Category: Uncategorized

A questionnaire score described as “denial” in methamphetamine research is not a direct measurement of dishonesty. A 2015 study examined recognition of drug-related problems and willingness to consider change. Its associations with cognition and brain connectivity do not establish whether an individual statement is truthful.

How was denial operationalized?

Dean and colleagues studied 58 non-treatment-seeking adults with methamphetamine dependence. The investigators used the eight-item Precontemplation subscale of the University of Rhode Island Change Assessment, instructing participants to answer about their methamphetamine use.

Higher scores reflected stronger disagreement that a problem required change. The measure therefore concerned a particular self-reported attitude toward drug problems. Deliberate deception, poor insight, disagreement with a diagnosis and low readiness for treatment cannot simply be assumed equivalent.

Testing occurred during brief abstinence

Participants tested positive for methamphetamine at study entry but underwent cognitive or imaging assessment following 4–12 days of abstinence, with negative tests for specified drugs at assessment. Forty-eight completed procedures as inpatients and ten as outpatients.

This was an observational study, without a controlled methamphetamine administration. It did not isolate the acute effect of a dose or relate a measured blood concentration to credibility.

The cognitive findings were selective

Higher Precontemplation scores were associated with lower overall cognitive performance after adjustment for age, sex, education and inpatient status, with p = 0.001. Follow-up analyses identified associations involving learning/memory and executive functioning.

The processing-speed/attention comparison was a nonsignificant trend, and working memory was not significantly associated with the score. Only 37 participants had the working-memory measures. A claim that denial was linked to deficits in every cognitive domain would overstate the results.

The imaging denominator was 21, not 58

A subset of 21 participants underwent resting-state functional MRI. Higher Precontemplation scores were associated with weaker connectivity involving selected rostral anterior cingulate and precuneus connections. Anterior insula connectivity was not significantly associated with the score.

Some associations no longer met significance after controlling for years of regular methamphetamine use. This sensitivity to covariate selection matters. The findings did not identify one stable “denial circuit” independent of exposure history.

The imaging subgroup also received a shorter cognitive battery because procedures changed. Although overlapping composite scores were strongly correlated in those with fuller testing, the study did not administer an identical complete battery to every participant.

What remains uncertain?

Cross-sectional correlations cannot show that methamphetamine caused reduced insight, that connectivity caused questionnaire responses, or that poor cognition caused denial. Pre-existing traits, exposure history and other influences may contribute.

The selected non-treatment-seeking sample and modest imaging subgroup limit generalization. A nonsignificant imaging comparison may reflect limited sensitivity rather than proof that the region is irrelevant. None of these observations validates an individual diagnostic cutoff.

Appropriate use of the evidence

The paper supports further investigation of cognitive and neural correlates of problem recognition. It does not permit a reviewer to infer dishonesty, intent, competence or intoxication from a drug history or questionnaire score. The exact construct measured must remain visible when scientific terminology is applied to an individual.

Primary scientific source

Dean AC, Kohno M, Morales AM, Ghahremani DG, London ED. Denial in methamphetamine users: Associations with cognition and functional connectivity in brain. Drug and Alcohol Dependence. 2015;151:84–91. doi:10.1016/j.drugalcdep.2015.03.004. Historical human observational study; original commentary on the supplied manuscript.

Years of schooling did not function as a straightforward proxy for cognitive ability in a 2012 methamphetamine study. The researchers found that educational attainment and measured cognition related differently in the dependence and comparison groups. That observation matters when education is used to “correct” a cognitive comparison.

What population generated the finding?

Dean and colleagues studied 36 non-treatment-seeking participants with methamphetamine dependence and 42 healthy comparison participants in Los Angeles. The methamphetamine group reported 4–15 days of abstinence at testing and had negative testing for the examined drugs on assessment days, apart from nicotine. No experimental methamphetamine dose was administered.

The investigators measured several cognitive domains and constructed a composite score. They also assessed vocabulary knowledge as a proxy for academic learning. Cognitive performance, vocabulary and years enrolled in school were different variables, rather than interchangeable measures of intelligence.

The education model produced a discrepancy

Using a model developed in the comparison group, researchers predicted educational attainment from cognitive performance and demographic characteristics. The dependence group had, on average, approximately 1.05 fewer years of schooling than that model predicted; the paired comparison had p = 0.003.

This is a discrepancy relative to a study-specific model. It does not mean every person with methamphetamine dependence lost exactly one year of education, nor does it identify what any individual’s education would have been without exposure.

