Muscle Strength, Cardio and Lung Fitness Cut Dementia Risk — UK Biobank Proteomics Reveal NfL, GDF15 and Hippocampal Pathways

Table of Contents

1. Key Highlights
2. Introduction
3. How the study combined population data, plasma proteomics and brain imaging
4. Distinct fitness domains, independent effects
5. Proteomics: different proteins, overlapping biological themes
6. Structural mediation: hippocampal volume explains part of the link
7. Who benefits most: sex and age differences
8. Estimating population impact: roughly one quarter of dementia cases linked to suboptimal fitness
9. Mechanistic picture: neuroinflammation, vascular health, and structural preservation
10. Translating findings into prevention strategies
11. Limitations and remaining questions
12. Practical takeaways for clinicians, policymakers and individuals
13. FAQ

Key Highlights

• Strong, independent associations between three fitness domains (handgrip strength, cardiorespiratory fitness, pulmonary function) and lower dementia risk were observed in 51,517 UK Biobank participants over 12 years; highest versus lowest tertiles showed hazard ratios of 0.50, 0.62, and 0.73 respectively.
• Plasma proteomics identified domain-specific molecular signatures—neurofilament light chain (NfL) dominating for muscular and cardiorespiratory fitness, inflammatory mediators including GDF15 for pulmonary function—and revealed converging neuroinflammatory and neurovascular pathways; hippocampal volume partially mediated the fitness–dementia link (3.7–10.1%).
• Population-attributable fraction modeling estimated suboptimal fitness may explain roughly 26% of dementia cases, emphasizing the potential for life-course prevention that targets multiple physical-fitness domains.

Introduction

Dementia now affects more than 55 million people worldwide and presents a mounting public-health burden as populations age. Identifying modifiable risk factors and the biological mechanisms that link them to neurodegeneration remains essential for prevention. A large, integrative analysis of 51,517 participants from the UK Biobank followed for up to 12 years offers one of the most detailed pictures yet of how different dimensions of physical fitness relate to incident dementia—and why.

This research brings three techniques together: traditional epidemiology to quantify risk, large-scale proteomics to map circulating molecular signals, and brain MRI to detect structural mediators. The result is a multilayered account showing that muscular fitness, aerobic capacity, and lung function each protect against dementia through partly distinct biological routes that converge on neuroinflammation, vascular integrity, and preservation of hippocampal structure. The findings carry direct implications for prevention strategies, clinical risk assessment, and the design of interventional trials that aim to reduce dementia incidence across populations.

How the study combined population data, plasma proteomics and brain imaging

Large observational cohorts can reveal population associations but rarely illuminate molecular or structural mechanisms. This study bridged those gaps by integrating three linked data streams in the same cohort.

• Epidemiological analysis: Investigators analyzed 51,517 UK Biobank participants with baseline measurements of handgrip strength, estimated cardiorespiratory fitness, and pulmonary function, then tracked incident dementia over a median follow-up of up to 12 years. Models adjusted for a wide range of confounders and produced hazard ratios comparing tertiles of fitness.
• Plasma proteomics: Plasma from study participants underwent high-throughput proteomic profiling. The analysis identified proteins whose concentrations correlated with each fitness domain and independently predicted dementia risk. Rather than a single biomarker, each fitness domain associated with a panel of 22–40 proteins that implicated specific biological pathways.
• Neuroimaging mediation: A subset of participants had brain MRI available. Structural measures—most importantly hippocampal volume—were tested as mediators in the pathway from fitness to dementia. The MRI findings quantified the extent to which preserved brain structure accounted for the association between higher fitness and reduced dementia risk.

Together, this design allowed the study to link behavioral exposures (fitness), circulating molecular signatures (proteins), and brain structure—providing a plausible mechanistic chain instead of isolated associations.

Distinct fitness domains, independent effects

The study treated physical fitness as multidimensional rather than a single construct. Three domains showed independent associations with dementia risk:

• Handgrip strength (muscular fitness): Participants in the highest third of grip strength had a 50% lower dementia risk compared with those in the lowest third (HR 0.50). Grip strength is a simple, validated proxy for overall muscle strength and functional capacity across the lifespan.
• Cardiorespiratory fitness (aerobic capacity): Higher estimated cardiorespiratory fitness associated with a 38% reduction in dementia risk for top versus bottom tertiles (HR 0.62). Aerobic fitness influences cerebral blood flow, metabolic resilience, and vascular health.
• Pulmonary function (respiratory capacity): Better lung function predicted a 27% lower dementia risk (HR 0.73) comparing highest to lowest tertiles. Pulmonary health affects oxygen delivery and systemic inflammation—both relevant to brain aging.

