Davis, Ann M.*; Bruce, Amanda S.†; Khasawneh, Rima*; Schulz, Trina‡; Fox, Catherine§; Dunn, Winifred‡
Feeding is one of the most important early skills that a child must learn, and the development of feeding skills is extremely complex and multifactorial (1). Many parents express frustration over young children being “picky eaters,” and feeding issues frequently become a struggle for families (2). “Picky eating” and other mild forms of feeding problems do not typically progress into serious feeding difficulties. Mild feeding difficulties occur in up to 25% to 45% of healthy children and 80% of children with developmental delays or chronic disease (3,4). Serious feeding difficulties requiring medical intervention occur in 3% to 10% of children (4). Premature infants are overrepresented among children with serious feeding difficulties (5), and improved preterm infant survival has increased the prevalence of feeding problems in older infants and toddlers (6). In addition, severe feeding problems may be more common in children with chronic medical conditions, occurring in 40% to 70% of this population (7,8).
Feeding difficulties are caused by structural abnormalities, neurological problems, cardiac conditions, Behavioral problems, and respiratory conditions, along with gastrointestinal motility and sensory abnormalities (9,10). The most common medical diagnostic categories of children presenting to a feeding team include neurological (62%), mechanical/structural (53%), behavioral (43%), cardiorespiratory (34%), and metabolic (12%), with most children falling into ≥2 diagnostic categories simultaneously (11). Children referred to a feeding team are often medically complicated (11). Neonates with long intensive care hospitalizations may lose opportunities for learning to eat and may associate eating with pain or discomfort (10). Premature infants are also more likely to present to a feeding team than children born at term (11). Gastrostomy and gastrojejunal tube feeding requirements may persist for months or years, resulting in chronic oral food refusal. As the prevalence of gastrostomy tube feeding has increased, so have the challenges associated with transitioning a child from tube to oral feeding (12,13).
Along with medical issues, sensory issues also affect feeding. Senses of smell, taste, touch, proprioception, vision, hearing, and sound filtering all contribute to the information the brain has to process and analyze during the feeding experience (14). Interactions between and within sensory modalities are relevant to food perception (15). During eating, some sensations are regarded as important and relevant, whereas others must be ignored. Specifically, researchers have found that smell (odor) can increase the intensity of taste (16). Touch represented as viscosity also decreases both sensitivity to (17–19) and intensity of (20–25) the 4 basic taste groups. Moreover, temperature of food between 22°C and 37°C sets taste thresholds at minimum for salt (26,27) as well as acid (HCl), bitter (quinine sulfate), and sweeteners (eg, dulcin) (27). Visual sense, with color as its main parameter, increases taste threshold and decreases taste discrimination (28–30). Red color enhances the sweet taste intensity in aqueous solutions (31).
The perceived quality and intensity of such sensations affect the brain's ability to set priorities for present focus versus things to ignore. Too much stimulation may draw the infant's attention and response toward a particular stimulus source rather than focusing on the other processes associated with eating, such as chewing or swallowing. Thus, problems affecting children's abilities to handle and process such multiple sensory stimuli render them susceptible to feeding problems or could complicate some preexisting feeding issues.
Surprisingly, there have been extremely few studies regarding the relation between sensory issues and feeding within clinical populations. In the only published study, Rommel et al (5) describe 700 children younger than 10 years who presented with feeding problems in a Belgian, multidisciplinary outpatient clinic. The purpose was to characterize the etiology of the feeding difficulties as medical, oral, or behavioral. Although they did assess for sensory issues, they relied on the clinical judgment of a speech-language pathologist rather than using a standardized measure. Results indicated that oral/medical problems were the major contributing factors for feeding issues in children younger than 2 years, whereas behavioral issues were the most prevalent after 2 years of age. The oral sensory-based feeding disorders corresponded with past medical intervention and nasogastric or gastrostomy tube feeding.
Despite strong evidence of a relation between sensory issues and feeding problems, there has been no global look at sensory issues faced by children who present to an outpatient multidisciplinary feeding team using a standardized measure of sensory processing abilities. This study describes the relation between sensory issues and medical complexity in a series of patients presenting to an outpatient multidisciplinary feeding team for evaluation, using a standardized measure of sensory processing abilities.
