Cognitive functioning of adults with Noonan syndrome: a case–control study


  • E. Wingbermühle,

    Corresponding author
    1. Donders Institute for Brain, Cognition and Behaviour, Centre for Cognition
    2. Behavioural Science Institute
    • Vincent van Gogh Institute for Psychiatry, Venray
    Search for more papers by this author
  • R. L. Roelofs,

    1. Vincent van Gogh Institute for Psychiatry, Venray
    2. Donders Institute for Brain, Cognition and Behaviour, Centre for Cognition
    3. Behavioural Science Institute
    Search for more papers by this author
  • I. van der Burgt,

    1. Department of Human Genetics
    Search for more papers by this author
  • P. M. Souren,

    1. Research Technical Support Group (RTOG), Radboud University Nijmegen, Nijmegen
    Search for more papers by this author
  • W. M. A. Verhoeven,

    1. Vincent van Gogh Institute for Psychiatry, Venray
    2. Department of Psychiatry, Erasmus University Medical Centre, Rotterdam
    Search for more papers by this author
  • R. P. C. Kessels,

    1. Vincent van Gogh Institute for Psychiatry, Venray
    2. Donders Institute for Brain, Cognition and Behaviour, Centre for Cognition
    3. Department of Medical Psychology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
    Search for more papers by this author
  • J. I. M. Egger

    1. Vincent van Gogh Institute for Psychiatry, Venray
    2. Donders Institute for Brain, Cognition and Behaviour, Centre for Cognition
    3. Behavioural Science Institute
    Search for more papers by this author

Corresponding author: E. Wingbermühle, MSc, Centre of Excellence for Neuropsychiatry, Vincent van Gogh Institute for Psychiatry, Stationsweg 46, 5803AC Venray, The Netherlands. E-mail:


Noonan syndrome (NS) is a genetic disorder characterised by short stature, facial dysmorphia, congenital heart defects and mildly lowered intellectual abilities. Research has mainly focused on genetic and somatic aspects, while intellectual and cognitive functioning has been documented scarcely. Also, to date studies have been primarily performed in children. This is the first study in which functioning within the major cognitive domains is systematically evaluated in a group of adults with NS and compared with a control group. Extensive neuropsychological assessment, including the domains intelligence, speed of information processing, memory (working memory, immediate recall and delayed recall), executive function and visuoconstruction, was performed in a sample of 42 patients with NS and 42 healthy controls, matched on age, sex and education level. In addition, subjective cognitive complaints were assessed with self-report questionnaires. On the domain speed of information processing patients performed worse than controls (P < 0.05). Furthermore, except for slightly better results on delayed recall in the patients with NS (P < 0.05), none of the other cognitive domains showed between-group differences. On the questionnaires, patients reported substantially more complaints about their own cognitive abilities than controls (P < 0.05). A lowered speed of information processing and relatively intact functioning in other cognitive domains characterises the cognitive profile of adult patients, in contrast to previous findings in children with NS, who seem to have more generalised cognitive deficits.

Noonan syndrome (NS) is a worldwide reported genetic disorder with an autosomal dominant inheritance pattern that was first described by Noonan and Ehmke (1963) and further elaborated upon by Noonan (1994) and Noonan and O'Connor (1995). The prevalence of NS is estimated to be between 1 in 1000 and 1 in 2500 live births (Allanson 2010; Mendez & Opitz 1985; Noonan 1994). Although clinical features may vary, NS is typically characterised by mild facial dysmorphisms including hypertelorism, ptosis, strabismus, downslanting palpebral fissures, low-set ears and a broad and sometimes webbed neck. In addition, NS is frequently associated with congenital heart defects and short stature (Allanson 2007; Van der Burgt 2007). Up to 90% of the patients have cardiac defects, particularly pulmonic valve stenosis (PVS), hypertrophic cardiomyopathy and atrial septal defects. Reported adult heights are about −2 SD, due to postnatally reduced growth (Noonan et al. 2003; Otten & Noordam 2009; Ranke et al. 1988). Other phenotypical characteristics include chest and spine deformities, cryptorchidism, lymphatic dysplasia and delayed pubertal development (Tartaglia et al. 2002; Van der Burgt 2007). Nowadays, the diagnosis is mostly established soon after birth and is based on a set of clinical criteria (Van der Burgt et al. 1994).

