Secondary Logo

Journal Logo

Original Articles–Liver and Nutrition

A Multicomponent Behavioral Program for Oral Aversion in Children Dependent on Gastrostomy Feedings

Byars, Kelly C.*,‡,§; Burklow, Kathleen A.*,‡,§; Ferguson, Kathleen*; O'Flaherty, Therese§; Santoro, Kathy§; Kaul, Ajay†,‡,§

Author Information
Journal of Pediatric Gastroenterology and Nutrition: October 2003 - Volume 37 - Issue 4 - p 473-480
  • Free



Feeding disorders are common in the pediatric population, with prevalence estimates ranging from 25% to 35% in developmentally normal children (1–4). Severe feeding problems are more prevalent (40%–70%) in children with developmental disabilities and chronic medical conditions (1,5,6). Feeding problems may arise from a broad range of etiologies, including neurodevelopmental disorders, disorders of appetite regulation, metabolic diseases, sensory defects, conditioned dysphagia, and anatomic abnormalities (1). Many feeding disorders initially develop as a result of an organic condition but are maintained over time by behavioral factors (7–10).

Establishing normal eating behavior in children who have required prolonged gastrostomy tube (G-tube) feedings is a clinical challenge because of the interplay between organic and nonorganic factors and because of the positive and negative impacts of common therapies. In a child with oral feeding resistance who fails to thrive, G-tube feedings may be essential to the delivery of basic nutritional requirements (11). Although G-tube feedings ensure sufficient nutrition and hydration, they may thwart the progress of oral feeding by decreasing the child's hunger-driven motivation to eat and by decreasing oral stimulation (12–15). Aggressive attempts to increase oral consumption may produce negative feeding experiences (choking, gagging, vomiting) and may worsen the conditioned aversion to oral feeding (13).

The utility of behavioral treatments for children with feeding problems who have undergone fundoplication with gastrostomy tube placement has not been fully explored. There is a critical need to delineate specific postoperative behavioral interventions that will promote optimal recovery in children with chronic food refusal who are dependent on feeding gastrostomy for nutritional support. In this study, we describe the outcomes of a multicomponent behavioral feeding program in a group of children with medically complex conditions; all of the children had Nissen fundoplication (NF) and G-tube placement. The treatment protocol emphasized appetite manipulation, time-limited behavioral treatment, and parent training. Clinical outcomes were assessed immediately after therapy and several months after treatment to evaluate the long-term effectiveness of therapy.



Nine children (four females) with a history of behavioral feeding resistance and G-tube dependence were admitted for intensive treatment to the Inpatient Feeding Program (IFP) at Cincinnati Children's Hospital Medical Center (CCHMC). This sample includes all of the IFP cases (from July 2000 to December 2001) referred for feeding problems after NF. The mean age of the patients was 3.1 ± 1.2 years (mean ± SD) (range, 1.8–5.5 years). All children had complicated medical histories that included NF with simultaneous G-tube placement. Mean age at the time of surgery was 11.6 ± 9.5 months (range, 4–29 months). Behavioral feeding resistance had precluded successful outpatient weaning from G-tube feedings. At entry into the IFP, G-tube feedings provided 85.4% ± 21% of daily caloric needs (range, 33%–100%). The mean duration of G-tube dependence was 2.2 ± 1.1 years (range, 1–4.4 years).


Preadmission Clinical Evaluation

Before admission, all children underwent a comprehensive evaluation with the Interdisciplinary Feeding Team (IFT) at CCHMC (16) to assess appropriateness for intensive behavioral feeding treatment. Consistent with published criteria for intensive treatment aimed at weaning from gastrostomy feeding, all children met the following clinical criteria: 1) resolution or stability of the medical problem necessitating NF, 2) absence of anatomic or functional impairment precluding safe oral feeding, 3) absence of oral-motor apraxia, 4) maintenance of adequate weight on G-tube feeding, 5) receptivity, motivation, and capacity of primary caregiver to participate actively in treatment, and 6) cognitive/developmental status adequate to allow a response to behavior therapy (≥12 months age equivalent) (14,15,17,18).