Patients were all children who presented at the University of Kansas Medical Center's multidisciplinary outpatient feeding clinic between 2004 and 2009. The feeding team consists of a physician, psychologist, dietician, and occupational therapist who see the patients simultaneously for an initial evaluation (and later for treatment, as appropriate). Patients are referred and refer themselves, and the purpose of the clinic is to help all children reach their full capacity regarding oral self-feeding. The parents of children who attend clinic for an initial evaluation are asked to complete several measures, one of which is the Short Sensory Profile (SSP). This retrospective chart review and all methodology for the study were approved by the human subjects committee.
The SSP is a 38-item caregiver-report measure designed to assess sensory processing problems in children 3 to 10 years of age (32), which takes approximately 10 minutes to complete and has been recommended for research protocols (33). Items are scored on a 5-point scale with higher scores indicating more impairment. The 7 subscales of the SSP are tactile sensitivity, taste/smell sensitivity, movement sensitivity, underresponsive/seeks sensation, auditory filtering, low energy/weak, and visual/auditory sensitivity. Scores result in categorizations of typical, probably difference (meaning the child's scores are more problematic than a typical child), and definite difference (meaning that the child's scores are in the clinical range). Internal consistency of the subscales ranges from 0.70 to 0.90 (33), and the measure has a reliability of 0.90 and discriminant validity >95% in identifying children with and without sensory modulation dysfunction (32). The total score is the best indicator of overall sensory dysfunction (32).
Medical Diagnostic Category
Patient charts were reviewed and all medical diagnoses were noted in a research database. These diagnoses were then independently categorized by 2 trained coders into 5 medical diagnostic categories, and agreement was assessed. Medical diagnostic categories included neurological, developmental, gastrointestinal, behavioral, and cardiorespiratory (Table 1). Categories are not mutually exclusive because a patient can have >1 diagnosis. These methods and categories have been previously validated (34).
All of the statistical analyses were performed using the SPSS version 18.0.3 (SPSS Inc, Chicago, IL). Interrater reliability of medical diagnostic category was calculated using the κ statistic. To begin, summary statistics were run on all SSP scores and on medical diagnostic categories. Because medical diagnostic categories were not mutually exclusive, a stepwise linear regression was run with SSP total score as the dependent variable and the 5 medical diagnostic categories as independent variables. To compare the SSP subscale scores by medical diagnostic category, a series of stepwise linear regressions were run with each subscale score as the dependent variable and medical diagnostic categories as the independent variables. A series of Pearson product-moment correlation coefficients were computed to assess the relation between medical diagnostic category and “definite difference” scores on the SSP. Because some children fell outside the recommended age range for the SSP, a 1-way analysis of variance was run on the SSP total score between children who were younger than 3 years and children who were over. There was not a significant difference between these 2 groups (F = 0.011, P = 0.916), so children of all ages were included in this article.
During the study period, 65 children older than 3 years attended a new clinic visit for a feeding team evaluation and had complete baseline measures and all were accepted into the present study. Children were an average of 45.77 months of age (SD 33.83) and 23 of them were girls (35.4%). See Table 2 for more information.
A total of 12 children (18.5%) had total scores in the typical range on the SSP, 9 scored in the probable difference range (13.8%), and 44 (67.6%) in the definite difference range. Results for the overall sample on the total sensory score, subscale scores, and by medical diagnosis are presented in Table 3.
Medical Diagnostic Categories
Agreement between coders of medical diagnostic categories was acceptable (κ = 0.89) and disagreements were resolved to arrive at the final coding used for analyses. Results indicate that 11 children (16.9%) did not have a medical diagnosis at the time of the initial appointment with the feeding team, necessitating further assessment. A total of 33 children (50.8%) had 1 diagnosis, 19 (29.2%) had 2 diagnoses, and 2 (3.1%) had 3 diagnoses, indicating this was a medically complicated population. The most common diagnostic category was developmental (n = 23), followed by gastrointestinal (n = 16) and neurological (n = 13; Table 1).
