Why "Gold-Standard" Phonological Therapy Alone Is Being Questioned
Structured phonological intervention — teaching letter-sound correspondence, phonemic awareness, decoding — is still the intervention with the deepest evidence base for dyslexia, and it remains the benchmark against which everything else gets measured. A 2025 protocol paper for a serious-games trial states this directly in its background section: speech and reading therapy is currently the gold-standard intervention for improving children's reading abilities. That's a background claim from the game's own investigators, not a systematic evidence review, but it matches what the rest of the literature in this post assumes. The same protocol immediately names the problem clinicians live with every day: intensive interventions are difficult to implement — kids need more sessions than caseloads, schedules, or funding allow.
That gap is part of what's driving interest in other approaches. A 2022 French trial protocol adds a second reason to look beyond any single intervention approach: dyslexia itself may not be one thing. Its stated rationale is aimed at remediation studies generally, not phonics specifically — most existing approaches assume dyslexia results from a single cause (phonological, visual-attentional, or cross-modal), when in fact many children show deficits across more than one of those domains at once. Five other approaches discussed below — a single-case protocol for children with high intellectual potential, transcranial direct current stimulation (tDCS), transcranial alternating current stimulation (tACS), a gamified home-practice app, and neurofeedback — are being tested against that backdrop, together with the multimodal training trial, for a total of six studies. They are not all asking the same kind of question. Some test whether adding a technique on top of standard therapy improves outcomes; the HIP-DD single-case study instead compares an intensive targeted protocol against standard speech therapy, and the serious-game trial tests whether therapy contact time can be safely cut in half. Supplementing, replacing, and reducing contact are three different clinical questions, and it matters which one a given study is actually asking.
Six Studies, Six Different Stages of Evidence
Before comparing what these approaches "show," it matters enormously whether a given study has produced any results at all. Of the six papers behind this post, four are trial protocols — public pre-registrations of a study design, published before a single outcome has been analyzed. Only two have completed data collection and reported findings.
- Multimodal cognitive training (France, 2022): protocol only, registered on ClinicalTrials.gov in 2019; no outcomes reported.
- HIP-DD single-case remediation protocol (2023, n=4): protocol only; no outcomes reported.
- hd-tDCS over V5/MT (2025): protocol only, registered 2023; no outcomes reported.
- Poppins Clinical serious game (2025): protocol only; recruitment expected to begin December 2024, completion by August 2025; no outcomes reported at the time of writing.
- Gamma-tACS combined with phonological training (2023): completed trial reporting a positive result.
- Neurofeedback systematic review (2025): completed synthesis of 12 studies reporting a null result.
That distribution matters. Exactly one modality here — tACS — has a completed trial reporting a measurable benefit, and exactly one — neurofeedback — has a completed synthesis concluding there isn't one. The other four (multimodal training, the HIP-DD single-case protocol, tDCS for V5/MT, and the serious-games noninferiority trial) simply haven't run far enough yet to tell you anything about efficacy. A protocol paper can tell you a great deal about how carefully a question is being asked; it cannot tell you the answer.
Multimodal Cognitive Training: Matching the Method to the Child's Profile — Before Knowing Which Piece Works
The French crossover trial is the most ambitious design in this set: 120 children aged 8 to 13 spend two months on baseline speech-language therapy, then rotate through three two-month blocks — phonological training, visual-attentional training, and cross-modal training — with two randomized groups of 60 receiving the phonological and visual-attentional blocks in reversed order. The main objective, as the protocol states it, is to test whether the three interventions together improve reading skills compared to baseline. Testing whether each individual component adds something beyond baseline, and whether delivery order changes the outcome, are listed as secondary objectives — not as findings, since the trial has not yet reported results. Reading comprehension, spelling, and self-reported disorder impact are the outcomes to watch for once results are eventually published.
A second protocol illustrates the sampling problem that runs through this entire field. A 2023 single-case design is testing a categorical-perception and rapid-naming remediation protocol in dyslexic children with high intellectual potential (HIP-DD), a population the authors note is poorly referenced in the literature. The study enrolls four children, followed for 30 weeks with weekly measurement, comparing an intensive targeted protocol against standard speech therapy within each child. Worth flagging: this protocol's ClinicalTrials.gov registration (NCT04028310) is identical to the multimodal cognitive-training trial above. Both appear to come out of the same French research group and registration record, so treat them as related pre-registrations rather than fully independent trials when you're tracking this literature. A four-person single-case design can generate genuinely useful hypotheses about a specific subgroup; it cannot tell you whether a protocol works for dyslexic children generally, and it was never designed to.
tDCS and tACS: Two Different Targets, Two Different Stages of Evidence
Two electrical-stimulation approaches are in play here, targeting different brain regions with different current types, and they sit at very different points in the evidence pipeline.
