Diabetes Research and Clinical Practice, 12 (1991) 143-148
0 1991 Elsevier Science Publishers
143
B.V. 0168-8227/91/$03.50
DIABET 00495
Relationship between erythrocyte sorbitol content and diabetic microangiopathy in patients with non-insulin-dependent diabetes mellitus: the study of a diet loading test Goji Hasegawa ‘, Yoshinobu Tsutsumi ‘, Shinichiro Aoki I, Manabu Sawada ‘, Junko Setoguchi ‘, Naoto Nakamura ‘, Koji Nakano ‘, Motoharu Kondo ’ and Takahiro Kanatsuna2 ‘First Department of Internal Medicine, Kyoto Prefectural University of Medicine and ‘Department
of Metabolic Disease, Kyoto
City Hospital, Kyoto. Japan
(Received 11 September 1990) (Revision accepted 11 January 1991)
Summary To determine whether erythrocyte sorbitol content could become an indicator of diabetic microangiopathy, we studied the relationship between the changes in erythrocyte sorbitol content in response to diet loading and diabetic microangiopathy in patients with non-insulin-dependent diabetes mellitus. The increase of change in erythrocyte sorbitol content (ASor) after diet loading (420 kcal) significantly correlated with that of plasma glucose levels (ABS). The patients with < 40 m/s of motor or sensory nerve conduction velocity (MCV and SCV) had significantly higher ASor and ASor/ABS values than those with > 40 m/s of MCV and SCV; nevertheless, there were no significant differences in ABS between the two groups. Furthermore there was a significant negative correlation between nerve conduction velocity and ASor and ASor/ABS values. On the other hand, the patients with nephropathy or retinopathy showed no significant increase in ASor or ASor/ABS compared with patients without these complications. The results demonstrated that ASor and ASor/ABS could become indicators of the presence or severity of diabetic neuropathy. Furthermore the more significant participation of alteration in the polyol pathway in the pathogenesis of neuropathy than of the other microangiopathies was suggested. Key words: Erythrocyte
sorbitol content;
Polyol pathway;
Introduction Accumulation of intracellular sorbitol, resulting from overactivity of the polyol pathway secondary --~ Correspondence to: Goji Hasegawa, First Dept. of Internal Kyoto Prefectural University of Medicine. Medicine, Kawaramachi-Hirokoji, Kamikyo-ku, Kyoto, 602, Japan.
Diet loading test; Diabetic microangiopathy
to hyperglycemia, has been suggested as being one of the etiologic factors in the pathogenesis of diabetic microangiopathy [ 1,2,3]. The activity of the sorbitol pathway in patients becomes readily accessible for measurement by the use of erythrocytes as a test material [4]. However, whether the erythrocyte sorbitol content could become an indicator of either the pres-
144
ence or the severity of diabetic complications is controversial, because it easily fluctuates with prevailing blood glucose levels [5]. To clarify whether or not the erythrocyte sorbitol content could become an indicator of the presence of diabetic microangiopathy, we studied the relationship between the erythrocyte sorbitol content in response to a diet loading test and diabetic microangiopathy in patients with noninsulin-dependent diabetes mellitus (NIDDM).
