Medicinal and combined medicinal/recreational cannabis use in California following the passage of Proposition 64

Cannabis policy in the United States has shifted over recent decades, with California first legalizing medicinal use in 1996 and adult-use in 2016 via Proposition 64. Despite federal Schedule I status, medicinal use has expanded, evidence for therapeutic benefits is mixed across conditions, and medical societies remain cautious. Following legalization, product availability and access have increased, potentially altering patient behaviors, sources of guidance, and relationships with healthcare providers. The purpose of this study was to explore patterns of THC-containing cannabis use among California adults who use for medicinal-only or combined medicinal plus recreational (M+R) purposes after Prop 64, including behaviors, symptoms targeted, perceptions of patient–provider relationships, information sources, and perceived effects.

Prior work and reviews (e.g., National Academies 2017) suggest potential benefits of cannabis for select conditions (e.g., neuropathic pain) while highlighting limited or inconsistent evidence for broader indications (chronic pain, spasticity, multiple sclerosis, IBS, mental health). FDA approvals are limited to specific cannabinoid-based pharmaceuticals (e.g., dronabinol, nabilone, cannabidiol for select indications). Professional societies generally urge caution due to insufficient evidence. Earlier surveys indicate growing medicinal use, with motivations including reducing prescription medications. Studies suggest M+R users are more likely to use inhaled products and less likely to consult healthcare professionals compared with medicinal-only users. However, many prior studies focused on specific patient populations (e.g., chronic pain), underscoring a need for broader, population-based assessments in a post-legalization context.

Design: Impact 64 used a three-tiered mixed-methods approach: (1) semi-structured qualitative interviews with 23 subject matter experts (SMEs); (2) a preliminary questionnaire (n=200); and (3) a detailed, population-matched online questionnaire targeting 5,000 California adults. Cannabis was defined as THC-containing products in the questionnaire. Recruitment and eligibility: Adults (≥21 years), California residents, English or Spanish fluent; exclusions included employment in cannabis industry, market research, or public relations. Quota sampling aligned participants with California’s 2020 census demographics. Screening captured demographics and cannabis history; participants were unaware of the study purpose at screening. From the screener, a subset completed the full questionnaire targeting: current users (use in past 3 months; goal n=4,000), former users (no use in ≥4 months; goal n=500), and non-users (goal n=500). Data collection: The full online questionnaire launched Dec 2, 2022 and closed Feb 6, 2023. Weighting: Iterative proportional fitting (rake weighting) matched screener data to California census demographics for age, gender, race/ethnicity, and income. Each full-questionnaire subgroup (current, former, non-users) was weighted to reflect that subgroup’s target demographics based on the screener. Statistical analysis: Descriptive statistics, chi-square tests, and multinomial logistic regression were conducted in SPSS v28 and JMP Pro v17. Significance threshold p<0.05. Analytic sample for this report: Of 15,200 screening respondents with cannabis history, 37% current users, 30% former, 33% non-users. A total of 5,178 completed the full questionnaire: 4,020 current users, 523 former users, 635 non-users. Among current users, 2,430 (61%) reported medicinal use; analysis here includes medicinal-only (n=711; 18% of current users) and M+R users (n=1,719; 43%). Recreational-only users were excluded.

