Does D-Amphetamine Increase Motivation Without Changing EEG Alpha Power? A Human–Mouse Experiment

Short answer: In a double-blind, within-person experiment involving 23 healthy adults, oral d-amphetamine at 10 mg and 20 mg increased how far participants continued in a progressively demanding joystick task, relative to placebo. The investigators did not detect a drug-related change in their proposed parietal alpha-power EEG marker (F(2,59) = 0.3, p = .76). In mice, the picture depended on dose and equipment: the highest tested dose reduced task completion in a 24-mouse EEG-tethered cohort, whereas 0.3 mg/kg increased completion in a separate 40-mouse untethered cohort. These are measures of experimental willingness to persist, not tests of a person’s blood amphetamine concentration, fitness to drive, or motivation in a particular event.

The original empirical study is Noback and colleagues, Amphetamine increases motivation of humans and mice as measured by breakpoint, but does not affect an Electroencephalographic biomarker (2024), DOI: 10.3758/s13415-023-01150-z. The essential forensic distinction is between an observed change in a defined laboratory behavior, the absence of a statistically established drug effect on a particular EEG index, and questions that the investigators never tested.

What did the investigators actually test?

They asked two related questions. First, would amphetamine change the point at which a subject elected to stop working for a reward that became progressively harder to obtain? Second, would the drug change a candidate brain-signal marker that tends to rise as subjects approach that stopping point? To investigate, the team used variants of a progressive-ratio breakpoint task in healthy humans and laboratory mice, with EEG recorded in the human experiment and one mouse cohort. A second mouse cohort performed without an EEG tether.

This was not a forensic validation study. It did not take timed post-dose venous whole-blood samples to establish analytical amphetamine concentrations, perform an on-road driving assessment, or determine whether any participant was legally impaired. It tested persistence in a particular task and a cortical electrical signal. Keeping the measured construct explicit is necessary because words such as “motivation,” “effort,” and “biomarker” can sound more general than the experiments justify.

How many independent humans and animals contributed?

The human study recruited 23 independently sampled healthy volunteers, aged 18 to 35 years, including 12 men and 11 women according to the study’s descriptive table. It used a randomized, double-blind, placebo-controlled, counterbalanced within-person design. Each person could be observed on three separate drug-condition visits. That creates repeated measurements, not 69 different human beings. EEG epochs and rotations of a joystick are additional nested observations within persons.

The mouse experiments involved two separate cohorts: 24 C57BL/6J mice with EEG equipment and 40 C57BL/6J mice tested without that equipment. They included males and females. Repeated drug sessions, touchscreen responses, EEG epochs, and reward deliveries must not be counted as independent animals. The study reports models with varying residual degrees of freedom, which is another reason not to infer that every endpoint contained the same number of complete observations. The second cohort was independent of the first and cannot be treated as a within-animal comparison of tethered and untethered conditions.

What dose and route were used in the human experiment?

The human active agent was d-amphetamine administered orally at 10 mg or 20 mg, compared with placebo on separate randomized visits about a week apart. Participants arrived after an overnight fast, completed screening procedures, ate a standardized breakfast, and underwent cognitive-neuroscience testing. The progressive-ratio task was performed approximately 150 minutes after the pill; preceding assessments began around 120 minutes after dosing. Dose here is a nominal administered quantity in milligrams, not a measured blood concentration.

The tested substance must not be silently renamed methamphetamine, lisdexamfetamine, methylphenidate, or mixed amphetamine salts. Although drugs within this broader pharmacological family may share some central nervous system actions, their disposition, metabolite relationships, doses, and clinical contexts are not interchangeable. The study did not produce a concentration–effect calibration curve applicable to human legal whole blood. Even a known administered dose does not establish what an individual’s venous whole-blood amphetamine concentration would be at a later forensic collection time.

What counted as “motivation” in the human task?

Participants repeatedly rotated an arcade joystick in a specified direction. Each completed stage required more rotations than the previous one: an example progression was 5, 15, 35, 70, then 120 rotations, with the requirement continuing upward. The subject received a computer signal of advancement and 50 task “points” when another level was completed. Importantly, those points had no monetary value. Participants were told they could discontinue and leave when they wished. Breakpoint was the highest level completed before disengagement.

