Abstract
Abstract
INTRODUCTION
Influenced by John Locke’s (1632–1704) theory that the senses are the building blocks of knowledge,Francis Galton (1822–1911) introduced the concept [1] that differences in sensory discrimination should be related to individual differences in cognitive ability and that these are inherited as is intelligence, albeit these traits were acknowledged to deteriorate somewhat with age. It has been shown that color discrimination is poor among patients with Alzheimer’s dementia relative to age-matched cognitively healthy individuals and that color discrimination performance are related to standard scores of cognitive function [2, 3]. The deficit could be pinpointed in patients with Alzheimer’s disease as differentially affecting discrimination in the blue and green area [4].
It is recognized that decreased visual function [5] is one of the non-motor dysfunctions of Parkinson disease (PD) and impaired cognitive function and dementia are cardinal signs of disease progression [6]. Poor color discrimination among patients with PD has long been recognized [7, 8]. Interestingly, in patients with PD, deterioration in color discrimination on a standard color discrimination evaluation, the Farnsworth-Munsell 100 hue test (FMHT) although correlated with age, is additionally correlated with greater impairment of motor function [3] (based on the standard Unified Parkinson’s Disease Rating Scale motor section), activities of daily life (Unified Parkinson’s Disease Rating Scale activities of daily living section), and worse pathologic psychiatric ratings [9]. It has also been shown that motor symptom severity in PD is correlated with performance on FMHT, especially with regard to the axial symptoms [5].
Nonetheless, it has been suggested that performance on the FMHT may reflect cognitive impairment more than purely color discrimination deficits in PD since it is also associated with white-matter alterations in right posterior brain regions [10]. In addition, a recent study showed that the results of color discrimination deficits in PD based on FMHT were also correlated with measures of visuospatial abilities and executive functions [10]. Interestingly, color discrimination on the FMHT among patients with LRRK2-associated PD was worse than those with idiopathic PD but family members who did not have LRRK2-associated PD still had worse color discrimination than controls, regardless ofmutation status [11].
In the past decade it has been shown that carrying one or two mutations in the β-glucocerebrosidase gene (GBA) that cause Gaucher disease (GD), as based initially on clinical evidence [12], is a risk factor for PD [13, 14]. Gaucher disease is a rare autosomal single-gene disorder with a predilection among Ashkenazi Jews for its non-neuronopathic form [in conjunction usually with an N370S mutation] that is characterized by lipid storage in the cells of the monocyte-macrophage system such as the bone marrow, spleen, and liver [15]. Autopsy studies of brains of patients with both GD and PD [16] reveal α-synuclein-positive Lewy bodies and inclusions. Moreover, in large-scale screening of patients with PD, there was a greater than expected frequency of GBA mutations [17]. In a multi-center investigation [18], it was shown that the Odds Ratio for any GBA mutation in PD patients versus controls was 5.43 across centers: as compared with patients who did not carry a GBA mutation, these presented earlier with PD motor signs and were more likely to have affected relatives.
Cognitive impairment [19] seems to characterize the early-onset and aggressive PD associated with GBA mutations [12]. In general, carriers of GBA mutations have a wide spectrum of PD phenotypes, but have been shown to suffer more often and more severely from a variety of non-motor symptoms such as neuropsychiatric complaints and autonomicimpairment [20].
Independently, patients with GD, as distinct from carriers of a GBA mutation, have been recognized as having some decrement in visuospatial performance (this is not equivalent to poor vision or visual discrimination) despite overall high-normal cognitive functioning [21].
The purpose of the present study was to assess color discrimination based on the FMHT among patients with GD and as compared to patients with GD who do not have PD (GD only), obligate carriers of GD (one GBA mutation) with and without PD, patients with PD who do not have GBA mutations, and healthy age-matched controls.
METHODS
This study was approved by the Shaare ZedekMedical Center Institutional Review Board in accord with the Declaration of Helsinki. Written informed consent was obtained from all participants.
