Nutritional Status of Individuals with Autism Spectrum Disorders: Do We Know Enough?^,

Abstract

The incidence of individuals with autism spectrum disorders (ASDs) is on the rise; therefore, well-timed screening is important. Given that this is a nutritionally vulnerable population, it is imperative to conduct a detailed nutritional assessment so that timely and intensive interventions can be recommended. This review article summarizes the research, focusing on the nutritional status of individuals with ASDs based on their anthropometric measurements, biomarkers, and dietary assessments. Research examining anthropometric measurements reveals an abnormally accelerated rate of growth among children with autism but shows inconsistent findings on the prevalence of overweight/obesity in comparison with typically growing children. Although dysregulated amino acid metabolism, increased homocysteine, and decreased folate, vitamins B-6 and B-12, and vitamin D concentrations have been proposed as possible biomarkers for an early diagnosis of ASDs, research investigating their association with age, gender, severity, and other comorbid psychiatric/nonpsychiatric disorders is lacking. There is consensus that children with autism have selective eating patterns, food neophobia, limited food repertoire, and sensory issues. Although inadequate micronutrient but adequate macronutrient intakes are increasingly reported, there are inconsistent results about the extent and type of nutrient deficiencies. Identification and development of nutritional assessment indicators that serve as early warning signs during routine practice beginning at birth and extending throughout the child’s growth are necessary. With this population aging, there is also a dire need to study the adult population. A more vigorous role by nutrition professionals is warranted because management of potential comorbidities and contributory factors may be particularly problematic.

Introduction

Autism or autistic spectrum disorder (ASD)³ is a wide-spectrum neurodevelopmental disorder encompassing impairments in social interaction, language, communication, and imaginative play. It also includes restricted, repetitive, and stereotyped patterns of behaviors, activities, and interests. Recent estimates are that 1 in 88 children suffer from ASDs, indicating a 78% increase from 2002, and are almost 5 times more common among boys (1 in 54) than girls (1 in 252) (1). Multiple genetic, environmental, and immunologic factors play a role in its pathogenesis. Increased awareness and improved diagnostic criteria may also be attributed to this jump rather than new environmental influences.

Research highlights that individuals with ASDs are nutritionally vulnerable because they exhibit a selective or picky eating pattern and sensory sensitivity that predisposes them to restricted intakes (2–8). This is further attenuated by dietary restrictions (such as gluten-free or casein-free diets) imposed by parents/caretakers as a therapeutic tool with the goal of improving behavior and/or gastrointestinal symptoms (9). Whether these individuals present malnutrition similarly or differently or more frequently than typical individuals is inconclusive. Because an individual’s nutritional status is a result of complex mechanisms and interactions, a detailed nutritional assessment by a registered dietitian is essential for developing guidelines specific for individuals with ASDs. The purpose of this review article, therefore, was to summarize the pertinent information regarding the nutritional status of this complex behavioral disorder.

Methods

Nutritional assessment by a registered dietitian often includes evaluation across 5 different domains popularly known as the anthropometry, biochemical, clinical, dietary, and environmental approach. Although anthropometry allows the measurement of body size, composition, weight, and proportions, the biochemical assessment involves measuring nutritional markers and indicators of organ function in biological specimens (blood, urine, feces, hair, nails, and tissue samples). The nutrition-focused clinical (also known as physical) exam assesses the patient for signs and symptoms consistent with malnutrition or specific nutrient deficiencies through inspection, palpation, percussion, and auscultation. The dietary assessment identifies the patient’s usual pattern of intake, food preferences (including ethnic, cultural, and religious influences), and use of alcohol, complementary and alternative medicine, and vitamin/mineral/herbal supplements (10). Finally, environmental factors such as socioeconomic status, social support systems, lifestyle, and social interactions affect nutritional status and are integral to nutritional assessment.

For the purpose of this review, research focusing on anthropometric, biochemical, and dietary assessment was studied. A literature search was performed using the electronic database PubMed to identify relevant research studies published in English during the past decade (since 2000). The search term combinations used were a population term (e.g., autism and autism spectrum disorder) and either an anthropometric term [e.g., height, weight, head circumference (HC), obesity, and overweight], a biochemical term (e.g., biochemical status, blood concentrations, and biomarkers), or a food-related term (e.g., food, feeding, diet, diet therapy, eating, nutrient, and nutrition). The lists of the articles obtained were manually searched for additional references, and all of the relevant studies have been included in this review under 3 broad categories: anthropometric, biochemical, and dietary assessment.

