Medical Hypotheses 81 (2013) 195–198

Contents lists available at SciVerse ScienceDirect

Medical Hypotheses

journal homepage: www.elsevier .com/locate /mehy

Autism, will vitamin D treat core symptoms?

0306-9877/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.mehy.2013.05.004

⇑ Tel.: +1 805 439 1075; fax: +1 805 439 1073.E-mail address: jjcannell@vitamindcouncil.org

John Jacob Cannell ⇑Vitamin D Council, 697 Higuera Street, Suite H, San Luis Obispo 93401, USA

a r t i c l e i n f o

Article history:Received 25 April 2013Accepted 11 May 2013

a b s t r a c t

No medication exists to treat the core symptoms of autism. However, some children spontaneouslyimprove and have optimal outcomes. Parents of autistic children who have access to swimming pool havereported summertime improvement in symptoms to me. A Japanese case report found the same summertimes improvements. If the cause of that summertime improvement could be identified, it may lead toeffective treatment. Vitamin D is highly seasonal with a summertime surfeit and a wintertime deficit.The hypotheses that the increased prevalence in the diagnosis of autism is due to better detection implythat parents, teachers and physicians of the 1950s, 60s, and 70s missed this non subtle diagnosis, anunlikely scenario. Recent research indicates that autism often first present itself during the second andthird year of life. This is a time when most toddlers have no known sources of vitamin D. Vitamin Dhas remarkable antioxidant, anti-inflammatory, and anti-autoimmune properties. In vitro, in vivo, andanimal experiments provide compelling data for vitamin D’s role brain proliferation, differentiation, neu-rotrophism, neuroprotection, neurotransmission, and neuroplasticity. It also upregulates glutathione,upregulates a suit of genes involved in DNA repair and raises the seizure threshold. Adequate, perhapspharmacological, doses of vitamin D may have a treatment effect in the core symptoms of autism.

� 2013 Elsevier Ltd. All rights reserved.

Paper

There is no treatment for the core symptoms of autism spec-trum disorder (ASD). However, some children spontaneously im-prove and have optimal outcomes [1]. Thus, it is possible thatsome unidentified factor in the environment may favorably affectthe course of the disease. A Japanese case report showed summer-time improvement in the core symptoms of autism [2]. Some par-ents of autistic children who have access to swimming pools in thesummer have reported dramatic seasonal improvements to me,with a surfeit of symptoms in August and a surplus of symptomsin February. If we could we identify the cause of the summertimeimprovement, it may lead us to the first known autism treatment.However clues to this factor may be found in epidemiological andprevalence data.

I remember in medical school, in 1975, one of my professorsinsisted that I see a child in the psychiatric clinic. My professor saidthat if I did not examine the child, I would likely miss the one caseof it I might see during my medical career. The child had autism.My professor was correct for a while in that I saw no autism inthe late 1970s. I owned a clinic in the coalfields of West Virginiaduring the 1980s and the clinic cared for hundreds if not thousandsof children but we never saw a case of ASD. However, the 1990sproved my professor wrong as ASD became increasingly common.

As of 2012, the prevalence of ASD in the USA was one in 88children. And it has risen dramatically in the last several decades,from one in 5000 children in 1975 to one in 88 children in 2012,a 600% increase in 30 years. While some of the increased preva-lence is clearly due to diagnostic substitution, such as substitutingthe diagnosis of autism for mental retardation, an unknown per-centage of the increase appears to be real. What accounts for thisincreased prevalence of ASD?

(Fig. 1) As ASD is highly heritable, geneticists say that theincreasing prevalence that is not due to diagnostic substitution issimply due to better detection. How, they ask, can a genetic diseaseincrease by 600% in 30 years, especially one in which the pheno-types do not often reproduce? So, the geneticists must say betterdetection explains the increase as they have no other possibleexplanation. Let’s examine what that better detection assertionactually implies.

Autism is not a subtle condition. The child I saw in 1975 wassimply impossible to miss. Children afflicted with autism usuallydo not speak, they are loners, they have impaired interpersonalrelationships, they frequently stim or do repetitive behaviors likeflapping their hands or hitting their heads, and they often havemajor temper meltdowns over minor changes in their schedule.In other words, the diagnosis is difficult to miss. However, that isexactly what the geneticists are telling us. They say that the schoolteachers, parents, and doctors of the 1950s, 60s, and the 70s simplymissed all these children. That explanation has no face validity.

The second implication of the better detection explanation forthe increasing prevalence has to do with parents. Currently,

Fig. 1. Autism Prevalence, Autism Speaks, http://www.ncbi.nlm.nih.gov/pubmed/22456193 (accessed 4.25.2013).

