Cognitive Processing Speed: The Neural Clock Speed

The CPU of the Self

In the “Guesswork Era,” we measured brain health through memory tests or subjective assessments of focus and attention. In the 2026 Consensus, we prioritise Cognitive Processing Speed (CPS) — the foundational bandwidth of the nervous system from which all higher cognitive functions depend.

Think of the brain as a distributed computing system. Executive function is the software — the capacity to plan, reason, and decide. CPS is the clock speed of the hardware — the rate at which raw sensory input is converted into processed signal, integrated across neural networks, and expressed as a motor response. Every higher cognitive function — memory retrieval, decision-making, language comprehension — runs on this substrate. When the hardware degrades, all software slows with it.

The mortality signal behind this framing is real. In the UK Biobank cohort (Zaccardi et al., Intelligence, 2017, n=54,019), slower reaction time was independently associated with all-cause mortality after adjustment for fluid intelligence, cardiorespiratory fitness, and lifestyle factors. Processing speed adds independent survival information beyond what physical fitness or general intelligence alone provide.

I. The Mechanism: Myelin, White Matter, and the Inflammaging Brake

CPS is physically determined by the health of the myelin sheath — the lipid-rich insulation coating axons throughout the central and peripheral nervous system. Myelination enables saltatory conduction: the electrical signal jumps between nodes of Ranvier rather than propagating continuously along the entire axon length, achieving conduction velocities up to 100 times greater than unmyelinated fibres.

When myelin integrity is compromised, signal propagation slows and desynchronises. The primary upstream drivers of white matter degradation in midlife are:

• Neuroinflammation: Chronic low-grade systemic inflammation (elevated hs-CRP, IL-6 — Pillar 04 markers) drives microglial activation in the CNS. Activated microglia in a chronic, unresolved state shift from a neuroprotective phenotype to a neurotoxic one — producing pro-inflammatory cytokines that directly damage oligodendrocytes (the cells that produce and maintain myelin). This is the mechanistic bridge between systemic inflammaging and neural processing velocity.

• Vascular White Matter Damage: The brain’s white matter tracts are supplied by small penetrating arteries that are highly sensitive to haemodynamic pulsatility. As established in our Vascular Age Briefing, arterial stiffening transmits high-amplitude pressure waves into the cerebral microvasculature — producing lacunar infarcts, white matter hyperintensities (WMH), and progressive disruption of the long-range connectivity networks that underpin processing speed. WMH burden is one of the strongest neuroimaging predictors of CPS decline in midlife cohorts.

• Glycation of Myelin Proteins: Advanced Glycation End-products (AGEs) — the direct downstream consequence of elevated HbA1c (see HbA1c Briefing) — directly modify myelin basic protein (MBP), impairing its structural stability and the integrity of the myelin sheath. Diabetic and pre-diabetic populations show measurably reduced nerve conduction velocity proportional to HbA1c elevation, with the effect present in peripheral nervous system conduction (a proxy for central white matter health) before clinical neuropathy develops.

• Mitochondrial Dependency: The CNS is the most metabolically demanding tissue in the body. Myelinated axons require continuous ATP supply to maintain the ion gradients that power action potential propagation. Mitochondrial dysfunction — driven by insulin resistance, oxidative stress, and sedentary behaviour — disproportionately impairs the high-energy-demand white matter tracts that mediate long-range neural communication and processing speed.

II. The Forge Range: Testing Context Determines Everything

This is the most important correction the Forge makes to how CPS targets are typically communicated. Reaction time is highly sensitive to the testing paradigm, hardware, and ecological context — and targets must be specified relative to a defined testing method.

The original article listed < 220 ms as the Forge Optimal reaction time for all adults. This requires critical revision:

The UK Biobank reaction time data (n=54,019, Zaccardi et al.) reports a median of 538 ms (IQR 480–613 ms) for complex visual recognition RT (matching symbols on a touchscreen). The MindCrowd study (n=75,666, Talboom et al.) reports simple visual RT averaging 270–300 ms for healthy young adults aged 18–25, declining to approximately 340–380 ms by age 50–60. Sub-220 ms performance represents elite laboratory conditions — roughly the 99th population percentile even in peak-age adults tested under optimal conditions. It is not a realistic optimisation target for the Forge readership.

The < 220 ms figure conflates two distinct measurement categories:

• Simple laboratory RT (single stimulus, immediate response, laboratory-grade hardware): typical range 150–300 ms in healthy adults aged 20–40. Elite athletes may reach 140–180 ms under optimal conditions.
• Ecological / consumer-tool RT (complex visual recognition, consumer device, real-world latency): typical range 300–650 ms across the adult lifespan.

Forge Editorial Note: The table below uses the MindCrowd + UK Biobank normative data as the reference standard. “Forge Optimal” targets the 25th percentile for each age bracket — below-average RT (i.e., faster than average). The Human Benchmark simple RT test (humanbenchmark.com) uses a simple visual paradigm and is the Forge-recommended consumer tool. Percentile interpretation must account for device and testing condition consistency.

