Pupillary light reflex in borderline diabetes mellitus

Diabetes Research und Clinicul Practice, 6 ( 1989) 89-94

89

Elsevier

DRC 00246

Pupillary light reflex in borderline diabetes mellitus Nobutoshi

Kuroda, Hiroshi Taniguchi, Shigeaki Baba and Misao Yamamoto’

Second Deparlmrnt of lnternul Medicine und ’ Depurtnlem of‘ Ophthaltnology.Kobr University School

(Received

Pupillary

9 February

light reflex; Cardiac

release: Oral glucose

1988, revision

beat-to-beat

tolerance

received

variation:

1I July 1988, accepted

Borderline

of Medicine. Kobe, Japan

I I July 1988)

diabetes mellitus; Early diabetes

mellitus; Early phase insulin

test

Summary

It is well known that in diabetes mellitus the early phase insulin release response to glucose becomes blunted. Besides, autonomic neuropathy develops frequently. The autonomic nerve function is related to insulin release in normal subjects. Therefore, in the present study the autonomic function was investigated in different degrees of the early phase insulin release response to glucose loading in borderline diabetes mellitus (B-DM) as well as the early stage of diabetes mellitus (E-DM). In our study the pupillary light reflex seemed to be more sensitive for the detection of autonomic neuropathy than the cardiac beat-to-beat variation in B-DM and E-DM. The light reflex, therefore, was used and the following results were obtained. (1) Autonomic neuropathy was detected even in B-DM, though it was slighter than in E-DM. (2) The parasympathetic nerve appeared to be impaired more than the sympathetic in B-DM and this was especially the case in subjects with relatively good insulin release response. Therefore, the suppression of the early phase insulin release in B-DM may be caused by this autonomic neuropathy already present. (3) In B-DM with reduced insulin release autonomic neuropathy seemed to be more advanced and it became more remarkable in E-DM. Particularly the sympathetic abnormality was loaded in this regard. These observations suggest that it is necessary to consider an involvement of autonomic nerve dysfunction for the understanding of the pathogenesis of B-DM.

Address

for correspondence:

ond Department of Medicine.

of Internal

Nobutoshi Medicine.

5-l. Kusunoki-cho

Kuroda,

M.D..

Kobe University

7 chome.

Chuo-ku,

School

Kobe 650,

Japan.

0168-8227/X9/$03.50

Sec-

‘(‘1 1989 Elsevier Science Publishers

Introduction

The early phase insulin release response to glucose is low or absent in diabetes mellitus (DM) and is

B.V. (Biomedical

Division)

90 also low in most cases of borderline

diabetes

tus (B-DM).

by a reduction

This may be explained

in /$cell number

and/or

insulin

content

melli-

in the pan-

ment in blood

glucose from the fasting

min after the glucose ( >0.5),

group

II

loading.

(0.2 <

level to 30

They were group

50.5)

and

group

I III

( 5 0.2) and four, six and seven people were included

creas or impairment release [I]. Regarding

of the mechanism of insulin the mechanism of insulin re-

in these three groups respectively.

The subjects with

lease, the autonomic

nervous

E-DM

3 years of onset

system plays an im-

portant

role [2]. The disturbance

nervous

system is often encountered

of this autonomic in diabetics

[3].

Thus, it is interesting to know the relationship between early phase insulin release and autonomic neuropathy in diabetes. Especially the autonomic nervous dysfunction in different degrees of the early phase insulin response to glucose is worth studying. since the disturbance of the autonomic nerve is well known to cause the suppression of insulin release

PI. In order to assess autonomic neuropathy, testing of the cardiovascular reflex such as heart rate variation and Schellong’s test is used widely. Besides, the recent development of infrared videopupillography has made objective and quantitative measurement of pupillary light reflex possible [4]. And this method has been now put into clinical practice as well to evaluate autonomic nerve function. Using this method, we have already reported a reduced pupillary area and light reflex in DM and B-DM [5,6]. In the present study we pushed forward this study to evaluate autonomic nerve function using heart rate variation during deep breathing (beat-tobeat variation, BBV) and pupillary light reflex in various degrees of early phase insulin release in BDM and the early stage of diabetes

(E-DM).

