In Short

Nitrate is a substance that enters groundwater — and from there, in part, drinking water — through the use of fertilizer in agriculture. A large Danish long-term study of 1.74 million people showed that higher nitrate levels in drinking water are linked to a slightly increased risk of colorectal cancer. In this paper we apply these findings to Germany: How many people here are affected, how many additional colorectal cancer cases might result, and what would it cost to fix the problem at its source? Our result: in Germany, around 3,060 colorectal cancer cases per year could be linked to nitrate in drinking water. Depending on the calculation model used, the resulting cost to society ranges between €900 million and €3.3 billion per year. Preventive measures in agriculture would be far cheaper than fixing the problem afterward with expensive technology at the waterworks.

1. What Is This About?

When farmers fertilize their fields, some of the nitrogen enters groundwater as nitrate — and from there often ends up in drinking water too. In Germany, nearly one in two people (about 45% of the population) drink water with a nitrate level above 10 mg per liter. Recent research suggests that even at this level, a slightly increased risk of colorectal cancer may already be present ^{1 2}.

Several studies have already tried to weigh the costs of stricter nitrate regulation against the potential savings for society ^{3 4 5}.

The most important basis for our calculation is a large Danish study (Schullehner et al., 2018) that followed 1.74 million adults over 33 years. It found that a statistically measurable link to colorectal cancer is detectable starting at a nitrate level of just 3.87 mg per liter — well below the permitted EU limit of 50 mg/L. A follow-up study (Jacobsen et al., 2024) calculated what this costs Danish society: around 317 million US dollars per year.

We apply this calculation method to Germany and estimate how many additional colorectal cancer cases and what costs could arise here as a result of nitrate in drinking water.

2. How Did We Go About This?

2.1 Where Does Our Data Come From?

Unfortunately, there is no official database for the whole of Germany that collects all drinking-water nitrate levels in one place. An overview page does exist with links to the testing bodies of the federal states ⁶, but we found a publicly accessible database with individual measurements only for Rhineland-Palatinate ⁷. However, that database does not allow measurements to be matched to specific supply areas.

That is why we have been collecting the data ourselves since 2016: we search the analyses published online by water utilities and enter the values, together with the relevant supply areas, into our own database (the “Wassertipps database”).

To do this, we linked postal-code areas with the utilities' drinking-water analyses. Because many utilities operate several smaller supply zones within a single postal code, we additionally mapped these zones using GIS (geographic information system) software.

For population figures per postal-code area, we initially used the 2011 census, later updated with the 2022 census ⁸.

This allows at least an approximate estimate of how many people are exposed to which nitrate levels in their drinking water.

Where a water utility published only an average value, we used that value. Where only a minimum and a maximum measured value were given, we calculated a mean from them.

Our data is not entirely complete: we have drinking-water data for about 80 million people in Germany, but for around 6 million of them the nitrate value specifically is missing.

That leaves data for around 74.2 million people in Germany for whom we have concrete nitrate values. The data covers the period from November 2024 to May 2026 ⁹.

One possible source of error: within individual postal-code areas there are sometimes smaller supply zones with different values, for which we do not have exact population figures. We looked at these cases more closely on their own ¹⁰.

This affected around 1,000 postal-code areas (about 20% of all the areas we recorded). Of these, however, only a small number were actually relevant to our calculation — 124 and 109 areas respectively, covering 1.57 and 1.43 million residents — where the value for the overall area and for the smaller zone fell on different sides of our thresholds (10 or 25 mg/l). We excluded these unclear cases from the calculation to stay on the safe side ¹¹.

According to our data, 32,538,291 people live with a nitrate level above 10 mg/l in their drinking water, of whom 2,985,085 are even above 25 mg/l.

2.2 How Many People Do We Consider, and How Many Colorectal Cancer Cases Are There?

We work with 71 million people in Germany for whom we have sufficiently good data (about 85% of the total population of 84 million). Of these, 32,538,291 people (45.8%) have more than 10 mg/l of nitrate in their drinking water, and 2,985,085 people (4.2%) have even more than 25 mg/l.

The Robert Koch Institute reported 56,280 new colorectal cancer diagnoses across Germany for 2022 ¹². Since we only consider part of the population, we scaled this figure down accordingly.

2.3 How Do We Know How Large the Risk Is?

Our most important basis is a European analysis from the MARCHES research project ¹³. Experts there re-examined seven existing review studies and the eleven individual studies underlying them in detail — and found significant errors: in two analyses, figures had been transcribed incorrectly, making the association appear weaker than it actually is. Others had accidentally counted the same study group twice. After correcting these errors, all analyses consistently show a statistically meaningful association.

