Theory of mind and epilepsy: What clinical implications?


  • Anna Rita Giovagnoli,

    Corresponding author
    1. Laboratory of Cognitive Neurology and Rehabilitation, Unit of Neurology and Neuropathology, Fondazione IRCCS Istituto Neurologico C. Besta, Milan, Italy
    • Address correspondence to Anna Rita Giovagnoli, Laboratory of Cognitive Neurology and Rehabilitation, Unit of Neurology and Neuropathology, Fondazione IRCCS Istituto Neurologico Carlo Besta, Via Celoria 11, 20133 Milano, Italy. E-mail:

    Search for more papers by this author
  • Annalisa Parente,

    1. Laboratory of Cognitive Neurology and Rehabilitation, Unit of Neurology and Neuropathology, Fondazione IRCCS Istituto Neurologico C. Besta, Milan, Italy
    Search for more papers by this author
  • Flavio Villani,

    1. Epilepsy Clinic and Experimental Neurophysiology, Fondazione IRCCS Istituto Neurologico C. Besta, Milan, Italy
    Search for more papers by this author
  • Silvana Franceschetti,

    1. Unit of Neurophysiology and Experimental Epileptology, Fondazione IRCCS Istituto Neurologico C. Besta, Milan, Italy
    Search for more papers by this author
  • Roberto Spreafico

    1. Epilepsy Clinic and Experimental Neurophysiology, Fondazione IRCCS Istituto Neurologico C. Besta, Milan, Italy
    Search for more papers by this author



Epilepsy can impair theory of mind (ToM), but the clinical significance of such a deficit is unknown. This study evaluated the influence of selective ToM deficits on self-appraisal, coping, and quality of life (QoL) in patients with focal epilepsy.


Data were collected from 66 patients with temporal or frontal lobe epilepsy, and from 42 healthy controls. The Faux Pas Task (FPT), Multiple Ability Self-report Questionnaire (MASQ), Coping Responses Inventory-Adult (CRI-Adult), and World Health Organization QoL 100 (WHOQoL 100) evaluated ToM, self-rated cognitive abilities, coping to stressful events, and QoL. Different tests and inventories assessed other cognitive functions, depression, and anxiety.

Key Findings

Patients were impaired in the recognition and comprehension of social faux pas. The FPT scores contributed to predict the MASQ, CRI-Adult, and WHOQoL overall scores; the comprehension of others' mental states and interactions score exerted a prominent influence.


In patients with focal epilepsy, selective ToM deficits may have clinical implications, with specific influence on self-appraisal, coping, and overall QoL. ToM evaluation may contribute in explaining some psychobehavioral difficulties and to plan nonpharmacological treatment.

Psychobehavioral problems frequently impair the health of patients with epilepsy (PWE; Fisher et al., 2000; Jehi et al., 2011; Mohammed et al., 2012). Depression, anxiety, obsessive-compulsive symptoms, psychosis, stigma, poor self-esteem, and coping difficulties were reported in 6–88% of PWE (Gramstad et al., 2001; Suurmeijer et al., 2001; Jones et al., 2003; Butterbaugh et al., 2005; Giovagnoli et al., 2006), showing a significant impact on quality of life (QoL; Fisher et al., 2000; Hermann et al., 2000; Tracy et al., 2007; Gois et al., 2011). Attention and memory deficits also proved to impair psychosocial adjustment (Shehata et al., 2009; Gois et al., 2011), but they did not predict QoL (Gois et al., 2011).

Theory of mind (ToM), an ability to mentally represent one's own and others' epistemic and affective mental states (Shamay-Premack & Woodroof, 1978; Baron-Cohen et al., 1994; Stone et al., 1998; Shamay-Tsoory & Aharon-Peretz, 2007), provides information for guiding one's own social conduct and understanding the behavior of others. Adults are credited with a ToM if they are able to recognize a false belief by realizing that another individual's mental states have no basis in reality or differ from representations pertaining to the self (Brass et al., 2009). Therefore, ToM allows one to distinguish real mental states from persuasion, metaphor, irony, lie, or delusion. Indeed, in neurologic or psychiatric conditions in the absence of epilepsy, ToM deficits showed a relationship to different psychopathologic symptoms (Gregory et al., 2002; Grisham et al., 2010; Koelkebeck et al., 2010; Mehl et al., 2010). In PWE, ToM deficits were strictly associated with age of seizure onset, presence of temporolimbic damage, and disease duration (Schacher et al., 2006; Giovagnoli et al., 2011), and they constituted a specific cognitive factor (Giovagnoli et al., 2011). In these patients, the clinical impact of ToM impairment was minimally considered (Farrant et al., 2005; Schacher et al., 2006; Shaw et al., 2007; Giovagnoli et al., 2011; Broicher et al., 2012).

