Laboratory Testing for Developmental Delay, Intellectual Disability, and Autism Spectrum Disorder

Content Review: February 2021 Last Update:

Developmental delay (DD) is defined as any significant lag in a child's physical, cognitive, emotional, or social maturity. Intellectual disability (ID) is characterized by broad impairment in cognitive and adaptive functioning, typically with an intelligence quotient (IQ) <70 diagnosed before 18 years of age. A global DD diagnosis often precedes a diagnosis of ID because neither cognitive skill nor IQ can be reliably assessed before age 6. Those with severe DD diagnosed before age 6 are most likely to develop ID.  

Autism spectrum disorders (ASDs) represent a neurodevelopmental continuum with varying degrees of social impairment, communication limitations, repetitive behaviors, and/or a restricted range of interests. ASDs are typically detected by 3 years of age based on parents’ and observers’ identification of abnormal interactions and behaviors.  Routine pediatric screening may identify a child with global delay and spur investigation of the underlying etiology.

Chromosomal microarray (CMA, also referred to as cytogenomic single nucleotide polymorphism [SNP] microarray) is the recommended first-line test for DD/ID or ASD of unknown etiology.    CMA offers the highest diagnostic yield (approximately 15-20%) in individuals with unexplained DD/ID, ASD, and multiple congenital anomalies (MCAs) and is preferred to chromosome analysis (karyotyping).  -  Multiple groups have made recommendations regarding additional testing for fragile X syndrome (FXS), inborn errors of metabolism (IEMs), X-linked intellectual disability, MECP2-related disorders, PTEN-related disorders, and/or chromosome analysis, depending on clinical findings and family history.    

As the technology evolves and test costs decline, whole genome sequencing (WGS), which can assess genetic sequences of nuclear and mitochondrial DNA and copy number variants (CNVs), or whole exome sequencing (WES), which can assess genetic sequences of the coding region of nuclear genes, but usually does not cover mitochondrial DNA and does not consistently identify CNVs, in combination with CMA, may become first-line testing for these conditions. 

Refer to the Testing for Genetic Syndromes Related to Developmental Delay, Intellectual Disability, and Autism Spectrum Disorder Algorithm for suggested tiered testing approaches.

Quick Answers for Clinicians

Which genetic test provides the highest diagnostic yield for developmental delay, intellectual disability, and autism spectrum disorders of unknown cause?

Chromosomal microarray (CMA) is the recommended first-tier diagnostic test for patients with developmental delay (DD), intellectual disability (ID), or autism spectrum disorders (ASDs) of unknown etiology. The diagnostic yield varies by patient population and the presence of comorbidities, but is estimated to be approximately 15-20% (approximately 10% higher than the detection rate by karyotype analysis in the DD/ID/ASD population). 

If clinical features or family history suggest a specific disorder, testing for that disorder before proceeding with CMA is recommended.  -  Refer to the Testing for Genetic Syndromes Related to Developmental Delay, Intellectual Disability, and Autism Spectrum Disorder Algorithm for more information.

Although not typically included in guidelines or billing algorithms, whole exome sequencing (WES) is gaining traction as a test for DD/ID/ASD. A 2019 meta-analysis asserts that WES has a higher diagnostic yield than CMA and should be a first-line test for neurodevelopmental disorders.  See Variants Detected by Different Methods table. WES does not reliably detect copy number variants (CNVs) and is usually performed after CMA.

What are the advantages and disadvantages of chromosomal microarray compared with G-banded chromosome analysis (karyotyping)?

The main advantage of chromosomal microarray (CMA), and the reason for the shift to CMA instead of karyotyping, is its superior resolution. CMA is able to identify submicroscopic deletions and duplications (smaller than approximately 10 Mb, the size of many of the deletions and duplications that lead to developmental delay/intellectual disability/autism spectrum disorders [DD/ID/ASDs]) that cannot be detected by karyotyping, which leads to a higher detection rate for patients with DD/ID, ASD, and multiple congenital anomalies (MCAs) of unknown etiology. This higher detection rate justifies the increased cost of CMA and explains why CMA is recommended as a replacement for chromosome analysis rather than as follow-up testing after a normal chromosome analysis result. Another benefit of CMA is that it will also identify regions of homozygosity, which can be scrutinized for autosomal recessive conditions and imprinting disorders. Despite these advantages, CMA (with a blood specimen) takes more time than stat chromosome analysis, does not detect balanced rearrangements, and does not characterize the location of detected copy number gains. Therefore, chromosome analysis is still preferred for some indications (eg, multiple miscarriages and suspected aneuploidy). See Variants Detected by Different Methods table.

