May 2018
Volume 59, Issue 6
Open Access
Visual Psychophysics and Physiological Optics  |   May 2018
Reduced Contrast Sensitivity is Associated With Elevated Equivalent Intrinsic Noise in Type 2 Diabetics Who Have Mild or No Retinopathy
Author Affiliations & Notes
  • J. Jason McAnany
    Department of Ophthalmology and Visual Sciences, University of Illinois at Chicago, Chicago, Illinois, United States
    Department of Bioengineering, University of Illinois at Chicago, Chicago, Illinois, United States
  • Jason C. Park
    Department of Ophthalmology and Visual Sciences, University of Illinois at Chicago, Chicago, Illinois, United States
  • Correspondence: J. Jason McAnany, University of Illinois at Chicago, Department of Ophthalmology and Visual Sciences, 1855 West Taylor Street, MC/648, Chicago, IL 60612, USA; jmcana1@uic.edu
Investigative Ophthalmology & Visual Science May 2018, Vol.59, 2652-2658. doi:10.1167/iovs.18-24151
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      J. Jason McAnany, Jason C. Park; Reduced Contrast Sensitivity is Associated With Elevated Equivalent Intrinsic Noise in Type 2 Diabetics Who Have Mild or No Retinopathy. Invest. Ophthalmol. Vis. Sci. 2018;59(6):2652-2658. doi: 10.1167/iovs.18-24151.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: To evaluate explanations for contrast sensitivity (CS) losses in subjects who have mild nonproliferative diabetic retinopathy (NPDR) or no diabetic retinopathy (NDR) by measuring and modeling CS in luminance noise.

Methods: Ten diabetic subjects with NDR, 10 with mild NPDR, and 10 age-equivalent nondiabetic controls participated. Contrast threshold energy (Et) was measured for letters presented in the absence of noise (Et0) and in four levels of luminance noise. Data were fit with the linear amplifier model to estimate inferred noise level within the visual pathway (Neq) and sampling efficiency (ability to use stimulus information optimally). Et0, Neq, and efficiency were compared to clinical characteristics.

Results: Neq was correlated with Et0 for the diabetic subjects (r = 0.93, P < 0.001) and ranged from normal to 12-times the upper limit of normal. ANOVA indicated significant differences among the subject groups for Et0 and Neq (both F > 11.92, P < 0.001). Et0 and Neq were elevated for the mild NPDR group compared to the control and NDR groups (all t > 3.89, P ≤ 0.001); the NDR and control groups did not differ significantly (all t < 0.61, P > 0.55). There were no significant efficiency differences among the groups (F = 1.29, P = 0.29). Neq was correlated significantly with disease duration, microperimetric sensitivity, and Pelli-Robson CS.

Conclusions: Elevated contrast threshold may be associated with increased intrinsic noise in early-stage diabetic subjects. Results suggest that noise-based CS measurements can provide important information about early neural dysfunction in these individuals.

Diabetic retinopathy (DR) is the most serious ocular complication of diabetes mellitus (DM) and is the leading cause of blindness among working-age adults.1 Although DR is commonly considered a vascular disorder and clinical staging is based on the severity of vascular abnormalities,2,3 there is mounting evidence that supports early neural dysfunction in these individuals. For example, contrast sensitivity (CS) has long been known to be reduced in diabetics who have not yet developed clinically-apparent retinopathy,48 and these CS deficits can become more severe as the disease progresses.7,9,10 Despite the sizeable literature showing CS losses in early-stage DR, the pathways and mechanisms underlying their reduced CS remain poorly understood. 
In a study8 designed to better understand the contrast processing pathways that are affected in early-stage DR, CS was measured under the “steady-pedestal” and “pulsed-pedestal” paradigms that have been shown to target CS mediated by the magnocellular (MC) and parvocellular (PC) pathways, respectively.1115 This study showed equivalent CS losses in the MC and PC pathways for diabetics who had no DR (NDR) and in diabetics who had nonproliferative DR (NPDR).8 In a recent study, structure-function relationships in diabetics who had no clinically-apparent retinopathy were studied using microperimetry (MP), a form of CS testing, and optical coherence tomography.16 This study showed that MP sensitivity losses were correlated with retinal ganglion cell layer thinning. Taken together, the existing literature indicates that CS losses may be associated with structural changes of the inner-retina that similarly affect the MC and PC pathways. 
We have previously shown that CS losses in retinitis pigmentosa, an inherited retinal degenerative disease, are highly correlated with estimated noise levels within the visual pathway.17 This finding is consistent with a previous proposal that most retinal diseases impair visual function by increasing internal noise.18 However, in a small sample of patients with age-related macular degeneration and CS loss, only one of four patients had an increased level of internal noise.19 Thus, the extent to which increased levels of noise within the visual pathway can account for CS losses in early-stage DR is unknown and is not readily predicted from the existing literature. 
A common approach to estimate the amount of noise within the visual pathway is to use the “equivalent input noise” method.20 According to this approach, contrast thresholds are measured against a uniform adapting field and also in additive white luminance noise. Threshold is plotted as a function of the externally added noise level, and the plots can be analyzed using the linear amplifier model (LAM), a common model of human performance in noise.1724 The LAM factors performance into two independent components: (1) equivalent intrinsic noise, which is an estimate of the amount of noise within the visual pathway, and (2) sampling efficiency, which represents the subject's ability to make use of stimulus information relative to an ideal observer.20 
In the present study, CS was measured in different levels of luminance noise in diabetic subjects who had either no clinically-apparent DR or have mild NPDR. The data were analyzed using the LAM to derive measures of equivalent intrinsic noise and sampling efficiency. We sought to define the relationships among these measures and to determine the extent to which internal noise elevations and efficiency reductions can account for CS deficits in early-stage DR. Additionally, these measures were compared to subject characteristics including age, disease duration, and glycated hemoglobin (HbA1c) percentage, as well as clinical measures including visual acuity, Pelli-Robson chart CS, and MP sensitivity. 
Methods
Subjects
Twenty subjects diagnosed with type 2 DM participated in the study. The diabetic subjects were recruited from the Retina and General Eye Clinics of the University of Illinois at Chicago Department of Ophthalmology and Visual Sciences. The stage of NPDR was graded by a retina specialist and the subjects were clinically classified as diabetic with NDR (n = 10) or diabetic with mild NPDR (n = 10), according to the Early Treatment of Diabetic Retinopathy Study (ETDRS) scale.2 Subjects classified as mild NPDR had one or more of the following vascular abnormalities: microaneurysms, hard exudates, cotton-wool spots, and/or mild retinal hemorrhage (equivalent to ETDRS level 35 or less2). Clinical characteristics including age, sex, visual acuity, Pelli-Robson chart CS, estimated diabetes duration, HbA1c percentage, and treatment history are provided in Table 1. All treatments were performed at least nine months before recruitment and no subject had clinically significant diabetic macular edema at the time of testing. 
Table 1
 