This arrangement operationalizes voluntary persistence under an escalating motor-response requirement. It does not independently measure conscientiousness, willingness to work at a job, reliability as a witness, criminal intent, or aggression. Physical facility with a joystick, fatigue, expected length of testing, reaction to the task rules, and the attractiveness of ending the session can also influence how long someone persists. The experimental definition is defensible because it was specified in advance, but transferring the word “motivation” to unrestricted real-world behavior requires separate validation.

Did 10 mg and 20 mg change human breakpoint?

Yes, as measured by this task. The reported mixed-model drug-condition result was F(2,42) = 7.3, p < .005. The authors reported that both oral doses, 10 mg and 20 mg, yielded greater breakpoint than placebo in their post-hoc comparisons (each p < .05). This supports a drug-associated increase in the number of stages participants completed before opting out under these experimental circumstances.

It does not show a validated change in every kind of motivation. The study did not furnish a human blood-drug assay, exposure-adjusted regression, or clinically validated threshold for interpreting a single test result. Nor does statistical significance by itself show that everyone responded in the same direction or that a particular individual continued longer because of a specific drug concentration. The plotted individual observations demonstrate variability; the inferential statistics summarize group-level within-person results.

Did the human findings differ by sex?

The researchers did not detect a significant sex main effect for human breakpoint (F(1,31) = 0.5, p = .493) or a drug-by-sex interaction (F(1,30) = 0.3, p = .599). These are observed nonsignificant results in a relatively small, healthy sample—not proof that all sexes have exactly identical responses or that a clinically important difference is impossible in a larger or different population.

Sex-specific questions are especially susceptible to overinterpretation when study samples are small and models contain repeated observations. A nonsignificant interaction can reflect a small true effect, insufficient precision, heterogeneity, or no true effect; the study does not establish equivalence. Nor may the measured joystick result be used as evidence about an individual person’s driving skill, response inhibition, risk-taking, or legally relevant mental state. Those are distinct endpoints.

What precisely was the proposed EEG marker?

The human EEG was recorded from a 64-lead scalp system. The analysis focused on parietal alpha-band power in the 8–12 Hz range, comparing activity near the beginning and end of the progressive-ratio task. Human recordings were divided into short epochs and summarized across the first and last 50 seconds. The investigators were interested in whether alpha power rose as participants approached the moment when they stopped working, and whether amphetamine altered that change.

EEG alpha power is an electrical physiological measure, not a chemical assay for amphetamine or dopamine. A band-limited scalp signal reflects combined neuronal activity and analysis choices; the observed difference across task periods is not synonymous with a calibrated “effort concentration.” The study’s proposed marker is a candidate correlate of task progression. It cannot reveal a person’s measured drug concentration, administration time, motive, or behavior in a nonexperimental setting by itself.

Did amphetamine change human alpha power?

The end-versus-beginning EEG comparison showed an overall time effect across conditions, F(1,28.24) = 33.4, p < .001, consistent with higher alpha power later in the task. Yet the key amphetamine-condition test on the end-minus-beginning difference was F(2,59) = 0.3, p = .76, with no significant sex main effect (p = .94) or drug-by-sex interaction (p = .50). The pattern therefore combined a behavioral response to amphetamine with an EEG measure that did not exhibit a statistically detected drug effect.

That contrast is the central scientific lesson: a candidate biomarker can vary with task performance without reliably tracking a pharmacological manipulation that changes the behavior. It would be inaccurate to say “amphetamine had no effect on the brain” merely because this particular frequency-band difference was nonsignificant. Likewise, it would be inaccurate to announce that alpha power validated an individual blood-drug or motivational cutoff. Other EEG signals, tasks, time windows, and pharmacological targets were not equivalent tests of this same marker.

What did the EEG-tethered mice do?

In the first animal cohort, 24 mice were trained to complete progressively increasing touchscreen responses for a strawberry-milk reward and performed with skull-mounted EEG equipment and a tether. Amphetamine was injected intraperitoneally approximately 30 minutes before testing at vehicle and 0.1, 0.3, or 1.0 mg/kg conditions. The protocol used repeated sessions separated by washout periods. This was a rodent food-reward procedure, not the human point-only joystick arrangement.

Breakpoints differed across the drug conditions, with F(3,56) = 49.9, p < .001. Notably, the highest dose, 1.0 mg/kg, reduced breakpoint relative to vehicle in the reported post-hoc comparison. The lower doses did not demonstrate the anticipated increase. Because the 1.0-mg/kg condition yielded too few trials for a reliable early-versus-late alpha estimate, it was excluded from that specific EEG analysis. Consequently, the EEG null result should never be misdescribed as a measured finding for all four mouse doses.

