Background Escalation of ultra-processed foods (UPFs) sales has been recorded in low-to-middle-income countries, including in South Asia. However, individual consumption levels and sociodemographic characteristics influencing UPF consumption remain largely unknown in South Asia. We aimed to quantify UPF consumption and investigate its sociodemographic correlates in South Asia.

Methods We analysed data from 60,714 participants in the South Asia Biobank that recruited adults in Bangladesh, Pakistan, Sri Lanka, North India, and South India. Dietary assessment was conducted using interviewer-led 24h dietary recalls with a South Asia-specific digital tool. Foods were classified by the degree of industrial processing using the NOVA classification. Adjusted two-part multivariable regression models examined associations between sociodemographic factors and any UPF consumption and quantity of UPF consumption in consumers.

Results In Bangladesh, Sri Lanka and North India, approximately 75% of the participants reported consuming any UPFs in the previous 24h while in South India and Pakistan this was 40%. Median contribution of UPFs to total energy among UPF consumers ranged between 17% in Pakistan, 15% in North India, and 13% in Bangladesh, Sri Lanka, and South India. Biscuits were a common source of UPF across all regions. Other commonly consumed UPFs among consumers included sweetened beverages in Pakistan, packaged salty snacks in South India, and breakfast cereals in Bangladesh. Diverse associations between sociodemographic factors and any UPF consumption were seen across regions. Younger age was associated with any UPF consumption in Pakistan and Sri Lanka whereas in Bangladesh and North India, older age was. In all regions except Bangladesh, female sex was associated with any UPF consumption. Higher education was associated with UPF consumption in Bangladesh (odds ratio 2.01; 95% confidence interval 1.71 to 2.35), Pakistan (1.69; 1.55 to 1.85), and North India (1.40; 1.13 to 1.73). Paid employment was not associated with UPF consumption in any region. Among UPF consumers, in all regions, UPF consumption was lower in married or cohabitating than in single people. In Bangladesh and Sri Lanka UPF consumption was higher in rural residents, while in Pakistan, consumption was higher in urban participants.

Conclusion Younger age, female sex, higher education, employment, and income exhibited associations with UPF, but this varied across South Asia. This heterogeneity should be considered when developing regionally specific interventions to support dietary public health. Our findings of regional consumption of specific UPFs, such as biscuits, breakfast cereals, sweetened beverages, and salty snacks, provide valuable insights for targeted interventions.