SSP Total Score by Medical Diagnostic Category
The stepwise multiple linear regression with SSP total score and the 5 medical diagnostic categories indicated that 2 diagnoses (behavioral, cardiorespiratory) explained a proportion of the variance that approached significance in SSP total score (R2 = 0.048, F(1,63) = 2.612, P = 0.081). As can be seen in Table 3, the percentage of subjects with definite difference scores on the SSP did differ by medical diagnostic category, with the highest being cardiorespiratory (91.7%), behavioral (71.4%), and gastrointestinal (70.6%). There was a significant relation between SSP total definite difference score and cardiorespiratory diagnosis (r = 0.206, P = 0.050) and behavioral diagnosis (r = −0.291, P = 0.009). Medical diagnosis by difference score is displayed in Table 4.
SSP Subscale Scores by Medical Diagnostic Category
Analyses on subscale scores indicated that for the tactile subscale, 2 diagnoses (cardiorespiratory, developmental) explained a significant portion of the variance (R2 = 0.154, F(2,64) = 5.629, P = 0.006). For the taste/smell subscale, only cardiorespiratory loaded significantly (R2 = 0.114, F(1,64) = 8.137, P = 0.006). None of the other domains (movement, underresponsive, auditory, low energy/weak, visual/auditory) loaded significantly onto diagnostic category.
The present study is the second to examine the possible co-occurrence of feeding problems and sensory issues in a pediatric sample, and the first to do so using a standardized measure of sensory processing, the SSP. Our sample of children presenting for evaluation at an outpatient multidisciplinary feeding clinic was medically complicated (32.3% had >1 diagnosis), which is consistent with previous reports (11). Other studies using the same medical diagnostic category coding scheme, however, have found differences in terms of the most common diagnostic categories. Our results indicate that the most common categories were developmental, followed by gastrointestinal and neurological, whereas previous research has suggested the most common categories were neurological, followed by mechanical/structural and behavioral. The reason for these differences is unclear but could be influenced by such factors as referral sources, variability in practices on the neonatal intensive care units and other services that provide care for these children, as well as variability in services available at different sites. For example, there is a well-known developmental screening and referral program colocated with our team, which could lead to a higher rate of developmental diagnoses at our site compared with the previous study.
Our study also found that a significant number of children presenting for a feeding evaluation had clinically significant scores on the SSP (67.6% of children had SSP scores in the definite difference range). The only previous study assessing sensory issues in children presenting for a feeding evaluation (5) suggested that 61% of children had an “oral diagnosis.” Given the fact that the “oral diagnosis” was based solely on clinician judgment and not using a standardized measure, it is surprising how close these 2 rates are and speaks to the clinical relevance of the SSP, and in particular a definite difference categorization on the total score. The previous study also indicated some differences in findings by age (5), but no such age differences were found in the present study, either for SSP total score or for medical diagnostic category.
Regarding the relation between medical diagnostic category and the SSP, our results indicate that the diagnostic categories of behavioral and cardiorespiratory were significantly related to SSP total score, as well as to the SSP definite difference score. This suggests that difficulties in sensory processing are significantly related to behavioral feeding problems such as dysfunctional feeder–child interaction, purposeful spitting at meals, and other behaviors that interfere with typical feeding development. Because this is the first use of a standardized measure such as SSP with a feeding population, it is not possible to compare these findings with previous research. Clinical judgment, however, would suggest that difficulties in sensory processing that are missed could lead to significant behavioral problems around mealtime, particularly as well-meaning therapists and caregivers attempt to coax the child to eat using typical feeding practices that are not sensitive to the child's existing sensory issues.
Our results also suggest that difficulties in sensory processing are significantly related to cardiorespiratory problems such as asthma, bronchopulmonary dysplasia, or pulmonic stenosis. Medicine has long been aware of the complicated, coordinated suck-swallow-breathe pattern required for oral eating (15). It is possible, and in fact likely, that children with cardiorespiratory problems are more likely to have difficulty navigating this suck-swallow-breathe pattern, which would lead to problems with choking, which definitely inhibits typical feeding development and, as our results would suggest, may also lead to increases in sensory processing problems.
