Impact of a Responsive Feeding Intervention on Children and Caregivers: A Nonrandomized,Repeated-Measures Study

Abstract

Importance: Limited research exists examining the long-term impact of a parent-mediated responsive feeding intervention on children’s mealtime skills and behaviors and caregiver stress and competence.

Objective: To examine children’s changes in mealtime skills and behaviors and to examine caregiver stress and competence.

Design: A nonrandomized repeated-measures study.

Setting: Home-based.

Participants: Twenty-five children who received more than 10% of their calories through tube feeding and from their parents.

Intervention: Caregivers participated in a home-based responsive feeding intervention directed at changing caregiver–child interactions during mealtimes. After receiving general caregiver education, interventions were delivered three to five times daily for 10 days, followed by 6 mo. of coaching sessions as needed.

Outcome and Measures: Clinical measures and the Children’s Eating Behaviour Questionnaire were used to measure mealtime skills and behaviors. The Parent Stress Index–Short Form–4 and the Parent Sense of Competence Scale were used to measure caregiver competency and stress.

Results: Seventeen caregivers completed the assessment battery for at least five of the six time points. All participants reduced tube use. Significant differences existed in children’s mealtime skills and behaviors over time with continued improvement with follow-up coaching. Caregivers continued to feel stressed while their competence improved.

Conclusions and Relevance: The intervention was effective at helping children transition to oral feeding and develop oral motor skills. It was not effective at reducing caregiver stress.

Plain-Language Summary: This study was conducted to determine whether responsive feeding interventions can help children who are tube fed learn to eat by mouth and develop skills. The results from this study support therapy feeding interventions that target caregiver–child interactions. It also suggests that children with motor impairments can still improve chewing, biting, utensil use, and hand use by participating in motivating and relevant activities. These improvements can occur with continued coaching only. The study showed that children’s mealtime skills and behaviors continued to improve after the intense intervention stopped and with coaching only. Caregivers may feel stressed about children’s eating, despite successful intervention. Occupational therapists who work with tube-fed children should focus interventions on caregiver–child interactions. Therapists should also monitor caregiver stress and mental health, even if the child learns to eat well.

The results of this study support occupational therapy feeding interventions that target caregiver–child interactions.

Infants and children with medical complexities often require enteral nutrition to survive, but transitioning to oral feeding can be challenging, even after medical stabilization (Edwards et al., 2016; Krom et al., 2017). Tube feeding, although essential, carries risks and may hinder the development of hunger and satiation recognition, which can impair self-regulation (Krom et al., 2017; Wilken et al., 2018). A multifaceted approach, combining behavioral, parent-directed, oral motor, and appetite-inducing interventions, is supported by research to aid in the transition from tube feeding to oral intake (Edwards et al., 2016).

Appetite-based interventions use hunger as a natural motivator for eating, and they help children develop self-regulation skills that are often delayed by prolonged tube use (Marinschek et al., 2019). These interventions, alongside responsive feeding practices, prioritize the child’s cues and emotional safety by encouraging healthy caregiver–child interactions (Garcia et al., 2022). Responsive feeding is an approach to feeding that is recommended to parents by the American Academy of Pediatrics (AAP; 2017), the World Health Organization (WHO; 2023), and the American Heart Association (Wood et al., 2020). Parents are coached to engage with their children at mealtimes and eliminate pressure to eat, instead following their child’s cues of hunger and fullness and prioritizing the child’s felt safety. Responsive feeding therapy (RFT) is a specific parent-mediated intervention approach based on responsive feeding that combines various evidence-based feeding interventions aimed at caregiver–child interaction. RFT involves prioritizing the child’s feeling of safety to allow for the “rediscovery” of internal cues (e.g., hunger, food curiosity) while simultaneously building skills, confidence, autonomy, and relationships (Pérez-Escamilla et al., 2017). Although there is literature on responsive feeding research, limited evidence exists that examines the effect of RFT on children’s mealtime skills and behavioral outcomes over time. Furthermore, although RFT did not arise from occupational therapy literature, RFT is an intervention approach aligned with occupational therapy theory and research. Henton (2018) proposed the relationship-based framework in occupational therapy to address needs in children with feeding problems. Like RFT, this framework emphasizes a holistic perspective of the child within their various contexts (e.g., home, community, school) and the relationships and interactions between the child and the child’s caregivers.

