Do salt cravings in children with autistic

disorders reveal low blood sodium

depleting brain taurine and glutamine?

reprinted (lightly revised) from

Medical Hypotheses 2011;77:1015–1021


Because boys are four times more likely than girls to develop autism, the role of male hormones (androgens) has received considerable scrutiny. Some researchers implicate arginine vasopressin, an androgen-dependent hormone from the pituitary that elicits male behavior. Elevated vasopressin is also the most common cause of low blood sodium (hyponatremia) – most serious in the brains of children. Hyponatremia causes astrocytes to swell, then release amino acids taurine and glutamine and their water to compensate. Taurine was the amino acid most wasted or depleted in urine of autistic children. Glutamine is a critical metabolic fuel in brain neurons, astrocytes, endothelial cells, and the intestines, especially during hypogly-cemia. Because glutamine is not thought to cross the blood–brain barrier significantly, the implications of low blood glutamine in these children are not recognized. Yet children with high brain glutamine from urea cycle disorders are rarely diagnosed with autistic disorders. Other common events in autistic children that release vasopressin are gastrointestinal inflammation, hypoglycemia, and stress. Signs of hyponatremia in these children are salt cravings reported online and anecdotally, deep yellow urine revealing concentration, and relief of autistic behavior by fluid/salt diets. Several interventions offer promise: (a) taurine to suppress vasopressin and replenish astrocytes; (b) glutamine as fuel for intestines and brain; (c) arginine to spare glutamine, detoxify ammonia, and increase brain blood flow; and (d) oral rehydration salts to compensate dilutional hyponatremia. This hypothesis appears eminently testable: Does your child crave salt? Is his urine deep yellow?


I think we need to conduct research as if we know this is an emergency.

Martha Herbert, 2009 [1].

High levels of ammonia in the blood of children with autistic disorders (ASD) were first detected in the early 1980s [2]. Cohen subsequently found high plasma ammonia in one autistic boy [3]; Filipek et al. detected significantly elevated plasma ammonia in a majority of 100 autistic children [4]. Wakefield et al. suggested their diseased intestines generate more ammonia than their impaired liver can clear, which enters the brain, provoking a form of hepatic encephalopathy [5].

Most ammonia (NH3) is generated as a byproduct of bacteria digesting proteins in the large intestine, and degrading the amino acid glutamine in the small intestine [6]. The liver uses the amino acid arginine to detoxify ammonia to urea, excreted in urine [7]. Ammonia that reaches the brain is trapped by astrocytes combining ammonia and the amino acid transmitter glutamate to form glutamine, with no transmitter activity but much osmotic activity [8]. Excessive glutamine causes astrocytes to swell, inducing compensatory release of the amino acid taurine and other osmolytes and their water to restore normal cell volume [9]. Persons with high brain ammonia from liver disease (e.g. cirrhosis) or inborn urea cycle disorders (UCD) have high concentrations of glutamine in astrocytes [10], inducing swelling and intracranial pressure, e.g. hepatic encephalopathy [8].

Plasma ammonia concentrations in children with UCD are five times greater than plasma ammonia in liver failure [11], elevating brain glutamine and impairing cognition. Yet UCD children are rarely diagnosed with autistic disorders or mood disorders [12–14]. This striking observation corroborates Wakefield’s speculation that glutamine is low in autistic brains [5] – based on low serum levels [15] and evidence that liver dysfunction impairs the astrocyte glutamate transporter. But if glutamine is low in autistic brains, why are astrocytes swollen [16]? Ammonia and other neurotoxins implicated in autistic disorders, e.g. mercury and arsenic, cause astrocytes to swell [17,18]. A more productive explanation may be chronic hyponatremia (low blood sodium), which drives water into astrocytes, inducing compensatory release of taurine, glutamine, and their water [19]. Why would children with autistic disorders be chronically hyponatremic? One obvious explanation is recurring diarrhea [20]. Another is high concentrations of water-conserving arginine vasopressin (AVP) [21]. A third is depletion of taurine, the inhibi-tory neurotransmitter that suppresses vasopressin [22].



[C]hildren in this cohort [urea cycle disorders] show other behavioral/emotional strengths, including a minimal percentage with previous diagnoses of Autism spectrum disorders, mood disorders, and other psychiatric disorders. Krivitzky et al. 2009 [12]

Low concentrations of the amino acid glutamine detected in plasma and platelets of children with autistic disorders [23–27] are not thought to affect brain glutamine concentrations because glutamine does not cross the blood–brain barrier significantly. Yet a sodium-dependent, concentration gradient-dependent transporter has been identified that moves glutamine from blood into astrocytes [28]. Glutamine is alternative fuel in mitochon-dria of brain neurons and astrocytes, especially during hypoglycemia [29], and primary fuel in rapidly replicating cells, e.g. blood vessel endothelial cells, intestinal enterocytes, and tumors [30,31]. This evidence invites several speculations: (a) low blood glutamine in autistic disorders may reflect brain glutamine concentrations, and (b) glutamine depletion may explain brain and gut malfunctions in these children.

