Pupil cycle time and contrast sensitivity in type II diabetes mellitus patients: a pilot study.

CONTEXT: Pupil cycle time (PCT) has been widely used for examination of ocular diabetic autonomic neuropathy (DAN). AIMS: The primary aim of this study was to evaluate the difference of contrast sensitivity according to PCT results, and the secondary aim was to determine the factors associated with PCT difference for type II diabetes patients. SETTINGS AND
DESIGN: A clinical pilot study that included a total of 60 eyes of 60 type II diabetes patients.
MATERIALS AND METHODS: We divided all patients into three groups according to PCT results. Group A and Group C were composed of patients who had upper one third PCT and lower one third PCT, respectively. We analyzed difference of age, diabetes duration, hypertension duration, mean best corrected visual acuity (BCVA), mean spherical equivalent (SE), HbA1C, glomerular filtration rate (GFR), stage of diabetic retinopathy, and Cardiac Autonomic Function Score (CAFS). Contrast sensitivity and decrease of visual acuity by glare were also estimated and analyzed for 28 eyes of 28 non-proliferative diabetic retinopathy (NPDR) patients. Group [NPDR] A and Group [NPDR] C were defined as those who had lower one third PCT and upper one third PCT, respectively. STATISTICAL ANALYSIS: Statistical analysis was done using SPSS 17.0 software. RESULTS AND
CONCLUSIONS: Each group contained 20 eyes. Significant differences between Group A and Group C were observed in duration of diabetes and CAFS (P ≤ 0.001 and P < 0.001, respectively). Contrast sensitivity in Group [NPDR] A was significantly more than in Group [NPDR] C at all cpds (all P ≤ 0.001). We found that PCT can influence contrast sensitivity or glare in diabetes patients and also confirmed a significant correlation of PCT with CAFS and duration of diabetes.

Diabetic autonomic neuropathy (DAN) can involve autonomic nervous systems, such as the cardiovascular system, gastrointestinal system, urogenital system, thermoregulation, and pupillary reflex. Symptoms can include tachycardia, painless myocardial infarction, orthostatic hypotension, gastroparesis, diarrhea, constipation, erectile dysfunction, neurogenic bladder, hypoglycemia, and sweating disturbance. However, despite its significant negative impact on survival and quality of life, DAN is among the least recognized and understood complications of diabetes mellitus.[1] Several diagnostic methods are useful for evaluation of systemic DAN. One of these is the cardiac autonomic function test, which estimates change of blood pressure or heart rate according to respiration or position change. Another test is pupil cycle time (PCT) for evaluation of ocular DAN. Since it was first introduced in 1978 by Miller et al, PCT has been widely used for examination of ocular DAN.[2] Because it has been known to precede cardiovascular DAN, PCT could be an early diagnostic tool for ocular DAN.[3] In our diabetic retinopathy clinic, we had often encountered some patients with DAN complaining about the symptoms of glare and reduced visual quality. Under this background, we performed this study. The primary aim was to evaluate the difference of contrast sensitivity according to PCT results, and the secondary aim was to determine the factors associated with PCT difference for type II diabetes patients.

Materials and Methods

A clinical pilot study was performed. Initially, 214 type II diabetic patients in our hospital were recruited for this study. All patients had been diagnosed as type II diabetes in our hospital from May 2007 to May 2009. Of these, we excluded some patients by exclusion criteria, which were as follows: 1) best corrected visual acuity (BCVA) less than 0.1 according to the Snellen eye chart, 2) cataract greater than NO2 or NC2 by LOCS III classification, 3) patients with systemic diseases that could affect pupil, 4) patients with a history of medication that might affect autonomic function, 5) patients with corneal opacity or dystrophy, uveitis, rubeosis iridis, glaucoma, vitreous hemorrhage, previous ocular trauma, or ocular surgery, 6) pseudophakic or aphakic patients, 7) patients who had been on full session panretinal photocoagulation (PRP), 8) patients less than 40 years old or more than 80 years old. Finally, this study was advanced to 60 eyes of 60 participating patients.

