Eczema and Dementia: Cohort Associations Did Not Match the Genetic Causality Test
Dr. Arsy Editorial education, checked November 9, 2026. This is not individual medical advice or independent clinical review. Cited researchers and organizations do not endorse #003.
A headline linking eczema to dementia can make an ordinary forgotten appointment feel frightening. The research does not justify that conclusion. A January 2025 paper used two different approaches—observational cohorts and genetic analysis—and they did not provide the same kind of evidence. Neither approach can diagnose a reader or turn childhood eczema into a forecast of late-life dementia.
What the review included
The systematic review and Mendelian-randomization study searched through March 30, 2024 and combined five longitudinal cohorts, reporting 12,576,235 participants. The cohorts came from national or regional records and included markedly different populations: for example, a UK cohort aged 60–100 and Korean participants over 40. Some used diagnostic codes and treatment records, while another included questionnaire-based eczema information.
A large combined sample improves precision but does not erase differences in age, diagnostic definitions, follow-up or recorded confounders. The paper’s 2025 publication date also does not mean its search included every subsequent study.
| Evidence | Result | Interpretation |
|---|---|---|
| Unadjusted all-cause dementia, five cohorts | HR 1.06; 95% CI 0.95–1.17; I² 94% | Not statistically significant; very heterogeneous. |
| Adjusted all-cause dementia, five cohorts | HR 1.15; 95% CI 1.07–1.23; I² 75% | A modest association remained after the original studies’ different adjustments. |
| Adjusted vascular dementia, three cohorts | HR 1.13; 95% CI 0.90–1.42 | The interval included no association. |
| Genetic instrumental-variable analysis: any-cause dementia | IVW OR 1.01; 95% CI 0.83–1.25; P=0.90 | Did not provide evidence for a significant causal effect. |
Adjustment and genetics answer different questions
The adjusted observational estimate was higher than the unadjusted pooled estimate. That does not mean adjustment proved causation: each cohort controlled a different set of measured factors, and pooled analyses can include differing available estimates. Medical-contact patterns, disease definitions and unmeasured differences can still matter. An HR of 1.15 is not a 15% absolute probability of dementia.
Mendelian randomization uses genetic variants associated with eczema as instruments rather than randomly prescribing eczema. Its assumptions—including that variants affect the outcome through the exposure rather than another route—can be difficult to verify. The genetic estimates were also nonsignificant for Alzheimer’s disease, vascular dementia and cognitive performance. That weakens a simple causal claim, but it does not prove that every possible eczema-related pathway is absent.
Do not let a subgroup become a personal prognosis
The adjusted moderate-to-severe subgroup estimate was 1.37 (95% CI 1.21–1.55), based on only two severity-stratified papers. The mild subgroup was not significant, with a wide interval. These categories were not a calibrated screening tool. The study did not test whether treating eczema prevents dementia, and it did not establish that a cosmetic moisturizer changes dementia risk.
A practical response to memory concerns
The National Institute on Aging distinguishes occasional forgetfulness from changes that interfere with daily tasks. Repeated questions, getting lost in familiar places or difficulty following familiar instructions warrant a clinical conversation. Sleep problems, medication effects, depression and other medical conditions can also affect memory; ask for assessment rather than assuming eczema explains the change.
For related but different issues, see our sleep overview, nursing cohort survey analysis, emotional-wellbeing guide and migraine evidence explanation. A familiar walk or social activity can remain part of a comfortable routine without being advertised as eczema-specific dementia prevention.
Dr. Arsy Moisturizer #003 supports cosmetic hydration, moisture-barrier support and ordinary dry-skin comfort. It is not an eczema medicine or a treatment or preventive product for cognitive impairment or dementia.