Should other testing be performed after, or in addition to, chromosomal microarray?

Multiple guidelines recommend FMR1 analysis for fragile X syndrome (FXS) (see FMR1 Testing for Fragile X Syndrome) in addition to chromosomal microarray (CMA). Based on clinical and family history as well as a dysmorphology exam, recommended testing may also include metabolic testing, X-linked disorder testing, and/or targeted testing for a specific suspected condition. Multiple guidelines recommend evaluation for inborn errors of metabolism (IEMs), which has a low diagnostic yield but significant benefit for patients. Finally, testing for MECP2-related or PTEN-related disorders may be indicated in addition to CMA (see Laboratory Testing). Neither cytogenomic single nucleotide polymorphism (SNP) microarray (CMA) nor chromosome analysis will detect sequence variants, triplet repeat expansions (as in fragile X syndrome), copy number variations (CNVs) outside of probe coverage, heterodisomy (uniparental disomy), or some other etiologies.     CMA will miss exon-level deletions and duplications. Therefore, additional and/or concurrent testing may be needed when a patient has DD/ID and/or ASD of unknown etiology. For more information on the limitations of CMA testing and additional guidance on testing to inform diagnosis after negative CMA, see the 2018 ACMG clinical practice resource. 

What role does chromosome analysis (standard karyotyping) play in the workup for developmental delay/intellectual disability or autism spectrum disorder?

Chromosomal microarray (CMA) has replaced karyotype analysis as the first laboratory test for developmental delay (DD), intellectual disability (ID), and autism spectrum disorders (ASDs) of unknown etiology.  -  CMA is also now the first-line test for patients with multiple congenital anomalies (MCAs). Although chromosome analysis was historically the first-line test for patients with DD/ID, it is currently only recommended as a first test for patients with apparent aneuploidy (eg, Down syndrome, trisomy 13, trisomy 18, Klinefelter syndrome, or Turner syndrome). Chromosome analysis is also indicated when there is a family history of chromosome rearrangement or multiple miscarriages because it can detect balanced chromosomal abnormalities, which CMA does not detect.    To evaluate a patient with MCAs/DD/ID/ASD who appears to have aneuploidy, chromosome analysis with reflex to CMA can be ordered.

Is there a role for FISH analysis that targets a specific microdeletion or microduplication syndrome?

The consensus is not clear or consistent regarding chromosomal microarray (CMA) versus targeted fluorescence in situ hybridization (FISH) analysis.   FISH probes assess a specific copy number variant (CNV) associated with a specific syndrome. CMA detects CNVs across the entire genome. FISH is less expensive than CMA, but if FISH analysis does not confirm the suspected diagnosis, CMA must also be performed. CMA can identify CNVs that are too small to identify via FISH, and tandem submicroscopic duplications may be easier to confirm by CMA than FISH. Metaphase FISH shows whether a duplicated region is at its normal location, whereas CMA does not. Thus, FISH metaphase analysis is often used to assess the relatives of the affected patient for balanced rearrangements.

When should comprehensive metabolic testing be considered for patients with developmental delay, intellectual disability, or autism spectrum disorder?

Guidelines recommend that specific testing for inborn errors of metabolism (IEMs) should be performed after reviewing newborn screening (NBS) results and assessing for clinical indicators. This recommendation is based on the benefit of diagnosing conditions for which treatments and/or interventions exist. Although the incidence of IEMs is low, the potential to improve outcomes by intervention and treatment after diagnosis is high,    and a diagnosis may enable dietary management or other treatment options. Not all metabolic conditions are included on NBS panels, which vary by state. Clinical features of an IEM may include developmental delays in isolation or in combination with autism, regression (neurodegeneration), failure to thrive, poor physical endurance/lethargy, episodic symptoms such as epilepsy and encephalopathy, multiple organ dysfunction, dietary selectivity, unusual odors, and hearing loss.

