KempyKolibri

You said:

> if you'd picked a slightly different set of CVD-correlated biomarkers you would have got exactly the opposite result.

So the evidence I’m looking for is empirics showing CVD correlated biomarkers that suggest a beneficial effect from consuming levels of sodium above recommended levels. Without that evidence then I don’t see why we should believe they would have got the opposite result if they picked other CVD biomarkers.

> but this does not suggest it's the long term cause of blood pressures going up from a normal 120 to a chronic 160 or 200 as we're seeing in tons of people

I’m not claiming that salt is the single cause of hypertension, but that doesn’t mean that the kind of reductions you see from salt reduction aren’t meaningful or contribute to those very high figures. It’s easy to dismiss 5mmHg as insignificant, but we generally see a 5mmHg reduction in sysBP translate to a ~10% reduction in CVD events. Considering how prevalent CVD is, that’s a pretty large effect size.

Chronic diseases are often overdetermined and stack - people have a poor diet which means they consume too much salt, they’re overweight and obesity, have T2DM or prediabetes, sedentary lifestyle, etc etc. The fact that we can’t point to a single one of these and say “this is the thing causing your 160/100 BP doesn’t mean we shouldn’t try to fix the individual factors. So sure, salt reduction seems to offer 5-6mmHg reduction, exercise 4-8, hypertensive drugs 10. But put them all together and that’s a massive change.

> Anyway, I don't have time at the moment to look through the 11 studies cited in that meta analysis, but if you pick the one or two that give the strongest evidence for salt causing CVD I'd look at them.

It’s just three trials (or four depending on how you count TOHP I & II). I think I’ve met my burden in terms of showing there’s evidence of high salt intake having adverse effects, I have no interest in forcing you to read them. Just trying to provide evidence if that’s something you’re seeking.

> I'm genuinely trying to figure this out myself as best I can, because I know way too many people close to me dealing with early stage CVD and diabetes.

I’m sorry to hear that. We certainly seem to be struggling with chronic lifestyle-related disease these days, though with GLP-1 RAs I’m a lot more optimistic than I was a few years ago.

> And a lot of them say they're working on it by avoiding meat and dairy and eggs and salt, and instead of that they end up eating more refined oils and refined flour and sugar.

Yeah depending on the composition of what they’re eating that doesn’t sound great. IMO Replacing butter with refined oils and whole grain flours/carbs - sure, solid move. Replacing meat with plant proteins is also a worthwhile step, and eggs don’t seem to be great for health in many respects. But fermented dairy seems to be a positive as far as CVD risk goes, so ethics aside, seems like a backwards step if they replacing yoghurt and cheese with sugar and white flour!

> I'm not saying we need to eat tons of salt, maybe it does have a minor effect, just that it's not the real culprit.

I’m very wary of trying to find “the culprit” for public health problems. It’s so rarely the case that a disease has a single aetiology, and in my experience the people who’ll tell you “it was the sugar all along”, “it was the seed oils all along”, “it was the glyphosate all along” have a book to sell. The reality is probably closer to it being a combination of several things. Not as sexy though, no publisher or influencer is interested in that view!

> The EU has already defined dairy items milk, butter, cream, yoghurt and cheese as "products secreted by mammary glands"

I think they should just label things more explicitly like this - accelerate veganism 100x when people in the supermarkets have to choose between “pressed soybeans” and “mammary gland secretions”.

Two different things. If you don’t want to eat animal products for ethical reasons but you want something high protein and/or similar in taste/texture to a beef burger, eating a vegetable stir fry isn’t going to cut it. That’s what these products are for.

> Who is "we"?

Me and anyone else reading this.

> While empirical evidence is valid only for the situation in which it was obtained, mechanism is universal.

Mechanisms are inferred from empirical evidence, I don't see how you can treat them as two separate categories. For example, in your crash test dummy analogy, verification through crash tests (with dummies) is empirical evidence. Yet under your framework, should we assume that it is valid only for the situation in which it was obtained - only for dummies, not people; in cars pushed towards walls in controlled situations, rather than on public roads?

