Associations between changes in cannabis use between assessments and respiratory symptoms across multiple assessments using generalised estimating equations
| Participants n | Mean observations per participant | Group | OR (95% CI) | p-value | |
| Cough | 985 | 3.8 | Nonuser | Ref. | <0.001 |
| Quitter | 1.17 (0.79–1.74) | ||||
| Starter | 1.49 (1.03–2.16) | ||||
| Continuing | 2.84 (2.09–3.85) | ||||
| Sputum | 985 | 3.8 | Nonuser | Ref. | <0.001 |
| Quitter | 0.91 (0.59–1.43) | ||||
| Starter | 1.91 (1.34–2.71) | ||||
| Continuing | 3.06 (2.24–4.16) | ||||
| Wheeze | 985 | 3.8 | Nonuser | Ref. | 0.006 |
| Quitter | 1.21 (0.83–1.77) | ||||
| Starter | 1.29 (0.90–1.84) | ||||
| Continuing | 1.72 (1.27–2.34) | ||||
| Dyspnoea | 985 | 3.8 | Nonuser | Ref. | 0.347 |
| Quitter | 1.14 (0.78–1.67) | ||||
| Starter | 1.19 (0.83–1.72) | ||||
| Continuing | 1.33 (0.96–1.85) |
Analyses were adjusted for sex, age of assessment, and tobacco use and asthma at that age. Odds ratios represent the odds associated with changes in cannabis use compared with the reference (nonuser) group.