False-positive breath-alcohol test after a ketogenic diet
and S Ro
Department of Forensic Chemistry, National Board of Forensic Medicine and University Hospital, Linko
ping, Sweden and
Department of Obesity Research, Karolinska Hospital, Stockholm, Sweden
A 59-year-old man undergoing weight loss with very low calorie diets (VLCD) attempted to drive a car, which was fitted with an
alcohol ignition interlock device, but the vehicle failed to start. Because the man was a teetotaller, he was surprised and upset by
this result. VLCD treatment leads to ketonemia with high concentrations of acetone, acetoacetate and b-hydroxybutyrate in the
blood. The interlock device determines alcohol (ethanol) in breath by electrochemical oxidation, but acetone does not undergo
oxidation with this detector. However, under certain circumstances acetone is reduced in the body to isopropanol by hepatic
alcohol dehydrogenase (ADH). The ignition interlock device responds to other alcohols (e.g. methanol, n-propanol and
isopropanol), which therefore explains the false-positive result. This ‘side effect’ of ketogenic diets needs further discussion by
authorities when people engaged in safety-sensitive work (e.g. bus drivers and airline pilots) submit to random breath-alcohol
International Journal of Obesity (2007) 31, 559–561. doi:10.1038/sj.ijo.0803444; published online 8 August 2006
Keywords: acetone; alcohol; breath-test; driving; ignition interlocks; VLCD
Obesity constitutes a serious threat to health and longevity
and among various treatment options, very low calorie diets
(VLCD) are frequently used.
Such diets provide essential
proteins and fats but negligible amounts of carbohydrates
and they typically furnish 800 kcal/day. After a few days of
dieting, fat becomes the main source of energy, and VLCD
regimens are consequently ketogenic.
Ketone bodies (acetone, acetoacetate and b-hydroxybuty-
rate) increase appreciably in the blood of people on VLCD.
Acetone is a water-soluble volatile product of metabolism
and is therefore exhaled in the breath and excreted in the
urine. Indeed, monitoring breath-acetone has been advo-
cated as a way to ensure that patients comply with their
The elimination half-life of acetone in man is fairly long
(15–25 h) and the biosynthesis and metabolic fate of this
endogenous metabolite are summarized in Figure 1.
ing ketonemia, the reduction pathway toward isopropanol
becomes a strong possibility and, indeed, this secondary
alcohol has been identified in blood of patients with
hyperglycemia and poorly controlled diabetes.
conversion of acetoacetate into b-hydroxybutyrate and the
reduction of acetone to isopropanol are both nicotinamide
adenine dinucleotide (NAD)-dependent redox reactions.
Moreover, administration of amino acids, precursors of
proteins, can accelerate the elimination of ethanol from
blood by enhancing activity of hepatic alcohol dehydrogen-
An increased ADH activity after eating high
protein diets might help to promote reduction of acetone to
Among various strategies to reduce drunk driving and
improve road traffic safety, the use of alcohol ignition
interlock devices shows great promise.
Such devices are
increasingly being fitted to buses and other public transpor-
tation vehicles as well as long-haul trucks and also in some
private cars, especially in Sweden.
Incentives to install
ignition interlock systems in private cars include lower
insurance costs and earlier return of the driving permit to
people convicted of drunk driving and especially to control
Most of the ignition interlock devices used today measure
alcohol (ethanol) in a person’s breath by electrochemical
oxidation. Endogenous breath volatiles like acetone are not
oxidized at the same electrode potential.
secondary alcohol isopropanol (2-propanol) is oxidized at a
slightly faster rate than ethanol and these two alcohols
cannot be distinguished.
Accordingly, if acetone is reduced
to isopropanol during ketonemia, there is a strong possibility
of false-positive results when ignition interlocks are used.
Indeed, the concentration threshold for a positive test and
Received 18 May 2006; revised 31 May 2006; accepted 2 June 2006;
published online 8 August 2006
Correspondence: Dr AW Jones, Department of Forensic Chemistry, National
Board of Forensic Medicine and University Hospital, Artillerigatan 12,
Linko¨ping 581 33, Sweden.
International Journal of Obesity (2007) 31, 559–561
2007 Nature Publishing Group All rights reserved 0307-0565/07
failure to start the engine is often set fairly low, correspond-
ing to a blood alcohol concentration (BAC) of 0.01–0.02 g/
100 ml (10–20 mg/100 ml).
We report a case of a 59-year-old man, body mass index
, who began a weight reduction program, partly
because of knee pains but also because he was a glider pilot
where weight is important. He used a Swedish textbook on
obesity treatment written by S Ro
ssner together with the
commonly used Swedish VLCD Nutrilett (Cederroths,
Stockholm, Sweden), 5 packets/day for 3 weeks, which is an
approved standard regimen. This treatment resulted in a
weight loss of 7 kg.
