Children in China with Autism Spectrum Disorders (ASD) are prone to vitamin A deficiency (VAD). The present study compared two vitamin A supplements (VAS) in two groups of children with ASD and VAD to explore a better VAS program for children with ASD.
A total of 138 3–8-year-old children with ASD (118 males and 20 females) were enrolled in this 6-month study. Of these 138 children, 82 who had VAD (ASD-VAD) were divided into two VAS groups that received the recommended VAS program (RNI-VAS) or a weekly dose of VAS (WD-VAS). The 56 children who had normal vitamin A levels (ASD-VAN) served as a control group. The Social Responsiveness Scale (SRS) was used to assess the severity of social impairment before and after the interventions. Their serum retinol (VA) and oxytocin (OXT) concentrations, the mRNA expression of retinoic acid receptors (RARs), and CD38 gene in peripheral blood was measured before and after the 6-month intervention.
The WD-VAS program increased VA levels better than the RNI-VAS program did (P < 0.01), and it significantly decreased SRS scores (P < 0.05). In addition, the change in VA was positively correlated with the change in mRNA levels in RARβ (r = 0.2441, P = 0.0092), the CD38 in PBMC (r = 0.2729, P = 0.0033), and the change in OXT concentration in serum (r = 0.3735, P < 0.0001). VA was also negatively correlated with changes in SRS scores across the three groups (r = −0.2615, P = 0.0026).
The WD-VAS might be more suitable for children with ASD and VAD than other interventions to improve both VA and social functioning, which may be mediated through the RARβ-CD38-OXT axis.
The term autism spectrum disorders (ASD) refers to a group of neurodevelopmental disorders characterized by social deficits and repetitive stereotyped sensory-motor behaviors . ASD is growing rapidly across the world, and the prevalence of ASD in China is 1% , leading to ASD becoming a group that cannot be ignored. Currently, the treatment for ASD relies primarily on behavioral interventions.
Children with ASD are prone to multiple comorbidities, which affects the rehabilitation and quality of life . Among them, nutrient deficiency is a common comorbid disease in children with ASD. A good nutritional status in early life is conducive to the neurological development of children , but the stereotyped eating behavior and sensory abnormalities of them increase the risk of nutrient deficiency, which may further aggravate the symptoms of ASD.
Vitamin A deficiency (VAD) is a global public health problem, and the WHO recommends a VA supplement (VAS) program of 200,000 International Units (IU) every 4–6 months for children 1–6 years old as a low-cost intervention to reduce child morbidity and mortality in countries in which VAD is endemic . In China, preschool-age children are at high risk of VAD, which has a prevalence rate of 2–10%. Furthermore, the proportion of marginal VAD (MVAD) in preschool children is much higher than VAD . A survey of the nutritional status of children with ASD found their intake of VA was less than 80% of the recommended intake . Our own research found the incidence of VAD and MVAD of preschool children with ASD in China was 77.9%, which was significantly higher than that of normal children. We also found that serum VA levels were significantly increased after 6 months of a recommended VAS program in children with ASD. However, only 3.2% of children with ASD and VAD (including MVAD) returned to the normal VA level (>1.05 μmol/L) following an RNI dosage . Notwithstanding the evidence to date, there is a lack of comprehensive and systematic clinical studies on suitable VAS programs for children with ASD, and the influences of such programs on ASD needs to be clarified.
Vitamin A (VA) is critical for the brain development, which is transported in the form of retinol in the blood, and function in retinoic acid (RA) in tissues . The RA exerts is critical role in brain development via the transcription factor retinoic acid receptors (RARs), which could activate the genetic transcription . Although VA interventions have been highly valued for reducing the blindness and mortality among preschool children, the attention to VA status and its relationship with the neurodevelopmental disorders such as ASD has just begun. Some studies have reported the relationships between RA-RAR signal and CD38-oxytocin (OXT) pathway on the pathogenesis of ASD based on its vital role in improvement of social disorders [10,11,12,13,14]. Avraham found oral beta-carotene supplements for newborn BALB/c and BTBR mice reduced stereotyped behavior and increased social interaction by increasing CD38 and OXT . An all-trans retinoic acid (atRA) intervention can upregulate CD38 mRNA transcription in immortalized lymphocytes of patients with ASD . A previous study of ours found that embryonic VAD may be a risk factor for autistic-like behaviors in rats, whose molecular mechanism may involve RARβ regulation of the CD38-OXT axis in the hypothalamus .
