Women kept a thinking-test edge from midlife on, but did not get dementia less often

About two in three people living with dementia are women. Part of the reason is simple: women live longer, and dementia becomes more common with age. Whether women are actually more likely to develop it at a given age is still debated. At the same time, many studies find that women do better than men on memory tests in midlife, which could hide early decline in women or, alternatively, give them a buffer.

A team at Imperial College London looked at both questions in Whitehall II, a long-running cohort of British civil servants. Their answer: women held a small edge on thinking tests from midlife on, but it did not go with a lower rate of dementia.

What they did

• Who: 5,920 participants who took thinking tests for the first time in 1997–99, at an average age of 55.6. Seven in ten were men and over 90% were white. They were tested again roughly every four to five years until 2015–16; 5,047 had at least one repeat test.
• The tests: short-term memory (hearing 20 short words and writing down as many as possible) and two kinds of verbal fluency, the ability to find words quickly: writing as many words beginning with "S" as possible in one minute, and as many animal names as possible in one minute.
• Dementia: diagnoses were taken from national hospital and mental-health records and from death records. This analysis covered 4,154 people, of whom 289 developed dementia.
• Accounting for other differences: the team adjusted for age, ethnicity, education, income, job grade and marital status, and then for blood pressure, diabetes, heart disease, weight, smoking, drinking, diet, sleep, exercise, cancer, depression and anxiety.

What they found

• A female edge, once background is counted. Without adjustment, women and men scored about the same on memory and men did better at naming animals. But men in this group were more often university-educated and in senior jobs. After adjusting for education, income, job and marital status, women scored higher than men on all three tests. Adding health and lifestyle factors did not change this.
• Faster decline in men's word-finding. Over up to 16 years, men's scores on the animal-naming test fell faster than women's. Men's "S"-word scores also fell faster, but this difference was weaker and did not pass the stricter statistical test the team set for checking many results at once.
• No difference in memory decline. Memory scores fell at a similar rate in both sexes.
• No difference in dementia. Men and women were diagnosed with dementia at similar rates, whichever factors the team adjusted for.
• Birth year and genes changed little. People born in 1935–44 and in 1945–54, and carriers and non-carriers of APOE4 (a common gene variant that raises the risk of Alzheimer's disease), showed broadly the same pattern; some hints of differences did not hold up after correction for multiple testing.

What to keep in mind

• Observational. The study shows associations between sex and test results; it cannot explain them. The authors mention possible reasons, such as hormones or "cognitive reserve", but did not test them.
• Not a typical population. Most participants were men, most were white, and all were civil servants, who tend to be healthier and better off than the general population.
• Few very old people. Testing started at about 55, so few participants reached 80 during the study. Some earlier studies found sex differences in dementia only at the very oldest ages.
• Relatively few dementia cases. With 289 cases, the authors note that the study may have lacked the power to detect a real difference in dementia risk.
• Limited tests and no biomarkers. Only memory and word-finding were measured, not attention or spatial skills, and there were no blood or brain-scan measures of Alzheimer's or blood-vessel damage.

Why it matters

Doctors often use thinking tests to spot early decline. If women start from a higher baseline, a woman whose memory is slipping may still score "normal" for a while. This study adds to evidence that women do better than men on some tests in midlife, and that men's word-finding may decline faster, but also that these test differences do not translate into a difference in dementia diagnoses, at least in this group. The authors call for studies that combine thinking tests with biological markers and life-course factors to understand why.