A speech-language pathologist treating "Tiffany," a 6-year-old first-grader, argued that speech-language services were no longer warranted because her speech was essentially normal, except for /r/ distortions. Many first-graders struggle with this sound and soon acquire it on their own, the SLP reasoned. Tiffany, however, was different from most first-graders with /r/ problems. She had produced her first word at age 3. As a preschooler, she was painfully aware that peers weren't able to understand her speech, and she preferred to play by herself. Her older brother had experienced the same extremely delayed speech onset and social isolation. And in her family, several others, including her mother and uncle, had a history of severe speech-sound disorder, most likely of genetic origin.
Would Tiffany's residual problems with /r/ go away on their own, as in the case of many other children, or would the familial speech disorder subtype in her family require additional services?
Speech-sound disorder, language impairment, dyslexia, stuttering, and hearing impairment are all primary communication disorders that have a genetic component-in other words, in some (but not all) cases, these disorders are caused by inherited or spontaneous DNA changes. Many other disorders of genetic origin, such as autism, clefting, galactosemia, and Fragile X syndrome, can negatively influence speech and/or language development.
Not long ago, genetic origins of speech and language disorders were the stuff of anecdotal evidence from family surveys, academic speculations, and a few plastic tubes with suspended DNA in research lab freezers.
Fast-forward to the present. In 2012, clinicians providing services to clients with communication disorders routinely make diagnostic and therapeutic decisions that can and should be informed by a new wealth of knowledge about genetic etiologies. The question is how clinicians can access and use this new knowledge effectively.
A good place to start is reviewing what we know about the genetics of primary speech and language disorders, and then considering some important practical implications and limitations. As SLPs, we can positively influence the disorder progression, and knowledge of genetic etiologies can inform clinical decision-making and resource allocation.
Genetics in Primary Communication Disorders
Until recently, speech and language disorders have been viewed as complex disorders produced by multiple-gene interaction. Participant samples selected for genetics studies consisted of affected children, their siblings, and, in some cases, the parents or even a multigenerational set of relatives; most samples represented many different families. Research in hearing impairment and autism has shown that, although there may be many causal genes, in some cases, these disorders result from disruptions in one or just a few genes in a given family, as recently reviewed (McClellan & King, 2010).
A new approach to studying the genetics of speech and language disorders is to study the disorder within individual families. If there are distinct subtypes of genetic etiology, this approach is more likely to capture causal genes and is more feasible now than in the past, given new technology such as high-efficiency ("next-generation") sequencing of parts of or even the entire genome.
To appreciate the challenges of discovering causal genes, however, we should remember that very few genes have been proven to cause speech or language deficits. These include the FOXP2 gene that, when disrupted, causes a complex speech and language disorder and structural brain differences (Lai, Fisher, Hurst, Vargha-Khadem, & Monaco, 2001), and the GNPTAB gene that is causal in stuttering (Kang et al., 2010). Studies have pointed to several candidate regions and genes that still need to be validated.
Children with language impairment may struggle to comprehend what is being said, to express their thoughts in correctly formed words and sentences, and to know the meanings of words. Intriguingly, children with language impairment show slowed processing speeds across a wide variety of linguistic and nonlinguistic tasks (Miller, Kail, Leonard, & Tomblin, 2001; Miller et al., 2006). There is evidence that children with language impairment fall into different subtypes, including one with lexical and syntactic deficits and one with semantic and pragmatic deficits (Conti-Ramsden, Crutchley, & Botting, 1997). Genome-wide searches for regions that are inherited along with the disorder (linkage analyses) in families with language impairment have pointed to candidate regions on chromosomes 12, 13, 16, and 19.
