Confabulations in a teenager with a right frontal hemispherotomy: Possible underlying mechanisms

Abstract

Objective: Confabulations, i.e. false memories without intention to deceive, can be observed in adults with frontal brain damage. Confabulations are typically associated with episodic memory and/or executive disorders although the severity of these impairments is highly variable. Confabulations may also be associated with emotional/motivational particularities, as a positive bias seems to prevail in such situations. Several distinct cognitive and socio-affective processes may account for the various types of confabulations and these issues remain open. Method: We present the case of a teenager with an early acquired frontal damage, referred for “strange lies.” Besides a standard neuropsychological assessment, we explored (1) source memory, using a reality-monitoring task, (2) episodic autobiographical memory, including both the recollection of the past memories and the ability to imagine future personal events, with an episodic future thinking task (EFT), (3) the sense of self, with questionnaires targeting self-representations, self-esteem and self-competence. Results: The results showed the expected source memory deficits and poor episodic future thinking, whereas episodic autobiographical memory was preserved, contrary to the episodic memory dysfunction usually evidenced in adult confabulators. The confabulations produced by this teenager displayed a clear positive bias, seemed to fit to personal/social goals and to wishful ideations, and were associated with an above average self-esteem and self-representation. Conclusions: These results support the hypothesis of self-identity and emotional regulatory roles of the confabulations. Although the literature on confabulating children/teenagers is nearly non-existent, a more systematic screening of confabulations should be conducted in order to avoid false interpretation to strange discourse or behavior.

Keywords:

• Confabulations
• episodic future thinking
• sense of self
• source memory
• reality monitoring

Introduction

Confabulations, i.e. false memories without intention to deceive, have been reported in adults with a variety of neurological conditions, and mainly occur in cases of amnesia or impaired executive functioning. The prevalence of confabulations in adults following brain damage remains however largely unknown, and reports of confabulations in children or teenagers following brain damage are nearly non-existent. We report here the case of a teenager presenting with confabulations, in the context of an early onset intractable frontal lobe epilepsy resulting in a right frontal hemispherotomy.

Frontal lobe epilepsy in children has been associated with various cognitive deficits, such as attention and executive dysfunction, memory impairments, motor skill and behavioral disorders (for a review, see Verche et al., 2018). The deficits seem rather similar between adults and children, although less severe in the latter, and frontal lobe resections in patients with intractable epilepsy do not seem to significantly alter cognitive abilities when comparing pre- and post-surgical results (Ferguson et al., 2021). Although memory disorders are frequently reported in cases of frontal lobe epilepsy, memory studies in this population are limited and results remain controversial. Recent studies argue for similar declarative memory functioning in frontal versus temporal epilepsies in children (Kibby et al., 2019; Smith, 2016) and for a variability of memory dysfunction in adults, depending on the epileptic focus and spread of the epileptic dysfunction (for a review, see Centeno et al., 2010). With regard to autobiographical memory, Thaiss and Petrides (2008) reported an absence of autobiographical memory deficit in seizure-free adult patients (following neurosurgery), whereas Rayner et al. (2015) evidenced poor autobiographical memory in a 5/9 adult patients with intractable frontal epilepsy. These results suggest that persisting epileptic dysfunction may interfere with the cerebral networks and cognitive processes involved in (autobiographical) memory but that frontal epilepsy may not alter (autobiographical) memory per se, as it normalizes after the suppression of the epileptic dysfunction.

Contrary to the paucity of studies focusing on episodic memory in the context of frontal lobe epilepsy, memory in the context of frontal lobe lesions has been more extensively studied. Thus, the prefrontal lobe has been linked to episodic memory (Levine et al., 1999; Wheeler et al., 1995, 1997) and poor autobiographical retrieval appears to correlate with impaired executive functioning (Della Sala et al., 1993). Patients with frontal lobe lesions are characterized not only by difficulties to recall past autobiographical events but also to pre-experience future events or, in other words, to mentally project themselves into the future (Bertossi et al., 2016; 2017; Ciaramelli et al., 2021), an ability referred to as episodic future thinking. Moreover, frontal lesions are typically associated with source memory deficits (Janowsky et al., 1989), i.e. to impairments regarding the attribution of the origin/context in which the information has been encountered. In particular, existing studies have highlighted the importance of the prefrontal cortex in the ability to retrieve contextual information linked to an event/item, and different regions of the prefrontal cortex (ventrolateral/dorsolateral) have been related to distinct contextual retrieval skills (Chapados & Petrides, 2015; Petrides, 2000).

Source memory includes multiple facets, and the literature currently differentiates between external and internal source memory. Whereas external source memory refers to the ability to associate the correct context in which an item was encountered (for example, the ability to know if the information was delivered by Paul or John), internal source memory instead refers to the ability to distinguish between internally or externally-generated actions (for example, the ability to discriminate between a real action or something that was imagined). This internal source memory is referred to as a “reality-monitoring” and appears to rely on the medial anterior prefrontal cortex (for a review, see Simons et al., 2017). Classically, reality-monitoring has been empirically tested using paradigms in which participants had to read a series of word overtly or covertly and then remember in which condition the word was read (Harvey, 1985).

Impaired reality-monitoring sometimes results in the production of confabulations, i.e. distorted/false souvenirs about oneself or the world, and has been reported in patients with ventral and/or orbital prefrontal lesions of various etiologies (Turner et al., 2008b). Considering the diversity of confabulations’ type (spontaneous versus question-provoked, acted or not, fantastic or mundane), their underlying cognitive processes may differ. The origin of confabulations thus remains debated and several hypotheses have been proposed (Dalla Barba & Kopelman, 2017), including the source amnesia hypothesis (Johnson, 1991; Moskovitch, 1989; Schacter et al., 1984), and the hypothesis of superimposed motivational and emotional dysregulations, as confabulations often appear to be tightly linked to topics affectively relevant for the patient (Burgess & McNeil, 1999; Fotopoulou et al., 2004; Turnbull & Salas, 2017). More precisely, individuals who confabulate seem to magnify the reality, to present themselves in rewarding social situations, suggesting a “positive bias” in the content of the confabulations compared what is actually experienced by the patient “in real-life” (Fotopoulou et al., 2008b).

