Evaluation of the Contribution of Natural Selection to Greater Cardiometabolic Disease Risk in South Asian Populations

This study suggests that while ancient or polygenic selection on standing variation likely contributed to the higher cardiometabolic disease risk in South Asian populations compared to Europeans, recent selective sweeps and gene-gene or gene-environment interactions are unlikely to be the primary drivers of this disparity.

The Gist

Imagine two groups of people, let's call them the "European Team" and the "South Asian Team." Both teams play the same game: living life. But when it comes to a specific set of health challenges—like heart trouble and Type 2 diabetes—the South Asian Team seems to get hit harder and faster, even when they are living in the same neighborhoods and eating similar foods as the European Team.

Scientists have long wondered: Is this because of their genes?

This paper is like a team of genetic detectives trying to solve that mystery. They wanted to know if natural selection (nature's way of editing the human "source code" over thousands of years) accidentally left the South Asian Team with a genetic setup that makes them more vulnerable to these diseases today.

Here is how they solved the case, using some simple analogies:

1. Checking the "Instruction Manuals" (Genetic Effects)

First, the detectives had to make sure the "instruction manuals" (genes) actually work the same way in both teams. If a gene says "store fat here," does it mean the same thing for a European and a South Asian person?

• The Finding: They checked thousands of genes and found that, for the most part, the instructions are identical. A gene that increases body fat in one team does the same thing in the other.
• The Analogy: Imagine two cars, a Ford and a Toyota. If you press the gas pedal in both, they speed up. The paper found that the "gas pedals" (genes) work the same way in both populations. This rules out the idea that the genes themselves are broken or work differently in South Asians.

2. Looking for "Evolutionary Scars" (Signatures of Selection)

If the genes work the same, why are the disease rates different? The detectives looked for "scars" left by evolution.

Imagine a long line of people waiting for a bus.

• Hard Sweeps (The "Stampede"): Sometimes, a new, super-useful mutation appears, and everyone rushes to grab it. This leaves a very distinct, long "scar" in the genetic line because everyone is carrying the exact same version of that gene. The paper looked for these "stampedes" using a tool called XP-EHH.
  • The Result: They found no stampedes. There were no signs of a sudden, recent rush to grab a specific gene that caused the difference.
• Soft Sweeps (The "Slow Drift"): Sometimes, nature doesn't pick a single new gene. Instead, it slowly nudges the frequency of many existing genes over a very long time. It's like a gentle breeze slowly pushing a pile of leaves in one direction, rather than a hurricane. The paper looked for this using a tool called FST.
  • The Result: They found clear evidence of this gentle breeze. Specifically, genes related to trunk fat (fat around the middle) and Type 2 diabetes showed signs that nature had slowly shifted their frequencies over thousands of years.

3. The "Hair Color" Comparison

To make sure their tools were working, they compared the disease genes to something we know for sure changed quickly: Black hair color.

• We know Europeans and South Asians have different hair colors because of strong, recent evolution.
• The tools successfully found the "hard sweep" scars in the hair color genes (the "stampede" happened there).
• But for the diabetes and fat genes, the tools only found the "gentle breeze" signals, not the "stampede" signals.

4. The Big Conclusion: The "Thrifty" History

So, what does this mean?

The paper suggests that the higher risk of heart disease and diabetes in South Asians isn't because they have "bad" genes or genes that work differently. Instead, it looks like nature slowly tweaked their genetic makeup over a very long time.

• The Metaphor: Imagine a family that lived in a place where food was scarce and unpredictable for thousands of years. Nature might have slowly selected for genes that helped them store fat very efficiently (like a squirrel storing nuts for winter). This was a survival advantage back then.
• The Twist: Today, in a world with plenty of food, that same "efficient storage" trait is now a liability, leading to diabetes and heart issues.
• The Evidence: The paper found that this "slow tweak" (soft selection) happened specifically for genes that control fat distribution (especially fat around the trunk) and diabetes risk.

What the Paper Does Not Say

• It does not say that South Asians are "genetically doomed." It just says their genetic history was shaped by different environmental pressures in the past.
• It does not say that lifestyle changes (diet, exercise) won't help. In fact, knowing the history helps explain why the body reacts the way it does, but it doesn't change the fact that modern habits matter.
• It does not claim that all the risk is genetic. The paper acknowledges that rare, unique genes and environmental factors still play a part, but the "common" genes shared by everyone show this specific evolutionary history.

In a nutshell: The South Asian population carries a genetic "legacy" of efficient fat storage, likely shaped by ancient environments. Nature didn't suddenly change their genes recently (no "stampede"), but it slowly adjusted them over millennia ("gentle breeze"). Today, that ancient survival trick is unfortunately a risk factor for modern diseases.

Technical Summary

Technical Summary: Evaluation of the Contribution of Natural Selection to Greater Cardiometabolic Disease Risk in South Asian Populations

Problem Statement
South Asian populations exhibit significantly elevated risks for cardiometabolic diseases, including Type 2 Diabetes (T2D) and cardiovascular disease, compared to European populations. This disparity persists even after migration to Western countries and adjustment for environmental factors, suggesting a potential genetic basis. While previous studies have largely failed to identify specific risk alleles with divergent effect sizes between these groups, it remains plausible that natural selection acting on adiposity-related traits in the evolutionary past has shaped the genetic architecture of these populations, contributing to current disease susceptibility. The challenge lies in distinguishing true signals of polygenic adaptation from confounding factors such as population stratification, gene-environment interactions (GxE), and differences in linkage disequilibrium (LD) patterns.