Distinguishing brain glutamine from glutamate by magnetic resonance spectroscopy (MRS), however, may require an ultra-high (7 Tesla) field. DeVito et al. detected reduced concentrations of combined glutamate/glutamine in the cortex and cerebellum of autistic boys [32]. Bernardi et al. reported reduced glutamate/glutamine in the right anterior cingulate cortex of high-functioning adults with autistic disorders [33]. Page et al. found glutamate/glutamine elevated in the amygdala-hippocampal region of autistic adults, and normal in the parietal region [34].

Magnesium is required to convert glutamate to glutamine [29], which may partly explain why magnesium/vitamin B6 supplements have helped almost half of ASD persons tested [35]. The effectiveness of risperidone (Risperdal), which helped 54% of ASD persons (but aggravated 20%) [35], also suggests brain glutamine may be low. Risperidone is thought to suppress serotonin and dopamine activity at synapses, but also stimulates glutamate uptake by astrocytes and activity of glutamine synthetase, the enzyme that converts glutamate and ammonia to glutamine [36].

The most compelling evidence for low brain glutamine in autistic disorders, however, must be the notable lack of autistic behavior in children with high brain glutamine from urea cycle disorders (UCD) [12,13]. But if brain glutamine is low in ASD children, why are their astrocytes swollen? One explanation may be an innate immune response [16]. Various neurotoxins implicated in autism (ammonia, mercury, arsenic) also cause astrocytes to swell [17,37]. Gabis et al. detected by MRS high concentrations of the compensatory osmolyte myoinositol in ASD brains, usually signifying astrocyte proliferation [38]. The most intriguing explanation for swollen astrocytes in autistic disorders – and low brain glutamine – must be chronic hyponatremia, which drives water into astrocytes, provoking compensatory release of taurine, glutamine, and their water [19]. Why would autistic children be chronically hyponatremic? Not only is diarrhea a recurring problem [20], these children appear to have high levels of arginine vasopressin [21,39] – most common cause of hyponatremia [40].


If a substance can cross a plasma membrane, it cannot exert an osmotic pressure across that membrane. The solute will equilibrate across the mem-brane instead of forcing water to move .... Heins & Zwingmann 2010 [41].

Because sodium ions (Na+) cannot cross cell membranes freely, and sodium is the most abundant ion in blood plasma and extracellular fluid (ECF), abnormal sodium concentrations in these fluids force water in or out of cells. High blood sodium (hypernatremia) draws water out of cells; low blood sodium (hyponatremia) drives water into cells. Hyponatremic enceph-alopathy arises when low blood sodium forces water into brain astrocytes, which swell and exert pressure. Children are most vulnerable to hyponatremic encephalopathy because their brains grow faster than their skulls, leaving less room for expansion; a child’s brain is adult-sized by six years, his skull not adult-sized until 16 years [40]. Autistic children have enlarged brains within a few months after birth [42], presumably aggra-vating the risk of hyponatremic encephalopathy.

But why would children with autistic disorders be chronically hyponatremic? Two common features of ASD are recurring diarrhea [20], and high levels of male hormones (androgens), to judge from hypermasculinization, preco-cious puberty, and other indications [43]. Geier and Geier found pre-pubertal boys and girls with ASD had clinical signs of hyperandrogenicity, and high concentrations of dehydroepiandrosterone (DHEA) and testos-terone in blood [44]. DHEA, a steroid precursor from the adrenal cortex, can form androgens androstenedione and testosterone, they pointed out, or be sulfated to the ‘‘normally favored storage molecule’’ DHEA sulfate (DHEAS) — major precursor of placental estrogens during pregnancy [45]. Impaired sulfation in these children causes more DHEA to become androgens and less to become DHEAS, they proposed [44] — a plausible explanation for the “extreme male brain” of autism [43,46].

Androgens stimulate the pituitary gland to release arginine vasopressin (AVP) [39,47], which causes the kidneys to reabsorb water. Diarrhea depletes sodium and water, inducing dehydration hyponatremia; vaso-pressin conserves water, inducing dilutional hyponatremia [11,48]. Moritz and Ayus [40]: “Most cases of hyponatremia are the result of increased AVP production ....” Momeni et al. detected high vasopressin in autistic disorders [21], which Emanuele suggested might signify vasopressin resistance [49]. Non-osmotic events that elevate vasopressin include hypoglycemia, gastroenteritis (diarrhea and vomiting), and stress [48] – common events in autistic children.

Symptoms of mild hyponatremia are headache, nausea and vomiting, lethargy, weakness, confusion, altered consciousness, agitation, and gait disturbances; severe hyponatremia can lead to seizures, coma, and death [40]. Koide concluded hyponatremia should be part of the differential diagnosis of convulsive seizures in autism [50]. Hiratani et al. [51] pointed out “water intoxication should always be considered when an autistic patient shows recurrent epileptic attacks or episodic strange behaviors with hyponatremia.”

The brain quickly responds to swelling and pressure from hyponatremia by extruding extracellular fluid into cerebrospinal fluid, and releasing potas-sium and sodium and their water. Release of organic osmolytes – myoinositol and amino acids – occurs slowly over the next few days because transporters first need to be synthesized [11]. Massieu et al. reported the primary amino acids released from the brain by chronic hyponatremia are taurine, glutamine, glutamate, and aspartate [19]. Because taurine suppres-ses vasopressin release [22], does taurine depletion make dilutional hyponatremia chronic?


