Does Inattention and Hyperactivity Moderate the Relation Between Speed of Processing and Language Skills?

Abstract

The causal role of speed of processing (SOP) in developmental language disorder (DLD) is unclear given that SOP has been implicated in other neurodevelopmental disorders such as attention-deficit/hyperactivity disorder. This study investigated associations between SOP, language, and inattention/hyperactivity in a U.K. epidemiological cohort (N = 528). Monolingual children from a range of socioeconomic backgrounds were assessed longitudinally; at ages 5-6 (2012/2013) and 7-8 years (2014/2015). Persistent weaknesses in SOP characterized children with DLD but did not predict language longitudinally. Ratings of inattention/hyperactivity moderated the association between SOP and language, indicating that SOP deficits are particularly detrimental for language when coupled with poor attention/hyperactivity. SOP may be a shared risk factor for DLD and inattention/hyperactivity or a general marker of neurodevelopmental disorder.

Figure 1

Standard z‐score differences between children with and without language disorder on measures of speed of processing in Year 1 (a) and Year 3 (b). Error bars are 95% confidence intervals. Bars that cross the zero midline indicate no significant group difference. Boxes to the left of the zero indicate poorer performance in the developmental language disorder group. RT = reaction time; RAN = rapid automatized naming.

Figure 2

Longitudinal autoregressive path model with cross‐lagged effect showing the relation between language (LANG) and speed of processing (SOP). Standardized path estimates and correlation coefficients are depicted by single and double‐headed arrows, respectively. Path weights and confidence intervals for the whole sample are shown outside the brackets (N = 499), those for the sample without children with known diagnoses are shown inside the brackets (N = 441). Paths significant at the 0.05 level are represented by solid lines; nonsignificant paths by dashed lines. EOWPVT = Expressive One Word Picture Vocabulary Test; ROWPVT = Receptive One Word Picture Vocabulary Test; SASIT = School‐Age Sentence Imitation Test; TROG = Test for Reception of Grammar; RT = reaction time; RAN = rapid automatized naming.

Figure 3

Measurement model for the continuously distributed dimensions of speed of processing, language, and inattention/hyperactivity (as measured by the Strengths and Weaknesses of Attention‐Deficit/Hyperactivity Disorder symptoms and Normal behavior scale [SWAN]) in Year 1 (N = 343). Standardized path estimates and correlation coefficients (with 95% CIs) are depicted by single and double‐headed arrows, respectively (see Table S5 for the SWAN latent variable path estimates: all loadings range between .72–.94). EOWPVT = Expressive One Word Picture Vocabulary Test; ROWPVT = Receptive One Word Picture Vocabulary Test; SASIT = School‐Age Sentence Imitation Test; TROG = Test for Reception of Grammar; RT = reaction time; RAN = rapid automatized naming; CFI = comparative fit index; TLI = Tucker–Lewis index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual.

Figure 4

Simple slopes and 95% CIs for language and speed of processing at three levels of the inattention/hyperactivity latent variable: high Strengths and Weaknesses of Attention‐Deficit/Hyperactivity Disorder Symptoms and Normal Behavior scale (SWAN) scores (+1 SD), that is, fewer symptoms of inattention/hyperactivity (dark grey dash); medium SWAN scores, that is sample mean (light gray); and low SWAN scores (−1 SD), that is, more symptoms of inattention/hyperactivity (black).

Figure 5

Longitudinal autoregressive path model with cross‐lagged effect showing the relation between speed of processing, language, and inattention/hyperactivity. Key standardized path estimates and correlation coefficients are depicted by single and double‐headed arrows, respectively (see Table S5 for standardized path estimates for the inattention/hyperactivity latent variable that range between .71–.94 and .65–.94 for Years 1 and 3, respectively). Path weights and confidence intervals for the whole sample are shown outside the brackets (N = 362), those for the sample without children with known diagnoses are shown inside the brackets (N = 318). Paths significant at the .05 level are represented by solid lines; nonsignificant paths by dashed lines. EOWPVT = Expressive One Word Picture Vocabulary Test; ROWPVT = Receptive One Word Picture Vocabulary Test; SASIT = School‐Age Sentence Imitation Test; TROG = Test for Reception of Grammar; RT = reaction time; RAN = rapid automatized naming.