Top Lang Disorders Vol. 31, No. 2, pp. 145–167 Copyright c 2011 Wolters Kluwer Health | Lippincott Williams & Wilkins Evidence for a Familial Speech Sound Disorder Subtype in a Multigenerational Study of Oral and Hand Motor Sequencing Ability

Beate Peter and Wendy H. Raskind

Purpose: To evaluate phenotypic expressions of speech sound disorder (SSD) in multigenerational families with evidence of familial forms of SSD. Method: Members of five multigenerational families (N = 36) produced rapid sequences of monosyllables and disyllables and tapped computer keys with repetitive and alternating movements. Results: Measures of repetitive and alternating motor speed were correlated within and between the two motor systems. Repetitive and alternating motor speeds increased in children and decreased in adults as a function of age. In two families with children who had severe speech deficits consistent with disrupted praxis, slowed alternating, but not repetitive, oral movements characterized most of the affected children and adults with a history of SSD, and slowed alternating hand movements were seen in some of the biologically related participants as well. Conclusion: Results are consistent with a familial motor-based SSD subtype with incomplete penetrance, motivating new clinical questions about motor-based intervention not only in the oral but also the limb system. Key words: common disease/common variant, common disease/rare variant, disorder subtype, familial phenotype, motor sequencing, motor speed, speech sound disorder SPEECH SOUND DISORDER (SSD) DEFINITION

Speech sound disorder (SSD) is a childhood Author Affiliation: Departments of Speech and communication disorder that interferes with Hearing Sciences (Dr Peter), Medicine (Dr Raskind), the development of speech sound production Psychiatry and Behavioral Sciences (Dr Raskind), in the absence of known causes (Pennington University of Washington, Seattle. & Bishop, 2009). The disorder can be char- The authors thank the families whose participation acterized by deficits in articulation, phono- made this study possible. Many thanks to the following undergraduate and graduate students for their assis- logical processing, and/or cognitive repre- tance with data collection and analysis: Leah Anderson, sentation of language (Lewis et al., 2006). Lynn Bak, Yayin Chen, Erica Gonzales, Mariya Legesse, Speech sound disorder is common, although Amelie Lehmkuhler,¨ Jonathan Mahaffie, David Ramm, and Nancy Yuan. The software for the key tapping task published prevalence rates vary; for instance, was designed by Elias Peter. The authors are grateful 1.1% in Australian school-age children (McK- to the following funding sources: NIDCD T32DC00033 innon, McLeod, & Reilly, 2007), 4% in US 6- (B. Peter), American Speech-Language-Hearing Foun- dation New Century Scholars Research Grant (B. Peter), year-olds (Shriberg, Tomblin, & McSweeny, and R01HD054562 (W. H. Raskind). 1999), and 15.6% in US preschoolers (Camp- Corresponding Author: Beate Peter, PhD, CCC-SLP, bell et al., 2003). Differences in prevalence Department of Speech and Hearing Sciences, Box estimates arise, in part, from differences in 354875, University of Washington, Seattle, WA 98195 the age of the children under study, criteria ([email protected]). for participation, and assessment methodol- DOI: 10.1097/TLD.0b013e318217b855 ogy. Speech sound disorder co-occurs with

145 146 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011 language impairment and dyslexia at rates catalogue of diagnostic criteria is not avail- greater than expected by chance (Pennington able. & Bishop, 2009; Peterson, McGrath, Smith, & As mentioned in the ASHA position state- Pennington, 2007). ment, planning and programming of motor A universally accepted SSD subtype clas- sequences in the oral motor system appears sification is not yet available. Several clas- to be disrupted by CAS. An example is the sification schemes coexist in the literature. finding that repetition rates of the trisyl- Examples include taxonomies based on er- labic sequence /pataka/ as well as trisyllable/ ror types, a phonetic-phonemic continuum, monosyllable rate ratios effectively distin- and suspected etiologies, as recently reviewed guished children with CAS from typical con- (Peter, 2010; Peter & Stoel-Gammon, 2008). trols, whereas there was no group differ- Subtypes based on familial or genetic findings ence in repetition rates for monosyllables have not yet been described. (Thoonen, Maassen, Gabreels, & Schreuder, 1999; Thoonen, Maassen, Wit, Gabreels, & Childhood apraxia of speech as a Schreuder, 1996). In a study of 11 children motor-based SSD subtype with SSD, timing accuracy of oral movements One proposed SSD subtype is childhood was correlated with timing accuracy in hand apraxia of speech (CAS), which is thought movements, and the lowest timing accuracy to interfere with motor planning and/or mo- in both motor systems was observed in chil- tor programming processes related to speech dren with the highest number of CAS charac- production. In its position statement on CAS, teristics (Peter & Stoel-Gammon, 2008). It is the American Speech-Language-Hearing As- possible that CAS has a genetic etiology. In a sociation (http://www.asha.org/docs/html/ study of 11 children with CAS, six had a fam- PS2007–00277.html) indicated a view of CAS ily history of speech and language difficulties as a “distinct diagnostic subtype of childhood (Thoonen, Maassen, Gabreels, Schreuder, & (pediatric) speech sound disorder” and de- de Swart, 1997). fined it as follows: In view of the fact that there is no univer- sal consensus regarding how to define SSD [CAS] is a neurological childhood (pediatric) subtypes, a genetic etiology would provide speech sound disorder in which the precision and consistency of movements underlying speech are a new SSD subtype classification based on impaired in the absence of neuromuscular deficits a biological model. Consequently, a central (e.g., abnormal reflexes, abnormal tone). CAS may question motivating this study was whether occur as a result of known neurological impair- distinct SSD subtypes aggregate in families, ment, in association with complex neurobehav- which would be consistent with a subtype- ioral disorders of known or unknown origin, or as specific genetic etiology. an idiopathic neurogenic speech sound disorder. The core impairment in planning and/or program- Toward SSD subtypes based on ming spatiotemporal parameters of movement se- genetic etiology quences results in errors in speech sound produc- It is thought that SSD in general has a ge- tion and prosody. netic component, but family-specific traits or Although this definition views all forms of causal genes have not yet been identified. CAS as a subtype of SSD, some researchers Higher SSD concordance rates in monozy- (Potter, Lazarus, Johnson, Steiner, & Shriberg, gotic versus dizygotic twins (Lewis, 1992), 2008; Shriberg, Potter, & Strand, 2010; Ter- high SSD heritability estimates (Bishop, 2002), band, Maassen, van Lieshout, & Nijland, 2011) and higher susceptibility in biological ver- hold a more narrow view in which primary, sus adopted children with affected parents nonacquired forms of CAS form one SSD sub- (Felsenfeld & Plomin, 1997) provided early type and secondary, acquired forms are out- evidence on the basis of behavioral observa- side the SSD definition. A universally accepted tions that SSD has genetic influences. Familial Speech Sound Disorder Subtype 147

