Dohme Pharmaceuticals Co Ltd for financial assistance and Dr. C.W.K. Lam for arranging the lipid assays.
References 1 Christensen CK, Mogensen CE. The course of incipient diabetic nephropathy: studies of albumin excretion and blood pressure. Diabet Med 1985;2:97-102. 2 Wiseman M, Viberti G, Mackintosh D, Jarret RJ, Keen H. Glycemia, arterial pressure and microalbuminuria in type 1 (insulin-dependent) diabetes mellitus. Diabetologia 1984;26:401-405. 3 Jensen T. Stender S, Deckert T. Abnormalities in plasma
10
concentrations of lipoproteins and fibrinogen in type 1 (insulin-dependent) diabetic patients with increased urinary albumin excretion. Diabetologia 1988;31:142-145. Kapelrud H, Bangstad HJ, Dahl-Jorgensen K, Berg K, Hanssen KF. Serum Lpfa) lipoprotein concentrations in insulin dependent diabetic patients with microalbuminuria. Br Med J 1991;303:675-678. Mogensen CE. Microalbuminuria predicts clinical proteinuria in maturity onset diabetes. N Engl J Med 1984;310: 356-360. Yudkin JS, Forrest RD, Jackson CA. Microalbuminuria as a predictor of vascular disease in non-diabetic subjects. Ishington Diabetes Survey. Lancet 1988;2:530-533. Haffner SM, Stern MP, Gruber MKK, Hazuda HP, Mitchell BD, Patterson JK. Microalbuminuria. Potential marker for increased cardiovascular risk factors in nondiabetic subjects? Arteriosclerosis 1990;10:727-731. Yeh J, for the Strong Heart Study Investigators. Albuminuria and cardiovascular disease (CVD) risk factors in American Indians. Circulation 1990;82(suppl III):620. Woo J, Lau E, Chan A, Cockram C, Swaminathan R. Blood pressure and urinary cations in a Chinese population. J Human Hypertens 1992;6:299-304. Cheung CK, Swaminathan R. Rapid, economical immunoturbidimetric method for microalbuminuria. Clin Chem 1987;33:204-205. Talameh Y, Wei R, Naito H. Measurement of total HDL, HDL, and HDL, by dextran sulphate-MgClz precipitation technique in human serum. Clin Chim Acta 1986; 158:33-41. Friedewald WT, Levy RI, Frederickson DS. Estimation of plasma low density lipoprotein cholesterol concentration without the use of the preparative ultracentrifuge. Clin Chem 1972;18:499-512. World Heath Organization. Diabetes Mellitus. WHO Tech Rep Ser 1985;727:103-108. Modan M, Halkin H, Almog S et al. Hyperinsulinaemia. A link between hypertension, obesity and glucose intolerance. J Clin Invest 1985:75:809-817.
