456 Independent causal effect of psoriasis on multiple sclerosis identified by Mendelian randomization

Abstract

Psoriasis and multiple sclerosis (MS) are physiologically distinct yet share multiple genetic signals and have overlapping pathogenesis. Previous studies suggest psoriasis patients are at greater risk of MS than the general population, and the two diseases also have overlapping comorbidities and traits (for example, they are both more prevalent in northern latitudes). To evaluate whether there is a direct causal relationship, we conducted Mendelian randomization (MR) analysis with six different techniques, using existing GWAS for psoriasis (11,024 cases and 16,336 controls) and MS (14,802 cases and 26,703 controls); we also addressed 10 potential confounding exposures, previously suggested to be associated with both diseases, by including GWAS data from body mass index, coronary artery disease, inflammatory bowel disease (IBD), type 1 diabetes (T1D), type 2 diabetes (T2D), asthma, rheumatoid arthritis, drinks per week, cigarettes per day, and vitamin D levels. Four of the MR techniques indicated a significant effect of psoriasis on MS (FDR<0.05) and two showed nominal significance, whereas the effect of MS on psoriasis was not significant in any of the MR analyses. When including all covariates nominally significant from at least one technique in a multivariable MR analysis, psoriasis still has a significant effect on MS (p=5.8×10-3, OR=1.04), independent of T1D (p=4.3×10-7, OR=1.05), T2D (p=2.3×10-3, OR=1.08), IBD (p=1.6×10-11, OR=1.11) and vitamin D (p=9.4×10-3, OR=0.75) in the multivariable analysis, while BMI and drinks/per week were not significant. By applying multiple different MR techniques to multiple comorbidities and traits, we have been able to reveal the most important modifiable risk factors and determine there is indeed an independent causal relationship between psoriasis and MS.