Table of Contents
Multilingualism and Brain Age: 13 Years Younger
Multilingualism and brain age: four-language speakers patterned 13 years younger on brain-connectivity measures. A conference finding, not a published study.
Four languages, thirteen years. On July 6, 2026, researchers presenting at the Federation of European Neuroscience Societies Forum in Barcelona reported that people who spoke four languages had brain-connectivity patterns corresponding to about 13 years younger than monolingual people. This was a conference presentation, not a peer-reviewed paper, and the main comparison rested on 144 people. The multilingualism and brain age link is one of the more interesting threads in cognitive ageing research, and it is also one of the easiest to overread. The researchers said so themselves.
Key Takeaways
- The study was presented at the FENS Forum in Barcelona on July 6, 2026 by a team from Spain, Chile, Argentina and Dublin, and reported by The Guardian the same day.
- Initial measurements covered 728 people of varying ages and language ability. The headline comparison used a matched subgroup of 144, with equal numbers speaking one, two, three or four languages.
- Reported brain-age gaps: about 6 years younger for two languages, 7 years for three, and 13 years for four, relative to monolinguals.
- The method was magnetoencephalography, with an AI model estimating the connectivity pattern typical for a given chronological age.
- Dr Lucia Amoruso said “higher language proficiency and earlier acquisition of a second language were also associated with more delayed brain ageing.” Association. The team could not rule out lifestyle and social engagement as explanations.
What “brain age” actually measures
Brain age is a model output, not a biological fact. A machine-learning model is trained on brain scans from many people of known chronological age, learns what a typical pattern looks like at each age, and then estimates an age for a new scan. The difference between that estimate and the person’s real age is the brain-age gap.
Cole and Franke described the approach in Trends in Neurosciences in 2017 as a class of “innovative brain ageing biomarkers,” and the framing is important: a brain-age gap is a summary statistic about how typical your scan looks, which has been found to correlate with health outcomes in large cohorts. It is not a measurement of how well your brain works, and it is not a clinical test.
The 2026 study used magnetoencephalography rather than MRI. MEG measures the tiny magnetic fields produced by electrical activity in neurons, with millisecond timing, which makes it good at capturing how regions coordinate rather than what they look like. So the quantity being modelled is functional connectivity, how brain regions talk to each other over time, not brain structure or volume.
That distinction matters when you hear “13 years younger.” It does not mean a 60-year-old four-language speaker has the brain of a 47-year-old. It means their pattern of inter-regional coordination resembled the average pattern of a 47-year-old in this model, on this instrument, in this sample.
Why the sample size is the number to watch
The study reports 728 people in its initial measurement group, which is respectable for MEG work. MEG scanners are rare and expensive, and sessions take time, so cohorts in the hundreds are normal rather than lazy.
The comparison that produced the headline, though, used 144 people split into four groups of equal size. That is 36 people per language count. With 36 people per cell, a single unusual participant moves the group average noticeably, and the confidence interval around “13 years” is likely wide. The press report does not give it, which is typical of conference coverage and is itself a reason for caution.
There is also a specific feature of the setting that cuts both ways. The research was conducted in the Basque region, where multilingualism is common and the languages involved were Spanish, Basque, French and English. A high-multilingualism region is the right place to find four-language speakers in sufficient numbers. It also means the four-language group is culturally and educationally distinctive in ways that are hard to fully match.
Professor Eef Hogervorst, commenting on the work, named the core confound directly: people who speak multiple languages “also engage in healthier lifestyles and/or have better access to other protective environments.” Education, income, social network size and physical activity all travel with multilingualism and all independently predict cognitive ageing.
How multilingualism and brain age fit with the rest of the evidence
| Study | Design | What it found | Main limitation |
|---|---|---|---|
| Bialystok, Craik & Freedman, Neuropsychologia, 2007 | Retrospective clinic records of patients with dementia | Bilingual patients showed symptom onset later than monolinguals | Retrospective; immigration and education differences hard to separate |
| Livingston et al., Lancet Commission, 2024 | Systematic review of modifiable dementia risk factors | Identified a set of life-course factors with population-level effects on dementia risk | Population attributable fractions are estimates, not individual guarantees |
| Sahakian, Perrin & Langley, Psychological Medicine, 2026 | Editorial reviewing evidence on reading for pleasure | Reading associated with better cognition and lower dementia risk across the lifespan | An editorial synthesis, not a new meta-analysis |
| FENS Forum presentation, July 2026 | MEG brain-age modelling, 144-person matched comparison | 13-year younger brain-age pattern in four-language speakers | Conference abstract; small cells; confounds acknowledged by authors |
Read down the “main limitation” column. The field has a consistent and honest problem: you cannot randomly assign children to grow up multilingual, so every study is observational, and observational studies of multilingualism are studies of everything that comes with it.
