Specific attentional impairments and complex visual hallucinations in eye disease

RESEARCH ARTICLE

Specific attentional impairments and complex visualy
hallucinations in eye disease
G Graham1, J Dean2, UP Mosimann3, C Colbourn1, R Dudley2, M Clarke4 and D Collerton2

1Doctorate in Clinical Psychology Programme, University of Teesside, Middlesbrough, UK2Institute of Neuroscience, Doctorate of Clinical Psychology, Newcastle University, Newcastle upon Tyne, UK3Institute for Ageing and Health, Newcastle University, Newcastle upon Tyne, UK4Eye Department, Claremont Wing, Royal Victoria Infirmary, Newcastle upon Tyne, UKCorrespondence to: D. Collerton, E-mail: [email protected]

yThis is the sole submission for publication of this work.

Copyr

Objective:To test the prediction by the Perception and Attention Deficit (PAD) model of complex visualhallucinations that cognitive impairment, specifically in visual attention, is a key risk factor for complexhallucinations in eye disease.

Methods: Two studies of elderly patients with acquired eye disease investigated the relationship betweencomplex visual hallucinations (CVH) and impairments in general cognition and verbal attention(Study 1) and between CVH, selective visual attention and visual object perception (Study 2). TheNorth East Visual Hallucinations Inventory was used to classify CVH.

Results: In Study 1, there was no relationship between CVH (n¼ 10/39) and performance on cognitivescreening or verbal attention tasks. In Study 2, participants with CVH (n¼ 11/31) showed poorerperformance on amodified Stroop task (p< 0.05), a novel imagery-based attentional task (p< 0.05) andpicture (p< 0.05) but not silhouette naming (p¼ 0.13) tasks. Performance on these tasks correctlyclassified 83% of the participants as hallucinators or non-hallucinators.

Conclusions: The results suggest that, consistent with the PAD model, complex visual hallucinations inpeople with acquired eye disease are associated with visual attention impairment. Copyright # 2010John Wiley & Sons, Ltd.

Key words: visual hallucinations; attention; Charles Bonnet; eye disease; PAD modelHistory: Received 17 March 2009; Accepted 23 February 2010; Published online 4 August 2010 in Wiley Online Library(wileyonlinelibrary.com).DOI: 10.1002/gps.2522

Introduction

Complex visual hallucinations (CVH) of people,animals or objects are a common phenomenon in awide range of clinical disorders, including dementia,delirium, psychosis and eye disease. A number oftheorists have proposed interactive models which arepotentially applicable across these disorders (Horo-witz, 1975; Diederich et al., 2005; Aleman and Laroi,2008). We have developed the Perceptual andAttention Deficit (PAD) model to explain the genesisof visual hallucinations in these disorders (Collertonet al., 2005). In PAD, CVH occurs when scene

ight # 2010 John Wiley & Sons, Ltd.

perception is constrained by a combination ofimpaired attentional binding and poor sensoryactivation of a correct proto-object (an object whichlies just under the threshold for perception) suchthat an appropriate but incorrect perception—thehallucination—can become conscious. Clinically,these impairments are manifest as poor visual objectattention and perception.The PAD model has received empirical support in

neurodegenerative disorders and in delirium in whichpoor executive and visual perceptual function areassociated with hallucinations (Ramirez-Ruiz et al.,2006; Ozer et al., 2007; Barnes and Boubert, 2008;

Int J Geriatr Psychiatry 2011; 26: 263–267.

