Appetite 133 (2019) 223–230

Contents lists available at ScienceDirect

Appetite
journal homepage: www.elsevier.com/locate/appet

Sensory Eating Problems Scale (SEPS) for children: Psychometrics and T

associations with mealtime problems behaviors
Laura Seiverlinga, Keith E. Williamsb,∗, Helen M. Hendyc, Whitney Adamsb, Stella Yusupovad,
Aleksandra Kaczore
a
Ball State University, USA
b
Penn State Hershey Medical Center, USA
c
Penn State University, Schuylkill Campus, USA
d
St Mary's Hospital for Children, Bayside, NY, USA
e
Queens College (City University of New York), USA

ARTICLE INFO ABSTRACT

Keywords: The present study developed the 22-item Sensory Eating Problems Scale (SEPS) to measure sensory aspects for
Sensory feeding problems children surrounding eating, documented psychometrics of SEPS subscales, and examined their association with
Food touch sensitivity mealtime behavior problems. Study participants were 449 caretakers of children referred to feeding clinics,
Gagging including children in three special needs status groups: autism spectrum disorder (ASD), other special needs, and
Overstuffing
no special needs. Caretakers completed surveys to report children's demographics, four measures of children's
Mealtime behavior problems
Diet variety
mealtime behavior problems, and five-point ratings for how often children showed various sensory feeding
reactions. Exploratory factor analysis of the sensory feeding items identified six SEPS subscales with acceptable
goodness-of-fit, internal reliability, and test-retest reliability: Food Touch Aversion, Single Food Focus, Gagging,
Temperature Sensitivity, Expulsion, and Overstuffing. ANCOVAs revealed that child demographics most asso-
ciated with higher SEPS subscale scores were younger age and special needs. Multiple regression analyses found
that children's mealtime behavior problems were most often associated with SEPS subscales of Food Touch
Aversion, Single Food Focus, Expulsion, and Overstuffing, with the set of six subscales explaining 18–44% of
variance in mealtime behavior problems. Suggestions for how clinicians and researchers may find the SEPS
useful for assessment and intervention are provided.

1. Introduction tactile sensitivity, taste/smell sensitivity, and visual/auditory sensi-

tivity subscales of the Short Sensory Profile (SSP; Dunn, 1999)
The etiology of childhood feeding problems is often complex and (Coulthard & Blissett, 2009). While these researchers found positive
likely due to multiple factors (Williams, Field, & Seiverling, 2010) such correlations between all three sensory subscales and the neophobia
as biological issues, oral motor deficits, and environmental con- measure, the highest correlation was between the four-item Taste/smell
tingencies. Sensory processing, or the ability to receive, integrate, and sensitivity subscale and the FNS. This positive correlation indicates
process sensory input, such as visual, olfactory, or gustatory informa- higher levels of sensory impairment or dysfunction were related to a
tion, has also been speculated to influence eating (Brown, Morrison, & greater reluctance to eat new foods. The taste/smell sensitivity subscale
Stagnitti, 2010). A growing literature has examined the association was also inversely correlated to both parent-reported measures of
between sensory processing and eating behavior in several populations portions consumed and preference for fruits and vegetables. A sub-
of children. sequent study involved having caregivers of 180 preschoolers complete
Among children with typical development, a few studies have ex- the 12-item Oral Sensory Processing subscale of the Sensory Profile (SP;
amined the relation between sensory processing and food neophobia. In Dunn, 1999) and the FNS (Johnson, Davies, Boles Gavin, & Bellows,
both of these studies, neophobia was measured using the Child Food 2015). While results of this subsequent study also showed impaired
Neophobia Scale (FNS; Pliner, 1994). In the first study, caregivers of 73 sensory processing was positively related to food neophobia, this study
children completed questionnaires which included the FNS and the found no association between sensory processing and fruit or vegetable

∗
Corresponding author. Penn State Hershey Medical Center, Feeding Program, 905 W. Governor Road, Hershey, PA, 17033, USA.
E-mail address: [email protected] (K.E. Williams).

https://doi.org/10.1016/j.appet.2018.11.008
Received 26 June 2018; Received in revised form 2 October 2018; Accepted 12 November 2018
Available online 14 November 2018
0195-6663/ © 2018 Elsevier Ltd. All rights reserved.
L. Seiverling et al. Appetite 133 (2019) 223–230