We examined basic information for 60 patients, including age, BCVA, mean spherical equivalent (SE), duration of diabetes mellitus, duration of hypertension, HbA1C, glomerular filtration rate (GFR), and stage of diabetic retinopathy. We estimated Cardiac Autonomic Function Score (CAFS) and PCT. We then divided all patients into three groups according to PCT results. Group A was composed of patients who had lower one third PCT, Group B of patients who had middle one third PCT, and Group C of patients who had upper one third PCT. We analyzed the difference of several factors, such as age, BCVA, SE, duration of diabetes, duration of hypertension, HbA1C, GFR, stage of diabetic retinopathy, and CAFS according to each group. Then, for achieving the primary aim of this study, we estimated contrast sensitivity and reduction of BCVA by glare effect in 28 non-proliferative diabetic retinopathy (NPDR) patients. Because contrast sensitivity could be influenced by progression of diabetic retinopathy, we excluded “no diabetic retinopathy (DR)” and “Proliferative diabetic retinopathy (PDR)” patients in order to reduce this bias. The NPDR patients who had lower one-third of PCT (n = 9) were classified as Group [NPDR] A. The NPDR patients who had upper one-third of PCT (n = 9) were classified as Group [NPDR] C.

Measurement of pupil cycle time

PCT was measured using the method described by Miller et al.[2] The patient was seated at a slit lamp in a dimly illuminated room and asked to look into far distance. A thin horizontally aligned beam of light with moderate intensity, measuring 9 mm in length and 0.5 mm in width was focused from below on the inferior pupillary margin, for initiation of pupil cycle constriction and dilation. PCT was measured by a hand-held electronic stopwatch measuring 1/100th of a second. The time taken by 90 cycles (three runs of 30 cycles each) in seconds was multiplied by 1000/90 to obtain PCT in milliseconds/cycle. All measurements were performed by one author.

Cardiac Autonomic Function Score

With help from an endocrinologist at our hospital, we examined cardiovascular autonomic function using the DICAN (Diabetic Cardiovascular Autonomic Neuropathy, Medicore Co. Ltd., Seoul, Korea). The test included five methods. (1) Beat-to-beat variation in heart rate during deep breathing: This test analyzed beat-to beat variation with subjects lying quietly and breathing deeply at approximately 6 breaths/minute. Using an instantaneous monitor, the difference between maximum and minimum heart rate was measured. (2) Beat-to-beat variation during the Valsalva maneuver: The subject was asked to blow into a mouthpiece connected to a manometer held at 40 mmHg pressure for 15 seconds; using a continuous electrocardiogram (ECG), the response could be expressed by the Valsalva ratio, which is the longest RR interval, which is duration of ventricular cardiac cycle, after the maneuver to the shortest RR interval during the maneuver. (3) Heart rate response to standing (RR30/RR15): The test can be simplified using a continuous ECG recording; the lengths of the RR intervals at beats 15 and 30 after standing provide the “30:15” ratio. (4) Systolic blood pressure response to standing: Postural systolic hypotension was detected using a cuff sphygmomanometer. (5) Blood pressure response to isometric handgrip: A simple test based on this reflex uses a handgrip dynamometer standardized at 40% of the maximum voluntary contraction with measurement of blood pressure. The difference of diastolic blood pressure was then confirmed[4-6] [Table 1].

Table 1

Standard sets for cardiac autonomic function score using diabetic cardiovascular autonomic neuropathy

Glare and contrast sensitivity

We examined contrast sensitivity using the Visual Capacity Analyzer (VCA, L2 informatique, Paris, France). It was estimated under a photopic condition of 100 cd/m2 and a mesopic condition of 30 cd/m2 using the Landolt rings at 3 cpd, 4.8 cpd, and 7.5 cpd. For evaluation of glare effect, we estimated reduction of BCVA when the glare condition was added.

The institutional review board at our medical center approved the study protocol prior to the start of recruitment. We received informed consent from all patients in the study. All statistical analyses were performed using SPSS software, version 17.0 (SPSS Inc., Chicago, IL, USA).