264 G. Graham et al.

Brown et al., 2009). However, its extension to otherdisorders, particularly eye disease, is disputed. Drawingon differences in the phenomenology of CVH in eyedisease and dementia, ffytche (2005, 2007) has arguedagainst the PAD model as a general theory of CVH,proposing instead that visual hallucinations in eyedisease reflect the effects of chronic visual deaf-ferentation. Consistent with this proposal, reducedvisual acuity is one of the more consistent risk factorsfor CVH (Khan et al., 2008; ffytche, 2009).There are two ways in which the PAD model could

account for CVH in eye disease. First, people with eyedisease tend to be old, and dementing illnesses (whichare associated with CVH) are common in later life.Around 15% of people with eye disease report CVHand around 10–20% of older people have dementia.CVH in eye disease might therefore simply reflect thecoincidence of another disorder which is a risk factorfor CVH. Previous studies (reviewed in ffytche, 2009)have found inconsistent relationships between com-plex hallucinations and cognition. We tested thishypothesis in Study 1 by looking at the relationshipbetween CVH and performance on modified cognitivescreening and verbal attentional tests, since both havebeen shown to be related to CVH in dementia andwould not be affected by low vision. The secondpossibility is that people with eye disease and CVHhave visual attentional impairments which are notdue to coincidental dementia. In the absence of adementing illness, it is known that specific areas ofthe brain responsible for focussed attention, namelythe prefrontal cortex, are prone to age-related decline(Shan et al., 2005). This second hypothesis was testedin Study 2 by looking at the relationship between CVHand performance on tests of visual attention and visualobject perception. Though these abilities are notentirely separable in practice, we used tasks whichwere primarily of visual attention (in which perform-ance is dependent on focussing on one aspect of aperception) and tests of primarily visual objectperception (in which performance is dependent onholistic recognition).

Methods

Participants

Eye disease participants were recruited as conveniencesamples through Low Vision Clinics at a hospitalOphthalmology department and a charitable supportorganisation in Newcastle and Gateshead. Inclusioncriteria for the eye disease group included: (a) a

Copyright # 2010 John Wiley & Sons, Ltd.

diagnosis of acquired eye disease (wet and dry variantage-related macular degeneration, glaucoma, cataracts,retinal vein obstruction and diabetic retinopathy);(b) significant visual impairment but not total visualloss; (c) fluency in the English language; (d) validinformed consent. Recruitment of participants withCVH was prioritised in order to get large enoughsample sizes for an adequately powered study. Eachstudy used different participants. Favourable ethicalopinions were given by Gateshead and Newcastle LocalResearch Ethics Committees.

Procedures

Participants’ visual impairment was classified on theWorld Health Organisation (WHO, 2003) seven pointcriteria (0¼ normal vision to 6¼ total blindness) usingbest near visual acuity assessed by either Landolt C orSnellen charts displayed at a distance of 40 cm.Demographic data was collected from notes andinterview.The North East Visual Hallucinations Inventory

(NEVHI) (Mosimann et al., 2008) was used to classifyparticipants as having CVH if they experienced at leastone complex visual hallucination per month. Allparticipants with eye disease reported simple visualhallucinations (dots, flashes and similar unformedvisual experiences).The Mini Mental State Examination (MMSE)

(Folstein et al., 1975) was used to screen for grossintellectual impairment. The tool was modified forvisually impaired people. The item which asksrespondents to obey the written sentence ‘close youreyes’ was changed to a verbal command. The itemasking participants to write a sentence of theirchoosing was changed to a verbal response. Finally,the figure copy item was omitted, thus allowing for amaximum score of 29.Study 1 looked at the relationship between CVH and

performance on cognitive screening (MMSE) andverbal attentional tests (letter and animal verbalfluency tasks (Tombaugh et al., 1999), and attentionalitems within the MMSE: serial 7s and spelling ‘World’backwards).Study 2 focussed on the relationship between CVH

and performance on tests of primarily visual attentionand tests of primarily visual object perception.Standard tests of visual object attention were modifiedin order to capitalise on residual vision in eye disease.First, the standard Stroop colour-word paradigm(Macleod, 1991) which is impaired in patients withCVH in Parkinson’s disease (Barnes and Boubert,