A 2025 protocol is testing high-definition transcranial direct current stimulation (hd-tDCS) over left V5/MT, a motion-sensitive visual area implicated in reading that most neurostimulation studies in dyslexia have overlooked in favor of language regions. The design is a triple-blind, within-subjects study in 36 children and adolescents aged 8 to 13, comparing active stimulation over V5/MT, active stimulation over a control site (V1), and sham stimulation, while measuring reading, eye movements, and EEG oscillations. Like the multimodal trial, this is a protocol: it describes what will be measured, not what was found.
A completed 2023 trial took a different target and a different current type — gamma-band transcranial alternating current stimulation (tACS) over bilateral auditory cortex — and it is the one genuine positive result in this entire set of sources. Children and adolescents with dyslexia (sample size not reported in the abstract available to us) received 10 sessions of gamma-tACS over five weeks, delivered concurrently with phonological and spelling training. The stimulation shifted the peak frequency of auditory gamma oscillations, phonemic processing skills improved, and spelling gains were still measurable four months after the intervention ended. That's a real, specified, combined protocol: dose (10 sessions, 5 weeks), target (bilateral auditory cortex), and pairing with therapy (concurrent phonological and spelling training) are all stated. It is also a single trial, in one sample, that has not yet been independently replicated.
Between these two studies, what you're looking at isn't a disagreement so much as two distinct research questions, pursued for different theoretical reasons: the tDCS protocol is probing a motion-sensitive visual pathway rarely studied in dyslexia, while the completed tACS trial targets auditory cortex specifically because of its established role in phonemic processing. Only one of the two has completed enrollment and reported outcomes. Neither study, on its own or together, establishes a standard dose, electrode placement, or supervision protocol that a clinician could point to — each simply reports what its own investigators chose to test, for their own study's purposes.
Gamifying Therapy: Can a Serious Game Substitute for Therapist Contact Time?
The most caseload-relevant question in this set comes from a 2025 protocol testing Poppins Clinical, a serious game combining rhythm and written-language exercises. The trial uses a noninferiority design — a specific statistical framework built to ask whether a more scalable option is not meaningfully worse, rather than whether it's better. Three hundred and six children with dyslexia will be randomized to either 20 minutes of daily home gameplay plus one reading-therapy session every two weeks, or one reading-therapy session every week with no game. Recruitment was expected to begin in December 2024, with study completion by the end of August 2025 — meaning at the time this trial was protocolized, no reading-accuracy, speed, or comprehension outcomes existed yet. The real question the design is built to answer isn't whether the game "works" in isolation; it's whether halving in-person session frequency, backed by daily gamified home practice, holds up against standard weekly care. That's a genuinely useful question for a field short on therapist hours — but it is, again, a question that hasn't been answered yet.
Neurofeedback: The Outlier With a Null Result
A completed 2025 systematic review pooling 12 articles on neurofeedback treatment for reading impairments is the least ambiguous finding in this entire post. Its background section frames neurofeedback as a technique with "studies published over the last decade" suggesting it "may serve as a substantial complement" to reading instruction — but that's a single framing sentence, not a claim about the size of the evidence base, and the review itself searched a 20-year window, not a 10-year one. What the review actually concluded, after pooling those 12 studies, was that neurofeedback techniques alone do not present a significant benefit in improving reading skills in the dyslexic population, and that there is currently no significant evidence to determine the efficacy of neurofeedback interventions in this population. The review also flagged heterogeneity among the 12 pooled neurofeedback studies specifically — different protocols, different training targets, different outcome measures — which the authors say complicated any unbiased determination of efficacy. That heterogeneity finding is specific to neurofeedback's larger evidence base; it can't be extended to the other five approaches in this post, four of which are single protocols with no other studies to be heterogeneous against. Unlike the other approaches discussed here, neurofeedback isn't sitting on an early positive signal waiting for a bigger sample; it has an actual completed synthesis, and that synthesis came up without one.
Why "Mixed Evidence" Isn't the Same as "No Evidence"
It's tempting to flatten all of this into "the evidence is mixed" and move on. Don't. Mixed evidence and no evidence require different responses from you, and heterogeneous protocols require different responses depending on what stage they're actually at.