Screening/Diagnostic Measures

Before treatment, the factors contributing to the child's feeding disorder were assessed and the severity of mealtime behavior problems was evaluated.

Behavioral Pediatric Feeding Assessment Scale (BPFAS)

The BPFAS (19) is a 35-item questionnaire developed to measure mealtime behaviors in young children (ages 9 months–7 years) and parent and child behaviors associated with poor nutritional intake. The BPFAS was used in this study to assess patients' behavioral problems during feeding. For each item, parents indicated the frequency of a specific mealtime behavior using a Likert-type scale (1 = never to 5 = always). Using a yes/no format, parents were asked to indicate whether the particular behavior was a problem for the family. The questionnaire yields four scores: Child Behavior–Frequency (i.e., sum of 5-point Likert-type scores; range: 5–125), Child Behavior–Problems (i.e., number of behaviors rated yes; range: 0–25), Parents' Feelings/Strategies–Frequency (i.e., sum of 5-point Likert-type scores; range: 10–50), Parents' Feelings/Strategies–Problems (i.e., number of behaviors rated yes; range: 0–10). Frequency scores reflect how often parent and child behaviors occur. Problem scores reflect the number of feeding behaviors the parents consider a problem. Higher scores suggest maladaptive feeding behaviors as well as problematic parental strategies for managing mealtime behavior. Parent and child scores for each domain may be combined to yield a total frequency score and a total problem score.

Psychometric analyses suggest that the BPFAS has adequate internal consistency for the behavior domains and the measure as a whole (alpha = .74 to .88) (19). The test-retest reliability coefficients (.82 to .85) measured during a 2-year interval reflect the stability of this assessment measure (19). Normal and medically impaired reference groups for this measure have been established (7). Both reference groups were composed of children with feeding aversion referred to a feeding and nutrition clinic at a tertiary care, teaching hospital serving Nova Scotia, Prince Edward Island, and New Brunswick, Canada. Subjects in both groups were children whose feeding problems had not responded to community-based treatment. The normal group had no known physical condition associated with feeding problems. The medically impaired group had organic conditions that were thought to influence the feeding difficulties (e.g., cardiorespiratory, renal, neurologic, gastrointestinal, structural, or metabolic disorders; cystic fibrosis; or disorders of the immune system) (7). In our study, the mother of each child completed the BPFAS during the initial evaluation with the IFT. Total frequency and problem scores were used as a pretreatment measure of the level of behavioral dysfunction during feeding.

Classification System for Complex Feeding Disorders (CSCFD)

The CSCFD was used to classify each child with regard to specific etiological factors (8). The CSCFD uses five categories common to the etiology of complex feeding disorders: structural abnormalities (cleft palate, esophageal stricture, gastroesophageal reflux disease), neurologic conditions (cerebral palsy, brain stem injury, nerve dysfunction), cardiorespiratory problems (tachypnea associated with bronchopulmonary dysplasia), metabolic dysfunction (hereditary fructose intolerance, dumping syndrome), and behavioral issues. Behavioral issues included psychosocial problems (poor environmental stimulation, maladaptive feeder-caregiver interactions), negative feeding behavior caused or maintained by internal or external reinforcement (selective food refusal, rumination), and emotional difficulties (choking phobia, conditioned emotional reactions, anxiety, depression). Two independent coders classified all subjects using the preestablished coding system for CSCFD (8). Interrater reliability was 100% for the sample. Percent agreement was calculated by dividing the number of coding agreements by the number of total codes identified.

