Arsenic in wine: A news update, but not a scientific one

The news this morning is full of pieces on Kevin Hicks, proprietor of a consumer-oriented wine analysis company called Beverage Grades, and the class action lawsuit he’s bringing against multiple California wineries for selling wine with arsenic concentrations far exceeding what the US Environmental Protection Agency allows in drinking water. Many of the reports are emphasizing the “wine may kill you” side of this story.

Mr. Hicks contacted me by way of an email and a suggested arsenic-in-wine story — for which he advised me I could pay him by check or PayPal — a little over a year ago, which prompted me to write this post on whether wine consumers should be concerned. My admittedly brief scan of the literature suggests that no new scientific research on wine and arsenic has been published in English since then. I’ll stand by what I wrote in 2014. The key points:

  • Arsenic is definitely found in wine. It’s also found in many other foods and beverages. Arsenic is, in fact, naturally present in water and soil, and unless you’re part of a special population, drinking water is your primary source of dietary arsenic.
  • Researchers have found evidence of higher arsenic intake in wine drinkers, but also in people who drink beer and who eat rice, fish, and/or Brussels sprouts. (Exemplar references here and here.)
  • The FDA regulates arsenic levels in apple and pear juice, but not (yet) in wine. Dietary arsenic isn’t well understood, and whether we have good evidence for the current “safe” cut-offs and what those cut-offs should be has been discussed for decades.
  • Our current best evidence indicates that arsenic in wine isn’t a health concern. It’s fair to say that every food and drink we consume brings minor amounts of potentially harmful substances into our bodies. Risk assessments say that the amount of arsenic in wine doesn’t pose a threat to consumers. (Exemplar references here and here)

The wines indicted in Hicks’ lawsuit weren’t a major part of the studies I’ve listed above. His data may show something amiss with these specific wines, but he hasn’t shared either his methods or his data. When I looked at Beverage Grades a year ago, I was disturbed by the complete lack of detail offered to back up awarding specific wines badges like “HealthyPour™.”

I’m uncomfortable with Kevin Hicks and Beverage Grades’ tactics of withholding rather than being transparent with information, damaging at the best of times and ironic in light of his accusations. If third-party labs can back up Hicks’ claims in the course of this lawsuit, we may well have something to talk about: ways to reduce arsenic levels in wine, new regulations, and/or renewed scrutiny of the EPA guidelines. But until then, the healthiest thing to avoid is likely the inflammatory news headlines.

How to replicate a wine from 1500 year-old grape seeds

How to replicate a 1500 year-old Negev wine: 

  1. Unearth 1500 year-old grape seeds from the famed Byzantine-era city Halutza. (Done; credit Israeli archaeologists.)
  2. Extract DNA; sequence. (This is the easy part.)
  3. Check sequence against known contemporary grape genomes (to infer evolutionary relationships, and to double-check that seeds really weren’t refuse from archeologist’s lunch before proceeding).
  4. Realize that DNA sequence from seeds doesn’t represent DNA sequence of original cultivated vines, because Vitis species are notoriously mutation-prone, Halutza vignerons would probably have worked with grafted vines and selected varieties the way we do now (cf. Theophrastus who wrote about grafting in Athens in the 3rd c. B.C.). Deal, because we can infer a lot about the parent vine from the seed DNA but the stuff we can’t infer can definitely change wine quality (cf. reason why every contemporary commercial grapevine is propagated by cutting and not from seed).
  5. Synthesize fresh, lab-manufactured DNA from the old sequence (there are companies for this).
  6. Get someone like Dr. Andy Walker at UC Davis to (get his grad students to) clone new DNA into grape stem cells.
  7. Plant lots of babies. Watch lots of them die. Wait.
  8. Try to find a climate similar to c. 500 AD Halutza for vine cultivation (consult archeological meteorologist?) Hope that soil types also have something in common, or try to replicate historically appropriate soil.
  9. Accept that since grapes are subject to lots of somatic mutations, the new grape vines may change a bit along the way and not be exactly like the original Negev wines. Keep dealing.
  10. Consult Dr. Patrick McGovern (biomolecular archaeologist specializing in fermented beverages; Dogfish Head collaborator/co-creator of Midas’s Touch, etc.; outstandingly nifty person) about mimicking historical winemaking. Do what he says.
  11. Hope that Dr. McGovern has advice on how to handle the problem of modern Saccharomyces cerevisiae being a darn sight different than whatever might have been around to ferment things in the Byzantine Negev.
  12. Option A: Backtrack 15-20 years and clone person using DNA from bones found at Halutza dig. Raise in a bubble (consult designers of The Truman Show) to mimic growing up in Byzantine Halutza. Make sure he doesn’t try any wine until the replicated stuff is ready, then get him to write tasting notes. Option B: Accept that even if we’ve perfectly replicated an ancient wine (see above caveats about genetic variation, yeast, etc), we won’t actually have replicated the wine because we, the drinkers, will be different, and the wine only fully exists in its drinking and enjoying via the participation of complex sensory and thinking apparatus attached to a subjective human being.
  13. Drink the darn wine anyway. Invite Robin Trento over to make dinner (ask her to bring her own garum). Look up retirement residence addresses for the journalists who were “ready for a taste of the Byzantine Empire’s favorite wine” back in 2015 and make sure they get a bottle.

