Dementia Prevention: The Three Modifiable Risk Factors That Matter Most

By the time a person receives a dementia diagnosis, the pathological changes in their brain have typically been accumulating for two decades or more. Amyloid plaques have been silently aggregating between neurons. Tau proteins have been tangling inside them. Inflammation has been smoldering through white matter. The diagnosis, in other words, is not the beginning of the disease. It is closer to its final chapter. This is precisely why dementia prevention research has become one of the most consequential frontiers in longevity medicine: the window for intervention is not in a neurology clinic at age 75, but in the choices, the biomarkers, and the biological environments maintained across the preceding decades of midlife.

The 2024 Lancet Commission on dementia prevention, intervention, and care identified 14 modifiable risk factors collectively responsible for approximately 45% of dementia cases worldwide [1]. That number is both humbling and empowering. It means that genetics, the factor most people fear when a parent develops Alzheimer's disease, accounts for a minority of the actual disease burden. The majority of dementia cases are, in principle, preventable or at least significantly delayable. Among the full list of modifiable factors, three emerge from the convergence of epidemiological breadth, mechanistic depth, and clinical evidence as the most critical levers: cardiovascular risk (specifically the triad of hypertension, metabolic dysfunction, and dyslipidemia), physical inactivity, and poor sleep quality. Each of these is not merely a statistical association. Each has a well-characterized biological pathway connecting it to neurodegeneration. And crucially, each is modifiable at any age, with meaningful effect on dementia-free longevity.

Why Dementia Begins in Midlife, Not Old Age

Alzheimer's disease and vascular dementia, the two most common forms, have long been framed as diseases of the elderly. This framing is clinically accurate but biologically misleading. Autopsy studies and longitudinal neuroimaging have consistently shown that amyloid deposition begins in some individuals as early as their 30s and 40s, proceeding silently for 15 to 25 years before cognitive symptoms emerge [2]. The clinical threshold for dementia, the point at which cognitive impairment becomes functionally apparent, reflects the depletion of what researchers call cognitive reserve: the brain's accumulated structural and functional capacity to absorb damage before showing dysfunction. A brain with greater reserve can tolerate more pathology before symptomatic collapse. A brain exposed to decades of vascular stress, metabolic dysregulation, and sleep deprivation depletes that reserve faster.

This reframing has profound practical implications. Midlife, defined in the dementia literature as roughly ages 40 to 65, is not too early to think about dementia prevention. It is exactly the right time. Blood pressure elevation in midlife, for instance, carries a substantially greater risk of late-life dementia than hypertension first developing after age 65 [1]. The same midlife sensitivity applies to metabolic health and physical fitness. The brain is not an organ that waits passively for old age to begin deteriorating. It responds in real time to the internal environment that the body provides, and that environment is shaped most powerfully during the decades of active working life.

The majority of dementia cases are, in principle, preventable or at least significantly delayable — and the window for intervention opens in midlife, not in a neurologist's office at 75.

Understanding this timeline changes the conversation from reactive treatment to proactive architecture. The question is not "how do we treat dementia?" but "what does a brain-healthy internal environment look like, and how do we maintain it across decades?" The three factors explored below are the most powerful tools available for building that environment.

Factor One: Cardiovascular and Metabolic Health

The brain consumes roughly 20% of the body's total energy despite comprising only about 2% of its mass. It is the most metabolically demanding organ in the body, sustained by an intricate vascular network delivering oxygen and glucose through 400 miles of blood vessels, the smallest of which are capillaries just wide enough for a single red blood cell to squeeze through in single file. When that vascular network is compromised, whether by hypertension, insulin resistance, or dyslipidemia, the consequences for neural tissue are not merely incidental. They are foundational to the dementia process.

Hypertension is the single most prevalent modifiable risk factor for dementia identified by the Lancet Commission [1]. Chronically elevated blood pressure damages the arterial wall through a process called endothelial dysfunction: the cells lining blood vessels lose their ability to dilate appropriately and begin to allow inflammatory molecules to infiltrate the vessel wall. In the brain's smallest vessels, this process manifests as white matter hyperintensities, small areas of ischemic damage visible on MRI, and silent microinfarcts. These accumulate over years, gradually eroding the connectivity between brain regions. The hippocampus, the brain structure central to memory formation, is particularly vulnerable because of its high metabolic demand and dependence on sustained cerebrovascular perfusion. A landmark analysis of data from the SPRINT MIND trial found that intensive blood pressure control (targeting systolic pressure below 120 mmHg) significantly reduced the incidence of mild cognitive impairment compared to standard treatment [3].