On the other hand, reality-monitoring skills (together with executive skills) have been related to the quality of episodic future thinking (Cole et al., 2013). This suggests that the ability to extract different elements from multiple contexts is necessary to mentally construct a coherent episodic future event. Interestingly, narratives of plausible future events in patients with bilateral medial prefrontal lesions are characterized by particularly poor self-references compared to patients with hippocampal lesion (Kurczek et al., 2015), suggesting an important role of the frontal regions in the incorporation of the self in future thinking. This report is in line with the literature reporting a specific role of ventromedial prefrontal regions in self-representation (Murray et al., 2012; Northoff, 2011; Northoff et al., 2006).

In this paper, we report the case of a teenager (NM) who was referred for a neuropsychological assessment by his psychotherapist because of a suspicion of confabulations a few years after a surgical resection for intractable frontal epilepsy.

The first aim of this study was thus to attest the presence of confabulations and to characterize them. Then, we were interested in the possible underlying mechanisms of the confabulations and assessed executive and memory skills. According to the literature, we made the hypothesis that NM would present with executive dysfunctions. We also expected that NM would present reality-monitoring impairments. More precisely, we made the hypothesis that he would display an internal reality-monitoring deficit, i.e. that he would fail to distinguish between internally- and externally-generated events. We thus used a speech monitoring task, in which NM had to distinguish and recognize words he had previously read silently or overtly. We expected an abnormal rate of source confusions (i.e. confusions between the silently versus overtly read words). Moreover, we intended to explore more precisely the episodic autobiographical memory, both in the past and in the future, using an episodic future thinking task (EFT). We expected to observe future thinking impairments, whereas autobiographic and episodic memory would be globally preserved, as reported in frontal lesions studies. Finally, we wondered whether a positive bias could be found in the confabulations, according to the dysregulation hypothesis, and whether this positive bias would also affect the self-representation and self-esteem of the patient.

Case report

This right-handed bilingual boy, without family history of psychiatric or neurologic disease, presented a first epileptic seizure at 3 years old (resulting in a change of hand laterality from left- to right-handed). Epilepsy became intractable and resulted in a surgical disconnection of the right anterior frontal lobe at the age of 9. He was then seizure free but the successive EEGs performed between 9 and 16 years of age showed the persistence of a very abundant anterior right epileptiform activity, diffusing to the left. Antiepileptic medication (Keppra) was thus maintained. A brain MRI with tractography performed at 16 years old revealed a status post right anterior hemispherotomy with involution of the frontal lobe, and with the persistence of rare tracts in the anterior portion of the corpus callosum. Posterior portion of the corpus callosum and splenium was preserved. A deviation of midline structures by 12 mm to the right in relation to the atrophy was reported.

Because of an important speech and language delay and poor emotional regulation with frequent fights with peers, NM has been in special education from the age of 4. He benefitted from speech and language therapy and from intermittent psychotherapy for several years. Although being impulsive and displaying poor insight, the psycho-affective assessments did not reveal any mental health disorder. Intellectual efficiency was reported as mildly impaired at 13 years old (WISC-V: TIQ = 53, VCI 55, VSI 64, PRI 64, WMI 69, PSI 79), with results rather close to the presurgical assessment (TIQ = 48 at 7 years old).

NM’s behavior deteriorated when he was 15 years old, during the COVID-19 pandemic period. He threatened his mother, and an aggravation of the aggressive behaviors with his peers was reported. Moreover, the teachers of the institution mentioned the apparition of strange “lies”: for example, the teenager pretended to have a 7-year-old son with his girlfriend. He also reported more trivial false memories, such as having purchased a red scooter and used it for trips with his friends (although he had no driving license and no money to buy a scooter), or having participated to music festivals with friends hundred kilometers away from his home. When confronted to the unrealistic nature of these events, NM seemed to be confused, accepting the fact that it was not possible but nevertheless expressing the intimate conviction that it was true and that he had really lived these events. This alerted his psychotherapist who referred this nearly 17 years old teenager for a neuropsychological assessment.

During the developmental history interview in the presence of his mother and then during the assessment, this teenager interacted and behaved appropriately, responded politely, willingly and informatively to the questions. He was however relatively immature and slightly passive. The venue of an unfamiliar person resulted in an unexpected shyness reaction (he didn’t look at her). Fatigability and distractibility occurred after 60 to 90 minutes. During the neuropsychological assessment, NM was very cooperative, eager to do well, but showed signs of anosognosia (finding everything “easy”). However, he acknowledged spontaneously having a tendency to get angry too quickly, a failure to refrain from fighting and hitting his peers when upset or when having the feeling to be provoked. Although feeling bad after such events, he would fail to control himself.

At the neuropsychological assessment (Table 1), results showed a progression of the intellectual efficiency over the years, now situated in the borderline range. Although language was fluent, intelligible and informative, there were persisting speech and language deficits, with a poor lexicon, a few semantic paraphasias and poor morpho-syntaxic and discursive skills. General semantic memory and lexico-semantic knowledge was also very poor (standard score = 3 for both the information and the vocabulary subtests, WISC V). Declarative memory for simple, isolated items (lists of words or figures) showed a normal encoding curve, preserved delayed recall and recognition skills. However, a few intrusions, false recognitions and interferences occurred, suggesting monitoring deficits. The results obtained in the Children Memory Scale (CMS) showed poor verbal memory indexes, whereas the visuo-spatial memory indexes were within the normal range. The language deficit may have interfered with the verbal memory tasks (poor understanding and thus poor recall of the stories for example). Interestingly, the patient reported difficulties in the Everyday Memory Questionnaire, whereas the ratings of his parents revealed average scores. The analyses showed however that the patient pointed to attentional difficulties rather than retrieval deficits. Inattention and poor verbal working memory was confirmed in the tests and questionnaires.

