Does Vitamin D Deficiency Increase the Risk of Autism Spectrum Disorder? Linking Evidence with Theory—A Narrative Review

Abstract

Introduction: Vitamin D plays a crucial role in brain health by providing antioxidant, anti-inflammatory, and neuroprotective benefits. It regulates neurotransmitters and neurotrophins that are essential for the development, maintenance, and functioning of the nervous system. Deficiency in vitamin D during pregnancy and early childhood can disrupt neurodevelopment, potentially contributing to autism spectrum disorder (ASD). The aim of this narrative review was to analyze the potential link between vitamin D deficiency and the development of ASD, as well as to explore the therapeutic benefits of vitamin D supplementation. Method: We performed a literature search across PubMed, EMBASE, Web of Science, and the Cochrane Library databases, reviewing observational studies, randomized controlled trials (RCTs), and meta-analyses for evidence of an association between vitamin D deficiency and ASD. Results: The results were mixed but promising, with most observational studies suggesting a positive link between vitamin D deficiency and ASD, though these findings were not consistently replicated in prospective studies or RCTs. In conclusion, the available data are insufficient to establish vitamin D deficiency as a definitive cause of ASD. Further RCTs, particularly during pregnancy and infancy, are needed to better understand the role of vitamin D in the etiology of ASD and its potential as a therapeutic intervention. Conclusions: The current available data are insufficient to support vitamin D deficiency as a definitive factor in the etiology of autism spectrum disorders. To translate this hypothesis into clinical practice, additional randomized controlled trials, particularly during pregnancy and early infancy, are needed.

1. Introduction

Autism spectrum disorder (ASD) is categorized as a neurodevelopmental disorder presenting with a lack of social interaction and communication and repetitive and stereotypical behaviors [1]. ASD is a polygenic multifactorial illness resulting from the interplay between genes and the environment [2]. The literature suggests that genetic factors play an important role in the pathogenesis of ASD [3,4]. However, only 25–30% of children with ASD have shown genetic abnormality, while 70% did not show any such abnormality, suggesting the role of environmental factors [4]. Environmental factors, including maternal folic acid and vitamin D (VD) deficiency, medication use during pregnancy, maternal smoking, infections during pregnancy, exposure to toxic substances, and vaccines, are associated with ASD [5].

Based on data from Cannell’s epidemiological survey, there was a higher prevalence of ASD in urban areas, at higher elevations, and in locations with more air pollution [6]. Air pollution, cloudy weather, seasonal variations, dark-skinned immigrants, migration to poleward latitudes, and nutrition are specific factors linked to both ASD and VD deficiency [7]. Through its effects on immune modulation, neurotrophic factors, neurotransmitters, and gene regulation, processes that are similarly dysregulated in ASD, vitamin D plays a crucial role in neurodevelopment [8]. In contrast to other non-genetic causes, VD insufficiency is a distinct and biologically plausible contributing factor.

A wealth of emerging studies links maternal and childhood VD deficiency to increased risk and severity of ASD, emphasizing its potential as an actionable and feasible focus for intervention. Considering the worldwide prevalence of VD deficiency, addressing this factor could potentially reduce ASD incidence and severity. This review examines existing evidence on VD deficiency as a modifiable risk factor for ASD, highlighting its role in pathophysiology and implications for prevention and treatment.

2. Materials and Methods

To explore and evaluate the broad range of topics related to the hypothesis that vitamin D deficiency may contribute to ASD, we opted to conduct a narrative review. Search strategies: We performed a literature search using PubMed, EMBASE, Web of Science, and the Cochrane Library databases. The search terms included a combination of MeSH terms and free-text keywords. The search strategy used the following Boolean operators: (“Autism” OR “Autism Spectrum Disorder” OR “ASD” OR “Autistic”)AND (“Vitamin D” OR “1,25-dihydroxyvitamin D3” OR “25-hydroxyvitamin D2” OR “25-hydroxyvitamin D3” OR “25(OH)D”) AND (“Supplements” OR “Supplementation” OR “Intake”).

Inclusion criteria included observational studies, including case–control, cohort, and register-based surveys, that assessed the association between gestational vitamin D (VD) deficiency and autism spectrum disorder (ASD) diagnosis in offspring. Additionally, we considered observational studies evaluating VD levels in neonates and children diagnosed with ASD. Randomized controlled trials (RCTs) investigating the effects of VD supplementation in neonates and children under 18 years with ASD were also included. Furthermore, we incorporated meta-analyses that examined the association between VD deficiency and ASD, as well as meta-analyses evaluating the efficacy of VD supplementation in ASD.

