Does Meditation Rewire Your Brain? What fMRI and MRI Studies Actually Show

Before you read. This post summarises published research and guidelines for general information. It is not medical advice; talk to your doctor before making decisions about tests, treatment or supplements.

Corrected in October 2026: added the largest randomized trial of MBSR and brain structure (Kral et al. 2022, no structural change) and rewrote the conclusions to match it; separated Davidson 2003 (MBSR, EEG, flu vaccine) from Lutz 2004 (monks, gamma EEG) and Hölzel 2010 from Hölzel 2011; corrected the Tang, Klimecki and MBCT details; removed an invented "~100 hours" figure, an unsourced "-30% stress" figure and an unsourced timeline.

"Meditation Changes Your Brain" — But What Does the Science Actually Show?

The story usually starts in Richard Davidson's lab at the University of Wisconsin-Madison. Two of its studies are often blended into one, but they are different studies:

• Davidson et al. (2003, Psychosomatic Medicine) studied healthy employees, not monks. An 8-week MBSR group (n = 25) was compared with a waitlist control group (n = 16). The MBSR group showed a larger increase in left-sided anterior EEG activation (a pattern linked to positive affect) and a larger antibody response to a flu vaccine. It did not use fMRI or measure brain structure.
• Lutz et al. (2004, PNAS) compared 8 long-term Tibetan Buddhist practitioners with 10 student controls. During a loving-kindness and compassion meditation, the practitioners produced high-amplitude gamma-band EEG synchrony. This was also an EEG study, not a structural one.

Since then, hundreds of studies have used fMRI, structural MRI, EEG and DTI (diffusion tensor imaging). Split by outcome, the evidence looks like this:

• Function and self-report: changes in brain activity during and after meditation training, and reductions in self-reported stress, are reported reasonably consistently.
• Structure: the claim that "8 weeks of meditation grows the hippocampus and shrinks the amygdala" is not well supported. It failed to replicate in the largest and most rigorously controlled trial so far (Kral et al. 2022, below).

One terminology point: gray matter density, volume and cortical thickness are measured with structural MRI, not fMRI. fMRI measures brain activity.

The Most Important Study: A Large Trial Found No Structural Change

Kral et al. (2022) — No Brain-Structure Change After MBSR

University of Wisconsin-Madison, Davidson's own lab (Science Advances, doi:10.1126/sciadv.abk3316)

Design: Two combined three-arm randomized controlled trials
Participants: 218 meditation-naive adults
Arms: 8-week MBSR (n = 75) / matched active control, a health-enhancement program (n = 73) / waitlist (n = 70)
Measurement: Structural MRI (gray matter volume, gray matter density, cortical thickness) before and after

Results:

• MBSR produced no structural brain change compared with either the active control or the waitlist group.
• This held at the whole-brain level and in the regions of interest taken from earlier MBSR studies (hippocampus, posterior cingulate, TPJ, cerebellum and others), and in the amygdala and insula.
• The authors describe it as "the largest and most rigorously controlled study to date" and report that they failed to replicate the earlier findings.
• One exception: participants who did more MBSR home practice showed a larger reduction in right amygdala volume relative to the active control. There was no difference between the group averages.

Why this matters: the authors note that in the same participants, MBSR did reduce amygdala reactivity to emotional stimuli, changed functional connectivity and increased self-reported mindfulness. The training worked; the functional effects were there, the structural effects were not. Read the earlier structural studies below with that in mind.

The Earlier Structural Studies: What They Reported and Their Limits

Study 1: Hölzel et al. (2011) — 8 Weeks of MBSR and Gray Matter

Massachusetts General Hospital / Harvard Medical School (Psychiatry Research: Neuroimaging)

Design: Controlled longitudinal study (not randomized; MBSR participants recruited from course enrollees)
Participants: 16 meditation-naive healthy adults in MBSR
Control group: 17 waitlist controls
Measurement: Structural MRI before and after (voxel-based morphometry)

Results (increases only):

Brain Region	Change	What the region is involved in
Left hippocampus	↑ Gray matter density	Learning, memory, emotion regulation
Posterior cingulate	↑ Gray matter density	Self-referential processing
Temporo-parietal junction (TPJ)	↑ Gray matter density	Perspective-taking
Cerebellum	↑ Gray matter density	Motor control and more

The right-hand column describes what these regions generally do; the study did not measure memory or empathy. These increases did not replicate in Kral et al. 2022.

