Psychiatric morbidity in children with KCNJ11 neonatal diabetes

Abstract Aims Mutations in the KCNJ11 gene, which encodes the Kir6.2 subunit of the pancreatic KATP channel, cause neonatal diabetes. KCNJ11 is also expressed in the brain, and ~ 20% of those affected have neurological features, which may include features suggestive of psychiatric disorder. No previous studies have systematically characterized the psychiatric morbidity in people with KCNJ11 neonatal diabetes. We aimed to characterize the types of psychiatric disorders present in children with KCNJ11 mutations, and explore their impact on families. Methods The parents and teachers of 10 children with neonatal diabetes due to KCNJ11 mutations completed the Strengths and Difficulties Questionnaire and the Development and Wellbeing Assessment. Strengths and Difficulties Questionnaire scores were compared with normative data. Diagnoses from the Development and Wellbeing Assessment were compared with known clinical diagnoses. Results Strengths and Difficulties Questionnaire scores indicated high levels of psychopathology and impact. Psychiatric disorder(s) were present in all six children with the V59M or R201C mutation, and the presence of more than one psychiatric disorder was common. Only two children had received a formal clinical diagnosis, with a further one awaiting assessment, and the coexistence of more than one psychiatric disorder had been missed. Neurodevelopmental (attention deficit hyperactivity disorder and autism) and anxiety disorders predominated. Conclusions Systematic assessment using standardized validated questionnaires reveals a range of psychiatric morbidity in children with KCNJ11 neonatal diabetes. This is under‐recognized clinically and has a significant impact on affected children and their families. An integrated collaborative approach to clinical care is needed to manage the complex needs of people with KCNJ11 neonatal diabetes.


Introduction
Mutations in KCNJ11, which encodes the Kir6.2 subunit of the K ATP channel, are the commonest cause of neonatal diabetes. These are important to diagnose because > 90% of people with these mutations can transfer from insulin treatment to an oral sulfonylurea, achieving excellent glycaemic control [1]. KCNJ11 is expressed in the brain as well as the pancreas [2], explaining why~20% of people with mutations in this gene have a neurological phenotype known as DEND (developmental delay, epilepsy and neonatal diabetes) syndrome [3]. Even those without an overt neurological phenotype have recently been shown to have attention deficits and developmental coordination disorder on neuropsychological testing [4].
Case reports and animal data suggest that KCNJ11 mutations may be associated with childhood psychiatric disorders. Two people with the commonest DEND mutation, V59M, have been reported to have attention deficit hyperactivity disorder (ADHD) [5]. Mouse models with V59M mutations targeted to neuronal tissue have replicated the hyperactive phenotype, and show increased exploratory behaviour and reduced anxiety behaviour consistent with the inattention and impulsivity reported in humans, suggesting a role for KCNJ11 in emotional regulation [6]. Autism (comprising impaired language and social interaction and restricted/repetitive behaviours) has also been reported in one person with the V59M mutation [7]. No previous studies have systematically assessed the psychiatric morbidity in people with KCNJ11 neonatal diabetes, or the impact that this has on families.
We aimed to characterize the types of psychiatric disorders present in children and adolescents with KCNJ11 mutations, and explore the impact of these on families.

Participants
We recruited 10 children from the UK (median age 8.5, range 6-17 years) with neonatal diabetes due to a mutation in the KCNJ11 gene.

Ethical approval
Ethical approval was obtained for the study from the National Research Ethics Service Committee South West-Exeter.

Recruitment and consent
Patients were recruited at a neonatal diabetes family event in Exeter. Valid informed consent was obtained from parents (at the family event) and teachers (at a later date).

Developmental and physical health history
Parents reported the ages at which their child achieved major fine and gross motor, social and speech and language milestones, their child's educational attainment, and any interventions required, for example, speech and language therapy, extra support at school. They also gave details of their child's physical and mental health history and current diabetes medication.

Psychiatric evaluation
Parents and teachers completed the Development and Wellbeing Assessment (DAWBA) and the Strengths and Difficulties Questionnaire (SDQ). The DAWBA is a standardized diagnostic interview that combines structured and semi-structured approaches to generate Diagnostic and Statistical Manual of Mental Disorders IV (DSM-IV) [8] psychiatric diagnoses on 5-17-year-olds. Parallel versions exist for parents and young people aged 11-17 years, with a briefer questionnaire for teachers. It covers common emotional, behavioural and hyperactivity disorders, as well as less common but sometimes more severe psychiatric disorders. Clinicians assess the data from all available informants to assign diagnoses according to DSM-IV. The initial DAWBA validation study showed excellent discrimination between community and clinic samples in rates of diagnosed disorder [9]. Since then, the DAWBA has been widely used in British national surveys and as a tool to aid clinical assessment in many countries.
The SDQ consists of 25 items relating to emotional symptoms, conduct problems, hyperactivity/inattention, peer relationships and prosocial behaviour. Scores across the first four subscales (5 items each) are summed to create a 'total difficulties score' that ranges from 0 to 40. The impact supplement measures distress and social impairment caused by the child's difficulties. The reliability and validity of the SDQ make it a useful screen for psychopathology in children [10].

Data analysis
The total difficulties, impact scores and psychiatric diagnoses were compared with normative data for~8000 school-age children from the British Child and Adolescent Mental Health Survey 2004 [11].

Clinical and developmental history
Parents reported high levels of developmental delay and learning difficulties (Table S1). Nine of ten children required intervention to assist education or development. Psychiatric morbidity (Table S1) was recognized: two children had a clinical diagnosis of autism, one was awaiting assessment due to probable autism and one had 'autistic tendencies'. Another child required psychological support for low mood.