- Check whether a paper is a protocol or a completed trial. Titles that include "protocol," "study protocol," or a future completion date are telling you the outcomes don't exist yet. Four of the six papers behind this post fall into that category.
- Check the sample and design. A 4-person single-case study and a 36-person within-subjects protocol are answering different kinds of questions than a 306-person randomized trial or a completed trial with an unreported sample size. None of these numbers alone justify a practice-wide policy change.
- Check whether "unsettled" means conflicting findings or simply that not enough studies have finished. For tDCS and tACS right now, it's mostly the latter — the two studies are pursuing different targets for different theoretical reasons, and only one has run to completion.
- Check whether a null result is really a null result. The neurofeedback review is not inconclusive the way protocol papers are inconclusive — it's a completed synthesis of 12 studies concluding no significant benefit. That's a different, and more decisive, kind of uncertainty than "we haven't finished collecting data yet."
What This Means for Your Caseload Today
If a parent or administrator brings you a headline about brain stimulation, a dyslexia app, or neurofeedback, here's what the six sources behind this post actually support you saying.
- Structured phonological and reading therapy remains the intervention with the deepest published trial support; nothing here proposes replacing it outright, though the HIP-DD and Poppins Clinical protocols are explicitly testing whether a different or reduced-contact model can match it.
- One completed trial (gamma-tACS) shows a real, specified benefit when paired with phonological and spelling training, with effects holding at four months — but it's one trial in one sample, and it hasn't been replicated.
- One completed systematic review (neurofeedback) concludes there currently isn't significant evidence it works on its own, despite background literature spanning roughly two decades.
- Four approaches — multimodal cognitive training, the HIP-DD single-case protocol, hd-tDCS over V5/MT, and the Poppins Clinical serious game — are still mid-protocol. There's nothing wrong with a well-designed protocol, but "a trial is underway" is not the same claim as "a trial found a benefit."
- For three of these five approaches, this source set contains exactly one study each, so noting that there's "no cross-study agreement" on dosing or targeting is really just a statement about how young this research is — not a finding about the field. Even the tACS trial, the one with a positive completed result, stands alone and unreplicated.
What not to promise: don't tell a family that any of these five approaches is validated as a standard of care, don't imply neurofeedback has "some evidence" when the most recent systematic synthesis found none, and don't let excitement about a protocol paper substitute for waiting on its results.
A Therapist's Checklist for the Next "Breakthrough" Dyslexia Intervention
The next study, app, or device that crosses your desk will come with a citation and probably a compelling one-page summary. Before you factor it into a conversation with a family or an IEP team, run it through this checklist.
- Is this a completed study or a protocol? Look for language like "study protocol," "will be recruited," or a future completion date — exactly the language in the Poppins Clinical and hd-tDCS papers reviewed above. If outcomes aren't reported yet, treat any efficacy claim built on it as speculative.
- What's the sample size, and does it match the claim being made? The HIP-DD study's 4-person case series supports a hypothesis about a specific subgroup, not a caseload-wide recommendation; the 306-person Poppins Clinical trial supports a much stronger claim once it reports results.
- Is there an active comparison group, or just a before/after? The hd-tDCS protocol's triple-blind, three-condition design and the multimodal trial's randomized crossover tell you more than the HIP-DD study's single-case ABCA design, which compares conditions only within the same four children.
- Has anyone tried to replicate it? The gamma-tACS trial is a single positive result, even a well-designed one — a first data point, not a consensus, and nothing in this source set replicates it yet.
- Does the study specify dose, target, and combination with existing therapy — and do other studies of the same technique agree with those parameters? If every study picks a different electrode site, training duration, or session count, standardization isn't possible yet.
- Is there a published systematic review, and what does it conclude about consistency across studies? The neurofeedback review is a useful model here: twelve studies, genuine heterogeneity, and an honest "not yet" conclusion.
- Has a professional body issued guidance? This is worth checking directly with ASHA, IDA, or your state or provincial licensing body before advising a family. None of the six studies behind this post address whether a professional organization has weighed in, so treat that as an open question to research yourself rather than something this evidence base can answer.
None of this is a reason to dismiss these approaches, and it isn't a reason to adopt them either. It's a reason to read past the abstract, notice which stage of the pipeline a given paper is actually in, and be the person in the room who can explain the difference between "a trial is testing this" and "a trial found this works."