196 J.J. Cannell / Medical Hypotheses 81 (2013) 195–198

parents of autistic children expend their time and treasure onexpensive and time consuming behavioral, occupational, andspeech therapy. The average out-of-pocket expenditure for familieswith an autistic child is around $6,000.00 per year or 14% of theparent’s annual income [3]. Thus, both financial expenses and timeexpenditure is substantial [4].

Many states report that they have shortages of ASD-relatedtreatment personnel, including shortages of behavioral therapists,speech-language pathologists, and occupational therapists [5].Thus, the better detection theorists would have us believe that par-ents today are expending huge sums of their income to treat chil-dren that they would have been ignored 30 year ago. Again, suchan explanation has no face validity.

More sober geneticists now admit that a significant percentageof the increased prevalence of autism reflects a true increased inci-dence [6]. But how can a highly heritable condition explode in inci-dence in one generation? Some geneticists now seek to explain itas a genetic-environmental interaction [7]. If so, what is the envi-ronmental factor?

For arguments sake, let’s say a broad autism genotype under-lays the broad autism phenotype [8]. About 5–10% of the adultpopulation has the broad autism phenotype, a personality struc-ture characterized by rigidity, obsessiveness, impaired socializa-tion, and aloofness [9]. Also, for arguments sake, let’s assume thisbroad autism genotype is unchanged in incidence. That is, thegenotype has not changed over time. However, something mustbe interacting with the broad autism genotype to produce anexploding incidence of ASD. What could that something be andwhat qualifying characteristics should any such candidate risk fac-tor fulfill?

First, that candidate risk factor must be involved in brain devel-opment and brain function. Second, it must ‘‘fit the curve,’’ that is,that candidate risk factor must have increased or decreased

dramatically over the last 30 years to explain the exploding inci-dence. Third, it must explain the epidemiology of ASD.

I believe that factor is a powerful but forgotten neurosteroid,calcitriol, a seco-steroid intimately involved in both brain develop-ment and brain function [10]. In vitro, in vivo, and animal experi-ments provide compelling data for calcitriol’s role brainproliferation, differentiation, neurotrophism, neuroprotection,neurotransmission, and neuroplasticity [11,12]. Calcitriol exertsits biological function not only by influencing cellular processes di-rectly, but also by influencing gene expression via a vitamin Dreceptor that resides on up to 10% of the genes in the codinghuman genome [13].

The only known substrate for calcitriol is 25(OH)D and the onlyknown substrate for 25(OH)D is vitamin D, either made in the skinor put in the mouth. Studies show that 25(OH)D levels have fallendramatically from 1994 to 2004 [14]. More than 60% of US childrenhad insufficient 25(OH)D levels in the National Health and Nutri-tion Examination Survey of 2001–2004 [15]. No studies existdetailing the change in 25(OH)D levels over the last three decadesbut certainly sun avoidance became commonplace during thattime. During that same time, there were no compensatory recom-mendations for higher oral intake to make up for what the sun wasno longer making in the skin. Thus, it appears that vitamin D ‘‘fitsthe curve’’, in that 25(OH)D levels were dramatically falling as aut-ism prevalence was dramatically increasing.

The epidemiology of autism includes the facts that autism ismore common in urban areas, high air pollution areas, cloudyareas, and at higher latitudes [16]. While not all studies agree, anumber of studies are also now showing that autism is more com-mon among the dark-skinned [17,18]. All these epidemiologicalfacts are explained by the vitamin D hypothesis.

Four studies have looked at vitamin D levels in autistic childrenor their mothers and all have found low levels (<30 ng/ml) in autis-

J.J. Cannell / Medical Hypotheses 81 (2013) 195–198 197

tic children [12]. One found no difference in vitamin D levelsbetween autistic children and boys with acute inflammation (aninadequate control), while the other three found differences, somesignificant and some not. One study found Somali mothers withautistic children had average vitamin D levels of 6.7 ng/ml, about30% lower than Somali mothers without autistic children. Abouta dozen papers have looked at vitamin D intake in autistic children,all finding that most autistic children do not meet vitamin D intakerequirements for their age.

More recently, Mostafa and AL-Ayadhi studied 50 autistic chil-dren, comparing them to 30 typically developing controls [19].They found that autistic children had significantly lower serumlevels of 25-hydroxy vitamin D than healthy children (P < 0.001)with 40% and 48% being vitamin D deficient and insufficient,respectively although the parents of the autistic children reportedthe same amount of sun exposure as did the parents of typicallydeveloping children. The authors also found that serum 25-hydro-xy vitamin D had significant negative correlations with the Child-hood Autism Rating Scale (P < 0.001). In addition, anautoantibody (anti-MAG) was identified in 70% of autistic patientsand serum 25-hydroxy vitamin D levels had significant negativecorrelations with serum levels of anti-MAG auto-antibodies(P < 0.001).