SDQ
Both parents (Fig. 1) and teachers (Fig. S1) reported high levels of psychopathology. Parent-reported median impact and median total difficulties Z-scores were 2.8 and 1.8 compared with the general school-age population (Z-score > 1.3 psychiatric evaluation suggested). Problems were most marked in emotional difficulties and hyperactivity. Prosocial behaviour scores were also reduced compared with the background population (median Z-scores -1 and -0.7 for parent and teacher report, respectively).

What's new?
• This is the first study to systematically assess psychiatric morbidity in people with KCNJ11 mutations, using validated, standardized diagnostic tools.
• The data show that KCNJ11 mutations, in addition to causing neonatal diabetes, also cause psychiatric disorders that are clinically unrecognized but have high impact on families.
• This research highlights the need for early assessment and an integrated and collaborative approach to clinical care in people with KCNJ11 neonatal diabetes.

DIABETICMedicine
Psychiatric morbidity in KCNJ11 neonatal diabetes P. Bowman et al.

DAWBA
Clinical evaluation of parents' and teachers' responses to the DAWBA showed that definite psychiatric disorder(s) were present in six of ten children, but only two had a clinical diagnosis, with a further one awaiting formal assessment ( Table 1). All children with either V59M (n = 4) or R201C (n = 2) had a definitive psychiatric diagnosis on the DAWBA. There was more than one disorder in four of six children, which was not recognized clinically. The prevalence of psychiatric disorder in British school children using the DAWBA is 10% [11], so the level of psychiatric morbidity is this cohort is higher than the background population (P = 0.0001 for a one-sample proportion test). Neurodevelopmental disorders were prominent (autism and ADHD). Three of ten children had both autism and ADHD; although DSM-IV criteria exclude ADHD as a diagnosis in the presence of autism, clinical practice and DSM-V have moved towards assigning both due to the high impact on families and need for clinical intervention. Anxiety   disorders were common, with five of ten children being diagnosed with at least one anxiety disorder. Three children had additional probable diagnoses, but we could not make definitive diagnoses based on the DAWBA. Case 6 (R201C) was assigned a probable diagnosis of autism (and is awaiting formal diagnostic assessment for autism by his local services), and Case 8 (I182V) had a probable diagnosis of other disruptive (conduct) disorder, but this related to behaviour more than 6 months previously which was now resolving. Finally, Case 9 (K170N) was supported by health psychology and a nurture group for low mood and self-esteem, but did not reach the DAWBA diagnostic threshold for an emotional disorder.

Discussion
Neurodevelopmental disorders (autism and ADHD) and/or anxiety disorders were present in all six children with sulfonylurea-treated neonatal diabetes due to V59M or R201C KCNJ11 mutations. Most of these psychiatric disorders had not been diagnosed in clinical practice.
Definitive psychiatric diagnosis occurring only in those with the V59M or R201C mutation is consistent with the previous literature. These are the two commonest mutations associated with neurological/developmental features, which are almost invariable in V59M and inconsistently reported in R201C [3]. There is a clustering in the type of psychiatric disorder; ADHD is present in all four children with the V59M mutation, consistent with previous reports of ADHD in people and hyperactivity, inattention and impulsivity in mice with the mutation [2,5,6]. Autism, found in four children, has previously been reported in a single patient [7]. The presence of anxiety disorders in four children differs from the reduced anxiety behaviour noted in the V59M mouse model [6].
One of the most striking features of the assessment process was the impact on families of the difficulties identified. In those most severely impaired, parents had become full-time carers for their children. Some families reported that their children needed more support, which suggests that awareness of the psychological problems faced by such families should be raised among healthcare professionals involved in their care. The complex pattern of needs that we identified requires a fully integrated and collaborative approach involving parents, carers, general practitioners, paediatric endocrinologists, occupational therapists, clinical/educational psychologists, teachers, special educational needs coordinators and child and adolescent mental health services.

Limitations of the study
Because of the rarity of the condition, the number of participants recruited was small. In addition, the families who attended our family day may not be representative of all people with KCNJ11 neonatal diabetes. Psychiatric difficulties could make patients more reluctant to attend a public meeting with considerable travelling or make them more likely to attend to seek advice. The total number of UK patients aged 5-17 years with KCNJ11 mutations at the time of the study was 21, therefore 48% of those eligible for inclusion in the study took part. Furthermore, our cohort did have significantly more V59M and R201C mutations than in the total UK paediatric cohort (60% vs. 26% P = 0.049). For these reasons, we have been unable to use this study to provide prevalence data on the psychiatric features associated with KCNJ11 mutations. However, our systematic and detailed assessments found considerable unrecognized psychiatric morbidity in this group.

Further work
Our research suggests that psychiatric morbidity predominantly affects people with V59M and R201C mutations and most of these mutation carriers are affected. A larger study assessing more patients with these and other mutations will give information on prevalence and the extent to which psychiatric morbidity forms part of a specific phenotype/genotype relationship. In addition, further studies are needed to assess the effects of sulfonylurea therapy on psychiatric symptoms in people with KCNJ11 mutations.

Conclusions/implications
Systematic assessment using standardized validated questionnaires reveals a range of psychiatric morbidity in children with KCNJ11 neonatal diabetes. This is under-recognized clinically and has a significant impact on affected children and their families. An integrated and collaborative approach to clinical care is needed to ensure early identification and appropriate management of the complex needs of people with diabetes due to KCNJ11 mutations.

Supporting Information
Additional Supporting Information may be found in the online version of this article: Table S1. Patients characteristics and clinical information Figure S1. Patient difficulties as shown by teacher-report SDQ scores.