Subjects and methods Seventeen people with B-DM, six patients with EDM and nine non-diabetics were subjected to the study. They were all male and aged between 41 and 58 years. DM and B-DM were decided by oral glucose tolerance test (OGTT, 50 g) based on the recommendations of the Japan Diabetic Society for criteria of DM and B-DM [7]. Moreover, B-DM was divided into three groups according to the degree of insulinogenic index, i.e., the increment in insulin from the fasting level divided by the incre-

included

diabetics

within

and with an insulinogenic index less than 0.2. They were all on a diet and none of them had cataract, diabetic

retinopathy

and symptoms

of autonomic

neuropathy. The insulinogenic index was 0.76 f 0.15 (mean f SEM), 0.26 f 0.03, 0.1 I * 0.02 and 0.14 i 0.03 in groups I, 11, III of B-DM, and E-DM respectively. That was significantly different among groups I, II and III, but similar between group III of B-DM and E-DM. The subjects’ ages were similar, i.e., 50.0 f 1.5, 48.5 * 3.1. 48.7 f 1.2. 48.9 f I.7 and 46.7 f 2. I years in the non-diabetic controls, groups I, II, 111 of B-DM. and E-DM respectively. All subjects had hemoglobin A, (HbA ]) levels within the normal range measured around the time of the examination of autonomic nervous function. They were 6.2 f 0.8, 6.7 * 0.4, 6.4 f 0.3 and 7.7 f 0.2% in groups I. 11, III of B-DM, and E-DM respectively. An electrocardiogram, after a 20-min rest in the supine position, was recorded during five successive maximal respiratory cycles with inspiration and expiration each lasting 5 s, using a computerized device developed in our laboratory (Autonomic R-105”, M.E. Commercial, Osaka, Japan), and the heart rates were calculated from the R-R intervals. The difference between the maximal and the minimal heart rate was obtained in one respiratory cycle and the mean of these difference values in five respiratory cycles was expressed as the BBV. The pupillary light reflex was investigated 15 min after adaptation to darkness using a computerized infrared videopupillograph (HTV-C301. Hamamatsu TV Co., Ltd., Hamamatsu. Japan) installing an open-loop photic stimulator with a LED of 20P 30 nit (cd/m2) for a duration of 1.0 s between 9:00 a.m. and 3:00 p.m. when the diurnal variation of the pupillary light reflex was stable [Xl. The pupillogram is illustrated schematically in Fig. I. In the present study, the pupil area prior to photic stimulus (Al). the amplitude of the constriction in re-

91

PHOTIC

STIMULI

4-p

AREA

VELOCITY

0 I

n B-DM

Fig. 2. Cardiac

beat-to-beat

variation

healthy controls,

groups

Control

ACCELERATION

tus (B-DM) Fig.

1.Pupillogram.

Al: pupil area prior to stimulus

amplitude

of constriction

stimulus

(mm); Tl: latency

initiation

of stimulus and the time of achievement

T2: l/2 constriction T5: 0.63 dilatation ity (mm’/s):

(mm’);

time (ms); VC: maximum constriction

dilatation

(BBV) in non-diabetic

mellitus

diabetes melli-

(E-DM).

The column

the mean and the bar SEM.

to the

of 10% of VC: time (ms);

constriction

velocity

acceleration

prior

between

time (ms); T3: total constriction

VD: maximum

maximum

Dl: pupil diameter

time (ms), i.e., period

E-DM

I, II and III of borderline

and early diabetes indicates

(mm’); A3:

m

(mm’k);

velocAC:

(mm’k’).

sponse to the stimuli of light flashes (A3), the maximum constriction veIocity (VC), the maximum dilatation velocity (VD) and the latency time (Tl) were measured by a computer. Furthermore, the pupillary constriction rate (A3/A 1 x 100) was calculated and investigated as a novel index. The value obtained from the pupillogram was expressed as mean z!z SEM. Statistical analysis of the data was made using Student’s t-test.