Looking at the SSP subscale scores, our results suggest that only 2 subscales (tactile and taste/smell) were significantly related to medical diagnostic category. In particular, tactile was related to cardiorespiratory and developmental, and taste/smell to cardiorespiratory alone. Given that cardiorespiratory was one of the diagnostic categories most related to SSP total score, it is interesting to note that it is the only medical diagnostic category related to 2 subscales: tactile and taste/smell. Unfortunately, the larger literature does not help to explain this finding. There have been studies of the SSP with autistic children (35,36), children with attention-deficit/hyperactivity disorder (37), and children with learning disabilities (38), but no reports of the SSP in children with other medical issues that may explain these subscale findings. Future research is needed to determine why there is a relation between certain SSP subscale scores and certain medical diagnoses (for tactile, cardiorespiratory and developmental, and taste/smell, cardiorespiratory).
Our study does not come without limitations. First, it is a study of children presenting for evaluation: a unique group of children among all those with feeding problems. Second, the children were all presenting at a single site, which may have resulted in unique referral patterns or other biases. The age range of the children may be considered larger than typical, given the standard deviation in our age range. A definite strength of the study is the use of the SSP, a validated and reliable measure of sensory processing in young children. The SSP is a caregiver-report measure, however, that some may see as a weakness as it relies on the parent to report their perspective; however, in this case it was the best measure available of sensory issues in children presenting to a feeding clinic. Parent report of problems is also important to consider because it drives treatment seeking and other key behaviors, and research has shown that parents are reliable and valid reporters of their child's behaviors (39). Although our study did have the necessary power to detect significant differences between groups, our sample was somewhat small, so larger studies should be conducted in the future. Finally, our sample size was sufficient for all of the analyses conducted, but it was not large enough to allow for item-by-item analyses, something that may be useful (35).
In summary, this first analysis of the SSP in a group of children with feeding problems suggests that their feeding issues do indeed correlate with elevated scores of sensory processing, and that elevated scores are significantly related to certain medical diagnostic categories (behavioral, cardiorespiratory). Future research needs to examine why children who fall into certain medical diagnostic categories are more or less impaired regarding sensory processing, as well as why children who fall into certain medical diagnostic categories are more or less impaired in specific areas of sensory processing such as tactile or taste/smell. As most feeding programs have inpatient, day-treatment, and outpatient programs, it would also be interesting to compare sensory processing problems across these different treatment venues and to look at the relation between sensory processing difficulties and outcome of treatment of feeding issues. For practitioners who work in feeding teams, these results imply that having a truly multidisciplinary approach that incorporates sensory assessment into typical clinical care is recommended because many of these children are facing sensory issues that likely affect their feeding development.
The authors thank Dr John Belmont for statistical guidance.
1. Stevenson RD, Allaire JH. The development of normal feeding and swallowing. Pediatr Clin North Am
2. Carruth BR, Skinner J, Houck K, et al. The phenomenon of the “picky eater”: a behavioral marker in eating patterns of toddlers. J Am Coll Nutr
3. Lefton-Grief MA, Arvedson JC. Pediatric feeding and swallowing disorders: state of health, population trends, and application of the International Classification of Functioning, Disability, and Health. Semin Speech Lang
4. Manikam R, Perman JA. Pediatric feeding disorders. J Clin Gastro
5. Rommel N, De Meyer AM, Feenstra l, et al. The complexity of feeding problems in 700 infants and young children presenting to a tertiary care institution. J Pediatr Gastroenterol Nutr
6. Stephens BE, Vohr BR. Neurodevelopmental outcome of the premature infant. Pediatr Clin North Am
7. Rudolph CD, Link DT. Feeding disorders in infants and children. J Pediatr Gastroenterol Nutr
8. Staiano A. Food refusal in toddlers with chronic diseases. J Pediatr Gastroenterol Nutr
9. Miller CK. Updates on pediatric feeding and swallowing problems. Curr Opin Otolaryngol Head Neck Surg
10. Zangen T, Ciarla C, Zangen S, et al. Gastrointestinal motility and sensory abnormalities may contribute to food refusal in medically fragile toddlers. J Pediatr Gastroenterol Nutr
11. Burklow KA, McGrath AM, Valerius KS, et al. Relationship between feeding difficulties, medical complexity, and gestational age. Nutr Clin Prac
12. Mason SJ, Harris G, Blissett J. Tube feeding in infancy: implications for development of normal eating and drinking skills. Dysphagia