[T]he degree of hyponatremia in preterm neonates is predictive of both increased salt appetite and dietary sodium intake in adolescence .... Moritz & Ayus 2009 [52]

Salt cravings in children with autistic disorders have been reported online [53,54] and anecdotally, though not in the medical literature. Primary causes of salt cravings are salt wasting, overhydration, and dehydration [55,56]. Dehydration inducing salt appetite seems counterintuitive, but salt helps dehydrated persons (often 3rd World children with severe diarrhea) expand blood volume by retaining water [55]. An unpublished study of taste sensitivity in autistic children found diminished sensitivity to salt and sucrose, hypersensitivity to bitter and sour [20]. Hyposensitivity induces overconsumption, thus often signifies deficiency [57]. Increased salt appe-tite has been reported in human adrenal insufficiency, where much sodium is lost in urine [56].

Hyponatremic encephalopathy aggravates hepatic encephalopathy

Experimental studies have shown that chronic hyponatremia exacerbates brain edema induced by ammonia.... It is plausible that any change in the state of cell hydration induced by a decrease in extracellular sodium or an increase in intracellular glutamine activates compensatory mechanisms and increase[s] the risk of astrocyte swelling in front of an additional osmotic stress.  Cordoba et al. 2010 [11].

Cordoba et al. reported rats made hyponatremic showed decreases of all three organic osmolytes in the brain – glutamine, myoinositol, and taurine. Elevating ammonia in rats also decreased brain myoinositol and taurine, but increased brain glutamine. Ammonia and hyponatremia increased brain glutamine but not as much as ammonia alone [58]. Heins and Zwingmann confirmed these findings [41]. Because accumulation of glutamine and its water in astrocytes is thought to be the decisive deficit in hepatic encephal-opathy [8], why would hyponatremia releasing glutamine from astrocytes aggravate encephalopathy? Cordoba et al. concluded loss of intracellular osmolytes via hyponatremia diminished the capacity of astrocytes to compensate further osmotic loads [11,58].


Pangborn found taurine was the amino acid most wasted or depleted in urine of autistic children [59]. Astrocytes swollen by hyponatremia or glutamine release taurine and its water to relieve swelling [9,60]. If children with urea cycle disorders have high brain glutamine and low taurine, and children with autistic disorders have low brain glutamine and low taurine, glutamine presumably protects against autistic disorders more than taurine. Yet taurine released by specialized astrocytes (pituicytes) suppresses vasopressin. Sensing low ion concentrations in ECF, these astrocytes release taurine to inhibit calcium (Ca2+) fluxes into neurons that release vasopres-sin. Hussy et al. concluded taurine depletion ‘‘abolishes the osmo-dependent inhibition of vasopressin release.’’[22].

Taurine is a critical brain osmolyte and inhibitory transmitter, regulator  of active intracellular Ca2+ [61], and magnesium complement [62]. Taurine is most vulnerable to abbreviated breastfeeding [63], dietary deficiencies of its precursors methionine and cysteine [64], impaired synthesis from deficiency of bioactive vitamin B6 (pyridoxal phosphate) [59], and preemp-tory requirements for sulfate and glutathione under toxic/oxidant attack. Mother’s milk is rich in taurine; cow’s milk is low after calves are weaned [63]. Many mothers of autistic children breastfed one week [65]. Schultz found longer breastfeeding was associated with a decreased likelihood of developing autism [66]. The Autism Research Institute recommended 250–500 mg/day of taurine for children, up to 2 g/day for adults and adult-sized children [67].


Another critical amino acid in autistic disorders is arginine – required to detoxify ammonia to urea in the liver, for arginine vasopressin, and as only substrate for nitric oxide (NO). High levels of nitrite in plasma of these children may reflect inducible nitric oxide responding to intestinal infection [68], depleting arginine as substrate for endothelial and neuronal nitric oxide, the brain’s primary vasodilators. Carrick and Carrick reported oral arginine calms unstable emotions and improves sociability dramatically in their adult son with autism [69]. Children with urea cycle disorders take supplemental arginine or its precursor citrulline to help the liver detoxify ammonia [12].


[I]t is possible that changes in mood and appetite are among the first noticeable manifestations accompanying sodium deficiency. Morris et al. [56].

High arginine vasopressin in autistic disorders [21] has not been confirmed, although suspected from high androgens and effects of AVP on male behavior [39,47]. Nor have swollen astrocytes [16] been confirmed, although they appear proliferated [38]. Swollen astrocytes compressing brain capillaries were proposed to explain low brain blood flow in ASD [70] and the benefit of fever [71]. Interestingly, infectious fever reduces appetite (anorexia), provoking rapid release of alanine and glutamine from skeletal muscles as metabolic fuel [72]. Glutamate/glutamine was normal-to-high by MRS in some ASD brains [34], low in others [32,33]. Plasma or urinary glutamine was high in ASD children with high blood ammonia [59]. Norenberg et al. contended excessive glutamine in brain mitochondria degrades to ammonia, impairing energy metabolism and the sodium pump [73]. Clearance of vasopressin from blood and degradation by the liver is impaired in cirrhosis [74]. Chronic stress [75] presumably elevates AVP. In summary, considerable evidence supports the hypothesis, though the key observation – high levels of arginine vasopressin in autistic children – remains unconfirmed.