Microbial terroir? The media gets it wrong again (surprise)

As has by now been widely publicized in wine circles and elsewhere, Dr. David Mills and graduate student Nicholas Bokulich of UC Davis have just published a journal article (in the Early Edition of the Proceedings of the National Academy of Sciences here) demonstrating that populations of bacteria and yeast associated with wine grapes vary geographically in organized and predictable ways. Bokulich collected samples across California, isolated bacteria and yeast from those samples, sequenced bits of their DNA, and then looked for patterns.

This is a beautiful, extremely strong study with useful implications. Unfortunately but unsurprisingly, the news headlines are getting much of it wrong. The New York Times article on the subject* sets up the long-standing American skepticism about terroir, then proclaims that “American researchers may have penetrated the veil that hides the landscape of terroir from clear view, at least in part. They have seized on a plausible aspect of terroir that can be scientifically measured – the fungi and bacteria that grow on the surface of the wine grape.” This is the kind of unscientific media hogwash that contributes to people on the street having balderdash-worthy ideas about how science works. I can’t blame the article’s author for being one more journalist who doesn’t understand science (or who knows better but still needs the sexy story). I can blame the NY Times for not having the sense or taking the effort to get someone with enough research knowledge to cover the story well (Inside Scoop SF did, though they do have Jon Bonné). You’re the NY Times; you folks can do this.

Problem #1: Bokulich and Mills’ findings don’t actually say anything about terroir as we typically talk about it: in terms of sensory impact. The paper describes regional variations. The paper doesn’t connect those variations with any element of wine quality, perceptible or otherwise.

Problem #2: This is not the first time researchers have attempted to quantify some element of terroir. Not even close. Geologists and pedologists (soil scientists) have done a lot of good work looking at soil structure, depth, aspect, and so forth. Other microbiologists have looked at differences in bacteria and yeast across space. Bokulich’s study is exceptionally strong, but it’s not as earth-shatteringly unique as the media are making it out to be.

Problem #3: A technical point, but Bokulich and Mills didn’t actually look at microbes on the surface of grapes. They collected grape musts, which for them meant “destemmed, crushed grapes, representing a mixed, aggregate sample of all grapes from an individual vineyard block” collected after ordinary stemming and crushing operations at the winery. The strength of looking at musts is in having a sample that reflects averaging across a vineyard block and does away with potentially idiosyncratic variations between individual grapes. The main weakness is that we’re a lot less sure of where the microbes came from. What if some of the microbes came from the winery equipment or from handling operations instead of being present on the grapes in the vineyard? Significant regional patterns correlated with environmental factors – precipitation and temperature, for example – but we still can’t actually pinpoint where those microbes are originating.

The NY Times article does a pretty good job of summarizing the original PNAS paper. Kudos to it’s author for talking a bit about the methods behind the findings, for observing near the end of the piece that “the Davis scientists still need to prove that these microbes affect the quality of the wine,” and for calling up Dr. Thomas Henick-Kling at Washington State University for a second opinion. The problem is in the headline and the first few paragraphs which are, of course, what get picked up and misconstrued by everyone else.

This is a fine example of a frequent pattern in news science coverage. Researchers publish a paper on a sexy topic like wine or cancer, and – being like other humans – their conclusions about the implications of their findings might take a few steps over the bounds of reasonability. The university and/or the academic journal puts out some kind of press release, highlighting the sexy bits and the in-our-dreams implications. Journalists pick up on the sexy bits and elaborate even sexier hooks and headlines around them. The hooks and headlines get picked up by less reputable news replicators and on Facebook and Twitter. And by now we’re wandering around smack-dab in the middle of unreasonableness territory.

A chicken-and-egg problem: do we get headlines like this because national science literacy is bad, or does poor science literacy stem (in part) from the uncritical quality of our media? Either way, there’s improving to be done here.

  • Wine-Searcher’s coverage has to get an honorable mention, not only for its especially unreasonable tag line – “New research suggests that bacteria and fungi could be as important in the expression of regionality as soil and climate” – but for referring to what we all know and love as Botrytis cinerea or Noble rot by the show-worthy name of its anamorph (another form of the same fungi), Botryotinia fuckeliana.