More recently, studies have addressed the rate assumption for only a subset of cases molecular genetics of speech problems. In of familial SSD. The common disease/rare rare cases, speech production difficulties of variant model assumes that multiple com- genetic origin are part of a syndrome. FOXP2 mon variants interact to confer disease sus- mutations cause moderate to severe diffi- ceptibility in a single individual. In the studies culties with speech, receptive and expres- mentioned here, the participants were chil- sive language, reading, writing, cognition, and dren with SSD and their siblings. Many dif- oral praxis (Fisher, Vargha-Khadem, Watkins, ferent families were represented in each sam- Monaco, & Pembrey, 1998; Lai, Fisher, Hurst, ple. The disorder criteria were general, requir- Vargha-Khadem, & Monaco, 2001; MacDer- ing low scores on standardized articulation mot et al., 2005; Vargha-Khadem, Watkins, Al- and phonology tests. It is possible that the cock, Fletcher, & Passingham, 1995; Watkins, samples contained several familial subtypes Dronkers, & Vargha-Khadem, 2002). Changes with varying expressions of SSD. One of the in this gene, however, do not explain non- studies reported differential linkage patterns syndromic SSD. In one study of 49 individuals among subgroups of sibling pairs, where the with disordered speech with apraxic charac- probands were preschool age (Stein et al., teristics, only three related individuals had a 2004). Of the 77 sibling pairs, 34 contributed protein-altering FOXP2 variant (MacDermot to linkage results for both multisyllabic word et al., 2005). repetition (MWR) and nonsense word repeti- As recent reviews have indicated (Lewis tion, skills that are characteristically impaired et al., 2006; Pennington & Bishop, 2009; Pe- in SSD; Nine of these sibling pairs were con- ter, 2010), the genetic mechanisms causing cordantly affected (i.e., both children in each common, nonsyndromic forms of SSD are pair had low MWR scores) and two were con- not yet well understood. Several studies con- cordantly unaffected (i.e., both children in ducted to investigate gene locations involved each pair had typical MWR scores). Given an in primary SSD have focused on candidate estimated SSD prevalence of 16% at preschool regions implicated in other disorders with age, the concordantly affected group may rep- comorbid speech deficits (Miscimarra et al., resent an SSD subtype with a phonological 2007; Smith, Pennington, Boada, & Shriberg, memory component, with an estimated preva- 2005; Stein et al., 2004, 2006). Study sam- lence rate of 1.2%, and in the concordantly ples have consisted of affected children and unaffected group with SSD without such a their siblings. Genetic linkage analyses tar- phonological memory component, that rate geted candidate regions on chromosomes 1, would be 0.4% (unadjusted for ascertainment 3, 6, and 15. These were selected because bias). There are, hence, compelling reasons to of suspected linkage to other disorders, such consider the common disease/rare variant as dyslexia and Angelman syndrome, which model that posits that rare variants in individ- share phenotypic traits with SSD. Evidence ual families, in aggregate, explain many cases was obtained for linkage of deficits in speech of SSD. It is possible that in a subset of familial articulation, which are pathognomonic for SSD cases, different genetic mechanisms may SSD, to some of these candidate regions, but be at work in different families, which would causal genes have not been identified to date. be consistent with SSD as a heterogeneous The authors’ collective findings led to the hy- disorder rather than a complex disorder. pothesis that SSD and dyslexia share a com- The common disease/rare variant hy- mon genetic basis but are also influenced by pothesis of SSD is in line with a general variants unique to each disorder. paradigm shift in thinking about genetic vari- The traditional view of SSD as a common ation in complex diseases (McClellan & King, and complex disorder (Stein et al., 2004), 2010). Inherited hearing loss is one example in accordance with the common disease/ of this type of genotype/phenotype associ- common variant model, may be an accu- ation (Dror & Avraham, 2009). Any one of 148 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011 dozens of genes, when disrupted, can cause The broad purpose of the current research hearing loss in a given family. Depending on project was to investigate familial forms the gene, the pattern of inheritance can be of SSD in multigenerational families using dominant or recessive, and the gene locus evidence from behavioral observations as can be autosomal, X linked, or mitochondrial. well as from DNA analyses. In this first study, Despite the divergent genetic etiology and we describe phenotypic expressions of SSD the different types of biochemical disruptions within single families, focusing on aspects in the inner ear across various families, the of motor processing in the oral and hand effect on the auditory system is the same, in motor systems. Future reports will focus on that hearing loss results. Similarly, many cases additional behavioral traits and molecular of inherited autism can be explained individ- genetic findings. ually by rare causal variants in a given family Specifically, this study addresses the role of (Bucan et al., 2009). Regarding SSD, the com- repetitive and sequential motor speeds in fam- mondisease/rarevariantmodel is consis- ilies where one or more children have been tent with the discrepancy between the strong diagnosed with SSD and where other biolog- evidence of a genetic etiology based on twin ical relatives also have a history of SSD. We (Lewis, 1992), adoption (Felsenfeld & Plomin, hypothesized that, in a subset of the families 1997), and heritability studies (Bishop, 2002) who participated in this study, limited motor on one hand and the lack of unambiguous speeds in repetitive and/or alternating move- linkage peaks across samples on the other. ments might be directly observable in cur- In the past, methodological challenges to rently affected children as well as in adults studies of SSD genetics in samples represent- with a childhood history of speech difficul- ing multiple families have included the likeli- ties whose speech has since normalized. Hand hood of heterogeneity, unknown penetrance tasks were included in this study because re- levels, unknown mode of inheritance, rapidly cent work shows close associations between changing speech error profiles as a result of rapid oral and hand motor performance. For therapy, and the inclusion of adult partici- instance, temporal aspects in a nonword im- pants, given that they typically have compen- itation task and a handclap task were signif- sated for their disability. This latter aspect may icantly correlated in a sample of young chil- explain why studies have investigated SSD in dren with SSD and typical controls (Peter & affected children and their siblings, not multi- Stoel-Gammon, 2008). Performance on a se- generational families, even though such fami- quential oral motor task of producing rapid lies typically offer substantially greater power repetitions of the trisyllable /pataka/ and a to detect shared inherited gene regions in af- rapid sequential hand motor task was asso- fected relatives, greater accuracy in gene lo- ciated with a latent dimension characterized calization, and greater accuracy of parameter by central processing speed in a large family estimation than sibling pairs and nuclear fam- sample ascertained through proband children ilies (Wijsman & Amos, 1997). Fortunately, with dyslexia (Peter, Matsushita, & Raskind, new assessment tools now make it possible 2010). It is possible that motor deficits in the to address the question of how to incorporate oral and hand systems characterize some SSD adults with a history of SSD into investigations subtypes in children with currently expressed of SSD genetics. In a recent study, 36 adults SSD, as well as related adults with a history of with a history of SSD obtained significantly SSD but normalized speech. lower scores in taxing speech tasks such as In repetitive motor speech tasks, such as imitating multisyllabic words, nonwords, and producing sequences of monosyllables, rate tongue twisters than 144 adults without such is limited by various factors, including speed a history (Lewis et al., 2007). It is unknown and distance of articulator excursion. Be- what other, if any, characteristics are retained cause of the nature of the task, the same by adults with resolved childhood SSD. articulators repetitively produce the airflow Familial Speech Sound Disorder Subtype 149 constriction underlying the acoustic signal. speech production process, but a universally Similarly, striking a computer key repetitively accepted subtype scheme is not yet avail- with the same finger is rate limited by speed able. If familial SSD characteristics could be and excursion factors and the fact that the described in affected children and related same finger is used to strike the key. The key adults with a history of SSD, this would con- itself contributes yet another rate limit with tribute to a novel, biologically based SSD sub- its depression depth, resistance, and rebound type scheme. Deficits in oral motor sequenc- speed. In alternating movements, by contrast, ing have been described in children with multiple agents are coordinated, resulting speech errors consistent with CAS, which is in more degrees of freedom and potentially a proposed SSD subtype. If motor sequencing shorter intervals. For instance, the disyllable deficits are systemic, that is, not just confined /pata/ involves bilabial lip closure alternating to the oral motor system but also observable with tongue/alveolar ridge contact, resulting in other motor systems, and if they persist as in two separate repetitive movement oscilla- such in adults with a history of a motor-based tions interleaved in time. A meta-analysis of form of SSD, this would constitute a subtype the norms for monosyllables and disyllables of SSD that is familial and likely of genetic in children aged 6–13 years (Fletcher, 1972) origin. This study, hence, addresses the fol- shows that syllable durations decrease in gen- lowing research questions: eral as a function of age and that durations 1. Are motor speeds in repetitive and al- of the disyllables are as long as, or longer ternating tasks associated across motor than, those in monosyllables in the younger systems? If so, this would confirm previ- children, but by age 11 years the disyllabic ous reports of cross-system correlations syllables are substantially shorter, indicating and strengthen our hypothesis that mo- a speed advantage in alternating oral move- tor speeds across systems are controlled ments. The same advantage for trisyllables is by a shared biological mechanism. approached, but not reached, by age 13 years. 