Does Methamphetamine Dependence Affect Remembering Future Intentions? A 65-Adult Prospective-Memory Study

Short answer: In an observational study of 65 adults, the 39 participants with a history of methamphetamine dependence obtained lower scores than 26 comparison participants on a laboratory test of remembering to carry out intended actions. Their median total Memory for Intentions Screening Test (MIST) score was 39 rather than 42 (p = .001), and their median number of coded errors was four rather than one (p < .001). Task-substitution and mistimed-action errors were more frequent. Yet the groups showed no statistically significant difference on a separate recognition test (p = .740), a simultaneous word-search task (p = .204), or a next-day telephone-intention task (p = .227). The investigators did not measure methamphetamine concentration in forensic human venous whole blood or establish that a particular person’s missed obligation, driving behavior, or memory of a witnessed event was drug-caused.

This distinction arises from an original primary study by Jennifer E. Iudicello, Erica Weber, Igor Grant, Michael Weinborn, Steven Paul Woods and the HIV Neurobehavioral Research Center Group, Misremembering Future Intentions in Methamphetamine Dependent Individuals, published in The Clinical Neuropsychologist, volume 25, issue 2 (2011). The investigation tested delayed intention execution, rather than drug concentrations or the accuracy of eyewitness descriptions. Those are different evidentiary questions.

What does prospective memory actually mean?

Prospective memory is the ability to remember that an intended action must be performed in the future. A person may remember the instruction when it is given, understand what it means, and still fail to carry it out when the correct time or circumstance arrives. Examples include remembering to return a telephone call after a meeting, submitting a report when a timer expires, or handing over a document after encountering a particular person. The relevant outcome is whether an already-formed intention is executed when the cue occurs.

That faculty should not be equated with recalling what happened yesterday. Remembering the details of an incident is ordinarily described as retrospective or episodic memory; remembering that one must call the investigator tomorrow is prospective memory. The two processes can overlap because someone must retain the content of the intended action, but they answer distinct questions. An omission to make a scheduled call cannot, by itself, establish that the caller’s account of an earlier event is false.

What was the narrow research question?

The investigators asked whether adults classified as methamphetamine dependent showed differences in a structured laboratory measure of future-intention performance relative to a comparison group, and whether the pattern was related to other cognitive measures or historical methamphetamine-use characteristics. They examined the distinction between cues defined by elapsed time and cues defined by events, as well as the types of mistakes made when an intention was missed or implemented incorrectly.

That design does not test whether a specified dose produces a specified behavioral effect. No participant was randomly assigned to receive methamphetamine, and no controlled acute drug challenge supplied a concentration–effect curve. “Methamphetamine dependent” was a clinical and historical classification, not an experimentally manipulated exposure or a statement that a measured quantity of methamphetamine remained in blood at the testing session.

Who were the independent biological participants?

The study compared 39 adults with methamphetamine dependence and 26 adults without that history, for 65 independently sampled human participants. The comparison was cross-sectional; the same people were not prospectively followed from before dependence through recovery. Each participant completed several instructions, trials, recognition items, and cognitive tests. Those repeated measurements are observations nested within people, not extra independent participants. The optional 24-hour intention is likewise a delayed outcome for a person already in the study, not a new subject.

Baseline comparability was limited. Approximately 92% of the methamphetamine-history group were men, compared with 42% of the comparison group (p < .001). Histories of cannabis dependence (28% versus 0%, p = .002), cocaine dependence (36% versus 8%, p = .017), and lifetime major depressive disorder (59% versus 31%, p = .026) also differed. Median mood-disturbance scores were 58 versus 36 (p = .034). Mean ages were 41.6 and 40.6 years, with no significant age difference (p = .742). These characteristics matter because the study did not independently manipulate substance history, mental-health history, or sex distribution.

What methamphetamine exposure was documented—and what was not?

Information about earlier methamphetamine use came from retrospective participant reports and clinical classification. In the methamphetamine group, the reported median time since last use was 105 days, with an interquartile range of 21.0 to 187.5 days; reported lifetime-use quantity had a median of approximately 1,058 grams. These are recalled historical street-product estimates and intervals, not witnessed dosing events or analytically verified doses of pure methamphetamine. The reported quantity has no validated conversion to parent-drug concentration at the day of testing.

The research included contemporaneous exclusion screening for alcohol and selected drug use, using breath testing and clinical urine screening. Such screening cannot be silently relabeled as a quantitative confirmation of parent methamphetamine in a legally collected human venous whole-blood specimen. A negative or qualifying clinical screen has different analytes, targets, matrices, thresholds, and evidentiary meaning from a properly validated forensic whole-blood assay. The study does not report a methamphetamine/amphetamine mass-spectrometric blood concentration, a blood specimen collection-to-use timeline, an enantiomeric d/l-methamphetamine analysis, or a whole-blood lower limit of quantification.