The Newcastle–Ottawa quality scale for case–control and cohort study assessment was used to evaluate the quality of case–control studies included in this study [9]. We included two case–control studies, two nested controls, two prospective cohort studies, and one register-based population survey assessing the association between maternal VD deficiency and the diagnosis of ASD in offspring. Five case–control studies evaluating the association of VD levels in neonates and ASD diagnosis and eleven case–control studies examining the VD deficiency in children and the diagnosis of ASD were included. All relevant RCTs of VD supplementation alone in children with ASD <18 years of age were included in our study. After the search and exclusion of duplication, 7 RCTs were included in our study. Two meta-analyses of case–control studies and three meta-analyses of RCTs of VD supplementation were also included.

We excluded studies published in languages other than English. Narrative reviews, expert opinions, commentaries, letters, and conference abstracts without original data were also excluded. Studies with incomplete or insufficient data, such as missing key outcome measures or lacking statistical analysis, were not considered. Research that did not use established diagnostic criteria for ASD (e.g., DSM or ICD classifications) or assessed VD levels without a control group was excluded. Duplicate studies or those using overlapping data without new insights were removed. Case reports and small case series with limited sample sizes, as well as studies with a high risk of bias, poor study design, or inadequate control for confounders, were also excluded.

3. Results

3.1. Linking Vitamin D with Autism

VD, in addition to regulating calcium and phosphorus metabolism, has an important role in neurodevelopment. Growing research indicates that vitamin D deficiency likely plays a role in the pathophysiology of ASD [8]. Concurrently, research has demonstrated the role of VD supplementation in treating the primary symptoms of ASD in children [10].

VD exists in two distinct forms: VD2 and D3. Vitamin D3 is mostly created in the skin by ultraviolet B (UVB) radiation from 7-dehydrocholesterol (7-DHC), whereas vitamin D2 is derived from food. VD is essential for the proper functioning of the neurological system. It functions via vitamin D receptors (VDR) found in the neurons and glial cells. VDR is expressed most highly in the hippocampus, thalamus, hypothalamus, substantia nigra, orbitofrontal cortex, cingulate, and amygdala. The presence of both VDR and the enzyme that converts 25 hydroxyvitamin D3 [25(OH)D3] to the active form 1,25 di-hydroxyvitamin D3 [1,25 5(OH2)D3] in the brain is important evidence supporting the involvement of VD in brain functioning. The active form, 1,25(OH2)D3, is suggested to be a neurosteroid hormone that stimulates cell division and proliferation. VD demonstrates anti-inflammatory and antioxidant properties. It plays a significant role in the expression of genes and proteins essential to neuron physiology, neurotrophin expression, intracellular calcium signaling, neurotransmitter release, and neuronal differentiation [11].

Studies indicate that VD plays a role in synaptic transmission, particularly chemical synapses [12]. Because VD deficiency increases the amounts of cholesterol in the presynaptic membrane and vesicles, it alters the fusion characteristics of the synaptic membrane and affects the efficiency of neurotransmitter release [13]. Conversely, VD treatment may partially restore vesicle fusion capacity [14]. Furthermore, VD affects the transcription of proteins such as complexin2 and synaptojanin1 which are important for neurotransmitter release [15,16]. VD directly boosts the activity of voltage-dependent calcium channels [17] and promotes the expression of calcium sensors such as syt1 and syt2 in the brain [18], thereby influencing synchronized transmitter release. As a result, VD deficiency can affect all those functions that rely on efficient neurotransmitter release and good neuronal connectivity such as cognition, learning, and behavior [19]. Research has shown abnormalities of neurotransmitters such as gamma-aminobutyric acid (GABA), glutamate, serotonin, dopamine, oxytocin, and acetylcholine in children with ASD [20]. Consequently, it seems reasonable to presume that VD deficiency is a contributing factor in the etiopathogenesis of ASD.

Vitamin D (VD) plays a crucial role in regulating neurotransmitter release and metabolism by modulating the expression of receptors, transporters, and enzymes. VD deficiency has been associated with a decline in the expression of the excitatory amino acid transporter (EAAT) and GABA transporter 3 (GAT3), resulting in impaired glutamate and GABA reuptake systems [21]. Additionally, VD deficiency reduces the expression of glutamate synthetase 1, glutamate decarboxylase, and GABA receptor mRNA [22,23,24].

GABA serves as the primary excitatory neurotransmitter in the developing brain, and disruptions in its function can affect critical neurodevelopmental processes such as cell differentiation, synapse maturation, neuronal migration, and proliferation [25]. Similarly, autism spectrum disorder (ASD) is hypothesized to arise from imbalances in the glutamatergic and GABAergic systems, leading to an altered excitatory/inhibitory equilibrium. This dysregulation, particularly in the medial prefrontal cortex, contributes to impaired information processing and dysfunctional social behavior [25].

Furthermore, magnetic resonance spectroscopy studies have revealed lower glutamate concentrations in the striatum of adults with ASD and ASD mouse models compared to controls [26]. Given its role in neurotransmitter synthesis, VD may help regulate excitatory and inhibitory signaling, thereby influencing neurodevelopmental and cognitive functions [27].

