NOVA was not built for Asia
The debate over ultra-processed food has arrived in Southeast Asia at speed. Governments from Manila to Singapore are reaching for classification tools and nutrient thresholds to address rising rates of obesity, type 2 diabetes, and diet-related chronic disease.
The science behind the concern is real. Research published in ScienceDirect in 2025 found that each ten per cent increase in ultra-processed food consumption is associated with a thirteen per cent increased risk of developing diabetes. The health trajectory in the region is serious, and governments are right to treat it as such.
The problem is the tool they are reaching for. The NOVA classification system, which underpins most current UPF policy discussion, was developed at the University of São Paulo in Brazil in 2009. It groups all food into four categories based on the degree and purpose of industrial processing, placing anything containing certain additives or modified ingredients into its highest-risk category, Group 4. The epidemiological work built on NOVA is substantial. The classification system itself, however, was not designed with Asia in mind, and applying it here produces results that confuse rather than clarify.
This is, in part, a belief problem. The idea that ultra-processed food constitutes a coherent, scientifically settled category has spread faster than the evidence supports. Policymakers, health advocates, and media have converged on NOVA as if the debate were closed. It is not. That gap between what the science actually shows and what the policy world believes it shows is where poor regulation is born, and Southeast Asia is now in its path.
Consider tofu. A protein staple eaten across East and Southeast Asia for well over a thousand years, tofu is routinely classified as a Group 4 ultra-processed food under NOVA when commercial varieties contain thickeners or flavourings. Tempeh, a fermented soybean product central to Indonesian diets, faces the same problem. Miso, fermented fish pastes, and canned products containing monosodium glutamate, an additive with deep roots in Asian culinary tradition, can all trigger NOVA's ultra-processed designation. A system that cannot reliably distinguish between a century-old fermented staple and an industrially formulated snack product is not ready for policy deployment across cultures it was never tested against.
A Brazilian framework applied to Filipino sardines
This matters at a practical level. In the Philippines earlier this year, industry analysis found that up to 95 per cent of commercially available packaged products would fail the Philippine Nutrient Profile Model thresholds as proposed, including affordable everyday staples such as canned sardines and instant noodles. That model has since moved from proposal to policy. The National Nutrition Council formally adopted it on 10 June as the national reference for identifying foods high in nutrients of concern, and public health coalitions are urging government to fast-track its application to warning labels, procurement rules, and taxes. A labelling or taxation regime that designates the majority of a low-income consumer's food supply as categorically unhealthy is not a public health tool. It is a blunt instrument pointed at the wrong target. And once deployed, the belief that it represents serious public health policy is hard to dislodge, regardless of what the evidence subsequently shows.
The academic community is already grappling with the science. A University of Copenhagen-led project due to report at the end of 2026 is working to restructure NOVA, with the explicit aim of incorporating nutritional content and food matrix analysis rather than processing level alone. The Institute of Food Technologists has noted that NOVA's failure to account for ingredients as a determinant of health impact represents a significant gap. These are scientific objections from within the research community, not industry lobbying. But they are not cutting through, because the belief has already formed and counter-evidence struggles to shift it once it has.
None of this should be read as a case for inaction. Two-thirds of Asia-Pacific consumers say they are worried about ultra-processed foods, and they are not wrong to be. Consumption is rising, particularly among younger and more urbanised populations. Diet-related disease has followed. The case for clear, evidence-based consumer information is strong. The problem is not the intention behind UPF regulation. It is the gap between the belief driving it and the evidence that should underpin it.
Belief is driving the bus. Evidence is running to catch up
What Southeast Asian consumers need is a framework built on their diets, their food cultures, and their specific nutritional challenges, not one retrofitted from a Brazilian research context. It would account for the nutritional role that fortified and processed foods play where micronutrient deficiency remains a genuine concern. It would treat fermented and traditionally processed products differently from industrially formulated ones. It would be calibrated against local dietary data rather than thresholds designed for a different hemisphere.
Building that framework requires health ministries, nutrition researchers, consumer groups, and the food industry to work from the same evidence base. Parts of industry are asking for exactly that. Food Industry Asia chief executive Matt Kovac has cautioned that the Philippine model is not a silver bullet, nor a one-size-fits-all solution. Regulatory harmonisation anchored to Codex Alimentarius gives the region a starting point, but Codex was not designed to replace locally-grounded nutritional research either.
The UPF debate in Southeast Asia is real and worth having. The region deserves better than a classification system whose Western-centric limitations academic researchers have been documenting for years. Getting the science right before locking in the policy is not a reason to delay. It is the only way to make the policy work. In the Philippines the classification is now settled; the implementing rules on labels, marketing, and taxes are not. That is where the evidence still has time to catch up. Diagnosis before prescription. That principle applies in medicine. It should apply in food regulation too.
