← Back

The skill nobody trains

The 400‑Millisecond Window

You train strength. You train cardio. But the thing that decides whether you catch yourself or hit the floor is measured in milliseconds, and almost nobody trains it.

The timeline of a stumble A timeline of a stumble: the foot catches at zero milliseconds, the brain registers the loss of balance at roughly 100 milliseconds, muscles fire at roughly 250 milliseconds, and the body is either caught or falling by roughly 400 milliseconds. THE ENTIRE DECISION HAPPENS IN HERE Toe catches Brain registers it Muscles fire Caught, or not 0 MS ~100 MS ~250 MS ~400 MS
Schematic. The exact numbers vary by person and by stumble; the point is the scale. A recovery step is not a decision you make. It is a reflex loop that either closes in time or does not.

Your foot clips the edge of a curb. Somewhere in the next half-second, your nervous system has to notice the problem, pick the right muscles, fire them in the right order, and put a foot down where your falling body needs it to be. Get that sequence out fast enough and you swear at the curb and keep walking. Get it out late and you are on the pavement.

That gap, between a stimulus hitting your senses and your body doing something about it, is reaction time. It is trainable, it is measurable, and it declines with age in almost exactly the way muscle mass does: quietly from your twenties, then faster after sixty.

Elite sport has treated it as a trainable quality for decades. Light boards, reflex apps, cued agility work: all standard. Outside of sport it almost never makes it into a training plan, which is strange, because for anyone over fifty it may be the single most consequential physical skill they own.

#1 Falls are the leading cause of injury-related death among US adults 65 and older, according to the CDC. Slow reaction time is a significant contributor.

The decline nobody notices until it costs something

Reaction time does not fail dramatically. It erodes. And the erosion usually shows up first as behaviour, not as an accident.

People whose reactions have slowed start editing their lives. They stop walking on uneven ground. They give up the sport they used to play. They take the elevator, they avoid the icy stretch, they stop carrying things down stairs. None of it feels like a decision. It feels like sense.

The trap is that this self-imposed restriction accelerates the very decline it is meant to protect against. Less varied movement means less balance challenge, less strength, less coordination, and a nervous system that gets fewer chances to practise reacting. The caution compounds the problem. It is the same cascade that turns one fall into lost independence.

How reaction time changes with age A schematic curve showing reaction time getting slower gradually from the twenties and more steeply after sixty. 20s40s 60s70s 80s SLOWER → Age Slope steepens after 60 Decline starts here
Schematic: the shape of the trend, not measured values. Research cited by CNN puts the onset of decline as early as the twenties, with acceleration after sixty.

It is a brain measurement too

Here is the part that makes this a longevity story rather than a balance story. Reaction time is not only a muscular quality. It is a readout of how fast your central nervous system takes information in and turns it into action: processing speed, essentially, with a motor output attached.

Processing speed happens to be among the cognitive abilities most vulnerable to age-related decline, and in some people that decline runs steeper and ends in dementia. Which raises the obvious question: if you train the speed, does the trajectory change?

The best attempt at an answer comes from the ACTIVE trial, an NIH-funded study that enrolled more than 2,800 older adults and tested three kinds of cognitive training: speed of processing, memory, and reasoning. The speed training was essentially a screen-based reaction drill, identifying visual targets fast with the difficulty climbing as you improve. Of the three arms, it was the only one to show a statistically significant protective effect. In the ten-year analysis, participants in the speed arm had roughly a 29% lower risk of developing dementia than untrained controls, and the figure quoted in recent coverage of the trial’s longer follow-up is around 25%.

Read this carefully

Screen-based speed training is not the same thing as catching a tennis ball, and one trial is not a settled science. What ACTIVE supports is a narrower claim: processing speed is trainable, and training it appears to leave a mark that lasts. Treat the dementia figure as promising, not proven.

What actually drives the adaptation

One principle governs all of this: the cue has to be unpredictable.

A memorised sequence, however complicated, trains execution. It does not train the react-and-respond loop, because there is nothing to react to. You already know what is next. Real stumbles do not announce themselves. Neither should your drills.

That is why the drills below all have randomness baked in: a recorded voice calling numbers you cannot anticipate, a ball coming off a wall at an angle you did not choose, a partner pointing left or right on a whim.

Measure yours right now

Interactive

Reaction tester

Tap to arm it. The panel will turn amber after an unpredictable delay. Tap again the instant it does.

Last
Best
Average

A phone screen adds its own lag, so treat this as a personal baseline rather than a clinical number. What matters is the direction it moves over a few weeks.

Four drills, no equipment budget

You can buy the same light boards and reflex apps the pros use, and they do give you structured programming and automatic tracking. They also cost money, often on subscription. None of the following costs anything beyond a tennis ball and some sticky notes.

Solo

Clock taps

Number twelve sticky notes and stick them on a wall in a clock formation. Record yourself on your phone calling out random numbers for 30 to 60 seconds, a few seconds apart. Play it back and slap each number as it is called.

5 rounds, short breaks. Re-record weekly so the sequence never becomes memory.

ANGLE = UNPREDICTABLE Solo

Wall bounce

Stand a few feet from a bare wall and throw a tennis ball at it, varying the angle every time. Catch the rebound before it bounces. Because you are changing the angle, you cannot pre-plan the catch.

3 rounds of 1 to 2 minutes. Step further back as you improve.

NO WARNING Partner

Ball drops

Stand an arm's length from a partner. They hold a tennis ball at shoulder height and let go without any signal. You catch it before the second bounce. Every success, take a small step back; every miss, step in again.

3 rounds of 1 to 2 minutes. The step-back is the progression.

~6 FT / 2 M APART Partner

Lateral shuffles

Set two markers about six feet apart, closer if that suits your fitness. A partner points left or right at random intervals and you shuffle that way. The randomness is the whole training effect; a metronome would ruin it.

3 rounds of 1 to 2 minutes. Shorten the distance rather than skipping the drill.

The dose

Pick two drills. Five to ten minutes, two or three times a week. That is the whole prescription, and it is small enough that the only real failure mode is not starting.

Log it, though. A line in your phone after each session: how fast you got through the clock taps, how many balls you dropped, whether the shuffle felt sharper or more controlled. Reaction gains are invisible day to day and obvious across a month, and the notes are the only way you will see them.

Then watch for where it shows up. Not in the drill, but in the curb you did not trip on, the glass you caught, the uneven trail you stopped avoiding.

Before you start

Some of these drills are cardiovascular. Talk to your doctor before beginning this or any new exercise programme, particularly if you have a history of falls, dizziness or heart trouble.

Sources