When people retired mattered little for dementia; part-time work and disability did

A popular idea holds that work keeps the brain sharp, so retiring early should speed up memory loss: "use it or lose it". Studies have been split on it. One reason may be that "retirement age" is too simple a measure. Two people who retire at 63 may have got there in very different ways: one after full-time work, the other after years of part-time jobs, unemployment or illness.

A team led from Washington University in St. Louis looked at the whole path from age 51 to 75 instead of a single retirement date. It found that the timing of retirement mattered little for later dementia risk, but the kind of work along the way, and long periods of disability, did. It is an observational study, so it shows links, not cause and effect, and a preprint, not yet peer-reviewed.

What they did

• Who: 13,879 people from the US Health and Retirement Study, a large national survey of Americans over 50 that interviews them about every two years. Everyone in the analysis was still in the study at age 76.
• Work histories: for every year from age 51 to 75, each person was placed in one of six states: working full-time, working part-time, retired, disabled, not working (mostly caring for family), or unknown. A method called sequence analysis compared these 25-year histories and sorted them into 10 groups of similar paths.
• Dementia: from age 76 on, a validated formula estimated each person's probability of having dementia at each interview, using memory tests and, for people too impaired to answer, interviews with a relative. It is an estimate, not a doctor's diagnosis.
• Comparison: the reference group was the largest one, people who worked full-time and then retired at 62 to 65 (3,783 people). The analysis took into account age, gender, race and ethnicity, birthplace, years of schooling, parents' schooling and childhood health.

What they found

The results are given as odds ratios: 1 means the same odds of dementia as the reference group, below 1 lower odds, above 1 higher odds.

• Retirement timing made little difference. People who retired from full-time work before 60 (odds ratio 1.03) or at 70 or later (1.05) had about the same odds as those who retired at 62 to 65.
• Part-time work went with lower odds. People who kept working part-time through most of the years from 51 to 75 had 0.62 times the odds (518 people). Those who moved from full-time to part-time work before retiring at 70 or older had 0.75 times the odds. A third part-time group, who retired at 62 to 65, pointed the same way (0.81), but that result could have been chance.
• Long disability went with higher odds. People who spent long stretches on disability had 1.53 times the odds (757 people). This group had, on average, less schooling and poorer health in childhood.
• Not working, mostly to care for family, made no clear difference. A group with mostly unknown work histories had higher odds, which the authors find hard to interpret.
• Checks with a different way of grouping the paths and a different dementia measure gave similar results.

What to keep in mind

• Who can keep working? The authors stress that the findings "should not be interpreted as causal". People who are sharper, better educated or healthier may be both more able to keep a part-time job and less likely to develop dementia. The link could also run backwards: early memory problems can push people out of work.
• Missing information. Because part of each work history was reconstructed from memory, the team couldn't account for income, marital status or the type of job, all of which may matter.
• Survivors only. Only people who lived to 76 and stayed in the study were included, so people who left work early because of serious illness and died young are missing.
• One country. It is a US study; the authors call for checks in other settings.
• Not yet peer-reviewed.

Why it matters

Many countries are raising the retirement age, partly with the hope that longer working lives will also keep minds healthier. This study finds little sign that simply retiring later from a full-time job is linked to lower dementia risk. The more interesting link was with part-time work in later life, which may combine some mental and social activity with less stress and more free time, though it may just as well mark people who were healthier to begin with. The link between long disability and higher risk points to a group that may need more support. Whether gradual, flexible retirement actually helps the brain is a question this kind of study can raise but not settle.