WHS Biomarkers
Inflammation, Cholesterol, Lipoprotein(a), and 30-Year Cardiovascular Outcomes in Women
Clinical Question
Do baseline levels of high-sensitivity CRP, LDL cholesterol, and lipoprotein(a) predict cardiovascular outcomes over 30 years in initially healthy women?
Bottom Line
Baseline high-sensitivity CRP, LDL cholesterol, and lipoprotein(a) levels independently predicted 30-year cardiovascular risk in healthy women, with the highest risk observed when all three biomarkers were elevated.
Major Points
- 30-year prospective follow-up of 27,939 initially healthy U.S. women enrolled in the Women’s Health Study
- High-sensitivity CRP, LDL cholesterol, and lipoprotein(a) each independently predicted cardiovascular risk
- Greatest risk was among women with all three biomarkers in the highest quintile (HR 2.63; 95% CI, 2.16–3.19)
- Hazard ratios attenuated slightly over time for CRP and LDL, but not lipoprotein(a)
- Supports use of combined biomarker stratification for long-term risk prediction
Design
Study Type: Prospective cohort study
Randomization:
Enrollment Period: 1992–1995
Follow-up Duration: 30 years
Centers: 1
Countries: USA
Sample Size: 27939
Analysis: Cause-specific Cox models, Fine–Gray competing risk models, stratified by quintile, adjusted for covariates and statin use in sensitivity analyses
Inclusion Criteria
- Initially healthy female health professionals
- Enrolled in Women’s Health Study between 1992–1995
- Provided baseline blood sample for biomarker assay
Exclusion Criteria
- Pre-existing cardiovascular disease at baseline
- Missing biomarker data
Arms
| Field | Women with baseline biomarker data |
|---|---|
| Intervention | Observational follow-up with no intervention |
| Duration | 30 years |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| First major adverse cardiovascular event (MI, revascularization, stroke, or CV death) | Primary | Reference: lowest quintile of each biomarker | CRP: HR 1.70; LDL: HR 1.36; Lp(a): HR 1.33 (Q5 vs Q1) | ||
| Combined effect of all 3 biomarkers in highest quintile | Secondary | No biomarker in Q5 | HR 2.63 (95% CI 2.16–3.19) | 2.63 | |
| Individual risk for stroke with all 3 biomarkers in Q5 | Secondary | No biomarker in Q5 | HR 1.68 (95% CI 1.14–2.48) | 1.68 | |
| Individual risk for coronary heart disease with all 3 biomarkers in Q5 | Secondary | No biomarker in Q5 | HR 3.71 (95% CI 2.94–4.68) | 3.71 | |
| Not reported | Adverse | No adverse event data extracted for this trial |
Subgroup Analysis
Stratified analyses by number of biomarkers in highest quintile, quintile-based risk curves, and biomarker interactions confirmed additive independent predictive effects
Criticisms
- Limited racial diversity (94% White participants)
- Female-only cohort limits generalizability to men
- Confidence intervals not adjusted for multiplicity
- No repeated biomarker measurements over 30 years
Funding
National Institutes of Health (HL043851, HL080467, HL099355, CA047988, CA182913)
Based on: WHS Biomarkers (The New England Journal of Medicine, 2024)
Authors: Ridker PM, Moorthy MV, Cook NR, ..., Buring JE
Citation: N Engl J Med 2024;391:2087–2097. doi:10.1056/NEJMoa2405182
Content summarized and formatted by NeuroTrials.ai.