Crucially, each domain retained predictive power when models included the other domains. That pattern indicates additive and partially independent protective effects rather than a single latent “fitness” factor driving the association. The independence of effects matters for prevention: improving one domain is beneficial, but simultaneous improvements across domains may yield greater protection.

Real-world example: An older adult who maintains a regular resistance program to preserve grip and leg strength, engages in weekly moderate-to-vigorous aerobic activity for cardiovascular fitness, and addresses smoking cessation and chronic obstructive pulmonary disease (COPD) for lung health is likely to derive broad-based protection across these independent pathways. The study suggests that such multidomain attention to fitness could substantially lower dementia risk.

Proteomics: different proteins, overlapping biological themes

Linking circulating proteins to both fitness and dementia generated key insights into underlying biology. The proteomic analysis detected 22–40 protein predictors per fitness domain and revealed distinct dominant signals:

• Neurofilament light chain (NfL) predominated for muscular and cardiorespiratory fitness. NfL is a marker of neuroaxonal injury; lower levels in fitter individuals suggest either reduced baseline neuronal damage or enhanced clearance mechanisms. NfL is widely used in neurology as a sensitive marker for neurodegeneration.
• Inflammatory mediators, including growth differentiation factor 15 (GDF15), were prominent for pulmonary function. GDF15 rises in systemic stress states and is associated with inflammation and mitochondrial dysfunction. Elevated GDF15 has been tied to cardiopulmonary disease and frailty.
• Across domains, proteins converged on neuroinflammatory and neurovascular pathways. That convergence indicates multiple upstream pathways—mechanical forces of muscle use, aerobic metabolic effects, and pulmonary-mediated systemic inflammation—may all funnel into common processes that affect brain health.

Interpreting domain-specific proteins The presence of different dominant proteins across fitness domains supports the idea of partly divergent mechanisms:

• Muscular fitness and NfL: Muscle strength may protect neurons indirectly through improved metabolic support, better glycemic control, and reduced frailty. Lower circulating NfL among stronger individuals is consistent with less ongoing neuroaxonal damage or a lower burden of subclinical neurodegeneration.
• Cardiorespiratory fitness and NfL: Aerobic fitness boosts cerebral perfusion, augments mitochondrial function, and fosters vascular remodeling. These benefits reduce neuronal stress and, thus, NfL release.
• Pulmonary function, inflammatory proteins and GDF15: Chronic lung impairment elevates systemic inflammation and hypoxic stress, which can injure the cerebrovasculature and trigger inflammatory cascades implicated in neurodegeneration.

The proteomic panels contained tens of proteins; the combined patterns implicated processes that are already recognized in dementia research—blood–brain barrier integrity, endothelial function, immune activation, and neuroaxonal integrity—but the study ties these molecular footprints to specific, modifiable fitness domains.

Clinical translation: proteomics as a bridge Plasma proteomics allows measurement of a broad set of circulating markers that may serve as early signals of risk or as intermediate outcomes in intervention trials. If a fitness program lowers a set of proteins linked to dementia—especially NfL or inflammatory mediators—it provides molecular evidence that the intervention is affecting disease-related biology even before cognitive decline emerges.

Structural mediation: hippocampal volume explains part of the link

Neuroimaging offered a complementary layer of evidence. Hippocampal volume emerged as a significant structural mediator between fitness and dementia. Quantitatively, hippocampal preservation accounted for roughly 3.7% to 10.1% of the association depending on the fitness domain. That proportion signals two points:

• Structural preservation is a real, measurable pathway: higher fitness correlates with larger hippocampal volumes, which in turn reduces dementia risk.
• Most of the protective effect is mediated by other mechanisms. The modest mediation percentages imply the fitness–dementia relationship also operates via systemic, vascular, inflammatory, and molecular channels not captured by hippocampal volume alone.