Table 1. Results of the clinical neuropsychological assessment.

Domain	Results	Tests
Intellectual efficiency	VCI = 70 (CI 95% = 65-83), VSI = 72 (CI 95% = 67-82) PRI = 72 (CI 95% = 67-81) WMI = 67 (CI 95% = 62-79) PSI = 77 (CI 95% = 71-89) TIQ = 66 (CI 95% = 61-74)	WISC V
Speech and language	Index = 63 Fluent and intelligible verbal productions but poor lexicon (∼7 years old), semantic paraphasias, morpho-syntaxic and discursive deficit	PPVT KABC II Frog story
Fine motor skills	Dominant hand (right): −1.4 SD Non dominant hand (left): −4.3 SD	Purdue Pegboard
Short-term & working memory	Digit span: SS = 3 (direct order:5, reversed order:3) Picture span: SS = 5	WISC V
Episodic memory	General memory Index: 58 Encoding/learning index: 63 Visual immediate/delayed memory index: 75/96 Verbal immediate/delayed memory index: 55/59	CMS
	Encoding: SS = 8, Delayed recall: NS = 7, Recognition: NS = 8 1 intrusion and 1 interference	Word list CMS
	Encoding: −1 SD, Recognition: 12/15 (1 omission, 2 false recognitions)	Rey’s 15 figures test
Semantic memory	Information: SS = 3 Vocabulary: SS = 3	WISC V
Executive and attentional functions	Verbal category fluency: −1 SD/Figural fluency: −3 SD Inhibition: −0.9 SD Flexibility: > PR16 Planification: Total move score: SS = 104 Total correct score: SS = 100 Total initiation time: SS = 90 Total problem solving time: SS = 100 Inattention: T-score = 62 (inconsistency and variability of RT) Sustained attention deficit (significant errors’ increase across blocks)	Fluency tests Stroop Color Trail Towers of London CPT III
Social cognition	Affect recognition: SS = 5 Fear recognition: PR =2-5 Anger recognition: PR <2 TOM contextual: 3/5 Failure to detect sarcasm and bluff Detects the faux-pas but fails to justify	NEPSY II NEPSY II Happé stories Baron-Cohen faux-pas
Questionnaires	Executive Behaviour: GEC = PR.78, MI = PR.78, BRI = PR.86 Emotional control (PR.98)^a Working memory (PR.92)^a Everyday memory General mean: 1.77 (-1.5 SD)/0.92 (-0.2 SD) Retrieval factor: 2 (-1.2 SD)/1.4 (-0.5 SD) Attentional factor: 2.25 (-2.55 SD)/0.5 (-0.1 SD) Temporal estimation (total correct = 21/34) Temporal orientation: 4/4 Temporal sequences: 5/6 Objective duration estimation: 2/10 Subjective duration estimation: 4/5 Anticipation: 6/9	BRIEF (parental report) EMQ-R (Royle & Lincoln 2008) Auto-/Parental Ratings QTE (Quartier)

^aBRIEF: The higher the percentile, the more severe the deficit. SS = standard score; SD = standard deviation; PR = Percentile Rank, WISC V: Wechsler Intelligence Scale for children, 5th edition; PPVT = Peabody picture vocabulary test; CMS = children memory scale; BRIEF = behavioral rating inventory of executive function; GEC = global executive composite, MI = metacognitive index, BRI = behavioral regulation index; CPT III = continuous performance test, 3rd edition; RT = reaction time; NEPSY II: Developmental Neuropsychological Assessment, 2nd edition; TOM = theory of mind; EMQ-R = Everyday Memory Questionnaire-revised form, QTE = Questionnaire d’Estimation Temporelle.

The executive functions assessment included tests of planification (Towers of London), flexibility (Color Trail Making Test), inhibition (Stroop test) and initiation (fluency tasks, in the verbal and non-verbal domain). The Behavioral Rating Inventory of Executive Functions (BRIEF) was moreover completed by the parents of the patient in order to collect more ecological data. This inventory allows to compute a Global Executive Composite (GEC) score, composed of 2 scales: the Behavioral Regulation Index (BRI) assessing aspects related to inhibition, flexibility and emotional control skills; and the Metacognitive Index (MI) reflecting initiation skills, working memory, planification, organization of material and monitoring ability. The results obtained in the executive tests (planification, flexibility, inhibition, verbal initiation) and in the everyday life executive behavior inventory (GEC, BRI, MI) were all situated within the normal range, except for the non-verbal fluency test (very few productions but no perseveration) and for the BRIEF emotional regulation subscale. NM was also significantly impaired at recognizing angry and fearful facial expressions. At last, severe dyschronia was noted, NM being unable to estimate the objective duration of everyday life activities (e.g. duration of toothbrushing: 3 hours), although the temporal orientation and other temporal judgements were preserved.

As the teachers and the psychotherapist reported possible confabulations, we adapted the Confabulation Battery (Dalla Barba et al., 2019) for the age and intellectual level of the patient. The obtained responses revealed preserved semantic autobiographical memory and adequate temporo-spatial orientation. Responses to episodic autobiographical questions were instead stereotyped and similar events were reported when asking what occurred last Christmas, at New Year’s Eve, or last Saturday evening (“I spent it with friends, we looked at a film, then we went to a concert with singers and then we went in a bar until 4 a.m.”). Many confabulations (8/10) were produced when semantic or episodic questions with an expected “I don’t know” response were asked (for example: What did you do on March 13th 2017? “I watched TV with my brother”; When did you go for the first time in a public swimming pool? “On May 7th 2015”; When was the music group Sexion d’Assaut formed? “March 6th 2007.” What was the job of Marylin Monroe’s father? “Film producer”) (Table 2).

Table 2. Results obtained in the experimental memory tasks and complementary questionnaires, compared to typically developing age-controls and IQ-matched controls.

