Proposed Article Structure
This topic is best organized around clinical relevance and causal inference: highlights, disease burden and rationale, study design and Mendelian randomization framework, key findings, mechanistic interpretation and mediation, clinical implications, strengths and limitations, and conclusion with citation and funding disclosure.
Highlights
High body mass index was associated with a higher genetic risk of vascular-related dementia across both one-sample and two-sample Mendelian randomization analyses.
The estimated effect size was clinically meaningful: in pooled one-sample Mendelian randomization, each 1-standard deviation higher BMI predicted an odds ratio of 1.63 for vascular-related dementia.
Blood pressure appeared to explain part, but not all, of the association. Systolic blood pressure mediated 18% and diastolic blood pressure 25% of the genetic effect of BMI on vascular-related dementia.
The findings reinforce excess adiposity and hypertension as modifiable prevention targets in dementia risk reduction, especially for vascular cognitive disease.
Background and Clinical Context
Dementia is increasingly understood as a heterogeneous syndrome rather than a single disease entity. Vascular-related dementia occupies an important clinical space because it arises from cerebrovascular injury, small-vessel disease, strategic infarction, or mixed neurodegenerative and vascular pathology. In practice, many older adults classified as having dementia have overlapping Alzheimer and vascular lesions, making prevention of vascular injury especially important.
Obesity has long been linked to cognitive decline and dementia in observational studies, but interpretation has been difficult. Conventional cohort analyses are vulnerable to confounding, reverse causation, and age-dependent effects. For example, body weight often falls in the prodromal phase of dementia, which can make lower BMI appear harmful in later life while masking the effects of longstanding excess adiposity in midlife. This has contributed to inconsistent literature, including reports of U-shaped or even paradoxical associations.
Mendelian randomization offers a way to strengthen causal inference. Because inherited genetic variants associated with BMI are assigned at conception and generally precede disease onset, they can act as instrumental variables that are less prone to confounding and reverse causation than standard observational analyses. In this context, the study by Nordestgaard and colleagues addresses a clinically important question: is high BMI a causal risk factor for vascular-related dementia, and if so, to what extent is the effect mediated through known metabolic and vascular intermediates such as blood pressure, lipids, glycemia, and inflammation?
Study Design and Methods
Overall design
This study integrated several complementary approaches: prospective cohort analyses, one-sample Mendelian randomization, two-sample Mendelian randomization, and Mendelian randomization mediation analyses. The investigators drew on individual-level population data from cohorts in the Copenhagen area and the United Kingdom, as well as summary-level data from consortia.
Population and data sources
The abstract indicates inclusion of general population studies from Copenhagen and across the United Kingdom, together with consortia-level genetic data. This combination is methodologically important. One-sample Mendelian randomization allows instrument-exposure and instrument-outcome relationships to be assessed within the same dataset, while two-sample Mendelian randomization leverages larger summary datasets to improve precision and examine robustness across analytic methods.
Exposure and outcomes
The primary exposure was body mass index, analyzed per 1-standard deviation genetically predicted increase. The primary outcome was vascular-related dementia. The study also considered Alzheimer’s disease and ischemic heart disease among main outcome measures, although the headline findings presented in the abstract focus on vascular-related dementia.
Mendelian randomization strategy
The investigators used multiple Mendelian randomization estimators, including inverse-variance weighted, weighted median, and weighted mode approaches. This matters because each estimator makes somewhat different assumptions about horizontal pleiotropy, meaning the possibility that genetic variants influence the outcome through pathways other than BMI. Directional consistency across methods increases confidence that the observed association is not an artifact of a single modeling assumption.
Mediation analysis
To explore potential pathways, the authors performed Mendelian randomization mediation analyses examining hypertension, hyperlipidemia, hyperglycemia, and low-grade inflammation. This step moves the work beyond a simple exposure-outcome association toward a more translational question: which modifiable downstream processes might account for the elevated dementia risk associated with higher adiposity?
Key Findings
Primary causal estimate from one-sample Mendelian randomization
In a meta-analysis of two one-sample Mendelian randomization studies, a 1-standard deviation higher BMI was associated with an odds ratio of 1.63 for vascular-related dementia, with a 95% confidence interval of 1.13 to 2.35. This estimate suggests a substantial relative increase in risk and is statistically compatible with a clinically relevant causal effect.
For clinicians, the magnitude is notable. A roughly 60% higher odds of vascular-related dementia per standard deviation increase in BMI is larger than what would usually be dismissed as a weak epidemiologic signal. Although absolute risk depends on age and competing mortality, the effect size is strong enough to support serious consideration of excess adiposity as part of dementia prevention strategy rather than solely cardiovascular risk management.
