INTRODUCTION

Autism Spectrum Disorder (ASD) is the name of a group of neurodevelopmental disorders with a complex multifactorial etiology. ASD is characterised by persistent abnormal social interactions (e.g., obvious impairment in eye-to-eye gaze, lack of joint attention) and restricted repetitive and stereotyped patterns of behavior, interests, and activities which start in very early life [¹]. ASD becomes apparent before the 24^th month of age and persists into adulthood, causing lifelong disability [²,³]. There are three distinct groups of ASD identified in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR); autistic disorder, Asperger syndrome and Pervasive Developmental Disorder-Not Otherwise Specified (PDD-NOS) [⁴]. This classification is reviewed in the new edition published in 2013 Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [¹]. The incidence of ASD is dramatically increasing year by year and about one child to 59 has been identified with ASD according to estimates from Centers for Disease Control and Prevention’s (CDC) Autism and Developmental Disabilities Monitoring (ADDM) Network [⁵]. ASD is reported to occur in all racial, ethnic, and socioeconomic groups and it is almost five times more common among boys than girls [⁶-⁸].

The majority of children with ASD has been found to present comorbid nutrition problems, especially some mealtime issues [⁹]. Generally, these problems tended to separate to three categories: (1) food selectivity based on type and texture, (2) food refusal, and (3) disruptive mealtime behaviors. Collectively, these reports indicate that children with ASD have more nutritional problems caused by mealtime behaviors and more narrow in the foods they accept [¹⁰]. Besides all these data, it has been reported that 25% of children with ASD have at least one gastrointestinal (GI) symptom (e.g., diarrhea, constipation, abdominal pain, and GI reflux disease). This strongly suggests that the routine treatment should be accompanied by the nutritional therapy in children with ASD [¹¹]. Parents, children’s first role model, have an important influence on the dietary behavior of their children [¹²,¹³]. Therewithal, the employment of women increases household income, and as a result, this positively affects the nutritional status of households, especially mothers. What’s more, studies indicate that there is a strong relationship between the maternal education level and children’s nutritional status [¹⁴]. So, mothers’ empowerment can affect children’s nutritional status because of the fact that mother’s diet quality is an important predictor for the child’s diet quality [¹⁵].

In the studies, the effects of the economic power, educational level and health literacy of the parents on the nutritional status of the children without specific diseases were evaluated. In addition, the relationship between nutritional status of children and some maternal features such as employment, order of birth and birth interval has been studied [¹⁶-²³]. The difference of our study from these studies is that it was conducted on children with ASD. In the literature, there is almost no data on the association between maternal characteristics (demographic features and nutritional status of mothers) and autism-related nutritional problems (meal behavior and nutritional status of children). This paper provides an opinion on whether there was a relationship between the maternal features and common eating disorders such as picky eating, food refusal and inadequate dietary intake in autistic children.

METHODS

This descriptive and cross-sectional study was conducted with autistic children and adolescents (6-19 years) who were registered to special education centers in Eskişehir, Turkey, and with their mothers in January-July 2016. PASS 11, a sample size software, was used for the determination of sample size. In determining the sample size, the mean dietary energy intake reported in a study (1938.7±426.3kcal) evaluating the nutritional status of autistic children was based [²⁴]. For 95% and 80% statistical power, the sample size required 71 and 44 subjects, respectively.

An invitation was sent to the families for “The informative meeting about the study” through the managers of special education centers. Clear explanations about the study design and rights of the participants were provided for the families attended the meetings. A total of 58 mothers agreed to participate in the study with a written informed consent in accordance with Declaration of Helsinki (World Medical Association).

