Search This Blog

Saturday, May 10, 2014

Diacetyl - What's the Real, Objective Danger?

This post posits that diacetyl alone may not be the killer that popular culture suggests.  Certainly I am not any sort of official "health" scientist but I think there are a lot of very, very "loose" ends which need a better explanation.

The question I want to ask is simple: is diacetyl really dangerous?  And what about someone vaping something that might contain minute traces of diacytel, i.e., from a flavor provider that might not label their ingredients properly.

In researching this post I discovered a DuPont 1936 patent on the details of an industrial process to create diacetyl (or butane-2,3-dione).

Diacetyl has been used in the food industry since, as best I can tell, at least the 1930's and probably well before.  Its used in brewing beer (see this and this as examples) and is naturally created by fermentation.  In 1983 the FDA deemed diacetyl as GRAS (Generally Regarded as Safe).

From this web site ( we find a rough timeline of how Bronchiolitis obliterans (popcorn worker's disease) was discovered (after roughly five decades of diacetyl used commercially):

March 12, 1985: NIOSH investigators visit an Indiana facility that produces flavors for bakeries, where two young, previously health, nonsmoking employees have been diagnosed with severe fixed obstructive pulmonary disease (consistent with bronchiolitis obliterans). Diacetyl is one of the chemicals commonly used at the facility.

February 16, 1993: Researchers for the German company BASF finish conducting a study in rats underwent a single 4-hour exposure to diacetyl vapors; all of them died at diacetyl concentrations > 23.9 mg/l.

May 19, 2000: Missouri Department of Health notifies OSHA that ten workers from one popcorn plant have bronchiolitis obliterans and asks OSHA to inspect the facility.

May 23, 2000: OSHA inspector visits the plant, but samples cannot be analyzed by OSHA’s laboratory.

August – November 2000: NIOSH investigates Missouri microwave popcorn facility; findings indicate that workers exposed to flavorings at the microwave popcorn plant are at risk for developing fixed obstructive lung disease.

December 2000: NIOSH issues interim recommendations to the Missouri microwave popcorn plant on ways to control workers’ exposure to the artificial butter flavoring

September 2001: NIOSH investigators return to the Missouri factory they studied to distribute materials describing investigation results, ongoing activities, and worker precautions.

September and December 2001: Attorney representing sick workers files complaints with OSHA, noting that worker health continued to decline after plant took measures recommended by NIOSH.

April 26, 2002: Scientists from NIOSH and the Missouri Department of Health publish an article in CDC’s Morbidity and Mortality Weekly Report describing their investigation; the article notes that "recent reports to CDC document bronchiolitis obliterans cases in the settings of flavoring manufacture" and that "preliminary animal studies at CDC suggest severe damage to airway epithelium after inhalation exposure to high air concentrations of a butter flavoring."


Fast forward to December of 2007.  Here we find this very interesting article in the Seattle-PI: "Flavor additive puts professional cooks at risk."

Now what's interesting is the assertions about the danger levels of diacetyl.

Let's look first at the EPA (see in the "The Pump Handle Blog" on diacetyl).

Here we see a fellow who eats two bags of microwave popcorn a day acquiring a case of bronchiolitis obliterans.  About six paragraphs down we see this "The exposure levels in his home were 0.5-3 parts per million after he popped the bags. This is comparable to levels in popcorn factories and evidently enough to damage his lungs."

0.5 to 3 ppm - So let's remember that.

(BTW - 1 ppm is roughly 1 drop in 13 gallons, 1 ppb (billion) is 1 drop in 13,000 gallons.)

Going back to our Seattle-PI article we find the following analyzing 22 of "the more popular cooking products. LabCor, a Seattle-based certified analytical laboratory frequently used by the Environmental Protection Agency."  (Note this is the level of "diacetyl in air samples captured from a heated skillet in which the products were tested.")
  • Two real butters were analyzed and diacetyl was found in a range of 7 ppm to almost 16 ppm.

  • In all the margarine and shortening products, levels of 7 ppm to almost 180 ppm were present.
  • A butter-flavored cooking spray released more than 164 ppm of diacetyl.
  • Butter-flavored cooking oils used by professional cooks ranged from 23 ppm to 234 ppm.
  • Two brands of oil for popping corn came in at 1,062 ppm and 1,125 ppm.
So these number vastly exceed the 0.5-3 ppm found in the home popcorn case.  (You can google others who have done similar studies - I found this detailed FDA report.)