Results/Current Status of Knowledge

Anthropometric assessment

Researchers have studied the anthropometric measurements, mainly height, weight, BMI, and HC, of children with autism and compared them either with typically developing controls or with a reference healthy population or those with other psychiatric conditions (Table 1). Findings on the prevalence of obesity have been quite inconsistent (2, 11), although some studies (12, 13) reported a somewhat similar prevalence of overweight (19% compared with 16%)/obesity (30.4% compared with 23.6%) between autistic individuals and individuals without autism; other studies (14–17) reported a higher prevalence of overweight among children with ASDs (13–20%). With increasing age, children with autism have shown a trend toward an increasing prevalence of underweight (17, 18) or overweight [2- to 5-y-old children compared with 6- to 11-y-old children: 14.2% compared with 50% (13) and 31.8% compared with 37.9%, respectively (15)] and were twice as likely to be obese as children without autism (12, 18). Nonetheless, overweight and obesity are as important a concern in children with autism as in the general population (12, 13), and their unusual dietary patterns and decreased access to opportunities for physical activity may be contributory factors.

TABLE 1.

A summary of studies highlighting the anthropometrics of individuals with ASDs¹

Study	Population	Group	Sample size, n	Age,2 y	Finding (ASD vs. control)
Retrospective chart review
Egan et al. (14)	United States	Children with ASDs	169	3.89 ± 0.1	Prevalence of overweight: 17.2% vs. 12.5%
		Children with Asperger disorder/PDD-NOS	104		Prevalence of obesity: 21.9% vs. 10.6%
Grandgeorge et al. (30)	France	Children with ASDs	422	Birth	HC, body length, and weight: similar at birth
		Typically developing children	153		Macrocephaly*
Fukumoto et al. (26)	Japan	Infants with autism	85	Birth–1	HC (boys): similar at birth, 6 mo,* and 12 mo**
		Control children	14,115
Mraz et al. (25)	United States	Children with ASDs	35	Birth–2	HC: birth–2 wk** and 10–14 mo*
		Healthy infants	37		Height/weight: 1–2 mo*
		National normative data (CDC)
Hazlett et al. (23)	United States	Children with autism	113	Birth–3	HC: normal at birth and 12 mo*
		Control children	189
Courchesne et al. (21)	United States	Children with ASDs	48	Birth–1	HC: birth** and 6–14 mo* (increased from 25th to 84th percentile)
Cross-sectional study
Curtin et al. (12)	United States	Children from NSCH (2003–2004)	85,272	3–17	Prevalence of obesity: 30.4% vs. 23.6%.
Bicer and Alsaffar (2)	Turkey	Children with autism	164	4–18	Prevalence of overweight/obesity: 58.5%
Suren et al. (20)	Norway	Children with ASDs	376	Birth–3	HC, weight, and length: similar at birth
		Population cohort	106,082		By 12 mo of age: boys, HC similar but greater variability, longer (by 1.1 cm), and heavier (by 300 g); girls, reduced HC (by 0.5 cm), similar length, and lighter (by 150–350 g)
Hyman et al. (17)	United States	Children with ASDs	3673	2–11	2- to 5-y olds: overweight/obese*
		Controls (NHANES, 2007–2008)	559		6- to 11-y olds: underweight*
Xiong et al. (15)	China	Children with autism	429	2–11	2- to 5-y olds vs. 6- to 11-y olds: at risk of obesity (31.8%/37.9%) and overweight (17.0%/21.8%)
Meta-analysis
Redcay and Courchesne (27)	United States			2.4–46	Brain size at birth is 13% smaller and at 12 mo is 10% greater; stabilized reaching normal range by adulthood

¹Values are ranges or means ± SDs for age. *Significantly greater, P < 0.05; **significantly less, P < 0.05. ASD, autism spectrum disorder; HC, head circumference; NSCH, National Survey of Children‘s Health; PDD-NOS, pervasive developmental disorder not otherwise specified.

²Age at the time of anthropometric measurements.

³BMI data on 362 individuals.

Studies consistently show that children with ASDs have a sex-specific atypical head growth pattern; at birth, they appear to have a normal or even slightly decreased HC, followed by an increase in the rate of growth of HC up until 12 mo of age before the typical age of clinical identification (19–28). Thereafter, relative to growth during the first year, there is a rapid deceleration in HC between 12 and 24 mo of age, such that it is normal compared with the controls (24, 27, 29). HC is relatively increased in comparison with height (24, 26, 30, 31); however, others report that HC is normal (25) or smaller (32) in relation to height. Many (19, 21, 28, 33) have proposed that this atypical and abnormally accelerated growth in HC is due to dysregulation of growth in general rather than of neuronal growth in the brain, and this may serve as an early indicator of vulnerability to autism in children and among their infant siblings with and without a history of autistic regression.

