Edited by: Liana Fattore, National Research Council (CNR), Italy
Reviewed by: Marilyn A. Huestis, Thomas Jefferson University, United States; Dennis Sholler, Johns Hopkins Medicine, United States
This article was submitted to Addictive Disorders, a section of the journal Frontiers in Psychiatry
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
In some states in the USA, the legalization of cannabis for adult and medical use has increased the availability and sales of cannabis products, such as extracts, that have a THC content >70% (
We systematically reviewed evidence on the degree to which people who use cannabis for recreational purposes can and do reduce their THC dose when using more potent products.
We included original studies published from 1973 to the date of our search (17 June 2020) if they reported quantitative data on behavior indicative of titrating THC doses from cannabis products (
We excluded studies of medicinal cannabis use and clinical/pharmacological studies where patients/participants were instructed to titrate their dose of cannabis consumption. Our review aimed to examine evidence on recreational cannabis use in non-supervised settings to better inform public health policy on the regulation of recreational cannabis use.
The search was conducted in PubMed and Embase with terms related to “Cannabis” AND 'Titration' in the title/abstract/related MeSH and Emtree explosion subject headings, with the “Humans” filter applied, as follows:
PubMed search: ((cannabis [tiab] OR marijuana [tiab] OR Cannabis [MeSH] OR Marijuana Use [MeSH] OR Marijuana Smoking* [MeSH])) AND ((titration [tiab] OR self-titration [tiab] OR self-titrating [tiab] OR self-titra* [tiab] OR titrant [tiab] OR titrat* [tiab] OR auto-titration [tiab] OR autotitration [tiab])) AND (humans[Filter]).
Embase search: (((cannabis:ti,ab OR marijuana:ti,ab OR “cannabis”/exp OR “marijuana use”/exp OR “marijuana smoking*”) AND (titration:ti,ab OR “self-titration”:ti,ab OR “self-titrating”:ti,ab OR “self titra*”:ti,ab OR titrant:ti,ab OR titrat*:ti,ab OR “auto titration”:ti,ab OR autotitration:ti,ab)) AND “human”/de) AND (“article”/it OR “article in press”/it OR “review”/it).
The supplementary search involved the authors' collection and a snowball search of secondary references identified from all relevant records from the database search and authors' collection. Two researchers carried out the screening, study selection, and data extraction.
Findings from experimental and observational studies and surveys were synthesized narratively on evidence of: (1) titration behavior (e.g., amount smoked, smoking topography) and (2) evidence of effective titration, defined as adjusting consumption when using high THC products to deliver the same THC dose or to achieve the same physiological, neurobehavioral, or psychological effects obtained from using a lower dose product.
We identified 197 records from the database search and 338 records from the supplementary search, from which, we screened 497 unique titles after exclusion of duplicates. After full-text screening (
Most studies were from the USA (
Experimental laboratory studies (
Summary of experimental (a), observational (b), and survey (c) studies on titration of recreational cannabis products by potency.