Insulin resistance and type 2 diabetes amplify dementia risk through overlapping but distinct mechanisms. The brain is not insulin-independent: insulin receptors are dense throughout the hippocampus and prefrontal cortex, and insulin signaling in neurons promotes synaptic plasticity and suppresses apoptosis (programmed cell death). When those receptors become resistant, as occurs systemically in metabolic syndrome, neurons lose a critical survival signal. Simultaneously, chronically elevated blood glucose generates advanced glycation end-products that stiffen arterial walls, impair mitochondrial function, and accelerate the production of amyloid-beta [4]. Some researchers now refer to Alzheimer's disease colloquially as "type 3 diabetes," a descriptor that remains contested but captures the mechanistic intimacy between metabolic dysfunction and neurodegeneration. People with type 2 diabetes carry approximately double the risk of developing Alzheimer's disease compared to metabolically healthy individuals [1].

Dyslipidemia, particularly elevated LDL cholesterol in midlife, adds a third layer of vascular risk. LDL particles that penetrate and oxidize within arterial walls trigger the formation of atherosclerotic plaques, which reduce cerebral blood flow and increase the risk of embolic events. But cholesterol's relationship to dementia is more nuanced than simple accumulation: the brain synthesizes its own cholesterol, which plays essential roles in synapse formation and myelin integrity. The relationship between peripheral lipid management and brain cholesterol homeostasis is an active area of research, with statins showing inconsistent results in cognitive outcomes trials despite clear cardiovascular benefits.

Taken together, these three cardiovascular and metabolic factors explain why the concept of cardiometabolic risk is inseparable from cognitive longevity. Treating the heart and treating the brain are not separate projects. They share biology. This is the clinical rationale for programs like continuous glucose monitoring to detect early insulin resistance and metabolic drift, tools like Healthspan's CGM Metabolic Protocol, or the use of GLP-1 receptor agonists, which have demonstrated improvements in both metabolic markers and systemic inflammation. For individuals managing weight-related cardiometabolic risk, programs like GLP-1 Longevity Care address not just body weight but the broader metabolic environment that feeds or starves the aging brain.

The Blood-Brain Barrier: A Checkpoint Under Siege

Underpinning all three cardiovascular risk factors is a structure that rarely features in popular health discussions but is central to neurodegeneration: the blood-brain barrier. This is not a metaphorical wall but a physical one, composed of tightly interlocked endothelial cells reinforced by pericytes and astrocytic end-feet, that regulates every molecule that passes from the bloodstream into neural tissue. It keeps out pathogens, neurotoxic metabolic byproducts, and peripheral immune cells. When it functions correctly, it also facilitates the clearance of brain waste, including amyloid-beta and tau, into the cerebrospinal fluid and ultimately into the lymphatic system.

Hypertension, insulin resistance, and chronic inflammation all progressively degrade blood-brain barrier integrity. Tight junction proteins, the molecular clamps that hold endothelial cells together, become disrupted. The barrier begins to leak. Peripheral inflammatory cytokines, normally excluded from neural tissue, gain access to the brain's microenvironment. Microglia, the brain's resident immune cells, shift into a chronic activation state that damages the very neurons they evolved to protect. Research using cerebrospinal fluid biomarkers has confirmed that blood-brain barrier breakdown is detectable years before amyloid accumulation becomes significant, suggesting it may be an early initiating event in the Alzheimer's cascade rather than a downstream consequence [5].

Blood-brain barrier breakdown is detectable years before amyloid accumulation becomes significant — suggesting it may be an initiating event in Alzheimer's disease, not a downstream consequence.

This mechanistic insight reframes vascular health not as a secondary concern for cognitive aging but as the foundational substrate. Preserving the structural integrity of the blood-brain barrier across decades of midlife is, in a very literal sense, preserving the brain's ability to protect itself. The next factor, physical activity, operates directly on this substrate.

Factor Two: Physical Activity and Cardiorespiratory Fitness

Few interventions in all of medicine have as broad and consistent a body of evidence for dementia prevention as regular aerobic exercise. The magnitude of protection is substantial: physically active individuals have a 30 to 45% lower risk of developing dementia compared to sedentary individuals, an effect size that rivals or exceeds most pharmacological interventions studied in prevention trials [1]. More strikingly, the benefit is not merely correlational. It is mechanistic, specific, and dose-responsive.

