An alternative approach to categorization focuses on the etiology of feeding disorders, most often involving a broad dichotomization between organic and nonorganic (a.k.a., functional) precipitants (Babbitt et al., 1994; Piazza, 2008). Organic factors refer to medical concerns that precipitate or coincide with the emergence of food rejection and/or growth failure; nonorganic (a.k.a., functional) factors involve environmental events (e.g., antecedents and consequences for feeding behaviors) that shape or strengthen refusal behaviors during meals. A summary of the most frequently cited organic and nonorganic factors is presented in Table 17.2.

A dichotomy of this nature is helpful in recognizing the frequent contribution of medical concerns in the emergence of feeding concerns. Estimates suggest that 40–70 % of children with chronic medical concerns experience feeding difficulties (Lukens & Silverman, 2014). For example, a chart review involving 72 children treated for feeding tube dependence at a hospital-based feeding program reported that 83 % of the children presented with oropharyngeal or GI abnormalities and 64 % had cardiac, pulmonary, neurological, or genetic conditions (Greer, Gulotta, Masler, & Laud, 2009). Of 103 children referred to an interdisciplinary feeding team, 74 % experienced both behavioral and structural, neurological, cardiorespiratory, or neurological issues (Burklow, Phelps, Schultz, McConnell, & Rudolph, 1998). A similar pattern was reported by Sharp et al. (2010) in a summary of treatment literature, with GERD representing one of the most frequent medical concerns associated with feeding problems. Medical issues such as GERD, food allergy, gastroenteritis, and/or structural abnormalities are posited to create an association between food and aversive/unpleasant consequences (e.g., pain, nausea, and/or fatigue)—making eating something to avoid versus a pleasurable activity connected with the alleviation of hunger cues (Hyman, 1994).

The organic versus nonorganic dichotomy also recognizes that feeding disturbances may occur among children with no clear physiological precursor. This includes a significant number of children with ASD and food selectivity. Most past reports of severe feeding problems in ASD do not coincide with obvious organic factors or GI etiology (Ledford & Gast, 2006). However, children with ASD do have more GI-associated complaints than typical children, representing one of the mo st frequently cited comorbidities in this population. A recent meta-analysis indicated that children with ASD were four times more likely to experience at least one GI symptom. Children with ASD were also three times as likely to experience constipation and diarrhea and more than twice as likely to complain about abdominal pain compared to peers (McElhanon, McCracken, Karpen, & Sharp, 2014). These GI problems, however, may have a behavioral etiology, such as poor dietary diversity involving high intake of processed foods and low intake of fiber-rich fruits and vegetables. High prevalence of behaviorally based toileting concerns in ASD may also contribute to GI symptoms, including absent or delayed acquisition of bowel training. In addition, data on other GI symptoms (e.g., GERD, food allergies) typically associated with organic pathology remain insufficient and there is no evidence suggesting a unique GI pathology in ASD to account for the emergence, prevalence, and topography of food selectivity observed in ASD. This has led to the hypothesis that aberrant feeding habits among those with ASD may be a manifestation of restricted interests, behavioral rigidity, sensory sensitivity , and/or perseveration (Ahearn et al., 2001).

In general, the distinction between organic and nonorganic factors holds limited utility as a diagnostic nosology, as it is now generally accepted that (1) organic issues (when present) often operate concurrently with functional factors to maintain feeding problems; (2) most feeding disorders involve multiple causal pathways; and (3) disrupted family functioning and maladaptive patterns of reinforcement often play a central role in long-standing feeding concerns (Babbitt et al., 1994; Sharp et al., 2010). This latter point emphasizes the role of learned behaviors in promoting escape from unpleasant feeding experiences and/or gaining attention from caregivers (Piazza et al., 2003). Among children with ASD, however, the exact aversive qualities of non-preferred foods (e.g., texture, taste, temperature, color, or smell) that contribute to the emergence of food selectivity in this popul ation remain unclear.

Davies and colleagues (2006) proposed a multiaxial diagnostic framework, entitled “Feeding Disorder between Parent and Child,” to capture the broader context in which feeding problems occur. This framework was built on the argument that existing diagnostic approaches (1) fail to capture relational and multisystemic processes involved in conducting meals with children; (2) focus exclusively on child factors that contribute to disruptions in feeding; and (3) are overly concerned with exclusionary criteria (e.g., other medical, structural, or psychiatric disorders). Davies et al. sought to address these limitations by emphasizing the parent–child feeding relationship and broader characteristics in which disordered eating occurs. This includes (1) assessing both child and parent factors that contribute to mealtime difficulties (i.e., medical, developmental, and behavioral); (2) identifying key aspects of the parent–child relationship impacted by feeding difficulties (e.g., level of caregiver stress/concern, severity of interactional difficulties); and (3) determining contributing psychosocial and environmental problems/stressors (e.g., problems in caregiving, domestic violence, socioeconomic concerns).

