Bilingualism and Alzheimer’s: does speaking Catalan and Spanish protect the brain?

What is cognitive reserve?

For example, imagine two cities connected by a main road.

The first city has only that road. If it closes, traffic is blocked.

The second has several alternatives. When one stops working, vehicles can take other routes and still reach their destination.

Similarly, cognitive reserve allows other routes to be used.

This does not mean the brain does not age or is free of disease. Rather, it can use its networks more efficiently, flexibly, or in a compensatory way to maintain functioning longer.

Several concepts are usually distinguished to help explain this:

Education, complex professional activity, social life, learning, cognitive exercise, and bilingualism have been studied as possible contributors to this reserve. However, they do not necessarily act in the same way or carry the same weight in everyone.

Sources: [1] Alzheimer’s & Dementia · Perspective · April 2026.

Why might speaking two languages train the brain?

When a bilingual person speaks Catalan, Spanish does not completely disappear from their brain.

Both linguistic systems can remain partly active. The brain must select the appropriate word, inhibit alternatives, maintain attention, and adapt to context.

For example, the following may come to mind:

• taula and mesa;
• finestra and ventana;
• feina and trabajo;
• records and recuerdos.

The person does not normally experience this as a major conscious effort. Over the years, however, it represents repeated practice in selection, control, memory, and flexibility.

Any potential benefit would therefore come from more than “knowing words.”

It may also involve:

• switching from one language to another;
• choosing which to use with each person;
• recognizing different grammatical structures;
• maintaining partly shared vocabularies;
• reading and writing in several codes;
• and participating in different linguistic communities.

A 2026 international review describes bilingualism as a cognitively complex activity involving attention and executive control mechanisms. However, it also warns that the supposed “bilingual advantage” does not appear universally across all tasks or populations.

Sources: [1] Alzheimer’s & Dementia · Perspective · April 2026.

Catalonia: an almost unique natural laboratory

In this context, Catalonia offers a particularly interesting scientific opportunity.

Many people have been exposed to Catalan and Spanish from an early age. Some use both every day; others understand Catalan perfectly but mainly speak Spanish; others switch depending on family, work, friends, or location.

There is therefore no single type of bilingualism.

For example, the age each language was learned, speaking and writing proficiency, frequency of use, home language, work language, need to switch, and actual balance between the two can vary.

The BILiARD project and ALFA cohort

The Universitat Oberta de Catalunya leads the BILiARDproject, which studies which features of Catalan-Spanish bilingualism might be related to better cognitive reserve.

The project analyzes data from 2,700 healthy people, mainly ages 45 to 74, participating in the Fundación Pasqual Maragall’s ALFA project. Most are children of people with Alzheimer’s, and approximately 32 % carry the APOE4 genetic variant, associated with a higher risk of developing the disease. Cognitive performance, habits, neuroimaging, and longitudinal changes will be studied.

It is therefore important to emphasize:

BILiARD is still a research project. Its final conclusions are not yet available.

Its value lies precisely in asking more than whether someone is “bilingual” or “monolingual.” It seeks to determine which form of bilingualism, level of use, and age of acquisition might matter most.

Sources: [2] UOC · BILiARD project.

The difference between active and passive bilingualism

For example, someone may understand Catalan perfectly, follow a conversation, and read it, but almost always reply in Spanish.

Someone else may use both languages every day and switch depending on the situation.

Recent research therefore distinguishes:

This distinction appears important because the hypothetical brain training would depend not only on accumulated knowledge, but on the repeated need to select, retrieve, and alternate between both linguistic systems.

However, it should not be treated as an on-off switch.

Between fully active and passive bilingualism lies a continuum: listening, understanding, reading, writing, speaking, alternating, translating, and switching with context.

Current science therefore tries to measure this continuum rather than put millions of people into two overly simple boxes.

Sources: [3] medRxiv · Preprint · 2026 · [4] medRxiv · Preprint · April 2026.

The first 2026 study: functioning better with the same biological Alzheimer’s

A 2026 preprint compared:

• 280 active bilinguals;
• and 287 passive bilinguals;

all Catalan and Spanish speakers with biomarker-confirmed Alzheimer’s disease.

Active bilinguals performed better on tests of attention and executive functions, language, visuospatial functioning, and visuomotor coordination.

The groups were also at a similar biological disease stage. This suggests actively bilingual people might maintain better cognitive functioning despite a comparable pathological burden.

The researchers also observed differences in cerebrospinal fluid and plasma biomarkers related to amyloid burden and neuroinflammation, raising the possibility that bilingualism is related both to resilience and to some degree of biological resistance.

What this does and does not allow us to conclude

However, caution is needed.

