Breathalyzers and diabetes: how do breathalyzers work?

Breath testing of drunk driving suspects has been such a commonplace practice for so many decades, most of us hardly think to consider how they work. “Breathalyzers” are just a ubiquitous part of our understanding of police work. But in light of Rochester’s Deputy Mayor Redon’s DWI arrest, it’s worth considering the technology behind the most common field science instrument in modern police work. Can they really be tripped up by something so common as diabetes? How do breathalyzers work?

The first thing to know is that “Breathalyzer” is like “Kleenex,” in that it is a brand name that has become shorthand for all products of a type. The original Breathalyzer was invented in the 1940’s by a Dr. Bob Borkenstein (giggle) for the Indiana State Police. The brand and the technology are much less used today. New technologies have pushed this most basic field unit aside in favour of more admissible evidence.

The Theory

Regardless of the technology, all breath tests are based on very simple principle. Alcohol does not break down in the body and does not get digested. It passes through the blood stream and eventually into the alvioli in the lungs, where it evaporates when it comes into contact with air. As your body takes in oxygen, it expels a tell-tale quantity of alcohol. The rate of evaporation is directly proportional to the rate of concentration in the blood, based on an 1800’s constant known as Henry’s Law.

Simply put, by measuring the amount of alcohol on your breath, police can correctly identify the amount of alcohol in your blood stream. It is a simple ratio of 1:2100, or 2,100 milliliters of breath containing the same alcohol as 1 milliliter of blood.

However, it has also been well-established for decades that certain types of blood alcohol measurement can be thrown off by the introduction of other chemicals. Specifically, acetone which is commonly found on the breath of diabetics has been shown to inflate BAC measurements.

The question is: how do you measure the amount of alcohol in the breath? Broadly, there are three types of alcohol test in common usage: chemical catalyst, infrared spectroscopy, and fuel cell technology.

Chemical Catalyst

This is the Breathalyzer system, also sometimes referred to as the “wet method,” because it relies on chemical reactions between the alcohol on your breath and potassium dichromate. Alcohol turns the normally orangeish potassium dichromate greenish, then a photocell compares the mixed chemicals with a “control group” of unmixed potassium dichromate. Officers are required to dial a knob to measure the change between the two vials and determine a driver’s blood alcohol.

This system is scientifically sound and unaffected by acetone, but legally suffers from the amount of human interaction. Since an officer is required to do the measurement, DWI cases in the past have been thrown out on suspicion of tampering. For this reason, the Breathalyzer has often been given the derisive name “Dial-a-Drunk.”

Infrared Spectroscopy

When light hits a molecule, the bonds between the various atoms vibrate. As they vibrate, they emit light, the color of which depends on how far apart the bonds get as they go back and forth. The higher the bounce, the further up that ROYGBIV scale we all know and love. By hitting an unknown substance with a predictable wavelength of light, the color change can be measured and, based on prior research, pegged to specific chemicals with known vibration rates.

This is the basis for all spectroscopy. And because these sub-microscopic light shows are so consistent and don’t require any operator interaction, they make a perfect field sobriety test. Hence IS blood alcohol testers are in extremely common use the world over.

However, it was identified in the 1970’s that acetone can interfere with a BAC test that uses spectroscopy. That is because acetone and alcohol have very similar vibration rates that can easily be confused with one another. It is estimated that the inflation can be as great as .06 BAC, which is significant.

The solution to this problem for law enforcement has long ago been to use multiple wavelengths of infrared light. Because while the single rate of vibration between alcohol and acetone may be similar, the difference between their reaction to two different wavelengths is not. Thus most modern BAC testers such as the Datamaster DMT are at least theoretically immune to this line of defense.

The New Frontier: Fuel Cell Testing

The last type of testing seems for now to have the best of both worlds: it is unaffected by interfering chemicals in any way, and requires no human measurement. That technology is a fuel cell, which like it’s budding use in automobile power, works to separate electrons from a substance. In this case however, it separates the subject’s breath and instead of using the electrons to power a car, it measures the number to determine if alcohol is present.

The only trouble, as recently as two years ago, with using this type of technology is that the fuel cells may not last very long or very consistently. In Minnesota, defense lawyers effectively halted their use because of these problems.

So, diabetes and DWI? No luck?

Sorry, no. While it is certainly true that the single most commonly-used technology in BAC testing by law enforcement is the one technology that is subject to this flaw, the kinks have long-since been ironed out. The New York State Troopers that pulled Mr. Redon over use Datamaster DMT breathalyzers for the court-admissible sobriety test and those devices definitely use multi-wavelength testing.


