In a recent article published in the BMC Medicine Journal, researchers performed Mendelian randomization (MR) phenome-wide association study (PheWAS) of insomnia symptoms using 11,409 outcome variables, derived and analyzed using an automated pipeline, PHESANT.
Study: Identifying the potential causal role of insomnia symptoms on 11,409 health-related outcomes: a phenome-wide Mendelian randomisation analysis in UK Biobank. Image Credit: fizkes/Shutterstock.com
Background
Insomnia, defined as continuous discontent with the quality or magnitude of sleep for a prolonged duration, could lead to multimorbidity, i.e., the incidence of multiple chronic health conditions.
It is often associated with poor quality of life (QoL), polypharmacy, and premature mortality. Since multimorbidity is an emerging global health threat, identifying its potential causes is rapidly becoming a research priority.
There is growing data that around six to seven percent of Europeans have insomnia, with 33 to 37% of people self-reporting its symptoms.
It is more prevalent in women and older people, and it is the second most prevalent mental health disorder after anxiety.
Yet, studies have not examined its potential causal effects across multiple health/disease outcomes. Earlier studies were mainly observational and focused on hypothesized mental or cardiometabolic outcomes, which did not reflect causal effects.
About the study
In the present hypothesis-free study (MR-PheWAS), researchers tested the causal effects of insomnia symptoms on many phenotypes using genetic variants, in this case, single-nucleotide polymorphisms (SNPs) robustly associated with self-reported insomnia symptoms.
They used 11,409 outcomes from the United Kingdom (UK) biobank covering data from 336,975 British participants to explore the causal effects of insomnia symptoms on multiple diseases- and health-related traits.
A linkage disequilibrium (LD) threshold of R2>0.001 helped the researchers cluster the GWAS-related SNPs into independent SNPs. Next, they generated a weighted genetic risk score (GRS) using 129 independent SNPs based on their allelic association with insomnia symptoms in the original GWAS.
The study outcomes included those obtained from responses to baseline and follow-up assessments. The formerly covered questions regarding height, weight, blood pressure, and bone density, while the latter enquired about accelerometer measurements and different scan reports, including brain and cardiac scans.
It also assessed biomarker measures from blood/urine samples. The researchers linked final health/disease outcomes from primary and secondary care to UK’s national cancer and death registers.
Study findings
The study cohort comprised 54% female and 32% male participants with an average age of 57. These people often self-reported insomnia symptoms, with 48% and 28% reporting these symptoms sometimes and usually, respectively.
The MR-PheWAS covered 11,409 GRS outcome variables, of which 437 met the potential causal effect(s) criteria. In total, the authors followed up on 71 outcomes using two-sample MR.
In the primary and follow-up MR analyses, the authors found proof of an antagonistic causal effect of insomnia symptoms on 30 new outcomes, never before explored using MR.
These were related to anxiety, respiratory, musculoskeletal, and digestive systems disorders, and body composition assessments.
Together, these findings supported that insomnia symptoms caused multimorbidity and raised the possibility of using insomnia treatments, such as cognitive behavioral therapy, to treat other antagonistic health-related outcomes; however, their combined effectiveness requires more investigation.
In prior non-MR scientific literature, the authors have extensively examined the bidirectional relationship between insomnia and headache.
However, they also found new evidence consistently across three phases of this study, the MR-PheWAS, the follow-up, and sensitivity analyses using two samples for several new outcomes, e.g., C-reactive protein (CRP) levels. Note 95% confidence intervals (CIs) barred the null in the first two phases.
Other newly identified health/disease-related outcomes included spondylosis, shoulder lesions, unspecified soft-tissue and joint disorders, diaphragmatic hernia, gastritis, oesophagitis, diverticular disease of the intestine, and bronchitis.
Strikingly, a positive correlation between insomnia and CRP levels, a marker of inflammation triggered by the immune system, provided further evidence that insomnia affects the immune system.
Findings related to insomnia and most outcomes were novel and unexplored in conventional epidemiology studies. Since diaphragmatic hernia, a congenital defect, could not be caused by insomnia; thus, these results, in some cases, violated the main assumptions.
Conclusions
Overall, the current study remarkably identified broad effects of insomnia symptoms on human health, including some new ones, replicated in subsequent analyses, indicating the possible role of insomnia in multimorbidity.
For instance, insomnia symptoms affect respiratory and soft-tissue disorders and digestive issues. The findings confirmed pre-identified outcomes, such as hyperglycemia, pain, mental health, and body composition, reiterating the significance of integrating insomnia treatments into other disease treatments.
In future work, researchers should analyze all individual outcomes extensively to confirm novel revelations, including methods for time-varying exposures and non-linear associations.
Cognitive behavioral therapy for insomnia linked to reduced postpartum depression