Dependent Measures: before Treatment to Follow-up

Dependent measures were obtained during three assessment points (pretreatment, posttreatment, and follow-up). Pretreatment measures were obtained on the day of admission, before the start of treatment. Posttreatment measures were obtained on day of discharge from the hospital (mean length of treatment, 11.4 ± 3.4 days; range, 5–16 days.). We obtained dependent measures for all patients (n = 9) at the pretreatment and posttreatment assessments. Follow-up dependent measures were obtained in six patients during a comprehensive IFT clinic visit scheduled 2 to 4 months after the child's discharge (mean follow-up, 3.1 ± 0.5 months; range, 2.4–3.6 months). Three families could not return for the IFT clinic-based follow-up because they lived more than 700 miles from the medical center. No weight measurements were obtained for these three patients. Telephone assessments to gather calorie data were completed for two of these three patients.

Caloric Intake

Two measures of the patient's caloric intake (oral and G-tube) were obtained at each of the three assessment points. Oral caloric intake at the pretreatment and follow-up assessments was estimated from parental diet diaries and diet recall obtained by a licensed registered dietitian. Oral intake at the posttreatment assessment was based on consumption observed and recorded during the last full day of inpatient treatment (three meals). G-tube caloric intake at the pretreatment and follow-up assessments was based on the parent's report of the volume of formula provided through the G-tube. G-tube caloric intake at the posttreatment assessment was calculated from the volume of feeding delivered during the last 24 hours of inpatient treatment. A licensed registered dietitian or a registered diet technician under the guidance of a registered dietitian computed calorie counts for all three assessment periods.

Weight and Height

Weight and height were measured at each of the three assessments. Linear height was measured with a wall-mounted stadiometer in patients older than 2 years, and with a length board in patients younger than 2 years old. Body weight was measured with a digital scale. Weight was expressed as percentage of ideal body weight for height (%IBW), which was calculated by taking the quotient of actual weight to weight at the 50th percentile for the child's actual height (20).


The IFP at CCHMC provides short-term behavioral treatment with a family-focused approach. During the admission, a team of behavioral therapists, including a licensed pediatric psychologist, managed all aspects of behavioral treatment. A gastroenterologist and registered dietitian monitored and managed the patient's medical and nutritional status throughout the admission. All patients followed a standard treatment protocol using a multicomponent intervention targeting the treatment variables described (see Fig. 1).

FIG. 1.:
Intensive feeding treatment process.

Appetite Inducement

Stimulating appetite drive is a critical motivational variable in our treatment program. To maximize the patient's appetite, G-tube feeding was reduced by more than 50% of the preadmission levels at the onset of treatment. G-tube supplementation was decreased further throughout the admission as oral intake improved. If diurnal G-tube feedings were clinically necessary, they were scheduled to optimize appetite by giving them immediately after oral feedings. Patient access to food was restricted to three daily treatment sessions. To maintain hydration, patients were permitted to drink water or juice diluted to 50% concentration with water ad libitum. All fluids were restricted 60 minutes before treatment sessions.

Medical Stability

Behavioral resistance (food refusal) at the initiation of treatment in conjunction with decreased G-tube feeding and food/liquid restriction resulted, as anticipated, in a reduced calorie and fluid intake, which produced weight loss at treatment outset. Patients were closely monitored throughout the admission to assess the impact on health of the intensive treatment and to make modifications in the treatment protocol, if necessary. Daily calorie counts, body weight, vital signs, and stool and urine output were used to monitor possible dehydration, electrolyte imbalance, orthostasis, or excessive weight loss.

Environment/Schedule of Intake

Patients were admitted to private inpatient rooms on a medical unit. Therapists conducted treatment sessions at 0800, 1200, and 1700 daily during the admission in a treatment room adjacent to the patient's bedroom. The treatment room was relatively devoid of distractions. All meals (treatment sessions) were limited to 25 to 30 minutes unless otherwise clinically indicated. Parents were instructed to wake their child 30 to 45 minutes before mealtime to ensure an adequate level of alertness during treatment sessions.

