Abstract
Children with feeding disorders often cannot or do not chew when presented with table food. Children with chewing deficits also often swallow the bite before masticating it appropriately, which we will refer to as early swallowing. In the current study, we evaluated a clinical protocol to increase chews per bite, assess mastication, and eliminate early swallowing with three children with feeding disorders. The current study adds to a small body of literature on chewing and mastication of children with feeding disorders. Suggestions for future research are also discussed.
For most typically eating children, the efficiency of chewing improves with experience and time. Immature chewing emerges at about 6 months of age. Most parents introduce small amounts of soft or meltable table food between 7 and 9 months of age. Parents then increase the amount of table food and introduce table foods that are more difficult to chew as they observe the child successfully manage the advanced textures (Volkert, Piazza, Vaz, & Frese, 2013). Most children consume a diet that is largely composed of table food by 24 months of age (Carruth, Ziegler, Gordon, & Hendricks, 2004; Illingworth & Lister, 1964), and fully mature chewing skills emerge by age 4 (Edwards & Martin, 2011).
By contrast, children with feeding disorders often do not chew when presented with table food or do not exhibit developmentally typical chewing patterns (Volkert et al., 2013). Although it is not clear why chewing behavior does not emerge in the typical developmental sequence in children with feeding disorders, Illingworth and Lister (1964) have suggested that children may fail to develop chewing skills if they are not presented solid food at 6 or 7 months of age, which they describe as the critical period for chewing. This hypothesis may be relevant to children with feeding disorders who often are not exposed to oral nutrition during this period due to medical problems (e.g., aspiration, prematurity) and/or refusal behavior (e.g., gagging, vomiting, head turning).
A child who consumes regular-textured food must be able to perform a complex chain of interdependent responses to be a successful feeder. The child must allow the bite into his or her mouth, move the bite to the molars, grind the bite until it is masticated, move the bite from the molars back to the tongue, and elevate the tongue and propel the bolus into the pharynx to swallow. Failure to successfully complete one response in the chain affects the child’s ability to complete subsequent responses in the chain. For example, if the child does not accept the bite, we cannot assess his or her chewing skills. Or, if the child is unable to move the bite to the molars once it enters the mouth, he or she will not have the opportunity to masticate the bite. Many children with feeding disorders do not exhibit some or all of the responses in the chain necessary to consume regular-textured food, and intervention is necessary for them to do so.
A limited number of studies have focused on teaching chewing skills to children with feeding disorders (Butterfield & Parson, 1973; Eckman, Williams, Riegel, & Paul, 2008). For example, Eckman et al. (2008) used oral-motor techniques, nonremoval of the bite, reinforcement, shaping, and fading to teach two children with a feeding disorder to chew. Although the treatment resulted in increased chewing, it was not clear which component(s) of the treatment was responsible for the increase or whether the increase resulted in effective mastication.
Volkert et al. (2013) evaluated the efficacy of least-to-most prompting plus praise with one child and descriptive prompts and praise with another child to increase chews per bite. This study was unique in that mastication, a product measure of chewing that indicated the efficacy of the child’s chewing behavior, was a dependent variable. During least-to-most prompting plus praise, the feeder prompted the child to “chew eight times” and counted each chew aloud once the bite entered the child’s mouth. If the child did not chew, the feeder progressed to a model and then a physical prompt. The feeder provided praise for chews that occurred independent of physical guidance. Although the treatment resulted in increased chews per bite, it was unclear whether praise, counting aloud, or least-to-most prompting was responsible for the increases in chewing. Therefore, the authors evaluated the effects of descriptive prompts plus praise on chews per bite and percentage of mastication for the second child. The feeder verbally prompted the child to chew a specific number of times (i.e., 10) and provided behavior-specific praise (“Good job chewing 10 times”) if he chewed the required number of times. Descriptive prompts plus praise increased chews per bite and percentage of mastication. To our knowledge, this was the first study to include a measure of mastication and to demonstrate that observable and measurable chewing behavior resulted in masticated food.
In Volkert et al. (2013), the therapist presented the bite to the child and initiated a three-step sequence to prompt the child to chew. The success of this procedure depended, in part, on the child keeping the bite in his or her mouth throughout the prompting sequence. The children in the current study often swallowed presented bites prior to initiating chewing or prior to masticating the bite, which we will refer to as early swallowing. It is difficult to teach a child to chew in the presence of a bite of food if the child swallows the bite as soon as it enters his or her mouth. Thus, it was necessary to include a treatment component to decrease early swallowing to teach these children to chew while a bite was in the mouth. In addition, the children in the current study exhibited near zero or low levels of chewing in baseline, which suggested they did not have or had limited chewing skills in their repertoire. Therefore, in the current study, we evaluated a clinical protocol to increase chews per bite, eliminate early swallowing, and assess mastication with three children with feeding disorders.
Method
Participants, Setting, and Materials
The participants were typically developing and had been admitted to a pediatric feeding disorders day-treatment program and then transitioned into weekly outpatient therapy. Boyd was a 4-year-old boy with a history of failure-to-thrive and diagnoses of diabetes insipidus, reflux, and gastrostomy-tube dependence. He received 18% of his caloric needs by mouth and 82% via gastrostomy tube. May was a 3-year-old girl with a history of failure-to-thrive and diagnoses of reflux and nasogastric-tube dependence. She received 1% of her caloric needs by mouth and 99% via nasogastric tube. Gary was a 2-year-old boy with a history of failure-to- thrive. He consumed approximately 70% of his caloric needs by mouth and did not have a feeding tube. He only consumed a few smooth foods (e.g., apple sauce, yogurt, baby-food fruits mixed with rice cereal) and whole milk from a bottle. On discharge from the day-treatment program, we increased each child’s intake to 100% of their caloric needs by mouth via 16 pureed foods (four fruits, four proteins, four starches, four vegetables) and liquid from a cup (Similac for Boyd, milk with Carnation Instant Breakfast for May, and Pediasure for Gary). Although all participants’ teeth had erupted typically, the participants had little to no experience consuming food (e.g., food cut into pieces) other than purees. The goal of the current treatment was to increase the texture from puree to table food cut into small pieces. A speech language pathologist or physician cleared all participants as safe oral feeders.
