Abstract
Introduction:
Alzheimer’s disease (AD) is characterized by amnesia, though non-memory-cognitive domains like visual are also affected. We planned to study frequency of visual dysfunctions in AD and their relationship with dementia severity.
Materials and Methods:
This study was conducted in the Cognitive clinic of Department of Neurology, Bangur Institute of Neurosciences, Kolkata, between January 2007 and December 2010. 55 patients of AD were evaluated by neurological and neuropsychological assessments and by special tests for visual dysfunctions.
Results:
Common visual dysfunctions were visuo-constructional (87.3%), visuo-perceptual (63.6%), object agnosia(47.3%), prosopagnosia (45.5%), visual hallucination (27.3%) and simultanagnosia (12.7%). Symptoms of ventral visual pathway dysfunction were more common than that of dorsal pathway. MMSE score and number of visual manifestations had a good correlation. Conclusions: Visual dysfunctions are common in AD, elicitation of which helps us to understand the cause of disability so that appropriate steps can be taken.
Introduction
Alzheimer’s disease (AD) is the commonest type of degenerative dementia, where the most frequent and prominent dysfunction is of episodic memory. Nonmemory-cognitive domains including that of the visual system are also affected which may hamper activities of daily living (ADLs) and magnify the effects of other cognitive dysfunctions. 1
Visual manifestations (VMs) are particularly common in posterior cortical atrophy, a clinical entity first described in 1988. 2 The syndrome may have various underlying etiopathogenic diagnoses such as dementia with Lewy bodies (DLBs), corticobasal syndrome, and Creutzfeldt-Jakob disease (CJD), but majority are due to AD. The latter, often called “visual subtype,” differs from typical AD by predominance of visual symptoms while memory symptoms are not so prominent. Visual symptoms have been reported to be the initial manifestation in some patients with AD. 3 Early-onset AD (EOAD) may present with visuoperceptual deficits, in contrast to the slowly progressive memory deficits seen in late-onset AD (LOAD). 4
The VM in typical AD is expected to be present in view of prominent involvement of association cortex in parieto–temporo–occipital area, still it is less extensively been studied. A study on 30 patients with probable AD showed that about 43% had complex visual complaints with agnosia in 57% and complete Balint’s syndrome in 20% of the cases. 5 Prevalence of VM varies widely across the tests, ranging from 0% to 58%. 6 It had also been seen that patients with worse dementia disability had the most complex visual disturbances. 5
The VMs in AD had been described to be either basic or complex. 1 The former includes abnormalities in peripheral vision, visual-evoked potentials (VEPs), and contrast sensitivity. Complex VMs include difficulties in construction, recognition of objects and faces, Balint’s syndrome, and hallucinations.
The pathology behind VM in AD is the involvement of the visual association areas with the relative sparing of the primary visual areas. Although there may be optic nerve degeneration in AD, with dropout of retinal ganglion cells and their axons, this optic neuropathy results in minimal visual impairment, 7 and most of the observed visual system abnormalities in AD probably result from disease in visual association cortex. This is supposed to be due to involvement of ventral and dorsal visual pathways. 8
However, knowledge about prevalence, correlates, and prognostic implications of VMs in typical AD is lacking. They had been primarily studied from the viewpoint of neuro-ophthalmology, which is however inadequate to explain the complex phenomena of disturbed visuocognition. Besides, most of the dementia screening and neuropsychological test batteries do not assess VM in detail.
In this background, we planned to study visual problems in AD in detail, to know the frequency of visual symptoms and to assess whether visual symptoms represent a single- or multidimensional phenomenon. We also planned to assess the frequency of VM due to either dorsal or ventral visual pathway involvement. Moreover, we tried to assess relationship of dementia severity with various visual symptoms and to know which visual symptom has strongest association with dementia severity. This is, to best of our knowledge, the first study from India in this regard.
