Journal of Psychological Abnormalities

Autism Spectrum Disorder (ASD) is characterized by a significant impairment in social interaction and communication, a wide range of rigid behaviours and sensory disintegration. These symptoms make ASD one of the most devastating neurodevelopmental disorders, causing major dysfunction in the individual’s daily living and having a negative impact on the families of affected children [1-3].

According to the latest update by the Centers for Disease Control and Prevention (CDC) in the US, about 1 in 68 children suffers from ASD, the rising trend of prevalence has also been reported worldwide [4]. However, in Hong Kong (HK) there is a lack of well-designed population-based prevalence and epidemiological study on ASD. Here, over 90% of children aged less than 2 years old attend Maternity and Child Health Centers (MCHCs) to receive vaccinations and developmental surveillance (DSS). Children are referred to the tertiary Child Assessment Center (CAC) for further evaluation of suspected developmental problems. Children with ASD are diagnosed by the paediatrician or clinical psychologist according to the standards of Diagnostic and Statistical Manual of Mental Disorders, 4^th Edition (DSM IV). An annual report by the Child Assessment Services (CAS) from the Department of Health revealed that there was an increase of 184% in ASD diagnosis from the years 1997 to 2005 [5]. Though CAS captured a great proportion of children diagnosed with ASD, their data remains an underestimate of the true prevalence in Hong Kong. This is due to quite a number of children nowadays being diagnosed in the private sector, and there are certainly children attending mainstream schools without a diagnostic label being given.

Preschool aged children who are diagnosed to have ASD are placed in rehabilitation services according to the degree of severity, where attention is focused mainly on behavioural, speech and physical management. No regular medical follow-up is given if no definitive cause is identified for this group of children. Once they are allocated a place for training, little contact will be made with paediatricians, psychiatrists or general practitioners. Hence, medical child health issues are overlooked. Our group looked into the health issues of this group of children and published a report on the dental health problems in children with ASD in Hong Kong [6]. The study found that the occurrence of dental caries was significantly higher in children of mothers with a low educational level and in children from low-income families.

Feeding is another mandatory health area affecting early childhood health and behavioural development for survival. Atypical eating behaviours and problems have become a growing concern among parents of children with ASD [1,7-9], Ledford and Gast reviewed 381 children with ASD from seven studies: all studies showed that food selectivity was the most significant feeding problem [8]. The wide variety of methodologies in these studies explains the wide-ranging prevalence rate of feeding difficulties from 46% to 89%. Typical feeding development was greatly affected by such eating behavioural difficulties.

Nutritional concerns about restrictive feeding behaviour were underinvestigated. Conflicting results were reported due to the lack of comparison groups and inconsistent dietary records. The latest metaanalysis by W.G. Sharp et al. reviewed 17 eligible studies, analyzing comparison groups of children with ASD and found that the subjects had significantly less intake of calcium and protein [9].

The feeding behavioural problems of children with Autism Spectrum Disorder appear to be relatively common in Hong Kong, though no reliable statistics are available. Thus, the present study aims to examine the prevalence of feeding problems in preschool children with ASD in HK and to arouse public and medical awareness of the health issues of this specific group of children.

Participants

All children aged 2 to 6 years diagnosed with confirmed or suspected Autism, Autistic features, Autism Spectrum Disorder, Asperger Syndrome, Pervasive Developmental Disorder by developmental paediatricians or clinical psychologists from public or private sectors based on the DSM-IV, were invited from 13 rehabilitation-training centres (Early Training Centers and Special Child Care Centers) of Heep Hong Society in the year of 2012-2013. Heep Hong Society is one of the rehabilitation organizations in Hong Kong. It operates 13 Special Child Care Centers (SCCC) for preschool children, and makes up one third of all SCCCs in Hong Kong. Written consent was obtained from parents. Basic demographic data was collected.

