Women who reach menopause unusually early already carry a higher long-term risk of dying sooner and of developing heart disease and dementia. A study published online August 12 in Menopause, the journal of The Menopause Society, asked a question clinical practice has largely skipped: does it matter what kind of life those women are living?
It does, by a wider margin than most clinical protocols would suggest. Among 19,912 UK Biobank women who went through early natural menopause, those in the most socially disadvantaged third had a 75% higher risk of dying from any cause than those in the least disadvantaged third over a median of nearly 11 years. Their risk of incident dementia was 83% higher.
The single factor that mattered most was not income, education, or neighborhood. It was being out of work.
Fourteen Measures of Disadvantage, Added Up
The research team at Anhui Medical University in China, with Wen Yang among three co-first authors, took an approach that differs from most work in this area. Rather than testing one social variable at a time, they scored 14 of them together, on the logic that disadvantages rarely arrive alone.
The measures spanned five domains: financial circumstances such as household income, employment status, and area-level income deprivation; educational attainment and area-level education deprivation; area-level healthcare deprivation; housing stability, local crime scores, and the natural environment; and social factors including living alone, social support, social activity, isolation, and emotional distress. Each was scored 0 or 1, summed, and the women were split into thirds.
Compared with the most favorable third, the least favorable third had a hazard ratio of 1.75 for all-cause mortality, 1.69 for cardiovascular death, 1.48 for incident cardiovascular disease, and 1.83 for incident dementia, after adjusting for age, body mass index, smoking, drinking, hypertension, diabetes, age at first period, number of births, and hormone therapy use.
The mortality findings were then reproduced in a separate American population, 1,732 women in the US National Health and Nutrition Examination Survey, scored on nine comparable measures. There, the least favorable group had a hazard ratio of 1.53 for all-cause mortality.
Losing a Job Mattered More Than Income or Education
When the researchers entered every social variable into a single model to see which carried independent weight, unemployment separated from the rest. Compared with women who were employed or retired, unemployed women had a hazard ratio of 1.91 for all-cause mortality in the UK cohort and 1.43 in the US cohort. For incident dementia, unemployment was associated with a hazard ratio of 1.98.
Cardiovascular disease behaved differently. There, household income under £31,000 was the strongest single contributor, at a hazard ratio of 1.36.
The finding sits alongside a large existing body of literature. A systematic review and meta-analysis of unemployment and mortality covering some 20 million people found that unemployment raised the risk of death from any cause by 63%, and a separate meta-analysis of longitudinal cohorts identified low social participation, infrequent social contact, and loneliness as predictors of dementia.
The authors are careful about mechanism, offering three routes rather than one: depression, which is common after job loss; psychosocial stress driving inflammation; and shifts in smoking and drinking that accumulate into chronic disease.
More Cancer Deaths, but Not More Cancer
The most striking result in the paper is a null one. Social disadvantage showed no association at all with developing cancer. The hazard ratio for incident cancer in the least favorable group was 0.99, effectively identical to the most favorable group.
Yet disadvantage was associated with a 35% higher risk of dying from cancer in the UK cohort, and in the smaller American cohort the estimate reached a hazard ratio of 2.61, based on just 22 cancer deaths in that group.
A pattern where the odds of getting a disease stay flat while the odds of dying from it rise points away from biology and toward what happens after diagnosis. The study did not measure screening uptake, time to diagnosis, stage, or treatment completion, so it cannot demonstrate that. But it is the kind of divergence that usually reflects the care pathway rather than the tumor.
Where the Study Runs Out of Certainty
Several limits deserve prominence, and the authors state them plainly.
This is observational research. It cannot establish that disadvantage causes worse outcomes, and reverse causation is a live concern: the disadvantaged group was already in worse health at baseline, more likely to smoke, to have a higher BMI, and to have hypertension or diabetes. Some of the unemployment signals may reflect illness that preceded the job loss.
Menopause age was self-reported years later, introducing recall bias. Social circumstances were measured only once at enrollment, so the study cannot track women whose circumstances changed. And because participants were surveyed a median of 15 years after menopause, women who died or became severely ill soon afterward never entered the cohort at all.
One definitional point is worth flagging. Premature and early menopause are conventionally defined as occurring before 40 and between 40 and 45. This study used a threshold of 48 to capture women below the population average, though a sensitivity analysis using the under-45 definition produced consistent results. Estimates of prevalence also vary: the paper cites about 8% in developed countries and 12% globally, while the accompanying press release gives roughly 6% worldwide.
The clinical argument the authors make is narrow and reasonable. Care for early menopause currently centers on hormone therapy and lifestyle advice. Dr. Stephanie Faubion, medical director of The Menopause Society, said in a statement that socioeconomic disadvantages "significantly amplify adverse health trajectories" in these women, and the study team argues that social assessment belongs inside that care rather than outside it.
Nothing here means a disadvantaged woman will have a poor outcome. These are population averages. Women who went through menopause early should discuss cardiovascular and cognitive risk with a clinician, who can weigh hormone therapy and other prevention against their individual history.
Key Questions Answered
What did the study measure?
It scored 14 social determinants of health together in 19,912 UK Biobank women with early natural menopause, then tracked deaths, cardiovascular disease, cancer, and dementia over a median of about 11 years.
Which social factor mattered most?
Unemployment, which carried the strongest independent association with both all-cause mortality and incident dementia. Low household income was the leading contributor to cardiovascular disease.
Why is the cancer finding unusual?
Disadvantage was not associated with a higher chance of developing cancer, but was associated with a higher chance of dying from it, a pattern that points toward diagnosis and treatment rather than disease onset.
Does this prove that social disadvantage causes these outcomes?
No. This is observational research, and the disadvantaged group was already in worse health at the start, so some of the association may run in the other direction.
How was early menopause defined?
As natural menopause before age 48, which is broader than the conventional threshold of 45. A sensitivity analysis using the under-45 definition produced similar results.
What should women who have had early menopause do?
Discuss cardiovascular and cognitive risk with a clinician. The study argues that clinicians should also ask about employment, income, and social support rather than treating menopause as a purely hormonal issue.