	NM results	Age-control norms: mean (standard deviation)	IQ-control norms: mean (standard deviation)
Confabulation questionnaire (adapted from Dalla Barba et al., 2019)	Semantic autobiographical memory: 10/10 Orientation: 9/10 “I don’t know” questions: 80% confabulations	NA	NA
Source memory (speech-monitoring task, from Debbane et al., 2010)	Nb correct recognition (target items): 40/48	33.16/48 (5.83)	32.58/48 (6.36)
	Nb false recognition: 11	4.89 % (2.66)	5.74 % (3.02)
	Percentage source confusions: 16/40 (40%)	28.41 % (2.66)	34.85 % (2.66)
	Internal misattributions: 47.6 %	27.83 % (13.24)	39.02 % (14.06)
	External misattributions: 31.5 %	29.14 % (13.53)	30.36 % (13.78)
Episodic Future Thinking Task (Feller et al., 2021)	Recollection (total = 8)
	- Specific episodic narratives: 8 (100%)	79.78% (25.04)	65.29% (25.79)
	- Extended narratives: 0 (0%)	12.59% (18.82)	18.07% (17.01)
	- Categoric narratives: 0 (0%)	7.63% (15.31)	16.64% (19.5)
	- Experiential index (sum): 36	32.09 (8.5)	18 (7.51)
	- Subjective appreciation index (sum): 25.5	24.85 (5.55)	15.46 (7.32)
	Production (total = 8)
	- Specific episodic narratives: 1 (12.5%)	58.06% (29.15)	48.99% (37.15)
	- Extended narratives: 3 (37.5%)	22.56% (17.61)	9.7% (13.53)
	- Categoric narratives: 4 (50%)	19.37% (22.31)	41.31% (37.87)
	- Experiential index (sum): 26	27.42 (8.6)	13.83 (6.44)
	- Subjective appreciation index (sum): 15	21.67 (6.24)	10.71 (4.83)
Self esteem	Rosenberg Self-Esteem Scale (Valliere &Vallerand	32.02 (4.88)	NA
	French adaptation, 1990): 32 points
	Social self esteem (Michaud et al., 2006)
	Self-competence:
	Self-assertion/social anxiety: 21/35 points	24.3/35 (5.5)
	Initiative/social communication: 32/35 points	26.7/35 (5.2)
	Extraversion/leadership: 16 points	13.6/21 (3.6)
	Self-disclosure/relational empathy: 19 points	16/21 (3.6)
	Social self-acceptance, popularity, relationships
	Social acceptance/popularity: 18/20 points	17.9/25 (3.7)
	Interest and social motivation: 18/20 points	16.1/20 (3.1)
	Success and motivation in love relationships: 14/20	14.3/20 (3.2)
Psychotic-like experiences (PQ-16)	Prodromal Questionnaire – short version 16 items (Ising et al., 2012): 0/16 A score ≥ 6 indicates the presence of psychotic-like experiences requiring a more in-depth clinical assessment	not clinical	NA

NA = not available.

We finally looked for eventual psychotic-like experiences with a standardized structured interview based on the Prodromal Questionnaire (Ising et al., 2012), which was unambiguously negative.

Informed consent according to the Declaration of Helsinki was obtained by the patient and his parents.

Procedure

Two experimental tasks were used in order to explore reality-monitoring and episodic autobiographic memory, including the recall of past events and episodic future thinking. We selected tasks that had been designed and successfully used with teenagers with borderline intellectual efficiency in other clinical settings (autism spectrum disorder/22q11.2 deletion syndrome) to ensure that our patient could perform the tasks adequately. For the two experimental tasks, NM’s performance was compared to the mean performance of same age typically developing participants (age-control norm) but also to the mean performance of same age youth with mild intellectual disability (IQ-control norm), as published in Debbane et al. (2010) and Feller et al. (2021). Of note, the mean intellectual level of the IQ-matched participants in these two papers was 79.12 (sd = 12.65) and 73.00 (sd = 16.49), respectively. The age- and IQ-control norms for the two paradigms are displayed in Table 2. Finally, self-reported questionnaires targeting self-esteem and the feeling of self-competence were administered to examine the presence of a possible positive bias regarding self-representation.

Reality-monitoring task

We used the speech-monitoring task adapted by Debbane et al. (2010) in which participants had to distinguish and recognize words and pseudo-words among distractors that were previously read silently or overtly.

In this task, NM was presented with 6 blocks of 8 items alternating between silent and overt reading conditions. Before each block, the patient was instructed to read the items silently or overtly. In total, 48 items (target) were presented, including 12 words read overtly, 12 words read silently, 12 pseudo-words read overtly, and 12 pseudo-words read silently. After a 15-minute delay (during which NM completed a visuo-spatial filler task), the patient had to recognize whether the items were part of the reading task he had just performed and, in case the item was recognized, whether the item had been read overtly of silently. A total of 72 items were presented during the recognition phase (48 previously presented items and 24 new distractors items).

The total number of correctly recognized items (maximum correct = 48) and the total number of false recognition (i.e. number of distractors identified as targets) were computed. Among the correctly identified targets, the percentage of source memory confusions (i.e. the percentage of misattribution to the silent versus overt condition) was also computed. Among these confusions, the number of external misattributions (i.e. number silently-read items reported as overtly-read, divided by the number of correct recognition of the silently-read items) and internal misattributions (i.e. number of overtly-read items reported as silently-read, divided by the number of correct recognition of the overtly-read items) was specified.

Episodic future thinking task (EFT)

We used an episodic future thinking task developed in our group (Feller et al., 2021) and adapted from d’Argembeau et al. (2008). In this task, NM had to recall personal past events (i.e. recollection condition) and to imagine plausible future events that could likely happen (i.e. production condition). For both temporal conditions (recollection and production), the instruction was to generate specific events (i.e. events that took place in a specific location on a specific day and lasted no longer than a day) based on a cue word. For example, NM was requested to recall a past personal event (recollection condition) or imagine a plausible future event (production condition) associated with the cue word “holidays.” An example of NM’s production is provided as Supplementary Material. The patient was free to choose any personal event that took place or could possibly take place in the future, and was specifically instructed to give as many details as possible, including olfactive/gustative, visual and auditive details but also thoughts, feelings and actions. In half of the narratives, the patient was instructed to recall/imagine a solitary event (non-social condition) and in the remaining narratives an event where he was accompanied by at least another person (social condition). The detailed instruction of the task is provided in Supplementary Material. In total, the patient had to tell 16 narratives (i.e. four for each word, one narrative per condition [temporal: recollection and production; social: non-social and social]).