Biochemical assessment

Analysis of biological specimens for nutrients and nutrient-related substances is imperative in the diagnosis of disease(s) before clinical symptoms are apparent. The knowledge obtained from such analysis helps evaluate treatment plans and monitor effectiveness. Findings on vitamins, minerals, amino acids, FAs, and metabolic markers of children with autism are highlighted in Tables 2–4.

TABLE 2.

Studies summarizing vitamin and mineral concentrations in individuals with ASDs¹

Case-controlled study	Population	Group	Sample size, n	Age, y	Finding (ASD vs. control)
Vitamins
Adams et al. (34)	United States	Autism	55	5–16	Vitamin C (plasma)*
		Neurotypical	44		Biotin (WB), pantothenic acid (WB), vitamin E (serum), and total carotenoids (plasma)**
					Vitamin B-6 (RBC)*: (ASDs = 3 SD of controls)
					Lipoic acid (plasma) and free choline (RBC): similar
					Total choline (RBC), FIGLU (urine), and N-methylnicotinamide (urine)*
Ali et al. (35)	Oman	Children with ASDs	40	3–5	Folate (serum)**
		Controls	40		Vitamin B-12 (serum)**
Al-Gadani et al. (39)	Saudi Arabia	Autistic children	30	3–15	Vitamin E (plasma)** and vitamin c (plasma): similar
		Healthy children	30
Adams et al. (40)	United States	Autism	35	3–9	Vitamin B-6 (plasma): 77% of ASD group had >2 SD above the median value of controls
		Controls	11
Minerals
Russo et al. (46)	United States	Individuals with autism	102	Birth to >60	Copper (plasma)*: 108.9 μg/dL vs. 86.5 μg/dL
		Neurotypical controls	18		Copper/zinc (plasma)*: 1.41 vs. 1.19
					Zinc (plasma): 80.5 μg/dL vs. 84.7 μg/dL
Adams et al. (34)	United States	Children with ASDs	55	5–16	Lithium (WB), calcium (RBC), and magnesium (serum, WB)**
		Neurotypical controls	44		Copper (WB, RBC), iron, potassium, phosphorus, and boron (RBC)*
Lakshmi Priya and Geetha (45)	India	Children with ASDs	45	4–12	Zinc (hair, nails): variable
		Healthy children	50		Copper, mercury, and lead (hair, nails)*
					Magnesium and selenium (hair, nails)**
Adams et al. (44)	United States	Children with ASDs	51	3–15	Iodine (hair)**: by 45%
		Mothers of children with ASDs	29		Lithium (hair)**: in those aged 3–6 y, and their mothers
		Neurotypical children	40		Potassium (hair)**: in those with low muscle tone
		Mothers of neurotypical children	25		Zinc (hair)*: in those with low muscle tone
					Chromium (hair)**: by 38% in those with pica

¹Values are ranges for age. *Significantly greater, P < 0.05; **significantly less, P < 0.05. ASD, autism spectrum disorder; FIGLU, formiminoglutamic acid; WB, whole blood.

TABLE 4.

Research studies on FA concentrations and metabolic markers in individuals with ASDs¹

Case-control study	Population	Group	Sample size, n	Age, y	Finding (ASD vs. control)
FA concentrations
El-Ansary et al. (59)	Saudi Arabia	Children with autism	25	3–15	Linoleic acid/arachidonic acid ratio, α-linolenic acid/DHA ratio, and EPA/arachidonic acid ratio (plasma)*
		Healthy children	16		EPA/DHA ratio: similar
					Arachidonic acid/DHA ratio (plasma)**
Vancassel et al. (60)	France	Children with autism	15	3–17	Total (n–6) PUFAs, plasma linoleic acid, and α-linolenic acid: similar
		Mentally retarded children	18	1–19	Total (n–3) PUFAs (plasma)**
Metabolic markers
Adams et al. (34)	United States	Children with ASDs	55	5–16	Sulfate (free, total), SAM, SAM/SAH ratio, and GSH (plasma)**
		Neurotypical controls	44		Plasma ATP, NADH (RBCs), and NADPH (RBCs)**
					Uridine, adenosine, GSSG, GSSG:GSH ratio, and nitrotyrosine (plasma)*
Geier et al. (54)	United States	Children with ASDs	38	2–16	GSH (plasma), and sulfate (total, free; plasma)**
		Neurotypical controls	>25		GSSG (plasma)*
Pasca et al. (37)	Romania	Children with autism	12	8.29 ± 2.76	Erythrocytes GSH-Px: similar; negatively correlated with total plasma homocysteine concentrations
		Controls	9	8.33 ± 1.82

¹Values are ranges or means ± SDs for age. *Significantly greater, P < 0.05; **significantly less, P < 0.05. ASD, autism spectrum disorders; GSH, reduced glutathione; GSH-Px, glutathione peroxidase; GSSG, oxidized glutathione; SAH, S-adenosylhomocysteine; SAM, S-adenosylmethionine.

