(Un)awareness of unilateral spatial neglect: A quantitative evaluation of performance in visuo-spatial tasks

www.sciencedirect.com

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2

Available online at

ScienceDirect

Journal homepage: www.elsevier.com/locate/cortex

Special issue: Research report

(Un)awareness of unilateral spatial neglect: Aquantitative evaluation of performance invisuo-spatial tasks

Roberta Ronchi a,b,c,*, Nadia Bolognini a,b,d, Marcello Gallucci a,d,Laura Chiapella e, Lorella Algeri e, Maria Simonetta Spada e andGiuseppe Vallar a,b,d

a Department of Psychology, University of Milano-Bicocca, Milano, Italyb Neuropsychological Laboratory, S. Luca Hospital, IRCCS Istituto Auxologico Italiano, Milano, Italyc Laboratory of Cognitive Neuroscience, Brain Mind Institute, School of Life Sciences, Ecole Polytechnique F�ed�erale de

Lausanne, Lausanne, Switzerlandd Milan Centre for Neuroscience, University of Milano-Bicocca, Piazza dell'Ateneo Nuovo 1, Milano, Italye Unit�a di Psicologia Clinica, Ospedale Papa Giovanni XXIII, Bergamo, Italy

a r t i c l e i n f o

Article history:

Received 23 December 2013

Reviewed 6 April 2014

Revised 6 June 2014

Accepted 6 October 2014

Keywords:

Unilateral left spatial neglect

Unawareness/anosognosia for left

neglect and hemiplegia

Right-brain damage

Evaluation of cognitive performance

* Corresponding author. Laboratory of CogF�ed�erale de Lausanne, Station 19, CH-1015,

E-mail address: [email protected] (Rhttp://dx.doi.org/10.1016/j.cortex.2014.10.0040010-9452/© 2014 Elsevier Ltd. All rights rese

a b s t r a c t

Right-brain-damaged patients with unilateral spatial neglect are usually unaware (anosogno-

sic) about their spatial deficits. However, in the scientific literature there is a lack of systematic

and quantitative evaluation of this kind of unawareness, despite the negative impact of ano-

sognosia on rehabilitation programs. This study investigated anosognosia for neglect-related

impairments at different clinical tasks, by means of a quantitative assessment. Patients were

tested in two different conditions (before and after execution of each task), in order to evaluate

changes in the level of awareness of neglect-related behaviours triggered by task execution.

Twenty-nine right-brain-damaged patients (17 with left spatial neglect) and 27 neurologically

unimpaired controls entered the study. Anosognosia for spatial deficits is not pervasive, with

different tasks evoking different degrees of awareness about neglect symptoms. Indeed, pa-

tients showed a largely preserved awareness about their performance in complex visuo-motor

spatial and reading tasks; conversely, theywere impaired in evaluating their spatial difficulties

in line bisection and drawing from memory, showing over-estimation of their performance.

The selectivity of the patients' unawareness of specific manifestations of spatial neglect is

further supported by their preserved awareness of performance at a linguistic task, and by the

absence of anosognosia for hemiplegia. This evidence indicates that discrete processes are

involved in theawaremonitoringof cognitiveandmotorperformance,whichcanbeselectively

compromised by brain damage. Awareness of spatial difficulties is supported by a number of

distinct components, and influenced by the specific skills required to perform a given task.

© 2014 Elsevier Ltd. All rights reserved.

nitive Neuroscience, BraLausanne, Switzerland.. Ronchi).

rved.

in Mind Institute, School of Life Sciences, Ecole Polytechnique

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2168

1. Introduction

“Anosognosia” (from the ancient Greek, “a” without, “nosos”

disease, “gnosis” knowledge) is a term, a neologism, intro-

duced in the neurological literature by the French neurologist

Joseph Babinski (Philippon & Poirier, 2009) to indicate a lack of

awareness (i.e., unawareness) of a neurological deficit, spe-

cifically, left hemiplegia (Babinski, 1914; Papagno & Vallar,

2003, for revision). Indeed, the right-brain-damaged patients

described by Babinski (1914) ignored or appeared to ignore

their paralysis, namely they presented with anosognosia for

left hemiplegia. Patients may also show indifference, lack of

concern for left hemiplegia, a condition termed by Babinski

anosodiaphoria (Prigatano, 2010). Here, we will use the terms of

“anosognosia” and “unawareness” as synonymous, as done in

the neuropsychological literature (Jenkinson, Preston, & Ellis,

2011; Langer, 2009; McGlynn & Schacter, 1989).

Anosognosia is not confined to neurological deficits,

namely motor, somatosensory and visual field disorders

(Prigatano, 2010; Prigatano & Schacter, 1991; Vallar & Ronchi,

2006). Although the majority of studies focus on unaware-

ness of neurological impairment (mainly motor deficits, but

also somatosensory and visual half-field disorders, see Bisiach

& Geminiani, 1991; Celesia, Brigell, & Vaphiades, 1997), pa-

tients may be also unaware of neuropsychological deficits,

including aphasia (Kertesz, 2010; Lebrun, 1987), memory dis-

orders (Akai, Hanyu, Sakurai, Sato, & Iwamoto, 2009), and

constructional apraxia (CA) (Rinaldi, Piras, & Pizzamiglio,

2010). With respect to spatial disorders, the clinical observa-

tion of right-brain-damaged patients with left spatial neglect

(Husain et al., 2001; Vallar, 1998; Vallar & Bolognini, 2014)

suggests that they are not aware of their defective perception

and exploration of the side of space contralateral to the side of

the hemispheric lesion (contralesional). The unawareness of

spatial deficits in the neglect syndrome is clinically relevant,

since it impacts negatively on activities of daily living, hence

on functional outcome after hospital discharge (Vossel,Weiss,

Eschenbeck, & Fink, 2013). So far, only a few studies have

examined anosognosia for spatial deficits in neglect experi-

mentally. Unawareness of left neglect dyslexia (see review in

Vallar, Burani, & Arduino, 2010), and of spatial neglect in

drawing was investigated by Berti, L�adavas, and Della Corte

(1996). In this seminal study, Berti et al. (1996) required right-

brain-damaged patients to perform an evaluation of their

reading and drawing performances; patients were asked to

report whether or not they had read correctly the sentence,

and if they were happywith their drawing performance, being

dichotomised as aware or unaware of left neglect dyslexia or

neglect in drawing, based on their response (“yes”, “happy”),

and on any subsequent confabulatory response. Berti et al.

(1996) also found that anosognosia for spatial neglect in sen-

tence reading and drawing could occur without anosognosia

for left hemiplegia (Bisiach, Perani, Vallar, & Berti, 1986). A

more recent study (Jehkonen, Ahonen, Dastidar, Laippala, &

Vilkki, 2000) showed a double dissociation between anosog-

nosia for left spatial neglect, assessed by asking patients to

report about their spatial difficulties with a yes/no response

(“Do you have any difficulties observing any part of the

space?” and if necessary, giving alternatives: left, right, both,

none), and anosognosia for left hemiplegia and, more gener-

ally, for the illness. As for its neural underpinnings, un-

awareness of left spatial neglect, indexed by the comparison

between the patients' self-rating [1 (severe) to 5 (no diffi-

culties)] of their performance in cancellation, line bisection,

drawing and text reading tasks, and that made by an external

observer, was associated with lesions of the angular gyrus of

the inferior parietal lobule, and the superior temporal gyrus in

the right hemisphere (Vossel et al., 2012). This lesion pattern,

largely overlappingwith that of spatial neglect itself (review in

Vallar & Bolognini, 2014), suggests a close relationship be-

tween the spatial processes dysfunctional in left neglect and

their monitoring, namely the cognitive process that allow to

check whether an activity or a task e here the spatial perfor-

mance e is carried out correctly.

Overall, the behavioural results from the literature about

the presence and the extent of anosognosia for spatial neglect

are not exhaustive, as previous researches have examined

only selective aspects of this topic, andmostly in a qualitative

way. In order to fill this gap in the literature, the present study

aimed at exploring the presence of anosognosia for spatial

neglect by assessing how neglect patients evaluate their

ability to perform a series of different clinical tasks,

commonly used for the diagnosis of spatial neglect. The

rationale of using different tasks is in agreement with the

concept of spatial neglect as a multi-componential syndrome

(Vallar, 1998), featured by dissociable pathological neglect

symptoms in different tasks (review in Vallar & Bolognini,

2014). Linear regression analyses were used to investigate

how the severity of the spatial deficit in each task impacted on

themonitoring of the spatial performance in that specific task.

We hypothesized that neglect patients could show unaware-

ness of the neglect-related performance deficit, indepen-

dently from the specific task considered. This hypothesis is in

line with a traditional, clinical, view of anosognosia as a

pervasive component of left spatial neglect. Some previous

evidence also supports this hypothesis. A perusal of the data

by Berti and collaborators (1996) suggests the absence of

double dissociation between anosognosia for neglect dyslexia

and for neglect in drawing (i.e., neglect patients were either

aware or unaware of both deficits); however, Berti et al. only

compared two tests in a qualitative descriptive fashion,

without highlighting the possible association between the

level of awareness and the actual performance at each task.

