AFL men take more head knocks per match than AFLW women. Per hour, the gap nearly vanishes

For most of the time Australian football has been worrying about head injuries, the underlying exposure has been estimated rather than measured. A study published in Sports Medicine – Open in February reports what instrumented mouthguards actually recorded across a season of elite men's and women's competition [s1].

What the study did

Instrumented mouthguard data were collected from Australian Football League (AFL) and AFL Women's (AFLW) players during 2023, using the HIT-IQ Nexus Gen III device, targeting 500 player-matches per sex [s1]. Crucially, match footage was synchronised with the mouthguard data so that every recorded event could be video-verified and the match play coded [s1].

That verification step is what separates this from raw sensor output. Mouthguard accelerometers register events that are not head impacts at all — a guard being inserted, chewed, or spat out. Without video confirmation, a dataset of "head acceleration events" is partly a dataset of mouthguard handling.

Three hundred and eleven players contributed data: 126 men, with a median of 3 matches per player (IQR 2–6), and 185 women, median 3 matches (IQR 1–4) [s1].

The headline comparison, and why the framing changes it

Men had a higher mean incidence of head acceleration events at or above 8 g per match than women — 5.7 (95% CI 5.0–6.4) versus 3.0 (95% CI 2.8–3.3) [s1].

Adjusted for time, the difference largely disappears: 3.6 per hour (95% CI 3.2–4.1) in men versus 3.1 per hour (95% CI 2.9–3.4) in women [s1]. AFLW matches are shorter than AFL matches, and a per-match comparison substantially reflects that. The per-hour figures are the ones that describe the game rather than its scheduling.

On magnitude, men recorded higher maximum single-impact peak linear acceleration per match — median 31 g versus 22 g in women (p < 0.001) — and higher maximum peak rotational acceleration, median 3.0 krad/s² versus 2.4 krad/s² (p < 0.001) [s1].

The inversion

The result that does not follow the pattern is what happens per impact rather than per match. Women experienced greater peak rotational acceleration per impact — median 1.4 krad/s² versus 1.2 krad/s² in men (p < 0.001) — while peak linear acceleration per impact was statistically similar (men median 14 g, women median 13 g; p = 0.330) [s1].

Rotational acceleration is the parameter most closely associated with brain injury mechanisms in biomechanical models, because it produces the shear strain that linear acceleration alone does not. A finding that the typical impact in the women's competition carries more rotational acceleration, while the extremes in the men's competition are higher, is not a contradiction. It describes two different distributions: men's exposure is characterised by a longer upper tail, women's by a higher central value.

The study does not establish why, and stating a mechanism would go beyond what it reports. What it does establish is that assuming women's exposure is simply a scaled-down version of men's is not supported by the data.

Where the impacts come from

The leading match events producing head acceleration events differed by competition: marking contests in men, at 1.0 per hour, and contested ball situations in women, at 0.8 per hour [s1].

By playing position, incidence and magnitude were generally comparable — with one clear exception. Men's ruckmen recorded a significantly higher incidence of head acceleration events, at 6.5 per hour, than forwards (3.6/hour), midfielders (3.2/hour) or defenders (2.7/hour) [s1].

A ruckman at 6.5 events per hour is absorbing roughly two and a half times the exposure of a defender in the same match. That is a position-specific occupational exposure, and it is the kind of finding that translates most directly into a rule or training change — the ruck contest is a discrete, coachable, legislatively adjustable part of the game.

What this does not show

The study measures exposure, not outcome. It does not report concussion diagnoses, symptom duration, or any longer-term neurological measure, and it cannot say what number of head acceleration events at what magnitude produces harm. The authors frame the findings as providing insight into impact exposure patterns that may inform targeted skill development and policy adjustments to enhance player safety [s1] — an explicitly forward-looking claim, not an established link to injury.

The sample is also thinner than the per-player totals suggest: a median of three matches per player means most participants contributed a small window of their season, and a player's individual exposure over a full year is not what was measured.

What to watch

The obvious next question is whether the ruck differential and the women's per-impact rotational finding replicate across further seasons, and whether either can be attached to concussion incidence rather than exposure alone. Instrumented mouthguards are now producing this kind of data across several collision sports; the value comes when exposure datasets and injury datasets are joined.

Sources

• [s1] Head Acceleration Events Measured by Instrumented Mouthguards in Elite Australian Football, Sports Medicine – Open, published online 16 February 2026. https://doi.org/10.1186/s40798-026-00982-6