Recent studies have found that the WHO-recommended single high-dose regimen of VAS is not ideal for improving serum VAD. It has a short-term, slight improvement in serum VA levels in patients with “subclinical” or mild to moderate VAD, but serum VA levels did not improve further after 2 months of administration [18, 19]. Although the maintenance of basic food intake (including fortified foods) combined with regular low-dose supplements significantly increased serum VA and reduce the incidence of VAD , the daily dietary VA intake of children with ASD is low and their VA level is more difficult to improve and sustain because of their high prevalence of feeding problems, which are caused by their stereotyped eating behaviors or sensory abnormalities . Thus, it is necessary to explore different supplemental dosages and time intervals for them.
The present study compared two vitamin A treatments in two groups of children with ASD and VAD, to investigate a better VAS program for children with ASD. Moreover, we assessed the change in vitamin A over 6 months in a group of children with ASD but not VAD that was only treated with behavioral therapy. In addition, we aimed to determine whether VAS could improve the core symptoms, especially social interaction, of children with ASD through the regulation of the RARβ -CD38-OXT signal.
Materials and methods
A total of 138 patients with ASD, including 118 boys and 20 girls, who were 3–8 years old were included in this study and were followed-up for 6 months. All children enrolled in the study were evaluated by clinical observations conducted by a professional developmental-behavioral pediatrician and diagnosed using the Diagnostic and Statistical Manual of Mental Disorders (DSM-V).
The research program was peer-reviewed and approved by the ethics review committee. The clinical trial is also registered with the China Clinical Trials Registry (ChiCTR-ROC-14005442). The parents and/or legal guardians of the participants signed written informed consent forms before the research procedures began. The exclusion criteria were 1) having other developmental disorders or a history of mental illness, such as Rett syndrome, 2) having cerebral palsy, epilepsy, or other congenital or hereditary diseases, 3) having a history of recent infections, or 4) using high doses of vitamin supplements in the past 6 months.
As shown in Fig. 1, all the children were divided into VA normal (ASD-VAN, n = 56) or VA deficiency (ASD-VAD, n = 82) groups using a serum VA cutoff of 1.05 μmol/L. The 82 children with VAD were randomly divided into two vitamin A supplementary programs: the recommended intake program (RNI-VAS, n = 46) or a weekly VAS dosage (WD-VAS, n = 36). The RNI-VAS group received a single oral supplement of 200,000 IU of VA at 6-month intervals, equivalent to 1100 IU/d. The WD-VAS group received a VAS program of 50,000 IU per week for 11 weeks, followed by 13 weeks of intermittent treatment, which was equivalent to 3000 IU/d. The dosage of the VAS program was mainly based on the VA Dietary Reference Intake of the Chinese Nutrition Society  and the WHO  recommendations for preschool children. And all children, including those in the ASD-VAN group, received necessary behavioral therapy during the study.
The anthropometric measurements included height and weight. A standardized Z-score conversion of the raw data were performed using WHO Anthro (within 5 years old) and WHO Anthroplus (beyond 5 years old) software. The Z scores for height and weight are referred to as ZHA and ZWA, respectively.
Assessment of ASD-related symptoms scale
All the children completed the Social Responsiveness Scale (SRS) at baseline and at the end of the study. The SRS scale is the most commonly used questionnaire for measuring the severity of autism symptoms in children and adolescents. An SRS score of 60–75 indicates mild to moderate autism, whereas a score of 76 or higher indicates severe autism [23, 24].
Blood collection and retinol detection
All the children underwent venipuncture before and after the 6-month intervention to collect 2 ml of venous blood, which was immediately centrifuged at 3000 rpm/min for 10 min at room temperature. The serum was used to detect the levels of retinol and oxytocin, and peripheral blood mononuclear cells (PBMC) were used to extract total mRNA to detect the expression levels of RARs and CD38.