Hallucinations and attention in eye disease 265

2008) was adapted. This task compares a participant’sresponse times in naming the ink colour of visualstimuli in a baseline condition (i.e. a list of colouredpatches) and an interference condition (a list ofcolour words printed in incongruously coloured inke.g. the word red printed in blue ink). The task wasmodified by increasing the size and simplifying theformat of the stimuli. In line with traditional Strooptasks, the difference in response times betweenthe baseline and interference trial was used as theperformance measure. Second, a novel measure ofvisual attention using mental imagery, the ImageryTask, was developed. Visual research suggests thatmental imagery draws on most of the same neuralmechanisms as visual perception (Kosslyn et al.,2001). Thus, since imagery and perceptual perform-ance are closely related, an imagery-based testprovided some solution to the confounding effectof poor vision on visual attention in that it shouldprovide a measure of attentional ability which isindependent of acuity. Each item asked participantsto summon up two memorable images (e.g. thecapital letters F and A), following which they wereasked questions which required them to focus onspecific parts of the images (e.g. ‘Which letter hasthe most horizontal lines?’). Accuracy of responsesdid not differ between groups and, as is usual inattentional tasks, time to complete the task was usedas the measure of performance. Copies of these tasksare available from the authors.Visual object perception was assessed in Study 2,

firstly using the Graded Naming Test (GNT, McKennaand Warrington, 1980) with its scoring adapted toprovide a purer measure of visual perception.Participants’ responses were marked as correct as longas their answers demonstrated that they had correctlyperceived the pictures (e.g. describing the Kangaroo asan animal from Australia who hops and carries its babyin a pouch) even if they could not provide the correctname. Second, the Silhouettes subtest of the VisualObject and Space Perception Battery was used as ameasure of visual object perception (Warrington andJames, 1991).Each of the attention and perception tests were

administered on a 33 cm� 21 cm visual display unitpositioned 40–60 cm from the participant.As these were novel or heavily adapted tasks, a

healthy control group was also recruited for Study 2 toprovide normative data. The control group wasrecruited through a range of community organisations,using the same inclusion criteria except that peoplewith uncorrected visual impairment or cognitiveimpairment were excluded.


Analyses

For both studies, eye disease participants with CVHwere compared to eye disease participants with noCVH in a cross sectional group design using t-tests orMann–Whitney exact tests as appropriate. In Study 2,a logistic regression was performed to assess therelationship between CVH and performance on thecognitive measures.

Results

Participants in the CVH groups reported a range ofhallucinatory images on the NEVHI characteristicof CVH in eye disease including faces and figuresof children, babies and adults, animals, objects andcartoon-type creatures. Participants had chronic,frequent CVH, with 90% hallucinating every weekand onsets between several months and 2 yearspreviously.Selected demographic information and results from

each of the measures are presented in Table 1. Age,education and use of medications did not differbetween the CVH and non-CVH groups, thoughanalyses in Study 2 showed the eye disease groups to bereliably older, less educated and more medicated thanthe control group.In both studies global cognitive functioning, as

measured by the MMSE, did not differ between theCVH, non-CVH or healthy control groups. Consistentwith previous findings, participants with CVH hadpoorer visual acuity than those who did not experienceCVH (effect size r¼ 0.8).In Study 1, there was no difference between CVH

and non-CVH groups on the verbal fluency tasks, or onserial 7s or spelling World backwards in the MMSE(data not shown).In Study 2, participants with CVH were slower to

complete all tasks though this only reached statisticalreliability in the Adapted Stroop interference score(effect size r¼ 0.6) and Imagery tasks (r¼ 0.4). Scoreson both object perception tasks were also lower in theCVH group, with differences reaching conventionallevels of significance on the GNT (r¼ 0.5), althoughnot on the Silhouette task (r¼ 0.1), perhaps because ofgreater variability in performance on the latter task.Logistic regression showed that combined perform-ance on the visual attention and visual objectperception measures correctly classified 90% of thenon-CVH group, 72.7% of the CVH group and 83.9%of the sample overall (p< 0.05). Negerkerke R2 (0.61)was indicative of a large effect size. Roa’s efficient score


Table 1 Comparisons of CVH and non-CVH groups

Study 1 Study 2

CVH (n¼ 10) Non-CVH(n¼29)

Control(n¼27)

