Originally published 2019-02-04.
A widely shared Australian study found that among people aged 40 and over, working up to roughly 25 hours a week was linked with better scores on memory and reasoning tests, while longer hours were linked with lower scores. Very long weeks, around 44 to 60 hours, went with scores below those of people who did not work at all. Newspapers turned that into the headline that people over 40 should work only three days a week. The study supports a more modest message: some work seems good for the ageing brain, and a great deal of it may not be.
Below is what the researchers measured, what they found, where the popular version goes too far, and what it might mean for people planning the second half of their working lives.
Who did the research
The paper, titled "Use it Too Much and Lose it? The Effect of Working Hours on Cognitive Ability," was written by Japanese economists Shinya Kajitani, Colin McKenzie of Keio University and Kei Sakata. It first appeared in 2016 in the Melbourne Institute Working Paper series and was later issued as a Keio University discussion paper.
The authors used Wave 12 of the Household Income and Labour Dynamics in Australia survey, known as HILDA, a long-running national study that follows thousands of households. Wave 12, collected around 2012, was the only round that included cognitive tests. Their sample was 2,965 men and 3,502 women, all aged 40 or older and with English as their first language.
How cognition was measured
Participants took three short tests, each capturing a different side of thinking:
- Backward digit span: repeating strings of numbers in reverse order, a test of working memory.
- Symbol digit modalities: matching symbols to numbers as fast as possible in 90 seconds, which tests attention, visual scanning and processing speed.
- National Adult Reading Test (25 words): pronouncing irregularly spelled words, which mostly reflects accumulated knowledge, sometimes called crystallised intelligence.
The researchers compared these scores with how many hours each person worked per week, while controlling for factors such as education, income, health, family situation and occupation.
What the study found
The relationship looked like an upside-down U. Moving from no paid work to part-time work went with better scores. Around 25 hours a week, the benefit peaked. Beyond that, each extra hour went with slightly lower scores. The peak varied by test: about 25 hours for men and 22 for women on the memory test, and nearer 30 and 27 hours on the speed test.
The most striking result came at the far end. Men working roughly 44 to 52 hours a week and women working about 44 to 60 hours scored lower than people who were not employed at all. The authors found no statistically significant difference between men and women in how hours affected cognition.
Some coverage at the time quoted a 30-hour threshold and a 55-hour danger point. The paper's own estimates are closer to 25 hours, with the negative zone starting in the mid-40s, which is a normal full-time week for many people.
Why work might help, and why too much might hurt
The idea behind the study is sometimes summed up as "use it or lose it." Work brings problems to solve, people to talk to, routines and reasons to learn new things, all of which are thought to keep mental skills in practice. Leaving work entirely removes many of those prompts.
Speaking to The Times, as reported by The Independent, McKenzie described work as "a double-edged sword, in that it can stimulate brain activity." Long hours, he added, bring fatigue and stress that may undermine the very abilities the work exercises. His suggestion was that part-time work in middle and older age could help people keep their cognitive edge.
There are plausible mechanisms on the downside too. Chronic stress, poor sleep and little time for exercise or social life are all associated with worse cognitive health, and very long working weeks tend to squeeze all three.
The limits of the evidence
This was a careful study, but it has clear boundaries, and the authors acknowledge several of them.
- A single snapshot. The data come from one survey wave, so the study compares different people at one point in time. It cannot show how any individual's thinking changed as their hours changed.
- Cause and effect are hard to separate. People with sharper minds may choose, or be able to keep, certain kinds of jobs and hours. The researchers used a statistical technique called instrumental variables to reduce this problem, but they note that some of their instruments may not be perfect.
- One country, one era. The results come from Australians in about 2012. Work culture, job types and retirement rules differ elsewhere.
- Hours are not the whole story. The study did not look at how demanding, varied or stressful each job was. Twenty-five hours of repetitive work and twenty-five hours of complex problem-solving may not have the same effect.
So the study does not show that everyone over 40 should switch to three days a week, or that full-time work damages the brain. It shows an association that fits a reasonable theory and deserves further testing.
Why it matters for retirement policy
Part of the reason the study drew attention is that many countries are raising the age at which people can claim a state pension, which means more people working into their late sixties. In the UK, the state pension age began rising from 66 to 67 in 2026, and a further rise to 68 is legislated for 2044 to 2046, affecting people born from the late 1970s onwards. A government review launched in 2025 is considering the timetable again.
If longer careers are coming, the study's authors argue, the shape of those careers matters. Flexible or reduced hours for older workers could help people stay productive, healthy and employed for longer, rather than burning out or stopping abruptly. Research on retirement and cognition is mixed, but the general direction points toward staying engaged rather than either overworking or withdrawing completely.
Practical takeaways
For individuals, the study offers food for thought rather than a prescription:
- If you are planning for later working life, phased retirement or part-time work may be worth exploring, both for income and for staying mentally active.
- If you regularly work very long weeks, protecting sleep, exercise and time with other people is sensible for many reasons, cognitive health among them.
- Activity outside paid work counts too. Volunteering, learning, hobbies and social contact all provide the kind of stimulation researchers think matters.
- Anyone noticing real changes in memory or concentration should talk to a doctor rather than putting it down to workload.
The headline was too neat, but the underlying point is worth keeping: for the brain, as for the body, the right amount of work seems to be somewhere between none and too much.
Sources: Kajitani, McKenzie and Sakata, Melbourne Institute Working Paper 7/16; Keio University Panel Data Research Center, DP2016-008; The Independent (McKenzie comments); Wikipedia: State Pension (United Kingdom).
COMMENTS