Cappell et al. ( |
Experimental, in lab, ( |
Experienced in cannabis use, aged 21–28, males | 0.8 vs. 0.4% or 0.2% THC flower (as cigarettes) | Total time with smoke in lungs, number of puffs, mean duration of puff, mean interval between puffs, estimated weight of material consumed, finger pulse, blood pressure, and conjunctival injection. | Effective titration of intake did not occur. Number of puffs and duration for which puffs were held in lungs did not differ as a function of THC concentration. |
No | No |
Behavioral tasks: pursuit rotor, verbal memory, and raw reaction time. | |||||||
Cappell and Pliner ( |
Experimental, in lab, ( |
Frequent or infrequent cannabis use, aged 18–29 (mean = 22), males | 1.45 vs. 0.73% or 0.36% THC flower (as cigarette) | Cigarette size (small and large), pulse rate, number/duration/intervals between inhalations. | There was some evidence of titration behavior with the amount of cannabis consumed increasing as potency decreased. | Yes | No |
Participants in the more potent conditions, however, self-administered more total THC, attaining the same subjective endpoint of intoxication. | |||||||
Domino et al. ( |
Experimental, in clinic, ( |
Experienced in cannabis use, aged 21–33, males | 2.9 vs. 0.5% (as cigarette) | Amount of cigarettes smoked/THC concentration; effects on size of palpebral fissure and pupil diameter; patellar reflex and heart rate; mood as assessed by Clyde Mood test scores. | After being asked to smoke as much as they could, participants in the higher concentrate condition had a higher increase in the amplitude of the patellar reflex and heart rate, blood pressure, pulse rate changes, and on self-reported mood. | No | No |
Perez-Reyes et al. ( |
Experimental, in lab, ( |
Experienced in cannabis use, aged 23–36, 50% males | 2.54 vs. 1.32 vs. 1.97% | Smoking time, number of puffs, length of puff, length of hold, interval between puffs, THC plasma concentration, peak subjective high, cardiac acceleration |
THC cigarette consumption was dose-dependent when comparing high to low THC content, but there was no evidence of effective titration in THC plasma levels, heart rate acceleration, or reported subjective high. | Yes | No |
Herning et al. ( |
Experimental, in lab, ( |
Experienced in cannabis use, mean age = 29, males | 3.9 vs. 1.2% (as cigarette) | Number of puffs, inter-puff interval, puff volume, puff duration, inhalation volume, inhalation duration, cumulative puff volume, cumulative inhalation volume, total smoking duration; physiological measures: heart rate, blood pressure, skin temperature, and expired CO; verbal self-report of subjective high. | The high potency cigarettes were smoked with more puffs and longer inter-puff intervals with greater inhaled volumes of air, thereby diluting the cannabis smoke. |
Yes | No |
Chait ( |
Experimental, in lab, ( |
Experienced in cannabis use, aged 19–33 (mean = 23), 80% males | 0.9 vs. 1.7 vs. 2.7% THC (as cigarette) | Amount of cigarettes smoked, cut-off time, expired carbon monoxide levels, heart rate, cigarette questionnaire (taste, harshness, draw), visual analog scales, Addiction Research Center Inventory (ARCI), mood |
The post-smoking increase in expired air carbon monoxide levels and psychological measures did not differ between the conditions. | No | No |
Heishman et al. ( |
Experimental, in lab, ( |
Experienced in cannabis use, aged 23–43 (mean age = 31), males | 2.7 vs. 1.3 vs. 0% | Heart rate, smoking topography (inter-puff interval, puff duration, puff volume, maximum flow rate/puff, average flow rate/puff); subjective report of drug effects; a cognitive battery measuring working memory, attention, and motor ability (digit-symbol substation task). | Participants in the high dose condition took smaller puffs, lesser inhalation volumes and shorter puff duration, but did not differ in other smoking topography measures. | Yes | No |
There was no effect on attention—digit span/symbol substation tasks results did not show a dose-response effect. However, subjective reports of dose-related effects of cannabis were obtained. | |||||||
Matthias et al. ( |
Quasi-experimental, in lab, ( |
Experienced in cannabis use, mean age = 23, males | 3.95 vs. 1.77 vs. 0% | COHb saturation, self-report subjective level of intoxication, volume and number of puffs and inter-puff intervals, inhaled volume, breath-holding time, respiratory THC retention, heart rate. | Participants in the stronger dose condition showed reduced intake of smoke and tar yield. |
Yes | No |
Hartman et al. ( |
Experimental, in lab, ( |
Used in past 3 months, at most three times/ week, aged 21–37, 72% males | Placebo (0.008%), low (2.9%), vs. high (6.7%) THC; ground bulk cannabis vaporized |
Blood and plasma cannabinoid analysis. | Of participants that completed all experimental sessions, 10 showed self-titration as indexed by maximum blood THC concentration (μg/L). | – | Mixed |
No behavioral measures of titration presented. | Low concentration cannabis sessions produced consistent max concentration and AUC values in participants, whereas high dose products did not. | ||||||
Hartman et al. ( |
