Why the Same THC Dose Behaves Differently at 30 and at 70

Plasma cannabinoid measurement answers a narrower question than most people expect. It tells you how much of a compound reached the bloodstream and how fast it was cleared. It does not tell you how a patient feels, and the correlation between the two is weak even in carefully controlled laboratory conditions.

In older adults, the practical concern is rarely that blood levels run wildly high. It is that the same blood level lands in a body with less physiologic reserve, more competing medications, and a lower tolerance for an afternoon of sedation.

Most published cannabinoid pharmacokinetic work was done in young, healthy, often cannabis-experienced volunteers. Applying those curves to a 74-year-old on four other medications is an extrapolation, not a measurement, and it is worth being honest about that when a patient asks what dose is right.

The single controlled pharmacokinetic study of oral THC conducted specifically in healthy adults over 65 enrolled twelve people. That is the evidence base. Everything beyond it is inference from general geriatric pharmacology, which is reasonable to use but should not be presented as established cannabinoid science.

Four compounds account for most of what shows up when plasma is analyzed after cannabis use. Delta-9-tetrahydrocannabinol is the parent psychoactive molecule. 11-hydroxy-THC is an active hepatic metabolite. 11-nor-9-carboxy-THC is the inactive terminal metabolite that persists longest. THCA, the acid form present in unheated plant material, generally reaches plasma in meaningful amounts only when raw or under-decarboxylated preparations are used.

The ratio between these compounds is route-dependent in a way that matters clinically. In a controlled comparison published in Clinical Pharmacology and Therapeutics, the plasma ratio of 11-hydroxy-THC to THC was roughly 1 to 10 or 1 to 20 after intravenous dosing, but approximately 0.5 to 1 up to 1 to 1 after oral dosing. Swallowed cannabis passes through the liver before reaching the circulation, generating proportionally far more of the active metabolite.

That is the pharmacologic reason an edible can feel qualitatively different from an inhaled dose of nominally equal size. The patient is not imagining it, and the difference is not a matter of potency alone.

The most directly relevant trial is a phase 1, randomized, double-blind, double-dummy, placebo-controlled crossover study published in European Neuropsychopharmacology in 2014 by Ahmed and colleagues at Radboud University Medical Center. Twelve healthy older adults, six of them men, with a mean age of 72 plus or minus 5 years, received single oral doses of 3 mg, 5 mg, or 6.5 mg of a tablet formulation of THC, or matching placebo. Data from eleven participants were analyzed.

THC was safe and well tolerated in this group. The most common adverse events were drowsiness, reported in 27 percent of cases, and dry mouth, in 11 percent. Participants reported more adverse events at 6.5 mg than at 3 mg, at 5 mg, or on placebo. Among those whose peak concentration fell inside the two-hour sampling window, maximum plasma THC ranged from 1.42 to 4.57 ng/mL and time to peak ranged from 67 to 92 minutes.

Two findings deserve emphasis. Inter-individual variability in plasma concentration was wide, meaning the same milligram dose produced substantially different exposures across a group of healthy people chosen for their similarity. And the pharmacodynamic effects observed were smaller than effects previously reported in young adults given comparable doses.

Cannabinoids are highly lipophilic. After intravenous THC, the apparent volume of distribution is on the order of 10 L/kg, which reflects extensive movement out of plasma and into lipid-rich tissue. Terminal half-life for THC in that same study ranged from 25 to 36 hours regardless of route or sex. Body composition therefore has a direct bearing on how a dose distributes, and body composition shifts with age.

The persistence of stored cannabinoid is measurable. In 28 chronic frequent cannabis smokers studied under monitored abstinence for up to 33 days, plasma THC remained at or above 2 micrograms per liter in 16 of 28 participants at 48 hours, and THC was still detectable in some participants after 30 days. That is a study of frequent users, not of older adults, but it demonstrates that the lipid compartment releases cannabinoid back into circulation over a long tail.

Genetics contribute at least as much as age to inter-individual spread. In 43 healthy volunteers given oral THC, carriers of two copies of the CYP2C9*3 variant showed median THC exposure roughly threefold higher than people with two copies of the common allele, alongside 70 percent lower concentrations of the carboxy metabolite and a trend toward more sedation. Clinicians are not genotyping cannabis patients, and this is a reminder that the variability seen in the clinic has real pharmacologic origins.

This question comes up constantly, and the honest answer is that controlled studies disagree. A randomized, double-blind, placebo-controlled crossover trial published in Drug and Alcohol Dependence in 2026 by Hall and colleagues compared vaporized weight-adjusted preparations: THC alone at 8 mg per 75 kg, THC plus CBD at 8 mg THC and 24 mg CBD per 75 kg, and placebo. Forty-eight participants were enrolled and 35 had complete pharmacokinetic datasets. The THC plus CBD condition produced significantly higher plasma THC, 11-hydroxy-THC, and carboxy-THC by log-transformed area under the curve and peak concentration.

An earlier randomized double-blind study published in Clinical Chemistry reached the opposite conclusion. Nine cannabis smokers received 5 and 15 mg oral THC, low-dose and high-dose oromucosal THC and CBD spray, and placebo. There were no statistically significant differences in peak concentration, time to peak, or area under the curve between matched oral THC and spray doses, which the authors read as evidence that CBD modulation of THC effects is not a pharmacokinetic phenomenon at those doses.