The scoring of the narratives was performed as follows: first, each narrative was classified in different categories, as described in d’Argembeau et al. (2008): (1) specific: events that took place in a specific location on a specific day, and lasted no longer than a day; (2) extended: events lasting more than a day; (3) categoric: collection of events that were not related to one another or did not refer to a specific period of time. Secondly, an Experiential Index was computed, reflecting the number of details provided for each narrative. Finally, a Subjective Appreciation Index (rated from 1 = not at all to 5 = extremely) was attributed for each narrative, reflecting how vividly the interviewer subjectively perceived the narrative. This index reflects the narratives’ vividness, imaginability. In addition to these scores, we also rated more qualitative aspects of the narratives. In the recollection condition, we reported the number of events that were clearly or potentially non-plausible, reflecting the presence of potential spontaneous confabulations. For each narrative, we also rated whether the valence was clearly positive, clearly negative or somewhat neutral. All the narratives were double scored. For the specificity and experiential index, potential discrepancies between the two raters were resolved by discussing the individual scoring and going back to the transcript of the narratives if necessary. For the Subjective Appreciation Index, a mean of the two ratings was computed in case of discrepancy. For additional information about the task and scoring, see Feller et al. (2021).

Self-esteem, feeling of self-competence and social self-acceptance questionnaires

We used the French version of the Rosenberg Self-Esteem questionnaire (Vallieres & Vallerand, 1990) to obtain a global measure of self-esteem. This questionnaire has been validated in 18-year-old participants and is composed of 10 questions rated on a 4-point scale (from “not at all agree” to “completely agree”). Secondly, we used the Social Self-Esteem questionnaire designed for young adults aged 17–20 (Michaud et al., 2006), which explores the feeling of self-competence in different interpersonal social contexts (16 items rated on a 7-point scale ranging from “not at all able” to “completely able”), and the subjective social value estimation, i.e. the feeling of self-acceptance, popularity and interest in social relationship (13 items rated on a 5-point scale ranging from “not at all agree” to “completely agree”). Within the self-competence dimension, four domains are covered: (1) self-assertion and social anxiety (e.g. “are you able to ask someone you like out on a date” or “feeling comfortable with unfamiliar people”); (2) initiative and social communication (e.g. “attract the interest and attention of others in a discussion” or “initiating social contacts rather than waiting for others to come to you”); (3) extraversion and leadership (e.g. “take charge and be the leader in social situations” or “participate in a group discussion in one of your classes”); (4) self-disclosure and relational empathy (e.g. “show others that you care and that you are interested in them”). As for the subjective social value dimension, three domains are covered: (1) social acceptance and popularity (e.g. “I find it difficult to make friends” or “I am popular with the other kids of my age”); (2) interest and motivation for interpersonal relationships (e.g. “even when I am very busy (school), I find it necessary to spend time talking and sharing with my friends”); (3) success and motivation in love relationships (e.g. “I would rather be single than in a relationship” or “in love, I am generally successful with girls/boys”).

Note that the questions of both questionnaires are rather simple and easy to understand. However, in order to exclude a bias due to poor sentence understanding/reading skills, we checked for the comprehension of each question with NM before completion.

Results

Reality-monitoring task

The results show preserved basic memory recognition skills, as shown by the high level of correct recognition of the target items (Table 2). However, our patient committed a large number of false recognitions (accepting a distractor item as a target item) and presented with major source deficits, confusing the condition in which the items had been encountered (silence vs overt condition). We applied Crawford’s guidelines for statistical analysis in single case studies (Crawford et al., 2010; Crawford & Garthwaite, 2002; Crawford & Howell, 1998) and the results were significantly lower than those of both age-control and IQ-control groups (Table 3).

Table 3. Statistical analyses for the experimental tasks, comparing the patient’s scores to an IQ-control sample (A) and to an age-control sample (B).

A
Test	IQ-Control sample			Case’s score	Significance test^a		Estimated percentage of the control population obtaining a higher/lower score than the case^b		Estimated size effect ZCC
	n	mean	SD		t	p	Point	95% CI	point	95% CI
Source Memory/False recognition Source Memory/Source confusions	19	5.74	3.02	11	1.698	.05	94.65	84.38-99.29	1.742	1.011-2.454
	19	34.85	2.66	40	1.887	.03	96.23	87.59-99.65	1.936	1.155-2.699
Autobio Memory/Episodic recollection Autobio Memory/Specific narratives (EFT)	24	65.29	25.79	100	1.319	0.1	89.98	78.28-97.09	1.346	0.782-1.895
	24	48.99	37.15	12.5	−0.962	0.17	17.293	7.15-31.36	−0.982	1.465-0.485
B
Test	Age-Control sample			Case’s score	Significance test^a		Estimated percentage of the control population obtaining a higher/lower score than the case^b		Estimated size effect ZCC
	n	mean	SD		t	p	Point	95% CI	point	95% CI
Source Memory/False recognition Source Memory/Source confusions	19	4.89	2.66	11	2.239	.01	98.098	92.20-99.92	2.297	1.419-3.158
	19	28.41	2.66	40	4.247	.000	99.97	99.79-100	4.357	2.870-5.831
Autobio Memory/Episodic recollection Autobio Memory/Specific narratives (EFT)	33	79.78	25.04	100	0.796	0.21	78.392	65.86-88.42	0.808	0.409-1.197
	33	58.06	29.15	12.5	−1.54	0.06	6.672	1.927-14.78	−1.563	2.06-1.046

^aCrawford and Howell (1998): the results are for a one-tailed test.

