Incidence of and Determinants for Substance Use Disorders Among Individuals With Long‐Term Prescription Opioid Use: A Population‐Based Cohort Study

Purpose To assess the incidence of substance use disorders (SUDs) after long‐term prescription opioid use (LTOU) and to identify socioeconomic and clinical risk factors associated with SUD among individuals with LTOU. Methods Cohort study using linked nationwide registers (2011–2019). We identified 114 916 individuals who used opioids for more than 3 months (LTOU) without previous LTOU or SUD diagnosis. Outcomes were any incident SUD diagnosed in primary or secondary care (ICPC‐2: P15‐P16, P18‐P19; ICD‐10: F10‐F16, F18‐F19) and opioid use‐related disorders (OUD) diagnosed in secondary care (ICD‐10: F11). We calculated age‐ and sex‐stratified incidence rates (IR), incidence rate ratios (IRR) and age‐standardized incidence rates (ASIR). Adjusted hazard ratios (aHR) were calculated using Cox proportional hazards regression. Results In total, 5.3% (6069/114916) were diagnosed with SUD (ASIR = 28.7 per 1000 person‐years), and males had higher IRs compared to females (IRR). Males had higher risk of SUD in both the younger (aHR = 1.59, 95% CI 1.47–1.72) and older (1.66, 1.54–1.78) age group. Low education (1.87, 1.66–2.11) and unemployment (1.26, 1.15–1.38) had the strongest association with SUD in the younger age group versus low income (1.37, 1.21–1.57) and living alone (1.53, 1.41–1.65) in the older age group. Previously diagnosed mental disorders and use of benzodiazepines‐ or benzodiazepine‐related drugs (BZDRs) were associated with SUD in both age groups (1.85, 1.71–2.01; 2.37, 2.18–2.57). Being male and having used BZDRs were the covariates strongest associated with OUD. Conclusions Being male, young, having low socioeconomic status, previous mental disorders or BZDR use were associated with SUD diagnosis among individuals with LTOU.