Fact or Fiction – Are Raw Cheeses Better?

Over a year ago I wrote an article on the safety of raw milk, and what can be done to mitigate its risks. As part of that article I reviewed a study that showed that consumers could not tell the difference between raw and pasteurized milk…and promised a post on whether the same was true of raw cheeses. Finally, I am delivering on that promise.

But first, let’s set the stage. There has long been a claim that cheeses made from pasteurized milk are inferior to those made of raw milk. These claims go back to the advent of pasteurization itself, and persist through to today. But are they true?

As you may guess, quality scientific data on this question is hard to come by. There are some very entrenched interests who want this to be true (raw cheese producers and raw milk proponents) and others who very much want this to be false (pasteurized cheese producers, public health agencies). Unfortunately, much of the science investigating this question is performed by these groups, often leading to studies that are designed to provide the desired answer rather than provide an empirical and impartial analysis.

What Would a Good Study of Raw Cheeses Look Like?

Before going into the data, we should spend a few seconds talking about what good versus bad experiments look like. A general rule for experimental design is to “hold all variables constant except the one you are investigating”. This allows you to be fairly certain that any differences you discover are due to the variable you manipulated, and not due to extraneous factors.

An example of bad design: There are many studies out there which acquire the same style of cheese from different manufactures – some raw cheeses, others pasteurized – and then subject them to tasting, chemical, or other analyses. While this is an easy way to run a study, it’s also fundamentally flawed. Even though the style of cheese is the same, you’re still looking at cheeses which likely have some differences in their recipes, cultures, preparation, and affiange (aging). Thus, if you find a difference, you don’t know if it was because raw versus pasteurized milk was used, or whether the differences are due to factors such as the cheesemakers skill level, differences in recipe formulation, the type and amount of culture used, how the cheese was aged, etc. As a example, this study found that raw milk cheeses are often perceived as more pungent, using this approach.

What does good design look like: So the above is bad design – but what would be good design? This is maybe obvious, but the ideal experiment would use the same milk, collected from the same animals, on the same day. A portion would be pasteurized, a portion left raw. Then, a single cheesemaker uses that milk to make raw versus pasteurized versions of the same cheese, using the exact same recipe and ingredients, making both cheeses at the same time. The cheese would then be aged under identical conditions, in the same cheese cave (or other aging location). Finally, the taste test or other analyses would be done blind – e.g. the taste testers, chemists, etc, doing the analysis don’t know which sample is which – and even better, they shouldn’t even know the nature of the experiment. This approach eliminates most possible sources of bias, as the only thing which changes is how the milk is treated.

To some this may seem excessive, but it’s worth pointing out that milk from different farms in the same region can lead to very different flavours in cheeses that are otherwise made the same. Which just goes to show how important it is to control all of these “extraneous” factors.

Another Brief Segue – Preferences are not Facts

One last point to touch on before going into the studies themselves is preferences. When we talk about “cheese X is better than cheese Y”, in most cases we are talking about consumer (or our own personal) preferences. These are not facts – you and I may like very different cheeses or beers. You’re not wrong for liking something different, you simply like something different.

So to some extent this entire article is somewhat moot. Even if there are customer-detectable differences between raw cheeses and pasteurized cheeses, which “taste better” will vary from person to person. Because of this, many studies don’t ask for preferences, but rather ask if tasters can reliably tell two cheeses apart, or will have trained testers rank the relative strength of specific tastes and aromas.

What Does the (good) Science Say?

As I alluded to above, the number of rigorous studies in this area are fairly limited, with 4 recent works meeting the fairly stringent standard I described above. I want to highlight these, as combined they tell an interesting story.

Study 1 – Cheddar for the masses

The first study looked at cheddar cheese production, in the same dairy, using raw versus pasteurized milk. The same milk, recipe, cultures, processing, and aging was used for both cheeses. The cheeses were then analyzed by a group of university students using a two-part analysis:

Triangle Test: In this test, made famous by the Brulosophy crew, the students were given two servings of one cheese, and one serving of the other. They were then tested on whether they could tell the samples apart. 85% of participants were able to tell the pasteurized and raw cheeses apart, pretty clearly showing that there was a detectable difference in taste and/or aroma between the raw and pasteurised cheeses.

Preference Test: Those participants who could tell the cheeses apart where then asked which one they preferred. 74% preferred the pasteurized cheese. So case closed – pasteurized cheeses is superior! Right?

Study 2: Big, Bold and Blue

The second study I want to discuss looks at the production of a Spanish Valdeón Blue Cheese, and used both chemical analyses and tasters to assess the impact of pasteurization. What they found was interesting – raw milk cheeses developed more free fatty acids (FFAs) as they aged. FFAs are made by the lipolysis of triglycerides (fats) in the cheese, by lipase enzymes from the milk. FFAs themselves are flavourless, but can be converted by bacteria into a range of flavour- and aroma-active compounds. These tend to be “bold” flavours, often described as “goaty” and “spicey”. But that was machines measuring chemicals, so what did tasters find?