The study is a preprint:

• it has not yet completed peer review;
• its design does not demonstrate causation;
• language use was classified using participants’ own reports;
• and social, educational, or life differences may not have been fully captured.

Overall, the result is promising.

Still, it does not show that starting to speak Catalan tomorrow will change Alzheimer’s biomarkers.

Sources: [3] medRxiv · Preprint · 2026.

The second study: symptoms 1.4 to 2.2 years later

Another 2026 preprint examined reported age at first symptom onset in Catalan-Spanish bilingual people at different stages of cognitive impairment.

After adjustment for education, sex, and disease severity, active bilingualism was associated with later symptom onset:

• 2.21 years in amnestic mild cognitive impairment;
• 1.42 years in amnestic dementia;
• 1.45 years in biomarker-confirmed Alzheimer’s disease.

These figures are interesting because they reduce some etiological uncertainty from earlier studies: the work did not rely only on clinical diagnoses, as one group had biological confirmation of Alzheimer’s.

A difference between groups, not a personal promise

However, these findings should not become an individual promise either.

For example, this does not mean every bilingual person will delay symptoms by exactly a year and a half.

Age at onset was also reconstructed retrospectively and may be influenced by patient or family memory, healthcare access, symptom awareness, education, occupation, and social differences.

It also remains a preprint.

The appropriate formulation is therefore not:

“Catalan delays Alzheimer’s by 2.21 years.”

It would instead be:

“In this sample, active use of Catalan and Spanish was associated with somewhat later symptom onset.”

Sources: [4] medRxiv · Preprint · April 2026.

The large European study: 86,149 people from 27 countries

In November 2025, Nature Aging published one of the largest studies of multilingualism and aging.

The researchers analyzed information from 86,149 people from 27 European countries and constructed a measure called biobehavioral age, based on cognitive, functional, educational, cardiometabolic, and sensory variables.

Multilingualism was associated with a lower likelihood of accelerated aging in both cross-sectional and longitudinal analyses. The authors also found a possible gradient: contexts with greater multilingualism showed more favorable trajectories.

What biobehavioral age measures

This highly interesting result nevertheless needs several qualifications.

Specifically, so-called biobehavioral age:

• is not the brain’s literal age;
• it is not obtained by directly examining how many neurons a person has;
• and it does not mean someone has physically become younger.

Some exposure to multilingualism was also estimated at the population or national level. Living in a multilingual society may be related to other factors: education, migration, social life, cultural activity, job opportunities, mobility, and socioeconomic characteristics.

The researchers made extensive statistical adjustments, but no observational study can completely eliminate every potential confounder.

The association is strong. However, causation has not yet been demonstrated.

Sources: [5] Nature Aging · November 2025.

Can four languages really make the brain look thirteen years younger?

In July 2026, researchers presented a study conducted in the Basque Country at the Federation of European Neuroscience Societies conference.

They first used magnetoencephalography and machine learning to build a “brain age” model from 728 people’s data. They then applied it to another group of 144 participants speaking between one and four languages.

According to the conference presentation:

• bilingual people showed brain patterns approximately six years younger;
• trilingual people, around seven;
• and those speaking four languages, up to thirteen years younger than monolingual people.

A more favorable association was also observed with greater language proficiency and earlier acquisition.

At first glance, the finding is spectacular.

What “younger” means to an algorithm

However, a “brain thirteen years younger” does not mean all its cells have become thirteen years younger.

It means an algorithm found functional connectivity patterns resembling those typically seen in younger people.

Also:

• the results were presented at a conference;
• they had not yet been published as a peer-reviewed article;
• the multilingual group was relatively small;
• and lifestyle or social participation differences could not be completely ruled out.

It is an excellent scientific curiosity.

Still, it is not yet a medical recommendation.

Sources: [6] FENS Forum · Conference presentation · July 2026.

Bilingualism does not automatically improve every ability

One problem in communicating science is turning a specific association into a general superpower.

For example, bilingual people have been said to have better attention, more memory, greater flexibility, better executive functioning, more creativity, and a globally superior brain.

However, the evidence does not support such a simple universal advantage.

A study published in Scientific Reports examined 94 people ages 18 to 83 using working memory, inhibition, and attention tests in auditory and visuospatial modalities.

It found no general advantages in most tasks. The association appeared specifically in visuospatial working memory among older participants with greater multilingual experience.

The researchers therefore concluded that possible cognitive effects may depend on age, the function studied, stimulus type, and specific features of language experience.

Therefore:

Being bilingual does not automatically make someone more intelligent, more attentive, or immune to impairment.