This information is relevant to understand whether the

presence of anosognosia for spatial neglect is linked to the

neglect-related behaviour: if this is the case, anosognosia for

spatial neglect should selectively affect the patients' aware-

ness of their neglect-related performance only in those tasks

where a spatial deficit is present.

There is also evidence, in other neuropsychological disor-

ders, that the patients' monitoring of their cognitive perfor-

mance may be more accurate after task execution (Ansell &

Bucks, 2006; Barrett, Eslinger, Ballentine, & Heilman, 2005).

By comparing how neglect patients evaluate their perfor-

mance before and after having executed a given task, we

aimed at verifying the possibility that patients showing ano-

sognosia for neglect in a specific task before performing it,

may be able (at least partially) to update their erroneous

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2 169

evaluation, showing improved awareness after task

execution.

Finally, the relationship between anosognosia for spatial

neglect and anosognosia for motor deficits was explored, to

verify previous evidence that monitoring processes of spatial

andmotor deficits are independent (Berti et al., 1996; Jehkonen

et al., 2000).

2. Materials and methods

2.1. Participants

Twenty-nine patients (14 males, 15 females; mean age:

56.7± 16.4years, range: 34e97;meaneducation: 11.8± 4.3 years;

range: 5e18)with righthemisphere lesions, asassessedbyCTor

MRI scan, participated in this study. The aetiology of the focal

lesion was vascular in 26 patients (14 ischaemic, 12 haemor-

rhagic; sub-acute and chronic phase,mean duration of disease:

4.5 months, range: .75e29months), and neoplastic in three. All

participants were right-handed, with no history or evidence of

previous neurological or psychiatric disorders. Global cognitive

efficiency was assessed using the Mini Mental State Examina-

tion (Folstein, Folstein, & Mchugh, 1975; Grigoletto, Zappala,

Andeerson, & Lebowitz, 1999) or a verbal reasoning task

Table 1 e Demographic and neurological data of 29 right-brain-

Sex/Age Education (years) Etiology/Lesion side

P1 F/75 13 I/T

P2 M/74 12 I/bg

P3 M/36 13 I/F-T-In

P4 M/40 8 H/F-T-P

P5 F/65 13 I/bg-ic

P6 F/34 13 I/F-T-bg

P7 M/46 17 I/Th-ic

P8 F/41 8 N/F

P9 F/52 8 I/bg

P10 M/34 13 H/bg-ic

P11 M/53 18 I/Sylvian region

P12 F/50 17 I-H/F

P13 F/66 5 N/F-T

P14 F/38 13 I/F-T-O-In-ic

P15 F/37 8 H/F-bg

P16 F/34 12 I-H/Sylvian region

P17 M/70 17 I/T-P-O

P18 M/71 18 I/Sylvian region

P19 M/54 17 I/bg-ic

P20 F/57 8 H/F-P-In-bg

P21 M/61 5 I/Sylvian region

P22 M/54 13 I/F-T-In

P23 F/97 18 I-H/F-P

P24 F/73 5 H/F-s-cort

P25 F/50 13 I-H/In-bg

P26 M/60 8 I/T-O

P27 M/77 13 H/bg

P28 F/74 8 H/F-P

P29 M/72 7 N/parasellar-T-bg

I: ischaemic lesion; H: haemorrhagic lesion; N: neoplastic lesion.

F: frontal; P: parietal; T: temporal; O: occipital; In: insula; ic: internal caps

M: left motor deficit; SS: left somatosensory deficit; V: left visual half-fie

anosognosia; PN: personal neglect; SP: somatoparaphrenia; þ: presence o

(Spinnler and Tognoni, 1987). Contralesional motor, somato-

sensory and visual half-field defects were evaluated by a stan-

dard neurological examination. Anosognosia for neurological

(motor, somatosensory, visual) deficits was assessed with a

short standardized interview (Bisiach, Vallar, Perani, Papagno,

& Berti, 1986), comprising ad-hoc questions for each type of

deficit. The score range was 0e3, with “0” corresponding to full

awareness about the deficit and “3” to a severe anosognosia,

unchangeable by the neurological demonstration of the deficit.

The patients' demographic and neurological data are re-

ported in Table 1.

Twenty-seven right-handed neurologically unimpaired

participants (13 males), matched for age (mean: 57.7 ± 15.5

years, range: 28e83) and years of education (mean: 11.6 ± 4.7

years, range: 2e18), served as controls (C).

2.2. Baseline evaluation: neuropsychological assessmentfor left spatial neglect

The presence/absence of left spatial neglect (Nþ/N�) was first

assessed by a baseline evaluation, done 7 days before the

experimental protocol (Day-7). The baseline assessment

included the following tests: line bisection (six lines; Ronchi,

Posteraro, Fortis, Bricolo, & Vallar, 2009), letter (Diller &

Weinberg, 1977), star (Wilson, Cockburn, & Halligan, 1987)

damaged patients.

Neurological deficit Associated deficit

M SS V AN PN SP

þ e e e e e

e e e e e e

e e e e e e

þ e e e þ e

þ e e e e e

þ e e e e e

þ e e e e e

þ e e e e e

þ e e e e e

þ þ e þSS e e

e e þ e e e

þ e e e e e

þ e e e e e

þ þ þ þSS-V e e

þ þ þ e e e

þ þ þ þSS-V e e

þ e þ e e e

þ þ þ þSS e e

þ þ þ þSS-V þ e

þ þ e þSS e e

e e þ e e e

e e e e e e

þ þ þ þM-SS-V þ þþ þ þ þSS-V þ e

þ þ þ þM-SS-V e e

e e þ e e e

þ e e e e e

þ þ þ þSS-V e e

þ þ þ þSS-V þ þ

ule; bg: basal ganglia; Th: thalamus; s-cort: sub-cortical.

ld deficit; e: left extinction to double simultaneous stimulation; AN:

f deficit; e: absence of deficit.

Fig. 1 e Lesions of Nþ patients drawn on standard MRI template with a 1-mm slice distance (voxels of 1 mm3) (MRIcro

software, see Rorden & Brett, 2000; www.mricro.com). Patients P13 and P29 not included, due to a neoplastic lesion (see

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2170

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2 171

and bell (Gauthier, Dehaut,& Joanette, 1989) cancellation tasks,

sentence reading (Pizzamiglio et al., 1992), drawing tests by

copy [a complex figure with five elements (Gainotti, Messerli,&

Tissot, 1972), two daisies (Halligan & Marshall, 1993), one daisy

(Ronchi et al., 2009), one butterfly], and from memory [a clock

(Ronchi et al., 2009), one daisy, one butterfly]. Patients were

classified as affected by left neglect (Nþ), if they showed a

defective performance (with reference to normative data or to

the controls' scores) on at least one task (between target

cancellation and line bisection; Bisiach, Bulgarelli, Sterzi, &

Vallar, 1983; Ferber & Karnath, 2001; Vallar & Perani, 1986).

Participants who did not meet this criterion were classified as

not affected by neglect (N�). Accordingly, the patients' sample

included 12 N� (P1eP12) and 17 Nþ (P13eP29) (see below). The

inspection of the scores of the N� patients indicated that they

did not show any sign of unilateral spatial neglect also in the

other tasks (i.e., drawings and reading), not used as inclusion

criteria in the present study. Following the clinical standard

procedure, after each administered task, the examiner asked

each patient if the task had been completed, and if the patient

considered that he/she had done everything had been

requested. All patients answered “yes” to the examiner'squestion, both when they showed signs of spatial neglect, and

when they did not. Based on this clinical observation, all

neglect patients could be considered “anosognosic” about their

neglect-related performance in neuropsychological tasks.

The lesion localisation of Nþ patients is shown in Fig. 1.

2.3. Experimental tasks

In the experimental section, both right-brain-damaged patients

and healthy participants were required to evaluate their per-

formance (see the Procedure below) relative to a list of tasks.

Six tasks were chosen to assess the presence of unilateral

spatial neglect, and the evaluation that patients did about their

performance was taken as an indication of the presence/

absence of anosognosia for neglect-related deficits in each

task. Clinically (both during the baseline and the experimental

assessments) all patients with left neglect showed “anosog-

nosia” for neglect itself, namely stated on the examiner'srequest that they had performed, and completed the task,

according to the instructions. Accordingly, we could not

compare patients based on the clinical evidence of anosog-

nosia for neglect, since all patients were clinically “anosog-

nosic”. We asked instead for amore quantitative, and possibly

more sensitive, task-related evaluation of performance. We

also asked participants for evaluating their performance in a

cognitive non-spatial linguistic task, to exclude that neglect

patients presented with a general impairment in evaluating

their cognitive performance. Finally, we assessed the presence

of motor deficits of the left (contralesional) upper limb,

including tasks that require participants doing direct move-

ments (DMs) with the arms, and executing a series of actions

for which both hands were necessary or only one was suffi-

cient (control condition for motor evaluation); also for these

Table 1); patient P28 scan images not available for mapping. Max

putamen and white matter underlying frontal, temporal and pa

coordinates of each transverse section are reported.

motor tasks, participants had to evaluate their performance, in

order to detect the presence of anosognosia for motor deficits.