consumption {Note: The four-item Taste/smell sensitivity subscale of problems relevant to children's feeding problems. Further, along the
the SSP is a subset of the Oral Sensory Processing subscale of the SP}. Oral Sensory Processing Subscale of the SP, a particular oral sensory
Children with tactile sensitivity, as measured with the SP, have also problem, such as sensitivity to texture, may be measured by only one or
been found to be more likely to gag on foods, less likely to try foods due two items.
to smell or temperature, and be less likely to eat unfamiliar foods than The purpose of the present study is to provide a more detailed as-
children who are not tactile sensitive (Smith, Roux, Naidoo, & Venter, sessment of sensory eating behavior than provided in previous studies
2005). by developing the Sensory Eating Problems Scale (SEPS) to identify
Sensory processing has been extensively examined in children with specific dimensions of oral sensitivity children show during feeding.
autism spectrum disorders (ASD) (for review, Ben-Sasson et al., 2009). Child demographics (age, gender, special needs status, BMI z-score)
Several studies have shown children with ASD exhibit more sensory associated with each SEPS subscale were examined along with the
processing dysfunction in areas of taste/smell sensitivity or oral sensory convergent validity between the SEPS and the Oral Sensory Processing
processing than children with typical development (Tomchek & Dunn, subscale from the latest version of the SP, the Sensory Profile-2 (SP-2;
2007; Wiggins, Robins, Bakeman, & Adamson, 2009). Researchers have Dunn, 2014). Finally, the study aimed to determine which SEPS sub-
also examined the relation between sensory processing and various scales were most associated with children's mealtime behavior pro-
measures of eating or intake in children with ASD. In one study, re- blems as reported by parents. It is hoped information from this measure
searchers examined the relation between sensory processing and eating could aid clinicians and caregivers to focus their interventions on the
problems in a sample of 97 children with ASD and found that greater oral sensitivity issues that are most likely to affect children's mealtime
sensory processing problems according to the Taste/smell, Tactile, and problems.
Visual/auditory subscales of the SSP were related to a greater number
of eating problems (Nadon, Feldman, Dunn, & Gisel, 2011). In a study 2. Method
examining the association between mealtime behavior problems, as
measured by the Brief Autism Mealtime Behavior Inventory (BAMBI; 2.1. Participants
Lukens & Linscheid, 2008), and the SSP, higher ratings of mealtime
behavior were significantly related to greater sensory dysfunction Participants of the present study included 449 caregivers and their
(Zobel-Lachiusa, Andrianopoulos, Mailloux, & Cermak, 2015). This children 24 months and older who were referred to hospital-based
study also found children with ASD had both higher levels of sensory feeding clinics in Pennsylvania and New York (67.9% male; mean
dysfunction and more mealtime problem behaviors than children with age = 69.59 months, SD = 38.84; mean body mass index, BMI z-
typical development. The findings of Zobel-Lachiusa and colleagues score = -0.03, SD = 1.44). The children had been referred to the clinics
were replicated in a subsequent study which found relations between for eating problems that included failure to gain weight, dependence on
the BAMBI and full-length version of the SP as well as significantly tube feeding or oral supplements, difficulties with texture or learning to
more mealtime problem behaviors and sensory dysfunction in ASD chew, limited diet variety, and mealtime behavior problems. Children's
compared to typically developing siblings and typically developing weight status was determined by clinic staff who measured the child's
peers (Shmaya, Eilat-Adar, Leitner, Reif, & Gabis, 2017). weight and height, then calculated the child's body mass index
Research has also started to explore sensory processing among (BMI = kg2/m) percentile and z-scores in comparison to children of
children identified with eating problems. In the first examination of their age and gender using online applications from the Children's
sensory processing in a clinical sample of 65 children with feeding Hospital of Philadelphia (http://stokes.chop.edu/web/zscore).
problems, Davis et al. (2013) found sensory processing as measured by Children were divided into three groups based on special needs status:
the SSP was impaired in 68% of the sample. They also found categories autism spectrum disorder, ASD (n = 156), other special needs including
of medical diagnoses (e.g. cardio-respiratory, developmental, gastro- intellectual disabilities, speech delay, attention deficit hyperactivity
intestinal) were related to performance on the Tactile and Taste/Smell disorder, cerebral palsy, traumatic brain injury, spinal bifida, or seizure
subscales of the SSP. A later study comparing 16 children identified disorders (n = 144), and no special needs (n = 149). Children's special
with failure to thrive and feeding problems with age-matched controls needs status was determined by review of patient records and caretaker
on the Infant/Toddler Sensory Profile (ITSP; Dunn & Daniels, 2002) interviews during an initial evaluation that included two or more pro-
revealed more sensory processing problems among the children with fessionals from speech pathology, nutrition, psychology, nursing, and
feeding problems (Yi, Joung, Choe, Kim, & Kwon, 2015). This study pediatric gastroenterology.
found the sensory processing impairments were limited to the Tactile, From the sample of 449 caregivers of feeding clinic children, one
Vestibular, and Oral subscales of the ITSP. Both studies noted their subset of 48 caregivers was randomly selected for examination of test-
limited sample sizes and called for further investigation of sensory retest reliability for SEPS subscales (72.9% male; mean age = 74.42
processing among children with feeding problems. months, SD = 38.17; mean BMI z-score = −0.11, SD = 1.36; 18 with
To date, much of the research exploring the relations between ASD, 12 with other special needs, 18 with no special needs). These 48
sensory processing and feeding or eating behavior has involved some caregivers rated SEPS items on two occasions before feeding clinic in-
version of Dunn's SP. While the original SP contains 125 questions in- terventions began for their children's eating problems, with SEPS sub-
volving multiple areas of sensory processing, the Oral Sensory scale scores calculated at each occasion to examine their correspon-
Processing subscale consists of only ten items. A subset of four of these dence (as described below in data analysis).
ten questions comprise the Taste/smell subscale on the 38-item SSP From the sample of 449 caregivers of feeding clinic children, an-
(Dunn, 1999). Similarly, the oral sensory processing subscale of the 48- other subset of 62 caregivers was randomly selected for examination of
item ITSP consists of five questions regarding eating. convergent validity of SEPS subscales with the Oral Sensory Processing
While child feeding problems have been found to be associated with subscale from the SP-2 (74.2% male; mean age = 62.69 months,
sensory processing problems, clinicians could benefit from more specifics SD = 29.62; mean BMI z-score = −0.26, SD = 1.30; 22 with ASD, 21
about the types of oral sensory problems most associated with feeding with other special needs, 19 with no special needs). These 62 caregivers
problems. For example, the Oral Sensory Processing subscale of the used a five-point rating (blank or 1 = almost never, 2 = occasionally,
Sensory Profile includes a collection of sensory problems associated with 3 = half the time, 4 = frequently, 5 = almost always) to respond to the
feeding difficulties such as gagging easily to food or utensils in the 86 items of the SP-2 (Dunn, 2014). Each child's score on the Oral
mouth, rejecting certain food tastes or smells, accepting limited food Sensory Processing subscale was calculated as the mean five-point
textures, biting the tongue or lips (Dunn, 2014), but without doc- rating given for the 10 items in that subscale so its association with each
umentation that identifies the separate dimensions of oral sensitivity SEPS subscale could be examined (as described below in data analysis).

