A Longitudinal Study of Cannabis Use Increasing the Use of Asthma Medication in Young Norwegian Adults

Jørgen G. Bramness; Tilmann von Soest


BMC Pulm Med. 2019;19(52) 

In This Article


Procedure and Participants

This study is based on data from the Young in Norway Study, described in more detail elsewhere.[29,30] A population-based sample of Norwegian adolescents was followed over a 13-year span with four data collections. The initial sample at the first time point (T1) was composed of 12,287 persons with a response rate of 97%. Only parts of the sample was invited for follow-up at later time points, and the cumulative response rate over all four data collection times for those who were eligible to be included at all data collection points was 69%. Participants were asked to give their consent to obtain information about them in various nationwide official registers such as the Norwegian Prescription Database (NorPD), and 90.0% consented to such linkage. In this study, we drew on the available data from 2602 individuals, 1145 males (44.0%) and 1457 females (56.0%). Survey data were collected at four times and mean age of the respondents at these data collection points was T1: 15.05 (SD = 1.96 in 1992), T2: 16.53 (1994), T3: 22.95 (1999) and T4: 28.48 years (2005–6), respectively.

Questionnaire data from the Young in Norway Study were linked to register data from the NorPD. Since 1 January 2004, all pharmacies in Norway are obliged, by law, to submit monthly electronic data on dispensed prescriptions to the Norwegian Institute of Public Health. The NorPD contains information on all prescription drugs, reimbursed or not, dispensed at Norwegian pharmacies to individual patients who live outside institutions.[31] The register contains information about all prescriptions, including the patients' unique identifiers (encrypted), gender, age, date of dispensing and drug information, including brand name and anatomical therapeutic chemical (ATC) code.[32] The data from the Young in Norway Study and NorPD were linked by Statistics Norway as a third party ensuring the anonymity of the responders. The survey data from Young in Norway Study or NorPD were not visible for Statistics Norway during the linkage prosess.


Cannabis use parameters were taken from the Young in Norway Study. Cannabis use was measured at T4. We categorized respondents into three groups according to their self-reported use of cannabis: those who had never used cannabis; those who had used cannabis at least once in their lifetime, but not in the last 12 months; and those who had used cannabis at least once during the last 12 months.

Gender, age, years of education and body mass index (BMI; kg/m2) were taken from reports at T4, while self-reported information on asthma ("Do you have asthma?" no/yes) and allergies ("Are you bothered by allergies?" no/yes) were taken from the earlier data collection at T1. Information on tobacco smoking habits were taken from T4, categorizing the responders into: those who had never smoked regularly; those who smoked regularly before, but not now; those who currently smoked sometimes, but not daily; and current daily smokers.

Information about all anti-asthma medication prescriptions between 2007 and 2015 were obtained through the NorPD, and we compared participants who did not obtain any prescriptions for these drugs in this time interval with participants who did. The interval between self-reported cannabis use and filling a prescription could thus vary from 1 to 9 years. The drugs studied were β2-adrenergic receptor agonists (ATC code R03A*) and glucocorticoids for inhalation (R03B*).

Statistical Analysis

Bivariate relationships between the explanatory variables and the outcome variable "prescriptions for anti-asthmatic drugs" were examined using chi square tests for categorical variables and Student's T-tests for continuous variables. We also examined how our main explanatory variable "cannabis use" was related to other explanatory variables by using chi square for categorical variables and ANOVA for continuous variables. In a new set of analyses we performed binary logistic regressions with anti-asthmatic drug prescriptions as outcome, presenting firstly unadjusted odds ratios (OR) with 95% confidence intervals (95% CI), then a model adjusting for gender, age, earlier self-reported asthma and allergies, and a final model adjusting additionally for level of education, BMI and smoking habits. P-values of less than 0.05 were considered statistically significant, but mostly exact p-values are presented.