Methodology
This cross-sectional study was conducted in the Cognitive clinic of Bangur Institute of Neurosciences (BIN), Kolkata, India, between January 2007 and December 2010. Individuals with suspected cognitive dysfunction were referred to this specialty clinic from the general neurology and psychiatry outpatient departments of BIN, which is a tertiary-level training hospital for neurology and allied disciplines in the eastern part of India. Details of demographic profile, family history, and progression of symptoms were collected by a multidisciplinary team, after taking informed consent from either the patient or the guardians. A thorough neurological, psychiatric, and cognitive examination was performed on each patient. Laboratory investigations including neuroimaging (computed tomography [CT], magnetic resonance imaging [MRI], and single-photon emission CT scan of brain in few), hemogram, blood biochemistry (renal, liver and thyroid function tests, homocysteine level), and serology (human immunodeficiency virus HIV-enzyme-linked immunosorbent assay and Venereal Disease Research Laboratory test) were done.
For cognitive assessment, we routinely use the Kolkata Cognitive Screening Battery (KCSB), 9 which is a scale in the local language, Bengali, validated for use in a similar population. The KCSB includes a 30-item screening tool Bengali-mental state examination (BMSE), along with tests of attention, fluency, new learning ability using a 10-word-list, and visuoconstructional ability. In many patients, repeat visits were advised within a month of the first workup to fill up the missing data, because many patients are unable to complete the whole battery of tests in 1 sitting. As the stimulus figure for copying was “diamond-within-a-square” and test for attention and concentration was counting days of the week backward in BMSE, patients were separately asked to copy the “intersecting pentagons” and to perform 100-7 test to calculate Mini-Mental State Examination (MMSE) score.
Screening ophthalmological evaluation including acuity of vision, field of vision, extraocular movements, and fundoscopy was done in all patients who were evaluated for VM.
Patients were asked to read from a paragraph and to write spontaneously.
In the 10-object naming scale, which was developed in control population, mean score was 9.3 (standard deviation [SD] 1.8). We considered the lowest 10th percentile score as the cutoff point for identifying the significantly impaired 10th of the population. In this case, the cutoff was 8. For the purpose of scoring, 1 point was given to each object correctly named or described semantically. Incorrect naming or semantic description was scored 0.
Object identification is tested and interpreted per Table 1.
Object Identification.
Face recognition was assessed using a card with pictures of 20 famous personalities from our sociocultural background. It was developed in the control population, where mean score was 16.1 (SD 3.2). We chose the lowest 10th percentile score as the cutoff point for identifying the significantly impaired 10th of the population. This cutoff was 12. For the purpose of scoring, 1 point was given to each photograph correctly named or described semantically. Incorrect naming or semantic description was scored 0. Prosopagnosia is differentiated from anomia by asking patients to elucidate the sphere of accomplishment or point out when name cue was given.
Unilateral visual stimulation, bilateral simultaneous visual stimulation, line bisection test, and A-cancellation tests were done to elicit and differentiate field defect and hemineglect (Table 2).
Hemineglect Versus Hemifield Defect.
Simultanagnosia was picked up by “modified A-cancellation test” (mixed large and small letter) and complicated large figure identification. We adopted another innovative test to pick up simultanagnosia. In this test, we wrote on the blackboard few letters, which in part simulate another letter. Examples are “P-D,” “B-D,” and “L-I.” Patients with simultanagnosia read “P” as “D” and so on when the stimulus was large, but read correctly when stimulus was small. Optic ataxia and oculomotor apraxia were evaluated by standard tests. Visuoperceptual deficits are checked by fragmented letter identification and dot counting.
Consecutive patients, provisionally diagnosed as AD according to Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition, Text Revision) criteria, during the evaluation period, were included in this study. Those patients were categorized as possible AD, where dementia could be due to another condition, and probable AD, where dementia could not be due to any other cause. Those patients were excluded from the study in which investigation, particularly imaging, showed features suggestive of other disease like prominent vascular burden. Each file was reviewed, and the relevant data were collected in a predesigned proforma. Exclusion criteria included the presence of prominent frontal symptoms at presentation or examination, focal neurodeficits, significant extrapyramidal signs, and prominent vascular burden in brain MRI.
Severity of AD was graded according to the MMSE scores as follows: ≤10 = severe; >10 to 20 = moderate; and >20 as mild. 10
Relationship between dementia severity as measured by MMSE and scores of individual visual symptoms was explored using Pearson correlation coefficient. In all cases, 2-tailed tests were used, P value of less than .05 being considered statistically significant. Statistical analysis was done with SPSS 16.0 for Windows software.
Exhaustive chi-square automatic interaction detection was utilized to identify clinical features (especially visual problems) that can be used in establishing the predictive model for severity of dementia.