Earlier reported onset of methamphetamine use was associated with fewer years of education and a larger discrepancy between predicted and actual attainment. The cross-sectional design cannot establish whether methamphetamine caused that difference. Family circumstances, socioeconomic conditions and other developmental factors remain relevant alternatives.

Vocabulary raised a different measurement issue

The investigators estimated schooling obtained before the reported onset of methamphetamine use. Thirteen of the 36 participants had begun using before completing their education. The estimated pre-use schooling variable predicted vocabulary better than total schooling in their regression analysis.

That finding should not turn vocabulary into a perfect measure of educational quality. Vocabulary also reflects language development, social context and other abilities. The authors acknowledged that pre-existing verbal differences could help explain the pattern.

What about parental education?

Maternal education was explored as another correlate. In the dependence group, its correlation with cognitive performance was numerically stronger than the correlation involving participants’ own schooling. However, the study was underpowered to formally compare those correlations.

A significant correlation for one variable and a nonsignificant correlation for another do not, by themselves, establish a significant difference between the correlations. The exploratory result does not validate a general rule replacing a person’s cognitive assessment with a parent’s schooling.

Why this matters for interpretation

Before adjusting for education, identify its intended role. Is it meant to estimate prior cognitive ability, educational opportunity or acquired academic knowledge? Those are different constructs. A variable influenced by exposure and social circumstances may not behave as a neutral baseline correction.

The sample was small and selected, socioeconomic status was not comprehensively measured, and some participants overlapped earlier research cohorts. This paper should not be counted as wholly independent replication of every study it cites.

The defensible conclusion is methodological: in this historical sample, educational attainment was an imperfect proxy for the cognitive constructs being considered. It does not establish an individual’s premorbid ability, brain injury, current intoxication or impairment during an earlier event.

Primary scientific source

Dean AC, Hellemann G, Sugar CA, London ED. Educational Attainment is not a Good Proxy for Cognitive Function in Methamphetamine Dependence. Drug and Alcohol Dependence. 2012;123:249–254. doi:10.1016/j.drugalcdep.2011.11.019. Historical human observational study; original commentary on the supplied paper.

A controlled 2021 study found that oral methamphetamine changed some connections between resting brain networks without producing a statistically significant change within networks under the study’s corrected analysis. “Altered brain connectivity” therefore needs a more precise description than a claim that the drug increased or decreased connectivity everywhere.

What did the experiment involve?

Malina and colleagues studied 22 healthy adults, including 12 women, aged 18–35. Each participant received 20 mg oral methamphetamine and placebo on separate sessions at least four days apart, in counterbalanced order under double-blind conditions.

The administered drug was prepared in an oral syrup. Participants were escorted to a 45-minute imaging session after a post-dose assessment; resting-state scans were collected as part of the larger imaging protocol. These were measurements after a defined acute dose in selected healthy volunteers, not scans diagnosing chronic methamphetamine injury.

What does functional connectivity measure?

The analysis examined relationships among time-varying functional MRI signals. Functional connectivity in this setting describes statistical coordination of signals. It is not a direct count of anatomical connections, a measure of nerve-fiber loss or a blood methamphetamine concentration.

The researchers used independent-component analyses of placebo scans to identify networks, then compared drug and placebo conditions. They distinguished connectivity among regions within a network from connectivity between separate networks. The final analysis included 23 resting-state networks.

The corrected within-network result was nonsignificant

Methamphetamine did not significantly alter within-network connectivity under the Bonferroni-corrected analysis. The paper also examined results at more liberal thresholds. Those exploratory or nominal findings should not replace the main corrected null result.

A nonsignificant result does not prove that within-network effects are impossible. The authors noted that a broad analysis with stricter significance thresholds may miss weaker effects. It does mean this experiment did not establish the broad within-network change under its primary corrected criteria.

Between-network changes went in both directions

Pattern	Connections reported
Increased connectivity	Five pairs involving thalamus or cerebellum and selected sensorimotor, postcentral or middle-temporal networks
Decreased connectivity	Two pairs connecting the lateral sensorimotor network with middle-temporal or posterior middle-temporal networks

The seven affected pairs were not seven new groups of participants. They were network-pair findings from the same 22-person experiment. Increased coordination with a subcortical network could coexist with decreased coordination between cortical networks.

Neither direction is inherently a clinical improvement or injury. Assigning a benefit or deficit requires evidence connecting the specific change to a meaningful functional outcome.

Subjective effects did not provide that bridge

Methamphetamine increased stimulant-like subjective ratings, heart rate and mean arterial pressure. The study did not find significant correlations between subjective responses and the examined connectivity changes.