In detail, the tasks administered to the participants were

the following:

2.3.1. Unilateral spatial neglect tests

- Line bisection (Ronchi et al., 2009). The taskwas tomarkwith

a pencil the mid-point of six horizontal black lines

(width ¼ 2 mm; length ¼ 10�15�25 cm, two lines for each

length), printed on an A4 sheet and presented in a random

fixed order. The score was the deviation of the participants'mark from the objective mid-point, measured to the

nearest mm; a positive score denoted a rightward

displacement, a negative score a leftward displacement.

- Star cancellation (modified from Wilson et al., 1987). Partic-

ipants were instructed to cross out 60 small stars (30 in

each side of the sheet), distributed among distracters (big

stars, letters, Italian words) on an A4 sheet.

- Letter cancellation (Diller & Weinberg, 1977). The task was to

cross out all of 104 ‘H’ letters (53 in the left-hand side, 51 in

the right-hand side of the sheet) printed on an A3 sheet,

distributed among other letter distracters.

- Complex drawing by copy (Fortis et al., 2010; Gainotti et al.,

1972). The task was to copy a complex figure with two

trees in the left-hand side, two pine trees in the right-hand

side, and a house printed in the centre of an A4 sheet.

- Clock drawing from memory (Ronchi et al., 2009). The task

was to draw frommemory the hours of a clock in a circular

quadrant (Ø ¼ 12 cm), printed on an A4 sheet.

- Sentence reading. Participants were given 12 sentences to

read. Each sentence was printed horizontally in black up-

percase letters (Arial, pt.14) in the centre of an A4 sheet and

presented to participants one at a time. The length of each

sentence varied from 7 to 15 words.

For healthy controls, the mean bisection error was

�.49 mm (SD: 3.3 mm, range: �8/þ5.8); the maximum differ-

ence between omission errors on the two sides of the sheet

(right-sided omissions minus left-sided omissions) was one in

the Star and two in the Letter cancellation tasks (Vallar,

Rusconi, Fontana, & Musicco, 1994); the maximum omission

score in the complex drawing task was of .5; the maximum

neglect-like errors in the clock drawing test was of 1; no

neglect-like errors (Ellis, Flude,& Young, 1987) in the sentence

reading task were recorded. One control participant (C4) did

not perform the clock drawing test.

2.3.2. Cognitive non-spatial linguistic test

- Phonemic verbal fluency (Novelli et al., 1986). This task

required producing as many words as they could, begin-

ning with a given letter (F-P-L) within 60 sec for each letter;

a score �16 is pathological (Novelli et al., 1986). All controls

performed normal at this test (scores �16).

imum overlap of 14 out of 17 Nþ patients was subcortical:

rietal regions. Montreal Neurological Institute (MNI) Z-

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2172

2.3.3. Upper-limb motor tests

- Direct movements (see Bisiach, Vallar, et al., 1986). Partici-

pants were asked to raise their right and left arms,

separately.

- Unimanual tasks (see Marcel, Tegner, & Nimmo-Smith,

2004). Participants received instructions to perform the

following actions with one hand: brush their teeth; comb

their hair; drink a glass of water; open a door; sign their

name.

- Bimanual tasks (see Marcel et al., 2004). Participants had to

perform the following actions by using both hands: shuffle

adeckof cards; separate two sheets glued in the centre; tie a

bowona cylindrical box; put a keyonakeychain; opena tin.

Control participants executed all motor tasks in a proper

way.

2.4. Procedure

During the experimental session (one week after the baseline

assessment, Day 1), before performing the experimental tasks

(i.e., unilateral spatial neglect, cognitive non-spatial linguistic

and upper-limb motor tasks; see details in the section 2.3),

participantswere asked to answer to the following question; for

every task,we varied the description of the activity accordingly:

- Pre-condition (PRE): “In your present state, how well can you

perform a task in which you have to… (e.g.: cross out all the small

stars on the sheet)?”.

Then, the patients were asked to execute each task, and

immediately at the end of each task to answer to the following

question:

- Post-condition (POST): “Howwell have you performed this task,

in which you had to…?”.

For the complex drawing task, scores related to left spatial

neglect and CA were computed separately. CA scores were

assigned to the drawings of Nþ and N� patients: 2 for each

element correctly oriented and drawn; 1 for each element

distorted or poorly oriented, but recognizable; 0 for each

element not recognizable and/or for the absence of the three-

dimensional component in the “house” element. Possible

differences in neglect severity were removed by computing

the proportion between neglect and CA scores (CAp), with the

CAp scores ranging from 0 (severe CA in all drawn elements)

to 1 (no CA in drawn elements).

Upper-limb motor function was scored as: 0 ¼ perfect

execution; 1 ¼ action performed with slowness/minimal

clumsiness; 2 ¼ action performed with a great effort;

3 ¼ action impossible to perform.

For each question, participants received instructions to

indicate their response on a vertical 18-cm Likert scale, shown

inFig. 2a. Each scalewas centredonanA4sheet, andsubdivided

in 7-points graduating in colour from the bottom (dark redwith

a “minus” sign ¼ 1, namely: Impossible to perform) to the top

(dark green with a “plus” sign ¼ 7, namely: Flawless

performance), through white at the mid-point of the scale

(score ¼ 4). The examiner instructed participants to indicate,

with their right index finger, the point on the scale that best

represented their ability to perform each task. Therefore, in

taskspoorlyperformeda lowerscore indicatedawarenessof the

impairment, a higher score unawareness of the impairment.

The experiment schedule is summarized in Fig. 2b.

2.5. Statistical analyses

The mean Likert evaluation scores of right-brain-damaged

patients were compared with their performance scores ob-

tained in each experimental task. To assess the correspon-

dence between the patients' evaluation and their performance,

linear regression analyses were conducted to assess whether

the level of performance in each task (independent variable)

was related to the corresponding evaluation scores. With the

regressions, we examined separately both the PRE- and the

POST-conditions, to verify how patients evaluated their per-

formance in each task, before and after its execution. In the

regression analyses, patients without (N�) and with (Nþ)

neglect were considered as a unique group, since inserting

their performance scores in the model (independent variable)

allows to detect the presence/absence of a pathological spatial

performance in the task; this, in turn, is inherently informative

about differences between N� and Nþ. Figs. 3e7 show the

scatterplots and regression lines of the relationship between

performance and evaluation in each task: to ease the com-

parison between plots, performance scores were converted in

T-scores, which are standardized scores. A score of 50 repre-

sents the mean (SD ± 10). A difference of 10 from the mean

indicates a difference of one standard deviation: thus, a score

of 60 is one standard deviation above themean,while a score of

30 is two standard deviations below the mean.

As data were not clearly fulfilling the regression assump-

tions, Spearman non-parametric correlation coefficients were

computed on the same variables involved in the regression

models, to assess the robustness of the results against

possible violations of these assumptions. If parametric and

non-parametric statistics bring to converging evidence

(Teuber, 1955), the conclusions are statistically more reliable:

indeed, the former are usually more informative and easier to

interpret, the latter more robust (Siegel & Castellan, 1988).

The results of the neurologically unimpaired participants

were not compared statistically with those of right-brain-

damaged patients, as we did not expect a substantial vari-

ability in their evaluation, because there was not substantial

variability in their (always good) performance. For this reason,

the control group served only as a “reference” group, in order to

rule out methodological flaws (i.e., lack of reliability of the

tasks, or the scale, or the method) in the measurement

procedure.

3. Results

3.1. Evaluation of neglect-related performances

Healthy controls showed always a good performance, and a

congruent positive mean evaluation before and after task

Fig. 2 e A: Seven-point Likert scale used by participants to evaluate their performance, graduating in colour from dark red

(task impossible to perform, score ¼ 1) to dark green (flawless performance, score ¼ 7). Before the experimental run, two

practice tasks were given: during practice, participants were asked how they would evaluate losing or winning a large sum

of money, by using two versions of the Likert scale, randomized between questions, namely a 7-points scale with coloured

circles (shown in the figure), and a 7-points scale with coloured squares. B: Experimental schedule. In each neglect task,

when the examiner posed the first (PRE) question, the test material was presented to each participant, in a central position,

to facilitate the understanding of the type of task. During the PRE-condition for the cognitive non-spatial linguistic task, one

verbal practice example about words starting with a specific letter (i.e., a letter different from the ones of the experimental

task) was given. Evaluation of performance in the POST-condition was always required immediately after task execution.

RBD: right-brain-damaged patients; C ¼ controls.

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2 173

execution (PRE-condition mean scores: Line bisection ¼ 5.9,

SD ¼ ±1.2; Star cancellation ¼ 6.5, SD ¼ ±.6; Complex drawing

by copy ¼ 6.2, SD ¼ ±1.1; Clock drawing from memory ¼ 6.5,

SD ¼ ±.6; Sentence reading ¼ 6.6, SD ¼ ±.6; Letter

cancellation¼ 6.7, SD¼ ±.5. POST-conditionmean scores: Line

bisection ¼ 6.1, SD ¼ ±.8; Star cancellation ¼ 6.9, SD ¼ ±.4;Complex drawing by copy ¼ 5.9, SD ¼ ±.9; Clock drawing from

memory ¼ 6.4, SD ¼ ±.9; Sentence reading ¼ 6.7, SD ¼ ±.4;Letter cancellation ¼ 6.8, SD ¼ ±.4). These findings indicate

that the evaluation method was reliable, under conditions of

no brain damage.