Not only does RFT align with occupational therapy theory by emphasizing the impact of the child’s context on participation, but it also aligns with research within occupational therapy that targets occupational therapy’s role in addressing interoception. Interoception is defined as one’s understanding and sense of internal states and physiology (Craig, 2003). Recent literature on interoception supports the role of occupational therapists in supporting participation for people whose interoceptive skills affect their participation (Brown & Dunn, 2023). Understanding internal needs such as hunger is an interoceptive sense and is necessary to maintain a feeling of safety and homeostasis (Schmitt & Schoen, 2022). Literature also suggests that stress at mealtimes, including caregiver–child interaction, influences children’s interoception and feeding skills (Filippetti, 2021). RFT is an intervention that targets a child’s interoceptive skills by focusing on the caregiver’s understanding and response to their child’s hunger cues.

Although RFT aligns with occupational therapy theory, research on its impact on children’s oral motor outcomes and caregiver stress remains limited. This study aims to answer whether RFT, appetite induction, and coaching improve children’s feeding skills and behaviors, as well as caregiver stress in a home-based protocol. We hypothesize that children’s mealtime skills and behaviors will improve, caregiver competence will improve, whereas caregiver stress remains high. By addressing eating, family stress, and child development, this research contributes to the broader understanding of successful mealtime participation.

Method

Design and Setting

This nonrandomized, repeated-measures feasibility study examined children’s mealtime skills and behaviors in a tube-weaning program and their caregivers’ stress and competence. Data were collected at evaluation (Time 1); Day 1 of wean (Time 2); Day 10 (wean end; Time 3); and at 1, 3, and 6 mo. The time between evaluation and Time 2 was between 1 and 9 mo., with a mean of 108 d. Four speech therapists and one occupational therapist, each with at least 3 yr of feeding experience and a master’s degree, delivered the intervention after completing mentorship training. Mentorship training took 3 to 6 mo and occurred before the study began; it consisted of shadowing, then delivering the intervention in a pair, and then leading a tube wean. Each participant was assigned two therapists during the intervention, each blinded to the assessment results. The team met weekly to discuss participant progress and rate participant skills and behaviors.

Participants

Participants who sought home-based tube-weaning interventions through a private practice were invited to participate in the study, and informed consent was obtained during the evaluation. Inclusion criteria were as follows:

▪ clearance by a physician to participate

▪ consumed more than 10% of their calories by tube

▪ demonstration of a safe swallow.

Although children with a diagnosis of autism could participate in this intervention, their treatment length was often longer and more variable, so they were excluded from this preliminary study. Children with developmental delays were not excluded. Participant characteristics are displayed in Table 1. Researchers relied on caregivers’ report of their child’s diagnoses. One caregiver disclosed that their child had a developmental delay, and four other caregivers reported genetic conditions. The University of Kansas Medical Center Institutional Research Board approved all study procedures and caregivers’ consent to participate in this study. Participants received a $50 gift certificate for completing the entire assessment battery.

Table 1.

Demographic Information (N = 17)

Characteristic	n	%
Age at evaluation, mo
0–12	5	30
13–24	5	30
25–36	3	17
37–48	1	6
49–60	2	11
≥61	1	6
Race
White	14	82.3
Asian	2	5.8
Unknown	1	5.8
Gender assigned at birth
Female	10	58.8
Male	7	41.2
Co-occurring conditions
Lung condition (e.g., bronchopulmonary dysplasia, tracheomalacia)
Prematurity	9	52.9
Genetic condition	8	47
Reflux vomiting	4	23.5
Cerebral palsy–hypotonia	4	23.5
Failure to thrive	3	17.6
Heart condition	3	17.6
Intrauterine growth restriction	3	17.6
Developmental delay	2	11.7
Co-occurring conditions
Lung condition (e.g., bronchopulmonary dysplasia, tracheomalacia)	9	52.9
Prematurity	8	47
Genetic condition	4	23.5
Reflux vomiting	4	23.5
Cerebral palsy–hypotonia	3	17.6
Failure to thrive	3	17.6
Heart condition	3	17.6
Intrauterine growth restriction	2	11.7
Developmental delay	1	5.8
Length of tube use, mo
≥24 mo	5	29.4
12–24	7	41.1
6–12 mo	3	17.6
3–6 mo	2	11.7
Household income
<50,000	1	5.8
100,000–199,999	11	64.7
≥200,000	5	29.4

