Sulphites in wine

Over the last few weeks I published a mini-series of posts on sulphites in wine, trying to go into a bit more depth than is usually found, and referring to my evidence base where possible. It is also perhaps a bit technical in places, but I hope it strikes the right balance at least for some winelovers.

The main thread of the posts starts with introductory material, including sulphite allergies, and a bit about typical sulphite concentrations, and limits specified by the various regulatory bodies. Then, after brief discussion of sulphite-induced headaches, which I see as a bit of a side-issue, I move on to the effect of sulphites on flavour. My posts are linked to below:

Clearing up a few points

Maximum and typical concentrations

Headaches

Effect on flavour

I must admit that when I started writing about sulphites, my view was that the issue was quite cut-and-dried, and they got far too much attention in the wine-world. I now see that sulphites are a lot more important than I thought, and not only for  wine-preservation and health reasons. The extent to which sulphites are good or bad is now not entirely clear to me, but I am convinced that sulphite-usage decisions are not to be taken lightly.

Sulphites in wine – effect on flavour

This was actually the post I intended to write a few weeks ago, but then I realised there are many poorly-understood aspects of sulphites, and decided first to embark on a mini-series of posts on the subject – clearing up a few points, maximum and typical concentrations and headaches.

Health-effects aside, most people seem to regard the primary function of sulphites to be the prevention of spoilage. It kills off bacteria and yeasts that can create nasty off-flavours, and works against oxidation, allowing the supposedly unsullied essential wine characteristics to shine through. And many people who object to the use of sulphites think and argue in the same arena, saying that the off-flavours add interest and character. However, it is true that some also praise the clarity and brightness of fruit flavours in natural wine. Terroir is mustered by both pro- and anti-sulphite factions to serve their separate causes: either saying the faults resulting from insufficient sulphur mask terroir, or that those qualities are actually a reflection of terroir, because the microorganisms that cause them are an essential component of it.

What is often common to both sides of the argument is a general failure, beyond the mere existence or absence of faults, to recognise the far-reaching consequences of sulphites on the organoleptic properties of wine. This point was clearly made in a recent SevenFifty Daily article, How Sulfites Affect a Wine’s Chemistry, which I make no apology for summarising below. I encourage you to read the whole article for further details and references.

There is a lot we still don’t know about how sulphites impact on wine chemistry, but research is starting to show that they affect a large number of chemical components in wine, and its organoleptic properties. Notably, sulphites act with oxygen and acetaldehyde to affect colour and mouthfeel, and aromatic compounds are also altered significantly. For example, a Sauvignon Blanc made reductively with sulphites in stainless steel tanks has a very different aromatic profile to one made in barrels with no sulphite additions. The effects are not only wide-ranging, but long-lasting. Research with Chardonnay shows that differing amounts of sulphite cause differences in the finished wine, even after several years of bottle age

Sulphite additions early in the winemaking process are particularly important. You either allow oxidative processes at that stage by not using sulphites, or must commit to fighting oxidation with sulphites throughout the winemaking process. Counterintuitively, winemaker experiments in vinifying with and without sulphites have shown that the “without” wines tend to have better long-term resistance to oxygen. They also tend to taste older when young, but show freshness of fruit as they get older, also exhibiting softer tannins, lighter colour, and more floral notes. On the other hand, reductive notes (not necessarily a bad thing) and cassis are more likely to be associated with sulphite additions.

The author of the SevenFifty Daily article seems to argue for the pragmatic approach of letting science decide how sulphites are to be used, with the goal of creating a stable wine that has the organoleptic properties intended by the winemaker. I certainly have some sympathy with that view, but on the other hand I also fully appreciate the ideological stance that added sulphites simply do not belong in wines. Should we use them simply because they are perceived by some to be beneficial? And if so, what other additions should be permitted on the same basis?

Irrespective of what we think, the presence or absence of sulphites in wine is an undeniably important issue – it affects stability in complex ways, and sulphites are a dangerous allergen for some – but also, as is becoming increasingly clear, its effect on flavour and mouthfeel can be profound, and cannot be ignored.

Sulphites in wine – headaches

Before I get back to posting more techy-stuff about sulphites in wine, here’s a brief diversion into the question of whether sulphites in wine can cause headaches.

Two things are clear. One is that, as far as I know, there is no scientific evidence that they do – we know that sulphites may result in a number of various symptoms in a small section of the population, but headaches is not one of them. The second is, despite that, quite a few people do believe that sulphites cause headaches, and have anecdotal evidence to support their belief. And of course a lack of scientific evidence for a proposition does not mean it is wrong.

The main problem with the anecdotal evidence is that wine contains another chemical, one that we know for sure causes headaches: alcohol. And there are also other chemicals that could conceivably do the same. So if you suffer from wine-related headaches how might you attempt to identify what is the culprit?