The corrected result is this: for every additional 10 milligrams of nitrate per liter, the risk of colorectal cancer rises by about 6 percent (technical term: relative risk of 1.06). The authors themselves consider this figure to be conservatively calculated, because in the lower range — up to roughly half the permitted limit — the risk rises more steeply, by around 15 percent per 10 mg/l. At very high levels the curve then flattens out again, though there is only limited data in that range, so this part remains uncertain.

We deliberately chose not to run a second, parallel calculation using the figures from the Danish study: that study is itself part of the European analysis, and its values are in any case very close to the combined result. Such a duplicate calculation would only create a false impression of added precision.

2.4 How Do We Calculate the Number of Additional Cases?

We use a recognized statistical method called the “Population Attributable Fraction” (PAF), which calculates what share of all colorectal cancer cases is probably attributable to nitrate exposure, given how many people are affected and how much the risk increases. Multiplying this share by the total number of colorectal cancer cases (47,570) gives the estimated number of nitrate-related cases.

We take each exposure level into account individually: someone with 3 mg/l in their water carries a smaller additional risk than someone with 30 mg/l. This is more precise than lumping everyone above 10 mg/l together.

A worked example illustrates this. Take the group with 10 to 25 mg/l in their drinking water — the largest exposed group in Germany. Their average level is about 17.5 mg/l.

We know that the risk rises by 6 percent for every 10 mg/l. At 17.5 mg/l, that is 1.75 times this step — which works out to a roughly 10.7 percent increased risk compared with someone with nitrate-free water. This group comprises 29.4 million people, or 39.7 percent of the population we consider.

Combining the two gives this group's contribution: 39.7% × 10.7% = 0.0426. In plain terms: if everyone were exposed at the level of this group, the overall rate of colorectal cancer would be about 4 percent higher.

We do the same calculation for every exposure level and add up the contributions:

From this total, a final percentage is calculated. The formula is 0.0688 divided by (0.0688 + 1) — which gives 0.0644, or 6.44 percent. This conversion is necessary because increased risks cannot simply be added together; they are spread across a correspondingly larger total number of cases.

The final step: 6.44 percent of 47,570 colorectal cancer cases gives 3,063 cases per year that are mathematically attributable to nitrate in drinking water. Well over two-thirds of these (around 1,900) come from the group with 10 to 25 mg/l — simply because that group is home to so many people.

2.5 What Does Treating Colorectal Cancer Cost?

For treatment costs, we rely on a German study using real health-insurance data from over 2,800 patients ¹⁴. Costs differ depending on the phase of the disease: the first year after diagnosis, the following years, and — if the disease is fatal — the final year of life. Adding these together and weighting by the survival rate (63% survive at least 5 years) gives an average of €40,879 per case.

An older study from 2014 ¹⁵ arrived at €61,520 per case. We use the newer, more up-to-date figures.

2.6 How Do We Put a Monetary Value on a Life Year Saved?

To calculate the cost to society, lost life years also need to be expressed in monetary terms — this is standard practice in health economics and is known as the “value of a life year” (VOLY). We use two different, well-documented values to show a realistic range:

(1) The EU standard value (€40,000): This value is used by the European Commission for air-quality calculations and comes from a large survey conducted in nine countries ¹⁶.

(2) A value specific to Germany (€173,868): This value comes from a study by the German Cancer Research Center (DKFZ), which evaluated many surveys on how much people are actually willing to pay for an additional year of life ¹⁷. This value is significantly higher than the EU standard value but compares well with other international studies.

Total costs per case consist of: treatment costs (€40,879) plus the value of lost life years plus other costs such as care and lost earnings (€22,247).

3. What Are the Results?

3.1 How Many Additional Colorectal Cancer Cases Result from Nitrate?

Our main calculation gives: around 3,063 colorectal cancer cases per year in Germany could be linked to nitrate in drinking water. Because the underlying risk estimate itself carries some uncertainty, the statistical range runs from 0 to 5,429 cases.

Where does this figure come from? Well over two-thirds of it (around 1,900 cases) comes from the group between 10 and 25 mg/l — that is 29.4 million people in Germany. The more highly exposed groups together account for about one-fifth, and the most heavily exposed group (over 50 mg/l) for almost nothing. This matters because the evidence base is thinnest exactly for the high values — so the core of our estimate rests on the better-supported range.

More relevant for policy than the total figure, though, is a different question: how many cases could be avoided if the legal limit were lowered? We calculated this for three target values (Table 3).