In comparison with the non-diabetic controls EDM showed low values in Al (29.0 f 1.7 vs. 19.7 f 2.0mm2; P < O.OOS), VD(12.3 f 1.0~~. 8.4 f 1.1 mm*/s; P -=c 0.05). A3 (15.1 f 0.9 vs. 10.9 * 0.9 mm2; P < 0.005) and VC (38.7 f 2.1 vs. 29.2 f 1.8 mm’js; P < 0.05) (Figs. 4. -7).

350

300

‘3

250

$ L t= 200

Results

There were no differences in BBV among the groups. The BBV was 12.4 & 2.0, 11.3 f 2.7, 12.8 f 1.4, 11.0 f. 1.8 and 11.0 f 1.6 beats/min in nondiabetic controls, groups I, II. III of B-DM, and E-DM, respectively (Fig. 2). In these groups Tl was 300.9 zt 5.9. 287.5 f 8.8, 293.0 f 4.3,294.8 zt 7.1 and 287.0 f 8.7 ms respectively. There were no differences between groups (Fig. 3).

0 E-DM

B-DM

Control

Fig. 3. Latency time (TI) of pupillary light reflex in non-diabetic healthy controls, groups 1. II and III of borderline diabetes mellitus (B-DM)

and early diabetes indicates

mellitus

(E-DM).

the mean and the bar SEM.

The column

92 p
t

pco.2

!

p
,

P
,

8 I

pco.1

!I

20

T c4

10

Q

I Fig. 4. Pupil area prior to stimulus reRex in non-diabetic borderline

diabetes

(E-DM).

healthy

mellitus (II-DM)

The column

E-DM

indicates

groups

mellitus

the mean and the bar SEM.

lxo.05 I

Fig. 6. Amplitude borderline (E-DM).

diabetes

PCO.1

healthy

controls,

indicates

groups

light re-

I, II. and III of mellitus

the mean and the bar SEM.

p
I I

pco.1

E-DM

(A3) of the pupillary

mellitus (B-DM) and early diabetes

The column

3

III

B-DM

of constriction

flex in non-diabetic

I

PCO.2

1 I

hght

I. II and 111 of

and early diabctcs

II

I Control

(Al) of the pupillary

controls.

-

0

m I3-LM

Control

pco.1

p
! I

1

pa1

’

1 I

I

20

I

‘;; l% n\ E 10 5

II

9

0

Control Fig. 5. Maximum

dilatation

reflex in non-diabetic borderline (E-DM).

diabetes

healthy

B-DM

Control

velocity (VD) of the pupillary controls.

mellitus (B-DM)

The column

E-DM

indicates

groups

light

1. II and III 01

and early diabetes

mellitus

the mean and the bar SEM.

B-DM

E-DM

Fig. 7. Maximum constriction velocity (VC) ofthe pupillary light reflex in non-diabetic healthy controls. groups 1, II and 111 of borderline (E-DM).

diabetes

mellitus (B-DM)

The column

indicates

and early diahctes

mellitus

the mean and the bar SEM.

93 pco.2 P
I I

P
I

'

p
‘wo.005

100

'

wo.01

pco.05

Discussion

I

’

-I

I 1

1 ’ ’

I

p40.1

I

I

B-DM

Control

Fig. 8. Constriction

rate (A3/Al

reflex in non-diabetic borderline (E-DM).

diabetes

healthy

x 100) of the pupillary

controls,

mellitus (B-DM)

The column

E-DM

indicates

light

groups I, II, and III of

and early diabetes

mellitus

the mean and the bar SEM.

B-DM showed a higher value of Al in groups I (33.3 f 3.2 mm2; P < 0.005), II (33.4 f 1.4 mm2; P < 0.005) and III (28.3 f 2.0 mm2; P < 0.01) and of VC in group II (35.5 f 2.2 mm2/s; P < 0.05) compared with E-DM (Figs. 4, 7). On the other hand, compared with the non-diabetic controls BDM tended to have low VC in group I (38.7 f 2.1 vs. 32.3 i 3.0 mm’/s; P < 0.1) and VD in group III (12.3 f 1.0~s. 9.5 If 0.9mm2/s; P < 0.1) (Figs. 5, 7). Of the three groups of B-DM group III tended to have the lowest Al value (Fig. 4). The pupillary constriction rate was 54.0 f 2.2, 40.4 f 4.2,39.8 f 2.2,45.7 f 3.5and59.1 f 2.9% in non-diabetic controls, groups I, II, III of B-DM, and E-DM, respectively (Fig. 8). The value in EDM was not different from that in non-diabetics. Compared with the value in non-diabetics as well as E-DM. that in each group of B-DM was lowered. Among the three groups of B-DM, group III tended to have the highest actual value.