This type of multifaceted approach recognizes that, as a relational process, disruptions in feeding may have detrimental outcomes beyond child factors (e.g., growth concerns) that are often the focus of clinical attention. This is particularly salient for children with ASD given existing concerns regarding high caregiver burden in this population (Fletcher, Markoulakis, & Bryden, 2012). Chronic feeding difficulties and related dietary concern s represent an additional source of strain on quality of life (Khanna et al., 2012), increasing child-rearing needs, parental stress, and social isolation. Children with ASD often exhibit a strong emotional response when presented with non-preferred food, including crying, disruption, and aggression (Sharp, Jaquess, & Lukens, 2013). As a result, severe food selectivity and related behavioral concerns often necessitate caregivers preparing multiple menus for each meal—one plate for the child with ASD and a separate menu that reflects the family’s diet. For children whose behavioral disruptions occur in response to the sight or smell of non-preferred foods, they often cannot sit at the table with the family and peers and thus miss out on opportunities to learn and enjoy social engagement (Nadon, Feldman, Dunn, & Gisel, 2011). Families are also more likely to miss organized activities (e.g., birthdays; family gatherings) that involve eating and experience reduced opportunities to eat at restaurants or social occasions , resulting in further isolatio n (Sharp, Berry et al., 2013).

The new psychiatric diagnosis Avoidant/Restrictive Food Intake Disorder, as outlined by the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5) (APA, 2013), provides a more comprehensive framework for capturing the heterogeneity of pediatric feeding disorders compared with previous formal diagnostic systems. The main diagnostic feature of the disorder involves avoidance or restriction of food intake as reflected by failure to meet nutrition and/or energy needs. The criteria specify that this may manifest as one (or more) of the following clinical indicators : (1) significant weight loss, (2) significant nutritional deficiencies, (3) dependence on enteral feeding or oral nutritional supplementations, and/or (4) marked interference with psychosocial functioning. A notable strength of this new definition is movement beyond a singular focus on growth failure to capture significant feeding disturbances, astutely recognizing that not all children with disordered eating will present with weight concerns. This is particularly salient for children reliant on artificial supports (e.g., a feeding tube) to meet their energy requirements, as well as cases where dietary variety (vs. volume) is the primary feeding concern—as is often the case in ASD. Other strengths include consideration of the broader impact on psychosocial functioning as argued by Davies et al. (2006) and recognition of the potential role of traumatic or painful events in conditioning food aversion, with a specific reference to medical conditions involving the GI tract.

With these strengths in mind, the diagnosis remains broad and nonspecific in regard to feeding topographies, which is a frequent criticism of formal diagnostic systems (Piazza, 2008; Sharp et al., 2010). It also provides minimal guidance regarding how to navigate the diagnostic process for certain at-risk populations. For example, determination of what constitutes significant weight loss or nutritional deficiency is left up to clinical judgment. Further, while recognizing rigid eating patterns and heightened sensory sensitivity as prominent in ASD, it also notes that the level of impairment may not meet the diagnostic threshold, although the only criteria specified is whether the eating disturbance requires specific treatment. Unfortunately, no guideline exists for determining when feeding intervention is warranted in pediatric populations. As such, avoidant/restrictive food intake disorder is best viewed as casting a more comprehensive diagnostic net with limited clinical utility for guiding the assessment proces s.

Compromised gross anthropometric parameters (i.e., height, weight, and body mass index [BMI]) is the most salient symptom of a feeding disorder likely to trigger attention in pediatric settings (Sharp, Berry et al., 2013). The use of anthropometrics as a primary clinical indicator makes pragmatic sense because height and weight are typically obtained as part of routine clinical care and assuring adequate energy intake is a critical consideration in supporting appropriate growth and development. The use of faltering growth as a proxy for a possible feeding disorder, however, biases detection towards children with food refusal vs. food selectivity and may help explain why feeding concerns in ASD are often overlooked in relation to other areas of clinical concern (McElhanon et al., 2014). Among children without ASD, feeding disorders most often involve severe restrictions in the volume of food consumed, leading to artificial supports (e.g., a feeding tube; oral formula supplementation) to support growth. In contrast, severe food selectivity in ASD most often involves deficits in dietary variety, not volume, and children with ASD typically consume enough food to meet gross energy needs. For example, Sharp, Berry et al. (2013) identified a pool of seven studies (involving 426 children with ASD) presenting information on growth status compared with typically developing peers. All seven studies reported significantly higher rates of feeding problems in children with ASD, but no statistically significant difference in growth status between groups. This pattern also holds true for children with ASD receiving intervention for food selectivity . Sharp, Jaquess, Morton, and Miles (2011) reported that only 2 out of 13 (15 %) children with ASD enrolled in an intensive day treatment program fell below the 5th percentile (weight for height). Similarly, Laud, Girolami, Boscoe, and Gulotta (2009) reported that only 7 out of 46 children (15 %) admitted for intensive feeding intervention met criteria for failure to thrive.