Results
A provisional diagnosis of AD was made in 70 (25.54%) of the total of 275 patients who were evaluated in our clinic during the study period. Of them, 14 patients were excluded because complete information were not available or because of diagnostic confusion. This includes 4 patients in whom primary visual problem was thought to be a confounding factor according to ophthalmological opinion. Another patient, with delusional disorder for the last 25 years, was also excluded from the analysis, as the underlying psychiatric illness may have interfered with the cognitive functions.
Tables 3 to 5 summarize clinical and demographic profile of 55 study patients.
Summary of Numerical Study Variables.
Abbreviations: MMSE, Mini-Mental State Examination; Std. deviation, standard deviation.
a Could not be assessed in 3 patients.
b Could not be assessed in 5 patients.
Summary of Categorical Study Variables.
Summary of the Frequency of Occurrence of Various Visual Symptoms.
Thus, we describe the data for 55 (35 males; 63.6%) patients with AD. In all, 31 patients (56.4%) had developed symptoms of dementia before 65 years (EOAD) and 24 (43.6%) at or after 65 years (LOAD).
A positive family history of dementia in first- or second-degree relatives was present in 17 (30.9%) individuals. All patients were right handed. The imaging in all 55 patients was corroborative with AD.
Most of our patients (n = 45; 81.8%) presented with forgetfulness of recent events with varying degrees of involvement in other cognitive domains, while only visual symptoms had been present in the initial stage in 1 (1.8%). Four (7.3%) patients had behavioral manifestations like apathy, depression, or anger outbursts at the start of illness; the other 5 (9.1%) had calculation difficulties, topographical disorientation, or deterioration in their planning and sequencing abilities. All except 3 (5.45%) developed significant forgetfulness at some point of time during the course of illness; yet, these 3 patients with only subtle recent memory deficits merited a diagnosis of AD, considering the pattern of topographical disorientation or language involvement and suggestive neuroimaging finding and lack of factors suggestive of other type of dementia on biochemical investigations and neuroimaging.
Visual acuity score were available in 35 of 55, and it was ≥6 of 9.
Visual Manifestations
Nine VMs were specifically evaluated in these patients, namely, object agnosia, prosopagnosia, simultanagnosia (±optic ataxia and oculomotor apraxia), visual hallucinations, visual field defect and neglect, alexia without agraphia, constructional, and perceptual deficits. Information about any VM was not available in 2 (3.6%) patients. At least 1 VM was present in 51 (92.73%), whereas at least 3 VMs were present in 60%. None among our sample had alexia without agraphia or the complete Balint’s syndrome. All the patients who were diagnosed to have simultanagnosia on the basis of modified “A-cancellation test” and complicated large figure identification test also failed in “partially simulating letter identification test.” The detailed frequencies of different VMs are shown in Table 5.
The 2 patients without any VM were having mild dementia; on the other hand, all patients with 4 or more VMs (n = 22; 41.6%) had either moderate (n = 7; 28% of all moderate dementias) or severe dementia (n = 15; 62.5% of all severe dementias).
Patients with mild dementia did not have object agnosia, simultanagnosia, field defect, or hemineglect. One-third of those with mild dementia had prosopagnosia, visual hallucinations, and impaired visuoconstructional function (VCF), while 16.7% had deficits in visuoperceptual function (VPF). Higher proportions of patients were having each VM in the moderate and severe group.
Mean MMSE score was 13.1 (±8.16). The numbers of patients with AD with mild, moderate, and severe dementia were 6 (10.9%), 25 (45.5%), and 24 (43.6%), respectively.
All the visual parameters except visual neglect are significantly correlated with MMSE score (Table 6).
Correlations Between BMSE Score and Visual Symptoms.
Abbreviations: BMSE, Bengali-mental state examination; MMSE, Mini-Mental State Examination; NS not significant; sig, significant.
a Correlation is significant at the .01 level (2-tailed).
b Correlation is significant at the .001 level (2-tailed).
It has been seen that chance of having severe dementia is 92.3% in the group having object agnosis, and this went to cent percentage if associated with initial forgetfulness (Figure 1).
Decision tree (exhaustive chi-square automatic interaction detection [CHAID]) analysis for predicting severity. The significant variables from Table 6 were used as predictors.