Six groups of 10 individuals each were enrolled: GD with PD (GD+PD); GD but no PD (GD only); GD obligate carriers (one GBA mutation) with PD; GD obligate carriers (one GBA mutation) without PD; PD with no GBA mutations (PD only); and healthy individuals (controls) who do not have GBA mutations and do not have PD.
All GD patients had been diagnosed with GD either by bone marrow diagnosis or low GBA enzyme activity plus had been analyzed for mutations by standard molecular diagnosis. All obligate carriers were assessed for GBA mutations most commonly seen among Caucasians and Ashkenazi Jews. All PD patients had been diagnosed by a movement disorder specialist or competent neurologist based on UK Brain Bank criteria.
The patients with GD+PD have been described previously with regard to severity of phenotype [22].: seven of these had cognitive impairment as part of the symptoms of PD but only three patients had signs of severe GD. The ten patients included in the study were all the living GD+PD patients in our clinic.
Age-matched participants [with and without PD and healthy] were found for the GD+PD cohort. Demographic characteristics were collected based on a simple questionnaire prior to administration of the FMHT. GD-specific information and PD-specific information were taken from patient files.
The FMHT was administered according to the manufacturer’s instructions under comparable conditions for all participants, including optimum lighting. The FMHT has been used for more than 40 years among the general population to separate persons with normal color vision into classes of superior, average, and low color discrimination ability and to measure the zones of color confusion among persons who are color defective. The test consists of four trays containing a total of 85 removable color reference caps (incremental hue variation) with anchor caps for each tray spanning the visible spectrum so that each tray is specific to a certain hue spectrum. No practice pre-test or re-testing was allowed. A total error score of 20–100 is the range of normal competence for discrimination.
Color vision abnormalities are based on errors in placing the caps in the correct sequence along the respective spectrum in each of the trays. Outcome on the FMHT are reported as the computer-generated TES (Total Error Score): the lower the absolute number, the better the performance. The TES is a measure of accuracy of an observer in arranging the caps so as to form a gradual transition in chroma between the two anchor caps; the higher the number of misplacements (total errors), the larger the TES. TES also reflects a qualitative feature by quantifying the distance of a cap placed erroneously from its correct placement. Means and 95th percentiles are available for the chronological age decades from the 30’s to the 70’s [23]. The TES is given as whole digits. The TES is combination of the results of the Major Radius and Minor Radius whose values were tabulated separately. Scoring of the results included the refinements of Vingrys & King-Smith [24] which has three values: [a] “angle” indicating the axis of confusion and therefore the type of color deficiency; [b] the “C-Index” for color Confusion-Index, which is a measure of the severity of the color deficit; and [c] the “S-Index” for the Scatter-Index, which is used to assess the degree of randomness or selectivity in the observer’s arrangement and is the ratio of the Major and Minor Radii. The Major Radius length may be used as an index of error [24].
When requested, one of us, an ophthalmologist(SS-T), performed additional eye examinations for patients who requested examinations for visual acuity and/or specific ocular complaints. These results were not included in the study except to verify that there was no problem with visual acuity.
Statistical analysis
Descriptive statistics were used for demographic data collected from the patient files including means, medians, etc. One-way analysis of variance (ANOVA) was employed to assess effect of age and gender which were considered confounding variables for TES; post hoc, the Levene’s test for equality of means was also applied. Other post hoc analyses were for multiple comparisons among and between groups and included the non-parametric Kruskal-Wallis test, Scheffe test, and the Pearson chi-square and Fisher’s exact tests for significance. Analysis of covariance (ANCOVA) was employed when combining groups (GD only versus GBA carriers; and PD only versus non-PD).
Results
Six groups of 10 persons each were tested. No subject had impaired vision (although the majority needed glasses).
The entire cohort of 10 patients (8 males) with GD+PD followed at the SZMC Gaucher Clinic who are mobile was included. GBA carriers were recruited from among obligate carriers whose mutations were previously identified; some of the PD patients with GBA mutations and all the patients with PD who did not have GBA mutations (PD only) were recruited from a referral clinic for Parkinson disease (NG).
