In plain language
Esophageal cancer in inland Brazil: what the study found
A summary for non-specialists, including residents of the municipalities mentioned. Versão em português.
By Clériston Almeida Capistrano, independent researcher · cledon.org
This is not a health alert. It summarizes an exploratory study based on public data. It raises questions for specialists to investigate; it does not say what causes the disease, does not measure anyone's risk and does not replace medical advice. Version 1.7 · November 14, 2026.
In short
- In 16 inland municipalities (9 in Minas Gerais, 5 in the South and 2 in Bahia), more people died of esophageal cancer than expected for their own state.
- The excess is not a one-off: for the 16 municipalities together, it has been present since 1996. From 2000 to 2014, years not used to select these municipalities, deaths were 1.9 times the expected number (almost double).
- Hospital records show the same pattern, and other cancers are not above expectation in these municipalities. This suggests something specific to esophageal cancer.
- The study did not find the cause. It shows where it is worth investigating habits already known to raise risk, such as very hot drinks, alcohol and smoking.
What the study did
We used the death certificates that Brazil's Ministry of Health publishes without names: 7,913,003 deaths from 2020 to 2024 in all 5,570 municipalities. For each municipality, we calculated how many cancer deaths would be expected if it followed its state's average, taking residents' age and sex into account (places with more older people naturally have more cancer). Then we compared the actual number with the expected one.
To avoid mistaking chance for a pattern, we required strong statistical evidence and checked the results in several ways: in other years (1996 to 2019), in the records of cancer hospitals, and against other types of cancer. We also set aside state capitals and municipalities with large hospitals, where the numbers are harder to interpret.
What we found
Esophageal cancer was the type with the most municipalities above expectation after these filters. In 12 of the 16 municipalities, the excess appeared again in 2015–2019, years not used to select them. In hospital records, the share of esophageal cancer among cancer cases in these municipalities is about 2.7 times their state's.
At the same time, in the same municipalities, deaths from lung cancer and all other cancers were below expectation (0.87 and 0.84 times). So it does not look like “more cancer in general”.
These are small municipalities, each with few deaths a year (from 2.2 to 7.4 on average). The number for any single municipality is imprecise; what matters is the set and the repetition over the years.
The 16 municipalities. The names appear because the data are public and already on the map. This is not a list of “dangerous places”.
| Municipality | State | Population | Deaths | Expected |
|---|---|---|---|---|
| Matina | BA | 10,330 | 12 | 2.4 |
| Palmas de Monte Alto | BA | 20,078 | 18 | 5.3 |
| Brasília de Minas | MG | 32,025 | 28 | 10.6 |
| Braúnas | MG | 4,441 | 11 | 2.0 |
| Capelinha | MG | 39,626 | 37 | 9.8 |
| Caraí | MG | 19,548 | 20 | 6.5 |
| Cipotânea | MG | 5,581 | 13 | 2.2 |
| Ervália | MG | 20,255 | 21 | 6.9 |
| Poté | MG | 13,666 | 19 | 5.2 |
| Santa Maria do Suaçuí | MG | 12,788 | 18 | 4.4 |
| Serro | MG | 21,952 | 22 | 6.7 |
| Mallet | PR | 13,428 | 17 | 3.9 |
| Prudentópolis | PR | 49,393 | 35 | 14.1 |
| Dom Feliciano | RS | 13,051 | 19 | 5.6 |
| Riozinho | RS | 4,473 | 13 | 2.1 |
| Itaiópolis | SC | 22,051 | 18 | 5.5 |
Esophageal cancer deaths from 2020 to 2024. Expected: how many there would be if the municipality followed its state average, given residents' age and sex. Population from the 2022 Census.
What this does not mean
- It does not mean that living in these municipalities causes cancer. The study compares places, not people. We do not know what the people who fell ill ate, drank or did.
- It does not measure your risk. The vast majority of residents will never have esophageal cancer.
- It is no reason to move or to panic. The precautions for these municipalities are the same as those recommended for all of Brazil (see below).
- Part of the increase may be better recording. In the 1990s, many deaths had no defined cause on the certificate (25.0% in these municipalities in 1996–1999, against 10.8% in 2010–2014). With better certificates, more cases get counted.
Why it might be happening
These are hypotheses, not conclusions. They come from comparing public characteristics of municipalities, which does not prove cause.
- In the South, the excess follows municipalities with more yerba mate extraction and tobacco farming. This fits older studies in the region on very hot maté (chimarrão), smoking and alcohol.
- In Minas Gerais, where chimarrão is rare, the excess follows more rural, lower-income municipalities that produce cachaça (sugarcane spirit). All 9 Minas Gerais municipalities on the list produced cachaça in 2006, against 64% of the state's municipalities. Producing cachaça is not the same as drinking it; that would need to be measured in people.
- More cases also appeared where more households use well or spring water. The study did not analyse the water. These households are more common in rural areas, and the association may reflect other aspects of rural life.
What is already known about esophageal cancer
According to Brazil's National Cancer Institute (INCA, in Portuguese), risk increases with very hot drinks (65 °C or more, such as scalding chimarrão, coffee or tea), alcoholic drinks (there is no safe level), cigarettes and other tobacco, excess weight and untreated reflux.
To protect yourself, INCA recommends: do not smoke, avoid alcohol, keep a healthy weight, treat reflux and drink hot beverages below 60 °C, waiting about five minutes after the water boils. The problem is the temperature, not the herb or the coffee.
See a health service if you have difficulty or pain swallowing, a feeling that food gets stuck, pain in the middle of the chest or unexplained weight loss. Early on, the disease usually causes no symptoms.
Frequently asked questions
I live in one of these municipalities. Should I worry?
There is no reason for alarm. The study does not measure anyone's risk. The same care recommended to everyone applies: do not smoke, avoid alcohol, do not drink scalding beverages, and see a health service if you have the symptoms above.
Can I keep drinking chimarrão?
What raises risk is a very hot drink, not the herb. Keeping the water below 60 °C (not boiling, or letting it cool a little) reduces that risk, according to INCA and the International Agency for Research on Cancer (IARC).
Is the water in my town dangerous?
The study did not test the water anywhere. It only saw that the excess is larger where more households use wells or springs, which is typical of rural areas. Questions about water quality should go to the municipal health department (health surveillance), which monitors drinking water.
My municipality is not on the list. Does that mean there is no problem?
Not necessarily. The study only flags municipalities with strong statistical evidence. Very small municipalities, capitals and places with large hospitals may have an excess without appearing on the list. The map shows every municipality.
Who did the study? Was it reviewed?
It was carried out by an independent researcher with no formal training in health sciences, with help from artificial intelligence for coding, writing and review. The data, code and results are open, and anyone can redo everything with a single command. The study has not yet been reviewed by independent specialists, which is what we are seeking now.
What happens next?
The next step is to take these results to epidemiologists and cancer services in the regions. Confirming a cause requires a study that talks to people, asking about drink temperature, alcohol, smoking, diet and wood-burning stoves. That can only be done with research institutions.
Learn more
- Cancer map of Brazil, with every municipality and cancer type.
- Technical reports: Phase 1 (signals), Phase 1b (hospital records and trends), Phase 1c (1996–2014), Phase 2 and Phase 2b (municipal characteristics).
- Deposited data, code and reports: doi.org/10.5281/zenodo.24195057.
- Official information on the disease: INCA, esophageal cancer (in Portuguese).
- Found an error or want to help? contato@cledon.org. Please do not send test results or personal health information.