In all tasks assessing spatial neglect, the evaluation scores

of right-brain-damaged patients (N ¼ 29) were submitted to

linear regression analyses, with task performance as the in-

dependent variable and evaluation score (in PRE- or POST-

condition) as the dependent variable. Results showed that

the correspondence between performance and evaluation

varied across tasks.

For the cancellation tasks, the correspondence between

performance and evaluation was weak before performing the

task, but it became significant after the taskwas performed, as

shown in Fig. 3. In particular, for Star cancellation (Nþ pa-

tients with defective performance: N ¼ 10), the linear

regression established that the omission percent scores were

weakly associated with the patients' evaluation in the PRE-

condition (beta ¼ �.337, F1,27 ¼ 3.45, p ¼ .074), but signifi-

cantly associated with their evaluation in the POST-condition

(beta ¼ �.445, F1,27 ¼ 6.67, p ¼ .016). For the Letter cancellation

test (Nþ patients with defective performance: N ¼ 12), no

significant association was found between the two variables

in the PRE-condition (beta ¼ �.088, F1,27 ¼ .21, p ¼ .650), while

the rate of omissions in this task predicted the evaluation

scores in the POST-condition (beta ¼ �.486, F1,27 ¼ 8.37,

p ¼ .007).

As shown in Fig. 4, performance in the Line bisection task

(Nþ patients with defective performance: N ¼ 10) was unre-

lated to the evaluation in both the PRE- (beta ¼ �.228,

F1,27 ¼ 1.49, p ¼ .233), and the POST- (beta ¼ .019, F1,27 ¼ .10,

p ¼ .923) conditions; similarly, performance at Clock drawing

frommemory (Nþ patients with defective performance:N¼ 6)

did not predict the evaluation scores in both the PRE-

(beta¼ �.037, F1,27 ¼ .04, p¼ .851) and the POST- (beta ¼�.264,

F1,27 ¼ 2.02, p ¼ .167) conditions.

For Complex drawing by copy (Nþ patients with defective

performance: N ¼ 10), we estimated the relationship between

performance and evaluation while keeping constant the

Fig. 3 e Scatterplots and linear regression lines of the relationship of performance in cancellation tasks [abscissa: omission

errors transformed in T scores to ease comparisons between plots (mean ¼ 50, SD ¼ ±10); higher T-scores correspond to

worse performance] with evaluation scores (ordinate: 1e7 Likert Scale) before (PRE), and after (POST) task execution in right-

brain-damaged patients with/without (Nþ/N) left spatial neglect.

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2174

possible intervening effects of CA. This was accomplished by

adding CAp scores (see section 2.4) in the model. As shown in

Fig. 5, neglect performance scores were weakly related to

evaluation in the PRE-condition (beta: e.415, F2,26 ¼ 3.83,

p¼ .074), becoming however related to the patients' evaluationin the POST-condition (beta ¼ �.528, F2,26 ¼ 5.45, p ¼ .020).

Neglect reading errors (Fig. 5) appeared significantly associ-

ated with evaluation scores in both the PRE- (beta ¼ �.483,

F1,27 ¼ 8.23, p ¼ .008), and the POST- (beta ¼ �.435, F1,27 ¼ 6.29,

p ¼ .018) conditions (Nþ patients with defective performance:

N ¼ 13). Table 2 shows a summary of the linear regression

results.

These findings were further corroborated by non-

parametric Spearman correlation analyses, showing signifi-

cant negative correlations between performance (i.e., neglect

error scores) and evaluation for the same tasks in which the

linear regression analyses had revealed statistically significant

relationships [Star cancellation: POST-condition (r ¼ �.368;

p¼ .050); Sentence reading: PRE-condition (r¼�.526; p¼ .003),

POST-condition (r¼�.393; p¼ .035); Letter cancellation: POST-

condition (r ¼ �.552; p ¼ .002)]. As for Complex drawing by

copy, both the correlations in the PRE- (r¼�.561; p¼ .002), and

the POST- (r ¼ �.577; p ¼ .001) conditions were significant, but

in this analysis the contribution of constructional disorders

was not taken into account. No other correlation proved to be

significant [Line bisection: PRE-condition (r ¼ �.182; p ¼ .344),

POST-condition (r ¼ �.112; p ¼ .564). Clock drawing from

memory: PRE-condition (r ¼ .032; p ¼ .869), POST-condition

(r ¼ �.292; p ¼ .124)].

To sum up, in a number of clinical tasks assessing left

spatial neglect, the evaluation that right-brain-damaged pa-

tients attributed to their performance was associated with the

degree of their neglect-related impairment, namely: the

higher was the number of spatial neglect errors, the lowerwas

the evaluation score. This relationship was significant in the

PRE-condition for the Sentence reading, with a tendency to-

wards significance for Star cancellation and Complex drawing

by copy. In the POST-condition, evaluation scoreswere related

to performance level in Star and Letter cancellation, Complex

drawing by copy, as well as in Sentence reading tasks. In the

case of complex visuo-motor (cancellation and drawing) tests,

the results in the POST-condition demonstrated that task

execution was able to update and improve the level of

awareness about neglect-related symptoms in these tasks. By

contrast, the evaluation of performance in Line bisection and

Clock drawing from memory tasks was never related to the

spatial performance level. As shown by the regression lines of

Line bisection (see Fig. 4), the lack of association between

Fig. 4 e Performance in line bisection (abscissa: deviation scores from the objective mid-point transformed in T scores;

higher T-scores correspond to worse performance) and clock drawing from memory (abscissa: neglect-like errors

transformed in T scores; higher T-scores correspond to worse performance) tasks and evaluation (ordinate: 1e7 Likert

Scale). For scatterplots, linear regression lines, and T scores, see Fig. 3

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2 175

evaluation and performance is due, on average, to over-

estimation of the actual performance: in fact, for “good” per-

formance, the evaluation of patients was, on average, “good”;

however, as performance decreased and became “patholog-

ical”, evaluation scores, on average, did not substantially

decrease.

3.2. Evaluation of performance in a non-spatialcognitive linguistic task: phonemic verbal fluency

Healthy controls showed normal performance and corre-

sponding positive evaluation (PRE-condition mean score ¼ 6,

SD ¼ ±.6; POST-condition mean score ¼ 5.5, SD ¼ ±.9) in the

phonemic verbal fluency task.

Eleven out of 12 N� participants and 7 out of 17 Nþ patients

scored within the normal range in the verbal fluency test,

while the remaining 11 patients (one N� and 10 Nþ) had a

defective performance (i.e., equivalent score ¼ 0) in this task.

For right-brain-damaged patients, linear regression analyses

with task performance (raw scores, i.e.: total number of cor-

rect words produced) as the independent variable and evalu-

ation score (in PRE- or POST-condition) as the dependent

variable were conducted. The patients' performance in the

verbal fluency task significantly predicted evaluation scores in

both the PRE- (beta¼ .597, F1,27 ¼ 14.96, p¼ .001), and the POST

(beta ¼ .504, F1,27 ¼ 9.20, p ¼ .005) conditions. Therefore, the

more words were produced by patients, the higher were their

evaluation scores. Fig. 6 shows the scatterplot of evaluation

and performance scores, and Table 2 a summary of the linear

regression results.

Non-parametric Spearman correlation analyses confirmed

a strong positive correlation between raw performance scores

and evaluation ratings [PRE-condition (r ¼ .634; p < .001);

POST-condition (r ¼ .547; p ¼ .002)].

To sum up, in a non-spatial linguistic task a positive as-

sociation between performance and evaluation was found in

right-brain-damaged patients. This pattern is clearly visible in

the regression lines of Fig. 6: on average, “good” performance

was associated with “good” evaluation, “pathological” per-

formance with a decrease in evaluation scores. These results

confirm that the capacity to evaluate their performance in a

cognitive task, not requiring spatial abilities, was not

compromised in neglect patients.

3.3. Evaluation of upper-limb motor performances

In the control group, the performance score was constant for

DM and unimanual (UNIM) actions, with all controls obtaining

Fig. 5 e Performance in reading (abscissa: neglect-type errors transformed in T scores; higher T-scores correspond to worse

performance) and complex drawing by copy (abscissa: neglect omission errors transformed in T scores; higher T-scores

correspond to worse performance) tasks and evaluation (ordinate: 1e7 Likert Scale). For scatterplots, linear regression lines,

and T scores, see Fig. 3.

Fig. 6 e Performance in phonemic verbal fluency (abscissa: number of produced words transformed in T scores; higher

T-scores correspond to better performance) and evaluation (ordinate: 1e7 Likert Scale). Dþ/D¡: patients with defective

(equivalent score ¼ 0)/preserved performance. For scatterplots, linear regression lines, and T scores, see Fig. 3.