Materials and Methods Patients: thirty patients with NIDDM (9 men, 21 women; mean k SE age 61.6 k 1.2years, range 47-69 years) with HbA,, levels greater than 7.0:” were tested. None of the subjects had any disease other than diabetes. No medications which were known to affect diabetic microangiopathy, had been given for at least one year before the study. The mean duration of diabetes was 14.2 + 1.5 years (mean + SE: range 3-32 years), obtained from the subject’s history. Fourteen patients received sulfonylurea agents, 11 received insulin, 2 received sulfonylurea agents with insulin and 3 were maintained on diet alone. STUDY
PROTOCOL
On the day of the study, the administration of insulin or oral hypoglycemics were delayed until after the diet loading tests were completed. After an overnight fast, blood samples were taken for the measurement of plasma glucose and erythrocyte sorbitol content at 0900 h. The standard breakfast meal consisted of bread, banana, margarine and milk (420 kcal; 64.4 g carbohydrate. 14.2 g fat, 9.6 g protein). This was eaten within a 15-min period from 0930 h, and blood samples were again taken at 1100 h and 1300 h. ANALYTICAL
PROCEDURE
Plasma was separated and plasma glucose was determined by the glucose oxidase method with a
Beckman glucose analyzer. The erythrocyte sorbitol content was measured by a fluorometric enzyme assay with packed erythrocytes according to the technique of Malone et al. 141. EVALUATION
OF DIABETIC
MICROANGIOPATHY
Evaluation of diabetic microangiopathy was performed within 2 months before or after the study. The grade of retinopathy was determined by ophthalmologists with ophthalmoscopy and fluorescent angiography, and was classified as no change, simple, preproliferative or proliferative. Nephropathy was evaluated by proteinuria or increase of urinary concentration of albumin without urinary tract infection. Urinary albumin levels of morning urine samples were measured by radioimmunoassay in cases of a negative reagent strip method. The examination was performed twice on different days and mean values were obtained. The patients with urinary albumin levels less than 20 mg/g . creatinine were judged as normal. Nerve conduction studies were performed by the same observer with a Medelec MS6 electromyograph using surface electrodes. Motor nerve conduction velocity (MCV) was measured in the posterior tibia1 nerves on the left side. Sensory conduction velocity (SCV) was measured in the sural nerve. All measurements were undertaken in a warm room with skin temperature between 32-34°C. The recordings of sural sensory measurements in 5 patients were technically unsatisfactory. CALCULATIONS
AND
STATISTICS
The increase of change in plasma glucose levels (ABS) and erythrocyte sorbitol contents (ASor) after diet loading were calculated by subtracting each value before eating (0900 h) from the value at 1100 h. Because erythrocyte sorbitol levels are known to correlate with ambient glucose levels, we calculated ASor/ABS to standardize the sorbitol accumulation in cells in response to various increases in plasma glucose levels. Statistical evaluation was performed using Student’s t-test
145
for unpaired data and linear regression analysis. Comparisons of distribution of sex and mode of therapy by the severity of diabetic microangiopathy were done by X2-test. Results are expressed as mean + SE.
%
200
i-p
ij
_' u E loo-
Results
-----____
,/
-30 ;
--A
,/ /C* 6-'.x
ln -20 g
s a
The profile of the erythrocyte sorbitol content after diet loading well reflected that of the plasma glucose level with their peak levels 1.5 h after the diet (Fig. 1). The erythrocyte sorbitol content before diet loading (fasting erythrocyte sorbitol content: F-Sor) significantly correlated with fasting plasma glucose levels (FBS) (r = 0.50; P < 0.01) and HbA,, levels (r = 0.53; P < 0.003). ASor significantly correlated with ABS (Y = 0.49;
8 -10 2 5 w 0
1lOOh
0900h
1300h0
Time
Fig. 1. The changes in erythrocyte sorbitol contents and plasma glucose levels with the diet loading test. The standard breakfast meal (420 kcal) was eaten within a 15-min period from 0930 h. 0 -, erythrocyte sorbitol content; plasma glucose level (n = 30). The data represent 0 -, mean f SE.
TABLE 1 The relationship between motor (MCV) and sensory nerve conduction diet loading test
velocity (SCV) and the parameters
MCV (m/s) 40 < 44.2 f 0.5 n (M/F) Age (years) Diabetes duration (years) Mode of therapy (n) diet su insulin insulin + su HbA,, (“/a) FBS (mg/dl) F-Sor (nmol/g Hb) BBS (mg/dl) ASor (nmol/g . Hb) ASor/ABS Su: sulfonylurea
obtained
from the
SCV (m/s) 40 > 37.9 f
0.5
40 < 44.9 f
0.8
40 > 33.3 f
1.8
17 (5/12) 61.9 & 1.4
13 (4/9) 61.2 t 2.0
16 (S/11) 61.2 k 1.4
9 (4/5) 62.2 f 2.9
15.1 f 2.0
13.1 + 2.2
15.8 f
2.2
15.8 + 2.2
0.3 7.2
1 2 6 0 8.2 + 0.4 121.3 of: 10.6
19.7 f 1.4 117.7 * 10.7
18.2 + 1.3 135.2 + 16.3
1 10 4 2 8.4 f 0.3 139.1 + 7.1
2 4 7 0 7.9 f 133.6 f
19.8 f 1.3 120.7 f 9.9
17.7 f 1.1 119.9 + 12.4
11.4 + 1.0 10.2 + 0.9
16.8 f 14.6 f
0.3 9.0
1.9** 1.4#
agents.