Why the hippocampus matters The hippocampus supports episodic memory and is among the earliest structures to atrophy in Alzheimer’s disease. Preserving hippocampal structure through exercise and vascular health aligns with the known neuropathology of common dementia syndromes. Still, the fact that hippocampal mediation explains only a minority of the effect reinforces that prevention must reach beyond neuroanatomy.

Contrast with clinical trials Some randomized trials of aerobic and resistance exercise have reported preserved hippocampal volume or slowed atrophy in older adults. The current study provides population-level corroboration, linking habitual fitness to hippocampal integrity and downstream dementia risk. However, the MRI mediation numbers emphasize that structural change is one of multiple actionable targets.

Who benefits most: sex and age differences

The analysis revealed stronger associations in women and in younger participants. That pattern deserves careful interpretation:

• Sex differences: Women showed larger protective associations between fitness domains and dementia risk. Biological differences in body composition, hormonal milieu, vascular aging, and susceptibility to neurodegeneration may contribute. Social and behavioral factors—differences in activity patterns, health care access, or comorbidity profiles—could also influence effect sizes.
• Age differences: Associations were stronger in younger participants within the cohort. Early-to-mid-life fitness may have greater preventive leverage, consistent with life-course models of dementia risk where midlife exposures set trajectories for late-life brain health.

These subgroup findings have implications for targeting prevention. Interventions in midlife and programs attentive to sex-specific needs and responses may yield higher impact. They also raise research priorities: understanding why women derive stronger protective associations could inform tailored programs and mechanistic studies.

Real-world implications: midlife interventions Population health strategies that promote strength training and aerobic fitness beginning in midlife—through workplace programs, primary-care counseling, community exercise initiatives, and policy measures that support active transport—could realize larger gains than interventions that start in late life alone.

Estimating population impact: roughly one quarter of dementia cases linked to suboptimal fitness

The investigators used population-attributable fraction modeling to estimate the proportion of dementia cases that might be avoided if everyone achieved higher fitness levels. The model suggested that suboptimal fitness accounts for approximately 26% of dementia cases.

Interpretation and caveats A 26% attributable fraction emphasises the potential public-health impact of fitness promotion. Several points temper the estimate:

• Attributable fractions assume causality and that risk factor modification will change risk as projected by observational associations. Randomized trials are required to confirm causal reductions in dementia.
• The figure aggregates across domains; real-world implementation would face barriers to achieving and sustaining higher fitness levels across entire populations.
• Confounding and reverse causation (incipient cognitive decline leading to lower activity) are concerns. The study used long follow-up and adjustments to limit these biases, but observational designs cannot eliminate them entirely.

Still, even conservative interpretations suggest substantial potential. For context, leading modifiable dementia risk factors—hypertension, smoking, physical inactivity, and obesity—are commonly targeted in public-health strategies. Recognizing muscular strength, aerobic fitness, and pulmonary health as separate and actionable contributors refines prevention priorities.

Policy relevance A prevention agenda that treats multiple fitness domains as key targets could be integrated into national dementia strategies, long-term care planning, and primary-care performance metrics. Funding for community-based strength and aerobic programs, smoking cessation and air-quality policies, and coverage for pulmonary rehabilitation would align with the mechanistic evidence presented.

Mechanistic picture: neuroinflammation, vascular health, and structural preservation

Putting the proteomics and MRI findings together yields a coherent mechanistic narrative with three intersecting themes:

1. Neuroinflammation and systemic immune signaling:
   • Fitness influences systemic inflammatory tone. Better pulmonary function is linked to lower inflammatory mediators such as GDF15; aerobic fitness and muscular strength also associate with anti-inflammatory profiles.
   • Chronic systemic inflammation alters microglial activation, promotes neurotoxic responses, and disrupts synaptic maintenance—all pathways implicated in dementia.
2. Neurovascular function and blood–brain barrier integrity:
   • Cardiorespiratory fitness enhances endothelial function and cerebral perfusion, reducing small-vessel disease and white matter injury.
   • Pulmonary impairment and systemic inflammation can harm the cerebrovasculature, increasing vulnerability to ischemia and contributing to mixed-pathology dementia.
3. Neuroaxonal integrity and structural preservation:
   • Lower circulating NfL in fitter individuals suggests reduced axonal injury. Protected hippocampal volume provides a structural correlate of cognitive resilience.
   • These influences combine to preserve cognitive networks and delay the clinical thresholds for dementia.

