The Science of Attention — How Creatine Supports Focus
You open the document you were meant to finish yesterday. You read the first line. You read it again. Somewhere between the second and third attempt, your brain has wandered off into a half-remembered conversation from last week and a sudden urge to check whether the parcel was delivered. You pull yourself back. Ten minutes later, the same thing happens.
You are not broken. You are also not uniquely bad at this. What you are running into is the basic architecture of human attention — a system that was never designed to hold a single fixed beam on a boring document for eight hours. It was designed to move.
The question is not why your focus drifts. The question is why anyone ever believed focus was a binary state — on or off, present or absent — rather than a dimension that shifts across the day, across tasks, and across the lifespan of a human brain.
Why attention is a spectrum, not a switch
For most of the twentieth century, attention was treated as a lightbulb. Either you had it or you did not. You were "paying attention" or you were "distracted." Teachers wrote the word "focus" in report card margins as if it were a moral trait.
The modern research tells a completely different story. Gary Aston-Jones and Jonathan Cohen's 2005 paper in the Annual Review of Neuroscience proposed what is now the dominant framework: the locus coeruleus–norepinephrine system operates in two modes. A phasic mode supports narrow, task-focused attention. A tonic mode supports broader, more exploratory scanning of the environment. Both are useful. Neither is the default. Your brain slides between them depending on context, reward, novelty, and fatigue.
This is why Amy Arnsten's work at Yale keeps returning to the same diagram: an inverted U. Too little norepinephrine in the prefrontal cortex and you are drifting. Too much and you are wired, anxious, unable to sustain a thought. Somewhere in the middle is the window where sustained attention is actually possible. That window is not a personality trait. It is a physiological state.
Russell Barkley and the dimensional model researchers have been making an adjacent point for years. ADHD, in their view, is not a qualitatively different condition from ordinary distraction. It is the extreme lower tail of a continuous distribution of executive regulation. A 2015 paper in the journal ADHD Attention Deficit and Hyperactivity Disorders argued directly that ADHD symptomatology is best conceptualised as a spectrum, supported by twin studies, brain imaging, and molecular genetics.
If you have read our piece on adult ADHD and what the signs actually look like, this framing will feel familiar. Attention is a dimension everyone sits on. Some people sit closer to one end. Most of us fluctuate across it depending on sleep, stress, blood sugar, and what we are trying to focus on.
What is actually happening when focus fails
Hold a phone number in your head while someone gives you directions. Read a sentence and keep track of the argument five sentences earlier. Write an email while filtering out a conversation at the next desk. Each of these asks the prefrontal cortex to keep information online while suppressing competing signals.
That work is expensive. The brain is roughly 2% of your body weight and consumes about 20% of your resting energy. Neurons maintaining ion gradients, firing action potentials, and recycling neurotransmitters at synapses burn through ATP at a rate almost nothing else in the body can sustain.
When that energy supply cannot keep pace with demand, attention degrades first. This is why sleep deprivation wrecks focus before it wrecks motor coordination. It is why the last hour of an intense cognitive task feels qualitatively worse than the first. It is why boredom is so exhausting — low-stimulation tasks require more effortful top-down control, not less, because there is no dopaminergic engagement to carry the brain along on its own momentum.
The bottleneck is not willpower. The bottleneck is bioenergetics.
Where creatine enters the conversation
Most people know creatine as a gym supplement. The brain research is older and quieter, but it is not new. The fundamental insight is that creatine — stored in cells as phosphocreatine — is the fastest ATP recycling system the body has. Every time a neuron fires, it burns ATP. Phosphocreatine donates a phosphate group back to ADP, regenerating ATP almost instantly, buying the cell time before slower mitochondrial pathways catch up.
The brain has its own creatine metabolism. Neurons express the creatine transporter SLC6A8, produce some creatine endogenously, and rely on a circulating pool for the rest. When the brain is under metabolic stress — sleep deprivation, prolonged mental effort, hypoxia, aging — phosphocreatine reserves fall. When reserves fall, cognitive performance follows.
