|Home | About | Journals | Submit | Contact Us | Français|
Dopaminergic degeneration affects both nigrostriatal projection neurons and retinal amacrine cells in Parkinson disease (PD). Parkinsonian retinopathy is associated with impaired color discrimination and contrast sensitivity. Some prior studies described associations between color discrimination deficits and cognitive deficits in PD, suggesting that contrast discrimination deficits are due, at least in part, to cognitive deficits in PD. We investigated the relationship between cognitive deficits and impaired contrast sensitivity in PD.
PD subjects, n=43; 15F/28M; mean age 66.5±8.2, Hoehn and Yahr stage 2.6±0.6, and duration of disease of 6.2±5.0 years underwent neuropsychological and Rabin contrast sensitivity testing.
Mean Rabin contrast sensitivity score was 1.34±0.40. Bivariate analyses showed significant correlation between Rabin contrast sensitivity scores and global cognitive z-scores (R=0.54, P=0.0002). Cognitive domain Z-score post hoc analysis demonstrated most robust correlation between Rabin scores and executive functions (R=0.49, P=0.0009), followed by verbal learning (R=0.44, P=0.0028), visuospatial (R=0.39, P=0.001) and attention z-scores (R=0.32, P=0.036).
Impaired contrast sensitivity in PD is robustly associated with cognitive deficits, particularly executive function deficits. These results suggest that contrast sensitivity may be a useful biomarker for cognitive changes in PD and may have implications for driving safety evaluations in PD.
Parkinson's disease (PD) is a neurodegenerative condition with characteristic motor and non-motor features, including cognitive and visual changes. The latter include impairments in contrast sensitivity and color discrimination . Retinal dopaminergic depletion is well established in PD (parkinsonian retinopathy) [2–6]. Retinal dopaminergic amacrine cells participate in light adaptation, spatial contrast sensitivity, color discrimination, and photoreceptor renewal [7, 8]. Recent studies using retinal optical coherence tomography showed significant retinal pathologies in PD, including thinning of the retinal nerve fiber layer . While retinal nerve fiber thinning largely reflects retinal ganglia cell pathology, Kaur et al. reported that retinal nerve fiber layer thinning correlated with impaired contrast sensitivity in PD . Another study of retinal nerve fiber layer thickness in PD, however, had a conflicting result .
There is evidence to suggest that cerebral deficits may contribute to visual function changes in PD. For example, Bertrand et al. suggested that color discrimination difficulties partially reflect cognitive difficulties, including decreased attention and visuospatial skills, and that color discrimination deficits were associated with the presence of brain white matter microstructural abnormalities . This may also relate to impaired neuronal function in the occipital cortex, which shows decreased glucose metabolism in subjects with PD compared to controls . The combination of mild cognitive impairment (MCI) and visuospatial dysfunctions predicts the development of dementia more so than isolated MCI in PD [14, 15] These observations also imply that some visual function changes share a common neural substrate with processes underlying cognitive impairment in PD, including parallel dopaminergic losses in the retina and the brain in PD .
Because of these conflicting observations, it is unclear whether contrast sensitivity is a purely retinal dysfunction or may also reflect central cognitive defects in PD. In the present study, we investigated the relationship between cognitive function and impaired contrast sensitivity in PD.
This cross-sectional study involved analysis of 43 subjects with PD (28 males, 15 females), mean age 66.5±8.2 (range 51–84), mean duration of disease of 6.2±5.0 years (range 0.5–20) and mean Hoehn and Yahr stage 2.6±0.6 (range 1–5). PD subjects met the UK Parkinson's Disease Society Brain Bank clinical diagnostic criteria . Subjects on cholinesterase inhibitor drugs were not eligible for the study.
Eighteen subjects were taking a combination of L-DOPA and dopamine agonists, 16 L-DOPA alone, 7 dopamine agonists alone and 2 subjects were not on dopaminergic drugs. Five subjects were taking entacapone, 6 rasagiline, 3 selegiline, 6 (low dose) amantadine but no subjects were taking typical anti-cholinergic drugs or memantine.
Each subject underwent a detailed cognitive examination following an approach previously reported to characterize cognitive impairment in PD . These tests included measures of verbal memory: California Verbal Learning Test ; executive/reasoning functions: WAIS III Picture Arrangement test , and Delis-Kaplan Executive Function System Sorting Test ; attention/psychomotor speed as absolute time on the Stroop 1 test ; and visuospatial function: Benton Judgment of Line Orientation test . Composite Z-scores were calculated for these different cognitive domains based on normative data. A global composite cognitive Z-score was calculated as the mean of the different domain Z-scores. Two subjects had mild dementia.