Given the demands on clinical care by epilepsy-related psychobehavioral difficulties, it is of great interest to learn more about their causes. ToM plays a role in determining different psychopathologic aspects in nonepilepsy patients. It is possible that ToM deficits contribute to cause the psychobehavioral distress and impoverishment of QoL of PWE. In particular, an impaired capacity to understand mental states and to comprehend interpersonal relations could affect self-appraisal and adaptation capacities. This study evaluated ToM with the goal of clarifying its relation to cognitive self-evaluation, coping with stressful events, and selective QoL domains. To our knowledge, this is the first investigation assessing such aspects in patients with epilepsy.


Subjects' description

Sixty-six patients with temporal lobe epilepsy (TLE; n = 54) or frontal lobe epilepsy (FLE; n = 12) were selected on the basis of clinical and electroencephalography (EEG) findings (Berg et al., 2010). Patients with major depression, psychosis, or mental retardation were excluded on the basis of psychiatric interview, school accomplishment, and the level of education attained. Magnetic resonance imaging (MRI) detected focal lesions congruent with the location indicated by clinical and EEG diagnosis in 48 patients with TLE and 10 patients with FLE. Medial temporal lobe sclerosis (MTLS) was found in 18 patients with left and 15 patients with right TLE. Dysplasia, cavernomas, or low-grade gliomas were detected in the medial temporal region in seven TLE patients, and in the lateral temporal area in eight TLE patients. Frontal lobe lesions included two cavernomas, two dysplasia, three posttraumatic atrophic lesions, and three vascular lesions. Patients with FLE and patients with TLE did not differ in disease duration, age of seizure onset, seizure frequency, and the number of antiepileptic drugs (AEDs).

Forty-two healthy controls, selected from the hospital staff and patients' relatives, underwent the neuropsychological testing. PWE and controls showed the same chronological age, gender distribution, and schooling (Table 1).

Table 1. Characteristics of PWE and controls
 All patients N = 66TLE patients N = 54FLE patients N = 12Healthy controls N = 42
  1. PWE, patients with epilepsy; TLE, temporal lobe epilepsy; FLE, frontal lobe epilepsy; AED, antiepileptic drug.

Schooling (years)11.79 ± 3.41 (5–18)11.91 ± 3.4711.25 ± 3.2511.81 ± 3.38 (3–18)
Age37.68 ± 9.98 (18–65)37.80 ± 9.2037.17 ± 13.4140–64 ± 12.61
Disease duration17.61 ± 12.84 (1–50)18.89 ± 13.0411.83 ± 10.53 
Age of seizure onset19.91 ± 14.06 (1–55)18.70 ± 13.4825.33 ± 15.92 
Monthly seizure frequency (previous 6 months)10.25 ± 16.23 (0–90)9.33 ± 11.6914.73 ± 30.57 
Number of AEDs2.13 ± 0.92 (1–5)2.13 ± 0.882.09 ± 1.14 
Focal lesions584810 

Neuropsychological measurements

ToM was evaluated using the Faux Pas Task (FPT; Stone et al., 1998). The FPT performance, which is normally solved at age 11–12, can be significantly impaired by TLE and FLE (Giovagnoli et al., 2011). It requires the recognition or exclusion of social faux pas (FP) in 20 short stories. After reading each story, a subject is asked the detection question: “Did anyone say something they shouldn't have said?” If they answer “yes,” they are asked four comprehension questions (“Who said something they shouldn't have said?,” “Why shouldn't he/she have said what he/she did?,” “Why did he/she say that?,” “How did he/she feel?”) that investigate the identification of the character making the FP, comprehension of mental states and social interactions before judging the behavior's adequacy, comprehension of intentions, and comprehension of affective mental states. A control question is asked at the end of each story to verify that the subject has kept in mind the contextual details. Two detection (FP recognition, non-FP exclusion), four comprehension, and two control scores are computed (score ranges: 0–10).