CVH (n¼11) Non-CVH(n¼20)

Age in years 71 (22) 78 (12) 68 (8)** 82 (6) 79 (7)Education in years 11 (3) 9 (1) 13 (3)** 11 (2) 12 (2)Degree of visual impairment (World HealthOrganisation Classification)

1.5 (1.3)* 0.5 (0.7) 0 (0)** 2.3 (1.3)* 1.3 (1.6)

Adapted MMSE (total score) 26.4 (1.4) 26.9 (1.7) 27.4 (0.7) 28.0 (1.2) 28.1 (0.9)Number of current medications 5.4 (3.8) 4.5 (2.7) 1.2** (1.3) 5.0 (2.5) 3.9 (2.8)FAS verbal fluency (age-educationadjusted score)

47 (26) 54 (34)

Animals verbal fluency (age-educationadjusted score)

44 (27) 43 (27)

Silhouette Naming Test (total correct) 21 (3)** 15 (5) 16 (6)Adapted Graded Naming Test (total correct) 28 (2) 15 (9)* 23 (8)Images Test (seconds) 39 (18) 82 (29)* 56 (29)Adapted Stroop Interference Task (seconds) 21 (18) 33 (18)* 15 (7)

All data are expressed as mean (SD).

*CVH group differs from non-CVH group at p< 0.05; **Control group differs from eye disease groups.

266 G. Graham et al.

statistics suggested that the Adapted Stroop Task(p< 0.001) had the most predictive strength, with theImagery task (p< 0.05) and Graded Naming Test(p< 0.05) also reaching statistical significance. Repla-cing scores on the cognitive tests with visual acuityreduced the proportion of cases correctly classified inthe logistic regression to 74% overall and the modelwas no longer statistically reliable (overall p> 0.05).Differences between hallucinating and non-hallucinat-ing groups on the Adapted Stroop Task (the mostpowerful discriminator) remained reliable in ananalysis of covariance when acuity was controlledfor (F(2,26)¼ 7.41, p¼ 0.03). Additionally, visualacuity was not reliably related to Stroop performanceor to Imagery performance in either the normal ornon-CVH groups, though it was in the CVH group(r¼�0.6, p< 0.05); suggesting that perceptual per-formance rather than poor eyesight more stronglydifferentiates hallucinating from non-hallucinatinggroups.

Conclusions

Consistent with four out of six previous studies, theresults of Study 1 suggest that CVH in eye disease aregenerally neither due to a coincident dementia nor ageneral dysexecutive syndrome. Neither total MMSEscores, nor performances on the verbal attentionaltasks which have been associated with CVH indementia differed between the groups.


Consistent with the PAD model, Study 2 suggeststhat performances on tests of visual attention andvisual perception are poorer in participants withCVH than in those without. It is expected thatpoorer eyesight would lead to worse performanceon visual perceptual tasks (the PAD model merelyrequires poor perceptual performance withoutrequiring these to be exclusively from sensory orcerebral impairments, though it is also the case thatcognitive performance was more closely relatedto hallucinations than is acuity), but might thesame be true of the attentional tasks? This cannot befully excluded as an explanation. However, groupdifferences remain after statistically controllingfor acuity, and the Stroop performance of thenon-CVH group is actually better than that of thecontrol group, perhaps because their labouredperception reduced the interference between wordsand colours. This suggests that poor vision per sedoes not account for poor performance onthe Stroop task. A further possibility, which cannotbe excluded by this data, is that the CVH group issimply slower.This study provides the first, preliminary, evidence

that a single mechanism may underlie hallucinationsin eye disease and dementia. Further work is neededto replicate these findings, to define more closely thenature of the attentional and perceptual impairmentsin eye disease and their inter-relationships, and toinvestigate the relationship between chronic visualloss and attentional function.


Key Points

� The cause of complex visual hallucinations in eyedisease is not known.

� Impaired visual attention is a proposed risk factorfor visual hallucinations in eye disease.

� This study has found that a specific impairment invisual attention is associated with complexhallucinations in patients with eye disease.