Two stubborn questions vex autism researchers and theorists [76]: (a) why is autistic behavior recognizable in many other disorders (non-specificity), and (b) why does autistic behavior look different in every child (hetero-geneity)? Many disorders show autistic behavior: ADHD, obsessive–compulsive disorder, language impairments, epilepsies, digestive, allergic, and immune disorders, metabolic disorders, toxic and infectious diseases, and genetic syndromes [76].

Heterogeneity complicates the search for interventions that help most children, not only some. Magnesium/vitamin B6, for example, one of the most effective, improves behavior in only about half of ASD persons tested [35]. This suggests subsets of autistic persons with characteristic metabolic deficits underlying their distinctive behavior. Yet although each autistic child’s behavior is unique, it is also recognizably autistic – why autistic behavior is obvious in other disorders.

One characteristic feature is that autistic behavior often remits, sometimes dramatically, under certain conditions, notably infectious fever and perceived emergencies [70,71]. Intriguingly, diets restricted to clear fluids (liquids and salt) a day or two before medical procedures relieved autistic behavior transiently [77]. These clues reveal autistic behavior is more reversible than structural – more “state” than “trait” [76]. Brain malfunctions that are not structural are called metabolic encephalopathies [11], two now familiar examples are hepatic encephalopathy from high brain ammonia and glutamine, reversible “up to a point” [8], and hyponatremic encephalopathy from salt wasting or overhydration [11]. Hyponatremia accompanies many illnesses [40,78]; is that why autistic behavior does also?

Another conundrum is the apparent coexistence of hyponatremia and high blood ammonia in ASD. Chronic hyponatremia reduces brain glutamine; ammonia accumulates it [11]. If ammonia becoming glutamine in the brain protects children with UCD from autism, why doesn’t it protect children with autism? One explanation is that hyponatremia depleting brain glutamine is more severe in autistic disorders than ammonia accumulating glutamine. Wakefield suspected the conversion of ammonia to glutamine was impaired in autistic disorders, so not all ammonia that reached the brain became glutamine [5].

If high brain glutamine protects against autistic behavior, oral arginine relieves autistic behavior dramatically (as does fever), fluid/salt diets relieve autistic behavior transiently, and taurine and glutamine are depleted, what links these vectors? First, hyponatremia depletes brain taurine and glutamine. Second, taurine depletion deregulates vasopressin. Third, transporters rely on sodium concentration gradients to move taurine and glutamine into brain cells against their concentration gradients [28]. Although arginine transport into the brain is largely independent of sodium gradients [28], supplemental arginine spares glutamine. Wu et al. [79]:

“[W]hen dietary levels of arginine are high, intestinal synthesis of citrulline from glutamine and glutamate may be inhibited for sparing of glutamine and glutamate for other metabolic pathways.”

One intriguing explanation for the dramatic benefit of fever is the proposal by Myers and Veale that the ratio of calcium ions to sodium ions in the hypothalamus determines the set point for body temperature. When fever raises the set point, brain Ca2+ levels fall and Na+ levels rise. This Ca2+/Na+ ratio is established prenatally [80]. Swollen astrocytes compressing brain capillaries [17] were proposed to explain low brain blood flow in autistic disorders [70,71]. Yet astrocytes in urea cycle disorders are more swollen (by glutamine) than astrocytes in autistic disorders, without provoking autistic behavior. Do arginine/citrulline supplements to detoxify ammonia in UCD children restore brain blood flow (via nitric oxide) in vessels compromised by swollen astrocytes? But then children with UCD might present with autistic behavior. A better explanation for low brain blood flow may be vasoconstriction by vasopressin [40].

In this scenario of vasopressin-induced chronic hyponatremia, what provokes autistic regression? Horvath and Perman reported autistic behavior usually first appeared between 12 and 18 months – about the time gastrointestinal symptoms (unless present at birth) first appeared [20]. Diarrhea and vomiting (gastroenteritis) deplete sodium and release vaso-pressin. But if gastroenteritis was present at birth, why no autistic behavior for a year or more? Multiple vaccinations between 12 and 18 months are undoubtedly stressful, elevating vasopressin. Acetaminophen for vaccine reactions may impair the liver’s clearance of vasopressin – and underlie the autism epidemic [46,81]. Intestinal colonization by harmful bacteria initiated by amoxicillin/clavulanate (Augmentin) [82–84], an oral antibiotic introduced in 1981, may also explain why the epidemic began in the early 1980s [59].

One might first simply ask: Does your child crave salt? Deep yellow urine may reveal concentration by vasopressin. Sodium, glutamine, taurine, arginine, and vasopressin may be measured in urine, blood, and in brain by MRS. Glutamine and taurine supplements may gradually replenish brain concentrations of these critical amino acids; oral arginine may help more dramatically [69]. Watermelon, rich in arginine and its precursor citrulline, may be a test and remedy for arginine depletion [85].

Treating hyponatremia is more complicated. Too rapid correction (via concentrated intravenous saline) has provoked osmotic demyelination in the brain. Attempts to suppress androgens or vasopressin seem inherently risky. Two less invasive strategies: (a) oral rehydration salts (ORS) (essentially salt and glucose) to compensate dilution by vasopressin [48]; and (b) supplemental taurine to suppress vasopressin physiologically [22]. ORS formulated to World Health Organization (WHO) specifications (to treat diarrhea in 3rd World children) are readily available [86]. Bardhan recom-mended adding glutamine to ORS [87]: ‘[G]lutamine is able to promote the absorption of sodium and water, even more effectively than glucose.’’ Although tumors consume glutamine avidly, glutamine supplements do not stimulate their growth [30]. Glutamine degrades to ammonia in the small intestine – a problem when blood ammonia is high, but perhaps useful when brain glutamine is low.