Meal Characteristics

The food texture and type offered (blended, chopped, or table food), seating arrangement (high chair versus table and chair), eating equipment (cups, spoons, forks, bibs), and level of caregiver assistance (spoon-feeding, loading utensil for child to self-feed, holding child's cup) were tailored to the patient's individual needs. Treatment was tailored to the functional feeding/eating skills consistent with the child's developmental status. For all patients, treatment goals included increasing the range and amount of foods eaten and reducing the level of G-tube support. The initial IFT evaluation determined the meal characteristics and specific treatment goals for each patient.

Behavioral Intervention

Even in cases in which feeding aversion is initially caused by a medical issue, mismanagement of feeding resistance may lead to maladaptive parent-child interactions (21). Thus, the foundation of our behavioral treatment was to re-establish adaptive mealtime interactions. Behavioral strategies (operant learning principles) used in our treatment protocol, including positive and negative reinforcement, shaping, discrimination, fading, and escape extinction have been well described and validated in the behavioral literature addressing feeding problems (4,9,10). An example of the application of positive reinforcement is the use of verbal praise or brief access to a preferred toy or activity immediately after achieving a targeted feeding behavior (a child accepts a mouthful of food). Alternatively, an example of escape extinction is when the therapist keeps the spoon at the child's lips (without the use of force) until the mouthful is accepted (22). A licensed psychologist designed, initiated, and monitored behavioral treatment at scheduled meals, with trained behavioral therapists (master's degree or higher in psychology) carrying out treatment until stable outcomes were achieved. The behavioral techniques used during treatment were specific to each patient and modified throughout treatment as the patient's behavior changed. Therapists followed a “least intrusive” approach to treatment. Some behavioral programs (14,23,24) use negative reinforcement techniques (physical guidance, jaw prompting, forced feeding) that require the feeder to physically guide food into the child's mouth while holding the child's jaw until the child accepts the food. Such procedures may exacerbate feeding problems (25,26) and cause complications, such as aspiration pneumonia (27). Furthermore, such procedures are often unacceptable to caregivers and may be associated with caregiver noncompliance (22). Thus, in our program therapists limited negative reinforcement procedures and used less-invasive escape extinction techniques to encourage active acceptance of oral intake.

Parent Training/Support

During the preliminary phase of treatment, parent training included didactic sessions and supportive counseling, as well as regular feedback on their child's progress. Didactic counseling addressed behavior management strategies (use of verbal communication/prompts and reinforcement strategies/differential attention). Supportive counseling addressed the emotional stress commonly experienced by caregivers during hospitalization. We found that caregivers often required support early in treatment to deal with their ambivalence regarding the progress made by their child during therapy sessions from which they were excluded.

Initially, the caregiver received instruction by unobtrusively viewing treatment sessions via hidden camera/closed circuit television. A trained behavioral therapist facilitated the observational sessions, reviewing and discussing with the caregiver the behavioral techniques being used during the child's mealtime. After the caregiver gained an understanding of the behavioral protocol and the child established stable treatment outcomes (child ate consistently at targeted levels), the caregiver was introduced into treatment sessions in three stages. Initially the caregiver simply observed treatment while seated in the treatment room with the behavioral therapist and the child. Next, the caregiver fed the child with direct guidance and supervision from the therapist in the treatment room. Finally, the caregiver fed the child independently while the therapist observed through the hidden camera and communicated with the caregiver using a remote “bug-in-the-ear” device. Thus, the caregivers could feed the child alone in the treatment room yet continue to receive guidance from the therapist.

During each transition in treatment, as the caregiver was given greater autonomy in feeding the child, it was not unusual that both the child and parent reverted transiently to previous maladaptive patterns of behavior. Thus, at each treatment transition, stable treatment outcomes were reestablished before moving to the next phase of treatment. Families were discharged from the program when initial treatment goals were met and the caregiver could feed the child independently with minimal guidance or supervision from the therapist. Subsequent to discharge, a licensed psychologist provided each family with maintenance therapy sessions to review and refine behavioral techniques and to assess the child's progress.