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2176

Fig. 7 e Performance in the motor direct movement and bimanual tasks (abscissa: error scores transformed in T scores;

higher T-scores correspond to worse performance) and evaluation (ordinate: 1e7 Likert Scale) before (PRE) and after (POST)

task execution in right-brain-damaged patients, with and without (Nþ and N¡) left unilateral spatial neglect. For

scatterplots, linear regression lines, and T scores, see Fig. 3.

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2 177

the maximum score of 0 (perfect execution). The bimanual

tasks (BIM)werewell performed (scores between 0 and 1); only

one control participant (C27) scored 2 in one BIM; this item

was not considered. Controls attributed themselves coherent

positive evaluation scores (PRE-condition mean scores:

DM ¼ 7, SD ¼ ±.1; UNIM ¼ 7, SD ¼ ±.1; BIM ¼ 6.7, SD ¼ ±.3.POST-condition mean scores: DM ¼ 7, SD ¼ ±.1; UNIM ¼ 7,

SD ¼ ±.1; BIM ¼ 6.8, SD ¼ ±.3).A series of linear regression analyses with motor perfor-

mance as the independent variable (score range: 0e3) and

evaluation score (in PRE- or POST-condition) as the dependent

variable were conducted in 28 out of the 29 right-brain-

damaged patients (one patient was not available for the

evaluation of motor functions). In patients, performance in

unimanual tasks was constant and flawless (all scores ¼ 0,

perfect execution): therefore, no further analyses were con-

ducted. Linear regression analyses showed that performance

in motor tasks, when a DM or BIM action was required, pre-

dicted the evaluation scores both in the PRE-condition [DM:

(beta ¼ �.878, F1,26 ¼ 87.35, p < .001); BIM: (beta ¼ �.726,

F1,26 ¼ 29.02, p < .001)] and in the POST-condition [DM:

(beta ¼ �.653, F1,26 ¼ 19.31, p < .001); BIM: (beta ¼ �.733,

F1,26 ¼ 30.27, p < .001)]. Therefore, the higher was the perfor-

mance score (i.e., poor motor performance, see section 2.4),

the lower was the patients' evaluation, both before and after

task execution. Fig. 7 shows the scatterplot of the patients'evaluation and performance scores (DM and BIM), and Table 2

a summary of the linear regression results.

Non-parametric Spearman correlations confirmed a nega-

tive correlation between performance and evaluation scores

[DM: PRE-condition (r ¼ �.910; p < .001), POST-condition (r

¼ �.754; p < .001); BIM: PRE-condition (r ¼ �.617; p < .001),

POST-condition (r ¼ �.738; p < .001)].

Overall, right-brain-damaged patients, as a group, were

aware about their motor impairments, attributing them-

selves, on average, low scores in the medium/negative part of

the Likert scale for the motor actions that they were unable to

perform (i.e., higher performance error scores). Conversely, in

the evaluation of cognitive performance (see sections 3.1 and

3.2) patients scored, on average, in the green upper part of the

Likert scale.

Finally, we also examined the evaluation ratings of the two

right-brain-damaged patients who presented anosognosia for

hemiplegia in the baseline evaluation (P23 and P25; see Table

1). Although the two (clinical and experimental) assessments

were different, it may be of interest to illustrate qualitatively

the specific experimental response patterns of these patients.

During the baseline assessment, P25 showed a mild anosog-

nosia for hemiplegia (score ¼ 1/3), becoming aware of the

motor impairment only after a specific question about the left

Table 2 e Linear regression analyses made on theevaluation score (dependent variable) and the level ofperformance (independent variable) at spatial neglect,non-spatial (linguistic), and motor tasks. The sample ofneurological patients is considered (N ¼ 29 for spatial andlinguistic tests; N ¼ 28 for motor tests). For spatial neglecttests and motor tasks, the erroneous performance isconsidered (greater the score, worst the performance); forthe non-spatial linguistic task, the accuracy performanceis considered (greater the score, better the performance).

Beta value

PRE-evaluation POST-evaluation

Unilateral spatial neglect tests

Star cancellation �.337 �.445a

Letter Cancellation �.088 �.486a

Line bisection �.228 .019

Clock drawing from

memory

�.037 �.264

Complex drawing

by copy

�.415 �.528a

Reading �.483a �.435a

Non-spatial linguistic task

Phonemic verbal

fluency

.597a .504a

Upper-limb motor tasks

Direct movement �.878a �.653a

Unimanual n.a. n.a.

Bimanual �.726a �.733a

a Statistically significant regression; n.a. ¼ not assessed as per-

formancewas constant (all scores¼ 0, perfect execution). The same

results were found excluding the two chronic patients (one N� and

one Nþ) from the patients' group.

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2178

upper limbs (Bisiach, Vallar, et al., 1986); P23 presented with a

severe anosognosia for hemiplegia (score: 3/3), as she was not

able to acknowledge the motor deficit also after the neuro-

logical demonstration by the examiner. In the experimental

assessment, P25 evaluated negatively her performance in DM

of the left arm (score PRE-condition: 2; score POST-condition:

3), and BIM actions (mean score PRE-condition: 2.6; mean

score POST-condition: 2.4), coherently with her failure in the

execution of themotor tasks.With respect to P23, when asked

to raise the left arm (DM), which was an action for her

impossible to perform, P23 evaluated this taskwith amedium/

positive score in the PRE-condition (score: 4), but she was

aware about her defective performance immediately after

task execution (score POST-condition: 1); however, she

demonstrated poor awareness about task performance in BIM

actions (mean score PRE-condition: 5.2; mean score POST-

condition: 4.4), even if she was completely unable to

perform the required actions.

4. Discussion

The main findings of this study, investigating quantitatively

anosognosia for left spatial neglect (Berti et al., 1996; Jehkonen

et al., 2000; Vossel et al., 2013), may be summarized as follows:

1) for some, but not all, tasks assessing spatial neglect, the

patients' evaluation of performance is related to the level of

spatial impairment in that specific task, suggesting that some

tasks may elicit more awareness of the pathological spatial

performance; 2) task execution improves the accuracy of the

patients' evaluation of performance; 3) the patients' ability to

correctly evaluate performance in other domains (linguistic

and motor) is preserved.

All neglect patients exhibited “anosognosia” for spatial

neglect at the clinical level, namely: when directly inquired,

they stated to have completed and accurately performed each

task assessing the presence of spatial neglect. Accordingly, we

did not compare patients based on the clinical evidence of

anosognosia, asking instead for a quantitative and more

sensitive task-specific evaluation. Based on this approach,

right-brain-damaged patients present with relatively pre-

served ability in evaluating their level of performance in

Cancellation, Sentence reading, and Complex drawing by copy

tasks. Conversely, they are impaired in correctly evaluating

their level of performance in Line bisection, a perceptual

visuo-motor task, and in Clock drawing from memory, a task

assessing representational neglect (Beschin, Cocchini, Della

Sala, & Logie, 1997; Lepore, Conson, Ferrigno, Grossi, &

Trojano, 2004), showing anosognosia for neglect-related

symptoms at these tasks. Together with the absence of

monitoring deficits in linguistic (non-spatial) andmotor tasks,

the presence of anosognosia for some, but not all, neglect-

related deficits suggests a task-specific account of anosog-

nosia for spatial neglect. This result disproves the initial hy-

pothesis of a pervasive presence of anosognosia for all

manifestations of neglect. By adopting, for the first time, a

quantitative and complete assessment of neglect perfor-

mance at different tasks, which involve different spatial

abilities, and by analysing the relationship between the pa-

tients' evaluation and their actual performance at each task,

we demonstrate that not all clinical tests are able to elicit the

same degree of awareness of neglect performance. The

finding of significant regression coefficients in some tasks, but

not in others, suggests that the different tasks assessing

spatial neglect used in this study evoke different level of

awareness, although they were not directly compared in this

respect. Hence, in line with the widely accepted view that

neglect is amulti-componential syndrome (Vallar& Bolognini,

2014), also anosognosia for neglect appears to be modular in

nature, and, therefore, dissociable across tasks assessing

different aspects of the syndrome.

However, at least for line bisection, another interpretation

of the defective evaluation of performance should be consid-

ered, related to possible differences in task difficulty (Vallar,

2000). In this view, the bisection task might be “easier”,

hence requiring less engagement of cognitive resources, such

as executive function and spatial working memory (Husain

et al., 2001; Malhotra et al., 2005), as compared with cancel-

lation and copy drawing tasks. However, it is noteworthy that

left spatial neglect may selectively affect the patients' per-formance in cancellation and bisection tasks: right-brain-

damaged patients may indeed show left neglect in cancella-

tion, but not in bisection, tasks, and vice-versa (Ferber &

Karnath, 2001; Halligan & Marshall, 1992; Marshall &

Halligan, 1995; Vallar & Bolognini, 2014), conjuring up a dou-

ble dissociation of deficits (Teuber, 1955; Vallar, 2000).