Vitamins.

Studies report decreased concentrations likely below the reference range of pantothenic acid [whole blood (WB)] (34), biotin (WB) (34), folate (serum) (35, 36), vitamin B-12 (serum and plasma) (35–37), vitamin D (25-hydroxyvitamin D and 1,25-dihydroxyvitamin D; serum) (38), and vitamin E (serum and plasma) (34, 39) in children with autism.

Elevated and unusually broad vitamin B-6 (RBC and plasma) concentrations have been reported. Adams et al. (34, 40) found that autistic children from the United States had vitamin B-6 concentrations [measured as RBC pyridoxal 5-phosphate (P5P) and total vitamin B-6 in plasma] >2 times SDs of their nonsibling neurotypical controls, and neither group was taking vitamin/mineral supplements. These findings indicate that children with autism may have a low activity of pyridoxal kinase that converts pyridoxal and pyridoxine into the active form P5P. This would ultimately result in high amounts of pyridoxal and, thus, total vitamin B-6, as well as low amounts of P5P, which is the active cofactor for several enzymatic reactions, including the formation of many key neurotransmitters. Broad distribution of RBC P5P also suggests that there is a subset of children who need more vitamin B-6 and a subset who have high concentrations of vitamin B-6. This explains the benefits of high-dose vitamin B-6 supplementation in individuals with autism with low RBC P5P (34). Inadequate intake of vitamin B-6 by autistic children in China was observed by Xia et al. (41), who also suggested that supplementation with vitamin B-6 can have positive benefits for behavior because it is required for brain development and function.

There are inconsistent findings on the amounts of vitamin C (plasma) and vitamin A/total carotenes/β-carotene (plasma) (34, 39, 42) as well as free/total choline (RBCs and plasma) (34, 43). Among children with ASD, 69–93% had choline intake less than adequate, and 18–30% had betaine intake <3.5 mg/kg. Lower choline and betaine intakes were also correlated with lower plasma choline and betaine concentrations (43). This suggests that the choline-betaine-homocysteine pathway for Met synthesis may be compromised in children with ASDs. Al-Farsi et al. (36) also showed that compared with controls, children with ASDs had significantly lower dietary intake of folate and vitamin B-12 that was consistently low in their serum as well. They also found a decrease in the activity of Met synthase, highlighting the fact that dietary and serum deficits of folate and vitamin B-12 in children with ASDs have functional consequences.

The precise reason for variations in vitamin concentrations is unclear, but may be attributed to differences in the measurement status (fasting compared with nonfasting) and also the country of study that affects the ethnicity, diet, and other factors.

Minerals.

Concentrations of minerals in WB, serum, plasma, RBCs, hair, and nails have been studied in children with ASDs (Table 2). Compared with controls and/or reference ranges of healthy children, children with autism have lower concentrations of lithium (WB and hair) (34, 44), calcium (RBCs and serum) (34, 38), magnesium (serum, hair, and nail) (34, 45), iodine (hair) (44), chromium (hair) (44), and selenium (hair and nails) (45).

Low concentrations of lithium in children with autism are particularly interesting because its deficiency has been linked to a wide range of psychiatric disorders including autism. The decreased activity of enzymes involved in growth factor signaling pathways and regulation of neurotransmitter levels caused by low concentrations of lithium suggest that low-level lithium supplementation may be beneficial for mood stabilization in this group (34).

Copper (WB, plasma, serum, RBCs, hair, and nails) (34, 45–47), phosphorus (RBCs) (34), boron (RBCs) (34), mercury (hair and nails) (45), and lead (hair and nails) (45) concentrations have been found to be elevated in children with autism. However, there are inconsistent findings on the amounts of potassium (RBCs and hair) (34, 44) and zinc (plasma, hair, and nails) (44–47), possibly because of differences in culture, diet, and geographic area.

The prevalence of iron deficiency and anemia in subjects with autism is reported between 24% and 32% and 8% and 16%, respectively (48–51). Because iron deficiency, with or without anemia, results in impaired cognition and developmental defects, iron deficiency in children with autism could further compromise their communication and behavioral impairments.

Amino acids.

Essential and nonessential amino acids studied in individuals with ASDs are shown in Table 3. Compared with controls, children with autism have increased concentrations of β-aminoisobutyrate or 3-aminoisobutyric acid, glutamate, Ser (plasma) (34), and Arg (plasma) (52), as well as Glu, Asn, Ala, and Lys (plasma) (53). On the other hand, they have lower concentrations of Met (serum) (36), Cys (plasma) (54), Gln (plasma) (53), and Trp (plasma) (34, 55).