We conducted sessions in a 4-m by 4-m room equipped with one-way observation and sound with May and Gary, and observers were seated in the session room. We conducted sessions with Boyd via telemedicine because the family lived out of state. For Boyd, observers were in a private room in the clinic and linked to the family via a secure web-conferencing platform (i.e., Polycom and then Adobe Connect); the family was in a private room at a local hospital, and we zoomed the camera in to see Boyd’s face, mouth, and upper torso clearly on the monitor. For all participants, materials included feeding utensils (e.g., spoons, chew tube), booster seat or high chair, table and chairs, scale, timer, gloves, and laptop computers. The chew tube was a 7.6-cm length of clear, flexible aquarium airline tubing that was 0.6 cm in diameter (see upper left panel of Figure 1).
Empty chew tube with 0.6 cm by 0.6 cm sample green bean (top left), chew tube with embedded 0.6 cm by 0.6 cm green bean (top right), placement of the chew tube on molars with embedded 0.6 cm by 0.6 cm green bean (bottom left), and half chew tube plus 0.6 cm by 0.6 cm by 5.1 cm strip of carrot (bottom right).
Data Collection and Interobserver Agreement (IOA)
Observers collected data on laptop computers. Observers scored a chew each time the child’s teeth and/or jaw completed one up-and-down motion with the teeth parted at least 1.3 cm following a verbal or model prompt while food was visible anywhere in the mouth except the center of the tongue or between the front teeth. We calculated chews per bite by dividing the number of chews by the number of bites that entered the child’s mouth. Observers scored the bite as masticated at the mastication check (described below) if the child had chewed the bite, the bite had broken into small pieces, and the pieces of food in the child’s mouth or chew tube were approximately 0.2 cm by 0.2 cm or smaller. Observers scored the bite as not masticated at the mastication check if any of the pieces of food in the child’s mouth or chew tube were larger than 0.2 cm by 0.2 cm. Observers scored early swallow at the mastication check if no food was visible in the mouth, and the absence of food was not the result of expulsion. We calculated percentage of masticated bites after dividing the number of bites masticated by the number of bites that entered the child’s mouth minus the number of early swallows. Observers scored mouth clean 45 s after the bite entered the child’s mouth during baseline, unless we indicate otherwise, or 30 s after the child masticated the bite during treatment when no food larger than rice size was in the child’s mouth, and the absence of food was not the result of expulsion. When we increased the bite size to larger than 0.6 cm by 0.6 cm, we altered the criterion for mouth clean to no food larger than pea size. We calculated percentage of mouth clean by dividing the number of bites with mouth clean by the number of bites that entered the child’s mouth. Two individuals collected data simultaneously, but independently, during 39%, 45%, and 37% of sessions for Boyd, May, and Gary, respectively.
Observers were individuals with bachelor’s, master’s, or PhD degrees in psychology, behavior analysis, or a related field. We trained data collectors to criterion (agreement of 80% or higher for three consecutive sessions) on the operational definitions using didactic instruction, rehearsal, and feedback. If mean IOA fell below 80% for any variable, the experimenter and observers discussed the operational definitions in detail following sessions until mean IOA again exceeded 80%. We calculated IOA by breaking the session into successive 10-s intervals. An agreement was both observers scoring at least one response or both observers scoring no responses in the interval. We then divided agreements by the sum of agreements and disagreements and converted this ratio to a percentage. Mean IOA was 98% (range: 58%-100%), 98% (range: 80%-100%), and 98% (range: 78%-100%) for chews for Boyd, May, and Gary, respectively. Mean IOA was 97% (range: 69%-100%), 97% (range: 71%-100%), and 98% (range: 77%-100%) for mastication for Boyd, May, and Gary, respectively. Mean IOA was 98% (range: 69%-100%), 93% (range: 73%-100%), and 94% (range: 78%-100%) for mouth clean for Boyd, May, and Gary, respectively.
Experimental Design
We used a multiple-probe design across foods to demonstrate experimental control. We conducted baseline probes periodically to determine whether the next phases of treatment were necessary.
With carrots and chicken for Boyd, we conducted baseline, least-to-most prompting with an empty chew tube (training), baseline, least-to-most prompting with a bite of food in a chew tube, least-to-most prompting with a strip of food on a half chew tube, baseline with a 15-s mouth check, and baseline with a 15-s mouth check and an increased bite size. With potatoes, we conducted baseline, least-to-most prompting with an empty chew tube (training), baseline, least-to-most prompting with a bite of food in a chew tube, least-to-most prompting with a strip of food on a half chew tube, least-to-most prompting with a bite of food, baseline with a 15-s mouth check, and baseline with a 15-s mouth check and an increased bite size. We conducted baseline probes at Session 37 for carrots, Sessions 44 and 63 for potatoes, and Session 61 for chicken. We conducted a least-to-most prompting with a bite-of-food probe at Session 65 for chicken.
With all foods for May, we conducted least-to-most prompting with an empty chew tube (training); baseline; least-to-most prompting with a bite of food in a chew tube; least-to-most prompting with a strip of food; least-to-most prompting, interruption, and differential reinforcement of alternative (DRA) behavior; and baseline with a 15-s mouth check. We implemented baseline probes at Sessions 54, 59, and 103 for chicken; Sessions 52 and 92 for green beans; Sessions 48, 49, and 108 for peaches; and Sessions 61, 65, 91, and 97 for potatoes.