Furthermore, these two tasks have different neural un-

derpinnings (Baier, Mueller, Fechir, & Dieterich, 2010; Daini,

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2 179

Angelelli, Antonucci, Cappa, & Vallar, 2002; Revill, Karnath, &

Rorden, 2011; Rorden, Fruhmann Berger, & Karnath, 2006;

Verdon, Schwartz, Lovblad, Hauert, & Vuilleumier, 2010). As

for spatial working and short-term memory there is evidence

that some patients with left neglect do not show such deficits

(Ronchi et al., 2009) and, vice-versa, i.e., right-brain-damaged

patients with defective spatial short-term memory do not

exhibit left neglect (De Renzi&Nichelli, 1975). The existence of

both behavioural and neural double dissociations argues

against an explanation based on a mere difference in task

difficulty of cancellation versus bisection tasks, inducing

different levels of awareness of the spatial abilities. Rather, it

points to a task-specific account of anosognosia for left spatial

neglect. Furthermore, any other difference in the patients'evaluation that can be attributed to a “subjective” perception

of task difficulty, independently from the presence of spatial

neglect, has been taken into account and removed by the in-

clusion of patients without neglect symptoms in the regres-

sion analyses.

With respect to the tests eliciting a greater degree of

awareness of the spatial deficit, Sentence reading, Complex

drawing by copy and Cancellation tasks trigger patients'awareness. As for reading, patients show a relative accurate

prediction of their performance before the task, and their

evaluation is accurate after task execution. Previous obser-

vations, based on a dichotomous classificatory approach,

indicate that some patients are unaware of neglect dyslexia

(Berti et al., 1996). By contrast, in our study, in which a

direct comparison between task's performance and evalua-

tion is done, overall, right-brain-damaged patients appear

to be able to monitor their neglect-related reading symp-

toms. Nevertheless, inspection of Fig. 5 shows that three

patients evaluate their performance in the POST-condition

with the maximum evaluation score (i.e., score ¼ 7, sub-

jective flawless performance), in spite of a very severe

neglect dyslexia (i.e., number of neglect reading errors

greater than 10 out of 12), therefore showing anosognosia

for the neglect-related reading deficit. Thus, although ano-

sognosia for neglect dyslexia can be observed in the indi-

vidual patient, we find an average trend of accuracy in

monitoring the neglect-related performance. It may be also

noticed that reading involves stimuli made up by multiple

components, with variable contributions from phonological,

lexical, and semantic systems [i.e., sequences of letters, that

may constitute a word, meaningless or not, though pro-

nounceable (a nonword); see Vallar et al., 2010]. Seen in this

perspective, written letter strings share some features with

the multiple-component stimuli used in target cancellation

and drawing tasks; reading has also a motor (eye move-

ment) component.

Neglect performance in tasks requiring complex visuo-

motor abilities and serial exploration, such as copy of a

drawing and target cancellation tests, are better monitored by

patients and their evaluation of performance is relatively

coherent with the actual task execution. For the Complex

drawing by copy, moreover, the influence of neglect errors on

the patients' evaluation survives alsowhen the contribution of

CA (Gainotti & Tiacci, 1970; Gainotti, 1985; Hier, Mondlock, &

Caplan, 1983a, 1983b; Kleist, 1934; Russell et al., 2010) is par-

tialled out.

All together, the present findings indicate that anosognosia

for neglect-related behavioursmay be selective, affecting only

specificmanifestations of unilateral spatial neglect. As neglect

performance may be dissociated in line bisection and

cancellation tests (Halligan & Marshall, 1992; Marshall &

Halligan, 1995), in the same way the patients' awareness

about performance in these two tasks can dissociate. This

specific “neglect of neglect” (a term which we would suggest

here) is in line with the view that the monitoring systems of

spatial functions can be highly multi-componential. Further-

more, within left spatial neglect, complex visuo-motor

(cancellation and copy drawing) and sentence reading tasks

appear to be able to elicit the greater level of awareness of

spatial deficit.

In agreement with our initial hypothesis, we found that the

evaluation of patients about their level of performance in

spatial tasks is adjustable by task execution. This conclusion

is suggested by different statistical results found when per-

formance levels were comparedwith evaluationsmade before

(PRE-) or after (POST-) task execution. This did not occur sys-

tematically, but selectively for a subset of the tasks consid-

ered, namely: target cancellation, and copy of a drawing.

Conversely, task execution is unable to update (and improve)

the patients' evaluation of their performance at the Line

bisection and Clock drawing from memory tasks. Even if

cancellation, drawing and bisection are all visuo-motor tests,

the visuo-motor feedback about the pathological neglect-

related performance seems more “salient” in target cancella-

tion and in drawing by copy, than that in Line bisection.

Moreover, cancellation and drawing tasks require a serial

exploration and continuous on-line inspection of the perfor-

mance, which is absent in line bisection. This evidence could

account for the difference in the role of task execution in

improving the level of awareness across different tests. The

possibility may be entertained that the patients' evaluation of

performance prior to task execution could have been influ-

enced by a generic knowledge about their abilities, stored in

memory representations in the pre-morbid condition.

Conversely, failure to perform the task may elicit an update of

their actual abilities, leading to a more accurate judgment

after execution (for related evidence about anosognosia for

hemiplegia see: Berti et al., 1996; Cocchini, Beschin, & Della

Sala, 2002; Marcel et al., 2004). Notably, patients had already

performed all experimental visuo-spatial tests in the baseline

assessment, but evaluation of performance was not required.

In line with these conclusions, patients with probable Alz-

heimer disease may show a pre-test over-estimation of visuo-

spatial abilities, which is reduced in the post-testing stage, at

variance with their evaluation of memory, for which the

pattern shown by patients is opposite, with anosognosia

emerging in the post-test (Barrett et al., 2005). There is also

evidence, however, that patients with probable Alzheimer

disease in an early stage are more aware of their memory

competences after exposure to a memory task (Ansell &

Bucks, 2006).

The pattern of results in the cognitive spatial versus non-

spatial tasks' evaluation further suggests that anosognosia

for left spatial neglect, when present, is a specific deficit in

spatial performance monitoring. In fact, right-brain-damaged

patients correctly estimate their competences in the control

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2180

(not assessing visuo-spatial abilities) verbal fluency test:

neglect patients with a pathological verbal fluency score

evaluate worst their performance, both before and after task

execution.

Finally, right-brain-damaged patients are aware of their

motor abilities, regardless of the presence of left spatial

neglect: when they are unable to perform the motor tasks,

their evaluations are strongly negative, in line with their

actual level of performance. As a group, patients show

awareness of motor deficits in both direct questions and

BIM. Patients' awareness of their motor inabilities may be

triggered by repeated failure on motor tasks, particularly

during physical rehabilitation. There is also evidence, how-

ever, that awareness for motor deficits may vary in relation

to the specific evaluation test (Cocchini, Beschin,

Fotopoulou, & Della Sala, 2010; Marcel et al., 2004; Nimmo-

Smith, Marcel, & Tegner, 2005; Orfei et al., 2007; Starkstein,

Jorge, & Robinson, 2010). A perusal of our patients' sample

indicates that in the baseline assessment one patient (P23)

presented with a severe anosognosia for hemiplegia, as

evaluated with a standard clinical score (Bisiach, Vallar,

et al., 1986). Notwithstanding the tasks administered dur-

ing the experimental protocol are different, possibly

prompting different results, P23's evaluation scores in DM

and BIM are indicative of unawareness of motor deficits,

with the evaluation scores in the medium/positive part of

the Likert scale. The only negative score this patient attrib-

uted to herself, coherent with the failure in raising the left

arm, was just after task execution, with motor performance

triggering a short-term modulation of awareness of the

motor defect. In the same patient, the negative feedback

provided by failed motor performance was not effective for

BIM, but only for DM request, indicating differences is

sensitivity between the two tasks for eliciting awareness of

hemiplegia (Marcel et al., 2004).

Indeed, as first found in a seminal group study (Bisiach,

Vallar, et al., 1986), unilateral spatial neglect and anosog-

nosia for left motor deficits following right-hemisphere

damage may occur independently of each other (Appelros,

Karlsson, & Hennerdal, 2007; Berti et al., 2005; Bottini,

Bisiach, Sterzi, & Vallar, 2002; Kortte & Hillis, 2009; Rode

et al., 1992; Rode, Perenin, Honore, & Boisson, 1998;

Starkstein, Fedoroff, Price, & Robinson, 1993). Our findings

provide evidence that also anosognosia for left spatial neglect

and anosognosia for left hemiplegia are independent deficits

(see also Jehkonen et al., 2000).