The discovery that disruptions in the FOXP2 gene cause a complex speech and language disorder, together with the fact that this gene is a transcription factor that regulates the functions of other genes, has led to the hypothesis that some of the genes under the control of FOXP2 might influence speech and language functions in more specific and less syndromic ways. This hypothesis was supported when a team in the United Kingdom (Vernes et al., 2008) chose CNTNAP2, one of the genes regulated by FOXP2, as a candidate gene for language impairment. In particular, the ability to imitate nonwords was strongly associated with markers in this gene. This study suggested that CNTNAP2 may influence isolated abilities important for language functions, whereas FOXP2 may influence multiple abilities simultaneously, perhaps because it regulates various other genes, which can have multiple effects.
People with dyslexia typically have difficulty sounding out words and recognizing words that do not follow standard spelling rules. Observations based on neural imaging are consistent with the idea that dyslexia has a biological origin: The brains of individuals with dyslexia show activation patterns during reading tasks that are different from the patterns exhibited by individuals without dyslexia. Of three left-hemisphere regions that are typically active during reading, two are underactive in many individuals with dyslexia, and some show more activation in the right hemisphere instead (Richlan, Kronbichler, & Wimmer, 2009). There are also differences in brain structures regarding the organization of the cortical neurons and the density and shape of left-hemisphere white matter and gray matter (for a review, see Catts, Kamhi, & Adlof, 2012).
People with dyslexia have difficulty perceiving discrete units of the speech stream such as phonemes and syllables. In fact, differences in event-related brain responses to sound have been found in newborn infants at familial risk for dyslexia. Children with dyslexia often have difficulty accessing their mental vocabulary quickly when shown a series of pictures for a rapid-naming task. There also is evidence that children with dyslexia, like children with language impairment, process information at slightly slower rates than peers without dyslexia (see Peterson & Pennington, 2012, for a summary of these findings).
The fact that not all children with deficits in auditory perception of speech stream units and/or rapid naming develop dyslexia is taken as evidence that dyslexia may result from a variety of underlying deficits such as a general learning disability, attention deficits, or reduced processing speeds. It is also possible that there are distinctly different subtypes of dyslexia.
Given its biologically based characteristics, it is not surprising to note that dyslexia runs in families and appears to be hereditary. Genome-wide linkage analyses have suggested nine candidate regions across the genome, labeled DYX1 through DYX9, and replicated in several studies. Within these, six candidate genes have been suggested, some of which are important during embryonic brain development. New potential candidate regions continue to be described. For instance, in a dyslexia family sample, performance during a rapid-naming task was associated with a region near to, but not identical with, DYX2 on chromosome 6 (König et al., 2011).
Speech Sound Disorder
Nearly every pediatric SLP regularly sees children who struggle to produce speech sounds correctly. Speech sound disorder is thought to have various subtypes. Children who consistently misarticulate certain sounds—for instance a lateral lisp during /s, z/—are said to have an articulation disorder, whereas children who produce errors only in certain contexts such as clusters or at the ends of words are said to have a phonologically based disorder. Another subtype is childhood apraxia of speech, characterized, in part, by highly unintelligible speech, multiple speech sounds including vowels produced incorrectly, and robotic-sounding prosody.
Molecular genetic studies in children with speech sound disorder have been attempted only fairly recently (Miscimarra et al., 2007; Smith, Pennington, & Shriberg, 2005; Stein et al., 2006). Based on the rationale that many preschoolers with speech sound disorder later struggle with learning to read, some of the genomic regions implicated in dyslexia (DYX1, DYX2, DYX5, and DYX 8) were selected as candidate regions for speech sound disorder, and linkage analyses within these narrow regions indeed showed evidence of linkage. These findings may indicate that children with speech sound disorder and dyslexia share not only some observable deficits but also some genetic risk factors.