High-performance liquid chromatography
Retinol is one of the major forms of VA metabolism and has become the diagnostic criteria for assessing VA levels in many clinical applications. Serum retinol concentration was estimated via high-performance liquid chromatography (HPLC) based on previously described methods . Briefly, deproteinization was conducted with dehydrated alcohol, and retinol extraction from serum was performed using hexane. Hexane was then removed through evaporation by nitrogen gas. A mobile phase mixture (methanol: water = 97:3) was used to dissolve the retinol residue. Then, the retinol was detected in the prepared sample with an HPLC apparatus (DGU-20As, Shimadzu Corporation, Kyoto, Japan) (C18, 315 nm). The entire procedure was performed by the same operator in a dark room to protect the serum from light. A serum retinol concentration greater than 1.05 μmol/L is defined as a normal VA level (VAN), and less than 1.05 μmol/L is defined as VA deficiency (VAD), and each assay was performed in triplicate.
Enzyme-linked immunosorbent assay
The serum was collected to determine the concentration of OXT using a commercial OXT ELISA kit (ENZO Life Sciences, New York, USA). The procedure was performed according to the manufacturer’s protocols. The color optical density absorbance was measured at a wavelength of 405 nm. The sample concentrations were calculated using a microtiter plate reader (Thermo, MA, USA), according to the relevant standard curves, and each assay was performed in triplicate.
Real-time PCR was used to assess the relative mRNA expression of RARs and the CD38 gene, using the following procedure. The mRNA was isolated using the Tri-Reagent (Sigma, MO, USA) in accordance with the guidelines and reverse transcribed into cDNA, and real-time PCR was performed with a SYBR Green mix (QIAGEN, MD, GER) and CFX Manager (BIO-RAD, California, USA). The primer sequences were designed by Primer 6.0 software (San Francisco, USA) and produced by the Beijing Genomics Institution (BGI, Beijing, PRC). The primer sequences for the PCR genes are listed in Table 1.
The statistical analysis and graphing of the data were performed with GraphPad Prism 7.0 software (San Diego, USA). The normal distribution of each dataset was tested by the D’Agostino and Pearson normality test. The data from more than two groups were analyzed by one-way ANOVA or nonparametric tests, based on the normality test of the data. The independent t-test was used to analyze the difference between two groups and the dependent (or paired) t-test was used to analyze within-group changes over time (before and after treatment). Pearson’s correlation (r) and linear regression were used to examine the relationship between changes in vitamin A levels and changes in serum oxytocin, SRS scores, and related gene expression. Two-tailed tests were used in all analyses with the significance value of <0.05.
As shown in Table 2, 138 eligible children were recruited and randomly assigned to three groups at baseline (56 subjects in the ASD-VAN group, 46 in the 1100 IU/d RNI-VAS group, and 36 in the 3000 IU/d WD-VAS group). There were no significant differences in their mean age, ZHA, ZWA, or the subscale scores or total scores of the SRS among the three groups. The baseline vitamin A levels of the children in the RNI-VAS and WD-VAS group were comparable (Table 2).
Variations in the serum retinol (VA) concentrations of the three groups
Figure 2 shows the serum VA concentrations of the three groups before and after the 6-month intervention. The VA level of the ASD-VAN group was significantly lower at 6 months than it was at baseline (Before vs. After, 1.278 ± 0.1728 μmol/L vs. 1.139 ± 0.2689 μmol/L, P = 0.0028). Among the 56 children in the ASD-VAN group, 15 children were newly diagnosed with VAD, accounting for 26.79% of that group. The RNI-VAS intervention significantly increased the VA concentration (Before vs. After, 0.839 ± 0.1413 μmol/L vs. 0.9816 ± 0.2882 μmol/L, P = 0.0061), as did the WD-VAS (Before vs. After, 0.809 ± 0.1772 μmol/L vs. 1.161 ± 0.2668 μmol/L, P < 0.0001).