2. Do repetitive and/or alternating motor Similarly, striking two keys by alternating be- speeds distinguish between adults with tween two fingers involves two repetition cy- and without a history of SSD in general? If cles, one for each finger. With intact motor se- so, motor deficits would represent one of quencing skill, individual syllables and tap in- the long-term sequelae of SSD in adults. tervals in an alternating task can be expected 3. Within individual families, do adults with to be just as short as, or shorter than, the inter- a history of speech difficulties exhibit vals in a repetitive task. Parallel to the norms similar motor rate limits as the related af- for monosyllable repetition rates in children, fected children in the same family? If so, norms for repetitive key tapping intervals in this would be consistent with a familial children show that interval durations decrease motor-based SSD subtype. as a function of age (Gualtieri & Johnson, To facilitate a meaningful comparison of 2006; Prigatano, Gray, & Legacy, 2008), then motormeasuresinchildrenandadults,anad- increase again in late adulthood (Bartzokis ditional goal of this study was to develop mo- et al., 2008); norms for alternating key strikes tor measures that were not readily available. are not available. In individuals with motor Thus, this article not only addresses ques- sequencing difficulty such as individuals with tions regarding familial SSD subtypes but also CAS, interval durations in an alternating task contributes methodological considerations of may be disproportionally longer than those in studying motor measures across the lifespan. a repetitive task, not only in the oral system METHOD as previously shown (Thoonen et al., 1996, 1999) but also in the hand motor system. Participants To summarize the motivation for this study, Data for this study were collected at the Uni- SSD is a common disorder affecting the versity of Washington Department of Speech 150 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011 and Hearing Sciences and with the approval the time of testing, his speech was charac- of the University of Washington’s Human terized by /r/ missing from his phonetic in- Subjects Institutional Review Board. Fami- ventory and a frontal lisp. His GFTA-2 SS lies were ascertained through a proband was 81 and KLPA-2 SS was 91. His younger child with a positive SSD history. Criteria for brother, code 1304, age 3;4, also had a participation specified proband child aged 5– speech delay and had received speech ther- 9 years, a positive family history of SSD as apy. His speech was characterized by /r/ miss- defined by at least two additional biological ing from his inventory, inconsistent cluster relatives with SSD, and the absence of neuro- reduction, velar fronting in initial position, logic impairments such as cerebral palsy, diag- [f]/θ substitution, and [d]/ð substitution. In nosed cognitive-communicative impairments addition, he produced many vowel errors, such as autism, impairments in oral function for example, [nef]/knife, [fada]/feather, and such as dysarthria, and impairments of oral [panselz]/pencils. His GFTA-2 standard score structures such as cleft palate. In each family, (SS) was 92 and KLPA-2 SS was 89. The boy’s both biological parents of the proband and ad- mother, code 1205, reported a childhood his- ditional relatives, such as siblings, grandpar- tory of difficulties learning to read and a ha- ents, great-grandparents, aunts, uncles, and bitual [f]/θ substitution even into adulthood; cousins, participated. no other speech difficulties were noted. The Study sessions took place in a quiet labora- boy’s father, code 1204, reported a child- tory room. In cases where participants were hood history of speech delay and speech ther- willing to participate in the full study protocol apy. His brother, code 1201, did not report a but were not able to travel to the University childhood history of speech difficulties but of Washington, data were collected in quiet his two sons, ages 9;3 (code 1301) and 6;4 study rooms located in clinic or library facili- (code 1302), had been diagnosed with severe ties near their homes. All sessions were video- SSD, whereas the boy’s mother, code 1202, and audio-recorded. did not have a history of speech difficulties. To date, five families (N = 57) participated. The proband’s father had a second brother, Thirty-nine participants underwent behav- code 1203, who reported that he was told as ioral testing and provided a family history and, a child that he sounded like he spoke with a with one exception, DNA. In all cases, the pos- foreign accent but he never received speech sibility that dialectal variation could explain therapy. The proband child’s paternal grand- observed speech production differences was mother, code 1102, did not report a history of considered and excluded. In what follows, the speech difficulties. The paternal grandfather families are described in terms of SSD history. was deceased and it was unknown whether Four-digit participant codes represent family he had a history of speech difficulties. code, generation number, and a two-digit indi- In Family 002, the proband child was a boy, vidual number. Codes are listed only for family age 8;4, code 2503, with a history of severe members who participated in the behavioral SSD. At the time of testing, his speech had testing and/or provided DNA. Where avail- greatly improved following intensive therapy able, standard scores (SS) from the Goldman- and was characterized only by distortions of Fristoe Test of Articulation–2 (GFTA-2; /r, l/. He had not undergone formal assess- Goldman, 2000) and the Khan-Lewis Phono- ment for CAS. His GFTA-2 SS was 86 and KLPA- logical Analysis–2 (KLPA-2; Khan & Lewis, 2 SS was 71. His two younger brothers, ages 2002) are reported, except in cases where no 6;5 and 5;5, codes 2504 and 2505, also had speech sound errors were observed. been diagnosed with severe SSD. At the time In Family 001, the proband child was a of this study, the 6-year-old brother’s speech boy, code 1303, age 5;11 (years;months) was characterized by absence of /r/ from the with a speech delay for which he had re- inventory and slight /l/ distortions. His GFTA- ceived speech therapy as a preschooler. At 2 SS was 101 and KLPA-2 SS was 85. When he Familial Speech Sound Disorder Subtype 151 was 2;11, a speech-language pathologist had Her GFTA-2 SS was 81 and KLPA-2 SS was observed difficulties with initiation of speech, 94. inconsistent vowel and consonant errors, lim- In Family 003, the proband was a boy, age ited phonetic inventory, and restricted ex- 5;11, code 3301, whose parents had concerns pressive vocabulary and concluded that these regarding his speech development, especially characteristics were consistent with CAS. The given a family history of SSD. He had not yet 5-year-old brother also was unable to produce undergone formal assessment. His speech was /r/ and, in addition, his speech was charac- characterized by absence of /r/ from the in- terized by palatal fronting, deaffrication, and ventory, [f]/θ substitution in medial and final [f]/θ substitution. His GFTA-2 SS was 79 and position, and [d]/ð substitution. His GFTA-2 KLPA-2 SS was 84. The same speech-language SS was 89 and KLPA-2 SS was 87. His father, pathologist who had assessed the 6-year-old code 3201, did not report a history of speech brother had assessed this brother at age 2;0 difficulties, but his mother, code 3202, had a and informally observed difficulty with oral childhood history of speech difficulties. One motor movements, concluding that the severe of the mother’s adult brothers, code 3204, speech delay was consistent with CAS as well. and one of her adult sisters, code 3203, re- Their 3-year-old sister, code 2506, had also ported childhood histories of speech difficul- been diagnosed with SSD and had received ties. No written records of their childhood therapy. Neither the mother, code 2405, nor speech therapy were available, but accord- the father, code 2504, reported childhood dif- ing to the parents, the brother, as a young ficulties with speech development. The ma- child, omitted final consonants. Her father, ternal grandmother, code 2303, reported se- code 3101, the proband’s maternal grandfa- vere speech difficulties in childhood but ther- ther, reported a childhood history of late de- apy had not been available to her at that time. veloping speech. Her mother, code 3102, the Her mother, code 2201, the proband’s mater- proband’s maternal grandmother, did not re- nal great-grandmother, did not report a child- port any childhood speech difficulties. The hood history of speech difficulties. Her hus- proband had several cousins on his mother’s band, code 2202, the proband’s deceased ma- side of the family, reportedly with speech de- ternal great-grandfather, had an unknown his- lays, who could not participate in the study tory of childhood speech difficulties, but his because they were not yet old enough or lived sister, code 2203, reported a history of diffi- out of the country. culties with speech. As a child, she had been In Family 004, the proband was a girl, age told that she was “tongue-tied.” Her speech at 9;0, code 4303, who had speech therapy as the time of the study was characterized by a a preschooler but whose speech had since frontal lisp. Her GFTA-2 SS was 59 and KLPA- normalized. Her younger sister, age 4;5, code 2 SS was 99. The maternal grandmother’s 4304, was unable to produce /r/ and she had a brother had a history of childhood speech dif- frontal lisp. Her GFTA-2 SS was 95 and KLPA-2 ficulties. His adult son, code 2401, reported a SS was 98. Her older brother, age 10;8, code childhood history of speech and language de- 4302, had no history of speech difficulties. lays. His son, age 7;0, code 2501, did not have The children’s father, code 4203, and mother, a diagnosed speech delay, and his speech was code 4202, did not report childhood speech characterized by inconsistent [f]/θ substitu- difficulties, but the paternal grandmother, tion. His GFTA-2 SS was 97 and KLPA-2 SS code 4102, reported receiving speech ther- was 101. His daughter, age 3;7, code 2502, apy as a child for what was explained to exhibited speech sound errors such as cluster her as tongue thrust, with tongue place- reductions, inconsistent and unusual speech ment consistent with a frontal lisp. Her adult sound substitutions, for instance, [v]/w, [v]/r, speech was still characterized by a frontal and [p]/kr, and a number of vowel errors, but lisp. She reported that one of her nieces she was able to produce /r/ and /l/ accurately. had also been treated for tongue thrusting, 152 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011 and several of this niece’s children wore or- The proband’s paternal aunt had two daugh- thodontic appliances to correct speech sound ters, ages 8;8 and 6;3, codes 5503 and 5504. distortions because of a forward tongue place- The 8-year-old girl had a history of severe SSD. ment. Detailed data from these relatives, how- According to a speech evaluation at age 3;10, ever, were not available. using the Clinical Assessment of Articulation Family 005 was unusual in that both par- and Phonology (Secord & Donohue, 2002), ents of the proband child, a girl, age 5;2, code her speech error patterns included consistent 5502, had histories of childhood speech diffi- cluster reduction; velar and palatal fronting; culties, and there were other relatives on both and frequent syllable reduction, vocalization, sides of the family with histories of speech gliding, and stopping. The six-year-old girl had difficulties. The proband child had previously undergone a speech evaluation at age 3;5. been diagnosed with a speech disorder con- Her speech sound substitutions ([t]/, [d]/, sistent with severe CAS and had been receiv- [f]/θ, [b]/v, [w]/r), omissions of final /r/, and ing speech therapy since the age of 3 years. cluster reductions (all /l/ and /r/ clusters) The CAS diagnosis was