The most studied mechanism is neurogenesis in the hippocampus. For most of the 20th century, neuroscience held that the adult brain does not generate new neurons. That consensus was overturned in the late 1990s with the discovery of adult hippocampal neurogenesis, and subsequent research has established that aerobic exercise is one of the most potent stimulators of this process [6]. The key mediator is brain-derived neurotrophic factor (BDNF), a protein that acts like fertilizer for neurons: it promotes their survival, their branching, and the strength of their synaptic connections. Blood and cerebrospinal fluid levels of BDNF rise reliably in response to sustained aerobic exercise, and in human trials, 12 months of moderate-intensity walking was sufficient to increase hippocampal volume by approximately 2%, effectively reversing one to two years of age-related hippocampal atrophy [6].

Exercise also acts on the brain's waste clearance system, known as the glymphatic system. This network of fluid channels surrounding cerebral blood vessels drives cerebrospinal fluid through brain tissue, flushing out metabolic waste including amyloid-beta and tau aggregates. Glymphatic flow depends on the pulsatile expansion and contraction of blood vessels, a hydraulic process driven partly by cardiovascular fitness. Higher cardiorespiratory fitness is associated with more efficient glymphatic clearance, while sedentary behavior and poor vascular tone are associated with glymphatic stagnation and accelerated amyloid accumulation [7].

Beyond aerobic exercise, resistance training contributes distinct neuroprotective effects. Skeletal muscle, increasingly recognized as an endocrine organ rather than simply a locomotive structure, secretes a family of proteins called myokines during contraction. Irisin, one of the best-characterized myokines, crosses the blood-brain barrier and directly upregulates BDNF expression in hippocampal neurons. Insulin-like growth factor 1 (IGF-1), released by muscle and liver in response to resistance exercise, promotes neuronal survival and suppresses neuroinflammation. Studies in older adults with mild cognitive impairment have found that six months of progressive resistance training improves executive function and episodic memory, with effects correlated with gains in muscle strength rather than simply cardiovascular fitness [8].

The practical implication is that optimal exercise for dementia prevention is not a single modality but a combination: sustained aerobic exercise for BDNF, glymphatic clearance, and cerebrovascular health, combined with progressive resistance training for myokine signaling, metabolic health, and preservation of muscle mass. VO2 max, the maximal rate of oxygen consumption during exercise, has emerged as one of the strongest single predictors of cognitive longevity, more predictive of late-life cognitive decline than any individual biomarker currently available. Each unit increase in VO2 max is associated with measurably better executive function and reduced dementia risk, independent of other factors [9].

This is not simply about achieving an athletic physique. It is about maintaining the biological machinery, cardiovascular efficiency, muscle mass, metabolic flexibility, that keeps the brain's internal environment clean, nourished, and structurally robust. For individuals seeking to structure this more formally, tools that support muscle preservation and metabolic health become relevant adjuncts to an exercise program. The Creatine + Protein Bundle addresses the nutritional substrate for resistance training adaptations, and programs like Healthspan's Longevity Optimization provide the clinical framework within which exercise goals, metabolic targets, and biomarker tracking integrate coherently.

Factor Three: Sleep Quality and Glymphatic Clearance

Sleep earned its place in the 2024 Lancet Commission as one of the newly identified modifiable risk factors, a recognition that reflects a decade of mechanistic research that has fundamentally changed neuroscience's understanding of what the sleeping brain is actually doing [1]. The conventional wisdom was that sleep was a period of neural rest, a consolidation phase for memories formed during the day. That picture is incomplete. Sleep is also when the brain runs its most intensive cleaning cycle, and the efficiency of that cycle may be the single most important determinant of amyloid and tau accumulation over decades.

The mechanism is the glymphatic system, introduced in the context of exercise above. During slow-wave sleep, the deepest phase of non-REM sleep, interstitial space between brain cells expands by approximately 60%, allowing cerebrospinal fluid to flush through neural tissue at dramatically higher volumes than during waking [10]. This hydraulic surge sweeps amyloid-beta and tau out of the brain's extracellular space, transporting them into the cerebrospinal fluid and ultimately into the lymphatic drainage system. A single night of sleep deprivation produces a measurable 25 to 30% increase in amyloid-beta concentration in the cerebrospinal fluid of healthy adults, as the overnight clearance cycle fails to run [11]. Extend that deprivation over years or decades of chronically poor sleep, and the cumulative amyloid load becomes a primary driver of Alzheimer's pathology.

A single night of sleep deprivation produces a 25 to 30% increase in cerebrospinal fluid amyloid-beta concentration — a window into what decades of poor sleep do to the aging brain.