Descriptive and maternal characteristics (gender, age, age at diagnosis, maternal age at birth, maternal education and employment status) and GI symptoms (diarrhea and abdominal pain) of the children were obtained with a general questionnaire. Also, in some evaluations, the children were divided into two groups as school-age children (6-12 years of age) and adolescents (over 12 years), as stated by the World Health Organization (WHO) [²⁵]. For the evaluation of nutritional status and meal behavior, the anthropometric measurements (height and body weight) and 24-hour dietary recall (24HR) were taken, and the Brief Autism Mealtime Behavior Inventory (BAMBI) was applied [²⁶]. Also, for the evaluation of orthorexia nervosa tendency of the mothers, the ORTO-15 questionnaire was applied with face-to-face interview [²⁷]. For the study, ethical approval was obtained from Gazi University Ethics Committee (dated January 15, 2016. Project n. 77082166-604.01.02).

Autistic children and adolescents who were diagnosed with a chronic disease or who have a special diet (i.e. gluten-free diet, casein-free diet, gluten- and casein-free diet, and ketogenic diet) were excluded from the study. The anthropometric measurements were taken by well-trained researchers according to the measurement protocols. The height of the participants was measured in frankfort plane by a stadiometer having 0.1cm sensitivity. The body weight was measured with a calibrated electronic scale having 0.1kg sensitivity [²⁸]. Attention was paid to being with a thin dress and without shoes for the body weight measurement. The Body Mass Index (BMI) was calculated as weight (kg) divided by height squared (m²) (BMI=weight/height²) and was classified according to the BMI cut-off points accepted by WHO [²⁹]. In addition, WHO AnthroPlus software was used to evaluate age- and gender-specific BMI.

Daily energy intake and food consumption was assessed using 24HR. So, the mothers were asked what their children consumed in the last 24 hours. To determine the amounts and portion size of the meals/foods “The Food and Meal Photo Catalog”, a photographic atlas including Turkish foods and meals, was used [³⁰]. Also, BEBIS, a food analysis software, was used to determine the dietary energy and nutrient intakes of the participants [³¹].

The BAMBI, 5-point Likert-type scale, was used to evaluate the meal behavior of the children. The scale, developed by Lukens and Lischeid [²⁶] in order to determine the behavioral and nutritional problems of children with ASD, consists of 18 items and each item includes five different options indicating the frequency of occurrence (1= never/rarely, 2= seldom, 3= occasionally, 4= often, 5= at almost every meal). In the Turkish validity and reliability study of the BAMBI, it was specified that the items 3, 9, 10 and 15 should be evaluated by reverse scoring [³²]. The BAMBI gives three subscores (limited variety, food refusal and disruptive behavior) and a total score. The total point of the items 10, 11, 13, 14, 15, 16, 17, and 18 refers to “limited variety” score (minimum 8, maximum 40 points). The total point of the items 1, 2, 4, 7, and 8 refers to “food refusal” score (minimum 5, maximum 25 points), and also the total point of the remaining items (3, 5, 6, 9, and 12) refers to “disruptive behavior” score (minimum 5, maximum 25 points). Finally, the total point of all these subscores refers to the total BAMBI score (minimum 18, maximum 90 points). Higher BAMBI subscores and total score indicate more negative autism specific meal behaviors.

The ORTO-15, 4-point Likert-type scale, was used to evaluate the orthorexia tendency of the mothers. The scale, developed by Donini et al. [²⁷], consists of 15 items and each item includes four different options indicating the frequency of occurence (1= always, 2= often, 3= sometimes, 4= never). The items 2, 5, 8, and 9 are evaluated by reverse scoring. Also, the items 1 and 13 are evaluated by scoring as “1= never, 2= always, 3= sometimes, and 4= often”. The minimum and maximum scores are 15 and 60 points respectively, and also a score greater than or equal to 40 points refers to “high orthorexia tendency”.

Statistical analysis was performed using Statistical Package for Social Sciences (SPSS) version 22.0 [³³]. Kolmogorov-Smirnov test was used to determine whether the data had normal distribution. The descriptive variables were expressed as “number (percentage)” or “mean±standard deviation” in Table 1. Mann Whitney U test was used to evaluate the statistical differences between the groups (maternal education, maternal employment status, and ORTO-15 groups) in terms of quantitative variables (BAMBI scores, anthropometric measurements, and dietary intake). Spearman’s rho correlation was used to evaluate the relationship between maternal features (maternal age at birth, breastfeeding duration, and ORTO-15 scores) and children’s characteristics (BAMBI scores, anthropometric measurements, and dietary intake). Presence of diarrhea and abdominal pain in the children according to the maternal features were evaluated by Pearson’s Chi-Square test. Boxplots were used to express the BAMBI scores of the mothers with high and low orthorexia tendency. Also, the quantitative variables (BAMBI scores, anthropometric measurements, and dietary intake) were expressed as “median (interquartile range)” in the comparison of the groups (maternal education and employment status). A p-value below 0.05 was considered statistically significant.