Experimental, in lab, ( |
Used in past 3 months, at most three times/ week, aged 21–37, 72% males | Placebo (0.008%), low (2.9%), or high (6.7%) THC; ground bulk cannabis vaporized vs. libitum for 10 min | Oral fluid THC concentration. | Max THC concentrations in oral fluid were higher in active (low and high) dose cannabis conditions than placebo. No difference in oral fluid THC were detectable between low and high dose overall or at any timepoint post-dose. Given that differences in blood and plasma were detected in some participants, this suggests a failure of oral fluid THC sensitivity. | – | Mixed |
Blood and plasma cannabinoid (also reported in Hartman et al. ( |
|||||||
Bidwell et al. ( |
Experimental (between-subjects), sample recruited |
Experienced flower or concentrate use, mean age = 28, 55–64% males | Concentrates (70 vs. 90%) or flowers (16 vs. 24%) | Plasma cannabinoids; subjective drug intoxication; mood |
THC exposure was significantly higher in the concentrates conditions. Neuro-behavioral outcomes did not differ by potency. | – | Mixed |
Freeman et al. ( |
Naturalistic observational, recruited by word-of-mouth and snowball sample, ( |
Used daily, mean age = 20, 74% males | Own cannabis products varying in potency (1–10) and type (skunk, resin, or herbal); samples analyzed for THC concentrations | Consumption behavior observed from participants smoking their own cannabis in front of the researcher. Self-reported subjective intoxication. Verbal IQ assessed using Wechsler Test of Adult Reading (WTAR). | There was a negative association between THC concentration and amount of cannabis used, but non-daily users were poor in potency estimation. |
Yes | Incomplete |
van der Pol et al. ( |
Naturalistic observational, “coffee-shops” and chain referral sample, (N=98) | Experienced in cannabis use, aged 19–32 (mean = 24), 75% males | Own products varying in THC concentration, and comparisons of 15.72 vs. 3.64% | Smoking topography measured using a portable device [puff volume, duration, inter-puff interval, average velocity (ml/second), peak flow (ml/second), time to peak puff velocity (ml)]. | Higher THC concentration was associated with lower inhalation volume and pace, but not with other topography measures, and positively associated with amount used. The sub-group who used the highest THC product (15.72%) inhaled less than users of average products (3.64%), but the inhalation only halved when the THC concentration was four times higher. | Mixed | Incomplete |
Reinarman ( |
Household survey, (San Francisco |
Experienced in cannabis use, mean age = 34–37, 53–59% males | “Stronger cannabis” | One self-report item: “When using stronger cannabis, do you use…” Less, Same, or More? | Seventy percent of participants self-reported that they use less when using stronger cannabis | Yes | – |
Korf et al. ( |
“Coffee-shops” field interviews, ( |
Smoked cannabis in last 30 days, mean age = 28, 79% males | Own products with dosage assessed using a prompt card showing 0.05, 0.10, 0.20, 0.30 g of cannabis/hash | Self-report of (a) consumption characteristics measured using validated tools, and (b) self-adjustment behaviors in the hypothetical situations that they were smoking more potent products. | Three broad types of cannabis users were identified with mixed results. The type who preferred milder cannabis reported compensating by inhaling less deeply and smoking less. However, the youngest group who consumed the highest monthly dose reported inhaling more deeply, and the oldest group did not report adjustments to intake. | Mixed | – |
There was mixed evidence of titration in experimental studies that were conducted in the 70–90's (
Some of these studies reported differences in smoking topography, such as taking smaller puffs, smaller inhalation volumes, shorter puff duration, longer inter-puff intervals, when using more potent cannabis products (
More recent studies have found some evidence of titration. Hartman et al. conducted an experimental study that evaluated the cannabinoid levels in blood and plasma after the use of vaporized cannabis that varied THC content, with and without alcohol consumption, while allowing
In the analyses of blood and plasma THC concentrations (
Bidwell et al. (
Blood THC and THC metabolite levels differed between the two forms of cannabis, with concentrates producing higher blood levels than flower (
Concentrate users achieved more than double the mean blood THC level of flower users (
Observational studies of cannabis users' behavior when using cannabis that varied in potency have shown mixed evidence of titration (
Freeman et al. (
The study found a negative relationship between THC concentration of the cannabis and the amount of cannabis added to their joints. This relationship was not influenced by the users' frequency of use. The THC levels of the cannabis products were positively correlated with participants' estimation of their potency but the correlation was low. The amount of cannabis consumed was not influenced by product type/potency and participants did not differ in their subjective levels of intoxication.
