Autism and Vitamin D by John J. Canneii, MD Any theory of autism's etiology must take into account its strong genetic basis, but also explain how genetics interacts with the environment to produce autism's unusual epidemiology. Activated vitamin D, calcitriol, is a potent neurosteroid hormone with critical roles in mammalian brain development. Calcitriol's physiology is unique among the steroid hormones, because normal steroid feedback kinetics does not regulate neural levels of calcitriol; human behavior does so. The apparent dramatic increase in the prevalence of autism and other neurodevelopmental disorders over the last 20 years corresponds with the medical profession's increasing advice to avoid the sun during that same time, advice that may have dramatically lowered brain calcitriol levels. Estrogen and testosterone have very different effects on calcitriol's metabolism, differences that may explain the striking male/female sex ratios in autism. Cognitive abilities are positively associated with vitamin D levels, even after correction for sun exposure. Autistic individuals have immunological abnormalities that are similar to those produced by vitamin D deficiency Calcttrioi downregulates production of inflammatory cytokines in the brain, which have consistently been associated with cognitive impairment. Severe maternal vitamin D deficiency leads to rat pups with increased brain size and enlarged ventricles, abnormalities similar to those found in autistic children. Vitamin D deficiency impairs glutathione metabolism, which may explain the link between oxidative stress, mercury accumulation, and autism. Multivitamins containing vitamin D reduce symptoms of autism and increase cognitive abilities in children. Consumption of vitamin D-containing fish during pregnancy reduces autistic symptoms in children. Most studies show autistic births are higher in the winter, especially in March, when vitamin D levels are their lowest.
TOWNSEND LETTER - APRIL 2008
Neurodevelopmental abnormalities appear to be higher among those with low vitamin D levels, such as African Americans, and maternal vitamin D deficiency is exceptionally common, especially among
"Always observe the physician with the same diligence as he the disease." John Donne(1623) African Americans, regardless of prenatal vitamin use. While future epidemiological, genetic, and interventional studies should test this theory, physicians should diagnose and treat vitamin D deficiency during pregnancy and childhood now. Introduction Arguably, the five most striking epidemiological aspects of autism are its high monozygotic (407o-90%) vs. dizygotic (0% -10%) twin concordance rates,' widely varying phenotypic expression even among monozygotic twins,^ striking male/female ratio (-3:1}, apparent increased prevalence in African Americans, and apparent rapid increase in incidence rates over the last 20 years. Whatever its genetic roots, and they are strong, autism hardly follows classic Mendelian inheritance. A useful theory of causation must explain all these discrepant epidemiological observations - and should do so parsimoniously. When a disease with strong genetic roots displays such peculiar epidemiology, it is reasonable to seek an explanation amongenvi ronmental genetic contri butors. While the predisposing autistic lesion is genetic, the above epidemiological observations indicate the environment is dramatically affecting phenotypic expression. That is, the environment
is altering the brain, epigenetically, or through gene-environment interactions. Environmental genetic contributors directly influence the genome, but the environment directly influences them as well, the neurosteroid hormones being one example. Furthermore, if current claims of increasing prevalence rates of autism and other neurodevelopmental disorders over the last 20 years (Figure 1)' are due to some actual increases in incidence and this seems increasingly likely^ - then it is reasonable to search for neurosteroids that have changed over the same time autism has increased. Furthermore, if a neurosteroid exists that profoundly affects brain development, whose metabolism is different in males and females, whose levels are lov^^er in African Americans than Whites, whose levels are directly affected by diet or human behavior, and whose brain levels have dramatically decreased in the last 20 years, surely that neurosteroid is a prime causation candidate. In a recent article discussing aulism and genetics, Herbert et :329-34. 100. Bodnar LM, Catov |M, Roberts )M. Racial/ethnic differences in the monthly variation oí preecUmpsia incidence. Am / Obstel Gyneco/. 2007 Apr;l96(4l:324.e1-5. 101. Scholl TO, et al. Use of m ultivitam in/mineral prenatal supplements; influence on the outcome of pregnancy, Am / fpidemiol. 1997|ul 15;146(21:1 34-41. tO2. Vahratian A, et al. Multivitamin use and the risk of preterm birth. Am / Epidemio/, 2004 Nov 1;160(91:886-92, 103. Mardones-Santander F, et al.Effect of a milk-based food supplement on matemal nutritional status and fetal growth m underweight Chilean women. Am I Clin Nutr. 1988 Mar;4 7(31:413-9, 104. Marya RK. Rathee S. Lata V, Mudgil S. Effects of viUmin D supplementation in pregnancy. Gyneco/ Ob.stet Invest. 1981;12(31:155-61. 105. Bodnar LM. et al. High prevalence of vitamin D insufficierKy in black dr>d white pregnant women residing in the rtorthern United States and their neonates. / Nufr. 2O07Feb;137(2l:447-52. 106. Ziegler EE. Hollis 8W, Nelson SE. leter |M. Vilamin D deficiency m breastfed infants in Iowa, Pediatrics. 2006 Aug;! 18(21:603-10. 107. Hollis BW, Wagnei CL. Vitamin D requirements during lactation: high-dose matemal supplementation as therapy to prevent hypovitaminosis D for both the mother and the nursing infant. Arn Í Clin Nutr. 2004 Dec;80(6 108. Nesby-O'Dell S. et al. Hypovitaminosis D prevalence and determinants among African Anf>erican and white women of reproductive age: third National Health and Nutrition Examination Survey, 1988-1994, Am I Clin Nutr. 2002 |ul;76(l 1:187-92. 109. lips P, et al. A global stucty ol vitamin D status and parathyroid function in postmenopausal women with osteoporosis: baseline dala from the multiple outcomes of raloxifene evaluation clinical trial. 1he louinal oi Clinical indocrinology and Metaboli'tm. 2O01;86: 1212-1221. n o . Heaney RP, et al. Calcium absr>rption varies wrihin the reference range for serum 2 5-hydroxy vitamin D lovtnil oi the American College oí Nutrition. 2003; 22: 142-146. t i l . Lappe |M, et al. Vitamin D and calcium supplementation reduces cancer risk: results of a randomized Irial. American lournai of Clinical Nutrition. 2007, in press. 112.flischoff-Ferrari HA. et al. Higher 25-hydro)iyvitamin D concentrations are associated with better lovwr-extremity function in berth active and inaclive persons aged > or - 6 0 y. American lournai oi Clinical Nutrition. 2004; 80: 752-758. 113. Heaney RP. The Vilamin D requirement in health and disease. / Steroid Biochem Mot Biol. 2t3O5 Jul 15; [Epub ahead of print). 114. Vieth R. What is the optimal vitamin D status for healthf Prog fliophys Mo/Bio/. 2006 Sep;92(1|:26-32. 115. Holick MF. The vitamin D epidemic and its heallh consequences. ( Nut/. 2005 Nov; 135(11):2739S-4aS.
TOWNSEND LETTER - APRIL 2008