The clearly beneficial role of [glutamine] when its level in the CNS remains within the physiological range may turn – and possibly does turn – into primarily detrimental when its production under hyperammonemic conditions exceeds its normal consumption. Albrecht et al. 2007 [60].

In conclusion (see Table 1 and Fig. 1), arginine vasopressin appears chronically elevated in children with autistic disorders, because of high androgens, recurring gastrointestinal inflammation, hypoglycemia, and stress, and depletion of astrocyte taurine – primary suppressor of vasopressin. Vasopressin conserves water at the kidneys, diluting sodium concentrations in blood (hyponatremia) and extracellular fluid. Diarrhea and vomiting in these children deplete sodium directly; mothers may be sodium-deficient. Hyponatremia drives water into astrocytes, provoking compensatory release of taurine, glutamine, and their water. Hyponatremia also reduces sodium gradients that transport taurine and glutamine into the brain. Dilutional hyponatremia by vasopressin may eventually become chronic by depleting its major inhibitor taurine.

Excessive androgens and taurine depletion at an early age may predispose to elevated vasopressin and chronic hyponatremia throughout childhood. Depletion of brain glutamine by hyponatremia may be equally decisive, to judge from glutamine’s protection against autistic behavior. Glutamine is prominent fuel in brain neurons, astrocytes, endothelial cells, and the intestines, especially during hypoglycemia. If low blood glutamine in these children reflects intestinal and brain glutamine, brain and gut malfunctions may be explained. The most promising interventions: taurine to replenish astrocytes and suppress vasopressin; glutamine to fuel intestines and brain; arginine/citrulline to spare glutamine, detoxify ammonia, and generate nitric oxide; and oral rehydration salts to compensate chronic hypona-tremia.


1. Children with autistic disorders have high androgen levels [43,44] which stimulate release of arginine vasopressin from the pituitary gland [21,39,47]

2. Excessive production of water-conserving vasopressin is the most common cause of hyponatremia (low blood sodium) [40]

3. Hyponatremia forces water into astrocytes, inducing compensatory release of taurine, glutamine, and their water to relieve swelling [11,19]. Children are more vulnerable to the brain swelling of hyponatremia because their brains grow faster than their skulls [40]. Autistic brains are enlarged soon after birth [42]

4. Taurine, primary osmotic inhibitor of vasopressin [22], was the amino acid most wasted or depleted in urine of autistic children [59]

5. Vasopressin is also released by non-osmotic events (gastroenteritis, hypoglycemia, stress) [48] common in these children

6. Brain glutamine and taurine transporters rely on the energy of sodium concentration gradients [28]

7. Children with autistic disorders have low blood glutamine [23–27] which may reflect brain glutamine [28,32,33]

8. Glutamine is a critical metabolic fuel in brain mitochondria [29] and the intestines [30]

9. Children with urea cycle disorders and high brain glutamine are rarely diagnosed with autistic disorders or mood disorders [12,13]

10. Salt cravings in autistic children have been reported online [53,54] and anecdotally. One mother was salt-deficient as a child

11. Fluid/salt diets for one or two days relieved autistic behavior transiently [77]

12. Vasopressin constricting cerebral vessels [40] may explain low brain blood flow in autistic disorders [70]

Fig. 1. A pathogenic scenario. Androgens may be high in these children before and after birth; mothers may be sodium-deficient, their infants taurine-deficient. Androgens provoke release of arginine vasopressin (AVP) from the pituitary, as do diarrhea/vomiting, stress, and hypoglycemia. AVP conserves water (H2O) at the kidneys, diluting sodium (Na+) in blood (hyponatremia) and extracellular fluid. Hyponatremia itself causes symptoms by reducing sodium gradients, expanding extracellular space, and driving H2O into astrocytes and myelin. Hyponatremia driving H2O into astrocytes also provokes compensatory release of taurine (TAU), glutamine (GLN), and H2O. TAU – primary brain osmolyte and inhibitory transmitter – suppresses calcium (Ca2+) fluxes into neurons that release AVP. Thus AVP, by depleting its primary suppressor TAU, becomes deregulated over time. TAU depletion also deregulates intracellular Ca2+ and magnesium (Mg2+). GLN depletion by chronic hyponatremia starves brain neurons, astrocytes, and endothelial cells. AVP constricting cerebral arteries/arterioles reduces brain blood flow.


I’m most grateful to James Harduvel of the Deschutes County Library in Bend, Oregon, for dedicated retrieval of the literature; Martha Herbert, who inspired this study; Jon Pangborn of ARI, for clues to the problem of ammonia; Eugene Kiyatkin of NIH, for literature and encouragement; William Ellis, for helping spread the word; and Helen Emily Couch, for everything else. Special thanks also to Jeff Bhavnanie of

My apologies to anyone offended by the word autistic to describe these children. Autistic describes behavior, not nature – reversible state, not static trait. If autistic behavior reverses, what is autism?