The Digit Span, Corsi Blocks Span, Short Story, Rey Complex Figure Delayed Recall (RCFDR), Attentive Matrices, Trail Making Test (TMT) A and B, Raven Colored Progressive Matrices (RCPM), Weigl Sorting Test, and Word Fluency Test (WFT) on phonemic or semantic cues evaluated different cognitive abilities (Giovagnoli et al., 2011).

Measurements of cognitive self-evaluation, coping, mood, and QoL

The Multiple Ability Self-report Questionnaire (MASQ; Seidenberg et al., 1994) comprises 38 questions that explore the subjective perception of cognitive abilities (language, visual-perceptual abilities, verbal and visual memory, attention/concentration). Each subscale contains eight questions, except for the visual-perceptual abilities subscale, which contains six. One to five points are attributed to each answer in relation to the frequency (almost always, often, sometimes, seldom, almost never) of specific difficulties. The total score (38–190) is the sum of all subscale scores, and gives an index of everyday failures: the higher the total score, the more severe the perceived impairment.

The Coping Responses Inventory–Adult Form (CRI-Adult; Moos et al., 1990) measures eight types of coping responses to stressful life events. Eight scales measure cognitive (logical analysis, positive reappraisal, cognitive avoidance, acceptance or resignation) or behavioral coping strategies (seeking guidance and support, problem solving, seeking alternative rewards, emotional discharges), distinguishing approach and avoidance styles. In responding to the CRI-Adult, a subject refers to a recent stressful circumstance, answering on a 4-point rating scale. Two scores (0–74) for cognitive and behavioral coping and a total score (0–148) are computed. The higher the score, the more frequent the use of the approach.

The World Health Organization QoL scale (WHOQoL 100; WHO, 1996) assesses 25 facets, each including four items. The facets are grouped into seven domains including different facets: Overall QoL (one facet), Physical (pain and discomfort, energy and fatigue, sleep and rest), Psychological (positive feelings, thinking, learning, memory and concentration, self-esteem, body image and appearance, negative feelings), Level of independence (and mobility, activities of daily living, dependence on medication or treatments, work capacity), Social relationships (personal relationships, social support, sexual activity), Environment (physical safety and security, home environment, financial resources, health and social care – accessibility and quality, opportunities for acquiring new information and skills, participation in and opportunities for recreation/leisure activities, physical environment – pollution/noise/traffic/climate, transport), and Influence of personal beliefs on life (one facet). Individual items are rated on a 5-point Likert scale, where 1 indicates low, negative perceptions and 5 indicates high, positive perceptions. A total score, sum of all item scores, and seven domain scores (the means of the facet scores within the domain multiplied by four, except for the Overall QoL and Influence of beliefs on life domains) are computed. The higher the score, the better the QoL dimension.

The Beck Depression Inventory (BDI, total score 0–39; Beck et al., 1996) and State Trait Anxiety Inventory (STAI 1 and 2; total scores 20–80; Spielberger, 1989) assess depression and anxiety. The higher the scores, the worse depression and anxiety.

Data analysis

Factor analysis of the test scores was used to assess the structure of the neuropsychological battery. Analysis of variance (ANOVA) with chronological age, gender, and schooling as covariates (ANCOVA) compared PWE and controls.

Preliminary Pearson product moment coefficient and regression analyses evaluated the relationship of the FPT scores to the demographic and clinical variables. Partial correlation analyses, with schooling and age of epilepsy onset as covariates, and multiple stepwise regression analyses determined the relationships of the MASQ, CRI-Adult, and WHOQoL 100 scores to the ToM factor. Subsequent partial correlation and regression analyses explored the relationships of the three inventory scores to the FPT scores.


Cognitive factors

Factor analysis of the neuropsychological test scores obtained by PWE yielded four factors (ToM, Attention and memory, Lexical search, Set shifting) (Table 2).

Table 2. Factor analysis of the neuropsychological test scores
 Theory of mindAttention and memoryLexical searchSet shifting
  1. FPT, Faux Pas Task; FP, faux pas; TMT, Trail Making Test; RCPM, Raven Colored Progressive Matrices; RCFDR, Rey Complex Figure Delayed Recall; WFT, Word Fluency Test.