That observation does not establish a universal separation between experience and brain function. It means these measured changes did not show a significant relationship in this sample. A participant’s subjective response cannot be inferred from an individual network-pair result using this study.

What limits its forensic use?

The sample was small and homogeneous. Resting-state measurements do not necessarily describe connectivity while a person drives, reasons or responds to an emergency. Residual movement remained a possible influence, and the study did not include a perfusion measurement to assess global blood-flow effects.

Network definition and analytic strategy also matter. Findings from this data-driven approach need not duplicate results from studies selecting particular seed regions in advance. Different stimulant drugs, doses and administration routes cannot be silently substituted for one another.

The experiment establishes specific group-level changes under controlled conditions. It does not supply an impairment cutoff, determine the cause of an individual’s behavior, or turn a scan into a measure of permanent damage.

Primary scientific source

Malina M, Keedy S, Weafer J, Van Hedger K, de Wit H. Effects of Methamphetamine on Within- and Between-Network Connectivity in Healthy Adults. Cerebral Cortex Communications. 2021;2:tgab063. doi:10.1093/texcom/tgab063. Original commentary on the supplied controlled human study; citation follows the paper’s author list.

A 2014 study found that an imaging response during a risk-taking task was associated with later self-reported drug use among treatment participants with methamphetamine dependence. It did not establish a clinically validated scan that can determine whether a particular person will relapse.

Start with the follow-up denominator

Gowin and colleagues scanned 68 people enrolled in treatment during early abstinence. Sixty-three were reached at approximately one year; five could not be tracked. Among the 63 followed participants, 18 reported relapse and 45 reported remaining abstinent from the drugs covered by the follow-up definition.

Thus, the reported 29% relapse fraction uses 18 of 63, not 18 of 68. The five missing follow-ups cannot be assumed abstinent. This was a selected treatment cohort, not an estimate of relapse among everyone using methamphetamine.

“Relapse” was broader than methamphetamine use alone

The follow-up asked about sedatives, hallucinogens, stimulants, marijuana and opiates; any reported use in those categories counted as relapse. The outcome should not be relabeled as exclusively confirmed methamphetamine recurrence.

Follow-up information came from interviews and was not confirmed with urine analyses. Initial screening procedures and a later interview outcome are different sources of evidence. A negative urine screen around scanning would not validate a year of subsequent abstinence.

The task contrasted safe and risky choices

Participants selected among 20-, 40- and 80-point options. The 20-point option was safe, while the larger options could produce gains or losses. Unbeknownst to participants, choosing the same option throughout would yield the same final point total.

The later-relapse and abstinent groups did not significantly differ in the frequency of risky choices, either overall or after losses. Their imaging responses nevertheless differed: the abstinent group showed a greater separation in insular activation between risky and safe decisions.

A null behavioral comparison does not demonstrate identical real-world decision-making. Equally, a regional imaging difference does not prove that one group visibly took more risks on the measured task.

Association and classification were promising but imperfect

The reported right-insula analysis had an area under the receiver-operating-characteristic curve of 0.76 in this cohort. AUC summarizes discrimination across thresholds; it is not a statement that 76% of every future population will be classified correctly.

Using a median split, the paper reported correctly identifying 14 of 18 participants who relapsed and 27 of 45 who remained abstinent. Those counts also show the errors: four later-relapse participants and 18 abstinent participants fell on the other side of the classification. The counts are clearer than treating the scan as a definitive prediction.

A Cox coefficient is not a person’s absolute probability

The time-to-event model reported an exponentiated coefficient of approximately 0.34 per standard-deviation increase in the risky-minus-safe insula signal. Although the paper used odds-ratio terminology in places, the analysis was a Cox proportional-hazards model. Its relative time-to-event association should not be treated as a 34% absolute probability of relapse.

The imaging feature and prediction analyses came from the same cohort. The paper did not establish external performance in an independent clinical population. Selection of imaging clusters, limited event counts and model complexity therefore matter when assessing possible optimism in the reported discrimination.

What the study does not justify

Co-occurring substance histories, a predominantly male sample, missing follow-ups and self-reported outcomes limit generalization. The findings do not show that an insular difference caused relapse, that treatment based on the scan prevents relapse, or that the result identifies intoxication or driving impairment.

The study offers a historical research signal worthy of further testing. It supplies no basis for treating an individual scan as proof of future conduct or substituting it for a complete clinical and behavioral assessment.