The visual items show strong interitem correlation when principle component analysis was applied. It was found that all items load with a single factor or component. Highest loading was with object agnosia and lowest with neglect. Therefore, visual symptoms may be considered to be a unidimentional phenomenon.
Relation Between Involvement of Dorsal and Ventral Pathways
Subanalyses were done considering only those patients with involvement of dorsal and ventral visual pathways, irrespective of abnormalities in visuoconstructional and VPFs. This was done because the latter is less specifically localizable, as multiple brain areas contribute to their proper performance.
Involvement of dorsal pathway was seen in 8 (14.55%) as evidenced by the presence of at least 1 subcomponent of Balint’s syndrome. Simultanagnosia (n = 7) was the commonest symptom in this subgroup. All had impaired VCF and VPF. All except 1 (87.5%) also had at least 1 agnosia. Although Gerstmann’s syndrome is due to dorsal pathway involvement, we have not presented the results here, as our study is primarily focused on visual symptoms.
Involvement of ventral pathway was seen in a total of 33 (60%) patients as evidenced by presence of object agnosia and/ or prosopagnosia (none of our patients had alexia without agraphia). Object agnosia was slightly more common than prosopagnosia (Table 2), while 18 had both types of agnosia. Only 7 (21.21%) among these patients with AD with ventral pathway involvement had a Balint’s subcomponent.
Discussion
This is the first study in this field among ethnic Indian patients. In our cohort, early-onset dementia constituted nearly half of the patients. This is quite unusual if we compare it with other reports. The reason of this finding is not clear; however, different genetic predisposition, dietary habit, or other yet unknown factors could be a reason for this. Moreover, in our society forgetfulness in the elderly patients is perceived as normal senescence, and the family members are reluctant to bring them to medical attention which might bring down the percentage of late-onset dementia.
This study evaluated the prevalence and clinical correlations of 9 visuocognitive symptoms and/or sign in patients with AD attending a tertiary-level hospital. To the best of our knowledge, this is the first study from India to look into this less studied phenomenon of AD. An advantage of this study is that it used scales in the local language validated for our population for most of the cognitive assessments.
First and foremost, it should be noted that EOAD outnumbers LOAD in our patient population. It has been shown in previous study 4 that EOAD may present with visuoperceptual deficits, in contrast to the slowly progressive memory deficits seen in LOAD. Thus, our findings may not be truly generalized. However, we only tried to emphasize the fact that VM should be searched for in more detail during evaluation of patients with AD. Although all the patients underwent basic ophthalmological examination, we could not exclude with certainty the possibility of having basic visual defects by contrast sensitivity and VEP.
More than 90% of our patients had at least 1 VM, and about two-thirds had at least 3. It suggests that these signs may be more common than obvious, probably because they are overshadowed by the presence of memory and language disturbances, and the reliance by clinicians on the latter for diagnosis of AD. The fact that we were actively searching for these signs would have also contributed to the discovery of their presence. Our frequency is closer to previous study, where visual agnosis was present in 57% of the cases and Balint’s syndrome in 20% of the cases. It is important for clinicians to be sensitized regarding the presence of VM in AD and to clarify diagnostic dilemma as there may be an overlap of these symptoms with DLB and CJD. Recognition of such symptoms in AD is necessary for early diagnosis 11 and early drug therapy, which becomes relevant in view of the expected development.
Our study showed a good correlation between the presence of VM and the severity of dementia. Similar correlation was found between impairment of visual function and overall cognitive impairment in AD. 5,8,12 Functional studies (positron emission tomography) have also shown greater reduction in cerebral metabolism in patients with AD with visual symptoms than those without. 13 Similarly, cerebral activities in left occipitotemporal area and bilateral inferior and middle temporal regions in AD have been inversely correlated with both prosopagnosia and impairment in performance of instrumental ADL (IADL). 14
Our study indicates that visual symptoms seem to be a unidimentional entity in dementia. Whether such entity may need a separate instrument for further assessment needs to be explored further. Since visual symptoms seem to club together in spite of apparent differences in terms of assessment and known structural involvement as per our present state of knowledge, further study might show whether any neuropathological or circuit correlate might account for the visual symptoms as a whole.
Visuoconstructional and visuoperceptual deficits may arise due to involvement of many different areas of brain. They were the commonest symptoms, underscoring and reflecting their relevance in diagnostic scales for AD. Agnosia for objects and faces was found to be common and may underlie disabilities in AD that are often attributed to language or memory deficits.