Study 1b: Hölzel et al. (2010) — The Amygdala Finding

The amygdala decrease often attributed to the 2011 study actually comes from a separate 2010 paper (Social Cognitive and Affective Neuroscience):

Participants: 26 stressed but otherwise healthy adults, 8-week MBSR
Control group: none
Result: Perceived stress (PSS) fell significantly, and people whose stress fell more
        showed a larger decrease in right basolateral amygdala gray matter density (a correlation)

With no control group and a correlational design, this does not show that meditation shrinks the amygdala.

Study 2: Lazar et al. (2005) — Long-Term Meditators

Massachusetts General Hospital (NeuroReport)

Design: Cross-sectional comparison
Participants: 20 Insight meditation practitioners (about 9 years of practice on average)
vs. 15 controls with no meditation or yoga experience
Measurement: Cortical thickness via MRI

Key findings:

Meditators had thicker cortex in:

• Right anterior insula (body awareness, interoception)
• Sensory cortices
• Parts of the prefrontal cortex

The thickness of two regions correlated with meditation experience.

The aging finding: The prefrontal difference was largest in older participants. In one prefrontal region (BA 9/10), the average thickness of 40–50-year-old meditators was similar to that of 20–30-year-old participants. The authors suggested meditation "might offset age-related cortical thinning".

Important caveat: a cross-sectional design cannot rule out
"people with thicker cortices are drawn to meditation, or stay with it".

Study 3: Tang et al. (2010) — 11 Hours of Training and White Matter

Dalian University of Technology (China) and University of Oregon (PNAS)

Design: Randomized, active control (relaxation training)
Participants: 45 University of Oregon undergraduates
Intervention: Integrative Body-Mind Training (IBMT)
Duration: 30 minutes per session over about a month, 11 hours in total
Measurement: DTI (Diffusion Tensor Imaging) — white matter

Result:

• Increased fractional anisotropy (FA) in white matter tracts connected to the anterior cingulate cortex (ACC), largest in the left anterior corona radiata
• No FA change in the relaxation group
• No gray matter change

Findings from the same group's other papers, often merged into this one:

• Reduced stress-related cortisol: Tang et al. 2007 (5 days × 20 minutes of IBMT)
• Heart rate variability and other autonomic changes: Tang et al. 2009
• Axial diffusivity changes: Tang et al. 2012

All of these are small studies with a few dozen participants each.

Study 4: The Default Mode Network

The Default Mode Network (DMN):
A set of brain regions active during:
• Mind-wandering
• Self-referential rumination ("Did I say something wrong?")
• Worry about future / regret about past
• Unfocused "default" brain activity

What meditation does to the DMN:

Brewer et al. (2011), Yale (PNAS):
• Experienced meditators vs meditation-naive controls
• fMRI during several meditation types
• Main DMN nodes were relatively deactivated in experienced meditators:
  - Medial prefrontal cortex
  - Posterior cingulate cortex (PCC)
• Authors' interpretation: consistent with less mind-wandering

This is a plausible neural correlate of the "quieter mind" meditators describe, but it is a cross-sectional comparison, so it cannot show cause.

Brain Region by Region: Where the Evidence Stands

Hippocampus: Memory and Stress Regulation

Structural evidence:
• Hölzel et al. (2011): left hippocampal gray matter density increase after MBSR (small, non-randomized)
• Luders et al. (2009): long-term meditators had larger right hippocampal volume (cross-sectional, 44 subjects)
• Kral et al. (2022): not replicated in the largest randomized trial

Note:
• None of these studies showed better memory directly

Amygdala: Fear and Stress Reactivity

Functional evidence:
• Kral's group reports reduced amygdala reactivity to emotional stimuli after MBSR
  in the same trial participants who showed no structural change
• Taylor et al. (2011): in beginners, a mindful state reduced left amygdala responses
  to emotional pictures; experienced meditators instead showed lower DMN activity

Structural evidence:
• Hölzel et al. (2010): amygdala gray matter decrease correlated with stress reduction (no control group)
• Kral et al. (2022): no group difference

Self-report:
• Lower perceived stress after MBSR (e.g., Hölzel 2010)

Prefrontal Cortex: Executive Control

Evidence:
• Lazar et al. (2005): parts of the prefrontal cortex thicker in long-term meditators (cross-sectional)
• Brefczynski-Lewis et al. (2007): see "Focused Attention" below