Where can I go for more information on the evaluation, including laboratory testing, of global developmental delay, intellectual disability, and autism spectrum disorders?

The American Academy of Pediatrics (AAP) global developmental delay (DD) and intellectual disability (ID) evaluation guideline  provides guidance for clinical evaluation of DD and ID, including medical and family histories; physical, dysmorphology, and neurologic exams; imaging; patient management and communications; and laboratory testing. The American College of Medical Genetics (ACMG) practice guidelines  outline the approach to a genetic diagnostic evaluation for patients with autism spectrum disorders (ASDs).

Indications for Testing

Laboratory evaluation of suspected DD, ID, or ASD should be considered for individuals with any of the following presentations   :

  • Failure to meet developmental milestones
  • IQ <70 and difficulty performing daily living activities
  • Comorbidities (eg, dysmorphic features, congenital anomalies) that may guide testing

Determining Diagnosis for Care Planning

Clinical evaluation of DD/ID and ASD will guide the laboratory testing strategy and should include  :

  • Medical history, including prenatal and birth histories
  • Family history with at least a three-generation pedigree
  • Physical and neurologic exam with particular attention to dysmorphology (minor anomalies)
  • Examination of behavior and neurologic symptoms

After the clinical evaluation, the judicious use of laboratory testing, imaging, and other techniques is recommended.    If a specific condition is suspected, consider targeted testing. If the etiology of DD/ID/ASD is unknown, proceed with tiered testing based on which tests have the highest diagnostic yield in the patient population.

Laboratory Testing

The American College of Medical Genetics and Genomics (ACMG) developed a tiered evaluation system to assist clinicians in the clinical genetic diagnostic evaluation of ASD.  The logic behind the tiered approach is that tests performed in higher (or earlier) tiers have a greater expected diagnostic yield, are less invasive, and provide better potential for intervention. The tiered approach also allows for customization to the clinical situation at hand.

The American Academy of Pediatrics (AAP) recommends a similar stepwise approach in the evaluation of children with ID or global DD.  The American Academy of Neurology (AAN) also supports laboratory testing in the evaluation of patients with global DD/ID. 

First-Tier Evaluation

The first-tier evaluation for DD/ID includes a detailed clinical evaluation, as described in the Determining Diagnosis for Care Planning section. If a specific disorder is suspected, targeted testing should be performed with appropriate follow-up. If no specific etiology is suggested, testing should proceed with CMA (and/or chromosome analysis in limited circumstances) and fragile X testing. Particularly for global DD/ID, consider testing for IEMs, which has a low yield but high benefit for diagnosed patients.   

As with DD/ID, the first-tier evaluation for ASD includes a detailed clinical evaluation and should include an audiogram.  Certain disorders that have firmly established associations with ASD (eg, Angelman syndrome and Prader-Willi syndrome) may be identified through disorder-specific testing.  If such a disorder is diagnosed, no further testing to identify the etiology of ASD is required.  Otherwise, proceed with tiered testing and/or consider referral for a medical genetics evaluation.

Chromosomal Microarray​

CMA is the preferred first-tier test for DD/ID, ASD, and MCAs in patients for whom the causal diagnosis is unknown. CMA offers a higher diagnostic yield (approximately 15-20%) for genetic testing of individuals with unexplained DD/ID, ASD, or MCAs than does chromosome analysis (approximately 3%, excluding Down syndrome and other recognizable chromosomal syndromes).    CMA detects syndrome-associated microdeletions and microduplications that are below the resolution of karyotype analysis. Cryptic CNVs that indicate unbalanced translocations may be identified. Compared with chromosome analysis, CMA provides far superior detail that facilitates the interpretation of results.

If CMA results are normal, review of other first-tier test results and consideration of second-tier testing and/or referral to a medical geneticist are recommended. If CMA results are abnormal, genetic counseling is recommended.