[1] Herbert M. The future of autism research. Keynote address at the Autism Society’s 40th National Conference on Autism Spectrum Disorders, July 23, 2009. <> (accessed 7/3/11).

[2] Pangborn JB. Personal communication 2010.

[3] Cohen BI. Infantile autism and the liver: a possible connection. Autism 2000;4:441–2.

[4] Filipek PA, Juranek J, Nguyen MT, Cummings C, Gargus JJ. Relative carnitine deficiency in autism. J Autism Dev Disord 2004;34:615–23.

[5] Wakefield AJ, Puleston JM, Montgomery SM, Anthony A, O’Leary JJ, Murch SH. Review article: the concept of entero-colonic encephalopathy, autism and opioid receptor ligands. Aliment Pharmacol Ther 2002;16:663–74.

[6] Williams R. Review article: bacterial flora and pathogenesis in hepatic encephalopathy. Aliment Pharmacol Ther 2006;25(Supp 1):17–22.

[7] Milner JA. Metabolic aberrations associated with arginine deficiency. J Nutr 1985;115:516–23.

[8] Brusilow SW, Koehler RC, Traystman RJ, Cooper AJ. Astrocyte glutamine synthetase: importance in hyperammonemic syndromes and potential target for therapy. Neurotherapeutics 2010;7:452–70.

[9] Benarroch EE. Neuron-astrocyte interactions: partnership for normal function and disease in the central nervous system. Mayo Clin Proc 2005; 80:1326–38.

[10] Felipo V, Butterworth RF. Neurobiology of ammonia. Prog Neurobiol 2002;67:259–79.

[11] Córdoba J, García-Martinez R, Simón-Talero M. Hyponatremic and hepatic encephalopathies: similarities, differences and coexistence. Metab Brain Dis 2010;25:73–80.

[12] Krivitzky L, Babikian T, Lee HS, Thomas NH, Burk-Paull KL, Batshaw ML. Intellectual, adaptive, and behavioral functioning in children with urea cycle disorders. Pediatr Res 2009;66:96–101.

[13] Anonymous MD. Personal communication 2011.

[14] Görker I, Tüzün U. Autistic-like findings associated with a urea cycle disorder in a 4-year-old girl. J Psychiatry Neurosci 2005;30:133–5.

[15] Anthony A, Waring R, Murch SH, Wakefield AJ. Glutamine and glutamate:glutamine ratios in children with regressive autism and enterocolitis: preliminary evidence for an entero-colonic encephalopathy. Gastroenterology 2001;120(Suppl.):A726 [abstract only].

[16] Vargas DL, Nascimbene C, Krishnan C, Zimmerman AW, Pardo CA. Neuroglial activation and neuroinflammation in the brain of patients with autism. Ann Neurol 2005;57:67–81.

[17] Aschner M, Allen JW, Kimelberg HK, LoPachin RM, Streit WJ. Glial cells in neurotoxicity development. Annu Rev Pharmacol Toxicol 1999;39:151–73.

[18] Lonsdale D, Shamberger RJ, Audhya T. Treatment of autism spectrum children with thiamine tetrahydrofurfuryl disulfide: a pilot study. Neuro Endocrinol Lett 2002;23:303–8.

[19] Massieu L, Montiel T, Robles G, Quesada O. Brain amino acids during hyponatremia in vivo: clinical observations and experimental studies. Neurochem Res 2004;29:73–81.

[20] Horvath K, Perman JA. Autism and gastrointestinal symptoms. Curr Gastroenterol Rep 2002;4:251–8.

[21] Momeni N, Nordström BM, Horstmann V, Avarseji H, Sivberg BV. Alterations of prolyl endopeptidase activity in the plasma of children with autistic spectrum disorders. BMC Psychiatry 2005;5:27–32.

[22] Hussy N, Brès V, Rochette M, et al. Osmoregulation of vasopressin secretion via activation of neurohypophysial nerve terminals glycine receptors by glial taurine. J Neurosci 2001;21:7110–6.

[23] Rolf LH, Haarmann FY, Grotemeyer KH, Kehrer H. Serotonin and amino acid content in platelets of autistic children. Acta Psychiatr Scand 1993;87: 312–6.

[24] Moreno-Fuenmayor H, Borjas L, Arrieta A, Valera V, Socorro-Candanoza L. Plasma excitatory amino acids in autism. Invest Clin 1996;37: 113–28.

[25] Zavala M, Castejón HV, Ortega PA, Castejón OJ, Marcano de Hidalgo A, Montiel N. Imbalance of plasma amino acids in patients with autism and subjects with attention deficit/hyperactivity disorder [article in Spanish, English abstract]. Rev Neurol 2001;33:401–8.

[26] Aldred S, Moore KM, Fitzgerald M, Waring RH. Plasma amino acid levels in children with autism and their families. J Autism Dev Disord 2003;33:


[27] Tirouvanziam R, Obukhanych TV, Laval J, et al. Distinct plasma profile of polar neutral amino acids, leucine, and glutamate in children with autism spectrum disorders. J Autism Dev Disord 2011. [Epub ahead of print].

[28] Bode BP. Recent molecular advances in mammalian glutamine transport. J Nutr 2001;131:2475S–85S.

[29] Stelmashook EV, Isaev NK, Lozier ER, Goryacheva ES, Khaspekov LG. Role of glutamine in neuronal survival and death during brain ischemia and hypoglycemia. Int J Neuroscience 2011. [Epub ahead of print].