The first direct evidence that this mattered for healthy adults came from Caroline Rae and colleagues at the University of Sydney. Their 2003 double-blind, placebo-controlled crossover trial in Proceedings of the Royal Society B gave 45 vegetarians 5 grams of creatine per day for six weeks. On Raven's Advanced Progressive Matrices and backward digit span — two of the cleanest tests of working memory and reasoning — the creatine group improved significantly (p < 0.0001). Digit span, the classic working memory measure, rose from roughly 7 items to 8.5. A single supplement moved a cognitive metric textbooks describe as largely fixed.
Airi Watanabe, Nobumasa Kato, and Tadafumi Kato at Keio University published an equally telling study in Neuroscience Research in 2002. Subjects given 8 grams of creatine per day for five days reported significantly less mental fatigue during a repetitive arithmetic task. Near-infrared spectroscopy showed reduced cerebral oxygenated haemoglobin, which the authors interpreted as more efficient oxygen utilisation. The brain was doing the same job with less metabolic strain.
What the 2024 sleep deprivation trial actually found
The most talked-about recent study comes from Ali Gordji-Nejad's team at Forschungszentrum Jülich, published in Scientific Reports in February 2024. Fifteen healthy adults pulled an overnight sleep deprivation protocol. Half received a single high dose of creatine monohydrate — 0.35 grams per kilogram of body weight, which for a 70 kg adult works out to about 25 grams — and half received placebo.
The results were striking. The creatine group showed measurable improvements in processing speed and short-term memory compared to placebo. The effect lasted up to nine hours and peaked around four hours after ingestion. Magnetic resonance spectroscopy confirmed that brain phosphocreatine levels actually rose during the sleep deprivation, which runs against what happens normally — a sleep-deprived brain sees phosphocreatine fall.
"Creatine supplementation is a promising intervention for a broad range of situations in which cognitive function may be transiently impaired, including mental fatigue, sleep deprivation, and potentially cognitive stress." — Ali Gordji-Nejad, Forschungszentrum Jülich
The dose in that trial is not a recommendation for general use. At 25 grams in a single hit, it is far above what any clinician would suggest for a long-term routine. But the mechanism it reveals is important: when the brain is under metabolic load, raising phosphocreatine reserves is a real intervention with a measurable cognitive payoff.
What the meta-analyses say and where they disagree
The enthusiasm should be tempered by the messiness of the broader literature. A 2024 systematic review and meta-analysis published in Frontiers in Nutrition pulled together 16 randomised controlled trials with 492 participants. It concluded that creatine supplementation produced significant improvements in memory and in attention time (how quickly participants completed attention-demanding tasks). Processing speed showed a signal. Overall executive function did not move reliably. A 2025 corrigendum walked back some of the attention-specific claims, which is a fair reminder that this is a live research area, not settled science.
Konstantinos Prokopidis's earlier meta-analysis found that creatine reliably improved memory in older adults aged 66 to 76, but not in younger adults who already have ample brain creatine reserves. This pattern — bigger effects when the brain is under stress or starting from a deficit — repeats across the literature.
David Benton and Rachel Donohoe's 2011 trial in the British Journal of Nutrition sharpened the story further. They gave 20 grams per day for five days to 128 young women, split between vegetarians and omnivores. Memory improved specifically in the vegetarians. Reaction time variability — one of the strongest behavioural markers of attentional instability — dropped in both groups. Less variability means a steadier attentional state, closer to the middle of that inverted-U window.
The emerging consensus is not that creatine is a nootropic miracle. It is that creatine helps most when the brain is stressed, depleted, or aging — and that the most consistent effects land on the cognitive processes that depend most on rapid ATP recycling: working memory, processing speed, and the steadiness of attention over time.
Why this maps onto the attention and memory dimensions
The CognitionType framework describes seven cognitive dimensions. Two of them are where creatine's effects concentrate, and the mapping is not coincidental.
Attention and rhythm is the regulatory dimension — how steadily your brain holds a focus state, how smoothly it transitions between tasks, how well it recovers from interruption. The locus coeruleus–norepinephrine system lives at the heart of this dimension, and it is an expensive system to run. When energy is tight, tonic firing patterns become noisy, phasic responses weaken, and attention fragments. Every study that shows creatine reducing reaction time variability or subjective mental fatigue is reporting on this dimension. The brain is not becoming more attentive in some mystical sense. It is becoming more able to sustain the metabolic cost of attention.