In general, it appears that most children with ASD and food selectivity are able to maintain at least minimally adequate anthropometric parameters despite restricted dietary variety. In fact, food selectivity in ASD may actually involve excessive intake of calories in some cases. This highlights the need to look beyond faltering growth as a means to quantify the impact of atypical patterns of intake in ASD. Evidence suggests that food selectivity in ASD places this population at risk for long-term nutritional or medical complications not captured by broad anthropometrics or analysis of overall energy intake. This includes vitamin and mineral deficiencies (Sharp, Berry et al., 2013) and compromised poor bone growth (Hediger et al., 2008). Selective eating patterns (e.g., complex carbohydrates and fats) may also increase the risk for diet-related diseases (e.g., obesity, cardiovascular disease). In a sample of 273 children with ASD, Egan et al. reported that 21.9 % had a body mass index (BMI) in the obese range, a rate that is higher than a nationally representative sample. Curtin et al. reported that the prevalence of obesity in children with ASD was 30.4 % compared with 23.6 % of children without ASD—corresponding to a 1.42 increased odds of obesity in this population. A recent large-scale chart review suggests that this trend extends into adulthood (Croen, Zerbo, Qian, & Massolo, 2014). When compared to non-ASD peers, adults with ASD experienced a 69 % higher incidence of obesity, 42 % greater risk of hypertension, and 50 % increase in diabetes. With the prevalence of ASD estimated at 1 in 68 children (CDC, 2014), high prevalence of feeding concerns and associated health concerns in ASD intensifies the need to develop and refine methods for detecting and remediating food selectivity in this population.

A summary of key research findings regarding feeding problems among children with and without ASD is presented in Table 17.3. Among children with ASD, the comparison highlights (1) high prevalence of food selectivity; (2) low probability of faltering growth; (3) enhanced risk of nutritional deficiencies and/or excesses; and (4) lack of evidence for medical concerns to account for the pattern and prevalence of feeding difficulties in this population. It also emphasizes the presence of significant mealtime behavior problems (e.g., crying; severe tantrums) in both groups and related caregiver stress; however, problem behaviors tend to be isolated to the presentation of non-preferred foods in ASD. Finally, there is decreased likelihood of significant experienced-based oral-motor deficits in this population. Children with ASD likely consume chewable foods, often in the form of table texture snacks and processed foods (e.g., crackers, chips, chicken nuggets). As a result, they are more likely to possess foundational oral-motor skills (e.g., adequate variety of tongue movement concomitant with mastication to safely move the food bolus to swallow) necessary for processing higher texture foods, but may experience difficulty with generalizing these skills to non-preferred foods, such as fruits and vegetables, that represent a significant texture change from preferred foods.

A breakdown of this nature is intended to provide a general roadmap to guide the assessment process , with each of these major areas—behavior, nutrition, oral-motor, and medical—representing key considerations when assessing a feeding disorder in ASD. It is not, however, meant to imply that all children with ASD will fit this nomothetic pattern, nor should it be viewed as discounting the importance of a detailed medical or oral-motor examination among children with ASD and food selectivity. On the contrary, medical screening and assessment of oral-motor function should be viewed as central to the assessment process given the high association between organic issues and feeding disorders in other pediatric populations. Case in point, food refusal and feeding tube dependence , particularly among cases involving a history of GERD, have been described in past reports of children with ASD. With this in mind, this probabilistic description affords scaffolding for a more detailed evaluation to determine the topography, etiology, and potential impact of atypical intake in ASD.

A common theme throughout the assessment process is the unique challenges associated with ASD, most notably communication barriers and intense emotional responses . This necessitates increased reliance on caregiver report and adaptation to existing methodologies. The importance of a multidisciplinary continuum of care is also emphasized by this model, with distinctive contributions from behavioral psychology, nutrition, speech language pathologist (SLP) or occupational therapist (OT), and medicine to fully capture the diagnostic complexity of a feeding disorder in ASD.