Intervention

Intervention consisted of two parts: an intensive home-based tube weaning (RFT) paired with appetite induction and caregiver coaching; and coaching after the intensive intervention ended. Between Time 1 and Time 2, caregivers had three, 1-hr virtual educational sessions, and the average time between Time 1 and Time 2 was 2 mo. Hunger induction—which consisted of reducing the participant’s calories by 10% per day—began 5 d before Time 2 and was guided by their physician and dietician. During the intensive portion of the intervention (Time 2–Time 3), coaching occurred three to five times daily virtually and in person. Mealtime coaching focused on supporting developmentally appropriate autonomy for the child, food choices, food presentations, mealtime interactions, and reading and responding to the child’s cues regarding mealtime comfort and when to finish the meal. Therapists monitored wellness by observing the child’s mood, wet diapers, stooling patterns, and energy levels, whereas the child’s physician monitored weight. Therapists were present during mealtimes for coaching, and sessions focused on responsive feeding (e.g., responding to cues, reducing stress). After Time 3, virtual coaching continued as needed through 6 mo., with a mean number of 17 sessions.

Measures

Families who were enrolled in the study completed the assessment measures at each assessment timepoint. The participant’s therapist collected data on mealtime skills and behavioral outcomes at each of the study’s time points. Mealtime skills were measured on a 7-point scale (1 = pharyngeal swallow, 2 = drinking from bottle or breast, 3 = straw use, 4 = purees, 5 = chewing, 6 = self-feeding by hand, and 7 = utensil use). Feeding behaviors were measured on a 5-point scale (1 = initiation, 2 = attention to eating, 3 = eating enjoyment, 4 = reflex/vomiting, and 5 = food received through tube). Mealtime skills and behaviors were operationalized on the basis of evidence on developmental milestones (e.g., Morris & Klein, 2000) and clinical expertise. Data were gathered through observation and parent report using the clinical checklist and occurred during follow-up coaching sessions. Therapists ensured reliability through fidelity checking during weekly meetings.

The Children’s Eating Behaviour Questionnaire (CEBQ) is a 35-item caregiver questionnaire that examines children’s responses to eating on the basis of caregiver report (Wardle et al., 2001) and has been widely used to research the eating behaviors of children with feeding disorders, which include overeating and picky eating (e.g., Ahmad et al., 2020). The questionnaire analyzes feeding across eight domains: Food Responsiveness (FR), Enjoyment of Food (EF), Emotional Overeating (EOE), Desire to Drink (DD), Satiety Responsiveness (SR), Slowness in Eating (SE), Emotional Undereating (EUE), and Food Fussiness (FF). Parents report their child’s behaviors on a Likert scale ranging from 1 (never) to 5 (always). The SR, SE, and EUE subscales have been associated with undereating, whereas the FR, EF, and EOE subscales have been associated with a higher BMI (Kininmonth et al., 2021).

We assessed parent stress using the Parent Stress Index–Short Form–4 (PSI–SF–4), a 36-item questionnaire assessing caregiver stress within three domains—Parent Distress, Parent–Child Dysfunctional Interaction, and Difficult Child—as well as a Total Stress outcome (Abidin et al., 2006). A total score can be derived by adding the three domains together. The PSI–SF–4 has been used in research to examine stress over time, within the context of tube weaning (e.g., Greer et al., 2008; Silverman et al., 2013), and within occupational therapy research (e.g., Poe & Martin, 2023). The PSI–SF–4 demonstrates internal consistency and stability in scores over time (Haskett et al., 2006). For the purposes of this study, only the Total Stress score was evaluated across time.

We assessed parent competence using the Parent Sense of Competence Scale (PSOC; Johnston & Mash, 1989), a 17-item Likert scale that examines caregiver competence across three domains: Satisfaction, Efficacy, and Interest (Gilmore & Cuskelly, 2009). The PSOC has been used in occupational therapy literature to examine the impact of an intervention on parent competence and to measure change (e.g., Dunn et al., 2012; Fabrizi & Hubbell, 2017); and it demonstrates good internal consistency, validity and reliability (Nunes et al., 2022; Ohan et al., 2000).