Diabetes and alcohol: is it a factor in Deputy Mayor Redon’s DWI?

On Wednesday night, March 19th, Rochester’s Deputy Mayor was pulled over for doing 70 in a 55 and later found by a breathalyzer to have a .13BAC. He was arrested for DWI, speeding and also for an unrelated expired inspection. On Thursday, spox for the Mayor’s Office made the statement that “alcohol and diabetes don’t mix,” per @ashleyzilka of News 8:

Reports go on to state that Redon was recently diagnosed with diabetes, the potential explanation for the arrest being that perhaps being newly diagnosed, he may not have known what the consequences of drinking would be. @rachbarnhart summed up nicely:

I was diagnosed diabetic about five or six years ago. And with the necessary caveat that I am certainly not a doctor, I think I can add a bit of context to the situation. But there’s a lot to parse, here:

  1. Is Redon Type I or Type II?
  2. How does having diabetes affect drinking alcohol?
  3. Can you, as suggested by one source, test positive for intoxication without being drunk when you are a diabetic?

On the issue of whether Redon is Type I or II, we can certainly make the educated guess that he’s Type II. Type I, or juvenile diabetes, is usually diagnosed when a person is young. I don’t think I’ve encountered anyone who told me they were diagnosed with Type I later in life, only Type II.

Also worth noting that he appears to have lost some weight. That’s a pretty classic sign of late-diagnosis diabetes.

The type of diabetes is significant for two main reasons. The first is that Type II diabetics, at least of the untreated variety, have a problem with too much glucose in the blood stream, not too little. With treatment, however, the short-term effect is an extremely wonky blood sugar level that does fluctuate pretty wildly. So it is entirely plausible that a recently-diagnosed, recently medicated person can experience sudden drops in blood glucose.

But the second, related reason Type II diabetes is significant is because those sudden drops in glucose are scary as hell. Imagine going from perfectly sober to three-shots-of-tequila buzzed in about five minutes – without drinking – and you’ll see what I mean. Even if you are “used” to that feeling, it still causes panic attacks, because low blood sugar and anxiety are also related. My thinking is that most people would just pull over and panic, not speed up to 70.

Our second question is equally important. How does alcohol affect diabetics? While it is true that alcohol generally lowers blood suger, the real answer isn’t quite that straight forward. One important clue is what you drank. Brews and ferments like beer and wine tend to have enormously high levels of sugar and carbs, which means that in the short term, your blood sugar levels will go up, not down. Distilled spirits like vodka or whiskey have exceptionally low sugar content, and therefore are likely to lower your blood sugar.

So, the closest thing to a straight answer is that, if Redon had a beer or two, he would not have appeared drunk. If he had hard alcohol, well,.. maybe.

The final, and to me most intriguing, question is the whether a person can fail a breathalyzer test because they have low blood sugar? Is that really possible? Like a lot of lawyer tricks, this comes down to a pretty ambiguous “maybe.” There is scientific data to back up the basic claim, so let’s start there.

Hypoglycemia, or low blood sugar, occurs in all people, not just diabetics. When hypoglycemia happens, the body turns to alternative fuel sources. It begins converting fatty acids in the body into acetone, among other chemicals, which can be used by the brain as a fuel source. This condition is called Ketosis.

In this state, a breathalyzer will recognize the acetone incorrectly as alcohol.  According to one study, subjects tested showed as much as a .06 BAC which was actually acetone, not alcohol:

Hypoglycemia as a cause of acetone in the breath is well known and research has demonstrated that diabetics can have levels of acetone in the breath sufficiently high to register false readings of .06.

However, here come the caveats.

First, while hypoglycemia happens to everyone, Ketosis does not. It takes a fairly sustained or routine lowering of blood sugar for your body to hit the panic button like that. Wikipedia notes specifically that ketosis is more common among Type I diabetics. That makes sense, because they have been struggling with blood sugar for their entire lives. It’s worth noting here that it is the acetone created by ketosis that causes diabetics to go into comas, not the lowered blood sugar itself. So in order for Mr. Redon to have shown signs of ketosis, he would have had to be well past the point of being able to drive.

Finally and most damagingly to his case, while acetone could potentially get read as alcohol in a breathalyzer, the study I link to above only shows an increased reading of .06BAC. Mr. Redon’s breathalyzer came in at .13BAC. If we subtract our “handicap,” we still come out at .07, which would still get you a DUI. And of course, he has admitted to drinking, anyway.

Drinking and driving is no joke. Diabetes is no joke. Mayor Warren has not had a good run so far, and this isn’t even her appointment. But trying to paper over wrong doing by spreading misinformation about diabetes is not going to help anyone.