The Lancet Commission on 21st century global threats to health identifies 17 threats that could shape human health through 2100. The list reaches far beyond the diseases we usually discuss in public health.
I see this report as an important cue for action. It asks a fundamental question: Are we heading in the right direction?
What are the threats?
The table ranks the Commission’s 17 threats by cumulative expected amenable disability-adjusted life years (DALYs) between 2025 and 2100. One DALY is one year of healthy life lost through early death or disability. Here, “amenable” is the difference between plausible worse and better scenarios: it estimates health loss that action could avert, not the number of people affected now or a prediction of what will happen. Values are in billions of DALYs. The ranges show the Commission’s 95% uncertainty intervals.
| Rank | Threat | Amenable DALYs (billions) |
95% uncertainty interval (billions) |
|---|---|---|---|
| 1 | Obesity, including related diet and physical activity risks | 23.1 | 19.5–26.9 |
| 2 | Low educational attainment | 16.1 | 12.2–22.8 |
| 3 | Income inequality | 13.8 | 6.47–28.8 |
| 4 | Malicious use of AI, including biological weapons | 10.4 | 0–15.6 |
| 5= | Nuclear conflict and nuclear famine | 9.35 | 0–24.5 |
| 5= | Poor mental health | 9.35 | 6.34–13.0 |
| 7 | Unhealthy ageing | 8.26 | 6.38–10.6 |
| 8 | Conventional conflict | 7.71 | 6.58–9.14 |
| 9 | Smoking | 7.06 | 5.93–8.38 |
| 10 | High blood pressure | 4.79 | 3.74–6.25 |
| 11 | Antimicrobial resistance | 3.44 | 2.84–4.06 |
| 12 | Naturally occurring pandemics | 2.39 | 0.90–7.09 |
| 13 | High alcohol use | 2.35 | 1.97–3.06 |
| 14 | Climate change | 1.89 | 1.32–2.48 |
| 15 | Air pollution | 1.64 | 1.33–1.95 |
| 16 | Declining development assistance for health | 1.50 | 1.20–1.95 |
| 17 | Childhood malnutrition | 1.02 | 0.67–1.62 |
How to read the DALYs: One DALY is one year of healthy life lost, either through premature death or time lived with disability. For obesity, 23.1 billion DALYs means that, across the world and accumulated from 2025 to 2100, the Commission estimates a difference of 23.1 billion healthy life years between plausible worse and better scenarios for obesity and related risks. It represents potential health loss that better action could avoid. It does not mean 23.1 billion people have obesity, 23.1 billion deaths, or 23.1 billion DALYs lost in a single year.
Source: Lancet Commission; figure and data transcription. Ranks use point estimates; overlapping uncertainty intervals mean many nearby positions are not firmly separated. These burdens overlap and must not be added together. The AI and nuclear estimates rely heavily on expert judgement. Climate-related displacement and other difficult-to-measure harms may not be fully captured. This is a comparison of potential gains from action, not a fixed order of political priorities.
Obesity comes first by this measure, while childhood malnutrition comes last among these 17. That does not make a malnourished child less urgent, nor does the uncertain figure for nuclear conflict make prevention optional. The setting, the people affected and the type of action needed still matter.
AI and the loss of trust
The inclusion of AI stands out to me. Public health has often discussed AI as a tool for research, forecasting and care. The Commission also asks us to consider its malicious use, including the development of biological weapons. I am concerned about another route to harm: AI can make false health claims easier to produce and spread at scale, weakening trust in health authorities.
We saw during COVID-19 that health misinformation can affect decisions. A randomised study in the US and UK found that exposure to false claims about COVID-19 vaccines reduced people’s intention to be vaccinated. That study was about online misinformation, not generative AI. Today’s AI tools could make the same problem harder to contain.
The United States offers a warning about how important public trust and clear communication have become. In 2025, the CDC moved COVID-19 vaccination to shared clinical decision-making. In January 2026, it narrowed the vaccines recommended for all children, with others placed in categories for higher-risk groups or shared decisions. The vaccines were not removed from access or insurance coverage. We cannot claim that AI misinformation caused these policy changes. They do show why the evidence, the reasons for policy decisions and the way authorities speak with the public all matter.
When trust is lost, even a sound public health measure becomes harder to carry out. Leaders must protect the integrity of health information as seriously as they protect health services.
Finding our true north
For too long, we have tended to organise health around treating illness after it occurs. Treatment matters, of course. But the threats in this report show that health is also determined by what happens in our schools, economies, food systems, cities, environment and political decisions.
Our true north in public and planetary health should be clear: protect the conditions that allow people to live well, today and in the future. That means reducing preventable illness, protecting the natural systems on which life depends, and ensuring that progress does not leave the most vulnerable behind.
The threats are connected. Inequality shapes who is exposed to harm and who can recover. Climate change affects food, water and living conditions. Conflict destroys both lives and the systems needed to protect them. Education affects people’s opportunities and their ability to make informed decisions. We cannot expect the health sector alone to resolve problems created across society.
A warning to world leaders
World leaders should be concerned. Many of these threats cross borders, outlast election cycles and cannot be managed by a single ministry. Yet governments often respond only when a crisis is already visible.
The task is to act earlier. Leaders need to put health into decisions on education, agriculture, energy, technology, security and the economy. They need reliable information, institutions able to plan beyond the next few years, and the willingness to cooperate even when national interests differ.
The scale of the threats is sobering. But the report also points to something within our control: choices made now can change the amount of harm people experience. The future is not simply something we forecast. It is also shaped by what we decide to prevent.
Using uncertainty wisely
Some of the Commission’s estimates have wide uncertainty intervals, especially for events whose timing and consequences are difficult to predict. We should be honest about that uncertainty. A point estimate should not be treated as a precise prediction of what will happen decades from now.
But uncertainty is no reason to wait. We do not need an exact count of future deaths or lost healthy years to reduce smoking, improve education, prepare for pandemics, protect the environment or strengthen safeguards around dangerous technologies. For risks with potentially devastating consequences, uncertainty should prompt serious preparation.
What matters is whether an estimate helps us ask better questions, set priorities and test whether our actions are making a difference.
The direction we choose
I would read these 17 threats as a prompt to correct our course. Public health must look beyond individual disease. Planetary health must become part of everyday decisions, rather than a separate concern discussed only at international meetings. And leaders must be judged by whether they leave people and the planet better protected for the next generation.
The report gives us a warning. Our response will show whether we have understood it.
Source: The Lancet Commission on 21st century global threats to health (2026), https://doi.org/10.1016/S0140-6736(26)01714-9.