Different routes, different doses, different populations, opposite conclusions. A patient who reports that a high-CBD product feels stronger or weaker than expected is describing something the literature has not resolved, and telling them otherwise would be overreach.

The 2026 vaporized cannabis trial set out explicitly to test whether cannabinoid pharmacokinetics are moderated by age, and found no strong evidence of age-related differences. That sounds like an answer until the age groups are examined: the comparison was between participants aged 16 to 17 and participants aged 26 to 29. It is a valuable adolescent versus young adult comparison. It says nothing about a 70-year-old.

No published controlled trial has directly compared plasma cannabinoid pharmacokinetics between young adults and adults over 65 under matched dosing conditions. The older-adult data and the young-adult data come from separate studies using different formulations, routes, sampling schedules, and analytical methods. Cross-study comparison of that kind can generate hypotheses. It cannot settle them.

Anyone who tells you confidently that older adults clear THC more slowly, or more quickly, is going beyond the published record. The physiologic rationale for expecting differences is sound. The measurement has not been done.

The practical implications do not depend on resolving the pharmacokinetic question. Start lower than feels necessary. Increase slowly, over days rather than within a session. Hold the route constant while titrating, because switching between inhaled and oral changes the metabolite profile and not merely the dose.

For an older patient specifically, the relevant risks are sedation, orthostatic symptoms, and additive effects with other central nervous system agents. Fall risk, not blood level, is the outcome worth protecting. A dose that produces a modest plasma concentration but a two-hour period of unsteadiness is the wrong dose regardless of what the pharmacokinetics look like.

Medication review comes before dose selection. Cannabinoids are metabolized by cytochrome P450 enzymes that handle a large share of commonly prescribed drugs, and older patients carry more of those prescriptions. That conversation belongs with a clinician who has the full medication list in front of them.

The individual pharmacokinetic studies cited here are methodologically sound: randomized, controlled, blinded where feasible, with validated mass spectrometry assays. Their limitation is not quality but scope. Sample sizes run from 9 to 48 participants, which is typical and appropriate for intensive pharmacokinetic sampling but leaves little room for subgroup analysis.

For the specific question of how age changes plasma cannabinoid handling, the evidence is weak by design rather than by execution. The older-adult data come from one 12-person trial. No study has enrolled young and older adults into the same protocol and compared them directly. That is a gap in the literature, not a controversy within it.

Cross-study comparison is doing more work here than it should. The oral THC data in older adults used a specific tablet formulation with a two-hour sampling window, which was too short to capture peak concentration in some participants. Comparing those numbers to inhaled data from a different laboratory using different analytics is suggestive at best.

The frequently repeated claim that older adults metabolize THC more slowly because of reduced hepatic enzyme activity is a plausible extrapolation from general geriatric pharmacology, not a finding from cannabinoid research. The most direct measurement available found smaller pharmacodynamic effects in older adults than had been reported in younger people, which is not what a simple slower-clearance model would predict.

None of this work shows that a particular dose is safe or effective for an older patient with a particular condition. Pharmacokinetic studies in healthy volunteers measure exposure, not therapeutic benefit, and healthy volunteers over 65 are not the same population as frail patients over 80.

The literature also does not establish a plasma concentration that corresponds to impairment, therapeutic effect, or safety to drive. The correlation between plasma THC and subjective or psychomotor effect is weak enough that attempts to define such a threshold have repeatedly failed.

Cannabis medicine has an evidence distribution problem. The population using medical cannabis skews older every year, while the population studied in controlled pharmacology skews young. That mismatch is not unique to cannabis, but it is unusually wide here because the research was historically designed around abuse liability rather than therapeutic use.

The CYP2C9 finding points toward where this field should go next. A threefold difference in exposure attributable to a single common genetic variant is larger than most of the age effects being speculated about, and it is measurable today. Pharmacogenomic stratification would likely explain more clinical variability than chronological age does.

In clinic, the age conversation almost never turns on metabolism. It turns on what a patient can afford to lose. A 28-year-old who feels foggy for an hour has an inconvenience. An 82-year-old who feels foggy for an hour has a fall risk, and a fall changes the rest of their life.

That is why I start older patients at doses that sound almost pointlessly small to people who have used cannabis recreationally. The goal is to find the bottom of the effective range and stop there, not to find the top of the tolerable range and back off. Those two approaches produce very different outcomes in the same patient.

I also tell older patients plainly that the research on their age group is thin. They tend to appreciate hearing it. It is a more useful basis for shared decision-making than false precision about blood levels.

Plasma cannabinoid levels vary widely between people at the same dose, and route of administration changes the metabolite profile more than the dose does. Direct comparisons between young and older adults have not been performed. For older patients, choose the starting dose based on fall risk and medication interactions rather than on pharmacokinetic reasoning the literature cannot yet support.

Carry forward three things: the metabolite ratio shifts substantially with route, inter-individual variability is large and partly genetic, and the direct age comparison has not been done. Leave behind any confident statement about how quickly older adults clear THC relative to younger ones. That claim circulates widely and does not have a controlled trial behind it.