A person’s age or sex cannot be used as a shortcut for estimating a blood amphetamine concentration from the 2007 study by Alan Wayne Jones. Among 300 apprehended drivers, increasing age had a weak statistical association with concentration, while the reported sex comparisons were not statistically significant. The study is useful for understanding population summaries, rather than predicting an individual result.
Who was included?
This historical observational study selected 300 Swedish suspected drug-impaired drivers whose testing identified amphetamine as the only psychoactive drug in blood. There were 246 men and 54 women. These were selected forensic cases, not a random sample of the public, a controlled dosing experiment, or a study of prescribed treatment.
The laboratory screened for specified drug classes and used more specific methods to confirm positive findings. Blood amphetamine was quantified by gas chromatography–mass spectrometry with a deuterated internal standard. The routine limit of quantitation was 0.03 mg/L. “Only drug identified” must be understood within the analytical scope; it is not proof that every possible substance was excluded.
A significant age association can still be weak
The reported age–concentration correlation was r = 0.18, with p < 0.05. The association was statistically significant, but its magnitude was small. Statistical significance addresses evidence against a specified null model; it does not mean age accurately predicts an individual’s concentration.
Squaring the reported correlation gives approximately 0.032, or 3.2%. As a descriptive calculation for a simple linear relationship, that illustrates the limited shared variation. It is not a validated prediction model, a causal estimate, or a measure of driving impairment.
Doses, routes and elapsed times were not assigned by the investigators. An observed association with age therefore cannot identify slower elimination as its cause. Differences in exposure histories, selection into the forensic sample and other unmeasured characteristics remain possible explanations.
Preserve the nonsignificant sex comparison
| Recorded group | Number | Mean blood amphetamine | Median |
|---|---|---|---|
| Men | 246 | 0.97 mg/L | 0.8 mg/L |
| Women | 54 | 1.11 mg/L | 1.0 mg/L |
The women’s mean was numerically higher, but the paper reported no statistically significant difference using either Student’s t-test or the Mann–Whitney test. Describing women as having a demonstrated higher concentration would overstate the result. Conversely, failure to find significance does not prove equivalence between the groups.
The unequal group sizes matter when assessing precision and generalizability. These summaries do not establish a biological explanation, a sex-specific dosing rule, or a correction factor for forensic interpretation.
A sample distribution is not a reference interval for safe driving
Across all 300 cases, the mean was 1.0 mg/L, the median 0.9 mg/L, and the reported 2.5th and 97.5th percentiles were 0.06 and 3.0 mg/L. Those percentiles describe this selected distribution. They are not therapeutic boundaries or a range in which drivers can be presumed safe.
The paper also examined clinical impairment in a separate, smaller retrospective sample of 70. That denominator should not be substituted for the 300-person demographic analysis or described as 300 clinical examinations. Age, concentration and an impairment classification are different variables.
What should an expert ask?
Identify which population generated the statistic, which analyte and specimen were measured, whether the comparison was significant, and whether prediction was actually validated. For a particular person, the measured result and its uncertainty take precedence over a demographic expectation. Dose, timing and functional conclusions each need additional support.
Scientific source
Jones AW. Age- and gender-related differences in blood amphetamine concentrations in apprehended drivers: lack of association with clinical evidence of impairment. Addiction. 2007;102:1085–1091. doi:10.1111/j.1360-0443.2007.01802.x. Original commentary on the supplied historical human observational study; no paper text, figures or source tables reproduced.
Leave a Reply