FPT FP recognition0.90   
FPT Question 10.93   
FPT Question 20.88   
FPT Question 30.76   
FPT Question 40.91   
Attentive Matrices 0.66  
TMT A −0.70  
TMT B −0.70  
Corsi Blocks Span 0.61  
RCPM 0.54  
Short Story 0.53  
RCFDR 0.64  
WFT on phonemic cue  0.89 
WFT on semantic cue  0.79 
Digit Span  0.74 
Weigl Sorting Test   0.72
FPT No-FP exclusion   0.70
Explained variance Total = 68.44%25.39%19.57%15.21%8.27%

Comparisons between PWE and controls

Tables 3 and 4 summarize the test and inventory scores. ANCOVA revealed a mild global influence for schooling on the FPT performances (Pillai's value = 0.11, F = 1.96, p = 0.08), with effects on FP recognition (F = 4.20, p = 0.04) and on the first (F = 4.91, p = 0.03), second (F = 7.43, p = 0.008), and third comprehension scores (F = 8.92, p = 0.004), whereas chronological age and gender had no influence. The analysis also revealed a significant group effect (Pillai's value = 0.26, F = 5.62, p < 0.001). PWE were significantly impaired compared with controls, with no difference between FLE and TLE patients.

Table 3. Mean FPT scores obtained by PWE and controls
 All patientsTLE patientsFLE patientsControlsPWE versus controls
  1. PWE, patients with epilepsy; FPT, Faux Pas Task; FP, faux pas; TMT, Trail Making Test; WFT, Word Fluency Test; RCPM, Raven Colored Progressive Matrices; RCFDR, Rey Complex Figure Delayed Recall.

FPT FP recognition9.41 ± 1.219.17 ± 1.539.42 ± 1.169.71 ± 0.55F = 4.25, p = 0.04
FPT non-FP exclusion9.58 ± 0.969.35 ± 1.179.83 ± 0.399.81 ± 0.67NS
Question 19.18 ± 1.448.89 ± 1.778.75 ± 1.549.67 ± 0.57F = 7.53, p = 0.007
Question 28.87 ± 1.678.37 ± 1.978.42 ± 1.739.64 ± 0.69F = 15.84, p < 0.001
Question 37.86 ± 2.396.83 ± 2.697.50 ± 1.689.29 ± 1.17F = 28.83, p < 0.001
Question 48.97 ± 1.608.54 ± 1.208.75 ± 1.369.60 ± 0.59F = 9.98, p = 0.002
Attentive Matrices54.44 ± 5.7854.52 ± 5.3450.00 ± 9.9255.60 ± 4,10NS
TMT A38.56 ± 19.2138.04 ± 21.5248.75 ± 25.4036.31 ± 12.60NS
TMT B106.39 ± 58.28109.30 ± 68.09108.33 ± 34.79102.1 ± 50.15NS
WFT on phonemic cues32.57 ± 12.4129.37 ± 12.3627.83 ± 12.1838.05 ± 10.67F = 16.32, p < 0.001
WFT on semantic cues38.39 ± 9.0536.52 ± 8.7433.83 ± 9.3642.10 ± 8.16F = 14.69, p < 0.001
RCPM31.66 ± 4.0431.31 ± 3.3830.17 ± 5.5232.52 ± 4.25NS
Weigl Sorting Test11.82 ± 2.2911.81 ± 2.2411.92 ± 2.4311.81 ± 2.36NS
Digit Span5.81 ± 1.025.72 ± 0.965.42 ± 1.166.02 ± 1.02NS
Corsi Blocks Span5.36 ± 1.035.22 ± 1.085.00 ± 1.485.64 ± 0.73F = 7.89, p = 0.006
Short Story14.13 ± 4.5013.14 ± 4.2213.13 ± 5.6715.69 ± 4.13F = 10.17, p = 0.002
RCFDR18.35 ± 6.4317.18 ± 6.7316.10 ± 7.3920.50 ± 5.18F = 12.55, p = 0.001
Table 4. Psychobehavioral inventory scores obtained by PWE
 Mean score ± SDMinimumMaximum
  1. PWE, patients with epilepsy; WHOQoL, World Health Organization Quality of Life; BDI, Beck Depression Inventory; STAI, State Trait Anxiety Inventory; CRI, Coping Responses Inventory; MASQ, Multiple Ability Self-report Questionnaire.