Most experts agree that any vitamin D effect must operate dur-ing gestation and that symptoms are present from birth onward.However, Rogers recently reviewed the developmental literatureand concluded that autism is not a disorder that is evident fromthe earliest months of life [20]. Rather, Rogers concluded it is a dis-order with a gradual onset of symptoms beginning during the first2–3 years of life.

Thus, the pathophysiology of ASD is compatible with a theorythat the environmental insult occurs during the first 12–24 monthsof life and not exclusively during gestation. While 25(OH)D levelshave clearly fallen during the last 3 decades, it is instructive toask yourself where a toddler gets his or her vitamin D. Most infantshave a source of vitamin D in formula, either because they are notbreast fed or because breast fed infants are often supplementedwith formula. But what happens after weaning?

Studies show that fewer and fewer infants are weaned on vita-min D enriched cow’s milk, as used to be the case, and more andmore onto fruit juice. In fact, a recent study found that only 14%of mothers thought cow’s milk should be used at weaning while84% thought fruit juice the best [21]. The other source of vitaminD during toddlerhood, sun exposure, has also been curtailed overthe last 30 years [22]. If toddlers don’t get it from diet, and notfrom sun exposure, then supplements are the only other alterna-tive. A 2006 study showed that only 31% of American children be-tween the ages of 12–24 months receive a supplement [23].

It may be that the broad autism genotype becomes ASD aftertoddlers fail to get adequate vitamin D during the ages of12–24 months. A recent study found that the prevalence of autismin the 50 states is inversely and strongly correlated with surfaceUVB readings [24]. The regression coefficient for solar UVB dosesand autism prevalence was �0.57. These results are compatiblewith ASD being caused by either gestational vitamin D deficiencyand/or early childhood vitamin D deficiency.

The early childhood vitamin D deficiency theory of ASD could beeasily studied by seeing if toddlers who are supplemented withvitamin D have lower rates of ASD. That would explain why Egyp-tian children with ASD had Autism Rating Scales that were inver-sely and strongly (R = �0.84) associated with serum 25(OH)D[25]. It also means that autism is preventable with physiologicaldoses of vitamin D.

Very recently, vitamin D has been found to have a treatmenteffect in some autoimmune diseases such as multiple sclerosis[26–29]. Autism has some autoimmune phenomena and at leasttwo anti-neural autoantibodies are directly and strongly associatedwith autism severity [30,21]. It is important to remember that vita-min D has as many mechanisms of actions as genes it regulates.Vitamin D has anti-inflammatory actions and anti-autoimmune ef-fects; it increases the seizure threshold, increases the percentage ofT-regulatory cells, protects mitochondria from oxidation, upregu-lates glutathione, and upregulates at least five DNA repair proteins[31]. Thus, vitamin D may be the seasonal factor and thus the coresymptoms of ASD may respond to adequate doses of vitamin D thatare given in a timely manner.

Conflict of interest

Dr. Cannell is Executive Director of the Vitamin D Council andreceives remuneration from Purity Products.

References

[1] Fein D, Barton M, Eigsti IM, Kelley E, Naigles L, Schultz RT, et al. Optimaloutcome in individuals with a history of autism. J Child Psychol Psychiatry2013;54(2):195–205.

[2] Hayashi E. Seasonal changes in sleep and behavioral problems in a pubescentcase with autism. Psychiatry Clin Neurosci 2001;55(3):223–4.

[3] Montes G, Halterman JS. Association of childhood autism spectrum disordersand loss of family income. Pediatrics 2008;121(4):e821–6.

[4] Kogan MD, Strickland BB, Blumberg SJ, Singh GK, Perrin JM, van Dyck PC. Anational profile of the health care experiences and family impact of autismspectrum disorder among children in the United States, 2005–2006. Pediatrics2008;122(6):e1149–58.

[5] Wise MD, Little AA, Holliman JB, Wise PH, Wang CJ. Can state earlyintervention programs meet the increased demand of children suspected ofhaving autism spectrum disorders? J Dev Behav Pediatr 2010;31(6):469–76.

[6] Schieve LA, Rice C, Yeargin-Allsopp M, Boyle CA, Kogan MD, Drews C, et al.Parent-reported prevalence of autism spectrum disorders in US-born children:an assessment of changes within birth cohorts from the 2003 to the 2007national survey of children’s health. Matern Child Health J 2012;16(Suppl.1):S151–7.

[7] Hallmayer J, Cleveland S, Torres A, Phillips J, Cohen B, Torigoe T, et al. Geneticheritability and shared environmental factors among twin pairs with autism.Arch Gen Psychiatry 2011;68(11):1095–102.

