Comparing behavioral treatment of feeding difficulties and tube dependence in children with cerebral palsy and autism spectrum disorder

Abstract

BACKGROUND:

Feeding disorders are multifaceted with behavioral components often contributing to the development and continuation of food refusal. In these cases, behavioral interventions are effective in treating feeding problems, even when medical or oral motor components are also involved. Although behavioral interventions for feeding problems are frequently employed with children with autism, they are less commonly discussed for children with cerebral palsy.

OBJECTIVE:

The purpose of this study was to compare the effectiveness of using applied behavior analytic interventions to address feeding difficulties and tube dependence in children with autism and children with cerebral palsy.

METHOD:

Children ages 1 to 12 years who were enrolled in an intensive feeding program between 2003 and 2013, where they received individualized behavioral treatment, participated.

RESULTS:

Behavioral treatment components were similar across groups, predominately consisting of escape extinction (e.g., nonremoval of the spoon) and differential reinforcement. For both groups, behavioral treatment was similarly effective in increasing gram consumption and in decreasing refusal and negative vocalizations. A high percentage of individualized goals were met by both groups as well as high caregiver satisfaction reported.

CONCLUSIONS:

Behavioral interventions for food refusal are effective for children with cerebral palsy with behavioral refusal, just as they are for children with autism.

Keywords

Cerebral palsy autism spectrum disorder inappropriate mealtime behavior food refusal pediatric feeding disorders tube dependence

1 Introduction

Feeding difficulties are commonly comorbid with cerebral palsy (CP), with up to 89% of individuals with CP exhibiting some degree of feeding difficulty (Fung et al., 2002; Reilly, Skuse, & Poblete, 1996; Sullivan et al., 2002). Level of motor impairment, often measured by the Gross Motor Function Classification System (Wood & Rosenbaum, 2000), is a good indicator of the presence and severity of these feeding difficulties (Fung et al., 2002; Sullivan et al., 2002; Weir et al., 2013). Oral motor deficits are prevalent in children with CP, with diagnoses of dysphagia common (Benfer et al., 2013; Santoro et al., 2012). Because of this, individuals with CP may present with hypotonia, weak sucking, persistent tongue thrusting, poor lip closure, poor saliva control, decreased tongue lateralization, immature chewing, and aspiration (Arvedson, Rogers, Buck, Smart, & Msall, 1994; Parkes, Hill, Platt, & Donnelly, 2010; Reilly & Skuse, 1992; Yilmaz, Basar, & Gisel, 2004), all of which may significantly impact oral feedings.

Likely due to the medical and oral motor relationship between CP and feeding, assessment and treatment of feeding disorders in individuals with CP has largely taken a medical and oral motor standpoint to date (Dahlseng et al., 2012; Pruitt & Tsai, 2009; Rogers, 2004; Santoro et al., 2012). Medically, feeding tubes are often recommended to ensure that individuals with CP who have feeding difficulties receive adequate nutrition (e.g., Sullivan et al., 2005). However, although feeding tubes are generally effective in treating the nutritional consequences of a feeding disorder (Rogers, 2004; Sullivan et al., 2005), they are not effective in treating the feeding disorder itself. Furthermore, caregivers often express dissatisfaction with the decision to use a feeding tube due to hindrance of oral feedings and oral motor progress, reduction in socialization opportunities (e.g., during mealtimes), interference with daily functioning, and stigmatization (Petersen, Kedia, Davis, Newman, & Temple, 2006). These are all legitimate concerns, but due to the severe consequences of malnutrition in children with a feeding disorder, it is understandable that a feeding tube is one of the primary interventions in cases of low oral intake and poor growth. With that being said, feeding tubes do not need to be used to the exclusion of other treatments to promote oral intake. Treatment from an oral motor perspective often focuses on improving oral motor skill deficits (e.g., lip closure, chewing), providing individualized positioning to improve swallowing and decrease aspiration, and making modifications to foods and drinks to increase the child’s ability to safely manage their consumption (Agency for Healthcare Research and Quality, 2013). However, evidence for the effectiveness of these interventions is lacking.

