The participants came from two studies, including 686 adolescents or young adults between 12 and 21 years old, who received care at the Children’s Health Center at Zuckerberg San Francisco General Hospital Center . The study was conducted from 2013 to 2014, prior to the legalization of recreational marijuana in California. Adolescents attending the CHC in 2013 included 62.4% Latino, 15.9% non-Hispanic Black , 11.0% Asian, 5.1% White and 5.5% other race/ethnicity children. In the first study, adolescents who had surplus urine during a clinic visit for both sick and well care were included. Urine samples were collected for clinical indications, including urinary tract or sexually-transmitted infection screening and diagnosis, abdominal pain evaluation, trauma and pregnancy screening. There was no direct patient contact, and after chart review, all patient identifiers were deleted from the database and research charts. This study was conducted anonymously and participants were not consented. In the second study, adolescents participated in a consented study in which they were asked to provide urine samples and complete a questionnaire regarding tobacco use, secondhand smoke exposure, and the use of marijuana products, including blunts, and alcohol. Participants were told that their study data would not be shared with their doctors or other health care providers. Each question had an option not to answer. Information on race/ethnicity, sex,plant bench indoor age and self-reported tobacco use history was collected on all participants. These studies were approved by the Institutional Review Board at the University of California San Francisco.
As described in previous publications in this study population, we categorized urine cotinine levels in urine as follows: Active smoking for values >30 ng/ml, significant SHS exposure as 0.25 to 30 ng/ml, and light SHS and/or thirdhand smoke exposure as 0.05 to 0.25 ng/ml . Including covariate categories of race/ ethnicity, sex and age , we present a descriptive analysis of THC detection frequencies and median for THC-COOH concentrations. We tested differences in THC-COOH detection frequency with the χ2 test or fisher’s exact test as appropriate. To evaluate whether absolute measured values of THCCOOH differed across covariates we used Kruskal-Wallis tests. Spearman’s correlations were used to describe the correlation of tobacco biomarkers with THC-COOH and this relationship between smoking status and THC-COOH detection was further assessed using logistic regression. Both an unadjusted and an age- and race category-adjusted model of the odds ratio for THC-COOH detection given active smoking are provided with 95% confidence intervals. All statistical analyses were carried out using SAS v. 9.4 . All statistical tests were considered significant at p < 0.05. For the full study group, prevalence of THC exposure based on THC-COOH increased with age, and was highest in Black and Mixed/Other groups, but did not differ by sex . Among those who were positive for THC-COOH, absolute levels of THC-COOH followed a similar pattern, with highest levels in older adolescents, in Blacks and Mixed/Other group, and in males. Based on a urine cotinine cut-point of 30 ng/ml, 66 participants were determined to be active tobacco cigarette smokers. Frequency distributions for total THC-COOH among smokers and non-smokers are shown in Fig 1. The prevalence of THC exposure was much higher among adolescent tobacco cigarette smokers compared to those who did not use tobacco . Among biochemically-determined cigarette smokers, 81.8% were positive for marijuana use by total THC-COOH and 71.2% were positive using the THC screen.
The proportion of THC-positives did not vary significantly by sex or age, but was highest in Black and Mixed/Other groups. The absolute levels of THC-COOH were much higher in cigarette smokers compared to non-smokers, and increased with age . Comparison by race/ethnicity was not possible due to small numbers in some groups. Using urine cotinine, we were also able to examine the relationship between extent of secondhand tobacco smoke exposure to THC use . Based on THCCOOH, the prevalence of THC exposure was 2.8% among those with no tobacco smoke exposure, 8.7% with light SHS exposure, 33.9% with heavy SHS exposure, and as noted above 81.8% in active smokers . As shown in Fig 3, across all levels of tobacco exposure, there was a moderately strong quantitative correlation between tobacco use and THC biomarkers, with r = 0.60, p <0.001 for cotinine and r = 0.55 for NNAL, respectively, vs total THC-COOH . The odds ratio for a urine cotinine level of > 30 ng/ml as a predictor of COOH-THC positivity was 18.9 unadjusted and 13.2 after adjusting for age and race . In a subset of participants we were able to compare self-report of various self-reported behaviors to THC biomarker levels . For those who reported marijuana use in the past three days , 91.7% were positive and 8.3% negative for total THC-COOH. For those who reported marijuana use in the past 3 months , 53.8% were positive and 46.2% negative for THC-COOH. Among all self-reported ever marijuana users , 77% had smoked marijuana in the form of blunts . For those who reported blunt use in the past 30 days , 74.1% were positive THC-COOH. For those who reported alcohol use in the past 3 months , 32.3% were positive; while among those who report alcohol use in the past 3 days , 54.5% were positive for THC-COOH. Our study provides novel information in several areas. We show that the commonly used urine immunoassay screen for THC exposure with a cut-off of 50 ng/ml substantially underestimates actual THC exposure, as measured by a sensitive chromatographic method.
