Surveys conducted in April 2020 found that Canadian adolescents aged 14 to 18 years increased their alcohol and cannabis use, and had increased feelings of depression and fear, which are associated with solitary substance use. To our knowledge, no study has examined changes in cannabis exposures in California since recreational legalization in November 2016, the institution of a recreational retail sales market in January 2018, or after the March 2020 statewide shelter-in-place orders intended to reduce the risk of exposure to COVID-19. Past research assessing unintended consequences of cannabis legalizations notes that existing studies are not generalizable to all populations and states; this is particularly relevant for California, which by itself constitutes the world’s largest cannabis market. Previous studies of cannabis exposures were completed before the COVID-19 pandemic and failed to capture exposure rates under pandemic conditions. Since Colorado first legalized recreational cannabis use in 2012, other states have followed and also implemented legalization of recreational cannabis. In this study we reviewed cannabis exposures in California, before and after legalization of recreational cannabis use, after the establishment of recreational retail sales, and during the first nine months of the COVID-19 global pandemic. We also classified product exposures by type to assess which might be associated with exposures among children, in light of popular media reports that have identified group overdoses among children involving cannabis gummies.CPCS serves California’s population of 40 million people, making it the largest poison control provider in the United States. We obtained reports of cannabis exposures from CPCS from January 1, 2010 to December 31, 2020. Inclusion criteria were human exposures to cannabis and cannabis containing products reported within California. Cases were identified by searching the CPCS database for American Association of Poison Control Centers codes relating to cannabis. We excluded calls from outside California.Exposures were defined as an “actual or suspected contact with any substance which has been ingested, inhaled, absorbed, applied to, or injected into the body, regardless of toxicity or clinical manifestation.”Case records were individually reviewed by one of four raters to verify that exposures were actually related to cannabis, to separate human from animal exposure calls, to validate the call involved an exposure rather than a request for information, to check whether exposures involved a single substance or multiple substances, and to detail the nature of the product involved in each exposure given that poison control centers until recently did not classify cannabis exposures beyond “marijuana” and route of exposure .

Records with unclear classifications were reviewed with three other authors . CPCS records were collected and managed using RED Cap, a secure, web-based software platform designed to collect and manage study data.CPCS coded 12,108 exposures from January 2010 to December 2020 as cannabis; 1,351 of these exposures did not meet inclusion criteria, as they were miscoded,trim trays involved animals, were calls from outside California, or were requests for information. Of the remaining 10,757 exposures, 20 percent involved someone under the age of six, 6 percent someone between the ages of six and twelve, 24 percent someone between the ages of thirteen and nineteen, and 50 percent an adult . Forty-four percent of exposures were female, and 56 percent were male. Additionally, 79 percent of the exposures involved ingestion, 18 percent involved inhalation, and 3 percent other routes including topical, rectal, parenteral, subcutaneous, or ophthalmic, as shown in Table 1. Although the total population of California grew by an estimated 6.1% from 2010 to 2020, with an increase of 22.5% in those under the age of 18 and a 6% increase in those under 5 years of age; calls to CPCS related to cannabis more than tripled over the same period. The number of cannabis exposure calls in proportion of all incoming calls is described in Table 2. Among children under the age of six years, 2,130 calls were assigned a code indicating the reason for exposure, of these, 2,107 were coded as unintentional exposures, zero as intentional, and the remaining 23 were coded as other . Among children aged six to 12 years, 625 calls were assigned a reason code, and of these 504 were coded as unintentional, 84 as intentional, and the remaining 38 as other. Our interrupted time series analysis first considered overall changes in exposures after legalization of use, initiation of retail sales, and after the COVID-19 shelter-in-place order. As noted in methods, ITSA coefficients represent estimated monthly increases or decreases in reported exposures after an intervention. Following recreational legalization in 2016, estimated monthly cannabis exposures increased significantly . Following the implementation of retail sales in 2018, cannabis exposures increased significantly as well . However, no significant change in cannabis exposures was observed following the shelter-in-place order. A graph of exposures over time is provided in Figure 1; detailed estimates are provided in Table 3. We continued by comparing changes in exposures for two age groups: those under thirteen years old and those thirteen years and older. Age is provided in CPCS records and this categorization follows AAPCC convention; exposures without information on age were excluded. Cannabis exposures in those under thirteen increased significantly both after recreational legalization and after the opening of the retail sales market , but not following the shelter-in-place order .