Age Group	UK Biobank Median RT	MindCrowd Simple RT Mean	Forge Optimal Target (≤ 25th percentile)
18–30	N/A (UKB ages 40–70)	~270–290 ms	< 240 ms
30–45	~510–540 ms (complex)	~290–320 ms	< 270 ms
45–60	~540–580 ms (complex)	~320–360 ms	< 300 ms
60+	~580–620 ms (complex)	~360–400 ms	< 340 ms

Trajectory is the primary signal. A 50-year-old with a stable 310 ms simple RT over 3 years of testing is performing better than a 50-year-old declining from 280 ms to 340 ms over the same period. The MindCrowd data shows 7 ms/year average decline — any individual maintaining sub-7 ms/year trajectory is outperforming the population aging curve.

Forge Verdict: CPS is the ultimate cross-system audit. A deteriorating reaction time trajectory — independent of any single test result — is a leading signal that one or more upstream systems are under strain: sleep architecture (Deep Sleep %), vascular health (Vascular Age), metabolic integrity (HbA1c), or systemic inflammation (hs-CRP). The RT number is the output; the protocol audit investigates the inputs.

III. The Forge Protocol: Neural Overclocking

Improving CPS is not achieved by practising reaction time tests. The primary levers are the upstream structural determinants of white matter integrity and neural metabolic capacity. Reaction time practice produces modest task-specific adaptation; protecting the biological substrate produces durable, system-wide improvement.

01. Myelin Substrate — The Lipid Foundation

The brain is approximately 60% fat by dry weight, and myelin is predominantly composed of specific lipids — phosphatidylcholine, sphingomyelin, and cholesterol. To maintain structural myelin integrity, the Forge prioritises:

• DHA (Omega-3): The dominant polyunsaturated fatty acid in neuronal membranes. Incorporated into phospholipid bilayers, it modulates membrane fluidity and signal transduction efficiency. Meta-analyses confirm DHA supplementation improves neural processing and reaction time in deficient populations. Target: 2–3g combined EPA/DHA daily, prioritising EPA:DHA ratios of ≥ 1:1 for combined neural and vascular benefit.
• Phosphatidylserine (PS): A phospholipid concentrated in brain cell membranes with a supportive role in neurotransmitter release and neuronal membrane function. Multiple double-blind RCTs confirm PS supplementation attenuates age-related CPS decline. Dose: 100–300 mg/day. Evidence grade: B — effect sizes are modest; most trials are in age-related decline populations, not healthy optimisers.

02. BDNF Induction — The Neuroplasticity Signal

Brain-Derived Neurotrophic Factor (BDNF) is the primary neurotrophin governing neuroplasticity, oligodendrocyte development, and white matter maintenance. BDNF drives the myelination and remyelination processes that directly determine axonal conduction velocity. The two highest-yield BDNF induction protocols:

• High-Intensity Interval Training (HIIT): Produces the largest acute BDNF elevation of any exercise modality — serum BDNF increases of 20–30% have been documented post-HIIT in multiple RCTs. The mechanism involves lactate-mediated BDNF expression and PGC-1α activation in hippocampal and cortical neurons. Recommended: 2 sessions per week, 20–30 minutes per session, maximal or near-maximal effort intervals.
• Sauna Exposure (≥ 80°C, 20 minutes): Heat stress elevates BDNF via heat-shock protein pathways and increased cerebral blood flow. The mechanistic evidence is solid; long-term trial data on CPS specifically is limited but directionally consistent.

03. Tactical Support — Evidence-Graded

• Creatine Monohydrate (3–5g daily): As established in our Muscle Mass Briefing, creatine’s primary mechanism is phosphocreatine buffering for ATP regeneration. In the brain — the most ATP-dependent organ — this translates to improved high-frequency neural firing capacity and reduced cognitive fatigue under load. Multiple RCTs confirm creatine supplementation improves cognitive performance on processing speed tasks, particularly under sleep deprivation or physiological stress. This is one of the better-supported neurological claims for creatine beyond athletic performance.

• Lutein and Zeaxanthin: These macular carotenoids accumulate in the brain’s occipital and frontal regions, not only in retinal tissue. In the COGITATE trial and related observational work (UK Biobank cognitive subgroup), higher brain lutein/zeaxanthin status was positively associated with white matter integrity and processing speed independent of age. The mechanism is likely antioxidant protection of myelin and reduction of neuroinflammatory oxidative load. Evidence grade: C for longevity-specific CPS outcomes; B for white matter integrity association. Food-first approach: egg yolks, dark leafy greens, and avocados are the highest-bioavailability sources.

• L-Theanine + Caffeine Stack: This is the best-evidenced acute cognitive performance stack in the supplement literature. Multiple double-blind RCTs confirm the combination produces improvements in reaction time, sustained attention, and accuracy compared to either compound alone or placebo. The mechanism: caffeine blocks adenosine receptors (reducing neural “noise” from fatigue signalling); L-theanine promotes alpha-wave activity (focused alertness without sympathetic arousal). Dose: 100–200 mg caffeine + 100–200 mg L-theanine (1:1 to 2:1 L-theanine:caffeine ratio). This is an acute performance stack, not a long-term CPS structural intervention.