RESULTS

Descriptive characteristics of participants are given in Table 1. The majority of the samples were male with the percentage of 81.0%. The mean age was 11.9±3.7 years and also ASD diagnosis age was 3.1±1.3 years. The mean body weight, height, BMI, and BMI Z-score values were 50.8±23.3kg, 149.0±18.9cm, 21.8±6.1kg/m², and 0.86±2.04, respectively. Also, the mean BAMBI scores of the samples were 41.1±8.6 (total), 22.0±5.6 (limited variety), 8.5±3.7 (food refusal), and 10.8±2.7 (disruptive behavior), respectively. The mothers were asked for age of birth, breastfeeding duration, educational level, and employment status. The mean age of birth and breastfeeding duration were determined as 28.3±6.4 years and 12.6±9.5 months, respectively. Besides, 44.8% of mothers had low educational level (<high school graduates) and 81.0% of them were unemployed. The mean ORTO-15 score of the mothers was 37.9±3.4 and it was found that 36.2% of the mothers had high orthorexia tendency.

The mean dietary energy, protein, carbohydrate, fat, dietary fiber, vitamin B6, magnesium and iron intakes were 1685.7±601.9kcal, 57.8±26.4g, 209.0±82.1g, 67.4±26.3g, 21.0±7.8g, 1.2±0.5mg, 233.2±102.0mg, and 9.9±4.0mg, respectively. In addition, as expected, dietary intakes of adolescents were found to be higher than school-age children (not shown in the tables).

The BAMBI scores of the children of mothers with low and high orthorexia tendency were given in Figure 1. There were no statistically significant differences between these two groups in terms of BAMBI total and subscores (p>0.05).

Note: (a) limited variety, (b) food refusal, (c) disruptive behavior and (d) total.

Figure 1 ORTO-15 in its two outcomes, respectively, reffering to the Brief Autism Mealtime Behavior Inventory (BAMBI). 

There were also no statistically significant differences between the children of mothers with low educational level (<high school graduate) and the children of mothers with high educational level (≥high school graduate) in terms of BAMBI scores, body weight, BMI, and BMI-Z values (p>0.05). The BMI values of the children were evaluated separately for school-age children and adolescents according to the mothers’ employment status and educational level, and there were no statistically significant differences between the groups. On the other hand, in relation to dietary intake, the children of mothers with high educational level had significantly higher vitamin B6 intake (p<0.05). However, the differences between two groups were not statistically significant in terms of dietary intake other than vitamin B6 (p>0.05). Also, when two age groups (school age and adolescence) were examined separately in terms of all nutrient intakes, there were no statistically significant differences between the mother groups (not shown in the tables) (p>0.05). The BAMBI (limited variety) scores of the employed mothers’ children were significantly higher than the scores of unemployed mothers’ children (p<0.05). In addition, there were no statistically significant differences between employed and unemployed mothers’ children in terms of the other BAMBI scores (food refusal, disruptive behavior, and total) (p>0.05) (Table 2).

Table 3 shows the correlations between maternal features (maternal age at birth, breastfeeding duration and ORTO-15 scores) and children’s features (BAMBI total and subscores, body weight meusurements, BMI values, and dietary intakes). Maternal age at birth was negatively correlated with dietary energy and protein intakes (p<0.05). Also, there were significantly positive correlations between breastfeeding duration and BAMBI scores (disruptive behavior and total) (p<0.05). However, the ORTO-15 scores of mothers were not significantly correlated with children’s features (p>0.05).

