Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Am J Psychiatry . Author manuscript; available in PMC: 2026 Apr 11. Published in final edited form as: Am J Psychiatry. 2026 Mar 4;183(4):260–268. doi: 10.1176/appi.ajp.20250336 Search in PMC Search in PubMed View in NLM Catalog Add to search Association of Cannabis Use Disorder Versus Other Substance Use Disorders With Psychiatric Conditions: A Propensity-Matched Retrospective Cohort Analysis Ryan C Nicholson Ryan C Nicholson , M.D., M.P.H. 1 Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore. Find articles by Ryan C Nicholson 1 , Una E Choi Una E Choi , M.D. 2 Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital, Boston. Find articles by Una E Choi 2 , Ramin Mojtabai Ramin Mojtabai , M.D., Ph.D. 3 Department of Psychiatry and Behavioral Sciences, Tulane University School of Medicine, New Orleans. Find articles by Ramin Mojtabai 3 , Johannes Thrul Johannes Thrul , Ph.D. 4 Department of Mental Health, Johns Hopkins University Bloomberg School of Public Health, Baltimore. 5 Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore. 6 Centre for Alcohol Policy Research, La Trobe University, Melbourne. Find articles by Johannes Thrul 4, 5, 6 Author information Article notes Copyright and License information 1 Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore. 2 Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital, Boston. 3 Department of Psychiatry and Behavioral Sciences, Tulane University School of Medicine, New Orleans. 4 Department of Mental Health, Johns Hopkins University Bloomberg School of Public Health, Baltimore. 5 Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore. 6 Centre for Alcohol Policy Research, La Trobe University, Melbourne. ✉ Send correspondence to Dr. Nicholson ( [email protected] ). # Drs. Nicholson and Choi contributed equally to this work. Issue date 2026 Apr 1. PMC Copyright notice PMCID: PMC13067382 NIHMSID: NIHMS2155230 PMID: 41776722 The publisher's version of this article is available at Am J Psychiatry Abstract Objective: The authors compared the risk of mental disorders between patients with cannabis use disorder (CUD) and those with other substance use disorders (SUDs). Methods: The TriNetX Research Network was queried to identify patients with SUDs and no preceding mental disorders and compare 1) adult patients with CUD only versus those with other SUDs, 2) pediatric patients with CUD only versus those with other SUDs, and 3) adult patients with CUD plus another SUD versus those with comorbid noncannabis SUDs. Propensity score matching was performed on demographic characteristics and 24 risk factors or comorbidities. Subsequent diagnosis of schizophrenia and other common mental disorders was assessed. Results: Compared to adults with other SUDs, those with noncomorbid CUD (N=345,903 for both cohorts) had a lower risk of schizophrenia (0.34% vs. 0.42%; relative risk [RR]=0.81, 95% CI=0.75, 0.88), depression (1.35% vs. 1.74%; RR=0.78, 95% CI=0.75, 0.81), and psychotic disorders (0.36% vs. 0.52%; RR=0.68, 95% CI=0.63, 0.73). Compared to pediatric patients with other SUDs, those with CUD (N=24,793 for both cohorts) had a higher risk of schizophrenia (0.29% vs. 0.19%; RR=1.52, 95% CI=1.06, 2.19), depression (1.65% vs. 1.27%; RR=1.30, 95% CI=1.13, 1.51), and anxiety disorders (8.13% vs. 6.71%; RR=1.21, 95% CI=1.14, 1.29). Compared to adult patients with other SUDs, those with CUD and a comorbid SUD (N=203,916 for both cohorts) had a decreased risk of schizophrenia (1.94% vs. 2.25%; RR=0.86, 95% CI=0.83, 0.90), depression (3.98% vs. 5.67%; RR=0.70, 95% CI=0.68, 0.72), bipolar disorder (4.23% vs. 5.60%; RR=0.76, 95% CI=0.74, 0.78), and anxiety disorders (16.20% vs. 21.36%; RR=0.76, 95% CI=0.75, 0.77). Conclusions: CUD-associated mental health risks varied by age and comorbid SUDs, possibly due to earlier onset of mental disorders in cannabis users or age-related differences in CUD effects. Cannabis use has risen in the United States as more states have legalized it ( 1 - 3 ). As of January 2025, 39 states had legalized medical cannabis use, and 24 had legalized recreational use ( 4 ). Previous studies have demonstrated that use was increasing even before widespread legalization in the United States ( 1 ), although legalization may be accelerating this trend ( 5 ). Cannabis use is particularly prevalent among young people in the United States, with prevalence highest in the 18–34 age group, at slightly over 25% in 2022 ( 6 ). Although lower than the rate of use among younger people, the rate among adults ages 35–49 increased between 2021 and 2022 (from 14.25% to 17.23%), highlighting that this trend is not isolated to a particular demographic. Health effects associated with cannabis use include negative impacts on cognition and memory and an increased number of fatal motor vehicle accidents following consumption ( 7 ). Of particular interest are the rising rates of cannabis use disorder (CUD) in states where cannabis is legal ( 2 ), as well as the interplay between CUD and mental disorders. Recent evidence suggests an association between CUD or other SUDs and the development of schizophrenia or other psychotic disorders ( 8 - 11 ). This phenomenon is not isolated to the United States; a link was demonstrated by researchers in Denmark who analyzed data for nearly 7 million individuals in that country from 1972 to 2021 ( 10 ). In a historical cohort study of Swedish military conscripts, Zammit et al. ( 12 ) determined that cannabis use was a potent risk factor for schizophrenia, and this relationship was not explained by other drug use. Despite differences in sociodemographic characteristics, policy, regulation, and culture between the United States and Europe, many of the concerns about cannabis are the same across these regions: increases in prevalence of use, product potency, and rates of CUD treatment ( 1 , 5 , 13 , 14 ). Groups particularly susceptible to mental disorders include younger individuals or males diagnosed with CUD, although the risk for subsequent new diagnosis of a psychiatric condition seems to also be linked to the age at first use and may be dose-dependent ( 8 ). Recent genetically informed studies, such as by Cheng et al. ( 9 ), suggest that there may be a causal link between CUD and mental disorders. Khokar et al. ( 15 ) posited that the same genetic determinants for schizophrenia make individuals more vulnerable to substance use. If this vulnerability develops before psychotic symptoms, then substance use may increase this susceptibility and increase risk for a psychotic disorder by damaging neural connections. Although recent research has focused extensively on the relationship between CUD and psychotic disorders, the question of whether, or how, the impact of CUD differs from the impact of other SUDs on psychotic disorders and other mental disorders remains underexplored. Given the high rates of co-occurrence of psychiatric conditions with SUDs and the unanswered questions regarding the difference in impact between cannabis and other substances, our aim in this study was to determine whether the risk of receiving a mental health diagnosis differs between patients diagnosed with CUD and those diagnosed with other SUDs. We performed a retrospective cohort analysis of patients diagnosed with an SUD before any mental disorder. Furthermore, given the high rates of polysubstance use among those with an SUD and the unique clinical characteristics and outcomes of this patient population ( 16 , 17 ), we performed subanalyses for patients with multiple SUDs. To our knowledge, this is one of the first large cohort comparisons of mental health outcomes between patients diagnosed with CUD and those diagnosed with other SUDs. Our findings may inform psychiatric care by identifying whether patients with certain SUDs, in particular CUD, require more extensive screening for psychotic or other mental disorders and highlight important considerations for cannabis policy and education. METHODS Database We conducted a retrospective cohort analysis using the TriNetX Research Network, a platform comprising medical and pharmacy claims as well as electronic health record data and vital statistics for over 130 million patients. Inpatient and outpatient data are included, as well as data from academic and nonacademic health care institutions across the United States. We queried the network to identify patients matching the inclusion criteria based on demographic characteristics and ICD-10 codes up to December 31, 2024. Only patients with data from the past 20 years were included in the cohorts; however, the length of the period of available data varies across patients. TriNetX complies with the Health Insurance Portability and Accountability Act (HIPAA), and data are de-identified per the standard outlined in Section 164.514(b)( 1 ) of the HIPAA Privacy Rule. To protect patient identities, only de-identified, aggregate data are provided within the platform, and patient counts for diagnoses with <10 patients total are obfuscated. The TriNetX platform features statistical tools for researchers built on Java 11.0.16, R 4.0.2, and Python 3.0.7. This report adheres to the applicable EQUATOR guide-lines. Cohort Construction We constructed cohorts of patients ≥18 years of age who had a diagnosis of an SUD but no prior mental disorders. We identified SUDs using ICD-10 codes: alcohol-related disorders (F10), opioid-related disorders (F11), cannabis-related disorders (F12), sedative-, hypnotic-, or anxiolytic-related disorders (F13), cocaine-related disorders (F14), other stimulant-related disorders (F15), hallucinogen-related disorders (F16), inhalant-related disorders (F18), and other psychoactive substance–related disorders (F19; i.e., undetermined drug use, polysubstance use, etc.) ( Figure 1 ). FIGURE 1. Diagram detailing cohort construction and comparisons for cannabis use disorder versus other substance use disorders, both comorbid and alone a . Open in a new tab a ADHD, attention deficit hyperactivity disorder; CUD, cannabis use disorder; SUD, substance use disorder. We constructed a primary cohort of patients diagnosed with CUD and no other SUDs and a comparison cohort consisting of patients diagnosed with any single SUD other than CUD. Given the concern for the effects of cannabis use on brain development ( 10 ), we performed an additional analysis of patients ≤17 years of age who were diagnosed with either CUD only or any other SUD. Our secondary focus was the risk of mental disorders among adult patients with co-occurring SUDs. We constructed two cohorts for this analysis: 1) patients with CUD co-occurring with at least one other SUD and no prior mental disorders, and 2) patients with at least two co-occurring SUDs other than CUD and no prior mental disorders. To compare risk between specific SUDs and CUD, in further analyses we stratified adult cohorts for the most common SUDs within the database: alcohol, cocaine, and opioid use disorders. Outcome Measures The observation period was from the time a patient met the inclusion criteria to when they no longer had any health information tracked in the TriNetX database or to the study end date. We identified outcomes using ICD-10 codes as determined by claims data, which are listed in Table S1 in the online supplement . Outcomes included the risks of the following diagnoses after an SUD diagnosis: schizophrenia, nonrecurrent depressive episodes, recurrent major depressive disorder (MDD), suicide attempts, bipolar disorder, phobic and anxiety disorders, attention deficit hyperactivity disorders (ADHD), borderline personality disorder, and nonmood psychotic disorders. We limited the outcomes of individual SUD-stratified analyses to schizophrenia, MDD, anxiety disorders, and other psychotic disorders. Statistical Analysis To compare risk for mental disorders, we utilized propensity score matching. We calculated propensity scores with a logistic regression model and matched them without replacement at a 1:1 ratio using a greedy nearest-neighbor algorithm with a caliper width set to 0.1 standard deviations of the propensity score. Patient order was randomized before matching to eliminate record order bias. We matched patients on characteristics including age, gender, race, and ethnicity. ICD-10 codes captured other social factors, such as income levels, history of abuse or intimate partner violence, and certain adverse childhood events, among others. We also matched on diagnoses per ICD-10 codes including nicotine dependence, newborns affected by maternal drug use, family history of mental or behavioral disorders, hemolytic anemias, bullous pemphigoid, celiac disease, chronic interstitial cystitis, and hyperthyroidism. We selected co-variates based on prior studies associating these diagnoses with SUDs and mental disorders ( 18 , 19 ). A complete list of ICD-10 codes and patient characteristics used for matching is provided in Table S1 in the online supplement . In our primary and secondary analyses, we compared cohorts for the risk of new diagnoses of mental disorders using risk ratios and 95% confidence intervals. We defined significance as a two-sided alpha <0.05. As TriNetX does not output p values for relative risk, we calculated these values using Wald tests. RESULTS Adults With CUD Versus Any Other SUD Before matching, there were 369,391 patients in the CUD-only cohort and 1,603,784 patients in the other-SUD cohort (see Table S2 in the online supplement ). After matching, there were 345,903 patients per cohort. The CUD-only cohort had lower risk of schizophrenia (0.34% vs. 0.42%; relative risk [RR]=0.81, 95% CI=0.75, 0.88), nonrecurrent depressive episodes (6.09% vs. 6.73%; RR=0.91, 95% CI=0.89, 0.92), recurrent MDD (1.35% vs. 1.74%; RR=0.78, 95% CI=0.75, 0.81), suicide attempts (0.06% vs. 0.10%; RR=0.59, 95% CI=0.50, 0.70), bipolar disorder (1.07% vs. 1.22%; RR=0.88, 95% CI=0.84, 0.92), and psychotic disorders (0.36% vs. 0.52%; RR=0.68, 95% CI=0.63, 0.73) ( Table 1 , Figure 2 ). Patients with CUD had a higher risk of subsequent diagnosis of anxiety disorders than patients with any other SUD (8.62% vs. 8.48%; RR=1.02, 95% CI=1.00, 1.03). TABLE 1. Relative risks of mental health disorders for adult and pediatric cohorts with cannabis use disorder only compared to other substance use disorders a Cohort Psychiatric diagnosis N % N % RR 95% CI p Adult CUD only (N=345,903) Adult other SUDs (N=345,903) Schizophrenia 1,179 0.34 1,448 0.42 0.81 0.75, 0.88 <0.0001 Nonrecurrent depressive episode 21,074 6.09 23,291 6.73 0.91 0.89, 0.92 <0.0001 MDD 4,677 1.35 6,029 1.74 0.78 0.75, 0.81 <0.0001 Suicide attempt 209 0.06 354 0.10 0.59 0.50, 0.70 <0.0001 Bipolar disorder 3,708 1.07 4,230 1.22 0.88 0.84, 0.92 <0.0001 Anxiety disorders 29,815 8.62 29,333 8.48 1.02 1.00, 1.03 0.041 ADHD 2,638 0.76 2,750 0.80 0.96 0.91, 1.01 0.12 Borderline personality disorder 399 0.12 370 0.11 1.08 0.94, 1.24 0.29 Psychotic disorders 1,227 0.36 1,805 0.52 0.68 0.63, 0.73 <0.0001 Adult co-occurring CUD+SUD (N=203,916) Adult co-occurring other SUDs (N=203,916) Schizophrenia 3,953 1.94 4,578 2.25 0.86 0.83, 0.90 <0.0001 Nonrecurrent depressive episode 28,040 13.75 39,338 19.29 0.71 0.70, 0.72 <0.0001 MDD 8,109 3.98 11,556 5.67 0.70 0.68, 0.72 <0.0001 Suicide attempt 785 0.39 1,239 0.61 0.63 0.58, 0.69 <0.0001 Bipolar disorder 8,631 4.23 11,425 5.60 0.76 0.74, 0.78 <0.0001 Anxiety disorders 33,041 16.20 43,563 21.36 0.76 0.75, 0.77 <0.0001 ADHD 3,394 1.66 4,771 2.34 0.71 0.68, 0.74 <0.0001 Borderline personality disorder 940 0.46 1,147 0.56 0.82 0.75, 0.89 <0.0001 Psychotic disorders 4,517 2.22 5,511 2.70 0.82 0.79, 0.85 <0.0001 Pediatric CUD only (N=24,793) Pediatric other SUDs only (N=24,793) Schizophrenia 73 0.29 48 0.19 1.52 1.06, 2.19 0.024 Nonrecurrent depressive episode 1,526 6.16 1,341 5.41 1.14 1.06, 1.22 0.0004 MDD 409 1.65 314 1.27 1.30 1.13, 1.51 0.0004 Suicide attempt 58 0.23 62 0.25 0.94 0.65, 1.34 0.73 Bipolar disorder 254 1.02 248 1.00 1.02 0.86, 1.22 0.80 Anxiety disorders 2,015 8.13 1,663 6.71 1.21 1.14, 1.29 <0.0001 ADHD 579 2.34 676 2.73 0.86 0.77, 0.96 0.0058 Borderline personality disorder 41 0.17 36 0.15 1.14 0.73, 1.78 0.58 Psychotic disorders 125 0.50 126 0.51 0.99 0.78, 1.27 0.95 Open in a new tab a The p values were calculated using Wald tests. ADHD, attention deficit hyperactivity disorder; CUD, cannabis use disorder; MDD, major depressive disorder; RR, relative risk; SUD, substance use disorder. FIGURE 2. Forest plot of relative risk of schizophrenia for comparisons of adults with cannabis use disorder, other substance use disorders, or both a . Open in a new tab a Error bars represent the bounds of 95% confidence intervals. CUD, cannabis use disorder; SUD, substance use disorder. Pediatric Patients With CUD Versus Any Other SUD Before matching, the pediatric CUD-only cohort consisted of 32,032 patients, and the pediatric other-SUD cohort consisted of 39,820 patients. After matching, each cohort comprised 24,793 patients (see Table S2 in the online supplement ). Compared to the other-SUD cohort, patients with CUD had increased risk of schizophrenia (0.29% vs. 0.19%; RR=1.52, 95% CI=1.06, 2.19), nonrecurrent depressive episodes (6.16% vs. 5.41%; RR=1.14, 95% CI=1.06, 1.22), recurrent MDD (1.65% vs. 1.27%; RR=1.30, 95% CI=1.13, 1.51), and anxiety disorders (8.13% vs. 6.71%; RR=1.21, 95% CI=1.14, 1.29) ( Table 1 , Figure 3 ). The CUD cohort had a lower risk of ADHD (2.34% vs. 2.73%; RR=0.86, 95% CI=0.77, 0.96). FIGURE 3. Forest plot of relative risks for adult and pediatric patients with cannabis use disorder only versus any other substance use disorder a . Open in a new tab a Error bars represent the bounds of 95% confidence intervals. Adults With Co-Occurring CUD Versus Co-Occurring Other SUDs Before matching, the cohort with co-occurring CUD had 230,275 patients and the cohort with co-occurring other SUDs had 359,726 patients (see Table S2 in the online supplement ). After matching, there were 203,916 patients in each cohort. The cohort with co-occurring CUD had a lower risk of new diagnoses of schizophrenia (1.94% vs. 2.25%; RR=0.86, 95% CI=0.83, 0.90), nonrecurrent depression (13.75% vs. 19.29%; RR=0.71, 95% CI=0.70, 0.72), recurrent MDD (3.98% vs. 5.67%; RR=0.70, 95% CI=0.68, 0.72), suicide attempts (0.39% vs. 0.61%; RR=0.63, 95% CI=0.58, 0.69), bipolar disorder (4.23% vs. 5.60%; RR=0.76, 95% CI=0.74, 0.78), anxiety disorders (16.20% vs. 21.36%; RR=0.76, 95% CI=0.75, 0.77), ADHD (1.66% vs. 2.34%; RR=0.71, 95% CI=0.68, 0.74), borderline personality disorder (0.46% vs. 0.56%; RR=0.82, 95% CI=0.75, 0.89), and psychotic disorders (2.22% vs. 2.70%; RR=0.82, 95% CI=0.79, 0.85) ( Table 1 ). Adults With CUD Versus Individual SUDs Before matching, there were 369,391 patients in the CUD cohort, 1,109,330 patients in the alcohol use disorder cohort, 52,555 patients in the cocaine use disorder cohort, and 200,820 patients in the opioid use disorder cohort (see Table S3 in the online supplement ). Postmatch counts were 318,361 patients, 52,551 patients, and 173,869 patients in the alcohol, cocaine, and opioid use disorder comparisons, respectively. Compared to the alcohol use disorder cohort, the CUD-only cohort had a lower risk of MDD (1.33% vs. 1.79%; RR=0.74, 95% CI=0.71, 0.77) and psychotic disorders (0.35% vs. 0.45%; RR=0.77, 95% CI=0.71, 0.83) ( Table 2 ). Compared to the cocaine use disorder cohort, the CUD-only cohort had a lower risk of schizophrenia (0.32% vs. 0.79%; RR=0.40, 95% CI=0.33, 0.48), MDD (1.16% vs. 1.44%; RR=0.80, 95% CI=0.72, 0.89), and psychotic disorders (0.34% vs. 0.53%; RR=0.64, 95% CI=0.53, 0.77). Patients with CUD had a higher risk of anxiety disorders compared to those with cocaine use disorder (7.48% vs. 6.47%; RR=1.16, 95% CI=1.11, 1.21). Compared to the opioid use disorder cohort, the CUD-only cohort had a higher risk of schizophrenia (0.25% vs. 0.22%; RR=1.17, 95% CI=1.02, 1.34) and a lower risk of MDD (1.40% vs. 2.39%; RR=0.58, 95% CI=0.56, 0.61) and anxiety disorders (9.54% vs. 11.08%; RR=0.86, 95% CI=0.85, 0.88). TABLE 2. Relative risks of mental health disorders for adults with cannabis use disorder only compared to adults with alcohol, cocaine, and opioid use disorders a Cohort Psychiatric diagnosis N % N % RR 95% CI p CUD only (N=318,361) Alcohol use disorder (N=318,361) Schizophrenia 1,049 0.33 1,022 0.32 1.03 0.94, 1.12 0.57 MDD 4,226 1.33 5,692 1.79 0.74 0.71, 0.77 <0.0001 Bipolar disorder 3,330 1.05 3,144 0.99 1.06 1.01, 1.11 0.021 Anxiety disorders 27,363 8.60 27,116 8.52 1.01 0.99, 1.03 0.27 Psychotic disorders 1,108 0.35 1,443 0.45 0.77 0.71, 0.83 <0.0001 CUD only (N=52,551) Cocaine use disorder (N=52,551) Schizophrenia 166 0.32 416 0.79 0.40 0.33, 0.48 <0.0001 MDD 607 1.16 758 1.44 0.80 0.72, 0.89 <0.0001 Bipolar disorder 411 0.78 804 1.53 0.51 0.45, 0.58 <0.0001 Anxiety disorders 3,928 7.48 3,400 6.47 1.16 1.11, 1.21 <0.0001 Psychotic disorders 177 0.34 278 0.53 0.64 0.53, 0.77 <0.0001 CUD only (N=173,869) Opioid use disorder (N=173,869) Schizophrenia 435 0.25 373 0.22 1.17 1.02, 1.34 0.029 MDD 2,430 1.40 4,162 2.39 0.58 0.56, 0.61 <0.0001 Bipolar disorder 1,763 1.01 2,373 1.37 0.74 0.70, 0.79 <0.0001 Anxiety disorders 16,591 9.54 19,293 11.08 0.86 0.85, 0.88 <0.0001 Psychotic disorders 499 0.29 464 0.27 1.08 0.95, 1.22 0.26 Open in a new tab a The p values were calculated using Wald tests. CUD, cannabis use disorder; MDD, major depressive disorder; RR, relative risk. DISCUSSION Our propensity-score-matching study used data from the TriNetX Research Network, a large database of electronic health records, to examine the risk of new mental disorder diagnoses among patients with SUDs, with a focus on the comparison of CUD to other SUDs in adult and pediatric populations. Results differed between adult and pediatric patients. Among pediatric patients, CUD was associated with a greater risk of schizophrenia, MDD, and anxiety disorders compared to other SUDs. Adult patients had considerable variation in risk, depending on SUD type. Compared to adults with noncannabis SUDs, those with CUD, either alone or co-occurring with another SUD, generally had a lower risk for all mental disorders, including MDD, suicide attempts, and psychotic disorders. However, among adults with a single SUD, those with CUD had no significant difference in risk of being diagnosed with ADHD or borderline personality disorder. The association between any SUD and mental disorders has been well-established by previous longitudinal studies in both Europe and the United States ( 10 , 12 , 20 - 22 ). These investigations are particularly salient given the rising use and legalization of cannabis ( 1 - 3 , 5 ). Whereas previous studies examined patients with single or comorbid SUDs compared to nonusers ( 15 , 21 , 23 ), we compared mental disorder outcomes between patients with different SUDs, and the patients in our sample were diagnosed with SUDs prior to mental disorders. Furthermore, many existing studies are based on historical data from periods before cannabis was legal; the majority of countries from which longitudinal studies originate have still not legalized recreational cannabis use ( 24 , 25 ). Advancing research requires data that are reflective of current trends in use and regulation, updated clinical guidelines, and increased potency of cannabis products. Although our study examined data from the United States, findings would likely be similar among European populations, given the similarities in legalization concerns and trends in use ( 1 , 5 , 13 , 14 ), although studies outside the United States are needed. Our finding of differential risks for mental disorders between various SUDs suggests that there may be a component of the mental health–SUD interaction that is being over-looked. While the previously established self-medication and primary addiction hypotheses have detailed the link between mental disorders and subsequent SUD development, we characterized the reverse relationship, finding that mental disorders vary in risk by type of preceding SUD and by age. When considered in conjunction with other research, such as Chiappelli and colleagues’ study that found that certain mental disorders may disproportionately lead people to favor certain substances ( 26 ), our findings suggest shared vulnerabilities, such as common genetic predispositions underlying specific comorbidities. Indeed, our results indicate that cannabis may have a greater role than opioids in the subsequent diagnosis of schizophrenia, possibly due to a greater role of highly cannabis-sensitive neural circuitry, such as the endocannabinoid system, in this disease than the opioid receptor system. This possibility is supported by recent results from the Adolescent Brain Cognitive Development (ABCD) Study that suggest a possible role of self-medication in the link between psychotic symptoms and cannabis use ( 27 ). More to this end is our finding that co-occurring noncannabis SUDs differ in mental health outcomes compared to CUD co-occurring with another SUD. A secondary analysis of clinical trial data by Liu et al. similarly found that co-occurring SUDs had a greater association with anxiety disorders than single SUDs ( 23 ). The reason for this is unclear, but it may be due to compounded neurocircuitry abnormalities resulting from the interaction of multiple SUDs. Previous studies demonstrated that changes to brain structure differed depending on the substance; for example, gray matter volume and white matter integrity were reduced in people with alcohol use disorder but not in those with CUD ( 28 ). This is informed by a study showing reduced cortical region thickness in synthetic cannabinoid users ( 29 ). Unlike the Δ 9 -tetrahydrocannabinol (Δ 9 -THC) in cannabis, synthetic cannabinoids are cannabinoid receptor full agonists, may act at other receptors, such as N -methyl- d -aspartate (NMDA) receptors, and have exceptionally stronger binding affinity to cannabinoid CB 1 and CB 2 receptors ( 30 ). Given that our co-occurring-CUD cohort had a lower risk for almost all mental health outcomes, it may be that the structural changes from SUDs compound one another because of drug potency or mixed mechanisms of action, or that the potency of typical cannabis products and Δ 9 -THC is insufficient to cause a similar effect. In our study, comorbid SUDs were operationalized as the presence of at least two SUDs. The actual number of SUDs, however, was not recorded. Additional research may seek to explore the impact of specific combinations or numbers of comorbid SUDs and monitor subsequent changes to brain structure. Our finding of increased mental health outcome risk in pediatric patients with only CUD suggests that the endocannabinoid system may be especially relevant in the unmasking or maintenance of conditions in this population. This is open to multiple interpretations. First, vulnerable individuals who develop mental disorders early on would be underrepresented in the adult samples. In other words, exposure to heavy cannabis use may expedite the onset of mental disorders in adolescents but deplete the pool of vulnerable individuals who reach adulthood, thus creating an impression of a protective effect of cannabis use in these adults. Regional differences in brain maturation may be another explanation ( 31 ). CB 1 , the most prevalent cannabinoid receptor in the brain, reaches maximal concentrations during adolescence and is concentrated in areas associated with higher-level cognition. Binding of CB 1 by THC may alter the critical prefrontal cortex maturation processes occurring during adolescence ( 32 ). It is also possible that exposure to cannabis unmasks a preexisting genetic liability to schizophrenia and mood disorders ( 33 , 34 ). Furthermore, developing areas of the adolescent brain involved in schizophrenia or MDD may be more affected by cannabis than the same regions of the adult brain. This possibility correlates with prior studies demonstrating a link between earlier age of first cannabis use and subsequent development of schizophrenia ( 10 ); combined with the effects of cannabis on cognition and academic achievement ( 7 ), it presents an argument for well-rounded education about cannabis use for younger people. Further research is needed to explore the causal links between CUD and development of comorbid mental disorders. Interestingly, the mental health consequences of CUD across age ranges are reflected in public attitudes toward cannabis legalization. A study from California indicated that legalization of cannabis was associated with a greater perception of positive health impacts among adults ( 35 ), whereas a Canadian study showed that legalization was associated with a greater perception of harms among youths ( 36 ). Strategies to address adolescent cannabis use may involve intervention in primary care settings, where youths are often asked about their drug use. Individuals reporting use may benefit from improved messaging to curb use, including by encouraging moderation of drug use as a form of harm reduction. Furthermore, policy initiatives may take a harm-reduction approach by setting limits on the potency of cannabis products, although quantitative studies are needed to identify a nonarbitrary limit. Preventing use before it begins is most important. Successful strategies may include encouraging parents who use cannabis to restrict their adolescents’ access to it at home, similar to encouraging the use of gun safes or locks, which are effective in reducing firearm-related injuries ( 37 ). Our study has several notable limitations. First, although the TriNetX database contains substantial clinical information, allowing for sophisticated cohort construction and analysis, it cannot quantify the severity of an SUD or the age at first drug use. In studies of SUDs, especially CUD, this variable is important in quantifying risk ( 10 ). Related to this, information regarding the types of cannabis product used was not recorded. Cannabis products vary considerably in active compounds, concentration, and route of administration ( 38 ), all of which need to be tracked in greater detail in electronic health records ( 39 ). Second, the database contains information from the past 20 years, but not all health care organizations contain the same length of patient information. Consequently, we lack information on mental disorders that might have developed years after SUDs in patients who were no longer treated by a health care organization tracked by TriNetX. Similarly, conditions that typically originate early in life, such as ADHD and borderline personality disorder, may start developing in childhood prior to SUDs but not be diagnosed until adulthood, or conditions during childhood may go undiagnosed if the patient’s medical history is not transferred across health care organizations. Our cohorts and analyses rely on the validity of ICD-10 codes for identifying SUDs and mental disorders among individuals who seek treatment and interact with the health care system. How well the coding corresponds to an SUD varies by condition, with validation studies demonstrating high specificities but sensitivities ranging from 47% to 83% ( 40 ). Individuals who do not seek treatment—a group that comprises the majority of those with SUDs and mental disorders—are not included in these cohorts. Lastly, we acknowledge that this study alone cannot establish a firm cause-and-effect relationship between CUD or other SUDs and a mental disorder. CONCLUSIONS After propensity score matching, we found that the risk for mental disorders differed between adult and pediatric patient populations diagnosed with CUD or other SUDs, and that adult patients diagnosed either with CUD alone or with CUD and a comorbid SUD had a lower risk for most mental disorders as compared to patients diagnosed with noncannabis SUDs. Differences in substance-specific addiction-reward pathways may contribute to differential risk for mental disorders among adults, whereas cannabis use in pediatric populations may result in greater susceptibility for schizophrenia and MDD because of vulnerabilities in the developing adolescent brain. The association of cannabis use with mental disorders requires further investigation with behavioral, neuroimaging, and pathology studies. Supplementary Material Supplement NIHMS2155230-supplement-Supplement.pdf (478.4KB, pdf) Acknowledgments Dr. Mojtabai and Dr. Thrul receive grant support from NIDA. This study was completed without external financial assistance or funding. Dr. Mojtabai’s contribution was partly supported by the Louisiana Board of Regents Endowed Chairs for Eminent Scholars program. Dr. Thrul’s contribution was partly supported by NIDA grant UM1DA059000. Dr. Mojtabai has received royalties from Wolters Kluwer for contributions to UpToDate, and he has served as a consultant for MindMed. The other authors report no financial relationships with commercial interests. REFERENCES 1. Compton WM, Han B, Jones CM, et al. : Marijuana use and use disorders in adults in the USA, 2002–14: analysis of annual cross-sectional surveys. Lancet Psychiatry 2016; 3:954–964 [ DOI ] [ PubMed ] [ Google Scholar ] 2. Lapham GT, Matson TE, Bobb JF, et al. : Prevalence of cannabis use disorder and reasons for use among adults in a US state where recreational cannabis use is legal. JAMA Netw Open 2023; 6:e2328934. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Hasin D, Walsh C: Trends over time in adult cannabis use: a review of recent findings. Curr Opin Psychol 2021; 38:80–85 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. State Medical Cannabis Laws. Washington, DC, National Conference of State Legislatures. https://www.ncsl.org/health/state-med-ical-cannabis-laws [ Google Scholar ] 5. Goodman S, Wadsworth E, Leos-Toro C, et al. : Prevalence and forms of cannabis use in legal vs illegal recreational cannabis markets. Int J Drug Policy 2020; 76:102658. [ DOI ] [ PubMed ] [ Google Scholar ] 6. Mattingly DT, Richardson MK, Hart JL: Prevalence of and trends in current cannabis use among US youth and adults, 2013–2022. Drug Alcohol Depend Rep 2024; 12:100253. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Karila L, Roux P, Rolland B, et al. : Acute and long-term effects of cannabis use: a review. Curr Pharm Des 2014; 20:4112–4118 [ DOI ] [ PubMed ] [ Google Scholar ] 8. Ortiz-Medina MB, Perea M, Torales J, et al. : Cannabis consumption and psychosis or schizophrenia development. Int J Soc Psychiatry 2018; 64:690–704 [ DOI ] [ PubMed ] [ Google Scholar ] 9. Cheng W, Parker N, Karadag N, et al. : The relationship between cannabis use, schizophrenia, and bipolar disorder: a genetically informed study. Lancet Psychiatry 2023; 10:441–451 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Hjorthøj C, Compton W, Starzer M, et al. : Association between cannabis use disorder and schizophrenia stronger in young males than in females. Psychol Med 2023; 53:7322–7328 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Hamilton I: Cannabis, psychosis and schizophrenia: unravelling a complex interaction. Addiction 2017; 112:1653–1657 [ DOI ] [ PubMed ] [ Google Scholar ] 12. Zammit S, Allebeck P, Andreasson S, et al. : Self reported cannabis use as a risk factor for schizophrenia in Swedish conscripts of 1969: historical cohort study. BMJ 2002; 325:1199. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Manthey J, Freeman TP, Kilian C, et al. : Public health monitoring of cannabis use in Europe: prevalence of use, cannabis potency, and treatment rates. Lancet Reg Health Eur 2021; 10:100227. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. ElSohly MA, Chandra S, Radwan M, et al. : A comprehensive review of cannabis potency in the United States in the last decade. Biol Psychiatry Cogn Neurosci Neuroimaging 2021; 6:603–606 [ DOI ] [ PubMed ] [ Google Scholar ] 15. Khokhar JY, Dwiel L, Henricks A, et al. : The link between schizophrenia and substance use disorder: a unifying hypothesis. Schizophr Res 2018; 194:78–85 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Lin LA, Bohnert ASB, Blow FC, et al. : Polysubstance use and association with opioid use disorder treatment in the US Veterans Health Administration. Addiction 2021; 116:96–104 [ DOI ] [ PubMed ] [ Google Scholar ] 17. Bunting AM, Krawczyk N, Choo T-H, et al. : Polysubstance use before and during treatment with medication for opioid use disorder: prevalence and association with treatment outcomes. J Subst Abuse Treat 2022; 143:108830. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Varese F, Smeets F, Drukker M, et al. : Childhood adversities increase the risk of psychosis: a meta-analysis of patient-control, prospective- and cross-sectional cohort studies. Schizophr Bull 2012; 38:661–671 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Eaton WW, Byrne M, Ewald H, et al. : Association of schizophrenia and autoimmune diseases: linkage of Danish national registers. Am J Psychiatry 2006; 163:521–528 [ DOI ] [ PubMed ] [ Google Scholar ] 20. van Os J, Bak M, Hanssen M, et al. : Cannabis use and psychosis: a longitudinal population-based study. Am J Epidemiol 2002; 156:319–327 [ DOI ] [ PubMed ] [ Google Scholar ] 21. Hall W, Degenhardt L: Cannabis use and the risk of developing a psychotic disorder. World Psychiatry 2008; 7:68–71 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Moore THM, Zammit S, Lingford-Hughes A, et al. : Cannabis use and risk of psychotic or affective mental health outcomes: a systematic review. Lancet 2007; 370:319–328 [ DOI ] [ PubMed ] [ Google Scholar ] 23. Liu S, Nwabueze C, Pan Y, et al. : Polysubstance use, mood disorders, and chronic conditions with anxiety in opioid patients. West J Nurs Res 2021; 44:1088–1099 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Søgaard TF, Nygaard-Christensen M, Frank VA: Danish cannabis policy revisited: multiple framings of cannabis use in policy discourse. Nordisk Alkohol Nark 2021; 38:377–393 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Marthinussen HF: Decriminalisation of drug use: the outlook for reform in Norway. Bergen J Crim Law Crim Justice 2018; 6:53–67 [ Google Scholar ] 26. Chiappelli J, Chen S, Hackman A, et al. : Evidence for differential opioid use disorder in schizophrenia in an addiction treatment population. Schizophr Res 2018; 194:26–31 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Osborne KJ, Barch DM, Jackson JJ, et al. : Psychosis spectrum symptoms before and after adolescent cannabis use initiation. JAMA Psychiatry 2025; 82:181–190 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Thayer RE, YorkWilliams S, Karoly HC, et al. : Structural neuro-imaging correlates of alcohol and cannabis use in adolescents and adults. Addiction 2017; 112:2144–2154 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Çolak C, Çelik ZÇ, Zorlu N, et al. : Cortical thickness and subcortical volumes in adolescent synthetic cannabinoid users with or without ADHD: a preliminary study. Noro Psikiyatr Ars 2019; 56:167–172 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Fattore L: Synthetic cannabinoids: further evidence supporting the relationship between cannabinoids and psychosis. Biol Psychiatry 2016; 79:539–548 [ DOI ] [ PubMed ] [ Google Scholar ] 31. Bava S, Tapert SF: Adolescent brain development and the risk for alcohol and other drug problems. Neuropsychol Rev 2010; 20:398–413 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Karlsgodt KH: Cannabis use in adolescence: vulnerability to cognitive and psychological effects. Biol Psychiatry Glob Open Sci 2023; 3:167–168 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. De Hert M, Wampers M, Jendricko T, et al. : Effects of cannabis use on age at onset in schizophrenia and bipolar disorder. Schizophr Res 2011; 126:270–276 [ DOI ] [ PubMed ] [ Google Scholar ] 34. Large M, Sharma S, Compton MT, et al. : Cannabis use and earlier onset of psychosis: a systematic meta-analysis. Arch Gen Psychiatry 2011; 68:555–561 [ DOI ] [ PubMed ] [ Google Scholar ] 35. Gali K, Winter SJ, Ahuja NJ, et al. : Changes in cannabis use, exposure, and health perceptions following legalization of adult recreational cannabis use in California: a prospective observational study. Subst Abuse Treat Prev Policy 2021; 16:16. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Nguyen HV, Mital S, Bornstein S: Short-term effects of recreational cannabis legalization on youth cannabis initiation. J Adolesc Health 2023; 72:111–117 [ DOI ] [ PubMed ] [ Google Scholar ] 37. Volano P, Bonne S, Duncan T, et al. : Prevention of firearm injuries with gun safety devices and safe storage: an Eastern Association for the Surgery of Trauma Systematic Review. J Trauma Acute Care Surg 2018; 84:1003–1011 [ DOI ] [ PubMed ] [ Google Scholar ] 38. Spindle TR, Bonn-Miller MO, Vandrey R: Changing landscape of cannabis: novel products, formulations, and methods of administration. Curr Opin Psychol 2019; 30:98–102 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Thrul J, Vandrey R: The Cannabis and Health Research Initiative will help integrate medicinal cannabis in healthcare. Nat Med 2024; 30:3394–3395 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. McGrew KM, Homco JB, Garwe T, et al. : Validity of International Classification of Diseases codes in identifying illicit drug use target conditions using medical record data as a reference standard: a systematic review. Drug Alcohol Depend 2020; 208:107825. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplement NIHMS2155230-supplement-Supplement.pdf (478.4KB, pdf) ACTIONS View on publisher site PDF (582.4 KB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top