Quality of life   
WHOQoL 100 total354.38 ± 64.043459
WHOQoL 100 influence of personal beliefs on life3.71 ± 0.882.005.00
WHOQoL 100 overall3.27 ± 0.781.05.0
WHOQoL 100 physical14.65 ± 2.508.3318.67
WHOQoL 100 psychological13.90 ± 2.349.818.6
WHOQoL 100 Level of independence14.13 ± 2.607.7518.75
WHOQoL 100 Social14.58 ± 2.189.6719.67
WHOQoL 100 Environment14.60 ± 2.089.0019.13
BDI4.56 ± 4.18015
STAI141.36 ± 10.702068
STAI243.08 ± 10.462064
Total100.97 ± 12.2973.25124.75
Cognitive48.46 ± 6.573460
Behavioral52.51 ± 7.9131.5068.25
Cognitive self-evaluation   
MASQ74.29 ± 21.878124

Separate ANCOVAs comparing the other test scores showed significant effects for schooling (Pillai's value = 0.29, F = 3.42, p = 0.001), chronological age (Pillai's value = 0.20, F = 2.06, p = 0.03), and group (Pillai's value = 0.22, F = 3.12, p = 0.001). PWE were impaired in memory and verbal fluency compared with controls.

Relationships of ToM to demographic and epilepsy-related variables

The non-FP exclusion score (r = 0.29, p = 0.02), total comprehension score (r = 0.28, p = 0.02), and the third (r = 0.28, p = 0.02) and fourth comprehension scores (r = 0.31, p = 0.01) correlated with age of seizure onset. The non-FP exclusion score (r = −0.26, p = 0.03), total comprehension score (r = −0.30, p = 0.01), and the second (r = −0.27, p = 0.03), third (r = −0.24, p = 0.05), and fourth comprehension scores (r = −0.33, p = 0.007) also correlated with disease duration, whereas none of the FPT scores correlated with seizure frequency and AED number. The second (r = 0.32, p = 0.01) and third comprehension scores (r = 0.30, p = 0.014) correlated with schooling. Multiple stepwise regression analysis entering the demographic and epilepsy-related variables revealed that the ToM factor was predicted by age of seizure onset (R2 = 0.08, F = 4.45, p = 0.04).

Relationship between ToM and self-rated cognitive abilities

The MASQ score showed no correlation with the demographic or epilepsy-related variables. Partial correlation analyses with schooling and age of seizure onset as covariates showed a correlation between the MASQ score and the FPT (r = −0.29, p = 0.02), Attention and memory (r = −0.24, p = 0.05), and Set shifting factors (r = 0.26, p = 0.04), and the BDI (r = 0.26, p = 0.04), STAI1 (r = 0.48, p < 0.001), and STAI2 scores (r = 0.39, p = 0.001). Subsequent regression analysis revealed that the MASQ score was predicted by the STAI1 (R2 = 0.25, F = 21.19, p < 0.001) and Attention and memory (R2 = 0.32, F = 14.89, p < 0.001) and ToM factors scores (R2 = 0.37, F = 12.19, p < 0.001).

When considering the FPT scores, the MASQ score correlated with the FP recognition and the first, second, third, and fourth comprehension scores, and it was predicted by the second comprehension score (Table 5).

Table 5. Correlations between the FPT scores and the MASQ, CRI-Adult, and WHOQoL 100 scores
 Faux pas recognitionIdentification of the characterComprehension of others' mental states and interactionsComprehension of intentionsComprehension of affective states
  1. FPT, Faux Pas Task; MASQ, Multiple Ability Self-report Questionnaire; CRI, Coping Responses Inventory; WHOQoL, World Health Organization Quality of Life.

MASQr = −0.33, p = 0.009r = −0.29, p = 0.02

r = −0.36, p = 0.004

R2 = 0.13, F = 9.75 p = 0.003

r = −0.31, p = 0.01r = −0.27, p = 0.03
CRI-Adult total r = −0.33 p = 0.02

r = −0.40, p = 0.005

R2 = 0.15, F = 8.95, p = 0.004

 r = −0.34, p = 0.02
CRI-Adult cognitiveR2 = 0.20, F = 6.09, p = 0.004 

r = −0.40, p = 0.005

R2 = 0.10, F = 5.44, p = 0.024

r = −0.03, p = 0.04r = −0.34, p = 0.016
CRI-Adult behavioralr = −0.28 p = 0.05r = −0.32 p = 0.024