The GI symptoms of children (diarrhea and abdominal pain) are given in Table 4. When the children were grouped according to the maternal features, there were no statistically significant differences between the groups in terms of diarrhea presence (p>0.05). In addition, the mothers were asked if the children had abdominal pain often. It was determined that the percentage of abdominal pain in employed mothers’ children was 54.5% and this percentage was 19.1% in unemployed mothers’ children (p<0.05).

DISCUSSION

In many studies, it was reported that there is a close relationship between eating habits, nutritional status and ASD. The studies focused on maternal nutrition as much as the nutritional status of the children with ASD [³⁴-³⁷]. Mostly, the nutritional status of mothers during pregnancy was emphasized in the studies investigating the effect of maternal nutrition on autism [³⁵-³⁷]. However, there are no studies investigating the effect of maternal eating disorder on the meal behavior of children with ASD in the literature.

In this study, the mean age of children with ASD, 81.0% of whom were male, was 12.0±3.7 years. Also, the mean heights were 150.5±19.1cm and 142.6±17.2cm in males and females, respectively. In Turkey Dietary Guideline (TUBER), the specified median height values were 149cm and 151cm for 12-year-old males and females, respectively [³⁸]. On the other hand, the mean body weight of males and females (51.1±24.1kg and 49.6±19.9kg respectively) was found to be higher than the median body weight values for 12-year-old children specified in TUBER (38.9kg for males and 41.2kg for females). In addition, it was determined that the daily dietary protein, carbohydrate and fiber intakes of children were higher than the adequate intake levels of protein (43.8g for males and 45.8g for females), carbohydrate (130g) and fiber (19g) specified in TUBER. Based on all these results, it is thought that the children with ASD do not have serious physical development retardation.

The BAMBI is a standardized assessment tool developed to evaluate meal behavior problems in children with ASD. In our study, the mean total BAMBI score was 41.1±8.6. Since there was not a control group in the study, children with ASD were not comparable with typically developing children in terms of BAMBI scores. However, in a study conducted by Zobel-Lachiusa et al. [³⁹], the mean total BAMBI score in children with ASD was found to be 44.39±10.83; on the other hand, this mean value was 30.08±7.90 in typically developing children (p<0.05). Because of BAMBI do not have cut-off points, the children with ASD could not been grouped in terms of meal behavior. Even so the BAMBI scores show that the children with ASD have more obvious meal behavior problems than typically developing children.

Studies conducted with mentally healthy children and their mothers show that maternal eating disorders and nutritional status may reflect the eating habits and nutritional status of children [⁴⁰-⁴²]. In this study, the fact that there was no statistically significant difference between the children of mothers with low and high orthorexia tendency in terms of BAMBI scores indicates that the mother’s orthorexia tendency was not influential on meal behavior in children with ASD. Therefore, it can be considered maternal eating disorders do not significantly affect meal behaviors of children with ASD.

Many studies reported that mother’s empowerment and autonomy based on educational and socioeconomic status have a direct positive effect on healthy eating habits and anthropometric indicators in children [⁴³-⁴⁸]. In a study, it was reported that the children whose mothers had salary from employment had a better Weight for Age Z score (WAZ) and Weight for Height Z score (WHZ), and it was also reported maternal educational status was correlated with WHZ of children (r=0.25 p=0.001) [¹²]. In this study, it was found that dietary energy, protein, fat, fiber, vitamin B6, iron and magnesium intakes were higher in autistic children of mothers with high educational level. However, there was a statistically significant difference between the groups only in terms of vitamin B6 intake. Likewise, mother’s educational level does not have an effect on meal behavior and anthropometric measurements of children with ASD. It is thought that maternal demographic factors have a limited effect on the nutritional status in children with ASD.