NS is caused by germline mutations (mutations in gametes) in the Ras-mitogen-activated protein kinase (Ras-MAPK) pathway. This signal transduction pathway is involved in cell-cycle control processes (Boutros et al. 2008; Zenker 2009). At present, while research in NS continues to identify new causative mutations, in approximately 75% of the patients a causative mutation is found in one of the following genes: PTPN11, SOS1, RAF1, KRAS, NRAS, BRAF, SHOC2, MEK1 (MAP2K1) and CBL (Tartaglia et al. 2011). Mutations in the PTPN11 gene are most common, accounting for over 50% of the cases. As the disorder is caused by germline mutations, it is inheritable (Allanson 2010). In 60% of the cases, mutations occur spontaneously (de novo) (Fakouri & Fakouri 1998; Shaw et al. 2007).

Cognitive functioning in NS

Recently, problems in affective information processing were demonstrated in patients with NS (Wingbermühle et al. 2012). Little is known, however, about the performance of patients with NS on cognitive domains such as memory, executive function and speed of information processing. To date, studies have mainly focused on somatic complications and only incidentally address intellectual (dis)abilities or behavioural characteristics. Also, studies have mostly examined children or adolescents with NS. Moreover, if cognitive functioning is reported, evaluations often result from single-case studies using only one or a few assessment instruments, mostly questionnaires completed by parents or teachers. In the occasional group studies, age ranges are usually wide, including both infants and adolescents, although cognitive performance develops and differs over age stages. Table 1 summarizes sample sizes and age ranges of previous studies on cognitive and behavioural functioning in patients with NS that have included some kind of neuropsychological assessment. Overall, these studies suggest a mildly lowered average intelligence, a delay in language and motor development, mild problems in selective and sustained attention as well as suboptimal executive function to be present in patients with NS (Wingbermühle et al. 2009). These kinds of cognitive problems may explain an increased need for special education in young patients with NS (Shaw et al. 2007). However, a systematic, controlled and multiple-sourced assessment of cognitive functioning in adults with NS has not been performed as yet.

Table 1. Summary and characteristics of studies on neurobehavioural functioning in Noonan syndrome
StudiesSample sizeAge range (years)Control groupType of neuropsychological assessment
  1. NS, Noonan syndrome; LS, LEOPARD syndrome; CS, Costello syndrome; CFC, cardiofaciocutaneous syndrome.

  2. a

    Verhoeven et al. (2008) and Wingbermühle et al. (2012) studied a subset of patients with NS included in this study.

Alfieri et al. (2011a)16 NS/4 LS5.0–11.1 (median = 9.6)NoFull examination
Alfieri et al. (2011b)18NS: 4–17 (mean = 10)Yes (n = 43)Visual processing paradigms
Alfieri et al. (2008)18NS: 6–36NoVisuomotor screening
Cesarini et al. (2009)29 NS/LS, 9 CS and 11 CFCNS/LS: 0.7–35.6Yes (n = 20, CS and CFC)Intelligence test
Collins and Turner (1973)271.5–61 (mean = 14.2)NoIntelligence test
Cornish (1996)34.2; 5.1; 24.6NoIntelligence test
Duenas et al. (1973)252–36NoIntelligence test (n = 15)
Ghaziuddin et al. (1994)113NoScreening
Horiguchi and Takeshita (2003)110NoFull examination
Krishna et al. (1977)137NoFull examination
Lee et al. (2005)484–16 (mean = 9)NoScreening
Mahendran and Aw (1989)130NoIntelligence test
Money and Kalus (1979)813–26 (mean = 15.8)NoScreening
Pierpont et al. (2009)654–18 (mean = 10)NoIntelligence test
Pierpont et al. (2010a)664–18 (mean = 10)NoFull examination
Pierpont et al. (2010b)671–24 (mean = 8.0)Yes (n = 22 CFC)Parent report questionnaires
Sarimski (2000)261–17 (mean = 4.3)NoParent report questionnaires
Troyer and Joschko (1997)28 and 9NoFull examination
Van der Burgt et al. (1999)357–18 (mean = 12.4)NoScreening
Verhoeven et al. (2004)119NoFull examination
Verhoeven et al. (2008)a2816–59 (mean = 29.6)NoFull examination
Wilson and Dyson (1982)17NoLanguage assessment
Wingbermühle et al. (2012)a4016–61 (mean = 29.1)Yes (n = 40)Full examination
Wood et al. (1995)212–16 (median = 9.3)NoParent report questionnaires