Despite feeding interventions for individuals with CP taking a largely medical and oral motor approach, it is widely accepted that feeding disorders in general are multifaceted with behavioral components often playing a role in the development, continuation, and exacerbation of feeding problems (Katz, Hyche, & Wingert, 2013; Piazza & Roane, 2009). In such cases, behavioral interventions are exceedingly effective in treating feeding disorders, even when those feeding disorders are also associated with medical or oral motor impairments (Bachmeyer et al., 2009; De Moor, Didden, & Korzilius, 2007). While these treatments are more commonly accepted and utilized with individuals with autism spectrum disorder (ASD) and other developmental disabilities in which a behavioral etiology is frequently identified (Laud, Girolami, Boscoe, & Gulotta, 2009; Matson & Fodstad, 2009; Sharp, Jaquess, Morton, & Miles, 2011; Volkert & Vaz, 2010), they are infrequently discussed in the presence of CP.

When behavioral interventions have been implemented for children with CP with comorbid feeding disorders, the results have been promising. Behavioral interventions, along with oral motor therapy and parent education, were effective in improving oral intake in eight children ages 18 months to 4.7 years with spastic diplegic CP (Clawson, Kuchinski, & Bach, 2007). The children accepted and swallowed bites and drinks quicker, consumed a larger amount of grams, and engaged in lower rates of refusal (e.g., head turns, batting at the spoon) with treatment. Additionally, percentage of tube feeds was able to be reduced by discharge. The purpose of the current study was to replicate and extend these findings by comparing the outcomes of applied behavior analytic interventions targeting feeding difficulties and tube dependence in children with CP to children with ASD, a population for which behavioral interventions are commonly used to address feeding difficulties, using a larger sample size. It was hypothesized that significant improvements in oral consumption, refusal, and negative vocalizations would occur similarly for both populations.

2 Method

2.1 Participants and setting

Children with a diagnosis of ASD or CP and gastrostomy tube dependence who were enrolled in an interdisciplinary intensive hospital-based feeding program between the years of 2003 and 2013 were recruited (N = 58). Prior to admission, all children were evaluated by an interdisciplinary team consisting of a developmental pediatrician, gastroenterologist, or nurse practitioner, behavioral therapist or psychologist, dietitian, and oral motor therapist (occupational therapist or speech-language pathologist), and the child had an interdisciplinary team containing those same disciplines throughout the child’s admission. In the initial interdisciplinary evaluation, the majority of caregivers reported complicated medical histories, chronic tube dependence, and a long history of a variety of previous failed treatment attempts (e.g., medications, outpatient behavioral psychology therapies, outpatient speech therapies, outpatient occupational therapies). During admission, behavioral psychology meal blocks were typically up to 1 hour, 3 times per day. An individualized behavioral treatment protocol was developed and evaluated for each child during these meal blocks. In addition to behavioral psychology meals, each child received two to four 30-minute sessions per week with a speech-language pathologist and with an occupational therapist; however, the data presented within are specific to the child’s behavioral psychology meal blocks only. Medical professionals and dietitians were also involved throughout the admission. It should be noted that during the admission, tube cuts were done isocalorically, such that tube deprivation/hunger provocation was not utilized.

Participants were divided into one of two groups according to diagnosis based upon records review by an interdisciplinary team during the child’s initial evaluation, admission, and discharge – an ASD group (n = 25) and a CP group (n = 33). If a child had both diagnoses, he or she was assigned to the CP group. Participants ranged in age from 20 to 148 months (M = 69.53, SD = 30.69) at the time of admission. Thirty-five (60%) of the children were male. Race distribution was as follows: White = 39, Black = 13, Hispanic = 2, Asian = 1, Other = 2, and Unknown = 1. Regarding level of service, 30 participants were in the intensive day treatment program (5 days per week), 25 were inpatient (7 days per week), and 3 were inpatient and then transferred to intensive day treatment. Average length of service was 46.77 days (SD = 12.00; Range = 18 to 73 days), and average waitlist time from initial evaluation to admission was 234.96 days (SD = 214.26). No statistically significant (p > 0.05 via chi square or t test analyses) diagnostic group differences were found for sex, race, age, level of service, days of service, or days on waitlist.

