Auditory brainstem response deficits in learning disorders and developmental language disorder: a systematic review and meta-analysis

Abstract

Although learning disorders (LD) and developmental language disorder (DLD) can be linked to overlapping psychological and behavioral deficits, such as phonological, morphological, orthographic, semantic, and syntactic deficits, as well as academic (e.g., reading) difficulties, they are currently separate diagnoses in the DSM-5 with explicit phenotypic differences. At a neural level, it is yet to be determined to what extent they have overlapping or distinct signatures. The identification of such neural markers/endophenotypes could be important for the development of physiological diagnostic tools, as well as an understanding of disorders across different dimensions, as recommended by the Research Domain Criteria Initiative (RDoC). The current systematic review and meta-analysis examined whether the two disorders can be differentiated based on the auditory brainstem response (ABR). Even though both diagnoses require hearing problems to be ruled out, a number of articles have demonstrated associations of these disorders with the auditory brainstem response. We demonstrated that both LD and DLD are associated with longer latencies in ABR Waves III, V, and A, as well as reduced amplitude in Waves V and A. However, multilevel subgroup analyses revealed that LD and DLD do not significantly differ for any of these ABR waves. Results suggest that less efficient early auditory processing is a shared mechanism underlying both LD and DLD.

Figure 1

Article selection process with numbers of exclusions.

Figure 2

Wave I amplitude forest plot. Positive values for Hedges’ g indicate larger amplitude for the disorder group relative to controls. Rows with decimal points after the article year indicate multiple analyses from the same paper, and correspond to the row order of the analysis in the systematic review Supplement A.

Figure 3

Wave I Latency forest plot. Positive values for Hedges’ g indicate longer latency for the disorder group relative to controls. Rows with decimal points after the article year indicate multiple analyses from the same paper, and correspond to the row order of the analysis in the systematic review table in Supplement A.

Figure 4

Wave III latency forest plot. Positive values for Hedges’ g indicate longer latency for the disorder group relative to controls. Rows with decimal points after the article year indicate multiple analyses from the same paper, and correspond to the row order of the analyses in the systematic review table in Supplement B.

Figure 5

Wave V Overall Amplitude effect forest plot. Negative values of Hedges’ g indicate lower amplitude in the disorder group relative to the control group. Rows with decimal points after the article year indicate multiple analyses from the same paper, and correspond to the row order of the analysis in the systematic review table in Supplement C.

Figure 6

Wave V Overall latency effect forest plot. Positive values of Hedges’ g indicate longer latency for the disorder group relative to controls. Rows with decimal points after the article year indicate multiple analyses from the same paper, and correspond to the row order of the analysis in the systematic review table in Supplement C.

Figure 7

Wave A/Vn Amplitude forest plot. Negative values of Hedges’ g indicate lower amplitude in the disorder group relative to the control group. Although subgroup and analyses of stimulus type were not possible due to the limited number of effect sizes in this analysis, the two statistically significant effects were from articles with two different stimulus types (Malayeri (2014).2 – speech, Gabr (2005) – clicks), so differences between papers were not likely due a difference between speech and non-speech stimuli. Rows with decimal points after the article year indicate multiple analyses from the same paper, and correspond to the row order of the analysis in the systematic review table in Supplement D.

Figure 8

Wave A/Vn latency forest plot. Positive values of Hedges’ g indicate longer latency for the disorder group relative to controls. Rows with decimal points after the article year indicate multiple analyses from the same paper, and correspond to the row order of the analysis in the systematic review table in Supplement D.