Subject Characteristics
Table 1
 
Subject Characteristics
Ten visually-normal nondiabetic control subjects also participated. The mean age of the control subjects did not differ significantly from that of either diabetic group (F = 0.35, P = 0.71). All control subjects had best-corrected visual acuity of 0.04 log minimum angle of resolution (MAR; equivalent to approximately 20/22 Snellen acuity) or better, as assessed with the Lighthouse distance visual acuity chart, and normal letter CS, as measured with a Pelli-Robson chart. The research followed the tenets of the Declaration of Helsinki and was approved by an institutional review board of the University of Illinois at Chicago. All subjects provided written informed consent. 
Stimuli and Instrumentation
The instrumentation has been described in detail elsewhere.25,26 In brief, stimuli were generated using the Cambridge Research Systems ViSaGe stimulus generator (Cambridge Research Systems, Ltd., Rochester, Kent, UK) and were displayed on a Mitsubishi Diamond Pro 2070 cathode ray tube monitor (Mitsubishi Electric Corp., Tokyo, Japan). The screen resolution was 1024 × 768 and the refresh rate was 100 Hz. The monitor, which was the only source of illumination in the room, was viewed monocularly through a phoropter with the observer's best refractive correction. A 3.0-mm artificial pupil was mounted on the eyepiece of the phoropter to control retinal illuminance. The luminance values used to generate the stimuli were determined by the ViSaGe linearized look-up table, which were verified by measurements made with a Minolta LS-110 photometer (Konica Minolta, Tokyo, Japan). The temporal characteristics of the display were confirmed using an oscilloscope and photocell. 
Test targets were 0.9 log MAR (approximately 20/160 Snellen equivalent) Sloan letters (C, D, H, K, N, O, R, S, V, and Z) that were constructed according to published guidelines.27 Previous work has shown that these 10 letters have similar thresholds for measurements made in the presence and absence of luminance noise.28 A letter was selected at random and presented either in the center of a uniform 50-cd/m2 luminance field or in the center of a noise field of the same mean luminance. The contrast (C) of the letter was defined as Weber contrast:  
\(\def\upalpha{\unicode[Times]{x3B1}}\)\(\def\upbeta{\unicode[Times]{x3B2}}\)\(\def\upgamma{\unicode[Times]{x3B3}}\)\(\def\updelta{\unicode[Times]{x3B4}}\)\(\def\upvarepsilon{\unicode[Times]{x3B5}}\)\(\def\upzeta{\unicode[Times]{x3B6}}\)\(\def\upeta{\unicode[Times]{x3B7}}\)\(\def\uptheta{\unicode[Times]{x3B8}}\)\(\def\upiota{\unicode[Times]{x3B9}}\)\(\def\upkappa{\unicode[Times]{x3BA}}\)\(\def\uplambda{\unicode[Times]{x3BB}}\)\(\def\upmu{\unicode[Times]{x3BC}}\)\(\def\upnu{\unicode[Times]{x3BD}}\)\(\def\upxi{\unicode[Times]{x3BE}}\)\(\def\upomicron{\unicode[Times]{x3BF}}\)\(\def\uppi{\unicode[Times]{x3C0}}\)\(\def\uprho{\unicode[Times]{x3C1}}\)\(\def\upsigma{\unicode[Times]{x3C3}}\)\(\def\uptau{\unicode[Times]{x3C4}}\)\(\def\upupsilon{\unicode[Times]{x3C5}}\)\(\def\upphi{\unicode[Times]{x3C6}}\)\(\def\upchi{\unicode[Times]{x3C7}}\)\(\def\uppsy{\unicode[Times]{x3C8}}\)\(\def\upomega{\unicode[Times]{x3C9}}\)\(\def\bialpha{\boldsymbol{\alpha}}\)\(\def\bibeta{\boldsymbol{\beta}}\)\(\def\bigamma{\boldsymbol{\gamma}}\)\(\def\bidelta{\boldsymbol{\delta}}\)\(\def\bivarepsilon{\boldsymbol{\varepsilon}}\)\(\def\bizeta{\boldsymbol{\zeta}}\)\(\def\bieta{\boldsymbol{\eta}}\)\(\def\bitheta{\boldsymbol{\theta}}\)\(\def\biiota{\boldsymbol{\iota}}\)\(\def\bikappa{\boldsymbol{\kappa}}\)\(\def\bilambda{\boldsymbol{\lambda}}\)\(\def\bimu{\boldsymbol{\mu}}\)\(\def\binu{\boldsymbol{\nu}}\)\(\def\bixi{\boldsymbol{\xi}}\)\(\def\biomicron{\boldsymbol{\micron