LWW/TLD TLD200113 July 27, 2012 1:55 Top Lang Disorders Vol. 32, No. 3, pp. 247–263 Copyright c 2012 Wolters Kluwer Health | Lippincott Williams & Wilkins Speech-Sound Disorders and Attention-Deficit/Hyperactivity Disorder Symptoms Barbara A. Lewis, Elizabeth J. Short, Sudha K. Iyengar, H. Gerry Taylor, Lisa Freebairn, Jessica Tag, Allison A. Avrich, and Catherine M. Stein Purpose: The purpose of this study was to examine the association of speech-sound disorders (SSD) with symptoms of attention-deficit/hyperactivity disorder (ADHD) by the severity of the SSD and the mode of transmission of SSD within the pedigrees of children with SSD. Participants and Methods: The participants were 412 children who were enrolled in a longitudinal family study of SSD. Children were grouped on the basis of the severity of their SSD as determined by their scores on the Goldman–Fristoe Test of Articulation and history of an SSD. Five severity groups were compared: no SSD, resolved SSD, mild SSD, mild–moderate SSD, and moderate–severe SSD. Participants were also coded for comorbid language impairment (LI), based on scores on a standardized language test. Pedigrees of children were considered to represent bilineal inheritance of disorders if there was a history for SSD on both the maternal and paternal sides of the family. Parents completed the ADHD rating scale and a developmental questionnaire for each of their children. Results and Conclusions: Children with moderate–severe SSD had higher ratings on the inattention and hyperactive/impulsivity scales than children with no SSD. Children whose family pedigrees demonstrated bilineal inheritance had higher ratings of inattention than children without bilineal inheritance. To determine the best predictors of ADHD ratings, multiple linear regression analyses were conducted. LI was more predictive of ADHD symptoms than SSD severity, bilineal inheritance of SSD, age, or gender. Findings support that LI rather than SSD is associated with ADHD. Key words ADHD, attention, genetics, language impairment, reading disorder, speech sound disorder PEECH SOUND DISORDERS (SSD), lan- S guage impairment (LI), and attention- deficit/hyperactivity disorder (ADHD) are common developmental disorders that can Author Affiliations: Departments of Psychological significantly impact academic success. The Sciences (Drs Lewis and Short and Mss Freebairn comorbidity of these disorders may substan- and Tag), Epidemiology and Biostatistics (Drs Iyengar and Stein and Ms Avrich), and Pediatrics tially add to the difficulties experienced by the (Dr Taylor), Case Western Reserve University, child and require a combined treatment ap- Cleveland, Ohio. proach. SSD and LI are often comorbid, with This research was supported by the National Institutes approximately 40%–60% of children with SSD of Health, National Institute on Deafness and Other at preschool (Shriberg & Kwiatkowski, 1994) Communication Disorders, Grant DC00528, awarded to Barbara A. Lewis. and 11%–15% of 6-year-old children with SSD presenting with LI (Shriberg, Tomblin, & Mc- The authors have no relevant financial or nonfinancial relationships to disclose. Sweeny, 1999). LI and ADHD are often co- morbid as well, with rates of comorbidity re- Corresponding Author: Barbara A. Lewis, PhD, De- partment of Psychological Sciences, Case Western Re- ported to be 30%–50% in children seen at serve University, 11635 Euclid Ave., Room 330, Cleve- ADHD clinics (Tannock & Schachar, 1996). land, OH 44106 ([email protected]). Although there are only a few reports on DOI: 10.1097/TLD.0b013e318261f086 the comorbidity of SSD and ADHD, these 247 Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. LWW/TLD TLD200113 July 27, 2012 1:55 248 TOPICS IN LANGUAGE DISORDERS/JULY–SEPTEMBER 2012 studies have suggested that LI, rather than of SSD) has on the presence of comorbid SSD, is associated with ADHD (McGrath, ADHD. Hutaff-Lee, Scott, Boada, Shriberg, & Penning- ton, 2007; Short, Freebairn, Hansen, & Lewis, 2008). THERELATIONSHIPOFADHDTOSSD Children with SSD have varied academic AND LI outcomes depending on the constellation of deficits that they exhibit. In our previous stud- The onset and prevalence of developmen- ies, we found that children with SSD in isola- tal disorders vary widely. SSD is highly preva- tion had better school-age outcomes on liter- lent in young children, with rates as high as acy measures than children with comorbid LI 15.6% of 3-year-old children (Campbell et al., (Lewis, Freebairn, & Taylor, 2002). We also 2003) declining to 3.8% of 6-year-old children demonstrated that children with SSD and co- (Shriberg et al., 1999). LI has reported preva- morbid conditions of LI or reading disorder lence estimates of 7.4% in kindergarten (males (RD) or both perform more poorly than chil- = 8%, females = 6%; Tomblin et al., 1997). dren with isolated SSD on measures of en- Prevalence estimates of ADHD vary from 1.9% dophenotypes, including phonological mem- to 17.8% of school-age children, with esti- ory, phonological awareness, and speeded mates differing depending on the nature