For those thirteen and older, there was no significant change over time. As a result, although exposures in children under thirteen were the minority in January 2010, by December 2020 they represented nearly half of all exposures, as shown in Figure 2. To assess possible changes in exposures by product type, we organized ingestion exposures by product type and grouped these into categories. We identified significant increases over in the number of exposures for gummies , candies , chocolate , dabs , edibles in the form of drinks , hemp, joints , blunts , cannabis oils , vapes , other edible products , and all other products between 2010 and 2020. However, there was no change in the trend of exposures for cookies, brownies, other edible baked goods, hash, plant products, or synthetic products. We aggregated these categories into nine broad product types: chocolate and candy, other edibles and drinks, gummies, brownies, cookies, and other baked goods, new technology , traditional products , oils, hemp products, and synthetics, then graphed exposures . Chocolate and candy, other edibles and drinks, and gummies increased from levels near zero prior to recreational legalization to thousands of exposures per year by 2020. For example, there were only 16 total reported gummies exposures in the six years between 2010 and 2015; these increased to 409 exposures in 2020 alone.We analyzed trends in cannabis exposures reported to the CPCS before and after the legalization of recreational cannabis in November 2016, the establishment of recreational retail sales in January 2018, and the institution of a statewide shelter-in-place order due to the COVID-19 pandemic in March 2020 and found that as expected, exposures increased following recreational legalization and the establishment of retail sales, consistent with previous studies. However despite expectations we did not find a significant change in cannabis consumption following the COVID-19 shelter-inplace order. This finding may reflect that only nine months of exposure cases following March 2020 were available at the time of this study. We also found that cannabis exposures in children under thirteen increased significantly following recreational legalization and initiation of retail sales but did not increase for teens and adults. As a result, although cannabis exposures were uncommon among young children in 2010, by 2020 they constituted nearly half of all exposures.Cannabis edibles such as gummies, candy, and other dessert-like products have been involved with increased use in younger users. Our detailed records review found that a common exposure after 2018 involved a child or group of children finding cannabis edibles that they perceived to be normal candy and consuming an entire package.

Particularly among the youngest children the primary reason for exposure was accidental ingestion, in which children or their caregivers mistakenly identified cannabis gummies as ordinary candy. Cases in which cannabis gummies and other edibles are mistaken for non-cannabis products may result from issues with packaging. Although California regulates the potency of cannabis edibles and requires opaque, resealable packaging, each edible can contain up to 10 mg of THC and each package up to 100 mg of THC; as a result, even a single gummy represents a high dosage for a naïve user,trimming trays particularly a child. By comparison, edible regulations in Canada, for example, place a limit of 10 mg of THC per package, even if it the package contains multiple edibles, as well as requiring plain packaging and larger warning labels. As a result, a child who accidently consumed an entire bag of cannabis gummies in Canada would likely be exposed to the same level of THC as one who consumed a single gummy in California. We note that Canadian regulations on packaging were instituted in 2020, so there is limited data to assess potential changes in pediatric exposures after this policy change. However, given reported confusion among both children and caregivers about whether candy products contain cannabis, instituting similar regulations such as plain packaging and lower doses per edible, or expanding on them by requiring individual packaging, offer potential for reducing the high levels of exposures among children. Our study has limitations. The data were drawn from a single state, limiting potential generalizability; however, California’s status as the most populous with the largest cannabis market allows us to assess trends that would not be possible in smaller areas. Moreover, the more granular data provided by CPCS made it possible to classify product types; these data are not available at the national level. Using poison control data only captures data volunteered by patients and providers and these may not capture general patterns of use. In addition, although CPCS seeks to create a case report linked to individuals rather than to group exposures, in some cases, multiple exposures were reported in a single record . As a result, these findings are likely to be underestimates of actual exposures.

Our classification of product types was limited by reporters, who may use a range of terms to describe cannabis products ; as a result, we were unable to categorize all exposures and may have failed to identify additional products associated with exposures. Finally, given that the study was observational in nature, we could only identify associations between cannabis exposures and policy interventions rather than establishing causality. Despite these limitations, the absence of other contemporaneous factors expected to increase exposures, as well as the consistency of these findings with prior research, suggest that recreational legalization and sales were associated with significant increases in exposures, particularly among children.Cannabis is the most commonly used illicit psychoactive substance in developed nations . While a majority of cannabis users do not report problems, 10–30% of those who ever use cannabis meet criteria for a lifetime history of cannabis abuse or dependence as defined by the fourth edition of the Diagnostic and Statistical Manual . Recently, changes to the diagnostic criteria for substance use disorder have been made in DSM-5 , including several for the diagnosis of cannabis use disorders . Across the broad range of substance use disorders,the distinction between abuse and dependence has been replaced by a unidimensional symptom count, with endorsement of 2 or more symptoms resulting in a DSM-5 diagnosis of substance use disorder ; the DSM-IV criterion of legal problems has been eliminated from the diagnostic repertoire; and a new criterion for the DSM-5, craving has been added. More specifically for cannabis, withdrawal is now a criterion. A wealth of psychometric evaluations in epidemiological and clinical samples support these recommendations; however, the impact of these revisions on the prevalence of cannabis use disorders under the new DSM-5 classification remains largely unexplored. A recent study of Australian adults found a modest reduction in the rate of cannabis use disorder with the transition from DSM-IV to DSM-5 , while another study of individuals with substance use disorders note damodest increase of 4% . Twin studies indicate that 50–60% of the variation in cannabis use disorders can be attributed to heritable influences.Despite this robust heritability estimate, association studies for cannabis use disorders have largely failed to identify genetic variants of significant and replicable effect.