Hallucinations and attention in eye disease 267

Conflict of interest

None declared.

Acknowledgements

The authors would like to thank the staff and usersof Ophthalmology services at the Royal VictoriaInfirmary, Newcastle, Sight Services, Gateshead, andthe Macular Degeneration Support Group for theirassistance with these studies.

References

Aleman A, Laroi F. 2008. Hallucinations: The Science of Idiosyncratic Perceptions.American Psychological Association: Washington.

Barnes J, Boubert L. 2008. Executive functions are impaired in patients with Parkin-son’s disease with visual hallucinations. J Neurol Neurosurg Psychiat 79: 190–192.


Brown LJE, McGrory S, McLaren L, et al. 2009. Cognitive visual perceptual deficits inpatients with delirium. J Neurol Neurosurg Psychiat 80: 594–599.

Collerton D, Perry E, McKeith I. 2005. Why people see things that are not there: anovel Perception and Attention Deficit model for recurrent complex visualhallucinations. Behav Brain Sci 28: 737–757.

Diederich NJ, Goetz CG, Stebbins GT. 2005. Repeated visual hallucinations inParkinson’s disease as disturbed external/internal perceptions: focused reviewand a new integrative model. Mov Disord 20: 130–140.

ffytche DH. 2005. Two visual hallucinatory syndromes. Behav Brain Sci 28: 763–764.

ffytche DH. 2007. Visual hallucinatory syndromes: past, present and future. Dial ClinNeurosci 9: 173–189.

ffytche DH. 2009. Visual hallucinations in eye disease. Curr Opin Neurol 22: 28–35.Folstein MF, Folstein SE, McHugh PR. 1975. Mini Mental State. A practical method of

grading the cognitive state of patients for the clinician. J Psychiat Res 12: 189–198.

Horowitz MJ. 1975. Hallucinations: an information-processing approach. In Hallu-cinations. Behavior, Experience and Theory. Siegel RK, West LJ (eds). John Wileyand Sons: New York; 163–195.

Khan JC, Shahid H, Thurlby DA. et al. 2008. Charles Bonnet syndrome in age-related macular degeneration; the nature and frequency of images in subjectswith end-stage disease. Ophthalmic Epidemiol 15: 202–208.

Kosslyn SM, Ganis G, Thompson WL. 2001. Neural foundations of imagery. Nat RevNeurosci 2: 635–642.

Macleod CM. 1991. Half a century of research on the Stroop effect: an integrativereview. Psych Bull 109: 163–203.

McKenna P, Warrington EK. 1980. Graded Naming Test. NFER-Nelson PublishingCompany: London.

Mosimann UP, Collerton D, Dudley R. et al. 2008. A semi-structured interview toassess visual hallucinations in older people. Int J Geriatr Psychiatr 23: 712–718.

Ozer F, Merai H, Hanoglu L. et al. 2007. Cognitive impairment patternsin Parkinson’s disease with visual hallucinations. J Clin Neurosci 14: 742–746.

Ramirez-Ruiz B, Junque C, Marti MJ. et al. 2006. Neuropsychological deficits inParkinson’s disease patients with visual hallucinations. Mov Disord 21: 1483–1487.

Shan ZY, Lui JZ, Saghal V,Wang B, Yue GH. 2005. Selective atrophy of left hemisphereand frontal lobe of the brain in old men. J Gerontol A Biol Sci Med 60: 165–174.

Tombaugh TN, Kozak J, Rees L. 1999. Normative data stratified by age and educationfor twomeasures of verbal fluency: FAS and animal naming. Arch Clin Neuropsychol14: 167–177.

Warrington EK, James M. 1991. Visual Object and Space Perception Battery Bury.Thames Valley Test Company: St Edmunds.

World Health Organisation. 2003 International statistical Classification of Diseases andRelated Problems. 10th Revision Chapter VII Blindness and Low Vision. World HealthOrganisation: Geneva.


Documents

Specific attentional impairments and complex visual hallucinations in eye disease