}}\)\(\def\bipi{\boldsymbol{\pi}}\)\(\def\birho{\boldsymbol{\rho}}\)\(\def\bisigma{\boldsymbol{\sigma}}\)\(\def\bitau{\boldsymbol{\tau}}\)\(\def\biupsilon{\boldsymbol{\upsilon}}\)\(\def\biphi{\boldsymbol{\phi}}\)\(\def\bichi{\boldsymbol{\chi}}\)\(\def\bipsy{\boldsymbol{\psy}}\)\(\def\biomega{\boldsymbol{\omega}}\)\(\def\bupalpha{\unicode[Times]{x1D6C2}}\)\(\def\bupbeta{\unicode[Times]{x1D6C3}}\)\(\def\bupgamma{\unicode[Times]{x1D6C4}}\)\(\def\bupdelta{\unicode[Times]{x1D6C5}}\)\(\def\bupepsilon{\unicode[Times]{x1D6C6}}\)\(\def\bupvarepsilon{\unicode[Times]{x1D6DC}}\)\(\def\bupzeta{\unicode[Times]{x1D6C7}}\)\(\def\bupeta{\unicode[Times]{x1D6C8}}\)\(\def\buptheta{\unicode[Times]{x1D6C9}}\)\(\def\bupiota{\unicode[Times]{x1D6CA}}\)\(\def\bupkappa{\unicode[Times]{x1D6CB}}\)\(\def\buplambda{\unicode[Times]{x1D6CC}}\)\(\def\bupmu{\unicode[Times]{x1D6CD}}\)\(\def\bupnu{\unicode[Times]{x1D6CE}}\)\(\def\bupxi{\unicode[Times]{x1D6CF}}\)\(\def\bupomicron{\unicode[Times]{x1D6D0}}\)\(\def\buppi{\unicode[Times]{x1D6D1}}\)\(\def\buprho{\unicode[Times]{x1D6D2}}\)\(\def\bupsigma{\unicode[Times]{x1D6D4}}\)\(\def\buptau{\unicode[Times]{x1D6D5}}\)\(\def\bupupsilon{\unicode[Times]{x1D6D6}}\)\(\def\bupphi{\unicode[Times]{x1D6D7}}\)\(\def\bupchi{\unicode[Times]{x1D6D8}}\)\(\def\buppsy{\unicode[Times]{x1D6D9}}\)\(\def\bupomega{\unicode[Times]{x1D6DA}}\)\(\def\bupvartheta{\unicode[Times]{x1D6DD}}\)\(\def\bGamma{\bf{\Gamma}}\)\(\def\bDelta{\bf{\Delta}}\)\(\def\bTheta{\bf{\Theta}}\)\(\def\bLambda{\bf{\Lambda}}\)\(\def\bXi{\bf{\Xi}}\)\(\def\bPi{\bf{\Pi}}\)\(\def\bSigma{\bf{\Sigma}}\)\(\def\bUpsilon{\bf{\Upsilon}}\)\(\def\bPhi{\bf{\Phi}}\)\(\def\bPsi{\bf{\Psi}}\)\(\def\bOmega{\bf{\Omega}}\)\(\def\iGamma{\unicode[Times]{x1D6E4}}\)\(\def\iDelta{\unicode[Times]{x1D6E5}}\)\(\def\iTheta{\unicode[Times]{x1D6E9}}\)\(\def\iLambda{\unicode[Times]{x1D6EC}}\)\(\def\iXi{\unicode[Times]{x1D6EF}}\)\(\def\iPi{\unicode[Times]{x1D6F1}}\)\(\def\iSigma{\unicode[Times]{x1D6F4}}\)\(\def\iUpsilon{\unicode[Times]{x1D6F6}}\)\(\def\iPhi{\unicode[Times]{x1D6F7}}\)\(\def\iPsi{\unicode[Times]{x1D6F9}}\)\(\def\iOmega{\unicode[Times]{x1D6FA}}\)\(\def\biGamma{\unicode[Times]{x1D71E}}\)\(\def\biDelta{\unicode[Times]{x1D71F}}\)\(\def\biTheta{\unicode[Times]{x1D723}}\)\(\def\biLambda{\unicode[Times]{x1D726}}\)\(\def\biXi{\unicode[Times]{x1D729}}\)\(\def\biPi{\unicode[Times]{x1D72B}}\)\(\def\biSigma{\unicode[Times]{x1D72E}}\)\(\def\biUpsilon{\unicode[Times]{x1D730}}\)\(\def\biPhi{\unicode[Times]{x1D731}}\)\(\def\biPsi{\unicode[Times]{x1D733}}\)\(\def\biOmega{\unicode[Times]{x1D734}}\)\begin{equation}\tag{1}C = ({L_L} - {L_M})/{L_M},\!\end{equation}
where LL is the luminance of the letter and LM is the mean luminance of the adapting (or noise) field (the letters were positive contrast, with letter luminance greater than adapting field luminance).  
The static noise field, which covered an area that was approximately twice as large as the letter, consisted of independently generated square checks with luminance values drawn randomly from a uniform distribution. Each noise check subtended 0.044 deg by 0.044 deg, which corresponds to three noise checks per letter stroke width, a value consistent with that used by others.18 The noise spectral density (N) was computed as the product of squared root mean square contrast and check area.29 N ranged from 0 to 4.8 × 10−5 deg2 in five steps, each separated by 0.8 log units. 
As illustrated in Figure 1, the target duration was 200 ms and the total noise duration was 400 ms. The target onset was delayed relative to the noise onset by 100 ms, and 100 ms of noise also followed the target offset. This type of asynchronous presentation is commonly used in visual noise-based studies.17,18,22,23,29,30 
Figure 1
 