Halving today's EU limit of 50 mg/l to 25 mg/l would affect 4.7 million people — and would mathematically avoid only 147 cases a year. At a target of 10 mg/l, 34.2 million people would be affected and 1,248 cases avoidable. At the 6 mg/l recommended by the expert panel, it would be 45.9 million people and 1,821 cases. At the Danish reference value of 9.25 mg/l, it would be 1,344 cases.

The reason for this large difference is simple: only 4.2 percent of people in Germany live above 25 mg/l, but 45.8 percent live above 10 mg/l. A rule that targets only the most heavily contaminated areas therefore achieves less than a tenth of what would be possible overall. The expert panel's report gained Europe-wide political attention within just a few months: environmental organizations cite it in calling for the EU Drinking Water Directive's limits to be brought in line with current cancer research and for the precautionary principle to be applied ³².

3.2 What Does a Single Case Cost Society?

Depending on which value is used for a life year, very different totals result per case: around €293,500 with the EU standard value, and over a million euros with the Germany-specific value. This large difference arises mainly from how highly the loss of life years is valued financially.

3.3 What Is the Total Annual Damage?

For the main calculation (3,063 cases), the estimate ranges from €899 million per year (with the EU standard value) to €3,261 million per year (with the Germany-specific value).

3.4 Does Prevention Pay Off Financially?

Yes, clearly — especially if you address the cause rather than fixing it later. To calculate this, we compare the annual benefit (i.e., the avoided health costs) with the costs a reduction in nitrate would require. These cost figures come from a comprehensive study by the German Environment Agency (UBA) titled “Quantifying the Agriculturally Caused Costs of Securing Drinking Water Supply” ¹⁸, which we explain in more detail in the next section.

On the prevention side we deliberately work with a range, because the official cost figure is incomplete (more on this in the next section): using the commonly cited €112 million per year, every euro invested returns €8 to €29. Calculating more realistically with around €292 million per year, it is still €3.10 to €11.20. Opting instead for expensive technical water treatment, which the UBA estimates at €580 to €767 million per year, still pays off — but only with a return of about €1.20 to €5.60 per euro invested.

¹ Low cost assumption: administrative burden of the 2017 Fertilizer Ordinance only.
² More realistic cost assumption: the 2020 Fertilizer Ordinance including crop-yield losses.
³ If technical treatment were required instead (€580–767 million/year according to the German Environment Agency).

3.5 Comparison with Denmark

In Germany, the share of the population with a high nitrate level is far greater than in Denmark: 45.8% compared with under 10%. This explains why a significantly higher share of colorectal cancer cases here is likely attributable to nitrate (6.44% versus 1.4%).

Important for a fair comparison: the Danish figures answer the question of how many cases would be avoidable at a given limit — not how many are attributable to nitrate overall. They must therefore be compared with our limit-based calculations. At the same limit (9.25 mg/l), 72 avoidable cases in Denmark compare with 1,344 in Germany — roughly nineteen times as many. Scaled up to all of Germany and valued using the Germany-specific figure, this amounts to a total damage of around €3.3 billion per year.

4. What Does This Mean? — Context and Open Questions

4.1 Why We Consider the Higher German Value Important

We rely mainly on the Germany-specific value (€173,868 per life year), because it comes directly from surveys of what people in Germany would actually be willing to pay for an additional year of life. It is significantly higher than the EU standard value (€40,000), but fits well with values from other wealthy countries and is only slightly above the international median (€164,409).

This higher value shows: relying only on the lower EU standard value probably significantly underestimates the actual health damage. For policy decisions in Germany, we therefore consider the Germany-specific value more informative.

4.2 What Would It Cost to Solve the Problem?

The basis for our cost calculation is a 2017 study by the German Environment Agency (UBA) titled “Quantifying the Agriculturally Caused Costs of Securing Drinking Water Supply” ¹⁸. For the first time, experts systematically calculated what groundwater nitrate contamination actually costs the general public. To do so, they analyzed data from five model regions (including the Oldenburg-East Frisian Water Board, RheinEnergie in Cologne, and the Rhenish-Westphalian Waterworks Company) and additionally surveyed water utilities across Germany.

The central finding of this study: there are two fundamentally different ways to deal with excess nitrate — and between them lies an enormous financial difference:

Route 1 — Prevention in agriculture: If less fertilizer is used from the outset, the nitrate problem never arises in the first place. A figure from the Federal Ministry of Food and Agriculture is often cited for this: €111.7 million per year. We looked at this figure more closely — and it falls short.