An easy first step would be to establish which wines are more likely to give you the worst headaches. Are sweet ones the worst? Followed by dry whites, and with red wines least likely to give you headaches? If so, that pattern is consistent with sulphites being to blame, because sweet wines are likely to contain more sulphites than dry ones, and white wines more than reds. On the other hand, if red wines are the worst, then it is more likely to be due to a chemical extracted from grape skins. A possible culprit is histamine – you could test for that by seeing if an antihistamine helps (but do check first that it is OK to consume alcohol with the drug). Or of course red wines might be giving you more headaches simply because they tend to contain more alcohol.

You could also compare your normal wines with ones that have particularly low sulphite levels. Generally speaking, a better quality wine, or something natural, biodynamic or natural, will be likely to have lower sulphite levels. But to be sure that you are comparing with a low sulphite wine you could go to the Raw Wine website and search for wines with levels under, say, 20 ppm. If lower sulphite wines give you fewer headaches it may still not be the sulphites themselves that are making the difference, as wines with low sulphite levels are probably low on other additives too. But on the other hand, if you like the low sulphite wine you might consider your problem solved anyway, so who cares?

Another approach would be to note if other sulphite-containing food and drink gives you headaches. Bright orange dried apricots are said to contain up to 1,000 ppm of sulphites. That’s a lot more than would be allowed in any wine, though to be fair you are likely to consume more wine in one session than apricots. Here’s a list I found of food and drink containing more than 100 ppm sulphites:

Dried fruits (excluding dark raisins and prunes)
Bottled lemon juice (non-frozen)
Bottled lime juice (non-frozen)
Wine
Molasses
Sauerkraut (and its juice)
Grape juices (white, white sparkling, pink sparkling, red sparkling)
Pickled cocktail onions

I feel a bit ill reading that list, but if you can pig-out on dried apricots, molasses, sauerkraut (and its juice) and pickled onions without getting a headache, probably sulphites are not responsible for your hangovers.

All the above suggestions are as unscientific as any assertion that sulphites cause headaches. The best they can achieve is to give you a little more understanding of how you personally react to sulphite additives, and I’m afraid that’s the best I can offer.

Sulphites in wine – maximum and typical concentrations

In pretty much all wine-producing countries, there are regulations to limit the sulphite content of wine. The limits are always expressed nominally in terms of total sulphur dioxide, but I suspect they all assume an analysis method that fails to account for some of the bound sulphur dioxide, as discussed in my previous post.

Let’s start by taking a look at sulphite limits in the European Union. The rather complex set of limits are very nicely summarised in a table in the document EU rules for organic wine production: Background, Evaluation and Further Sector Development:

If you feel motivated to check these various EU limits or get more details, the table provides you with the relevant document numbers, which can be found by a web search. But it is not a task to be tackled lightly – I once set out to extract the various sulphite limits from the 1999 regulations and nearly lost the will to live.

The US regulations on the other hand have the advantage of simplicity: for wines with no organic-credentials the limit is 350 ppm (or mg/l – the units are practically identical). That figure is widely quoted, but I am afraid I have not been able to track down the actual regulatory document. I had more success with finding an authoritative-looking document for organic wines in the USA: Organic Wine: Oversight, Labeling & Trade. That document covers two categories: wine made with organic grapes and organic wine, where organic wine has stricter rules. For organic wine, added sulphites are not allowed at all – though the wine will still contain naturally occurring sulphites, and probably need a sulphites warning on the label, as discussed in my previous post. However, for wine made with organic grapes sulphite additions are allowed, providing there is no more than 100 ppm sulphur dioxide in the finished wine. In the EU by the way, wine made with organic grapes is not a special category: the grapes must be farmed organically as claimed, but all winemaking regulations are as for conventional wines in the above table.

For allowable sulphite levels in other countries, The Oxford Companion to Wine reads: “In South Africa, the limit is 150 mg/l for dry reds, 160 mg/l for dry white, rosé, and sparkling, and between 200 and 300 mg/l for sweet wines depending on style and level of sweetness. Argentina: 130 mg/l for dry reds, 180 mg/l for dry white and rosé wines and sweet reds, 210 mg/l for sweet white and rosé. Chile: 300 mg/l for all dry wines and 400 mg/l for sweet wines.”