VD may promote the synthesis of neurotrophins such as the nerve growth factor (NGF) and the glial-derived neurotrophic factor (GDNF) [28], which are important for the growth and function of neurons [29]. However, one meta-analysis [30] indicated that children with ASD had considerably greater levels of NGF than controls, indicating that the role of neurotrophins in ASD is not well understood. Nevertheless, studies have indicated that VD increases the expression of neurotrophins [31], which has positive effects on the developing brain; yet, there is little experimental data regarding children with ASD.

Children with ASD have higher levels of inflammation. ASD children are reported to have elevated proinflammatory cytokines such as interleukin IL-6, IL-8, and interferon-gamma (IFN-γ) [32]. Research employing mice as model organisms for autism has demonstrated that elevated levels of IL-6 affect synaptic transmissions and result in traits similar to autism, such as decreased social engagement and difficulties with learning [33]. A systematic review of normally developing children and adolescents concluded that VD status negatively correlated with inflammation [34]. Similarly, another systematic review of immune cell studies found that VD supplementation consistently reduced IL-6 and IL-8 [35]. Additionally, VD helps to increase the levels of glutathione peroxidase 1 (GP1), which lowers oxidative stress [36] and helps regulate glutathione redox imbalance, which is another contributing factor to ASD [37]. A summary of actions of VD on brain addressing various etiological factors implicated in ASD is depicted in Figure 1

3.2. Evidence of the Relationship Between Vitamin D Deficiency and ASD

Animal studies have reported behavioral abnormalities in the offspring of VD-deficient mice. The offspring of deficient mice showed a lack of dam/pup interactions, like maternal licking and grooming [38]. As juveniles or adults, these offspring showed social and behavioral deficits and stereotyped behaviors similar to ASD symptoms [39,40]. Hyperactivity, impulsivity, and sensory sensitivity were also reported in the offspring of vitamin D-deficient rats [41].

A strong correlation was found between maternal VD deficiency and an increased risk of ASD in offspring. Observational studies reported a varied risk of autism depending on the season of birth and latitude [42]. Conception during fall in the northern hemisphere region was linked to a 6% higher risk of autism compared with summer births [Odds Ratio = 1.06, 95% CI (1.02–1.10)] [42,43,44]. Being the child of a dark-skinned immigrant in a cold country increased the likelihood of developing ASD [45,46,47]. These results made VD deficiency a strong contender to account for these findings. The observational studies of the mother’s serum VD levels during pregnancy and the assessment of the offspring’s risk of an ASD diagnosis are shown in

Table 1. Overview of cross-sectional studies showing an association between maternal vitamin D deficiency and ASD.

First Author/Year	Study Design	Sample Size	Trimester	Sample and Assessment Method	Vitamin D Levels ng/mL	Diagnosis	Outcome
Chen et al. (2016) [48]	Case–control China	ASD: 68 Controls: 68	11–13 weeks	Serum	<25—deficiency >50—sufficient	DSM-5	Significant association (p—0.001 and OR: 3.99 (2.58–7.12)
Magnuss et al. 2016 [49]	Register-based total population study Sweden	N = 509 4–17 years	-	Serum	<25—deficiency	ICD-10	Significantly increased offspring risk of ASD OR = 2.51, (95% CI 1.22–5.16)
Lee et al. (2019) [51]	Nested case–control Sweden	ASD: 449 Controls: 574	10.9 (9.3–13.0) weeks	Dried blood spot LC-MS	<25—deficiency >50—sufficient	ICD-10 and DSMIV	VD deficiency was associated with 1.58 times higher odds of ASD (95% CI: 1.00, 2.49) as compared with controls.
Windham, et al. (2020) [52]	Case–control California	ASD—534 Controls—421	Mid-pregnancy	Serum LC-MS	<50—deficiency >75—sufficient	DSMIV-TR	No association OR-0.79 (95% CI 0.49–1.3)
Souranderetal 2021 [53]	Nested case–control study Finland	ASD—1558 Controls—1558	First and early second trimester	Serum CMI	<30—deficiency	ICD-10	Increased risk of ASD in VD deficiency (1.44, 95% CI 1.15–1.81, p = 0.001

Autism spectrum disorder (ASD), chemiluminescence microparticle immunoassay (CMI), confidence interval (CI), Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV), Diagnostic and Statistical Manual of Mental Disorders-IV-TR (DSM-IV-TR), International Classification of Diseases (ICD-10), liquid chromatography–tandem mass spectrometry method (LC-MS), Odds Ratio (OR), Social Responsiveness Scale (SRS), and vitamin D (VD).

[48,49,50,51,52,53,54]. Most studies showed a negative correlation between maternal serum 25 hydroxy vitamin D (25(OH) D) levels and a diagnosis of ASD [48,49,50,51,53], except for a few [52,54]. Nonetheless, a prospective study using Mendelian randomization did not find compelling evidence linking a mother’s vitamin D levels during pregnancy to an offspring’s diagnosis of ASD [54]; see

Table 2. Overview of longitudinal studies showing an association between maternal vitamin D deficiency and ASD.