Its fairly easy to find any number of sites that talk about the range (in ppm) of diacetyl in various popcorn and flavoring preparation plants (see this and this and this as an examples).  All relate values similar to the 0.5 to say 98ppm.  The last link (this) claiming "exposure of mice to 200 and 400 ppm diacetyl via inhalation for 6 hours per day over 5 days caused death."

Note this "200 to 400 ppm" is in line with the exposure level in the kitchen for butter flavored cooking oils and oils for popping corn indicated with Seatle-PI.

So why aren't the cooks and theater workers dying?

Well, for one thing a human male, on average, might weigh around 50kg (from this), or 50,000g as compared to a mouse at maybe 15g (see this).

Thus a human weighs about 3,333 times more than a mouse.

Since there are no vast die-offs in the kitchens (with perhaps 3.5 million cooks) and untold numbers of teenagers working in movie houses selling popcorn one has to believe that this difference in scale means much more diacetyl is probably necessary to kill humans.

So 3,333 x 3 ppm would be 10,000 parts per million or a 1% atmosphere of diacetyl.  Makes perfect sense (in the case of pop corn workers) you could inhale this much if you are bent over a large heated tank.

At least that's one explanation.

Another issue is if, according to the FDA report above, the Ridgeway plant, which started making microwave popcorn in 1986 apparently never had any evidence of health issues until 2002 (when the investigation was conducted). 

Now maybe there were issues and no one noticed - though there were no mysterious deaths or other activities to draw attention.

Or maybe something else changed in the plant (remember diacetyl has been in use since at least the 1930s) that in concert with diacytel began causing problems.

So the more interesting question is this: what is the objective danger of diacetyl?

Subjectively everyone can scream in panic that "diacetyl is killing us..." - but what is the real danger?

Clearly from 1986 to 2002 when the problems began there were no reported issues.

As best I can tell diacetyl has killed "a few scores of people" - some in the Netherlands and others in the US (see this OSHA publication).  Interestingly all around the same period of time.

Perhaps something changed in the industry relating to how microwave popcorn was made?  A sort of "new best practice?"

But back to diacetyl.

Automobiles kill about 100 a day (based on the best research I could do).

The CDC says 443,000 a year (or about 1,200 a day) die from smoking.

Objectively, at least compared other causes of death, diacetyl kills in the range of a rounding error on automobile deaths.

So what about vaping?

It gets more interesting here.

The other day I spoke with someone who mentioned that rigorous testing for diacetyl was a requirement in the new, proposed FDA vaping regulations.

There are two ways to look at this.

One, millions of people exposed to relatively large (objective ppm values) diacetyl every day do not get sick or die and have not for decades.

If vaping provides diacetyl at or below the levels seen in normal, everyday cooking then it seems unlikely, excluding other issues which I will mention below, that diacetyl is an issue.

If, on the other hand, vaping provides diacetyl at a rate above these cooking levels this it seems that it should be a concern.

Another concern is less is known about diacetyl substitutes (which are derived from, what else, diacetyl).  What would be used if no diacetyl is present?  A substitute of course.

Diacetyl is also present in cigarettes at high levels.  Is less diacytel better than more? (Yes, its burnt there so perhaps its not related - though no one has suggested its related to, say, cancer of the lungs...)

Other compounds such as acetoin can convert into diacetyl, i.e., there are compounds that may contain compounds that convert into diacetyl.

Another interesting issue is that much of this seems tied to very little initial work at the popcorn plants other than simply collecting data.

Much of diacetyl has supposedly been removed from various cooking products as a result of this investigation.  

Though potentially contradicting this is this link suggests biochemists are busily genetically engineering microbes to produce diacetyl.

(The BASF study isn't necessarily meaning full - many things breathed in a 1% atmosphere will kill you.)

My guess is that in the panic to find the "witch" for trial the real culprit escaped notice:  something to do with diacetyl, e.g., an interaction with another chemical, heating, a process change, etc.  Certainly problems were detected in other diacytel-using plants.  But there doesn't appear to be a uniform distribution of problems - both over time and location.

No comments:

Post a Comment