Data supported a positive causal association between narcolepsy and schizophrenia and a reverse causal relationship between narcolepsy and major depressive disorder.
Using a fixed-effects meta-analysis, research suggests that genetically predicted narcolepsy may increase the risk of schizophrenia, and that major depressive disorder (MDD) may be causally related to narcolepsy. The study failed to identify a causal relationship between narcolepsy and attention-deficit hyperactivity disorder (ADHD).1
Published in the Journal of Psychiatric Research, the analysis comprised of genetic variables for schizophrenia (n = 77,096; cases: n = 33,640; controls: 43,456), MDD (n = 500,199; cases: n = 170,756; controls: n = 329,444) and ADHD (n = 55,374; cases: n = 20,183; controls: n = 35,191) from 3 large genome-wide association (GWAS) studies released by the PGC. Four genome-wide single nucleotide polymorphisms (SNPs) significantly associated with narcolepsy were included as instrumental variables for the discovery stage.
Led by senior investigator Changgui Kou, Department of Epidemiology and Biostatistics at the Jilin University in China, a bidirectional Mendelian randomization (MR) analysis was used to examine the causal relationship between narcolepsy and these psychiatric disorders. Although results for the causal estimation of narcolepsy in schizophrenia did not reach the significance threshold (P <.008), evidence (P <.05) suggested that genetically predicted narcolepsy was associated with a high risk of schizophrenia (inverse weighted [IVW]: OR, 1.059; 95% CI, 1.007-1.115; P = .026). As mentioned, no significant effect of narcolepsy was observed on MDD (IVW: OR, 1.011; 95% CI, 0.994-1.028; P = .200) or ADHD (IVW: OR, 1.028; 95% CI, 0.961-1.098; P = .423).
READ MORE: Vagus Nerve Stimulation for Narcolepsy Shows Improvement in Daytime Sleepiness
In the discussion section of the research, the study authors wrote, "investigating the relationship between narcolepsy and psychiatric disorders may lead to better clinical identification and precision pharmacological approaches due to their high comorbidity and symptom overlap. As an illustrative instance, we propose that narcolepsy could serve as a potential pre-diagnostic indicator for SCZ. Therefore, conducting regular assessments of individuals with narcolepsy could play a crucial role in effectively helping identify SCZ in clinical practice."
Narcolepsy and Schizophrenia:
Narcolepsy and Major Depressive Disorder (MDD):
Narcolepsy and ADHD Relationship:
In the validation stage, 23 and 2 SNPs were identified as candidate instruments for the UK Biobank and FinnGen datasets, respectively, following PRESSO outlier test correction. Here, results mirrored what was previously seen, as narcolepsy appeared to increase the risk of schizophrenia in the UK Biobank set (IVW: OR, 4.554; 95% CI, 2.041-10.163; P <.001) and FinnGen data set (Wald ratio: OR, 1.072; 95% CI, 1.037-1.108; P <.001). Notably, the significance and magnitude of narcolepsy in schizophrenia remained (OR, 1.070; 95% CI, 1.041-1.100; P <.001) after a meta-analysis that combined the estimators of 2 stages.
The positive association identified between MDD and narcolepsy was further replicated in the UK Biobank dataset (IVW: OR, 1.024; 95% CI, 1.008-1.040; P = .003); however, no significant evidence was found in the FinnGen MR analysis (IVW: OR, 2.733; 95% CI, 0.704-10.609; P = .146). Using a meta-analysis of both stages the combined OR of MDD on narcolepsy was 1.024 (95% CI, 1.009-1.040; P = .002). Reverse MR analysis of the validation stage and the meta-analysis suggested that schizophrenia and ADHD increased the risk of narcolepsy.
The study had some limitations, including the fact that the FinnGen narcolepsy dataset did not have sufficient SNPs to perform a genome-wide analysis with statistical significance. In addition, the analysis comprised of individuals of European ancestry and thus, makes it difficult to generalize the data to other populations. Lastly, the MR analysis at the discovery stage suffered from a small F value, which weakened the causal effectiveness because the 3 MR assumptions were not met.