2.2 Dependent variables

Bachelor’s- and master’s-level feeding therapists compiled data via a systematic retrospective chart review. Therapists were trained and met unit criteria for reliability to collect real-time data on observable defined mealtime behaviors via laptop computers and calculate grams consumed during meal blocks. Clinic standard is to collect interobserver agreement (IOA) data for all behavioral dependent variables for a minimum of 33% of sessions with a minimum of 80% agreement. Refusal was defined as making contact with the therapist’s arm during a presentation, covering mouth, or turning head 45 degrees or more away from the spoon or cup. Refusal was collected via frequency and converted to a per trial measure (frequency of occurrence/number of bite presentations). Negative vocalizations were defined as crying, screaming, whining, swearing, or making negative statements about the meal. Negative vocalizations were collected via duration by using immediate onset and 3 s offset criteria (i.e., the behavior was scored as being present immediately once it occurred and considered to be absent following 3 consecutive seconds without engagement in the behavior). The duration was then converted to a percentage of total session duration. Consumption was represented by Total Grams. Solid and liquid grams were calculated by subtracting post-weight grams from pre-weight grams and accounting for spill. Clinic standard is for meals to consist of a protein, starch, vegetable, and fruit to ensure adequate variety, and a nutritious caloric drink. Refusal and negative vocalizations were evaluated across two time periods, Initial Treatment and Final Treatment: the average of the first and last five sessions of feeding therapy during which the food or liquid was presented until accepted or a pre-specified time interval elapsed. Grams were evaluated across these same two time periods in addition to during Caregiver Baseline (upon admission, the average of three meals conducted in which the caregiver(s) were instructed to feed their children as they typically did at home) and Baseline (the average of the last five baseline sessions therapists conducted prior to implementing initial treatment, or the average of all baseline sessions therapists conducted if less than five sessions were conducted).

Other variables included caregiver reports of mealtime problems and program satisfaction, percentage of goals met, and weight in kilograms. Weight was collected at Initial Evaluation, which is the initial outpatient interdisciplinary screening visit, and upon admission and discharge from the program. Caregivers completed the Children’s Eating Behavior Inventory (CEBI) upon admission and discharge. The CEBI is a 40-item caregiver report measure intended to assess eating and mealtime problems (Archer, Rosenbaum, & Streiner, 1991). Two scores are derived from this measure: a) the Total Eating Problems score which measures the frequency of 19 different eating behaviors through the use of a 5-point rating scale, and b) the Total Perceived Problems score which asks caregivers to evaluate whether or not each behavior presents a problem for the family. Test-retest reliability has been reported at 0.87 for the Total Eating Problem score and 0.84 for the percentage of items perceived to be a problem. Internal consistency reliability ranged from 0.58 to 0.76 depending on family structure. The Total Eating Problems score was evaluated in the present study upon admission and discharge. Caregiver Satisfaction: Upon discharge, caregivers completed a structured questionnaire to assess program satisfaction. Overall satisfaction score (on a Likert type scale from 1-5) was the average score based on questions related to effectiveness of the behavioral approach to treatment, as well as the competency of the behavioral psychology staff throughout admission. Goals Met was calculated as a percentage of goals met at discharge. Individualized goals were set during an interdisciplinary steering meeting (typically in the second week of the admission) and included defined measurable goals targeting increasing acceptance, swallowing, variety, volume, texture and independence, decreasing refusal and negative vocalizations, training caregivers in the protocol, and generalizing the protocol to the natural environment.