Results

Each group contained 20 eyes. Mean ages of Group A, Group B, and Group C were 56.30 ± 5.66 years, 55.30 ± 3.88 years, and 58.55 ± 9.22 years, respectively. Mean SEs of Group A, Group B, and Group C were 0.05 ± 1.53 D, –0.46 ± 1.32 D, and 0.26 ± 1.60 D, respectively. Mean HbA1C values of Group A, Group B, and Group C were 8.95 ± 0.51%, 8.61 ± 0.56%, and 9.19 ± 0.61%, respectively. Mean GFRs of Group A, Group B, and Group C were 70.39 ± 10.57 ml/min/1.73 m2, 80.31 ± 8.28 ml/min/1.73 m2, and 66.06 ± 13.68 ml/min/1.73 m2, respectively. Mean duration of diabetes mellitus of Group A, Group B, and Group C was 8.00 ± 2.29 years, 9.55 ± 2.01 years, and 14.30 ± 1.45 years, respectively. Mean duration of hypertension of Group A, Group B, and Group C was 9.30 ± 3.96 years, 8.10 ± 5.77 years, and 9.80 ± 2.75 years, respectively. Mean PCTs of Group A, Group B, and Group C were 898.05 ± 35.24 ms, 942.75 ± 30.01 ms, and 1032.80 ± 53.51 ms, respectively.

Age, hypertension, HbA1C, and GFR between Group A and Group C showed no significant difference. Duration of diabetes and CAFS showed significant difference between Group A and Group C (P ≤ 0.001 and P <0.001, respectively; by Mann-Whitney U test) [Table 2]. Correlation coefficient between PCT and duration of DM was 0.905. Correlation coefficient was 0.538 between PCT and CAFS. Both results showed significant correlation [Fig. 1a and b]. PCT was analyzed according to the stage of diabetic retinopathy for a total of 60 eyes. PCTs for no DR, NPDR, and PDR were 911.13 ± 61.59 ms, 960.03 ± 49.30 ms, and 990.81 ± 70.13 ms, respectively (P = 0.008; by Kruskal–Wallis test) [Table 3].

Table 2

Clinical characteristics of study subjects

Figure 1

(a) Scatter of pupil cycle time and diabetes mellitus duration. Correlation coefficient was 0.905 and P value by Spearman correlation efficient test was less than 0.001. (b) Scatter of PCT and Cardiac Autonomic Function Score. Correlation coefficient was 0.538 and P value by Spearman correlation efficient test was less than 0.001

Table 3

Pupil cycle time and cardiac autonomic function score in no diabetic retinopathy, non-proliferative diabetic retinopathy and proliferative diabetic retinopathy

In order to find out the primary outcomes of this study, we analyzed contrast sensitivity with VCA and BCVA reduction by glare effect for 28 eyes of NPDR. Fig. 2a shows contrast sensitivity of Group [NPDR] A and Group [NPDR] C under the photopic condition. These resulted in 1.696 versus 1.618 at 3 cpd, 1.565 versus 1.470 at 4.8 cpd, and 1.035 versus 0.901 at 7.5 cpd, respectively, and all P values by Mann–Whitney U-test were less than 0.001. Fig. 2b shows contrast sensitivity of Group A and Group C under the mesopic condition. These resulted in 1.465 versus 1.380 at 3 cpd, 1.113 versus 0.938 at 4.8 cpd, and 0.573 versus 0.535 at 7.5 cpd, respectively, and P values were all <0.001. Reduction of BCVA by glare effect was –0.13 ± 0.24 in Group [NPDR] A and –0.42 ± 0.20 in Group [NPDR] C. The difference was significant (P = 0.026 by Mann–Whitney U-test) [Table 4].

Figure 2

(a) Contrast sensitivity of Group [NPDR] A and Group [NPDR] C under the photopic condition. These resulted in 1.696 versus 1.618 at 3 cpd, 1.565 versus 1.470 at 4.8 cpd, and 1.035 versus 0.901 at 7.5 cpd, respectively, and all P values by Mann–Whitney U-test were less than 0.001. (b) Contrast sensitivity of Group A and Group C under the mesopic condition. These resulted in 1.465 versus 1.380 at 3 cpd, 1.113 versus 0.938 at 4.8 cpd, and 0.573 versus 0.535 at 7.5 cpd, respectively, and all P values by Mann–Whitney U-test were less than 0.001

Table 4

Amount of best corrected visual acuity reduction between Group [NPDR] A and Group [NPDR] C

Discussion

Clinical symptoms of autonomic neuropathy generally do not occur during the early period of diabetes mellitus. However, subclinical autonomic dysfunction can occur within 1 year of diagnosis in type II diabetes patients and within 2 years in type I diabetes patients.[7] If subclinical DAN is neglected, it can give rise to symptoms such as diarrhea, constipation, erectile dysfunction, silent myocardial infarction, and even sudden death.[8910] Therefore, diabetes patients should be evaluated for diagnosis of subclinical DAN, even when they have not experienced clinical symptoms. Cardiovascular function test and PCT are two representative methods for evaluation of DAN.