Illustration of the sequence of stimulus presentation. A letter, presented for 200 ms, was added to white static luminance noise that was presented for 400 ms. Noise for 100 ms preceded and followed the target presentation.
Figure 1
 
Illustration of the sequence of stimulus presentation. A letter, presented for 200 ms, was added to white static luminance noise that was presented for 400 ms. Noise for 100 ms preceded and followed the target presentation.
Procedure
Prior to all measurements, the pupil of the tested eye was dilated with 2.5% phenylephrine hydrochloride and the display was viewed through the 3-mm artificial pupil. A 30-second period of adaptation to a uniform 50-cd/m2 field preceded each session, and a brief warning tone signaled the start of each stimulus presentation. The subject's task was to identify the letter presented, which was entered by the experimenter. No feedback was given. Only letters from the Sloan set were accepted as valid responses. In the event that the subject selected a letter that is not included in the Sloan set, the subject was asked to select a different letter. Contrast threshold for letter identification was measured using a 10-alternative forced-choice staircase procedure. An initial estimate of threshold was obtained by presenting a letter at a suprathreshold contrast level and then decreasing the contrast by 0.3 log units until an incorrect response was recorded. Following this initial search, log contrast threshold was determined using a two-down, one-up decision rule, which provides an estimate of the 76% correct point on a psychometric function.31 Each staircase continued until 12 reversals had occurred, and the geomean of the last four reversals was taken as contrast threshold. Excluding the initial search, the staircase length was typically 35 to 40 trials, which produced stable measurements. For each subject, contrast threshold was measured in the absence of noise and in the presence of each of the four external noise levels (N), with the order of noise level randomized. 
Analysis
Data were analyzed using the LAM, a standard model of visual performance in noise, as follows. First, contrast threshold measurements were converted to log threshold signal energy (Et), which was computed as the integral of the squared signal function.29 Next, log Et was plotted as a function of log N and the data were fit with the following equation20:  
\begin{equation}\tag{2}\log {E_t} = \log (k) + \log (N + {N_{eq}}),\!\end{equation}
where k and Neq are free parameters that were adjusted to minimize the mean squared error between the data and the fit. The subject's equivalent intrinsic noise (Neq) is given directly by equation 2, and sampling efficiency is reciprocally related to k of equation 2.20,32 The base-10 log of Neq and efficiency are plotted in the figures.  
Clinical Measurements
Pelli-Robson chart CS was measured according to the manufacturer's guidelines. HbA1c was measured from a blood sample provided immediately before testing. MP sensitivity was measured with an Optos SLO/microperimeter (Optos, Inc., Marlborough, MA, USA) that is described in detail elsewhere.33 Briefly, MP sensitivity was measured from the same eye in which CS, Neq, and efficiency measurements were obtained. Subjects were instructed to fixate centrally on a cross, and a small spot of light (0.4 deg) was presented for 200 ms. The spot contrast was changed on successive trials to measure MP CS (1/contrast threshold). MP CS was measured at 28 locations throughout the central 12° of the visual field. The overall MP CS value was calculated as the mean of the 28 MP CS measurements. 
Results
Figure 2 presents log Et as a function of log N for the control subjects (the gray region represents the range of control values), the NDR (top), and the mild NPDR (bottom) subjects. The log CS equivalents of the log Et values are shown on the right y-axis. The curves represent the least-squares best fit of equation 2 to each subject's data (all data points were included in the fit). The values of Et measured in the absence of noise (Et0; the leftmost points in Fig. 2) varied among the NDR subjects (∼10-fold difference among the subjects); all but two NDR subjects had Et0 within the control range. In comparison, Et measured in the highest level of noise (the rightmost points in Fig. 2) was highly similar among the NDR subjects and all were within the control range. Thus, the functions for the diabetic subjects tended to converge at the highest noise level tested. 
Figure 2
 
Log Et as a function of log N for the patients with NDR (top) or mild NPDR (bottom) compared to the control range (gray region).
Figure 2
 
