Complex OSA or OSA type III

Children with anatomical and/or functional abnormalities of the upper airways represent a very specific population which is at high risk of OSA. This type of OSA is also called complex OSA or OSA type III . As in normally developing children, OSA may be caused or associated with a hypertrophy of the adenoids and the tonsils. However, the main cause of upper airway obstruction is an anatomical or functional reduction of the upper airway calibre. Indeed, children with facial malformations, such as craniofacial stenosis, Pierre Robin Sequence, Treacher-Collins syndrome, micrognathia, choanal atresia, or any other type of malformation of the maxilla or mandibula, but also children with storage diseases, such as mucolipidoses or mucopolysaccharidoses, congenital bone disease, such as achondroplasia or pycnodysostosis, Prader Willi syndrome, or Down syndrome, present an increased risk of OSA due to a reduced upper airway calibre or patency . Most of these diseases are rare diseases but the cumulative number of diseases and the prevalence of OSA in each condition are both extremely high. OSA is not restricted to an age range and may occur at any age, from infancy to adulthood. Even after treatment, the risk of OSA persists lifelong, with the consequent need for a follow up. In the pediatric context, this is especially important until the end of the pubertal growth spurt . Symptoms, clinical examination, and sleep questionnaires are insufficiently sensitive and specific as screening tools to document OSA and its severity in these patients, underlining the need of systematic sleep studies . OSA is usually severe, with high risk of residual OSA after upper airway surgery, even if preceded, and orientated by drug-induced sleep endoscopy (“DISE”) .

Neurocognitive dysfunction in OSA type III

• •
  

  What are the challenges?

Neurocognitive dysfunction and behavioural problems associated with OSA are less well studied in children with OSA type III. This may be explained by several factors. Firstly, numerous children with OSA type III do not have systematic sleep studies. This may be explained by the lack of awareness of the importance of systematic sleep studies in children with rare and/or genetic disorders such as congenital bone diseases (including achondroplasia, pycnodysostosis, or other skeletal dysplasias), storage diseases, or 22q11 deletion (DiGeorge syndrome, velocardiofacial syndrome). Even when guidelines or recommendations are available, they may not be appropriate regarding the optimal timing for formal sleep breathing assessment. Indeed, for children with Down syndrome, the American Academy of Pediatrics recommends referral to a pediatric sleep laboratory for all children aged between 3 and 4 years . But recent studies showed that OSA may occur at a younger age in Down syndrome, and even during the first months of life . The limited access to sleep studies is also a challenge. The lack of sleep studies makes it thus impossible to have objective data on the potential association between OSA and a neurocognitive dysfunction and behavioural problems in these children.

The analysis of the link between OSA and neurocognition and behaviour is complex in this population. Patients may have sensory deficits with hearing problems in Treacher Collins syndrome, mucopolysaccharidoses, or achondroplasia. Patients with cleft palate or velopharyngeal insufficiency may have language problems that may impair communication and school performance. Visual problems may occur in children with faciocraniostenosis. Children with congenital bone disease or storage disease may have motor deficits that may hinder writing and mobility with the need for a wheelchair.

Intellectual disability and behavioural problems may be an integral part of the disease such as in Down syndrome, mucopolysaccharidosis, Prader Willi syndrome, or 22q11 deletion. In these children, intellectual disability is thus usually attributed to the underlying disease without recognition that part of the patient’s deficit may be aggravated or caused by associated OSA. Questionnaires assessing neurocognitive function, intelligence, and language and behaviour, are not always adapted and usable in children with neurocognitive dysfunction. The Vineland Adaptive Behavior Scales, Second Edition (VABS-II) is specifically designed for patients with intellectual disability but other questionnaires are less appropriate for children with intellectual disability. Also, a large number of different questionnaires have been used, which makes it difficult to compare the results of different studies. The age of the patients differs also between studies, with very few studies that have been performed in very young children. As OSA may be more severe and more common in young children, this may have an impact the association between sleep-disordered breathing and neurocognitive dysfunction. Concerning the effect of sleep-disordered breathing treatment on neurocognitive and behavioural improvement, the benefit may be greater in infants and young children, due to a shorter duration of sleep problems and a greater capacity for improvement during a critical window of rapid somatic and neuropsychological growth.

