Using Physical Food Transformation to Enhance the Sensory Approval of Children with Autism Spectrum Disorders for Consuming Fruits and Vegetables

Abstract

Introduction:

Children with autistic spectrum disorders (ASD) commonly exhibit aberrant tactile, taste, and smell sensitivity to foods. Food transformation that alters the appearance, texture, and temperature of foods is promising for solving this sensitivity problem.

Objectives:

This study evaluated the effect of physical transformation of fruits and vegetables (FV) to snacks to enhance ASD children's sensory approval for eating them.

Design:

This was a quasiexperimental design study. It lasted for 4 weeks with three episodes of FV exposure to the ASD children per week. The original condition was conducted in the first and fourth week, while intervention was performed in the second and third week.

Settings:

Food exposure took place in schools in a quiet room.

Subjects:

This study recruited 56 ASD children.

Intervention:

Food transformation constituted changing bananas into ice-cream, zucchinis and sweet potatoes into chips, apples and kiwis into popsicles, and carrots into juice.

Outcome Measures:

FV acceptance and habitual consumption were measured at baseline and postintervention.

Results:

ASD children increased their consumption of all FV, but only banana consumption was statistically significant from 55.3 to 78.0 g (p < 0.05). For habitual consumption, parents reported increases in all FV consumption for all three measured fruits and 2 of 3 measures of vegetables (pre vs. post mean for bananas: 2.4 vs. 2.9, apples 2.6 vs. 3.2, kiwis 2.4 vs. 2.9, zucchinis 1.9 vs. 2.5, and sweet potatoes 1.8 vs. 2.4; p < 0.05).

Conclusion:

Physical changes of foods may improve impaired sensory processing of ASD children to promote their FV acceptance. Project Code: RG 55/2019-2020R

Introduction

Selective eating is characterized by a reluctance to consume certain foods and/or an adherence to particular food preferences.¹ Although this eating problem is commonly present in children, the situation is significantly more severe in children with autism spectrum disorders (ASD).² When compared with typically developed children, ASD children exhibit a significantly lower daily intake of fruits and vegetables (FV).³ They also tend to experience more feeding problems,⁴ including those that are behavioral (e.g., neophobia, and food selectivity by type and texture) and/or skill-based (e.g., difficulties chewing and swallowing), with a frequency that is nearly five times greater compared with their peers. Furthermore, in comparison with normal children on indices of body weight, body height, and dietary intake (in terms of energy levels, and carbohydrate and fat consumption) across studies, children with ASD show higher rates of nutritional deficits, raising their risk of malnutrition and long-term adverse health outcomes.^4,5

Studies have indicated that ASD children have significantly lower fruit variety (mean: 6.5 vs. 9.4, p < 0.001) and vegetable variety (mean: 5.9 vs. 10.8, p < 0.001) than children with typical development.⁶ A longitudinal investigation⁷ found that 55.6% and 62.3% of fruit items and vegetable items, respectively, were refused by ASD children (mean age = 6.8) surveyed with a food frequency questionnaire (FFQ). Although these percentages improved when they were older (mean age = 13.2), FV refusal percentages still remained the highest among all FFQ items.⁷ This result does not fulfill the World Health Organization⁸ recommendation of consumption of >400 g of FV per day to reduce the risk of certain noncommunicable diseases and improve overall health. With such limited consumption of FV, they would have lower nutrient intake and resultant slower growth and development.⁹

It has been reported that oral sensory oversensitivity in children with ASD may contribute to their food selectivity, resulting in restrictive eating behaviors.⁶ The problem of selective eating could be related to their behaviors and developmental delay caused by neurodevelopmental disorders. Children with ASD do not perceive and respond to tactile, olfactory, and visual/auditory information in the same way as typically developed children. Their sensory disintegration is possibly linked to their restrictive eating behaviors, because they are also found to have problems in distinguishing olfactory and taste sensations, causing inappropriate responses to sensory stimuli.¹⁰ Indeed, they can become unable to cope with food consumption, which is a complex multisensory experience comprising food appearance, odor, taste, and texture. Some extant literature reported that these children eat few vegetables, do not want different foods physically touching one another, and refuse certain foods on the basis of taste, texture, smell, and even temperature.^7,11 For typically developed children, research demonstrates that touch and taste exposure constitute effective interventions for food selectivity.^12,13 With visual sensitivities, typically developed children's expectations of taste and texture help them to accept new foods. Children with ASD, on the other hand, may overreact to visual stimuli of foods and recall unpleasant memories of taste or texture repertoire that stop them from even trying the foods.¹⁴ In addition, ASD children's sensory defensiveness may limit their exploration through touch, and their more pronounced food preferences may affect their tactile sensitivity, which impedes the effectiveness of repeated exposure to induce food acceptance among them. It was found that this improper sensory processing resulted in fewer than 20 food varieties in their dietary repertoire.^7,14