Log Et as a function of log N for the patients with NDR (top) or mild NPDR (bottom) compared to the control range (gray region).
The lower panel shows marked variation among the mild NPDR subjects for Et0 (∼40-fold difference among the subjects) and that most patients had abnormally elevated Et0. In fact, only two mild NPDR subjects had an Et0 value that fell in the range of normal. Et measured in the highest level of noise showed less variation among the subjects (∼7-fold difference among the subjects), with six subjects having Et values falling outside of the control range. However, the elevation of Et values measured in the highest noise level was generally small for these six subjects. The log Et versus log N functions tended to be shifted upward and rightward by approximately equal amounts for the patients, relative to the controls. The direction of the shift is somewhat ambiguous for the subjects who had the largest threshold elevations. For these subjects, threshold was nearly independent of the luminance noise power, suggesting that higher levels of luminance noise are necessary to make the knee-points of the curves apparent. Nevertheless, elevated internal noise can be inferred for these subjects, as an efficiency loss would have shifted the curves vertically, resulting in a marked threshold elevation in high luminance noise, which was not observed. 
In Figure 3, the values of Neq that were derived from the LAM are plotted as a function of Et0 for the control subjects (black diamonds), NDR subjects (green squares), and mild NPDR subjects (red circles). Equivalent log CS in the absence of noise (CS0) is plotted along the top x-axis; the vertical and horizontal gray regions demarcate the normal range of Et0 and Neq, respectively. Only one NDR subject had a log Et0 value that was outside of the control range, and the elevation for this subject was relatively small (a factor of 1.6 higher than the upper limit of normal). Three NDR subjects had elevated Neq values (ranging from a factor of 1.1–3.0 above the upper limit of normal). Thus, two NDR subjects had small Neq elevations, despite having normal Et0. In comparison, seven mild NPDR subjects had elevated log Et0 that were elevated by as much as 1.2 log units (more than a factor of 15) above the upper limit of normal. Likewise, these seven mild NPDR subjects had elevated values of Neq. There was a statistically significant correlation between log Neq and log Et0 for the diabetic subjects (r = 0.93, P < 0.001; n = 20) and for the control subjects (r = 0.87, P = 0.001; n = 10). 
Figure 3
 
Log Neq versus log Et0 for the control subjects (black diamonds), NDR patients (green squares), and mild NPDR patients (red circles). The gray regions demarcate the range of normal log Neq (horizontal region) and normal Et0 (vertical region).
Figure 3
 
Log Neq versus log Et0 for the control subjects (black diamonds), NDR patients (green squares), and mild NPDR patients (red circles). The gray regions demarcate the range of normal log Neq (horizontal region) and normal Et0 (vertical region).
One-way ANOVA was performed to statistically compare the values of log Neq and log Et0 for the control and diabetic groups. The ANOVA indicated a significant effect of subject group for both log Et0 (F = 13.82, P < 0.001) and for log Neq (F = 11.92, P < 0.001). Log Et0 was significantly greater for the mild NPDR group than for the control and NDR groups (both t > 4.47, P < 0.001), but there was no significant difference between the NDR and control groups (t = 0.15, P = 0.88). Likewise, log Neq was significantly greater for the mild NPDR group than for the control and NDR groups (both t > 3.89, P ≤ 0.001), but there was no significant difference between the NDR and control groups (t = 0.61, P = 0.55). Thus, mild NPDR subjects typically had elevated contrast threshold values (poor CS) that were correlated with their internal noise levels. 
In Figure 4, log efficiency is plotted as a function of log Et0 for the control and DM subjects. The right y-axis shows the linear efficiency equivalents of the log efficiency values, and the top x-axis shows the log CS0 equivalent of the log Et0 values. The vertical and horizontal gray regions demarcate the normal range of Et0 and sampling efficiency, respectively. The efficiency values for the NDR subjects were generally within the range of normal. Three of the mild NPDR subjects had slight efficiency losses, less than 0.25% on average. ANOVA indicated that log efficiency for the DM groups was not significantly different from that of the controls (F = 1.29, P = 0.29). Additionally, there was no significant correlation between log efficiency and log Et0 for the patients (r = −0.22, P = 0.35; n = 20) or for the controls (r = −0.54, P = 0.11; n = 10). 
Figure 4
 
Log sampling efficiency versus log Et0 for the patients and control subjects. Efficiency values are shown along the right y-axis on a linear scale. The gray regions demarcate the range of normal log efficiency (horizontal region) and normal Et0 (vertical region). Other conventions are as in Figure 3.
Figure 4
 
Log sampling efficiency versus log Et0 for the patients and control subjects. Efficiency values are shown along the right y-axis on a linear scale. The gray regions demarcate the range of normal log efficiency (horizontal region) and normal Et0 (vertical region). Other conventions are as in Figure 3.
Log Et0, Neq, and efficiency were compared to standard clinical measures including age, HbA1c percentage, disease duration, visual acuity, Pelli-Robson chart CS, and MP sensitivity, with the results presented in Table 2. The table lists the Pearson correlation values for the two patient groups combined, with statistically significant correlations designated in bold with asterisks. Log Et0 was correlated significantly with log Neq, as discussed above, as well as with disease duration, MP sensitivity, and Pelli-Robson chart CS. The pattern of findings for Neq was similar to that of Et0 (significant correlations with disease duration, MP sensitivity, and Pelli-Robson chart CS), as expected given the high correlation between Neq and Et0. Interestingly, efficiency was correlated with Pelli-Robson chart CS and log MAR visual acuity (VA). The correlation between efficiency and log MAR VA, in particular, is somewhat surprising, because this is not a CS-based measure and there was relatively little variation in visual acuity among the subjects (range of approximately 20/16–20/25). Multiple regression models were developed as an additional approach to determine the relationship between the experimental measures (Neq, Et0, and efficiency) and the clinical measures (HbA1c, age, and disease duration). Consistent with the results presented in Table 2, the models indicated that only disease duration was a statistically significant, independent predictor of Neq (coefficient = 0.03; t = 2.81, P = 0.01) and Et0 (coefficient = 0.04; t = 3.07, P = 0.01). 
Table 2
 
