🧬 biology

Sex-Specific Polygenic Risk Scores for Cardiovascular Disease: A Multi-Ethnic Analysis Using the 1000 Genomes Project

This multi-ethnic analysis of the 1000 Genomes Project demonstrates that sex-stratified polygenic risk scores for cardiovascular disease reveal significant sex-specific genetic architectures, with females exhibiting higher genetic risk scores than males across diverse ancestral groups, thereby challenging current sex-blind assessment paradigms.

Original authors: Smmrithi Ravindran, Vanathi Gopalakrishnan

Published 2026-06-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Smmrithi Ravindran, Vanathi Gopalakrishnan

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Idea: A "One-Size-Fits-All" Mistake

Imagine you are trying to predict who might get a flat tire on a long road trip. For years, doctors have used a single "risk formula" for everyone, regardless of whether they are driving a heavy truck or a light sports car. They assume the rules of the road are the same for everyone.

This paper argues that when it comes to heart disease, using a single formula for men and women is like using a truck manual to fix a sports car. The authors found that men and women carry different "genetic blueprints" for heart health, and ignoring these differences means we are likely missing the warning signs for women.

The Investigation: Checking the Genetic "Ingredients"

The researchers didn't look at sick people; they looked at a massive, global database of healthy people's DNA (the 1000 Genomes Project). Think of this database as a giant library containing the genetic "cookbooks" of 2,500 people from five different parts of the world: Africa, Europe, East Asia, South Asia, and the Americas.

They focused on five specific "chapters" in these cookbooks that are known to affect heart health (related to cholesterol, inflammation, and artery health). They asked a simple question: "Do men and women have different amounts of the 'risk ingredients' in these chapters?"

The Discovery: Women Have More "Risk Ingredients"

The answer was a clear yes.

The Score: The team created a new way to count these ingredients, called a Polygenic Risk Score (PRS). Instead of just counting how many risk genes you have, they counted how the frequency of those genes differs between men and women.
The Result: When they ran the numbers, women had a 17.3% higher "risk score" than men.
The Analogy: Imagine a jar of marbles where red marbles represent heart risk. The study found that if you look at the jars belonging to women, they are significantly fuller of red marbles than the jars belonging to men.

The "Very High Risk" Group

The most striking finding was about the people at the very top of the risk scale.

If you look at the top 5% of people with the highest genetic risk, 71% more women fell into this "very high risk" category compared to men.
The Metaphor: It's like a smoke detector. The old system (sex-blind) wasn't loud enough for women. It was quiet enough that many women with a lot of "smoke" (genetic risk) were told, "You're fine," while men with less smoke were told, "Be careful." The new system turns up the volume for women, revealing that many are actually in a much more dangerous zone than previously thought.

Does This Happen Everywhere?

The researchers checked if this was just a fluke in one specific group. They looked at people from five different ancestral backgrounds (African, European, East Asian, South Asian, and American).

The Result: In every single group, women had higher risk scores than men.
The Significance: This is like finding that the same pattern of rain happens in the desert, the jungle, and the mountains. It proves that this isn't a local weirdness; it's a fundamental biological rule that applies to humans everywhere.

Why Is This Happening? (The "Why" Behind the "What")

The paper suggests a biological reason for this.

The Hormone Shield: Men tend to get heart disease earlier in life. Women often stay protected by estrogen (a hormone) until menopause.
The Catch-up: Once that protection fades, women's risk shoots up. The authors suggest that women might be carrying a "backup load" of genetic risk factors that are kept in check by hormones. When the hormones leave, that genetic load becomes visible.
The Analogy: Think of it like a dam holding back water. Men's dams might be smaller, so water (heart disease) leaks out early. Women's dams are stronger (hormones), holding back a massive reservoir of water (genetic risk). When the dam finally breaks (menopause), the flood is huge.

What the Paper Actually Says (and Doesn't Say)

It is important to stick to what the authors explicitly claim:

What they found: There are measurable, consistent genetic differences between men and women regarding heart disease risk. Women carry a higher genetic burden.
What they built: They created a new "Sex-Stratified" scoring method that accounts for these differences. They also developed a clinical questionnaire tool (which they have filed for patent protection) to help integrate this data.
What they did NOT do: They did not test this on patients who actually had heart attacks. They did not prove that using this new score will save lives yet. They only proved that the genetic math is different for men and women.
The Warning: The authors admit that because they used a global database of healthy people, they haven't confirmed that these higher scores actually lead to more heart attacks in real life. That is the next step for future research.

The Bottom Line

This paper is a wake-up call for the "one-size-fits-all" approach to heart health. It suggests that by ignoring sex differences, we have been systematically underestimating the genetic risk for women. Just as we wouldn't use a men's shoe size for a woman, the authors argue we shouldn't use a man's genetic risk formula for a woman.