First Author/Year	Study Design	Sample Size	Trimester	Sample and Assessment Method	Vitamin D Levels ng/mL	Diagnosis	Outcome
Vinkhuyzen et al. (2018) [50]	Prospective cohort Netherland	ASD: 62 Controls: 3895	Mid-gestation: 18.1–24.9 weeks Neonatal serum	Serum LC-MS	<25—deficiency >50—sufficient	SRS	Significantly high SRS scores (p = 0.001)
Dowd et al. (2022) [54]	Prospective cohort study UK	Mothers: 7689	Midpoint of each trimester: 7 weeks, 20 weeks, and 34 weeks	Serum LC-MS	<25—deficiency >50—sufficient	ICD-10 SCDC	No asssociation [OR = 0.9895% CI = 0.90–1.06),] Mendelian randomization suggested no causal effect [OR = 1.08, 95% CI = 0.46–2.55)]

Autism spectrum disorder (ASD), confidence interval (CI), International Classification of Diseases (ICD-10), liquid chromatography–tandem mass spectrometry method (LC-MS), Odds Ratio (OR), Social Responsiveness Scale (SRS), and Social Communication Disorder Checklist (SCDC).

.

Studies conducted with neonates [50,55,56,57,58] ato assess the association between VD deficiency and ASD are shown in

Table 3. Overview of cross-sectional studies showing an association between neonatal vitamin D deficiency and ASD.

First Author/Year	Study Design	Sample Size (Both Boys and Girls)	Sample and Assessment Method	Vitamin D Levels ng/mL	Diagnosis/Outcome Measure	Outcome
Fernell et al. (2015) [55]	Case–control Sweden	ASD: 58 Controls: 57	Dried blood Spot LC-MS	ASD: 24.0 ± 19.6 Controls: 31.9 ± 27.7	DSM-5	Significantly low VD levels in a group of newborn children, who later developed ASD.
Lee et al. (2019) [51]	Nested case–control Sweden	ASD: 449 Controls: 574	Dried blood spot LC-MS	ASD: 10.28 ± 5.96 Controls: 10.64 ± 6.31	ICD-10 and DSM-IV	VD deficiency was associated with 1.33 (CI 1.01, 1.74) times higher odds of ASD as compared with controls.
Schmidt et al. (2019) [57]	Case–control USA	ASD: 357 Controls: 234	Dried blood Spot LC-MS/MS	ASD: 32.04 ± 14.96 Controls: 33.08 ± 15.72	ADOS-2, ADI-R	No significant association OR: 0.98 (0.63–1.51).
Windham et al. (2019) [58]	Case–control	ASD: 563 Controls: 436	Dried blood Spot LC-MS/MS	ASD: 34.04 ± 14.1 Controls: 33.72 ± 11.85	DSM-IV-TR	No significant association OR: 0.96 (0.65–1.43).

Autism Diagnostic Interview-Revised (ADI-R), Autism Diagnostic Observation Schedule, (ADOS-2), autism spectrum disorder (ASD), Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV), Diagnostic and Statistical Manual of Mental Disorders-IV-TR (DSM-IV-TR), Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-5), International Classification of Diseases (ICD-10), liquid chromatography–tandem mass spectrometry (LC-MS/MS), Social Responsiveness Scale (SRI), and vitamin D (VD).

Three case–control studies [50,51,55] of neonates showed increased risk, while another three [56,57,58] did not show a significant increase in risk of ASD, and two longitudinal studies failed to show any such association; see

Table 4. Overview of longitudinal studies showing an association between neonatal vitamin D deficiency and ASD.

First Author/Year	Study Design	Sample Size (Both Boys and Girls)	Sample and Assessment Method	Vitamin D Levels ng/mL	Diagnosis/Outcome Measure	Outcome
Vinkhuyzen et al. (2017) [50]	Prospective cohort Netherland	ASD: 62 Controls: 3895	Serum LC-MS	-	SRS	VD-deficient individuals had significantly high SRS scores p = 0.001
Ali et al. (2019) [56]	Prospective study cohort Canada	ASD: 26 Controls: 2500	Serum: A two-step chemiluminescence assay	-	ADOS-2/DSM-IV/ DSM-5	No increase in risk of ASD RR: 1.06 (0.95–1.18)

Autism Diagnostic Observation Schedule, (ADOS-2), autism spectrum disorder (ASD), Diagnostic and Statistical Manual of Mental Disorders (DSM-IV)-IV, Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-5), liquid chromatography–tandem mass spectrometry (LC-MS/MS), Social Responsiveness Scale (SRI), vitamin D (VD), and risk ratio (RR).

. Most case–control studies [59,60,61,62,63,64,65,66,67,68,69] of children have shown a significant association between VD deficiency and ASD, except for two (see

Table 5. Overview of studies showing an association between vitamin D deficiency and ASD in children.