Still, it might contribute to certain forms of flexibility or compensation, especially with aging.

Sources: [7] Scientific Reports · volume 16 · 2026.

Why do some studies find benefits and others not?

The review published in Alzheimer’s & Dementia in April 2026 concludes that the role of bilingualism in cognitive aging remains uncertain and that many discrepancies stem from methodological differences.

First, studies do not all mean the same thing by “bilingual.”

For example, someone may be considered bilingual if they:

• learned two languages in childhood;
• studied a second language at school;
• migrated as an adult;
• understand a language but barely speak it;
• uses each language in completely separate settings;
• or switches constantly between the two.

Age of acquisition, proficiency, frequency, literacy, language similarity, need to switch, education, occupation, socioeconomic status, immigration, social life, and healthcare quality can also vary.

How language and evaluation matter

There is also a highly relevant clinical problem:

In which language is the cognitive test administered?

Someone may have more vocabulary, speed, or confidence in one language than another. If evaluation takes place only in the less-used language, impairment could appear greater than it really is.

Sources: [1] Alzheimer’s & Dementia · Perspective · April 2026.

Does bilingualism really prevent dementia?

We cannot yet say so.

Retrospective studies often find that bilingual people receive a diagnosis or show their first symptoms at an older age.

However, longitudinal studies have not consistently demonstrated:

• a lower incidence of dementia;
• a slower rate of decline;
• or less accumulation of pathology.

These results therefore fit the cognitive reserve model.

Bilingualism might allow someone to function better for longer even though the biological disease has already begun.

For example, imagine a dam holding back water.

The dam does not stop the water level from rising.

It delays the moment when the water overflows.

Bilingualism may therefore delay clinical expression without necessarily stopping the neurodegenerative process.

Delaying symptoms and preventing disease are not the same.

Both would be valuable.

However, scientifically, they must be distinguished.

Sources: [1] Alzheimer’s & Dementia · Perspective · April 2026.

Can being bilingual “hide” Alzheimer’s?

The word hide can create a misleading interpretation.

It does not mean bilingualism deceives the physician or prevents a person from receiving a diagnosis.

Rather, greater cognitive reserve might allow performance to be maintained longer as pathology progresses.

When reserve can no longer compensate, symptoms begin to become visible.

This raises two clinical consequences:

However, this hypothesis does not mean bilingual people should worry more.

Any progressive change in memory, orientation, or functioning warrants evaluation, even if the person maintains strong cultural knowledge and verbal skills.

Sources: [1] Alzheimer’s & Dementia · Perspective · April 2026.

Does actively using both languages matter?

Overall, recent data point in that direction.

However, listing a second language on a résumé or having studied it for several years does not appear enough.

The hypothetical benefit might be greater when the person:

• regularly converses in both languages;
• reads and writes;
• switches depending on the person they are speaking to;
• maintains a functional vocabulary;
• and uses both languages in meaningful settings.

This does not mean listening to or understanding a second language has no value.

Repeated active use may require more intense cognitive work than passive exposure.

The 2026 Catalan-Spanish studies defined active bilinguals as people regularly using both languages in everyday life, while passive bilinguals were exposed to Catalan but mainly used Spanish.

Sources: [3] medRxiv · Preprint · 2026 · [4] medRxiv · Preprint · April 2026.

Does it matter that Catalan and Spanish are similar languages?

This is a very interesting scientific question.

Both are Romance languages and share many words, structures, and sounds. This proximity may facilitate learning, but also create more competition between similar alternatives.

For example, when speaking, the brain may need to distinguish between similar words:

• possible and posible;
• important and importante;
• memòria and memoria;
• família and familia.

Some researchers therefore propose that managing closely related languages might require particularly fine control.

However, it has not yet been demonstrated that speaking two similar languages offers more protection than speaking distant languages. Current projects specifically investigate how similarity, proficiency, and switching frequency matter.

We therefore cannot currently rank “languages that are better for the brain.”

Sources: [6] FENS Forum · Conference presentation · July 2026 · [1] Alzheimer’s & Dementia · Perspective · April 2026.

Is learning a language after 50 or 60 useful?

It can be very worthwhile.

However, two claims must be distinguished.

Longitudinal trials capable of demonstrating that starting a second language in adulthood reduces dementia risk are still lacking.

Still, learning can offer independent benefits: cognitive stimulation, social contact, a sense of progress, cultural participation, confidence, purpose, and access to new activities.

The WHO includes staying cognitively and socially active among measures associated with lower risk of cognitive impairment. Learning a language can be an appealing way to serve both functions, although it is neither a specific nor a guaranteed preventive intervention.