What a parent should actually take from this
Learn the language for the language’s reasons
The strongest case for raising a child multilingual has never been dementia risk fifty years out. It is talking to grandparents, reading books in their original, working across borders, and having a second way of seeing the world. Those benefits are immediate, certain and uncontested. Treat any cognitive-ageing bonus as a possible extra rather than the point. Our overview of the cognitive research on bilingual kids covers what the childhood evidence does and does not support.
Note what the study says about timing and proficiency
Amoruso’s comment about higher proficiency and earlier acquisition is the most practically useful line in the coverage, and also the least reported. If the effect is real, it appears to scale with how well and how early you learn rather than with how many languages you can name. A child who genuinely uses two languages daily is a better fit to this pattern than one who has had four years of classroom exposure to three.
Resist the dose-response reading
Six years for two languages, seven for three, thirteen for four looks like a curve that rewards adding more. With 36 people per group, it could equally be noise around a single effect. If a follow-up study with larger groups reproduces the jump between three and four, that becomes a genuinely strange and interesting finding. Until then, do not plan a curriculum around it.
Put it beside the other lifestyle findings
The honest version of lifestyle-and-brain research is that several ordinary activities correlate with better cognitive ageing, and reading for pleasure has similar support from the 2026 Psychological Medicine review we cover in reading and dementia risk. None of them is a treatment. All of them are things worth doing anyway. That is a much calmer and more defensible place to stand than “learn four languages to save your brain.”
What not to do
Do not push a reluctant child into a third or fourth language on the strength of a conference abstract. Forced language instruction that a kid resents produces neither fluency nor whatever benefit fluency might carry, and it costs hours that could go to something they would actually keep doing. If your family has a real reason for a language, that reason is sufficient. If it does not, this study is not one.
What to Watch For Over the Next 3 Months
- Week 4: Watch for the study appearing as a peer-reviewed paper with confidence intervals and the full statistical model. Conference abstracts routinely change substantially before publication.
- Month 2 red flags: Apps and language programmes citing “13 years younger” in advertising. That number came from 36 people in a matched subgroup and was presented with caveats the marketing will remove.
- Month 3 self-check: Ask yourself whether you can explain what brain age is without using the phrase “younger brain.” If you can say “it’s how typical your scan looks compared to an age model,” you have it right.
Frequently Asked Questions
Does learning a language really slow brain ageing?
The evidence is associational and consistent with slowing, but nobody has shown it causally. You cannot randomise children into multilingual upbringings, so the field relies on observational comparisons where education, income and social engagement are tangled with language. The honest summary is that multilinguals tend to show better cognitive ageing markers, and why remains open.
Is my child too old to start a second language?
No. The research in this area, including the 2026 presentation, suggests earlier acquisition and higher proficiency associate with larger effects, which is an argument for starting sooner rather than a cutoff. Adults learn languages successfully all the time; the accent gets harder, the comprehension does not.
Does an app count?
For the purposes of this research, what was measured was proficiency in languages people actually speak, not exposure. Twenty minutes a day on an app can build real proficiency over years if it is paired with use. A streak on its own is not what the study measured.
How many languages should we aim for?
There is no evidence-based target, and the apparent jump at four languages rests on 36 people. Pick languages your family has a reason to use. Two used well is a better goal than four studied badly.
Should I worry if we only speak one language at home?
No. Monolingual families produce healthy, cognitively sharp people in enormous numbers, and the 2024 Lancet Commission on dementia prevention identifies many modifiable factors that have nothing to do with language. Hearing loss, physical activity, education, social contact and vascular health all appear in that literature with stronger evidence than language count.
About the author
Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.
Sources
- Bawden, A. (2026). “Learning another language appears to slow brain ageing, scientists say.” The Guardian, 6 July 2026. https://www.theguardian.com/education/2026/jul/06/learning-another-language-appears-to-slow-brain-ageing-scientists-say
- Cole, J. H., & Franke, K. (2017). “Predicting Age Using Neuroimaging: Innovative Brain Ageing Biomarkers.” Trends in Neurosciences, 40, 681–690. https://doi.org/10.1016/j.tins.2017.10.001
- Bialystok, E., Craik, F. I. M., & Freedman, M. (2007). “Bilingualism as a protection against the onset of symptoms of dementia.” Neuropsychologia, 45(2), 459–464. https://doi.org/10.1016/j.neuropsychologia.2006.10.009
- Livingston, G., et al. (2024). “Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission.” The Lancet, 404. https://doi.org/10.1016/S0140-6736(24)01296-0
- Sahakian, B. J., Perrin, V., & Langley, C. (2026). “Reading and associations with brain health, cognition, well-being, and mental health.” Psychological Medicine, 56, e262. https://doi.org/10.1017/S0033291726105625
- Wikipedia contributors. (2026). “2026 in science — July.” Wikipedia. https://en.wikipedia.org/wiki/2026_in_science