The Rabin contrast sensitivity test was performed using a small illuminator cabinet (Precision Vision®, La Salle, IL, USA). The test was administered with the best optical correction at a distance of 4 meters with the room lights turned off and the illumination box light on. With one eye occluded, each subject would start reading letters from left to right and continues row-by-row, down the chart. The count of the total letters missed was used to determine the log contrast sensitivity score (in Michelson contrast) using the manufacturer-supplied table. The procedure was repeated for the other eye and values were bilaterally averaged. A log contrast sensitivity value of less than 1.30 was below the normative threshold. The test was performed with the patient on their usual dopaminergic medications.
The study was approved by the Institutional Review Boards of the University of Michigan and Ann Arbor Department of Veterans Affairs medical centers. Written informed consent was obtained from all subjects.
Standard pooled-variance t or Satterthwaite's method of approximate t tests (tapprox) were used for group comparisons between subjects with abnormal or normal range contrast sensitivity. χ2 testing were performed for comparison of proportions between groups. Pearson correlation coefficients were computed to study the relationship between bilaterally averaged linear measures of contrast sensitivity and the cognitive measures. Shapiro-Wilk test confirmed normal distribution of the Rabin contrast sensitivity measure. Analysis of covariance was performed to evaluate the relationship between the linear Rabin contrast sensitivity measure and cognitive performance while accounting for confounder variables, including gender, age and duration of disease. Analyses were performed using SAS version 9.3, SAS institute, Cary, North Carolina.
Mean Rabin contrast sensitivity score was 1.34±0.40 (range 0.6–2.0), mean global cognitive Z-score was −0.44±0.91 (range −3.03 to 0.98). There were 23 subjects with abnormal and 20 subjects with normal range Rabin contrast sensitivity scores. Table 1 shows mean (± SD) values of demographic, clinical, cognitive variables in the patients with PD with abnormal and normal range contrast sensitivity. PD subjects with impaired contrast sensitivity were older and had more severe cognitive impairment compared to subjects with normal range contrast sensitivity. There was a gender trend effect with more men than women having impaired contrast sensitivity.
Bivariate analyses showed significant correlation between linear Rabin contrast sensitivity scores and global cognitive Z-scores (R=0.54, P=0.0002); Figure 1) but not with UPDRS total motor scores (R=−0.22, P=0.17),
Analysis of covariance was performed to further evaluate the main regression effect between contrast sensitivity scores and cognition while controlling for potential confounders (gender, age, and disease duration). The overall model using Rabin sensitivity scores as the dependent outcome parameter was significant (F(4,38)=4.21, P=0.0006) with significant independent effects for global cognitive composite Z-scores (F=12.3, P=0.0012) independent of effects for age (F=0.9, P=0.37), gender (F=0.02, P=0.88) and duration of disease (F=0.03, P=0.86). Similar results were obtained after excluding the two subjects with mild dementia. Similar results were also obtained after entry of LED as additional covariate in the model.
Cognitive domain Z-score post hoc analysis demonstrated most robust correlation between Rabin scores and executive functions (R=0.49, P=0.0009), followed by verbal learning (R=0.44, P=0.0028), visuospatial (R=0.39, P=0.001) and attention z-scores (R=0.32, P=0.036).
We found a significant association between impaired contrast sensitivity and cognitive deficits in PD. Impaired contrast sensitivity most robustly correlated with executive function, verbal learning and visuospatial deficits compared to attention function deficits. These findings extend previous study findings reporting correlation between impaired color discrimination and cognitive deficits in PD . Bertrand et al., however, found that impaired color discrimination most robustly correlated with decreased attention and visuospatial skills, whereas our impaired contrast sensitivity findings mainly reflected deficits in executive functions, verbal learning and visuospatial functions. These discrepancies may reflect differences in cognitive processing involved with tests of color discrimination versus tests of contrast sensitivity, or differences in the cognitive testing modalities themselves. Other mechanisms may also play a role as contrast sensitivity vision could represent a more binary visual determination and decision making (i.e., detection versus no detection) compared to more complex perceptual functions involved in hue difference detection with color discrimination tests. The robust effect of executive functions may relate to the necessity to make an executive assessment whether the faintly printed letters are or are not seen. The robust correlation with verbal learning is unexpected as the Rabin test does not intrinsically rely on stored verbal information. One prior study describes a significant correlation between cognitive impairment and contrast sensitivity in subjects with Alzheimer's and MCI, who tend to have more amnestic cognitive problems, compared with adults with cognitive complaints . Our findings suggest an overlap in neural circuitry involving verbal learning functions and contrast sensitivity detection. For example, the hippocampus provides important outflow to the posterior cingulate, which is at the center of higher order multimodal processing of cortical visual information . It is also conceivable that subjects may try to memorize letters from the test in an apparent attempt to improve performance based on recognition rather than pure reading. An alternative explanation may be that pathological changes in the retina have been linked to neurodegenerative conditions, including retinal deposition of β-amyloid or α-synuclein [25, 26]. Contrast sensitivity has been noted to decrease performance on neurocognitive testing in normal older adults . It is possible that the loss of retinal dopamine causes decreased contrast sensitivity originating in the retina, leading to or exacerbating decreased cognitive test performance. However, we would expect such cognitive effects to be limited to tasks involving vision. It is also possible that disruption of retino-cerebral network connectivity may underlie or contribute to cognitive changes associated with visual function deficits.