It covers only the administrative and compliance burden of the 2017 Fertilizer Ordinance — things like larger manure storage, additional soil sampling, and documentation requirements. What is completely missing is the actually most expensive consequence: crop-yield losses from reduced fertilizer use. For the revised 2020 Fertilizer Ordinance, the Ministry itself later quantified these losses — at €208 million per year, in addition to about €84 million in administrative costs ¹⁹. Realistically, then, one should expect around €292 million per year, roughly two and a half times as much.

That is why we show both variants in Table 6. Even with the higher, more realistic figure, prevention still clearly pays off: instead of €8 to €29, the return is €3.10 to €11.20 per euro invested. In fairness, the counterargument should also be mentioned: the fertilizer rules do not just protect our drinking water — they simultaneously protect rivers and lakes, air, climate, and biodiversity. The colorectal-cancer benefit we calculate here is therefore only part of what these costs achieve overall — our calculation is, in that sense, more cautious than optimistic.

Route 2 — Treating the water afterward at the waterworks: If, instead, one waits and only later removes nitrate from the water using technical plants, the German Environment Agency estimates this at €580 to €767 million per year — roughly five to seven times the cost of prevention. The range results from how strict a target value is set and how favorable local conditions are. Water customers would foot this bill: in affected regions, water prices would rise by 55 to 76 cents per cubic meter (an increase of 32 to 45 percent) — up to €134 more per year for a family of four.

Important to understand: this figure of €580 to €767 million is a projection for the case that nationwide treatment would be required — not the sum of costs actually being incurred today. In the five model regions studied by the German Environment Agency, none of the utilities was operating nitrate treatment at the time of the study; instead, they relied on workaround strategies such as relocating wells or blending contaminated with uncontaminated water.

Plants dedicated specifically to nitrate removal do exist in Germany, however — they are just rare. The best-known example is in Aschaffenburg: a biological denitrification plant has been operating there since December 2000, lowering the nitrate level of the extracted groundwater from around 52 milligrams per liter to about 25³³. At this scale, the plant is unique in Germany. In recent years, new technical approaches have also been advanced: Veolia's CARIX ion-exchange process now runs in more than 20 German waterworks³⁴, and in Groß-Umstadt in Hesse, the TZW: DVGW Water Technology Centre has worked with the town to develop a reverse-osmosis concept for nitrate removal that has been in routine operation since 2021³⁵.

The fact that such plants remain the exception is mainly for practical reasons: biological denitrification is the only process that removes nitrate specifically, but it is considered demanding to operate and requires elaborate post-treatment of the water. Where utilities do treat water today, they therefore usually use reverse osmosis or ion exchange — processes that remove other substances along with nitrate.

For context: according to the 2024 Nitrate Report, 25.6 percent of monitoring sites specifically located in agriculturally dominated areas exceed the permitted limit of 50 mg/l. Across the larger monitoring network of over 8,000 sites, the figure is 15 percent ²⁹. Around three-quarters of the total possible costs fall on the preventive route — making it clearly the cheaper option. The German Environment Agency has been making the same point for decades: “Prevention is cheaper than repair.”

Our calculation confirms this: depending on the cost assumption, prevention returns 3 to 29 times what it costs. Even the more expensive after-the-fact treatment still pays off if calculated realistically — only in the most unfavorable case (lowest life-year value, highest treatment costs) does it come out roughly cost-neutral.

One important caveat here: these comparison figures set two things side by side that are not quite equivalent. On the cost side are today's fertilizer rules — but how much these actually reduce nitrate levels is simply not known. The monitoring program meant to document exactly that was launched in 2019 and is still being built up. On the benefit side, by contrast, is the achievement of specific limit values, which would require stricter and more expensive measures. These figures should therefore not be read as an evaluation of the Fertilizer Ordinance, but as a comparison of scale: the avoidable health damage is many times larger than what even the most expensive agricultural measures currently under discussion would cost.

4.3 A Clever Dual Solution: Removing PFAS and Nitrate from Water at the Same Time?

Since January 2026, water utilities have also had to comply with strict PFAS limits — PFAS being persistent chemicals often referred to colloquially as “forever chemicals.” Because PFAS are already present in groundwater, many utilities have no option but technical treatment. This raises the question: could plants be built so that they solve both problems — PFAS and nitrate — at once, instead of requiring a second retrofit later?

The answer: only certain filtration systems (so-called membrane processes, such as reverse osmosis) can actually filter out both at the same time. The widely used CARIX process, which works well against nitrate, does not help against PFAS — different technology is needed there.

However, three caveats apply: membrane plants use more energy and are more expensive to operate. In addition, PFAS and nitrate often do not occur in the same regions — PFAS tends to cluster around industrial sites and airports, nitrate around intensive farming areas. And even the cleverest dual solution does not change the basic finding: preventing nitrate at the source remains cheaper in the long run than any form of technical after-treatment.

