For most of the twentieth century, a quiet assumption ran through both neuroscience textbooks and popular thinking about aging: that the brain reached something like a finished state by early adulthood, and that the decades after that were a long, slow departure from that peak.

The assumption was never quite that blunt in the scientific literature. Researchers knew that some forms of plasticity continued across life. But the framing — the implicit story — was of a brain that matured, stabilized, and then declined. Plasticity was something that happened in childhood. What came after was maintenance, and eventually, loss.

In 2026, a wave of research has complicated that picture in ways that are both more specific and more useful than the broad claim that the adult brain “can change.” It can, and does — but the mechanisms, the timescales, and the conditions vary enormously depending on age, sex, physiology, and experience. What follows is an account of what several of this year’s studies found, what they could and could not establish, and why the differences between types of plasticity matter more than the headline.

What Does "Plasticity" Mean for a 72-Year-Old Brain?

The word plasticity covers several distinct things: the growth of new synaptic connections, changes in the volume and density of specific brain regions, shifts in how efficiently a network communicates, and the brain’s capacity to recruit alternative pathways when a familiar one is damaged or strained. These are not the same process, they do not peak or decline at the same time, and they are not equally responsive to the same inputs.

A 2026 study published in Nature npj Aging examined 353 cognitively normal older adults with a mean age of 72.77 years. The researchers measured multiple dimensions of physical fitness — cardiorespiratory capacity via VO2max, muscular strength, and body composition — alongside structural MRI, blood biomarkers, and cognitive tests (Schwarck et al., Nature npj Aging, 2026).

The findings were specific rather than global. Higher VO2max was associated with greater thickness of the medial temporal lobe and larger gray matter volume in regions involved in memory. VO2max also explained additional variance in verbal memory performance beyond what was accounted for by tau burden — a marker of Alzheimer’s-related pathology — suggesting that cardiorespiratory fitness was contributing something to cognitive performance over and above the accumulation of pathological protein.

Muscular fitness told a different part of the story. Greater muscle strength was associated with fewer perivascular spaces in the basal ganglia — tiny fluid-filled channels around blood vessels that, when enlarged, are thought to reflect impaired glymphatic clearance. The blood biomarkers measured included BDNF (brain-derived neurotrophic factor), VEGF (vascular endothelial growth factor), and Cathepsin B, each a candidate mechanism by which physical activity might influence brain tissue.

The study cannot establish that exercise caused any of these structural features — this was a cross-sectional design with no randomized assignment. Older adults who maintained high cardiorespiratory fitness throughout their lives differ from sedentary adults in dozens of ways that the researchers could not fully control. What the study does show is that the associations between physical fitness and brain structure are still legible at a mean age of nearly 73, and that they map onto specific regions and mechanisms rather than appearing as a uniform effect across the brain.

The Molecular Relay Between Muscle and Brain

One question the structural studies naturally raise is mechanistic: if exercise is associated with brain structure, what is doing the carrying? How does a working muscle communicate with neurons?

A 2026 narrative review in the European Journal of Translational Myology examined what is currently understood about the irisin-BDNF signaling pathway in aging populations (Ersever, EJTM, 2026). The chain runs roughly as follows: during physical exercise, a protein called PGC-1α is activated in muscle cells. PGC-1α drives the expression of a membrane protein called FNDC5, which is then cleaved to release irisin into the bloodstream. Irisin can cross the blood-brain barrier, where it appears to stimulate BDNF production — particularly in the hippocampus, a region central to memory formation.

This pathway has been studied in both animal models and humans, though the human evidence remains less complete. The reviewer noted that different exercise modalities — aerobic, resistance, and combined training — produce distinct patterns of irisin and BDNF response, and that the magnitude of the effect changes with age. Older adults show lower baseline irisin levels and a different response curve compared to younger adults, which the author connects to both sarcopenia (the age-related loss of muscle mass) and the elevated risk of cognitive decline.

The limitation here is that narrative reviews synthesize existing literature without the formal bias controls of a systematic review or meta-analysis. They are useful for tracing a mechanism, but they can over-represent published positive findings and under-represent null results. The irisin-BDNF pathway is real and measurable, but the question of how large and durable its contribution to brain plasticity is in living humans — as opposed to in cell culture or rodents — remains open.

What the molecular research does clarify is that the brain’s capacity to respond to physical input does not simply disappear in older adulthood. The responsiveness changes — the dose, the modality, and the timing may all matter differently at 70 than at 30 — but a signaling pathway that links muscle activity to hippocampal BDNF production is still operating in aging tissue.

The Hormonal Transition and the Brain's Vulnerability Window

The research on exercise and fitness has largely been conducted across mixed-sex samples or primarily in men. A separate line of 2026 research has focused on women specifically, and on the neurobiological changes that accompany the hormonal transitions of perimenopause and postmenopause.

A 2026 narrative overview in the International Journal of Molecular Sciences examined the roles of aging, estrogen, nutrition, and neuroinflammation in women’s cognitive health and their connection to Alzheimer’s disease risk (Lephart, Weber, and Hedges, IJMS, 2026). The authors describe estrogen not as a reproductive hormone with incidental effects on the brain, but as a molecule that influences neuronal function, vascular tone, anti-inflammatory signaling, and mitochondrial efficiency in the brain directly. The decline in estrogen during perimenopause and the sustained low levels in postmenopause affect brain regions that include the prefrontal cortex, the hippocampus, and the amygdala — areas involved in working memory, verbal processing, and emotional regulation.