224
L. Seiverling et al. Appetite 133 (2019) 223–230

2.2. Item development Table 1

Results from exploratory factor analysis showing six dimensions of the 22-item
To develop the questions included in the SEPS, two areas of the SEPS for 449 caretakers of feeding clinic children.
literature were reviewed. First, existing measures of sensory processing # Item Factor loading
(e.g., Baranek, David, Poe, Stone, & Watson, 2006; Dunn, 2014; Schoen,
Miller, & Sullivan, 2014) were examined for items related to eating. Food Touch Aversion
7. My child has a clear dislike for food touching his/her lips. .878
Second, articles referring to “sensory” and eating problems were re-
8. My child does not like if food touches his/her teeth. .847
viewed to evaluate the types of sensory problems described (e.g., 19. My child gets upset when food or liquid touches his/her .777
Johnson, Davies, Boles, Gavin, & Bellows, 2015). In addition to a review lips.
of the literature, speech pathologists, occupational therapists, and be- 15. My child has difficulty touching food with his/her fingers. .647
Single Food Focus
havioral therapists working at two feeding programs were interviewed
18. My child will refuse entire food categories (e.g., all fruits, .756
regarding sensory issues encountered in their clinical work. A list of 46 all vegetables).
questions regarding sensory eating issues was developed and these 4. My child will eat one food for weeks or months at a time. .720
questions were included in a pre-intervention survey. 3. My child accepts only one flavor of a certain type of food .668
(e.g., strawberry yogurt).
35. My child avoids mixed textures of foods (e.g., spaghetti .653
2.3. Procedures and measurement
and meatballs).
Gagging
The 449 caregivers of feeding clinic children filled out a pre-inter- 16. My child has a sensitive gag reflex. .775
vention survey to provide children's demographics, respond to four 40. My child will gag or vomit at the sight of new food. .663
12. My child gags when food touches his/her tongue. .663
measures of mealtime behavior problems, and to the 46-item SEPS
27. My child gags when a spoon is placed directly on his/her .587
shown by their children using a five-point rating (0 = never, tongue.
1 = rarely, 2 = sometimes, 3 = often, 4 = always). The subset of 48 Temperature Sensitivity
caregivers whose responses were used for test-retest reliability of SEPS 11. My child will only eat foods that are room temperature. .821
subscales were given the SEPS measure on a second pre-intervention 9. My child is sensitive to food temperature. .706
2. My child will only eat foods that are warm. .644
occasion. The subset of 62 caregivers whose responses were used to
10. My child will only eat foods that are cold. .640
examine convergent validity for SEPS subscales also completed the Oral Expulsion
Sensory Processing subscale of the SP-2 (Dunn, 2014). 30. My child spits out food or liquid. .774
To measure children's mealtime behavior problems, caregivers were 14. My child uses his or her fingers to take food out of his/her .763
asked to respond to the 10-item Brief Assessment of Mealtime Behavior mouth.
13. My child expels food or liquid. .656
in Children (BAMBIC; Hendy, Seiverling, Lukens, & Williams, 2013), Overstuffing
using a five-point rating for how often each mealtime behavior occurred 24. My child attempts to swallow large pieces of food. .877
(blank or 1 = never/rarely, 2 = seldom, 3 = occasionally, 4 = often, 25. My child attempts to swallow bites of food without .746
5 = almost every meal). The BAMBIC includes three subscales of chewing.
21. My child overstuffs his/her mouth with food. .745
mealtime behavior problems as perceived by parents (Food Refusal,
Limited Variety, Disruptive Behavior). The present sample showed ac-
Excluded Items:
ceptable internal reliability for the three BAMBIC subscales: Food Re-
fusal (α = 0.78, M = 2.81, SD = 1.24), Limited Variety (α = 0.71, 1. My child has a preference for crunchy foods.
M = 4.02, SD = 0.98), Disruptive Behavior (α = 0.75, M = 1.68, 5. My child has a preference for salty foods.
6. My child has a strong preference to use food condiments.
SD = 0.94). Additionally, to measure children's diet variety as the
17. My child has a strong preference for certain food textures.
number of 80 common foods rejected, caregivers reported how often 20. My child refuses to hold utensils.
their children consumed each food (blank or 0 = never, 1 = daily, 22. My child has a preference for sweet foods.
2 = weekly, 3 = monthly), then each child's score for # 80 foods re- 23. My child exhibits facial grimaces when swallowing.
jected was the number “never” eaten. 26. My child chews using his/her front teeth.
28. My child prefers drinking over eating.
29. My child excessively wipes his/her tongue or lips when residue is present.
2.4. Data analysis 31. My child takes very small bites of food during meals.
32. My child gets upset if he/she gets dirty during a meal.
The primary goal for data analysis was to develop the SEPS and 33. My child regurgitates swallowed food.
34. My child has a preference for spicy foods.
examine psychometrics of its dimensions/subscales. Prior to ex-
36. My child avoids drinking from an open cup.
ploratory factor analysis, any of the 46 items showing “floor effects” 37. My child gags or vomits when he/she smells a food cooking.
(with 75% or more of participants giving a 0 = “never” rating) or 38. My child refuses to sit at the table while others are eating.
“ceiling effects” (with 75% or more of participants giving a 4 = “al- 39. My child cannot be in the kitchen when the family is preparing food.
ways” rating) were eliminated from consideration. For remaining items, 41. My child only eats from one particular bowl/cup/utensil.
42. My child is sensitive to food color.
exploratory factor analysis was conducted on the ratings given by the
43. My child has a preference for flavorful foods.
caretakers of the 449 children (See Table 1). The factor analysis was 44. My child takes a drink after every bite during meals.
conducted using varimax rotation to identify dimensions of sensory 45. My child has a preference for non-crunchy foods (e.g. yogurt).
sensitivity as conceptually clear and distinct from each other as pos- 46. My child gags or vomits when watching others eat.
sible. Each subscale of the SEPS was required to include at least three
items, with all items showing factor loadings of at least 0.50 for that
subscale only, and with all subscales showing internal reliability values: relative Chi2 with values of 2.00 or less suggesting good fit and
(Cronbach's α) of 0.70 or higher. Such strict criteria were used to in- values of 5.00 or less suggesting acceptable fit, CFI with values of 0.95
crease the probability that SEPS subscales would show acceptable or above suggesting good fit and values of 0.90 or above suggesting
psychometrics including goodness-of-fit, test-retest reliability, and acceptable fit, and RMSEA with values of 0.05 or less suggesting good
convergent validity with the Oral Sensory Processing subscale of the SP- fit and values of 0.08 or less suggesting acceptable fit (Arbuckle, 2007).
2 (Dunn, 2014). Psychometric examination of the SEPS subscales for the 449 caregivers
Psychometric examination of the SEPS for the sample of 449 care- also included calculation of internal reliability (Cronbach's α). Ad-
givers included calculation of a recommended set of goodness-of-fit ditionally, test-retest reliability was examined for the smaller subset of