Screening/Diagnostic Measures

Behavioral Pediatric Feeding Assessment Scale (BPFAS)

The total frequency and problem scores from the BPFAS were used as a measure of severity of mealtime behavior problems before treatment (see Table 1). t tests revealed that total frequency and problem scores for our sample were comparable to scores for the reference medical group (P = 0.123 and P = 0.161, respectively) and significantly higher than those for the normal group (P = 0.000 and P = 0.003, respectively). Posttreatment data from the BPFAS were not obtained as part of our assessment protocol.

Behavioral Pediatrics Feeding Assessment Scale *

Classification System for Complex Feeding Disorders (CSCFD)

Etiologic factors in our patients included: behavioral (N = 9), structural (N = 9), cardiorespiratory (N = 4), metabolic (N = 4), and neurologic (N = 3). Specific medical diagnoses are summarized in Table 2.

Patient characteristics and medical diagnoses

Dependent Measures

Caloric Intake

Oral and G-tube caloric intake at the pretreatment, posttreatment, and follow-up assessments are presented in Table 3. At the posttreatment assessment, patients were receiving 84% of their nutritional needs (64% orally; 20% by G-tube) on a restricted meal schedule. As noted previously, this program accomplishes appetite stimulation via a meal schedule that limits oral feeding to 3 times daily and fluid intake to water and diluted juice ad libitum between meals, with complete fluid restriction for 60 minutes before meals. At the time of discharge, four of nine (44%) patients had completely discontinued G-tube feeding. After discharge, and following the establishment of a meal routine at home, children were no longer prohibited from taking food between meals. Indeed, children were provided scheduled snacks, as appropriate, to increase opportunities for additional caloric intake. At follow-up, all of the children were receiving 100% of their calculated nutritional needs (88% orally; 12% by G-tube). G-tube feeding had been discontinued completely in six (67%) children. Paired-sample t tests revealed that increases in oral intake and decreases in G-tube support were statistically significant from baseline to the posttreatment assessment (t = −7.513, df = 8, P < 0.0001 and t = 9.812, df = 8, P < 0.0001, respectively) and from baseline to the follow-up assessment (t = −7.15, df = 7, P < 0.0001 and t = 7.145, df = 7, P <0.0001, respectively).

Calorie intake and percent ideal body weight

Weight and Height

Weight and height measurements were used to calculate %IBW at each of the assessment points (Table 3). As anticipated, patients experienced some weight loss during treatment because of the rigors of the intensive intervention. Small weight gains were observed at the initial follow-up assessment. Weight status was reevaluated at 7 ± 1.8 months after discharge to document long-term outcomes. Stability in weight was observed (mean %IBW, 91.4% ± 8.8%; N = 5). Thus, throughout the entire assessment period the patients maintained adequate weight while undergoing considerable change in their route of caloric intake.


Clinical Outcomes

Our prospectively designed treatment study supports the use of biobehavioral treatment in children with a history of G-tube dependence and feeding aversion unresponsive to outpatient therapy. There were significant increases in oral intake and significant reductions in G-tube supplementation in all of our patients. At discharge, 44% of the sample had been successfully weaned from gastrostomy feedings. At follow-up, 67% (six of nine) were completely weaned from G-tube feedings and were taking 100% of their nutritional needs by mouth. Weight gains were small. Nonetheless, the measurement of %IBW suggested that nutrition was adequate at each assessment point. Long-term stability in weight status was observed in most of the sample. The observation that patients did not return to ideal body weight after intensive treatment suggests that the weight of these patients while receiving high levels of G-tube nutrition may have been elevated beyond their genetic growth potential.

Advances in Behavioral Feeding Treatment

This study extends the literature supporting behavioral treatment for feeding problems in children with medically complex conditions who are dependent on G-tube feedings (13–15,17,28–30). Our treatment outcomes support the feasibility and clinical importance of several treatment variables, including 1) the use of dependent measures that more accurately inform clinical decisions regarding caloric intake and weight status; 2) a systematic approach to appetite stimulation that can be replicated in future research; 3) short-term intensive treatment, and 4) follow-up evaluation to support maintenance of treatment outcomes after discharge from treatment.