As far as biodynamic wine is concerned, Demeter certifies it internationally, and their document Standards for Demeter/Biodynamic Wine regulates the use of sulphites. The stated aim is that sulphur dioxide be restricted to the absolute minimum, but then the document proceeds to specify limits that are more lax than the US organic regulations. For different types of wine, the maximum allowable total sulphur dioxide at bottling in mg/l is:

But what about so-called natural wines? As you probably know there is no official definition or certifying body, but we can take a look at the list of wines made by the accredited growers and makers of Raw Wine – Isabel Legeron’s platform for the promotion of natural wines. By searching the list using filters provided on the website, it is possible to get a feeling for sulphite levels in wines deemed to be natural. The database also records which wines have added sulphites, but sadly you cannot use that as a search criterion. Here are the results of a search on 6th June 2019:

NATURAL WINES
Total sulphur dioxide (ppm) Number of wines
0 468
1-10 925
11-20 929
21-30 957
31-40 739
41-50 504
51-60 341
61-70 196
71-80 5
81-90 2
91-100 2

I think some of the quoted sulphur dioxide analysis results in this table need to be taken with a pinch of salt. For example, when you look at the data in more detail it is obvious that some testing laboratories round the concentrations to the nearest integer multiple of 5 or 10 ppm. Note also the high number of wines with precisely zero sulphites, despite it being commonly stated that fermentation necessarily creates sulphites in all wines. I have seen 10 ppm referred to as a detectable level, implying that measuring anything less than that is problematic, so perhaps that explains all the wines at 0 ppm? But regardless of such quibbling, I think we can conclude from the table that most natural wines contain less than 40 ppm, and that there are fewer and fewer natural wines at levels increasing from 40 to 70 ppm.

So now we know the sulphite concentrations in some natural wines, thanks largely to sulphites being one of the obsessions of the natural wine community, but getting sulphite concentrations for wines in general seems to be more difficult. The assumption is often that the makers of cheap wines zap their wines with as much sulphite as they can get away with, to compensate for poor fruit quality, and for closer control of the winemaking process to create a consistent product that meets an expected flavour profile. On the other hand, higher quality producers are expected use less sulphites, as they take more care to select healthy fruit, are willing to put more effort into low-intervention winemaking, and are also more tolerant of variation in the end product. I personally think there must be some truth in that characterisation, but sadly cannot demonstrate it with numbers. More certainly, it is the case that red wines need less added sulphites than white, because the tannin in red wines will also provide protection against oxidation, and sweeter wines will tend to need more added sulphites, because the sugar binds sulphur dioxide, rendering it a lot less effective.

To get some feeling for typical sulphite concentrations in wines that are not claimed to be natural, I think we could do a lot worse than look at the maximums in the first table of this post. EU regulations are typically designed to reflect existing practice rather than to effect change, so I think it is reasonable to assume that most wines have sulphite levels approaching, but comfortably within, the specified limits for the different styles of wine. On the other hand, producers who particularly favour low-intervention methods (even if they do not identify with the natural wine movement) would be closer to around 30 ppm, and other high-quality wines would have intermediate concentrations.

So what are we to make of all this? To be honest I am not at all sure. One might hope that maximum sulphite levels were specified according to some sort of objective assessment of health risk, but I am not convinced we know enough about sulphite allergies to do that – there is not even agreement about what proportion of people are affected by sulphites (as mentioned in my previous post). So what we have are rules based on a mish-mash of current practice and ideology. In my opinion, the best that can be said for the current situation is that consumers can exercise a degree of choice about their exposure to sulphites – based on their world-view and how they personally perceive health risks.

Sulphites in wine – clearing up a few points

I studied a bit of chemistry at university so I know what a sulphite is, I thought. It’s an ion that has a single sulphur atom, three oxygens and a double negative charge. And I most definitely would not make the common error of confusing sulphite with sulphate, which is spelled with an a rather than an i, and has four oxygen atoms – a totally different beast.

The usual story with sulphites is that they are added to wine to work against oxidation, and kill yeast and bacteria, but the downside is that some people are allergic to them, and in the US and EU they require a warning label if present at 10 ppm (parts per million – very similar to mg/l) or more. They are also anecdotally associated with headaches, and thought by some to be a Bad Thing simply because they are not Natural. However, at the next level of sophistication you might also be aware that some sulphite content actually is natural, as it is a fermentation product, and thus present in every wine even if not added artificially.

So far so good, but when you start poking around a bit more, asking what sulphite concentrations actually mean, and how many of us are allergic to sulphites, it suddenly gets rather more murky. What people call sulphites are not necessarily the sulphites a chemistry undergraduate is confident about, and it is not clear precisely what people are allergic to. I don’t claim to have a complete overview of all these issues, but will try here to clarify what I can.