What did the mouse EEG measure—and not show?

The mouse EEG analysis examined the same broad alpha-band idea but used the first and last five task epochs, rather than the first and last 50 seconds selected for the human analysis. In the included vehicle, 0.1-mg/kg, and 0.3-mg/kg conditions, the model did not establish a significant drug effect on the last-minus-first alpha change: F(2,49.3) = 1.4, p = .26. A vehicle-only comparison reported a later-than-earlier increase, Wilcoxon z = 2.0, p = .045.

The broader EEG model also reported no significant time main effect (p = .19) or drug main effect (p = .09) in the tested subset, though a sex main effect appeared (p = .01). These findings are not logically contradictory: a selected within-vehicle nonparametric comparison and a larger model address different contrasts with different uncertainty. Nor does a small p value for a single comparison establish a sufficiently specific, sensitive, and reproducible individual drug-testing biomarker. The study did not provide such diagnostic validation.

What changed when the second mouse cohort was not tethered?

To investigate an unexpected negative result, the investigators trained a distinct cohort of 40 mice and tested them without the EEG tether. They compared vehicle with 0.3 mg/kg intraperitoneal amphetamine, using the same broad progressive-ratio endpoint but without simultaneous EEG recordings. In this cohort, amphetamine increased breakpoint, F(1,30) = 11.4, p = .002. This was more consistent with the human direction of change than the tethered cohort’s lower-dose observations.

The outcome does not prove that the tether alone caused the discrepancy. The second cohort consisted of different animals and lacked concurrent EEG measurement; cohort differences, sampling variability, adaptation to the apparatus, and other features can contribute. The comparison is informative and generates a plausible equipment-interference hypothesis, but it is not a randomized same-animal experiment assigning tether exposure while holding every other factor fixed. Accordingly, claims of exact mechanistic causation from the tether would exceed the experimental design.

Were the untethered mice simply faster to respond or collect rewards?

The untethered experiment included additional timing endpoints. It reported no significant amphetamine effect on mean choice latency (F(1,30) = 0.7, p = .418) or on mean reward latency (F(1,30) = 1.6, p = .211). A higher breakpoint therefore was not accompanied by a statistically detected difference in either of these particular latency measures in the reported models.

Those null findings help delimit what “increased motivation” means. More completed reward requirements cannot automatically be reexpressed as faster movement, faster reward retrieval, generalized arousal, or enhanced cognitive accuracy. Likewise, unchanged mean latency does not prove that every relevant movement or motor variable was equivalent: the investigators acknowledged limitations in measuring motor behavior during the tethered task. A forensic interpretation should describe breakpoint and each latency independently instead of collapsing them into one unqualified behavioral conclusion.

Source-grounded evidence: the human experiment

Source-grounded evidence: the two animal cohorts

What does the apparent human–mouse agreement actually establish?

Both oral d-amphetamine in people and a selected intraperitoneal dose in the untethered mice increased the number of escalating task requirements completed. That agreement provides limited evidence that the progressive-ratio task can detect a stimulant-related change in persistence across two species. The important qualifications are that the tasks were not identical, rewards differed, administration routes and dose units differed, and the EEG-tethered cohort gave a different direction at its highest dose. The data support a bounded translational observation, not a universal drug response.

The authors also found that an alpha-power change associated with progression toward quitting did not

How should a forensic expert distinguish the measured matrices and constructs?

Human specimens: the participants underwent screening, including qualitative urine testing, but this experiment did not report calibrated post-dose amphetamine identification or quantification in legally collected venous whole blood, separated plasma, serum, urine, oral fluid, hair, or breath. A screening test obtained before an experiment is not a later concentration–time series. A dose in milligrams cannot be substituted for a specimen concentration in ng/mL; neither an EEG trace nor joystick response is a chemical confirmation.

Animal specimens: mice received injected drug, and local or scalp electrical activity was recorded with electrodes. These are administered-dose and physiological recordings; the source did not establish human post-dose blood levels or a validated trans-species drug concentration threshold. Animal food reinforcement and human noncash points are different experimental rewards. Clinical and case facts: no actual traffic scene, crash, eyewitness memory, psychiatric outcome, workplace conduct, or named person’s motive was tested. The difference between group-level task behavior and an individual case conclusion must remain explicit.

What limitations matter when using this paper in court or professional education?

