The objective assessment of visual contrast sensitivity by pattern reversal visual evoked potentials in diabetes

The ObjectiveAssessmentof Visual ContrastSensitivity by Pattern ReversalVisual EvokedPotentialsin Diabetes

V. Martinelli

ABSTRACT

M. Lacerenza

Recent studies have described abnormalities of visual evoked potentials and pattern electroretinography in diabetics without retinopathy. The visual contrast sensitivity, determined by psychophysical tests, has proved to be abnormal in diabetic patients with and without clinical retlnopathy. In this study we evaluated contrast sensitivity function using both electrophyslologic and psychophysical methods. The objective assessment of functional visual contrast sensitivity was superior to psychophysical evaluation in the detection of contrast sensitivity alterations. No relationships were found between contrast sensitivity dysfunction and abnormalities of pattern electroretinography or fluorescein angiography. Our data suggest that functional visual deficits might precede background retinopathy and that the involvement of fovea1function is early and very frequent in diabetic patients, even if they have normal visual acuity. (The Journal of Diabetic Complications 2;1:44-46, 1966.)

M. Merenda F. Meschi L. Somazzi G. Comi

Department of Neurology and Pedia tries, Scientific Ins tit&e S. Ra f faele, and the Department of Ophthalmology, L. Sacco Hospital, University of Milan, Italy

INTRODUCTION Retinal microangiography has long been considered the only ophthalmologic complication in diabetes. Recent studies1-3 have revealed abnormalities in pattern reversal visual evoked potentials (VEP) and pattern reversal electroretinography (PERG) in diabetic patients with normal visual acuity and withoutfluorangiographic signsof retinopathy. These electrophysiologic findings are thought to be expressions of a subclinical involvement of the optic pathways in diabetes’ not detected by such conventional visual acuity measurements as Snellen octotypes. Snellen octotypes are of limited value for the assessment of visual function since they only evaluate the ability to detect high contrast stimuli, whereas in some visual system pathologies it is contrast sensitivity (CS) that is involved, especially in the early stages. In diabetes, contrast sensitivity (CS) function, as assessed by psychophysical tests, is frequently altered in patients both with and by without retinopathy. 4,5 However, assessment of the CS threshold psychophysical tests is subjective and can be criticized. For these reasons we decided to evaluate CS function in a group of diabetic patients using an objective electrophysiologic test.6,7 In addition, VEP, PERG, CS tests, and fluorescein angiography were done to determine which portions of the visual systems of these patients were functioning abnormally.

PATIENTS

Reprint requests: Vittorio Martinelli, Istituto Scientific0 S. Raffaele, Dipartimento di Neurologia, via Olgettina 60, 20132 Milano, Italy.

44

AN0 METHODS

Thirty insulin-dependent diabetic teenagers (16 boys and men, and 14 girls and women, mean age 16.6 years, range lo-20 years) took part in the study. They were selected as having a Snellen visual acuity of 1 .O or more, after correction where necessary. The duration of the disease ranged from 1 to 17 years (mean 6.9 years), and the mean level of total stable glycosylated hemoglobin was 11.8% (SD f 2.7). All patients were investigated for retinopathy by fluorescein angiography performed as follows: 2.5-5 ml of sodium fluorescein were injected

45

CONTRASTSENSITIVITYFUNCTIONIN DIABETES

intravenously and after a given time photographs of the fundi were taken (Topcon equipment). VEP and PERG were performed with MK7 Amplaid equipment, and Pattern Reversal Stimulation accomplished with a black and white checkerboard television screen. For VEP the check sizes were 11, 5.2, and 2.6 mm, with respective visual angles of 30 min and 15 min (21 mm and 11 mm for PERG). The reversal rate was 1.6 Hz. The PERG was evaluated at 10000 and 50% contrast levels. Objective electrophysiologic evaluation of contrast sensitivity (VEPCST) was determined by the evaluation of VEP latency at three different contrast levels: 50%, 25%, and 10% with checksof 2.6 mm; theoverall luminescence was60cd/m2. Normal values of VEP-CST were assessed in 16 healthy subjects, matched for sex and age. In diabetic patients values exceeding the mean value + 2.5 SD of normal controls were considered abnormal. VEP were recorded by a midoccipital electrode placed 5 cm above the external occipital protuberance and referred to a vertex electrode using monocular full-field stimulation. The major positive peak (PlOO wave) latency and the N75-PlOO amplitude were measured. PERG were recorded simultaneously in both eyes by a gold foil electrode inserted under the lower eyelid. An earlobe electrode was used as reference. We measured the peak latency of the b wave and the a-b peak to peak amplitude. Psychophysical contrast sensitivity function was determined by using Cambridge Low Contrast Gratings. In this test, gratings aregenerated by computer, as described by Della Sala.’ This is a new forced-choice test that allows a quick and reliable determination of contrast sensitivity at a spatial frequency of 4 cycles/ degree. A series of eight pairs of plates, one containing the gratings and one blank, was presented in random order at a total of three times for each eye. The first pair of the series has gratings with high contrast which serves as an example, while the gratings in subsequent plates decrease in contrast. The observer was forced to choose whether the top or the bottom plate contained the gratings, guessing if necessary. The score range was O-24, and the Psychophysical-Contrast Sensitivity Test (P-CST) was considered abnormal when at least one eye had a performance worse than the mean value for the controls minus 2.5 SD. This examination was performed in only 16 of our patients.