Correlation Matrix Showing Associations With Clinical Parameters
Table 2
 
Correlation Matrix Showing Associations With Clinical Parameters
MP CS was compared among the three subject groups using 1-way ANOVA. The ANOVA indicated significant effects of subject group (F = 4.08, P = 0.03). Tukey pairwise comparisons indicated a significant reduction in mean MP CS for the mild NPDR group compared to the control group (t = 2.31, P = 0.03). However, the mean MP CS values for the control and NDR groups did not differ significantly (t = 1.15, P = 0.26). 
Discussion
The goal of the present study was to test the hypothesis that reduced CS is associated with elevated levels of internal noise in diabetic subjects who have NDR or mild NPDR. Neq was elevated in most individuals with mild NPDR (70% of the patients had Neq that was outside of the control range), but even diabetics who had no clinically-apparent retinopathy could have elevated Neq (30% of the sample). Contrast thresholds measured in the absence of noise were also elevated in most of these mild NPDR subjects, and the threshold elevations were highly correlated with Neq, but not efficiency. The finding of elevated Neq with generally normal sampling efficiency suggests that high levels of noise within the visual system may, at least in part, limit CS in diabetic patients. 
In addition to the strong correlation with CS measured in the absence of noise, Neq was also significantly correlated with disease duration, Pelli-Robson chart CS, and MP sensitivity. The significant correlation between Neq and disease duration can likely be explained on the basis that subjects with longer disease duration tended to fall into the mild NPDR group, as opposed to the NDR group. All three measures of CS performed in the present study (Et0, Pelli-Robson chart CS, and MP sensitivity) were significantly correlated. Although this may be expected, it should be noted that these three independent CS measures differ in many respects, including stimulus size, type, duration, retinal location tested, adaptation level, and response task (detection versus letter identification). 
At present, the source of increased Neq in this sample of diabetic subjects is uncertain, but there are a number of potential explanations. For example, elevated Neq in early-stage DR could be due to inner-retina cell dysfunction or death that has been demonstrated in patients who have early-stage DR3436 and in animal models of DR.37,38 In addition to inner-retina abnormalities, previous reports in human patients3941 and in animal models of DR42,43 have also provided evidence for outer-retina abnormalities (i.e., photoreceptor dysfunction), which may also lead to elevated internal noise. The relative contributions of inner- and outer-retina sources of noise to Neq elevations in diabetics require further investigation. 
Elevations in Neq are typically thought to reflect increased levels of neural noise within the visual system, but subjects who have optical defects due to cataracts19,44 and senescent optical changes21,45 can also have elevated values of Neq. In our sample of diabetic subjects, it is unlikely that optical defects underlie elevated Neq, because subjects who had more than minimal cataracts were not recruited, all subjects were optically corrected to minimize low-order aberrations, and the stimuli were viewed through a 3-mm artificial pupil that also minimized ocular aberrations. Consequently, it is more likely that elevated Neq in our sample of diabetics is related to neural rather than optical sources. 
One limitation of the LAM is the assumption that contrast processing is linear and that noise is additive. If the assumption of linearity does not hold, then elevations in Neq could be due entirely, or in part, to nonlinear factors.46 Potential nonlinear factors include changes in sampling, gain, or inhibition within a gain control process.46 Future work is needed to evaluate potential explanations (e.g., gain changes or internal noise elevations) for CS loss in diabetics. For our purposes, however, the distinction between elevated internal noise and a gain change has minimal practical value. That is, attributing the source of CS loss to a gain abnormality, rather than an internal noise elevation, does little to further our understanding of the retinal sites and mechanisms underlying CS loss in diabetics. Additionally, we caution against an overly literal interpretation of elevated Neq (e.g., there is no evidence that diabetics experience constant “visual snow” akin to white noise). Rather, we propose that the findings of elevated Neq and generally normal efficiency provide a useful descriptor of the pattern of CS abnormality observed in external noise, allowing for comparisons to other patient populations. Additional work is needed to fully understand the sites and mechanisms underlying CS loss in early-stage DR. 
From a practical viewpoint, our results emphasize that diabetics can have early neural abnormalities, even in patients who have good visual acuity. Thus, visual noise-based CS measurements can provide important additional information about visual dysfunction in diabetics that cannot be obtained from visual acuity measurements alone. Longitudinal studies are needed to determine whether noise-based CS measurements are useful for predicting disease progression. For example, following the NDR subjects over time is of particular interest to determine if disease progression is more likely in the subset of patients who had elevated Neq, compared to the NDR subjects who had normal Neq
Acknowledgments
Supported by National Institutes of Health research Grants R01EY026004 (JJM), P30EY001792 (core grant), UL1TR002003 (Center for Clinical and Translational Science), an unrestricted departmental grant, and a Dolly Green Scholar award (JJM) from Research to Prevent Blindness. 
Disclosure: J.J. McAnany, None; J.C. Park, None 
References
Klein BE. Overview of epidemiologic studies of diabetic retinopathy. Ophthalmic Epidemiol. 2007; 14: 179–183.
Davis MD, Fisher MR, Gangnon RE, et al. Risk factors for high-risk proliferative diabetic retinopathy and severe visual loss: Early Treatment Diabetic Retinopathy Study Report #18. Invest Ophthalmol Vis Sci. 1998; 39: 233–252.
Wilkinson CP, Ferris FLIII, Klein RE, et al. Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales. Ophthalmology. 2003; 110: 1677–1682.
Banford D, North RV, Dolben J, Butler G, Owens DR. Longitudinal study of visual functions in young insulin dependent diabetics. Ophthalmic Physiol Opt. 1994; 14: 339–346.
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 Ophthalmol. 1985; 69: 136–142.