A recent genetic study to include multigenerational families with speech sound disorder (Peter & Raskind, 2011) showed that some of the adult relatives of children with childhood apraxia of speech had slowed syllable durations during repetition of multisyllables (e.g., /pata/), compared to monosyllables (/papa/), and an analogous deficit was observed during a task involving tapping two computer keys in alternation, compared to tapping a single key repetitively. The first genome-wide linkage analysis in speech sound disorder was attempted in one of these families, using the motor deficit during alternating oral and hand tasks as the variable of interest. The analysis revealed four new regions of interest (Peter, Matsushita, & Raskind, 2012), one of which overlapped with a new dyslexia region on chromosome 6 (König et al., 2011). This new approach to speech sound disorder genetics has led to genome-wide next-generation studies in progress.
Co-occurrence of Speech and Language Disorders
Speech, language, and/or reading disorders co-occur at slightly higher rates than expected by chance, which implies that there may be shared causes between any two or all three of these disorders. Consistent with this hypothesis are findings that, in a sample of families with language impairment, showed suggestive associations between a dyslexia candidate gene and measures of language and reading ability (Rice, Smith, & Gayán, 2009), and targeted studies in children with speech sound disorders that found evidence of linkage in the target regions, which had previously been implicated in dyslexia. It is thought that several of the candidate genes for dyslexia influence the formation of brain structures during embryonic development, and differences in brain structures have been observed in individuals with dyslexia.
Brain imaging studies have not yet been attempted widely in individuals with speech sound disorder. It would be interesting to see if similar gene-phenotype correspondences emerge here. The candidate regions for language impairment surprisingly do not overlap with those for dyslexia. Recently, a candidate gene in the DYX2 region, CYP19A1, was investigated in samples of families with dyslexia, language impairment, and/or speech sound disorder (Anthoni et al., 2012). The gene was found to be associated with measures of speech, language, and reading. The authors show that CYP19A1 defects in mice lead to structural brain disorganization and suggest that variants of this gene may influence speech, language, and reading ability in humans because of its global role in the development of brain regions crucial for communication functions.
Clinical Dos and Don'ts
Unlike genetic forms of hearing impairment, which often can be detected by newborn hearing screenings and for which many causal genes have already been identified, speech and language impairments do not manifest physically until much later in childhood. Only very few causal gene variants can be identified in genetic testing. Nonetheless, clinicians can use current knowledge of genetic risks to enhance clinical management.
- Ask about family history. Most physicians include family history of common diseases in the patient intake forms. A patient's family history provides an estimate of the patient's own medical risk factors, which can lead to advice on preventive measures, an earlier diagnosis of a looming disease, and a more targeted course of treatment. Similarly, an initial speech-language assessment should include information about relevant family history. For instance, a clinician evaluating a child due to concerns about difficult-to-understand speech should ask whether other family members have had any difficulty with speech, language, learning to read, or hearing. This information may be valuable when selecting a specific diagnosis such as childhood apraxia of speech, deciding whether a wait-and-see approach is justifiable, educating parents about enhancing a child's environment with opportunities to hear and practice speech clearly, and planning a course of intervention. Note, however, that information about others in the family does not replace careful observation and direct assessment of the client. Note also that genetic disorders can appear spontaneously without any family history.
- Ask about any health concerns. In some cases, a child's struggles with speech and/or language are secondary to another diagnosis such as autism, CHARGE syndrome, or fragile X syndrome. SLPs suspecting such a syndrome based on observations of the child's physical presentation or communication behaviors may be able to refer the child for helpful medical consults.
- Identify at-risk children early. In families with several members who have a communication disorder of a known genetic origin, it is possible in some cases to determine whether a new member of the family carries the risk genotype. Keep in mind, however, that the presence of a risk genotype does not automatically predict the emergence of a disorder. Gene changes can have variable penetrance rates, which are not yet known for all causal genes that have been identified. In many families with a communication disorder, the causal gene is unknown. In these cases, young children should be carefully observed for any early signs of the disorder. Tiffany's youngest brother, for example, had not produced any consonantal babble until age 13 months, a sign of delayed prespeech development and an early indication that he was at risk for the speech disorder in this family.