Built on the current evidence that atypical oral sensory processing is associated with higher rates of food refusal and less consumption of FV in children with ASD, strategies are needed to improve their sensory experiences to enhance diet adequacy and variety. Especially for FV, strategies for changing food sensory characteristics may increase the probability for children with ASD to accept certain foods. In this study, we hypothesized that physical food transformation (i.e., appearance and texture) of meals to snacks without any ingredient changes will enhance ASD children's sensory approval for eating a variety of FV.

Methods

Study setting

The study took place in the schools of the recruited children. Two activity rooms in each school were arranged, one for food sample preparation and the other for children's food exposure and measurement. The food exposure room was a quiet room with good ventilation, with the temperature set in the range of 23°C–25°C, so that the food quality, temperature, appearance, and texture were preserved. The children received food exposure between 2:00 pm and 3:00 pm to prevent hunger or satiety factors from affecting their appetite for FV. The sequences and types of fruits were rotated during the 4 weeks to minimize sequence effects.

Study design

This was a quasiexperimental study of 4 weeks duration with three episodes of FV exposure to the participants per week. The intervention was performed in the second and third week. The effectiveness of the intervention was assessed by comparing the food intake in the first and fourth week.

Participants and sampling

Students studying in three special schools were recruited by convenience sampling. Children studying in these special schools in Hong Kong were those diagnosed with ASD level 2 based on the Diagnostic and Statistical Manual of Mental Disorders: 5th Edition by healthcare professionals. Inclusion criteria were children with ASD 8–15 years of age, and children who could speak and understand Cantonese so as to follow the simple instructions and commands that were included in the study. In consideration of safety, children having a personal history or a family history of food allergy, food intolerance, or chewing or swallowing problems were excluded from the study.

Intervention

Each of the three episodes of food exposure was performed on Monday, Wednesday, and Friday, respectively, of each week. In each week, six foods (three fruits and three vegetables) were used, which included apples, kiwis, bananas, carrots, zucchinis, and sweet potatoes. For each episode, one fruit and one vegetable were prepared and served to the participating children. In the first week and the fourth week, all food samples were served with their original appearance and texture. The food samples were only cut into smaller pieces to facilitate the ASD children to try and eat the food more easily. In the second and third weeks, the food samples were transformed by varying their appearance with different food preparation methods, which might also change the texture and temperature of the foods. Such food transformation was designed to engage children's interest. For example, transformations included turning bananas into ice cream, zucchinis and sweet potatoes into chips, apples and kiwis into popsicles, and carrots into juice. Since the principle of the food transformation was based on the concept of healthy eating, no additional ingredients or condiments were added to the FV. Details concerning the transformation of FV to healthy snacks are provided in Appendix Table A1.

Outcome measurements

Since the participating children were exposed to fruit or vegetable samples in their original forms in the first and fourth week, the consumption in these weeks served as the comparison for the intervention weeks.

FV acceptance

In this study, if a child put food in his or her mouth and swallowed it, this was considered acceptance of the food. The original food samples were prepared in portions of ∼50 g each, with an accuracy of up to 1 g. The weight of the food sample, before and after the exposure for each child, was measured with an electronic balance. From the difference in weight of the food sample, the percentage of its acceptance by each child was calculated after each exposure.

Habitual FV consumption

To better elucidate the habitual FV consumption of the participating children, pre- and postquestionnaires were collected from their parents in the first and final (fourth) week of the study. The frequencies of habitual FV consumption, and those of the six sampled FV, in particular, were assessed by a 5-point Likert-type scale, with 1 representing “never,” and 5 representing “always.”

