Language modality and development

Studying language development in individuals who do not communicate verbally can help us understand how language modality (the way language is conveyed) relates to the brain and vice versa (4:15). Language delays are one of the earliest emerging endophenotypes (biomarker subsets) of autism. McFayden explains that standard models of language development employ the “wait and see” method, which is strictly behavior-based. Sometimes, practitioners wait too long and miss opportunities for the most effective early interventions (5:36). By investigating non-verbal language modality across time, the speaker continues, we can discover where language phenotypes begin to diverge from the neurotypical population and, therefore, find the most effective timing for early interventions (6:00). 

Siblings of autistic children (Sib-ASD) are 3 to 4 times more likely to experience language delays than siblings of non-autistic children. Similarly, 20 – 30% of Sib-ASD receive an autism diagnosis (7:00). Therefore, younger siblings of children with autism are considered a high likelihood (HL) group. In McFayden’s ongoing study, her team observes white matter development across time for groups at high likelihood (HL) and low likelihood (LL) for autism diagnosis (2:25). The speaker emphasizes that white matter measurements allow a mechanistic approach to language development, which may be visible earlier than behavioral presentations (7:25). 

White matter

McFayden describes white matter as the portion of the brain that is myelinated or enclosed in myelin. This fatty lipid-protein gives white matter its color (8:09). These myeline sheaths act as electrical insulators that support signal transduction in the brain and, as such, the formation of functional circuits. Therefore, white matter is critical to understanding how information is communicated throughout the brain (8:25). White matter is measured using an MRI to record fractional anisotropy (FA), an index of the rapid maturation of fiber pathways in the brain. Higher FA relates to increased age, myelination, and fiber maturation (9:15). 

White matter in autism

One of the earliest differences documented in autism is atypical organization of white matter, in some cases evident by six months post-natal (10:30). Behavioral implications of white matter organization discussed in the literature include social cognition, behavioral control, repetitive behaviors, and sensory interests. However, no work to date has mapped white matter development to language despite these occurring at similar times (11:40). 

White matter in language

White matter has been linked to expressive and receptive language in the neurotypical population, specifically in the arcuate fasciculus and corpus callosum. Generally, higher axon insulation (myelination) corresponds to higher language scores (12:10). McFayden reiterates the importance of evaluating the relationship between white matter and language to understand the pathophysiology and inform effective and timely language interventions for autism (13:00). 

Study: White matter and language development over time

The speaker outlines her teams’ ongoing study to evaluate longitudinal relationships between white matter microstructure in infants at HL (high likelihood) and LL (low likelihood) for autism at 6, 12, and 24 months. Specifically, they aim to answer two questions (13:30): 

• Which white matter structures are related to expressive and receptive language in the first two years postnatal?
• Do relationships between white matter and language over time differ as a function of the likelihood group? If so, when?

Data and analysis

Study data are gathered from the Infant Brain Imaging Study (IBIS), an ongoing (15 years and counting) investigation that follows younger siblings of autistic and neurotypical children at 6, 12, and 24 months (14:00). The speaker and her team separated 461 participants into three groups (15:00):

• HL-ASD: received a diagnosis at 24 months
• HL-Neg: did not receive a diagnosis
• LL-Neg: siblings of neurotypical children 

Researchers used the Mullen Scales of Early Learning to derive values for expressive and receptive language (EL/RL). They recorded FA values from nine common brain connection pathways to measure white matter volume (16:15). The speaker outlines analytical methods, highlighting linear mixed-effect modeling, piecewise linear estimates, and autoregressive correlation structure (17:10). White matter and language were considered main effects and used to assess how changes in white matter and language impact certain groups compared to others (18:25). 

Results 

Analysis revealed a unique significant relationship between language and the arcuate fasciculus (19:35). McFayden presents a data plot where participant FA values are plotted against expressive language measures at each time scale. Preliminary findings show a positive relationship between FA and expressive language in the HL-ASD group. This relationship diverges from the HL-Neg group at twelve months and the LL-Neg group at six months (20:53). The relationship between white matter and expressive language is steeply positive in the HL-ASD group compared to HL-Neg and LL-Neg groups (22:47). In a model of predictive language scores based on collected data, the HL-ASD group also breaks away in the third FA percentile (23:15).

Researchers did not find any significant interactions between FA values and receptive language for any group (24:06). However, group trends for receptive language exhibit the same patterns observed in the expressive language plots. The speaker asserts that, although the relationship did not reach statistical significance, the similarities in receptive and expressive language trends compared to white matter warrant further attention (25:00). 

Summary and discussion

McFayden reiterates the significant relationship between white matter and language in autism and notes the differences between expressive and receptive language (25:40). A robust positive relationship between language and white matter volume emerges in autism between six and twelve months postnatal and remains through the two-year mark. She explains that HL-ASD infants with the smallest gains in FA values exhibit the least gains in language over time and overall lowest language scores by 24 months. These findings underscore the importance of appropriately timed interventions (26:40). Because findings were specific to the right arcuate (expressive language), we can conclude that expressive and receptive language (left arcuate) do not have the same tractography (27:40). The speaker mentions evidence suggesting that left acuate lateralization does not happen to the same extent in autism, which could reflect right-dominant language processing (29:12). 

Future directions and key takeaways

Future investigations should use a smaller time window (e.g., 18 months) and larger HL-ASD samples to more fully understand the temporal aspect of this relationship. The speaker also suggests consistently observing brain-behavior relationships through school age (30:45). Finally, HL-ASD groups could be further parsed out according to language trajectory to reveal more subtle differences in the language-white matter endophenotype (31:28). 

McFayden reemphasizes that the development of white matter in infants who later receive an autism diagnosis is linked to lower expressive and receptive language in the first two years of life. These specific language relationships emerge as early as twelve months, which is well before diagnoses are generally made (32:20). The speaker asserts that these findings support the idea that the “wait and see” method is not a good approach. She notes that this is a new field of study and suggests incorporating these data into future observation and diagnosis methods to ensure proper intervention timing (33:00). 

McFayden provides thanks and acknowledgments before the Q&A, where she discusses preterm babies (40:30), ongoing studies and where to sign up (42:00), potential subgroups and language trajectories (47:00), and much more.