3 Results

A mixed model repeated measures Analysis of Variance (ANOVA) was conducted to evaluate differences in consumption in grams between group (CP and ASD) and within group over time (caregiver baseline, baseline, initial treatment, final treatment). There was a significant main effect for time (Wilks’ Lambda = 0.27, F (3, 50) = 44.57, p < 0.001, partial η² = 0.728), but not for the group and time interaction (Wilks’ Lambda = 0.94, F (3, 50) = 1.10, p = 0.359, partial η ² = 0.062) or group (F (1, 52) = 0.47, p = 0.498, partial η ² = 0.009). Simple contrasts revealed that final treatment consumption was higher than consumption at caregiver baseline (F (1, 52) = 48.75, p < 0.001, partial η ² = 0.484; M Difference = 121.91, SE = 17.46, 95% CI 74.02 – 169.81), baseline (F (1, 52) = 125.43, p < 0.001, partial η ² = 0.707; M Difference = 176.64, SE = 15.77, 95% CI 133.38 – 219.90), and initial treatment (F (1, 52) = 81.62, p < 0.001, partial η ² = 0.611; M Difference = 136.44, SE = 15.10, 95% CI 95.01 – 169.86). On average, oral intake increased by over 6 ounces irrespective of diagnosis. Figure 1 depicts consumption means (in grams) and 95% confidence intervals for each time period by group. Regarding liquid consumption at discharge, chi square analyses revealed no statistically significant diagnostic group differences, χ² (1, N = 58) = 2.95, p = 0.086. Twenty-one participants in each diagnostic group (64% of CP, 84% of ASD) were consuming liquids at discharge. Regarding highest solid texture consumed at discharge, chi square analyses revealed no statistically significant diagnostic group differences, χ² (1, N = 50) = 5.42, p = 0.367. Number of participants consuming each texture was as follows: Puree = 19, Junior = 11, Wet ground = 6, Fork Mashed = 1, Regular = 6, Combination = 10.

Fig.1

Total Grams Consumed by Diagnostic Group (CP = Cerebral Palsy; ASD = Autism Spectrum Disorder) across Time (Caregiver Baseline, Baseline, Initial Treatment, and Final Treatment): Means and 95% Confidence Intervals.

A mixed model repeated measures Multivariate ANOVA (MANOVA) was conducted to evaluate differences in refusal and negative vocalizations between group (CP and ASD) and within group over time (initial treatment to final treatment). There was a significant main effect for time (Wilks’ Lambda = 0.73, F (2, 48) = 8.73, p = 0.001, partial η ² = 0.267), but not for the group and time interaction (Wilks’ Lambda = 0.98, F (2, 48) = 0.44, p = 0.647, partial η ² = 0.018) or group (Wilks’ Lambda = 0.994, F (2, 48) = 0.14, p = 0.866, partial η ² = 0.006). Univariate tests revealed significant reductions in refusal (F (1, 49) = 4.46, p = 0.040, partial η ² = 0.083) and negative vocalizations (F (1, 49) = 13.862, p = 0.001, partial η ² = 0.221). On average, percentage of negative vocalizations decreased by 13, and food refusal decreased by 13 occurrences per bite presented, irrespective of diagnosis. Means, standard errors, mean differences, and confidence intervals are depicted in Table 1 for refusal and negative vocalizations upon initial and final treatment.

Table 1

Refusal per Trial and Percentage of Session with Negative Vocalizations in Initial Treatment and Final Treatment:\\ Means, Standard Errors, Mean Differences, and Confidence Intervals

	Mean (Standard Error)
	[95% Confidence Interval]
	Refusal per Trial	% Negative Vocalizations
Means
Initial Treatment	13.06 (5.93) [1.14, 24.99]	16.32 (3.70) [8.88, 23.76]
Final Treatment	0.54 (0.18) [0.17, 0.90]	3.25 (0.93) [1.39, 5.11]
Mean Differences
Initial to Final Treatment	12.53 (5.93)^* [0.61, 24.44]	13.07 (3.51)^* [6.02, 20.13]

Note: Bonferroni adjustments were applied for multiple comparisons. ^*The mean difference is significant.