Data Analysis

Researchers used descriptive statistics to summarize demographic characteristics (Table 1) and display means and standard deviations for all variables at each time point (Table 2). After assessing assumptions of normality, we used a repeated-measures analysis of variance (ANOVA) with least significant difference post hoc analysis to assess time effects. Given the exploratory nature of the study, significance was set at .05 without corrections for familywise error (Portney & Watkins, 2009), prioritizing the detection of meaningful differences over Type 1 error control. Missing data from the CEBQ, PSI–SF–4, and PSOC were imputed by averaging responses to compute total and scale scores. We also calculated effect sizes to highlight clinically meaningful differences, especially given the small sample size.

Table 2.

Means and Standard Deviations for Measures at All Time Points (N = 17)

Measure	M (SD)
	Time 1	Time 2	Time 3	Time 4	Time 5	Time 6
Clinic measure
n	17	17	17	17	17	17
% feeding by tube	3.71 (0.99)	3.53 (1.01)	1.94 (1.35)	1.18 (1.19)	0.53 (0.80)	0.47 (0.94)
Reflux vomiting	1.82 (1.29)	1.53 (1.18)	0.29 (0.47)	0.12 (0.33)	0.12 (0.33)	0.06 (0.24)
Biting	1.88 (1.93)	2.06 (1.89)	3.47 (1.66)	4.53 (1.74)	5.29 (1.05)	5.53 (1.07)
Chewing	1.65 (1.62)	1.82 (1.55)	3.68 (1.63)	4.12 (1.69)	4.94 (1.44)	5.18 (1.51)
Hand use	2.94 (1.92)	3.41 (1.62)	4.29 (0.99)	4.94 (0.90)	5.35 (0.61)	5.59 (0.62)
Utensil use	2.18 (2.16)	2.29 (2.05)	3.29 (1.49)	3.88 (1.36)	4.47 (1.13)	5.00 (0.79)
CEBQ subscale
n	16	17	16	15	16	16
Food Responsiveness	1.33 (0.40)	1.55 (0.51)	1.69 (0.49)	1.93 (.59)	2.25 (0.66)	2.13(0.64)
Emotional Overeating	1.08 (0.22)	1.13 (0.22)	1.19 (0.25)	1.32 (0.22)	1.53 (0.46)	1.50 (0.32)
Enjoyment of Food	2.33 (0.85)	2.68 (0.92)	3.30 (0.71)	3.70 (0.68)	3.83 (0.83)	3.81 (0.62)
Desire to Drink	1.83 (0.88)	2.31 (1.02)	2.27 (1.01)	2.29 (0.72)	2.77 (0.85)	2.96 (0.85)
Satiety Responsiveness	3.45 (0.86)	3.19 (0.61)	3.38 (0.66)	3.25 (0.73)	3.33 (0.53)	3.36 (0.54)
Slowness in Eating	3.75 (0.98)	3.50 (0.83)	3.58 (0.57)	3.75 (0.81)	3.38 (0.71)	3.52 (0.70)
Emotional Undereating	2.67 (1.47)	2.78 (1.39)	3.15 (1.37)	3.40 (1.12)	3.46 (1.11)	3.63 (1.22)
Food Fussiness	3.10 (0.71)	2.89 (0.48)	2.80 (0.54)	3.03 (0.61)	2.83 (0.49)	3.01 (0.41)
PSI–SF–4 subscale
n	17	17	16	15	17	16
Total Stress	128.47 (16.28)	129.47 (18.16)	138.81 (18.77)	133.07 (20.13)	138.35 (18.60)	143.40 (15.26)
Parent Distress	40.71 (8.89)	40.59 (9.59)	46.50 (9.25)	44.47 (9.10)	47.65 (10.26)	50.13 (10.42)
Difficult Child	43.00 (6.40)	43.06 (7.60)	45.19 (6.55)	42.93 (9.08)	44.41 (6.20)	45.94 (5.67)
Parent–Child Dysfunctional Interaction	48.94 (5.29)	49.47 (5.80)	51.56 (5.92)	49.93 (5.48)	50.47 (5.52)	52.27 (3.56)
PSOC subscale
n	17	17	16	15	17	16
Total Stress	48.76 (5.18)	49.29 (5.82)	49.19 (4.82)	51.27 (6.19)	50.12 (4.69)	51.56 (6.22)
Efficacy	23.06 (4.55)	22.71 (5.28)	24.13 (5.21)	24.27 (4.38)	23.59 (5.18)	26.69 (3.36)
Satisfaction	16.82 (3.43)	17.41 (3.79)	16.00 (5.25)	17.60 (5.77)	16.94 (4.46)	15.25 (5.73)
Interest	8.88 (1.93)	9.18 (1.33)	9.06 (1.24)	9.40 (1.35)	9.59 (1.50)	9.63 (1.78)

Note. CEBQ = Children’s Eating Behaviour Questionnaire; PSI–SF–4 = Parent Stress Index–Short Form–4; PSOC = Parent Sense of Competence Scale.