To summarize, the existence of anosognosia for neglect-

related performance confined to specific tasks supports the

view that the monitoring systems for spatial competences are

discrete, and may be selectively damaged, in line with previ-

ous evidence across different domains (Barrett et al., 2005;

Leicht, Berwig & Gertz, 2010; Marcel et al., 2004; Nielsen,

1938; Spinazzola, Pia, Folegatti, Marchetti, & Berti, 2008; Von

Hagen & Ives, 1937). When anosognosia for neglect symp-

toms is detected, the patients' evaluation of their spatial per-

formance cannot be traced back to a non-specific effect of

brain damage (as shown by the accurate evaluation of right-

brain-damaged patients without neglect), or to a general

compromised capacity to evaluate a cognitive performance

(as shown by the patients' evaluation of performance in the

verbal fluency test). Finally, the features of the tasks and

stimuli are able to elicit and modulate the level of awareness

of the defective performance: neglect patients are indeed

more aware of their neglect-related symptoms in the sentence

reading test, independently from the feedback provided by the

failure to execute the task. On the other hand, the feedback

about a pathological spatial performance elicited by task

execution is able to induce a more accurate awareness of the

spatial deficits, but only in cancellation and drawing by copy

tasks. These conclusions are based on statistically significant

versus non-significant correspondences between evaluation

and performance in these tests, rather than on the statistical

comparison between pre-versus post-evaluation ratings of

performance, and among tests. To summarise, awareness of

neglect-related behaviours occurs mainly in reading and

visuo-motor spatial tests, involving multiple-component

stimuli (e.g., multiple targets, letter strings, drawings), and

serial exploration.

Acknowledgements

The work was supported in part by the ‘Fondo di Ateneo’

grant, from the University of Milano-Bicocca to G.V. and N.B.,

and the Ricerca Corrente from the IRCCS Istituto Auxologico

Italiano, Milano.

r e f e r e n c e s

Akai, T., Hanyu, H., Sakurai, H., Sato, T., & Iwamoto, T. (2009).Longitudinal patterns of unawareness of memory deficits inmild Alzheimer's disease. Geriatrics and GerontologyInternational, 9(1), 16e20.

Ansell, E. L., & Bucks, R. S. (2006). Mnemonic anosognosia inAlzheimer's disease: a test of Agnew and Morris (1998).Neuropsychologia, 44(7), 1095e1102.

Appelros, P., Karlsson, G. M., & Hennerdal, S. (2007). Anosognosiaversus unilateral neglect. Coexistence and their relations toage, stroke severity, lesion site and cognition. European Journalof Neurology, 14(1), 54e59.

Babinski, J. (1914). Contribution �a l’�etude des troubles mentauxdans l’h�emipl�egie organique cerebrale (anosognosie). RevueNeurologique, 27, 845e848.

Baier, B., Mueller, N., Fechir, M., & Dieterich, M. (2010). Linebisection error and its anatomic correlate. Stroke, 41(7),1561e1563.

Barrett, A., Eslinger, P. J., Ballentine, N. H., & Heilman, K. M. (2005).Unawareness of cognitive deficit (cognitive anosognosia) inprobable ad and control subjects. Neurology, 64(4), 693e699.

Berti, A., Bottini, G., Gandola, M., Pia, L., Smania, N., Stracciari, A.,et al. (2005). Shared cortical anatomy for motor awareness andmotor control. Science, 309(5733), 488e491.

Berti, A., L�adavas, E., & Della Corte, M. (1996). Anosognosia forhemiplegia, neglect dyslexia, and drawing neglect: clinicalfindings and theoretical considerations. Journal of theInternational Neuropsychological Society, 2(5), 426e440.

Beschin, N., Cocchini, G., Della Sala, S., & Logie, R. H. (1997). Whatthe eyes perceive, the brain ignores: a case of pure unilateralrepresentational neglect. Cortex, 33(1), 3e26.

Bisiach, E., Bulgarelli, C., Sterzi, R., & Vallar, G. (1983). Linebisection and cognitive plasticity of unilateral neglect ofspace. Brain and Cognition, 2(1), 32e38.

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2 181

Bisiach, E., & Geminiani, G. (1991). Anosognosia related tohemiplegia and hemianopia. In G. P. Prigatano, &D. L. Schacter (Eds.), Awareness of deficit after brain injury (pp.17e39). New York: Oxford University Press.

Bisiach, E., Perani, D., Vallar, G., & Berti, A. (1986a). Unilateralneglect: personal and extrapersonal. Neuropsychologia, 24(6),759e767.

Bisiach, E., Vallar, G., Perani, D., Papagno, C., & Berti, A. (1986b).Unawareness of disease following lesions of the righthemisphere: anosognosia for hemiplegia and anosognosia forhemianopia. Neuropsychologia, 24(4), 471e482.

Bottini, G., Bisiach, E., Sterzi, R., & Vallar, G. (2002). Feelingtouches in someone else's hand. NeuroReport, 13(2), 249e252.

Celesia, G. G., Brigell, M. G., & Vaphiades, M. S. (1997). Hemianopicanosognosia. Neurology, 49(1), 88e97.

Cocchini, G., Beschin, N., & Della Sala, S. (2002). Chronicanosognosia: a case report and theoretical account.Neuropsychologia, 40(12), 2030e2038.

Cocchini, G., Beschin, N., Fotopoulou, A., & Della Sala, S. (2010).Explicit and implicit anosognosia or upper limb motorimpairment. Neuropsychologia, 48(5), 1489e1494.

Daini, R., Angelelli, P., Antonucci, G., Cappa, S. E., & Vallar, G.(2002). Exploring the syndrome of spatial unilateral neglectthrough an illusion of length. Experimental Brain Research,144(2), 224e237.

De Renzi, E., & Nichelli, P. (1975). Verbal and non-verbal short-term memory impairment following hemispheric damage.Cortex, 11(4), 341e354.

Diller, L., &Weinberg, J. (1977). Hemi-Inattention in rehabilitation.The evolution of a rational remediation program. InE. A. Weinstein, & R. P. Friedland (Eds.), Hemi-inattention andhemisphere specialization (pp. 62e82). New York: Raven Press.

Ellis, A. W., Flude, B. M., & Young, A. W. (1987). “Neglect dyslexia”and the early visual processing of letters in words andnonwords. Cognitive Neuropsychology, 4(4), 439e464.

Ferber, S., & Karnath, H. O. (2001). How to assess spatialneglecteline bisection or cancellation tasks? Journal of Clinicaland Experimental Neuropsychology, 23(5), 599e607.

Folstein, M., Folstein, S. E., & Mchugh, P. R. (1975). “Mini-mentalstate”. A practical method for grading the cognitive state ofpatients for the clinician. Journal of Psychiatric Research, 12(3),189e198.

Fortis, P., Maravita, A., Gallucci, M., Ronchi, R., Grassi, E., Senna, I.,et al. (2010). Rehabilitating patients with left spatial neglect byprism exposure during a visuomotor activity. Neuropsychology,24(6), 681e697.

Gainotti, G. (1985). Constructional apraxia. In J. A. M. Frederiks,G. W. Bruyn, & H. L. Klawans (Eds.), Clinical neuropsychology(Vol. 1(45), pp. 491e506). Amsterdam: Elsevier SciencePublishers, B.V.

Gainotti, G., Messerli, P., & Tissot, R. (1972). Qualitative analysis ofunilateral spatial neglect in relation to laterality of cerebrallesions. Journal of Neurology Neurosurgery and Psychiatry, 35(4),545e550.

Gainotti, G., & Tiacci, C. (1970). Patterns of drawing disability inright and left hemispheric patients. Neuropsychologia, 8(3),379e384.

Gauthier, L., Dehaut, F., & Joanette, Y. (1989). The Bells Test: aquantitative and qualitative test for visual neglect.International Journal of Clinical Neuropsychology, 11(2), 49e54.

Grigoletto, F., Zappala, G., Andeerson, D. W., & Lebowitz, B. D.(1999). Norms for the mini-mental state examination in ahealthy population. Neurology, 53(2), 315e320.

Halligan, P. W., & Marshall, J. C. (1992). Left visuo-spatial neglect:a meaningless entity? Cortex, 28(4), 525e535.

Halligan, P. W., & Marshall, J. C. (1993). When two is one: a casestudy of spatial parsing in visual neglect. Perception, 22(3),309e312.

Hier, D. B., Mondlock, J., & Caplan, L. R. (1983a). Recovery ofbehavioral abnormalities after right hemisphere stroke.Neurology, 33(3), 345e350.

Hier, D. B., Mondlock, J., & Caplan, L. R. (1983b). Behavioralabnormalities after right hemisphere stroke. Neurology, 33(3),337e344.

Husain, M., Mannan, S., Hodgson, T., Wojciulik, E., Driver, J., &Kennard, C. (2001). Impaired spatial working memory acrosssaccades contributes to abnormal search in parietal neglect.Brain, 124(5), 941e952.

Jehkonen, M., Ahonen, J. P., Dastidar, P., Laippala, P., & Vilkki, J.(2000). Unawareness of deficits after right hemisphere stroke:double-dissociations of anosognosias. Acta NeurologicaScandinavica, 102(6), 378e384.

Jenkinson, P. M., Preston, C., & Ellis, S. J. (2011). Unawarenessafter stroke: a review and practical guide to understanding,assessing, and managing anosognosia for hemiplegia.Journal of Clinical and Experimental Neuropsychology, 33(10),1079e1093.

Kertesz, A. (2010). Anosognosia for aphasia. In G. P. Prigatano(Ed.), The study of anosognosia (pp. 113e122). Oxford: OxfordUniversity Press.

Kleist, K. (1934). Gehirnpathologie. Leipzig: J. A. Barth.Kortte, K., & Hillis, A. E. (2009). Recent advances in the

understanding of neglect and anosognosia following righthemisphere stroke. Current Neurology and Neuroscience Reports,9(6), 459e465.