A mixed model repeated measures ANOVA was conducted to evaluate between group (CP and ASD) and within group over time (admission to discharge) differences in CEBI scores. There was a significant main effect for time (Wilks’ Lambda = 0.85, F (1, 39) = 7.10, p = 0.011, partial η ² = 0.154), but not for the group and time interaction (Wilks’ Lambda = 0.99, F (1, 39) = 0.41, p = 0.525, partial η ² = 0.010) or group (F (1, 39) = 2.85, p = 0.100, partial η ² = 0.068). CEBI scores decreased significantly (M Difference = 6.61, SE = 2.10, 95% CI = 1.35 – 9.87) from admission (M = 106.37, SE = 1.94) to discharge (M = 100.76, SE = 1.85) irrespective of diagnosis.

A mixed model repeated measures Multivariate Analysis of Covariance (MANCOVA) was conducted to evaluate differences in weight (in kilograms) between group (CP and ASD) and within group over time (initial evaluation, admission, and discharge), adjusting for total number of days from initial evaluation to discharge. There was a significant main effect for increases in weight over time (Wilks’ Lambda = 0.79, F (2, 44) = 5.82, p = 0.006, partial η ² = 0.209; adjusted Wilks’ Lambda = 0.32, F (2, 44) = 47.73, p < 0.001, partial η ² = 0.685) and for the interaction of increases in weight over time and total days (Wilks’ Lambda = 0.78, F (2, 44) = 6.08, p = 0.005, partial η ² = 0.217), but not for the group by weight increases in time interaction (Wilks’ Lambda = 0.96, F (2, 44) = 0.81, p = 0.450, partial η ² = 0.036) or group (F (1, 45) = 0.02, p = 0.904, partial η ² = 0.000). Irrespective of diagnosis, pairwise comparisons revealed that discharge weight (M = 18.24, SE = 0.77) was higher than weight at initial evaluation (M = 16.41, SE = 0.78; M Difference = 1.83, SE = 0.20, p < 0.001, 95% CI 1.34 – 2.32) and admission (M = 17.47, SE = 0.79; M Difference = 0.76, SE = 0.14, p < 0.001, 95% CI 0.42 – 1.11), and admission weight was higher than initial evaluation weight (M Difference = 1.07, SE = 0.21, p < 0.001, 95% CI 0.54 – 1.59).

Additional variables were examined. Diagnostic groups did not differ in percentage of goals met during the admission (t (53) = – 1.19, p = 0.238). Percentage of goals met on average were high, as participants with CP met 82.19% (SE = 5.22) of goals and participants with ASD met 90.43% (SE = 4.24) of goals. Diagnostic groups did not differ in caregiver satisfaction (t (50) = 0.41, p = 0.682). Caregiver satisfaction on average was high for participants with CP (M = 4.48, SE = 0.13) and participants with ASD (M = 4.41, SE = 0.09). Regarding solid and liquid treatment protocol components (e.g., escape extinction such as nonremoval of the spoon; positive reinforcement such as differential reinforcement/attention, noncontingent reinforcement; punishment such as response cost), chi square analyses revealed no significant differences (p > 0.05) based on diagnostic group.

4 Discussion

The present study examined the effectiveness of intensive applied behavior analytic treatment for participants with CP versus ASD with tube dependence. No significant differences were found based on diagnostic group of CP or ASD for any variables examined in the present study. Behavioral treatment was equally effective in participants with CP and ASD in increasing oral consumption and decreasing refusal and negative vocalizations. For both groups, caregiver report of feeding difficulties significantly improved, and satisfaction and percentage of goals met (e.g., increasing acceptance, swallowing, variety, volume, texture and independence, decreasing refusal and negative vocalizations, training caregivers in the protocol, and generalizing the protocol to the natural environment) were both high. Participants did not differ based on diagnostic group on demographic variables (e.g., gender, age, race, level of service), treatment components utilized, presence of liquids at discharge, or discharge texture. Thus, applied behavior analytic interventions were effective for participants with CP, just as they are for those with ASD. These results are consistent with those found by Clawson and colleagues (2007). Due to the effectiveness of such interventions, it is advised that behavioral treatments are also utilized with children with CP with feeding difficulties, in addition to focusing on the children’s oral motor and medical comorbidities. In addition to the positive findings found within the current study, implementing behavioral treatments and strategies with children with CP who have feeding difficulties and oral motor skill deficits may also positively impact progression with oral motor skills as decreasing refusal will likely provide oral motor therapists with better opportunities to effectively implement oral motor therapy.