Primary scientific source

Gowin JL, Harlé KM, Stewart JL, Wittmann M, Tapert SF, Paulus MP. Attenuated Insular Processing During Risk Predicts Relapse in Early Abstinent Methamphetamine-Dependent Individuals. Neuropsychopharmacology. 2014;39:1379–1387. doi:10.1038/npp.2013.333. Original commentary on the supplied historical prospective follow-up study.

A lower imaging response to a pleasant stimulus does not automatically mean a person experienced less pleasure. In a 2013 study of recently abstinent people with methamphetamine dependence, researchers found differences in brain activation during gentle touch while reported pleasantness and simple task accuracy did not significantly differ from controls.

Who was studied?

May and colleagues compared 25 treatment participants meeting then-used DSM-IV criteria for methamphetamine dependence with 17 healthy comparison participants. The dependence group included 19 men and six women; the comparison group included ten men and seven women.

The treatment participants were required to be 15–120 days sober, with a reported mean of approximately 41 days. No participants reported withdrawal symptoms during study participation. This was a historical observational comparison during abstinence, not an experiment administering a measured methamphetamine dose.

The stimulus and the task were different components

Researchers used a soft brush to stroke marked areas on the forearm or palm at about two centimeters per second. Visual cues indicated when touch was approaching. During imaging, participants also responded to arrows pointing left or right in a simple continuous-performance task.

The design separated anticipation from receipt of touch and compared the two skin locations. Functional MRI measured blood-oxygenation-level-dependent responses. Participants rated the experience using visual analog scales, including pleasantness and intensity.

The hypothesized difference in pleasantness was not found

Measured endpoint	Reported finding
Subjective pleasantness	No significant group difference
Subjective intensity	No significant group difference
Arrow-task accuracy	No significant group difference
Regional imaging responses	Group differences in selected regions and conditions

The investigators had expected the dependence group to describe the touch as less pleasant. That prediction was not supported. A summary saying participants “could not enjoy normal touch” would replace a null subjective finding with an unsupported behavioral claim.

The simple arrow task also did not establish an overall accuracy deficit. Imaging differences and measured task performance should be reported separately.

The imaging findings depended on the phase

Across the examined conditions, the dependence group showed lower activation in regions including anterior insula, dorsal striatum and thalamus. The striatal comparison also changed with task phase: activation was higher in the dependence group during anticipation but lower during touch itself.

That phase-specific pattern does not fit a claim that every reward-related region was simply less active at all times. The three-way interaction involving group, phase and skin location was not significant. The study also found no significant correlations between days of abstinence and the examined task, rating or regional activation outcomes.

Those null results do not establish that abstinence has no effect. The study observed people at one assessment, rather than repeatedly measuring recovery within each person.

Alternative explanations remain important

All treatment participants had a history of regular nicotine use, compared with 29% of controls. The groups also differed in education and histories of other substance use. These differences limit attribution of the imaging pattern specifically to methamphetamine.

The authors noted the limited repetitions of touch, the simplicity of the performance task, the predominantly male sample and the timing of subjective ratings after scanning. Such features affect what the design could detect.

What can this evidence establish in a case?

The study supports an association between treatment-group membership and selected brain responses during this experimental task. It did not validate a diagnostic scan for an individual, determine permanent brain injury, measure acute intoxication or establish driving impairment. No administered dose or blood-concentration rule can be recovered from this observational design.

Its useful lesson is the need to keep neural responses, subjective experience and objective performance distinct. A difference on one level should not be promoted into an unmeasured deficit on another.

Primary scientific source

May AC, Stewart JL, Migliorini R, Tapert SF, Paulus MP. Methamphetamine Dependent Individuals show Attenuated Brain Response to Pleasant Interoceptive Stimuli. Drug and Alcohol Dependence. 2013;131(3):238–246. doi:10.1016/j.drugalcdep.2013.05.029. Historical human observational imaging study; original commentary on the supplied manuscript.

Blood collected while a patient is alive can provide valuable evidence in a later death investigation, but a hospital serum ethanol result and a forensic whole-blood result are not automatically interchangeable. Saitman and colleagues found strong correlation alongside significant bias in 39 paired case comparisons. That combination makes the specimen and method essential parts of the interpretation.

These were antemortem specimens analyzed in death casework

The investigators reviewed a reported 15-month period of San Diego County Medical Examiner toxicology work. Of 2,321 cases, 280 involved hospital antemortem specimens, 59 screened positive for ethanol, and 39 had sufficient suitable original material for the quantitative comparison. The specimens were collected at hospital admission before death, rather than paired autopsy blood drawn later.