Memory and sequencing — working memory — is the other dimension that keeps showing up in the creatine data. Backward digit span, paired-associates learning, Raven's matrices: these are all working memory tasks, and they all depend on sustained prefrontal activity holding information online in the face of interference. Working memory is metabolically expensive, limited in capacity, and the first system to buckle under fatigue. It is also the system Rae's original 2003 trial targeted, and it is the system where effect sizes are most consistent across studies.
A third dimension is worth naming in passing. Emotional regulation interacts with both. Creatine has accumulating evidence in treatment-resistant depression, particularly in women — a 2012 trial by Perry Renshaw's group at Utah showed augmentation effects when creatine was added to SSRI therapy. Mood and attention are not separate rivers. When emotional regulation is strained, attentional regulation takes the hit. If you have read about emotional dysregulation and how feelings can move faster than thoughts, the overlap is exactly this energetic substrate.
What 'enough' creatine looks like
The standard protocol used in most brain-focused studies is 5 grams of creatine monohydrate per day, taken for at least four weeks before effects stabilise. This is the same dose used in sports performance research, and it is what the European Food Safety Authority and most regulatory bodies treat as the reference dose.
Brain creatine stores rise more slowly than muscle stores. Rae's original trial ran for six weeks. More recent imaging studies suggest that eight weeks at 5 grams per day produces a roughly 5 to 10 percent increase in brain creatine levels in healthy adults. People with lower baseline levels — vegetarians, older adults, women — see larger relative increases. People who already eat substantial amounts of red meat see smaller ones, though still meaningful.
Doses above 10 grams per day produce stronger effects in some studies, particularly under metabolic stress. The Jülich sleep deprivation trial used a single large bolus for acute effect. These higher doses are not without cost — creatine is cleared through the kidneys, and anyone with existing renal issues should not be self-medicating with it.
Vegetarians and vegans see the largest effects because their dietary creatine intake is minimal. The brain compensates partially through endogenous synthesis, but the gap between synthesis and optimal supply is where exogenous creatine can move the needle.
Where this fits into a broader approach to focus
Creatine is not the only lever, and it is almost never the first one that matters. Sleep is. Cardiovascular fitness is — we covered the mechanism in detail in our piece on how exercise changes your cognitive profile. Protein sufficiency, iron status, and omega-3 levels each contribute to the raw materials a prefrontal cortex needs to do its work.
Creatine is an upstream metabolic support, not a substitute for any of those. What it does well is raise the ceiling on how long the brain can sustain attention under load. If your focus collapses after two hours of concentrated work, creatine is a reasonable thing to investigate. If your focus collapses five minutes into a boring meeting, the problem is probably not bioenergetic — it is motivational, and no supplement fixes that.
The deeper point is that treating "focus" as a single thing obscures what is actually going on. Attention has structure. It has a dimensional shape that varies across people and across days. The lever that helps you depends on which part of that shape is most constrained.
CognitionType is a useful tool for understanding your own cognitive profile across seven dimensions, including attention and rhythm, memory and sequencing, and emotional regulation. It is designed to help you see which parts of your cognitive architecture are working fluently and which are running hot — and to match interventions like nutrition, movement, and daily structure to the dimensions that actually need them. It is an informational profile, not a clinical diagnosis, and it works best alongside professional evaluation when that is warranted.
The simple version
Attention is not a switch. It is a spectrum governed by the balance of neural gain, the metabolic cost of sustaining prefrontal activity, and the moment-to-moment interplay between norepinephrine and dopamine. Everyone sits somewhere on that spectrum. Where you sit shifts across the day, across the lifespan, and under the pressure of stress, sleep loss, and ageing.
Creatine is one of the few nutritional interventions with direct, mechanistic, replicated evidence for supporting the energetic substrate that attention depends on. It will not fix a boring meeting or a job that does not engage your interest-based nervous system. It will, in the conditions where the brain is running short on ATP, make sustained cognitive work slightly less costly and slightly less fatiguing.
That is a narrower claim than the supplement industry likes to make. It is also, for once, a claim the research actually supports.
CognitionType is an informational assessment, not a clinical diagnosis. If you suspect ADHD, a mood disorder, or another clinical condition affecting your attention, we encourage you to seek formal evaluation from a qualified clinician. A cognitive profile is a complement to clinical assessment, not a replacement.