Two-sample Mendelian randomization supports the signal
The two-sample Mendelian randomization analyses yielded similar results. The odds ratio for vascular-related dementia per 1-standard deviation higher BMI was 1.54 using the inverse-variance weighted method, 1.87 using the weighted median method, and 1.98 using the weighted mode method. The corresponding 95% confidence intervals were 1.10 to 2.16, 1.22 to 2.85, and 1.21 to 3.22, respectively.
These estimates are important for two reasons. First, all point estimates were above 1.0, and all confidence intervals excluded the null. Second, the stronger effect estimates with weighted median and weighted mode approaches may suggest that the signal persists even when some genetic instruments are potentially invalid, which argues against the association being wholly driven by pleiotropic bias.
Analyses with expanded genetic instruments were directionally consistent
The authors report that analyses using extended numbers of genetic variants were directionally consistent. Even without all numerical details in the abstract, this strengthens the internal coherence of the study. In Mendelian randomization, consistency across instrument sets is a useful check on robustness because it reduces concern that findings hinge on a narrow set of loci.
Observational versus genetic pattern
An especially interesting aspect of the paper is the contrast between observational and genetic associations. Observationally, the relation between BMI and vascular-related dementia was described as U-shaped, whereas genetically the relation was linear. This distinction is biologically and epidemiologically plausible. A U-shaped observational curve may reflect illness-related weight loss, frailty, smoking confounding, or selective survival, while the genetic analysis better captures lifelong predisposition to higher adiposity. In practical terms, the Mendelian randomization results argue that the harmful side of the BMI-dementia relationship is driven by higher BMI, not by a truly protective effect of overweight.
Mechanistic Interpretation and Mediation
Blood pressure as a partial mediator
The most clinically actionable mechanistic finding is that blood pressure mediated part of the BMI effect on vascular-related dementia. Systolic blood pressure accounted for 18% of the genetic effect, with a 95% confidence interval of 10% to 61%, while diastolic blood pressure accounted for 25%, with a 95% confidence interval of 13% to 75%.
These data align with established cerebrovascular biology. Excess adiposity contributes to sympathetic activation, sodium retention, vascular stiffness, endothelial dysfunction, and renin-angiotensin-aldosterone system activation, all of which can raise blood pressure. Hypertension then promotes white matter injury, lacunar infarction, microbleeds, blood-brain barrier dysfunction, and reduced cerebral perfusion reserve, all central features of vascular cognitive impairment.
At the same time, mediation was incomplete. Most of the BMI effect was not explained by the blood pressure measures alone. This implies that obesity likely contributes to vascular-related dementia through additional pathways, potentially including insulin resistance, dyslipidemia, thrombosis, sleep-disordered breathing, systemic inflammation, adipokine signaling, and direct effects on cerebral small vessels or neurovascular coupling.
What about lipids, glycemia, and inflammation?
The abstract emphasizes blood pressure as the significant mediator and does not report comparable mediation estimates for hyperlipidemia, hyperglycemia, or low-grade inflammation. That absence is itself informative. It suggests either weaker evidence for mediation through those pathways or insufficient precision to quantify it robustly. Clinically, this does not mean those pathways are irrelevant, only that hypertension emerged as the clearest and most quantifiable conduit in this analysis.
Clinical and Public Health Implications
Implications for dementia prevention
The study supports a prevention model in which excess adiposity is not merely associated with vascular dementia but contributes causally to it. This matters because obesity is common, modifiable, and often present decades before dementia manifests clinically. The findings therefore reinforce the idea that dementia prevention should begin well before old age and should include cardiometabolic risk reduction as a brain health strategy.
For clinicians in primary care, neurology, geriatrics, and endocrinology, the practical message is straightforward: sustained management of obesity and hypertension may have relevance beyond myocardial infarction and stroke prevention. These exposures may also influence the lifetime risk of vascular cognitive decline and dementia.
How should this affect current practice?
The study does not by itself establish that intentional weight loss reduces dementia incidence, nor does it define the best intervention window. However, it adds causal support to existing guideline-consistent care: treat elevated BMI as an important long-term vascular risk factor, and aggressively identify and control high blood pressure. Midlife is likely to be especially important, given the long latency of cerebrovascular and neurodegenerative pathology.
These findings also complement broader dementia prevention frameworks, such as those emphasizing control of hypertension, diabetes, smoking, physical inactivity, and hearing loss. In that sense, obesity may function both as an independent risk factor and as an upstream amplifier of other recognized hazards.