Sources: [8] World Health Organization · [1] Alzheimer’s & Dementia · Perspective · April 2026.

Seven ways to make languages a real cognitive activity

1. Speak, rather than only study

Conversation requires retrieving information in real time.

2. Change contexts

Using one language with family and another in cultural, professional, or social activities may also broaden the networks involved.

3. Read somewhat challenging texts

For example, do not read only headlines or short messages. Fiction, essays, or the press require sustained attention and context.

4. Write

Preparing an email, diary, reflection, or story requires more active retrieval than recognizing words.

5. Learn new vocabulary

Even in our own native language, we can keep expanding our repertoire.

6. Use the language with other people

The social component may be as relevant as the linguistic one.

7. Keep the challenge appropriate

An activity that is too easy requires no adaptation. An impossible task, however, creates frustration and abandonment. The goal is therefore progressive difficulty.

Academic perfection is not necessary.

Cognitively, making mistakes, correcting yourself, and trying again are part of learning.

Languages do not replace better-supported preventive measures

Being bilingual does not compensate for:

• untreated hypertension;
• poorly controlled diabetes;
• smoking;
• physical inactivity;
• harmful alcohol use;
• social isolation;
• uncorrected hearing loss;
• or chronically poor sleep.

The WHO identifies hypertension, diabetes, obesity, depression, vision or hearing loss, poor sleep, head injury, stroke, tobacco, harmful alcohol use, physical inactivity, isolation, and environmental pollution as factors associated with higher dementia risk.

A set of habits, rather than isolated protection

The best-supported priorities are therefore:

• staying physically active;
• not smoking;
• avoiding harmful alcohol use;
• following a balanced diet;
• controlling blood pressure, blood glucose, and cholesterol;
• treating hearing problems;
• maintaining social connections;
• getting adequate sleep;
• and continuing cognitively meaningful activities.

Bilingualism can be added to this set. However, it cannot replace it.

Sources: [8] World Health Organization.

The clinical detail that can change a diagnosis: language of evaluation

Imagine someone who has spoken Catalan with family all their life but worked mainly in Spanish for forty years.

In which language should a memory test be administered?

And a naming test?

Also, how should an interview about emotional symptoms be approached?

The answer is not always to choose only one.

An evaluation of a bilingual person should therefore establish:

• the language learned first;
• the currently dominant language;
• the family language;
• reading and writing proficiency;
• frequency of use;
• professional context;
• and any recent changes.

A healthy person may take somewhat longer to find a rarely used word in one of their languages. This does not automatically mean neurocognitive impairment.

At the same time, disease may affect each language differently depending on when it was acquired and how much it was used.

The 2026 international review warns that testing in an area’s main language without considering the patient’s actual dominant language or the instrument’s validation can introduce evaluation bias.

Language is not an administrative detail. It is part of the clinical history.

Sources: [1] Alzheimer’s & Dementia · Perspective · April 2026.

Which memory lapses are normal, and which should cause concern?

Occasionally forgetting a name, taking time to find a word, or entering a room without immediately remembering why can happen to healthy people.

For example, fatigue, anxiety, lack of sleep, pain, overload, medication, alcohol, and divided attention can contribute.

Progressive changes affecting life

It is worth seeking evaluation when changes are progressive and begin to affect functioning.

The WHO lists possible early signs including:

• forgetting recent events;
• repeatedly misplacing objects;
• becoming disoriented in familiar places;
• losing track of time;
• having difficulty solving problems;
• following conversations;
• finding words;
• or doing previously familiar tasks.

Apathy, withdrawal, irritability, personality changes, anxiety, or loss of interest may also appear before or alongside memory loss.

A memory complaint therefore does not necessarily mean Alzheimer’s.

However, progressive difficulty interfering with life warrants investigation.

Sources: [8] World Health Organization.

Memory, anxiety, depression, and sleep: why differential diagnosis is needed

Not everyone who forgets things is developing dementia.

Attention is the gateway to memory.

If someone is:

• ruminating;
• anxious;
• depressed;
• sleeping four hours;
• sedated by medications;
• using alcohol or cannabis;
• or under extreme overload,

they may register information less effectively and later feel they have forgotten it.

In other situations, however, there may be mild cognitive impairment, a neurodegenerative disease, a vascular disorder, an endocrine problem, a nutritional deficiency, or medication effects.

An appropriate evaluation should not choose in advance between “it is all emotional” and “it is all neurological.”

It should therefore investigate both possibilities.

Sources: [8] World Health Organization.

A person can lose words without losing their identity

Language changes are particularly painful.

Someone may forget a word, mix languages, or progressively return to the language learned in childhood.