The present study has some limitations, particularly due to its retrospective nature. Participants were assigned to diagnostic groups based on chart review; therefore, information regarding the qualifications of those diagnosing the participants and what diagnostic assessments were completed is not available. Additionally, the severity of both diagnoses is unknown. Although all children in the CP group had been previously diagnosed with CP, motor impairment measures were not used to categorize children further. Future research should examine differences in severity of motor impairment in individuals with CP to assess how severity of impairment affects behavioral treatment outcomes. This could have impacted the outcomes of this study, especially because children admitted into the feeding program are those children who had first been evaluated by the interdisciplinary team and recommended for such services. However, in general, common reasons for children with tube dependence to not be recommended for the interdisciplinary intensive hospital-based feeding program during the evaluation include outpatient services being recommended as more appropriate, or because the child has been deemed unsafe to swallow solids and liquids based on the results of a swallow study. In the latter case, behavioral interventions are not recommended. If a child’s ability to safely swallow solids or liquids is unknown at the time of evaluation due to unclear swallow study results related to refusal during the study, services are recommended to assist in preparing the child for a repeat study to determine safety. These services may be offered through outpatient appointments or within the interdisciplinary intensive hospital-based feeding program depending on the child.

Additionally, total gram consumption within the Baseline and Initial Treatment phases, which was represented by Total Grams, was partially influenced by therapist-directed guidelines regarding number of programmed bites and time caps. During these phases, a pre-determined number of bites was presented with a set time cap. The volume was kept consistent across the Baseline and Initial Treatment phases to ensure that the treatment was the mechanism of change rather than an alteration in bite number or another component. Having said that, the volume of food presented during Caregiver Baseline was based on caregiver discretion and representative of each child’s consumption upon admission, and the volume of food presented during Final Treatment was guided by a dietitian based on the individual child’s needs. Therefore, the significant increase in volume consumed between Baseline and Initial Treatment, and also between Caregiver Baseline and Final Treatment, demonstrates the effectiveness of the behavioral intervention in total gram consumption. Also, it is important to note that in Final Treatment compared to Caregiver Baseline, a variety of foods (i.e., protein, starch, vegetable, and fruit) are included in the meal.

Prospective, systematic, and well-controlled studies are needed to further evaluate patient characteristics and, in particular, individual components of an intensive behavior analytic and interdisciplinary approach to increasing oral intake in children who are dependent on tube feeding. Within these studies, information regarding percent tube dependence at admission and discharge should be evaluated and compared between groups. Although oral consumption data were collected in the present study, thereby providing some insight into tube dependence, exact calculations of tube dependence were not collected. In order for this factor to be considered in future studies, a dietitian should be involved to perform appropriate caloric calculations. Future research should also further examine differences in texture and independence (e.g., self-feeding and drinking). A thorough evaluation of advancement in texture was not conducted due to caregivers sometimes using inappropriate textures in home baseline and, though food textures were manipulated for appropriateness prior to pre-treatment sessions, complete texture assessments were not always conducted prior to implementation of sessions. Especially due to the motor impairments associated with CP, this information may be valuable to caregivers and worth investigating. Finally, follow-up data should be collected to determine whether gains made with behavioral treatments for both populations are maintained over time.

Conflict of interest

The authors declare they have no conflicts of interest.

References

Agency for Healthcare Research and Quality. (2013). Effective health care program comparative effectiveness review number: Interventions for feeding and nutrition in cerebral palsy. AHRQ Publication No. 13-EHC015-EF). Retrieved from www.effectivehealthcare.ahrq.gov/reports/final.cfm.