Epidemiological evidence corroborates the mechanistic data. Adults who consistently sleep fewer than six hours per night in midlife have been found to have a 30% higher risk of developing dementia compared to those sleeping six to eight hours, with the association persisting after controlling for depression, cardiovascular risk, and socioeconomic factors [12]. Sleep-disordered breathing, primarily obstructive sleep apnea, is a particularly high-impact risk factor: the intermittent hypoxia (periodic oxygen desaturation) that characterizes apnea episodes activates inflammatory cascades in the brain, disrupts slow-wave sleep architecture, and suppresses glymphatic flow during exactly the periods when clearance should be most active. Estimates suggest that obstructive sleep apnea may be present in 30 to 50% of older adults with mild cognitive impairment, though it frequently goes undiagnosed [13].

The sleep-dementia relationship is bidirectional and self-reinforcing in a clinically important way. Poor sleep accelerates amyloid accumulation. Accumulating amyloid disrupts the neural circuits that regulate sleep, particularly those generating slow-wave oscillations in the prefrontal cortex. The result is a feedback loop in which poor sleep quality worsens the pathological burden that in turn worsens sleep quality. Identifying and treating sleep disruption before cognitive symptoms appear is therefore not merely about subjective wellbeing. It is about interrupting a cycle that, if left unchecked, progressively depletes cognitive reserve.

Circadian rhythm integrity is inseparable from sleep quality in this context. Light exposure, meal timing, exercise timing, and core body temperature cycles all entrain the brain's master clock, the suprachiasmatic nucleus. When circadian rhythms are disrupted, as occurs with shift work, irregular sleep schedules, or excessive evening light exposure, slow-wave sleep architecture deteriorates even when total sleep duration is preserved. This circadian dimension of sleep health explains why sleep hygiene, the structured optimization of sleep timing, environment, and behavioral cues, is not a soft lifestyle recommendation but a neurobiological intervention with measurable effects on the brain's nightly maintenance program.

How the Three Factors Converge: The Inflammation-Clearance Axis

Cardiometabolic dysfunction, physical inactivity, and poor sleep do not operate through independent pathways that happen to converge on dementia risk. They form a mutually reinforcing biological system in which each factor amplifies the others, and in which the common downstream currency is neuroinflammation and impaired amyloid clearance.

Elevated blood pressure degrades the blood-brain barrier, allowing peripheral inflammatory signals into neural tissue. Physical inactivity allows visceral adipose tissue to expand, increasing the secretion of pro-inflammatory adipokines that cross that compromised barrier. Poor sleep fails to clear the amyloid and tau aggregates that microglial cells would otherwise detect and remove, triggering instead a chronic, low-grade microglial activation that generates further inflammatory cytokines, which in turn further disrupt sleep architecture. The system is not three parallel risk factors. It is a triangle of biological feedback in which weakness in any corner destabilizes the others.

This convergence has an important clinical implication: interventions that address multiple corners of this triangle simultaneously produce synergistic benefits that exceed what any single intervention achieves alone. The most striking evidence for this comes from the FINGER trial (Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability), a landmark multi-domain randomized controlled trial in which participants at elevated dementia risk received simultaneous interventions targeting diet, exercise, cognitive training, and vascular risk management. After two years, the intervention group showed 25% better cognitive performance than the control group overall, with specific improvements of 83% in executive function and 150% in processing speed [14]. The whole was dramatically greater than the sum of its parts.

The FINGER results have since been replicated in modified forms across multiple countries through the World Wide FINGERS Network, consistently showing that multi-domain lifestyle interventions outperform single-domain approaches and that the effect is largest in individuals with higher baseline cardiovascular risk [1]. This is the clearest evidence that dementia prevention is not about optimizing one variable but about maintaining the integrity of an interconnected biological system.

Emerging Evidence and the Limits of Current Knowledge

The strength of the evidence for these three factors varies by mechanism. The epidemiological associations are robust, replicated across dozens of cohort studies and multiple national populations. The mechanistic understanding, particularly for sleep and glymphatic clearance, has advanced rapidly but remains largely derived from animal models, with human translation still ongoing. Randomized controlled trial evidence for dementia incidence endpoints is limited, partly because these trials require decades of follow-up and face the practical impossibility of controlling all lifestyle variables in a free-living population.

Several important caveats apply to the metabolic risk factor cluster. The protective effect of blood pressure reduction on cognitive outcomes, while supported by SPRINT MIND, was not fully corroborated in all subgroup analyses, and the optimal blood pressure target for cognitive protection remains debated [3]. The insulin resistance and dementia relationship is well-established epidemiologically, but whether pharmacological glycemic control translates to cognitive protection in the absence of lifestyle change is less clear. Metformin, for instance, has shown mixed results in cognitive outcome studies despite its strong cardiometabolic profile and mTOR-modulating properties. Its role specifically in dementia prevention remains an active area of investigation.

