Physiological versus risk-relevant lipid remodelling across the menopausal transition: FMP-anchored trajectories, male benchmarking, LDLR genetics, risk calibration and incident MACE in 273,036 UK Biobank women

Genedy, Nader and Zouwail, Soha 2026. Physiological versus risk-relevant lipid remodelling across the menopausal transition: FMP-anchored trajectories, male benchmarking, LDLR genetics, risk calibration and incident MACE in 273,036 UK Biobank women. Journal of Clinical Lipidology 10.1016/j.jacl.2026.07.022

PDF - Accepted Post-Print Version Available under License Creative Commons Attribution. Download (1MB)

Abstract

Background Cardiovascular risk factors accelerate around the final menstrual period (FMP), yet clinicians lack an operational way to separate expected reproductive-ageing physiology from risk-relevant lipid divergence, and age-weighted calculators may under-recognise midlife women. We aimed to define, benchmark and stratify menopausal lipid remodelling against incident major adverse cardiovascular events (MACE). Methods We analysed 273,036 UK Biobank women (63,939 premenopausal; 165,276 postmenopausal; 43,821 uncertain, excluded from the primary contrast). Linked analyses used NMR metabolomics (n=222,487), cardiac MRI (n=42,033), oestradiol (n=57,999), an age-matched male reference cohort (n=228,900), serial Lp(a) (n=6,104 paired), and a Welsh FSH-confirmatory cohort (n=330). Incident-event analyses used a base cohort of 224,256 women (16,961 MACE) and an age-as-timescale Cox lipid cohort of 151,812 women (11,237 events). We integrated FMP-anchored trajectories, male-matched signal classification, treatment-adjusted change, counterfactual oestradiol mediation, one-sample Mendelian randomisation (MR) with LDLR ClinVar stratification, event-type dissociation, marker-added discrimination, guideline cutpoint calibration, and an internal sex-specific calibration stress test. Results Postmenopausal women had higher LDL-C (3.23 to 3.72 mmol/L; +15.3%), ApoB (+13.5%), non-HDL-C (+16.8%) and triglycerides (+29.9%); treatment adjustment raised the LDL-C and ApoB estimates to +19.8% and +17.5%. Male-matched analysis classified ApoB, LDL-C, non-HDL-C and remnant cholesterol as menopause-linked and TC:HDL and TG:HDL as age-driven. LDL-C and ApoB accelerated around the FMP (breakpoints 4.32 and 4.48 years). Oestradiol mediated 61.5% of the LDL-C difference. MR supported LDL-C causality (causal HR 1.196 [1.018-1.374]) with a null menopause interaction (P=0.826); LDLR carriers started higher but did not rise faster (Δ +11.5% vs +15.3%). ApoB tracked IHD (HR 1.125 [1.104-1.146]) but not higher HF risk (HR 0.916 [0.889-0.943]). In postmenopausal women, TC:HDL and TG:HDL, not ApoB, added short-term discrimination. At age 58, ApoB ≥1.2 g/L corresponded to 10.88% 10-year MACE risk in men versus 4.34% in postmenopausal women; a sex-neutral internal model over-predicted women (observed/expected 0.65-0.78) and under-predicted men (1.19). Ten-year MACE risk ranged from 1.09% (premenopausal, ApoB <1.2) to 5.89% (postmenopausal, ApoB ≥1.2). Conclusions Menopausal lipid remodelling comprises an expected physiological envelope and a risk-relevant, ApoB-rich, remnant-enriched divergence that predicts IHD. Because static sex-neutral cutpoints mis-estimate female risk and LDLR carriage is additive rather than synergistic, transition-era prevention needs sex-specific absolute-risk interpretation and treatment-aware relative-change monitoring. The proposed bands extend SWAN and align with ESC/EAS 2025 and AHA/ACC 2026 but are hypothesis-generating and require prospective validation.

Item Type:	Article
Date Type:	Published Online
Status:	In Press
Schools:	Schools > Medicine
Publisher:	Elsevier
ISSN:	1933-2874
Date of First Compliant Deposit:	4 September 2026
Date of Acceptance:	14 July 2026
Last Modified:	04 Sep 2026 11:45
URI:	https://orca.cardiff.ac.uk/id/eprint/189402

Actions (repository staff only)

Edit Item