First Author	Study Design	Sample Size/Age	Sample and Assessment Method	Vitamin D Levels ng/mL	Diagnosis/Outcome Measure	Outcome
Saad et al. (2016) [59]	Case–control cross-sectional Egypt	ASD: 122 Controls: 100 Age ASD: 5.09 ± 1.42 year Controls: 4.88 ± 1.30 year	Serum ELISA	ASD: 18.02 ± 8.75 Controls: 42.51 ± 9.48	DSM-IV-TR	Significantly low VD in ASD children (p ≤ 0.0001)
Coşkun et al. (2016) [60]	Case–control Turkey	ASD: 85 Controls: 82 Age ASD: 43.4 ± 25.3 month Controls: 47.1 ± 14.2 month	Serum ELISA	ASD: 79.5 ± 25.9 Controls: 65.1 ± 23.9	DSM-5	Significantly low VD in ASD children (p ≤ 0.0001)
Basheer et al. (2017) [61]	Case–Control India	ASD: 40 Controls: 30 Age ASD: 3–12 year control: 3–12 year	Serum LC-MS/MS	ASD: 13.5 ± 4.7 Controls: 12.7 ± 4.7	DSM-5 ADI-R	No significant difference between groups p = 0.462 OR: 0.88 (0.14–5.63)
Desoky et al. (2017) [62]	Case–control cross-sectional Egypt	ASD: 60 Controls: 40 Age ASD: 7.03 ± 2.34 year Controls: 7.91 ± 3.21 year	Serum ELISA	ASD: 18.63 ± 10.8 Controls: 45.9 ± 8.85	-	Significantly low VD in ASD children (p = 0.001)
Altun et al. (2018) [63]	Cross-sectional Turkey	ASD: 60 Controls: 45 Age ASD: 5.8 ± 2.7 year Controls: 6.7 ± 2.5 year	Serum ELISA	ASD: 13.79 ± 1.03	DSM-IV-TR	Significantly low VD in ASD children (p < 0.001)
Arastoo et al. (2018) [64]	Cross-sectional Iran	ASD: 31 Controls: 31 Age ASD: 9.17 ± 2.11 year Controls: 9.31 ± 2.09 year	Serum ELISA	ASD: 9.03 ± 4.14 Controls: 15.25 ± 7.89	DSM-IV ADI-R	Significantly low VD in ASD children (p = 0.001) OR: 12.273 (1.447–104.101)
El-Ansary et al. (2018) [65]	Cross-sectional Saudi Arabia	ASD: 28 Controls: 27 Age ASD: 7.0 ± 2.34 year Controls: 7.2 ± 2.14 year	Plasma ELISA	ASD: 95.63 ± 26.63 Controls: 140.43 ± 17.68	DSM-IV-TR	Significantly low VD in ASD children (p < 0.001)
Alzghoul et al. (2018) [66]	Case–control cross-sectional Jordan	ASD: 83 Controls: 106 Age ASD: 5.08 year Controls: 5.02 year	Serum LC-MS/MS	ASD: 23.4 Controls: 37.5	DSM-5	Significantly low VD in ASD children OR: 9.896 (4.605–21.264)
Chtourou et al. (2019) [67]	Case–control Tunisia	ASD: 40 Controls: 43 Age Controls: 4.76 ± 1.08 year	Serum -	ASD: 17.13 ± 9.65 Controls: 21.34 ± 8.1	DSM-5	No significant difference between groups (p = 0.03)
Petruzzelli et al. (2020) [68]	Case–control Italy	ASD: 54 Controls: 36 Age ASD: 6.87 (±3.92) year Controls: 11.28 (±4.44) year	Serum CMI	ASD: 35 (64.8) Controls: 12 (33.3)	DSM-5 ADI-R ADOS-2	A significant association between ASD and VD deficiency (p = 0.006)
Sengenc¸ et al. 2020 [69]	Case–control Turkey	ASD: 100 Controls: 100 Age ASD: 5.95, 3.13 year Controls: 6.68, 3.8 year	Serum ELISA	ASD: 42.86, 19.84 Controls: 48.57, 22.36	DSM-5	Significantly low VD in ASD children (p = 0.037)

Autism Diagnostic Interview-Revised (ADI-R), Autism Diagnostic Observation Schedule (ADOS-2), autism spectrum disorder (ASD), chemiluminescence microparticle immunoassay (CMI), Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV), Diagnostic and Statistical Manual of Mental Disorders-IV-TR (DSM-IV-TR), Diagnostic and Statistical Manual of Mental Disorders-5 (DSM-5), enzyme-linked immunosorbent assay (ELISA), liquid chromatography–tandem mass spectrometry (LC-MS/MS), and vitamin D (VD).