Still, wherever the obligation to treat for PFAS already exists, factoring in nitrate reduction as well can make sense — solving two problems with one investment.

4.4 Differences Between the German Federal States

Nitrate levels differ considerably by region and federal state (averaging between 1 and 15 mg/l). We did not factor these regional differences into our nationwide calculation separately.

4.5 What Our Figures Cannot Capture

1. We apply foreign research findings to Germany. This is the biggest weakness of our calculation. The risk values come from international studies, mainly from Denmark — there simply is no comparable German study. Whether the same relationships hold true here cannot therefore be verified. Compounding this, the starting situation is very different: in Denmark, fewer than 10% of people have more than 10 mg/l in their water, compared with 45.8% here. We are therefore calculating in an exposure range for which comparatively little data exists abroad. That said, because our main calculation is not based on a single study but on a synthesis of many studies, this objection is somewhat mitigated.

2. Our water data is a snapshot in time. It comes from the years 2024 to 2026. How heavily individual people were exposed over their entire lifetime cannot be reconstructed from this — even though that is exactly what matters for the development of cancer (see point 5).

3. Uncertain treatment costs: The figures we use (2012–2016) do not yet fully account for newer, more expensive cancer therapies. The true cost could lie somewhere between €40,879 and €61,520 per case.

4. The value of a life year may now be higher: Our Germany-specific value comes from older surveys (1995–2015). Given today's higher level of prosperity, it would probably be closer to around €210,000 rather than €173,868 — but we cautiously use the older, published value.

5. Colorectal cancer develops over many years: From the first changes in the bowel to a fully developed cancer, it often takes 10 to 25 years ²¹. The Danish study therefore also accounts for nitrate exposure 20 to 35 years before diagnosis. Our current data from 2024–2026 cannot capture this longer-past exposure — which makes our estimate more likely to be cautious than overstated.

6. We consider only a single disease. Colorectal cancer is currently the only health outcome for which the evidence base is strong enough to calculate reliably. For other possible consequences, however, there are indications that cannot yet be expressed in figures — for example, premature births, low birth weight, birth defects, and cancer in children. A review study also found a clear association with stomach cancer. Because none of this is included in our calculation, the damage we have estimated is more likely a lower bound ¹³.

7. The evidence base is especially thin at high nitrate levels. Most studies group their most heavily exposed category very broadly — for example, as “over 9.25 mg/l.” This category then contains people with both 10 and 50 mg/l, without any way of knowing the proportions. For an analysis, however, this group must still be assigned a numerical value — and that assignment is inevitably somewhat arbitrary. It is therefore currently not possible to say reliably how steeply the risk actually rises above 10 mg/l. For this reason we deliberately calculate using the cautious variant; the true effect is likely to be higher ¹³.

8. We don't know how effective the fertilizer rules actually are. On the cost side, we calculate using the measures of the Fertilizer Ordinance. However, how much these actually reduce nitrate levels in groundwater is not reliably documented anywhere. The stricter requirements apply only in certain designated areas in any case, and in several federal states these areas even became smaller when the boundaries were redrawn. The monitoring program meant to document the effect has been under development since 2019. On top of that, nitrate takes many years to reach groundwater — so changes in fertilizer use only show their effects over decades.

9. We are probably missing small water utilities. Our data comes from the published analyses of water utilities. Small utilities publish less often — so the 6.1 million people for whom we lack nitrate values likely belong predominantly to small supply areas. This matters because, across Europe, small utilities exceed the legal limit considerably more often than large ones (around 95% versus 99.7% compliance) and are mostly located in rural agricultural areas, where exposure is highest. Our figures are therefore more likely to underestimate the true exposure than to overstate it.

An intriguing idea that is not yet proven: does today's rise in colorectal cancer among young people have anything to do with nitrate?

The German Cancer Research Center has found that colorectal cancer among younger adults (ages 20 to 39) has increased over the past 20 years ²². In the United States this trend is even stronger. Researchers explain this partly by the fact that people born from around 1950 to 1960 onward appear to carry a higher risk — likely due to influences they were exposed to early in life. Coincidentally, this same period saw the massive expansion of synthetic fertilizer use in agriculture, which, with some delay, led to higher nitrate levels in groundwater. People who grew up from birth with higher nitrate levels in their drinking water are now reaching precisely the age at which the observed increase occurs.

