Get access

Validity of Transdermal Alcohol Monitoring: Fixed and Self-Regulated Dosing


  • This research was supported by a grant from Alcohol Monitoring Systems, Inc., NIDA grants K08DA016314, DA09842, DA11015, DA12845, NIMH grant 5T32MH15442, and NIH grant M01#RR00051. Portions of this paper were previously presented at the annual meeting of the Research Society on Alcoholism (2005). The authors would also like to thank Dr. Boris Tabakoff and Dr. Robert Booth.

Reprint requests: Joseph Sakai, MD, Division of Substance Dependence, University of Colorado School of Medicine, 4200 East Ninth Ave, Box C268-35, Denver CO 80262; Fax: 303-315-0394; E-mail:


Background: To study the validity of transdermal assessment of alcohol concentration measured by a lightweight, noninvasive device.

Methods: Subjects wore a 227-g anklet that sensed transdermal alcohol concentrations (TACs) every 15 to 30 minutes, downloading results to a remote computer each day. Twenty-four subjects entered a laboratory and received a dose of 0, 0.28, or 0.56 g/kg of ethanol. Breath alcohol concentrations (BrAC) and TAC were measured every 15 to 30 minutes Twenty others [10 alcohol dependent (AD) and 10 not (NAD)] in the community who wore the anklet for 8 days kept a drinking log and provided a BrAC sample each day.

Results: In the laboratory, no zero-dose subject, and every subject receiving alcohol, had alcohol-positive TACs. The device distinguished low- and high-alcohol–dosing groups using peak (t14=3.37; p<0.01) and area under the curve (t14=3.42; p<0.01) of TACs. Within dosing groups, average TAC curves were broader (right-shifted) and had lower peaks than average BrAC curves. For community participants, self-reported number of drinks (t18=−3.77; p<0.01), area under the TAC curve (t9.5=−3.56; p<0.01), and mean TAC (t9.9=−3.35; p<0.01) all significantly distinguished the AD and NAD groups. However, individual transdermal readings were not reliably quantitatively equivalent to simultaneously obtained breath results.

Conclusions: Within the limits of the laboratory study, the device consistently detected consumption of approximately 2 standard drinks. On average, the device shows discriminative validity as a semiquantitative measure of alcohol consumption but individual readings often are not equivalent to simultaneous BrACs.