With potatoes for Gary, we implemented baseline, least-to-most prompting with an empty chew tube (training), least-to-most prompting with a bite of food in a chew tube, least-to-most prompting with a strip of food, least-to-most prompting with a strip of food on a half chew tube, least-to-most prompting with a strip of food, least-to-most prompting with a bite of food, and least-to-most prompting with an increased bite size. With chicken, pears, and green beans, we implemented baseline, least-to-most prompting with an empty chew tube (training), baseline, least-to-most prompting with a bite of food in a chew tube, least-to-most prompting with a strip of food on a half chew tube, least-to-most prompting with a strip of food, least-to-most prompting with a bite of food, and least-to-most prompting with a bite of food and an increased bite size (green beans only). We conducted baseline probes at Sessions 32 and 60 for potatoes, Sessions 99 and 106 for chicken, Session 72 for pears, and Session 64 for green beans. We conducted least-to-most prompting with bite-of-food probes at Session 41 for potatoes, Session 40 for pears, and Sessions 42 and 45 for green beans.
General Procedure
The child’s mother served as feeder. We cut the food into small pieces, carefully monitored the child, and conducted brief baselines to minimize the risk of choking or aspiration if the child swallowed the bite without chewing. We did not observe choking or other behavior associated with aspiration (e.g., wet vocal sounds) during the analysis. The target foods were canned potato, fresh cooked carrots (Boyd) or canned green beans (May and Gary), canned peach (May) or pear (Gary), and grilled chicken. We sent Boyd’s mother a kit that included a ruler, chew tubes, and foam 3-D examples of the bite sizes to aide her with food preparation. We prepared the food at the clinic for May and Gary.
The child and mother attended weekly 1-hr outpatient appointments in which we conducted multiple sessions with brief breaks between sessions. Sessions consisted of five bite presentations of a single food (e.g., chicken), or we ended the session after 10 min if the mother was not able to present all five bites. The mother conducted one to seven sessions during each 1-hr appointment. The number of sessions varied because we worked on other feeding skills (e.g., self-feeding) during the appointment. The size of the bite was 0.6 cm by 0.6 cm, and the mother presented the bite on a small Maroon spoon, unless we specify otherwise. The mother presented one bite to the child’s lips approximately every 45 s to 60 s with the verbal prompt, “take a bite.” The mother provided praise if the child accepted the bite within 5 s. If the child did not accept the bite within 5 s, the mother held the utensil (spoon, chew tube, strip of food) at the child’s lips until the child opened his or her mouth such that the mother could deposit the bite into the child’s mouth (nonremoval of the spoon, chew tube, or strip of food; Hoch, Babbitt, Coe, Duncan, & Trusty, 1995).
The mother conducted a mastication mouth-check or mouth-clean check by prompting the child to “Show me, ahh” while modeling an open mouth. If the child did not open, the mother placed the spoon to the lips and turned the spoon 90° to prompt the child to open his or her mouth while saying, “Show me.” Observers scored whether the child masticated, did not masticate, or early swallowed the bite at the initial mastication check. The mothers provided praise for mouth clean and presented the next bite. If Boyd and Gary packed (converse of mouth clean) the bite, the mother said “Swallow your bite” and immediately presented the next bite. The procedure for packed bites differed for May across baseline and treatment and is described further below. If the child was packing after the mother presented the fifth bite, the mother conducted a mouth check every 45 s and prompted the child to “Swallow your bite” until no food larger than a grain of rice was in the mouth or 10 min elapsed from the start of the session. If food was in the child’s mouth at the 10-min mark, the mother removed the food from the mouth with the spoon. The mother provided no differential consequence for inappropriate behavior. If the child initiated conversation, the mother conversed with him or her. Prior to the first session of an appointment, the mother explained the session contingencies. The mother did not reiterate the contingencies unless they changed during the appointment. A speech therapist approved the procedure.
Baseline and baseline probes
The purpose was to evaluate the child’s chews per bite and ability to masticate a bite of food. Once the bite entered the child’s mouth, the mother prompted the child to “chew and swallow” and provided praise if the child chewed the bite 10 (Boyd and Gary) or 12 (May) times prior to the mouth-clean check. The mother counted the child’s chews silently. The mother conducted a mastication mouth-check 45 s, 30 s, or 15 s after the bite entered the mouth of Boyd, May, or Gary, respectively. Boyd’s mother conducted the mastication mouth-check at the same time as the mouth-clean check; thus, observers also scored mouth clean or pack for Boyd at this check. The mothers of May and Gary conducted a mouth-clean check 15 s or 30 s, respectively, after the initial mastication mouth-check, and observers scored mouth clean or pack for May and Gary at this check. Packing was zero for carrots and potatoes and a mean of 13% for chicken for Boyd and zero for Gary across all foods during baseline. Therefore, Boyd very rarely and Gary never had more than one bite of food in his mouth at a time, and the mother followed the general procedure. May packed more frequently during baseline. Although mean packing was only 8% for potatoes and peaches, packing was 49% for chicken and green beans. Therefore, if she packed the bite, May’s mother prompted her to “Swallow your bite” every 45 s and waited to present the next bite. When May had no food larger than a grain of rice in her mouth, May’s mother then presented the next bite.
The mother provided no differential consequence if the child expelled the bite, and observers scored whether the expelled bite was masticated if this occurred prior to the initial mastication mouth-check.