Di Leo MA, Caputo S, Falsini B, et al. Nonselective loss of contrast sensitivity in visual system testing in early type I diabetes. Diabetes Care. 1992; 15: 620–625.
Ghafour IM, Foulds WS, Allan D, McClure E. Contrast sensitivity in diabetic subjects with and without retinopathy. Br J Ophthalmol. 1982; 66: 492–495.
Gualtieri M, Bandeira M, Hamer RD, Damico FM, Moura AL, Ventura DF. Contrast sensitivity mediated by inferred magno- and parvocellular pathways in type 2 diabetics with and without nonproliferative retinopathy. Invest Ophthalmol Vis Sci. 2011; 52: 1151–1155.
Howes SC, Caelli T, Mitchell P. Contrast sensitivity in diabetics with retinopathy and cataract. Aust J Ophthalmol. 1982; 10: 173–178.
Hyvarinen L, Laurinen P, Rovamo J. Contrast sensitivity in evaluation of visual impairment due to diabetes. Acta Ophthalmol (Copenh). 1983; 61: 94–101.
Leonova A, Pokorny J, Smith VC. Spatial frequency processing in inferred PC- and MC-pathways. Vision Res. 2003; 43: 2133–2139.
McAnany JJ, Alexander KR. Contrast sensitivity for letter optotypes vs. gratings under conditions biased toward parvocellular and magnocellular pathways. Vision Res. 2006; 46: 1574–1584.
McAnany JJ, Alexander KR. Spatial frequencies used in Landolt C orientation judgments: relation to inferred magnocellular and parvocellular pathways. Vision Res. 2008; 48: 2615–2624.
Pokorny J. Review: steady and pulsed pedestals, the how and why of post-receptoral pathway separation. J Vis. 2011; 11 (5): 7.
Pokorny J, Smith VC. Psychophysical signatures associated with magnocellular and parvocellular pathway contrast gain. J Opt Soc Am A Opt Image Sci Vis. 1997; 14: 2477–2486.
Montesano G, Gervasoni A, Ferri P, et al. Structure-function relationship in early diabetic retinopathy: a spatial correlation analysis with OCT and microperimetry. Eye (Lond). 2017; 31: 931–939.
McAnany JJ, Alexander KR, Genead MA, Fishman GA. Equivalent intrinsic noise, sampling efficiency, and contrast sensitivity in patients with retinitis pigmentosa. Invest Ophthalmol Vis Sci. 2013; 54: 3857–3862.
Pelli DG, Levi DM, Chung ST. Using visual noise to characterize amblyopic letter identification. J Vis. 2004; 4 (10): 6.
Kersten D HR, Plant GT. Assessing contrast sensitivity behind cloudy media. Clin Vis Sci. 1988; 2: 143–158.
Pelli DG, Farell B. Why use noise? J Opt Soc Am A Opt Image Sci Vis. 1999; 16: 647–653.
Bennett PJ, Sekuler AB, Ozin L. Effects of aging on calculation efficiency and equivalent noise. J Opt Soc Am A Opt Image Sci Vis. 1999; 16: 654–668.
McAnany JJ, Alexander KR. Contrast thresholds in additive luminance noise: effect of noise temporal characteristics. Vision Res. 2009; 49: 1389–1396.
McAnany JJ, Alexander KR. Spatial contrast sensitivity in dynamic and static additive luminance noise. Vision Res. 2010; 50: 1957–1965.
Pardhan S. Contrast sensitivity loss with aging: sampling efficiency and equivalent noise at different spatial frequencies. J Opt Soc Am A Opt Image Sci Vis. 2004; 21: 169–175.
Hall C, Wang S, Bhagat R, McAnany JJ. Effect of luminance noise on the object frequencies mediating letter identification. Front Psychol. 2014; 5: 663.
Hall CM, McAnany JJ. Luminance noise as a novel approach for measuring contrast sensitivity within the magnocellular and parvocellular pathways. J Vis. 2017; 17 (8): 5.
Nation Research Council Committee on Vision. Recommended standard procedures for the clinical measurement and specification of visual acuity. Report of Working Group 39. Adv Ophthalmol. 1980; 41: 103–148.
Hall C, Wang S, McAnany JJ. Individual letter contrast thresholds: effect of object frequency and noise. Optom Vis Sci. 2015; 92: 1125–1132.
Legge GE, Kersten D, Burgess AE. Contrast discrimination in noise. J Opt Soc Am A. 1987; 4: 391–404.
Manahilov V, Calvert J, Simpson WA. Temporal properties of the visual responses to luminance and contrast modulated noise. Vision Res. 2003; 43: 1855–1867.
Garcia-Perez MA. Forced-choice staircases with fixed step sizes: asymptotic and small-sample properties. Vision Res. 1998; 38: 1861–1881.
Pelli DG, Burns CW, Farell B, Moore-Page DC. Feature detection and letter identification. Vision Res. 2006; 46: 4646–4674.
Anastasakis A, McAnany JJ, Fishman GA, Seiple WH. Clinical value, normative retinal sensitivity values, and intrasession repeatability using a combined spectral domain optical coherence tomography/scanning laser ophthalmoscope microperimeter. Eye (Lond). 2011; 25: 245–251.
van Dijk HW, Kok PH, Garvin M, et al. Selective loss of inner retinal layer thickness in type 1 diabetic patients with minimal diabetic retinopathy. Invest Ophthalmol Vis Sci. 2009; 50: 3404–3409.
van Dijk HW, Verbraak FD, Kok PH, et al. Decreased retinal ganglion cell layer thickness in patients with type 1 diabetes. Invest Ophthalmol Vis Sci. 2010; 51: 3660–3665.
van Dijk HW, Verbraak FD, Kok PH, et al. Early neurodegeneration in the retina of type 2 diabetic patients. Invest Ophthalmol Vis Sci. 2012; 53: 2715–2719.
Gastinger MJ, Kunselman AR, Conboy EE, Bronson SK, Barber AJ. Dendrite remodeling and other abnormalities in the retinal ganglion cells of Ins2 Akita diabetic mice. Invest Ophthalmol Vis Sci. 2008; 49: 2635–2642.
Gastinger MJ, Singh RS, Barber AJ. Loss of cholinergic and dopaminergic amacrine cells in streptozotocin-diabetic rat and Ins2Akita-diabetic mouse retinas. Invest Ophthalmol Vis Sci. 2006; 47: 3143–3150.
Elsner AE, Burns SA, Lobes LAJr, Doft BH. Cone photopigment bleaching abnormalities in diabetes. Invest Ophthalmol Vis Sci. 1987; 28: 718–724.
Forooghian F, Stetson PF, Meyer SA, et al. Relationship between photoreceptor outer segment length and visual acuity in diabetic macular edema. Retina. 2010; 30: 63–70.
Holopigian K, Greenstein VC, Seiple W, Hood DC, Carr RE. Evidence for photoreceptor changes in patients with diabetic retinopathy. Invest Ophthalmol Vis Sci 1997; 38: 2355–2365.
Bogdanov P, Corraliza L, Villena JA, et al. The db/db mouse: a useful model for the study of diabetic retinal neurodegeneration. PLoS One. 2014; 9: e97302.
Park SH, Park JW, Park SJ, et al. Apoptotic death of photoreceptors in the streptozotocin-induced diabetic rat retina. Diabetologia. 2003; 46: 1260–1268.
Pardhan S, Gilchrist J, Beh GK. Contrast detection in noise: a new method for assessing the visual function in cataract. Optom Vis Sci. 1993; 70: 914–922.
Pardhan S, Gilchrist J, Elliott DB, Beh GK. A comparison of sampling efficiency and internal noise level in young and old subjects. Vision Res. 1996; 36: 1641–1648.
Baldwin AS, Baker DH, Hess RF. What do contrast threshold equivalent noise studies actually measure? Noise vs. nonlinearity in different masking paradigms. PLoS One. 2016; 11: e0150942.
Figure 1
 
