The brain ages in layers: the “entry” layer of the sensory cortex thickens, while the deep layers become thinner

A paper published in Nature Neuroscience shows how aging affects the layers of the sensory cortex differently in humans and mice. In older adults, the “entry” layer IV appears thicker and more myelinated, while the deep layers (V–VI) become thinner, despite an overall increase in myelin. In tissue and calcium experiments on mice, sensory neuronal activity increased with age, and the density of PV interneurons, a likely “compensator” for maintaining excitation/inhibition balance, increased. In other words, the cortex ages not uniformly, but in layers.

Background

• What is usually thought about brain aging. They often say "the cortex thins with age" - and this explains everything. But this is an average picture for the entire thickness of the cortex, without taking into account that the cortex is a "layered cake" with different tasks for each layer.
• What remained unclear was whether the cortex ages uniformly, or whether each layer has its own path. Especially in the sensory cortex, where the fourth layer (layer IV) receives input from the thalamus (the “input port”) and deeper layers send commands downstream. Early work hinted at layer-by-layer shifts, but direct, high-resolution human data were scarce.
• Why it’s easier to study this now. 7-T MRI methods with layer-by-layer analysis of structure and function, as well as quantitative myelin maps (qT1, QSM) have emerged. They can be compared with experiments on mice — from two-photon “calcium” imaging of neuronal activity to histology. This “human ↔ mouse” design allows us to check whether aging really occurs in layers, and is not simply “averaged” across the entire cortex.
• Clues from models. In animals, sensory responses often increase with age, and inhibitory interneurons with the protein parvalbumin (PV) are often rewired — these are the “brake” cells that keep the network from “overexciting.” If their density or function changes, the network can compensate for age-related shifts in input signals.

What did they do?

A team from DZNE (Germany), the Universities of Magdeburg and Tübingen and partners compared young and old groups of people using ultra-high-field 7-T MRI: they measured layer thickness, myelin proxy (qT1) and magnetic susceptibility (QSM), as well as functional responses to tactile stimulation of the fingers. In parallel, two-photon calcium imaging was performed in the barrel cortex of mice and post-mortem myelin analyses were performed. This “bilingual” design (human ↔ mouse) allowed us to compare aging patterns at the layer level.

The main findings - in simple words

• Layer IV (the input channel) is larger and more myelinated in older adults, with extended sensory input signals. The deeper layers are thinner, although they also show signs of greater myelination. The normal "average cortical thickness" masks these differential shifts, so layer-specific metrics are more informative.
• The “borders” of finger maps (areas with low myelin between finger representations) are preserved with age—no clear boundaries were found in degradation.
• Mice showed greater sensory neuronal activation and a higher density of PV interneurons (the “brake” cells) with age, which may serve as compensation to keep networks from “running wild.” Cortical myelin in mice showed age-related dynamics, including an increase in adulthood and a decrease in old age (inverted U-curve).

Why is this important?

• Not everything is about "thinning". Yes, the cortex is thinner in older people on average, but this "average" hides the key: different layers change differently. For diagnostics and science, it is more accurate to look at the profile by layers, and not just the overall thickness.
• Neurobiological implications. Layer IV thickening/myelination and increased PV inhibition appear to be an adaptation in mouse models: input signals are longer and wider, and the system adds “brakes” to curb overactivation. This helps explain why some older adults show enhanced sensory responses without overt evidence of loss of inhibition.
• Bridge to the clinic: Layer-specific approaches may shed light on how normal aging differs from diseases where other layers and mechanisms are affected – for example, in Alzheimer’s or multiple sclerosis, other levels and types of myelin/interneurons are more involved.

Details to look out for

• In one dataset, humans had a total hand thickness of ≈2.0 mm in S1, and the difference between ages was about –0.12 mm – but the key point is that it was the deep layers that contributed, while the middle layer thickened.
• The authors found no clear evidence of weakened inhibition in older adults at the BOLD level; instead, in mouse single-neuron recordings, they observed increased inhibitory co-activation and an increase in PV+ cells, consistent with the idea of compensation.
• In press materials, the study is presented as evidence of “layered” aging of the cortex and that the human cortex ages more slowly than previously thought, at least in the somatosensory zone, because some layers retain or even increase structural “resources.”

Authors' comments

Here is what the authors themselves emphasize (based on the meaning of their discussion and conclusions):

• Aging is not a “uniform thinning,” but a layer-by-layer restructuring. They see shifts in different directions: the “entry” layer IV in older people looks thicker and more myelinated, while the deep layers make the main contribution to the overall thinning of the cortex. Therefore, average metrics across the entire thickness of the cortex hide key changes - you need to look “layer by layer.”
• Sensory input is stretched, the network adapts. Thicker/more myelinated layer IV in the elderly is associated with longer sensory inputs; in a mouse model, sensory neuronal activity is enhanced and the proportion of PV interneurons increases, a likely compensation mechanism to maintain excitation/inhibition balance.
• Deep layers are a vulnerable spot in aging. According to their data, it is the deep layers that explain age-related thinning and changes in functional modulation, while the middle layers can show opposite shifts. Hence the conclusion: different layers have different aging trajectories, and they cannot be reduced to one “average curve”.
• Implications for clinical practice and methods. The authors advocate layer-specific optics: such metrics will help to more accurately distinguish normal aging from diseases (where other layers/mechanisms are affected) and to better interpret high-density (7T) MRI — both structural and functional data.
• The strength of the work is the human↔mouse “bridge.” The combination of 7T MRI in humans with calcium imaging and histology in mice yielded a consistent picture across layers. This, according to the authors, increases the reliability of the interpretation of human findings and suggests mechanisms (myelin, PV interneurons) that can be tested further.
• Limitations—and where to dig next. The human study is cross-sectional (not the same participants over time) and focused on the primary somatosensory cortex; longitudinal studies, other cortical areas, and comparisons with clinical groups are needed. It is also important to clarify the extent to which the 1:1 mechanisms in mice are transferable to humans.

In short, their position: the brain ages “layer by layer,” and this is visible both in the structure (myelin, thickness) and in the network’s operation; the “input” and “output” of the cortex change differently, and some of the effects appear to be adaptive. This changes the approach to diagnostics and the study of age-related changes.

Limitations and the next step

The work is cross-sectional (different people, not the same ones over time) and focuses on the primary somatosensory cortex; the mechanism of differences between species (human ↔ mouse) also requires clarification. Longitudinal layer-specific studies are ahead, and testing how this “layered signature” changes in neurodegenerative and demyelinating diseases.