Genetic, cardiac and psychosocial influences

NS results from point mutations in genes encoding upstream components of the Ras-MAPK pathway. Since this signalling cascade affects metabolic functioning, including cell and tissue growth and placement, dysfunction may very well result in abnormalities in brain development and function. Manifestations of NS in the human central nervous system have, however, not extensively been studied. Potential neurological disorders that have been described in NS are epilepsy, craniosynostosis, hydrocephalus and Arnold Chiari Malformation (Wingbermühle et al. 2009). Animal studies have showed abnormal central nervous system development and mental retardation to be linked to signalling disturbances in the Ras-MAPK pathway (Gauthier et al. 2007; Samuels et al. 2009). In humans, lowered intellectual and adaptive functioning has been associated with dysregulation of the Ras-MAPK cascade, with mutations more downstream resulting in more severe impairments than mutations in upstream components of the pathway (Cesarini et al. 2009; Pierpont et al. 2010b). Cardiac defects associated with NS may also influence cognitive functioning. The aetiology of cognitive deficits in congenital heart disease (CHD) is thought to be multi-factorial and includes preoperative, operative and postoperative factors. The most profound intellectual and cognitive impairments, in particular with respect to executive functions such as set shifting, planning, problem solving and calculation, have been described for cyanotic conditions (Daliento et al. 2006). Nonetheless, acyanotic cardiac disorders, which are more common in NS, may also reduce cognitive performance (Mahle 2001; Miatton et al. 2006; Snookes et al. 2010). However, the only study about cardiac involvement in cognitive function in NS, Pierpont et al. (2009) did not find a significant relation between the severity of heart disease and both verbal and nonverbal abilities in 65 children with NS.

Given the somatic history of most patients with NS, with frequent hospital visits during their childhood, higher stress levels and an increased prevalence of psychopathology (i.e. mood and anxiety disorders) might be expected, which in turn may be associated with cognitive impairments. Depressive symptoms are known to occur more frequently in patients with CHD, and increased severity of cardiac problems has been associated with more anxiety and fears (Green 2004; Gupta et al. 2001; Lip et al. 2003). Older children and adolescents with CHD seem to be more vulnerable for internalizing and, to a lesser extent, externalizing behavioural problems than younger children (Karsdorp et al. 2007).

In children and adolescents with NS, short stature and deviant physical appearances may increase the risk of being bullied by peers, which, in turn, may result in increased levels of psychological distress. In their study of healthy sixth graders, Nishina et al. (2005) found frequent peer harassment to result in higher levels of depression, anxiety, loneliness and a lower self-esteem. Kumpulainen (2008) described bullying as a clear predictor of both concurrent and future psychiatric disorders, with heightened levels of anxiety in (male) bully-victims in young adulthood.

In summary, both genetic and somatic make-up, as well as psychosocial aspects may affect cognitive functioning in NS. Therefore, this study aims to systematically examine the cognitive profile of adults with NS, using both extensive neuropsychological assessment methods as well as self-report measures of cognitive complaints.

Materials and methods


In this study, 42 adult patients with NS and 42 control participants were included. The Department of Human Genetics of the Radboud University Nijmegen Medical Centre referred patients for cognitive and behavioural evaluation. Participation was voluntary and written informed consent was obtained from all participants and/or from their legal representatives. The study was approved by the Institutional Review Board of Vincent van Gogh Institute for Psychiatry in accordance with the Declaration of Helsinki.

The patient group is to be considered as a convenience sample, composed of consecutively referred patients with NS. Inclusion criteria for the patients were a confirmed clinical diagnosis of NS (scoring system Van der Burgt et al. 1994) and an age of 16 years or older. The age of the included patients ranged from 16 to 61 (mean = 30.67, SD = 13.38; 24 females). The education level ranged from 2 (only primary school completed) to 7 (academic degree), according to the Dutch educational system (see Duits & Kessels 2006).