based in part on the were ruled to be at the low end of the nor- presence of 32 of 49 characteristics from the mal range for her age. Apraxia Profile (Hickman, 1997) and a sever- The proband had two male fourth cousins ity rating of severe impairment in the areas of who were brothers. One of them, age 14, code simple phonemic level, complex phonemic 5505, had a diagnosis of Asperger syndrome level, and spontaneous length in the Kaufman and the other, age 10, code 5506, had been Speech Praxis Test for Children (Kaufman, diagnosed with severe speech difficulties con- 1995), conducted when she was 4;11. At the sistent with disrupted praxis in addition to time of the study, the proband’s consonant hemiplegia, possibly related to a suspected inventory was substantially reduced. She sub- stroke suffered at birth. These two brothers stituted [d] for most obstruents and consonant participated in the full protocol but their oral clusters (e.g., [dØbɔl ]/shovel; [daυn]/clown; motor scores were excluded from the present [dipɔ]/zipper). Phonological process analy- study due to concerns about confounding ef- sis revealed the presence of the following fects from their primary diagnoses. The 10- processes: deletion of final consonants, stop- year-old boy did not participate in the rapid ping, cluster simplification, liquid simplifica- naming tasks. tion, palatal and velar fronting, cluster reduc- The proband’s mother, code 5402, re- tion, and initial voicing. In some cases, sev- ported childhood speech difficulties that had eral processes converged on a single target. resolved with therapy. Her mother, code For instance, [dØbɔl]/shovel shows the ef- 5304, the proband’s maternal grandmother fects of three phonological processes, palatal reported no speech difficulties and neither fronting, stopping, and initial voicing, all act- did her father, code 5303, the proband’s ma- ing on the target /ʃ/. Her GFTA-2 SS was 42 ternal grandfather, although his brother had and KLPA-2 SS was 41. Of note, this girl had received speech therapy as a child for a brief difficulty with saliva control. Her brother, age period. The proband’s mother’s half-sister, 7;9, also had a history of speech difficulties age 14, code 5401, had a history of mild but they were less severe and largely resolved speech delay and had received therapy for by the time of testing following therapy, with help with /r/. stimulable [f]/θ substitution being the only residual error. His GFTA-2 SS was 90 and Protocol KLPA-2 SS was 89. Articulation was assessed with the GFTA-2. The proband’s father, code 5403, and his The GFTA-2 quantifies consonant errors on sister, code 5404, both had childhood his- the single word level by phoneme and word tories of speech difficulties, and so did the position. Phonological processes were quan- proband’s paternal grandmother, code 5308. tified with the KLPA-2, using the word Familial Speech Sound Disorder Subtype 153 productions from the GFTA-2. Two ques- LabView (National Instruments Corporation, tionnaires about hand preference (adapted Austin, Texas). Ten trials were administered, from “The assessment and analysis of hand- five for the right hand and five for the left, edness: the Edinburgh inventory,” by R. starting with the right hand regardless of a C. Oldfield, 1971, Neuropsychologia, 9(1), participant’s handedness status, and switch- pp. 97–113.) and educational and develop- ing hands after each trial to minimize fatigue. mental backgrounds (generated specifically Each trial began with clicking on a button to for this study) were filled out. Parents pro- startthetrial,followedbyawaitperiodthat vided the requested information for their chil- varied randomly between 2 and 4 s and a vi- dren. The data collection further included sual start signal. In the alternating version of DNA and a variety of additional behavioral this activity, participants were instructed to tasks, not further detailed here, covering as- tap two different computer keys with the in- pects of speech production, language ability, dex and middle fingers of the same hand, go- reading, spelling, phonological memory, and ing back and forth between the two keys as verbal and nonverbal processing. fast as possible. The left and right arrow keys, For the present study, motor data were which are separated by the down arrow key, available from 11 children (9 ever affected) were programmed to receive the inputs from and 25 adults (12 ever affected), 36 par- this activity. For both types of the key tapping ticipants together. The term “ever affected” task, the instructions included a brief model refers to individuals who had a history of SSD, and the keyboard was positioned such that whether or not they showed evidence of SSD participants could reach the target keys com- atthetimeoftesting.Thistermisusedin fortably for each trial. In all study sessions, contrast with “never affected,” which refers the finger tapping task was administered be- to individuals without a history of SSD. fore the oral motor task. To investigate motor processing in the oral and hand systems, repetitive and alternating Data reduction movement tasks were administered. Follow- Data were reduced and analyzed by a team ing the methods in Fletcher (Fletcher, 1972), consisting of the first author and qualified participants were instructed to produce series undergraduate and graduate students. Sylla- of monosyllables (/pa, /ta/, /ka/), disyllables ble durations were measured using the freely (/pata/, /taka/), and trisyllables (/pataka/) as available acoustic software Praat (Boersma, fast and as accurately as they could. Each trial 2001), version 5.1.25. Short durations were was preceded by a brief model and a prac- interpreted as consistent with rapid motor tice run. At least 20 productions of the mono- speed, although it is acknowledged that short syllables, 15 of the disyllables, and 10 of the durations can also be achieved with relatively trisyllables were collected. In the case of in- small articulator excursions, not necessarily accurate production in the disyllable task, for representing rapid movement. As in previ- instance saying /papata .../ instead of /pata- ous studies of nonword repetition and max- pata .../, the inaccurately produced syllables imum syllable repetition rates (Peter & Stoel- were included in the calculation of average Gammon, 2008; Thoonen et al., 1996, 1999), syllable duration. the first token in a series was excluded from To measure maximum key tapping rates analysis to minimize nonlinear initiation ef- during a repetitive task, participants were fects, and the last token prior to an inhalation instructed to tap the spacebar of a laptop was excluded to minimize final lengthening computer using their index finger as many effects and also because, in open syllables, times as possible during a 10-second interval, the vowel endpoints can not always be es- following published protocols of this activ- tablished reliably because of variations in the ity (Gualtieri & Johnson, 2006). Tap intervals acoustic environment. Inhalations were ex- were recorded using a program designed with cluded. Most adults completed the target set 154 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011 of syllables within one breath, but many chil- lifespan are not available, the norms from dren were not able to do so. Whether or not the 13-year-olds were used for all individuals the tokens were produced in one breath had 13 years and older, even though it is possi- minimal, if any, effects on the syllable dura- ble that these norms slightly underestimate tions because inhalations and first and last syl- age-adjusted oral motor speeds in the older lables in a breath group were excluded from adults (Peter et al., 2010). To observe relative the analysis. deficits in sequencing places of articulation in Norms were only available for some of the the /pata/ task, the z scores from the disyllable measures of interest. For instance, norms for durations were subtracted from the z scores the mono- and disyllabic repetitions are avail- from the monosyllable durations. A large dis- able for ages 2;6 (years;months) through 6 crepancy was interpreted as faster monosyl- years (Robbins & Klee, 1987) and 6–13 years lable rates than disyllable rates and, hence, a (Fletcher, 1972), and they show decreasing relative deficit in motor sequencing. mean syllable duration rates as a function of Key tap intervals were obtained from the age. One goal of this study, hence, was to LabView output files and averaged for each develop measures that can be used in chil- participant. Similar to the syllable repetition dren as well as adults. In addition to raw syl- task, raw durations were converted to z scores lable durations, averaged per participant, du- by using norms for ages 5–7 years (Gray, Liv- ration ratios for monosyllables and multisyl- ingston, Marshall, & Haak, 2000) and 8–83 lables were calculated to factor out the raw years (Gualtieri & Johnson, 2006). The norms speeds. A ratio greater than 1 indicated that for the younger children were available for the multisyllables had shorter durations than the dominant and nondominant hand, and the monosyllables, implying shorter intervals those for older children and adults were avail- and, by inference, more rapid movement in able for the right and left hand; therefore, the the multisyllables than in the monosyllables, z scores averaged for both hands were used which is consistent with intact sequencing for normalizing the present data. The norms skills as typically seen in older children. For for the key tapping task also differed in that the purposes of the present study, the ratio the data for young children were obtained of the monosyllable /pa/ and the disyllable using a special key tapping device, whereas /pata/ was selected because young children those for the older children and adults were with SSD frequently have difficulty produc- obtained using a computer keyboard; there- ing /k/. Norms for ages 6–13 years (Fletcher, fore, norms for the younger children were 1972) show that sequencing ability in children imputed by comparing both scales at the increases as a function of age, as /papa, pata/ age overlap. Similar to the syllable repetition ratios range from 0.99 for age 6 years to 1.2 tasks, a ratio of repetitive and alternating tap for age 13 years. In an attempt to factor out intervals was calculated. Because norms for in- age, z scores were calculated for each partic- tertap intervals from alternating key tapping ipant using these norms (Fletcher, 1972). For were not available, a z-score difference could the participants younger than 6 years, norms not be calculated. Table 1 summarizes the mo- were imputed using norms for children aged tor measures with respect to the tasks, mea- 2;6 (years;months) through 6 years (Robbins sured ability, and standardization. & Klee, 1987). For children aged 6 years, these two scales differ by 53 ms, likely because of methodological differences. The norms for Reliability the older children were based on time-by- Approximately 15% of the motor data and count measures, whereas the norms for the 50% of the data from the standardized speech younger children were based on count-by- testing were checked for reliability. Standard time measures of 3-s intervals. Because norms scores from the articulation testing in cur- for older children and adults throughout the rently affected children differed, on average, Familial Speech Sound Disorder Subtype 155

