Children with ASD form a heterogeneous group, with varying degrees of difficulties in social interaction, social communication, and rigidity, and with varying psychiatric comorbidity (APA
2013). A relation between ASD and feeding problems has been identified early in the history of ASD research (e.g. Kanner, 1943 in Ledford and Gast
2006). A systematic review by Sharp et al. (
2013a,
b) estimated that children with ASD have a fivefold probability of having feeding problems compared to children without ASD. Roughly 44% to 89% of children with ASD have feeding problems (Cermak et al.
2010; Seiverling et al.
2018). For this reason, it has been argued that the presence of severe, atypical or chronic feeding problems should alert professionals in the direction of underlying ASD (Keen
2008), in children as well as in adults (Dell’Osso et al.
2018). Also, the co-authors of this article report from clinical experience that a-specific symptoms, such as problems with sleeping and eating, are often the first to emerge in young children, in many cases before the onset of more specific symptoms of ASD. This illustrates how both scientific studies and clinical experience suggest that feeding problems are frequently comorbid with ASD.
Definitions and Symptoms
Systematic reviews on feeding problems in ASD have provided robust evidence of significant feeding problems in this population (Ledford and Gast
2006; Sharp et al.
2013a,
b). However, there is currently no consensus on terminology with regard to definitions of feeding problems and their symptoms. Terms that are often used are ‘neophobia’ (systematic rejection of novel food), ‘picky eating’ (low appetite, fussy behavior or sensory problems), ‘feeding disorder’ (cases with nutritional, organic or emotional consequences) and ‘feeding difficulties’ (an umbrella term indicating that there is a problem with feeding of some sort) (see Kerzner et al.
2015). Generally, the terms ‘feeding difficulties’ and ‘feeding problems’ are commonly used to specifically refer to problems in (early) childhood which do not necessarily lead to significant nutritional deficiencies, weight loss, or psychosocial problems for the child (yet).
The most prevalent symptom for children with ASD is
food selectivity (Cermak et al.
2010; Marí-Bauset et al.
2014), which means that they can be very fastidious about which foods they accept and which they refuse. In a study by Seiverling and colleagues (
2018), a sample of children with ASD showed more food selectivity by texture and type than a non-ASD sample with language delays (around 24% compared to below 10% in the non-ASD sample). According to caregivers, selectivity can be triggered by various characteristics of the food, such as texture, color, taste, smell and temperature (Williams et al.
2000). It is also known that a lack of food variety—which is the result of food selectivity—is predominant (Nadon et al.
2011). Food selectivity is evident as early as 15 months and although it generally decreases with age, it can persist into adolescence and even adulthood. Research indicates that many children with ASD also have a preference for food with a softer texture (Schreck et al.
2004). In general, these children prefer sweet or fatty foods and refuse fruit and vegetables (Suarez et al.
2014; Vissoker et al.
2019).
Another category of feeding problems in children with ASD is related to their
feeding behavior. Parents of children with ASD report examples of disturbing behaviors, such as walking away from the table, whining, and yelling (Matson and Fodstad
2009). Other problems that have been mentioned are throwing or dumping foods, difficulty eating at restaurants, and also aggression and tantrums during eating (Provost et al.
2010). Children with ASD often have rituals around food (Vissoker et al.
2015) and show signs of anxiety (Kim et al.
2000). Eating anxiety can be related to compulsive rituals, avoidance of certain types of food (Twachtman-Reilly et al.
2008), and the insistence on specific methods of food preparation, food types and mealtime rules (Matson and Fodstad
2009; Zandt et al.
2007).
An additional kind of feeding problems that is reported is related to an
atypical way of eating, which involves gagging (Provost et al.
2010), pica, overeating (Vissoker et al.
2015), rapid eating, vomiting, regurgitation, and rumination (Nicholls and Bryant-Waugh
2009; Seiverling et al.
2018). In general, it should be stressed that although feeding problems are generally more prevalent in children with ASD, their degree, expression and development are very idiosyncratic.
Causes and Consequences
Several factors have been identified to contribute to the etiology of feeding problems in children with ASD. First, the stereotypical and rigid behavior and interests which are characteristic of the disorder have often been associated with the emergence of feeding problems (Vissoker et al.
2015). For instance, Johnson and colleagues (
2014) reported a strong association between parental report of feeding problems and the severity of stereotypical behavior. This may also be connected to specific food preferences and idiosyncratic requests and rules these children often have, with regard to the preparation and presentation of the food and choice for particular commercial brands (Ahearn et al.
2001). Children with ASD often have problems with changes in situations and contexts and have a need for consistency. Because meals and eating situations often vary from day to day, this can be challenging for these children (Kuschner et al.
2015).
A second factor is the existence of sensory processing problems in children with ASD (Nadon et al.
2011; Niedźwiecka et al.
2020; Zobel-Lachiusa et al.