*P < 0.05, **P < 0.02, # P < 0.01 vs 40 m/s < of MCV or SCV; mean + SE.
1 9 5 1 8.3 k 139.5 k
11.6 f 10.6 +
1.2 1.1
18.7 + 2.P 15.0 + 1.6* -
146 TABLE 3
TABLE 2 Correlations parameters
between
Parameters
nerve conduction
MCV
velocity
and the
The relationship between urinary albumin excretion and the parameters obtained from the diet loading test Urinary albumin (mg/g creatinine)
scv
r
P
r
P
0.14 0.12 0.14 0.32 - 0.48 - 0.43
ns ns ns ns 0.01 0.02
0.07 0.12 0.13 -0.17 - 0.64 -0.51
ns ns ns ns 0.002 0.01
20 > HbA,c
FBS F-Sor ABS ASor ASor/ABS n
30
25
ns: not significant.
P < O.Ol), whereas
ASor did not correlate with FBS, HbA,, levels or F-Sor. The relationship between diabetic microangiopathy and the various parameters obtained from the diet loading test are shown in Tables 1, 2, 3, 4 and 5. Mean MCV and SCV in the patients tested were 41.4 + 0.7 m/s and 40.7 k 1.4 m/s, respectively. So, in the study of neuropathy, we defined the boundary at 40 m/s of nerve conduction velocity and divided the patients into two groups. No sex and age effects on the severity of each microangiopathy were found. Duration of diabetes and the distribution of mode of therapy did not differ according to the severity of neuropathy and nephropathy. However, the patients with preproliferative-proliferative retinopathy had diabetes of longer duration than those without retinopathy (P < 0.05). Furthermore, the patients who received insulin therapy were more frequently in these groups with preproliferative and proliferative retinopathy (P < 0.025, $-test). The patients with <40 m/s of MCV and SCV had significantly higher ASor and ASor/ABS values than those with values > 40 m/s for MCV and SCV; nevertheless there were no significant differences in ABS between the two groups. Furthermore, in contrast to the lack of correlation between nerve conduction velocity and ABS, there was a significant negative correlation between nerve conduction velocity with A Sor and
n (M/F) Age (years) Diabetes duration (years) Mode of therapy (n) diet SU
insulin Insulin + su HbA,, (Pb) FB S (mg/dl) F-Sor (nmol/g . Hb) ABS (mg/dl) ASor (nmol/g . Hb) ASor/ABS
20 <
19 (6/13) 60.9 + 1.3 13.0 + 2.0 3 9 5 2 8.1 132.2 19.1 118.5 12.3 11.5
f f + f + f
0.3 5.8 1.2 9.2 1.1 1.3
Su: Sulfonylurea agents. a Including 3 patients with clinical proteinuria.
11 (3/S)’ 62.6 + 2.4 16.4 f 2.0 0 5 6 0 8.3 + 144.6 + 18.6 + 123.6 f 16.2 f 13.1 +
0.3 11.3 1.1 14.0 2.2 0.9
Mean _+SE.
ASor/ABS (Table 2). The results indicate that sorbitol is more easily accumulated in erythrocytes in response to the increase of plasma glucose levels with the progression of diabetic neuropathy. The patients with nephropathy (> 20 mg/g . creatinine of urinary albumin excretion) showed higher ASor values than those without nephropathy, although there were no significant differences. Because there were no significant differTABLE 4 Correlations parameters
between
urinary albumin
excretion
and the
Parameters
r
P
HbA,, FBS F-Sor ABS ASor ASoriABS
0.21 0.28 0.24 0.28 0.29 0.006
NS NS NS NS NS NS
n = 27 (excluding 3 patients with clicical proteinuria).