When we further analyzed the effect of the two VAS programs on the VA recovery rate (serum retinol above 1.05 μmol/L), we found the VA recovery rate was 45.7% (21/ 46) in the RNI-VAS group and 63.9% (23/36) in the WD-VAS group. The recovery rate in the WD-VAS program was slightly higher than that of the RNI-VAS regimen, but the difference was not statistically significant (χ2 = 2.701, P = 0.100), possibly because of the relatively small sample sizes.
Figure 2b shows the differences in the VA levels of all three groups at 6 months relative to their baseline levels. The children in the ASD-VAN group (no VA intervention) had a decrease in serum VA levels between baseline and 6 months. The increased level of VA in the WD-VAS group was significantly higher than it was in the RNI-VAS group at 6 months (RNI-VAS vs. WD-VAS, 0.143 ± 0.3115 μmol/L vs. 0.346 ± 0.3025 μmol/L, P = 0.0036).
Effects of different doses of VA on serum OXT and SRS scores
Figure 3 shows the difference in OXT levels and SRS scores between baseline and 6 months. Compared to the ASD-VAN group, the 3000 IU/d VA supplement group (WD-VAS program) significantly improved the level of serum OXT (ASD-VAN, −6.736 ± 38.39 pg/mL; RNI-VAS, 4.349 ± 31.54 pg/mL; WD-VAS, 25.04 ± 49.56 pg/mL; ASD-VAN vs. WD-VAS, P = 0.0022, Fig. 3a). The WD-VAS program also decreased SRS scores (ASD-VAN, 1.808 ± 20.79; RNI-VAS, −2.682 ± 25.58; WD-VAS, −10.14 ± 15.07; ASD-VAN vs. WD-VAS, P = 0.0292, Fig. 3b). However, the changes in the OXT levels (ASD-VAN vs. RNI-VAS, P = 0.2452; RNI-VAS vs. WD-VAS, P = 0.3318) and SRS scores (ASD-VAN vs. RNI-VAS, P > 0.9999; RNI-VAS vs. WD-VAS, P = 0.1804) of the RNI-VAS group were not statistically different from those of the ASD-VAN group or the WD-VAS group.
Correlations of changes in measures after the intervention
The correlation analysis included changes in the measures before and after the intervention in all the children in the study. As shown in Fig. 4a, the change in serum VA between baseline and 6 months had a significant negative correlation with the change in SRS scores (P = 0.0026, r = −0.2615). Moreover, there were significant positive correlations between the changes in serum VA and OXT levels (P < 0.0001, r = 0.3735), and the change in the mRNA expression levels of RARβ (P = 0.0092, r = 0.2441) and CD38 (P = 0.0033, r = 0.2729) genes (Fig. 4b–d). There was no significant correlation between VA changes and RARα, RARγ, OXT, or OXTR mRNA expression (Supplementary File S1).
This is the first study that has used a weekly VAS program and compared its effects with RNI-VAS for ASD children comorbid with VAD. The weekly WD-VAS program was slightly better for the recovery for serum VA than the RNI-VAS regimen was. However, there were still 36.1% of children who did not reach a normal vitamin A level in the WD-VAS group, compared to 63.9% in the RNI-VAS group, which was in consistent with our previous studies . All these findings strongly indicate that ASD children in preschool age are high-risk groups for VAD and there are several related reasons for this. First, rapid growth and neurodevelopment at this age require adequate VA. Second, WD-VAS reduced the social impairment of ASD(SRS scores), but no obvious improvement in stereotyped behavior was observed during the intervention. Stereotypical eating behavior may be the least likely to be improved, leading to persistent insufficient intake of dietary VA. And it is hard to maintain normal VA levels merely rely on nutritional supplements. Therefore, the preschool age children with ASD should be regularly monitored for nutritional management.
We found significant correlations between VA level fluctuations and changes in OXT levels and social function in children with ASD. In addition to improving serum vitamin A, the WD-VAS program improved the social functioning of children with ASD, which might be superior to the RNI-VAS regimen. Moreover, we found significant correlations between changes in vitamin A levels and changes in the RARβ-CD38-OXT axis and social function in children with ASD (Fig. 4). Therefore, reducing VAD in children with ASD may contribute to a reduction in both malnutrition and core symptoms of ASD (especially social impairment). Even though the WD-VAS program brought some minor changes in serum VA and the social functioning of children with ASD and VAD, larger doses of VAS may be needed to achieve satisfactory VA recovery rate. If it is determined that it is helpful for treating the disease, taking a therapeutic amount instead of a preventive amount of vitamin A might be necessary.