TABLE 3.

Studies summarizing amino acid concentrations in individuals with ASDs¹

Study	Population	Group	Sample size, n	Age, y	Finding (ASD vs. control)
Case-control study
Adams et al. (34)	United States	Children with ASDs	55	5–16	Glutamate, β-aminoisobutyrate, and Ser (plasma)*
		Neurotypical controls	44		Trp, Phe, Tyr, and taurine (plasma)**
Kałużna-Czaplińska et al. (56)	Poland	Children with autism	34	4–11	Homocysteine (urine)*
		Neurotypical healthy children	21
Geier et al. (54)	United States	Children with ASDs	38	2–16	Cys (plasma)**
		Neurotypical controls	>25		Taurine (plasma)**
Aldred et al. (53)	United Kingdom	Patients with autism/Asperger disorder	23	4–29	Glu, Lys, Phe, Asn, Tyr, and Ala (plasma)*
		Parents of ASD patients	32	24–58	Gln (plasma)**
		Siblings of ASD patients	13	9–30	Other amino acids: normal
		Controls		4–16
Retrospective chart review
Arnold et al. (55)	United States	Children with autism on unrestricted diet	26	<5	Met (plasma)**: for those on unrestricted diets
		Children with autism on gluten- and casein-restricted diet	10		Val, Ile, Leu, Tyr, and Lys (plasma)**: for those on restricted diets
		Control children having developmental delay but not autism	24

¹Values are ranges for age. *Significantly greater, P < 0.05; **significantly less, P < 0.05. ASD, autism spectrum disorder.

There are inconclusive findings about the amounts of taurine (plasma) (34, 52, 54), Phe (plasma) (34, 53), Tyr (plasma) (34, 53, 55), and homocysteine (serum, plasma, and urine) (35–37, 56).

Lower concentrations of plasma Tyr and Trp in children with autism may likely impair serotonin synthesis, which plays an important role in neurogenesis and neurotransmission (34, 55). Decreased protein intake as evidenced by Zimmer et al. (57) and/or impaired digestion of protein into amino acids may be the contributory factors to decreased Trp concentrations. Elevated plasma glutamate may be linked to the behavioral issues associated with autism and also increased vitamin B-6 requirements for conversion of glutamate to Gln. Some children with autism (those with low Trp and Phe) may benefit from increased protein intake, use of digestive enzymes containing proteases, and/or Trp or 5-hydroxytrptophan and Phe supplements (34). Small sample size, absence of neurotypical controls, age differences between the autism and control groups, nonfasting biochemical estimations, and lack of sensitive methods of estimations (particularly for homocysteine) may be the contributory factors to inconsistent and contradictory results.

FAs.

Studies reported significantly higher plasma EPA (20:6n–3):arachidonic acid (20:4n–6), linoleic acid (18:2n–6):arachidonic acid, α-linolenic acid:DHA (22:6n–3), and n–3 to n–6 FA ratios and lower n–3 FAs, particularly DHA and the arachidonic acid:DHA ratio, in children with ASDs (39, 58–60) (Table 4).

Metabolic markers.

Significantly higher concentrations of a number of markers of oxidative stress, namely, oxidized glutathione (plasma) (34, 54), the oxidized glutathione:reduced glutathione (GSH) ratio (plasma) (34), nitrotyrosine (plasma) (34), superoxide dismutase (RBC) (39), and malonyldialdehyde (plasma) (61), have been reported in children with ASDs (Table 4). Although plasma catalase activity was similar in the austistic and control groups (39), the GSH peroxidase activity in plasma and erythrocytes was either higher (39) or similar (37). Al-Gadani et al. (39) reported that compared with age-matched controls, autistic patients had 54.4% increased concentrations of lipid peroxides and, thus, increased oxidative stress. This could lead to free radical formation that extracts hydrogen atoms from lipoproteins and membrane phospholipids, specifically PUFAs, causing lipid peroxidation that could be correlated to mitochondrial dysfunction in autistic patients. Lower plasma concentrations of antioxidants/antioxidant proteins such as glutathione (39), GSH (plasma) (34, 54), transferrin (serum) (61), and ceruloplasmin (serum) (61) suggest the possibility of antioxidant supplementation for early intervention with this pediatric population. Significantly lower concentrations of S-adenosylmethionine (SAM; plasma) (34) SAM:S-adenosylhomocysteine ratio (SAM:SAH; plasma) (34), and total and free sulfate (plasma) (34, 54) and a lower as well as significantly higher concentrations of uridine and adenosine (plasma) (34), demonstrate impaired capacity for methylation and increased oxidative stress in children with autism. Abnormal iron and copper metabolism (61) and decreased detoxification capacity, particularly of mercury (54), have also been reported in children with autism.