Poor sleep quality, even without associated sleep-disordered breathing, may by itself cause or aggravate behavioural problems, poor concentration, and impair executive function and memory. As sleep-disordered breathing is often associated with poor sleep quality, it is difficult to establish their respective roles in the patient’s neurocognitive functioning.

Of note, in practice, behavioural problems may limit the cooperation of the child for a sleep study and require technicians and staff having a training and expertise in these more challenging patients . Ultimately, because of a lack of cooperation, a sleep study will not possible in the least cooperative patients .

Moreover, the number of diseases that have been studied with regard to the consequences of OSA is low compared to the extremely large number of diseases which are associated with an increased risk of OSA.

• •
  

  Neurocognitive dysfunction in OSA type III

Type of studies

Most of the studies analysing the link between OSA and neurocognitive dysfunction and behaviour and language problems are case-control or cross sectional studies. To our knowledge, few interventional studies, comparing the status of the patients before and after OSA treatment, are available.

Disorders that have been studied

The majority of studies that analyzed the link between OSA and neurocognitive dysfunction in children with OSA type-III concern children with Down syndrome.

The parents of 80 children with Down syndrome, aged 36–70 months completed the Behavior Rating Inventory of Executive Function − Preschool Version (BRIEF-P) . Of these children, 69 had a domiciliary cardiorespiratory polygraphy. As compared to the children without OSA, the 41 children with OSA scored lower for working memory, emotional control, and shifting, defined by the ability to shift attention in response to situational demands or to shift between simples response sets following clear verbal instructions. Another study in 41 children with Down syndrome correlated PSG data to neurocognitive evaluation . Some scores of the Raven’s coloured progressive matrices (CPM), which is a non-verbal test evaluating abstract reasoning and fluid intelligence, correlated negatively with the respiratory arousal index, and the obstructive AHI (OAHI) tended to correlate positively to the Reiss behavioural problems . Of the 16 patients who had a follow up visit after treatment, some displayed an improvement in their neuropsychological scores. A pilot study in 30 children with Down syndrome observed a correlation between the conceptual score of the Adaptive Behavior Assessment System® —Second Edition (ABAS) with the OAHI, showing some link between OSA severity and adaptive functioning, particularly in the area of communication skills . However, no correlation was observed between the overall adaptive score (general adaptive composite, GAC) and OSA severity . A cardiorespiratory polygraphy was performed in 22 typically developing preschoolers and 22 age-matched children with Down syndrome . Cognition was assessed by using the Mullen Scales of Early Learning, behaviour by using the Strengths and Difficulties Questionnaire (SDQ), and the MacArthur Communicative Development Inventory (MCDI) measured language level. Children with Down syndrome scored worse on all scales of cognition and behaviour but no correlation was observed with sleep disordered breathing. Surprisingly, for children with Down syndrome sleep disordered breathing was associated with increased language understanding and use of actions and gestures on the MCDI.

A study compared 19 children with Down syndrome and OSA documented on a PSG and 12 age-matched children with Down syndrome without OSA . Patients with OSA had significantly lower mean verbal intelligence score and lower cognitive flexibility. In addition, those with OSA had a significant increase in light-stage sleep at the expense of slow-wave sleep. In 25 school-aged children with Down syndrome, PSG data was correlated to a battery of neuropsychological tests . At baseline, there was no correlation between OSA and neurocognitive functioning. However, total sleep time and sleep efficiency correlated to cognitive ability and comprehension, and the time in slow-wave sleep correlated with tests of achievement and adaptive behaviours. At follow up, the 5 patients who achieved successful sleep-disordered treatment had an improvement in adaptive behaviour, visual-motor integration, achievement and attention, as compared to those who did not tolerate treatment. This study showed that sleep quality and sleep architecture, and in particular slow-wave sleep, were better predictors of neuropsychological function than OSA assessed on the AHI .

