Somewhat sad to see Marion Nestle, with whom I almost always agree, linking, without question or comment, to an article in a pet-food trade journal which suggests that insect protein is a key solution to a sustainable pet food industry. The article contains some eye-opening numbers for the pet food business in the US and globally, as well as some dubious claims about pet health; they are not properly sourced, so I’m not going to bother to address them.

I do, however, take issue with this:

“In terms of sustainability, the key point is that this isn’t just greenwashing,” said Hobbs. “Insect protein in pet food truly has a significant positive impact.”

It is just greenwashing.

Hobbs is Aaron Hobbs, executive director of the North American Coalition for Insect Agriculture (NACIA), so Mandy Rice-Davies applies. The key point, which neither he nor Marion Nestle seem to have appreciated, but which you will because you listened to the recent episode on insects as food (for people and their pets), is that the “waste” that insects are reducing is usually a feed product that could be given direct to livestock and, in some cases, people and their pets.

Premium-priced insect-based pet food might assuage the consciences of some pet owners, but it is unlikely to do anything at all for food waste.

Yesterday’s graph from USDA is really interesting. It shows that the February 2022 recall of formula milk in the US, which compounded the supply chain difficulties of Covid, was associated with a striking increase in the number of infants fully and partially breastfed (and a drop in the number fully formula fed).

The survey covers only people enrolled in the USDA’s Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), so it isn’t comprehensive, but it represents a more vulnerable sector of society. It also shows that while formula offers the benefit of convenience, when push comes to shove many women are able to breastfeed more. Of course I would like to see a more detailed breakdown that factors in things like the need to be out working, but I find these results encouraging.

These findings also suggest a way out for mothers faced with what today’s Guardian said are “historically high” prices for formula in the UK.

The Competition and Markets Authority (CMA) found in November that the average price of infant formula had risen by 25% in the past two years and families could save more than £500 over the first year of a baby’s life by switching to cheaper powders.

They could save even more by breastfeeding more and for longer.

Syndicated to the mothership.

Perhaps this will help while away a few minutes as you wait for a new episode to drop. One question for each of this year’s 18 episodes. There will be prizes, but I want to try and ensure that they are things you actually want, so probably tokens or gift cards of some sort. You have until midnight GMT on 1 January 2024 to submit your entry

NB: I need an operational email to ensure I can tell you if you are a winner. That is all I will use it for.

The available figures on honey adulteration are pretty alarming: 46% of samples in the EU, 100% of honey exported from the UK, more than a quarter of Australian samples “of questionable authenticity”. However, as Matt Phillpott pointed out in a recent episode of Eat This Podcast, one of the great difficulties honey poses is that it is so variable. All of the many “natural” components of honey vary from batch to batch, hive to hive, season to season, so that while a specific “unnatural” chemical might unambiguously signal adulteration, other kinds of evidence are a lot less cut and dried.

That kind of uncertainty spurred the Government Chemist, the UK’s official food inspector, to look at the evidence behind a claim in the Daily Mail that “Supermarket brands of honey are ‘bulked out with cheap sugar syrups made from rice and corn’”. Behind that headline, with its exculpatory quotation marks, the Mail back-pedalled slightly[1] and allowed the supermarkets to cast doubt on both the results and the robustness of the tests, carried out by a German laboratory.

What to do

The Mail’s article was based on a report from the German company, which did not contain all the data from the tests, but rather an opinion on what those tests showed. To some extent, that’s what the supermarkets were arguing against, saying that their own tests were more reliable, or more trustworthy, or something.To try and settle the matter the Government Chemist posed two questions:

1) Is it acceptable to report an adverse interpretation without exhibiting all the supporting data? (2) How may a valid overarching authenticity opinion be derived from a large partially conflicting dataset?

Two papers in the journal Science of Food offer answers.[2]

Opacity

In the first, the researchers looked at the details in three sample Certificates of Analysis (CoA) from the German laboratory, focussing on the specific measurements, and they pick up several inconsistencies. This is not the place to go into detail — there’s plenty in the papers — just to note that the laboratory described some samples as “compliant” even though the measurements were not different to those from samples described as “non-compliant”. These differences point to the larger problem of identifying adulterated honey: honey’s natural variability, made worse by disagreement between different analytical techniques.

The paper cites, for example, a small study in which Danish beekeepers sent 14 samples (definitely not adulterated) to two different companies, including the German laboratory. Both companies said that 4 of the 14 were adulterated. The beekeepers say that is impossible (allowing for the possibility of emergency feeding sugar solution ending up in the honey). The companies disagreed on another four samples; one found evidence of possible fraud, the other did not. The beekeepers asked for a different analytical test to be carried out on seven of the eight samples flagged as possibly adulterated; all seven were found to be “not adulterated and without feed residues”.