Regions often discussed:
• Dorsolateral PFC (dlPFC): Working memory, planning, cognitive flexibility
• Ventromedial PFC (vmPFC): Emotion integration, self-referential processing
• Anterior cingulate cortex (ACC): Attention regulation, conflict monitoring

Insula: Interoception and Empathy

The insula processes:
• Internal bodily states ("How does my body feel?")
• Integration of emotion and physical sensation
• Empathy — sharing others' emotional states

Evidence in meditators:
• Increased cortical thickness in the right anterior insula (Lazar et al., 2005; cross-sectional)
• Greater insula activation to distress sounds during compassion meditation
  in experts than in novices (Lutz et al., 2008; fMRI)

Corpus Callosum: Left-Right Connections

The corpus callosum connects the left and right hemispheres.

Evidence:
• Luders et al. (2012): 30 long-term meditators vs 30 controls; callosal FA and thickness
  larger in meditators, mainly in anterior sections (significant only with fine-grained,
  point-wise measures; cross-sectional)
• Tang et al. (2010): FA increases in the genu and body of the corpus callosum after 11 h of IBMT

The authors themselves note that longitudinal studies are needed to tell whether
this is caused by meditation or predates it.

How Long Before You See Effects?

An earlier version of this post had a week-by-week timeline of brain changes. It was not based on specific studies, so it has been removed. What the research does support:

• 5 days (20 min/day): better attention-task scores, improved mood and lower
  stress-related cortisol vs relaxation training (Tang 2007, 40 students)
• About a month (11 h total): white matter FA increase (Tang 2010, 45 students)
• 8 weeks of MBSR: lower self-reported stress in several studies;
  no structural change in the largest randomized trial (Kral 2022)
• Years of practice: structural differences in cross-sectional studies,
  which cannot show that meditation caused them

Types of Meditation: Different Practices, Different Brain Effects

Neuroscience research distinguishes several meditation types with somewhat different neural profiles.

Focused Attention (FA) Meditation

Practice: Sustaining attention on a single object (usually breath)
When mind wanders, notice and return attention

Primary neural targets:
• Dorsal attention network (sustained attention)
• Anterior cingulate cortex (monitoring for mind-wandering)
• Dorsolateral PFC (executive attention control)

Best for:
• Improving concentration and sustained attention
• Reducing distractibility

Representative fMRI finding:
• Brefczynski-Lewis et al. (2007): activation in attention regions followed an inverted U.
  Experts averaging ~19,000 hours of practice showed MORE activation than novices;
  experts averaging ~44,000 hours showed LESS.

Open Monitoring (OM) Meditation

Practice: Non-reactive, non-judgmental awareness of all arising experience
No single focus — monitoring the field of awareness itself

Primary neural targets:
• Insula (interoceptive awareness)
• Default Mode Network regulation
• Frontoparietal networks (metacognition)

Best for:
• Reducing rumination and worry
• Emotional regulation
• Self-awareness and metacognition

Loving-Kindness / Compassion Meditation (LKM)

Practice: Cultivating warm feelings and goodwill toward self and others
Progressive: self → loved ones → neutral → difficult people → all beings

Key studies:
• Fredrickson et al. (2008): six group sessions over 7 weeks increased daily positive
  emotions, which in turn increased personal resources (e.g., social support) and
  reduced depressive symptoms
• Lutz et al. (2004): long-term practitioners showed high-amplitude gamma EEG synchrony
  during loving-kindness/compassion meditation (EEG)
• Lutz et al. (2008): during compassion meditation, experts showed greater insula
  activation to distress sounds than novices (fMRI)
• Klimecki et al. (2013, 2014): EMPATHY training increased anterior insula and anterior
  midcingulate activation along with negative affect; COMPASSION training instead increased
  activation in medial orbitofrontal cortex, ventral striatum and pregenual ACC, along with
  positive affect

Best for:
• Self-criticism
• Building positive affect
• Empathic distress and burnout risk (particularly relevant for caregivers, healthcare workers)

MBSR vs. MBCT: What's the Difference?