Dietary assessment

A qualitative and quantitative assessment of dietary intake helps determine adequacies and inadequacies in nutrient intake, which is a marker of the nutritional status of an individual. Based on the existing literature, for the purpose of this review, studies on dietary intake have been categorized as those focusing on a child’s eating behavior that undoubtedly affects their dietary intakes (Table 5) and those highlighting actual nutrient intakes (Table 6).

TABLE 5.

Studies on the eating behavior of individuals with ASDs that influence their nutrient intake¹

Study	Population	Group	Sample Size, n	Age, y	Method of data collection	Finding (ASD vs. control)
Bandini et al. (7)	United States	Children with ASDs	53	3–11	Modified YAQ FFQ	Food refusal*: 41.7% vs. 18.9% of foods offered, especially vegetables
		Typically developing children	58		3-d Food record	Limited food repertoire*: 19.0 vs. 22.5 foods
						No. of foods not offered: 0–90 vs. 3–46 foods
Emond et al. (8)	United Kingdom	Children with ASDs	79	0.5–4.5	Feeding and FFQ	Difficult to feed, slow-feeders, and very choosy*: from 15 to 54 mo
		Typically developing controls	12,901			Food variety score: similar (at 6 mo; from 15 mo**)
						More likelihood of eating different meals from their family by 24 mo
Martins et al. (66)	Australia	Children with autism or ASDs	41	2–12	Feeding and eating behavior report	Self-feeding skills and mother’s control of their child’s feeding practices**: than typically developing children
		Typically developing siblings of children with autism	14			Neophobic*
		Typically developing children with no sibling with a disability	41
Schreck et al. (67)	United States	Children with autism	138	7–9.5	Eating problems	Feeding issues and acceptance of foods of low texture*
		Typically developing children	298			Range of foods consumed (except starches)**
Williams et al. (63)	United States	Children with ASDs	100	1.8–10	Survey reports	Picky eaters, 67%; food neophobia, 69%; limited variety, 60%; and rituals surrounding eating, 46%
						Exhibited food selectivity based on texture (69%), appearance (58%), taste (45%), smell (36%), and temperature (22%)

¹Values are ranges for age. *Significantly greater, P < 0.05; **significantly less, P < 0.05. ASD, autism spectrum disorder; YAQ, Youth/Adolescent Questionnaire.

TABLE 6.

Studies highlighting the nutrient intakes of individuals with ASDs¹

Study	Population	Group	Sample size, n	Age, y	Method of data collection	Finding (ASD vs. control)
Bicer and Alsaffar (2)	Turkey	Children with autism	1642	4–18	3-d Food records and feeding assessment survey	Fiber, calcium, zinc, iron, vitamins A and B-6, and folate intakes: inadequate by majority
						Sodium intake: above upper limit
Hyman et al. (17)	United States	Children with ASDs	3673	2–11	3-d Food records	Energy, vitamins A and C, zinc, and phosphorous intakes**
		Controls (NHANES, 2007–2008)				Fiber; choline; calcium; vitamins A, D, and K; magnesium; phosphorous; and potassium intakes: majority did not meet RDA
Zimmer et al. (57)	United States	Children with autistic disorder	22	8.2 ± 3.2	FFQ	Protein, calcium, and vitamins B-12 and D intakes**
		Typically developing children	22	8.1 ± 3.3		Magnesium*
Emond et al. (8)	United Kingdom	Children with ASDs	79	0.5–4.5	Feeding and FFQ	Energy, total fat, carbohydrate, protein, iron, and calcium intakes: similar (at 38 mo)
		Typically developing controls	12,901			Vitamins C and D intakes** (at 38 mo)
Xia et al. (41)	China	Children with ASDs	111	2–9	3-d Dietary recall	Calories, protein, vitamin E, and niacin: DRI met
						Fat; folic acid; vitamins A, B-1, B-2, B-6, and C; calcium; iron; zinc; and magnesium: DRI unmet
Herndon et al.4 (73)	United States	Children with ASDs	46	4.7 ± 1.2	3-d Sequential diary	Fiber, calcium, iron, and vitamins E and D intakes: majority consumed less than RDA in either group
		Typically developing children	31	4.9 ± 1.4		Vitamin B-6 intake*
Lockner et al. (6)	New Mexico	Children with ASDs	20	3–5	3-d Food records	Carbohydrate, protein, folate, iron, and vitamins B-6 and C intakes: majority consumed more than RDA in either group
		Typically developing children	20			Fiber, calcium, and vitamins E and A: majority consumed less than RDA in either group

¹Values are ranges or means ± SDs for age. *Significantly greater, P < 0.05; **significantly less, P < 0.05. ASD, autism spectrum disorder.