• Prevalence and demographics: 61% of current users reported medicinal use (medicinal-only or M+R). Medicinal-only users were more likely female (OR 1.6, p<0.001), have children in the household (OR 1.5, p<0.001), and initiated cannabis later (mean 34 vs 23 years, p<0.001). Monthly spending was lower for medicinal-only ($127, SD 176) vs M+R ($186, SD 235; p<0.001).
• Frequency and product types: Very frequent use (multiple times/day) was lower in medicinal-only vs M+R (25% vs 47%, p<0.001); occasional use was higher (39% vs 21%, p<0.001). M+R more often used most product types (e.g., dried flower 65% vs 34%; vaping 43% vs 22%; dabbing 24% vs 11%; beverages 18% vs 11%); edibles 53% vs 48%*; topical and oil/tinctures similar (ns).
• MMIC: Current MMIC holder ~33–34%; past cardholder more common in M+R (37% vs 28%**); never had MMIC more common in medicinal-only (38% vs 29%).
• Reasons for use: Pain common in both (69% M+R vs 62% M*). M+R more likely to use for stress (75% vs 49%), anxiety (72% vs 53%), sleep (69% vs 57%), PTSD (24% vs 18%), decreased appetite (24% vs 10%). Aggregated categories: any physical symptom 89% M+R vs 85% M; any mental symptom 88% M+R vs 62% M**. M+R more often sought the high (75% vs 42%), and were more likely to treat multiple symptoms (e.g., 4 symptoms: 32% M+R vs 18% M).
• Patient–provider relationship: Perceived acceptance of medicinal use by primary doctor was similar (68% M+R vs 66%). Acceptance of recreational use higher in M+R (56% vs 44%, p=0.0001). Comfort discussing cannabis higher for M+R (83% vs 75%, p=0.0001); primary doctor awareness higher in M+R (73% vs 66%, p=0.0001). About 59–61% used cannabis instead of prescription medications; among those using for sleep, pain, and PTSD, substitution (instead of prescriptions) was significantly more likely.
• Information sources: Internet (57% M+R vs 44% M**), family/friends (52% vs 47%), budtenders (42% vs 28%), health professionals (~26–27%), workplace (10% vs 7%*). Among MMIC holders (n=829), 65% obtained recommendations from primary care; 34% from alternative providers/online.
• Sources of cannabis and licensure: Dispensaries were primary source (84% M+R vs 72% M**); delivery (45% vs 28%); family/friends (36% vs 22%); grow own (14% vs 10%*). Licensed outlets were very/extremely important to most, more so for medicinal-only (72% vs 66%, p<0.01). Overall, 94% reported using licensed dispensaries.
• Perceived effects: Positive effects reported by both groups. Physical health improvements were higher in medicinal-only (76% vs 69%); mental health, emotional health, relationships, and clear head/focus were more often reported by M+R (e.g., mental health 88% M+R vs 76% M; emotional health 88% vs 80%; relationships 62% vs 55%; focus 62% vs 59%**).
• Adverse effects: M+R more likely to report finances (21% vs 17%), paranoia (20% vs 12%), fatigue (20% vs 16%), memory loss (17% vs 10%), lack of motivation (24% vs 14%), weight gain (20% vs 14%), and dependency (9% vs 6%**). Earlier initiation (<17 years) was associated with many adverse outcomes.
• Locations/activities: Use at home common (93–94%). M+R more likely to use at someone else’s home (39% vs 15%), parties/indoor public (38% vs 12%), outdoor public (27% vs 9%), work (10% vs 5%), in car (5% vs 2%), during recreational creative activities (50% vs 34%), and while drinking alcohol (36% vs 23%). Only ~4% used with children present (not tabulated).
• Driving after use: Medicinal-only users reported longer wait times before feeling safe to drive across product types. For flower: an hour or less (29% M+R vs 19% M**); wait until next day (10% M+R vs 18% M**). For edibles: an hour or less (16% M+R vs 20% M*); 2–4 h (28% M+R vs 11% M**); next day (24% M+R vs 35% M**). For vaping/dabbing: an hour or less (27% M+R vs 19% M**); next day (12% M+R vs 18% M*).

Findings address how adult Californians using cannabis for medicinal-only versus combined medicinal plus recreational purposes differ in demographics, initiation age, frequency and modes of use, reasons for use, information and supply sources, perceived benefits and harms, and risk-related behaviors (e.g., concurrent alcohol use, driving). Medicinal-only users were more often female, initiated later, spent less, used less frequently, prioritized licensed dispensaries, and waited longer before driving. M+R users reported a broader range of symptoms (especially mental health) and more adverse effects, consistent with higher frequency and diverse product use. Both groups primarily relied on non-clinical sources for guidance, and many substituted cannabis for prescription medications, underscoring gaps in clinical engagement. These results are relevant for public health messaging, harm reduction, and clinical practice, highlighting the need for provider education and patient–provider communication to guide safer, evidence-informed medicinal use in a legalized market.

Cannabis use for medicinal purposes is common among California adults post-Prop 64. Medicinal-only users differ from M+R users across key domains, including demographics, use patterns, motivations, sourcing, and perceived effects. The study underscores the importance of bolstering patient–provider relationships, expanding provider training on cannabis counseling (indications, dosing, risks, harm reduction), and ensuring access to reliable information. Future research should: (1) examine dose, potency, and product formulation effects on benefits/harms; (2) explore factors underlying driving-risk perceptions and behaviors; (3) assess storage and child-access prevention; (4) evaluate outcomes of integrating cannabis counseling into primary care; and (5) expand analyses to younger populations and jurisdictions with different legal contexts.

• Generalizability is limited to California; findings may not apply to states without legal access or with different cannabis markets and attitudes.
• The study included only adults ≥21 years; results do not inform underage use patterns.
• Online data collection may under-represent individuals without reliable internet/phone access, potentially excluding some marginalized groups.
• Self-report data are subject to recall and social desirability biases.
• The study did not assess product potency/dose, storage practices, or child access, which could influence outcomes and safety behaviors.