Langer, K. G. (2009). Babinski's anosognosia for hemiplegia inearly twentieth-century French neurology. Journal of the Historyof Neurosciences, 18(4), 387e405.

Lebrun, Y. (1987). Anosognosia in aphasics. Cortex, 23(2), 251e263.Leicht, H., Berwig, M., & Gertz, H. J. (2010). Anosognosia in

Alzheimer's disease: the role of impairment levels inassessment of insight across domains. Journal of theInternational Neuropsychological Society, 16(3), 463e473.

Lepore, M., Conson, M., Ferrigno, A., Grossi, D., & Trojano, L.(2004). Spatial transpositions across tasks and responsemodalities: exploring representational allochiria. Neurocase,10(5), 386e392.

Malhotra, P., J€ager, H. R., Parton, A., Greenwood, R., Playford, E. D.,Brown, M. M., et al. (2005). Spatial working memory capacity inunilateral neglect. Brain, 128(2), 424e435.

Marcel, A. J., Tegner, R., & Nimmo-Smith, I. (2004). Anosognosiafor plegia: specificity, extension, partiality and disunity ofbodily unawareness. Cortex, 40(1), 19e40.

Marshall, J. C., & Halligan, P. W. (1995). Within- and between-taskdissociations in visuo-spatial neglect: a case study. Cortex,31(2), 367e376.

McGlynn, S. M., & Schacter, D. L. (1989). Unawareness of deficitsin neuropsychological syndromes. Journal of Clinical andExperimental Neuropsychology, 11(2), 143e205.

Nielsen, J. M. (1938). Gerstmann syndrome: finger agnosia,agraphia, confusion of right and left and acalculia. Archives ofNeurology and Psychiatry, 39(3), 536e559.

Nimmo-Smith, I., Marcel, A. J., & Tegner, R. (2005). A diagnostictest of unawareness of bilateral motor task abilities inanosognosia for hemiplegia. Journal of Neurology Neurosurgeryand Psychiatry, 76(8), 1167e1169.

Novelli, G., Papagno, C., Capitani, E., Laiacona, M., Vallar, G., &Cappa, S. F. (1986). Tre test clinici di ricerca e produzionelessicale. Taratura su soggetti normali. Archivio di PsicologiaNeurologia e Psichiatria, 47(4), 477e506.

Orfei, M. D., Robinson, R. G., Prigatano, G. P., Starkstein, S.,Rusch, N., Bria, P., et al. (2007). Anosognosia for hemiplegiaafter stroke is a multifaceted phenomenon: a systematicreview of the literature. Brain, 130(12), 3075e3090.

Papagno, C., & Vallar, G. (2003). Anosognosia for left hemiplegia:Babinski's (1914) cases. In C. Code, C.-W. Wallesch, Y. Joanette,

c o r t e x 6 1 ( 2 0 1 4 ) 1 6 7e1 8 2182

& A. R. Lecours (Eds.), Vol. 2s. Classic cases in neuropsychology(pp. 171e189). Hove, East Sussex: Psychology Press.

Philippon, J., & Poirier, J. (2009). Joseph Babinski. A biography.Oxford, England: Oxford University Press.

Pizzamiglio, L., Antonucci, G., Judica, A., Montenero, P.,Razzano, C., & Zoccolotti, P. (1992). Cognitive rehabilitation ofthe hemineglect disorder in chronic patients with unilateralbrain damage. Journal of Clinical and ExperimentalNeuropsychology, 14(6), 901e923.

Prigatano, G. P. (Ed.). (2010). The study of anosognosia. Oxford:Oxford University Press.

Prigatano, G. P., & Schacter, D. L. (Eds.). (1991). Awareness of deficitafter brain injury. Clinical and theoretical issues. Oxford: OxfordUniversity Press.

Revill, K. P., Karnath, H. O., & Rorden, C. (2011). Distinct anatomyfor visual search and bisection: a neuroimaging study.NeuroImage, 57(2), 476e481.

Rinaldi, M. C., Piras, F., & Pizzamiglio, L. (2010). Lack of awarenessfor spatial and verbal constructive apraxia. Neuropsychologia,48(6), 1574e1582.

Rode, G., Charles, N., Perenin, M. T., Vighetto, A., Trillet, M., &Aimard, G. (1992). Partial remission of hemiplegia andsomatoparaphrenia through vestibular stimulation in a caseof unilateral neglect. Cortex, 28(2), 203e208.

Rode, G., Perenin, M. T., Honore, J., & Boisson, D. (1998).Improvement of the motor deficit of neglect patients throughvestibular stimulation: evidence for a motor neglectcomponent. Cortex, 34(2), 253e261.

Ronchi, R., Posteraro, L., Fortis, F., Bricolo, E., & Vallar, G. (2009).Perseveration in left spatial neglect: drawing and cancellationtasks. Cortex, 45(3), 300e312.

Rorden, C., & Brett, M. (2000). Stereotaxic display of brain lesions.Behavioural Neurology, 12(4), 191e200.

Rorden, C., Fruhmann Berger, M., & Karnath, H. O. (2006).Disturbed line bisection is associated with posterior brainlesions. Brain Research, 1080(1), 17e25.

Russell, C., Deidda, C., Malhotra, P., Crinion, J. T., Merola, S., &Husain, M. (2010). A deficit of spatial remapping inconstructional apraxia after right-hemisphere stroke. Brain,133(4), 1239e1251.

Siegel, S., & Castellan, N. J. (1988). Nonparametric statistics for thebehavioral sciences. New York: McGraw-Hill.

Spinazzola, L., Pia, L., Folegatti, A., Marchetti, C., & Berti, A. (2008).Modular structure of awareness for sensorimotor disorders:evidence from anosognosia for hemiplegia and anosognosiafor hemianaesthesia. Neuropsychologia, 46(3), 915e926.

Spinnler, H., & Tognoni, G. (1987). Standardizzazione e taraturaitaliana di test neuropsicologici. The Italian Journal ofNeurological Sciences, 6(Suppl. 8), 1e120.

Starkstein, S. E., Fedoroff, J. P., Price, T. R., & Robinson, R. G. (1993).Denial of illness scale: a reliability and validity study.Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 6(2),93e97.

Starkstein, S. E., Jorge, R. E., & Robinson, R. G. (2010). Thefrequency, clinical correlates, and mechanism of anosognosiaafter stroke. The Canadian Journal of Psychiatry, 55(6), 355e361.

Teuber, H.-L. (1955). Physiological psychology. Annual Review ofPsychology, 6, 267e296.

Vallar, G. (1998). Spatial hemineglect in humans. Trends inCognitive Sciences, 2(3), 87e97.

Vallar, G. (2000). The methodological foundations of humanneuropsychology: studies in brain-damaged patients. InF. Boller, J. Grafman, & G. Rizzolatti (Eds.), Handbook ofneuropsychology (Vol. 1, pp. 305e344). Amsterdam, TheNetherlands: Elsevier.

Vallar, G., & Bolognini, N. (2014). Unilateral spatial neglect. InA. C. Nobre, & S. Kastner (Eds.), Oxford handbook of attention (pp.972e1027). Oxford: Oxford University Press.

Vallar, G., Burani, C., & Arduino, L. S. (2010). Neglect dyslexia: areview of the neuropsychological literature. Experimental BrainResearch, 206(2), 219e235.

Vallar, G., & Perani, D. (1986). The anatomy of unilateral neglectafter right hemisphere stroke lesions. A clinical CT/Scancorrelation study in man. Neuropsychologia, 24(5), 609e622.

Vallar, G., & Ronchi, R. (2006). Anosognosia for motor and sensorydeficits after unilateral brain damage: a review. RestorativeNeurology and Neuroscience, 24(4e6), 247e257.

Vallar, G., Rusconi, M. L., Fontana, S., & Musicco, M. (1994). Tretest di esplorazione visuo-spaziale: taratura su 212 soggettinormali. Archivio di Psicologia Neurologia e Psichiatria, 55(4),827e841.

Verdon, V., Schwartz, S., Lovblad, K. O., Hauert, C. A., &Vuilleumier, P. (2010). Neuroanatomy of hemispatial neglectand its functional components: a study using voxel-basedlesion-symptom mapping. Brain, 133(3), 880e894.

Von Hagen, K. O., & Ives, E. R. (1937). Anosognosia (Babinski),imperception of hemiplegia. Bulletin of the Los AngelesNeurological Societies, 2, 95e103.

Vossel, S., Weiss, P. H., Eschenbeck, P., Saliger, J., Karbe, H., &Fink, G. R. (2012). The neural basis of anosognosia for spatialneglect after stroke. Stroke, 43(7), 1954e1956.

Vossel, S., Weiss, P. H., Eschenbeck, P., & Fink, G. R. (2013).Anosognosia, neglect, extinction and lesion site predictimpairment of daily living after right-hemispheric stroke.Cortex, 49(7), 1782e1799.

Wilson, B., Cockburn, J., & Halligan, P. W. (1987). Behaviouralinattention test. Titchfield, Hants: Thames Valley TestCompany.