This distinction limits what the study can show about postmortem ethanol formation. It demonstrates the utility and complications of recovering predeath specimens; it did not experimentally measure microbial ethanol production in decomposing bodies.

Hospital results were from serum, whereas the forensic laboratory generally analyzed whole blood and used serum in four cases. Most received whole-blood specimens were in EDTA tubes, although sodium citrate and sodium fluoride tubes also occurred. Specimens were stored at 4°C after arrival at the medical examiner’s office. The maximum interval between hospital and forensic testing was ten days. A uniform venipuncture site was not specified.

Different methods can produce correlated but biased results

The hospital method was enzymatic alcohol-dehydrogenase testing, measuring a spectrophotometric response associated with NADH production. Forensic testing used headspace gas chromatography with flame-ionization detection. The forensic screen used a single calibrator and one column; subsequent quantitation used three calibrators, a different column, duplicate testing and averaged results. These were distinct analytical stages.

The overall R² between hospital and forensic results was 0.942. Yet forensic results showed an average negative bias of 14.1% relative to hospital results, and the paired difference was statistically significant (P < 0.0001). Larger percentage differences occurred below 0.10 g/dL. The authors attributed the findings potentially to matrix, method and calibration differences; the design did not isolate one cause for the entire bias.

Accordingly, 14.1% is a study finding, not a universal correction factor. It would be inappropriate to subtract that percentage from every hospital result without specimen- and method-specific support.

Sample identity can matter more than the regression

One discrepant case had a hospital value of approximately 0.34 g/dL and a forensic value of approximately 0.20 g/dL. Excluding that case increased R² to 0.981. The authors suspected different samples may have been analyzed, but did not prove a switch. This uncertainty should remain visible rather than being treated as a confirmed laboratory error or silently discarded.

Hospitals differed in collection practice, tube selection and analytical procedure. The article also describes contamination examples outside the 39-case ethanol cohort, including isopropanol and material associated with separator tubes. Those examples are reasons to investigate a questionable result, not estimates of contamination prevalence in the study.

Performance characteristics do not resolve every interpretive issue

The forensic method reported a detection limit of 0.005 g/dL and quantification limit of 0.02 g/dL. Over nine months, control results averaged 0.077 g/dL for a 0.081 g/dL target and 0.198 g/dL for a 0.202 g/dL target, with CVs of 2.2% and 1.5%, respectively. These describe analytical performance under the reported conditions, not certainty about specimen identity or the concentration at an earlier event.

A useful case review preserves collection time, specimen type, tube and storage history, original laboratory records and exact units. It separates the measured antemortem concentration from any later postmortem change, and separates chemical concentration from impairment or cause of death. Our discussion of different alcohol-measurement compartments provides related context.

Primary source

Saitman A, Estrada J, Fitzgerald RL, McIntyre IM. Comparative analysis of hospital and forensic laboratory ethanol concentrations: A 15 month investigation of antemortem specimens. Journal of Forensic and Legal Medicine. 2015;33:23–27. doi:10.1016/j.jflm.2015.03.012.

Postmortem “serum” should not automatically be treated as the serum or plasma collected from a living patient. Hoffman and colleagues studied the supernatant produced by centrifuging postmortem central blood and found cannabinoid ratios that differed from clinical blood-to-plasma relationships. Their results caution against importing a living-person conversion factor into a postmortem report.

What the quotation marks around “serum” mean

In this 25-decedent observational study, central blood was collected from the cardiac vein into glass tubes, including serum-separator tubes without preservatives. The centrifuged supernatant from the latter was designated central “serum.” The label describes how the specimen was produced; it does not demonstrate that its composition matches antemortem plasma.

The series was selected using cannabinoid screening. The detailed methods specify peripheral-blood ELISA with a THCCOOH cutoff of 30 ng/mL. Specimens were obtained at autopsy, on average 23.4 hours after death, with a range of 9.08–49.3 hours, and stored at 4°C for up to six weeks. Only “serum” specimens with low visual hemolysis scores, 1 or 2 on a five-point scale, were included. Results should not be extended without qualification to heavily hemolyzed or decomposed specimens.

The ratios differed by analyte

The median central blood-to-“serum” ratio was 1.0 for THC, with an interquartile range of 0.8–1.2, and 0.8 for THCCOOH, with an interquartile range of 0.7–0.9. A ratio below one means that the denominator, “serum,” had the higher concentration. Reversing the ratio would reverse the intended interpretation.