The brief autism mealtime inventory (BAMBI) - Chinese version

The Brief Autism Mealtime Inventory (BAMBI) is the standardized measure developed for mealtime behaviour specific to the ASD population. It consists of 18 items, and is a parent self-report questionnaire. A frequency score can be calculated using the total score of the 18 Likert responses. Higher scores represent more problematic mealtime behaviour. The BAMBI has three domains: ‘Limited Variety’, ‘Food Refusal’ and ‘Features of Autism’. The ‘Limited Variety’ domain consists of eight items related to limited food preferences, ‘Food Refusal’ has five items related to rejection of food, and the ‘Features of Autism’ has five items related to behavioural characteristics or associated features of Autism. It indicates how often their child engages in a particular eating behaviour with response options ranging from 1 for ‘never’ to 5 for ‘always’ with a neutral midpoint. The internal consistency for the 18 items questionnaire of BAMBI was reported as high (Cronbach’s coefficient alpha = 0.88), with good test-retest reliability (r = 0.87, p < 0.01) and inter-rater reliability (r = 0.78, p < 0.01) [10]. BAMBI has been translated and back translated into the Chinese version. A total frequency score was obtained. Caregivers indicated their perception of problems in daily living according to each symptom by answering “Yes” or “No” to each question.

Statistical analysis

Descriptive data was presented as percentages for discrete variables and as the mean for continuous variables. Trends were identified using cross-tabulation to assess the linear-by-linear relationships among different ethnic groups of parents who reported problem eating behaviours. All statistical analyses were performed using PASW statistics 21.0 (SPSS Inc., Chicago, IL, USA). All statistical tests were two-sided and a P-value 0.05 was considered statistically significant.

Ethics

The present study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving human subjects/patients were approved by the Joint Chinese University of Hong Kong and New Territories East Cluster Clinical Research Ethics Committee.

A total of 194 questionnaires were collected from 13 Heep Hong centers, 177 of them were valid with less than 4 unanswered items. All children recruited were Hong Kong Chinese, of which 78.0% were boys and 22.0% were girls with the mean age of 5.0 (2.9-6.3) (Table 1). The distribution of the family income, paternal and maternal education level were all comparative to the Hong Kong general population as reported by the Hong Kong Statistics Department [11].

Table 1: Characteristics of the pre-schoolers with autism.

The survey results were compared with independent data from Japan and Indonesia. Among the 177 Hong Kong, 31 Japanese and 13 Indonesian parents reported eating problem behaviours of their Autistic children in terms of ‘Limited Variety’, ‘Food Refusal’ and ‘Features of Autism’ in similar proportion with no statistic difference demonstrated (Table 2) [12].

Table 2: Parents report of eating behaviors as problem.

BAMBI mean raw scores

The mean raw scores of the three feeding behaviour domains were: “Food Refusal” 8.1 (+/- 2.9); “Features of Autism” 9.7 (+/- 2.6); “Limited Variety” 19.9 (+/- 4.9). The highest mean raw score belonged to factor of “Limited Variety” of food. Frequent scores higher than 3 indicating “very frequent” or “always” were classified as significant.

Symptoms of “limited variety of food”: In the Hong Kong study, there were a number of interesting findings. Thirty-nine percent of the children were reported as being unwilling to try new foods. 42.4% of children disliked certain foods and refused to eat them. A quarter of them did not accept variety of food. A quarter of the subjects preferred eating crunchy foods. Ten percent of children preferred to have food served or prepared in particular way and preferred only sweet foods. Close to half (45.8%) of the caregivers found the symptoms of “limited food variety” problematic to their daily functioning.

Symptoms related to features of autism: There were 38.9% of children reported to have problems remaining seated for the completion of eating a meal. A quarter of them were not flexible about mealtime routines. Twenty percent of them refused food that required a lot of effort in chewing. Ten percent of them were aggressive during mealtimes and 5% of them would display self-injurious behaviour during mealtimes. A quarter to 30% of caregivers found that these symptoms were problematic to their daily living.

“Food refusal”: A quarter of the subjects were reported as turning away from food and 22% would spill out food that they had eaten. Thirteen percent to 17.5% of them would close their mouths tightly when food was presented, cry or scream and were disruptive during mealtimes.

The BAMBI score was significantly and positively correlated with the number of feeding behaviours that caregivers found problematic, i.e., perception of problems were correlated with the frequency of eating behaviours (p < 0.001). As reported, 34% of our subjects required at least 2 caregivers for supervision during the mealtime period. The severity of feeding behaviour was not correlated to family income, paternal or maternal educational level. The duration of training in Heep Hong centers also did not correlate to the BAMBI frequency score and the perception of problems by their caregivers.