[30] Souba WW. Glutamine: a key substrate for the splanchnic bed. Ann Rev Nutrit 1991;11:285–308.

[31] Brosnan JT. Interorgan amino acid transport and its regulation. J Nutr 2003;133:2068S–72S.

[32] DeVito TJ, Drost DJ, Neufeld RW, et al. Evidence for cortical dysfunction in autism: a proton magnetic resonance spectroscopic imaging study. Biol Psychiatry 2007;61(4):465–73.

[33] Bernardi S, Anagnostou E, Shen J, et al. In vivo 1H-magnetic resonance spectroscopy study of the attentional networks in autism. Brain Res 2011;1380:198–205.

[34] Page LA, Daly E, Schmitz N, et al. In vivo 1H-magnetic resonance spectroscopy study of amygdala-hippocampal and parietal regions in autism. Am J Psychiatry 2006;163:2189–92.

[35] Autism Research Institute. Parent ratings of behavioral effects of biomedical interventions. <> (accessed 12/4/10).

[36] Quincozes-Santos A, Bobermin LD, Kleinkauf-Rocha J, et al. Atypical neuroleptic risperidone modulates glial functions in C6 astroglial cells. Prog Neuropsychopharmacol Biol Psychiatry 2009;33:11–5.

[37] Lumsden CE. Pathogenetic mechanisms in the leucoencephalopathies, in anoxic-ischaemic processes, in disorders of the blood and in intoxications. In: Vinken GW, Bruyn PJ, editors. Handbook of Clinical Neurology v. 9. New York: American Elsevier; 1970. p. 572–663.

[38] Gabis L, Huang Wei, Azizian A, et al. 1H-magnetic resonance spectroscopy markers of cognitive and language ability in clinical subtypes of autism spectrum disorders. J Child Neurol 2008;23:766–74.

[39] Carter CS. Sex differences in oxytocin and vasopressin: implications for autism spectrum disorders? Behav Brain Res 2007;176:170–86.

[40] Moritz ML, Ayus JC. New aspects in the pathogenesis, prevention, and treatment of hyponatremic encephalopathy in children. Pediatr Nephrol 2010;25:1225–38.

[41] Heins J, Zwingmann C. Organic osmolytes in hyponatremia and ammonia toxicity. Metab Brain Dis 2010;25:81–9.

[42] Courchesne E, Carper R, Akshoomoff N. Evidence of brain overgrowth in the first year of life in autism. JAMA 2003;290:337–44.

[43] Baron-Cohen S, Knickmeyer RC, Belmonte MK. Sex differences in the brain: implications for explaining autism. Science 2005;310:819–23.

[44] Geier DA, Geier MR. A clinical and laboratory evaluation of methionine cycletranssulfuration and androgen pathway markers in children with autistic disorders. Horm Res 2006;66:182–8.

[45] Barker EV, Hume R, Hallas A, Coughtrie WH. Dehydroepiandrosterone sulfotransferase in the developing human fetus: quantitative biochemical and immunological characterization of the hepatic, renal, and adrenal enzymes. Endocrinol 1994;134:982–9.

[46] Good P. Did acetaminophen provoke the autism epidemic? Altern Med Rev 2009;13:364–72. [republished on this site]

[47] Harony H, Wagner S. The contribution of oxytocin and vasopressin tovmammalian social behavior: potential role in autism spectrum disorder. Neurosignals 2010;18:82–97.

[48] Neville KA, Verge CF, O’Meara MW, Walker JL. High antidiuretic hormone levels and hyponatremia in children with gastroenteritis. Pediatrics 2005;116:1401–7.

[49] Emanuele E. Autism conceptualized as a vasopressin resistance syndrome. Med Hypotheses 2006;66:1245.

[50] Koide H. Three cases of hyponatremia during administration of pimozide [Article in Japanese, English abstract]. No To Hattatsu 1991;23: 502–5.

[51] Hiratani M, Munesue T, Terai K, Haruki S. Two cases of infantile autism with intermittent water intoxication due to compulsive water drinking and episodic release of antidiuretic hormone (SIADH) [Article in Japanese, English abstract]. No To Hattatsu 1997;29:367–72.

[52] Moritz ML, Ayus JC. Hyponatremia in preterm neonates: not a benign condition. Pediatrics 2009;124:1014–6.

[53] Web MD. Unusual salt craving. <> (accessed 7/3/11).

[54] Autism Blogger. Salt cravings. <> (accessed 7/3/11).

[55] De Luca LA, Vendramini Jr RC, Pereira DT, et al. Water deprivation and the double-depletion hypothesis: common neural mechanisms underlie thirst and salt appetite. Braz J Med Biol Res 2007;40:707–12.

[56] Morris MJ, Na ES, Johnson AK. Salt craving: the psychobiology of pathogenic sodium intake. Physiol Behav 2008;94:709–21.

[57] Beauchamp GK, Bertino M, Moran M. Sodium regulation: sensory aspects. J Am Diet Assoc 1982;80:40–5.

[58] Córdoba J, Gottstein J, Blei AT. Chronic hyponatremia exacerbates ammonia-induced brain edema in rats after portacaval anastomosis. J Hepatol 1998;29:589–94.