). Moreover, two meta-analyses demonstrated that children with ASD had considerably lower serum concentrations of 25(OH) D than typically developing (TD) children [70,71]; see

Table 6. Overview of metanalysis showing an association between Vitamin D deficiency and ASD.

First Author	Study Design	Population Characteristics	Serum Vitamin D Estimation Method	Diagnosis	Result
Wang T et al. 2015 [70]	11 case–control	870 ASD patients and 782 healthy controls	Serum plasma ELISA, RIA, LC-MS/MS HPLC	DSM IV(TR)	Significant difference between the ASD group and control group (WMD = −8.63; 95% CI (−13.17, −4.09), p = 0.0002).
Wang Z et al. 2020 [71]	34 studies	Total participants = 20,580 (Asia, America, Europe, Africa)	Serum, plasma, or dried blood spot ELISA, RIA, LC-MS/MS HPLC	DSM-IV, DSM-IV-TR, ADOS, ADIR, DSM-V, ICD-9, ICD-10, ICD-F84.0	Vitamin D concentration of the ASD group was 7.46 ng/mL lower than that of the control group (95% CI: −10.26; −4.66 ng/mL, p < 0.0001.
	26 case–control studies	Assessing Vitamin D in children 1792 ASDs 1969 controls
	3 case–control studies and 2 nested case–control studies of neonates	Assessing Vitamin D in neonates (2687 ASDs 3574 controls)
	1 nested case–control	Maternal vitamin D concentration of the ASD and control groups (517 ASDs, 642 controls)
	2 cohort studies	Investigated the OR/RR for ASD incidence after being exposed to early-life vitamin D deficiency (5442 neonates, 3957 pregnant women)

Autism Diagnostic Interview-Revised (ADI-R), Autism Diagnostic Observation Schedule (ADOS), autism spectrum disorder (ASD), confidence interval (CI), Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV), Diagnostic and Statistical Manual of Mental Disorders-IV-TR (DSM-IV-TR), Diagnostic and Statistical Manual of Mental Disorder-5 (DSM-5), enzyme-linked immunosorbent assay (ELISA), high-performance liquid chromatography (HPLC), International Classification of Diseases-9 (ICD-9), International Classification of Diseases-10 (ICD-10), International Classification of Diseases-F84.0 (ICD-F84.0), liquid chromatography–tandem mass spectrometry (LC-MS/MS), and radioimmunoassay (RIA).

. However, the conclusions of these studies shall be interpreted cautiously. The generalizability of these findings is limited because most studies had small sample sizes and were geographically limited. Some studies used secondary care records to derive vitamin D status and did not address the confounders like immigration, disadvantaged social status, lifestyle factors, psychiatric illness in mothers, and psychotropic drug use. Many studies did not measure other nutrients that have a role in neurodevelopment and could have been the actual cause of the increased risk of ASD. Few vitamin D studies have used the same cut-off values for deficiency in both mothers and neonates. The scales used to measure ASD such as the Social Responsiveness Scale (SRS) can show higher scores in children with other childhood communication disorders. Moreover, VD deficiency is inseparable from other modifiable risk factors, making it difficult to assume a causal relationship.

3.3. Evidence of Improvement with Vitamin D Supplementation

Multivitamin intake in pregnancy has shown a reduced risk of ASD [72]. However, evidence to support the role of VD supplementation for ASD treatment is less clear. Feng et al. [73] included 285 healthy controls and 215 children with ASD (mean age 5.1 years old) in an open-label study. The Aberrant Behavior Checklist (ABC) and Childhood Autism Rating Scale (CARS) were used to evaluate the symptoms of ASD. When the ASD children received VD supplementation for three months at a dose of 150,000 IU/month i.d. + 400 IU/day, their scores on the ABC and CARS considerably improved. However, most randomized control trials (RCTs) have shown no significant improvement in ASD with VD supplementation; see

Table 7. Overview of RCTs of vitamin D supplementation for treating ASD.

