Can a FAAH Gene Variant Predict How Amphetamine Feels? What 159 Volunteers Actually Showed

Answer: In a blinded, three-session experiment involving 159 analyzed healthy adults, two genetic markers in the fatty acid amide hydrolase (FAAH) gene were associated with differences in self-reported arousal and fatigue after a 10-mg oral dose of d-amphetamine. The associations were not established at the 20-mg dose. Other markers and several related outcomes were nonsignificant. The study did not determine anyone’s amphetamine blood concentration, demonstrate a reliable genetic test for individual intoxication, or measure driving impairment. A subjective stimulant response and a legally relevant toxicology result answer different questions.

The underlying original experiment was reported by Dlugos, Hamidovic, Hodgkinson, Goldman, Palmer, and de Wit in Neuropsychopharmacology (2010), “More Aroused, Less Fatigued: Fatty Acid Amide Hydrolase Gene Polymorphisms Influence Acute Response to Amphetamine.” The complete primary publication is accessible through its original research DOI. The following analysis preserves the experiment’s distinct sample sizes, dosing conditions, questionnaire outcomes, significant associations, and genuine null findings.

What is the narrow scientific question?

Can variation in a gene associated with endocannabinoid metabolism explain part of the difference between how healthy people report feeling after a controlled amphetamine dose? The investigators focused on FAAH, which encodes fatty acid amide hydrolase. The enzyme participates in degradation of certain endogenous lipid messengers. The research hypothesis was that differences in this pathway might change subjective stimulant response.

This was not a study of cannabis administration, cannabis testing, or THC concentration. Nor was it a forensic validation of genetic profiling. The proposed mechanism concerns endogenous signaling and variation in response to d-amphetamine. A mechanistic rationale does not itself confirm that a particular variant caused the observed difference, and the paper did not measure each participant’s brain endocannabinoid concentrations or FAAH enzyme activity during dosing.

How many independent people were studied?

The investigators recruited 162 healthy adults, described as 72 women and 90 men, ages 18–35. Three participants lacked a complete set of Profile of Mood States questionnaires and were excluded from the principal mood analyses. The reported analytic population was therefore 159 independent individuals, approximately 44% female. Each individual attended multiple dosing visits, but those visits were repeated observations within a person rather than new participants. There were not 477 independent research subjects.

The study limited enrollment to volunteers of self-reported Caucasian origin and applied ancestry-informative markers to examine ancestry structure. Participants also had medical, psychiatric, body-mass, medication, and substance-use screening. The narrow population helped the investigators reduce some sources of variation, but it restricts immediate generalization to older adults, people with substance-use disorders, patients using prescribed stimulants chronically, and ancestry groups not represented in this sample.

What was actually administered?

Participants received placebo, 10 mg oral d-amphetamine, and 20 mg oral d-amphetamine in randomized order under double-blind crossover conditions. Sessions were conducted at least 48 hours apart. A separate subset attended a fourth session involving 5 mg, but those lower-dose observations were not included in the reported three-condition analysis. Thus, a statement about an effect at 5 mg would not be a result of this principal data set.

At each qualifying session, a capsule was administered after baseline evaluation. Researchers obtained subjective and physiological measures at approximately 30, 60, 90, 150, and 180 minutes after ingestion. They reported the main subjective drug effects becoming apparent around 60 minutes and often peaking between 90 and 120 minutes. These are measurement times in a controlled experiment, not an independently validated clock for reconstructing an unobserved dose in litigation.

Which test measured arousal and fatigue?

The principal endpoint came from the Profile of Mood States (POMS), a questionnaire capturing the participant’s report of current feeling. The researchers used a composite arousal score calculated as anxiety plus vigor minus fatigue and confusion. More positive arousal can therefore reflect multiple combinations of self-reported states. It is not a direct measurement of electroencephalographic wakefulness, physical performance, reaction-time accuracy, or clinical fitness to drive.

Repeated ratings were summarized with a trapezoidal area under the curve, or AUC, over the observation period. That AUC is an integrated questionnaire-score response, not the pharmacokinetic area under a plasma or whole-blood concentration–time curve. The distinction matters: identical abbreviations can describe quantities that cannot be converted into one another. A mood AUC has no amphetamine ng/mL unit.

What was measured biologically?

A blood specimen was collected for DNA genotyping. Investigators studied four single-nucleotide polymorphisms (SNPs): rs6703669, rs3766246, rs324420, and rs2295633. They also estimated combinations of linked variants called haplotypes. Genotyping was performed without knowledge of the corresponding mood outcomes. The experiment included qualitative breath and urine screening before sessions to check study eligibility and instructed participants to avoid alcohol, nicotine, caffeine, and other drugs beforehand.

The presence of a blood draw should not be misread as blood amphetamine toxicology. The article did not report quantitatively measured post-dose d-amphetamine in legally relevant venous whole blood, physically separated plasma, or serum. It did not report a clinical parent-drug or metabolite LC-MS/MS concentration series, concentration-to-effect calibration, or individual blood-to-saliva conversion. There is consequently no scientifically supported way to insert a forensic blood level into this experiment’s results.