2015). The epidemiological-based study of Jussila and colleagues (
2020) reported that the prevalence of sensory abnormalities was almost 54% in a sample of 8-year old children with ASD. Clinic-based studies have provided estimations between 69 and 95% (Baranek et al.
2006; Dellapiazza et al.
2018; Tomchek and Dunn
2007). Epidemiological studies screening for ASD probably have a lower percentage due to a larger group with less severe ASD symptoms than participants in clinical studies (Jussila et al.
2020). In any case, sensory problems are common in children with ASD and can be related to the emergence of feeding problems (Vissoker et al.
2015). This is particularly the case in relation to an increased sensitivity to certain texture of the food, which contributes to increased selectivity (Cermak et al.
2010; Johnson et al.
2014; Lane et al.
2010). For instance, Smith and colleagues (
2005) reported that in a sample of children with ASD, children with greater sensory hypersensitivity showed more problems with textures of the food and showed more food refusal than children without problems in sensory processing. A longitudinal study of Suarez et al. (
2012,
2014) showed that this relation between sensory processing problems and feeding problems is stable across a 2-year period. Sensory processing problems lead to discomfort during eating which may cause anxiety. Suarez et al. (
2014) argue that the stereotypical behavior may stem from an attempt to make the meals more predictable and to reduce anxiety. Smith (
2016) investigated the patterns of sensory processing with the Sensory Profile and found six out of eight sections to be predictive of eating behaviors in children with ASD. This included tactile, taste/smell, and movement sensitivity, but also under-responsiveness and low/weak energy, suggesting that the total spectrum of sensory processing abnormalities can be associated with feeding abnormalities.
Third, it has been suggested that problems in social communication also contribute to the development of feeding problems in children with ASD. These children are generally less socially motivated to participate in meals (while mealtimes are typically social in nature), are less sensitive to verbal reinforcement and are less competent to imitate positive feeding habits (Johnson et al.
2014). For instance, Postorino et al. (
2015) reported that in a sample of children with ASD, children with more limitations in social communications had more feeding problems than children with fewer of such limitations. The problems in social communications may contribute to disruptive mealtime behavior. It may also be speculated that the combination of processing both sensory and social stimuli at the same time lead to feeding problems. Other contributing factors to the etiology of feeding problems in children with ASD are gross motor and fine motor impairments (Kaur et al.
2018), intellectual disability (Matson et al.
2012), and gastro-intestinal dysfunction, which are all more prevalent in children with ASD than in the general population (Williams et al.
2000).
Detecting Feeding Problems
Consequences of persistent feeding problems in childhood should not be underestimated. In the general population, these problems increase the risk of malnutrition (such as undernutrition), suboptimal or stunted growth, and developmental and cognitive delays (Sharp et al.
2013a,
b). In the case of ASD, the study of Hyman et al. (
2012) reported that the majority of children are deficient in fiber, choline, calcium, and vitamin D and K. Kral et al. (
2014) reported increased waist circumferences and waist to height ratio, putting children with ASD at risk for being overweight. Also, feeding problems do not only have negative consequences for the children themselves, but also for their parents. Caregivers of children with feeding difficulties show higher levels of parental stress (Jones and Bryant-Waugh
2012; Sharp et al.
2013a,
b), including parent–child conflict and parent-spouse stress (Kuschner et al.
2017).
Because of the prevalence of feeding problems in children with ASD and their negative consequences for all concerned, pediatricians and health care professionals should be alerted to detect feeding problems in early childhood in order to timely refer for further diagnosis and treatment. Feeding problems are complex in nature and involve many different aspects—such as oral motor skills, feeding history, and feeding behavior—which should be taken into account for diagnosis and treatment (Sanchez et al.
2015). Diagnostic procedures in clinical practice usually consist of reviewing anamnestic information and other parental reports, and conducting a physical examination and a feeding observation (Arvedson
2008). Without adequate diagnosis and treatment, feeding problems tend to persist throughout childhood (Suarez et al.
2014) and adulthood, and may in some cases lead to eating disorders (Westwood and Tchanturia
2017). Early detection can be accomplished by using a screening instrument based on caregiver-reported feeding problems during routine check-ups in the general population. Many existing questionnaires—such as the
Behavioral Pediatrics Feeding Assessment Scale (BPFAS; Crist and Napier-Phillips
2001), the
Children’
s Eating Behavior Questionnaire (CEBQ; Wardle et al.
2001), and the
Mealtime Behavior Questionnaire (MBQ; Berlin et al.
2009)—are less suitable for quick identification because of their length. For this reason, the
Montreal Children’
s Hospital Feeding Scale (MCH-FS) was developed consisting of only 14 items (Ramsay et al.
2011). The questionnaire is based on the observation that clinical and non-clinical groups show similar behaviors around feeding, but that children with feeding difficulties show these behaviors at a higher frequency (Crist and Napier-Phillips
2001). The questionnaire measures seven main constructs: parental concern, family reactions, compensatory strategies, appetite, mealtime behaviors, oral sensory behavior, and oral motor behavior. The MCH-FS has been validated for French, English and Dutch children, and has been demonstrated to have good sensitivity and specificity (Sanchez et al.