First, the human biological sample was modest and comprised screened healthy adults rather than patients with substance use disorder, sleep loss, traumatic injury, or complex medication regimens. Second, a repeated-condition study estimates population-average experimental effects; it does not supply a calibrated individual decision rule. Third, the animal cohorts differed both in sampling and equipment, so the proposed tether explanation remains tentative. Fourth, the high mouse dose produced too few observations for part of the EEG analysis, so that signal was not established for the entire dose range.

Fifth, EEG alpha-band results depend on electrode placement, epoch selection, reference procedures, signal processing, and the definition of early and late trial windows; they do not automatically represent drug-specific biomarkers. Sixth, mouse operant persistence, the human points task, and measured response latencies are related but distinct operational outcomes. Seventh, no analytical chemistry result connected dose to forensic whole blood, and the investigators did not establish sensitivity, specificity, decision thresholds, or person-level impairment outcomes. These constraints do not diminish the research; they identify what evidence would be needed for different claims.

Frequently asked questions

1. Does this paper prove amphetamine increases motivation in every person?

No. It reports a statistically significant increase in one escalating joystick task under controlled conditions among 23 healthy volunteers. “Motivation” was operationalized as the last task level completed, and individuals can vary. The study cannot establish that a particular person was more motivated in a job, incident, or legal dispute.

2. Was the drug methamphetamine?

No. Humans received oral d-amphetamine at 10 or 20 mg on separate visits; mice received injected amphetamine expressed in mg/kg. Methamphetamine, lisdexamfetamine, and methylphenidate are different drugs or formulations. The paper must not be represented as a direct trial of those other substances.

3. Did alpha EEG power increase during the task?

In humans, alpha power was higher near the end of the task across conditions, with an overall time effect p < .001. In vehicle-treated mice, a selected comparison also showed an increase. However, the specific drug effect on the human alpha difference was nonsignificant (p = .76), and the analyzable tethered-mouse drug effect was nonsignificant (p = .26).

4. Why did 1 mg/kg amphetamine lower the mouse breakpoint?

That decrease was observed in mice working while attached to EEG apparatus. The experiment did not isolate one definitive mechanism: high-dose behavioral changes, apparatus burden, motor effects, prior training, and other cohort-related factors are possible contributors. The largest dose also left too few task trials for the planned alpha comparison.

5. Why did 0.3 mg/kg help the mice without EEG tethers?

In the independent untethered cohort, that dose increased completed response requirements (p = .002). The observation is consistent with the idea that equipment or experimental context can influence task results. But distinct animals were used, so the data do not conclusively prove that simply removing a tether causes the change.

6. Does the experiment provide a whole-blood amphetamine impairment cutoff?

No. Neither legal venous whole-blood concentrations nor a validated driving or impairment endpoint was measured. Administered 10- or 20-mg oral doses are not interchangeable with ng/mL results in forensic blood, and parietal alpha power is not an amphetamine-specific analytical assay.

7. Can unchanged EEG mean amphetamine had no behavioral effect?

No. Behavioral breakpoint increased significantly in the human task even though the drug comparison for this one EEG-derived alpha metric was nonsignificant. It is equally wrong to claim that a nonsignificant alpha comparison proves two conditions are identical. Both outcomes must be described as measured, with their study-level uncertainty.

Practical interpretation for scientific readers

The defensible summary is narrow: under the tested conditions, oral d-amphetamine increased human progressive-ratio persistence without a statistically established effect on the selected parietal alpha EEG difference; mouse breakpoint changed with dose and cohort setup, while the measured drug–EEG comparison was also nonsignificant. This evidence is useful for thinking critically about the specificity and validation of proposed physiological biomarkers. It cannot replace independent documentation of chemical identity, matrix, specimen collection time, analytical method, or real-world behavior in a forensic case.

For related interpretive principles, consult this site’s published guides to biological specimen matrices, confirmatory drug analysis, and therapeutic concentration monitoring. These topics explain why analytical measurement, experimental physiology, and clinical or legal inference require separate evidence.

Original primary research: Noback M, Bhakta SG, Talledo JA, Kotz JE, Benster L, Roberts BZ, Nungaray JA, Light GA, Swerdlow NR, Brigman JL, Cavanagh JF, Young JW. Amphetamine increases motivation of humans and mice as measured by breakpoint, but does not affect an Electroencephalographic biomarker. Cognitive, Affective, & Behavioral Neuroscience. 2024;24(2):269–278. https://doi.org/10.3758/s13415-023-01150-z.