Strengths of the Study
The study has several notable strengths. First, it uses Mendelian randomization, a design well suited to probing causality when randomized trials are impractical or impossible. Second, it combines one-sample and two-sample analyses, increasing methodological triangulation. Third, use of multiple Mendelian randomization estimators improves robustness to different forms of instrument invalidity. Fourth, the mediation analyses increase translational value by highlighting blood pressure as a tangible intervention target.
Another strength is the focus on vascular-related dementia rather than all-cause dementia alone. This is important because obesity may be expected to show stronger causal links with cerebrovascular pathology than with pure Alzheimer pathobiology. Disease-specific analyses can therefore reveal effects that may be diluted in broader dementia phenotypes.
Limitations and Cautions
Despite the strength of the causal framework, several limitations deserve attention. Mendelian randomization depends on instrumental variable assumptions that can never be proven fully. Horizontal pleiotropy remains a potential concern, even when sensitivity methods are used. The consistency across inverse-variance weighted, weighted median, and weighted mode analyses is reassuring but not definitive.
Phenotype definition is another issue. Vascular-related dementia can be difficult to diagnose accurately in routine datasets because mixed pathologies are common and diagnostic coding may vary across healthcare systems. Misclassification could bias results, although this would often tend to attenuate true associations.
The interpretation of BMI also has limitations. BMI is a pragmatic population measure of adiposity but does not distinguish fat distribution, visceral adiposity, muscle mass, or metabolic health status. Future genetic studies using waist-to-hip ratio, body fat percentage, or imaging-based adiposity traits may refine risk attribution further.
Generalizability may also be constrained by the ancestry composition of the contributing cohorts, which is not detailed in the abstract but often remains predominantly European in large Mendelian randomization studies. Applicability to more diverse populations requires confirmation.
Finally, mediation estimates should be interpreted carefully. Mediation in Mendelian randomization is informative but complex, particularly when pathways overlap biologically and when mediators such as blood pressure change across the life course.
Expert Commentary
This study fits well with an emerging view of dementia as partly preventable through vascular and metabolic risk modification. It is especially persuasive because it addresses a long-running controversy in observational epidemiology: whether obesity itself contributes to dementia or whether the association is mainly confounded by age, frailty, and prodromal weight loss. The genetic evidence here supports a genuine harmful effect of higher BMI on vascular-related dementia.
Importantly, the work should not be overextended to imply that all dementia subtypes share the same adiposity biology. The title and data point specifically to vascular-related dementia. That distinction matters in counseling, research planning, and therapeutic prioritization. Future studies separating vascular dementia, mixed dementia, and biomarker-defined Alzheimer disease will be valuable.
Conclusion
The study by Nordestgaard and colleagues provides strong genetic evidence that higher BMI is a causal risk factor for vascular-related dementia. Across one-sample and two-sample Mendelian randomization analyses, each 1-standard deviation increase in genetically predicted BMI was associated with materially higher odds of vascular-related dementia. Elevated systolic and diastolic blood pressure explained part of this relationship, identifying hypertension as an important downstream mediator and prevention target.
For clinicians and health systems, the broader implication is clear: obesity management and blood pressure control belong not only to cardiometabolic care but also to long-horizon brain health strategy. For researchers, the next steps include clarifying subtype specificity, identifying additional mediating pathways, and determining whether effective long-term weight reduction translates into measurable reductions in vascular cognitive impairment and dementia incidence.
Funding and ClinicalTrials.gov
The abstract does not report a ClinicalTrials.gov registration number, which is expected given the observational and genetic epidemiology design rather than an interventional trial. Specific funding details are not provided in the supplied abstract and should be confirmed from the full article.
Citation and Selected References
Nordestgaard LT, Luo J, Emanuelsson F, Leyden G, Sanderson E, Davey Smith G, Christoffersen M, Afzal S, Benn M, Nordestgaard BG, Tybjærg-Hansen A, Frikke-Schmidt R. High Body Mass Index as a Causal Risk Factor for Vascular-Related Dementia: A Mendelian Randomization Study. The Journal of Clinical Endocrinology and Metabolism. 2026 May 19;111(6):e1681-e1694. PMID: 41568975. URL: https://pubmed.ncbi.nlm.nih.gov/41568975/
Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396(10248):413-446.
Emdin CA, Khera AV, Kathiresan S. Mendelian randomization. JAMA. 2017;318(19):1925-1926.
Skoog I, Gustafson D. Hypertension, hypertension-clustering factors and Alzheimer’s disease. Neurol Res. 2003;25(6):675-680.
Iadecola C, Duering M, Hachinski V, Joutel A, Pendlebury ST, Schneider JA, et al. Vascular cognitive impairment and dementia: JACC scientific expert panel. Journal of the American College of Cardiology. 2019;73(25):3326-3344.