What's Happening?
South Asians are significantly underrepresented in global health databases, with less than 1% participation in genome-wide association studies between 2005 and 2025, compared to over 86% of European ancestry. This data gap is critical as advances in artificial
intelligence and machine learning increasingly rely on vast genomic and health data to detect diseases, predict risks, monitor patients, and tailor treatments. Integrated biobanks, such as the U.K. Biobank, have transformed biomedical research by combining genomic information with electronic health records, environmental exposures, and lifestyle data. However, the lack of diverse data means that tools built on European-heavy datasets are less accurate for South Asian populations, who face higher rates of conditions like type 2 diabetes, cardiovascular disease, and asthma. Experts like Bhramar Mukherjee of Yale School of Public Health emphasize that this neglect denies South Asians the opportunity to attain maximal health.
Why It's Important?
The underrepresentation of South Asians in global health databases has profound implications for medical research and public health. Polygenic risk scores, which estimate genetic risk for diseases, have been found to be less accurate when applied to South Asian populations. This inaccuracy stems from the fact that most predictions about gene expression and cell function are inferred from European datasets, limiting the ability to understand disease mechanisms and identify relevant drug targets for South Asians. Diagnostic thresholds, risk scores, and prediction models developed predominantly from European populations may not be suitable for South Asian individuals without validation and recalibration using local data. Furthermore, South Asia is a highly diverse region, and treating it as a single genetic block obscures important differences, as evidenced by the GenomeIndia Project which identified over 40 million genetic variants unique to the Indian population. This data gap ultimately translates into a health gap, where the populations most in need of accurate health tools are underserved.
What's Next?
To address this critical data gap, a recent perspective in The Lancet Regional Health – Southeast Asia proposes building greater regional collaboration among existing biobanks and cohorts in South Asia. The authors advocate for an infrastructure that ensures South Asian researchers and institutions retain a meaningful role in how their data are used, including safeguards such as co-authorship, joint intellectual property, technology transfer, training, and infrastructure support. The goal is to create a system where existing datasets can communicate, historically overlooked populations are included, and data-generating researchers share in scientific benefits. This collaborative approach is deemed vital to prevent the region from being excluded from the genomic revolution and to ensure that future AI models and clinical tools do not perpetuate existing biases at a larger scale. While the situation is challenging, experts believe it is not too late to make significant progress.
Beyond the Headlines
The issue of data diversity in global health databases extends beyond mere statistics, touching upon ethical and equity concerns. The historical concentration of research funding and infrastructure in high-income countries has left low- and middle-income countries, including those in South Asia, with inadequate resources for large-scale genomic studies. This imbalance dictates whose health problems are prioritized and whose evidence informs health policy. The lack of diverse data in AI models raises ethical questions about algorithmic bias and its potential to exacerbate health disparities. If AI tools are trained on skewed data, they will inevitably reproduce and amplify those biases, leading to less effective or even harmful health interventions for underrepresented groups. Addressing this requires not only scientific collaboration but also a fundamental shift in global health funding and research priorities to ensure equitable access to the benefits of genomic medicine and AI-driven healthcare.