MBSR (Mindfulness-Based Stress Reduction):
• Developed by Jon Kabat-Zinn, UMass Medical School
• 8 weekly 2.5-hour group sessions + one full day of practice + daily home practice
• Focus: Stress reduction, chronic pain, health conditions
• Evidence: self-reported stress reduction; early structural MRI findings
  did not replicate in the largest randomized trial (Kral 2022)

MBCT (Mindfulness-Based Cognitive Therapy):
• Adapted for recurrent depression prevention
• Combines MBSR with cognitive therapy elements
• Kuyken et al. (2016) individual patient data meta-analysis (9 trials, 1,258 patients):
  lower risk of depressive relapse within 60 weeks vs. no MBCT
  (hazard ratio 0.69, i.e. about 31% lower relative risk);
  HR 0.79 vs. active treatments including antidepressants;
  the benefit looked larger in people with more severe residual symptoms

For brain health:
MBCT has the stronger clinical evidence specifically for preventing depressive relapse.
MBSR has a broader evidence base for stress and general well-being.
Neither has convincing evidence for structural brain change.

Practical Guide: Starting a Practice

How Much Practice?

There is no established "dose" that produces structural brain change.
For reference, MBSR participants in the Kral 2022 trial averaged about
32 hours of home practice over 8 weeks (range 2–85 hours), and still showed
no group-level structural change.

The practical answer: pick an amount you can sustain. Regular short practice
is easier to keep up than occasional long sessions.

Building the Practice

Week 1-2: Foundation
• 5-10 minutes daily, consistent time
• Simple breath focus (count breaths 1-10, restart at 1)
• Use an app (Insight Timer, Headspace, Waking Up)
• Don't worry about "doing it right"

Week 3-4: Expansion
• 10-15 minutes daily
• Add body scan 2-3x per week
• Brief mindful moments throughout day (eating, walking)

Week 5-8: Deepening
• 20 minutes daily
• Try OM practice alongside FA
• Formal sitting practice, not just "while doing something else"

Month 3+: Maintenance and refinement
• 20-30 minutes daily
• Explore different traditions and approaches
• Retreat or workshop to deepen practice

What's Normal (and What Isn't)

Normal experiences in meditation:
• Lots of mind-wandering (especially early on) ✓
• Sleepiness during practice ✓
• Restlessness and discomfort ✓
• Not feeling "meditative" or peaceful ✓
• Noticing how busy the mind is ✓
(This noticing IS the practice)

Requires attention or stopping:
• Intense, distressing emotions that don't settle
• Dissociation or unreality feelings
• Resurgence of trauma-related material
• Increasing anxiety or panic in practice
(These are reasons to seek guidance, not push through)

Recommended Apps

Headspace:
• Beginner-friendly, structured curriculum
• Good research base (Headspace has funded peer-reviewed studies)

Calm:
• Strong sleep meditation library
• Variety of guided sessions

Insight Timer:
• Free (large free library)
• Community features
• Huge variety of teachers and styles

Waking Up (Sam Harris):
• More philosophical / contemplative angle
• Strong on theory and understanding
• Good for people who want to understand what they're doing

Limitations and Honest Caveats

Methodological issues in the field:
• Most early structural studies had a few dozen participants
• Many lacked randomization or used only waitlist controls (hard to blind)
• Long-term meditator studies are cross-sectional (self-selection)
• Publication bias toward positive results
• Heterogeneous meditation interventions across studies

What this means:
When a large, randomized trial with an active control was run (Kral 2022),
the structural findings did not replicate. Functional and self-report
effects remain better supported than structural ones.

Clinical caution:
• Meditation is not a first-line treatment for severe mental illness
• Can be destabilizing for some trauma survivors
• Should complement, not replace, professional treatment when indicated
• "Meditation-related adverse events" are real, if uncommon

Conclusion

What the evidence reasonably supports:

• Brain activity patterns (fMRI, EEG) differ during and after meditation training
• Programs like MBSR reduce self-reported stress
• MBCT lowers the risk of depressive relapse in recurrent depression (HR ~0.69 in the Kuyken 2016 meta-analysis)

What it does not yet support:

• "8 weeks of meditation grows your hippocampus and shrinks your amygdala." The largest randomized trial with an active control found no structural change.
• That the brain-structure differences seen in long-term meditators were caused by meditation

There are good reasons to try meditation as a way of handling stress. Just be aware that "it physically rebuilds your brain" runs ahead of the evidence.

---

For the stress-brain connection: How Chronic Stress Physically Shrinks Your Hippocampus

For the cortisol-memory mechanism: Cortisol and Memory: How Stress Hormones Affect Your Ability to Remember