FMR1 Testing for Fragile X Syndrome

FXS is the most common form of heritable ID. This X-linked condition presents as severe DD in males and as mild ID in females. Apparently spontaneous cases with negative family history arise due to the expansion of premutations in female carriers. The AAP recommends that all children who present with global DD/ID of unknown etiology be tested for FXS.  The AAN evidence report also supports testing patients of both sexes with unexplained DD/ID.  The ACMG recommends FMR1 analysis for boys with ASD and for girls with a consistent family history and phenotype.  The AAN has found that FMR1 testing has a combined yield of at least 2% in male and female patients who have mild DD/ID.  Timely diagnosis supports access to early intervention and enables timely reproductive risk counseling. For test specific information, see the Fragile X Syndrome Test Fact Sheet.

Chromosome Analysis (Karyotyping)

G-banded chromosome analysis is no longer recommended unless aneuploidy is suspected. Chromosome analysis has a lower diagnostic yield than CMA because it typically only detects genomic imbalances greater than 5-10 Mb. It is only cost-effective in patients whose symptoms are consistent with a specific chromosome abnormality such as Down syndrome, trisomy 13, or trisomy 18, or in certain other special cases (eg, family history of multiple miscarriages).    

Metabolic and/or Mitochondrial Disorder Testing

Metabolic and/or mitochondrial disorder testing may be considered after reviewing newborn screening (NBS) results and assessing for clinical indicators that may suggest a metabolic disorder (eg, failure to thrive, unusual odors, hearing loss, and episodic symptoms). The AAP, AAN, and ACMG support the consideration of screening for metabolic conditions in children who present with DD/ID/ASD. Many metabolic tests are available at a relatively low cost, and, despite the low prevalence of inherited metabolic conditions, the potential for improved outcomes after diagnosis and treatment is high.   

Second-Tier Evaluation


If no etiology is found for ASD, second-tier testing includes MECP2 sequencing (for Rett syndrome) in all females and MECP2 duplication testing in males with suggestive phenotypes. 

Second-tier testing for DD/ID includes MECP2 full gene analysis in females. The AAN suggests MECP2 testing in girls with severe global DD/ID, and the AAP emphasizes the need for complete MECP2 testing for females with global DD/ID.   Both groups emphasize that MECP2 testing in these patients is indicated even in the absence of classic Rett syndrome features.


Patients of both sexes who have head circumferences >2.5 standard deviations above the mean should have PTEN gene analysis if the ASD etiology is still unknown after initial evaluations. 

Other Testing

Although not typically recommended by clinical guidelines or included in billing algorithms, WES is increasingly being used in DD/ID/ASD.  WES does not reliably detect CNVs and is generally performed after CMA.

The AAP, ACMG, and AAN recommend consideration of testing for X-linked ID in males with a suggestive family history.

Brain magnetic resonance imaging (MRI) is recommended for patients with microcephaly, macrocephaly, or abnormal neurologic exams.

Development of a plan for follow-up, including potential reevaluation, is recommended. 

Refer to the Testing for Genetic Syndromes Related to Developmental Delay, Intellectual Disability, and Autism Spectrum Disorder Algorithm for suggested testing strategies.

Diagnostic Yields of Genetic Testing

The following approximate diagnostic yields are expected in the genetic evaluation of ASDs and global DD/ID:

Testing Approximate Diagnostic Yield (Global DD/ID) Approximate Diagnostic Yield (ASD)
CMA 15-20% a


4.2-24.5% b

WESc,d,e 30%  3.1-28.6% b
FXS >2% (mild-moderate DD/ID)  1-5% 
MECP2 1.5% (females with moderate-severe DD/ID)  4% (females) 
PTEN 5% (of those tested with head circumferences >2.5 SDs) 
Karyotyping 3-4%  (excluding Down syndrome) 3% 
Other (eg, IEM, X-linked disorders) Up to 10%  10% 

aAverage of 33 studies.

bStratified by presence of dysmorphism, congenital anomalies, and structural brain abnormalities.

cDiagnostic utility of WGS is equal to or greater than WES.

dTrio-based sequencing (sequencing of proband and parents concurrently) is more sensitive and analytically efficient than proband-only sequencing.

eData sharing and periodic WES reanalysis increase diagnostic yield.

SDs, standard deviations

Sources: Michelson, 2011 ; Schaefer, 2013 ; Miller, 2010 ; Tammimies, 2015 ; Retterer, 2016 

Comparison of Variants Detected by Different Methods of Genomic Tests

Although traditional chromosome analysis, WES, and CMA assess variants across the genome, some variant categories may only be detected by a single test method.

