When brain development and brain injury collide

During adolescence, the brain is not simply growing. Its circuits are being remodeled. Some connections are strengthened while others are removed. Microglia, the brain’s resident immune cells, participate in this process by engulfing selected synapses, the connections neurons use to communicate. In the adolescent prefrontal cortex, researchers have directly observed microglia participating in this normal process of synaptic pruning in rats (Mallya et al., 2019).

That creates an intriguing possibility. If repetitive mild traumatic brain injury (rmTBI) activates microglia while those same cells are helping remodel developing circuits, could injury disrupt the process?

The evidence is suggestive, but incomplete.

In 2022, Eyolfson and colleagues compared adolescent and adult male and female mice exposed to rmTBI. The results were not uniform. Injured adolescent males and females spent significantly less time investigating a novel object, suggesting impaired working memory. Microglial density decreased in the motor cortex of injured males but not females. Most interestingly, dendritic spine density, an indicator of structures where many excitatory synapses form, decreased in the agranular insular cortex but increased in the motor cortex of adolescent males (Eyolfson et al., 2022).

That last result matters. If rmTBI simply caused excessive pruning everywhere, spine density should not move in opposite directions across different regions. The developing brain appears to respond to injury in a more complicated, circuit-specific way.

Another piece of the story comes from the complement system, a collection of immune proteins that can mark material for removal.

In adult mice, Alawieh and colleagues (2021) found ongoing complement activation three months after TBI. Microglia were engulfing complement-tagged synapses, and this chronic response was associated with progressive cognitive decline. Even when complement inhibition was delayed until two months after injury, it interrupted the degenerative response and reversed cognitive decline.

Here the authors argue that these findings support considering TBI “a chronic rather than an acute disease condition” (Alawieh et al., 2021). That distinction matters. The biological consequences of an injury may continue long after the initial impact.

A 2025 experiment brought the complement question closer to repetitive mild injury. Mallah and colleagues used a model involving 12 mild closed-head impacts and followed the resulting pathology and cognition for up to 21 days after the final injury. They found substantial changes in local immune-cell activity, including microglial activation and increased expression of complement receptors and phagocytosis proteins, the cellular process of engulfing material. Treatment with CR2-Crry, an inhibitor targeted to sites of complement activation, reduced injury-associated changes and protected against cognitive impairment (Mallah et al., 2025).

The pieces seem to fit. Adolescent brains are actively remodeling synapses. rmTBI alters microglia and dendritic spines. Complement can drive microglial synapse removal after brain injury.

But there is a problem.

No experiment has yet directly shown that adolescence makes complement-dependent synaptic pathology after rmTBI different from the same injury in adulthood.

And complement itself may not work the same way everywhere. Salter and colleagues (2026) tested microglial pruning in the mouse hippocampus and found that removing complement component C3 did not significantly impair microglial engulfment of the excitatory synapses they studied. Even during experimentally induced neuroinflammation, pruning persisted without C3.

Importantly, all of this evidence comes from animal models, rats in the Mallya study and mice in the others, so the human adolescent story remains inferred from animal data rather than directly demonstrated.

So complement cannot simply be treated as the brain’s universal “pruning switch.”

A better hypothesis is a developmental collision, where injury-triggered immune signaling intersects with normal, age-, sex-, and region-specific brain remodeling.

Depending on where and when that collision occurs, the result could be excessive synapse removal, failed removal, or compensatory remodeling rather than one uniform loss of connections.

The next experiment is clear. Give adolescent and adult animals the same controlled repetitive mild injuries. Measure complement deposited on synapses, microglial engulfment, synaptic density, and cognition, and then inhibit complement. If blocking complement selectively restores adolescent synaptic and behavioral outcomes, we would finally have evidence for a developmental vulnerability mechanism.

Until then, the vulnerability window remains a compelling hypothesis.

That is not a failure of the idea. It is precisely where the next experiment begins.