In a review of the behavioral and medical literature to date, there are only five studies that specifically address the behavioral treatment of dependence on gastrostomy feedings in children who have undergone fundoplication and placement of G-tube (9,14,15,31,32). All five studies report improvement in oral intake with behavioral therapy. Complete weaning from feeding gastrostomy was reported in three of the five studies (9,14,15). Duration of behavioral treatment ranged from 2 weeks to 17 months, with outpatient treatment being more protracted than intensive inpatient treatment. These findings are promising, but these reports all have limitations. Study measures lacked specificity regarding the patient's calorie consumption and weight status. Two of the five studies reporting changes in oral intake used either grams (9) or fluid ounces (31) consumed, without any consideration of the calorie content of food. Whereas each of these variables can be used as a measure of treatment efficacy, they lack the specificity necessary to inform nutritional recommendations.

Our approach, which documents the caloric content of foods consumed, as opposed to simple mass or volume, provides a more specific measure of intake that can be used to make decisions about reduction in G-tube feedings. The only studies (31,32) reporting weight gain documented only the absolute weight gain without consideration of the child's weight for height. Our protocol used %IBW as a measure of weight status, which provides clinically important information about the appropriateness of a child's weight as compared with absolute weight gain/loss during treatment. The appropriateness of caloric intake and growth are critical variables that must be closely monitored during intensive treatment. In addition, although most of the previous studies adequately describe the technical aspects of behavioral treatment, the procedures for appetite stimulation were not clearly delineated. We have provided very specific treatment procedures that can be replicated in future behavioral treatment studies. Only two studies (31,32) provided any case-specific follow-up data to document the persistence of the outcomes seen during treatment. Our observations provide evidence that the positive effects of intensive treatment continued for a significant time after inpatient discharge.

Because pediatric feeding disorders are heterogeneous with respect to etiology and physical symptoms (33), most of the research in this area has been based on single subject designs among children with a broad range of physical and developmental disabilities. In contrast, the cases described in this study have several common etiological factors. All children had a history of behavioral feeding resistance that precluded successful outpatient treatment. In addition, all cases had complicated medical histories that included gastroesophageal reflux disease, fundoplication surgery, and G-tube placement to maintain adequate nutrition.

This study highlights several benefits of inpatient hospitalization for intensive feeding treatment. In our experience, children with longstanding G-tube dependence respond more readily to behavioral feeding treatment when an aggressive weaning protocol is used. Rapid weaning promotes hunger that is necessary to overcome learned oral feeding aversion. Inpatient treatment provides an appropriate context for significant calorie intake restrictions that cannot be safely accomplished at home in the absence of medical monitoring. In addition, multiple daily treatments provide a consistent setting for diminishing severe mealtime behavioral resistance that cannot be provided on an outpatient basis. Furthermore, extensive parent education and graduated integration into treatment allows caregivers to gain the knowledge and parenting skills needed to overcome maladaptive feeding patterns. Finally, inpatient hospitalization decreased the need for protracted treatment that is common with outpatient treatment programs.

Limitations and Future Directions

Although the findings of this study are promising, there are a number of factors that should be considered before generalizing the results. Strict criteria were used for inclusion in the study. Each child had undergone comprehensive evaluation and had been medically cleared for intensive treatment. It is possible that intensive behavioral treatment may not be as effective in children with ongoing medical issues that continue to complicate oral feeding behavior. In all cases, families participating in treatment were motivated to make behavioral changes and were actively involved in treatment. Thus, behavioral treatment may not be as effective in families that are not so committed to making significant changes in the family system.