During dieting, the man discovered that an alcohol
ignition interlock device, installed in an official company
car, indicated that he had consumed alcohol and the vehicle
failed to start. This was confusing because the man was a life-
long teetotaller and was therefore both surprised and upset
by the result. As he had been supervising private aviation
he had access to a second breath-alcohol analyzer, which
indicated a simultaneous BAC ranging from 0.01 to 0.02 g/
In an attempt to understand the reason for the positive
breath-test result, which obviously caused some discomfort
and practical problems, the man contacted the Obesity Unit
(Karolinska University Hospital, Stockholm, Sweden) for
advice, and the mechanism was elucidated. Although we
did not have the opportunity to measure acetone and
isopropanol directly in this subject, the most plausible
explanation for the positive breath-alcohol test is reduction
of acetone to isopropanol, which then undergoes electro-
Most countries enforce statutory BAC limits above which
it is an offence to drive a motor vehicle. These limits differ
between countries owing to tradition, lifestyle and political
The punishable BAC limits for driving range
from as low as 0.02 g/100 ml in Norway and Sweden to
0.05 g/100 ml in most European countries and 0.08 g/100 ml
in UK, Ireland, USA and Canada. It seems important
therefore to consider the consequences of ketogenic diets
when blood- and breath-alcohol tests are interpreted in a
Suspected drunk drivers first submit to a roadside breath-
alcohol screening test and if this is positive they provide
either an evidential breath-alcohol test or a blood specimen
is taken for laboratory analysis. Breath-alcohol screening
tests incorporate electrochemical detectors similar to those
used in the ignition interlock device and therefore respond
to isopropanol. By contrast, most evidential breath-testing is
performed by multifilter infrared analysis and these are
programmed to abort the test if acetone is detected on the
suspect’s breath above a certain threshold value.
the half-life of isopropanol (t
¼ 3–5 h) is much shorter than
that of acetone (t
¼ 15–25 h), it is hard to envisage finding
elevated concentrations of isopropanol without concomi-
tant high concentrations of acetone. However, evidential
breath-alcohol analyzers based on electrochemical oxidation
cannot distinguish ethanol from isopropanol and this
resulted in a false-positive test after VLCD. An apparent
BAC of 0.02 g/100 ml seems likely according to the present
The reduction of acetone to isopropanol is not a problem
with blood-ethanol determination because gas chromato-
graphy is used and this highly specific method can resolve
ethanol from both acetone and isopropanol under normal
In conclusion, we suggest that people on ketogenic diets
run the risk of false-positive breath alcohol tests owing to
reduction of acetone to isopropanol. People on VLCD need
to be warned about this artifact when alcohol ignition
interlock devices are used. This possibility also warrants
consideration in connection with workplace alcohol testing
and screening of drunk drivers with electrochemical sensors.
Both the manufacturers of ignition interlock devices and
government agencies that monitor performance and admin-
ister sanctions should consider these problem. Technological
improvements might be possible, for example, by measuring
not only the final reading but also the kinetics of the
detector response to different alcohols.
There was no external funding for preparing this article and
neither author considers there to be any conflicts of interest.
1 Wyatt SB, Winters KP, Dubbert PM. Overweight and obesity:
prevalence, consequences, and causes of a growing public health
problem. Am J Med Sci 2006; 331: 166–174.
2 Bravata DM, Sanders L, Huang J, Krumholz HM, Olkin I, Gardner
CD et al. Efficacy and safety of low carbohydrate diets: a
systematic review. JAMA 2003; 289: 1837–1850.
Figure 1 Biosynthesis and metabolic fate of acetone after ketogenic diets.
FFA ¼ free fatty acids; ADH ¼ alcohol dehydrogenase; NAD
and NADH are
oxidized and reduced forms, respectively, of the coenzyme nicotinamide
adenine dinucleotide; CYP2E1 ¼ cytochrome P450 isozyme.
VLCD cause false-positive alcohol test
AW Jones and S Ro¨ssner
International Journal of Obesity
3 Shah P, Isley WL. Ketoacidosis during a low-carbohydrate diet.
N Engl J Med 2006; 354: 97–98.
4 Beisswenger BG, Delucia EM, Lapoint N, Sanford RJ, Beisswenger
PJ. Ketosis leads to increased methylglyoxal production on the
Atkins diet. Ann NY Acad Sci 2005; 1043: 201–210.
5 Musa-Veloso K, Likhodii SS, Cunnane SC. Breath acetone is a
reliable indicator of ketosis in adults consuming ketogenic meals.
Am J Clin Nutr 2002; 76: 65–70.
6 Jones AW. Elimination half-life of acetone in humans; case reports
and a review of the literature. J Anal Toxicol 2000; 24:8–10.
7 Kalapos MP. On the mammalian acetone metabolism: from
chemistry to clinical implications. Biochim Biophys Acta 2003;
8 Jones AE, Summers RL. Detection of isopropanol in a patient with
diabetic ketoacidosis. J Emerg Med 2000; 19: 165–168.
9 Bailey DN. Detection of isopropanol in acetonemic patients not
exposed to isopropanol. Clin Toxicol 1990; 28: 459–466.
10 Jones AW, Andersson L. Biotransformation of acetone to
isopropanol observed in a motorist involved in a sobriety control.
J Forensic Sci 1995; 40: 686–687.
11 Lisander B, Lundvall O, Tomner J, Jones AW. Enhanced rate of
ethanol elimination from blood after intravenous administration
of amino acids compared with equicaloric glucose. Alcohol Alcohol
2006; 41: 39–43.
12 Beirness DJ, Marques PR. Alcohol ignition interlock programs.
Traffic Inj Prev 2004; 5: 299–308.
13 Bjerre B. An evaluation of the Swedish ignition interlock
programme. Traffic Inj Prev 2003; 4: 98–104.
14 Falkensson M, Jones AW, So
rbo B. Bedside diagnosis of alcohol
intoxication with a pocket-size breath-alcohol device: sampling
from unconscious subjects and specificity for ethanol. Clin Chem
1989; 35: 918–921.
15 Jones AW. Medicolegal alcohol determinations – blood or breath
alcohol concentration? Forensic Sci Rev 2000; 12: 23–48.
VLCD cause false-positive alcohol test
AW Jones and S Ro¨ssner
International Journal of Obesity