With respect to genetic subtypes, 22 patients had a confirmed mutation in the PTPN11 gene, five in SOS1, one in KRAS and one in SHOC2. In seven patients, no known mutation could be found and in six patients, mutation analyses had not been performed or completed yet.

A heart defect was present in 21 of 40 patients (in two patients, no information was available about cardiac condition). Twenty-one patients had a PVS, of whom five had PVS combined with another heart condition (minor septal defects). Five patients had a non-PVS heart condition only, i.e. hypertrophic cardiomyopathy (1), arrhythmia (1), or atrial septal defect (3). Eighteen patients with NS reported a history of or current need for special education and 35 reported a history of being bullied.

A control group was included consisting of community-dwelling volunteers, who were recruited via the network of the researchers involved in the process of data collection. Control participants were recruited based on education level primarily, and subsequently matched to the group of patients with NS on education level, age and sex using the patients' group characteristics (supported by statistical between-group testing of these variables). Exclusion criteria for controls were: a diagnosis of a genetic or neurological disorder, and the presence of a current psychiatric disorder. Their age ranged from 18 to 56 (mean = 34.19, SD = 14.28; 25 females). The level of education ranged from categories 3 to 6. No significant differences were found on age (t(82) = 1.17, P = 0.25, d = 0.25), sex (U = 861.0, Z = 0.22, P = 0.83), or education level (U = 722.5, Z = 1.54, P = 0.12) between patients with NS and control participants. None of the controls reported a history of or current need for special education and 17 reported a history of being bullied. Minor cardiac insufficiencies were present in two controls.


Neuropsychological testing

Neuropsychological tests were selected to assess the major cognitive domains. In total, 14 measures were obtained, covering five cognitive domains: intelligence, speed of information processing, memory (with the subdomains working memory, immediate recall and delayed recall), executive function and visuoconstruction. Trained assessors administered the tests. The order of the tests was not fixed, but depended on clinical logistics and patient characteristics. In most cases, testing took place on two consecutive days.


The full scale IQ of the third edition of the Wechsler Adult Intelligence Scales (WAIS-III) and the estimated IQ of the Dutch version of the National Adult Reading Test (NART) were used as measures of intelligence (Schmand et al 1992; Wechsler 2005).

Speed of information processing

To test information processing speed, three tasks were included. The raw score on the WAIS-III subtest digit symbol coding (DSC) served as an index of motor and mental speed. During this test, the subject is asked to pair numbers and symbols in accordance with presented examples. In addition, the raw score of the WAIS-III subtest symbol search (SS) was included, a task in which target symbols have to be detected in a stimulus set as quickly as possible. Both DSC and SS have to be performed under time pressure, and the combined scores of these tests constitute the processing speed index of the WAIS-III (Wechsler 1997). Furthermore, two of three parts of the Stroop Color Word Test (CWT) were used to measure mental speed. To this end, the average time used to complete part I (quickly reading of colour names) and part II (quickly naming of colours) was computed (Stroop 1935).


Memory was divided into three subdomains: working memory, immediate recall and delayed recall. To assess working memory, digit span and letter-number sequencing (LNS) of the WAIS-III were administered. In digit span, series of numbers of increasing length have to be repeated in forward and backward order. In LNS a string of combined letters and numbers has to be repeated, while placing them in numerical and alphabetical order. Episodic memory was measured by the immediate and delayed recall conditions of both the Rey Auditory Verbal Learning Test (RAVLT) (Rey 1941; Van der Elst et al. 2005) and the Rey-Osterrieth Complex Figure Test (Rey CFT) for visual memory (Rey 1964). In the RAVLT, the participant is asked to repeat a word list, presented in five trials. Subsequently, the information has to be recalled after 15 min. In the Rey CFT, the participant is asked to copy a complex geometrical figure, followed by an immediate recall condition and a delayed recall trial after 10 min.