of naming (Lewis, Avrich, Freebairn, Taylor, the sample recruited (Greenhill, Benton, & Iyengar, & Stein, 2011). Furthermore, we have Tirmizi, 2003). reported that children with SSD and comor- Children with combined SSD and LI com- bid LI are rated more poorly by their parents prise approximately 7% of the general popu- on an ADHD rating scale (ARS) than children lation (Johnson et al., 1999) and appear to be with isolated SSD (Short et al., 2008). In that at increased risk for ADHD. Reilly, Cunning- study, we examined three groups of partici- ham, Richards, Elbard, and Mahoney (1999) pants: children with SSD alone (n = 71), chil- reported that more than 30% of children diag- dren with SSD and comorbid LI (n = 77), and nosed with SSD + LI met criteria for ADHD. typical children (n = 72). Children with SSD Estimates of comorbid ADHD may vary de- + LI were more likely to have been diagnosed pending on whether the sample is clinically as- with ADHD by a health care provider, and certained or population based. Rates of ADHD to have reading problems and spelling prob- were 5 times higher in the language-disabled lems, than children with SSD alone or typi- population obtained from a speech and lan- cal children. Although the ARS was the best guage clinic than in a sample of children from predictor of ADHD, the combination of the the general population without a diagnosis ADHD rating score and a score on the Clin- of LI. Similarly, rates were quite high in a ical Evaluation of Language Fundamentals 3 prospective sample of Canadian kindergarten (CELF-3) improved the prediction of ADHD. children (Beitchman, Nair, Clegg, Ferguson, This suggests that language abilities are pre- & Patel, 1986), with 16% of children diag- dictive of ADHD over and above the ARS. nosed with both SSD and LI simultaneously However, in these studies the severity of SSD diagnosed with ADHD, as compared with and the family history for disorders were not only 6% of typical kindergarten children diag- considered. This article extends these findings nosed with ADHD. In a later study, Beitchman by examining differences in ADHD ratings in et al. (1989) reported that 59% of children children with resolved, mild, mild–moderate, with both SSD and LI had comorbid attention or moderate–severe SSD with and without deficits, whereas only 5% of children with SSD comorbid LI. Furthermore, we consider the alone demonstrated these attentional prob- impact that family history for SSD reported lems. Many children diagnosed with ADHD for both parents (i.e., bilineal transmission have both articulation and language problems Copyright © 2012 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. LWW/TLD TLD200113 July 27, 2012 1:55 Speech-Sound Disorders and ADHD 249 (Baker & Cantwell, 1992), with expressive One method of subgrouping children with language difficulties noted more frequently in SSD has been to group children on the ba- children with ADHD than receptive language sis of the severity of SSD (Lewis et al., 2011). difficulties (Barkeley, DuPaul, & McMurray, Traditionally, clinical severity ratings of SSD 1990). can be identified as mild, moderate, or se- In addition, a study by McGrath et al. (2007) vere depending upon the speech–language examined ADHD symptoms as rated by par- pathologist’s (SLP’s) judgment of the child’s ents and teachers in 108 children aged 4– intelligibility of conversational speech or the 7 years with SSD who were also diagnosed number of speech sound errors that the child with or without LI. Children with combined exhibits on a standardized articulation test. SSD and LI had higher rates of inattention Another measure of severity, which is more than children with SSD alone. It should be objective, is the Percentage of Consonants noted, however, that no differences were Correct (PCC) developed by Shriberg and found in the children with and without LI Kwiatkowski (1982). The PCC includes the on the hyperactivity/impulsivity ratings. An number of consonants produced correctly in unexpected finding was that children with a continuous speech sample divided by the SSD and LI whose speech disorder had nor- total number of consonants in the sample. malized tended to have higher rates of inat- Severity of SSD is classified as mild (85%–100% tention than children with persistent SSD. correct), mild–moderate (65%–85% correct), This suggests heterogeneity within the SSD moderate–severe (50%–65% correct), or se- population. One hypothesis proposed by Mc- vere (<50% correct). Despite the widespread Grath et al. to account for their findings usage of severity ratings to make treatment is that children with comorbid SSD and LI decisions, however, clinicians do not al- whose SSD resolves may exhibit more gen- ways agree in their perceptual ratings of a eral language learning disabilities that are ac- child’s speech (Flipsen, Hammer, & Yost, companied by inattention disorders, whereas 2005). children with persistent SSD may demon- An issue that is not yet resolved is whether strate a more circumscribed phonological children with more severe forms of SSD deficit.