Relationship Between Handedness and Language Function in Autism

Friday, May 15, 2015: 11:30 AM-1:30 PM
Imperial Ballroom (Grand America Hotel)
L. M. McCue1, A. R. McMichael2, L. H. Flick1, D. L. Williams3 and T. E. Conturo4, (1)St. Louis University, St. Louis, MO, (2)Washington University, St Louis, MO, (3)Duquesne University, Pittsburgh, PA, (4)Washington University, St. Louis, MO

Language impairment is an important core symptom in autism spectrum disorder (ASD).  It is imperative to understand its biological mechanisms in order to find early predictors of language dysfunction (1) and devise early treatments.  Handedness-language relations may give insight into biological mechanisms of language in ASD.  Handedness and language may be linked in various ways.  First, the hemispheres used for cognitive and motor aspects of language depend on handedness. In right-handers, structural language is strongly left-hemisphere dominant (2,3), but in left-handers is weakly hemisphere-dominant (4) and frequently  bi-hemispheric (3). Language prosody is typically right-hemispheric (e.g., 5,6,7).  In contrast, covert/overt motor speech output uses left premotor/motor areas in right-handers, and vice-versa (8).  Thus, depending on handedness some aspects of language may require interhemispheric connectivity and information flow, which may be compromised in ASD.  Second, the presence/absence of hand preference may indicate degree of hemispheric specialization.  Decreased hemispheric specialization may be a cause of language impairments in ASD.  Thus, handedness-language relationships may provide a window into two possible mechanisms of language impairment: decreased interhemispheric connectivity, and decreased hemispheric specialization. There are few studies of handedness-language relationships in autism (9-11), which are typically limited by too few left-handers.


Test for associations between handedness (right/left/ambiguous) and language in a large sample of children with well-verified ASD from simplex and multiplex families.


We analyzed data from the Autism Genetic Resource Exchange (AGRE) database.  Because handedness presents at ~5 years (12), we chose a minimum age of 7 (to reduce effects from any handedness delay) and a maximum age of 17.  Our sample was N=403 children with ASD.  Children were grouped into normal (OLANG=0) and abnormal (OLANG>0) language on the basis of the ADOS Overall Level of Non-Echoic Language.  Children with inadequate language for assessment (OLANG=8) were excluded.  Handedness was coded as left (N=35) vs right, and then as ambidextrous (N=30) vs hand-dominant.  The non-independence of observations inherent with familial data  was accounted for using PROC SURVEY procedures (SAS v9.4).


Among children with ASD, right-handers had  2.18 times the odds of abnormal language compared  to left-handers (odds-ratio=2.18; 95%-confidence-interval 1.07-4.44; p=0.03). There was no association between language and ambidexterity  (odds-ratio=0.65; 95%-confidence-interval: 0.31-1.58; p=0.30).  The decreased language function in right-handers could be explained by decreased connectivity from right-hemispheric (prosodic) language areas to left-hemispheric motor speech areas (consistent with lack of prosody in ASD). In contrast, in left-handers, language areas are more commonly bilateral, and language areas can thus access right-hemispheric motor speech areas without interhemispheric transfer.


Decreased language function in right-handers versus left-handers with ASD suggests decreased interhemispheric connectivity as a mechanism.  The lack of ambidexterity effect suggests that reduced hemispheric specialization is a less likely mechanism. These results show the utility of hand preference data for interpreting interhemispheric connectivity and hemispheric specialization.


[1]Stone & Yoder (2001); [2]Knecht et.al (2000a); [3]Pujol et.al (1999); [4]Knecht et.al (2000b); [5]Shapiro & Danlye (1985); [6]Behrens (1989); [7]Pell (2007); [8]Longcamp et.al (2005); [9]Tsai (1983); [10]Escalante-Mead et.al (2003); [11]Lindell & Hudry (2013); [12]McManus et.al (1988).