General treatment procedure
The mother followed the general procedure unless otherwise stated. During conditions with the tube or strip of food, the mother alternated between presenting the tube or strip to the right and left sides of the child’s mouth. Once the mother placed the tube or strip between the child’s molars or the 0.6-cm by 0.6-cm bite entered the child’s mouth, the mother prompted the child to “chew (target number) times” and counted each chew aloud. The target number of chews is described below for least-to-most prompting with an empty chew tube (training) and was 10 (Boyd and Gary) or 12 (May) for the remaining treatment conditions unless otherwise stated. The mother used least-to-most prompting (verbal, gestural, physical) to prompt the child to chew. If the child did not chew within 5 s of the verbal prompt or stopped chewing for 5 s before completing the target number of chews, the mother modeled chewing and prompted the child to “Chew like this.” After the model prompt, if the child did not chew within 5 s or stopped chewing for 5 s before completing the target number of chews, the mother physically guided the target number of chews or the remaining chews. Physical guidance consisted of the mother placing one hand on top of the child’s head just above the forehead and placing the thumb of her other hand horizontally under the child’s chin while gently providing upward and downward pressure to move the child’s jaw. The mother kept the tube or strip of food in the child’s mouth by holding it between the index and middle fingers of the hand positioned under the chin. The mother provided verbal praise if the child completed the target number of chews without physical guidance.
After the child completed the target number of chews, the mother removed the tube or strip of food from the child’s mouth and conducted a mastication check. If the bite was not masticated, the mother re-presented the tube or strip of food to the child and instructed the child to chew four more times using least-to-most prompting followed by a mastication check. The mother repeated the four-chew prompting and mastication-check sequence until the child masticated the bite. Once the child masticated the bite, the mother removed the masticated bite from the tube, strip, or mouth and placed the bite onto the spoon unless we specify otherwise.
The mother presented the spoon with the masticated bite to the child using the nonremoval procedure. The mother said, “you can swallow your bite” when the bite entered the child’s mouth. The mother conducted a mouth-clean check 30 s after the masticated bite entered the child’s mouth, the observers scored mouth clean or pack, and the mother presented the next bite. For May, this was a change in procedure if she was packing the previous bite, as the mother did not present the next bite during baseline. However, May never packed potatoes, green beans, or peaches during this condition, so the change for these foods did not affect the data. May did pack chicken during this condition; thus, the data for chicken are limited to some extent as a result of this change.
If the child expelled the masticated bite, the mother re-presented it. If the masticated bite contacted an unclean surface following expulsion, the mother mashed an equivalent amount of food with her finger or spoon to resemble a masticated bite before re-presenting the bite.
Least-to-most prompting with an empty chew tube (training)
The purpose was to teach the child one component of chewing behavior, to raise and lower the jaw such that the child’s molars met and exerted pressure on the tube. A second purpose was to teach the child to comply with the prompt to “chew x times.” The sessions consisted of five presentations of the tube.
Initially, we presented the tube to Boyd without prompting any chewing and to May with a prompt to chew one time to initiate training with a low response-effort requirement. For Gary, we conducted one session with a prompt to chew 10 times to set the initial chew requirement. He chewed a mean of six times; thus, we initiated chew training with a prompt to “chew six times.” After three sessions, Gary’s mean for chews per bite was four; thus, we decreased the chew requirement to four.
We advanced the chew requirement every three consecutive sessions if mean acceptance (i.e., the chew tube entered the child’s mouth between the molars within 5 s of presentation) was at least 80%, mean number of inappropriate mealtime behavior (e.g., head turns, bats, blocks) was five or below, and mean chews per bite was at least 80% of the chew requirement. For example, we advanced the chew requirement from two to three if mean chews per bite was 1.6 (0.8 × 2) or more. We did not have to decrease the chew requirement for any participant except as described above for Gary initially. We ended training after Boyd and Gary were chewing 10 times per bite, and May was chewing 12 times per bite. These data are not depicted on the figure but are available on request.
Least-to-most prompting with a bite of food in a chew tube
The purpose was to teach the child to masticate and then swallow a bite of food, that is, to link the behaviors of chewing and swallowing. To reduce the probability of the child making the error of swallowing the bite before he or she masticated it, we placed the bite in the tube.
The mother used a tube that we had slit lengthwise (see upper right panel of Figure 1). The mother placed the bite in the tube by pulling the tube apart along the slit, positioning the bite at the end of the tube, and releasing the tube to re-form around the bite and used the general treatment procedure with the following differences.
If the bite of food came out of the tube before the child completed the target number of chews, the mother retrieved the bite from the child’s mouth if possible, placed the bite back in the tube, and continued the prompting procedure described previously. The bite presentation ended, however, if the child swallowed the bite before the mother could retrieve it from the child’s mouth. In this case, the mother presented a new bite in the tube approximately 45 s from when the bite entered the child’s mouth.
Least-to-most prompting with a strip of food on a half chew tube
We conducted this condition with Boyd and Gary. Children with chewing deficits often have difficulty moving food from the tongue to the molars and vice versa and keeping the food on the molars while chewing. Because the mother placed the strip of food on the half tube on the child’s molars and held it there, the child had the opportunity to chew directly on the food without having to control the position of the food in the mouth. The half tube reduced the probability of food migrating into the child’s mouth because when pieces broke off the strip, they tended to remain on top of the tube. In addition, the half tube provided a stable base for the strip, which allowed the mother to position and control the strip on the half tube more easily relative to the strip alone.
The mother presented a 0.6-cm by 0.6-cm by 5.1-cm strip of food on a tube cut in half lengthwise (see bottom right panel of Figure 1) and used the general treatment procedure with the following exceptions. During the mastication check, if a 0.6-cm by 0.6-cm piece of masticated food had broken off the strip and was in the child’s mouth, the mother removed the strip and half chew tube from the child’s mouth and prompted the child to swallow the food in his or her mouth. If a piece of masticated food larger than 0.6 cm by 0.6 cm had broken off the strip and was in the child’s mouth, the mother removed the strip, half chew tube, and pieces of food from the child’s mouth; placed a 0.6-cm by 0.6-cm piece of the masticated food on the spoon; presented it to the child; and prompted the child to swallow. If the total amount of masticated food attached to the strip plus masticated food that had broken off the strip and was in the child’s mouth was equal to 0.6 cm by 0.6 cm, the mother removed the strip, half chew tube, and pieces of food from the child’s mouth. She then broke off the masticated piece attached to the strip, placed the masticated piece from the strip plus the masticated food from the mouth onto the spoon, presented it to the child, and prompted the child to swallow. If the 0.6-cm by 0.6-cm piece of masticated food was attached to the strip, the mother removed the strip and half chew tube from the child’s mouth, broke off the masticated piece attached to the strip, presented it to the child, and prompted the child to swallow. If the mother observed that a piece of unmasticated food had broken off the strip and was in the child’s mouth during the chew-prompting sequence, the mother removed the strip, half chew tube, and piece of food from the child’s mouth; placed the piece onto the half chew tube with the strip; put them back in the child’s mouth; and re-initiated the chew-prompting sequence where she had left off.