Which genetic associations were statistically significant?

Genotype-by-drug analyses found associations for rs3766246 and rs2295633 on the composite arousal outcome, with reported p-values of .010 and .018, respectively. The same two variants were associated with the fatigue subscale, each at p = .009. In dose-specific follow-up analysis, the stronger differences were observed after 10 mg and not established after placebo or 20 mg.

For rs2295633, the published figure distinguished 67 participants with C/C, 70 with C/T, and 22 with T/T genotypes. The C/C group showed greater questionnaire arousal and less fatigue following 10 mg than the other genotype groups. These group comparisons describe averages and variation across volunteers. They do not establish that every C/C individual becomes more alert, less fatigued, or less impaired.

Why is the 10-mg finding not a universal dose rule?

The report’s dose-specific follow-up tests gave arousal p = .038 for rs3766246 and p = .021 for rs2295633 at 10 mg. The authors did not establish corresponding significant genotype differences at 20 mg. That observation supports a hypothesis that a lower-dose response may expose differences masked at higher doses; it does not prove a biological ceiling, a dose threshold, or a predictable clinical response at every intermediate dose.

The administered doses were fixed amounts rather than body-weight-adjusted exposures. Even two people who swallow the same 10-mg dose may have different absorption, distribution, elimination, medication histories, and measured concentrations. This study cannot quantify those differences because it lacked contemporaneous quantitative drug assays. It should not be used to assign an unknown defendant or patient a concentration from genotype alone.

What were the important nonsignificant findings?

The rs6703669 marker did not show a significant genotype-by-drug association with arousal or fatigue. The functional missense marker rs324420 likewise did not reach significance for the main arousal interaction (p = .208) or the overall fatigue interaction (p = .091). The researchers explored a 10-mg fatigue comparison for rs324420 and reported a significant post-hoc result, but that narrower finding does not erase the nonsignificant overall interaction. The statistical hierarchy should remain visible.

The investigators reported no significant genotype-by-drug effects on the other POMS components of anxiety, vigor, or confusion considered independently in their post-hoc examination. Nor did the tested SNPs or haplotypes show a detected association with heart-rate, systolic-pressure, or diastolic-pressure AUC responses. A lack of statistical significance is not evidence that biological effects are impossible; it means the tested association was not established in this sample by the reported analysis.

What did haplotype analysis add?

Linked variants may be inherited together. The authors reconstructed three common FAAH haplotypes using rs3766246, rs324420, and rs2295633, then related estimated haplotype dosage to fatigue and arousal after the 10-mg condition. The original paper used an adaptive permutation approach to provide corrected empirical p-values for the haplotype associations. Those are not tests of chemically measured drug or endocannabinoid concentrations.

The CCC haplotype was associated with lower fatigue ratings, and TAT with higher fatigue ratings. TCT was not significantly associated with fatigue. Importantly, none of these three haplotypes achieved a statistically significant association with the composite arousal endpoint. Thus, the haplotype findings cannot legitimately be presented as a uniformly positive validation of both outcomes.

Why can correlated markers complicate a causal explanation?

Several SNPs in this region are correlated through linkage disequilibrium. A statistical association at one position does not prove that position is the biochemical driver of a mood difference. The rs324420 variant has a known amino-acid change, whereas the other tested variants may tag nearby biological features. The authors discussed possible gene-expression or enzyme-activity pathways, but did not directly test each causal step in the dosed human volunteers.

A single experimental cohort cannot disentangle every linked sequence variant, downstream expression difference, environmental history, or inherited factor correlated with a questionnaire response. The study did use ancestry markers to check for some population structure, which is a meaningful methodological strength. That check does not create validation in unstudied ancestry groups or prove that an unmeasured mechanism explains the association.

Were participants experimentally intoxicated in a forensic sense?

They received known d-amphetamine doses in a controlled research setting. That supports a causal comparison of administered drug versus placebo for the average subjective drug response. But “intoxication” in a legal or clinical question may require additional evidence: observed behavior, an applicable offense definition, contemporaneous circumstances, actual analyte measurements, timing, and method-specific analytical quality. Those components were not all measured here.

Self-reported fatigue reduction is not the same as restored vigilance in a demanding real-world task. A volunteer may feel energized yet perform poorly in divided attention, hazard detection, or a complex sequence of judgments. Conversely, a genotype category with no significant arousal association cannot be presumed unaffected by amphetamine. This experiment has no validated genotype-to-driving-risk classifier, no simulator outcome, and no real roadway observations.

How should blood, urine, and genomic samples be kept separate?

In a case file, a DNA result and a toxicology concentration are different laboratory findings even if both originate from a tube labeled “blood.” Genomic typing identifies inherited sequence variation. A legal whole-blood amphetamine measurement, if performed, would require a separately established specimen, validated analytical method, calibration, controls, quantitation, and reporting units. Serum or plasma results would involve different specimen preparation and potentially different concentration relationships.