The risk of developing diabetes mellitus is high when the earIy phase insulin secretion in the oral glucose tolerance test is low [9]. Therefore, we studied three different degrees of this insulin response in B-DM, i.e., relatively good, poor and the intermediate response group, determined on the basis of their insulinogenic index. The present study did not reveal any significant difference in BBV between these B-DM groups and non-diabetic controls or E-DM. Thus, the detection of autonomic neuropathy in BDM and E-DM seems to be difficult when it is examined with only BBV measurement. On the other hand, using the pupillary light reflex, B-DM, compared with non-diabetics, showed relatively low VC and VD in groups I and III respectively, while EDM, compared with non-diabetics, was significantly high in Al. A3, VC and VD. These results indicate the usefulness of the pupillary light reflex for the detection of autonomic neuropathy in B-DM as well as E-DM and also suggest that autonomic neuropathy is present even in B-DM, though it is slight compared with that in E-DM. Besides relatively low VC and VD in groups I and III respectively compared with non-diabetics, a tendency to a low Al was observed in group III compared with groups I and II. A decrease in Al and VD is considered to be due mainly to the sympathetic disturbance, and that in A3 and VC to the parasympathetic dysfunction [ 10,111. This consideration suggests that the parasympathetic is more easily impaired than the sympathetic nervous system during a period when the early phase insulin release response is still relatively good, and that the sympathetic nervous system seems to become disturbed when the release response worsens. With respect to the latency time in the pupillary light reflex using an open-loop photic stimulator, it has been reported that a delay was not observed in diabetics without neuropathy compared to controls [12], and furthermore, the present study showed that neither was such a delay observed in E-DM and B-DM. In the present investigation, the pupillary constriction rate was calculated as a new index. It did

94 not differ between the non-diabetic controls and EDM, but was significantly low in B-DM. In B-DM. furthermore, group,

constriction tion

the poor

i.e., group

response higher

rate than the other two groups.

Reduc-

seems to be due largely

constriction

rate

to a decrease,

in A3. Since the pupillary

rate is considered sympathetic

secretory a relatively

of the pupillary

significantly,

insulin

III, exhibited

to reflect the balance

and the parasympathetic

in B-DM though

in-

constriction between functions,

the the

present observations suggest that parasympathetic impairment is more common in B-DM. In contrast. E-DM exhibited an elevation in the pupillary constriction rate in spite of a reduction of both A I and A3, because the decrease in the former was greatet than that in the latter. This suggests that the sympathetic function appears to be impaired rather more easily than the parasympathetic. Though the identification of the accurate time of onset of DM and of the prediabetic period is difficult, it is known that autonomic neuropathy frequently occurs almost simultaneously with the development of carbohydrate metabolism dysfunction. Pfeifer et al. found low beat-to-beat variation and small Al in both non-insulinand insulin-dependent diabetes mellitus within 12 and 24 months of the onset respectively. compared with controls [13]. According to our present study, the pupillary light reflex had worsened in diabetics within 3 years of the onset and the decrease in the pupillary constriction rate in B-DM as well. It is well known that the parasympathetic and the sympathetic nerves are involved in the stimulation and inhibition of insulin secretion respectively [14-161. The predominant impairment of the parasympathetics as shown by the present investigation further indicates that the suppression of early phase insulin release in B-DM may be due to the underlying autonomic disturbance. Thus, it is necessary to consider an involvement of autonomic nerve dysfunction for the understanding of B-DM with special reference to the deficient or absent early phase insulin release.