225
L. Seiverling et al. Appetite 133 (2019) 223–230

48 caregivers by calculating a Pearson correlation between subscale failed to show factor loadings of at least 0.50 on any dimension (items
scores provided on two pre-intervention occasions, with each SEPS #6, #20, #23, #28, #29, #32, #33, #41, #42, #43, #44). Ten items
subscale score defined as the mean five-point rating for items within the were eliminated because they loaded on dimensions with fewer than
subscale. Finally, convergent validity of the SEPS was examined for the three items (items #1, #5, #17, #22, #26, #31, #36, #38, #39, #45)
subset of 62 caregivers by calculating Pearson correlations between (See Table 1).
each SEPS subscale and the Oral Sensory Processing subscale of the SP- In examination of psychometrics, the SEPS showed acceptable
2 (Dunn, 2014). Because items on the Oral Sensory Processing subscale goodness-of-fit values for the sample of 449 caregivers: relative
include various feeding sensory problems (gagging, food textures, food Chi2 = 2.63, CFI = 0.90, RMSEA = 0.060. Also, each of the SEPS sub-
smells), we anticipated that a number of the new SEPS subscales would scales revealed acceptable internal reliability with Cronbach's α scores:
be found significantly associated with it. Food Touch Aversion (0.85), Single Food Focus (0.70), Gagging (0.73),
Another goal for data analysis was to examine child demographics Temperature Sensitivity (0.72), Expulsion (0.71), and Overstuffing
associated with increased risk for each dimension/subscale of the SEPS (0.71). Table 2 shows bivariate correlations for the six SEPS subscales.
for the primary sample of 449 caregivers and their children. Using each Additionally, each SEPS subscales revealed significant test-retest cor-
SEPS subscale score as the dependent variable (again defined as the relations for scores calculated at two pre-intervention occasions for the
mean rating for items in the dimension), a 2 × 3 ANCOVA was con- smaller subset of 48 caregivers, although test-retest reliability did not
ducted to compare the subscale score across two genders (male, female) reach the traditionally recommended 0.70 for three of the six subscales:
and across three special needs groups (ASD, other, none), with child age Food Touch Aversion (r = 0.76, p = .000), Single Food Focus (r = 0.78,
and BMI z-score as covariates. To avoid Type I errors and maintain p = .000), Gagging (r = 0.52, p = .000), Temperature Sensitivity
strong statistical power to detect significant variable relationships, we (r = 0.51, p = .000), Expulsion (r = 0.67, p = .000), and Overstuffing
used a more conservative criterion of p < .01 rather than the tradi- (r = 0.81, p = .000). Finally, support for convergent validity of the
tional p < .05 in these repeated ANCOVAs. SEPS subscales was documented in significant correlations between the
The final goal for data analysis was to determine which SEPS sub- Oral Sensory Processing subscale of the SP-2 (Dunn, 2014) and five of
scales were most associated with mealtime behavior problems for the the six SEPS subscales for the smaller subset of 62 caregivers: Food
449 clinic children. Measures of four mealtime behavior problems in- Touch Aversion (r = 0.38, p = .003), Single Food Focus (r = 0.57,
cluded the three BAMBIC subscales (Food Refusal, Limited Variety, p = .000), Gagging (r = 0.31, p = .014), Temperature Sensitivity
Disruptive Behavior) and # of 80 common foods rejected. Using each of (r = 0.39, p = .002), Expulsion (r = 0.39, p = .002), and Overstuffing
these four mealtime behavior problems as the outcome variable, mul- (r = −0.073, p = .570).
tiple regression analysis was conducted with SEPS subscales as possible
predictor variables. To avoid Type I errors and maintain strong statis-
tical power to detect significant variable relationships, we used a more 3.2. Child demographics associated with SEPS subscales
conservative criterion of p < .01 rather than the traditional p < .05 in
these multiple regression analyses. The 2 × 3 ANCOVAs (Table 3) to examine children's demographics
(gender, special needs, age, BMI z-score) associated with SEPS subscales
3. Results reported by the 449 caregivers revealed significant effects for age, with
younger children showing more Food Touch Aversion (r = −0.19,
3.1. SEPS subscales and psychometrics p = .000), Gagging (r = −0.13, p = .007), Temperature Sensitivity
(r = −0.16, p = .001), and Expulsion (r = −0.34, p = .000). These
Prior to exploratory factor analysis, three of the 46 items were ANCOVAs also revealed significant effects for special needs status, with
eliminated from consideration (items #34, #37, #46) because of “floor post-hoc t-test comparisons showing that children with ASD and other
effects” in caregiver ratings, with 75% or more of them reporting their needs had more Food Touch Aversion, Expulsion, and Overstuffing than
children “never” showed that reaction. No “ceiling effects” were seen in did children with no special needs, and children with other needs had
caregiver ratings. Then, exploratory factor analysis of responses by the less Single Food Focus than the other two special needs groups (Table 4
449 caregivers of feeding clinic children to the 43 remaining items shows SEPS subscales separately for the three special needs status
produced the new 22-item SEPS with six subscales: Food Touch groups). Finally, the ANOVAs found no significant main effects for
Aversion (4 items), Single Food Focus (4 items), Gagging (4 items), gender, main effects for BMI z-score, or gender X special needs inter-
Temperature Sensitivity (4 items), Expulsion (3 items), and action effects for any of the six SEPS subscales (using p < .01).
Overstuffing (3 items). Eleven items were eliminated because they