In previous studies, advanced maternal age was found to be associated with increased risk of ASD [⁴⁹]. According to the results of the 6- and 66-month Taiwan Birth Cohort Study, the risk of being diagnosed with ASD was increased by the maternal age being over 40 years old [⁵⁰]. Also, Rubenstein et al. [⁵¹] reported that advanced maternal age is a risk factor for ASD, but it was not independently associated likely, because it is a consequence of maternal education and other sociodemographic features. Unlike, the mean maternal age at birth was not very high (28.3±6.4 years) in our study. This may be because the sample size in this study is relatively small compared to the other studies. Furthermore, maternal age was significantly associated with daily dietary energy and protein intake in children with ASD. In the literature, no data was found on the relationship between maternal age and nutritional status of children with ASD. However, in a study conducted with typically developing children and their mothers, advanced maternal age was reported to be associated with malnutrition in children [⁵²].

There is convincing evidence that breastfeeding has positive effects on ASD [⁵³,⁵⁴]. In a meta-analysis including the results of seven studies, it was reported that the children with ASD were significantly less likely to have been breastfed than typically developing children (Odds Ratio (OR)=0.61, %95 Confidence Interval (CI)=0.45-0.83, p=0.002), therefore breastfeeding could provide protection against ASD [⁵³]. In a case-control study evaluating the association between breastfeeding and ASD, the absence of breastfeeding when compared to breastfeeding for more than six months was associated with a significant increase in the risk of ASD (OR=2.48, %95 CI=1.42-4.35) [⁵⁴]. Besides, in a study by Soke et al. [⁵⁵], the reported mean breastfeeding duration of the children with ASD was 7.3±7.2 months and that of control children was 9.3±7.2 months. In another study, Boucher et al. [⁵⁶] reported that the mean breastfeeding duration in the children with ASD was 6.9±5.1 months. It is well-known that breastfeeding is a traditional behavior in Turkish society. So, it is nearly impossible to report that the higher breastfeeding durations in our study (12.6±9.5 months) is associated with ASD.

Comorbidities often accompany ASD, and GI dysfunction is among the most frequently cited comorbidities [⁵⁷,⁵⁸]. The prevalence of GI dysfunction in the children with ASD range from 9 to 70% [⁵⁹-⁶³]. A meta-analysis reported that the children with ASD experienced GI symptoms such as diarrhea (OR=3.63; 95% CI=1.82-7.23) and abdominal pain (OR=2.45; 95% CI=1.19-5.07) significantly more than control children [⁶⁴]. In another study, the percentages of children with ASD who experienced diarrhea and abdominal pain in the last three months were 13% and 5.1%, respectively. This rate was 1.6% for both symptoms in typically developing children [⁶⁵]. In this study, the percentages of the children who experienced diarrhea and abdominal pain at least once a month were 37.9% and 25.9%, respectively. In addition, maternal features were not found to be effective on GI symptoms in the children with ASD. It is thought that, GI symptoms were more frequent in this study, because all of the children had severe autism characterized by highly visible lack of communication skills, very limited social interactions, and extreme difficulty coping with unexpected changes.

Orthorexia tendency of the mothers was not associated with meal behavior, anthropometric measurements and dietary intake of the children with ASD in our study. The ORTO-15, developed by Donini et al. [²⁷] based on the Bratman’s test, was used to determine the orthorexia tendency of mothers. Varga et al. [⁶⁶] reported that the validity of orthorexia assessment instruments, including ORTO-15, was not convincing. Because the Bratman’s test includes items that are not unique to orthorexia [⁶⁷]. It is also based on clinical experience and its validity has never been checked. The limitations of the Bratman’s test are making the face-validity of the ORTO-15 doubtful. So, further studies are needed to clarify the appropriate diagnostic methods for orthorexia.

CONCLUSION

Within the scope of the study, we examined the effects of some maternal features including the maternal age at birth, educational level, employment status, breastfeeding duration, and orthorexia tendency on meal behaviors, dietary intakes, anthropometric measurements and some GI symptoms in children with ASD. We concluded that maternal features are not associated with children’s meal behavior in this sample. However, it may be mentioned that the maternal age and educational level have a very limited effect on the dietary intake in children with ASD. More comprehensive longitudinal studies with large sample sizes are needed to investigate if maternal eating disorders and demographic factors were associated with meal behavior and nutritional status in autistic children.