(188) Association between antidepressants and risk of withdrawal syndrome in illicit substance users: nationwide case-case-time-control study

Background: Substance use is linked to an extensive range of medical challenges, but appropriate treatment for non-opioid substance use is not available currently. Venlafaxine may reduce non-opioid substance withdrawal but the effects have not been evaluated. However, the findings are inconsistent and limited due to the strict eligibility criteria used in clinical trials. Given the urgent need for effective treatments for non-opioid substance use disorders, it is important to evaluate the effectiveness of venlafaxine in these patients through real-world evidence.

Objectives: To investigate the association between venlafaxine use and the risk of substance withdrawal in patients with non-opioid substance use

Methods: We conducted a case-case-time-control study to address issues of time invariant confounding and protopathic bias. We used Taiwan’s NHIRD to retrieve medical record and linked with Taiwan Illicit Drug Issue Database to identify illicit substance use records. We included patients arrested for illicit substance use with incident withdrawal syndrome between 2010 and 2019. The exposure of interest was venlafaxine. For each patient, we compared antidepressants during hazard periods (day -8 to -67 before the withdrawal event) with reference period (days -248 and -307). We estimated odds ratios with 95% confidence intervals by conditional logistic regression to evaluate the association between recent use of antidepressants and the risk of withdrawal syndrome. Additionally, we selected mirtazapine as a control exposure to perform a negative control analysis because the mechanism of mirtazapine is different from that of venlafaxine.

Results: The crossover analyses included 4,643 current cases. Participants’ average age on the index date was 39.5 years (SD 8.7) with 84.1% men and 88.3% having low income. With adjustment by the control crossover analysis consisting of future cases, we found the use of venlafaxine was associated with reduced risk of substance withdrawal (OR 0.35; 95% CI 0.13 to 0.96). However, we found no association between recent use of venlafaxine and all-cause mortality (OR 1.08; 95% CI 0.55 to 2.14). The results remained consistent throughout a series of sensitivity analyses, including various lengths of time period for hazard or reference period or consideration on the opioid users. We did not find a reduced risk of substance withdrawal from mirtazapine use.

Conclusions: Our findings indicated an association between the use of venlafaxine and a decreased risk of withdrawal syndrome, whereas the use mirtazapine did not show this association. This study provides a strong ground for clinicians to consider the use of venlafaxine to reduce patients’ substance withdrawal.