The proteomic signals act as fingerprints of these pathways. Their domain specificity implies that different forms of exercise or interventions might engage particular mechanisms more strongly—strength training could prioritize neuroaxonal protection, while aerobic conditioning might improve vascular function.

Practical consequences for trial design Interventional trials can use domain-specific proteomic signatures and imaging outcomes as intermediate endpoints. For example, trials of resistance training may include serial NfL measurements and hippocampal volumetry; aerobic training trials could focus on endothelial biomarkers and perfusion imaging. Using multiomics and imaging shortens the time to detect biologically meaningful effects compared with waiting for dementia endpoints.

Translating findings into prevention strategies

The evidence supports a broad-based prevention strategy that targets multiple fitness domains across the life course. Key practical recommendations follow from the study’s findings and existing exercise science:

• Preserve and build muscle strength: Regular resistance training two to three times per week—progressive, multi-joint exercises targeting major muscle groups—maintains grip strength and functional reserve. Community programs and home-based regimens can scale access.
• Improve cardiorespiratory fitness: Moderate-to-vigorous aerobic activity accumulating to at least 150 minutes weekly reduces cardiometabolic risk and improves cerebral blood flow. Structured programs, active commuting, and intervals can all lift aerobic capacity.
• Protect pulmonary function: Smoking cessation has immediate and long-term benefits for lung health and systemic inflammation. For people with chronic respiratory disease, guideline-driven pulmonary rehabilitation, optimized inhaler therapy, and vaccination reduce exacerbations and preserve function.
• Integrate approaches: Programs that combine strength, aerobic, and breathing exercises may provide additive or synergistic effects. The presence of independent protective associations across domains justifies combined interventions.
• Start early: The stronger associations in younger participants indicate that midlife or earlier interventions will have larger cumulative effects on dementia risk.

Health systems should treat fitness promotion as prevention medicine. Primary-care clinicians can incorporate brief screening (grip dynamometry, walking tests) and targeted referrals. Employers and municipalities can promote active design and accessible fitness options, which disproportionately benefit people with limited resources.

Barriers and equity Social determinants shape access to fitness opportunities. Low-income communities face fewer safe spaces for exercise, limited access to recreational facilities, and higher burdens of smoking and air pollution. Policies that reduce inequality—affordable community fitness programs, subsidized pulmonary rehabilitation, and smoke-free environments—could magnify the population-level benefits suggested by the study.

Limitations and remaining questions

Large integrative studies provide detailed portraits but cannot resolve all causal and practical questions. Important limitations and open issues include:

• Observational design: Associations do not prove causation. Reverse causation (preclinical dementia reducing fitness) and residual confounding are possible despite adjustments and long follow-up.
• Measurement specifics: The study used baseline fitness measures and did not capture longitudinal changes in fitness throughout follow-up. Understanding how midlife trajectories and later-life declines relate to risk requires repeated measures.
• Proteomic complexity: Proteomics identifies associations but not direct causal links. Some proteins may be markers rather than mediators. Experimental validation is needed to establish whether changing protein levels through interventions reduces dementia risk.
• Generalizability: The UK Biobank population has known selection biases (healthier and less deprived than the general population). Effect sizes and attributable fractions may differ in more diverse or disadvantaged groups.
• Clinical translation: Whether interventions that raise specific proteins or imaging measures translate into reduced clinical dementia incidence remains an empirical question for randomized trials.

Research priorities Key next steps include randomized trials that test whether targeted exercise programs alter proteomic signatures and imaging markers and reduce cognitive decline; longitudinal studies that map fitness trajectories across decades; mechanistic experiments that manipulate candidate proteins such as GDF15 to test causal roles; and implementation research to scale multifaceted fitness interventions equitably.

Practical takeaways for clinicians, policymakers and individuals

For clinicians:

• Assess physical fitness as part of midlife preventive care. Simple measures—grip dynamometry, timed walk tests, spirometry—offer actionable information.
• Recommend combined approaches: resistance training, aerobic activity, and pulmonary health optimization for patients at risk.

For policymakers:

• Invest in community infrastructure and programs that promote equitable access to strength and aerobic exercise and smoking-cessation services.
• Integrate fitness promotion into dementia prevention strategies and chronic-disease management.