We think that vitamin A activation/metabolism-related genetic or molecular abnormalities may coexist with the background of autistic genetic abnormalities in children with ASD, which may lead to disruptions in synaptic plasticity and may be involved in the development of ASD. Several studies have reported vitamin-A-related genetic and molecular abnormalities in individuals with ASD, including decreased retinoic acid , beta-carotene , and RALDH1 [11, 12] levels in serum, as well as genetic variants in the retinoic acid response element (RARE)  and retinoic acid induced 1(RAI1) . Our previous study also found embryonic VAD may be a risk factor for autistic-like behaviors in rats, and the related molecular mechanism may involve the RARβ-CD38-OXT axis in the hypothalamus of rats . In addition, the present study observed a significant increase in the levels of RARβ-CD38-OXT axis of ASD children after a WD-VAS intervention, further suggesting that VAS programs may increase active retinol metabolites—RA and its receptor RARβ expression—changing the levels of downstream CD38 and OXT expression, to alleviate the social disorder of children with ASD.
This study still has several limitations that should be noted. First, studies with a broader range of dosage are necessary to determine an optimal VAS dose for children with ASD. Second, the follow-up time of this study was only half a year. The long-term effects of VAS and the effects of sustained VAS need to be studied further. Third, the number of cases included in this study is limited, which may be the reason why some measures that changed after the VAS intervention did not yield statistically significant differences. Based on the exploratory findings of this study, it is necessary to conduct large-scale clinical randomized controlled studies that measure more variables to explore our scientific hypotheses in depth. For example, future studies could include detailed assessments of stereotyped behavior, dietary intake records, and eating behavior intervention records.
In conclusion, preschool-aged children with ASD are prone to VAD, and nutritional management is necessary for children with ASD to overcome the effects of malnutrition. A WD-VAS program for 6 months can effectively increase serum retinol concentrations and levels of the RARβ-CD38-OXT axis to alleviate the impairment of social interaction in children with ASD. Further large-scale clinical research and studies of specific mechanisms also need to be implemented.
Lord C, Elsabbagh M, Baird G, Veenstra-Vanderweele J. Autism spectrum disorder. Lancet. 2018;392:508–20.
Sun X, Allison C, Wei L, Matthews FE, Auyeung B, Wu YY, et al. Autism prevalence in China is comparable to Western prevalence. Mol Autism. 2019;10:7.
Moody L, Chen H, Pan YX. Early-life nutritional programming of cognition—the fundamental role of epigenetic mechanisms in mediating the relation between early-life environment and learning and memory process. Adv Nutr (Bethesda, Md). 2017;8:337–50.
WHO. Vitamin A Supplements: a guide to their use in the treatment and prevention of vitamin A deficiency and xerophthalmia. 2nd ed. WHO; 1997.
WHO. Global prevalence of vitamin A deficiency in populations at risk 1995–2005. WHO; 2009.
Sun C, Xia W, Zhao Y, Li N, Zhao D, Wu L. Nutritional status survey of children with autism and typically developing children aged 4-6 years in Heilongjiang Province, China. J Nutr Sci. 2013;2:e16-e.
Liu X, Liu J, Xiong X, Yang T, Hou N, Liang X, et al. Correlation between nutrition and symptoms: nutritional survey of children with autism spectrum disorder in Chongqing, China. Nutrients. 2016;8:294.
Shearer KD, Stoney PN, Morgan PJ, McCaffery PJ. A vitamin for the brain. Trends Neurosci. 2012;35:733–41.
Blomhoff R, Green MH, Berg T, Norum KR. Transport and storage of vitamin A. Science. 1990;250:399–404.
Carmona-Mora P, Walz K. Retinoic acid induced 1, RAI1: a dosage sensitive gene related to neurobehavioral alterations including autistic behavior. Curr Genomics. 2010;11:607–17.