Brief Autism Mealtime Behavior Inventory

Brief Autism Mealtime Behavior Inventory (BAMBI) is a validated tool developed to measure mealtime behavior of children with ASD, and is a parent-reported questionnaire.¹⁵ Three domains, that is, “food refusal,” “features of autism,” and “limited variety,” were measured by BAMBI. It comprised 18 items, each of which had its own frequency scores, which summed up to the total scores for the questionnaire. The 5-point Likert responses for the 18-items ranged from 1 meaning “never,” to 5 meaning “always.” The descriptions of four items were stated negatively and, therefore, the scores for the responses to these questions were required to be rearranged in reverse order when totaling the scores. Higher total scores indicated more problematic mealtime behavior for the children concerned. The number of items for the domains of food refusal, features of autism, and limited variety were 8, 5, and 5, respectively. BAMBI was tested and determined to have high internal consistency (Cronbach's coefficient α = 0.88), good test/retest reliability (r = 0.87, p < 0.01), and acceptable inter-rater reliability (r = 0.78, p < 0.01).² In this study, BAMBI was translated into traditional Mandarin Chinese for local caregivers to use. Backward translation was performed to ensure that the original meaning of the questionnaire was maintained.

Compliance with ethical standards

No conflicts of interest were involved in this study. Prior approval of the study was obtained from the Ethics Committee of the principal investigator's serving university. All parents/guardians of the participants were informed of the purpose and nature of the study, and their informed written consent was obtained.

Data collection procedures

Before the intervention, the children's demographic data and consumption frequencies of FV in the past 3 months were collected. The BAMBI questionnaire was administered to parents at the baseline to understand their children's mealtime behavior before the intervention.

Food samples were served to children in white plastic bowls to minimize any color effect of food containers on their food sensory processing. The children were also given forks to eat the foods on separate tables. No coercion was applied, and the researcher was present during the exposure sessions to monitor their behavior as a safety precaution. If a participant did not eat or stopped eating his or her food samples for 20 min, the research interpreted that the participant had no intention to try or eat any more of the food, and would measure the weight of the remaining food samples and record the actual amount of acceptance of the food samples.

Statistical analysis

All data were analyzed using SPSS software version 20. The descriptive data were presented as frequencies, percentages, means, and standard deviations. For the inferential data analysis, the paired t-test was used to compare the pre- and postresults from the questionnaire. Analysis of variance was utilized to analyze food acceptance of the ASD children during the 4 weeks of food intervention. For all tests used, the significance level was set at 0.05.

Results

Regarding the demographic characteristics of the 56 participants, their age ranged from 8 to 15, with mean ± standard deviation (SD) age of 10.7 ± 2.2, and 49 (87.5%) were male and 7 (12.5%) were female. Recruited children studying in the special schools in Hong Kong were assessed to have moderate-grade intellectual disability.

FV acceptance after intervention

Table 1 shows that the children's percentage acceptance of all food in its original form in week 4 was higher than in week 1. The mean difference in the pre/post comparison of bananas was significant (p < 0.05), while that for apples, kiwis, zucchinis, carrots, and sweet potatoes was not.

Table 1.

Children's Percentage of Consumption of Six Sampled Foods after Two Sessions of Transformed Sensory Exposure

	Baseline mean % (SD)	Postintervention mean % (SD)	Mean difference	p
Banana	55.3 (48.1)	78.0 (40.1)	22.7	<0.05
Apple	57.0 (49.4)	60.0 (47.3)	2.8	>0.05
Kiwi fruits	49.0 (48.2)	52.1 (48.5)	3.4	>0.05
Zucchini	46.9 (47.8)	49.8 (48.8)	2.9	>0.05
Carrot	30.9 (39.5)	38.3 (47.0)	7.4	>0.05
Sweet potato	37.0 (42.4)	43.5 (46.7)	6.5	>0.05

Paired sample t-tests.

SD, standard deviation.

