The benefits of
mental math, measured.

Mental arithmetic is unusual among puzzle types in having randomised trials behind it, and the trials tested something close to a daily habit: a few minutes of simple sums, most days, for months. The results were modest, specific and real. Here is what they showed.

The learning therapy trials

A research group at Tohoku University led by Ryuta Kawashima has spent two decades testing a programme they call learning therapy: reading text aloud and solving simple arithmetic problems as quickly as possible, for a short session each day. In a 2008 trial, 124 community-dwelling adults aged 70 to 86 were randomised to six months of the programme or to a control group.^[2] The training group improved on measures of frontal lobe function and processing speed that had not been trained directly, which is the transfer effect most brain-training studies fail to find.

A 2016 trial by the same group randomised 64 healthy older adults to six months of reading aloud and simple calculation or to a waiting list.^[1] The training group improved on inhibition (the ability to suppress an automatic response), verbal episodic memory, focused attention and processing speed. Again, these were not the tasks being practised.

Why these results carry weight

Both studies were randomised, which rules out the selection problem that limits large surveys such as PROTECT. Both measured outcomes on tasks different from the training, which is the transfer test the 2016 Simons review found most brain-training evidence failing.^[3] And both used a schedule that ordinary people can keep: short sessions, most days, for months. That last feature fits the broader finding that practice spread over many short sessions is retained better than the same practice massed together.^[4]

Why they should not be oversold

The samples were small by modern standards, the participants were Japanese older adults, and the programme combined reading aloud with arithmetic so the contribution of the sums alone is unknown. The 2016 trial’s control group did nothing, so some of the benefit may reflect engagement and expectation rather than arithmetic specifically. The effects were measured on laboratory tasks, not on how participants managed their lives. Independent replication in other populations would strengthen the case considerably.

What this means for a daily sum or two

The trials point to a specific recipe: simple problems, done quickly, a little every day. Brisk’s two daily maths questions are a very small dose of that recipe. The mental maths trainer offers a sixty-second version closer to what the trials used, and the practice guide explains the methods that make simple sums fast. We do not claim any of this protects your brain. We do note that of all the puzzle types, this is the one whose habit-sized dose has been put through a randomised trial and passed.