THC’s ratio was not significantly different from 1.0 (P = 0.82). THCCOOH’s ratio was significantly below 1.0 (P < 0.001) and below the THC ratio (P = 0.02). Individual variability was greater for THC: the reported coefficients of variation were 38% for THC and 16% for THCCOOH. A median of one is therefore not a guarantee of equality in a particular pair.

Strong correlation is not a validated conversion rule

Central blood and “serum” concentrations were strongly correlated, with R² values of 0.953 for THC and 0.994 for THCCOOH. Correlation describes how values varied together in the sample. It does not eliminate systematic specimen differences or uncertainty in an individual prediction.

The quantitative method used solid-phase extraction and GC–MS selected-ion monitoring after screening. Blood-method detection and quantification limits were 1 ng/mL for both compounds. THC was above the lower quantification limit in 20 cases, whereas THCCOOH was quantified in all 25. Precision ranged from 1.27–8.11% CV and accuracy from 95.6–105%. Those analytical characteristics do not themselves validate conversion between postmortem “serum,” peripheral blood and antemortem plasma.

Storage and container effects complicate equivalence

In the accompanying preserved-versus-unpreserved central-blood comparison, preserved THC averaged 12.5% higher; preserved THCCOOH averaged 18.3% lower. The authors described changes below 20% under the observed refrigerated conditions. This was not proof of stability under every temperature, storage duration or decomposition condition.

The authors also identified possible retention of cannabinoids in separator gel and the effects of hemolysis as limitations. All collection tubes were glass, so the study did not establish equivalent results for every container material. Additional cannabinoids, including 11-hydroxy-THC and glucuronidated metabolites, were not included.

How to evaluate a proposed conversion

First identify the actual material, anatomical source, additives, processing and storage. Then ask whether the proposed conversion was validated for that combination and analyte. This study does not support a universal multiplication factor for reconstructing antemortem THC. It also does not establish impairment or cause of death from a cannabinoid concentration.

The broader specimen principle appears in our discussion of why measurements from different biological compartments are not interchangeable. The cannabinoid ratios here remain specific to the postmortem study.

Primary source

Hoffman MA, Trochta A, Gary RD, Fitzgerald RL, McIntyre IM. An evaluation of postmortem concentrations of Δ9-tetrahydrocannabinol (THC) and 11-nor-9-carboxy-Δ9-tetrahydrocannabinol (THCCOOH). Forensic Science International. 2020;315:110414. doi:10.1016/j.forsciint.2020.110414.

A central-to-peripheral blood ratio describes a difference between two postmortem sampling sites. It does not directly measure the change from life to death. In a 25-decedent study, Hoffman and colleagues found modest median ratios for THC and THCCOOH, but individual variation and the absence of paired antemortem measurements prevent a universal correction back to a living-person concentration.

Which cases and collection sites were studied?

The methods describe selection by a positive peripheral-blood cannabinoid ELISA screen with a THCCOOH cutoff of 30 ng/mL. Although the paper’s introductory description refers to central-blood screening, the detailed inclusion method specifies peripheral blood. This was a selected cannabinoid-positive series, not a random sample of deaths.

Peripheral blood came from the iliac vein; central blood was collected from the cardiac vein. Glass tubes with sodium fluoride and potassium oxalate and tubes without additives were collected. Samples were stored at 4°C for up to six weeks. The mean death-to-collection interval was 23.4 hours, ranging from 9.08 to 49.3 hours. For unwitnessed deaths, the time found and pronounced could substitute for the actual death time, limiting precision of the postmortem interval.

Screening and quantitative measurement were separate

The ELISA result selected cases. Subsequent solid-phase extraction and GC–MS in selected-ion monitoring measured THC and its inactive metabolite THCCOOH separately. The validated blood method reported detection and quantification limits of 1 ng/mL for both analytes, precision coefficients of variation of 1.27–8.11%, accuracy of 95.6–105%, and average recoveries of 90% for THC and 93% for THCCOOH.

THC exceeded the lower quantification limit in 20 of the 25 cases; THCCOOH was quantified in all 25. Those different denominators matter when comparing the two analytes.

Median ratios were modest, but findings were not identical

The median central-to-peripheral ratio was 1.1 for THC, with an interquartile range of 0.9–1.4. For THCCOOH it was 1.3, with an interquartile range of 1.2–1.5. THC ratios were not significantly greater than 1.0 (P = 0.13); THCCOOH ratios were significantly greater than 1.0 (P < 0.001). Thus, a nonsignificant result for THC should not be generalized to the metabolite.

