The diagnosis of Autism: one, no one, 100.000 autisms

Autism (or autistic) Spectrum Disorders (ASDs) refer to 3 conditions: 

• Autistic Disorder 
• Asperger Syndrome 
• Pervasive Developmental Disorder not otherwise specified(PDD-NOS) 

These conditions are classified as Pervasive Developmental Disorders in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) together with Childhood Disintegrative Disorder and Rett Syndrome (see figure).

ASDs are typically characterized by social deficits, communication difficulties, stereotyped or repetitive behaviors and interests.
Although these diagnoses share some common features, individuals are affected in various degree of severity.

How are ASD diagnoses made?
Sydromic vs. Etiologic Diagnoses

The diagnosis of ASD is a syndromic diagnosis. Syndromic diagnoses, as opposite to etiologic diagnoses, are based only on observed symptoms without any reference to the underlying biological cause. Because they consider only the visible manifestation they are also called descriptive diagnoses. 

Syndromic diagnoses, although less rigorous and more susceptible to errors than etiologic diagnoses, are the only option when the underlying cause of the disorder is unknown. When it is not possible to identify any cause the clinician can base his diagnosis only on the observed symptoms (the distinction between signs and symptoms is here not considered as not relevant for this article). 

Example:

Migraine is a highly common syndrome with an unknown cause. The diagnose of Migraine (or of one of its numerous subtypes) is based on the symptoms reported and observed. The diagnosis provides a label to summarize the characteristics of the disorder but, being cause of Migraine unknown, does not inform on the presence of any specific underlying biological condition.

Syndromic diagnoses are labels to nominate a cluster of symptoms. 
They do not tell us anything about the cause of the observed disorder and neither imply that the disorder has a unique cause (as a matter of fact, it is commonly accepted that a syndrome (such as migraine) may have different causes in different individuals.

Symptoms are not grouped in a syndrome because of an underlying biological condition but because of a shared impression that these symptoms often co-occur. Syndromes are not natural entities but the result of a statistical (for this reason the DSM is the Diagnostic and Statistical Manual of Mental Disorders) analysis of groups of symptoms that has been see usually (not always) appear together. 

As for any other symdromic diagnosis the diagnosis of Autism Spectrum Disorders is only based on the behavioral characteristics of the patient.  

During the diagnostic process the clinician will obtain a behavioral description of the person (what does this person do or not do?)  and compare it to the defying criteria for the suspected disorder provided by the DSM-IV. If criteria are met the diagnostic label is assigned.

When necessary clinicians may employ additional means beyond the simple observation of the patient: questionnaires, tests and structured interviews to the patient or, especially when the person is not able to answer, to the parents. 

These instruments may help the clinician to obtain a more complete and accurate characterization of the patient but, no matter what instruments are used, a diagnosis of ASD always result from the comparison of the behavioural characterstics of the person to the DSM criteria.

This standardized procedure assures that every time a diagnosis is made is made on the same criteria and therefore refer to the same set of characteristics. 

Esample:

During the visit the professional observes that the child does not use verbal communication and verifies the lack of verbalizations by interviewing the parents about the child's communication skills. The clinician has in this way collect enough information to conclude that one of the criteria for the diagnosis of Autistic Disorder is met (i.e. qualitative impairments in communication). At this point he may carry on the assessment collecting information about the other characteristics of Autistic Disorder (impaired social interaction, presence of repetitive stereotyped behaviors). If the minimum number of criteria are met a diagnosis of autism will be assigned.

A symdromic diagnoses is a summary label for a set of observed behavioural characteristics. Assigning a label does not add anything to our knowledge of the nature of the problems that have been labeled. We decided to call the presence of restrictions in communication and social interaction and atypical, repetitive behaviors 'ASD, just as we decided to call a roof on top of a closed wall provided with door and windows 'house'.

Labels are quicker way to communicate but do not tell us how and why houses have been built or why some individuals struggle to communicate and interact in functional ways.

Like for every label the value of syndromic diagnoses is primarly communicative. Diagnostic labels allow clear and rapid communication between researchers, services providers and practitioners.

 

What do we know about causes of ASDs?

Supporting evidence of the genetic basis to ASDs is strong, as indicated by the recurrence risk in families and twin studies.

About 10–20% of individuals with an ASD has a recognized genetic condition, most commonly: fragile X, tuberous sclerosis, neurofibromatosis (more than 20 genetic and metabolic conditions responsible for ASDs have been listed). Further microscopically visible chromosomal alterations and rare mutations in synaptic genes have been reported in ∼5% of cases. Recently new genetic tests have revealed submicroscopic deletions and duplications in 5%–10% of ASD cases (see box)

Findings from the researh:

The ever-improving available genetic tests allow to discover always new submicroscopic structural variations in people with ASD. Marshall and colleagues (2008), for example, performed a last generation genetic assessment for structural abnormalities in 427 individuals diagnosed with ASD. Structural variations (copy number variation, including deletion and duplication, translocation, inversion) of chromosomes was discovered in more than 50% of the cases. Betancur (2011) carried out an exhaustive review of the research genetics literature in an attempt to collate all genes and recurrent genomic imbalances that have been implicated in the etiology of ASD. The study provided data on 103 disease genes and 44 genomic loci reported in subjects with ASD or autistic behavior.

Taken together, these findings show that ASDs are etiologically heterogeneous and that 

 "that autism is not a single clinical entity but a behavioral manifestation of tens or perhaps hundreds of genetic and genomic disorders" (Betancur, 2011).

In other words we are currently grouping many different biological conditions under the same clinical label. 

At present high cost of genetic evaluation and the fact that the information obtained typically do not change interventions for the patient (that mostly remains behavioral) limit the number of cases in which a genetic evaluation to obtain a genetic diagnosis is attempted. 

However identification of separate biological conditions can greatly improve research of biomedical treatments (e.g. diets, integrators, medications) that today are impaired by generic syndromic diagnoses. Etiologic diagnoses may also positively affect behavioral treatments allowing early diagnoses and better prognoses.  

References

Betancurin, (2011). Etiological heterogeneity in autism spectrum disorders: more than 100
genetic and genomic disorders and still counting, in Brain Research, 1380:42-77 (http://www.hal.inserm.fr/docs/00/54/98/73/PDF/Betancur_Brain_Res_2010.pdf)

Schaefer, GB, Mendelsohn NJ, (2008). Genetics evaluation for the etiologic diagnosis of autism spectrum disorders. In Genetic Medicine 10(6):464. (http://www.ncbi.nlm.nih.gov/pubmed/18197051)
 

Marshall CR, (2008). Structural Variation of Chromosomes in Autism Spectrum Disorder. In American Journal of Human Genetic; 82(2): 477–488. (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2426913/)