Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jun;21(6):434-448.
doi: 10.1016/j.tics.2017.03.008. Epub 2017 Apr 8.

Neural Noise Hypothesis of Developmental Dyslexia

Free PMC article

Neural Noise Hypothesis of Developmental Dyslexia

Roeland Hancock et al. Trends Cogn Sci. .
Free PMC article

Erratum in


Developmental dyslexia (decoding-based reading disorder; RD) is a complex trait with multifactorial origins at the genetic, neural, and cognitive levels. There is evidence that low-level sensory-processing deficits precede and underlie phonological problems, which are one of the best-documented aspects of RD. RD is also associated with impairments in integrating visual symbols with their corresponding speech sounds. Although causal relationships between sensory processing, print-speech integration, and fluent reading, and their neural bases are debated, these processes all require precise timing mechanisms across distributed brain networks. Neural excitability and neural noise are fundamental to these timing mechanisms. Here, we propose that neural noise stemming from increased neural excitability in cortical networks implicated in reading is one key distal contributor to RD.

Keywords: excitability; glutamate; neural oscillation; neurogenetics; reading.


Figure 1 (Key Figure)
Figure 1 (Key Figure). The neural noise hypothesis
Schematic of the neural noise hypothesis, illustrated through two genetic pathways known to affect neural noise within the domain of auditory processing and their downstream consequences on reading. Some genetic risk factors, such as DCDC2 mutations, increase neural noise through a direct effect on glutamatergic signaling and hyperexcitability. Other genetic risk factors, such as KIAA0319 mutations, may disrupt the formation of local excitatory-inhibitory circuits thereby increasing neural noise. There are likely other risk genes that act through similar pathways. Excess neural noise disrupts neural synchronization across multiple scales, leading to deficits in low-level temporal auditory processing, and the oscillatory neural processes that sample and encode sensory information. Loss of synchronization and precise neural spike timing also impairs multisensory integration. Ultimately, the downstream effects of neural noise lead to impairments in phonological awareness and multisensory integration, two key components of reading development. Although we focus our discussion on the consequences of neural noise in the auditory domain, similar consequences are predicted in the visual domain, ultimately impacting orthographic processing and reading. This speculative pathway is shown on the right. Dashed boxes and arrows reflect more speculative links in need of further study. Processes in light text are not discussed in detail in the main text. Numbers by arrows refer to supporting references in the main text.
Figure 2
Figure 2. Possible mechanisms of neural noise in RD
Schematic illustration of excitation-inhibition imbalance within a local network of excitatory pyramidal cells (grey triangles) and inhibitory GABAergic interneurons (blue circles) with feedback inhibition (red) and corresponding trial-by-trial spiking activity above. (a) Typical excitation-inhibition balance. In a balanced excitation-inhibition regime, stimuli evoke excitatory (green) synaptic conductances followed by comparable inhibitory conductances (red) within a few milliseconds. Together these conductances regulate spike activity to produce neural activity that is precisely timed with respect to the input. (b) Consequence of increased glutamatergic activity. Animal models suggest RD risk genes modulate excitatory postsynaptic activity, which may result in greater and more temporally extended increases in excitatory conductances and reduced spike timing precision, in the absence of compensatory increases in feedback inhibition to restore excitation-inhibition balance. (c) Consequence of abnormal neural migration. RD risk genes have also been linked to disrupted neural migration and dendritic spine formation on pyramidal cells. One possible consequence of this is disrupted feedback connectivity (dashed arrows) between GABAergic interneurons and pyramidal cells, reducing feedback inhibition available to dampen neural activity and again producing temporally extended responses. Figure based on [110]. See also Box 1.

Similar articles

See all similar articles

Cited by 15 articles

See all "Cited by" articles