The authors frame what they call a “window of vulnerability” during the menopausal transition, when the combination of declining estrogen and rising inflammatory burden may make the brain more susceptible to pathological protein accumulation. The four pillars they examine — aging per se, estrogen loss, dietary factors (particularly omega-3 fatty acids and phytoestrogens), and neuroinflammation — interact rather than acting independently. Estrogen modulates inflammatory signaling, inflammatory activity accelerates neurodegeneration, and dietary inputs influence both.

This is a narrative overview, which carries the same limitation as the Ersever review: it synthesizes rather than formally weighs evidence. The authors are not claiming that menopause causes Alzheimer’s disease, or that estrogen supplementation prevents it — those are contested questions the review explicitly does not resolve. What the synthesis argues is that estrogen’s decline represents a biological event with direct consequences for brain plasticity and resilience, not simply a reproductive transition with downstream mood effects.

For the study of lifelong brain plasticity, the women’s hormonal transition data adds a dimension that the fitness literature often omits: that the capacity for plasticity is not just a function of age in years, but of the hormonal and inflammatory milieu that different people carry at different points in their lives.

When Adaptation Becomes a Liability: Stress and Maladaptive Plasticity

Not all plasticity is beneficial. The brain changes in response to chronic stress and adverse experience just as it changes in response to exercise — but the direction of the change is different.

A 2026 paper in Frontiers in Molecular Neuroscience examined stress-related neuroplasticity as a developmental risk factor, using functional neurological disorder (FND) as a case study (Mavroudis et al., Frontiers in Molecular Neuroscience, 2026). FND encompasses neurological symptoms — weakness, tremor, sensory disturbance — that arise from brain dysfunction rather than structural damage. The researchers proposed that early-life stress and adverse experiences produce lasting neurobiological alterations in systems governing emotional processing, interoception, and motor control, and that these alterations represent a form of maladaptive plasticity.

The mechanism they describe is one of overlearning under adversity: a nervous system shaped by chronic threat becomes hyperresponsive to ambiguous signals, particularly those involving the body. What begins as an adaptive response to a dangerous environment — heightened vigilance, amplified interoceptive sensitivity — can persist as an error-prone prediction model long after the original threat has passed. The brain, in other words, learns the lesson too well and continues to act on it inappropriately.

This is not a new idea — the effects of early adversity on brain development have been documented for decades. What the 2026 paper adds is a detailed account of how predictive coding models and contemporary neuroplasticity frameworks might explain the specific pattern of vulnerability in FND. The authors are careful to note that their model is theoretical and that the causal chain from early adversity to maladaptive plasticity to neurological symptoms has not been experimentally established in humans.

For a broader understanding of lifelong brain change, the FND research makes visible something the fitness and exercise literature tends to bracket: the brain that arrives at midlife or late adulthood was shaped by decades of experience before it got there. Its plasticity — however measured — is not a property of the brain in isolation, but the outcome of everything it has already adapted to.

Limitations: What the 2026 Research Does and Doesn't Establish

Several limitations run across the studies described above, and they are worth naming plainly rather than passing over.

First, most of the structural and biomarker findings come from cross-sectional designs. They compare groups of people who differ in a variable of interest — fitness, hormonal status, stress history — at a single point in time. They cannot establish that any of the observed brain differences were caused by the variable in question, because the groups likely differed in many other ways that the studies did not fully measure or control for. A longitudinal design — following the same people over years and measuring both the exposure and the brain outcome — is necessary before causation can be claimed.

Second, the molecular evidence for mechanisms like the irisin-BDNF pathway is substantially stronger in animal models than in human studies. Rodent hippocampal tissue responds to irisin in well-characterized ways; the translation to the human brain, across the complexity of whole-body physiology and longer timescales, involves assumptions that have not yet been empirically validated at the same resolution.

Third, the studies described here used different populations, different age ranges, different fitness measures, and different brain imaging protocols. It is not straightforward to combine their findings into a single picture of “what the adaptable brain looks like across a lifetime.” Each finding applies to its specific population and study design.

Finally, none of this research — including the systematic review by AbuQeis et al. on exercise, cognitive function, and brain health across the lifespan (PRISMA 2020, ibrain, 2026) — allows group-level findings to be translated into predictions about any individual. A person who maintains high cardiorespiratory fitness at age 72 has a different statistical profile from a sedentary peer, on average. Whether any specific individual’s brain is more plastic, more resilient, or less vulnerable to any specific outcome is not something the data can say.

The honest summary of what 2026 has taught us about the adaptable brain is narrower than the headline: physical fitness continues to associate with specific structural brain features in older adulthood; molecular pathways linking exercise to brain function remain operative across the lifespan; women’s hormonal transitions constitute a biologically significant window for brain plasticity and risk; and the brain’s capacity to adapt to adversity can become a source of dysfunction as well as resilience. Each of these is meaningful and specific. None of them, taken alone, resolves the question of whether a given person’s brain can meaningfully change in a given direction.