Table 1. Measures of Repetitive and Alternating Motor Speeds

Standardization Standardization Variable Measured Ability Source Age

/pa/ Syllable Duration Raw speed of repetitive (ms) oral movement /pa/ z Score Age-adjusted speed of Robbins & Klee (1987), 2;6–6;11, repetitive oral Fletcher (1972) 6;0–13;11 movement /pata/ Syllable Duration Raw speed of alternating (ms) oral movement /pata/ z Score Age-adjusted speed of Robbins & Klee (1987), 2;6–6;11, alternating oral Fletcher (1972) 6;0–13;11 movement /pa, pata/ Duration Ratio Oral sequencing ability based on raw durations /pa, pata/ z Score Oral sequencing ability Robbins & Klee (1987), 2;6–6;11, Difference based on Z scores Fletcher (1972) 6;0–13;11 Rep. Key Tap Intervals Raw speed of repetitive (ms) finger movement Rep. Key Tap Interval Age-adjusted speed of Gray et al. (2000), 5;1–7;11, z Score repetitive finger Gualtieri & Johnson 8;1–83;11 movement (2006) Alt. Key Tap Intervals Raw speed of alternating (ms) finger Movement Rep./Alt. Key Tap Finger sequencing ability Interval Ratio based on raw durations

by 3.25 SS. Differences more than 1 SS were hand motor system were calculated. This was reconciled prior to phonological analysis by done separately in the three types of variables, reviewing the video records. The mean sylla- raw durations in ms, repetitive/alternating du- ble durations from the mono- and disyllabic ration ratio, and z score. Statistical signifi- production task differed by less than 1 ms. cance was Bonferroni adjusted for multiple Because the key tap intervals were captured testing, although it should be noted that the with a software program rather than retrieved variables are not independent of each other and analyzed manually, they were considered and the Bonferroni correction is more conser- accurate and reliable. The program’s data cap- vative than necessary. Seven tests were car- ture mechanism was tested prior to the study ried out, arriving at an adjusted alpha of .0071. sessions. To observe potential linear and/or nonlinear effects of age on the variables of interest, each of these measures was checked by visual in- Statistical analysis spection of scatter plots and with multiple To investigate the associations among the regression models using age in months and repetitive and alternating motor measures in a squared term of age in months. Significant the oral and hand motor systems (research correlations with the linear age term were in- question 1), correlations among and between terpreted as an indication that the measured the temporal measures in the oral and the variable increased or decreased as a function 156 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011 of age, whereas significant correlations with crease in adults as a function of age. For both the quadratic age term indicated an increas- variables, regression models confirmed these ing, then decreasing, trajectory or a decreas- trajectories with greater associations for the ing, then increasing, trajectory as a function linear age term than for the squared age term of age. (Table 2). Similarly, alternating interval dura- To investigate whether motor speeds distin- tions from the disyllabic /pata/ series and the guish between adults with and without a his- key tapping task involving two separate keys tory of SSD regardless of family-specific sub- were correlated significantly (r = .55, p = types (research question 2), one-tailed t tests .0008). As in the case of the repetitive move- between these two groups were conducted ment task, alternating durations from the oral for each of the variables of interest. To in- and the hand task followed a closely over- vestigate whether limits in motor control of laid quadratic trajectory across the lifespan speed and/or sequencing characterize adults (Figure 2). As in the variables from the repet- with a history of speech difficulties in simi- itive movement tasks, regression models con- lar ways as biologically related children (re- firmed these trajectories with greater associ- search question 3), variables of interest were ations for the linear age term than for the compared between ever affected and never squared age term (Table 2). affected participants in families with affected Regarding the durational ratios of intervals children who showed motor speed deficits. from repetitive and alternating movement, which are independent of raw durations, the RESULTS ratio of the durations from the monosyllabic and disyllabic task was not correlated signifi- Associations among and between oral cantly with the ratio of the durations from the and hand task performance scores repetitive and alternating key task (r = .30, Correlations were computed for three p = .0912). The repetitive/alternating du- types of variables—durations in ms units, ra- ration ratio from the key tapping task fol- tios of repetitive/alternating unit durations, lowed a quadratic trajectory with some out- and z scores. Of interest were correlations liers across the lifespan (Figure 3) with the within each of the two motor domains highest ratios and, hence, the greatest speed (oral, hand) as well as cross-domain correla- advantage in the alternating task, seen in tions (Table 2). Strongest within- and cross- participants between the ages of 18 and domain correlations were seen in the raw 57 years. The majority of participants 18 durations, followed by z scores and repeti- years and older exceeded a ratio of 1, in- tive/alternating durational ratios dicating a speed advantage for the alternat- Within the oral motor system, the syllable ing mode. Five of the six participants who durations from the monosyllabic (/pa/) and did not show this speed advantage were the disyllabic (/pata/) repetition tasks were members of Families 002 or 005 (Figure 3) significantly correlated (r = .71, p < .0001). and biologically related to the probands. It is Within the hand motor system, the intertap in- not clear what role the age of participants terval durations from the repetitive and alter- 2201, 2203, and 5308 played in these low nating task also were correlated significantly scores. Regression models showed that age (r = .85, p < .0001). in months was correlated with this ratio to a Across motor systems, durations from the greater extent than the squared term of age in monosyllabic /pa/ series and the repetitive months, consistent with a strong increase of key tapping task were correlated significantly sequencing ability with increasing age and a (r = .81, p < .0001). Durations in both mea- decrease in older adults. sures followed a closely overlaid quadratic tra- The analogous trajectory from the mono- jectory across the lifespan (Figure 1), showing syllable/disyllable ratios was not as clearly a steep decrease in children and a slight in- quadratic (Figure 4; Table 2), although it Familial Speech Sound Disorder Subtype 157 ∗ ∗ .0001 .0001 6.19 5.77 KT Rep., < < Alt. Ratio − ∗ .0001 .0001 KT Alt. 1.00 5.70 4.85 < < Dur. (ms) − − Score 0.13 0.07 z KT Rep. − − ∗ ∗ ∗ .85 .0001 .0001 .8970 .0001 .9460 1.00 8.52 KT Rep. < < < Dur. (ms) − ) From Multiple Regression Models Evaluating the p Score Diff. 1.00 0.04 z /pa, pata/ − Scores ( t .30.0912 1.00 Ratio 1.72 0.75 6.81 1.00 /pa, pata/ − z .53 1.48 Score 1.00 1.35 /pata/ − Oral Hand ∗ ∗ ∗ .0001 .5970.0001 .1490 .1880 .9660 .0950 .4570 .0008 .55 /pata/ 1.00 4.81 < < Dur. (ms) − z ) for Variables of Interest and p ∗ Score .96 .55 .0008 .33.0646 1.00 /pa/ 1.27 4.34 − ∗ ∗ ∗ ∗ (ms) .0001 .0001 .3450 .0001 .2120 .0001 .81 .71 6.10 1.00 7.43 < < < < /pa/ Dur. − − r r r r r Score Score: t z z r Score: t r r r z Score 1.00 z ) indicate statistical significance. ∗ (mo): (ms): Diff.: (ms): Ratio: p KT Rep. p KT Alt. Dur. p Squared Age p /pata/ p p /pa/ /pata/ Dur. (ms): /pa, pata/ Ratio: /pa, pata/ p KT Rep., Alt. p Pairwise Correlation Coefficients ( . Asterisks ( Domain Variable Oral /pa/ Dur. (ms): Hand KT Rep. Dur. Age Age (mo): Effects of Age in Months and Squared Age in Months Table 2. Note 158 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011

Figure 1. Durations (ms) from the monosyllable repetition task (/pɑ/) and the repetitive key tapping task as a function of age in months. showed a steep increase in the children and tio greater than 1 would be expected for adolescents. Given the norms (Fletcher, 1972) adults. Most adults showed this speed advan- showing that by age 11 years children pro- tage for alternating oral movements. Seven duce disyllables with shorter durations than of the 10 participants who did not show monosyllables, a monosyllable/disyllable ra- this advantage were biological relatives of the

Figure 2. Durations (ms) from the disyllable repetition task (/pɑta/) and the alternating key tapping task as a function of age in months. Familial Speech Sound Disorder Subtype 159

Figure 3. Ratios of repetitive/alternating key tapping (KT) durations as a function of age in months. Labels are participant codes for adults with ratios <1. probands in Families 002 or 005. It is not clear 4203) failed to reach a ratio greater than 1 in what role the ages of participants 2201, 2203, both tasks, consistent with motor sequencing and 5202 played in these low ratios. Four of deficits in both the oral and the hand task. the adult participants (2201, 2203, 2303, and Neither the linear nor the quadratic term of

Figure 4. Ratios of monosyllable and disyllable durations as a function of age in months. Labels are participant codes for adults with ratios <1. 160 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011

