Of 211 metabolites associated with lacunar stroke, two survived Mendelian randomisation

Design

case-control metabolomics with genome-wide association and bidirectional Mendelian randomisation

Population

2,408 UK participants - 1,456 with MRI-confirmed lacunar stroke, 952 hospital controls

Primary outcome

association of 250 serum metabolites with lacunar stroke, imaging markers and cognition

Effect

211 metabolites associated observationally; glycine odds ratio 0.704 (95% CI 0.585-0.848) for isolated lacunar infarct, medium-LDL cholesterol proportion 0.541 (0.401-0.730) for multiple infarcts or leukoaraiosis

Serum from 1,456 patients with MRI-confirmed lacunar stroke and 952 hospital controls, recruited across UK stroke centres, was assayed for 250 metabolites by NMR spectroscopy. Observational analysis found 211 of them significantly associated with lacunar stroke - a number that tells you mainly how many things move together in vascular disease.

Bidirectional Mendelian randomisation, which uses genetic variants as instruments to test whether an association is likely causal, supported just two. Glycine was inversely associated with isolated lacunar infarcts (odds ratio 0.704, 95% CI 0.585 to 0.848), and the proportion of cholesterol to total lipids within medium LDL was inversely associated with the multiple-infarct or leukoaraiosis subtype (0.541, 0.401 to 0.730). Notably each survived for a different subtype, which supports the long-argued position that these are distinct pathologies rather than a severity spectrum. Five further metabolites tracked executive function and processing speed observationally, and omega-3 proportion was implicated in two diffusion tensor imaging markers.

This is mechanism, not measurement. Nothing here is a test to order, and glycine is not a supplement to recommend on this evidence. What it changes is the framing of small vessel disease as a metabolic as well as a haemodynamic process, and it gives two specific targets for the studies that come next.

• Continue to manage lacunar stroke on blood pressure, smoking, diabetes and antiplatelet therapy
• Do not order metabolomic panels - none of this is a clinical test
• Note the subtype distinction: isolated lacunar infarct and multiple infarcts with leukoaraiosis behaved differently
• Record Fazekas grade in reports; the subtype split in this study depends on it
• Treat glycine supplementation claims that cite this paper as unsupported

Why it matters

It supplies evidence that the two lacunar subtypes differ in mechanism, not only in extent.

Don't overread it

Mendelian randomisation supports a causal direction; it does not show that changing glycine levels would prevent a stroke.

The statistics, in plain English

The gap between 211 observational associations and two that survive Mendelian randomisation is the point of the paper. Observational metabolite associations are heavily confounded - diet, renal function, inflammation and the stroke itself all move many metabolites at once. Mendelian randomisation uses inherited variants, fixed at conception, as a proxy for lifelong exposure, which removes most reverse causation. It has its own assumptions, chiefly that the genetic instrument affects the outcome only through that metabolite, and those assumptions cannot be fully verified.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.