As expected, Hong Kong Chinese preschoolers with Autism had poor food variety scores. Some of them appeared to be flexible in their food choices, while others were much more limited and seemed to restrict their intake. Our findings are also consistent with previously published reports on problematic eating behaviours: there is a high prevalence of limited variety of food behaviour such as unwillingness to try new food, a dislike of certain food and refusing to eat them [7,12-15].

Our findings were confirmed by BAMBI that Japanese, Indonesian and Hong Kong Chinese preschoolers had similar eating behaviour problems [12]. The prevalence of mealtime behavioural problems reported in Japanese and Indonesian preschool children with ASD was similar to our findings in Hong Kong regardless of their ethnic origin: the insistence on food being prepared in a certain way; difficulty remaining seated until the meal was finished; and inflexible mealtime routines.

Difficulties with social interactions are also characteristic of Autistic children. Lack of age-appropriate social exchanges and opportunities to observe role models on appropriate mealtime behavior may make it difficult for a child with Autism to learn the socially appropriate use of utensils and self-feeding skills [9]. The combination of these factors and communication deficits may make it difficult for a child to maintain a nutritionally adequate diet. Although profound growth retardation are not reported, diet quality and nutrient intake may be affected by this kind of disordered eating behaviours [16]. In addition, their rigidity and repetitive behaviour leading to food restrictions might contribute to long term nutritional consequences. A recent American study reported the frequency of selective eating and nutritional deficiency among 22 children with Autism with mean age of 8 and among an age-matched typically developing control group [17]. Children with Autism ate fewer food varieties on average than typically developing children (33.5 vs. 54.5 foods, p < 0.001). It suggested that selective eaters with Autism are at greater risk of inadequate nutritional intake, the findings of the majority of these studies are consistent with nutrient intake compared to non-selective eaters with Autism and a control group. Conclusions drawn from these descriptive reports of limited nutrition and high levels of behavioral feeding problems in children with Autism were limited by the variety of methodology and limited sample size.

Children with Autism demonstrate a wide range of more problematic eating behaviours than the general population [10,18]. The reason for selective eating habits among children with Autism has not been thoroughly investigated, but rigid adherence to rituals and routines seen as a core feature of autism is one hypothesized explanation. Sensory integration dysfunction may also play a role in problematic eating behaviours. Bennetto et al. found that children with ASD have problems correctly identifying taste and olfactory sensations suggesting that over or under responsiveness to sensory stimuli may also contribute to the high prevalence of feeding difficulties among this population [19]. This is an area that requires further investigation.

In our study, nearly 30-50% of parents were very stressed by the feeding mealtime behaviour. Every day, the preschool children would have 3 main meals and 2 snack times, carers or parents needed to encounter daily cumulative stress on their aggressive mealtime behaviour including screaming, crying and disruptive behaviours. Moreover, 34% of the children need at least two caregivers at each meal which indicates the mandatory need for intervention to this specific targeted feeding problem in order to save the psychological integrity of both the child and the one who cares them. These kinds of behaviours are not related to training duration or family background. Hence, children with ASD should be screened or evaluated for feeding mealtime behaviour and should be specifically referred to a multidisciplinary feeding program which should be under the supervision of physicians, nutritionists, psychologists, occupational therapists and speech therapists.

There are several limitations to this descriptive study. Although all recruited subjects were suffering from Autism Spectrum Disorder, they were all attending special child care centers which only provided rehabilitation services to children with a moderate to severe degree of the disorder. The relationship of eating mealtime behaviour might not be fully reflected for children with milder ASD. Furthermore, in terms of the overall reliability of BAMBI, some parents failed to complete all the questions; 91% of the questionnaires were valid with less than 4 unanswered items. The generalization of these results may be questionable. In addition, the children in this study were all referred from one single non-government organization within one single year, so it is unclear how this sample reflects the larger population of children with feeding problems. Further larger scale study recruiting all children with different degrees of ASD severity would be worth conducting.

BAMBI (Chinese Version) can be used on the Chinese population to assess the feeding and mealtime behaviour of children with ASD. Limited variety of food is the most frequent presentation of feeding problems in Hong Kong preschool children with ASD and causes a lot of stress to the caregivers. These feeding behavioural problems are not improved by current training. We recommend concentrating specifically on addressing the feeding behaviour though a multidisciplinary approach and by incorporating dietary advice, which will be beneficial to children with ASD and their families.

We would like to thank all the children, parents and professionals who participated in this study.