Urine screening may demonstrate a qualifying pattern of exposure or study eligibility but usually cannot by itself reconstruct a contemporaneous brain effect or legal impairment. Here the pre-session urine and breath collections were screening safeguards, not the primary endpoints. For a fuller framework, see the independently published guides to biological matrices and confirmatory drug testing. Neither guide converts this genetic association into a blood drug level.

What do these results imply for pharmacogenetics?

The paper offers a plausible research lead: inherited variation around FAAH may contribute to between-person differences in the experience of stimulant-related fatigue. That is more limited than an assertion that genotyping can select an individualized dose, diagnose stimulant impairment, forecast addiction, or establish the credibility of a witness. The sample was healthy and young, had one acute course per condition, and did not include a prospective clinical decision rule or independent external test cohort.

The authors themselves emphasized replication and the small contribution likely to be made by any one genetic locus. Multiple testing also warrants caution: four markers, multiple POMS subscales, doses, and follow-up contrasts create numerous opportunities for nominally significant results. The authors applied particular statistical procedures, including corrections for some post-hoc and haplotype comparisons, but the manuscript should not be misrepresented as a prospective, independently validated clinical assay.

What does a case-specific expert need beyond this paper?

A scientifically responsible interpretation begins with the exact question. If the issue is whether amphetamine was present, examine the specimen, analytical validation, target analytes and any identification or quantitation records. If the issue is recent exposure, examine collection time, dosing history and matrix limitations. If the question is impairment, consider relevant independently collected behavioral observations and the governing scientific and legal standard rather than substituting genotype or self-report scores.

For a medication-response question, account for drug identity, formulation, regimen, adherence, interacting conditions, and clinical follow-up. A genotype may eventually add context when appropriately validated, but it should not overrule a measured concentration or directly observed task performance. The therapeutic drug monitoring guide explains why a timed, physically analyzed drug result and a subjective response require different interpretive steps.

What findings are strongest and weakest?

The strongest supported narrow statement is an association in this sample between two linked FAAH markers and particular questionnaire outcomes after 10 mg oral d-amphetamine, accompanied by observed average drug effects versus placebo. The reproducible design details—the randomized within-person comparison, specified timepoints, genetic assay, and explicit null tests—help define that statement without stretching it into an individualized rule.

The weakest extrapolations would be to infer how an unrelated person drove, acted, felt, or metabolized amphetamine based solely on their genotype. Neither the study nor any one statistical p-value supplies a sensitivity, specificity, positive predictive value, individual legal cutoff, or validated testimony about a specific event. The distinction between an explanatory research association and an operational forensic method is essential.

Frequently asked questions

1. Does a FAAH genotype prove that someone was impaired by amphetamine?

No. The reported outcomes were questionnaire ratings after known oral doses, not an individually validated impairment criterion. No driving performance or case-specific intoxication classifier was tested.

2. Did the study test methamphetamine or a mixture of stimulant drugs?

No. The active drug was orally administered d-amphetamine at 10 mg and 20 mg. It should not be silently relabeled methamphetamine, mixed amphetamine salts, or lisdexamfetamine.

3. Were 159 people tested once, or were there 477 independent subjects?

There were 159 independently analyzed participants, each observed under three crossover conditions. The multiple visits improved within-person comparison but did not triple the independent sample size.

4. Does AUC here mean blood concentration over time?

No. This AUC integrated repeated Profile of Mood States scores. It is not the pharmacokinetic AUC of measured plasma or whole-blood amphetamine and has no drug-concentration unit.

5. Was the functional rs324420 marker significant for the main outcomes?

The overall rs324420 arousal and fatigue interactions were not statistically significant at the paper’s main threshold. A narrower exploratory 10-mg fatigue comparison was reported significant. Both statements are needed to avoid misleading selectivity.

6. Why were haplotype fatigue associations stronger than arousal associations?

The reconstructed CCC and TAT haplotypes had empirically corrected significant fatigue associations, while none of the haplotype arousal comparisons reached significance. The composite arousal score combines several domains, and the study did not establish a single causal sequence variant explaining the difference.

7. Can these data reveal a person’s exact last dose, legal blood level, or likelihood of a crash?

No. The protocol did not provide quantitative post-dose drug concentrations, an individualized last-use algorithm, real driving outcomes, or crash follow-up. Those questions require different, independently validated evidence.

Primary study and scientific scope

This article is an original interpretation of Dlugos AM, Hamidovic A, Hodgkinson CA, Goldman D, Palmer AA, and de Wit H, “More Aroused, Less Fatigued: Fatty Acid Amide Hydrolase Gene Polymorphisms Influence Acute Response to Amphetamine,” Neuropsychopharmacology 35:613–622 (2010). Review the complete original publication for the full methods, figures, questionnaire scales, tables and statistical footnotes. The paper concerns a selected population under controlled oral dosing; any individual legal or medical inference must be evaluated against that person’s independently documented circumstances.