The importance of sleep quality was also observed in a cross-sectional study on 29 infants with Down syndrome with a mean age of 40 months who were compared to 20 age-matched typically developing children . Sleep quality was assessed on actigraphy during 5 consecutive days. Infants with Down syndrome were then classified in “good sleepers” or “poor sleepers” according to a sleep efficiency on the actigraphy above or below 80 %. Children with Down syndrome had a language delay as compared to typically developing children, but those with a poor sleep efficiency had significantly lower scores for a vocabulary production total score, a word total score, the percentage of combining words, sentence length, vocalization, and length of longest utterance than children with Down syndrome who had a good sleep efficiency. A study comparing 44 children with Down syndrome to typically developing children showed a lower spindle activity in children with Down syndrome . Spindle activity was not correlated to Child Behavior Check List (CBCL) subscales but correlated negatively with OSA-18 emotional symptoms and caregiver concerns and C Fast activity correlated also negatively with daytime function and total problems. For the authors, reduced spindle activity likely underpins the more disrupted sleep and may be associated with reduced daytime functioning and also an early biomarker for an increased risk of developing dementia later in life in children with Down syndrome .

Very few studies assessed the impact of OSA treatment on neurocognitive functioning in children with Down syndrome. In 23 children with Down syndrome and sleep-disordered breathing aged 5 to 16 years, treatment of sleep-disordered breathing improved sleep-disordered breathing symptoms, sleep disturbance, and quality of life but had no demonstrable impact on daytime behaviour or function . A study in 24 children with Down syndrome and sleep-disordered breathing showed that treatment of sleep-disordered breathing improved the severity of the disease on a follow up PSG. But despite an improvement of quality of life by parental reports, treatment of sleep-disordered breathing had no demonstrable impacts on sleep quality, behaviour, or daytime functioning .

To our knowledge, there are no studies assess the effect of continuous positive airway pressure on the neurocognitive development in children with Down syndrome and severe OSA. Even if some studies reported an excellent adherence to CPAP in children with Down syndrome , adherence may be difficult or impossible due to behavioural problems . In these patients, hypoglossal nerve stimulation (HNS) has shown to be an interesting alternative to CPAP. Indeed, a small case series of 9 selected adolescents and young adults with Down syndrome and severe OSA, showed that HNS was associated with a significant decrease in the AHI and an improvement in all domains of neurocognition, namely intelligence, adaptive skills, expressive vocabulary, communication working memory, attention regulation, processing speed and quality of life .

Neurocognitive dysfunction and OSA has also been studied in some adults with Down syndrome. In a study of 12 adults with Down syndrome and OSA, a significant relationship was found between the AHI and visioperceptual skills evaluated by the Raven Progressive Matrices . In another study, the parents of 29 adults with Down syndrome completed a questionnaire on OSA symptoms and neurocognitive function in their children. A negative correlation was observed between parental rating of OSA symptoms and executive function and verbal fluency . And finally, another study observed a correlation between sleep problems and associated comorbidities such as anxiety, depression and dementia .

In conclusion, studies in children with Down syndrome show inconsistent results, with a trend for a preponderance of sleep quality over OSA severity influencing neurocognitive function and behaviour. The potential benefit of OSA correction on neurocognitive improvement is still uncertain and warrants further studies, especially in young infants who are at a critical period of neurological development.