Least-to-most prompting with a strip of food
We conducted this condition with May and Gary. The mother placed the strip of food on the child’s molars and held it there; however, without the stability of the half tube, the strip tended to move on the child’s molars. The mother could assist the child by re-positioning the strip, but the child also had to use his or her tongue to maintain the strip on the molars, particularly as the strip began to break into pieces. We did not implement this condition with Boyd because he began chewing during the baseline following least-to-most prompting with a strip of food on a half chew tube.
The mother presented a 0.6-cm by 0.6-cm by 5.1-cm strip of food (no chew tube) using the general treatment procedure with the following differences. During the mastication check, if only a 0.6-cm by 0.6-cm piece of masticated food had broken off from the strip and was in the child’s mouth, the mother removed the strip from the child’s mouth and prompted the child to swallow the food in his or her mouth. If the total amount of masticated food in the child’s mouth and/or on the strip was equal to 0.6 cm by 0.6 cm as described above, the mother removed the food from the child’s mouth, placed the masticated food on the spoon, presented it to the child, and prompted the child to swallow.
If food had broken off the strip and was less than 0.6 cm by 0.6 cm, whether masticated or not, and the child had not masticated any part of the strip, the mother removed the food from the child’s mouth, placed the strip in the child’s mouth, and re-initiated the chew-prompting sequence as described above until the child masticated a piece of food equivalent to 0.6 cm by 0.6 cm.
Least-to-most prompting with a bite of food
We conducted this condition with Boyd for potatoes and with Gary for all foods. The purpose was to determine whether the child could chew and masticate a bite of food when the mother used least-to-most prompting. The mother presented one 0.6-cm by 0.6-cm bite on a spoon to the child and then used the general treatment procedure with the following exceptions. For Gary, the mother also prompted him to “move it over to your molars” when the bite entered his mouth. Once Gary moved it to his molars, she prompted him to “chew 10 times.” If he moved the bite over to his molars within 5 s of the bite entering his mouth, the mother provided praise. The mother then resumed the general treatment procedure. The bite remained in the child’s mouth during the mastication mouth check. If the child swallowed the bite prior to the mastication mouth-check, the mother presented a new bite approximately 45 s from when the bite entered the child’s mouth. We also conducted probes of this condition with Boyd (with the exception of the move-it-over prompt) and Gary.
Least-to-most prompting with an increased bite size
This condition was identical to least-to-most prompting with a bite of food except the bite size was 0.6 cm by 0.6 cm by 1.3 cm. We conducted this procedure with Gary for potatoes and green beans. We did not increase the bite size of chicken and pears because the caregiver wanted to work on self-feeding instead.
Least-to-most prompting with a bite of food, interruption, and differential reinforcement of alternative behavior
For May, she began to pull her jaw forward so that her lower teeth protruded past her upper teeth (similar to an under bite) while chewing, which we will refer to as an incorrect chew. We also observed anecdotally that May masticated bites in about 8 rather than 12 chews. We implemented the procedure described for least-to-most prompting with a bite of food, with the following modifications. If May exhibited an incorrect chew, the mother inserted a rubber-coated baby spoon between May’s lips; said “no, you need to chew right like this”; and modeled a chew. If May exhibited an incorrect chew again, her mother inserted a rubber-coated baby spoon between May’s lips; placed the empty tube in her mouth; said “no, you need to chew right like this”; and then guided May to chew eight times or the remaining number of chews. If May chewed eight times without using an incorrect chew, independent of physical guidance, her mother provided her with a preferred toy for 30 s.
Baseline with a 15-s mouth check
We conducted this procedure with Boyd and May to evaluate whether the child could chew and masticate a bite of food in the absence of least-to-most prompting. The procedure was identical to baseline except that the mother conducted the mastication check approximately 15 s after the bite entered the child’s mouth. We decreased the time of the check in an effort to better observe whether the child masticated the bite prior to swallowing.
Baseline with a 15-s mouth check and an increased bite size
We conducted this condition with all foods for Boyd. The procedure was identical to baseline with a 15-s mouth check, except for the bite size was 0.6 cm by 0.6 cm by 1.3 cm.
Results
For Boyd, mean chews per bite (top of Figure 2) for carrots (top panel) was 0.07 (range: 0-0.2) during baseline before and after least-to-most prompting with an empty chew tube, 10 (range: 7.4-13) during least-to-most prompting with a bite of food in a chew tube, 0.6 during the baseline probe at Session 37, 10.9 (range: 10-11.8) during least-to-most prompting with a strip of food on a half chew tube, 11.6 (range: 5.4-14) during baseline with a 15-s mouth check, and 11.2 (range: 8.4-13.8) during baseline with a 15-s mouth check and an increased bite size. Mean chews per bite for potatoes (middle panel) was 0.6 (range: 0.4-1.4) during baseline before and after least-to-most prompting with an empty chew tube, 11.1 (range: 9.4-17.6) during least-to-most prompting with a bite of food in a chew tube, 2.6 during the baseline probe at Session 44, 9.3 (range: 6.2-10.8) during least-to-most prompting with a strip of food on a half chew tube, 7.2 during the baseline probe at Session 63, 9.9 (range: 8.8-10.4) during least-to-most prompting with a bite of food, 9.9 (range: 7-12.8) during baseline with a 15-s mouth check, and 7.3 (range: 4.2-9.8) during baseline with a 15-s mouth check and an increased bite size. For chicken (bottom panel), mean chews per bite was 1.8 (range: 0.6-3) before and after least-to-most prompting with an empty chew tube, 16.1 (range: 10-24.5) during least-to-most prompting with a bite of food in a chew tube, 4.8 during the baseline probe at Session 61, 22.8 during least-to-most prompting with a bite-of-food probe at Session 65, 11.6 (range: 10-14.8) during least-to-most prompting with a strip of food on a half chew tube, 11.5 (range: 6.4-19.6) during baseline with a 15-s mouth check, and 16.9 (range: 12.8-21.8) during baseline with a 15-s mouth check and an increased bite size.