If you name proxy experiments that support your views (crash tests) as mechanisms and ones that don't (SSB replacement with NNS RCTs) as "empirical evidence is valid only for the situation in which it was obtained" then sure, everything you want to believe is supported by sound science and everything you don't isn't. But the view itself seems to contain a logical contradiction, so you're dead before you've even got off the ground.

I would understand mechanistic evidence in the domain of health science to be in vitro and animal studies. Even if we were to grant that mechanism is universal in this field (which I wouldn't, we frequently see heterogenous results even within the same exposures on the same mouse models, for example), there are thousands of mechanisms that come together to influence the outcomes we actually care about. This is why when we look at translation rates of mechanisms to outcomes in humans we typically see rates below 5% (and is also why pharmaceuticals that work perfectly in animal models barely ever make it to market in humans).

Going back to the evidence you've cited in support of your intervention - the first two (the only ones in humans) are neither looking at NNSs nor an intervention on banning them. So it doesn't meet your own goalpost for "if we introduce a public health policy, then we need to take human behavior and adherence rates into account". In the rationing example, you have an entirely different context - one in which people literally cannot purchase large amounts of sugar. This would not be the case if we were to ban NNS today.

Your third study was in mice which, as discussed, has an incredibly low chance of actually translating into human outcomes. I don’t find “we have evidence in RCTs that NNSs are beneficial but there’s this mouse study that says otherwise so let’s ban them” a convincing argument.

So again, any actual evidence in support of your proposed intervention? How do we know, for example, that banning NNSs won't just lead to higher sugar consumption and adverse outcomes, since we know from RCTs that substituting SSBs for NNSs improves health outcomes? If all those consuming your banned substance now switch to SSBs instead of their NNSs, congratulations, you've just worsened health outcomes.

> You could say that this suggests higher sodium increases cardiovascular disease risk but that seems like a stretch when if you'd picked a slightly different set of CVD-correlated biomarkers you would have got exactly the opposite result.

What’s the evidence for this?

> Even if the results had been more convincing, the methods are not - it's an extremely short term study and looks only at correlative biomarkers and not at actual health outcomes

You're going to have to pick your poison here - when you're after long term data on dietary interventions showing hard outcomes it's highly unlikely you'll ever see this in the form of RCTs that you're looking for (i.e. where you _only_ alter salt consumption). That's why we look for converging lines of evidence - biomarkers/soft outcomes from RCTs and hard outcomes from prospective cohort studies, for example. When we look at this for salt, we consistently see lower salt = lower adverse outcomes.

That said, when we meta-analyse RCTs we do actually have sufficient power to see improvements on hard outcomes. In this meta (https://doi.org/10.1016/S0140-6736(11)61174-4) we see a 29% reduction in cardiovascular events in the 7 months to 11.5 years in normotensives in RCTs which looked exclusively at salt reduction. I wouldn't call 11.5 years short term, nor cardiovascular disease events a soft outcome. So surely this ticks all your boxes?

Ok, so we already have your standard: in order for you to support a public health intervention we need to see evidence that such an intervention results in positive outcomes in a free living population with ad libitum consumption, regardless of the evidence in controlled settings.

We have your proposed intervention: “In the context of school lunches (which is discussed here) [NNS] absolutely need to be banned, same as added sugars.”

So now we need evidence supporting this intervention sufficient to meet your own goalposts. Do you have it?

So DASH-sodium (https://www.jacc.org/doi/10.1016/j.jacc.2021.03.320) seems like it meets those goalposts? As does TOHP (https://doi.org/10.1136/bmj.39147.604896.55).

When we scale out to what I believe to be a superior intervention - replacement of sodium with potassium, we have really robust data. The SASS trial (https://www.nejm.org/doi/full/10.1056/NEJMoa2105675) showed reductions in stroke and CVD incidence from consuming salt substitute, along with reduction in all cause mortality.