Methodology
The study employs a multi-step genomic approach to evaluate whether natural selection has driven cross-population genetic differentiation between South Asians (SAS) and Europeans (EUR) for traits related to cardiometabolic disease.

1. Data Harmonization and Consistency Checks:

   • The authors utilized large-scale multi-ancestry GWAS meta-analyses (including the Type 2 Diabetes Global Genomics Initiative, GIANT, and UK Biobank) to minimize bias arising from using European-only discovery datasets.
   • Sentinel variants were selected via cross-ancestry fixed-effects meta-analysis.
   • A heterogeneity test (Cochran's Q) was performed to confirm that genetic effect sizes for sentinel variants were largely consistent between SAS and EUR populations. This step was crucial to rule out widespread GxE or gene-gene (GxG) interactions as the primary driver of differential disease prevalence.

2. Selection Statistics and Enrichment Tests:

   • FST Enrichment: The study tested for an enrichment of high $F_{ST}$ (fixation index) values at trait-associated loci compared to a null distribution of LD score and minor allele frequency (MAF) matched SNPs. This method, adapted from Guo et al., assesses whether allele frequency differences exceed expectations under genetic drift.
   • Haplotype-Based Statistics: To distinguish between different modes of selection, the authors calculated XP-EHH (cross-population extended haplotype homozygosity) and XP-nSL (cross-population number of segregating lengths by site). These statistics are sensitive to recent selective sweeps (hard sweeps on new mutations vs. soft sweeps on standing variation).
   • Controls: Height and black hair colour were included as positive controls. Height represents a polygenic trait with debated selection signals, while black hair colour is a known target of strong, recent directional selection in Europeans.

3. Directional Analysis:

   • Polygenic scores (PS) were calculated to determine the direction of genetic differentiation, comparing mean scores in SAS and EUR populations against the combined mean.

Key Results

• Genetic Effect Consistency: After adjusting for multiple testing, genetic effects for adiposity traits (Trunk Fat Mass, Trunk Fat Percentage, Body Fat Percentage, BMI) and T2D were found to be largely consistent between South Asian and European populations. Only one BMI-associated SNP (rs1229984 in ADH1B) showed significant heterogeneity, which was partially explained by alcohol consumption interactions but did not account for all variance.
• $F_{ST}$ Enrichment:
  • Significant enrichment of $F_{ST}$ was observed for Type 2 Diabetes, Trunk Fat Mass (TFM), and Black Hair Colour.
  • Enrichment was also noted for Trunk Fat Percentage (TFP) and Body Fat Percentage (BFP), though these did not survive strict multiple testing correction in all sensitivity analyses.
  • No significant enrichment was found for BMI, Appendicular Lean Mass (ALM), or Height.
• Haplotype-Based Statistics:
  • No significant enrichment was found for XP-EHH or XP-nSL for any of the cardiometabolic or adiposity traits.
  • In contrast, Black Hair Colour showed robust enrichment for both XP-EHH and XP-nSL, confirming the method's ability to detect recent, strong selective sweeps. Height showed modest XP-nSL enrichment that did not replicate in sensitivity analyses using South Asian reference panels.
• T2D Clusters: When analyzing T2D-associated SNPs grouped by mechanistic clusters (Smith et al.), significant $F_{ST}$ enrichment was most pronounced in the Lipodystrophy 2 cluster, as well as Lipodystrophy 1 and Liver-Lipid clusters (though the latter two did not survive multiple testing correction).
• Direction of Differentiation: Polygenic score analysis indicated that South Asian populations have higher mean genetic scores for fat distribution traits, BMI, and T2D compared to the combined mean, while Europeans have higher scores for Height and ALM.

Key Contributions

1. Evidence for Adaptive Differentiation: The study provides empirical evidence that the cross-population genetic differentiation in T2D and regional fat distribution (specifically trunk fat) is consistent with natural selection, rather than solely genetic drift.
2. Mechanism of Selection: The absence of enrichment in haplotype-based statistics (XP-EHH/XP-nSL) alongside the presence of $F_{ST}$ enrichment suggests that the observed differentiation is not driven by recent, strong "hard" selective sweeps. Instead, the results support a model of ancient selection or recent polygenic selection acting on standing variation (soft sweeps), where subtle, coordinated allele frequency shifts occurred over longer timescales.
3. Regional vs. General Adiposity: The findings highlight that selection signatures are stronger for regional fat distribution measures (trunk fat) than for general measures (BMI/Body Fat Percentage), aligning with clinical observations that visceral adiposity is a primary driver of cardiometabolic risk in South Asians.
4. Common vs. Rare Variants: The results suggest that a substantial component of the excess cardiometabolic risk in South Asians is attributable to cosmopolitan genetic variants (shared across ancestries) that have undergone differentiation, rather than being driven exclusively by rare, ancestry-specific variants.

Significance and Claims
The paper claims that natural selection has played a moderate role in shaping the genetic architecture of fat distribution and T2D risk in South Asian populations relative to Europeans. Specifically, the authors interpret the $F_{ST}$ enrichment without haplotype signatures as evidence that the increased disease risk is partly a byproduct of evolutionary pressures acting on adiposity-related traits in the past, likely involving soft sweeps on standing variation.

The authors explicitly state that while these findings support a genetic basis for the disparity, they do not quantify the strength of selection or the exact evolutionary timescale. They caution that the results are consistent with, but do not prove, that these evolutionary pressures contribute to the current higher disease burden. Furthermore, the study emphasizes that while common variants show signs of differentiation, rare ancestry-stratified variants likely also contribute to the risk, necessitating larger South Asian-specific GWAS and whole-genome sequencing efforts to fully resolve the genetic architecture of cardiometabolic disease in these populations.