Chews per bite (top panel) and percentage mastication (bottom panel) for Boyd.
Observers did not have an opportunity to score mastication (bottom of Figure 2) for carrots (top panel) during baseline sessions and during the baseline probe at Session 37 because Boyd swallowed early on all bite presentations. Mean percentage mastication for carrots was 81% (range: 25%-100%) during least-to-most prompting with a bite of food in a chew tube, 77% (range: 20%-100%) during least-to-most prompting with a strip of food on a half chew tube, 95% (range: 80%-100%) during baseline with a 15-s mouth check, and 97% (range: 80%-100%) during baseline with a 15-s mouth check and an increased bite size.
Observers did not have an opportunity to score mastication for potatoes (middle panel) during baseline sessions and during the baseline probe at Session 44 due to early swallowing. Mean percentage mastication for potatoes was 82% (range: 0%-100%) during least-to-most prompting with a bite of food in a chew tube, 80% (range: 0%-100%) during least-to-most prompting with a strip of food on a half chew tube, 100% during the baseline probe at Session 63, 93% (range: 80%-100%) during least-to-most prompting with a bite of food, 95% (range: 80%-100%) during baseline with a 15-s mouth check, and 97% (range: 80%-100%) during baseline with a 15-s mouth check and an increased bite size. Mean percentage mastication for chicken (bottom panel) was zero during baseline, 49% (range: 0%-100%) during least-to-most prompting with a bite of food in a chew tube, 100% during the baseline probe at Session 61, zero during the least-to-most prompting with a bite-of-food probe at Session 65, 61% (range: 0%-100%) during least-to-most prompting with a strip of food on a half chew tube, 88% (range: 0%-100%) during baseline with a 15-s mouth check, and 62% (range: 20%-100%) during baseline with a 15-s mouth check and an increased bite size. Mean levels of mouth clean for all foods was 92% (range: 60%-100%) during baseline and 94% (range: 20%-100%) during subsequent phases and baseline probes (data not shown).
For May, mean chews per bite (top of Figure 3) with chicken (top panel) was 0.2 (range: 0-0.6) during baseline after least-to-most prompting with an empty chew tube; 3.6 (range: 2.6-4.6) during baseline probes at Sessions 54 and 59; 11.9 (range: 0-23) during least-to-most prompting with a bite of food in a chew tube; 9.4 (range: 2.2-14) during least-to-most prompting with a strip of food; 0.8 during the baseline probe at Session 103; 8.5 (range: 7-12.6) during least-to-most prompting, interruption, and DRA with a bite of food; and 16 (range: 9.4-22.4) during baseline with a 15-s mouth check. With green beans (second panel), mean chews per bite was 1.7 (range: 0-6.8) during baseline after least-to-most prompting with an empty chew tube; 12 (range: 11.8-12) during least-to-most prompting with a bite of food in a chew tube; 0.2 during the baseline probe at Session 52; 11.8 (range: 9.8-13.8) during least-to-most prompting with a strip of food; 0 during the baseline probe at Session 92; 7.6 (range: 5-8.4) during least-to-most prompting, interruption, and DRA with a bite of food; and 12.4 (range: 3.8-19.6) during baseline with a 15-s mouth check. For peaches (third panel), mean chews per bite was 2.3 (range: 0-10) during baseline; 0.5 (range: 0.4-0.6) during baseline probes at Sessions 48 and 49; 12 (range: 10.4-13.2) during least-to-most prompting with a bite of food in a chew tube; 12.9 (range: 10.2-14.6) during least-to-most prompting with a strip of food; 0 during the baseline probe at Session 108; 8.1 (range: 6.4-11) during least-to-most prompting, interruption, and DRA with a bite of food; and 9.5 (range: 6-14.2) during baseline with a 15-s mouth check. For potatoes (bottom panel), mean chews per bite was 1.5 (range: 0-6.4) during baseline after least-to-most prompting with an empty chew tube; 11.6 (range: 9.6-12) during least-to-most prompting with a bite of food in a chew tube; 0 during baseline probes at Sessions 61 and 65; 12.1 (range: 12-12.6) during least-to-most prompting with a strip of food; 4.4 (range: 3.8-5) during baseline probes at Sessions 91 and 97; 8 (range: 7.6-8.4) during least-to-most prompting, interruption, and DRA with a bite of food; and 12.1 (range: 2.8-21.6) during baseline with a 15-s mouth check.
Chews per bite (top panel) and percentage mastication (bottom panel) for May.