The demographic characteristics of the six groups are presented in Table 1. The Fisher’s exact test was used to assess differences between groups because of gender representation, which showed a significant point probability (linear-by-linear association) of 0.059 because of the over-representation of males in the GD+PD group. There was a significant difference between groups (p < 0.001) because of age (older in the PD only and the GBA carriers with PD). Post hoc the Fisher’s Exact test showed significant difference for gender (p = 0.006).
Table 2 presents the means and standard deviations, and ranges of the TES and other values in the 6 groups. One-way ANOVA between groups showed significant differences between groups (p < 0.001). Mann-Whitney U-testing showed only a trend in differences in TES between the GD only group and the GD+PD group (n = 10 each; p = 0.075); and between the GD only and the GBA carriers only (n = 10 each; p = 0.105). However, in comparison of the GD only and the PD only, there was a significant difference (p = 0.007).
In terms of the relative positioning of the mean values, the highest mean TES was in the PD only group, the lowest in the GD only group; mean angle (primary axis of color confusion) was highest in the healthy group, lowest in PD only and GD+PD groups; mean S-Index was highest in the GD+PD, lowest in GBA carriers+PD; mean C-Index (which normalizes results to a perfect arrangement) was highest in PD only, lowest in GD only; mean Major Radius was highest in PD only, lowest in GD only; and mean Minor Radius was highest in the healthy group, lowest in the GD only. The GD+PD group means for these values was between the extremes of the GD only group and the PD only group.
Table 3 presents PD groups versus non-PD groups (n = 30 each). The Student’s t-test (two-tailed) for comparison of means of PD groups versus non-PD groups showed a significant effect for age (p = 0.036) and for TES (p = 0.017).
Table 4 presents groups sub-divided between the GD groups and the GBA carriers groups (n = 20 each), i.e., each with and without PD. Using two-way ANOVA analysis, there was a significant difference because of the effect of PD (n = 20 each; p = 0.018).
DISCUSSION
Consistent with previous studies, color discrimination was decreased in patients with PD [7, 8]. However, this study introduces a few new findings that relate to the impact of Gaucher disease in the context of PD-related color discrimination. These findings may imply a “protective effect” on color discrimination in those GD patients who also have PD which probably is due to a sensory mechanism rather than a cognitive effect. On the other hand, it is not to be gainsaid that patients with Gaucher disease (i.e., without PD) putatively have high-normal cognitive functioning [21], so that this too may be of consequence if the contention is that color discrimination is related to cognitive ability. If the hypothesis is correct that color discrimination in PD is a marker of disease severity [5], then, based on the findings in the current study, one might posit that GD (two mutations) is more protective of aggressive PD-related deterioration than one GBA mutation, but that is not the case with respect to cognitive decline in GD+PD. The seminal finding of lowest TES among patients with GD only (also relative to the healthy persons) was unexpected; on the other hand, GBA carriers’ scores for TES were poorer than the GD only (and comparable to those of healthy controls).
Patients with GD+PD performed significantly better than patients with PD only in TES and in the various Vingrys & King-Smith values. This was further highlighted by the fact that, on average, patients with GD (without and with PD) had lower TES than GBA carriers (without and with PD), the latter of whom (GBA carrier+PD) have been shown to have cognitive decline based on autopsy material of white matter deficits [25] as in GD+PD [16].
The patients with GD only had the best performance (lowest TES) on the FMHT (also better than healthy controls). We posit that the significantly better performance of the GD only group is because of a sensory mechanism rather than because of better cognitive function than that of the other groups. Bearing in mind that in addition to visual acuity (which was not impaired in any of the participants), color discrimination may be influenced by cognitive function which in patients with GD+PD may be among the first and most notable functions to be affected [19, 26], it is instructive that GD only patients performed better than the GBA carriers without PD, and that both the sub-groups of GD+PD and GBA carriers+PD performed better than the PD only group on virtually all outcome measures. Thus, these comparisons putatively control to some extent for the possible confounding effects of cognitive function both intra- and inter-group, by comparing persons with GD with each other and carriers of GBA mutations witheach other.