Variants Detected by Different Genomic Testing Methods
CMA WESa Karyotype Analysis
Chromosomal CNVs (deletions/duplications) Detected Not detectedb Only large deletions and duplications ≥10-15 Mb
Sequence variants (eg, missense and point mutations) Not detected Detected Not detected
Aneuploidy Detected Not detectedb Detected
Balanced chromosome rearrangements Not detected Not detected Detected
Unbalanced translocations Detectedc Not detectedb Detects unbalanced translocations ≥10-15 Mb
Trinucleotide repeat expansion disordersd Not detected Not detected Not detected
Aberrant methylation Not detected Not detected Not detected
Low-level mosaicisme Not detected Not detected Detected (aneuploidy)f
Triploidy Detected (with SNP analysis) Not detected Detected
Consanguinity Detected (with SNP analysis) Not detected Not detected
Uniparental disomy Detected (in some cases) Not detectedb Not detected
Exon-level deletions and duplicationsg Not detected Not detected Not detected
Intergenic variants Not detectedh Not detected Detected if ≥10-15 Mbi
Regions of homology Not detected (redundancy impedes analysis)j Detected at reduced resolution (homology impedes analysis) Detected if ≥10-15 Mb

aWGS detects categories of variants not assessed by WES.

bMay be included in testing and development of WES pipeline.

cCMA cannot provide structural (positional) information associated with genomic imbalance.

dTesting that targets the specific disorder is required.

eThe limit of detection for mosaicism varies depending on the size and type of genomic imbalance.

fLevel of detection depends on number of cells/colonies counted.

gAdditional testing methods, such as MLPA, may be required.

hSNP probes target unique sequence coding regions. Copy number probes provide coverage of noncoding regions, but variants in intergenic regions may not be interpreted and/or reported, depending on size and genomic content.

iMay identify variants involving noncoding DNA that need additional assessment to determine clinical significance.

jThe limit of detection varies depending on the size of region and probe coverage.

MLPA, multiplex ligation-dependent probe amplification

ARUP Laboratory Tests

First-Tier Testing


Tandem Mass Spectrometry/Electrophoresis/Spectrophotometry/ Gas Chromatography-Mass Spectrometry/Liquid Chromatography-Tandem Mass Spectrometry/Quantitative Liquid Chromatography-Tandem Mass Spectrometry, Genomic Microarray (Oligo-SNP Array), Polymerase Chain Reaction/Capillary Electrophoresis

Panel includes cytogenomic SNP microarray; fragile X (FMR1) with reflex to methylation analysis; acylcarnitine quantitative profile (plasma), mucopolysaccharides screen (electrophoresis and quantitation, urine), organic acids (urine), creatine disorders panel (serum/plasma and urine), and amino acids quantitative by LC-MS/MS (plasma)

AAP and ACMG guidelines suggest testing for fragile X disorder in males only

For individuals with phenotype suggesting aneuploidy or history of multiple miscarriages

Metabolic Testing

Panel includes acylcarnitine quantitative profile (plasma), mucopolysaccharides screen (electrophoresis and quantitation, urine), organic acids (urine), creatine disorders panel (serum/plasma and urine), and amino acids quantitative by LC-MS/MS (plasma)

Exome Sequencing


Medical Experts



Nicola Longo, MD, PhD
Professor, Pediatrics; Adjunct Professor of Pathology (Clinical), University of Utah
Chief, Medical Genetics Division; Medical Director, Biochemical Genetics and Newborn Screening, ARUP Laboratories


Rong Mao, MD, FACMG
Professor of Pathology (Clinical), and Co-Director of Laboratory Genetics and Genomics Fellowship, University of Utah
Medical Director, Molecular Genetics and Genomics, ARUP Laboratories


Marzia Pasquali, PhD, FACMG
Professor of Pathology, Adjunct Professor of Pediatrics, University of Utah School of Medicine
Section Chief, Biochemical Genetics; Medical Director, Biochemical Genetics and Newborn Screening, ARUP Laboratories


Reha Toydemir, MD, PhD, FACMG
Associate Professor of Pathology (Clinical), and Adjunct Assistant Professor of Pediatrics, University of Utah
Medical Director, Cytogenetics and Genomics, ARUP Laboratories