First Author/Year	Study Design	Type of Participants Age/Gender	Treatment Details and Duration	Change in Level of Vitamin D	ASD Severity Measure (M)	Result
						Before- After the Change	Between-Group Comparison
Azzam (2015) [74]	Prospective, case–control	n = 21 Age: 2–12 year M:F: 16:5	2000 IU/day of vitamin D3 for 6 months (n = 10) No supplement (n = 11) for cases	47 ± 20 to 71 ± 35 (nmol/L) 69 ± 41 to 70 ± 36 (nmol/L)	CARS ATEC	Decrease Decrease	NS
Kerley 2017 [75]	RDBPC	n = 38 Age: <18 year M:F: 33:5	2000 IU vitamin D3 (n = 18) for 20 weeks Placebo (n = 20)	58.4 ± 17.9 to 86.1 (nmol/L)	ABC SRS	Decrease Decrease	NS
Mazahery (2019) a New Zealand [76]	RDBPC	Age: 2.5–8 year M:F: 60:13	2000 IU/day (n = 28) Omega 3 (n = 28) Placebo (n = 30) 12 months	-	ABC	Decrease in hyperactivity and irritability	Significant reduction in irritability and hyperactivity (−5.2 ± 6.3 vs. −0.8 ± 5.6, p = 0.047).
Mazahery (2019) [77]	RDBPC	n = 73 Age: 2.5–8 year M:F: 60:13	2000 IU/day vitamin D3 (n = 19) omega-3 (n = 23) Vitamin D3 and omega 3 (n = 15) Placebo (n = 16) For 12 months	68 ± 21 to an increase by 95 ± 14 (nmol/L)	SRS SPM	Decrease Decrease	NS
Javadfar (2020) [78]	RDBPC	n = 43 children with ASD Age: 3–13 year M:F: 36:7	300 IU/kg/day (Max. 6000 IU/day) of vitamin D3 for 15 weeks (n = 22) Placebo n = 21	8.19 ± 6.78 to 39.10	CARS ATEC ABC-C subscale	Decrease Decrease Decrease	Significant decrease in CARS AND ATEC scores No significant difference in ABC domains
Moradi (2020) [79]	RDBPC	n = 100 Age: 6–9 year M:F: 100:0	300 IU/kg/day (max. 5000 IU/day) of vitamin D3 (n = 25) Perceptual-motor exercises (n = 25) Exercises and vitamin D3 (n = 25) Placebo (n = 25) 3 months	12.6 to 24.36	SS-GARS-2	Decrease	Significant decrease in stereotypes (p = 0.01)
Ansari et al., 2020 [80]	RDBPC	n = 40 Age: 6–14 year M:F: 100:0	50,000 IU/week or 50,000 IU/2 weeks (vitamin D3) for 10 weeks (n = 10) Placebo (n = 10)	11.12 to 31.60	GARS-2	Significant decrease in stereotypes

Aberrant Behavior Checklist (ABC). Aberrant Behavior Checklist–Community (ABC-C), Autism Treatment Evaluation Criteria (ATEC), Childhood Autism Rating Scale (CARS), female (F), Gilliam Autism Rating Scale (GARS-2), international unit (IU), male (M), number of subjects (n), not significant (NS), Randomized Double-Blind Placebo-Controlled (RDBPC) study, Sensory Processing Measures (SPMs), Social Responsiveness Scale (2SRS-2), and Stereotypy Subscale of the Gilliam Autism Rating Scale-2 (SS-GARS-2).

. Most RCTs showed potential bias due to uncertain allocation concealment and inadequate blinding [74,75,76,77,78,79]. Most studies showed no improvement in the core symptoms of ASD between groups, while there was a decrease in symptoms within groups, before and after the administration of the VD supplement [74,75,76,77,78,79]. Two RCTs have demonstrated a reduction in hyperactivity [76,78], while no change in irritability and sensory symptoms [79] was noted in the ASD group compared to placebo. Additionally, supplementing with VD has been demonstrated to lessen stereotypes [79,80]. Serotonin and dopamine abnormalities have been documented in children diagnosed with autism [81]. Studies have also reported that children with autism exhibit stereotypic behavior due to serotonin and dopamine abnormalities [82]. Since VD regulates serotonin synthesis, its supplementation can improve stereotypes [83]. According to meta-analysis (

Table 8. Overview of meta-analysis showing the effect of vitamin D supplementation on symptoms of ASD.

First Author	Study Design	Population Characteristics	Vitamin D Supplementation	Outcome Measure	Result
Song et al., 2020 [84]	4 RCTs 3 included in meta-analysis	- (New Zealand, China, Ireland) Age 2–10 years Both male and female	Dose range 800 IU/day to 2000 IU/day Duration 20 weeks–12 months	ABC SRS CARS DD-CGAS	No significant improvement (SMD = −0.46, 95% CI: −0.87 to −0.05; p = 0.03
Li et al., 2020 [85]	5 RCTs 3 included in meta-analysis	349 participants New Zealand Age 2–12 years Both male and female	Dose: 2000 IU/day Duration = 5–12 months	ABC SRS CARS SPM ATEC DD-CGAS SSGARS-2	Social interaction: No difference (pooled MD: −1.54; 95% CI: [−4.09, 1.01]; p = 0.24) Communication: No difference (pooled MD: −0.05; 95% CI: [−1.79, 1.69]; p = 0.96) Repetitive restrictive behavior: No difference (pooled MD: 0.85; 95% CI: [−0.33, 2.02]; p = 0.16) Hyperactivity: A significant difference (pooled MD: −3.20; 95% CI: [−6.06, −0.34]; p = 0.03) Irritability: No difference (pooled MD: −2.31; 95% CI: [−6.08, 1.46]; p = 0.23),
Zhang 2023 [86]	8 RCTs 6 included in meta-analysis	266 participants (New Zealand, Iran, China, Ireland) Age 2.5−14 years Both male and female	Dose range: 800 IU/day to 50,000 IU/week and 2000 IU/day Duration 10 weeks to 12 months	ABC SRS CARS	No difference in core symptoms (pooled MD: −8.74; 95% CI: −17.45, −0.03; p = 0.05) Social interaction: No difference (pooled MD: −0.07; 95% CI: −1.70, 1.57; p = 0.93) Communication: No difference (pooled MD: −0.04; 95% CI: −1.19, 1.10; p = 0.94), Stereotypes: Significant difference between the intervention and placebo groups (pooled MD: −1.39; 95% CI: −2.7, −0.07; p = 0.04) Irritability: No difference (pooled MD: −1.79; 95% CI: −4.42, −0.85; p = 0.18) Hyperactivity: No difference (pooled MD: −1.35; 95% CI: −4.37, 1.67; p = 038)