I have a real salt taste - if I had no care for health outcomes I’d absolutely cover my food in it! However I think the evidence in favour of reducing or, ideally, replacing it with potassium-heavy substitutes is really convincing.

Exactly. This isn’t a study that shows us UPF adds anything over what we already understand about dietary composition.

I’m not an absolutist on this myself, if convincing evidence comes out then I’m happy to change my mind. I have no horse in this race besides wanting people to stay healthy. I’m just not currently convinced the UPF category adds much to our existing dietary guidelines.

Avoiding UPFs can serve as a reasonable heuristic as it can help you avoid HFSS foods, but you could also easily make some adverse switches following such advice.

I’m not sure that study really supports the inference you’re making re: calories and weight gain. The UPF diet had ~15g fibre/1000kcal less than the MP diet, so for an individual consuming 2500kcal/day there’s a 150kcal difference of energy per day that doesn’t seem to be accounted for by the studies methods (could be missing this in their calcs though?).

So yeah, if we don’t equate things like this, things look bad for UPFs. But what is the “UPF” category adding here. We don’t need it to know that if more calories are consumed as fibre in an otherwise calorie-equated diet, you’ll lose weight relative to the low fibre diet.

Ok, so if I understand correctly, your argument is: in order for you to support a public health intervention we need to see evidence that such an intervention results in positive outcomes in a free living population with ad libitum consumption, regardless of the evidence in controlled settings.

So what’s the evidence that banning artificial sweeteners leads to positive outcomes in a free living population, considering you said: “I'm in favor of banning artificial sweeteners”?

Let me get this straight: your epistemic framework is such that when presented with RCTs in humans showing positive effects, observational studies in humans showing null findings (likely because of poor adjustment models) and negative associations in rats, we should conclude “artificially sweetened is every bit as bad as sugar sweetened”?

Let’s do this one at a time, starting with salt.

What’s your goalpost for evidence here, I.e. what would it take to convince you that salt consumption above the levels indicated in dietary guidelines is harmful?

Oh, so was your point that there’s no reason to feed them to children, but there are good reasons for NNSs for gen pop?

You have to be very specific in the intervention you’re either looking at in an RCT or modelling in a prospective cohort study. We wouldn’t expect adding NNSs to a diet to improve much (perhaps some benefit from carbonated ones on body mass). We need to investigate/model replacement of SSBs with NNSs.

When we do that we pretty consistently see benefits. Good overview as a response to the WHO position paper here that goes over that evidence base: https://mailchi.mp/b30c80ddf8ba/who-as

> and yet when Americans set foot on European soil the first thought they have is 'everyone is so thin'.

I’m not sure how this interacts with the point I’m making.

> if there's any misstep here, it's not focused enough on sugar (which is a hard drug for kids) .

Sure, free sugars are associated with risk (which is reflected in dietary guidelines). Plenty of UPFs are not (which is why they generally aren’t flagged as a concern in dietary guidelines).

There’s an excellent discussion of the most recent RCT on UPF foods with its author here (https://sigmanutrition.com/episode577/).

Not only is the study itself very interesting, but it does a great job going over the evidence landscape on UPFs and food at the current point in time.

Dr Dicken is very clear that at the moment, in his view, they have insufficient evidence to make policy decisions like this, a view that has been echoed by pretty much all academic institutions when they’ve helped inform recent dietary guidelines.

Making big sweeping moves based on flimsy evidence is a good way to make people wary of following dietary guidance at all. After all, what’s the point of listening to “the science” if they turn around in 5 years and say “whoops, turns out we were wrong”?

This is a misstep IMO.

Yep, this seems like quackery.

Wholemeal bread with soya lecithins? Evil UPF, ban it.

Artisinally produced sourdough using refined flour with tons of salt but no lecithins? Delightful, fill your boots.