Why do blood, plasma, serum, and urine need separate descriptions?

Human venous whole blood contains the cellular and liquid fractions as collected. Plasma is the liquid separated from anticoagulated blood, and serum is obtained after clotting and separation. Values from these matrices are not interchangeable without analyte-specific empirical comparison and a suitable validation basis. Urine is an excretory specimen that may support detection of exposure within a matrix-dependent window but ordinarily does not directly quantify contemporaneous whole-blood drug concentration or behavioral effect.

Oral fluid, hair, sweat, vitreous humor, milk, breath, and tissue are additional distinct matrices. None is converted into a legal human venous whole-blood concentration merely by discussing a substance or a positive screening category. This particular investigation measured neuropsychological task performance, not quantitatively assayed parent methamphetamine in any of those forensic matrices. HIV clinical laboratory variables in participants with HIV do not become measurements of methamphetamine simply because they involve blood or plasma.

How did the thirty-minute intentions test work?

The investigators used the research version of the Memory for Intentions Screening Test, or MIST. The core laboratory exercise lasted about 30 minutes and incorporated four time-based and four event-based intentions. Participants were also engaged in an ongoing word-search activity. Successfully executing an intention therefore required detecting or monitoring the relevant cue while occupied with another task, retrieving what was supposed to be done, and taking the correct action at the appropriate point.

A time-based intention depends on temporal monitoring, such as acting after a specified interval. An event-based intention depends on encountering a defined trigger in the environment. Those distinctions are experimentally useful, but neither class automatically corresponds to a legal duty, workplace requirement, or real-world witness situation. Instructions, cues, distractions, salience, motivation, and consequences outside the laboratory may differ substantially from the MIST.

The investigators also administered a separate recognition exercise after the timed laboratory trials, to examine whether the intended actions were recognizable when prompted. A next-day instruction asked participants to call the examiner roughly 24 hours later and report how many hours they had slept. That delayed call was scored separately and did not contribute to the main MIST summary score.

Original numerical evidence table 1: Prospective-memory performance

These p-values arise from observed group comparisons and do not establish the probability that methamphetamine caused a given error in a particular person. The study’s summary-score standardized effect size was reported as Cohen’s d = 0.87, and the total-error comparison as d = 1.19. Neither effect size is a sensitivity, specificity, blood-drug threshold, or validated classification probability for a forensic examinee.

Why can two groups have the same median and still differ statistically?

For event-based cues, the reported median was seven for both groups, yet the distributions differed (p = .013). A median compresses observations into a single midpoint and may conceal shifts in dispersion or the counts falling below and above that midpoint. For mistimed responses, both group medians were zero, but the methamphetamine-history group had a broader positive upper quartile and a statistically different error distribution. A statement that “both groups had zero errors” would be inaccurate because it would substitute a median for every participant’s raw performance.

Nor should p-values be read as the magnitude of impairment. A low p-value on a nonparametric test indicates evidence against the specified null model under that test’s assumptions. It does not report the percentage of people who were affected or whether a missed call or delayed action outside the study was caused by any substance. The study’s sample size and group composition also limit direct generalization to women, to more recently dosed people, or to forensic populations.

Was the deficit specifically greater for time-based instructions?

Not on the formal differential test. Both groups, overall, performed worse on time-based than on event-based intentions (cue-type main effect, F(1,63) = 24.3, p < .001). There was also an overall methamphetamine-history group effect (F(1,63) = 11.8, p = .001). But the group-by-cue-type interaction was not statistically significant (F(1,63) = 0.97, reported p > .10). Separate significant comparisons in each cue category do not establish that the size of the group association reliably differs between categories.

This is a particularly important distinction for legal or administrative reasoning. One cannot fairly convert these data into the claim that methamphetamine-dependent individuals uniquely fail at time-specific duties but succeed at event-triggered duties. The more circumscribed observation is that performance differed on both classes of laboratory intentions, while a specific difference in their relative group effects was not established.

What are substitution and timing errors?

The study classified errors by what the participant actually did. Task substitution meant performing an action different from the prescribed intention; a person might give an action when a verbal response was required, or vice versa. Loss of time meant acting at the wrong moment despite producing an otherwise correct response. The source also distinguished no response—no action made—and loss of content, when the cue was noticed but the intended instruction was incompletely retained.

Within the methamphetamine group, the investigators classified 68% of substitution errors as intrusions of a novel, incorrect response, and 11% as repetitions of an action already completed. About 71% of mistiming errors came from time-cued intentions, and 87% of those timing errors were made too early. These percentages describe the composition of errors within the relevant subset; they are not 68%, 71%, or 87% of all 65 participants, and they do not imply that an individual making an early response was acutely under the influence.