Central concentrations averaged 23% higher for THC and 22% higher for THCCOOH than peripheral concentrations. These are group averages, not corrections applicable to every case. One tabulated THC ratio was 4.05, illustrating how an individual pair can differ from the median.

The interval did not provide a reliable redistribution clock

The relationship between ratio and postmortem interval was weak: R² was 0.02 for THC and 0.21 for THCCOOH. These data do not justify dating exposure or calculating a case-specific redistribution amount from elapsed hours. Nor do they establish behavior beyond the approximately 9–49-hour collection window studied, including advanced decomposition.

The authors interpreted the results as indicating minimal potential for redistribution in this series while recognizing inconsistent findings in the literature and substantial individual variability. Because both specimens were postmortem, a ratio near one cannot establish that neither site changed after death.

From chemical findings to a defensible conclusion

The paper expressly cautions against multiplication factors for estimating antemortem THC concentrations and against determining impairment from blood concentrations alone. THC detection and THCCOOH detection also are different findings; the inactive metabolite should not be treated as a direct measure of current psychoactive effect.

For a death investigation, identify collection sites, timing, preservation, analytical limits and other case evidence before attributing impairment or causation. The distinction between a confirmed chemical result and a behavioral inference is also explained in our laboratory-records discussion, using a different drug-specific example.

Primary source

Hoffman MA, Trochta A, Gary RD, Fitzgerald RL, McIntyre IM. An evaluation of postmortem concentrations of Δ9-tetrahydrocannabinol (THC) and 11-nor-9-carboxy-Δ9-tetrahydrocannabinol (THCCOOH). Forensic Science International. 2020;315:110414. doi:10.1016/j.forsciint.2020.110414.

A vitreous drug finding may document detection without permitting a reliable estimate of peripheral blood concentration. Hubbard and colleagues demonstrated this distinction in 89 autopsy cases involving seven acidic drugs: detection rates and blood–vitreous relationships differed markedly by compound. The study offers evidence for selected alternative-matrix testing, not a general conversion formula.

The study was selected casework, not a prevalence survey

Over two years, the investigators included cases already tested for these drugs in peripheral blood that also had sufficient vitreous. Blood was collected at autopsy from the visually identified common iliac vein, up to 20 mL, into glass containing 100 mg sodium fluoride and 20 mg potassium oxalate. Vitreous from both eyes was stored in glass without preservative. Both were kept at 4°C until analysis. Individual death-to-collection intervals were not reported, and antemortem concentrations were unavailable.

The seven compounds were acetaminophen, ibuprofen, naproxen, salicylic acid, carbamazepine, phenobarbital and phenytoin. The total of 89 decedents should not be mistaken for 89 observations of each drug.

Detection depended on the analyte

Drug	Blood findings	Vitreous findings
Acetaminophen	50	50
Naproxen	17	17
Ibuprofen	9	4
Carbamazepine	8	8
Phenobarbital	4	4
Phenytoin	3	2
Salicylic acid	4	4

A negative vitreous result therefore did not exclude blood detection of ibuprofen or phenytoin. Conversely, matching detection counts for another drug do not prove that the concentrations were interchangeable.

Correlation and equality answer different questions

Vitreous concentrations were generally lower. Differences reached statistical significance for acetaminophen, carbamazepine, naproxen and ibuprofen, but not for salicylic acid, phenobarbital or phenytoin. The latter groups each involved four or fewer blood-positive cases; nonsignificance is not evidence that the matrices are equivalent.

Carbamazepine showed an R² of 0.995, and phenobarbital and phenytoin each exceeded 0.95. Those results suggest a relationship worth investigating, but the tiny phenobarbital and phenytoin subsets limit predictive confidence. Among the nonprescription analgesics, acetaminophen had an R² of 0.718 and naproxen 0.322. A high correlation, moreover, can coexist with systematically lower vitreous concentrations.

One naproxen pair, 8.1 mg/L in blood and 25 mg/L in vitreous, was treated as an outlier and excluded from subsequent analyses. That exclusion is material to interpreting the reported relationships; it should not disappear when the study is summarized.

The analytical limits are part of the finding

Analysis used HPLC with diode-array detection, not mass spectrometry. Identification incorporated relative retention time within ±0.05 and an ultraviolet library match of at least 95%. Calibration used at least three nonzero levels, R² ≥ 0.98, and calibrator/control acceptance within 20% of target.