[8] Losh M, Piven J. Social-cognition and the broad autism phenotype: identifyinggenetically meaningful phenotypes. J Child Psychol Psychiatry2007;48(1):105–12.

[9] Sasson NJ, Lam KS, Childress D, Parlier M, Daniels JL, Piven J. The broad autismphenotype questionnaire: prevalence and diagnostic classification. Autism Res2013;20.

[10] Cannell JJ. Autism and vitamin D. Med Hypotheses 2008;70(4):750–9.[11] Deluca GC, Kimball SM, Kolasinski J, Ramagopalan SV, Ebers GC. The role of

vitamin D in nervous system health and disease. Neuropathol Appl Neurobiol2013;21.

[12] Kocovska E, Fernell E, Billstedt E, Minnis H, Gillberg C. Vitamin D and autism:clinical review. Res Dev Disabil 2012;33(5):1541–50.

[13] Carlberg C, Seuter S, Heikkinen S. The first genome-wide view of vitamin Dreceptor locations and their mechanistic implications. Anticancer Res2012;32(1):271–82.

[14] Ginde AA, Liu MC, Camargo Jr CA. Demographic differences and trends ofvitamin D insufficiency in the US population, 1988–2004. Arch Intern Med2009;169(6):626–32.

[15] Kumar J, Muntner P, Kaskel FJ, Hailpern SM, Melamed ML. Prevalence andassociations of 25-hydroxyvitamin D deficiency in US children: NHANES2001–2004. Pediatrics 2009;124(3):e362–70.

[16] Cannell JJ. On the aetiology of autism. Acta Paediatr 2010;99(8):1128–30.[17] Dealberto MJ. Prevalence of autism according to maternal immigrant status

and ethnic origin. Acta Psychiatr Scand 2011;123(5):339–48.[18] Eyles DW, Vitamin D. And autism: Does skin colour modify risk? Acta Paediatr

2010;99(5):645–7.[19] Mostafa GA, Al-Ayadhi LY. Reduced serum concentrations of 25-hydroxy

vitamin D in children with autism: relation to autoimmunity. J Neuroinflamm2012;17(9):201.

[20] Rogers SJ. What are infant siblings teaching us about autism in infancy?Autism Res 2009;2(3):125–37.

[21] Walker RB, Conn JA, Davies MJ, Moore VM. Mothers’ views on feeding infantsaround the time of weaning. Public Health Nutr 2006;9(6):707–13.

198 J.J. Cannell / Medical Hypotheses 81 (2013) 195–198

[22] Hall HI, Jorgensen CM, McDavid K, Kraft JM, Breslow R. Protection from sunexposure in US white children ages 6 months to 11 years. Public Health Rep2001;116(4):353–61.

[23] Briefel R, Hanson C, Fox MK, Novak T, Ziegler P. Feeding infants and toddlersstudy: do vitamin and mineral supplements contribute to nutrient adequacyor excess among US infants and toddlers? J Am Diet Assoc 2006;106(1 Suppl.1):S52–65.

[24] Grant W, Cannell J. Autism prevalence in the United States with respect tosolar UV-B doses: an ecological study. Dermato-Endocrinol 2013;5:9–14.

[25] Mostafa GA, Al-Ayadhi LY. Reduced serum concentrations of 25-hydroxyvitamin D in children with autism: Relation to autoimmunity. J Neuroinflamm2012;17(9):201.

[26] Soilu-Hanninen M, Aivo J, Lindstrom BM, Elovaara I, Sumelahti ML, Farkkila M,et al. A randomised, double blind, placebo controlled trial with vitamin D3 as

an add on treatment to interferon b-1b in patients with multiple sclerosis. JNeurol Neurosurg Psychiatry 2012;83(5):565–71.

[27] Derakhshandi H, Etemadifar M, Feizi A, Abtahi SH, Minagar A, Abtahi MA, et al.Preventive effect of vitamin D3 supplementation on conversion of opticneuritis to clinically definite multiple sclerosis. Acta Neurol Belg 2012;19.

[28] Berlanga-Taylor AJ, Ramagopalan SV. Vitamin D and multiple sclerosis: what isthe clinical impact? Expert Opin Med Diagn 2013;9.

[29] Coimbra CG. Personal communication (4.22.2013).[30] Mostafa GA, Al-Ayadhi LY. The relationship between the increased frequency

of serum antineuronal antibodies and the severity of autism in children. Eur JPaediatr Neurol 2012;16(5):464–8.

[31] Cannell J, Grant W. What is the role of vitamin D in autism? Dermato-Endocrinol 2013;5:18–9.