Executive function

To measure the ability to inhibit a prepotent response, an interference score derived from the Stroop CWT test was used. Performance on the third trial (naming the colour of the ink of colour names printed in incongruent colours) was compared with performance on the previous trials by computing a ratio score (the time used to complete card III divided by the average of the time used to complete card I and II). Furthermore, the Tower of London, a multistep executive task, was included to assess strategy planning and problem solving behaviour (Ganzevles et al. 1994; Shallice 1982). In this task a participant is asked to transfer beads on three pegs, according to a set of rules. The total score was used in this study.


The raw score on the copy trial of the Rey CFT described earlier was used as an outcome measure to assess the domain of Visuoconstruction.

Assessment of subjective cognitive functioning

To gain insight into the experience of patients with NS of their own cognitive status, two self-report questionnaires were included. To assess perceived ability to plan, organise and monitor behaviour in daily life, the Dysexecutive (DEX) Questionnaire was administered. The DEX is a 20-item questionnaire containing items that address everyday executive problems that have to be rated on a five-point scale (Wilson et al. 1996).

The Symptom Checklist-90-Revised (SCL-90-R) is a self-report questionnaire that consists of 90 items, rated on a five-point scale (Arrindell & Ettema 2003). Two indices were used. First, the total score provides an estimate of the global level of psychological stress of the participant. Second, the subscale cognitive-performance difficulty specifically addresses cognitive complaints.

Statistical analyses

For each cognitive domain, raw test scores were transformed into standardised Z scores based on the total group's distribution (taking together controls and patients with NS; see e.g. Brands et al. 2006). Subsequently, domain scores were derived by calculating the mean of the standardised test scores that contributed to each of the cognitive domains. To examine differences between patients with NS and controls, domain scores were compared by general linear model (GLM) multivariate analyses of variance and standardised effect sizes (partial eta squared) were computed. For significant domains, subsequent analyses were performed on the separate subtests in the domain (Bonferroni-corrected; effect-size: Cohen's d). Additionally, the performance of individual patients and controls were compared with available normative data (age and/or education corrected), expressed in standardised Z scores (normative mean = 0, SD = 1). Subsequently, individuals were classified as impaired on a specific subtest if their performance was more than 1.5 Z score below the normative mean (Lezak et al. 2012). Nonparametric Mann–Whitney U-tests were performed to compare the frequency of impaired participants in both groups. Furthermore, the levels of cognitive complaints of the two groups were compared with independent t-tests. Finally, Pearson's correlations were computed between the score on the DEX and the domain score speed of information processing, and between cognitive performance difficulty SCL-90-R and speed of information processing.


Table 2 presents the mean raw scores, the cognitive domain scores, as well as the normative Z scores of the two groups. One patient was not able to complete the SCL-90-R; she was admitted to a psychiatric hospital elsewhere towards the end of the assessment and was not able to finalise all the questionnaires. Finally, one control did not complete the SCL-90-R and the DEX was not administered in two controls, due to time restrictions.

Table 2. Cognitive domain scores, mean raw test scores, normative Z scores and % impaired for controls and patients with NS
Cognitive domain and testsPatients with NSControls
Mean domain/ raw scoreSD domain/raw scoreMean normative Z scoreaN (%) impaired (i.e. Z < 1.5)Mean domain/raw scoreSD domain/raw scoreMean normative Z scoreaN (%) impaired (i.e. Z < 1.5)
  1. Notes: Negative values indicate worse performance.

  2. WAIS-III, Wechsler Adult Intelligence Scale – Third Edition; NART, National Adult Reading Test; RAVLT, Rey Auditory Verbal Learning Test; Rey CFT, Rey-Osterrieth Complex Figure Test; FSIQ, full scale IQ; Stroop I, II, mean part I and part II of the Stroop Color Word Test.

  3. a

    Normative data for WAIS-III (n = 588) from Wechsler (2005), NART (n = 601) from Bouma et al. (2012), RAVLT (n = 1680) from Van der Elst et al. (2005), Rey CFT (n = 117) from Gallagher and Burke (2007), Stroop Color Word Test (n = 585) from Schmand et al. (2003), and Tower of London (no normative data for this version available).

  4. *P < 0.05; **P < 0.01.

