The Climate Footprint of the Health System

These days I read everywhere what a huge climate burden agriculture is, and livestock farming in particular. My personal experience with a diet that runs counter to the usual dietary guidelines (ketogenic and fibre-free) has so far had nothing but positive effects on my health. Many studies and experiments point in a similar direction. So it makes sense to take a look at what sits on the other side of that statistic - to look at human-caused emissions holistically, so to speak:

The Basic Assumption

Let’s assume for a moment that the current dietary guidelines are precisely not unambiguously healthy - then the follow-up costs are worth a look. It should be common knowledge that the health system bears high costs from chronic non-communicable diseases - so-called “diseases of civilisation”. So the health system as a whole is worth a look. And if these diseases are diet-related (as described in several articles on butterbei.de), nutrition offers a direct means of better prevention and avoidance - which would have an immediate effect on the health system’s carbon footprint.

The Climate Inside the Health System

Looking at internationally available figures, it quickly becomes clear that individual countries’ health systems are set up very differently - the NGO Health Care Without Harm⁽¹⁾ puts the global share at 4.4% of emissions, with markedly higher shares in the industrialised nations (Germany⁽²⁾ 5.2%, considerably more, though internationally by far not the front-runner⁽³⁾).

These figures, incidentally, refer to the health system in the narrower sense. Not included are:

care that becomes necessary beyond acute treatment because of chronic diseases
home care, which doesn’t show up in monetary statistics anyway

So these are minimum values.

It would be similarly difficult to put a number on the quality of life lost to chronic illness and the treatments it requires - but that’s not an aspect for this article.

Attributing by Disease

Attributing emissions by disease - and especially to the diseases of civilisation - is difficult, and has not yet been done in a citable way. Their financial share of spending, however, is extraordinarily high - and once you factor in that chronic diseases also affect acute treatments, for example through poorer wound healing, you can sense that non-communicable diseases make up the lion’s share of the emissions. Let me recall the basic assumption: diseases of civilisation are diet-related - meaning they can be not merely reduced but avoided through appropriate nutrition.

With one example we can look more closely at why this attribution is so complicated:

Type 2 Diabetes

Probably the best-researched chronic disease is type 2 diabetes (T2D). With serious and not uncommon complications such as cardiovascular disease, kidney damage, blindness, amputations, strokes and cancer, T2D is a front-runner among the chronic diseases, and one that markedly raises the likelihood of other chronic diseases occurring.

It’s often said: “Diabetes alone accounts for up to 10% of the costs of the German health system"⁽⁴⁾, or “4 in 10 adults have diabetes or pre-diabetes.”

But it depends on what you count towards those costs - otherwise you can end up double-counting:

In 2020, 7.4 billion euros⁽⁵⁾ were billed under the This code is used to classify diseases → more for diabetes. That includes no complications - for example cardiovascular disease, which is billed under other codes.
In 2009, a study reported 21 billion euros⁽⁶⁾ in costs including complications - that is, costs caused by patients who also had a diabetes diagnosis.
Also in 2009, a study reported 48 billion euros⁽⁷⁾ in treatment costs for diabetes patients, which includes all treatments - even unrelated ones - and therefore certainly overshoots. The figure is nonetheless defensible, since diabetes complicates any treatment - for example through poorer wound healing.

All of these costs refer to Germany, and have certainly changed over the years, so they aren’t directly comparable. They illustrate the difficulty of assigning costs to these multimodal chronic diseases.

A nice parallel to Dr Bernstein⁽⁸⁾: the “law of small numbers” would help us here too. If the spending and effort were negligibly small, no one would look at these figures - and no one would want to undertake the effort of attribution in the first place.

Which Emissions Can Be Saved?

We can consider two different scenarios for this. By way of example, we’ll stick with T2D again:

Preventing T2D

A 2024 study⁽¹⁰⁾ compares the remaining-lifetime emissions of people with a pre-diabetes diagnosis (people who are, therefore, already no longer healthy) in whom progression to full T2D was halted by intervention, against those in whom it wasn’t.

It turned out that the remaining-lifetime emissions can be reduced by about 67%: emissions remain because of the longer lifespan, but without the need for treatment they are considerably lower than those during a lifespan shortened by T2D by about 6 years with treatment.

Interestingly, cardiovascular events were even 11% higher in the prevention group than in the control group - presumably because they had those 6 extra years to develop such complications, and had already built up the risk during their pre-diabetes phase. The savings came, among other things, from kidney damage (73% fewer) and eye diseases (59% fewer).

Managing T2D Better

A clinical standard in the treatment of T2D is being “well-adjusted” - which refers not to the patient’s mental outlook, but to the control of blood sugar, for example through medication and insulin.

In one study⁽¹¹⁾ concerned with this adjustment, the researchers examined the effects of adjusting to an HbA1c value of 7%.

The emission reduction for these uncured patients is, despite all this, still on the order of 1-1.5 t CO₂ equivalent per patient - though, if I’m interpreting it correctly, over a simulation period of 50 years, not per year. All in all an unimpressive approach with an equally unimpressive saving.

T2D in Remission

If, however, you look at other reports (some of them case studies) where a remission of T2D is achieved within a very short time through a low-carb or ketogenic diet, the savings for the health system could well be far greater than the studies mentioned above certify.

Virta Health⁽¹²⁾ provides resources and experience with this approach. There you can also see that blood sugar returns to markedly lower and healthier ranges within days. Since T2D is diagnosed via the average blood sugar over 3 months (measured as HbA1c), a remission nonetheless takes longer than the normalisation of blood sugar and insulin levels.

For this paragraph I have no concrete figures - but it should be clear that preventing both pre-diabetes and full T2D leads to the elimination of all the treatment emissions in question.

Thoughts from Within the Health System

We’ve already seen one “clinical standard” in the treatment of chronic diseases above - I’ve found yet another angle for looking at the health system’s emissions: in 2026 the BKK Dachverband published a “Report on the Climate Impact of Health Care"⁽¹³⁾. In it, various disease areas are listed explicitly, and “prevention” appears in a few places too. You can’t accuse the report of not wanting to do prevention at all, but almost everything I read in it seems to refer to a different notion of prevention than the one I hold. Its choice of words fits this⁽¹⁴⁾: what it seems to mean here is the secondary and tertiary prevention of hospitalisations, complications and acute treatments in people who are already ill, rather than primary prevention - preventing the illness itself. Which leaves the report doing the work of making system-internal efficiency wear prevention’s clothes.

When I look at the report’s author, “Prof. Dr Oliver Gröne, OptiMedis”, that’s understandable too: OptiMedis makes its money partly from caring for patients - it would be surprising if the suggestion to shrink the customer base came from there⁽¹⁵⁾.

The Health Effects of the Health System

The whole set of analyses can be pushed to its logical extreme: a 2024 study⁽¹⁶⁾ looked at the health effects attributable to the health system’s emissions - that is, the harm it does to exactly the health it is responsible for.

But that gets a bit too meta for me to analyse any further. Especially since the kind of prevention I have in mind is meant to ensure that these emissions drop massively.

My Conclusion

If I get to choose where my climate budget is spent on my behalf, then I’d rather pick the tasty health prevention that provides a long “healthspan” than the treatment of chronic diseases, which markedly reduce quality of life even when they aren’t being actively treated.

Now, this too isn’t entirely uncontested, and meat is being blamed for more and more of the world’s ills - an argument that, in light of various studies, publications and personal experience, I absolutely cannot share. I’m aware that this is a position far from the mainstream. Normally I’d like to see studies here - but what I see is almost stronger than a study, and is entirely ordinary scientific practice: there are enough concrete counterexamples to refute established (mostly epidemiological, low-quality) studies.

As an example: when epidemiological studies claim that meat causes diabetes, but enough individual case reports exist in which diabetes is cured⁽¹⁷⁾ with a ketogenic, carnivore diet, that shows that the claimed causal link is decidedly doubtful, and that meat by no means prevents blood-sugar control.

Back to the emissions: in the end, a pleasantly habitable planet matters to me too - I just don’t share the easily refuted arguments that cast plants (which have caused me a great many problems in my diet) as the cure-all and animals as the incarnation of evil, and that simply ignore the emissions from other diet-dependent sources (the health system).