TABLE

1

Mean Values of VEP Latency Between Normal and Diabetic Controls Controls (N = 16) Mean

Latency PlOO Latency PlOO Latency PlOO Latency PlOO Latency PlOO Psychophysical CST test

30’ 15’ 7.5’ 50% 7.5’ 25% 7.5’ 10%

’ PlOO not recognizable tPlO0 not recognizable *Controls: n = 55. SDiabetics: n = 16

107.36 111.52 124.23 129.71 141.03 21.65$

Dlabatlcs (N=36)

S.D.

Mean

S.D.

p<

4.50 5.20 7.5 10.10 ii.58 2.39

113.15 i la.59 133.06 139.07’ 149.20t 19.844

4.86 5.90 11.36 14.76 la.31 2.21

0.001 0.001 0.001 0.005 0.05 0.001

in three eyes. in 26 eyes.

diabetic and normal controls was significant for the contrast levels (Table 1); at the lowest contrast level PlOO was undetectable in 28 eyes, while it could not be detected in only one control eye. P-CST was abnormal in four of 16 patients. The mean score for patients on P-CST was significantly different from that for the control group (19.84 versus 21.65; p < 0.001). The relationship between VEP-CST and P-CST in 16 patients (32 eyes) was examined at 25% and 10% contrast levels. At 25% contrast, both VEP-CST and P-CST were abnormal in three eyes, only VEP-CST was abnormal in four eyes, and only P-CST was abnormal in one eye. At 10% contrast, VEP-CST was abnormal in all the four eyes that had abnormal P-CST and in another 12 eyes with normal P-CST. We found no significant differences in the PERG mean latency between diabetics and normal controls. Mean PERG amplitude was significantly reduced in diabetic patients with all types of stimulation. Abnormalities of PERG amplitude were detected in ten patients. When compared with the same spatial frequency stimulations (30 min check size), all patients with prolonged VEP latency had abnormal PERG amplitude. Fluorescein angiographic examination showed signs of background retinopathy (microaneurysms) in only three subjects.

DISCUSSION RESULTS At 5000 contrast, the PlOO latency of VEP was abnormal in two patients when thecheck sizestimulation was at 30’, in five patients when it was 15’, and in seven when it was 7.5’. In diabetic patients, the mean values of Pl 00 latency were 113.15 + 4.86 msec for 30’ check size (107.36 + 4.50 in normal controls), 118.59 & 5.90 for 15’ check size (111.52 f 5.20 in normal controls), and 133.06 -I 11.36 for 7.5’ check size (versus 124.23 i 7.50 in normal controls). VEP-CST was evaluated only for the 7.5’ check size. The number of patients with abnormal VEP-CST were as follows: seven (3 monocular and 4 binocular) at 50% contrast, nine (4 monocular and 5 binocular) at 25%, and 19 (5 monocular and 14 binocular) at 10% contrast. The difference in the mean values of VEP latency between

In diabetic patients, the frequency of abnormal VEPs increased with increasing spatial frequency of stimulation, reaching a value of 23% at 6 cycles per degree. The frequency of abnormalities in CS function became even higher when the results of the electrophysiologic contrast sensitivity examinations were also considered; the VEPCST were abnormal in 5000 of the eyes examined, while the frequency of P-CST abnormalities was much less ( 13%) There may be false positives among the results for the VEP-CST, because at low contrast levels the cooperation of the patient is essential for a reliable VEP response. As one of our normal controls had an undetectable VEP at 10% contrast, wecan assume that patient cooperation was usually even less. The psychophysical evaluation of

MARTINELLIET AL.

contrast sensitivity is, instead, very specific, since all patients with abnormal P-CST also had abnormal VEPCST. Contrast sensitivity abnormalities were unrelated to retinal microangiography as assessed by fluorescein angiography, probably because diabetic retinopathy is often highly localized. We have investigated contrast sensitivity function at spatial frequencies of 4 or more cycles per degree, suitable for measuring the fovea1 function within the central 2 degrees of the visual field. Our CS data, the frequent abnormalities in PERG and the observation that VEP abnormalities are more frequent at a spatial frequency of 4 cycles per degree suggest that fovea1 function is particularly involved in diabetes.

ACKNOWLEDGMENT This study was supported by the CNR Project, “Preventive and Rehabilitative Medicine,” Subproject “Degenerative Diseases of the Nervous System”; Project N. 8500719. 56.115.

REFERENCES 1. Puvanendran K, Devathasan G, Wong PK: Visual evoked responses in diabetes. J Neural Neurosurg Psychiatry 46: 643-7, 1963. 2. Cirillo D, Gonfiantini E, De Grandis D, Bongiovanni L, Robert JJ, Pinelli L: Visual evoked potentials in diabetic children and adolescents. Diabetes Care 7~273-5. 1984. Martinelli V, Merenda M, Natali Sora M, Meschi F, Beccaria L, Comi G: Correlation between patternelectroretinographyand visual evoked potential in diabetes. Electroenceph Ciin Neurophysiol66:S64, 1987. Hyvarinen L, Laurinen P, Rovamo J: Contrast sensitivity in evaluation of visual impairment due to diabetes. Acta Ophthalmol61:94-101, 1983. Della Sala S, Bertoni G, Somazzi L, Stubbe F, Wilkins AJ: Impaired contrast sensitivity in diabetic patients with and without retinopathy: a new technique for rapid assessment. Br J Ophthalmol69:136-42, 1985. Collier A, Mitchell sensitivity function 1985.

JD: Visual evoked potential and contrast in diabetic retinopathy. t3r Med J 291:248,

Howe JW, Mitchell KW: The objective assessment of contrast sensitivity function by electrophysiological means. Br J Ophthalmol68:626-38, 1984.