r = −0.30, p = 0.04

R2 = 0.14, F = 8.26, p = 0.006

WHOQoL 100 Total score R2 = 0.21, F = 8.80, p < 0.001

r = 0.30, p = 0.016

R2 = 0.07, F = 4.65, p = 0.03

r = 0.25, p = 0.04 
WHOQoL 100 Overall  

r = 0.29, p = 0.02

R2 = 0.06, F = 4.37, p = 0.04

WHOQoL 100 Physical  r = 0.28, p = 0.02

r = 0.32, p = 0.01

R2 = 0.09, F = 6.37, p = 0.01

WHOQoL 100 Psychological  r = 0.30, p = 0.016r = 0.34, p = 0.005 R2 = 0.07, F = 4.68, p = 0.03 
WHOQoL 100 Level of independence R2 = 0.20, F = 7.93, p = 0.001

r = 0.34, p = 0.006

R2 = 0.09, F = 6.33, p = 0.01

r = 0.31 p = 0.01 
WHOQoL 100 Social   r = 0.21, p = 0.09 
WHOQoL 100 Environmentr = 0.26, p = 0.04r = 0.26, p = 0.04r = 0.31 p = 0.01

r = 0.26, p = 0.04

R2 = 0.06, F = 14.08, p = 0.05


Relationships between ToM and coping

Fifty-one of 66 PWE completed the CRI-Adult inventory. The CRI-Adult cognitive (r = −0.36, p = 0.01), behavioral (r = −0.37, p = 0.01), and total scores (r = −0.42, p = 0.003) correlated with the ToM factor. The CRI-Adult cognitive score also correlated with the Lexical search factor (r = −0.35, p = 0.01) and BDI score (r = 0.32, p = 0.02), whereas the CRI-Adult behavioral score correlated with chronological age (r = −0.34, p = 0.016). None of the CRI-Adult scores correlated with demographic or epilepsy-related variables. Two separate regression analyses entering the cognitive factors, BDI, and STAI scores revealed that the CRI-Adult cognitive score was predicted by the Lexical search factor (R2 = 0.16, F = 9.12, p = 0.004) and BDI scores (R2 = 0.24, F = 7.57, p = 0.001), whereas the CRI-Adult total score was predicted by the ToM factor score (R2 = 0.16, F = 9.18, p = 0.004). Regression analysis of the CRI-Adult behavioral score entering the cognitive factors, BDI and STAI scores, and chronological age revealed an influence for the ToM factor (R2 = 0.15, F = 9.03, p = 0.004) and chronological age (R2 = 0.28, F = 9.09, p < 0.001).

When considering the FPT scores, the CRI-Adult cognitive score was predicted by the second comprehension and FP recognition scores, the behavioral score by the second comprehension score, and the total score by the second comprehension score (Table 5).

Relationships between ToM and QoL

The WHOQoL 100 Overall (r = 0.29, p = 0.02), Level of independence (r = 0.25, p = 0.04), and Environment domain scores (r = 0.29, p = 0.02) correlated with the ToM factor score, whereas none of the QoL domain scores correlated with the Attention and memory, Lexical search, or Set shifting factor scores. None of the WHOQoL 100 scores correlated with the demographic or epilepsy-related variables, whereas they significantly correlated with the BDI (total: r = −0.031, p = 0.01; Overall: r = −0.40, p = 0.001; Physical: r = −0.27, p = 0.03; Psychological: r = −0.48, p < 0.001; Level of independence: r = −0.50, p < 0.001; Social: r = −0.24, p = 0.06; Influence of personal beliefs on life: r = −0.27, p = 0.031; Environment: r = −0.27, p = 0.034), STAI1 (total: r = −0.48, p < 0.001; Overall: r = −0.40, p = 0.001; Physical: r = −0.43, p < 0.001; Psychological: r = −0.60, p < 0.001; Level of independence: r = −0.50, p < 0.001; Social: r = −0.35, p = 0.05; Influence of personal beliefs on life: r = −0.51, p < 0.001; Environment: r = −0.47, p < 0.001;), and STAI2 scores (score: r = −0.60, p < 0.001; Overall: r = −0.60, p < 0.001; Physical: r = −0.50, p < 0.001; Psychological: r = −0.72, p < 0.001; Level of independence: r = −0.54, p < 0.001; Social: r = −0.47, p < 0.001; Influence of personal beliefs on life: r = −0.53, p < 0.001; Environment: r = −0.50, p < 0.001).

Subsequent regression analyses entering the four cognitive factors showed that only the ToM factor predicted the Overall domain score (R2 = 0.08, F = 5.20, p = 0.03). Regression analyses adding the BDI and STAI scores revealed that the WHOQoL 100 total score (R2 = 0.38, F = 39.07, p < 0.001) and the Overall (R2 = 0.33, F = 33.03, p < 0.001), Level of independence (R2 = 0.30, F = 27.01, p < 0.001), Influence of personal beliefs on life (R2 = 0.30, F = 26.99, p < 0.001), Physical (R2 = 0.24, F = 22.97, p < 0.001), Psychological (R2 = 0.53, F = 73.15, p < 0.001), Social (R2 = 0.22, F = 17.82, p < 0.001), and Environment domain scores (R2 = 0.35, F = 34.95, p < 0.001) were predicted by the STAI2 score. The Level of independence domain score also related to the BDI scores (R2 = 0.35, F = 17.28, p < 0.001), whereas the Overall domain score showed mild correlations with the ToM (r = 0.24, p = 0.053) and Attention and memory factor scores (r = 0.24, p = 0.057).

In 51 PWE, cumulative regression analyses entering the cognitive factors, BDI, STAI, MASQ, and CRI-Adult scores, showed that the Overall domain score was predicted by the STAI2 (R2 = 0.40, F = 32.84, p < 0.001) and ToM factor scores (R2 = 0.49, F = 24.41, p < 0.001).

When considering the FPT scores, the WHOQoL 100 total score and the Level of Independence domain score were predicted by the second and first comprehension scores; the Overall domain score by the second comprehension score; and the Physical, Psychological, and Environment domain scores by the third comprehension score (Table 5).


This study evaluated the relationships of ToM to self-appraisal, coping, and QoL in patients with focal epilepsy with the goal of clarifying the clinical significance of ToM impairment. ToM deficits proved to predict self-rated cognitive functioning, coping strategies to stressful events, and Overall QoL perception. To our knowledge, this is the first study assessing such aspects in epilepsy, providing evidence that ToM may have psychobehavioral implications.

The preliminary results of the study (as revealed by comparisons between PWE and healthy subjects and factor analysis) maintained the following: (1) focal epilepsy originating from the temporal or frontal lobe can impair ToM and (2) ToM deficits represent a specific pattern distinct from other cognitive deficits (Farrant et al., 2005; Schacher et al., 2006; Giovagnoli et al., 2011). The results also confirmed that, although the FPT scores correlated with education, age of seizure onset, and disease duration, age of seizure onset only predicted ToM performance, explaining 8% of its variance (Giovagnoli et al., 2011).

A new result was that the capacity to recognize and comprehend real mental states and interpersonal interactions may contribute to explain the variance of cognitive self-rating, suggesting that ToM adequacy is important to a correct estimation of one's own functioning. A crucial component of ToM is the comprehension that minds can take different perspectives on the world. Correct solution of ToM tasks requires to distinguish between mental representations held by the self and by the others (Brass et al., 2009). Although the FPT does not directly assess the comprehension of one's own mental states, it requires a subject to separate the facts described to them from the mental states they attribute to the characters and their own point of view. Therefore, a subject who understands real mental states, maintaining a proper self-versus-other distinction, should also maintain self-awareness including a congruent appreciation of their cognitive functioning. In line with previous studies, self-rated cognitive functioning also related to the level of state anxiety (Giovagnoli et al., 1997; Sawrie et al., 1999; Baños et al., 2004; Giovagnoli, 2012) and the Attention and memory factor (Seidenberg et al., 1994; Giovagnoli, 2012).

In 51 of 66 PWE, ToM contributed to predict coping. The ToM factor explained 16% of the variance of the CRI-Adult total score and 15% of the variance of the CRI-Adult behavioral score, although the latter was also determined by chronological age (13%). The CRI-Adult cognitive score was predicted by the Lexical search factor and BDI scores. When considering distinct FPT performances, the CRI-Adult total, behavioral, and cognitive scores related to the second comprehension score, which explained 10–15% of their variance, and the cognitive score was also associated with the FP recognition score (10%). This suggests that, in PWE, the understanding of others' mental states may contribute to adaptation, thereby influencing coping strategies. In this regard, the comprehension of real mental states and social interactions (as expressed by the second comprehension score) appears the most relevant ToM performance. A well-preserved capacity to appreciate interpersonal communication and relational dynamics in a small social group may be important to analyze stressful or sudden circumstances and to solve new problems. Therefore, understanding other's thoughts and relations could help find adequate cognitive or behavioral strategies to face personal and interpersonal stressors. It is worth noting that the CRI-Adult cognitive score was predicted by the Lexical search factor and BDI scores, suggesting that well-preserved strategic search and language, together with adequate mood, are important to the choice of coping solutions.