Observers did not have an opportunity to score mastication (bottom of Figure 3) for chicken (top panel) during baseline probes at Sessions 59 and 103 because May swallowed early on all bite presentations. Mean percentage mastication was 13% (range: 0%-40%) during baseline; 100% during the baseline probe at Session 54 based on one opportunity to score mastication; 76% (range: 0%-100%) during least-to-most prompting with a bite of food in a chew tube; 53% (range: 0%-100%) during least-to-most prompting with a strip of food; 88% (range: 20%-100%) during least-to-most prompting, interruption, and DRA with a bite of food; and 96% (range: 80%-100%) during baseline with a 15-s mouth check. Observers did not have an opportunity to score mastication for green beans (second panel) during baseline probes at Sessions 52 and 92 due to early swallowing. Mean percentage mastication was zero during baseline; 93% (range: 60%-100%) during least-to-most prompting with a bite of food in a chew tube; 86% (range: 50%-100%) during least-to-most prompting with a strip of food; 98% (range: 80%-100%) during least-to-most prompting, interruption, and DRA with a bite of food; and 90% (range: 33%-100%) during baseline with a 15-s mouth check. With peaches (third panel), observers did not have the opportunity to score mastication during baseline and baseline probes at Sessions 48, 49, and 109 due to early swallowing. Mean percentage mastication was 89% (range: 80%-100%) during least-to-most prompting with a bite of food in a chew tube; 67% (range: 0%-100%) during least-to-most prompting with a strip of food; 86% (range: 40%-100%) during least-to-most prompting, interruption, and DRA with a bite of food; and 100% during baseline with a 15-s mouth check. Observers did not have the opportunity to score mastication for potatoes (bottom panel) during six out of seven baseline sessions and during baseline probes at Sessions 61, 65, and 97 due to early swallowing. Mean percentage mastication was 100% during the one baseline session in which observers had the opportunity to score it; 100% during least-to-most prompting with a bite of food in a chew tube; 89% (range: 60%-100%) during least-to-most prompting with a strip of food; zero during the baseline probe at Session 91; 100% during least-to-most prompting, interruption, and DRA with a bite of food; and 98% (range: 80%-100%) during baseline with a 15-s mouth check. Mean levels of mouth clean for all foods was 76% (range: 0%-100%) during baseline and 95% (range: 40%-100%) during subsequent phases and baseline probes (data not shown).
For Gary, mean chews per bite (top of Figure 4) for potatoes (top panel) was 0.1 (range: 0.8-1) during baseline before least-to-most prompting with an empty chew tube, 9.6 (range: 7.6-10.8) during least-to-most prompting with a bite of food in a chew tube, 0.2 during the baseline probe at Session 32, 11.2 (range: 10-13.2) during least-to-most prompting with a strip of food, 8.0 (range: 6.4-9.8) during least-to-most prompting with a strip of food in a half chew tube, zero during the baseline probe at Session 60, 8.5 (range: 3.4-10) during the return to least-to-most prompting with a strip of food, 9.2 (range: 5-10) during least-to-most prompting with a bite of food, and 10 (SD = 0) during least-to-most prompting with an increased bite size. Mean chews per bite for chicken (second panel) was 0.3 (range: 0-0.8) during baseline before and after least-to-most prompting with an empty chew tube, 9.4 during least-to-most prompting with a bite-of-food probe at Session 41, 10.2 (range: 8.4-11.6) during least-to-most prompting with a bite of food in a chew tube, 11.2 (range: 9.2-14.8) during least-to-most prompting with a strip of food on a half chew tube, 10.2 (range: 10-10.4) during least-to-most prompting with a strip of food, 4.7 (range: 3-6.4) during baseline probes at Sessions 99 and 106, and10.1 (range: 10-10.8) during least-to-most prompting with a bite of food. Mean chews per bite for pears (third panel) was 0.7 (range: 0-1.6) during baseline before and after least-to-most prompting with an empty chew tube, 5 during least-to-most prompting with a bite-of-food probe at Session 40, 10 (range: 9.8-10) during least-to-most prompting with a bite of food in a chew tube, 10 (SD = 0) during least-to-most prompting with a strip of food on a half chew tube, 0.4 during the baseline probe at Session 72, 10.4 (range: 9.8-11.6) during least-to-most prompting with a strip of food, and 10 (range: 8-11.2) during least-to-most prompting with a bite of food. Mean chews per bite for green beans (fourth panel) was 0.8 (range: 0-2.2) during baseline before and after least-to-most prompting with an empty chew tube, 4.1 (range: 0.2-8) during least-to-most prompting with a bite-of-food probes at Sessions 42 and 45, 9.3 (range: 7.6-10) during least-to-most prompting with a bite of food in a chew tube, 0 during the baseline probe at Session 64, 10 (SD = 0) during least-to-most prompting with a strip of food on a half chew tube, 10 (SD = 0) during least-to-most prompting with a strip of food, 10 (SD = 0) during least-to-most prompting with a bite of food, and 10 (SD = 0) during least-to-most prompting with an increased bite size.
Chews per bite (top panel) and percentage mastication (bottom panel) for Gary.
Observers did not have an opportunity to score mastication (bottom of Figure 4) for potatoes (top panel) during the baseline probe at Session 32 because Gary swallowed early on all bite presentations. Mean percentage mastication was 10% (range: 0%-20%) during baseline, 96% (range: 80%-100%) during least-to-most prompting with a bite of food in a chew tube, 39% (range: 0%-67%) during least-to-most prompting with a strip of food, 100% during least-to-most prompting with a strip of food on a half chew tube, 0% during the baseline probe at Session 60, 100% during the return to least-to-most prompting with a strip of food, 94% (range: 67%-100%) during least-to-most prompting with a bite of food, and 95% (range: 80%-100%) during least-to-most prompting with an increased bite size. Observers did not have an opportunity to score mastication for chicken (second panel) during the baseline probe at Session 106 due to early swallowing. Mean percentage mastication was zero during baseline before and after least-to-most prompting with an empty chew tube, 67% during least-to-most prompting with a bite-of-food probe at Session 41, 88% (range: 60%-100%) during least-to-most prompting with a bite of food in a chew tube, 77% (range: 20%-100%) during least-to-most prompting with a strip of food in a half chew tube, 100% during least-to-most prompting with a strip of food, 0% during the baseline probe at Session 99, and 99% (range: 80%-100%) during least-to-most prompting with a bite of food. Mean percentage mastication for pears (third panel) was zero during baseline before and after least-to-most prompting with an empty chew tube, 50% during the least-to-most prompting with a bite-of-food probe at Session 40, 97% (range: 80%-100%) during least-to-most prompting with a bite of food in a chew tube, 100% during least-to-most prompting with a strip of food on a half chew tube, 33% during the baseline probe at Session 72, 83% (range: 60%-100%) during least-to-most prompting with a strip of food, and 86% (range: 0%-100%) during least-to-most prompting with a bite of food. Observers did not have an opportunity to score mastication for green beans (fourth panel) during least-to-most prompting with bite-of-food probes at Sessions 42 and 45 and during the baseline probe at Session 64 due to early swallowing. Mean percentage mastication was 0% during baseline before and after least-to-most prompting with an empty chew tube, 89% (range: 60%-100%) during least-to-most prompting with a bite of food in a chew tube, 100% during least-to-most prompting with a strip of food on a half chew tube, 96% (range: 80%-100%) during least-to-most prompting with a strip of food, 100% during least-to-most prompting with a bite of food, and 100% during least-to-most prompting with an increased bite size. Mean levels of mouth clean for all foods was 51% (range: 0%-100%) during baseline and 97% (range: 60%-100%) during subsequent phases and baseline probes (data not shown).