Hao Z, Wu Q, Li Z, Li Y, Li Q, Lai X, et al. Maternal exposure to triclosan constitutes a yet unrecognized risk factor for autism spectrum disorders. Cell Res. 2019;29:866–9.
Pavăl D, Rad F, Rusu R, Niculae A-Ş, Colosi HA, Dobrescu I, et al. Low retinal dehydrogenase 1 (RALDH1) level in prepubertal boys with autism spectrum disorder: a possible link to dopamine dysfunction? Clin Psychopharmacol Neurosci. 2017;15:229–36.
Xu X, Li C, Gao X, Xia K, Guo H, Li Y, et al. Excessive UBE3A dosage impairs retinoic acid signaling and synaptic plasticity in autism spectrum disorders. Cell Res. 2018;28:48–68.
Zhou W, Li S. Decreased levels of serum retinoic acid in chinese children with autism spectrum disorder. Psychiatry Res. 2018;269:469–73.
Y A, EM B, M D, WA A, L V, A A, et al. Beta-carotene as a novel therapy for the treatment of “Autistic like behavior” in animal models of Autism. Behav Brain Res. 2019;364:469–79.
Riebold M, Mankuta D, Lerer E, Israel S, Zhong S, Nemanov L, et al. All-trans retinoic acid upregulates reduced CD38 transcription in lymphoblastoid cell lines from Autism spectrum disorder. Mol Med. 2011;17:799–806.
Lai X, Wu X, Hou N, Liu S, Li Q, Yang T, et al. Vitamin A deficiency induces autistic-like behaviors in rats by regulating the RARβ-CD38-oxytocin axis in the hypothalamus. Mol Nutr Food Res. 2018. https://doi.org/10.1002/mnfr.201700754.
Mason J, Greiner T, Shrimpton R, Sanders D, Yukich J. Vitamin A policies need rethinking. Int J Epidemiol. 2015;44:283–92.
Bento C, Matos AC, Cordeiro A, Ramalho A. Vitamin A deficiency is associated with body mass index and body adiposity in women with recommended intake of vitamin A. Nutricion Hosp. 2018;35:1072–8.
Ribeiro-Silva Rde C, Nunes IL, Assis AM. Prevalence and factors associated with vitamin A deficiency in children and adolescents. J Pediatr. 2014;90:486–92.
Bandini LG, Curtin C, Phillips S, Anderson SE, Maslin M, Must A. Changes in food selectivity in children with autism spectrum disorder. J Autism Dev Disord. 2017;47:439–46.
Chinese Nutrition Society, Dietary Guidelines for Chinese (People’s Medical Publishing House, China; 2016).
Bölte S, Poustka F, Constantino JN. Assessing autistic traits: cross-cultural validation of the social responsiveness scale (SRS). Autism Res. 2008;1:354–63.
Chan W, Smith LE, Hong J, Greenberg JS, Mailick MR. Validating the social responsiveness scale for adults with autism. Autism Res. 2017;10:1663–71.
Liu J, Liu X, Xiong XQ, Yang T, Cui T, Hou NL, et al. Effect of vitamin A supplementation on gut microbiota in children with autism spectrum disorders—a pilot study. BMC Microbiol. 2017;17:204.
Fregeau B, Kim BJ, Hernández-García A, Jordan VK, Cho MT, Schnur RE, et al. De novo mutations of RERE cause a genetic syndrome with features that overlap those associated with proximal 1p36 deletions. Am J Hum Genet. 2016;98:963–70.
This research was funded by the National Nature Science of Foundation of China (81770526, 81771223), the Key Project of Guangdong Province (2018B030335001) and Guangzhou City (202007030002), and the Scientific research innovation program for graduate students of Chongqing (CYB 17107).
Conflict of interest
The authors declare that they have no conflict of interest.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Lai, X., Zhang, Q., Zhu, J. et al. A weekly vitamin A supplementary program alleviates social impairment in Chinese children with autism spectrum disorders and vitamin A deficiency. Eur J Clin Nutr 75, 1118–1125 (2021). https://doi.org/10.1038/s41430-020-00827-9