Rabin contrast sensitivity testing can be performed and scored in minutes and provides a unique, multi-domain assessment of visual impairment and executive cognitive functions. Both of these neurologic capacities have a strong influence on driving safety in PD and other age-related neurologic conditions [28, 29]. The current American Academy of Neurology (AAN) guidelines on the evaluation and management of driving risk in dementia  highlight the need for simple bed-side tests that can assess driving safety in populations of subjects with and without dementia. Our analyses suggest that impaired contrast sensitivity performance in PD may associate with executive dysfunction regardless of dementia status and even after controlling for the effects of age, gender and disease duration. Incorporation of the Rabin testing into natural history studies of both PD and normal aging may allow for the validation of useful and non-invasive clinical predictor of age-related driving performance.
This study has limitations, including absence of a systematic ophthalmological or retinal assessment . However, none of our subjects had color blindness, diplopia or known glaucoma and all subjects were able to read. The cross-sectional design of our study does not allow us to assess whether changes in contrast sensitivity may predict conversion to dementia. Furthermore, contrast sensitivity testing was performed while the patients were on dopaminergic medication. However, if a dopaminergic medication visual effect were present it would likely have underestimated the effect size of our findings in this study. It is conceivable that the presence of visual function deficits by itself may bias assessment of some of the cognitive testing that depend on reading or visuoperceptual functions. However, our findings of auditory verbal learning cannot be explained by possible visual function bias as no visual assessment of the learning is involvement. We do not have data available of contrast sensitivity and cognitive testing in non-PD control subjects. Therefore, the specificity of our findings cannot be assessed.
We conclude that cognitive impairment, in particular decreased executive and verbal learning functions, may be an important correlate of impaired contrast sensitivity in PD. Contrast sensitivity deficits and cognitive impairments may reflect shared retinal and cerebral processes. Given the strong correlation between contrast sensitivity deficits and cognitive domain deficits in this largely non-demented cohort, contrast sensitivity tests may be a useful biomarker for cognitive domain changes in PD and may have implication for driving safety evaluations in PD. Additional cross-sectional and longitudinal studies will be necessary to evaluate this possibility.
The authors thank Christine Minderovic, Cyrus Sarosh, Virginia Rogers, the PET technologists, cyclotron operators, and chemists, for their assistance. This work was supported by the Department of Veterans Affairs [grant number grant number I01 RX000317]; the Michael J. Fox Foundation; and the NIH [grant numbers P01 NS015655, P50 NS091856 and RO1 NS070856].
Dr. Muller has research support from the NIH, Michael J. Fox Foundation and the Department of Veteran Affairs.
Dr. Kotagal receives funding from the NIH (P30AG024824 KL2), VA Health Systems (IK2CX001186 and AAVA GRECC), and the Blue Cross & Blue Shield of Michigan Foundation.
Dr. Frey has research support from the NIH, GE Healthcare and AVID Radiopharmaceuticals (Eli Lilly subsidiary). Dr. Frey also serves as a consultant to AVID Radiopharmaceuticals, MIMVista, Inc, Bayer-Schering and GE healthcare. He also holds equity (common stock) in GE, Bristol-Myers, Merck and Novo-Nordisk.
Dr. Albin serves on the editorial boards of Neurology, Experimental Neurology, and Neurobiology of Disease. He receives grant support from the National Institutes of Health and the Michael J. Fox Foundation. Dr. Albin serves on the Data Safety and Monitoring Boards of the PRIDE-HD and LEGATO-HD trials (ICON/Teva), and a phase 1 trial in HD sponsored by IONIS.
Dr. Bohnen has research support from the NIH, Department of Veteran Affairs, and the Michael J. Fox Foundation.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosures The authors declare no conflict of interest relevant to this work.
Dr. Ridder has nothing to disclose.