Which observed outcomes were genuinely nonsignificant?

Several direct measurements did not show a statistically significant group difference. The groups had identical recognition medians of eight, with a reported p = .740. The word-search activity had p = .204. “No response” errors had p = .411 and “loss of content” errors had p = .075. Completion of the separate 24-hour call was observed in 21% of the methamphetamine-history group and 35% of comparisons, p = .227. These are actual tested null results, not outcomes that went unmeasured. They do not establish practical equivalence, because nonsignificance alone is not a prespecified equivalence test.

The recognition comparison is especially informative: when a test presented explicit answer choices, the study did not find evidence of a group difference, even though executing intentions at the proper point differed. That contrast is consistent with the distinction between retaining an instruction and activating it spontaneously at the right moment. However, eight-of-eight recognition medians in both groups also suggest possible ceiling limitations. It would overreach to say the research proves every form of retrospective memory was intact.

Original numerical evidence table 2: Null findings and secondary analyses

For the table’s group comparisons, the first numerical value corresponds to participants with a methamphetamine-dependence history and the second to the comparison group. Medians are used where relevant; percentages are actual group-completion proportions. Spearman coefficients concern associations within only the 39 methamphetamine-history participants, not new or independently recruited subjects.

Did the findings persist after adjustment for measured differences?

The researchers examined a regression including sex, lifetime major depressive disorder, alcohol dependence, and total mood disturbance. Methamphetamine-history status remained associated with overall prospective-memory performance (p = .009), and the reported adjusted model explained approximately 20% of variance (adjusted R² = .20; overall model p = .003). This offers evidence that the observed association was not entirely explained by the specific covariates included in that model.

It does not provide randomized causal identification. Other measured baseline characteristics did not necessarily permit the same multivariable contrast because the comparison group included no participants with certain diagnoses, including HIV infection. In the methamphetamine group alone, tests of sex, HIV, HCV, and other substance-use disorders did not identify statistically significant overall prospective-memory differences (reported p values greater than .10). In a subgroup of only 39, absence of a significant difference is not proof that those conditions cannot affect prospective memory.

How did executive functions relate to intention performance?

Within the methamphetamine-history group, greater prospective-memory scores correlated with a composite of executive-function tests (Spearman rho = .41, p = .03). Other tested associations with a learning composite (rho = .26, p = .16), delayed recall (rho = .19, p = .31), and processing speed (rho = .17, p = .35) were not statistically significant. The investigators’ later regression of individual executive tests identified Trail Making Test Part B as the only significant individual contributor in that model, which had adjusted R² = .29 and model p = .02.

This supports a cautious research inference that performance on intended-action tasks may be associated with flexibility, monitoring, or executive-control processes in the studied group. It is not a validated algorithm for deciding whether a particular person knowingly disregarded a duty, had a neurological disorder, or was medically or legally incompetent. The paper reports relationships among imperfect behavioral measurements, not an anatomical or pharmacological demonstration of the mechanism causing every mistake.

Did reported lifetime dose, years of use, or last-use interval predict the score?

Within the 39 methamphetamine-history participants, earlier reported age of first use correlated with lower MIST summary performance (Spearman rho = .43, p < .01 for age at onset versus score). Other history variables—the number of days reported since last use, overall duration of use, and reported lifetime quantity—were not significantly related to the prospective-memory score (reported p values greater than .10). A reported median abstinence period of 105 days is thus a descriptive group characteristic, not evidence of a validated concentration-elimination curve or of a known drug-free period for every participant.

Age at first use is itself a retrospective report and can correlate with other life-course and clinical factors. The correlation cannot isolate developmental neurotoxicity, assign biological age of injury, or demonstrate how one individual’s future performance would change if an exposure happened earlier or later. Similarly, a nonsignificant association involving historical grams consumed does not show that methamphetamine dose has no biological consequences; those grams were not chemically assayed individual exposures in a randomized or prospective pharmacokinetic design.

Why is the next-day telephone result important?

The separate 24-hour intention asked participants to telephone the examiner and report hours slept. Only a minority in either group completed it according to the study’s rules: 21% of the methamphetamine-history group and 35% of the comparison group, p = .227. The difference was not significant, so the investigators did not establish a group contrast on that specific delayed task. Low overall completion also suggests that this measure operated under different demands and likely had limited ability to discriminate among individuals in this sample.

A missed call has multiple possible explanations: understanding or misunderstanding the instruction, forgetting, changed circumstances, motivation, telephone access, competing obligations, or a deliberate choice. The laboratory record does not identify which explanation caused a particular participant’s outcome. The task is useful as a naturalistic extension of the experiment, not a demonstrated detector of intoxication or of the reasons behind an actual missed appointment or legal deadline.