13. Senez C, Guys JM, Mancini J, et al. Weaning children from tube to oral feeding. Childs Nerv Syst
14. Arvedson JC, Brodsky L. Pediatric Swallowing and Feeding: Assessment and Management.
2nd ed. Albany, NY: Singular Publishing Group; 2002.
15. Verhagen JV, Engelen L. The neurocognitive bases of human multimodal food perception: sensory integration. Neurosci Biobehav Rev
16. Frank RA, Ducheny K, Mize S. Strawberry odor, but not red color, enhances the sweetness of sucrose solutions. Chem Senses
17. Mackey A, Valassi K. The discernment of primary tastes in the presence of different food textures. Food Technol
18. Mackey A. Discernment of taste substances as affected by solvent medium. Food Res
19. Stone H, Oliver S. Effect of viscosity on the detection of relative sweetness intensity of sucrose solutions. J Food Sci
20. Arabie MA, Moskowitz HR. The effects of viscosity upon perceived sweetness. Percept Psychophys
21. Pangborn RM, Trabue IM, Szcezesniak AS. Effects of hydrocolloids on oral viscosities and basic taste intensities. J Texture Stud
22. Pangborn RM, Gibbs ZM, Tassan C. Effect of hydrocolloids on apparent viscosity and sensory properties of selected beverages. J Texture Stud
23. Christensen CM. Texture-taste interactions. Cereal Food World
24. Christensen CM. Effects of solution viscosity on perceived saltiness and sweetness. Percept Psychophys
25. Izutsu T, Taneya S, Kikuchi E, et al. Effect of viscosity on perceived sweetness intensity of sweetened sodium carboxymethylcellulose solutions. J Texture Stud
26. Pangborn RM, Chrisp RB, Bertolero L. Gustatory salivary and oral thermal responses to solutions of sodium chloride at four temperatures. Percept Psychophys
27. McBurney DH, Collings M, Glanz M. Temperature dependence of human taste responses. Physiol Behav
28. Pangborn RM. Influence of color on the discrimination of sweetness. Am J Psychol
29. Pangborn RM, Hansen B. The influence of color on discrimination of sweetness and sourness in pear-nectar. Am J Psychol
30. Mega JA. Influence of color on taste thresholds. Chem Sens Flav
31. Johnson JL, Clydesdale FM. Perceived sweetness and redness in colored sucrose solutions. J Food Sci
32. McIntosh DN, Miller LJ, Shyu V. Development and validation of the short sensory profile. In: Dunn W, ed. Sensory Profile User's Manual.
San Antonio, TX: Psychological Corporation; 1999:59–73.
33. Dunn W. The Sensory Profile: User's Manual. San Antonio, TX:Psychological Corporation; 1999.
34. Burklow KA, Phelps AN, Schultz JR, et al. Classifying complex pediatric feeding disorders. J Pediatr Gastrol Nutr
35. Tomchek S, Dunn W. Sensory processing in children with and without autism: a comparative study utilizing the short sensory profile. Am J Occup Ther
36. Rogers SJ, Hepburn S, Wehner E. Parent reports of sensory symptoms in toddlers with autism and those with other developmental disorders. J Autism Dev Disord
37. Dunn W, Bennett D. Patterns of sensory processing in children with attention deficit hyperactivity disorder. Occup Ther J Res
38. Dove S, Dunn W. Sensory processing patterns in learning disabilities. Occup Ther Early Intervention Preschool Schools
39. Bricker D, Squires J. The effectiveness of parental screening of at-risk infants: the monitoring questionnaires. Topics Early Child Spec Educ