On the other hand, in a subset of adolescents in whom use of marijuana was queried, misreporting was relatively low, with only 4.6% and 1.8% of those who reported no marijuana use in the past 3 days and 3 months, respectively, biochemically positive for THC exposure. Thus, the routine urine THC screen has a high degree of specificity, but only a moderate degree of sensitivity. The THC screen presumably has lower sensitivity because many urines had THC-COOH values between 3 and 50 ng/ml. It should be noted that other immunoassays might have lower cut-off values and have a higher levels of sensitivity. We found that 25% of adolescents had biochemical evidence of marijuana use, which is slightly higher than the 15 to 17% self-reported prevalence of past month use reported in the Monitoring the Future study and in a recent survey in Washington State , but similar to the 24% biochemically-assessed exposure in adolescents attending a hospital clinics by Silber et al in Washington D.C. in 1987 . Consistent with other studies,greenhouse rolling racks we found an increased prevalence of marijuana use with increasing age and in Blacks, compared with other racial/ethnic groups. Black adolescents in particular have a much higher use of use of blunts compared to other groups, which may account for the higher prevalence of THC exposure . We found an extremely high biochemically-determined prevalence of THC use in active cigarettes smokers. A strong association between tobacco and marijuana use was expected based on other studies of self-reported behaviors, but the strength of the association was remarkable. The odds ratio of active cigarette smoking predicting THC use was 13.2 . Thus, cigarette smoking was a strong surrogate marker for marijuana use. Whether this remains the case as the prevalence of cigarette smoking declines in youth in the future and the extent to which the use of electronic cigarettes in youth are predictive of marijuana use remain to be determined. Another novel finding was the association between the level of SHS exposure, determined by urine cotinine, and biochemically-determined prevalence of THC use, with a progressive increase in prevalence from no exposure to light, and to heavy SHS exposure, and to active smoking. A limitation of our study is that we cannot tell if the source of nicotine exposure is cigarette smoking or use of electronic cigarettes. Additionally, we cannot determine if cotinine levels consistent with high levels of SHS exposure actually represent low-level nicotine exposure from SHS or intermittent non-daily smoking vaping, or whether the low levels of cotinine reflect the use of blunts. One laboratory study reported no detectable plasma nicotine after three puffs of a blunt, suggesting that cotinine levels attributable to blunt use would be low .
Across all categories of tobacco exposure as indicated by either urine cotinine or NNAL, there was a moderately strong quantitative correlation between tobacco biomarkers and level of THC exposure. The pharmacologic bases and health implications of this association remain to be explored.As expected, a history of use of blunts in the past 30 days was associated with a high prevalence of biochemically-determined marijuana use . A history of alcohol use in the past 3 days was also associated with a high prevalence of marijuana use . The concordance of marijuana and alcohol use has been well described in other studies . Limitations of our study include the use of a convenience sample of adolescents seeking medical care in an urban public hospital outpatient clinics in one city. Our study was conducted before recreational use of marijuana was legalized in California. The prevalence of marijuana use might be even higher now that recreational marijuana is legal. Further, we did not collect data on level of dependence on tobacco or marijuana. One of the more controversial questions in drug policy today is whether the trend toward legalizing marijuana for medicinal and adult recreational use could increase illicit marijuana use among young people . Since 1996, 28 U.S. states and the District of Columbia have legalized the production and sale of marijuana for medicinal use , and eight have legalized marijuana for adult recreational use. Medical marijuana laws could potentially increase the availability of marijuana and reduce perceptions of its harmfulness, leading more young people to try it. State medical marijuana laws include regulations that protect young people from illegally obtaining marijuana . But if these restrictions are not carefully enforced, young people could gain increased access to marijuana through the diversion of medical marijuana into illegal markets, which could also lower its price . Marijuana use by young people is associated with lasting detrimental changes in cognitive functioning of the developing brain, and poor educational and occupational outcomes . Use increases the risk of unintentional injuries and auto fatalities, mood and psychotic disorders, and drug dependence, especially when use is initiated at a young age . Long term marijuana smokers have a disproportionate burden of upper respiratory illnesses, including chronic bronchitis and some cancers, and an increased risk of cardiovascular disease . Medical marijuana producers and retailers are promoting new, more potent products such as oils often used as inhalants with tetrahydrocannabinol concentrations ranging from 40 to 70 percent . They are also developing new products that appeal to youth, such as packaged edibles and candies, that may increase the hazard of overdose due to their relatively slow rates of absorption and perceived intoxication . These products could increase the risk of overdose in young people, who tend to be less experienced users with low tolerance levels. Studies have reported little evidence that medical marijuana laws increase marijuana use among young people , although well controlled studies consistently report increased consumption in adults . Using the Youth Risk Behavior Survey, researchers have compared high school students’ consumption before and after medical marijuana laws were enacted, finding no evidence of rising consumption on a national basis . A national study of 12–17 year old found that medical marijuana laws had no causal impact on consumption , as did a carefully controlled national study of 13–18 year olds . Wen et al. reported a five percent increase in the likelihood of trying marijuana among 12–20 year olds who dwell in states with medical marijuana laws, but the study was limited by the need to pool such a broad range of ages. Prior studies have ignored or been unable to detect age related variation in the impact of medical marijuana laws by pooling children aged 12–17, or even 12–24, or by studying particular age groups in isolation. Age-related variation is important to capture because young peoples’ access to marijuana and their developmental susceptibility to drug related harms differs by age .