Although in this study, those with only PD were significantly older than those in other groups, and decrements in both visual acuity and/or discrimination in the blue-yellow range [27] may be attributed to an age effect alone, the comparable effect in patients with PD in two other groups who were slightly younger, also showed PD to be an independent predictor of poor color discrimination. Moreover, the FMHT is generally not considered to be greatly influenced by age after middle age [28, 29] until the last decades; in fact, mean TES are characterized by a U-pattern by age decades [23]. In this study, the participants were in their 60’s. However it is true that discrimination of blue-yellow deteriorates after the 30’s [23].
Similarly, the fact that the original GD+PD group had eight males [and indeed there was a significant effect for gender] might be construed as a limitation of this study. However, it has been shown that there is generally no effect of gender on the FMHT [28, 30].
Another putative limitation of this study is sample size. The choice of sample size of 10 patients per group was imposed on the study because there were only 10 patients with GD+PD in our Gaucher referral clinic (of ∼500 adult patients). It was for this reason that we chose pair-wise matching to increase statistical relevance if not necessarily statistical power. Nonetheless, larger comparator groups might increase the statistical significance even in pair-wise comparisons. Hence, it may prove to be the case that with greater numbers per group, there would be smaller standard deviations per group and the differences between GD only and GD+PD or between GBA carriers without and with PD, respectively would then be altered. However, the findings of performance of GBA carriers+PD as intermediate between that of GD+PD patients and that of PD only patients shows some consistentpatterning.
In an early study among obligate carriers of other lysosomal disorders including metachromatic leukodystrophy [31], carriers’ performance was poorer than otherwise healthy persons’ performance in neurocognitive testing particularly with regard to spatial function. Poorer performance among the carriers was apparently correlated with relatively lower levels of the disease-specific enzyme in that study. We too had noted in neurocognitive testing (unpublished) that pediatric GBA carriers performed in an intermediate manner between pediatric siblings with GD (whose performance was better) and pediatric siblings who had no GBA mutations (whose performance was relatively poorer). Thus, the finding of intermediate function, both in cognitive ability and in sensory discrimination, of GBA obligate carriers relative to those who have GD and those who do not have GD mutations, has been previously noted in GD and also in other lysosomal disorders.
Nearly a decade ago, various groups identified susceptibility for PD among patients with GD [32] and also among GBA obligate carriers of a single mutation [14] as well as among other first-degree relatives [33]. The evidence for increased risk of the early-onset PD as described in the seminal case report [12] is now accepted to be greater among those with “non-mild” (i.e., non-N370S) mutations [34] and among rare variants in non-Ashkenazi ethnicities as well [35]. Yet, the actual form of PD in patients and carriers of GBA mutations, while characterized variously in the literature, has implied that the features do not necessarily coincide with classic PD.
The current findings provide a new window into the opaque box of parkinsonism in patients with GD and carriers of GBA mutations because we now speculate that cognition and sensory discrimination may be differentially affected. Whereas in PD patients, performance on the FMHT is correlated with measures of visuospatial ability and executive function [10], patients with GD who have been shown to have imperfect visuospatial function but high cognitive function [21] performed relatively better than all others in this study on the FMHT, and patients with GD+PD performed better than the PD only patients. The (apparent) independent preservation of the sensory modality of color discrimination which may be a harbinger of non-motor PD severity versus deterioration of cognitive ability in GD+PD, may generate new hypotheses as to the epigenetic effects of β-glucocerebrosidase deficiency on the PD phenotype. This hypothetical construct is all the more cogent in light of recent findings that color vision impairment in patients with PD increased dementia risk whereas the sensory condition of olfactory dysfunction which is often considered as prodrome, did not [36]. Nonetheless, additional color vision testing using newer computer-based tests which quantify color vision in several dimensions may be warranted to more clearly describe the color discrimination in patients with GD and/or PD.
CONFLICT OF INTEREST AND SOURCES OF FUNDING
No author reports any conflict of interest regarding the performance and/or results of this study. No special funding was received for this study.