Archer

L. A.

, Rosenbaum

P. L.

, & Streiner

D. L.

(1991). The Children’s Eating Behavior Inventory: Reliability and validity results. Journal of Pediatric Psychology, 16(5), 629–642. doi: 10.1093/jpepsy/16.5.629.

Arvedson

, Rogers

, Buck

, Smart

, & Msall

(1994). Silent aspiration prominent in children with dysphagia. International Journal of Pediatric Otorhinolaryngology, 28(2), 173–181. doi: 10.1016/0165-5876(94)90009-4.

Bachmeyer

M. H.

, Piazza

C. C.

, Fredrick

L. D.

, Reed

G. K.

, Rivas

K. D.

, & Kadey

H. J.

(2009). Functional analysis and treatment of multiply controlled inappropriate mealtime behavior. Journal of Applied Behor Analysis, 42(3), 641–658. doi: 10.1901/jaba.2009.42-641.

Benfer

K. A.

, Weir

K. A.

, Bell

K. L.

, Ware

R. S.

, Davies

P. S. W.

, & Boyd

R. N.

(2013). Oropharyngeal dysphagia and gross motor skills in children with cerebral palsy. Pediatrics, 131(5), e1553–e1562. doi: 10.1542/peds.2012-3093.

Clawson

E. P.

, Kuchinski

K. S.

, & Bach

(2007). Use of behavioral interventions and parent education to address feeding difficulties in young children with spastic diplegic cerebral palsy. NeuroRehabilitation, 22(5), 397–406. http://www.iospress.nl/journal/neurorehabilitation/.

Dahlseng

M. O.

, Andersen

G. L.

, DA Graca Andrada

, Arnaud

, Balu

, De la Cruz

, & Júlíusson

P. B.

(2012). Gastrostomy tube feeding of children with cerebral palsy: Variation across six European countries. Developmental Medicine and Child Neurology, 54(10), 938–944. doi: 10.1111/j.1469-8749.2012.04391.x.

De Moor

, Didden

, & Korzilius

(2007). Behavioural treatment of severe food refusal in five toddlers with developmental disabilities. Child: Care, Health and Development, 33(6), 670–676. doi: 10.1111/j.1365-2214.2007.00734.x.

Fung

E. B.

, Samson-Fang

, Stallings

V. A.

, Conaway

, Liptak

, Henderson

R. C.

, & Rosenbaum

(2002). Feeding dysfunction is associated with poor growth and health status in children with cerebral palsy. Journal of the American Dietetic Association, 102(3), 361–373. doi: 10.1016/S0002-8223(02)90084-2.

10.

Katz

R. M.

, Hyche

J. K.

, & Wingert

E. K.

(2013). Pediatric feeding disorders: Feeding children who can’t or won’t eat. In Caballero

(Ed.), Encyclopedia of Human Nutrition (Third Edition) 21–27. Waltham: Academic Press.

11.

Laud

R. B.

, Girolami

P. A.

, Boscoe

J. H.

, & Gulotta

C. S.

(2009). Treatment outcomes for severe feeding problems in children with autism spectrum disorder. Behavior Modification, 33(5), 520–536. doi: 10.1177/0145445509346729.

12.

Matson

J. L.

, & Fodstad

J. C.

(2009). The treatment of food selectivity and other feeding problems in children with autism disorders. Research in Autism Spectrum Disorders, 3(2), 455–461. doi: 10.1016/j.rasd.2008.09.005.

13.

Parkes

, Hill

, Platt

M. J.

, & Donnelly

(2010). Oromotor dysfunction and communication impairments in children with cerebral palsy: A register study. Developmental Medicine and Child Neurology, 52(12), 1113–1119. doi: 10.1111/j.1469-8749.2010.03765.x.

14.

