Smoking and Epstein\u2013Barr virus infection in multiple sclerosis development

Anna Karin Hedström,Lars Alfredsson,Tim Waterboer,Jesse Huang,Tomas Olsson,Ingrid Kockum,Nicole Brenner,Jan Hillert,Julia Butt

doi:10.1038/s41598-020-67883-w

Abstract

It is unclear whether smoking interacts with different aspects of Epstein–Barr virus (EBV) infection with regard to multiple sclerosis (MS) risk. We aimed to investigate whether smoking acts synergistically with elevated EBNA-1 antibody levels or infectious mononucleosis (IM) history regarding MS risk. Two Swedish population-based case–control studies were used (6,340 cases and 6,219 matched controls). Subjects with different smoking, EBNA-1 and IM status were compared regarding MS risk, by calculating odds ratios (OR) with 95% confidence intervals (CI) employing logistic regression. Potential interaction on the additive scale was evaluated by calculating the attributable proportion due to interaction (AP). Current and past smokers had higher EBNA-1 antibody levels than never smokers (p < 0.0001). There was an additive interaction between current smoking and high EBNA-1 antibody levels (AP 0.3, 95% CI 0.2–0.4), but not between past smoking and high EBNA-1 antibody levels (AP 0.01, 95% CI − 0.1 to 0.1), with regard to MS risk. An interaction also occurred between current smoking and IM history (AP 0.2, 95% CI 0.004–0.4), but not between past smoking and IM history (AP − 0.06, 95% CI − 0.4 to 0.3). Current smoking increases EBNA-1 antibody levels and acts synergistically with both aspects of EBV infection to increase MS risk, indicating that there is at least one pathway to disease in which both risk factors are involved.

Highlights

It is unclear whether smoking interacts with different aspects of Epstein–Barr virus (EBV) infection with regard to multiple sclerosis (MS) risk
Since the logistic regression model is inherently multiplicative, presence of an interaction term implies departure from multiplicativity, which has no direct relevance for the issue of whether or not sufficient-cause interaction is p resent[17]
Ever smoking was associated with increased MS risk regardless of EBNA-1 status

Summary

Introduction

It is unclear whether smoking interacts with different aspects of Epstein–Barr virus (EBV) infection with regard to multiple sclerosis (MS) risk. We aimed to investigate whether smoking acts synergistically with elevated EBNA-1 antibody levels or infectious mononucleosis (IM) history regarding MS risk. Current smoking increases EBNA-1 antibody levels and acts synergistically with both aspects of EBV infection to increase MS risk, indicating that there is at least one pathway to disease in which both risk factors are involved. The first study observed a positive interaction on the multiplicative scale between smoking and elevated EBNA-1 IgG levels with regard to MS risk[4]. Using two Swedish population-based case–control studies comprising 6,340 cases and 6,219 controls, we aimed to investigate the interplay between smoking and different aspects of EBV infection [elevated EBNA-1 IgG levels and infectious mononucleosis (IM) history] with regard to MS risk. Comments Pooled analyses from two case–control studies (Sweden and Tasmania) and one prospective study (United States)

Objectives

Methods

Results

Conclusion