Table 2
Bivariate correlations of study variables for 449 caretakers of feeding clinic children.
Food Touch Aversion Single Food Gagging Temperature Expulsion Overstuffing (BAMBIC) Food (BAMBIC) (BAMBIC) # 80 Foods
Focus Sensitivity Refusal Limited Variety Disruptive Rejected
Behavior

Food Touch Aversion .259*** .438*** .336*** .324*** -.004 .462*** .174*** .265*** .421***
Single Food Focus .208*** .296*** .099 -.104* .161** .649*** .027 .435***
Gagging .301*** .426*** .128** .285*** .074 .137** .205***
Temperature Sensitivity .193*** .080 .179*** .127** .148** .109*
Expulsion .193*** .504*** .066 .358*** .068
Overstuffing .048 -.161** .180*** -.208***
(BAMBIC) Food Refusal .222*** .555*** .263***
(BAMBIC) Limited .077 .378***
Variety
(BAMBIC) Disruptive .112*
Behavior
# 80 Foods Rejected

*p < .05, **p < .01, ***p < .001.

226
L. Seiverling et al. Appetite 133 (2019) 223–230

Table 3 Problems of Disruptive Behavior (BAMBIC) were associated with SEPS

Results from 2 × 3 ANCOVAs to examine child demographics associated with Food Touch Aversion, Expulsion, and Overstuffing. Problems of a high
six SEPS subscales for 449 caregivers of feeding clinic children, comparing # of 80 foods rejected were associated with SEPS Food Touch Aversion,
across gender, special needs status (ASD, Other, None), with age and BMI z- Single Food Focus, and Overstuffing. The multiple regression analyses
score as covariates.
revealed that only the SEPS subscales of Gagging and Temperature
Effect F (df) p partial eta2 effect size Sensitivity were not significantly associated with any of the four mea-
sures of mealtime behavior problems included in the present study (See
Food Touch Aversion
Table 5).
Age 24.11 (1, 441) .000 .052
BMI z-score .04 (1, 441) .852 .000
Gender 3.15 (1, 441) .076 .007
Special needs 9.02 (2, 441) .000 .039 4. Discussion
Gender X special needs .37 (2, 441) .692 .002
Single Food Focus
Age .03 (1, 441) .860 .000
The 22-item SEPS developed in the present study offers clinicians
BMI z-score 4.27 (1, 441) .039 .010 and researchers a new measure to examine more specific sensory eating
Gender .89 (1, 441) .346 .002 problems than can be captured with previous measures, all of which
Special needs 5.29 (2, 441) .005 .023 were derived from Dunn's Oral Sensory Processing subscale of the SP
Gender X special needs 1.68 (2, 441) .189 .008
(Dunn, 1999, 2014; Dunn & Brown, 1997; Dunn & Daniels, 2002). The
Gagging
Age 7.76 (1, 441) .006 .017 six specific aspects of sensory eating problems included in the SEPS
BMI z-score 3.55 (1, 441) .060 .008 include Food Touch Aversion, Single Food Focus, Gagging, Tempera-
Gender .43 (1, 441) .513 .001 ture Sensitivity, Expulsion, and Overstuffing. Thus, the SEPS is wider
Special needs 1.34 (2, 441) .263 .006 ranging and more detailed in the measurement of sensory eating pro-
Gender X special needs 1.68 (2, 441) .188 .008
Temperature Sensitivity
blems than existing measures.
Age 14.25 (1, 441) .000 .031 Results from the present study for child demographics associated
BMI z-score .08 (1, 441) .779 .000 with SEPS subscales support and extend the existing literature. For
Gender 2.09 (1, 441) .149 .005 example, present results found three SEPS subscales (Food Touch
Special needs 1.50 (2, 441) .224 .007
Aversion, Expulsion, Overstuffing) were greater in children with ASD
Gender X special needs 1.16 (2, 441) .315 .005
Expulsion and other special needs, consistent with past research showing children
Age 67.29 (1, 441) .000 .132 with ASD exhibit more problems with sensory processing and
BMI z-score 2.13 (1, 441) .145 .000 Overstuffing than children without special needs (Seiverling, Towle,