Habitual FV consumption

Table 2 shows that after the intervention, the children's habitual fruit consumption was significantly more frequent (p < 0.05), and the frequencies of their consumption of bananas, apples, and kiwis increased significantly (p < 0.05). Concerning the children's habitual vegetable consumption, however, there was no significant change in frequency (p > 0.05). The results revealed that the children's consumption of zucchinis and sweet potatoes became significantly more frequent (p < 0.05). Moreover, their frequency of consuming carrots also increased, although not significantly (p > 0.05).

Table 2.

Parents' Reported Children Fruit/Vegetable Consumption in the Past 3 Months

	Baseline mean (SD)	Postintervention mean (SD)	z	p
Fruit consumption	2.8 (1.6)	3.1 (1.8)	−3.1	<0.05
Banana	2.4 (1.2)	2.9 (1.6)	−3.2	<0.05
Apple	2.6 (1.3)	3.2 (1.6)	−3.4	<0.05
Kiwi fruits	2.4 (0.9)	2.9 (1.2)	−2.1	<0.05
Vegetable consumption	3.0 (1.7)	3.0 (1.8)	0	>0.05
Zucchini	1.9 (0.8)	2.5 (1.3)	−2.1	<0.05
Carrot	2.3 (1.0)	2.6 (1.3)	−1.0	>0.05
Sweet potato	1.8 (0.8)	2.4 (1.1)	−2.8	<0.05

A 5-point Likert scale: 1 = never/rarely; 5 = at almost every meal.

SD, standard deviation.

Brief Autism Mealtime Behavior Inventory

Overall, the mean scores of the three feeding behavior domains of food refusal, features of autism, and limited variety were 8.7 (SD = 0.39), 9.7 (SD = 0.38), and 21.8 (SD = 0.97), respectively. As can be seen, the domain of “limited variety” had the highest scores. Within the “food refusal” domain, only a few (n = 8, 8.9%) parents reported that their children cried or screamed more frequently than occasionally during mealtimes or turned their faces or bodies away from food (n = 6, 10.7%). Moreover, one-fifth (n = 12, 20%) of parents reported that their children expelled food eaten more than occasionally, and one-half of them thought that this constituted a problem. In the “features of autism” domain, the most frequently reported feeding problem was their children's aggression during mealtimes (n = 11, 19.6%). Ten (17.8%) parents reported that their children were frequently inflexible about mealtime routines. Concerning the “limited variety” domain, one-fifth of parents (n = 12, 20%) reported that their children were often unwilling to try new foods, and one-third (n = 20, 35.8%) of parents stated that their children disliked certain foods more than occasionally Eighteen (32.2%) parents reported that their children preferred the same foods at each meal more than occasionally, and almost one-half (n = 25, 45%) of parents stated that their children preferred crunchy foods more than occasionally. Twenty-three (41.1%) parents reported that their children insisted on the sameness of foods more than occasionally, and approximately one-third (n = 16, 30%) of the parents reported that their children preferred sweet foods.

Discussion

Prevailing evidence shows that food selectivity is a common eating problem among children with ASD. Their food refusal not only puts their health at greater risk of poorer growth and development, but also constitutes a stressful experience for their families and/or caregivers. In Chan et al.'s study,¹⁶ more than one-third of the surveyed children were unwilling to try new foods, and almost one-half of the surveyed caregivers had a perception that limited food variety affected the children's daily functioning. In our study, the percentage of ASD children who preferred crunchy foods reached 50%, which inspired us to examine ways to create healthy snacks from FV for the intervention. With a crunchy texture as the predominant preference for children with ASD, transforming FV into chips would make FV more appealing to them.

In spite of the difficulties encountered, it is encouraging to learn that the relationship between sensory processing and eating problems was established in a few previous studies.^10,11 Research focusing on sensory modalities related to improving sensory sensitivity, however, was rare. In fact, sight exposure occurs before taste exposure during food presentation, and appealing presentations that attract and motivate typically developed children to make their first attempt could be ineffective for most children with ASD. We believe that transforming food to familiar snacks that are popular among both typically developed children and children with ASD assists to remove taste barriers for continuous and follow-up taste exposure for children with ASD.