2015). The Dutch normative study on children between the ages of 6 months and 3 years showed that the Cronbach’s alpha was 0.84 for the total score, suggesting a robust internal consistency. Evidence was found for a meaningful latent variable structure with two factors—(1) negative mealtime behaviors and (2) negative causes and consequences—but the high correlation between these two factors suggested that a one-factor solution is sufficient for rapid identification of feeding problems. Earlier studies have investigated the use of the MCH-FS in various clinical groups, such as Down Syndrome, children with premature birth and children with cleft palates. The question is whether this general instrument is also useful for children with other types of problems, such as ASD.
It may be the case that feeding problems in children with ASD manifest themselves differently than in the general population. For instance, these children’s preferences and rigidities are argued to be more persistent than in the general population (Bandini et al.
2010; Schreck and Williams
2006). For this reason, the
Brief Autism Mealtime Behaviorist Inventory (BAMBI) was developed as a short, standardized questionnaire specifically designed to assess feeding problems in children with ASD (Lukens and Linscheid
2008). The instrument aims to assess symptoms unique to the population, such as ritualistic and repetitive behavior during meals (Schreck and Williams
2006) and sensory feeding problems (Cermak et al.
2010). The 18-item version was shown to reflect three underlying factors: limited variety, food refusal and features of autism (Lukens and Lindscheid
2008). The third factor is scored rarely in typically developing children, making the instrument less suitable for use in the general population. In a more extensive study on the BAMBI, the number of items was reduced to 15 with four underlying factors: food selectivity, food refusal, disruptive mealtime behaviors, and mealtime rigidity (DeMand et al.
2015), and according to the authors, this should make the instrument suitable to be used more broadly as a screening measure. However, since the BAMBI was developed specifically for children with ASD and the 4-factor structure is intended to have clinical utility, it may not be the first choice to be used in the general population.
When inspecting the items of both questionnaires, there seems to be a large overlap between the MCH-FS and the BAMBI with regard to assessing sensory motor problems, food refusal, negative mealtime behavior and selectivity. For this reason, we aimed to address the question of whether the MCH-FS, as a short general screener of feeding problems, is also useful for the detection of feeding problems in children between the ages of 1 to 6 years with ASD.
There are indications that in the general population, symptoms of feeding problems ‘build up’ in early childhood as a function of age. For instance, it has been shown that parents of toddlers report a higher incidence of feeding problems than parents of infants (Wright et al.
2007). The problems tend to exacerbate due to the complex interactions between physical, psychological and social factors (e.g. Field et al.
2003; Lindberg et al.
1996; Rommel et al.
2003). The finding that feeding problems seem to increase with age was also corroborated in the normative study for the Dutch version of the MCH-FS in children between the ages of 6 months to 3 years (Van Dijk et al.
2011). It is unknown whether such an age effect also exists in children with ASD, but it seems that this may be much less the case. For instance, the validation study of the BAMBI by Castro et al. (
2019) with a large sample of children between ages 7 and 11 years, did not show any age effect. The effect of age was also non-significant in the study of Leiva-García et al. (
2019) on the association between feeding problems and oral health status in children with ASD between the ages of 6 and 18 years. In a study of Allen et al. (
2015) in children between ages 2 and 5 years, feeding problems were shown to be associated with ASD symptoms, behavior problems, sleep problems, and parenting stress, but not with child age. In contrast, some studies did report age effects. For instance, in the study of Nadon et al. (
2011) in which children aged 3 to 12 years with ASD were compared with their typically developing peers, older children had fewer problems than younger children. In addition, the study by Beighley et al. (
2013) reported a slight downward trend for food selectivity severity across childhood for a combined sample of children with different levels of ASD severity. The study by Vissoker et al. (
2015) showed that children with ASD between 3 and 7 years of age showed more ritualistic behavior around food than children between 2 and 3 years. However, other eating problems (such as chewing and swallowing problems, avoidance, and food selectivity) did not show any age-related differences. In conclusion, although only few studies address age-related changes in feeding problems in children with ASD, there is some evidence suggesting that the effect of age in children with ASD is marginal or small compared to that in typically developing children. The existence of such systematic age differences is relevant for the use of any general screening instrument. If certain problems occur more frequently in any specific age group, it may indicate that these are to some extent developmentally appropriate and transient. Previous studies suggest that feeding problems in typically developing children are somewhat more prevalent around the age of 2 and 3 years. The question is whether such a trend is also visible in the development of children with ASD. Considering the various differences between the two groups relevant to eating characteristics, this is not expected to be the case.
Because of the prevalence of feeding problems, healthcare providers are encouraged to include screening of feeding problems as part of routine medical examinations, both in the general population of young children (Ramsay
2011) and children with ASD (Sharp et al.
2013a,
b). The aim of the current study is to evaluate the use of the MCH-FS in a sample with ASD.