- Educate the client and the client's parents. Clinicians caring for individuals with a communication disorder of genetic origin can provide valuable information, including education on typical development and abilities so that clients and parents can recognize signs of a communication disorder early. For example, Tiffany's 3-year-old sister could say many understandable words. Her parents assumed that her speech was developing normally because Tiffany and her older brother had only begun producing words by that age. It turned out that the 3-year-old sister had a severe speech sound disorder, just not quite as severe as her older siblings. The parents needed help recognizing the signs of the disorder so that they could pursue treatment. Clinicians should also provide advice on creating an environment rich in opportunities to thrive. Professional preventive intervention may not always be an option when a young child is at familial or even genetic risk for a speech or language disorder. In these cases, parents can learn to observe their child carefully for early signs of delays and provide an environment rich in opportunities for verbal communication and/or interaction with written language.
- Initiate appropriate referrals for genetic counseling. SLPs are not licensed to provide genetic counseling, but genetic counselors are. These professionals, typically holding master's degrees in their specialized field, can answer questions about the medical aspects of disorders, the prognosis, and the probability that subsequent children in a given family with a certain disorder will also have that disorder. The National Society of Genetic Counselors promotes genetic and genomic services. Its website provides an overview of the scope of the profession, as well as information on locating a genetic counselor with expertise in certain subspecialties such as neurogenetics, metabolic disorder, psychiatric genetics, pediatric genetics, newborn screening, complex disorders in adults, and specialty diseases. In contrast to secondary causes of speech and language deficits, primary speech and language disorders are not yet an established area of clinical practice for genetic counselors, but as more genetic discoveries are made, the scope of practice in genetic counseling is constantly expanding. SLPs and audiologists are often the first professionals to notice signs of a genetic syndrome and can advise clients to consult a genetic counselor, typically by asking their physician for a formal referral.
- Be aware of the limitations of mail-order genetic screens. Personal genomics providers (direct-to-consumer genetics companies) incorporate a limited number of risk markers for primary speech and language impairments in their genetic analysis programs (for example, selected markers on the FOXP2 gene, which influences speech and language development, and on the KIAA019 gene, which influences dyslexia; markers for syndromic diseases that influence speech and language later in life, such as multiple sclerosis and Parkinson's disease). Because causal genes have not yet been identified in most types of primary communication disorders, no genetic screen can conclusively detect causal variants in someone's DNA. For hearing impairments, the story is slightly different. Of 100 genes associated with nonsyndromic hearing loss, three (GJB2, GJB6, and SLC26A4) account for more than a third of congenital cases, and these are typically included in direct-to-consumer genetic screens. Nonetheless, failure to detect a causal variant is a possibility with this type of method.
Into the Future
Causal genes for speech and language disorders will likely be identified and validated in the near future. In the process, we may learn that speech and language disorders result from various different gene disruptions in different families. Armed with this knowledge, we can be more informed in our treatment approaches with clients such as our earlier example of Tiffany, the 6-year-old with the familial speech disorder that is traumatizing her social development as a preschooler.
When Tiffany's SLP recommended ending treatment, her parents made the case that her older brother and mother had both needed speech-language treatment to acquire all speech sounds. Unfortunately, Tiffany could not be re-qualified for services under state regulations. She received speech-language materials for home practice, and her parents were advised to have her re-evaluated for eligibility if her speech did not improve.
To help children like Tiffany and her siblings, clinicians, researchers, and legislators need to embark on a new journey. Once causal genes are identified, we can build a disorder subtype catalog based on genetic causes and associated features. Then we can identify children with risk genotypes at very early ages.
Our challenge will be to develop early intervention approaches that reliably lessen or eliminate the expression of the disorder. For children like Tiffany's youngest brother, this approach may take the form of stimulating babble and vocal play and providing tactile support for first words.
By comparing the efficacy of early intervention in very young children to that of intervention in older children with a diagnosed disorder, we can inform health care policy regarding resource allocation. But all this is still the stuff of the future.