Results

Twenty-five participants initially enrolled in the study. Of the initial 25 participants, 1 dropped out of the program for medical reasons, and 6 completed data collection at four or fewer time points. One participant completed all data collection but resumed tube feeding after an adverse medical event unrelated to the study and so was excluded from analysis. Seventeen participants completed surveys for at least five of the six data collection times and were included in analysis. All 17 participants completed coaching sessions through 6 mo.

Mealtime and Behavioral Outcomes

All participants reduced tube use by the end of the study. Six participants were tube-free by the end of the intensive intervention and continued to be tube-free. Ten additional participants continued to show improved oral intake with coaching intervention only. One participant made improvements in oral feeding between Time 1 and Time 6; however, because of illness, their tube use between times was variable. Feeding and behavioral outcomes were complete for 17 participants. ANOVA results showed significant differences in mean scores over time for biting, F(5, 80) = 33.386, p ≤ .001, η² = .676; chewing, F(5, 80) = 41.572, p ≤ .001, η² = .772; hand use, F(5, 80) = 18.881, p ≤ .001, η² = .541; utensil use F(5, 80) = 23.804, p ≤ .001, η² = .598; percent feeding by tube F(5, 80) = 46.718, p ≤ .001, η² = .745; and reflux vomiting, F(5, 80) = 23.983, p ≤ .001, η² = .600. Means for biting, chewing, hand use, and utensil use increased over time, whereas means for percent feeding by tube and reflux vomiting decreased over time. Mean motor skills improved for the 3 participants with a diagnosis of hypotonia–cerebral palsy (Figure 1).

Figure 1.

Means for child and caregiver outcomes.

CEBQ

Twelve participants completed the CEBQ at all time points; completion of items for computing subscales ranged from 11 to 12 participants. The ANOVA results showed significant differences in mean scores over time for half of the CEBQ subscales: FR, F(5, 50) = 4.908, p ≤ .001, η² = .329; EOE, F(5, 50) = 4.613, p = .002, η² = .316; EF, F(5, 50) = 14.620, p ≤ .001, η² = .594; and DD, F(5, 50) = 2.509, p = .042, η² = .201. The ANOVA showed no significant difference in mean scores over time for FF, F(5, 55) = 2.283, p = .059; SR, F(5, 50) = .608, p = .694; SE, F(5, 50) = 0.820, p = .541; and EUE, F(5, 50) = 1.209, p = .309.

PSI–SF–4

Fourteen participants had complete data at all timepoints on the PSI–SF–4. The ANOVA results showed a significant increase in differences in mean scores for Total Stress, F(5, 60) = 3.923, p = .004, η² = .246; Parent Distress, F(5, 60) = 6.332, p ≤ .001, η² = .345; and Parent–Child Dysfunctional Interaction, F(5, 60) = 2.796, p = .025, η² = .193. There was no significant difference for Difficult Child, F(5, 60) = 1.017, p = .416.

PSOC

Fourteen participants had complete data at all timepoints on the PSOC. Over time, ANOVA results showed no significant difference in mean total PSOC scores, F(5, 60) = 0.969, p = .444; and the subscale scores for Satisfaction, F(5, 60) = 0.586, p = .710; and Interest, F(5, 60) = 1.773, p = .132. However, there was a significant difference over time in the PSOC Efficacy subscale, F(5, 60) = 3.583, p = .007, η² = .230.

Within-Timepoint Differences

Significant differences in timepoints of interest are displayed in Table 3. Specifically, we were interested in differences between the following:

Table 3.