Discussion
The goal of the current investigation was to describe a clinical protocol for increasing chewing and assessing mastication of three children with feeding disorders. We established baseline levels of chewing and mastication with three or four foods per child. The baseline data showed that chews per bite were low across all foods for all participants. Even when chewing occurred, either the child did not masticate the bite, or the child swallowed the bite early. Therefore, the data suggested that the goals of treatment should be to increase chews per bite, eliminate early swallowing, and increase levels of mastication.
We addressed the challenge of teaching the multiple-component behavior of chewing in several ways. Recall that prior to admission to our day-treatment program, these children refused most food. During day treatment, we established acceptance at 90% or greater with a variety of pureed table foods. Therefore, when we initiated chew training, the children were highly compliant with the prompt to “take a bite” and were familiar with the presented foods, albeit at a pureed texture. High levels of acceptance maintained when we introduced the tube and the foods at a higher texture during chew training. Using the tube in chew training allowed us to temporarily eliminate the interdependence of the responses of lateralizing the bite to the molars, grinding the bite until it was masticated, lateralizing the bite from the molars to the tongue, and elevating the tongue and propelling the bite into the pharynx to swallow.
With Boyd and May, we were able to return to baseline with a 15-s mouth check in the final phases of the analysis. Boyd and May exhibited stable levels of chewing and mastication, which suggested that they were able to chew and masticate the bite in the absence of the least-to-most-prompting procedure. In fact, we were able to increase the bite size for all foods for Boyd. With Gary, we implemented least-to-most prompting with the bite of food in the final phases of the analysis, and he exhibited stable levels of chewing and mastication. In addition, we were able to increase the bite size for potatoes and green beans for Gary.
This study is important from a technological standpoint for several reasons. First, we attempted to provide the reader with a precise, detailed, and replicable description of the chew-training protocol, which has not been available in the literature up to this point. Second, we continued to refine our operational definitions for chewing and mastication, which is also a novel contribution. A limitation of our operational definition of chewing was that it did not include measurement of lateralization or rotary chewing, two behaviors that are important for development of a mature chew. Because we did not measure lateralization, we do not know whether our protocol affected lateralization.
Another limitation was that we conducted a mastication check for Boyd and May, at 45 s and 30 s, respectively, after the bite entered the child’s mouth during baseline. In the final phases of the analysis, we repeated the baseline condition, but we conducted the mastication check 15 s after the bite entered the child’s mouth. In baseline, both children swallowed early, before observers could score mastication. It is possible that observers could have scored mastication if we had conducted a mastication check at 15 s during the initial baseline. Note, however, that we conducted a mastication check for Gary at 15 s in baseline, and we were still not able to determine whether Gary had masticated the bite due to his early swallowing. In the return to baseline with the 15-s mouth check for Boyd and May, early swallowing occurred at low levels, and observers were able to score mastication.
We do not know whether least-to-most prompting with an empty chew tube was necessary or whether we could have initiated chew training with food in the tube. This question could be explored in future research. With Boyd and May, we used a low response requirement (e.g., prompting the child to chew one time on the tube) as the starting point for empty chew-tube training. With Gary, we used a data-based method of identifying the starting point. We do not know whether one method of initiating training is better than another, and this should be a topic for future investigations. In addition, because we used a multi-component intervention, we do not know which components were responsible for the increases in chews per bite or mastication. Another limitation of the study is that we did not re-present expelled bites in baseline, but we did in subsequent phases. Boyd and May exhibited little to no expulsion throughout the analysis. By contrast, Gary expelled bites periodically during baseline. It is possible that chews per bite would have increased if we had simply re-presented bites to Gary in baseline. However, Gary also exhibited high levels of early swallowing during baseline, and we did not want to encourage swallowing of unmasticated bites of food by placing the expelled bite back into Gary’s mouth repeatedly.
We discharged Boyd and May from our feeding program after 30 and 27 months of treatment, respectively (8 weeks of day treatment and 118 and 106 weeks, respectively, of weekly outpatient follow up), and their feeding tubes had been removed. Gary is nearing discharge. For all children, we systematically increased bite size, introduced additional foods, and advanced to self-feeding age-appropriate portions.
The current study adds to a small body of literature on the chewing and mastication of children with feeding disorders. These data are important because we have described a clinical protocol for teaching children to chew and masticate bites of food. The study is also important because we have described a method for measuring progress relative to chewing and mastication. Future research should focus on improving the demonstration of functional control for the protocol, providing further details on when to use specific components of the protocol and refining the operational definitions.
Footnotes
Acknowledgements
Thanks to Lynn Farrell, Jana Frese, Haley Hall, and Laura Hartman for their assistance with this study.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.