Illustration of the sequence of stimulus presentation. A letter, presented for 200 ms, was added to white static luminance noise that was presented for 400 ms. Noise for 100 ms preceded and followed the target presentation.
Figure 1
 
Illustration of the sequence of stimulus presentation. A letter, presented for 200 ms, was added to white static luminance noise that was presented for 400 ms. Noise for 100 ms preceded and followed the target presentation.
Figure 2
 
Log Et as a function of log N for the patients with NDR (top) or mild NPDR (bottom) compared to the control range (gray region).
Figure 2
 
Log Et as a function of log N for the patients with NDR (top) or mild NPDR (bottom) compared to the control range (gray region).
Figure 3
 
Log Neq versus log Et0 for the control subjects (black diamonds), NDR patients (green squares), and mild NPDR patients (red circles). The gray regions demarcate the range of normal log Neq (horizontal region) and normal Et0 (vertical region).
Figure 3
 
Log Neq versus log Et0 for the control subjects (black diamonds), NDR patients (green squares), and mild NPDR patients (red circles). The gray regions demarcate the range of normal log Neq (horizontal region) and normal Et0 (vertical region).
Figure 4
 
Log sampling efficiency versus log Et0 for the patients and control subjects. Efficiency values are shown along the right y-axis on a linear scale. The gray regions demarcate the range of normal log efficiency (horizontal region) and normal Et0 (vertical region). Other conventions are as in Figure 3.
Figure 4
 
Log sampling efficiency versus log Et0 for the patients and control subjects. Efficiency values are shown along the right y-axis on a linear scale. The gray regions demarcate the range of normal log efficiency (horizontal region) and normal Et0 (vertical region). Other conventions are as in Figure 3.
Table 1
 
Subject Characteristics
Table 1
 
Subject Characteristics
Table 2
 
Correlation Matrix Showing Associations With Clinical Parameters
Table 2
 
Correlation Matrix Showing Associations With Clinical Parameters
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×