Does the ageing brain become ‘noisier’?

By Fenying Zang

Older people often respond less consistently when making decisions. But does a similar behavioural pattern occur in other species, and what happens to the activity of individual neurons as the brain ages? Researchers at Leiden University, working with collaborators from the International Brain Laboratory (IBL) and University of California, Los Angeles, analysed the activity of more than 18,000 neurons in mice spanning early adulthood to old age. They found that older mice performed slightly worse and had more variable response times in a visual decision-making task. At the neural level, ageing was associated with higher firing rates across many recorded regions and greater variability after a visual stimulus appeared—but the patterns depended on the brain region and the time at which neural activity was measured.

Image adapted from Zang et al. (2026), CC BY 4.0. Task schematic adapted from International Brain Laboratory et al. (2025), CC BY 4.0.

What does a ‘noisier’ brain mean?

If the same image is shown several times, the brain does not respond in exactly the same way each time. At the level of an individual neuron, this can be seen in the number of electrical impulses, or spikes, it produces: the number may vary from one presentation to the next. Such trial-to-trial variability is a normal feature of neural activity and is not necessarily meaningless or harmful “noise”. In this study, the researchers examined one specific form of it: variation in the number of spikes produced by individual neurons across repeated trials.

This form of neural variability is one component of the broader idea of neural “noise”. For decades, researchers have proposed that the ageing brain becomes “noisier” and that this may contribute to age-related changes in perception and decision-making. Testing this idea directly has been difficult. Previous single-neuron studies have generally included relatively small samples, concentrated on one or two brain regions, and often recorded from animals that were anaesthetised or passively viewing stimuli. They also rarely examined how neural variability changes from before to after sensory information arrives while an animal is actively making a decision.

Comparing behaviour and neural activity across adulthood

To address these limitations, the study combined previously released IBL recordings with new recordings that extended the dataset to older mice. All data were collected using the same standardised behavioural and neural-recording protocols. The mice saw an image on either the left or right side of a screen and turned a wheel to report its location.

Older mice performed slightly worse overall. Their response times also matched a pattern widely reported in human ageing research: older mice showed more variable response times across trials. Their average response times, however, were not slower.

While the mice performed the task, the researchers recorded neural activity using Neuropixels probes—tiny devices that can simultaneously record from many individual neurons. After standardised quality-control procedures, the resulting dataset included recordings from more than 18,000 neurons across 16 brain regions, including areas involved in visual processing, movement, memory and decision-making.

Ageing alters how neural variability changes after a stimulus

The neural recordings revealed that age did not affect every aspect of neural activity in the same way. The researchers first examined firing rates—the frequency with which neurons produced electrical impulses. When all recorded neurons were considered together, firing rates increased with age both before and after the visual stimulus. This overall increase concealed substantial regional differences. Firing rates were higher in older animals in areas of visual and motor cortex, striatum, midbrain and hippocampus, but lower in the two recorded thalamic areas.

A different pattern emerged when the researchers examined neural variability. Before the stimulus appeared, neural variability did not change systematically with age. After the stimulus appeared, however, neural responses were more variable overall in older mice, with the clearest regional evidence in the recorded midbrain regions and one thalamic area.

Neural variability typically reduces after a stimulus appears, meaning that responses become more consistent across repeated trials than they were beforehand. This phenomenon, known as “variability quenching”, is thought to reflect the brain settling into a more reliable state for processing incoming information. The researchers found that this reduction became weaker in older mice, particularly in visual and motor cortex, the striatum and one thalamic area.

Not simply a ‘noisier’ brain

Taken together, the study provides a qualified answer to the question of whether the ageing brain becomes “noisier”. Ageing was associated with greater single-neuron variability after the stimulus appeared, but not beforehand. The usual reduction in variability following stimulus onset also became weaker with age. At the same time, the direction and strength of the effects differed across brain regions. The ageing brain therefore did not simply become uniformly more variable. 

Although the behavioural and neural differences were observed during the same task, the study did not test whether one explained the other. It therefore does not establish that altered neural variability causes less consistent behaviour. Instead, it provides a large-scale, multi-region survey of age-related changes in single-neuron activity and variability, together with an openly available dataset that can support future investigations of how these neural changes relate to behaviour.

The open-access study, Age-related changes in behavioural and neural variability in a decision-making task, was published in Nature Communications.

Leave a comment