For individuals and caregivers:

• Strength training and aerobic activity provide complementary benefits for brain health; aim to include both.
• Preserve lung health—avoid smoking, get recommended vaccinations, and manage chronic respiratory conditions—to reduce systemic inflammation and support cognition.
• Start early and stay consistent. Midlife measures matter for late-life outcomes.

Each action targets specific mechanisms revealed by the study: muscular strength relates to neuroaxonal integrity, aerobic fitness to perfusion and vascular health, and pulmonary function to systemic inflammation. Together they form a coherent prevention framework.

FAQ

Q: Does this study prove exercise prevents dementia? A: The study provides strong, multi-layered observational evidence that higher muscular strength, cardiorespiratory fitness, and pulmonary function associate with lower dementia risk and links these associations to proteomic and structural brain markers. Observational data cannot conclusively prove causality; randomized trials that demonstrate cognitive or incidence benefits following fitness interventions remain the gold standard. The weight of evidence, including randomized trials showing cognitive benefits of exercise in certain populations, supports fitness as a promising preventive strategy.

Q: Which fitness domain should I prioritize for brain protection? A: All three domains—muscular strength, aerobic capacity, and pulmonary function—show independent protective associations. The most prudent approach is to incorporate both resistance and aerobic exercise into a routine and to protect pulmonary health by avoiding smoking and managing respiratory disease. Combining domains likely offers the broadest protection.

Q: What are NfL and GDF15, and why do they matter? A: NfL (neurofilament light chain) is a protein released into blood and cerebrospinal fluid when neurons or axons are damaged. Lower circulating NfL among fitter individuals suggests less ongoing neuroaxonal injury. GDF15 (growth differentiation factor 15) is a circulating mediator elevated in stress, inflammation, and certain cardiopulmonary conditions; it featured prominently in the pulmonary-fitness proteomic signature. Both serve as biomarkers that link peripheral physiology to brain health.

Q: How much of the fitness effect is due to hippocampal preservation? A: Hippocampal volume mediated approximately 3.7% to 10.1% of the association between fitness and dementia. This indicates hippocampal preservation is one pathway among several—most of the protective effect likely operates through other molecular and vascular mechanisms.

Q: Does pulmonary function matter even for non-smokers? A: Yes. Pulmonary function influences oxygen delivery and systemic inflammation regardless of smoking status. Environmental exposures, respiratory infections, and chronic lung diseases can affect pulmonary health and thereby impact the pathways linked to dementia.

Q: What age is best to start focusing on fitness for dementia prevention? A: The study found stronger associations in younger participants, suggesting that midlife fitness matters. Starting in midlife—or earlier—maximizes cumulative benefits. However, benefits accrue at any age; improving fitness later in life still supports cognition and overall health.

Q: Can proteomic testing guide individual prevention? A: Proteomic profiling provides insights at the research and trial level. Currently, routine proteomic testing for dementia prevention is not established in clinical practice. However, biomarkers such as NfL are increasingly used in neurology for risk stratification and monitoring; future trials may validate proteomic panels as intermediate endpoints to guide personalized prevention.

Q: What next steps will confirm these findings? A: Randomized controlled trials that test multidomain fitness interventions, measure proteomic and imaging endpoints, and follow cognitive outcomes will be decisive. Mechanistic studies that manipulate candidate mediators and implementation research to scale equitable programs will help translate findings into population health gains.

Q: Are there risks to increasing fitness? A: Exercise poses low risk for most people when programs are adapted to individual capability. Resistance training and aerobic activity may result in musculoskeletal strain if not progressed safely. Individuals with chronic conditions should seek tailored programs and professional guidance. Pulmonary rehabilitation and smoking-cessation interventions are evidence-based and safe under clinical supervision.

Q: How should health systems respond? A: Health systems should integrate fitness assessment into routine care, offer referrals to community-based exercise and rehabilitation programs, and prioritize policies that reduce disparities in access. Funding for trials that use proteomic and imaging markers as intermediate outcomes will accelerate evidence-based prevention strategies.

The study reinforces a clear message: maintaining strength, aerobic fitness, and lung health contributes to brain resilience through identifiable molecular and structural pathways. A prevention strategy attentive to all three domains, implemented broadly and equitably across the life course, has the potential to reduce dementia incidence meaningfully.