Psychiatric disorders, including depressive disorders, schizophrenia (SCZ), bipolar disorder (BIP), autism spectrum disorder (ASD), attention deficit/hyperactivity disorder (ADHD), and other mental disorders, have an age-standardized prevalence of >12,000 per 100,000 people and represent leading causes of disease burden worldwide (GBD 2019 Mental Disorders Collaborators, 2022). Common psychiatric disorders, such as SCZ, major depressive disorder (MDD) and BIP, exhibit a complex genetic basis characterized by high polygenicity and shared genetic architecture (Grotzinger et al., 2022; Lee et al., 2021).

Lung cancer (LC) has consistently ranked among the most prevalent cancers worldwide for several decades and remains a major contributor to cancer-related mortality (Bray et al., 2018; Torre et al., 2015). Tobacco smoking is the primary environmental risk factor associated with the development of LC (Malhotra et al., 2016; Torre et al., 2015). Notably, smoking is highly prevalent among individuals diagnosed with psychiatric disorders, including SCZ, DEP, BIP, and insomnia (INS) (de Leon and Diaz, 2005; Fornaro et al., 2021; George et al., 2012; Leger et al., 2022; Mathew et al., 2017). Compared with the general population, patients with SCZ (Lichtermann et al., 2001; McGinty et al., 2012; Pettersson et al., 2020), DEP (Dalton et al., 2002; Knekt et al., 1996; Wang et al., 2020), MDD (Hung et al., 2014), and INS (Peeri et al., 2022) have shown a higher incidence of LC, however, conflicting findings have also been observed (Nordentoft et al., 2021; Trudel-Fitzgerald et al., 2020). These findings raise questions regarding the causal relationships between major psychiatric disorders and the risk of developing LC, as well as the potential role of smoking as a mediator in these associations. Epidemiological studies have difficulty addressing these questions due to the inherent limitations in controlling both known and unknown confounders, potential measurement biases, and the issue of reverse causation (Edwards et al., 2015; Hammerton and Munafò, 2021; Yarmolinsky et al., 2018). Moreover, conducting randomized controlled trials to explore such causal inference would be ethically impractical.

Mendelian randomization (MR) utilizes germline genetic variants as instrumental variables (IVs), which serve as proxies for modifiable exposures such as smoking, to infer a causal relationship between exposures/risk factors and health outcomes such as LC. Genotypes, which are pairs of genetic variants/alleles at the same genomic position, are randomly segregated from parents to offspring and remain unchanged after conception. Thus, MR is less prone to confounding and reverse causation, enhancing its validity in investigating causal associations (Davey Smith and Hemani, 2014; Davies et al., 2018).

Here we present our findings from assessing the genetic correlation between eight psychiatric disorders, including SCZ, DEP, MDD (Wray et al., 2018), BIP, ASD, ADHD, INS, and neuroticism (NTC), and LC. In addition, we conducted comprehensive MR analyses to examine causal associations of psychiatric disorders with LC. Furthermore, we investigated the genetic associations between the mentioned disorders and smoking behavior, as well as the potential role of smoking in the causal relationships between psychiatric disorders and LC.