To our knowledge, this is the first study to explore the relation between the presence of anti-NMDA receptor antibodies and ASD in a systematic way. Nearly a quarter of the patients’ group (26.4%) had tested positive for anti-NMDA receptor antibodies.
Previous research has shown that about 30–69% of ASD children have neuroinflammation or encephalitis . Specifically, the so-called anti-brain autoantibody may damage fetal or children’s brain cells, eventually leading to children falling into an autistic or regressive state . In the current study, only one-third of the patients proved to have anti-NMDA antibodies, this could be related to the type of sampling and the choice of the antibody to study.
Several autoantibodies have been reported to occur in the ASD population with varying frequency. There were many early studies investigating this possibility. About 58% were found to have anti-myelin basic protein . Circulating antibodies towards neuronal and glial filament proteins have also been reported in autism . Also, circulating antibodies against the caudate nucleus were found in sera of 49% of 68 autistic children .
Frye et al. , found that 75.3% of their studied ASD population have folate receptor autoantibodies. 62.5% of a studied ASD population had positive serum anti-neuronal antibodies and 74% tested positive for anti-ganglioside antibodies and both were correlated with more severe symptoms of autism [24, 25].
There were also reports of maternal anti-brain antibodies in mothers of autistic children . It also has been reported that 25% of autistic children may have elevated anti-nuclear antibodies . Increased serum anti endothelial cell antibody levels were observed in 63% of autistic children . The rates reported in the current study (26.4%) are close to the lower range of other reported autoantibodies. This may be explained by the different courses that an anti-NMDA antibody pathology may take whether an acute and totally regressive course or a subacute and lesser regressive course than other antibodies.
A suggested further research in this area is to investigate the co-occurrence of these antibodies in the same population which may shed some light on the interaction between these autoantibodies and their combined effect.
The current study found that, adequate gross and fine motor development were correlated with positive anti-NMDA receptor antibodies. This may suggest that the anti-NMDA receptor antibody (if proven as a possible etiology to a subset of autism disorder) causes a milder type of brain affection in autism, or possibly due to partial recovery from anti-NMDAR encephalitis as a natural course of the illness .
Previous research demonstrated that the degree of motor development predicted the subsequent development of language in children with autism . Toe walking and age when first sitting predicted the severity of stereotyped behavior . Both severity of stereotyped behaviors and better degree of language development had a statistically significant correlation with the positive anti-NMDA receptor antibody as shown in Table 3.
Better speech stage compared to the rest of the autism population (phrase speech rather than preverbal/single word) was correlated with having positive anti-NMDA receptor antibodies. Furthermore, the positive anti-NMDAR antibody was correlated with a more calibrated score and age coefficient on the Vineland adaptive behavior scale. The current finding contrasts with previous research that demonstrated persistent cognitive impairment of patients with NMDA encephalitis . This could be related to the severity of the encephalitis, as none of the parents reported flu-like symptoms at the onset of disease, and different age and presentation of the current sample.
The basal ganglia have been linked to involuntary movements and stereotyped behaviors . Also, several antigens as targets to autoantibodies have been linked to various basal ganglia pathology and its resulting disorders including DPPX: dipeptidyl-peptidase-like protein-6, GFAP: glial fibrillary acidic protein, mGluR1: metabotropic glutamate receptor 1, and NMDAR: N-methyl-d-aspartate receptors . This may explain how the anti-NMDAR antibody is correlated with the severity of stereotyped behavior due to its action on the basal ganglia.
The kits used are directed towards the NR2A subunit of the NMDA receptor. NR2A and NR2B mRNAs are prominent in CA1 and CA3 pyramidal cells. NR2A is predominant in corticolimbic regions and is expressed mainly post-natal . This may point out that the area in question may be the prefrontal cortex or part of the limbic system such as the anterior cingulate cortex.
From the integration of these data, we can conclude that the anti-NMDAR antibody affects the severity of, and is correlated to, stereotyped behaviors more than social interaction, and both more than the language domain in autism.
Positive anti-NMDAR was correlated with abnormal EEG findings, particularly with epileptic discharge rather than slowing. This agrees with the general notion that the NMDAR pathology is correlated with abnormal EEG findings , though EEG patterns of the delta brush were more commonly reported in this study than the epileptic discharge.
One of the important parameters that did not have a significant correlation with the presence of the anti-NMDA receptor antibody was the occurrence of regression in developmental milestones. This contrasts with what was initially hypothesized. This may point to the need for detection of anti-NMDAR antibodies earlier, e.g., in mothers or in amniotic fluid as previously suggested for other antibodies by . The antibody may have been present from birth while exerting its effects at different stages or periods of brain development. Also, including a larger sample considering the acknowledged ratio between regression and non-regression groups of autism may enhance the statistical significance measurement of anti-NMDAR in autism. Moreover, recall bias cannot be excluded due to the cross-sectional stage sampling of the current study.
In comparing the clinical profile of the studies’ anti-NMDA receptor antibody with other autoantibodies,  did not find a correlation of folate receptor autoantibodies with developmental profile, occurrence of regression, type of regression, or EEG changes. Other studies correlated the presence of autoantibodies with more severity of ASD symptoms such as [13, 24]. This may seem to contradict the finding of this study where the positive antibody group had a better language level than the negative group. However, this may be explained by the notion that the effect of the anti-NMDA receptor antibody is related to its effects on the developing brain rather than a “quantitative” autoimmune pathology. Also, the more specific brain regions that this antibody acts upon as mentioned above that the receptor in question is concentrated more in certain brain regions.
Limitations of this study include a relatively small sample size which was due to premature termination of the study due to the COVID-19 pandemic. Also, the use of ELISA kits for the detection of NMDAR ab instead of the cell-based assay is due to higher expenses of the cell-based assay. The use of qualitative kits for anti-NMDA receptor antibodies was also a limitation, as a quantitative analysis would have provided better correlations with severity and other parameters. Other inflammatory markers in blood which could have been done, such as ESR (erythrocyte sedimentation rate), ANA (anti-nuclear antibody), and cytokine panel, again was not done due to expenses.
Future research in relation to this study may include functional brain imaging in correlation to the population with positive anti-NMDAR antibodies and prospective studies monitoring the severity of symptoms of the autism disorder with the level of anti-NMDAR antibody. Also, research targeting the presence of anti-NMDAR antibodies in sera of mothers with children with autism and research targeting the effect of immunological therapies such as steroids, IVIG, and plasmapharesis in children with autism testing positive for anti-NMDAR antibodies.