Skip to content
About Contact
Engevity NewsScience & health

What dyslexia research has settled, and what it hasn't

A 2026 review calls explicit phonics instruction a settled question in dyslexia research, while causes and brain-scan markers stay contested. A 2014 meta-analysis of 22 trials and a 2026 replication study mark the line between the two.

What dyslexia research has settled, and what it hasn't
What dyslexia research has settled, and what it hasn't

Dyslexia research has settled two big questions, what dyslexia looks like and which teaching helps. A 2026 review in the journal Dyslexia by Christy Austin and colleagues calls the case for explicit, systematic phonics instruction firmly settled, while causes and brain-based tests remain openly contested. The settled part already helps classrooms; the open part calls for patience.

That split matters for families and schools. The settled side points to specific teaching; the unsettled side warns against shortcuts sold as fixes. The literacy section follows how reading science reaches classrooms, and dyslexia sits at the center of that traffic.

Sorting settled from unsettled is a reader skill in itself. The question of whether peer review may catch less bad science than readers assume applies here, because dyslexia has drawn more than its share of bold claims. Named evidence, dated and checked, is the filter that helps.

What has research settled?

Two points hold firm. First, dyslexia means trouble with accurate, efficient word reading that intelligence does not explain, according to the Austin team’s 2026 account of the field. Second, step-by-step phonics teaching helps; the 2026 review uses the phrase firmly settled for that case. For related coverage, see How peer review works, and what it doesn't catch.

What dyslexia is not matters too. Reading disability is not the same thing as dyslexia, the authors note, because language disorders can produce poor reading by a different route. Word-reading trouble comes first in dyslexia, and comprehension problems often follow as a downstream cost.

The evidence trail is long. A 2014 meta-analysis in PLoS ONE by Katharina Galuschka and colleagues pooled 22 randomized trials and found phonics instruction was the only approach whose effect on reading and spelling reached statistical significance. Every other approach tested came in weaker.

Where does the evidence stop?

Causes remain open. The International Dyslexia Association’s 2025 definition, discussed in the Austin review, now lists genetic, neurobiological, and environmental influences together, with phonological problems common but not universal. No single cause explains every case, and the edges of the label still get argued.

Brain scans add caution. A 2026 study in Human Brain Mapping led by Alison Schug of Georgetown University retested gray matter differences in 262 children from a large national sample, and the differences failed to replicate. The authors ask readers to treat older scan findings with care.

Timing is a live question as well. Early oral language weaknesses often foreshadow literacy challenges in the 2025 definition, which points to the preschool years as a window worth study. How early screening should run, and for whom, stays unsettled in the same documents.

What about the claims sold as fixes?

The Austin review warns that misunderstandings have fed fad and pseudoscientific treatments with no support from reading research. Its list names coloured overlays and lenses, dyslexic fonts, vision therapy, and action video game training. None carries the evidence that phonics instruction holds.

The pattern behind the list repeats. Each product starts from a claim about cause, such as visual stress, and sells a matching remedy. When the cause claim fails a trial, the remedy falls with it. Structured literacy keeps its lead because its trials hold.

How the evidence stacks up

The three sources differ in strength. The 2014 meta-analysis pooled randomized trials, the strongest design for treatment claims. The 2026 review synthesizes definitions and instruction studies. The 2026 scan study ran a discovery and replication pair, which is why its null result carries weight.

Limits remain. The trials in the 2014 pool were and mostly lasted months, not years. The scan sample covered one age range of children. None of these papers closes every question, and none claims to.

Conclusion: settled core, open edges

The settled core is practical. Dyslexia shows up as slow, effortful word reading, and structured phonics teaching holds the strongest trial record. The edges stay open: causes multiply, brain markers wobble, and fad fixes keep circling. Dated, named evidence remains the best filter for the rest.

Sources

  1. Dyslexia: What It Is, Symptoms, Treatment & Types
  2. Dyslexia - Wikipedia
  3. Dyslexia Basics - International Dyslexia Association
  4. What is Dyslexia? - Yale Dyslexia
  5. Dispelling Myths and Misunderstandings About Defining and Remediating Dyslexia — Dyslexia (via Europe PMC)
  6. A Discovery and Replication Study of Dyslexia Does Not Reveal Reproducible Gray Matter Volume Differences — Human Brain Mapping (via Europe PMC)
  7. Effectiveness of treatment approaches for children and adolescents with reading disabilities: a meta-analysis of randomized controlled trials — PLoS ONE (via Europe PMC)

More from our brands

Part of the VUGA Network

Frequently Asked Questions

What is dyslexia, in plain terms?
Dyslexia is a specific learning disability marked by trouble reading words accurately and at a normal speed, and it appears across a range of severity. The International Dyslexia Association’s 2025 definition ties it to word-reading and spelling difficulty, with comprehension and writing problems often following later.
Does research support special dyslexia fonts or colored lenses?
No. A 2026 review in the journal Dyslexia lists coloured overlays and lenses, dyslexic fonts, vision therapy, and action video game training among treatments without support from reading research. Structured literacy instruction that teaches phonics explicitly holds the strongest evidence base for helping students with dyslexia.
Do brain scans diagnose dyslexia?
Not reliably. A 2026 study in Human Brain Mapping tested gray matter differences in 262 children and found the differences failed to replicate between its two samples. The authors advise caution toward older scan findings, and no scan-based test can replace behavioral reading assessment today.