The link between OSA and neurocognitive function has been poorly studied in other children at risk for OSA type-III. A study in 21 primary school-aged children with Pierre Robin Sequence and 21 age-matched controls assessed sleep by means of a PSG and neurocognitive outcome using the Wechsler Intelligence Scale for Children-Fifth Edition . All children with Pierre Robin Sequence had an effective treatment of OSA and had neurocognitive function within the normal range. Similar results were observed in 87 children with isolated Pierre Robin sequence or Pierre Robin sequence associated with Stickler syndrome who were aged 15 months to 6 years . OSA was treated by prone position for 77 % of the patients, CPAP for 16 %, and a tracheotomy for 6 %. On the French version of the Child Development Inventory, for 71 %, the developmental evolution was within normality, 26 % were at high risk of delay, and only 3 % were at very high risk of delay. The developmental quotients for expressive language and language comprehension were lower in the patients than the general population, but an improvement was noticed with the children’s growth. Of note, children whose mothers had low education levels were more at risk than the others.

Children with cleft lip and/or palate (CL/P) are at risk of OSA and language difficulties. An observational follow-up study evaluated the neurocognitive outcome (Bayley Scales of Infant and Toddler Development, Third Edition; BSID-III), quality of life (Infant/Toddler Quality of Life Questionnaire; ITQOL), and growth at 3 years of age in 33 children with CL/P . The mean AHI in infancy was 23.9 ± 18.0 events/h. Mean group BSID-III scores fell within the standardized normal range for all domains; however, language scores were lower than control children. PSG variables in infancy correlated with outcomes at 3 years of age; a lower percentage of Rapid Eye Movement (REM) sleep was associated with a lower cognition score and more obstructive events were associated with a lower global behaviour ITQOL score. The authors concluded that neurocognition, quality of life, and growth measures from children with CL/P fall within a normal range but scores in the language domain are lower than controls. Sleep and respiratory elements of sleep-disordered breathing in infancy appear to modify these outcomes at 3 years of age.

22q11.2 deletion syndrome (22q11.DS) is a neurogenetic disorder caused by a microdeletion in chromosome 11 whose phenotype includes both physical and neuro-psychiatric manifestations. In order to analyse the link between sleep and neurocognitive impairment, subjective sleep quality was assessed by the Pittsburgh Sleep Quality Index (PSQI) and cognitive functioning was assessed by the Penn computerized neurocognitive battery (CNB) in 23 adult 22q11.DS (Di George syndrome) patients and 24 healthy controls . 22q11.2DS patients had significantly worse sleep quality scores than the controls, unrelated to the psychiatric or physical comorbidities common to 22q11.2DS. Several domains of the CNB, such as the complex cognition score and the episodic memory score, were associated with poorer sleep quality, suggesting that cognitive impairments in 22q11.2DS may be at least partially explained by poor sleep quality.

Patients with Prader Willi syndrome are at high risk of OSA and psychiatric manifestations . The link between OSA and neurocognitive function has not been studied in children but one study included some adolescents . One hundred and twenty two patients with Prader Willi syndrome, mean age 19.3 ± 8.4 years, participated in a remote (web-based) assessment of neurobehavioral traits, including psychosis-risk symptoms (Prodromal Questionnaire-Brief Version; PQ-B) and sleep behaviours (Pittsburgh Sleep Quality Index). Subjectively-reported sleep quality, nap frequency, sleep duration, sleep disturbance, and daytime dysfunction were significant predictors of psychosis-risk symptom frequency and severity and sleep disturbance ratings were the strongest predictors of psychosis-risk symptoms.

In conclusion, in numerous genetic and/or rare disorders that are associated with an increased risk of OSA, OSA but probably more importantly poor sleep quality or (micro)-altered sleep architecture are associated with an increased risk of neurocognitive impairment or abnormal behaviour. Neurocognitive delay may be an integral part of these disorders, making it challenging to distinguish what is related to the primary disease and what to the additional deleterious consequences of OSA, sleep-disordered breathing, and abnormal sleep quality. A more systematic assessment of sleep and neurocognitive function is mandatory in these children in order to diagnose and treat OSA and sleep-disordered breathing as early and as efficiently as possible. Indeed, we should not miss a potential therapeutic window for improvement in these more severely affected children with disabilities.