What were the main sources of uncertainty?

The comparison groups were modest in size, observational, and materially unequal in sex and several lifetime clinical characteristics. The control group came from a separate prospective-memory research context; recruitment and selection mechanisms may therefore differ. Some behavioral contrasts were significant, others were not, and many measures were examined. Reporting a suite of p-values without asserting a single universally valid deficit is essential to preserve the original result structure. Multiple testing, differences in task distributions, and participant selection complicate efforts to extrapolate any one number.

The researchers did attempt selected covariate adjustment and investigated several co-occurring diagnoses, but statistical adjustment cannot recreate unmeasured counterfactuals or eliminate all differences between groups. The absence of quantitative drug assays also makes it impossible to associate the tested behaviors with a measured parent-methamphetamine concentration or a legally relevant sampling interval. No drug dosing, blinded acute placebo-control comparison, serial toxicokinetic specimens, on-road driving, accident investigation, or eyewitness event-recall test was performed.

What can a forensic toxicologist legitimately conclude?

A defensible scientific summary is that, in one small observational human sample, a history of methamphetamine dependence was associated with lower MIST prospective-memory performance and more substituted or mistimed responses. Recognition and several other comparisons were nonsignificant, and the formal cue-type interaction did not demonstrate disproportionate vulnerability of time-based cues. The findings cannot determine the presence, parent/metabolite amount, time since use, or pharmacological effect of methamphetamine in a specific person’s legally collected human venous whole blood.

In forensic casework, an expert must distinguish proof of analytical exposure from proof of contemporaneous functional impairment and, separately, proof that impairment caused the event being litigated. The MIST provides neither a blood toxicology result nor event-specific causal evidence. It also does not test whether a witness accurately remembers a collision, assault, statement, or interaction. Those conclusions require their own evidentiary foundation; historical group contrasts are not substitutes for an individual examination and the actual record of the event.

Frequently asked questions about methamphetamine and remembering intentions

Does this study show that everyone with a methamphetamine-dependence history forgets future tasks?

No. The study reports differences in group distributions, not a rule that applies to every person. Performance ranges overlapped, median values compress individual differences, and the participants were selected for research rather than sampled as a representative legal population. It did not determine whether any particular person’s omission was caused by past use.

Were time-based intentions selectively worse than event-based intentions?

The history group scored less favorably on both categories, but the formal group-by-cue-type interaction was not significant (F(1,63) = 0.97, p > .10). One cannot infer a statistically confirmed selective time-cue deficit simply because separate group comparisons were significant.

Did the study show that participants had lost the content of the instructions?

Not as a significant overall group effect on the separate recognition measure. Recognition medians were eight in both groups (p = .740), and loss-of-content errors did not significantly differ (p = .075). Recognition after prompting and spontaneous performance at a delayed cue are not the same task.

Can a qualitative urine test establish the venous whole-blood methamphetamine level that affected memory?

No. Urine screening, chemical confirmation, and quantitative parent-drug measurement in legally collected venous whole blood are different analytical procedures. This study did not provide a validated whole-blood methamphetamine value, dose–effect curve, or concentration threshold tied to any MIST score.

Did self-reported abstinence duration explain the group deficit?

Reported days since last use were not significantly correlated with the MIST summary score within the methamphetamine-history sample (reported p > .10). A retrospective median last-use interval is not a chemically verified individual abstinence timeline or proof of causal recovery.

What happened on the separate 24-hour phone-call intention?

Completion was 21% in the history group and 35% in comparisons, a nonsignificant contrast (p = .227). The task was separate from the primary summary and had low completion in both groups. It cannot determine why a specific person later missed a real appointment, call, or deadline.

Can these results establish an impaired driver, unreliable eyewitness, or cause of a particular incident?

No. No controlled drug administration, quantitative forensic blood drug assay, on-road driving outcome, eyewitness event-recall validation, or incident-specific causation analysis was carried out. The observed research associations are relevant background science, not individual adjudications.

Bottom line: a specific group association, not a forensic concentration test

The most informative combination of findings is the observed disadvantage in executing future intentions, particularly substitution and timing mistakes, alongside genuine null group differences in recognition, word-search activity, the next-day call, and the cue-type interaction. The study supports treating prospective intention execution as a distinct performance domain rather than equating it with all memory, all judgment, or all real-world behavior. The source is Iudicello and colleagues’ original 2011 prospective-memory investigation. Its 65 participants contributed cognitive observations, not a validated forensic dose, human venous whole-blood methamphetamine concentration, intoxication determination, driving test, or individual event-causation finding.