^bCrawford & Garthwaite (2002).

Within the source memory confusions, the self-monitoring analyses showed a higher tendency to make internal misattributions, whereas the percentage of external misattributions was similar to the control groups (Table 2). Our patient was thus more prone to report that an item presented in the overt condition had been encountered in the silence condition.

Episodic future thinking task (EFT)

The percentage of the patient’s specific/extended/categoric narratives in the recollection and production conditions is provided in Table 2. A sum score for the Experiential Index and the Subjective Appreciation Index in the recollection and production conditions is also provided in Table 2.

In the recollection condition, NM produced specific narratives only. Of the eight narratives, one was rated as clearly non-plausible (the patient recalled a New Year’s Evening party where he went out with friends and came back home very late) and one was rated as potentially non-plausible (the patient recalled receiving a steering wheel for the Playstation and managing to reproduce a very advanced trick in a game by copying a youtuber). He recalled four clearly positive events (two of them corresponded to the (potentially) non-plausible events described above), two somewhat neutral events, and two clearly negative events. NM’s experiential index and subjective appreciation index sum scores were comparable to the age-control group. Of note, most of the details provided by NM in the narratives were actions (16), followed by thoughts/emotions (10) and sensory (visual or auditory) (10) details. NM referred to himself in all the narratives (using “I” or “me” pronouns).

By contrast, in the production condition, NM produced only one specific narrative. This percentage is smaller than in the age-control or IQ-control group, although it did not reach significance (Table 3). NM produced five clearly positive events and three somewhat neutral events. Of note, the content of the narratives in the production condition was rather mundane and did not involve highly unrealistic events. Whereas the Experiential Index sum score was comparable to the age-control group, the Subjective Appreciation Index was somewhat intermediate compared to the age-control and IQ-control groups (Table 2). Of note, most of the details provided by NM in the narratives were actions (16), followed by thoughts/emotions (7) and sensory (visual) (5) details. As in the recollection condition, NM referred to himself in all the narratives (using “I” or “me” pronouns).

To sum up, NM showed a preserved ability to recall specific events of the past but produced a high rate (2/8 narratives) of spontaneous potential confabulations during the task. Interestingly, these confabulations occurred only when he recalled positively-valenced episodes. Later verification with his mother revealed that the content of these 2 episodes was clearly confabulated. The richness and imaginability of the narratives were comparable to what was reported in typically developing and IQ-marched controls. On the opposite, NM had a specific difficultly to generate specific plausible future events but most of them were positively-valenced. Self-references were present in all the narratives.

Self-esteem, feeling of self-competence and social self-acceptance questionnaires

The results showed that our patient had a good global self-esteem, as reported in the validated French adaptation of the Rosenberg questionnaire. The scores were similar to those reported by typically developing young adults.

In the social self-competence and subjective social value questionnaire, our patient even tended to show a better feeling of self-competence compared to young men, especially in the domains of social communication, relational empathy and leadership. High ratings in the domain of social self-representation were also documented regarding the social motivation to relationships.

Discussion

To our knowledge, detailed report of confabulations in youth with neurological conditions is nearly non-existent (for a review, see Christensen & Sorensen, 2019). This case report thus aimed to look for possible underlying mechanisms of confabulations in a teenager. We first explored the memory and executive domains, as adult literature classically points to memory and/or executive dysfunction in such situations of false memories production. In our patient, the executive and episodic/autobiographical memory scores obtained were below the mean but nevertheless in the normal range. In addition, compared to his global intellectual efficiency, which was two standard deviations below the mean (IQ around 70), the results obtained in the executive and memory tests/questionnaires appeared comparatively “better,” as they were situated at only one standard deviation below the mean, making it difficult to argue for a specific executive or episodic memory disorder. Thus, in a context of relatively and globally preserved episodic memory and executive functions, this teenager presented with spontaneous and question-provoked confabulations after a right frontal hemispherotomy for intractable epilepsy.

Interestingly, the question-provoked confabulations were mostly elicited by the “I don’t know” questions of the Confabulation Battery, either in the semantic or the episodic domain. This pattern of confabulation has been reported in patients without specific frontal damage and/or without executive disorder (Dalla Barba & Decaix, 2009). This type of false narratives has been labelled “hypermnesic confabulations” and their underlying mechanism remains unclear. In addition, when questioned on autobiographical personal events, our patient tended to produce relatively frequent events (going to a bar with friends) as if it was a specific, unique episode. He thus attributed repetitive events (or habits) to specific past or future personal episodes. This type of confabulation has been classified as “habit confabulation” (Dall Barba et al., 2020). Our patient finally produced “memory fabrication” confabulations to a smaller extent, which consists in the production of plausible episodic memories. The spontaneous confabulations were sometimes fantastic, but were mostly mundane. They were rather stereotyped and the scenes evoked by the patient systematically depicted him in rewarding social contexts, putting him in self-relevant affective and social situations. This positive bias was also evidenced in his self-representation, with scores on the self-esteem and self-competence questionnaires being in the normal range or even higher than in the general population. The experimental testing highlighted a severe reality-monitoring deficit, with a tendency to commit more internal than external misattributions. Autobiographical memory was globally preserved, whereas episodic future thinking was poor.

Our patient presented with early onset frontal epilepsy (at the age of 3), and a right hemispherotomy was performed at the age of 9 because of intractable epilepsy. Although the boy no longer presented with seizures after the neurosurgery, the EEG showed persistent epileptic propagation to the left frontal lobe, suggesting a persisting long-lasting bilateral frontal damage/dysfunction. The clinical picture of this teenager differs from the florid and fantastic confabulations associated to severe cognitive and behavioral deterioration reported in 2 young adults with right frontal epilepsy (Fujikawa et al., 2016). The clinical observations and assessment of our patient clearly failed to show signs of mental and behavioral deterioration, and/or psychotic features, as described in Fujikawa’s case report. Our specific psychiatric screening for signs of psychosis was moreover negative.

An intriguing aspect concerns the lack of executive dysfunction in this patient. We cannot completely exclude a flexibility impairment, as an impaired score was obtained in the figural fluency test (although no deficit was found in the Color Trail Test and no alteration was reported in the flexibility scale of the behavioral inventory). Furthermore, NM showed significant difficulties in “hot” emotional executive functions, as revealed by the BRIEF emotional regulation subscale and during the developmental history interview (frequent fights/altercations with peers). A more extensive executive assessment, notably the use of rule-detection tests (such as the Wisconsin card sorting test or the Brixton test), may have been interesting in this case, especially because they have been shown to best correlate with the presence of confabulations in adults (Burgess & McNeil, 1999).

Interestingly, our patient also differs from confabulators with later acquired frontal lobe lesions regarding memory skills. Although the literature describes memory impairments of variable severity in brain-lesioned confabulators (Turner et al., 2008a), our patient did not present with significant memory impairment, especially in the context his general borderline intellectual efficiency. The low results obtained in some of the complex verbal memory tasks (in particular in stories recall) seemed rather linked to his language lexical-semantic processing deficit and not to a genuine memory disorder. Easier tasks (regarding the linguistic constraints, such as word lists learning) were in the average range, the Everyday Memory Questionnaire was not suggestive of any memory impairment (both self- and parent-reported administration), and the episodic autobiographical memory assessment was within the normal range (the patient was able to recall specific past events during the EFT task). Our case study also differs from the provoked, memory recall confabulations associated to a severe episodic and semantic anterograde amnesia and a one year retrograde amnesia in a 9 years-old patient, in the context of a traumatic brain injury (Meguro et al., 1999).

An interesting point concerns the origin of the oral language deficit, and the potential role of the neurological dysfunction to this specific alteration. Our patient was born in Switzerland and raised in two languages (French and Serbo-Croatian) from birth. He followed school in French. NM’s parents had a good socio-educative level (chemists) and were integrated in the Swiss community. The bilingual status and/or socio-cultural aspects unlikely accounts for the poor oral language development (and the low verbal IQ). Moreover, there was no genetic susceptibility for language developmental disorder in the family. The participation of the early frontal epileptic dysfunction in the oral language deficit seems highly plausible in this situation (Dubois et al., 2004; Mayor-Dubois et al., 2008). Interestingly, the mother reported a change in the laterality at 3 years old (from left- to right-handed) at the onset of the epilepsy. One could wonder whether oral language was represented in the right hemisphere in this originally left-handed boy and thus was impacted by the early onset and persistent right hemispheric epileptic dysfunction and the later right frontal hemispherotomy. We have no evidence of direct links between the poor oral language skills (especially poor lexico-semantic knowledge) and confabulations in our patient, as he did not produce “semantic” confabulations when asked to describe the meaning of words and/or the attributes of objects/animals, contrary to a patient with aphasia described by Sandson et al. (1986). However, besides the very poor semantic representations of words, the impaired semantic memory of our patient may have contributed to the difficulty to provide and construct detailed plausible events, as suggested by Janssen et al. (2022).

An interesting question concerns the moment of apparition of the confabulations. Although the patient had a very early onset frontal brain dysfunction, confabulations became noticeable by adolescence only. This delay may result from the well-known late maturation of the (pre)frontal lobe and executive functions until late adolescence (Anderson et al., 2001; Lenroot & Giedd, 2006). Thus, although the cerebral damage occurred in early development, the apparition of the confabulations several years later may account for the “growing into functional deficit” hypothesis, i.e. the fact that cognitive deficit can remain silent for years and become apparent later in development (Cacucci & Vargha-Khadem, 2019). From another point of view, adolescence is a period of life which corresponds to challenges in the domain of individuation, as well as a search for a social and individual identity. The spontaneous confabulations of our patient appeared to present with a self-related motivational bias, and seemed to fit to personal/social goals and to wishful ideations. As suggested by Burgess and McNeil (1999), his confabulations looked like « attempts to sustain an emotionally important aspect of his self-identity.” It remains possible that confabulations were present earlier during development but remained unnoticed. However, one can wonder whether these confabulations appeared in the context of affective dysregulation, as claimed by Thurnbull and Salas (2017). In the case of our patient, confabulations were reported in parallel with a worsening of affective dysregulation occurring in the context of the COVID-19 pandemic period, when the patient was 15 years old.

Confabulations have recently been reported in children with neurodevelopmental affections, such as autism disorder spectrum (ASD) (Spitzer et al., 2017) or fetal alcoholic syndrome (FAS) (Brown, 2017). Spontaneous confabulations were reported in ASD, whereas individuals with FAS seemed rather prone to confabulate in stressful situations such as law inquiry. The confabulations’ type clearly differed from each other and the underlying cognitive mechanisms may thus be different, but in both situations the patients presented mainly with executive dysfunctions and/or reality-memory impairments rather than with episodic memory disorder. Our patient showed largely preserved executive skills, as assessed with tests and everyday behavioral questionnaires. The only deficit appeared in behavioral emotional regulation, our patient displaying impulsivity and over-reactivity in affectively challenging and distressing situations.

In line with what has been previously reported in the literature (Fotopoulou et al., 2008a and 2008b), NM’s spontaneous confabulations were positively-biased, with a systematic tendency to present himself in rewarding social situations. Two confabulations also appeared during the EFT, when the patient was asked to recall a personal past event associated with a cue word. The elicitation of confabulations in a similar context was previously reported in a case report of a patient with Korsakoff’s syndrome (Dalla Barba et al., 1990). Interestingly, these two confabulations were both positively-biased, in contrast with the produced “true” memories that were both positively- and negatively-valenced. As suggested by Fotopoulou (2008), this positive bias is consistent with Conway and Pleydell-Pearce’s model (2000), claiming that memory retrieval is guided by personal goal and wished for self-representations. Interestingly, NM showed normal or above-average scores on self-esteem and self-competence measures compared to typically developing youth, despite the presence of learning difficulties and frequent conflicts with peers. One cannot suspect anosognosia to be responsible for these results, as NM’s ratings on other questionnaires (e.g. scores on the Everyday Memory Questionnaire) were borderline to mildly impaired and were more severe than those of his parents. The positive bias was thus restricted to the self-identity representation. In that regard, one could tentatively suggest that the presence of these confabulations has a “protective” effect on the self-identity representation for NM, as also suggested previously by other authors (Conway & Loveday, 2015).

Our patient presented with a severe the reality-monitoring impairment, as has been typically reported in other confabulating brain-damaged patients (Johnson, 1991). These results are congruent with the hypothesis of source amnesia as being part of the origin of confabulations (Johnson, 1991; Moskovitch, 1989; Schacter et al., 1984). One could wonder whether the reality-monitoring deficit in our patient was accounted for by his borderline intellectual efficiency. However, the scores of our patient were clearly and specifically impaired compared to the IQ-control group performance. Besides, reality filtering seems to develop early in childhood, as this mechanism has been shown to be efficacious in typically developing children from the age of 7 years (Liverani et al., 2017) and the mental level of our patient is higher than that. Reality-monitoring (and more broadly source memory) impairments are typically reported in patients with frontal lobe lesion (Janowsky et al., 1989), and is thus not surprising when taking into account the right frontal hemispherotomy and persisting left frontal epileptic dysfunction of our patient.

In addition, even if not significantly different from the control groups (because of the large S.D. in the published samples), NM experienced marked difficulties to produce specific future events (only one specific event was produced). To the best of our knowledge, only two studies examined episodic future thinking in adults with Korsakoff syndrome who presented with confabulations (El Haj et al., 2019; Janssen et al., 2022) and also described a poor specificity of narratives. Of note, the content of these future narratives was rather mundane, which was also the case for NM, indicating that the presence of confabulations is not necessarily associated with the production of unrealistic events. According to El Haj et al. (2019), there could be a direct association between the presence of confabulations and impaired episodic future thinking, as the presence of distorted/fabricated memories could alter the future thinking process. However, future studies are clearly required in this domain to clarify the exact nature of this link. In the case of NM, his inability to produce specific future events was in sharp contrast with the relatively preserved (autobiographical) episodic memory abilities described above. There are several candidate cognitive mechanisms to explain episodic future thinking impairments in the context of relatively preserved episodic memory (for a review, see Schacter et al., 2017). First, a number of studies have highlighted the role of executive functioning in the production and recombination of event-related details (e.g. de Vito et al., 2012; Summerfield et al., 2010). In the case of NM, it is unlikely that executive impairments influenced performance on the episodic future thinking task, as his scores on the tests assessing a broad range of executive functions were mostly in the average range. Secondly, in patients with frontal lesions, such as adults with fronto-temporal dementias, it has been suggested that impairments in future thinking might be the reflect of a broader impairment in self-reflection (Irish & Piolino, 2016). Again, it is unlikely that this is the primary mechanism in the case of NM, as we did not find a lack of self-reference in the narratives, contrary to what was previously reported in adult patients with bilateral frontal lesions (Kurczek et al., 2015). A third mechanism pertains to the role of reality-monitoring, as suggested by Schacter et al. (2017). Indeed, as episodic future thinking relies on the ability to retrieve and recombine elements of past experiences to form plausible future events, the presence of reality-monitoring difficulties might alter this reconstruction process. As described above, NM’s performance on the reality-monitoring paradigm was severely impaired and could therefore sustain this hypothesis. Finally, Irish and Piolino (2016) have proposed that semantic memory is also important for future thinking, as semantic knowledge provides a general conceptual framework for future thinking. In the case of NM, semantic knowledge was severely impaired (as evidenced by the low scores on the Information and Vocabulary subtests of the WISC-V). However, as NM did not attend the regular school system, it is difficult to conclude whether low scores on these tests truly reflects impaired semantic memory or a lack of opportunity to be in contact with the topics that are probed in these subtests. A more detailed assessment of semantic memory would have been needed to further examine the relevance of the semantic scaffolding hypothesis in this context (see La Corte et al., 2021).

Conclusion

To our best knowledge, this is the first study providing an in-depth examination of confabulations, reality-monitoring, and episodic autobiographical memory (recollection of past events and future episodic thinking) in a teenager with an early frontal brain dysfunction/lesion, together with an analysis of self-representation and self-esteem. The results show the expected reality monitoring deficits and poor episodic future thinking, with a relative preservation of autobiographical memory and executive skills, contrary to the episodic memory dysfunction evidenced in adult confabulators. The spontaneous and provoked confabulations produced by this teenager displayed a clear positive bias, seemed to fit to personal/social goals and to wishful ideations, and were associated with an above average self-esteem and self-representation. These results support the hypothesis of a self-identity and emotional regulatory role of the confabulations.

Confabulations in children and or adolescent with neurological conditions or following early brain injury have been rarely reported. Whether this reflects a very exceptional situation in development (eventually linked to plasticity), or a lack of assessment/awareness of this clinical manifestation by child neuropsychologists remains open. More systematic investigations are clearly warranted in such populations. Indeed, confabulations place the child/teenager at high risk for social/professional integration failure and/or psychiatric mis-diagnosis, and should be specifically investigated in case of behavioral complaints (“lies,” strange acting). Thus, in case of suspected confabulations, a screening with an adapted version of the Dalla Barba’s confabulations questionnaire (adapted to the age of the patient) and containing “I don’t know” questions is highly recommended. Moreover, as the confabulations may be mundane, the events and autobiographical memories reported by the patient should be checked with the parents and/or with the teachers. At last, in case of confabulations, psycho-educative counseling is recommended for an appropriate management of the child at home and at school.

Acknowledgements

We would like to thank Clémence Feller for the double scoring of the Episodic Future Thinking task.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

The author(s) reported there is no funding associated with the work featured in this article.