Intelligence0.021.11−0.659 (21%)*−0.050.64−0.702 (5%)
     WAIS-III FSIQ92.1914.65−0.529 (21%)91.148.65−0.594 (10%)
     NART-IQ88.3613.80−0.789 (21%)88.408.22−0.773 (8%)
Speed of information processing−0.21*1.01−0.665 (12%)0.210.64−0.211 (2%)
     WAIS-III digit symbol coding66.1217.37−0.467 (17%)*70.3811.640.081 (2%)
     WAIS-III symbol search30.719.96−0.5910 (24%)*34.198.30−0.282 (5%)
     Stroop I, II58.37**12.46−0.938 (19%)52.027.70−0.443 (7%)
Working memory0.001.06−0.333 (7%)−0.040.76−0.382 (5%)
     WAIS-III letter-number sequencing10.203.10−0.113 (8%)9.122.18−0.466 (14%)
     WAIS-III digit span13.553.44−0.489 (21%)14.293.04−0.293 (7%)
Immediate recall0.140.91−0.265 (12%)−0.170.69−0.575 (12%)
     RAVLT – immediate recall50.0510.38−0.024 (10%)44.7410.31−0.6510 (24%)
     Rey CFT – immediate recall19.547.60−0.459 (22%)18.315.81−0.499 (21%)
Delayed recall0.18*0.85−0.115 (12%)−0.210.77−0.518 (19%)
     RAVLT – delayed recall10.692.97−0.116 (14%)9.313.17−0.599 (21%)
     Rey CFT – delayed recall20.427.34−0.0610 (24%)17.736.20−0.439 (21%)
Executive functioning0.100.910.011 (2%)−0.100.600.301 (2%)
     Tower of London27.834.2025.713.06
     Stroop interference1.780.360.011 (2%)1.720.240.301 (2%)
Visuoconstruction0.121.10−0.7312 (29%)−0.120.88−0.9014 (33%)
     Rey CFT – copy30.395.37−0.7312 (29%)29.234.30−0.9014 (33%)

GLM multivariate analyses of variance showed an overall worse performance in the patient group (F7,76 = 2.49, P = 0.023, ηp2 = 0.19). A significant difference was found in two domains, speed of information processing (F1,82 = 5.15, P = 0.026, ηp2 = 0.059) and delayed recall (F1,82 = 4.80, P = 0.031, ηp2 = 0.055). No significant group differences were found for the domains intelligence (F1,82 = 0.120, ηp2 = 0.001), working memory (F1,82 = 0.051, ηp2 = 0.001), immediate recall (F1,82 = 3.19, ηp2 = 0.037), executive function (F1,82) = 1.31, ηp2 = 0.016) or visuoconstruction (F1,82 = 1.21, ηp2 = 0.015).

For the significant domains speed of information processing and delayed recall, separate between-group comparisons on the individual neuropsychological measures were performed (independent t-tests; Bonferroni-corrected). Regarding speed of information processing, a significantly worse performance of the patient group compared with controls was found on the Stroop CWT (t(82) = 2.81, P = 0.006, Cohen's d = 0.61), but not on WAIS-III digit symbol coding (t(71.65) = 1.32, d = 0.29) or WAIS-III symbol search (t(82) = 1.74, d = 0.38). With respect to delayed recall, post hoc analyses did not show any significant differences on the two subtests in the domain (Rey CFT-delayed recall (t(78.10) = 1.80, d = 0.40; RAVLT-delayed recall (t(82) = 2.06, d = 0.45).

Patients with NS reported significantly more subjective executive problems on the DEX (NS mean 26.5, SD = 14.0; control mean 20.5, SD = 9.6; t(70.6) = 2.25, P = 0.028, d = 0.50). They also reported significantly higher levels of complaints on the subscale cognitive-performance difficulty of the SCL-90-R (NS mean 18.4, SD = 7.4; control mean: 15.0, SD = 4.6; t(66.9) = 2.52, P = 0.014, d = 0.55). No significant difference was found between the group with NS and controls on the SCL-90-R total score (NS mean: 146.1, SD = 41.8; control mean: 132.6, SD = 34.2; t(77.0) = 1.60, d = 0.35). Correlations between DEX or cognitive-performance difficulty SCL-90-R and the domain score speed of information processing were not statistically significant (r = −0.14 and r = −0.16, respectively).


