Evaluating the causal relationship between five modifiable factors and the risk of spinal stenosis: a multivariable Mendelian randomization analysis

Abstract

Background Spinal stenosis is a neurological disorder related to the compression of the spinal cord or nerve roots, and its incidence increases yearly. We aimed to use Mendelian randomization (MR) to investigate the causal relationship between several modifiable risk factors and the risk of spinal stenosis. Methods We obtained genome-wide association study summary data of large-sample projects (more than 100,000 individuals) from public databases. The data were associated with traits, including years of schooling (educational attainment) from the IEU OpenGWAS Project, smoking behavior (never vs. initiation) from the IEU OpenGWAS Project, body mass index (BMI) from the UK Biobank, length of mobile phone use from the UK Biobank, time spent watching television (TV) from the UK Biobank, and spinal stenosis from FinnGen biobank. Spinal stenosis was used as the outcome, whereas the other four traits were used as exposures. Inverse variance weighted (IVW) regressions were used as a primary to estimate the causal-effect size. Several sensitive analyses (including consistency, heterogenicity, and pleiotropy analyses) were conducted to test the stability and reliability of causal estimates. Results Univariable MR analyses showed that genetically predicted higher educational attainment (IVW; odds ratio (OR) = 0.606; 95% confidence interval (CI): 0.507–0.724; P = 3.37 × 10−8) and never smoking (IVW; OR = 1.388; 95% CI [1.135–1.697]; P = 0.001) were negatively correlated with the risk of spinal stenosis. Meanwhile, a higher BMI (IVW; OR = 1.569; 95% CI [1.403–1.754]; P = 2.35 × 10−8), longer time spent using a mobile phone (IVW; OR = 1.895; 95% CI [1.306–2.750]; P = 0.001), and watching TV (IVW; OR = 1.776; 95% CI [1.245–2.532]; P = 0.002) were positively associated with the risk of spinal stenosis. Multivariable MR analysis indicated that educational attainment (IVW; OR = 0.670; 95% CI [0.465–0.967]; P = 0.032) and BMI (IVW; OR = 1.365; 95% CI [1.179–1.580]; P = 3.12 × 10−5) were independently causally related to the risk of spinal stenosis. Conclusion Our findings supported the potential causal associations of the five factors (educational attainment, smoking behavior, BMI, length of mobile phone use, and watching TV) with the risk for spinal stenosis. While replication studies are essential, these findings may provide a new perspective on prevention and intervention strategies directed toward spinal stenosis.