Important: so far this is only a temporal coincidence, not proof. Nitrate is virtually never discussed in the scientific literature as a cause of early-onset colorectal cancer — the explanations usually put forward there are things like an unhealthy diet, obesity, lack of exercise, frequent antibiotic use, and changes in the gut microbiome. And these very factors also changed substantially over the same period — so the effects cannot simply be teased apart. Concluding a causal link from the coincidence of two trends over time would be a classic logical fallacy. We therefore present this idea explicitly as an open question for future research — not as evidence.

4.6 What Are the Policy Implications?

In June 2025, an important oversight instrument for agriculture was abolished (the so-called Nutrient Flow Balance Ordinance, or Stoffstrombilanzverordnung). At the EU level, the trend points the same way: since February 2026, certain processed manure-based fertilizers may be used even above the previous ceiling of 170 kilograms of nitrogen per hectare ³⁰. Germany's Federal Constitutional Court will hear arguments on 16 June 2026 on whether the abolition was lawful. Given our results — every euro invested in prevention returns, depending on the cost assumption, between €3.10 and €29 — this decision appears hard to justify from a health-economics standpoint ²³.

A look at Denmark shows what good research can achieve. The Danish Ministry of Environment convened an international expert panel in 2024 — prompted by exactly the study our own calculation is built on. In November 2025, this panel presented its report ²⁷: a slightly elevated colorectal cancer risk exists well below today's permitted limit of 50 mg/l. Its recommendation is therefore to lower the limit to 6 mg/l (technically: 1.35 mg/l of nitrate-nitrogen). For Denmark, the panel estimates this would avoid 57 to 90 colorectal cancer cases per year. This is backed by a new 27-year Danish study covering more than 2,200 cancer cases: it found no link between nitrate from food and colorectal cancer, but did find one between nitrate in drinking water and colon cancer starting at around 9 mg/l ²⁸. This undercuts the obvious objection that diet, not water, is really to blame ²⁴.

The comparison is revealing: in Denmark in 2024, around 90 percent of drinking water was under 5 mg/l ²⁵ — so the pressure to act there arises at an exposure level that only a minority reach here. If the recommended value of 6 mg/l were applied to Germany, our calculation suggests around 1,821 colorectal cancer cases could be avoided per year — roughly twenty to thirty times the Danish figure. At a less strict target of 10 mg/l it would still be 1,248 cases, while at 25 mg/l it would be only 147.

Important for context: the colorectal-cancer damage we have quantified here is only one specific piece of a far larger economic and environmental problem caused by over-fertilization. A study commissioned by the German Association of Energy and Water Industries (BDEW) puts the avoidable environmental costs — for water, air, health, and climate combined — resulting solely from the legally permitted excess fertilization (averaging around 25 kg of nitrogen per hectare too much) at roughly €3 billion per year ²⁶. The same study also explains, on technical grounds, why the current fertilizer rules fall short of protecting water bodies. Our calculated health costs should therefore be understood as a conservative lower bound within a considerably larger overall problem.

5. Conclusion

Based on our estimate, around 3,060 colorectal cancer cases per year in Germany are linked to nitrate in drinking water. The resulting cost to society is therefore around €3.3 billion per year. Preventive measures in agriculture would clearly pay off: with a 3- to 29-fold return on every euro invested, depending on how the cost of the fertilizer rules is assessed.

These figures should, however, be read with one important caveat. They rest on applying foreign research findings to Germany. Whether this is valid cannot currently be verified — especially since the starting situations differ so greatly: in Denmark, fewer than 10 percent of people are above 10 mg/l, while here it is nearly half. We are therefore calculating in a range for which comparatively little research exists.

Our figures are therefore not an official record but an estimate of the order of magnitude. And that is precisely their real purpose: to show that this question is important enough to finally be properly investigated in Germany. What is missing is a study modeled on the Danish example — one that tracks, for a large number of people, the nitrate levels they were exposed to over their lifetime, and links this to cancer-registry data. All the building blocks already exist: cancer registries, a dense network of drinking-water analyses, and — as this work shows — the ability to bring the two together spatially. What is missing is only the systematic combination of the two. There is another reason in favor of this, too: worldwide, research is short of data specifically on people with higher nitrate levels — in most countries, there simply are not enough of them. In Germany, however, 32.5 million people live with more than 10 mg/l in their drinking water. For comparison: in New Zealand, where intensive agriculture is likewise the main source, researchers estimate around 100 colorectal cancer cases and 41 deaths per year from nitrate in drinking water among a population of around 5 million ³¹. A German study could therefore close exactly the gap that exists internationally — its value would extend far beyond Germany.

Denmark shows where this can lead: first a long-term study, then a cost calculation, then an expert commission — and, at the end, the recommendation to lower the limit from 50 to 6 mg/l. In Germany, where far more people are affected, this first step is still missing. Regardless, what our calculation shows remains valid: groundwater protection should be strengthened — through more binding fertilizer rules and larger drinking-water protection areas. Because even in the most cautious calculation, the avoided health damage is greater than the cost of prevention.

Sources

• Krisch, T (2026). Nitrate in Drinking Water and Colorectal Cancer in Germany: A Population Attributable Fraction Analysis with Health-Economic Valuation https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7635458
• Picetti R, Deeney M, Pastorino S et al. (2022). Nitrate and nitrite contamination in drinking water and cancer risk: A systematic review with meta-analysis. Environmental Research, 210, 112988. https://pubmed.ncbi.nlm.nih.gov/35217009/ ↩
• Schullehner J, Hansen B, Thygesen M, Pedersen CB, Sigsgaard T (2018). Nitrate in drinking water and colorectal cancer risk: A nationwide population-based cohort study. International Journal of Cancer, 143(1), 73–79. https://onlinelibrary.wiley.com/doi/abs/10.1002/ijc.31306 ↩
• Temkin AM, Evans SR, Manidis T, Campbell C, Naidenko OV (2019). Exposure-based assessment and economic valuation of adverse birth outcomes and cancer risk due to nitrate in United States drinking water. Environmental Research, 176, 108442. https://www.sciencedirect.com/science/article/pii/S001393511930218X ↩
• Grinven, Rabl, de Kok (2010). Estimation of incidence and social cost of colon cancer due to nitrate in drinking water in the EU: a tentative cost-benefit assessment. Environmental Health, 9, 58. https://link.springer.com/article/10.1186/1476-069X-9-58 ↩
• Jacobsen R, Hansen B, Schullehner J (2024). Health-economic valuation of lowering nitrate standards in drinking water related to colorectal cancer in Denmark. Science of the Total Environment, 905, 167182. https://www.sciencedirect.com/science/article/pii/S0048969723059958 ↩
• Federal-state internet access to the listed testing bodies under the German Drinking Water Ordinance (TrinkwV). https://www.lgl.bayern.de/downloads/zqm/doc/internetzugang_untersuchungsstellen.pdf ↩
• Rhineland-Palatinate drinking water information system. https://wasserportal.rlp-umwelt.de/auskunftssysteme/trinkwasserinformationssystem ↩
• ArcGIS Hub, postal-code areas in Germany. https://hub.arcgis.com/datasets/esri-de-content::postleitzahlengebiete-in-deutschland/explore ↩
• List of surveyed postal-code regions in Germany with nitrate values. https://wassertipps.de/files/nitratexposition_de_0526.csv ↩
• List of surveyed sub-regions of the postal-code regions in Germany with nitrate values. https://wassertipps.de/files/nitratexposition_de_teilgebiete_0526.csv ↩
• List of sub-regions with deviating nitrate values. https://wassertipps.de/files/nitratexposition_de_ausnahmen_0526.csv ↩
• GEKID/RKI (2025). Cancer in Germany for 2021–2023. Centre for Cancer Registry Data, Berlin. krebs_in_deutschland_2025.pdf ↩
• Schullehner J, Donat-Vargas C, Villanueva CB (2025). Selected drinking water nitrate exposure-response functions with baseline incidences for EVA. MARCHES Deliverable D2.2, Horizon Europe Grant No. 101095430, 27 January 2025. Corrected exposure-response function: RR = 1.06 (95% CI 1.00–1.11) per 10 mg/L; steeper trend in the lower concentration range: RR ≈ 1.15 (1.08–1.26). MARCHES D2.2 (CORDIS) ↩
• Heisser T, Simon A, Hapfelmeier J, Hoffmeister M, Brenner H (2022). Treatment Costs of Colorectal Cancer by Sex and Age: Population-Based Study on Health Insurance Data from Germany. Cancers (Basel), 14(15), 3836. https://pmc.ncbi.nlm.nih.gov/articles/PMC9367511/ ↩
• Haug U, Engel S, Verheyen F, Linder R (2014). Estimating Colorectal Cancer Treatment Costs: A Pragmatic Approach Exemplified by Health Insurance Data from Germany. https://pmc.ncbi.nlm.nih.gov/articles/PMC3929363/ ↩
• Desaigues B et al. (2011). Economic valuation of air pollution mortality: A 9-country contingent valuation survey of value of a life year (VOLY). Ecological Indicators, 11(3), 902–910. https://www.sciencedirect.com/science/article/abs/pii/S1470160X10002116 ↩
• Schlander M, Schaefer R, Schwarz O (2017). Empirical Studies On The Economic Value Of A Statistical Life Year (VSLY) In Europe: What Do They Tell Us? https://www.researchgate.net/publication/320524786 ↩
• Oelmann M, Czichy C, Scheele U et al. (2017). Quantifizierung der landwirtschaftlich verursachten Kosten zur Sicherung der Trinkwasserbereitstellung [Quantifying the agriculturally caused costs of securing drinking water supply]. German Environment Agency (UBA) Texte 43/2017. UBA Texte 43/2017 (PDF) ↩
• Ordinance Amending the Fertilizer Ordinance and Other Provisions (2020). Bundesrat printed matter 98/20. Recurring compliance costs approx. €84 million/year; additional costs from crop-yield losses due to a 20% reduction in fertilizer requirements in nitrate-affected areas: €208 million/year. Bundesrat printed matter 98/20 (PDF) ↩
• Distribution of nitrate concentration by federal state, 2026. https://www.nitratdatenbank.de/trinkwasser/aktuelle-trinkwassersituation.html ↩
• Wan EYF, Yu EYT, Chin WY et al. (2023). The application of lag times in cancer pharmacoepidemiology: a narrative review. https://www.sciencedirect.com/science/article/pii/S104727972300090X ↩
• German Cancer Research Center (DKFZ) (2025). Colorectal cancer also rising among young adults in Germany — but clearly below US levels. Press release. DKFZ press release ↩
• BUND (2025). Statement on the abolition of the Nutrient Flow Balance Ordinance. Friends of the Earth Germany (BUND). BUND statement (PDF) ↩
• International expert group (2025). Evaluation of the parametric value for Nitrate in drinking Water. Report commissioned by the Danish Ministry of Environment, December 2025. Recommendation: lowering the nitrate limit from 50 mg/L to 6 mg/L. Expert group report ↩
• MARCHES project, Aarhus University (2025). International experts recommend lower limit value for nitrate in drinking water. News item, 18 December 2025. MARCHES news item ↩
• Taube F (2021). Expert assessment of the new fertilizer legislation (DüngeG, DüV, AVV GeA) of 2020 in Germany from the perspective of drinking-water protection. Study commissioned by the BDEW. BDEW study (PDF) ↩
• Halldorsson TI et al. (2025). Evaluation of the parametric value for Nitrate in drinking water. Report to the Danish Ministry of Environment and Gender Equality, 24 November 2025. ↩
• Erichsen DW et al. (2025). Source-specific nitrate and nitrite intake and association with colorectal cancer in the Danish Diet, Cancer and Health Cohort. Environment International, August 2025. ↩
• Federal Ministry of Food and Agriculture / Federal Ministry for the Environment (2024). Nitrate Report 2024. Report pursuant to Directive 91/676/EEC, reporting period 2020–2022. ↩
• European Commission (2026). Directive (EU) of 9 February 2026 amending Directive 91/676/EEC as regards the use of certain fertilizers derived from animal excrements (RENURE). ↩
• Richards J et al. (2022). Nitrate contamination in drinking water and colorectal cancer: Exposure assessment and estimated health burden in New Zealand. Environmental Research, 204, 112322. ↩
• Greenpeace European Unit (2026). Europe: Nitrate in drinking water. How agricultural pollution and outdated standards threaten public health. Media Briefing, April 2026. Cited as evidence of the political reception of the expert report, not as a source for the figures used here. ↩
• Stadtwerke Aschaffenburg (2022). Wasserwerk Aschaffenburg – treatment stages. Denitrification plant in operation since December 2000; nitrate concentration in drinking water approx. 25 mg/L. Stadtwerke Aschaffenburg ↩
• Veolia Water Technologies Deutschland (2021). CARIX process for central softening and nitrate removal of drinking water; more than 20 plants operated by municipal water utilities in Germany since the early 2000s. Veolia Newsroom ↩
• Mattheß B, Müller U (2023). Nitratentfernung mit Umkehrosmose: vom Konzept zur Umsetzung. DVGW energie | wasser-praxis 09/2023, 20–24. TZW: DVGW Water Technology Centre, practice partner town of Groß-Umstadt; plant in routine operation since 2021. TZW project page ↩

AI-Based Tools Used

Claude (Anthropic) was used as a tool in the preparation of this working paper. Responsibility for the selection, review, and factual accuracy of all information rests with the author. All calculations were verified against the cited primary sources; every cited source was checked in the original.

Tools used: Claude (various Sonnet and Opus models), Anthropic PBC, https://claude.ai — period of use: June to October 2026.