We've let nutrition policy become controlled by fad diet book authors and the results aren't pretty...

As I wrote elsewhere:

If someone is habitually consuming sugar sweetened beverages, replacing those with ASBs will, the evidence strongly suggests, reduce your risk of obesity and various chronic diseases.

We can say "just don't consume either" but we have decades of attempting such policies that shows people don't work that way. Someone who wants to drink a can of coke will drink a can of coke, why would we ban the healthier option?

If someone is habitually consuming sugar sweetened beverages, replacing those with ASBs will, the evidence strongly suggests, reduce your risk of obesity and various chronic diseases. We can say "just don't consume either" but we have decades of attempting such policies that shows people don't work that way.

Someone who wants to drink a can of coke will drink a can of coke, why would we ban the healthier option?

Thanks for getting back to me! Doesn’t look like it’s likely to be that then.

Good luck to you, hope you keep on top of it!

What I’m arguing against is the form of the argument you’re making when you say:

> If the data here is correct and inflammation is a better indicator of heart disease than cholesterol levels, it should raise questions whether cholesterol levels are correlative rather than causal.

But, as with the asbestos and smoking example, I don’t think an argument with this structure is valid. What I’m arguing against is your claim that the existence of one exposure/outcome relationship with a stronger association entails casting doubt on other exposures’ relationship with the same outcome.

Does that make sense, and do you have a good argument as to why we should believe that it casts doubt?

Where we agree is that oxidation of LDL is part of the atherogenesis pathway.

However, I’m not aware of any evidence showing that higher oxLDL vs higher LDL entails greater risk. IIRC there is a paper that compared oxLDL to ApoB where ApoB was more strongly associated with risk, which doesn’t seem like it would be expected on your hypothesis.

The response to retention hypothesis and the mechanistic evidence supporting it point to oxidation being inevitable once ApoB tagged lipoproteins are trapped in the arterial wall. That being the case, it would be moot whether it entered the wall in an oxidised state or not.

So yes, oxLDL is one of the early stages of atherogenesis, but I’m not aware of any evidence that having LDL pre-oxidised before it’s trapped vs regular LDL increases risk.

I’m always open to new evidence that would change my view, but that’s my understanding of the research landscape at the moment.

Why would that raise questions about whether cholesterol levels are correlative rather than causal?

Presumably one of “time spent working in asbestos factories” and “cigarette pack years” is more predictive of lung cancer incidence than the other. Does this being the case mean that either smoking or inhaling asbestos particles isn’t causally implicated in lung cancer? If the answer is “no”, then presumably you can see why the inference you’re making is faulty.

Welcome! Thanks for being understanding.

The claim I was addressing was that we don’t have evidence of statin efficacy for primary prevention. My point was that JUPITER seems to be evidence this isn’t the case.

> Age-standardized cardiovascular mortality dropped steadily from 1975 to 2010 (with no particular discontinuity when statins were introduced), and has not budged since 2010.

Why would we favour cross sectional data over that produced by more rigorous methodologies?

> the tiny effect that statins claim (25-35%, under particular conditions)

25-35% reduction in one of the west’s leading killers is tiny? Wild.

Good luck to you, hope you keep on top of it. Do you know if you have high lp(a) out o curiosity? Curious what drives these events in cases such as yours.

I think that’s a disservice to quacks - most of them are far better at dressing up their claims in something believable sounding. Gundry is just stark raving mad, it blows my mind that anyone gives him the time of day.

We could grant all that for the sake of argument but it wouldn’t interact with anything I said.

I’m talking specifically about the claim that because individuals with a low exposure to something don’t necessarily have a given outcome, there is no causal pathway between the exposure and the outcome. It’s not clear to me how what you’ve written speaks to anything regarding that claim.

Not trying to be snarky, I just want to be really clear about the point I’m trying to make here.

Discordances like this between LDL-c and ApoB are one of the reasons many lipidologists are pushing for LDL-c measurement to be replaced with ApoB.