Significant Mean Differences for Main Outcomes at Baseline, After Immediate Treatment, and at 6 Mo

Measure	Time 1–Time 2			Time 2–Time 3			Time 2–Time 6
	MD	SE	p	MD	SE	p	MD	SE	p
Behavioral or oral motor
% feeding by tube	0.176	0.356	.627	1.588	0.333	<.001	3.059	0.326	<.001
Reflux vomiting	0.294	0.294	.332	1.235	0.304	<.001	1.471	0.273	<.001
Biting	−0.176	0.095	.083	−1.412	0.374	.002	−3.471	0.471	<.001
Chewing	−0.176	0.095	.083	−1.853	0.358	<.001	−3.353	0.428	<.001
Hand use	−0.471	0.322	.163	−0.882	0.401	.043	−2.176	0.386	<.001
Utensil use	−0.118	0.225	.608	−1.000	0.309	.005	−2.706	0.435	<.001
CEBQ subscale
Food Responsiveness	−0.182	0.176	.326	−0.091	0.169	.602	−0.424	0.191	.051
Emotional Overeating	−0.091	0.127	.492	0.023	0.092	.810	−0.273	0.102	.059
Enjoyment of Food	−0.318	0.276	.276	−0.682	0.172	.003	−1.045	0.278	.004
Desire to Drink	−0.545	0.243	.049	0.242	0.186	.221	−0.424	0.331	.228
PSI–SF–4 subscale
Total Stress	−4.615	9.95	.144	−9.154	1.87	<.001	−9.692	4.78	.047
Defensive Responding	−0.385	1.08	.727	−4.385	0.91	<.001	−5.846	1.66	.004
Parent Distress	−0.846	1.59	.604	−6.385	1.22	<.001	−8.538	2.56	.006
PSOC subscale
Efficacy	0.462	0.637	.482	−0.432	0.645	.069	−3.154	0.960	.007

Note. CEBQ = Children’s Eating Behaviour Questionnaire; MD = mean difference; PSI = Parent Stress Index–Short Form–4; PSOC = Parent Sense of Competence Scale.

• ▪ Time 1 and Time 2: the baseline period when parents received three educational sessions in RFT but children had not begun frequent coaching sessions and no direct tube-related intervention was provided

• ▪ Time 2 and Time 3: immediate intervention effects when tube-related nutrition was significantly reduced and therapists provided intense, daily intervention

• ▪ Time 2 and Time 6: effects of the intervention over time.

For clinically measured mealtime and behavior outcomes, there were no significant differences from Time 1 to Time 2 during the baseline period, but there were significant improvements in all outcomes from Time 2 to Time 3 and from Time 2 to Time 6. For the three participants with a diagnosis of hypotonia–cerebral palsy, mean scores for chewing, biting, utensil use, and hand use improved from Time 2 to Time 3 and from Time 3 to Time 6. indicating that the intervention provided both immediate and long-term benefits, regardless of the amount of continued coaching that families received. Post hoc testing of significant outcomes of the CEBQ showed a significant change in DD scores from Time 1 to Time 2, as well as significant changes in EF scores from Time 2 to Time 3 and from Time 2 to Time 6, whereas FR scores only changed significantly from Time 2 to Time 6, and EOE scores did not change significantly for any of the time points of interest. The PSI–SF–4 Total score and the scores of two subscales (Defensive Responding and Parent Distress) did not change significantly during the baseline period (Time 1–Time 2) but increased significantly from Time 2 to Time 3 and from Time 2 to Time 6, indicating immediate change after the intervention that sustained over time. The Parent Efficacy subscale of the PSOC was only significant from Time 2 to Time 6.

Discussion

Transitioning from tube to oral feeding can be challenging for both children and caregivers. Although the primary goal of weaning is oral eating, other developmental and psychosocial outcomes may emerge. In this study, we aimed to assess changes in children’s mealtime skills and behaviors, as well as parental competence and stress, after a two-part intervention. We hypothesized that children’s mealtime skills and behaviors would improve, whereas caregiver competence would improve even as they remained stressed. Results showed that the intervention effectively helped children transition to oral eating and develop positive mealtime skills and that continued coaching supported further mealtime skill development. However, caregiver stress increased over time, despite a rise in caregiver competence, suggesting that feeding children can remain stressful, despite increasing caregiver competence.

Children’s Outcomes

This study supports the principle of occupational therapy in that children developed underlying skills through occupational participation. Children’s eating behaviors and skills improved significantly as tube use decreased, suggesting that the intervention helped them develop mealtime and behavioral skills. Although the intervention did not focus directly on chewing or other skills (e.g., biting, utensil use), these skills emerged for most of the participants, even after the intensive intervention ended, which supports previous research that occupation-based interventions and coaching promote the development of underlying skills (e.g., Laverdure & Beisbier, 2021). Caregiver competence also improved as children maintained oral intake and skills postintervention.

