Low-Intensity Cycling Beat Moderate Exercise in One Small Trial

Easy cycling may help you feel sharper before a mentally demanding task—but one small study does not justify replacing moderate exercise with a leisurely spin. In a 2025 randomized trial, low-intensity cycling produced faster Stroop reaction times and greater parietal-cortex connectivity than moderate cycling or sedentary rest in healthy adults aged 20 to 40. The result runs against the usual assumption that moderate intensity is the cognitive sweet spot. It is also based on a single acute session with roughly 16 completers per group.

That combination makes this an interesting signal, not a training prescription. The study offers a specific question worth testing: might a low-arousal exercise dose improve some attention-demanding tasks more than a moderate dose, at least immediately afterward? It does not show that low-intensity exercise improves long-term cognition, prevents cognitive decline, or beats moderate training for brain health.

A surprising result in a direct comparison

Shen et al. (2025) randomly assigned healthy young adults to one session of low-intensity aerobic cycling, moderate-intensity cycling, or sedentary rest. The study enrolled 58 participants and retained approximately 16 completers per arm. Participants completed a Stroop task, which measures response speed and the ability to manage conflicting information, while researchers used a 63-channel functional near-infrared spectroscopy (fNIRS) system to measure cortical hemodynamics.

The low-intensity group had significantly faster Stroop reaction times and stronger parietal functional connectivity than both the moderate-intensity and rest groups. The moderate group did not significantly outperform rest on the parietal measures. The publication details the direction and statistical significance of these findings, but the available report does not provide a confidence interval or effect estimate for the central between-group result. That omission limits any judgment about the likely size or practical importance of the advantage.

The parietal finding is unusual. Much of the exercise-and-cognition fNIRS literature concentrates on the prefrontal cortex, the area often examined during demanding executive tasks. A 2025 systematic review by Jung et al. focused on prefrontal oxygenation during high-intensity exercise. Shen and colleagues instead point to posterior parietal networks, which contribute to attention and task control. A different brain region and a different task can produce a different apparent intensity optimum.

Low intensity beat moderate intensity on one immediate Stroop-and-fNIRS comparison. It did not establish a universal cognitive ranking of exercise zones.

Why easier exercise could help in the short term

One plausible explanation involves the locus coeruleus-norepinephrine system, which helps regulate alertness. Too little arousal can leave attention sluggish; too much can interfere with demanding cognitive control. Very light exercise may raise arousal without imposing the cardiovascular and perceived stress of harder work.

That account has some experimental support. In Kuwamizu et al. (2023), healthy adults performed very-light exercise while researchers measured fNIRS signals and pupil dynamics. Greater pupil dilation during exercise predicted post-exercise dorsolateral prefrontal activation and a Stroop conflict benefit. Pupil size is an imperfect proxy for locus-coeruleus activity, so this study supports an arousal-based explanation rather than proving the pathway. Still, it gives the low-intensity result a biologically plausible context.

Exercise responses also need not follow a neat straight line. A crossover fNIRS study by Park et al. (2022) tested several submaximal walking intensities during a fine motor-cognitive dual task. Cortical activation changed nonlinearly across intensities. That finding does not replicate Shen's parietal result, but it weakens the idea that every additional increment of effort should yield a predictable increment in cortical engagement.

Moderate and high intensity still have evidence behind them

The new result should not be read as evidence that moderate exercise harms cognition. In a crossover fNIRS study of 27 healthy young adults, Zheng et al. (2022) found that acute moderate aerobic exercise improved working-memory performance alongside cortical activity changes. Working memory is not the same endpoint as Stroop reaction time, and the study did not test low intensity against moderate intensity. It does show that moderate exercise can benefit a different executive function measure in a similar-age population.

Results at higher intensities are also mixed rather than uniformly poor. In young adults, Shao et al. (2023) found that both acute HIIT and Tabata training improved inhibitory control and prefrontal activation relative to a control session. High intensity can therefore support immediate executive performance under some protocols, even if it may challenge prefrontal oxygenation in others.

At the upper end of the intensity range, a 2025 study of college students with internet dependency found moderate aerobic exercise produced better executive-control outcomes than high-intensity exercise. The study did not include a low-intensity condition, but Fan et al. (2025) adds another reason to avoid a one-size-fits-all answer. Its population also differs from Shen's healthy young adults, and it examined prefrontal activity and executive control rather than parietal connectivity and Stroop speed.

Recovery context can matter too. Xu et al. (2025) studied sedentary college students assigned to high-intensity interval exercise plus meditation, high-intensity exercise alone, or control. The combined intervention improved cognitive outcomes and altered prefrontal hemodynamics. Because meditation was part of the tested intervention, the study cannot isolate the effect of high-intensity exercise itself. It does, however, show why exercise intensity alone may not explain every post-exercise cognitive result.

The methodological brakes are real

Shen's trial is distinctive because it directly compared low and moderate cycling while measuring both behavior and brain-related signals. Its small size is the central problem. With approximately 16 completers per arm, chance imbalances can influence results, false-positive findings become more likely, and observed effects often look larger than they would in a larger replication. For illustration, a conventional two-group comparison designed to detect a medium effect size of d = 0.5 at 80% power and a 0.05 alpha level would require roughly 52 participants per group. That calculation is not a result from Shen's trial, but it puts the sample size in perspective.

The trial also measured one exercise bout and one executive-function task. Faster Stroop responses can reflect better conflict processing, but they can also reflect a transient rise in alertness or response speed. Without N-back, Flanker, task-switching, accuracy, or repeated follow-up measures, the study cannot tell us whether low-intensity cycling broadly improves executive function or simply suited this particular task at this particular moment.

fNIRS adds useful physiological information but not a direct readout of neural firing. Aerobic cycling changes scalp blood flow, which can affect fNIRS signals. Emphasizing deoxyhemoglobin may reduce some systemic contamination compared with oxyhemoglobin, but it cannot eliminate it. Motion during cycling can also distort posterior signals, and the available abstract does not allow an independent assessment of artifact correction. Participants and assessors also cannot be fully blinded to exercise intensity, leaving room for expectations to influence reaction time.

Acute focus is not a long-term training plan

The evidence says little about what intensity best supports cognition after weeks or months of training. In a short-term randomized study, Hu et al. (2021) found that two weeks of HIIT in sedentary women changed motor-cortex plasticity measures and improved executive-function outcomes. Those adaptations are not directly comparable with the immediate effects of one easy cycling session.

Likewise, a 2025 systematic review of randomized trials in healthy older adults examined chronic exercise interventions and executive function but did not identify a low-intensity advantage over moderate intensity. Older adults also differ substantially from Shen's 20-to-40-year-old participants. No direct evidence here supports using low-intensity exercise instead of moderate training to preserve cognition in aging, mild cognitive impairment, or neurological disease.

What to do with this finding

If you are a healthy young adult and want a low-friction pre-work or pre-study routine, very light to low-intensity cycling is a reasonable experiment. Its practical appeal is obvious: it may increase alertness without leaving you sweaty, depleted, or late for your meeting. Track whether it improves your own speed and accuracy on the work that matters to you.

Do not infer that moderate exercise is inferior, or that hard exercise makes you cognitively dull. The current evidence points to task-specific, region-specific, and context-dependent effects. Shen et al. add a valuable direct comparison and a plausible parietal-network hypothesis. A larger preregistered trial, with multiple cognitive tasks, carefully verified intensity targets, short-separation fNIRS channels, and repeated sessions, would tell us whether the result survives contact with a real sample size.