The methods list lower quantification limits of 2 mg/L for acetaminophen and ibuprofen, 2.5 for carbamazepine, 5 for naproxen, phenobarbital and phenytoin, and 20 for salicylic acid. Yet some tabulated concentrations fall below those listed limits. The authors also acknowledge incomplete vitreous validation, including uninvestigated extraction efficiency and appropriate detection limits. Low tabulated numbers therefore should not be presented as fully validated quantitative performance below the stated limits.

What a case interpretation can support

The study supports drug-specific evaluation of vitreous as an alternative specimen. It does not establish antemortem concentrations, prove absence of redistribution, or identify impairment or cause of death from a vitreous number alone. The laboratory’s actual method and validation must support the proposed inference. Our guide to screening and confirmation records explains why analytical stage and reporting scope matter.

Primary source

Hubbard JA, Navarrete AL, Fitzgerald RL, McIntyre IM. Acidic Drug Concentrations in Postmortem Vitreous Humor and Peripheral Blood. Journal of Analytical Toxicology. Published online 2019;45(1), 2021. doi:10.1093/jat/bkz076.

Vitreous humor can support postmortem drug screening when blood is unavailable, but a negative vitreous result does not reliably exclude drugs detected in blood. In a paired study of 51 autopsies, Metushi, Fitzgerald and McIntyre found fewer drug findings and fewer distinct compounds in vitreous than in peripheral whole blood. Their evidence supports an alternative qualitative specimen, with important limits on quantitative interpretation.

What specimens and method were actually compared?

The investigators collected approximately 20 mL of peripheral blood from the visually identified left common iliac vein during autopsy. Blood tubes contained 100 mg sodium fluoride and 20 mg potassium oxalate. Approximately 5 mL of vitreous was collected from both eyes into glass without preservative. Both matrices were stored at 4°C. The article does not provide a case-by-case death-to-collection interval, so the results cannot define performance across all postmortem intervals.

Samples underwent protein precipitation and solid-phase extraction followed by full-scan electron-impact GC–MS. Identification required relative retention time within ±0.08 of the specified criterion and at least a 70% mass-spectral library match. This was a broad screening procedure with quantitative or semiquantitative capability for selected compounds; it was not a universally validated confirmatory assay for every drug.

Fewer findings are an important specimen-selection result

The blood specimens yielded 209 aggregate findings involving 71 different compounds; vitreous yielded 169 findings involving 60 compounds. These are findings across 51 cases, not 209 or 169 independent people. They also are not universal sensitivity estimates: the study did not establish an independent exposure reference standard for every possible analyte.

Morphine appeared in blood in 10 cases and vitreous in five. Seven-aminoclonazepam appeared in two blood specimens and no vitreous specimens. By contrast, diphenhydramine appeared in 11 specimens of each matrix. Such differences show why an alternative specimen must be evaluated by analyte and method rather than accepted or rejected as a single category.

Concentrations cannot simply be transferred between matrices

Among drugs eligible for quantitative comparison, blood concentrations were significantly higher for trazodone and diazepam (P < 0.05). The reported blood and vitreous values were 1.42 ± 0.57 versus 0.15 ± 0.05 mg/L for trazodone, and 0.37 ± 0.06 versus 0.13 ± 0.01 mg/L for diazepam. Trends toward higher vitreous concentrations of oxycodone, hydrocodone and doxylamine were not statistically significant. They should not be converted into claims that those drugs consistently concentrate in the eye.

What the validation leaves unresolved

Water and phosphate-buffered saline were unsatisfactory substitutes for vitreous calibrators in preliminary recovery work. The investigators therefore used pooled drug-free human vitreous. Even so, extensive vitreous validation was not performed: extraction efficiency for every compound and limits of detection and quantification were not established. The method was not optimized for opiates, and gabapentin quality-control performance was inadequate for optimal quantitation.

These limitations matter most when a negative finding is offered as proof of no exposure, or a screening estimate is presented as a precise concentration. Neither inference follows automatically. Detection establishes an analytical finding in the collected specimen; this study did not validate a concentration threshold for impairment or establish that a detected drug caused death.

Applying the study to a laboratory report

Identify the exact specimen, collection site, storage conditions, analyte and reporting threshold. Determine whether the result is qualitative, semiquantitative or supported by a separate quantitative method. For the broader distinction between analytical stages, see our discussion of screening, confirmation and laboratory records; its methamphetamine example does not supply vitreous validation.

Primary source

Metushi IG, Fitzgerald RL, McIntyre IM. Assessment and Comparison of Vitreous Humor as an Alternative Matrix for Forensic Toxicology Screening by GC–MS. Journal of Analytical Toxicology. 2016;40:243–247. doi:10.1093/jat/bkw009.