Saliva and Forehead Sweat Could Detect Recent Cannabis Use in Impaired Drivers | Research

The takeaway

Among 198 injured drivers, THC was detectable in saliva and forehead sweat from most cannabis-positive individuals, with sweat slightly more sensitive than saliva, offering non-invasive alternatives to blood testing.

Read this if you want to understand how roadside cannabis testing works and why saliva is preferred over urine for detecting recent use.

Forehead sweat detected THC in 73% of cannabis-positive drivers

What the researchers found

Researchers collected blood, urine, saliva, and forehead sweat simultaneously from 198 injured drivers in Strasbourg, France. Of 22 drivers who tested positive for cannabis metabolites in urine, 14 had detectable THC in saliva and 16 had detectable THC in forehead sweat.

Saliva THC concentrations ranged from 1 to 103 ng per collection device, while forehead wipe concentrations ranged from 4 to 152 ng per pad. Neither metabolite (11-OH-THC and THC-COOH) was detected in saliva or sweat, only the parent compound THC.

The advantage of saliva and sweat over urine is that they detect THC itself rather than metabolites, better indicating recent use and likely current impairment. The non-invasive collection was practical for roadside situations. However, limitations included smaller specimen volumes and the absence of sufficiently sensitive immunoassays for roadside screening.

Why it matters

This study advanced the development of non-invasive cannabis impairment testing for drivers. By detecting parent THC rather than metabolites, saliva and sweat testing better reflects recent use than urine, where metabolites can persist for days to weeks after last use.

The numbers in context

198 drivers. 22 urine-positive. 14 saliva-positive (64%). 16 sweat-positive (73%). Saliva THC: 1-103 ng. Sweat THC: 4-152 ng.

How the study worked

Cross-sectional study of 198 injured drivers at a Strasbourg emergency hospital. Four specimen types collected simultaneously. GC-MS analysis for THC in saliva (Salivettes) and forehead wipes (cosmetic pads).

What this study cannot tell us

Only 22 cannabis-positive drivers in the sample. Saliva and sweat missed some urine-positive drivers (36% and 27% false negative, respectively). No immunoassay suitable for field screening existed at the time. THC can deposit in the oral cavity from smoking, confounding the distinction between systemic and local contamination.

How to read the evidence

A forensic toxicology study with simultaneous multi-specimen collection. Good methodology but small number of cannabis-positive cases.

When this study was published

Published in 2000. Oral fluid testing technology has improved substantially, with commercial devices now available for roadside use in multiple countries.

The bigger picture

This early work on alternative matrices contributed to the development of oral fluid drug testing devices now used in some jurisdictions for roadside cannabis screening. The principle of detecting parent THC for recent use remains central to impaired driving enforcement.

Questions still open

Have roadside saliva THC tests improved since 2000? What THC concentration in saliva correlates with impairment? Can sweat testing distinguish between users and secondhand exposure?

Common questions

Can saliva tests detect recent cannabis use?

Yes. Saliva detects THC itself (not metabolites), which is a better indicator of recent use than urine tests that can stay positive for days to weeks.

Is sweat testing better than saliva?

In this study, forehead sweat detected slightly more cannabis-positive drivers (73%) than saliva (64%), but both were useful alternatives to blood testing.

Read the original research

Detection of cannabis in oral fluid (saliva) and forehead wipes (sweat) from impaired drivers.

Journal of analytical toxicology, 24(7), 557-61

View on PubMed ↗

Citation

Kintz, P; Cirimele, V; Ludes, B. (2000). Detection of cannabis in oral fluid (saliva) and forehead wipes (sweat) from impaired drivers.. Journal of analytical toxicology, 24(7), 557-61.