The family may therefore feel they are losing part of who the person was.

However, a person with cognitive impairment is not reduced to memory difficulties or a diagnostic label. Diagnosis can help organize care and treatment, but should never sum up the whole person or erase their life story.

This also has implications for care.

Whenever possible, it is worth:

• communicating in the language in which the person feels safest;
• avoiding correcting every error;
• using clear sentences;
• preserving their preferences;
• and maintaining meaningful language activities.

Language is more than a cognitive function.

It also contains memories, humor, intimacy, affection, belonging, songs, and ways of relating to the world.

Sources: [1] Alzheimer’s & Dementia · Perspective · April 2026 · [8] World Health Organization.

The evidence traffic light

Sources: [1] Alzheimer’s & Dementia · Perspective · April 2026 · [3] medRxiv · Preprint · 2026 · [4] medRxiv · Preprint · April 2026 · [5] Nature Aging · November 2025 · [6] FENS Forum · Conference presentation · July 2026 · [7] Scientific Reports · volume 16 · 2026 · [8] World Health Organization.

An integrative perspective: language as part of a protective life

An integrative perspective does not turn bilingualism into a miracle therapy.

It places it within an ecosystem.

For example, speaking several languages may involve learning, flexibility, conversation, cultural belonging, community life, reading, travel, and intergenerational contact.

Each of these elements may help maintain a mentally active life.

However, the brain also needs movement, sleep, cardiovascular health, hearing, nutrition, relationships, purpose, and appropriate treatment of depression, anxiety, and other illnesses.

Perhaps a language’s protective effect comes from more than managing vocabularies.

It may also come from what that language allows someone to do:

Talk with their children, maintain friendships, participate in the community, read, learn, and keep part of their identity alive.

My clinical reading of bilingualism and Alzheimer’s

The hypothesis is plausible

Managing several languages involves prolonged, repeated cognitive practice.

The new Catalan-Spanish data are particularly interesting

They include people with Alzheimer’s biomarkers and distinguish active use from passive exposure.

The actual size of the benefit remains uncertain

Figures vary across studies and cannot be directly applied to each person.

Causation has not been demonstrated

For example, bilingual people may differ in education, social activity, migration, occupation, and other factors.

The cognitive advantage is not universal

Not all studies find improvements in attention, memory, or executive control.

Bilingualism may act more as reserve than prevention

It may delay clinical expression without necessarily preventing the disease.

Learning a language remains worthwhile for many reasons

These include cognitive, cultural, social, and emotional reasons. However, it should not be marketed as a guaranteed medical intervention.

The sentence that best captures the situation would therefore be:

Speaking two languages may not stop Alzheimer’s from arriving, but might help the brain keep more routes open for longer.

Bilingualism and Alzheimer’s: languages are memory too

On September 11, we can look at language from another perspective.

As cultural heritage, but also as an everyday brain experience.

Every conversation, language switch, and retrieved word requires the brain to select, remember, inhibit, and adapt.

New studies in Catalan and Spanish speakers suggest this sustained exercise might contribute to cognitive reserve and delay the visible appearance of some symptoms.

However, we still cannot say it prevents Alzheimer’s.

Science therefore continues working to separate association and causation, knowledge and use, active and passive bilingualism, cognitive reserve and less pathology, and linguistic effects and social context.

While stronger answers are awaited, one conclusion does seem reasonable:

Using a language does more than keep a way of communicating alive. It maintains the relationships, memories, and learning built through it.

Learning Catalan.

Keeping Spanish.

Reclaiming a family language.

Starting English at sixty.

Talking with grandchildren.

Reading, writing, and conversing.

None of this, however, guarantees dementia will not develop.

Still, it can all be part of a more connected, cognitively active, meaningful life.

Ultimately, protecting a meaningful life is always a good investment for the brain.

Evaluation of memory, concentration, and emotional health

A clinical evaluation can help reconstruct the course, review mental health, sleep, medication, and medical factors, and determine whether a specific neurological or neuropsychological evaluation is needed.

Integrative psychiatry consultation with Dr. Marc Agraz

Evaluation of cognitive and emotional symptoms, depression, anxiety, sleep, psychiatric medications, and behavioral changes. In-person care in Lleida and online consultations when clinically appropriate, coordinating with neurology, neuropsychology, or other specialties when indicated.

A psychiatric consultation does not replace neurological evaluation when a neurodegenerative disease is suspected. It can help with differential diagnosis and organizing an appropriate referral.

Frequently asked questions about bilingualism and Alzheimer’s

Prevention and language use

Learning and number of languages

Brain age, memory, and evaluation

Habits and when to seek care

Scientific sources and resources