²Food records were analyzed for 115 children.

³Food records were analyzed for 252 children.

⁴Results were excluded for those on gluten-free and casein-free diets.

Children's eating behavior that influences nutrient intake.

Between 46% and 89% of children with ASDs exhibit nutritional challenges (62). This is particularly concerning because ASDs now present as a public health problem. Common feeding concerns reported include 1) difficulty accepting new foods and resisting novel experiences that extends to tasting, trying new foods (neophobia), and throwing foods (6, 57, 63–66); 2) difficulty with transition to textures, especially during infancy, leading to late acceptance of solid foods (4, 8); 3) restricted or selective intake based on food category, color, texture, consistency, appearance, taste, smell, brand, packaging, and food temperature (2–8, 63–69); 4) difficulty with mealtime presentations, e.g., specific plate and cutlery or positioning of food on a plate (3, 67, 70); 5) increased sensory sensitivity (tactile) leading to rejecting foods because of aversion to temperature, texture, or other food characteristics or restricting intake to food of preferred, tolerable, and manageable textures (4, 66, 71); 6) disruptive mealtime behaviors such as not staying seated at mealtimes or not eating with the family (4); 7) persistently wanting the same foods (4) or that food be made in a certain way (70); and 8) pica behavior (8).

Although children with ASDs are often labeled as “picky eaters” (6, 63), this habit was not associated with a lack of appetite (63). Children with ASDs more likely accept foods of low texture and higher energy density (3, 67). Interestingly, Bandini et al. (7) highlighted the facts that not only did children with ASDs eat notably fewer types of foods, they were offered fewer foods, and the absolute number of foods they refused and foods refused as a percentage of those offered were more than those in healthy children, particularly for vegetables. Even among adults with intellectual disabilities, those with autism had more severe feeding and mealtime challenges such as food selectivity and refusal than did those without autism (72).

Nutrient intake of individuals with ASDs.

Although children with ASDs eat a considerably smaller variety of foods, authors report no overall differences in their total calories, carbohydrates, or fat intakes (5, 6, 8, 65, 73, 74), suggesting that their satiety mechanisms are not impaired. Protein intake was adequate, lower, or quite similar to that of typically developing children (8, 41, 57, 65, 73–75), and nondairy protein intake was increased in children with ASDs (57, 73). Children with ASDs reportedly ate fewer vegetables and more energy-dense foods (65, 68). In contradiction to anecdotal reports of high carbohydrate consumption, fiber intake was inadequate in a considerable number of children with ASDs (2, 6, 17, 73).

Compared with typically developing children, a substantial number of subjects with ASDs had inadequate intakes of calcium (6, 7, 17, 41, 57, 73, 75), iron (73, 75), zinc (17, 41, 75), vitamin A (6, 17, 41, 75), vitamin D (7, 57, 68, 73), vitamin E (6, 73), riboflavin (75), vitamin C (41, 68), vitamin B-12 (36, 57, 75), folic acid (36, 41), and choline (17, 43). Vitamin B-6 intake was either significantly higher (73) or not met (41) by children with ASDs. Children with ASDs were reportedly found to consume fewer vegetables, salads, and fresh fruit in a study by Emond et al. (8), which used prospective dietary data from a population-based sample, and Bandini et al. (7) found that children with ASDs had a greater number of nutrients with inadequate intakes compared with typically developing children. This is also reflected by deficient concentrations of vitamin A (plasma) (34, 42), vitamin C (plasma) (39), folate (serum) (35, 36), biotin (WB) (34), calcium (RBC and serum) (34, 38), zinc (plasma) (47), and iron (48) in children with ASDs.

We believe that dietary data may be inaccurate due to parental bias because it is possible that parents who consented to participate in the research studies may have been more concerned about nutrition and feeding behaviors in their children such that their dietary patterns might be different from the general population of children with ASDs.

Summary, Discussion, and Future Research

A steep rise in the prevalence of individuals with ASDs (1) and their increased standardized mortality rates between 1.9 and 5.6 (76) pose enormous public health implications. This has stimulated intense research around its potential etiologic factors and consequences so that corrective measures can be taken. Much attention has been given to the child population, but with the number of middle-aged and elderly people with ASDs growing, research needs to be focused on adult populations. Although the nutritional status of individuals with ASDs has been extensively studied, there are many shortcomings in the literature. Lack of direct anthropometric assessments, standardized biochemical assays, use of proxy indicators for dietary data, small sample size, and age disparities among the groups are some of the limitations of the current published literature.

