Despite growing awareness of the mental health damage caused by air pollution, the epidemiologic evidence on impact of air pollutants on major mental disorders (MDs) remains limited. We aim to explore the impact of various air pollutants on the risk of major MD.

This prospective study analyzed data from 170 369 participants without depression, anxiety, bipolar disorder, and schizophrenia at baseline. The concentrations of particulate matter with aerodynamic diameter ≤ 2.5 μm (PM2.5), particulate matter with aerodynamic diameter > 2.5 μm, and ≤ 10 μm (PM2.5–10), nitrogen dioxide (NO2), and nitric oxide (NO) were estimated using land-use regression models. The association between air pollutants and incident MD was investigated by Cox proportional hazard model.

During a median follow-up of 10.6 years, 9 004 participants developed MD. Exposure to air pollution in the highest quartile significantly increased the risk of MD compared with the lowest quartile: PM2.5 (hazard ratio [HR]: 1.16, 95% CI: 1.09–1.23), NO2 (HR: 1.12, 95% CI: 1.05–1.19), and NO (HR: 1.10, 95% CI: 1.03–1.17). Subgroup analysis showed that participants with lower income were more likely to experience MD when exposed to air pollution. We also observed joint effects of socioeconomic status or genetic risk with air pollution on the MD risk. For instance, the HR of individuals with the highest genetic risk and highest quartiles of PM2.5 was 1.63 (95% CI: 1.46–1.81) compared to those with the lowest genetic risk and lowest quartiles of PM2.5.

Our findings highlight the importance of air pollution control in alleviating the burden of MD.

Depression is a debilitating mental disorder that often coexists with anxiety. The genetic mechanisms of depression and anxiety have considerable overlap, and studying depression in non-anxiety samples could help to discover novel gene. We assess the genetic variation of depression in non-anxiety samples, using genome-wide association studies (GWAS) and linkage disequilibrium score regression (LDSC).

The GWAS of depression score and self-reported depression were conducted using the UK Biobank samples, comprising 99,178 non-anxiety participants with anxiety score <5 and 86,503 non-anxiety participants without self-reported anxiety, respectively. Replication analysis was then performed using two large-scale GWAS summary data of depression from Psychiatric Genomics Consortium (PGC). LDSC was finally used to evaluate genetic correlations with 855 health-related traits based on the primary GWAS.

Two genome-wide significant loci for non-anxiety depression were identified: rs139702470 (p = 1.54 × 10−8, OR = 0.29) locate in PIEZO2, and rs6046722 (p = 2.52 × 10−8, OR = 1.09) locate in CFAP61. These associated genes were replicated in two GWAS of depression from PGC, such as rs1040582 (preplication GWAS1 = 0.02, preplication GWAS2 = 2.71 × 10−3) in CFAP61, and rs11661122 (preplication GWAS1 = 8.16 × 10−3, preplication GWAS2 = 8.08 × 10−3) in PIEZO2. LDSC identified 19 traits genetically associated with non-anxiety depression (p < 0.001), such as marital separation/divorce (rg = 0.45, SE = 0.15).

Our findings provide novel clues for understanding of the complex genetic architecture of depression.

The role of neurological proteins in the development of bipolar disorder (BD) and schizophrenia (SCZ) remains elusive now. The current study aims to explore the potential genetic correlations of plasma neurological proteins with BD and SCZ.

By using the latest genome-wide association study (GWAS) summary data of BD and SCZ (including 41,917 BD cases, 11,260 SCZ cases, and 396,091 controls) derived from the Psychiatric GWAS Consortium website (PGC) and a recently released GWAS of neurological proteins (including 750 individuals), we performed a linkage disequilibrium score regression (LDSC) analysis to detect the potential genetic correlations between the two common psychiatric disorders and each of the 92 neurological proteins. Two-sample Mendelian randomisation (MR) analysis was then applied to assess the bidirectional causal relationship between the neurological proteins identified by LDSC, BD and SCZ.

LDSC analysis identified one neurological protein, NEP, which shows suggestive genetic correlation signals for both BD (coefficient = −0.165, p value = 0.035) and SCZ (coefficient = −0.235, p value = 0.020). However, those association did not remain significant after strict Bonferroni correction. Two sample MR analysis found that there was an association between genetically predicted level of NEP protein, BD (odd ratio [OR] = 0.87, p value = 1.61 × 10−6) and SCZ (OR = 0.90, p value = 4.04 × 10−6). However, in the opposite direction, there is no genetically predicted association between BD, SCZ, and NEP protein level.

This study provided novel clues for understanding the genetic effects of neurological proteins on BD and SCZ.