One study limitation that has implications for future behavioral research is the lack of posttreatment measures of behavioral feeding resistance. Obtaining BPFAS data at the initial assessment has been standard practice in our clinical program to screen patients for potential behavior problems. However, until recently we had not formally assessed behavioral feeding problems subsequent to treatment. The overall success of the program may be questioned, even in the face of improved calorie consumption, if the mealtime behavior of the child continues to be disruptive and stressful for the family. Our follow-up clinical contact with families indicated that improvements in oral intake were associated with overall improvements in mealtime behavioral functioning. However, future studies should incorporate formal measures of parental stress and child mealtime behavior problems into pretreatment, posttreatment, and follow-up assessments so that these variables can be quantified using psychometric instruments.

It also should be noted that we screened all candidates with a behavioral measure that was developed in a population of children from a geographically and culturally distinct region. The influence of these variables on parent perceptions of feeding problems and feeding practices is not known. Care should be taken when making any conclusions about the comparability of these groups. In addition, when interpreting these data, the children's developmental status and comorbid health conditions should not be overlooked. All of our patients had physical and medical problems that may have contributed to their feeding problems. Nonetheless, clinical evaluation suggested that none of these conditions precluded safe oral feeding. Furthermore, although some children in the sample had mild to moderate developmental delay based on history and clinical observation, no formal neurodevelopmental evaluation using psychometric measures was obtained. Experienced therapists clinically considered the patients' developmental status when developing and implementing treatment. Patients referred for treatment demonstrated sufficient appreciation of cause/effect relationships and receptive language to respond to operant learning principles. More formal measures of neurocognitive status would be helpful in future studies to potentially differentiate between treatment responders and nonresponders based on neurocognitive factors.

In summary, our findings support the efficacy of short-term intensive feeding treatment when a multicomponent approach is used. Critical treatment targets included in this program included appetite stimulation, intensive short-term behavioral feeding treatment, medical and nutritional management, and parental skills training. Our clinical experience suggests that each of these treatment variables plays a critical role in treatment. These data emphasize the need for a biobehavioral approach when evaluating and treating medically complicated feeding disorders.