147 TABLE 5 The relationship
between diabetic retinopathy Retinopathy
n (M/F) Age (years) Diabetes duration (years) Mode of therapy (n) diet SU
insulin insulin + su HbA,, (%) FBS (mg/dl) F-Sor (nmol/g Hb) ABS (mg/dl) ASor (nmol/g Hb) ASor/ABS
and the parameters (- )
obtained from the diet loading test
Retinopathy
(+ )
P
s
16 (S/11) 61.9 + 1.5
14 (4/10) 61.2 f 1.9
6 (214) 64.8 +
1.9
12.1 + 2.1
16.7 + 2.0
12.8 f
3.3
8 (2/6) 58.5 +
2.1
19.6 + 2.0* #
3 9 3 1 8.0 + 0.3 135.6 f 6.9
0 5 8 1 8.4 f 0.3 138.0 + 9.2
0 4 1 1 8.1 + 0.4 123.8 + 12.2
0 1 I 0 8.7 f 0.3 148.6 f 12.5
18.6 * 1.4 108.8 + 9.0
19.2 + 1.0 133.6 + 12.1
19.2 f 1.8 118.7 + 19.8
19.3 f 1.2 144.8 f 15.0*
12.3 f 1.3 12.1 f 1.4
15.4 f 12.1 +
1.8 1.1
12.5 + 12.0 f
1.2 1.9
11.5 + 2.8 12.2 f 1.3
( - ), no findings of diabetic retinopathy; S, simple diabetic retinopathy; P, preproliferative (n = 7); and proliferative diabetic retinopathy (n = 1). * P < 0.05 vs retinopathy (- ); # P < 0.025 X*-test. When the comparison was made among retinopathy (-), S and P. Mean it: SEM.
ences in ASor/ABS between the two groups, the high ASor value may be a consequence of high ABS levels (Table 3). This same pattern seen in nephropathy was also noted in the comparison between patients with and without retinopathy (Table 5). HbA,,, FBS levels and F-Sor did not differ between the groups for each microangiopathy (Tables 1, 3 and 5).
Discussion The erythrocyte sorbitol content before diet loading correlated with FBS and Hb,, levels, indicating that it reflects median term glycemic control. The change in erythrocyte sorbitol contents after diet loading correlated well with the change in plasma glucose levels, as demonstrated in earlier studies [6,7]. Furthermore, the erythrocyte sorbitol content before diet loading (F-Sor) did not reflect the
presence or severity of diabetic microangiopathy. Previous studies using chemically-induced diabetic rats showed a significant correlation between erythrocyte sorbitol content and that of retina and peripheral nerve [8,9]. These findings suggest that the polyol pathway activity in erythrocytes can provide information about that in tissues with diabetes-associated complications. However, the erythrocyte sorbitol content does not reflect the presence or severity of diabetic complication, probably because of its rapid in vivo turnover. We found that sorbitol readily accumulated in erythrocytes in response to increases in plasma glucose levels with the progression of diabetic neuropathy. It has been reported that the rate of sorbitol breakdown is not reduced in the diabetic state [5]. Therefore this may be caused by increased aldose reductase activity and/or altered kinetic properties of aldose reductase resulting from factors such as lowered Km of the enzyme for glucose [ lo]. Aldose reductase in erythrocytes
148
is activated at higher glucose concentrations via its intracellular metabolite, glucose-6-phosphate [ 111. In addition to the glucose concentration, our results suggest that unknown factors, genetic or secondary to metabolic disturbance, which increase aldose reductase activity exist in patients with diabetic neuropathy. These results are partially in agreement with the recent data described by Aida et al. [ 121, in which they used the ratio of erythrocyte sorbitol to blood glucose as an indicator of polyol pathway activity. The contribution of alteration in the polyol pathway to the pathogenesis of nephropathy or retinopathy has been reported [ 1-3, 131. However, in this study, the significant increase in sorbitol accumulation in erythrocytes, as described above with neuropathy, was not observed in patients with nephropathy or retinopathy. The cause of the diabetic microangiopathy cannot be explained only by the overactivity of the polyol pathway, because it may be attributed to the interaction of multiple metabolic, genetic and environmental factors. Therefore the difference might indicate greater participation of overactivity of the polyol pathway in the pathogenesis of diabetic neuropathy than of the other microangiopathies. In conclusion, we demonstrated that A Sor and ASor/ABS, obtained from the diet loading test, could become indicators of the presence or severity of diabetic neuropathy. Furthermore, it is speculated that alteration of the polyol pathway plays a significant role in the pathogenesis of diabetic neuropathy. Further studies are needed to determine whether ASor and ASor/ABS are also useful for selecting patients at high risk of developing neuropathy.