Similarly, while the sleep deprivation and amyloid accumulation data are striking, most human studies showing single-night effects are conducted in highly controlled laboratory settings with acute total sleep deprivation, which does not perfectly model the chronic partial sleep restriction that characterizes real-world poor sleep in millions of midlife adults. Whether treating obstructive sleep apnea with continuous positive airway pressure (CPAP) reduces dementia incidence remains under investigation, with observational studies suggesting benefit but randomized trial evidence still limited [13].

These limitations do not undermine the case for action. They define its shape. The evidence is strong enough to justify aggressive management of cardiometabolic risk, structured exercise across multiple modalities, and serious attention to sleep quality as a first-line dementia prevention strategy. The uncertainty lies primarily in the margins: in precise dose-response curves, in the relative timing and sequencing of interventions, and in whether emerging pharmacological adjuncts can safely amplify the gains achieved through lifestyle.

Translating Evidence Into Practice: A Longevity Framework

The practical architecture of dementia prevention looks different from the architecture of acute disease treatment. There is no single pill, no definitive procedure, no biomarker threshold crossed that signals safety. Instead, it is a continuous, decades-long maintenance of four interlocking biological conditions: clean and responsive cerebral vasculature, efficient metabolic processing, adequate muscle mass and cardiovascular fitness, and nightly glymphatic clearance through restorative slow-wave sleep.

Cardiometabolic risk management begins with knowing the numbers. Blood pressure, fasting glucose, HbA1c, fasting insulin, and lipid particle counts (not just total cholesterol) provide a detailed picture of the vascular environment the brain is inhabiting. For individuals showing early metabolic drift, continuous glucose monitoring offers a granular view of glycemic variability that static blood tests miss. SGLT2 inhibitors, which reduce glucose reabsorption by the kidney and have demonstrated cardiovascular protective effects, are under investigation for their potential neuroprotective properties, with Healthspan's SGLT2 Protocol offering a clinically supervised framework for their use in appropriate candidates.

Exercise prescription for cognitive longevity should include at minimum 150 minutes per week of moderate-intensity aerobic exercise, with progressive resistance training two to three times per week. VO2 max testing provides an objective baseline against which progress can be measured, and emerging evidence suggests that even relatively modest gains in VO2 max over one to two years produce measurable cognitive benefits in midlife adults. Maintaining skeletal muscle mass through adequate dietary protein is an underappreciated component of this picture: sarcopenia, the age-related loss of muscle, reduces myokine output and metabolic flexibility, two properties essential to the exercise-brain axis.

Sleep optimization, the third pillar, begins with an honest assessment of sleep architecture rather than simply duration. Home sleep testing can identify obstructive sleep apnea in individuals who may have never recognized their symptoms. Sleep diary analysis can reveal circadian misalignment. Cognitive behavioral therapy for insomnia (CBT-I) remains the most evidence-supported intervention for chronic insomnia, superior to pharmacological sleep aids in long-term outcomes. For individuals with complex sleep disruption alongside hormonal changes, particularly in perimenopause where progesterone decline directly reduces GABAergic sleep-promoting activity, hormone support may play a meaningful adjunct role, an area addressed by programs like Women's Hormone Health.

What unites all three pillars is the concept of biological environment. The brain does not exist in isolation. It inhabits a vascular, metabolic, and hormonal milieu that is shaped hour by hour, year by year, by how the body is maintained. Dementia is not an inevitable destination. It is, for many people, the cumulative consequence of a biological environment that has drifted too far from the conditions under which the brain thrives. Restoring and maintaining those conditions is both the science and the practice of dementia-free longevity.

Conclusion: The Decades That Decide

The search for a pharmaceutical cure for Alzheimer's disease has consumed billions of research dollars and produced, after decades of effort, a handful of modestly effective amyloid-clearing antibodies that slow decline in early-stage disease but do not reverse it. The absence of a cure is not a reason for nihilism. It is an argument for earlier and broader action on the factors that determine whether pathology accumulates to the threshold of disease in the first place.

The 45% of dementia cases attributable to modifiable risk factors is not a ceiling; it is a floor. It represents the fraction explainable by the factors researchers have so far measured systematically. The true preventable fraction is likely higher. And of those factors, three stand above the rest in their combination of mechanistic clarity, epidemiological weight, and clinical actionability: the cardiometabolic environment that the brain's vasculature depends on for oxygen and glucose delivery, the physical activity that produces BDNF, myokines, and glymphatic hydraulics, and the sleep quality that runs the brain's nightly maintenance cycle. These are not soft suggestions. They are the biological prerequisites for a brain that remains capable, curious, and connected into the eighth and ninth decades of life. The decades that decide are not 70 and 80. They are 40, 50, and 60. The choices made there write the neurology of old age.