Improving Speech Before Baby Speaks

Researcher discovers variations in genes that may cause childhood apraxia of speech, which could speed identification and treatment

Typically, children can make the sounds of their language well enough to be understood by the age of four – but this isn’t the case for kids with severe forms of speech sound disorders.

“Some of them are so difficult to understand that even their own parents are left to guess what they’re saying,” said Arizona State University (ASU) researcher Beate Peter, PhD, CCC-SLP, and assistant professor at ASU’s College of Health Solutions. “You can imagine how frustrating that is for the children and their families.”

That is why it is Peter’s goal to find genetic changes that influence how children learn to communicate, focusing much of her recent work on severe speech sound disorders.

“Currently, a child with a severe speech sound disorder might get a diagnosis at around age two or three at the earliest, but by that time they’ve missed out on tons of opportunities to communicate,” Peter said.

Identifying children at genetic risk for speech sound disorders much earlier may now be easier than before, owing to Peter’s recent discovery of two main genes relevant for childhood apraxia of speech (CAS), one of the most severe forms of speech sound disorder.

Childhood Apraxia

CAS is a motor speech disorder that involves difficulty translating the intention to speak into actual speech movements.

“If the brain is the conductor of a large orchestra, then the players are the lips, the vocal folds, the different parts of the tongue.all of them need to be orchestrated in a precisely timed manner,” analogized Peter. “For kids with CAS, it’s like the message that the conductor tries to send is all messed up and the result is garbled speech, just like you would wind up with horrible music if the conductor can’t get the message to the players.”

Though many kids have difficulty learning to talk, it’s a different case for children with CAS. They make sound errors that are not noted in other kinds of speech disorders. For instance, they might leave out the very first consonant in a word or only know how to make a few of the 23 consonants of English and use an easy sound like “h” for the rest. Some children with CAS also struggle with learning to read and write or have difficulties with their fine motor system. These complications make it even more important to detect risks for CAS as early as possible.


According to Peter, the biggest breakthrough in speech and language genetics was the discovery of the FOXP2 gene in a large multigenerational family in the UK in 2001. Genetic mutations in the FOXP2 gene are thought to cause CAS in certain families, but these cases are very rare. In the following years, several CAS candidate regions and genes were found. Then, in 2014, Peter discovered a CAS candidate gene called BCL11A in a child with CAS who was missing this gene on one of the two copies of chromosome 2. Peter’s most recent paper, published in PLOS ONE in April 2016, is really the first speech genetic study in multigenerational families since the discovery of the FOXP2 gene to find actual genetic mutations thought to cause CAS. “In that sense, it is a real novelty,” Peter remarked.

Data Collection

For this study, Peter amassed a huge amount of data from two separate multigenerational families with high occurrences of CAS. Peter tested each family member for two to three hours to thoroughly assess their oral and fine motor systems and detect residual effects of CAS in adults who no longer had a speech problem. Peter also obtained a DNA sample from each individual, which she sent to Dr. Wendy Raskind’s genetics lab at the University of Washington where she previously worked as a postdoctoral trainee.

Next, carefully measured amounts of DNA were sent to the lab at the University of Washington’s Center for Mendelian Genomics. For each DNA sample, Peter was given the genotypes for a single nucleotide polymorphism panel, which is a text file with over half a million lines of information spanning all chromosomes. Using a statistical procedure called linkage analysis, Peter determined which pieces of the chromosomes were inherited along with the speech disorder in each family. This was a huge amount of information – the first text file Peter analyzed had over 15 million lines!

The second piece of data Peter analyzed was the exome sequence of several people in each family. The exome sequence contains all of the pieces of DNA that code for proteins. Peter looked for harmful exome variations that were shared only by family members with CAS. The final step was to check the DNA of everyone in the family for these variations. Just as Peter thought, people with CAS had the variations and those without CAS did not have the variations.


In one family, Peter identified a main candidate gene called CDH18 on chromosome 5. This gene influences synaptic adhesions and, in this family, likely interacts with some other genes on another chromosome that influence brain development and function. Interestingly, CDH18 and most of the other candidate genes are highly expressed in the cerebellum, a brain region that regulates movement and plays a huge role in speech and language processing.

In the second family, Peter found just one variation on chromosome 4 in a gene called ZGRF1. Not much is known about this gene except that it is also expressed in the cerebellum and has a similar function to SETX, a known apraxia candidate gene. “It’s very exciting to think about, because these two families have a similar speech disorder but for completely different genetic reasons,” Peter remarked.

Because these gene variants were found on different chromosomes – chromosome 5 and 17 in one family, and chromosome 4 in the other family, Peter concluded that CAS is caused by different genes in different families, and in some cases, different genes can have effects that pile up on individual people. Peter plans to further investigate whether or not the two main genes she found cause CAS in other families.

In the future, families might be able to identify if their child is at risk for speech disorders as an infant through genetic testing. “If we keep this work going and create a denser catalogue of genetic risks that can be checked for, then we will make it possible for kids to benefit from earlier attention to the problem and also from early interventions that still need to be developed and tested for efficacy,” said Peter. Because traditional treatments are not really possible until a diagnosis is made, Peter’s dream for the future is to develop very early interventions. These early interventions would support infants as they start to vocalize, babble, and say their first words. If all goes as Peter imagines, this type of intervention might help improve speech later on, but at the moment, is still “the stuff of the future.”

According to Peter, speech disorders caused by genetic changes are very under-researched as of yet, but she hopes this field will continue to grow. Concrete findings are difficult to obtain because these disorders vary so highly among families. Additionally, very few people have the dual training that Peter has in both speech disorders and genetics.

Peter plans to expand her future research with the use of an electroencephalogram (EEG) to look for connections between genetic variations and the time course of brain activity. She also plans to continue using magnetic resonance imaging (MRI) to identify brain areas that are active during certain speech activities. Ultimately, Peter hopes to find a strong connection among genetic variations, the way the brain works, and outward observable speech language activity.

About The Author