Figure 5. /pɑ,pɑta/ z score differences as a function of age in months. Labels are participant codes for differences >1. age was associated with age to a significant Regarding measures suitable for evaluat- extent (Table 2). ing familial or genetic effects on repetitive Regarding standardized scores within the and sequential motor speeds across the lifes- oral motor system, z scores from the monosyl- pan, these findings indicate that raw dura- labic and disyllabic task were correlated sig- tions are strongly influenced by age, decreas- nificantly (r = .55, p = .0008). Across motor ing in childhood and adolescence and increas- systems, the correlation of z scores from the ing again in older adulthood. Therefore, raw repetitive oral task (/pa/) and the repetitive durations should not be used to evaluate famil- key tapping task did not reach statistical sig- ial or genetic effects on motor speeds across nificance (r = .33, p = .0646). Age-adjusted the lifespan. As measures of sequential motor norms are not available for the alternating key ability, durational ratios based on repetitive tapping task. The regression models (Table 2) and alternating tasks should be interpreted by show that age was not associated with any of taking the lifetime trajectories into account, the z scores to a significant extent, consistent because it appears that the ratios increase with what would be expected when using in childhood, then decrease in later adult- age norms to convert raw durations z scores. hood as a function of age, even when raw Figure 5 shows the distribution of the z score durations are factored out. The z scores from difference from the mono- and disyllabic task the oral and hand tasks appear to control for as a function of age. Of 12 scores that ex- most of this age effect and, hence, are use- ceeded 1 SD, indicating a substantial deficit ful for evaluating familial or genetic effects in oral motor sequencing, six came from Fam- on repetitive and alternating motor speeds in ily 002, two from Family 004, and four from children and adults. Because z scores for alter- Family 005. Table 2 summarizes the pairwise nating key tapping are not available, the repet- correlations among the variables of interest itive/alternating interval ratio can be used to and the multiple regression statistics from the determine divergences from expectations at a two age terms. given age. Familial Speech Sound Disorder Subtype 161

Motor speeds in adults with and without z scores for the disyllables from those from a history of SSD monosyllables, however, showed a relative None of the variables of interest distin- deficit in the motor sequencing element in- guished significantly between adults with and herent in the disyllabic task. A corresponding without a history of SSD in general. For the deficit was not observed in the key tapping monosyllable durations in ms, t (p) =−1.29 task, where repetitive/alternating ratios were (.8904); the disyllable durations in ms, t (p) = between 0.95 and 1.13, although it is acknowl- −.34 (.3690); the ratio from the mono- and di- edged that age norms for this variable are syllable durations, t (p) =−1.01 (.1609); the not available. Neither parent had a history of z score difference from the mono- and disylla- speech difficulties, but the unaffected mother, ble durations, t (p) = .67 (.2532); the key tap- code 2405, showed a similar discrepancy be- ping repetitive tap intervals in ms, t (p) =−.74 tween z scores from the monosyllable and di- (.2326); the key tapping alternating tap inter- syllable tasks and so did the unaffected ma- vals in ms, t (p) =−.85 (.2036); the ratio from ternal great-grandmother, code 2201. In addi- the repetitive and alternating intervals, t (p) = tion, three affected adults, codes 2303, 2203, .00 (.4991); and the z score from the repeti- and 2401, showed very long disyllable dura- tive key tapping intervals, t (p) = .84 (.2049), tions and large z score discrepancies between respectively. Group differences for the indi- mono- and disyllable durations ranging from vidual z scores from the oral tasks were not 2.97 to 4.60. Two of these, codes 2303 and computed because in the adults, they were 2203, also had very low repetitive/alternating linear transformations of the raw durations. key tapping ratios, indicating that their alter- Analogous group differences in the children nating intervals were much longer than their were not computed because the majority had repetitive intervals, consistent with slower se- a history of SSD. quencing speeds, whereas the third affected adult, code 2401, showed intact finger se- quencing skills. As shown in Figures 3, 4, and Familial limits of motor speeds 5, several members of this family, all biologi- None of the affected children showed cally related to the proband, had duration ra- deficits in repetitive motor speeds, whether tios that deviated from expectation for age in in the key tapping task as documented with both types of motor task and large z score z scores of repetitive tapping durations, nor in differences in the mono- and disyllable task. the monosyllable repetitions as documented Similarly, the proband child in Family 005, with the analogous z scores. Slowed repetitive code 5502, who had severe speech deficits speeds in both motor tasks were observed in consistent with CAS, showed slowed sequenc- the adult participants 2203 and 3204. Slowed ing in the syllable repetition task. Her mono- repetitive key tapping but not slowed mono- syllabic z score was substantially above ex- syllable rates were seen in the adult partic- pectation for her age and the disyllabic z ipants 3202 and 4204. Slowed monosyllable score was less than half a standard deviation rates but not slowed key tapping rates were below the mean. The z score difference of seen in the adult participants 4302 and 5308. 1.38 documented a relative deficit in her se- Affected children in two families, 002 and quencing ability. Her key tapping also showed 005, showed limits in motor sequencing slowed alternating durations compared to speeds. In Family 002, the three brothers– repetitive durations, although an age-adjusted codes 2503, 2504, and 2505—all with histo- interpretation cannot be offered. Her brother, ries of severe SSD and two, with speech histo- code 5501, also affected with SSD but less ries consistent with CAS, had short monosyl- severelyso,showedthesamerelativestrength lable durations and their disyllable durations of repetitive speed and limited sequencing were within a quarter of a standard deviation speed in both motor systems, although, again, from the population mean. Subtracting the age-adjusted norms of the alternating key tap 162 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011 intervals are not available. In the paternal in the oral and the finger task varied signif- branch of the family, both the father, code icantly as a function of age. A steep decline 5403, and the grandmother, code 5408, pro- in duration across childhood (indicating im- duced the disyllables with substantially longer proving motor facility) and a slight increase in syllable durations than the monosyllables. The older adulthood were observed as a function grandmother also showed the same slowed of age. Ratios of repetitive/alternating dura- motor sequencing during the key tapping task tions, which were constructed as a measure of but the father did not. A third cousin of the sequencing ability independent of actual dura- father, code 5406, did not report a history tions, also showed an effect of age, with an in- of speech difficulties but showed relatively crease of sequencing ability across childhood slower speeds in the alternating version of and a slight decrease in older adulthood. This the key tapping and syllable repetition tasks, shows that relative motor sequencing ability than those in the repetitive version. As was varies across the lifespan even when raw du- the case for Family 002, deviations from ex- rations are factored out. The z scores from pectation for age in both types of motor task the mono- and disyllable task and the repeti- are evident in Figures 3, 4, and 5 for several tive key tapping task showed the least effect members of this family, all biologically related of age. The z score difference from the mono- to the proband. and disyllable task was interpreted as a robust In the maternal branch of Family 005, none measure of oral motor sequencing across the of the related family members, whether ever lifespan that can be used to evaluate familial SSD affected or not, showed repetitive or al- or genetic effects on motor sequencing abil- ternating motor speed deficits in the key tap- ity. An analogous measure is not available for ping task. Z scores from the repetitive task the key tapping task, but divergences of the were 0.11 or higher, and ratios of repeti- repetitive/alternating ratio given expectation tive/alternating durations were greater than for age can be used to estimate deficits in fin- 1.16, consistent with intact finger sequenc- ger motor sequencing across the lifespan. It ing ability. Regarding disyllable durations, should be kept in mind, however, that this the proband’s mother, code 5402, and great- sample contained individuals with deficits in grandmother, code 5202, both had z scores repetitive and alternating motor speeds. To above expectations. The mother produced fully investigate the effect of age on motor the disyllables at much greater speeds than speeds, norming data should be generated in expected, whereas the great-grandmother a typical sample of adults. showed substantially slower speeds, resulting Regarding the role of age in repetitive fin- in a large z score difference. ger motor speeds, previous studies (Bartzokis et al., 2008; Gualtieri & Johnson, 2006; Pri- DISCUSSION gatano et al., 2008) have shown that raw key tapping speeds increase in childhood and The purpose of this study was to investigate decrease with advancing age in adults. The familial subtypes of SSD in five families with results from the present study are consis- multigenerational histories of SSD. The focus tent with these findings and extend them to was on the role of motor speeds in the oral the motor speech system, although our sam- and hand motor systems. ple contained individuals with SSD histories, which may have influenced performance on Suitable measures of repetitive and this task. Normative data for repetitive mo- alternating motor speeds across the tor speech speeds, available only for children lifespan age 2;6 through 13;11, reveal an increase in An evaluation of the effects of age on the speed as a function of age. Our results are variables of interest showed that raw dura- consistent with this trend and also show a de- tions of repetitive and alternating movement crease in speed with advancing age in adults, Familial Speech Sound Disorder Subtype 163 similar to the findings in the finger motor cidental or whether they reflect systemic rate literature, although it is acknowledged again limits inherent in the central and/or periph- that in some of the adult participants, it is not eral nervous system. In the youngest children, certain whether the repetitive motor rates are however, key tapping intervals were substan- influenced by age, a familial or genetic effect, tially longer than their syllable durations, es- or both. pecially in the alternating mode. This discrep- Regarding the role of age in alternating mo- ancy may, in part, result from the fact that tor speeds, norms for disyllables (Fletcher, young children have smaller hands than older 1972) show that by 11 years, children develop children and adults, placing them at a physical a speed advantage for alternating places of disadvantage with respect to the dimensions articulation, presumably due to progressively of the key depression depth and distance be- greater oral motor sequencing ability. Analo- tween two different keys. It is also possible gous norms for alternating key strikes are not that finger motor sequencing skills develop available. This study shows not only that alter- later than oral motor sequencing skills in chil- nating movement speeds in a finger task and dren. Environmental factors may play a role as in a syllable repetition task follow a quadratic well. Younger children have less experience trajectory (i.e., increase of sequencing ability pressing computer keys than older children during childhood and a decline with advanc- and adults, but they have accumulated con- ing age in adulthood) but also that a similar siderable motor practice in the oral system in quadratic trajectory is likely for the ratio of the process of speech acquisition, including repetitive and alternating intervals in a hand during the babble phase. task, independent of raw durations. Our sam- ple, however, contained several children and Correlations among measures of oral adults with histories of apraxic-like speech dif- and hand motor speeds ficulties, which may have confounded the ef- We found that interval durations in the fect of age on the ratios. A decrease of alternat- oral and the hand motor system were pos- ing movement speeds in adults was observed itively and significantly correlated with each in our study of global processing speeds in other, even showing highly similar interval du- a large sample of families with dyslexia (Pe- rations. This was the case for both the repet- ter et al., 2010), where the decreases were itive and alternating movements. Ratios of noted not only for timed measures of fin- repetitive/alternating durations and z scores ger motor and oral motor sequencing but showed a trend toward these cross-modal as- also rapid naming of stimuli with and with- sociations but not with statistical significance. out category switches and alphabet writing. Inthecaseofthez scores, this suggests that Age adjustments for performance on mea- participant age may, in part, account for the sures of phonological memory and articula- strengths of the cross-modal associations seen tion in adults with monotonically increasing in the raw durational measures. A larger par- raw scores as a function of age have recently ticipant sample would be necessary to probe been presented (Stein et al., 2010). Norms for cross-modal associations with controls for age older children and adults throughout the lifes- and motor deficits as well as for familial rela- pan (in populations that do not have familial tionships. histories of speech or language impairments) Cross-modal associations in hand and finger should be developed for measures whose raw tasks are consistent with our previous findings scores follow a linear or quadratic trajectory in a small sample of children with and with- across the lifespan. out SSD, where timing accuracy was corre- It was noted that the raw interval durations lated in a hand and a nonword imitation task in the finger and oral task were highly sim- (Peter & Stoel-Gammon, 2008), and a large ilar to each other in most participants. It is sample of families with evidence of familial unknown whether these similarities are coin- dyslexia, where z scores from a timed finger 164 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011 sequencing and a timed motor speech task slowed alternating movement rates showed loaded on a general factor characterized by evidence of familial aggregation. In Families global processing speed (Peter et al., 2010). 002 and 005, speech-language pathologists Given the associations between the oral and had previously concluded that the speech hand motor system during repetitive and se- characteristics of affected children were con- quential tasks, it is reasonable to ask whether sistent with CAS, a proposed SSD subtype children with speech disorders characterized with deficits in sequential motor processes. by motor sequencing difficulties might also In both families, the general mode of inheri- show these difficulties in other motor systems tance, based on affectation status, is consis- and whether biologically related adults with tent with autosomal dominance with cases a history of such speech disorders also show of nonpenetrance, such as the proband’s deficits in oral and/or hand sequencing tasks. mother in Family 002. Slowed oral sequencing rates characterized the affected relatives, and Motor speeds in adults with a history slowed finger sequencing was seen in some of SSD but not all of the affected relatives, indicating Overall, the adults with a history of SSD that oral sequencing speeds are more closely did not show evidence of repetitive or alter- associated with the presence of the SSD sub- nating motor deficits, compared to the never type in these families than finger sequencing affected adults. Therefore, it is unlikely that speeds. Slowed oral and/or finger sequenc- motor deficits are among the long-term seque- ing seen in some unaffected relatives, 2405, lae of all familial forms of SSD in this sample. 5202, and 5406, may indicate that sequencing Uniform motor sequelae across families would deficits are a predisposing but not determin- be consistent with the common disease/rare ing factor for SSD. variant model, where a common disease is in- Of all participants, the proband in Family fluenced by common and shared genetic vari- 005 had the most severe speech difficulty as ants. The absence of motor sequelae shared documented with the lowest GFTA-2 score. across all families allows for the possibility of Because both of her parents had positive fa- finding these sequelae within some individual milial SSD histories, it is possible that she families. inherited a risk allele from each of her par- ents, together disrupting her speech develop- Family-specific SSD subtypes ment more severely than her brother, who The families were selected for this study be- may have only inherited one of the risk al- cause they showed evidence of familial SSD, leles. The paternal branch of the family pro- but we found that phenotypic expression vided more evidence for a motor-based disor- varied between the families. In Family 004, der than the maternal branch did. The exact difficulty with /r/ and fronted tongue place- nature of the speech difficulties in the mater- ment for /s, z/ characterized the speech of nal branch could not be determined with the a young family member, consistent with for- available data. ward tongue placement as also reported in Taken together, the results from this study her grandmother and several other biological are consistent with the common disease/rare relatives. This may represent a subtype of SSD variant model in that we describe family- with oral structural components, although it is specific SSD subtypes based on behavioral acknowledged that direct data were not avail- data. Subtypes include a familial SSD form able from some of the members of this fam- with a structural component, a familial form ily. In Families 1 and 3, other traits, not yet of motor-based SSD, and potentially at least probed, may characterize the inherited SSD one familial form that does not fit either of subtype. the other two profiles. In one of these forms, Slowed repetitive movement rates did not motor sequencing difficulties in the oral mo- appear to aggregate in any of the families but tor system, as captured in a z score difference Familial Speech Sound Disorder Subtype 165 and ratios of monosyllable and disyllable rep- from genotypic investigations. Future studies etition durations, were indicative of this SSD also should be conducted to attempt to repli- subtype. Corresponding sequencing difficul- cate the results from this study in additional ties in the key tapping task, as estimated family samples with evidence of familial SSD. from ratios of repetitive/alternating durations, They should test our results in additional syl- were seen less frequently in ever affected in- lable types and with additional measures of dividuals. These results are consistent with sequencing including accuracy of attempted Thoonen et al.’s (1996, 1999) findings that the tokens. Sequencing ability in oral tasks and alternating but not repetitive oral movements perhaps also in the finger tasks may be an ap- distinguish between children with praxis- propriate component phenotype for molecu- related speech deficits and typical peers. Fur- lar genetic studies of SSD. thermore, they expand on the observation Future studies also should address clinical by Thoonen et al. (1997) that some children questions related to motor sequencing ability. with praxis-related speech deficits have a fam- If deficits in motor sequencing are found to ily history of speech and language difficulties be a core characteristic of certain SSD sub- by showing that motor sequencing difficulties types with a genetic etiology, then future re- characterize not only the affected children but search should investigate whether therapy di- also many of the ever-affected relatives. Given rected at speech sequencing can ameliorate that difficulty with motor sequencing was ob- the severity in affected children. It may further served in the proband’s unaffected mother in be of interest to investigate whether therapy Family 002 and a distantly related unaffected targeting sequencing in other motor systems relative, code 5406, in Family 005, it is pos- can influence a central sequencing deficit and sible that difficulty with motor sequencing is thereby support the effectiveness of speech a risk factor contributing to the full expres- therapy. Utmost care should be taken to fol- sion of SSD in combination with other factors low the mandates of evidence-based practice including environmental influences. in designing therapy for children with familial motor-based SSD subtypes. Because the cause- effect relationship between general motor se- Future Studies and Clinical Implications quencing deficits and motor-based SSD sub- Regarding the participants from the present types is not clear, interventions targeting mo- study, in future studies, we plan to address tor sequencing with the goal of ameliorating the role of additional phenotypes, including speech deficits should not be implemented phonological memory, verbal and nonverbal unless evidence for the efficacy of such inter- processing, and reading, and also the results ventions becomes available.

REFERENCES

Bartzokis, G., Lu, P. H., Tingus, K., Mendez, M. F., Richard, family-based study point to novel autism susceptibility A., Peters, D. G., et al. (2010). Lifespan trajectory of genes. PLoS Genetica, 5(6), e1000536. myelin integrity and maximum motor speed. Neuro- Campbell, T. F., Dollaghan, C. A., Rockette, H. E., biology of Aging, 31(9):1554–1562. Paradise, J. L., Feldman, H. M., Shriberg, L. D., et al. Bishop, D. V. (2002). Motor immaturity and specific (2003). Risk factors for speech delay of unknown ori- speech and language impairment: Evidence for a com- gin in 3-year-old children. Child Development, 74(2), mon genetic basis. American Journal of Medical Ge- 346–357. netics, 114(1), 56–63. Dror, A. A., & Avraham, K. B. (2009). Hearing loss: Mech- Boersma, P. (2001). Praat, a system for doing phonetics anisms revealed by genetics and cell biology. Annual by computer. Glot International, 5(9/10), 341–345. Review of Genetics, 43, 411–437. Bucan, M., Abrahams, B. S., Wang, K., Glessner, J. T., Felsenfeld, S., & Plomin, R. (1997). Epidemiological and Herman, E. I., Sonnenblick, L. I., et al. (2009). Genome- offspring analyses of developmental speech disor- wide analyses of exonic copy number variants in a ders using data from the Colorado Adoption Project. 166 TOPICS IN LANGUAGE DISORDERS/APRIL–JUNE 2011

Journal of Speech, Language, and Hearing Research, pleiotropy influencing speech and language: Analysis 40(4), 778–791. of the DYX8 region. Human Heredity, 63(1), 47–58. Fisher, S. E., Vargha-Khadem, F., Watkins, K. E., Monaco, Pennington, B. F., & Bishop, D. V. (2009). Relations A. P., & Pembrey, M. E. (1998). Localisation of a gene among speech, language, and reading disorders. An- implicated in a severe speech and language disorder. nual Review of Psychology, 60, 283–306. Nature Genetics, 18(2), 168–170. Peter, B. (2010). New frontiers in understanding speech Fletcher, S. G. (1972). Time-by-count measurement of sound disorder: Unraveling the mysteries of genetic diadochokinetic syllable rate. Journal of Speech and causes. In Harrison, A. E. (Ed.), Speech disorders: Hearing Research, 15(4), 763–770. Causes, treatment and social effect. New York: Nova. Gray, R. M., Livingston, R. B., Marshall, R. M., & Haak, R. Peter, B., Matsushita, M., & Raskind, W. H. (2010). Global A. (2000). Reference group data for the Reitan-Indiana processing speed in children with low reading abil- Neuropsychological Test Battery for young children. ity and in children and adults with typical reading Perceptual and Motor Skills, 91(2), 675–682. ability: Exploratory factor analytic models. Journal of Goldman, R. F., M. (2000). Goldman-Fristoe Test of Artic- Speech and Hearing Disorders, DOI: 10.1044/1092- ulation 2. Circle Pines: American Guidance Service. 4388(2010/10-0135). Gualtieri, C. T., & Johnson, L. G. (2006). Reliability and Peter, B., & Stoel-Gammon, C. (2008). Central timing validity of a computerized neurocognitive test battery, deficits in subtypes of primary speech disorders. Clin- CNS Vital Signs. Archieve Clinical Neuropsychology, ical Linguist Phoniatrica, 22(3), 171–198. 21(7), 623–643. Peterson, R. L., McGrath, L. M., Smith, S. D., & Hickman, L. A. (1997). Phonological disability in chil- Pennington, B. F. (2007). Neuropsychology and ge- dren: Studies in language disability and remedia- netics of speech, language, and literacy disorders. Pe- tion (2nd ed). New York: Elsevier. diatric Clinics of North America, 54(3), 543–561, vii. Kaufman, N. R. (1995). Kaufman Speech Praxis Test for Potter, N. L., Lazarus, J. A., Johnson, J. M., Steiner, R. children. Detroit, MI: Wayne State University Press. D., & Shriberg, L. D. (2008). Correlates of language Khan, L., & Lewis, N. (2002). Khan-Lewis phonological impairment in children with galactosaemia. Journal analysis (2nd ed). Circle Pines, MN: American Guid- of Inherited Metabolic Disease, 31(4), 524–532. ance Service. Prigatano, G. P., Gray, J. A., & Legacy, J. (2008). Predictors Lai, C. S., Fisher, S. E., Hurst, J. A., Vargha-Khadem, F., of quantitative and qualitative Halstead finger-tapping & Monaco, A. P. (2001). A forkhead-domain gene is scores in low socioeconomic status school-age chil- mutated in a severe speech and language disorder. dren. Child Neuropsychology, 14(3), 263–276. Nature, 413(6855), 519–523. Robbins, J., & Klee, T. (1987). Clinical assessment of Lewis, B. A. (1992). Pedigree analysis of children with oropharyngeal motor development in young children. phonology disorders. Journal of Learning Disabili- Journal of Speech and Hearing Disorders, 52(3), ties, 25(9), 586–597. 271–277. Lewis, B. A., Freebairn, L. A., Hansen, A. J., Miscimarra, Secord, W., & Donohue, J. (2002). Clinical assessment L., Iyengar, S. K., & Taylor, H. G. (2007). Speech and of articulation and phonology. Greenville, MS: Super language skills of parents of children with speech Duper Publications. sound disorders. American of Journal Speech Lan- Shriberg, L. D., Potter, N. L., & Strand, E. A. (2010). Preva- guage Pathology, 16(2), 108–118. lence and Phenotype of Childhood Apraxia of Speech Lewis, B. A., Shriberg, L. D., Freebairn, L. A., Hansen, In Youth with Galactosemia. Journal of Speech, Lan- A. J., Stein, C. M., Taylor, H. G., et al. (2006). The guage, and Hearing Research. DOI: 10.1044/1092- genetic bases of speech sound disorders: Evidence 4388(2010/10-0068). from spoken and written language. Journal of Speech, Shriberg, L. D., Tomblin, J. B., & McSweeny, J. L. (1999). Language, and Hearing Research, 49(6), 1294–1312. Prevalence of speech delay in 6-year-old children and MacDermot, K. D., Bonora, E., Sykes, N., Coupe, A. M., comorbidity with language impairment. Journal of Lai, C. S., Vernes, S. C., et al. (2005). Identification of Speech, Language, and Hearing Research, 42(6), FOXP2 truncation as a novel cause of developmental 1461–1481. speech and language deficits. American Journal of Smith, S. D., Pennington, B. F., Boada, R., & Shriberg, Human Genetics, 76(6), 1074–1080. L. D. (2005). Linkage of speech sound disorder to McClellan, J., & King, M. C. (2010). Genetic heterogeneity reading disability loci. Journal of Child Psychology in human disease. Cell, 141(2), 210–217. and Psychiatry and Allied Disciplines, 46(10), McKinnon, D. H., McLeod, S., & Reilly, S. (2007). The 1057–1066. prevalence of stuttering, voice, and speech-sound Stein, C. M., Lu, Q., Elston, R. C., Freebairn, L. A., Hansen, disorders in primary school students in Australia. A. J., Shriberg, L. D., et al. (2010). Heritability Esti- Language Speech Hearing Services School, 38(1), mation for Speech-Sound Traits with Developmental 5–15. Trajectories. Behavior Genetics, 41(2), 184–191. Miscimarra, L., Stein, C., Millard, C., Kluge, A., Cartier, Stein, C. M., Millard, C., Kluge, A., Miscimarra, L. E., K., Freebairn, L., et al. (2007). Further evidence of Cartier, K. C., Freebairn, L. A., et al. (2006). Speech Familial Speech Sound Disorder Subtype 167

sound disorder influenced by a locus in 15q14 region. Thoonen, G., Maassen, B., Wit, J., Gabreels, F., & Behavior Genetics, 36(6), 858–868. Schreuder, R. (1996). The integrated use of max- Stein, C. M., Schick, J. H., Taylor, H. G., Shriberg, L. D., imum performance tasks in differential diagnostic Millard, C., Kundtz-Kluge, A., et al. (2004). Pleiotropic evaluations among children with motor speech dis- effects of a chromosome 3 locus on speech-sound orders. Clinical Linguistics & Phonetics, 10(4), disorder and reading. American Journal of Human 311–336. Genetics, 74(2), 283–297. Vargha-Khadem, F., Watkins, K., Alcock, K., Fletcher, P., Terband, H., Maassen, B., van Lieshout, P., & Nijland, L. & Passingham, R. (1995). Praxic and nonverbal cogni- (2011). Stability and composition of functional syner- tive deficits in a large family with a genetically trans- gies for speech movements in children with develop- mitted speech and language disorder. Proceedings of mental speech disorders. Journal of Communication The National Academy of Sciences of The United Disorders, 44(1), 59–74. States of America, 92(3), 930–933. Thoonen, G., Maassen, B., Gabreels, F., & Schreuder, R. Watkins, K. E., Dronkers, N. F., & Vargha-Khadem, (1999). Validity of maximum performance tasks to F. (2002). Behavioural analysis of an inher- diagnose motor speech disorders in children. Clinical ited speech and language disorder: Compari- Linguistics & Phonetics, 13(1), 1–23. son with acquired aphasia. Brain, 125(Pt 3), Thoonen, G., Maassen, B., Gabreels, F., Schreuder, R., 452–464. & de Swart, B. (1997). Towards a standardised assess- Wijsman, E. M., & Amos, C. I. (1997). Genetic analysis ment procedure for developmental apraxia of speech. of simulated oligogenic traits in nuclear and extended European Journal of Disorders of Communication, pedigrees: summary of GAW10 contributions. Genet- 32(1), 37–60. ica Epidemiology, 14(6), 719–735.