The ToM factor, in particular the ability to comprehend others' mental states and interactions, contributed in predicting the Overall QoL perception. Although circumscribed, this relation indicates a positive influence for mind reading on subjective well-being. The more efficient the understanding of others' minds and interactions, the better the perception of QoL. An adequate comprehension of real mental states and interpersonal dynamics could avoid redundant thoughts or actions, helping make behavior fluid, or enhancing the feelings of belonging to a social group. A negative influence of ToM impairment on QoL parallels the effects of poor mood, stigma, and low-self-esteem on personal well-being and social integration (Gramstad et al., 2001; Suurmeijer et al., 2001; Giovagnoli et al., 2006). In line with the previous studies, QoL primarily related to anxiety and depression (Fisher et al., 2000; Hermann et al., 2000; Gramstad et al., 2001; Suurmeijer et al., 2001; Giovagnoli et al., 2006; Tracy et al., 2007; Gois et al., 2011).

Together, these findings resemble the results of previous studies of frontotemporal dementia, where ToM deficits related to psychobehavioral alterations, as expressed by the Neuropsychiatric Inventory (Gregory et al., 2002), obsessive-compulsive disorders (Grisham et al., 2010), schizophrenia (Koelkebeck et al., 2010), or schizophrenia-related social maladjustment (Mehl et al., 2010), supporting the clinical significance of ToM impairment. The present findings may contribute to clarification of the psychosocial maladjustment caused by epilepsy (Hermann et al., 2000; Reynders et al., 2005; Gois et al., 2011; Broicher et al., 2012). In keeping with Broicher et al. (2012), these results demonstrated no association between ToM and depression, maintaining that high-order cognitive abilities and mood are distinct aspects. On the contrary, the discrepancy in the relationship between ToM and QoL (not documented by Broicher et al., 2012) may reflect differences in QoL measurement. The QOLIE 30 (Vickrey et al., 1993) used by Broicher et al. (2012) is an epilepsy-specific inventory, whereas the WHOQoL 100 is a generic measurement. The WHOQoL 100, validated in a large population with a variety of health states, provides transversal non–disease-related information (WHO, 1996). Compared with disease-specific inventories, it may give more information on psychobehavioral aspects, allowing more precise comparisons between QoL and ToM.

In patients with TLE, Reynders et al. (2005) found no relationship between the judgment of social trustworthiness and psychobehavioral distress, but they did not assess ToM. In other patients with TLE, Gois et al. (2011) showed that social adjustment was influenced by attention and memory, whereas in patients with idiopathic epilepsy, Shehata et al. (2009) revealed that aggressive behavior and psychotic personality traits correlated with comprehension and memory performances. The present study extends such findings, demonstrating that, in patients with focal epilepsy, ToM may explain some psychobehavioral alterations more consistently than other cognitive functions. The association between ToM and psychobehavioral alterations may reflect damage to common neural substrates that underlie ToM and emotional-behavioral control (Shamay-Premack & Woodroof, 1978; Baron-Cohen et al., 1994; Stone et al., 1998; Shamay-Tsoory & Aharon-Peretz, 2007). Neural activity supported by the medial temporal and prefrontal cortical areas and their connecting pathways can be damaged by TLE and FLE, explaining a coexistence of ToM impairment and psychobehavioral alterations (Schacher et al., 2006; Giovagnoli et al., 2011).

To conclude, selective ToM deficits provoked by focal epilepsy may have a clinical impact, with particular effects on self-appraisal, coping, and overall QoL perception. Further studies are needed to confirm these results. Clarifying ToM alterations may help plan nonpharmacologic treatment, such as cognitive training, metacognitive therapy, psychotherapy, or work/social training. Cross-cultural comparisons between neuropsychological and psychobehavioral assessment used in different epilepsy clinics may validate any result.


The authors thank Serena Oliveri, Anna Tarallo, and Valentina Manfredi for helping with the neuropsychological tests, Chiara Paterlini for scoring the WHOQoL 100 and CRI-Adult inventories, and all of the patients for collaborating to the study.


The authors declare no conflict of interests. We confirm that we have read the Journal position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.