Early recognition of variation, if any, in the anthropometric measurements of individuals with autism from those of the healthy population will serve as an inexpensive, noninvasive, and objective method of nutritional status assessment. Abnormally accelerated rate of growth may be an early warning signal of vulnerability to autism not only among children with autism (21) but also among their infant siblings (29). Some believe this growth abnormality is a nonspecific expression of biological abnormalities found in these disorders (77). Whether the growth abnormalities are hereditary or are associated with psychiatric problems, predominantly ASDs, or due to multiple developmental disabilities needs to be addressed. Unusual dietary patterns, sedentary lifestyle, and decreased opportunities for physical activity may be the contributory factors for being overweight/obese among children with autism (13, 16). The observed variability in the prevalence estimates of obesity may be explained by diverse environmental factors, different definitions, lack of standardized measures of ASDs, or nonadjustment of comorbid conditions. Therefore, obesity prevention and management approaches for this at-risk group need further consideration.

The identification of specific biochemical correlates will increase reliability of nutritional assessment of individuals with autism. Researchers (35, 56) propose that lower folate, vitamins B-6 and B-12 concentrations, and hyperhomocysteinemia and hyperhomocysteineuria could be used as possible biomarkers for early diagnosis of ASDs. The broad distribution of vitamin B-6 in children with autism is very fascinating. Folate and vitamin B-12 deficiencies are known to arrest children’s development even in the absence of anemia and are linked to megaloblastic anemia, neuropathy, and demyelination of the central nervous system, respectively. Children with ASDs hail from a family background of dysregulated amino acid metabolism, which could be one of the underlying biochemical origins of ASDs (53). The significant positive correlation between 25-hydroxyvitamin D and calcium also supports the hypothesis that autism is a vitamin D deficiency disorder (38). Research investigating whether the concentrations of different biomarkers are associated with age, gender, developmental level, severity, and other comorbid psychiatric/nonpsychiatric disorders or are generalizable to all of the disorders in the spectrum is lacking. Methodologic issues with biochemical assays, fasting status, medication use, and dietary practices followed need consideration while resolving these findings. Nutrient deficiencies compounded by feeding difficulties exhibited by this population shed light on the importance of supplementation with these essential nutrients as well as antioxidants.

Lack of comprehension, communication, and memory makes recalling, noting, and estimating quantities a challenge for individuals with ASDs and is a barrier to collecting valid dietary data. With the aid of proxy indicators, particularly, parents or caretakers, researchers have gathered data on their dietary intakes and habits. Selective eating patterns, persistence to food presentations, limited food repertoire, food neophobia, sensory issues, and mealtime behavior issues are reported more often in children with autism than in those with other developmental disabilities or in typically developing children (2, 4–8, 63–65). Because food selectivity is more common in younger children than in older children, research focused on the role of selective eating and methods of treatment at an earlier age is warranted. Despite these challenges, macronutrient deficiencies are often not an area of major concern in children with autism (5, 6, 8, 65, 73, 74). Even though children with ASDs consume a less varied diet, giving them fortified foods rather than additional vitamins or food supplements may help meet their nutrient needs. In addition, because parental diet influences a child’s food choice, the environment in the home should be addressed, and more variety should be incorporated into the family meal. Furthermore, diet particularly disadvantaged by feeding issues affects the motor, cognitive, and behavioral development of children with autism; therefore, greater focus should be on early intervention and methods to prevent the emergence of feeding problems.

It is too early to propose a nutrition assessment tool for individuals with ASDs because of lack of research on larger populations and the varied methodologies used in the existing studies. However, based on existing research, the anthropometric, biochemical, and dietary markers listed in Figure 1 emerge as a potential screening tool for children with autism. Anecdotal and clinical evidence emphasizes the need for a more active role by nutrition professionals in this case because problems are often overlooked, because picky eating is common in neurotypical children and children with autism around the age when autism is diagnosed. Higher health costs and resources are involved because of the special medical conditions of children with ASDs, and, thus, improving their nutritional status before the contributory factors and comorbid/chronic conditions arise is of utmost importance. Although certain symptoms of ASDs are present early in life, ASDs are usually diagnosed after the age of 3 y, highlighting the need for methods to facilitate early diagnosis. Thus, research should identify anthropometric and biochemical indicators that can serve as early warning signs and be included in routine practice that begins at birth and continues throughout the child’s development. Dietary assessments should be corroborated with anthropometric and laboratory data to get a true picture of the nutritional status of children with ASDs. This will help these children benefit from timely and intensive interventions and in adulthood expand their options for living independently. Therefore, this area of research warrants focused attention and effort.

FIGURE 1.

Potential markers for screening children with autistic spectrum disorders.