In our patients, symptoms/signs suggestive of dorsal and ventral visual pathways involvement occur either simultaneously or in isolation. In those patients where this involvement was in isolation, sole ventral pathway involvement (26 of total 33 ventral pathway involvement) is more common than sole dorsal pathway involvement (1 of 8 dorsal pathway involvement). In a study of 30 patients with AD, object recognition defect was found in 57% when compared to Balint’s syndrome in 20% of the patients. 5 Similar observations about predominant ventral pathway involvement were made in other studies. 12,15 A study from ophthalmological point of view concluded that areas with decreasing order of vulnerability are pattern vision, spatial vision followed by motion, and flicker perception 6 also supporting more agnostic symptoms than visuospatial impairment like simultanagnosia. However, few recent studies on posterior cortical atrophy (PCA) have shown greater dorsal pathway involvement. 16 - 18 The reason behind this apparent contradictory observation in typical AD and PCA, which is mostly composed of AD, is not obvious. We can only presume that in typical AD, the pathology starting in medial temporal area spread to contagious neocortical temporal area before going to the relatively distal parietal area. Similar observation was made by Arriagada et al 19 based on neuropathological evidence of progression in spatial distribution over time. Arnold et al 20 also found that inferior and middle temporal gyrus (area 20 and 21) had a greater density of neuropathological markers, particularly neurofibrillary tangles than posterior parietal cortex (area 7 and 40) or posterior middle temporal gyrus and anterior occipital cortex (parts of area 39, 18, and 19). Thus, typical AD differs from PCA that does not start from medial temporal area.
In contrast to their prevalence of 20% 5 in 1 study, we did not find a single patient with complete Balint’s syndrome in our AD group, although we got subcomponent involvement in 14.55% of patients. This subcomponent involvement is understandable, as different anatomical functional module subserving them are differentially involved in degenerative dementia at least in the early to mid stage of the disease. This pattern of involvement is in contrast to that of stroke. Interestingly, all the patients who were diagnosed to have simultanagnosia on the basis of modified “A-cancellation test” and complicated large figure identification test also failed in “partially simulating letter identification test.” This simple test may be incorporated in assessing visuospatial deficits in AD.
Hemifield defect and hemineglect were found in 9.1% and 7.3% of patients, respectively, in our study. Few reports are available in this regard in typical AD and visual variant of AD. Homonymous field defect (17) and hemineglect (13) have been reported in visual variant of AD, 21 and homonymous quadrantanopia was found in AD. 22
The VH was not uncommon (27.3%) in our sample of AD. Similarly, in a study of 178 patients of AD, 13% had VH. 23 This highlights the fact that presence of VH and symptoms suggesting involvement of posterior areas of brain do not automatically imply a diagnosis of DLB.
Our study has certain limitations. India being a linguistically diverse country, this finding needs to be compared in other linguistic population in the country before making the finding as generalized. In certain cases, administering the tests for VM was hampered by concurrent presence of comprehension defect and attention problem, which has led to a varying number of missing data in each of the components studied. Use of a control group may have made it easier to generalize our findings, but again our aim was only to describe the VM in this group of patients with AD. Finally, we have been limited by the lack of neuropathological facilities for confirmation of diagnosis of AD.
Conclusion
It is evident from our study that VMs are common in AD. Frequency of VMs has got good correlation with severity of dementia. Visual symptoms seem to be a unidimentional entity in dementia. Visual symptom in typical AD is more due to ventral pathway involvement. Presence of individual subcomponent of Balint’s syndrome is a rule rather than exception in AD. “Partially simulating letter identification test” might be handy to pick up simultanagnosia. Presence of visual hallucination (commonly suggest DLB) and hemifield defect (commonly suggest stroke) does not exclude the possibility of AD. It is intriguing to consider that in the initial stage, AD may have different clinical syndromes pertaining to various cognitive domains, for example, language, memory, behavioral or frontal, and of course visual. These distinctions are obviously blurred with progression of disease when the patients actually come to a clinician. It is important to recognize the different “syndromes” early so that early treatment could be started. This information that VM is common in AD will also help the clinician as well as caregivers to understand many of the patients’ disabilities, so that unnecessary consultation from ophthalmologists can be avoided.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.