[59] Pangborn JB. Introduction to the diseases of autism and laboratory testing options. In: Pangborn JB, Baker SM. Biomedical Assessment Options for Children with Autism and Related Problems. San Diego, CA: Autism Research Institute; 2002. p. 1–130.

[60] Albrecht J, Sonnewald U, Waagepetersen HS, Schousboe A. Glutamine in the central nervous system: function and dysfunction. Front Biosci 2007;12:332–43.

[61] Huxtable RJ. Expanding the circle 1975–1999: sulfur biochemistry and insights on the biological functions of taurine. Adv Exp Med Biol 2000;483: 1–25.

[62] Durlach J, Durlach V. Speculations on hormonal controls of magnesium homeostasis: a hypothesis. Magnesium 1984;3:109–31.

[63] Laidlaw SA, Kopple JD. Newer concepts of the indispensable amino acids. Am J Clin Nutr 1987;46:593–605.

[64] Awapara J. The metabolism of taurine in the animal. In: Huxtable R, Barbeau A, eds. Taurine. New York: Raven Press; 1976. 1–19.

[65] Tanoue Y, Oda S. Weaning time of children with infantile autism.

J Autism Dev Disord 1989;19:425–34.

[66] Schultz ST, Klonoff-Cohen HS, Wingard DL, et al. Breastfeeding, infant formula supplementation, and autistic disorder: the results of a parent survey. Int Breastfeed J 2006. <> (accessed 5/23/10).

[67] Autism Research Institute. Treatment Options for Mercury/Metal Toxicity in Autism and Related Developmental Disabilities: Consensus Position Paper. February 2005. <> (accessed 6/14/11).

[68] Sweeten TL, Posey DJ, Shankar S, McDougle CJ. High nitric oxide production in autistic disorder: a possible role for interferon-gamma. Biol Psychiatry 2004;55:434–7.

[69] Carrick, Bob, Dee. The Use of Amino Acid Supplements with Autism and OCD. <> (accessed 2/6/11).

[70] Helt M, Kelley E, Kinsbourne M, et al. Can children with autism recover? If so, how? Neuropsychol Rev 2008;18:339–66.

[71] Good P. Letter in response to Helt et al. 2008. Does fever relieve autistic behavior by improving brain blood flow? Neuropsychol Rev 2011; 21:66–7.

[72] Wannemacher Jr RW, Pekarek RW, Bartelloni PJ, Vollmer RT, Beisel WR. Changes in individual plasma amino acids following experimentally induced sand fly fever virus infection. Metabolism 1972;21:67–76.

[73] Norenberg MD, Rao KV, Jayakumar AR. Mechanisms of ammonia-induced astrocyte swelling. Metab Brain Dis 2005;20:303–18.

[74] Solis-Herruzo JA, Gonzalez-Gamarra A, Castellano G, Muñoz-Yagüe MT. Metabolic clearance rate of arginine vasopressin in patients with cirrhosis. Hepatology 1992;16:974–9.

[75] Goodwin MS, Groden J, Velicer WF, et al. Cardiovascular arousal in individuals with autism. Focus Autism Other Dev Disabl 2006;21:100–23.

[76] Herbert MR. Treatment-guided research. Helping people now with humility, respect and boldness. Autism Advocate 2008:8–14.

[77] Herbert MR. SHANK3, the synapse, and autism. N Engl J Med 2011; 365:173–5.

[78] Tisdall M, Crocker M, Watkiss J, Smith M. Disturbances of sodium in critically ill adult neurologic patients: a clinical review. J Neurosurg Anesthesiol 2006;18:57–63.

[79] Wu G, Bazer FW, Davis TA, et al. Arginine metabolism and nutrition in growth, health and disease. Amino Acids 2009;37:153–68.

[80] Myers RD, Veale WL. Body temperature: possible ionic mechanism in the hypothalamus controlling the set point. Science 1970;170:95–7.

[81] Schultz ST, Klonoff-Cohen HS, Wingard DL, et al. Acetaminophen (paracetamol) use, measles-mumps-rubella vaccination, and autistic disorder: the results of a parent survey. Autism 2008;12:293–307.

[82] Fallon J. Could one of the most widely prescribed antibiotics amoxicillin/clavulanate ‘‘augmentin’’ be a risk factor for autism? Med Hypotheses 2005;64:312–5.

[83] Greenberg D, Hoffman S, Leibovitz E, Dagan R. Acute otitis media in children: association with day care centers – antibacterial resistance, treatment, and prevention. Paediatr Drugs 2008;10:75–83.

[84] Adams JB, Johansen LJ, Powell LD, Quig D, Rubin RA. Gastrointestinal flora and gastrointestinal status in children with autism – comparisons to neurotypical children and correlation with autism severity. BMC Gastroenterol 2011;11:22–34.

[85] Collins JK, Wu G, Perkins-Veazie P, et al. Watermelon consumption increases plasma arginine concentrations in adults. Nutrition 2007;23:261–6.

[86] Oral rehydration salts (ORS): (a) earlier WHO formulation (more glucose, no artificial sweeteners) available at REI or Jianas Brothers <> (accessed 8/5/11). (b) present WHO formulation (less glucose, artificial sweeteners) available as HYDRAssist.

[87] Bardhan PK. Improving the ORS: Does glutamine have a role? J Health Popul Nutr 2007;25:263–6.