The first thing tasters found was that some flavours increased in intensity over time – “spicy/rustic” notes became more intense, as did sourness and the perception of salt. But the second thing tasters found, and which was rather surprising, was that there was no significant differences in the taste of the pasteurised versus raw cheeses.

So score one for untrained university students, whom apparently out-performed trained testers in discerning raw versus pasteurized cheese!

Jokes aside, this result is not as unexpected as you may think. Blue cheeses tend to be rather intensely flavoured to begin with. Is it really surprising that modest differences in one group of flavour compounds wasn’t detectable against such a bold background?

Study 3: It’s Not Always About Flavour

The third study is a bit of an aside, and rather than looking at flavour, looked at other aspects of cheese making. The impact of pasteurization on cheese making using both cow and goats milk was assayed. They found that goats milk consistently had more protein and fat than did cows milk, and that pasteurization of both led to higher yields when cheese was made. This doesn’t help our understanding of the impact of pasteurization on flavour, but it is interesting to note that pasteurization may help to improve cheese yield.

Study 4: Experienced Tasters Sometimes Succeed

Finally, we’re onto the last study. This study looked at the impact of pasteurization on the produciton of the traditional Iranian cheese Liqhvan. This was a well-performed study where everything was kept the same in the cheese production other than pasteurization, and a number of chemical, microbiological, and flavour analyses were performed. This study looked at a broad range of flavour compounds that are produced by microbes, including soluble nitrogen compounds (e.g. biogenic amines) and FFA-derived flavour compounds. To my surprise, only FFAs were increased in raw cheeses. The other tested flavour compounds increased in concentration as all the cheeses were aged, but there were no differences in the levels detected between raw and pasteurized cheese.

Experienced tasters were able to find small differences in the flavour of the pasteurized versus raw cheeses (see the flavour wheel, above). The specific flavours detected did not change with pasteurization, but the general intensity of the flavour was higher for the raw cheeses. While subtle, these differences were significantly different – at least to trained tasters.

But I left this study to the end of this section for a reason, as it comes back to my first article on the topic of raw milk. I will let this study authors speak for themselves, rather than paraphrasing: However, coliforms and Escherichia coli were seen in raw milk cheeses until the last day of ripening…since the presence of E. coli makes the cheese inedible, it seems that the pasteurization of milk is mandatory for the production of this type of cheese.

Even after 60 days of aging, there were dangerous microorganisms present in the cheese. While I won’t talk about this more here, as I covered it exhaustively in a previous post, raw milk products are among the highest risk foods (in terms of the risk of foodborne illness), so great care needs to be taken with these products.

The Bigger Picture

There are a few things we can learn from these studies:

Flavour Differences: There are subtle differences in the production of one group of flavour/aroma-active compounds in raw versus pasteurized cheeses. These are the free fatty acids, which can provide a range of flavours from nutty, to goaty, to “spicy”. In at least some cheeses, this chemical difference can be detected by untrained and trained testers. But outside of the FFAs, there are no instrument- or human-detectable differences in any other class of flavour/aroma-active compounds. Whether these differences are detectable by humans depends on the style of cheese, with more intense cheeses masking these differences. Not only that, but these “missing” flavours are present in pasteurized cheese, they just develop more slowly.

The Nature of the Flavour Differences: One thing that is quite interesting about the difference in flavour between raw and pasteurized cheeses is that the FFAs are largely a product of the milk itself, and not of the microorganisms in the cheese. FFAs are released from fats in the milk by the enzyme lipase. While some bacteria make lipase, the primary source of it in cheesemaking is from the milk. Lipase is partially inactivated by pasteurization, so a concurrent decrease in the flavours made from this enzyme products is not unexpected in these cheeses. But this also raises an interesting point: cheese makers often add additional lipase to their cheeses, either to counter the loss of lipase during pasteurization, or to drive FFA production to higher levels. This suggests that it may be possible to eliminate the flavour and aroma differences between pasteurised and raw cheeses simply by adding lipase to any pasteurised milk used in cheesemaking.

I was not able to find any studies making a comparison between raw milk versus pasteurised + lipase-enhanced milk. But it’s an interesting avenue for investigation.

Preference is Subjective: Raw milk advocates claim that raw milk produces superior tasting cheeses – without exception. And this claim clearly fails, as in one study a large majority of tasters preferred the pasteurized version of a cheese. Food preferences are personal, individual, and entirely subjective.

Yield: While there are flavour differences, at least in some styles of cheese, there are other factors of interest. The increase in yield is not likely of interest to the home cheesemaker, but is of some importance to commercial producers, where any improvement in efficiency counts.

Safety: On a more serious note, we should mention the risk of foodborne illness. I covered this in detail before, but its worth pointing out that of the four studies, only one looked for the presence of potential pathogens. And that study found them exclusively in the raw milk cheese. This is consistent with the studies I summarized in my previous post, where raw milk products were associated with a 300-to-1200 times increase in the rates of foodborne illness than were pasteurised products. Given the marginal differences in flavour and aroma between pasteurised and raw cheeses, and the potential to use additional aging or lipase to overcome these differences, I’d argue that producing cheese from raw milk is simply an unnecessary hazard.