Gender 4.34 (1, 441) .038 .010 Hendy, & Pantelides, 2018; Shmaya et al., 2017; Zobel-Lachiusa et al.,
Special needs 9.53 (2, 441) .000 .041
Gender X special needs .90 (2, 441) .407 .004
2015). Present results also found four SEPS subscales (Food Touch
Overstuffing Aversion, Gagging, Temperature Sensitivity, Expulsion) associated with
Age 4.21 (1, 441) .041 .009 younger age. It is possible these sensory feeding problems attenuate
BMI z-score 5.25 (1, 441) .022 .012 across the course of development as children are repeatedly exposed to
Gender .01 (1, 441) .907 .000
various tastes, temperatures, and textures of food. This would be an
Special needs 15.42 (2, 441) .000 .065
Gender X special needs .96 (2, 441) .383 .004 area of future research with the SEPS using non-clinic samples.
Results from the present study revealed four of the six SEPS sub-
scales were significantly associated with the four measures of mealtime
behavior problems considered (three BAMBIC subscales, # of 80
3.3. SEPS subscales associated with children's mealtime behavior problems common foods rejected). Similarly, past research has found children's
problems with (non-food related) sensory processing associated with a
The multiple regression analyses (Table 5) revealed that four of the more limited diet variety (Chistol et al., 2018; Coulthard & Blissett,
six SEPS subscales were significantly associated with two or more 2009), and with higher scores on the BAMBI (Shmaya et al., 2017;
mealtime behavior problems in children (measured with three BAMBIC Zobel-Lachiusa et al., 2015), an earlier version of the BAMBIC included
subscales, # of 80 foods rejected), with the set of six SEPS subscales in the present study. Specifically, the four SEPS subscales most asso-
explaining 18–44% of variance in these mealtime problems. More ciated with our measures of mealtime behavior problems were Food
specifically, problems of Food Refusal (BAMBIC) were associated with Touch Sensitivity, Single Food Focus, Expulsion, and Overstuffing, with
more SEPS Food Touch Aversion and Expulsion. Problems of Limited the full set of six SEPS subscales explaining 18–44% of the variance in
Variety (BAMBIC) were associated with SEPS Single Food Focus. these mealtime behavior problems.

Table 4
Descriptive Statistics for Six SEPS Subscales, Shown Separately for Children from Three Special Needs Groups: ASD (N = 156), Other Special Needs (N = 144), No
Special Needs (N = 149). Subscale Scores are Means of Five-Point Ratings (0 = Never to 4 = Always).
ASD OTHER NONE

M (SD) range % with score of 0 M (SD) range % with score of 0 M (SD) range % with score of 0
(never) (never) (never)

Food Touch Aversion 1.17 (1.08) (0.00–4.00) 25.6% .96 (.99) (0.00–4.00) 27.1% .66 (.91) (0.00–3.75) 42.3%
Single Food Focus 2.48 (1.00) (0.00–4.00) 2.6% 1.98 (1.09) (0.00–4.00) 6.3% 2.39 (1.07) (0.00–4.00) 2.7%
Gagging .95 (.91) (0.00–3.75) 26.9% 1.05 (.84) (0.00–3.25) 17.4% .95 (.85) (0.00–3.25) 22.1%
Temperature Sensitivity 1.28 (.87) (0.00–4.00) 12.8% 1.09 (.87) (0.00–3.00) 18.1% 1.09 (.87) (0.00–3.00) 18.8%
Expulsion 1.43 (.94) (0.00–3.67) 7.7% 1.58 (.99) (0.00–4.00) 8.3% 1.20 (.87) (0.00–3.67) 17.4%
Overstuffing 1.22 (1.03) (0.00–4.00) 21.2% .85 (.86) (0.00–3.67) 31.3% .62 (.79) (0.00–3.33) 48.3%

227
L. Seiverling et al. Appetite 133 (2019) 223–230

Table 5
Results from multiple regression analyses to examine six SEPS subscales as predictors of feeding problems reported by 449 caretakers of feeding clinic children.
(BAMBIC) FOOD REFUSAL (BAMBIC) LIMITED VARIETY (BAMBIC) DISRUPTIVE BEHAVIOR # 80 FOODS REJECTED