IV. Actionable Resilience: The Audit

1. Establish a Baseline Using a Consistent Tool, Under Consistent Conditions. Use Human Benchmark (humanbenchmark.com/tests/reactiontime) — a free, browser-based simple visual RT test with a validated, consistent stimulus paradigm. Test 5 trials, take the average (excluding the first trial as a warm-up), and record the result. Test monthly under identical conditions: 30–60 minutes after waking, 1 cup of coffee consumed at the same relative time, same device, same testing environment. The tool and conditions must remain constant for serial data to be comparable.

2. Track 3-Month Rolling Average, Not Individual Results. Single RT readings vary by ±20–40 ms due to attention fluctuation, caffeine state, and testing variability. A 3-month rolling average is the signal; any individual result is noise. A persistent shift in rolling average of >15 ms in either direction is a meaningful trend.

3. Cross-Reference with HRV. Autonomic state directly modulates neural processing efficiency — sympathetic dominance (low HRV) impairs prefrontal cortical function and slows processing speed measurably. The HRV Briefing established that a 20% drop below rolling HRV baseline predicts performance impairment. On days when HRV is suppressed, expect 15–30 ms RT inflation — do not treat this as structural cognitive decline; treat it as an operational state indicator.

4. The Grip-Speed Correlation — Applied Correctly. The original article stated: “If your grip is strong but your speed is low, the issue is likely neural noise rather than structural decay.” This framing is directionally interesting but not established at the level of a clinical decision rule. The accurate application: grip strength and CPS are correlated because they share upstream structural determinants (white matter integrity, mitochondrial capacity, vascular health, inflammation). When grip and CPS are both declining simultaneously, the upstream audit is warranted. When they diverge, the divergence itself is the data point — it may indicate a domain-specific issue (e.g., isolated peripheral neuropathy affecting grip but not central processing, or vice versa).

5. A >10% Drop Over 30 Days Is a System Audit Trigger. If your 30-day rolling RT average has worsened by more than 10%, run a cross-system check before attributing it to a primary neural cause: (a) has Deep Sleep % dropped below your baseline? (b) has HRV been suppressed for more than 7 consecutive days? (c) has there been any significant dietary change affecting HbA1c trajectory? CPS is downstream of all three — the RT number is the warning light, not the fault location.

References

• Zaccardi F. et al., Intelligence (2017): “Reaction time, cardiorespiratory fitness and mortality in UK Biobank: An observational study.” n=54,019, UK Biobank. Slower RT independently associated with all-cause mortality after adjustment for fluid intelligence, cardiorespiratory fitness, and lifestyle factors. DOI: 10.1016/j.intell.2017.11.001
• Talboom J.S. et al., npj Aging and Mechanisms of Disease (2021): “Two separate, large cohorts reveal potential modifiers of age-associated variation in visual reaction time performance.” MindCrowd n=75,666 + UK Biobank n=158,249 = 233,000+ combined. Simple visual RT slows ~7 ms/year from age 20; smoking, stroke history, and Alzheimer’s family history as independent modifiers. DOI: 10.1038/s41514-021-00067-6
• Singh-Manoux A. et al., BMJ (2012): “Timing of onset of cognitive decline: results from Whitehall II prospective cohort study.” n=10,308 (Whitehall II). Cognitive decline including processing speed measurable from age 45 — not just in late life. DOI: 10.1136/bmj.e7622
• Trollor J.N. et al., Molecular Psychiatry (2012): “Systemic inflammation is associated with MCI and its subtypes: the Sydney Memory and Ageing Study.” TNF-α and IL-6 independently associated with lower processing speed and white matter lesion load. DOI: 10.1038/mp.2010.143
• Wen W. et al., Neurobiology of Aging (2011): “White matter hyperintensities: longitudinal study on aging and dementia.” White matter hyperintensity burden as the primary MRI correlate of processing speed decline in aging cohorts.
• Rawlings A.M. et al., JAMA Neurology (2017): “Diabetes in Midlife and Cognitive Change Over 20 Years.” HbA1c elevation in midlife independently associated with steeper 20-year cognitive decline including processing speed — Atherosclerosis Risk in Communities Study, n=13,351. DOI: 10.1001/jamaneurol.2017.0862
• Consortium of UK Biobank Cognitive Resilience GWAS, Genes (2022): 330,097 participants; RT as the most age-sensitive cognitive measure in UK Biobank; 13 independent genetic loci for preserved processing speed identified. DOI: 10.3390/genes13010122
• Consensus 14 Metadata: “Cognitive Processing Speed trajectory as Neural Resilience anchor — bidirectional cross-system signal with HRV, Deep Sleep %, Vascular Age (PWV), and HbA1c.”