Acknowledgement We thank Tomoko Nomura Yamada for their technical help.
and
Shigemi
References 1 Cogan, D. G., Kinoshita, J. H., Kador, P. F. et al. (1984) Aldose reductase and complications of diabetes. Ann. Intern. Med. Intern. Med. 101, 82-91. 2 Beyer-Mears, A., Cruz, E., Edelist, E. and Varagiannis, E. (1986) Diminished proteinuria in diabetes mellitus by sorbinil, an aldose reductase inhibitor. Pharmacology 32, 52-60. 3 Ghahary, A., Luo, J., Gong, Y., Chakrabarti, S., Sima, A. A. F. and Murphy, L. J. (1989) Increased renal aldose reductase activity, immunoreactivity, and mRNA in streptozocin-induced diabetic rats. Diabetes 38, 1067-1071. 4 Malone, J. I., Knox, G., Benford, S. and Tedesco, T. A. (1980) Red cell sorbitol; an indicator of diabetic control. Diabetes 29, 861-864. 5 Nagasaka, U., Fujii, S. and Kaneko. T. (1988) Human erythrocyte sorbitol metabolism and to role of sorbitol dehydrogenase. Diabetologia 3 1, 766-770. 6 Lapolla, A., Poli, T., Valerio, A. and Fedele, D. (1985) Is red cell sorbitol content a good marker of glycemic control in diabetic patients?. Diabetes Res. 2, 283-286. 7 Price. D. E., Grant, P. J.. Airey. C. M., Stickland, M. H., Kemp, J. and Wales, J. K. (1987) Red cell sorbitol levels during hyper and hypoglycaemic clamp studies in insulindependent diabetic patients. Diabetic Med. 4, 229-232. 8 Hotta, N.. Kakuta, H., Fukasawa, H. et al. (1985) Effects of fructose-rich diet and the aldose reductase inhibitor, ONO-2235, on the development ofdiabetic neuropathy in streptozotocin-treated rats. Diabetologia 28, 176-180. Y Yoshida, T., Nishioka, H., Yoshioka, K., Nakano. K., Kondo, M. and Terashima, H. (1987) Effect of aldose reductase inhibitor ON0 2235 on reduced sympathetic nervous system activity and peripheral nerve disorders in STZ-induced diabetic rats. Diabetes 36. 6-l 3. IO Srivastava, S. K., Ansari. N. H., Hair, G. A., Awasthi, S. and Dan, B. (1986) Activation of human erythrocyte, brain, aorta, muscle, and ocular tissue aldose reductase. Metabolism 35. 114-l 18. II Srivastava, S. K., Hair, G. A. and Das, B. (1985) Activated and unactivated forms of human erythrocyte aldose reductase. Proc. Nat]. Acad. Sci. U.S.A. 82, 7222-7226. 12 .4ida, K., Tawata, M., Shindo. H. and Onaya, T. (1990) Clinical significance oferythrocyte sorbitol-blood glucose rations in type II diabetes mellitus. Diabetes Care 13. 461-467. 13 Hotta, N., Sakamoto, N., Fukuda, M. et al. (1990) Epalrestat can truly prevent diabetic retinopathy in clinical double-blind study. Diabetes 39 (Suppl. 1 ), 61A.