Beta t p Beta t p Beta t p Beta t p

Food Touch Aversion .314 6.87 .000 .050 1.17 .243 .206 4.04 .000 .361 7.91 .000
Single Food Focus .045 1.07 .285 .654 16.85 .000 -.034 .73 .466 .353 8.43 .000
Gagging -.039 .83 .410 -.080 1.83 .069 -.110 2.09 .036 .049 1.06 .291
Temperature Sensitivity -.008 .18 .860 -.051 1.28 .200 .053 1.11 .269 -.108 2.51 .012
Expulsions .419 9.41 .000 .038 .93 .352 .306 6.15 .000 -.054 1.23 .220
Overstuffing -.019 .48 .634 -.080 2.14 .033 .129 2.85 .005 -.157 3.89 .000

R2 = .34 R2 = .44 R2 = .18 R2 = .33

F(6, 431) = 37.64 F(6, 431) = 55.53 F(6, 431) = 15.70 F(6, 442) = 36.77
p = .000 p = .000 p = .000 p = .000

4.1. Implications for clinical practice independent corroboration of children's diet variety (such as from
medical staff, from the child's other caregivers, from direct observation
Clinicians seeing children with feeding problems could use the SEPS of food presentations made to children with their responses recorded).
to identify specific sensory issues the children have in an eating context, Further, although caregivers indicated the feeding problems of their
which may suggest referral for further assessment. For example, if a children in the pre-intervention survey, they were not asked to report
parent or primary care provider (e.g., pediatrician) reports elevated on the severity of each feeding problem. Researchers should examine
scores for the SEPS subscale of Overstuffing, the treating clinician may the relationship between the SEPS and how it relates to clinical severity
conduct further assessment to determine if the overstuffing could be the in future studies.
result of oral motor dysfunction. As part of a comprehensive evaluation, While the goal of the present study was to develop a sensory mea-
the SEPS may also provide information useful in treatment planning. sure specific to eating, future work could also focus on further refine-
For example, clinicians may consider interventions based on systematic ment of the SEPS with the development of cut off scores and inclusion
desensitization for children reported to gag when exposed to novel of a sample of children not identified with eating problems.
foods (Lakkakula, Geaghan, Zanovec, Pierce, & Tuuri, 2018; Paul, Additionally, future research might examine use of the SEPS to evaluate
Williams, Riegel, & Gibbons, 2007; Werthmann et al., 2015). treatment-related changes in sensory processing.

4.2. Study limitations and directions for future research Declarations of interest

While test-retest correlations in the present study were found to be None.

significant (and above r = 0.50) for all six SEPS subscales, three of them
were below the traditionally expected .70 (Gagging, Temperature
Sensitivity, Expulsion). One possible reason may have been that parents Funding support
viewed the initial survey as determining whether their children re-
quired services and so they exaggerated their responses at that point. None.
Alternatively, although the first presentation of the survey was almost
always completed by the child's mother, clinic staff did not monitor that Acknowledgement
the same caregiver completed the second presentation of the survey.
An additional limitation of the present study was its use of caretaker We appreciate the input from the multidisciplinary clinical team at
estimates for children's diet variety in how many of the 80 common St Mary's Hospital for Children's Center for Pediatric Feeding Disorders
foods they accepted or rejected. Future research should include regarding development of the Sensory Eating Problems Scale.

Appendix A

Sensory Eating Problems Scale (SEPS)

(Please circle the option that best describes your child)

1. My child will only eat foods that are warm.

0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
2. My child accepts only one flavor of a certain type of food (e.g., strawberry yogurt).
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
3. My child will eat one food for weeks or months at a time.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
4. My child has a clear dislike for food touching his/her lips.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
5. My child does not like if food touches his/her teeth.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
6. My child is sensitive to food temperature.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
7. My child will only eat foods that are cold.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
8. My child will only eat foods that are room temperature.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always

228
L. Seiverling et al. Appetite 133 (2019) 223–230

9. My child gags when food touches his/her tongue.

0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
10. My child expels food or liquid.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
11. My child uses his or her fingers to take food out of his/her mouth.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
12. My child has difficulty touching food with his/her fingers.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
13. My child has a sensitive gag reflex.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
14. My child will refuse entire food categories (e.g., all fruits, all vegetables).
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
15. My child gets upset when food or liquid touches his/her lips.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
16. My child overstuffs his/her mouth with food.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
17. My child attempts to swallow large pieces of food.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
18. My child attempts to swallow bites of food without chewing.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
19. My child gags when a spoon is placed directly on his/her tongue.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
20. My child spits out food or liquid.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
21. My child avoids mixed textures of foods (e.g., spaghetti and meatballs).
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always
22. My child will gag or vomit at the sight of new food.
0 = Never 1 = Rarely 2 = Sometimes 3 = Often 4 = Always

Scoring:

Food Touch Aversion = (4 + 5 + 12 + 15)/4

Single Food Focus = (2 + 3 + 14 + 21)/4
Gagging = (9 + 13 + 19 + 22)/4
Temperature Sensitivity = (1 + 6 + 7 + 8)/4
Expulsion = (10 + 11 + 20)/3
Overstuffing = (16 + 17 + 18)/3

Appendix B. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.appet.2018.11.008.