For the CEBQ, FR, EF, and EOE subscale scores increased significantly. These subscales have also been linked to children with a higher BMI (Kininmonth et al., 2021). For children who have been chronic “undereaters,” scoring a significant increase on a scale associated with children who have a higher BMI is an interesting outcome. It is also possible that caregivers reported that their child was emotionally overeating and responding to food more because their child was now eating even when upset, when, previously, they would only eat in “perfect” conditions, if at all. Furthermore, the CEBQ subscale scores that have been associated with anorexia (FR, EF, and EOE) were not significant over time, indicating that there was no increase in signs of restrictive eating, because anorexia-related scores were not significant.

Caregiver Outcomes

The results of this study suggest that caregivers of tube-fed children continue to experience stress even after their child learns to eat and with continued coaching. Although caregiver stress increased over time, so did caregiver competence, suggesting that although caregivers felt more stress, they also felt more competent in their caregiving role. This is unsurprising, considering that caregivers moved from relying on the tube to trusting their child to self-feed. It also highlights the need for continued monitoring of caregivers of children with medical needs and that they may be at higher risk for posttraumatic stress (Carmassi et al., 2020).

Descriptive statistics also show that there was an increase in caregiver stress between Time 2 and Time 3, a decrease in stress between Time 3 and Time 4, and then a steady increase again between Times 4 and 6 (Figure 1). Although there was no significant difference in parental efficacy over time, there was a significant difference in the PSOC Efficacy subscale, indicating that caregivers felt more competent in their parenting role over time. It is also important to note that oral intake and skills were maintained once the intensive treatment was over, which may indicate that the parent competence extended to creating and maintaining a mealtime context that was conducive to oral intake, despite caregiver stress.

Limitations

Although 25 families initially completed the evaluation and consent, only the data from 17 children were analyzed because of dropouts. Furthermore, of the 17 caregivers, some failed to answer every question on the caregiver assessments, leaving a sample of 12 caregivers with fully completed CEBQs and 14 fully completed PSI–SF–4 and PSOC assessments. The small sample size limited our ability to generalize results. Future studies with a larger sample size should compare group outcomes for age and extended coaching. Another limitation was that researchers did not use a standardized assessment for mealtime skills and behaviors, primarily because no standardized assessment was used widely at the time of data collection. Instead, the authors used a measure that was developed at the clinic and familiar to the clinicians collecting the data. Next, whereas most studies that examine the efficacy of a tube-weaning program track children’s weight, this study did not, limiting the ability to measure whether RFT can support weight gain. Finally, the sample demographic was not heterogenous. The sample was a convenience sample from a clinic, and insurance coverage was inconsistent, which meant that families often used out-of-pocket funds to pay for their services. Of the 17 families, only one reported an income of less than $50,000; all other families earned over $100,000 annually. Thus, this intervention may not be transferable to families with fewer resources.

Implications for Occupational Therapy Practice

Some tube-fed children require intensive therapy to learn to eat by mouth. Few studies have focused on how parent-mediated, responsive feeding paired with appetite induction affects the long-term skills and behaviors of children and their caregivers. The results of this study have the following implications for occupational therapy practice:

▪ Therapists should target caregiver–child interaction during mealtimes. Often, therapists target feeding skills when the immediate focus of therapy should always begin with child’s comfort, the child–caregiver relationship, and contextual aspects of mealtime.

▪ While more research needs to occur, there is promising evidence that therapists should use responsive feeding during mealtime coaching sessions with families. Clinicians who choose to use this approach should carefully document treatment outcomes and monitor caregiver stress.

▪ Children can develop underlying motor skills by participating in meals, even if the underlying skills are not directly addressed.

▪ Feeding is stressful, and caregivers are stressed during and after their children learn to eat.

▪ Intensive tube weaning can occur in home and community contexts.

▪ Interventions that focus on the child–parent interaction and child’s interoceptive skills may be effective at helping children develop underlying feeding and oral motor skills. However, larger, and more rigorous studies need to occur.

▪ Consider family resources. Therapists should consider the demographics and resources of this study compared with their own.

Conclusion

In this study, we examined the impact of RFT on children’s mealtime skills and behaviors and their caregiver’s sense of stress and competence. Current literature on RFT’s impact on mealtime participation is limited. This study provides preliminary evidence that children’s mealtime skills and behaviors can improve after the intensive intervention ends, with therapist coaching. This study also suggests that, although caregivers of tube-fed children can develop competence because of this intervention, they may remain stressed in the long term.