The Government Chemist’s paper concludes that “The summary opinion of the reporting laboratory … was unequivocally that the samples were noncompliant. However our critical examination of the CoA data reveals a much more nuanced picture from which it is currently difficult to draw a definitive opinion on the authenticity of the samples examined.”

Laboratories should make their detailed analyses available, rather than their summary judgements.

Forensics

The second paper offers a potential way forward, by looking at the balance of probabilities and presenting conclusions suitable for a legal decision. That seems sensible given that, at least in the UK, food law is a matter for criminal justice. This approach, known as evaluative reporting, follows three basic steps.

First is a question for the analytical laboratory: is the test sample typical or atypical compared to a reference set of typical samples? One value is unlikely to be enough, and even a typical value does not let the sample off the hook as it may be that the adulterant is too dilute to be detected, or may not be detectable at all with the chosen technique. Likewise, a single atypical value does not condemn the sample, but should trigger extra tests, as in the Danish beekeepers’ study.

Second is to put some sort of number on the likelihood that an atypical result is indeed the result of adulteration rather than chance. This is the likelihood ratio, a statistical measure that compares the probability of the result being true under the two conditions, actual adulteration vs an “accident” not the result of adulteration. If the two probabilities are roughly the same, then clearly there is not much support for the charge that adulteration is to blame. If the probability of the accidental hypothesis is much lower than the probability of the adulteration hypothesis, support for the “truth” of adulteration is stronger.

A key point is that one can convert the likelihood ratio into a verbal form, from “moderate support” where the probability of the accidental hypothesis is up to 100 times less than the adulteration hypothesis to “extremely strong support” where the accidental hypothesis is more than 10,000 times less likely than the adulteration hypothesis.

Stage three then would be a decision, based on the all the evidence, to do something about the sample.

Yes, but …

That’s all well and good, and the paper offers a couple of worked examples to show why, for instance, a summary that the German laboratory based on its measurement of the enzyme diastase was flawed. Using a likelihood ratio requires having a good estimate of the probabilities of certain markers being present in genuinely unadulterated samples. Caramel, for example, may be added by adulterers to mimic the dark colour of forest honey. It could also, perhaps, be present in unadulterated samples, though it is hard to imagine how. In any case, to be able to offer strong or very strong support for the proposition that caramel signals adulteration, it would have to be undetected in at least 1000 samples of pure honey.

The battle against fraudulent honey will thus need, in addition to new analytical techniques to detect new cunning adulteration, the development of agreed databases representing typical measurements for a wide range of honeys. To that end, a month ago the Government Chemist published a Protocol for the Collection of Honey Reference Samples, a global first. The hope is that the protocol “will help in standardising how authenticity databases are built and curated and will lead to more trust in them”.

That’s unlikely to stop adulteration, which remains far too profitable, or satisfy supermarkets, who all want to proclaim their own monitoring as superior. However it might, eventually, with the recommendations of how to interpret and present the data, bring some adulterators to a form of justice.

In the episode on Jewish Food in Rome I made much of the fact that many Jewish Roman dishes are found in restaurants across Rome and beyond. Not just carciofi alla giudia but others that are probably not recognised as Jewish, like cicoria ripassata and aliciotti con l’invidia. By contrast, one that has stayed in the former Ghetto, and indeed at a single location as far as I know, is pizza ebraica.

A slice of Jewish pizza is a chunky bar about 10cm long that is a dense confection of almonds and pine nuts with dried and candied fruit and raisins, baked very dark and crunchy. It is sold by the kosher bakery Boccione, which these days generally has a line out the door. A little goes a long way.

Sweet “pizza” is not an historic abomination, unlike the Nutella-topped creations across the river in Trastevere. Bartolomeo Scappi has recipes in his 16th century treatise, and the word pizza can encompass all sorts of dishes. Pizza ebraica is more formally known as pizza di beridde, traditionally baked to celebrate the bris, or circumcision, of a baby Jewish boy. Beridde is the Roman Jewish dialect for bris.

Boccione, despite being tiny, is unmissable even if there isn’t a line to point the way. It is on the corner as you enter the main street, via del Portico Ottavia, at the far end from the synagogue. And, unsurprisingly, the building has been there a long time, longer even than the 200 years that the bakery has been there.

I was looking at a 1752 engraving by Giuseppe Vasi of the Piazza Giudia, which shows the Ghetto gate to the right, and my own photograph of Boccione, taken on a recent “research” visit. It’s a little tricky to get oriented, because the big fountain is no longer there, having been moved about 50 metres closer to the river in 1930 by order of Pope Pius XI. On the left of the engraving, though, is that distinctive chopped-off corner of the building and the number 3 and a Latin inscription. The words on the engraving look slightly different, but we all know artists edit what they see while the camera, of course, never lies.

So there it is, the forno del Ghetto, unique source of the stunningly good Jewish pizza. You can of course make it yourself, and dubious reverse-engineered recipes can be found online, but seriously, you owe it to yourself to buy a slice of the genuine article.