References 490–495.
Dunn, W., & Daniels, D. B. (2002). Initial development of the infant/toddler sensory
profile. Journal of Early Intervention, 25, 27–41.
Arbuckle, J. L. (2007). AMOS 16.0 user's guide. Spring House, PA: Amos Development Hendy, H. M., Seiverling, L., Lukens, C. T., & Williams, K. E. (2013). Brief assessment of
Corporation. mealtime behavior in children (BAMBIC): Psychometrics and association with child
Baranek, G. T., David, F. J., Poe, M. D., Stone, W. L., & Watson, L. R. (2006). Sensory characteristics and parent responses. Children's Health Care, 42, 1–14.
experiences questionnaire: Discriminating sensory features in young children with Johnson, S. L., Davies, P. L., Boles, R. E., Gavin, W. J., & Bellows, L. L. (2015). Young
autism, developmental delays, and typical development. The Journal of Child children's food neophobia characteristics and sensory behaviors are related to their
Psychology and Psychiatry and Allied Disciplines, 47, 591–601. food intake. Journal of Nutrition, 145, 2610–2616.
Ben-Sasson, A., Hen, L., Fluss, R., Cermak, S. A., Engel-Yeger, B., & Gal, E. (2009). A meta- Lakkakula, A., Geaghan, J., Zanovec, M., Pierce, S., & Tuuri, G. (2018). Repeated taste
analysis of sensory modulation symptoms in individuals with autism spectrum dis- exposure increases liking for vegetables by low-income elementary school children.
orders. Journal of Autism and Developmental Disorders, 39, 1–11. Appetite, 55(2), 226–231.
Brown, T., Morrison, I. C., & Stagnitti, K. (2010). The convergent validity of two sensory Lukens, C. T., & Linscheid, T. R. (2008). Development and validation of an inventory to
processing scales used with school-age children: Comparing the sensory profile and assess mealtime behavior problems in children with autism. Journal of Autism and
the sensory processing measure. New Zealand Journal of Occupational Therapy, 57, Developmental Disorders, 38, 342–352.
56–65. Nadon, G., Feldman, D. E., Dunn, W., & Gisel, E. (2011). Association of sensory processing
Chistol, L. T., Bandini, L. G., Must, A., Phillips, S., Cermak, S. A., & Curtin, C. (2018). and eating problems in children with autism spectrum disorders. Autism Research and
Sensory sensitivity and food selectivity in children with autism spectrum disorder. Treatment, 2011, 541926. https://doi.org/10.1155/2011/541926.
Journal of Autism and Developmental Disorders, 48(2), 583–591. Paul, C., Williams, K. E., Riegel, K., & Gibbons, B. (2007). Combining repeated taste
Coulthard, H., & Blissett, J. (2009). Fruit and vegetable consumption in children and their exposure and escape prevention: An intervention for the treatment of extreme food
mothers. Moderating effects of child sensory sensitivity. Appetite, 52, 410–415. selectivity. Appetite, 49(3), 708–711.
Davis, A., Bruce, A. S., Khasawneh, R., Schulz, T., Fox, C., & Dunn, W. (2013). Sensory Pliner, P. (1994). Development of measures of food neophobia in children. Appetite, 23,
processing issues in young children presenting to an outpatient feeding clinic: A 147–164.
retrospective chart review. Journal of Pediatric Gastroenterology and Nutrition, 56, Schoen, S. A., Miller, L. J., & Sullivan, J. C. (2014). Measurement in sensory modulation:
156–160. The sensory processing scale assessment. American Journal of Occupational Therapy,
Dunn, W. (1999). Sensory profile user's manual. San Antonio, TX: Psychological 68, 522–530.
Corporation. Seiverling, L., Towle, P., Hendy, H., & Pantelides, J. (2018). Prevalence of feeding pro-
Dunn, W. (2014). Sensory profile 2: User's manual. Bloomington, MN: Pearson. blems in young children with and without autism spectrum disorder: A chart review
Dunn, W., & Brown, C. (1997). Factor analysis on the Sensory Profile from a national study. Journal of Early Intervention, 40(4), 335–346.
sample of children without disabilities. American Journal of Occupational Therapy, 51, Shmaya, Y., Eilat-Adar, S., Leitner, Y., Reif, S., & Gabis, L. V. (2017). Meal time behavior

229
L. Seiverling et al. Appetite 133 (2019) 223–230

difficulties but not nutritional deficiencies correlate with sensory processing in Wiggins, L. D., Robins, D. L., Bakeman, R., & Adamson, L. B. (2009). Brief report: Sensory
children with autism spectrum disorder. Research in Developmental Disabilities, 66, abnormalities as distinguishing symptoms of autism spectrum disorders in young
27–33. children. Journal of Autism and Developmental Disorders, 39, 1087–1091.
Smith, A. M., Roux, S., Naidoo, N. R., & Venter, D. J. (2005). Food choices of tactile Williams, K. E., Field, D. G., & Seiverling, L. (2010). Food refusal in children: A review of
defensive children. Nutrition, 21, 14–19. the literature. Research in Developmental Disabilities, 31, 625–633.
Tomchek, S. D., & Dunn, W. (2007). Sensory processing in children with and without Yi, S. H., Joung, Y. S., Choe, Y. H., Kim, E. H., & Kwon, J. Y. (2015). Sensory processing
autism: A comparative study using the short sensory profile. American Journal of difficulties in toddlers with nonorganic failure-to-thrive and feeding problems.
Occupational Therapy, 61, 190–200. Journal of Pediatric Gastroenterology and Nutrition, 60, 819–824.
Werthmann, J., Jansen, A., Havermans, R., Nederkoorn, C., Kremers, S., & Roefs, A. Zobel-Lachiusa, J., Andrianopoulos, M. V., Mailloux, Z., & Cermak, S. A. (2015). Sensory
(2015). Bits and pieces. Food texture influences food acceptance in young children. differences and mealtime behavior in children with autism. American Journal of
Appetite, 84, 181–187. Occupational Therapy, 69, 1–8.

230