1. Rudolph CD, Link DT. Feeding disorders in infants and children. Pediatr Clin North Am 2002; 49( 1):97–112.
2. Beautrais Al, Ferguson DM, Shannon FT. Family life events and behavioral problems in pre-school aged children. Pediatrics 1982; 70:774–9.
3. Forsyth BW, Leventhal JM, McCarthy PJ. Mother's perceptions of feeding and crying behaviors. Am J Dis Child 1985; 139:269–72.
4. Linscheid TR. Behavioral treatment of feeding disorders in children. In: Watson TS, Gresham, FM, eds. Behavioral Treatment of Childhood Disorders. New York: Plenum; 1998:357–68.
5. Stevensen RD. Feeding and nutrition in children with developmental disabilities. Pediatr Ann 1995; 24( 5):255–50.
6. Mackay LE, Morgan AS, Bernsteun BA. Swallowing disorders in severe brain injury: risk factors affecting return to oral intake. Arch Phys Med Rehabil 1999; 80:365–71.
7. Crist W, Napier-Phillips A. Mealtime behaviors of young children: a comparison of normative and clinical data. J Dev Behav Pediatr 2001; 22( 5):279–86.
8. Burklow KA, Phelps AN, Schultz JR, et al. Classifying complex pediatric feeding disorders. J Pediatr Gastroenterol Nutr 1998; 27( 2):143–7.
9. Babbitt RL, Hoch TA, Coe DA, et al. Behavioral assessment and treatment of pediatric feeding disorders. J Dev Behav Pediatr 1994; 15( 4):278–91.
10. Kerwin ME. Empirically supported treatments in pediatric psychology: severe feeding problems. J Pediatr Psychol 1999; 24( 3):193–214.
11. Dellert SF, Hyams JS, Treem WR, et al. Feeding resistance and gastroesophageal reflux in infancy. J Pediatr Gastroenterol Nutr 1993; 17:66–71.
12. Morris SE. Development of oral-motor skills in the neurologically impaired child receiving non-oral feedings. Dysphagia 1989; 3:135–54.
13. Luiselli JK, Luiselli TE. A behavior analysis approach toward chronic food refusal in children with gastrostomy-tube dependency. Top Early Child Spec Educ 1995; 15( 1):1–18.
14. Blackman JA, Nelson CL. Rapid introduction of oral feedings to tube-fed patients. J Dev Behav Pediatr 1987; 8( 2):63–7.
15. Blackman JA, Nelson CL. Reinstituting oral feedings in children fed by gastrostomy tube. Clin Pediatr (Phila) 1985; 24( 8):434–8.
16. Miller CK, Burklow KA, Santoro K, et al. An interdisciplinary team approach to the management of pediatric feeding and swallowing disorders. Child Health Care 2001; 30( 3):201–18.
17. Gutentag S, Hammer D. Shaping oral feeding in a gastronomy tube-dependent child in natural settings. Behav Modif 2000; 24( 3):395–410.
18. Kedesdy JH, Budd KS. Childhood Feeding Disorders: Biobehavioral Assessment and Intervention. Baltimore: Brookes; 1998.
19. Crist W, McDonnell P, Beck M, et al. Behavior at mealtimes and the young child with cystic fibrosis. J Dev Behav Pediatr 1994; 15:157–61.
20. Himes JH. Anthropometric Assessment of Nutritional Status. New York: Wiley-Liss; 1991.
21. Stark LJ, Jelalian E, Powers SW, et al. Parent and child mealtime behavior. J Pediatr 2000; 136( 2):195–200.
22. Hoch TA, Babbitt RL, Coe DA, et al. Contingency contacting: combining positive reinforcement and escape extinction procedures to treat persistent food refusal. Behav Modif 1994; 18( 1):106–28.
23. Ives CC, Harris SL, Wolchik SA. Food refusal in an autistic type child treated by a multicomponent forced feeding procedure. J Behav Ther Exp Psychiatry 1978; 9( 1):61–4.
24. Riordan MM, Iwata BA, Finney JW, et al. Behavioral assessment and treatment of chronic food refusal in handicapped children. J Appl Behav Anal 1984; 17( 3):327–41.
25. Palmer S, Thompson RJ, Linscheid TR. Applied behavior analysis in the treatment of childhood feeding problems. Dev Med Child Neurol 1975; 17( 3):333–9.
26. Riordan MM, Iwata BA, Wohl MK, et al. Behavioral treatment of food refusal and selectivity in developmentally disabled children. Appl Res Ment Retard 1980; 1(1-Suppl-2):95–112.
27. Perske R, Clifton A, McLean B, et al., eds. Mealtimes for Severely and Profoundly Handicapped Persons: New Concepts and Attitudes. Oxford: Random House; 1977.
28. Farrell DA, Hagopian LP, Kurtz PF. A hospital- and home-based behavioral intervention for a child with chronic food refusal and gastrostomy tube dependence. J Dev Phys Dis 2001; 13( 4):407–18.
29. Benoit D, Wang EE, Zlotkin SH. Discontinuation of enterostomy tube feeding by behavioral treatment in early childhood: a randomized controlled trial. J Pediatr 2000; 137( 4):498–503.
30. Linscheid TR, Tarnowski KJ, Rasnake LK, et al. Behavioral treatment of food refusal in a child with short-gut syndrome. J Pediatr Psychol 1986; 12( 3):451–9.
31. Lamm N, Greer RD. Induction and maintenance of swallowing responses in infants with dysphagia. J Appl Behav Anal 1988; 21( 2):143–56.
32. Luiselli JK. Cueing, demand fading, and positive reinforcement to establish self-feeding and oral consumption in a child with chronic food refusal. Behav Modif 2000; 24( 3):348–58.
33. Rudolph CD. Feeding disorders in infants and children. J Pediatr 1994; 125(6 Pt 2):S116–24.

Behavior therapy; Gastroesophageal reflux; Enteral nutrition; Fundoplication; Childhood feeding disorders

© 2003 Lippincott Williams & Wilkins, Inc.