This is the first study to examine the performance of adults with NS on all major cognitive domains, compared with that of matched controls. Patients with NS showed a significantly worse performance in the domain speed of information processing. No between-group differences were found on any of the other cognitive domains, except for a slightly better performance of patients with NS on the subdomain delayed recall. The level of complaints concerning their own cognitive capacities was significantly higher in the patients compared with controls. Patients with NS experienced more cognitive problems in their daily lives than controls, although the overall level of distress did not differ between the two groups. Also, no relation was found between the performance on measures of speed of information processing and the level of subjective complaints of distress. Concerning the background variables, slightly lowered intelligence levels were found in the adults with NS as compared with the average population, which is consistent with results of the majority of previous studies. Moreover, a higher need for special education was noted in our sample (43% of the patients with NS vs. 0% of the controls), which has been reported by others as well (Shaw et al. 2007; Van der Burgt et al. 1999; Wood et al. 1995).

Patients showed a lower speed of information processing than the control group. This finding was mainly driven by a highly significant and medium to large sized difference in performance on the Stroop CWT (reading and naming speed on part I and part II), a task that does not require psychomotor responses. Moreover, a significantly higher proportion of patients than controls showed impairments on the other two subtests for processing speed, WAIS-III-DSC and WAIS-III-SS, compared with the normative mean. Occasionally, lowered scores have been reported on time-limited tests in children with NS (Horiguchi & Takeshita 2003; Troyer & Joschko 1997; Van der Burgt et al. 1999), which have typically been interpreted as manifestations of attentional impairments. Notably, a lack of selective or sustained attention has been observed in children with NS, mainly by their parents (Lee et al. 2005; Sarimski 2000; Wood et al. 1995). Our adult group did not show deficits in selective attention as part of cognitive control (response inhibition as measured with the Stroop CWT interference score), and observation during testing did not show any signs of inattention. Possibly, the mental slowness may be associated with metabolism, in particular with the energetic limitations due to mitochondrial dysfunction that is reported to be manifest in syndromes of the Ras-MAPK pathway, including NS (Kleefstra et al. 2011). So far, however, the exact pathophysiological mechanisms for this association remain unclear, and the hypothesis of energetic limitations as a cause of speed problems is still highly speculative.

Apart from speed of information processing, patients with NS did not perform worse than controls on any of the other cognitive domains, in contrast to previous studies in children. For example, mild decrements in working memory and learning have been reported in children with NS (Horiguchi & Takeshita 2003; Troyer & Joschko 1997; Van der Burgt et al. 1999). Although measures of episodic memory function are usually thought to be intact in these children, Alfieri et al. (2011a) recently showed reduced verbal recall memory performance and normal recognition in a group of children with NS or LEOPARD syndrome. In the adults examined in this study, no differences were found between patients with NS and controls with respect to working memory or immediate recall on episodic memory tasks. The performance on the subdomain delayed recall was even slightly better in the patients with NS than in controls, although subsequent analyses did not show any subgroup differences on the individual subtests in this domain. These findings regarding memory functions seem to be in contrast with those of Alfieri et al. (2011a), and with animal studies that have shown recall problems in relation to activations in the Ras-MAPK cascade (Sweatt 2001, 2004).

Van der Burgt et al. (1999) reported weaknesses in planning and organisation abilities in children with NS, using an extensive intelligence battery. Again, these results could not be corroborated in our adult sample, as our patients with NS performed at control level on measures of working memory or executive function (planning and response inhibition). Interestingly, although no differences were found in the performance on executive tests, patients reported more subjective executive problems than controls on questionnaires. It could be argued that these complaints are associated with executive impairments that were not addressed by the current tests, such as a deficit in concept shifting. However, as our aim was to investigate all major cognitive domains, we could not examine every aspect in detail because of time limitations. Furthermore, in a previous study from our group in a smaller sample of patients with NS (n = 33) that included the Wisconsin Card Sorting Test, a widely used concept shifting task, no deficits were found compared with normative data (Wingbermühle et al. 2009). Maybe, the mental slowness we found underlies the experience of executive problems in daily life. Also, it has been shown that subjective estimates of people's own cognitive status are often incongruent with objective measures (e.g. Goverover et al. 2005; Koerts et al. 2012).