This study hypothesized possible pathways for FV transformation to improve consumption. Repeated exposure that evidently facilitated typically developed children to accept new foods was determined to be ineffective for a group of ASD kindergarten children to increase regular vegetable consumption.¹⁷ Since children with ASD have lower tactile sensitivity as compared with typically developed children, this model hypothesized that FV, when transformed to snacks, would exhibit spontaneous changes in appearance, texture, and temperature, which would result in alterations in tactile, taste, and smell sensitivity due to neurological changes. These sensory processing signal changes, once occurred, would possibly enhance ASD children's acceptance and future consumption of original forms of FV.

This study bridges the gap in elucidating the linkage between the physical characteristics of foods after transformation and the sensory processing of children with ASD. If appearance, texture, and temperature of food could affect ASD children's consumption, this change in physical food characteristics may possibly change their sensory processing by altering their tactile and taste/smell sensitivity. Concretely, this study found that the acceptance of all sampled FV increased, although only that of bananas was significant. The results supported the hypothesis that transforming FV to appealing snacks facilitates ASD children's taste attempts by reducing the barriers of tactile and taste/smell sensitivity.

This study also revealed that habitual fruit consumption of children with ASD improved postintervention, as reported by their parents. After the intervention, habitual consumption of bananas, apples, and kiwis significantly increased. Although such positive results were not found in the habitual consumption of vegetables, the habitual consumption of zucchinis and sweet potatoes did rise significantly. The combined findings supported the hypothesis that the transformation of bananas, apples, kiwis, zucchinis, and sweet potatoes into popular snacks increased ASD children's consumption of them. If this intervention is generalized to other food items, the variety of foods for children with ASD could gradually expand in the future.

Limitation

The improvements in food consumption were encouraging, and the lack of statistical significance may have been due to the small sample size. This research only used three fruits and three vegetables to design children's snacks based on the texture of each type of food. In addition, the skills and home appliances used in preparing the snacks were simple, but the same practices may not be easily generalized to all types of foods. Innovative and creative ideas for making snacks based on FV should be individualized, and thus the contribution of food technologists and food scientists is needed to improve and expand the ASD food menu. Another limitation was that this study did not include a control group for comparison. To address this limitation, future research should incorporate additional food items in the intervention and a control group for more robust evaluations.

Conclusion

Food transformation of FV into appealing snacks increases not only ASD children's acceptance, but also their habitual consumption of these foods. Physical changes of foods may improve the impaired sensory processing of children with ASD to promote their FV acceptance, thus enriching their food variety intake.

Author Contributions

L.M.Y.C. designed the study, wrote the protocol, conducted the literature searches, and drafted the article. Q.P.S.L. performed the statistical analysis and interpretation. S.S.M.F. reviewed the article and provided supervision. All authors have approved of the final article.

Footnotes

Acknowledgments

The authors express their sincere gratitude to the three special schools for providing invaluable support in the data collection.

Author Disclosure Statement

All authors declare that there are no conflicts of interest and no commercial associations regarding the submitted article.

Funding Information

No funding sources were involved in this research project.

Appendix Table A1.

Process in Preparing Original and Transformed Fruits and Vegetables

Original food	Preparation of food in OF group	Serving temperature	Transformed appearance	Preparation of food in TF group	Serving temperature
Apple	Raw apples were peeled and cut into large slices	20°C–23°C	Popsicle	Raw apple slices were peeled to remove skin. They were cut and blended with blender until the apple slices become smoothie. Smoothie was filled into popsicle molds and covered with sticks. They were stored in a refrigerator at −20°C until the smoothie was frozen	−20°C to 0°C
Orange	Raw oranges were peeled and cut into large slices	20°C–23°C	Popsicle	Raw orange were peeled to remove skin. Slices were blended and then filtered to extract the orange flesh and juice. They were filled into popsicle molds and covered with sticks. They were stored in refrigerator at −20°C until the flesh and juice were frozen	−20°C to 0°C
Banana	Banana was peeled and cut into shorter pieces of 4 cm length each	20°C–23°C	Ice cream	Raw banana was peeled to remove skin. The banana was cut into smaller units which were put in the refrigerator at −20°C. They were frozen for 12 h. Before administration to the students, the frozen banana was put into a pressor, and the banana was pressed slim and flew continuously out of the machine. They were swirled into a transparent container	−20°C to 0°C
Sweet potato	Sweet potatoes were cut into smaller size and steamed at 100°C for 20 min. Sweet potatoes were cooled down at room temperature	20°C–23°C	Chips	Sweet potatoes were steamed at 100°C for 20 min. They were cut into 0.5-cm-thick pieces. They were then put in a food dehydrator with temperature set at 58°C for 15 h. The chips were then cooled down at room temperature	20°C to 23°C
Pumpkin	Pumpkins were cut into smaller sizes and were steamed at 100°C for 20 min. Pumpkins were cooled down at room temperature	20°C–23°C	Chips	Pumpkins were steamed at 100°C for 20 min. They were cut into 0.5 = cm-thick pieces They were then put in a food dehydrator with temperature set at 58°C for 15 h. The chips were then cooled down at room temperature	20°C to 23°C
Carrot	Raw carrot was cut into 0.5-cm-thick and 8-cm-long pieces	20°C–23°C	Juice	Carrot were grated into small pieces. They were then put into a masticating juicer. Carrot juice was pressed in a transparent cup for serving	20°C to 23°C