A small beam or slit of light focused on the pupillary margin will induce regular, persistent oscillations of the pupil. PCT is the period of this cycle. Theoretically, PCT is concerned not only with afferent signals, such as retina, optic nerve, and optic tract, but also with efferent signals, such as oculomotor nerve, ciliary ganglion, and iris sphincter muscle. If any part of this pupillary reflex arc is injured, PCT can be prolonged. Previous studies have reported on several characteristics of prolongation of PCT. Alio et al,[11] reported that sympathetic denervation may occur more dominantly than parasympathetic denervation in diabetes patients with more than 10 years duration. Tadayuki et al,[12] reported on histologic differences of iris muscle cells between diabetes patients and normal controls. Three characteristics of iris muscle cells in diabetes patients included concentric lamellar appearance, many pigment granules, and many lipid droplets. These findings were more frequent in patients who had a longer duration of diabetes and poor glycemic controls. They were also more frequent in dilator muscles than in sphincter muscles.

According to previous reports, because the cardiac autonomic function test and PCT have significant correlation, they could be used as complimentary diagnostic tools for DAN.[13] In our study, we were also able to confirm that PCT showed significant correlation with CAFS Fig. 1b After analysis of 30 eyes in type II diabetes patients, Datta et al,[14] reported that PCT showed a statistically significant difference between controls and patients with PDR. In our study, we were also able to confirm that PDR patients had longer PCT than NPDR or no DR patients.

Contrast sensitivity has been considered as a useful tool for determination of visual quality in patients who complain of visual discomfort, but have good visual acuity.[15] Good retinal image quality and retinal illumination are possible through proper reduction in pupil size diameter, which is maintained by normal pupillary reflex with normal autonomic function. Juan et al,[16] reported that contrast sensitivity could be enhanced when the surrounding luminance, operated by proper pupil size, is about 75% of the target luminance. In our study, we examined contrast sensitivity and decrease of visual acuity with glare in NPDR patients. Group [NPDR] C showed lower contrast sensitivity than Group [NPDR] A in 3 cpd, 4.8 cpd, and 7.5 cpd [Fig. 2], and Group [NPDR] C showed a greater decrease of BCVA with glare than Group [NPDR] A. According to this result, we might think that prolonged PCT could have an influence on reduction of contrast sensitivity and cause a decrease of visual acuity by glare.

The tinted lens could be a useful tool for management of patients with low contrast sensitivity or glare. Use of a tinted lens may be effective for a decrement of the chromatic aberration effects, a brightness increment, scattering reduction, or a decrement of lenticular fluorescence.[17-20] Because the decrement of blue radiation arriving at the eye could reduce the most scattering effects, the effect is strongest with yellow and orange filters. de Fez et al,[21] also reported that because it is capable of enhancing contrast sensitivity with minimum disturbance to the user's chromatic vision, use of the yellow filter would be more advisable. Even though most tinted filters reduce stimulus luminance, retinal contrast increments have been found, and such increments probably play a role in the feeling of improved visual quality reported by some subjects.[192223]

In conclusion, first, we found that PCT difference could influence visual quality or glare in diabetes patients. From these results, we think that if diabetes patients without severe cataract or PDR complain of glare, particularly in a bright environment, regardless of their visual acuity, it may be caused by the prolongation of PCT. Examination of contrast sensitivity and prescription of tinted lenses may then be necessary. Second, we confirmed that PCT had significant correlation with CAFS, diabetes duration, and stage of diabetic retinopathy, as mentioned in previous reports.

This clinical pilot study could show the likely success and feasibility of a much larger study investigating the association between PCT and contrast sensitivity in type II diabetes patients. Even though we tried to recruit as many samples as possible, we admit that small samples of this study may cause weakening in statistical results. Larger prospective studies will be needed for the assurance of our conclusion.