Aberrant Behavior Checklist (ABC), Autism Treatment Evaluation Criteria (ATEC), Childhood Autism Rating Scale (CARS), confidence interval (CI), Disabilities–Children’s Global Assessment Scale, (DD-CGAS), Mean Difference (MD), randomized controlled trial (RCT), Sensory Processing Measures (SPM)), Social Responsiveness Scale (SRS), and Stereotypy Subscale of the Gilliam Autism Rating Scale-2 (SSGARS-2).

), the core symptoms of ASD did not improve [84,85,86], while hyperactivity [85] and stereotypes [86] decreased significantly in children receiving VD supplementation. Most RCTs were conducted on small sample sizes and were limited to a specific geographical area, limiting the generalizability. Based on these results, it is too premature to conclude that VD supplementation can treat ASD. More research is required to determine which children can benefit from VD supplements and what dosage has the best chance of working while posing the least risk of negative side effects.

Studies regarding the prophylactic administration of VD in pregnant women to prevent ASD are very few. Stubbs et al. [87] reported a reduced risk of ASD in the children of mothers supplemented with VD. A dose of 5000 IU/day was given to pregnant women with a previous child with an ASD diagnosis, and newborns were given 1000 IU/day from birth until three years. The children of supplemented mothers had a fourfold less chance of developing ASD as compared to non-supplemented mothers [87]. In a more recent study, high-dose vitamin D3 supplementation from pregnancy week 24 until 1 week postpartum did not reduce the overall risk of autism [88]. However, the findings of these studies shall be interpreted with care due to the small sample size and varied duration of VD administration among women [87,88]. Some women were taking VD before pregnancy while others were prescribed VD at different trimesters [87].

In conclusion, the data available for VD prophylaxis in mothers to prevent ASD in children are currently insufficient.

4. Discussion

The existing evidence about VD deficiency in the etiopathogenesis of ASD and its supplementation in treatment is inadequate. A recent study by Gao et al. [89] reported no causal link between VD deficiency and ASD, while reverse analysis showed ASD leading to VD deficiency. In actuality, reverse causality could account for the substantial body of observational studies demonstrating a positive correlation between VD deficiency and ASD. Our discussion will focus on VD functions—like neuroprotection, inflammation, and gene expression—and their rationale for effectiveness in treating ASD.

The neuroprotective effect of VD may not be sufficient to treat ASD patients, as the neurotransmitters, synaptic connections, and the shaping of neural circuitry do not holistically depend on the role of VD [90] in neurodevelopment. The mere fact that VD can easily pass through the blood–brain barrier (BBB) does not mean it can affect brain cells. Its role in neurodevelopment depends on other factors such as the residual amount of VD in the brain, its metabolism, and its utilization by brain cells [91].

ASD patients have been shown to have increased neuroinflammation characterized by an increase in proinflammatory cytokines [92]. Vitamin D may suppress proinflammatory cytokines and enhance anti-inflammatory cytokines [93]. However, VD cannot fully address the extensive neuroinflammatory abnormalities observed in ASD.

A wide range of genetic abnormalities with complex mechanisms of gene regulation and expression were noted with ASD [94]. While VD plays an important role in gene expression, it cannot solely change the complex gene regulatory system in ASD. In addition, epigenetic modifications such as DNA methylation and histone modification are noted in ASD, and VD has a limited role in epigenetic regulation [95].

More research is warranted to conclude the role of VD, its safe dose, and the timing of treatment initiation in ASD patients. The initiation of VD at an appropriate time when neurodevelopment is taking place in the fetus is the key to preventing brain lesions in ASD. More RCTs are required to determine the efficacy of VD supplementation in children with ASD while controlling for all potential confounders.

5. Conclusions

Our summary shows mixed results about the effectiveness of VD supplementation in ASD. For clinicians, this suggests a potential complementary intervention for ASD management, especially in cases of vitamin D deficiency. Therefore, it should be considered alongside other treatments for ASD, such as behavioral therapy and pharmacotherapy. Dosage, duration, and patient-specific factors must be carefully managed to avoid toxicity.