OF, original food; TF, transformal food.

References

Jacobi

, Agras

, Bryson

, Hammer

. Behavioral validation, precursors, and concomitants of picky eating in childhood. J Am Acad Child Psy, 2003; 42:76–84.

Schreck

, Williams

, Smith

. A comparison of eating behaviours between children with and without autism. J Autism Dev Disord, 2004; 34:433–438.

Levy

, Hyman

. Complementary and alternative medicine treatments for children with autism spectrum disorders. Child Adolesc Psychiatr Clin N Am, 2008; 17:803–820.

Ledford

, Gast

. Feeding problems in children with autism spectrum disorders a review. Focus Autism Other Dev Disabil, 2006; 21:153–166.

Sharp

, Berry

, McCracken

, et al. Feeding problems and nutrient intake in children with autism spectrum disorders: A meta-analysis and comprehensive review of the literature. J Autism Dev Disord, 2013; 43:2159–2173.

Chistol

, Bandini

, Must

, et al. Sensory sensitivity and food selectivity in children with Autism Spectrum Disorder. J Autism Dev Disord, 2018; 48:583–591.

Bandini

, Curtin

, Phillips

, et al. Changes in food selectivity in children with autism spectrum disorder. J Autism Dev Disord, 2017; 47:439–446.

World Health Organization. Increasing fruit and vegetable consumption to reduce the risk of noncommunicable diseases. January 10, 2017. Online document at: www.who.int/elena/titles/fruit_vegetables_ncds/en, accessed July 31, 2020 .

Taylor

, Northstone

, Wernimont

, Emmett

. Macro- and micronutrient intakes in picky eaters: A cause for concern?. Am J Clin Nutr, 2016; 104:1647–1656.

10.

Bennetto

, Kuschner

, Hyman

. Olfaction and taste processing in autism. Biol Psychiatry, 2007; 62:1015–1021.

11.

Smith

, Roux

, Naidoo

, Venter

DJL

. Food choices of tactile defensive children. Nutrition, 2005; 21:14–19.

12.

Lakkakula

, Geaghan

, Zanovec

, et al. Repeated taste exposure increases liking for vegetables by low-income elementary school children. Appetite, 2010; 55:226–231.

13.

Chung

LMY

, Fong

SSM

. Repeated exposure to and appearance alteration of fruits and vegetables to increase their appeal to and consumption by school-age children: A pilot study. Health Psychol Open, 2018; 5:1–10.

14.

Nadon

, Feldman

, Dunn

, Gisel

. Mealtime problems in children with Autism Spectrum Disorder and their typically developing siblings: A comparison study. Autism, 2011; 15:98–113.

15.

Lukens

, Linscheid

. Development and validation of an inventory to assess mealtime behavior problems in children with autism. J Autism Dev Disord, 2008; 38:342–352.

16.

Chan

DFY

, Yu

CCW

, So

, et al. Mealtime behavioral problems in Hong Kong Chinese preschoolers with autism spectrum disorder. J Abnorm Psychol, 2016; S1:004.

17.

Kim

, Chung

, Jung

. Effects of repeated food exposure on increasing vegetable consumption in preschool children with autism spectrum disorder. Res Autism Spectr Disord, 2018; 47:26–35.