Technical Summary: Sex-Specific Polygenic Risk Scores for Cardiovascular Disease

Problem Statement
Current polygenic risk scores (PRS) for cardiovascular disease (CVD) exhibit poor clinical agreement, with approximately 20% of individuals being simultaneously classified as both high-risk and low-risk depending on the specific method used. A critical methodological limitation identified in recent evaluations is that existing PRS approaches treat sex merely as a covariate rather than stratifying analyses by sex. This "sex-blind" approach fails to account for well-documented sex differences in CVD incidence, outcomes, and underlying genetic architecture. Consequently, current models may systematically mask biological differences, leading to inaccurate risk stratification, particularly for female patients who often present with atypical symptoms and face delays in diagnosis.

Methodology
This study utilized genotype data from the 1000 Genomes Project Phase 3, comprising 2,509 individuals (1,237 male, 1,272 female) across five ancestral groups: African (AFR), European (EUR), East Asian (EAS), South Asian (SAS), and Admixed American (AMR).

Target Regions: The analysis focused on five cardiovascular-relevant genomic pathways: APOE (lipid metabolism), 9p21 (atherosclerosis), CRP (inflammation), ADIPOQ/CDH13 (metabolism), and IL-6 (inflammation).
Variant Identification: Sex-stratified allele frequencies were calculated. Variants exhibiting an allele frequency difference greater than 2% between males and females were identified as "sex-differential."
PRS Construction: Unlike traditional PRS weighted by disease effect sizes from case-control GWAS, this study constructed a score specifically to quantify sex-differential genetic architecture. Weights were derived from allele frequency-based odds ratios (OR) reflecting the magnitude of sex differences at each locus. A 5-SNP sex-stratified PRS was constructed using one index SNP per pathway.
Statistical Analysis: Chi-square tests assessed independence between sex and allele frequency. Significance was tested using independent-samples t-tests, with effect sizes quantified by Cohen's d.
Validation: The robustness of the findings was assessed via 70/30 train-test splits and 10-fold cross-validation. Ancestry-stratified analyses were performed to evaluate generalizability across populations.

Key Results

Identification of Sex-Differential Variants: A total of 182 sex-differential variants were identified across four of the five pathways (APOE, 9p21, CRP, and ADIPOQ/CDH13). No sex-differential variants were detected in the IL-6 pathway. The CDH13 region showed the most extensive sex-differential architecture (67 variants).
PRS Disparity: Female participants exhibited a mean PRS that was 17.3% higher than male participants ( $p = 1.65 \times 10^{-8}$ , Cohen's $d = 0.23$ ). The 95% confidence intervals for male and female means were non-overlapping, confirming a true population-level difference.
Risk Stratification: Using a threshold of the 95th percentile for "Very High Risk," 6.4% of female participants were classified as high-risk compared to 3.7% of males. This represents a 71% relative increase in the proportion of females classified as very high risk, translating to approximately 27 additional high-risk females per 1,000 screened.
Multi-Ethnic Consistency: Directional sex differences (females having higher PRS) were consistent across all five ancestral groups. Statistical significance was reached in African ( $p=0.017$ ) and East Asian ( $p=0.018$ ) populations, with consistent trends observed in European, South Asian, and American groups.
Index SNPs: Five index SNPs were selected based on the strongest sex-differential signals: rs7258841 (APOE), rs10965183 (9p21), rs2501352 (CRP), rs111339851 (CDH13), and rs1426810 (ADIPOQ). Notably, rs7258841 in the APOE region achieved genome-wide significance under strict Bonferroni correction, while the others reached nominal significance.

Significance and Claims
The authors claim that these findings challenge the prevailing paradigm of sex-blind genetic risk assessment. The study provides empirical evidence that failing to account for sex-specific genetic architecture may systematically underestimate the cardiovascular genetic burden in female patients.

Biological Plausibility: The finding of higher genetic risk scores in females is presented as biologically coherent with the observation that postmenopausal women experience accelerated CVD risk, potentially matching or exceeding that of age-matched men. The genetic architecture may capture the substrate for female-specific risk factors such as pre-eclampsia, gestational diabetes, and polycystic ovary syndrome.
Clinical Implications: The authors suggest that current sex-blind PRS may contribute to the underdiagnosis and undertreatment of CVD in women. They propose that future clinical guidelines should consider sex-stratified thresholds for genetic risk interpretation, analogous to sex-adjusted blood pressure thresholds.
Methodological Contribution: The study demonstrates that frequency-based weighting can robustly capture sex-differential genetic signals across diverse ancestries, offering a new approach to understanding genetic risk that complements traditional effect-size-based PRS.

Limitations and Future Directions
The authors maintain a modest stance regarding the immediate clinical application of these specific scores. They acknowledge that the 1000 Genomes Project lacks CVD outcome data, meaning the study cannot validate whether these genetic differences translate to differential disease incidence. Furthermore, aside from the APOE signal, the remaining findings are described as "hypothesis-generating" pending external replication in large, outcome-linked biobanks (e.g., UK Biobank, All of Us). The authors emphasize that future work must integrate these sex-differential signals with established disease-associated variants to create clinically actionable tools.