Benzodiazepine Use and Incident Cancer – A Population-based Cohort Study

Yochai Schonmann

Keywords: Benzodiazepines, malignancy, cancer, population-based, routinely collected data

Background:

Both long-term and sporadic use of benzodiazepines and Z-drugs are associated with various adverse effects, yet they are frequently prescribed for insomnia and anxiety in primary care. Although a link between their use and cancer incidence has been suggested, long-term population-based data is needed.

Research questions:

Does the use of benzodiazepines and related drugs predict future incident cancer?

Method:

A population-based historic cohort study was conducted, following all Clalit members aged 40-70 from 2006 to 2022. Cox regression models were fitted to assess the association between sporadic and chronic benzodiazepine use at baseline (based on cumulative claims during 2005) and subsequent incident cancer. Results were reported both in crude form and adjusted for age, sociodemographic characteristics (sex, ethnicity, socioeconomic status, peripherality), and baseline smoking and obesity.

Results:

We included 1,027,582 Clalit members without any malignancy diagnosis at baseline. The mean age was 53.0 years (±SD=8.1), and 52% were female. Of the participants, 102,619 (10.0%) were sporadic users of benzodiazepines at baseline, and 29,507 (2.9%) were chronic users, with at least nine monthly claims of a benzodiazepine in 2005. Over a median follow-up of 17.0 years, there were 153,956 cases of new cancer diagnoses. Sporadic use of a benzodiazepine was associated with a slightly increased incidence of cancer (adjHR 1.07, 95% CI 1.05-1.08, P<0.0001), while chronic use was associated with a higher risk (adjHR 1.13, 95% CI 1.10-1.16, P<0.0001), compatible with a linear association (P<0.0001). The association remained robust regardless of age but was stronger among younger cohort participants. Subsequent analysis confirmed the same association for both anxiolytic and hypnotic medications but not for the use of Z-drugs (adjHR 1.03, 95% CI 0.99-1.07, P<0.0001 for sporadic use, and adjHR 0.99, 95% CI 0.90-1.08, P<0.0001 for chronic use).

Conclusions:

The use of benzodiazepines, but not Z-drugs, was associated with incident cancer, with evidence supporting a dose-response relationship.

Points for discussion:

#12