References I Kanazawa,

Y. and Kasuga,

M. (1985) Abnormality

sulin secretion

and its action.

(Eds.). Clinical

Diabetes.

2 Kuzuya.

Kodansha,

N., Yanaihara,

(1980) Insulin secretion Matsumiya 3 Niakan, autonomic 4 Utsumi.

a preliminary

Kodansha,

Tokyo.

Metabolism

report.

pp. 27-31. T.

In: S. Fujii and H.

Y. and Comstock,

neuropdthy.

Society

T. and Kaneko,

and its regulation.

T. (1978) Open-loop

of in-

Diabetic

Tokyo,

N.. Kuzuya.

(Eds.). Insulin.

E.. Hardti,

In: Japan

pp. 3X-49.

J.P. (1986) Diabetic 35, 2244234.

infrared

videopupillography

Acta Sot. Ophthalmol.

-

Jpn. 82, 315-

321. 5 Taniguchl.

H., Inui, M., Ejiri, K. et al. (1985) Reduced

lary area and light reflex in diabetes

to. H. Min and S. Baba (Eds.). Current and

Experimental

Medica.

Amsterdam.

6 IKuroda, pillary

Aspects

Y.. Yamashiro.

ceedings of the 40th Annual

Excerpta

Research.

for diabetes

Y. et al. (1987) Pu-

diabetes

mellitus.

In: Pro-

Meeting of the Japan

Society of

p. 109 (abstract).

N. (1970) The report

standard

in Clinical

Mellitus.

pp. 35&35X.

N., Morimoto.

Neurovegetative

Topics

of Diabetes

light reflex in borderline

7 Kuzuya,

pupil-

mellitus. In: N. Sakamo-

of a committee

of diagnostic

mellitus in glucose tolerance

test. J. Jpn

Diab. Sot. 13, l-7. 8 Ulsumi. diurnal

T.. Onishi. variation

Y. and Hashimoto.

under

open-loop

T. (1978) Pupillary

condition.

Adv. Neural.

Sci. 22. 6 1S-623. Y Kosaka.

K.

( 1985) lnsuflicienl

and H. Matsumiya

(Eds.),

insulin secretion.

Insulin.

Kodansha.

In: S. Fujii Tokyo,

pp.

23-27. 10 Hashimoto. topically

T.. Utsumi, administrated

pillary dynamics

T. and Onishi, adrenergics

under open-loop

Y. (1978) Effect of

and cholinergics condition.

mol. Jpn. 29, I 13% 1147. II Hashimoto, T.. Utsumi, T. and Namba, topically

administrated

on pupillary

dynamics

lia Ophthalmol. 12 Yamba.

cholinolytic

on pu-

Folia Ophthal-

K. (1979) Effects of

and adrenolytic

under open-loop

agents

photic stimulus.

Fo-

Jpn. 30, 1008-1015.

K.. Utsumi,

T. and Kitazawa.

mellitus and pupil (a preliminary

A. (1980) Diabetes

report).

Acta Sot. Ophthal-

mol. Jpn. X4, 398-405. I3 Pfeifer. M.A.. Weinberg. tonomic

neural

subjects.

Diabetes

14 frimble,

D.L. et al. (1984) Au-

in recently

diagnosed

diabetic

Care 7, 447453.

E.R.. Berthoud,

ind Rcnold,

C.R., Cook,

dysfunction

H.R.. Siegel, E.G., Jeanrenaud.

A.E. (1981) Importance

iation of the pancreas

of cholinergic

for glucose tolerance

B.

inner-

in the rat. Am. J.

Physiol. 241, E337-E34l. 15 qishi, S.. Seino. Y., Ishida. H. et al. (1987) Vagal regulation >finsulin, glucagon and somatostatin .al. J. Clin. Invest. 79. 1191-l 196. I6 Esrunicardi, F.C.. Sun. Y.S., Druck. 1. and Andersen.

secretion

in vitro in the

P.. Goulet,

R.J., Elahi.

D.K. (1987) Splanchnic

)f insulin and glucagon secretion ‘reas. Am. J. Surg. 153, 34-40.

neural regulation

in the isolated

human

pan-