A similar study in the Netherlands by van der Pol et al. found mixed evidence on whether experienced cannabis users could successfully titrate their THC doses (
van der Pol et al. (
A survey comparing patterns of cannabis use in San Francisco and Amsterdam is often cited as evidence for titration (
Korf et al. (
This review found mixed evidence on how successful cannabis users were in adjusting their dose of more potent cannabis to achieve the same delivery of THC or the same desired psychoactive effects. Older experimental studies found little evidence for titration but often used cannabis with much lower THC levels that differed minimally between conditions. More recent experimental studies of
Observational studies found weak evidence that cannabis smokers reduced their THC doses when using cannabis products with higher levels of THC. In surveys, there were self-reported changes in cannabis use but no assessments were made of whether these produced differences in the THC dose consumed or in its physiological or psychological effects.
The question of most relevance to cannabis policy is whether the users of higher THC products do, in fact, titrate their doses. Epidemiological surveys of adverse effects reported by cannabis users suggest that users of more potent cannabis products incompletely adjust their THC doses. In these surveys, consumers of higher THC cannabis products report more negative consequences than users of less potent products (
There are supportive trends in ecological data. In the USA emergency, hospital, and poisoning center presentations related to cannabis have increased along with the increased use of high THC cannabis products after cannabis legalization (
This review was severely limited by the dearth of rigorous studies on whether people who use cannabis can effectively titrate their doses of higher potency cannabis. The recent rapid increase in THC potency in cannabis products on the market makes it difficult to compare the findings of early studies that used very low THC cannabis products by comparison with cannabis products now consumed.
There may also have been changes over time in the characteristics of people who use cannabis and in their frequency of use. Tolerance develops with the frequency of cannabis intake so cannabis effects will differ between the occasional users often studied in laboratories and the daily cannabis users who account for most of the cannabis consumed (
Routes of administration have also changed over time. Although we did not restrict our search to studies of any specific route of administrations, all the studies we included were of inhaled cannabis products. Methods and ease of titration between different routes of cannabis administration may vary. Vaporization and smoking provide similar cannabinoid delivery (
Some early laboratory studies of cannabis consumption assessed the relationship between blood concentrations of THC and the effects of cannabis (
Smoking topography was measured in some studies, with some authors arguing that it is difficult to assess titration without these measures (
This review was restricted to papers written in English. The predominance of studies from North America may limit the generalisability of these results. The marketing of high THC content products in the USA may have global impacts as online markets are increasingly popular and merchants accessible through online crypto-markets in the USA are prepared to ship cannabis products worldwide (
Our review excluded studies of cannabis when used to alleviate symptoms of chronic medical or mental health conditions. Future research is needed that monitors the prevalence of medical use and assesses the extent to which medicinal cannabis users titrate their doses.
There is an urgent need for larger and better controlled experimental and observational studies of the extent to which cannabis users can and do titrate their THC doses when using more potent cannabis products, such as, cannabis extracts and high potency cannabis flower. This research is needed to inform policymakers on how to reduce harms from the use of high potency cannabis products. It may indicate the need for caps on the potency of cannabis products or higher taxes on more potent cannabis products to discourage their heavy use (
WH and JL: design and conception. JL, DS, and DD: acquisition and analysis of data. JL and DD: first draft. All authors: interpretation of data, subsequent drafts, revision for important intellectual content, final approval, and agreement to be accountable for the work.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We would like to acknowledge the assistance of librarian Miranda Newell, The University of Queensland, for her assistance with the search.
The Supplementary Material for this article can be found online at: