Health Service Use Among Young Adults With a History of Adolescent Cannabis Use.

Abstract

This cohort study investigates the associations between distinct patterns of cannabis use in adolescence and medical care utilization for mental and physical health conditions in young adulthood.

Key Points

Question

Is adolescent cannabis use associated with medical care utilization for mental and physical health conditions in young adulthood?

Findings

In this cohort study of 1591 individuals followed up to age 23 years, adolescents who initiated cannabis use before 15 years of age and used frequently had increased odds of medical care utilization for both mental and physical conditions in young adulthood, compared with adolescents without cannabis use. Adolescents with late-onset use had no increased odds for mental health care utilization but did exhibit higher odds for physical conditions.

Meaning

Evidence from this study indicated that early and frequent cannabis use was associated with medical care utilization later in life, underscoring the importance of delaying or reducing adolescent cannabis use.

Abstract

Importance

Adolescent cannabis use is a substantial public health concern given its associations with adverse mental and physical health outcomes. Understanding how distinct use patterns are associated with medical care utilization in young adulthood is critical for prevention.

Objective

To examine the association between patterns of adolescent cannabis use and medical care utilization for mental and physical health conditions in young adulthood.

Design, Setting, and Participants

This population-based birth cohort study linked to population-wide administrative medical care databases was conducted in the Province of Québec, Canada. Participants included individuals recruited for the Québec Longitudinal Study of Child Development, followed up from birth (1997-1998) to 23 years of age (2021). Data were analyzed November 2023 to February 2025.

Exposure

Self-reported past 12-month cannabis use at ages 12, 13, 15, and 17 years.

Main Outcomes and Measures

Medical care utilization for any mental disorder (including common mental disorders and substance-related disorders), suicide-related behaviors, and any physical health condition (including respiratory diseases, injuries and poisoning, and other physical diseases) between ages 18 and 23 years. Analyses were adjusted for 32 individual, family, and community-level confounders measured from birth to younger than 12 years of age using overlap weights.

Results

Data for 1591 individuals (818 female [51.4%]; mean [SD] age at first exposure assessment, 12.1 [0.3] years) were analyzed. Three distinct adolescent cannabis use patterns were identified via group-based trajectory modeling: nonuse (948 [59.6%]), late-onset use (318 [20.0%], initiating after age 15 years), and early-onset and frequent use (325 [20.4%], initiating before age 15 years). In fully adjusted analyses, individuals with early-onset and frequent use had significantly higher odds of medical care utilization for any mental disorder (odds ratio [OR], 1.51 [95% CI, 1.10-2.08]), common mental disorders (OR, 1.57 [95% CI, 1.12-2.21]), any physical health condition (OR, 1.86 [95% CI, 1.30-2.67]), injuries and poisoning (OR, 1.41 [1.05-1.89]), and other physical diseases (OR, 1.47 [95% CI, 1.08-1.98]), compared with individuals with no use. Individuals with late-onset use did not differ significantly from those with nonuse for mental health outcomes (OR, 1.13 [95% CI, 0.80-1.58]) but had higher odds of medical care utilization for any physical health condition (OR, 1.63 [95% CI, 1.16-2.28]).

Conclusions and Relevance

Findings of this birth cohort study indicated that early-onset and frequent cannabis use was associated with greater medical care utilization for both mental and physical health conditions in young adulthood. These findings support the relevance of delaying initiation and reducing intensity of cannabis use during adolescence.

Article notes

Untitled section

Received 2025 May 21; Accepted 2025 Aug 13; Collection date 2025 Oct.

Author notes

Author Contributions: Drs Martínez and Orri had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.Concept and design: Martínez, Chadi, Castellanos-Ryan, Vergunst, Séguin, Tremblay, Geoffroy, Orri.Acquisition, analysis, or interpretation of data: Martínez, Chadi, Castellanos-Ryan, Vergunst, Dorais, Vitaro, Temcheff, Boivin, Côté, Geoffroy, Orri.Drafting of the manuscript: Martínez, Orri.Critical review of the manuscript for important intellectual content: All authors.Statistical analysis: Martínez, Castellanos-Ryan, Dorais, Orri.Obtained funding: Martínez, Castellanos-Ryan, Vergunst, Séguin, Vitaro, Temcheff, Tremblay, Boivin, Côté, Geoffroy, Orri.Administrative, technical, or material support: Martínez, Vitaro, Côté, Geoffroy.Supervision: Chadi, Côté, Orri.Conflict of Interest Disclosures: Drs Chadi, Séguin, Vitaro, Temcheff, Tremblay, Boivin, Côté, Geoffroy, and Orri reported receiving grants from the Canadian Institutes of Health Research during the conduct of the study. Dr Martínez reported receiving a Canadian Institutes of Health Research postdoctoral fellowship and a publication support grant from the Services and Public Health Analysis Platform of the Réseau québécois sur le suicide, les troubles de l’humeur et les troubles associés during the conduct of the study. Drs Côté, Geoffroy, and Orri reported holding a Canada Research Chair. Dr Geoffoy reported receiving funding from the Fonds de recherche du Québec – Société et culture for the Papageno Suicide Prevention Team. No other disclosures were reported.Funding/Support: The Québec Longitudinal Study of Child Development was supported by funding from the Ministère de la Santé et des Services sociaux, Ministère de la Famille, Ministère de l’Éducation and Ministère de l’Enseignement supérieur (Québec ministries), the Lucie and André Chagnon Foundation, the Institut de recherche Robert-Sauvé en santé et en sécurité du travail, the Research Centre of the Sainte-Justine University Hospital, the Ministère de l’Emploi et de la Solidarité sociale, and the Institut de la statistique du Québec. Additional funding was received by the Fonds de Recherche du Québec – Santé, the Fonds de Recherche du Québec – Société et Culture, the Social Science and Humanities Research Council of Canada), and the Canadian Institutes of Health Research (including grant 114984).Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.Data Sharing Statement: See Supplement 2.

Introduction

Adolescence is a critical developmental period marked by risk-taking behaviors and heightened sensitivity to substance-related neurobiological changes.¹ Cannabis is among the most commonly used substances by adolescents in Canada and worldwide.² Meta-analytical evidence associates regular adolescent use of cannabis with increased odds of substance-related problems, suicidality, and depression in young adulthood.^3,4 Early-onset use (before 15 years of age) has been associated with higher risk of psychosis,⁵ and longer-term use has been associated with physical health issues (eg, respiratory symptoms) that may emerge in adolescence.⁶ In Canada, where nonmedical cannabis was legalized in 2018, 12% of adolescents in 7th to 9th grade (typically 12 to 15 years of age) reported past-year cannabis use, highlighting the emergence of use during early adolescence.⁷ Widespread adolescent cannabis use, rising product potency,⁸ and increasing cannabis-related harms⁹ make adolescent cannabis use a substantial public health concern.¹⁰

Although not all adolescents are equally susceptible, a subset who begin using early and frequently appear especially vulnerable to subsequent health and social issues,¹¹ reporting increased odds of anxiety and depression,^{12,13,14,15,16,17} higher risk of cannabis and alcohol-related harms,^{12,13,14,15,16,18,19,20,21,22,23,24} and worse physical health.^16,25 Yet previous studies often overlook critical family and environmental factors measured in early childhood that may confound the association between adolescent cannabis use and health outcomes^22,24 and primarily rely on self-reported health outcomes, which may not fully capture severe health conditions requiring medical care.

To address these gaps, we used data from the Québec Longitudinal Study of Child Development (QLSCD),²⁶ a population-based birth cohort linked to population-wide administrative medical care databases in Québec, Canada. The QLSCD provides comprehensive early-life information at the individual, familial, and environmental level,²⁶ while administrative data offer objective, detailed records of medical care utilization within a universal health care system. By clarifying how different adolescent cannabis use patterns are associated with subsequent medical care utilization for mental and physical health conditions in early adulthood, these findings may inform targeted prevention strategies and guide resource allocation to mitigate long-term harms.

Methods

The QLSCD was approved by the Research Ethics Board of the Institut de la statistique du Québec (ISQ), Sainte Justine Hospital Research Centre and Montreal West Island Integrated University Health and Social Services Centre. Ethical approval for the overall linkage project was obtained by the Montreal West Island Integrated University Health and Social Services Centre’s Research Ethics Board. Analyses for this report were approved by McGill University’s Faculty of Medicine and Health Science Research Ethics Board. Written informed consent was obtained from participants or parents at each data collection. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.²⁷

Participants

Participants were drawn from the QLSCD, a population-based birth cohort of 2120 children born in Québec (1997-1998).²⁶ The sampling strategy and eligibility criteria are described elsewhere.²⁶ Data were collected at multiple time points from infancy to young adulthood, including waves at ages 0.5, 1.5, 2.5, 3.5, 4, 5, 6, 7, 8, 10, 12, 13, 15, and 17 years. The current study included data for 1591 participants (75.0% of the original cohort) who provided at least 1 data point on adolescent cannabis use between ages 12 and 17 years. Exclusion was primarily due to unavailability for follow-up. Administrative medical data linkage was available for 100% of the analytic sample.

Exposure

Adolescents reported past-year cannabis use at ages 12, 13, 15, and 17 years. Responses ranged from “I didn’t” (0) to “every day” (6) (eTable 1 in Supplement 1).

Outcomes

The ISQ provided access to QLSCD data linked to anonymized administrative medical care databases, including physician claims (Regie de l’assurance maladie du Québec), inpatient discharge records (Maintenance et exploitation des données pour l’étude de la cientèle hospitalière), and emergency department admissions (Banque de données communes des urgences [BDCU]), providing primary and secondary International Statistical Classification of Diseases, Ninth Revision (ICD-9) and ICD-10 diagnoses from birth (1997-1998) to age 23 years (2021) (BDCU available from 16 years of age) (eTable 2 in Supplement 1). We created dichotomous indicators of any medical care use between ages 18 and 23 years. Main outcomes were any mental disorder, including common mental disorders (depressive, anxiety, or adjustment disorders), severe mental disorders (bipolar or schizophrenia spectrum and other psychotic disorders), and substance-related disorders (alcohol-, cannabis-, and other drug-related disorders), and any physical health condition, encompassing diseases of the respiratory system, injuries, and poisoning, and other physical diseases. Secondary outcomes included common mental disorders, substance-related disorders, suicide-related behaviors (including suicide attempt and suicidal ideation), diseases of the respiratory system, injuries and poisoning, and other physical diseases. The other physical diseases category included conditions (eg, gastrointestinal tract, neurological, infectious, and cardiovascular) not captured in the prior groups but identified in the literature as potentially associated with cannabis use.⁶ A complete list of ICD codes is provided in eTable 2 in Supplement 1.

Outcome selection was guided by substantive knowledge (eTable 2 in Supplement 1) and used a minimum of 10 cases per exposure stratum to ensure a stable model estimation and to comply with ISQ identity-protection guidelines. This threshold was applied only to define outcome-specific analytic samples and did not reflect the number of parameters estimated in outcome models. Linkage to administrative medical care databases was performed using participants’ unique health insurance numbers.

Covariates

We identified 32 preexposure individual, family, and community-level confounders using a directed acyclic graph informed by substantive expertise (eAppendix 1 and eFigure 1 in Supplement 1). Table 1^{28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45} details the confounders extracted from the QLSCD, including the survey waves used for their assessment and descriptions of each variable. These confounders, measured via validated instruments, capture a wide range of early-life vulnerabilities from birth to younger than 12 years of age. Additional data from administrative medical care databases included sex (male or female, as recorded at birth in health records, which does not include gender identity) and utilization of medical care before age 12 years for any mental disorder, neurodevelopmental disorders (including disabilities, disorders of psychological development, attention-deficit/hyperactivity disorder, stereotyped movement disorders, and tic disorders), disturbances of conduct and emotions occurring in childhood or adolescence, and any physical health condition. Management of the administrative medical care databases followed the same approach outlined for the outcomes (eTable 2 in Supplement 1). Self-reported use of tobacco (past 30 days, scored from 1 [“never used”] to 5 [“every day”]) and alcohol (past 12 months, scored from 1 [“I didn’t use] to 7 [“every day”]) were assessed at ages 12, 13, 15, and 17 years. Parental alcohol use was coded on an ordinal frequency scale: 7, every day; 6, 4 to 6 times per week; 5, 2 to 3 times per week; 4, once a week; and 1, less than once a month. Socioeconomic status is expressed as a z score from a composite deprivation index.

Table 1.Study Confounders Extracted From the Québec Longitudinal Study of Child Development (QLSCD)

Confounder	Survey wave (child age, y)^a										Brief description^b
Confounder	0.5	1.5	2.5	3.5	4.0	5.0	6.0	7.0	8.0	10.0	Brief description^b
Maternal age	Yes	No	No	No	No	No	No	No	No	No	Maternal age when child was ~ 5 mo old: ≤24 y, 25 to 29 y, 30 to 34 y, and ≥35 y
Parental ethno-cultural origins^c	Yes	No	No	No	No	No	No	No	No	No	Parents Canadian-born or born outside Canada
Family type	Yes	No	No	No	No	No	No	No	No	No	Intact, 2-parent families or blended/single-parent families
Household socioeconomic status	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes	NLSCY,²⁸ composite score, mean (SD): 0 (3) of parental education, occupational status, income
Neighborhood conflict^d	Yes	Yes	No	No	Yes	No	Yes	No	Yes	Yes	NQ,²⁹ PMK-reported perception of neighborhood unsafety and lack of social cohesion (n = 5)
Maternal prenatal substance use^e	Yes	No	No	No	No	No	No	No	No	No	NLSCY and DIS,^30,31 maternal-reported use of tobacco, alcohol, or illegal drugs during pregnancy
Parental substance use^f	Yes	Yes	Yes	Yes	No	No	No	No	No	No	NLSCY and DIS,^30,31 parental-reported use of tobacco, alcohol, or illegal drugs during the past 12 mo
Maternal depression	Yes	Yes	No	Yes	No	Yes	No	Yes	No	Yes	CES-D,³² maternal self-report of depressive symptoms (n = 6)
Family dysfunction^g^,^h	Yes	Yes	No	No	No	No	Yes	No	No	No	CMHFFS,³³ PMK-reported family dynamics, emotional connection, and behavioral guidance (n ~ 6)
Positive parenting practices^g	No	No	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	NLSCY and PPS,^30,34 PMK-reported supportive interactions and nurturing care (n ~ 3)
Coercive parenting practices^g	No	No	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	NLSCY and PPS,^30,34 PMK-reported negative and hostile interactions (n ~ 3)
Consequent parenting practices^g	No	No	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	NLSCY and PPS,^30,34 PMK-reported application of consistent rules for similar behaviors (n ~ 4)
Parental monitoring^g	No	No	No	No	No	No	No	No	No	Yes	PMS,³⁵ PMK-reported knowledge of child’s peer network, communication, and oversight (n = 5)
Verbal skills	No	No	No	Yes	No	Yes	Yes	No	No	Yes	PPVT-R,³⁶ age-corrected and standardized scores, mean (SD): 100 (15)
Internalizing behaviors^g	No	No	No	No	No	No	Yes	Yes	Yes	Yes	SBQ,^37,38 teacher-rated anxiety (n = 4), emotional distress (n ~ 5), and withdrawal (n ~ 4)
Externalizing behaviors^g	No	No	No	No	No	No	Yes	Yes	Yes	Yes	SBQ,^37,38 teacher-rated aggression (n = 10), opposition (n = 4), and inattention/hyperactivity (n = 9)
Social skills^g	No	No	No	No	No	No	Yes	Yes	Yes	Yes	EDI,³⁹ teacher-rated sociability (n ~ 9) and responsibility (n ~ 6)
Victimization^g	No	No	No	No	No	No	Yes	Yes	Yes	Yes	SRVS,⁴⁰ self-reported verbal, physical, and relational peer victimization (n ~ 5)
Academic performanceⁱ	No	No	No	No	No	No	No	Yes	Yes	Yes	NLSCY,⁴¹ teacher-rated overall academic performance (n = 1)
Depressive symptoms^g	No	No	No	No	No	No	No	No	No	Yes	CDI,⁴² self-reported depressive symptoms (n = 8)
Deviant behaviors^g	No	No	No	No	No	No	No	No	No	Yes	NLSCY and QNTS,^43,44 self-reported risk-taking and rule-breaking behaviors (n = 9)

Statistical Analysis

Group-based trajectory modeling (GBTM), a finite mixture approach,⁴⁶ was used to identify subpopulations with distinct patterns of cannabis use across adolescence. Unlike variable-centered approaches that model continuous variation on a single dimension (eg, either cumulative use, age of onset, or intensity of use),⁴⁷ GBTM uses a person-centered approach that identifies qualitatively distinct usage profiles that reflect real-world heterogeneity in timing and intensity, enabling more precise identification of at-risk subgroups.

GBTM used the maximum likelihood estimation, retaining cases with at least 1 data point. Nonrandom attrition was assumed to occur at a constant rate within each group. Males and individuals from households of lower socioeconomic status were more likely to have missing follow-up cannabis data within the study sample; thus, parameters were estimated using a conditional estimator, with these covariates used to model group membership and dropout probability. Count and zero-inflated count distribution models were tested for the cannabis data. Time was measured in years (12-17) in the statistical model. The mean ages at each wave were 12.1, 13.1, 15.1, and 17.2 years, with SDs of 0.2, indicating minimal variability in time intervals across individuals.

An automated algorithm optimized the bayesian information criterion to select the optimal number of trajectories (1-5 groups) and the best polynomial order (constant, linear, or quadratic). We evaluated 40 combinations. Final model selection followed the Nagin recommended criteria: highest bayesian information criterion and probability correct model, mean posterior probabilities at least 0.7, odds of correct classification 5.0 or higher, entropy 0.700 or higher, and substantive interpretability.⁴⁶ To confirm that the class solution did not converge to a local maximum of the maximum likelihood distribution, the model was rerun with 100 sets of randomly generated starting values.⁴⁸

For missing covariate data, multiple imputation through fully conditional specification was used to handle arbitrary missing value patterns.⁴⁹ We generated 20 imputed datasets with 30 burn-in iterations each. To investigate the association between patterns of cannabis use and medical care utilization, we fitted unadjusted and overlap-weighted logistic regression models, reporting odds ratios (ORs) and their 95% CIs. In overlap-weighted models, weights were derived from a multinomial propensity score model including 32 early-life confounders. These weights strongly balanced the covariate distributions across exposure groups without including them directly in the outcome model, thereby mitigating confounding while avoiding overfitting.⁵⁰ The overlap-weighted analysis targeted individuals with similar probabilities of belonging to different cannabis use trajectories (ie, the population in clinical equipoise).⁵⁰ Confounders were assessed for multicollinearity by ensuring a variance inflation factor lower than 10 in the exposure model. Covariate balance after weighting was assessed using standardized mean differences (SMDs), with values lower than 0.1 indicating adequate balance.⁵⁰ Inverse-probability weights were applied in all estimations to account for differences between the study sample and the complete cohort based on sex and household socioeconomic status.

Considering sex-dependent differences in cannabis use responses,⁵¹ as well as the complex associations among different types of substance use during adolescence—in which evidence suggests cannabis use sometimes precedes tobacco or alcohol initiation—⁵² we examined whether sex or adolescent tobacco or alcohol use modified the associations between patterns of adolescent cannabis use and medical care utilization for any mental disorder or any physical health condition using interaction terms.

All analyses were conducted from November 2023 to February 2025 in Stata 18.0 (Stata Corp LLC), with the Stata code used for the analysis available online.⁵³ Significance thresholds were set at P < .05, and all tests were 2-tailed.

Results

Among 1591 participants (818 female [51.4%], 773 male [48.6%]) followed from birth to 23 years of age, the mean (SD) age at first cannabis exposure assessment was 12.1 (0.3) years, with 348 participants (21.9%) born to mothers younger than 24 years, and 298 participants (18.7%) living in step- or single-parent households during early childhood (Table 2).

Table 2.Characteristics of the Participants Included in the Study

Characteristic	Participants, No. (%)
	Trajectories of cannabis use during adolescence
	No use	Late-onset use	Early-onset and frequent use	Total	P value^a
Total	948 (59.6)	318 (20.0)	325 (20.4)	1591 (100)	NA
Sex
Female	467 (49.3)	164 (51.6)	187 (57.5)	818 (51.4)	.04
Male	481 (50.7)	154 (48.4)	138 (42.5)	773 (48.6)	.04
Internalizing behaviors, mean (SD)^b	2.4 (1.5)	2.1 (1.3)	2.1 (1.2)	2.3 (1.4)	.001
Depressive symptoms, mean (SD)^b	1.4 (1.4)	1.4 (1.5)	1.4 (1.4)	1.4 (1.4)	.52
Externalizing behaviors, mean (SD)^b	1.7 (1.8)	1.8 (1.8)	1.9 (1.7)	1.8 (1.8)	.10
Deviant behaviors, mean (SD)^b	0.7 (1.0)	0.9 (1.2)	1.0 (1.1)	0.8 (1.0)	<.001
Verbal skills, mean (SD)^c	98.5 (13.4)	101.6 (11.2)	100.2 (10.2)	99.5 (12.4)	<.001
Academic performance, mean (SD)^d	3.1 (1.4)	2.8 (1.2)	3.2 (1.3)	3.0 (1.3)	.001
Social skills, mean (SD)^b	7.7 (1.4)	7.8 (1.3)	7.5 (1.2)	7.7 (1.4)	.02
Victimization, mean (SD) ^b	3.3 (1.8)	3.5 (1.9)	3.8 (1.8)	3.4 (1.9)	<.001
Any mental and behavioral disorder^e^,^f	180 (19.0)	40 (12.6)	55 (16.9)	275 (17.3)	.03
Any physical condition^g	460 (48.5)	155 (48.7)	140 (43.1)	755 (47.5)	.24
Lifetime (adolescence) tobacco use	176 (23.6)	208 (65.4)	297 (95.8)	681 (49.6)	<.001
Lifetime alcohol use	645 (79.6)	314 (98.7)	318 (98.8)	1277 (88.1)	<.001
Cannabis use during the past 12 mo, mean (SD)^h	0.0 (0.0)	2.1 (1.3)	3.3 (1.9)	1.2 (1.7)	<.001
Maternal age, y
<24	180 (19.0)	67 (21.1)	101 (31.1)	348 (21.9)	<.001
25-29	324 (34.2)	90 (28.3)	91 (28.0)	505 (31.7)
30-34	314 (33.1)	108 (34.0)	99 (30.5)	521 (32.7)
≥35	130 (13.7)	53 (16.7)	34 (10.5)	217 (13.6)
Canadian-born mother	586 (62.1)	213 (67.4)	234 (72.9)	1033 (65.3)	.001
Canadian-born father	565 (64.9)	214 (72.1)	206 (70.5)	985 (67.5)	.03
Step- or single-parent family	175 (18.5)	46 (14.5)	75 (23.1)	296 (18.7)	.02
Household socioeconomic status, mean (SD), z score	−0.0 (0.9)	0.1 (0.9)	−0.3 (0.9)	−0.0 (0.9)	<.001
Neighborhood conflictⁱ	1.8 (0.6)	1.8 (0.6)	1.8 (0.6)	1.8 (0.6)	.51
Family functioning^b	1.7 (1.4)	1.6 (1.4)	1.8 (1.5)	1.7 (1.5)	.23
Positive parenting practices, mean (SD)^b	6.3 (0.8)	6.1 (0.8)	6.3 (0.8)	6.2 (0.8)	.02
Coercive parenting practices, mean (SD)^b	2.6 (0.9)	2.6 (0.8)	2.7 (0.9)	2.6 (0.9)	.19
Consequent parenting practices, mean (SD)^b	7.1 (1.2)	7.1 (1.2)	6.9 (1.2)	7.0 (1.2)	.03
Parental monitoring at age 10, mean (SD)^b	8.9 (1.0)	9.0 (0.8)	8.8 (0.9)	8.9 (1.0)	.04
Maternal depression, mean (SD)^b	1.4 (1.1)	1.3 (1.1)	1.4 (1.0)	1.4 (1.1)	.14
Maternal prenatal tobacco use	204 (21.6)	67 (21.2)	118 (36.5)	389 (24.6)	<.001
Maternal prenatal alcohol use, mean (SD)^j	0.5 (0.8)	0.5 (0.8)	0.6 (0.8)	0.5 (0.8)	.48
Maternal prenatal drug use^k	NA	NA	NA	NA	.08
Maternal tobacco use
Not really	727 (76.8)	239 (75.2)	195 (60.0)	1161 (73.0)	<.001
Used occasionally	33 (3.5)	9 (2.8)	14 (4.3)	56 (3.5)
Used everyday	187 (19.7)	70 (22.0)	116 (35.7)	373 (23.5)
Maternal alcohol use mean (SD)^j	1.6 (1.5)	1.8 (1.6)	2.1 (1.6)	1.7 (1.6)	<.001
Maternal drug use	34 (3.6)	12 (3.8)	19 (5.8)	65 (4.1)	.20
Paternal tobacco use
Not really	627 (71.6)	216 (71.1)	161 (54.4)	1004 (68.0)	<.001
Used occasionally	24 (2.7)	13 (4.3)	10 (3.4)	47 (3.2)
Used everyday	225 (25.7)	75 (24.7)	125 (42.2)	425 (28.8)
Paternal alcohol use, mean (SD) ^j	3.1 (1.9)	3.6 (1.9)	3.6 (1.9)	3.3 (1.9)	<.001
Paternal drug use	56 (6.4)	23 (7.6)	39 (13.2)	118 (8.0)	.001

Patterns of Adolescent Cannabis Use

We identified 3 patterns of adolescent cannabis use using GBTM with a count distribution model (Figure 1). The estimated parameters for each trajectory are presented in eTable 3 in Supplement 1. The identified groups were (1) individuals who never used cannabis during adolescence (nonuse, 948 [59.6%]); (2) individuals who started using cannabis after 15 years of age and by age 17 years used cannabis less than once a month (late-onset use, 318 [20.0%]); and (3) individuals who began using cannabis before 15 years of age and used cannabis at least once per month (early-onset and frequent use, 325 [20.4%]). The 3-group solution had a probability correct model of 0.99. All groups identified by GBTM had a mean posterior probability of at least 88.1%, and odds of correct classification equal or higher than 5.0, and an entropy of 0.719, indicating good model fit.⁴⁶ After rerunning the model with different sets of starting values, the model consistently converged and led to the same maximum likelihood estimations 91.0% of the time.

Figure 1.Patterns of Cannabis Use During AdolescenceOrange dashed lines indicate 95% confidence intervals of the respective trajectory. Frequency codes for cannabis use during the past 12 months were categorized as 0, “I didn’t use”; 1, “just once”; 2, “less than once a month or occasionally”; 3,”about once a month”; 4, “during the weekend or 1 to 2 times a week”; 5, “more than 3 times a week”; and 6, “every day.” GBTM indicates group-based trajectory modeling.

Child and Family Characteristics Across Patterns of Adolescent Cannabis Use

Individuals with early-onset and frequent cannabis use displayed more deviant behaviors (mean [SD], 1.0 [1.1] vs 0.7 [1.0] for nonuse and 0.9 [1.2] for late-onset use), experienced greater peer victimization (mean [SD], 3.8 (1.8) vs 3.3 [1.8] for nonuse and 3.5 [1.9] for late-onset use), and were notably more likely to have used tobacco themselves (297 [95.8%] vs 176 [23.6%] for nonuse and 208 [65.4%] for late-onset use) (Table 2). In addition, individuals with early-onset and frequent use were more likely to be female (187 [57.5%] compared with 138 [42.5%] male). In contrast, individuals with no use exhibited somewhat higher internalizing behaviors (mean [SD], 2.4 [1.5] vs 2.1 [1.3] for late-onset use and 2.1 [1.2] for early-onset and frequent use). Individuals with early-onset and frequent use had more maternal prenatal tobacco exposure (118 [36.5%] vs 204 [21.6%] for nonuse and 67 [21.2%] for late-onset use), higher daily paternal tobacco use (125 [42.2%] vs 225 [25.7%] for nonuse and 75 [24.7%]) for late-onset use) and drug use (39 [13.2%] vs 56 [6.4%] for nonuse and 23 [7.6%] for late-onset use), and increased parental alcohol use (mean [SD] score, 3.6 [1.9] vs 3.1 [1.9] for nonuse and 3.6 [1.9] for late-onset use) as well as overall lower household socioeconomic status (mean [SD] score, −0.3 [0.9] vs −0.0 [0.9] for nonuse and 0.1 [0.9] for late-onset use).

Patterns of Adolescent Cannabis Use and Medical Care Utilization for Mental and Physical Health Conditions

Primary Outcomes

In unadjusted analyses, individuals with early-onset and frequent use had higher odds of receiving medical care for any mental disorder (unadjusted OR [uOR], 1.65 [95% CI, 1.24-2.19]) and any physical condition (uOR, 2.07 [95% CI, 1.50-2.85]) compared with individuals with no use (Figure 2). After balancing the distribution of confounders using overlap weighting (Figure 3), adjusted estimates were slightly attenuated but remained increased for having any mental disorder (adjusted OR [aOR], 1.51 [95% CI, 1.10-2.08]) and any physical condition (aOR, 1.86 [95% CI, 1.30-2.67]) (Figure 2). For individuals with late-onset use, the aOR for having any mental disorder was 1.13 (95% CI, 0.80-1.58), and for having any physical condition, the aOR was 1.63 (95% CI, 1.16-2.28), indicating an association only for the latter.

Figure 2.Patterns of Cannabis Use During Adolescence and Medical Care Utilization for Mental Health and Physical Health Age 18 to 23 YearsOrange squares represent unadjusted odds ratios (ORs), and blue squares represent adjusted ORs estimated using overlap weighting to address confounder imbalance. Inverse probability weighting was applied to account for sampling differences between the study sample and the complete cohort. Data for medical care use were rounded to the nearest multiple of 5 in accordance with confidentiality requirements of the Institut de la statistique du Québec and compared with publicly available statistics.

Figure 3.Confounder Balance in the Study SampleAbsolute (pairwise) standardized mean differences for comparisons between the groups exposed to cannabis use vs the nonexposed group. Dashed lines denote the commonly accepted threshold of 0.1 for standardized mean differences indicating adequate confounder balance; orange diamonds, unweighted estimates; blue diamonds, overlap-weighted estimates (after applying overlap weights to balance confounders). SES indicates socioeconomic status.

Secondary Outcomes

For specific mental disorders (Figure 2A), individuals with early-onset and frequent use had significantly increased odds of receiving medical care for common mental disorders (uOR, 1.70 [95% CI, 1.26-2.29]; aOR, 1.57 [95% CI, 1.12-2.21]), but not for substance-related disorders (uOR, 1.95 [95% CI, 1.20-3.17]; aOR, 1.71 [95% CI, 0.99-2.94]) or suicide-related behaviors (uOR, 1.82 [95% CI, 1.04-3.20]; aOR, 1.56 [95% CI, 0.81-2.97]). For specific physical conditions (Figure 2B), individuals with early-onset and frequent use showed higher unadjusted odds of having respiratory disease (uOR, 1.43 [95% CI, 1.07-1.92]), injury and poisoning (uOR, 1.63 [95% CI, 1.25-2.12]), and other physical disease (uOR, 1.65 [95% CI, 1.26-2.16]). After overlap weighting, odds remained significantly increased for injury and poisoning (aOR, 1.41 [95% CI, 1.05-1.89]) and other physical disease (aOR, 1.47 [95% CI, 1.08-1.98]) but not for respiratory disease (aOR, 1.26 [95% CI, 0.91-1.75]). Among individuals with late-onset use, adjusted odds for having injury and poisoning (aOR, 1.21 [95% CI, 0.89-1.63]) and other physical disease (aOR, 1.32 [95% CI, 0.98-1.77]) were not statistically significant.

Results from the nonimputed sample were consistent across key outcomes (eTable 4, eFigure 2 in Supplement 1). No statistically significant interactions by sex or adolescent tobacco or alcohol use and patterns of adolescent cannabis use were found (eAppendix 2 and eTable 5 in Supplement 1).

Discussion

In this longitudinal birth cohort study linking adolescent self-reports of cannabis use to objective medical care databases from Québec, Canada, we found that adolescents who initiated cannabis use early and continued consistently accessed more medical care for both mental and physical conditions in young adulthood, compared with their peers who did not use cannabis. These associations persisted even after rigorous adjustment for a wide range of preexposure confounding factors, including early individual, familial, and community-based vulnerabilities.

Our findings extend previous research showing that early-onset and frequent cannabis use was associated with higher odds of conditions such as common mental disorders^{12,13,14,15,16,17} and substance-related disorders^{12,13,14,15,16,18,19,20,21,22,23,24} in adulthood. Beyond mental health, participants with early-onset and frequent cannabis use showed increased odds of medical care use for physical health conditions, notably injuries and poisonings. These increased risks may reflect neurocognitive and behavioral changes associated with early-onset and frequent cannabis exposure,^{54,55,56,57,58} including acute intoxication symptoms and withdrawal-related responses, as well as the broader availability of potent cannabis products that can lead to accidental overdose.⁵⁶ Although we did not observe consistent associations with chronic respiratory diseases, it is possible that these outcomes will emerge with longer follow-up, given that many physical diseases have longer latency periods.^6,59 Lifestyle factors associated with early cannabis use, such as poor diet and insufficient sleep,^16,25 may compound these risks over time. These findings underscore the importance of ongoing, systematic monitoring of youth cannabis use and associated health outcomes, especially given the increased accessibility, diversity, and potency of cannabis products that continue to expand in many jurisdictions in North America and internationally.⁶⁰

Unlike most prior investigations relying on self-reported health outcomes,^{13,15,18,21,22,23,24} our study leveraged data from comprehensive medical registries, thereby capturing clinically significant conditions that required medical intervention. While previous studies suggested that both early and late adolescent cannabis use could lead to poor health outcomes,^{12,13,14,15,16,18,20,21,22,23,24} we found that individuals with late-onset use did not differ significantly from those with nonuse in receipt of mental health–related medical care, yet they did exhibit higher odds of physical health conditions (including injuries and poisoning). Even after accounting for a wide range of early-life and family-level confounders, early-onset and frequent use remained associated with later medical care utilization, raising the possibility that cannabis exposure during a period of rapid development may be associated with adverse outcomes beyond what can be fully explained by preexisting vulnerabilities. This finding highlights the critical role of early and frequent cannabis use in shaping long-term health trajectories, suggesting that efforts to delay initiation or reduce frequency of early use may help mitigate potential long-term harms.^61,62,63,64 However, the increased risk of physical conditions observed even among individuals with late-onset use reinforces the importance of monitoring this group to better characterize their long-term health care needs.

The absence of significant associations in some of our adjusted models, such as those for suicide-related behaviors, may partly reflect limited statistical power and relatively low base rates of such events. However, it is also plausible that cannabis use initiated in later adolescence has weaker or no association with certain outcomes, consistent with developmental literature on early exposure risk.^3,4,12,15 Nonetheless, the patterns observed warrant continued monitoring, given the potential severity of these outcomes. Moreover, no significant interactions by sex or by co-occurring adolescent tobacco or alcohol use were found, indicating that the adverse consequences of early-onset and frequent cannabis use on medical care utilization are broadly similar across these subgroups. Longer-term studies are needed to investigate whether this pattern is sustained or evolves into adulthood. Future studies should also investigate potential mechanisms explaining these associations, including assessing the role of continuing cannabis use in the postadolescent years.

Finally, we found that individuals with early cannabis use differed substantially from those with late use and those with no cannabis use in their exposure to early-life risk factors, including parental substance use,^22,24 which is known to influence mental health and substance use trajectories in offspring.²² Such findings underscore the importance of also addressing underlying vulnerabilities in prevention strategies and policy formation. These early-life determinants are not only confounders but also upstream factors associated with adolescent cannabis exposure.^22,24 Population-level interventions that support families with young children may represent a viable path toward reducing cannabis-associated harms across development.^62,63

Limitations

Several limitations should be noted. First, despite rigorous adjustments for various confounding factors, residual confounding remained possible given unmeasured genetic predispositions, which may have influenced both cannabis use and health outcomes. Second, selective attrition was more prevalent among certain subgroups, particularly male participants and participants from lower socioeconomic backgrounds, which may have influenced the distribution of confounders in ways that partly explained the unexpected childhood profiles (eg, no cannabis use showing relatively higher internalizing behaviors. Third, although we used medical records to improve the validity of participant outcome data, some diagnostic codes (particularly for physical injuries and poisoning) were not routinely or consistently used, potentially leading to an underestimation of the true magnitude of associations. Fourth, our data did not capture subclinical symptoms or functional impairments that did not result in medical visits, potentially underestimating the broader impact of adolescent cannabis use on daily functioning and health trajectories. Fifth, adolescents with earlier or more frequent cannabis use may avoid medical care services due to stigma or other risk factors, thus reducing detection of relevant conditions. Sixth, self-reported measures of cannabis use may be subject to underreporting or recall bias. In addition, the generalizability of these findings may be constrained by the cohort’s geographic and temporal context—participants were adolescents prior to the 2018 legalization of nonmedical cannabis in Canada—and their patterns of use and associated risk may not reflect those observed in postlegalization cohorts. Moreover, increasing product potency, changing usage patterns, and differences in accessibility of medical services may further limit applicability to current populations.

Conclusions

In this population-based birth cohort study, early-onset and frequent cannabis use during adolescence was associated with increased medical care utilization for both mental and physical health conditions in young adulthood. Participants with late-onset cannabis use also showed increased utilization of medical care for specific physical conditions, highlighting risks across onset patterns. By integrating repeated adolescent self-reports with objective administrative medical records, and accounting for early-life confounders, this study offers insight into cannabis-associated medical care burdens in a universal health care setting. As access to high-potency cannabis expands, efforts to delay initiation and reduce frequency of adolescent use should remain a public health priority. Identifying adolescents with early risk profiles and attending to psychosocial and familial factors may help inform strategies to address future health care needs.

Untitled section

References

Untitled section

References

1.Steinfeld MR, Torregrossa MM. Consequences of adolescent drug use. Transl Psychiatry. 2023;13(1):313. doi: 10.1038/s41398-023-02590-4
2.Summary E. World Drug Report 2023. United Nations Office on Drugs and Crime UN. Accessed December 2024. https://www.unodc.org/res/WDR-2023/WDR23_Exsum_fin_DP.pdf
3.Silins E, Horwood LJ, Patton GC, et al. ; Cannabis Cohorts Research Consortium . Young adult sequelae of adolescent cannabis use: an integrative analysis. Lancet Psychiatry. 2014;1(4):286-293. doi: 10.1016/S2215-0366(14)70307-4
4.Gobbi G, Atkin T, Zytynski T, et al. Association of cannabis use in adolescence and risk of depression, anxiety, and suicidality in young adulthood: a systematic review and meta-analysis. JAMA Psychiatry. 2019;76(4):426-434. doi: 10.1001/jamapsychiatry.2018.4500
5.Matheson SL, Laurie M, Laurens KR. Substance use and psychotic-like experiences in young people: a systematic review and meta-analysis. Psychol Med. 2023;53(2):305-319.
6.Tuvel AL, Winiger EA, Ross JM. A review of the effects of adolescent cannabis use on physical health. Child Adolesc Psychiatr Clin N Am. 2023;32(1):85-105. doi: 10.1016/j.chc.2022.07.005
7.Alcohol and drug use among students in Canada, 2023-24: key findings from the Canadian Student Alcohol and Drugs Survey. Government of Canada. 2025. Accessed December 2024. https://www.canada.ca/content/dam/hc-sc/documents/services/canadian-student-tobacco-alcohol-drugs-survey/2023-2024-key-findings/2023-2024-key-findings.pdf
8.Wilson J, Freeman TP, Mackie CJ. Effects of increasing cannabis potency on adolescent health. Lancet Child Adolesc Health. 2019;3(2):121-128. doi: 10.1016/S2352-4642(18)30342-0
9.Myran DT, Imtiaz S, Konikoff L, Douglas L, Elton-Marshall T. Changes in health harms due to cannabis following legalisation of non-medical cannabis in Canada in context of cannabis commercialisation: a scoping review. Drug Alcohol Rev. 2023;42(2):277-298. doi: 10.1111/dar.13546
10.Fischer B, Jutras-Aswad D, Robinson T. How has non-medical cannabis legalization served the health and welfare of under-age (adolescent) youth in Canada? Lancet Reg Health Am. 2024;35:100773. Accessed June 12, 2024 doi: 10.1016/j.lana.2024.100773
11.Scheier LM, Griffin KW. Youth marijuana use: a review of causes and consequences. Curr Opin Psychol. 2021;38:11-18. doi: 10.1016/j.copsyc.2020.06.007
12.Juon HS, Fothergill KE, Green KM, Doherty EE, Ensminger ME. Antecedents and consequences of marijuana use trajectories over the life course in an African American population. Drug Alcohol Depend. 2011;118(2-3):216-223. doi: 10.1016/j.drugalcdep.2011.03.027
13.Epstein M, Hill KG, Nevell AM, et al. Trajectories of marijuana use from adolescence into adulthood: environmental and individual correlates. Dev Psychol. 2015;51(11):1650-1663. doi: 10.1037/dev0000054
14.Lynne-Landsman SD, Bradshaw CP, Ialongo NS. Testing a developmental cascade model of adolescent substance use trajectories and young adult adjustment. Dev Psychopathol. 2010;22(4):933-948. doi: 10.1017/S0954579410000556
15.Thompson K, Merrin G, Ames M, Leadbeater B. Marijuana trajectories in Canadian youth: associations with substance use and mental health. Can J Behav Sci. 2018;50(1):17-28. doi: 10.1037/cbs0000090
16.Tucker JS, Ellickson PL, Orlando M, Martino SC, Klein DJ. Substance use trajectories from early adolescence to emerging adulthood: a comparison of smoking, binge drinking, and marijuana use. J Drug Issues. 2005;35(2):307-332. doi: 10.1177/002204260503500205
17.Brook JS, Lee JY, Brown EN, Finch SJ, Brook DW. Developmental trajectories of marijuana use from adolescence to adulthood: personality and social role outcomes. Psychol Rep. 2011;108(2):339-357. doi: 10.2466/10.18.PR0.108.2.339-357
18.Scholes-Balog KE, Hemphill SA, Evans-Whipp TJ, Toumbourou JW, Patton GC. Developmental trajectories of adolescent cannabis use and their relationship to young adult social and behavioural adjustment: a longitudinal study of Australian youth. Addict Behav. 2016;53:11-18. doi: 10.1016/j.addbeh.2015.09.008
19.Windle M, Wiesner M. Trajectories of marijuana use from adolescence to young adulthood: predictors and outcomes. Dev Psychopathol. 2004;16(4):1007-1027. doi: 10.1017/S0954579404040118
20.Flory K, Lynam D, Milich R, Leukefeld C, Clayton R. Early adolescent through young adult alcohol and marijuana use trajectories: early predictors, young adult outcomes, and predictive utility. Dev Psychopathol. 2004;16(1):193-213. doi: 10.1017/S0954579404044475
21.Brook JS, Zhang C, Leukefeld CG, Brook DW. Marijuana use from adolescence to adulthood: developmental trajectories and their outcomes. Soc Psychiatry Psychiatr Epidemiol. 2016;51(10):1405-1415. doi: 10.1007/s00127-016-1229-0
22.Boden JM, Dhakal B, Foulds JA, Horwood LJ. Life-course trajectories of cannabis use: a latent class analysis of a New Zealand birth cohort. Addiction. 2020;115(2):279-290. doi: 10.1111/add.14814
23.Brown TL, Flory K, Lynam DR, Leukefeld C, Clayton RR. Comparing the developmental trajectories of marijuana use of African American and Caucasian adolescents: patterns, antecedents, and consequences. Exp Clin Psychopharmacol. 2004;12(1):47-56. doi: 10.1037/1064-1297.12.1.47
24.Taylor M, Collin SM, Munafò MR, MacLeod J, Hickman M, Heron J. Patterns of cannabis use during adolescence and their association with harmful substance use behaviour: findings from a UK birth cohort. J Epidemiol Community Health. 2017;71(8):764-770. doi: 10.1136/jech-2016-208503
25.Ames ME, Leadbeater BJ, Merrin GJ, Thompson K. Patterns of marijuana use and physical health indicators among Canadian youth. Int J Psychol. 2020;55(1):1-12. doi: 10.1002/ijop.12549
26.Orri M, Boivin M, Chen C, et al. Cohort profile: Quebec Longitudinal Study of Child Development (QLSCD). Soc Psychiatry Psychiatr Epidemiol. 2021;56(5):883-894. doi: 10.1007/s00127-020-01972-z
27.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147(8):573-577. doi: 10.7326/0003-4819-147-8-200710160-00010
28.Willms J, Shields M. A measure of socioeconomic status for the National Longitudinal Study of Children. Statistics Canada. 1996.
29.McGuire JB. The reliability and validity of a questionnaire describing neighborhood characteristics relevant to families and young children living in urban areas. J Community Psychol. 1997;25(6):551-566. doi: 10.1002/(SICI)1520-6629(199711)25:6<551::AID-JCOP5>3.0.CO;2-S
30.Growing up in Canada: issued by Human Resources Development Canada. Ottawa: Statistics Canada. 1996. Accessed September 8, 2025. https://publications.gc.ca/site/eng/65048/publication.html
31.Helzer JE, Robins LN. The diagnostic interview schedule: its development, evolution, and use. Soc Psychiatry Psychiatr Epidemiol. 1988;23(1):6-16. doi: 10.1007/BF01788437
32.Radloff LS. The CES-D Scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1(3):385-401. doi: 10.1177/014662167700100306
33.Cadman D, Rosenbaum P, Boyle M, Offord DR. Children with chronic illness: family and parent demographic characteristics and psychosocial adjustment. Pediatrics. 1991;87(6):884-889. doi: 10.1542/peds.87.6.884
34.Strayhorn JM, Weidman CS. A Parent Practices Scale and its relation to parent and child mental health. J Am Acad Child Adolesc Psychiatry. 1988;27(5):613-618. doi: 10.1097/00004583-198809000-00016
35.Stattin H, Kerr M. Parental monitoring: a reinterpretation. Child Dev. 2000;71(4):1072-1085. doi: 10.1111/1467-8624.00210
36.Dunn LM. Peabody Picture Vocabulary Test-Revised. American Guidance Service; 1981.
37.Tremblay RE, Pihl RO, Vitaro F, Dobkin PL. Predicting early onset of male antisocial behavior from preschool behavior. Arch Gen Psychiatry. 1994;51(9):732-739. doi: 10.1001/archpsyc.1994.03950090064009
38.Collet OA, Orri M, Tremblay RE, Boivin M, Côté SM. Psychometric properties of the Social Behavior Questionnaire (SBQ) in a longitudinal population-based sample. Int J Behav Dev. 2023;47(2):180-189. doi: 10.1177/01650254221113472
39.Janus M, Offord DR. Development and psychometric properties of the Early Development Instrument (EDI): a measure of children’s school readiness. Can J Behav Sci. 2007;39(1):1-22. doi: 10.1037/cjbs2007001
40.Ladd GW, Kochenderfer-Ladd B. Identifying victims of peer aggression from early to middle childhood: analysis of cross-informant data for concordance, estimation of relational adjustment, prevalence of victimization, and characteristics of identified victims. Psychol Assess. 2002;14(1):74-96. doi: 10.1037/1040-3590.14.1.74
41.National Longitudinal Survey of Children and Youth . Survey instruments: overview of survey instruments for 1998-1999 data collection: cycle 3. Accessed September 17, 2025. https://www150.statcan.gc.ca/n1/en/pub/89f0078x/89f0078x1999003-eng.pdf?st=mdOLYm75
42.Kovacs M. The Children’s Depression, Inventory (CDI). Psychopharmacol Bull. 1985;21(4):995-998.
43.National Longitudinal Survey of Children and Youth . Survey overview for 2004/2005 data collection cycle 6. Statistics Canada. Accessed September 8, 2025. https://www23.statcan.gc.ca/imdb-bmdi/pub/document/4450_D2_T9_V2-eng.pdf
44.Boivin M, Brendgen M, Dionne G, et al. The Quebec Newborn Twin Study at 21. Twin Res Hum Genet. 2019;22(6):475-481. doi: 10.1017/thg.2019.74
45.Desrosiers H. Early childhood characteristics and school readiness: the importance of a family’s social support. Institut de la Statistique du Québec. 2013;Nov(18):1-16. https://statistique.quebec.ca/en/fichier/no-18-early-childhood-characteristics-and-school-readiness-the-importance-of-social-support-for-families.pdf
46.Nagin D. Group-Based Modeling of Development. Harvard University Press; 2005. doi: 10.4159/9780674041318
47.Laursen BP, Hoff E. Person-centered and variable-centered approaches to longitudinal data. Merrill Palmer Q (Wayne State Univ Press). 2006;52(3):377-389. doi: 10.1353/mpq.2006.0029
48.van de Schoot R, Sijbrandij M, Winter SD, Depaoli S, Vermunt JK. The GRoLTS-checklist: guidelines for reporting on latent trajectory studies. Structural Equation Modeling: A Multidisciplinary Journal. 2017;24(3):451-467. doi: 10.1080/10705511.2016.1247646
49.Lee KJ, Carlin JB. Multiple imputation for missing data: fully conditional specification versus multivariate normal imputation. Am J Epidemiol. 2010;171(5):624-632. doi: 10.1093/aje/kwp425
50.Li F, Thomas LE, Li F. Addressing extreme propensity scores via the overlap weights. Am J Epidemiol. 2019;188(1):250-257.
51.Cooper ZD, Craft RM. Sex-dependent effects of cannabis and cannabinoids: a translational perspective. Neuropsychopharmacology. 2018;43(1):34-51. doi: 10.1038/npp.2017.140
52.Agrawal A, Budney AJ, Lynskey MT. The co-occurring use and misuse of cannabis and tobacco: a review. Addiction. 2012;107(7):1221-1233. doi: 10.1111/j.1360-0443.2012.03837.x
53.OrriLab/Health-use-cannabis-QLSCD . GitHub. Accessed September 17, 2025. https://github.com/OrriLab/Health-use-cannabis-QLSCD
54.Patel J, Marwaha R. Cannabis Use Disorder. 2025. Accessed January 25, 2025. https://www.ncbi.nlm.nih.gov/books/NBK538131/
55.Volkow ND, Baler RD, Compton WM, Weiss SRB. Adverse health effects of marijuana use. N Engl J Med. 2014;370(23):2219-2227. doi: 10.1056/NEJMra1402309
56.National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Board on Population Health and Public Health Practice; Committee on the Health Effects of Marijuana: An Evidence Review and Research Agenda. Injury and death. In: The Health Effects of Cannabis and Cannabinoids: The Current State of Evidence and Recommendations for Research. National Academies Press (US); 2017. Accessed December 6, 2024. https://www.ncbi.nlm.nih.gov/sites/books/NBK425742/
57.Koivisto MK, Puljula J, Levola JM, et al. Adolescent alcohol and cannabis use as risk factors for head trauma in the Northern Finland Birth Cohort study 1986. Eur J Public Health. 2023;33(6):1115-1121. doi: 10.1093/eurpub/ckad151
58.Dahlgren MK, Sagar KA, Smith RT, Lambros AM, Kuppe MK, Gruber SA. Recreational cannabis use impairs driving performance in the absence of acute intoxication. Drug Alcohol Depend. 2020;208:107771. doi: 10.1016/j.drugalcdep.2019.107771
59.National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Board on Population Health and Public Health Practice; Committee on the Health Effects of Marijuana: An Evidence Review and Research Agenda. Respiratory disease. In: The Health Effects of Cannabis and Cannabinoids: The Current State of Evidence and Recommendations for Research. National Academies Press (US); 2017. Accessed January 25, 2025. https://www.ncbi.nlm.nih.gov/books/NBK425753/
60.Murray RM, Hall W. Will legalization and commercialization of cannabis use increase the incidence and prevalence of psychosis? JAMA Psychiatry. 2020;77(8):777-778. doi: 10.1001/jamapsychiatry.2020.0339
61.Georgie J M, Sean H, Deborah M C, Matthew H, Rona C. Peer-led interventions to prevent tobacco, alcohol and/or drug use among young people aged 11-21 years: a systematic review and meta-analysis. Addiction. 2016;111(3):391-407. doi: 10.1111/add.13224
62.Kuntsche S, Kuntsche E. Parent-based interventions for preventing or reducing adolescent substance use - a systematic literature review. Clin Psychol Rev. 2016;45:89-101. doi: 10.1016/j.cpr.2016.02.004
63.Vermeulen-Smit E, Verdurmen JEE, Engels RCME. The effectiveness of family interventions in preventing adolescent illicit drug use: a systematic review and meta-analysis of randomized controlled trials. Clin Child Fam Psychol Rev. 2015;18(3):218-239. doi: 10.1007/s10567-015-0185-7
64.Lemstra M, Bennett N, Nannapaneni U, et al. A systematic review of school-based marijuana and alcohol prevention programs targeting adolescents aged 10–15. Addict Res Theory. 2010;18(1):84-96. doi: 10.3109/16066350802673224

Studies

Conditions

Cannabinoids and active compounds

Related research

Abstract

Key Points

Question

Findings

Meaning

Abstract

Importance

Objective

Design, Setting, and Participants

Exposure

Main Outcomes and Measures

Results

Conclusions and Relevance

Article notes

Untitled section

Author notes

Introduction

Methods

Participants

Exposure

Outcomes

Covariates

Statistical Analysis

Results

Patterns of Adolescent Cannabis Use

Child and Family Characteristics Across Patterns of Adolescent Cannabis Use

Patterns of Adolescent Cannabis Use and Medical Care Utilization for Mental and Physical Health Conditions

Primary Outcomes

Secondary Outcomes

Discussion

Limitations

Conclusions

Untitled section

References

Untitled section

References

Associated Data

Supplementary Materials

Additional materials