The sulphite ion (top left), sulphur dioxide molecule (top right), and the two tautomers of the bisulphite ion (bottom)

Let me start by talking through some of the points made on Ben Rotter’s excellent webpage on sulphur dioxide. Sulphur dioxide can be introduced into wine using any of a number of approved additives but, regardless of which one is used, the same set of sulphur-related entities will result:

  1. Sulphur dioxide – gas molecules in solutions
  2. Sulphite ions – as described at the start of this post
  3. Bisulphite ions – like sulphite but with a hydrogen, and less charge
  4. Unstable compounds – formed by bonding with various other chemicals
  5. Stable compounds – notably the product of reacting with acetaldehyde

Even though only number 1 in the list is actually sulphur dioxide, 1-3 are often referred to as free sulphur dioxide, and 4 and 5 as bound sulphur dioxide. The total sulphur dioxide includes both the free and the bound forms. Any protective properties are largely lost in the bound forms, with free sulphur dioxide doing the vast majority of the good work. In fact, it is the molecular sulphur dioxide that is most effective form but, depending on the pH, there may not be much of it present as a fraction of the total.

You may have noticed that from talking about sulphites we quietly slid into the subject of sulphur dioxide. You will find that is quite common in discussions of this subject, and the two are often seen as being practically synonymous. It does not help that when someone says sulphur dioxide they could be referring to

  1. Actual sulphur dioxide molecules,
  2. Free sulphur dioxide,
  3. Bound sulphur dioxide (unlikely perhaps),
  4. Total sulphur dioxide

In fact you could add a fifth, which is the sulphur dioxide measured by a specific procedure. More on this later, but let’s first take a look at sulphur compound allergies.

The first point to make on the subject of allergies is that people can have adverse reactions to several different sulphur compounds, and just because you have problems with one does not mean you need to avoid contact with all of them. In the food and drink industry it is common to talk about adding sulphites, and the product as containing sulphites. Accordingly, from a medical point of view, what is of interest to us here is usually referred to as a sulphite allergy or intolerance, as it is a reaction to food or drink that contains these sulphites. Also, my impression is that people take the practical view that all sulphited products finish up containing the same range of chemicals, as listed above, and do not bother to distinguish between the individual chemicals when gathering data about allergies.

There are many possible reactions to sulphites, both true allergies (due to an over-sensitive immune response) and other intolerances. The most severe response is anaphylaxis, which is rare but very serious and possibly lethal, while the most common is the worsening of any asthma symptoms, also potentially lethal in some cases. Other less serious possible reactions are hives and allergic rhinitis. But as far as I know there is no hard evidence that sulphites cause headaches. Estimates of the numbers of people affected by adverse reactions vary a lot. One estimate is that 1% of the population is affected, of which 5% also suffer from asthma; while another source says 5% of asthmatics are sensitive to sulphites, compared with 1% of the rest of the populations; and a third source claims that in America less than 0.05% of the whole population is affected.

Speaking of inconsistency in terminology, you might also note that although food scientists and medical people tend to talk about sulphites, and the warning text on wine labels reads “contains sulphites”, the threshold level for warnings is defined in terms of total sulphur dioxide content, a label warning being required for 10 ppm or more. In fact a specific procedure (or one that gives similar results) is mandated for measuring sulphur dioxide: the optimised Monier-Williams distillation-titration procedure.

Looking at methods similar to optimised Monier-Williams (here and here), it seems that the first step is to acidify, which converts all the free sulphur dioxide, and also a fixed proportion of bound sulphur dioxide, to the molecular form of sulphur dioxide, and then the sulphur dioxide gas content is measured. So it seems likely that the concentration levels determined for comparison with the 10 ppm limit, are expressed in terms of molecular sulphur dioxide gas, rather than the ionic forms. Another conclusion is that, as not all the bound sulphur dioxide is converted to the molecular form for measurement, the procedure will not actually give the true total sulphur dioxide content – it will be somewhat less, but greater than the amount of free sulphur dioxide.

Why is the limit set to 10 ppm? I could not find a definitive answer, but 10 ppm is referred to as being a detectable amount, so it seems that it is linked to the practicalities of measurement, rather than the level at which some people might react badly. As the mandated procedure under-measures, it would anyway seem silly to start fretting too much about the science behind the limit of 10 ppm. Perhaps the best that can be said about the regulations is that they are capable of being enforced consistently, and offer a degree of protection to those who are allergic.

I can only assume that the sulphite concentrations often bandied about by natural wine advocates also refer to sulphur dioxide as measured by the optimised Monier-Williams method, but I have never seen it stated. You will typically see sulphite concentrations quoted if a low-intervention producer choses to add sulphites, but otherwise an informal description of a low-intervention wine will often only say “no added sulphites”. Just remember that there will still be sulphites in a no-added-sulphites wines, and quite likely more than 10 ppm. This may be fine if you are mainly concerned about the natural-credentials of the wine, but is definitely something to bear in mind if you have a serious sulphite allergy.

So that is my best shot at explaining more precisely what sulphites actually are, how they are regulated, and to what extent they are responsible for allergies. I have done my best to stick to facts to the extent I could establish them, and keep my opinions to myself, but I might be more indulgent in future posts now I know a bit more what I am talking about. If I’ve got something wrong, or could have explained something better, please let me know, and I will try to correct or improve what I have written.