Petersen

M. C.

, Kedia

, Davis

, Newman

, & Temple

(2006). Eating and feeding are not the same: Caregivers’ perceptions of gastrostomy feeding for children with cerebral palsy. Developmental Medicine and Child Neurology, 48(9), 713–717. doi: https://dx-doi-org.web.bisu.edu.cn/10.1017/S0012162206001538.

15.

Piazza

C. C.

, & Roane

H. S.

(2009), Assessment of pediatric feeding disorders Assessing Childhood Psychopathology and Developmental Disabilities (pp 471–490): Springer.

16.

Pruitt

D. W.

, & Tsai

(2009). Common medical comorbidities associated with cerebral palsy. Physical Medicine and Rehabilitation Clinics of North America, 20(3), 453–467. doi: 10.1016/j.pmr.2009.06.002.

17.

Reilly

, & Skuse

(1992). Characteristics and management of feeding problems of young children with cerebral palsy. Developmental Medicine and Child Neurology, 34(5), 379–388. doi: 10.1111/j.1469-8749.1992.tb11449.x.

18.

Reilly

, Skuse

, & Poblete

(1996). Prevalence of feeding problems and oral motor dysfunction in children with cerebral palsy: A community survey. The Journal of Pediatrics, 129(6), 877–882. doi: 10.1016/S0022-3476(96)70032-X.

19.

Rogers

(2004). Feeding method and health outcomes of children with cerebral palsy. The Journal of Pediatrics, 145(2), S28–S32. doi: 10.1016/j.jpeds.2004.05.019.

20.

Santoro

, Lang

M. B. D.

, Moretti

, Sellari-Franceschini

, Orazini

, Cipriani

, & Battini

(2012). A proposed multidisciplinary approach for identifying feeding abnormalities in children with cerebral palsy. Journal of Child Neurology, 27(6), 708–712. doi: 10.1177/0883073811424083.

21.

Sharp

W. G.

, Jaquess

D. L.

, Morton

J. F.

, & Miles

A. G.

(2011). A retrospective chart review of dietary diversity and feeding behavior of children with autism spectrum disorder before and after admission to a day-treatment program. Focus on Autism and Other Developmental Disabilities, 26(1), 37–48. doi: 10.1177/1088357609349245.

22.

Sullivan

, Juszczak

, Bachlet

A. M. E.

, Lambert

, Vernon-Roberts

, Grant

H. W.

, & Thomas

A. G.

(2005). Gastrostomy tube feeding in children with cerebral palsy: A prospective, longitudinal study. Developmental Medicine & Child Neurology, 47, 77–85. doi: 10.1111/j.1469-8749.2005.tb01095.x.

23.

Sullivan

, Juszczak

, Lambert

, Rose

, Ford-Adams

, & Johnson

(2002). Impact of feeding problems on nutritional intake and growth: Oxford Feeding Study II. Developmental Medicine and Child Neurology, 44(7), 461–467. doi: 10.1111/j.1469-8749.2002.tb00307.x.

24.

Volkert

V. M.

, & Vaz

P. C. M.

(2010). Recent studies on feeding problems in children with autism. Journal of Applied Behavior Analysis, 43(1), 155–159. doi: 10.1901/jaba.2010.43-155.

25.

Weir

K. A.

, Bell

K. L.

, Caristo

, Ware

R. S.

, Davies

P. S.

, Fahey

, & Boyd

R. N.

(2013). Reported eating ability of young children with cerebral palsy: Is there an association with gross motor function? Archives of Physical Medicine and Rehabilitation, 94(3), 495–502. doi: 10.1016/j.apmr.2012.10.007.

26.

Wood

, & Rosenbaum

(2000). The gross motor function classification system for cerebral palsy: A study of reliability and stability over time. Developmental Medicine and Child Neurology, 42(5), 292–296. http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1469-8749.

27.

Yilmaz

, Basar

, & Gisel

E. G.

(2004). Assessment of feeding performance in patients with cerebral palsy. International Journal of Rehabilitation Research, 27(4), 325–329. doi: 10.1097/00004356-200412000-00013.