Finally, no differences were found between patients and controls on a test for visuoconstructive ability. In children with NS, research with cognitive-experimental paradigms has suggested visuomotor difficulties to be present (Horiguchi & Takeshita 2003; Lee et al. 2005; Troyer & Joschko 1997). Specifically, within the perceptual domain, the study of Alfieri et al. (2011b) showed impairments in form coherence in children with NS of varying age. It could be speculated that visuoconstruction is only compromised in NS during childhood, but no longer in adulthood. However, Alfieri used cognitive paradigms that are not applicable in clinical practice yet and their findings should be replicated with a dedicated multi-method assessment to objectify cognitive functioning of adults with NS in the domain of visuoconstruction.

In addition to the previously suggested, speculative energetic limitations, there are alternative explanations for the underlying deficit in information processing speed. For instance, heart defects may result in an increased risk for circulatory problems, with cognitive slowness as a result. In NS, however, the associated heart defects are usually fairly mild and, unless complications occur, largely asymptomatic. In line with this, Pierpont et al. (2009) did not find severity of heart defects to be associated with mental capacities in children with NS. Another, indirect, explanation for the speed problems may be in an effect of being bullied on cognition. The high rates of a bullying history in patients with NS can lead to anxiety, which may moderate cognitive development. However, recent research comparing the same patients and controls as examined in this study, did not show any differences in general levels of ‘anxiety’ (SCL-90-R subscale), although levels of social distress were significantly higher in the patients with NS (Wingbermühle et al. 2012).

Comparison of the performances of our groups with normative means shows a large number of impaired performances in various domains. It should be noted, however, that this is the case for both patients and controls, which is related to the fact that norms do not always adjust for schooling effects (Lezak et al. 2012). This emphasizes the need for the inclusion of specific control groups matched with respect to age and education level, to prevent overestimation of cognitive deficits in patients with lower intellectual ability.

Strengths of our study are the extensiveness of the neuropsychological assessments, evaluating all relevant cognitive domains, as well as the inclusion of a carefully matched control group. While a larger sample size would possibly have allowed for even more detailed subgroup analyses, it should be noted that neurobehavioural studies in NS are typically performed in smaller samples (see Table 1). Also, inspection of the effect sizes of the non-significant findings indicates that our study had sufficient statistical power resulting in reliable (non)results. Besides replication studies, preferably including more cognitive domains such as language and perception, future studies should in particular focus on the aetiological mechanisms of the mental slowness in adults with NS, which may be related to factors such as genetic influences on brain development (e.g. neuroimaging of cortical thickness and white-matter integrity) or somatic aspects (for example heart disease). As our sample consisted of patients with varying IQ levels from below average to higher ability, it would be interesting to stratify patients with NS on IQ level in future, larger studies. Moreover, there is a need for longitudinal studies to evaluate developmental change across the lifespan in NS. Systematically comparing specific patients groups with different NS mutation types is also recommended. Finally, future control groups should also include patients with other genetic disorders, preferably with Noonan-like genotypes and phenotypes (e.g. cardiofaciocutaneous syndrome, Turner syndrome), as well as young adults with a history of heart disease.

In conclusion, cognitive functioning in adults with NS is characterised by impaired speed of information processing, but not in any of the other cognitive domains, taking age and education level into account. Despite the relatively intact cognitive profile in adults with NS, patients frequently report cognitive complaints, indicating that neurocognitive and psychological assessment is important in clinical management of patients with NS. The extent and profile of cognitive impairment in adults with NS seems to be different from that of children with NS, who appear to have more generalized cognitive deficits. This may point towards a developmental delay rather than impairments that remain present during the lifespan.


This research is part of a collaborative project of the research group ‘Psychopathology and Genetics’ of the Radboud University Nijmegen and the Vincent van Gogh Institute for Psychiatry, Venray, The Netherlands. The authors would like to thank the Dutch Noonan Syndrome Foundation for their support in recruiting participants. The work of Margery Schrijvers, MSc and Marloes Wiltink, MSc of Radboud University Nijmegen in the data collection for the control group is gratefully acknowledged. Finally, the authors thank Regina Bökenkamp, MD, for her comments on the manuscript with respect to cardiac functioning in Noonan syndrome. The authors have declared that there is no conflict of interest.