Abstract
Background: The association between lifestyle factors and Multiple Sclerosis (MS) disease severity and progression has been investigated to a lesser extent compared with susceptibility to the disease. Objective: We aimed to assess the impact of lifetime alcohol and cigarette smoking load on MS severity. Methods: Design: a cross-sectional study. Three hundred fifty-one patients consecutively admitted to the Department of Neurology were asked to complete the "Questionnaire of Lifestyle" (part of the European Prospective Investigation into Cancer and Nutrition project). An estimation of the cumulative lifetime cigarette smoking and alcohol load was calculated as the weighted sum of the mean number of cigarettes smoked and standard alcoholic drinks consumed per day at different ages. The measure of exposure was expressed in terms of pack-year and drink-year. Disease severity was estimated by the Multiple Sclerosis Severity Score (MSSS). Logistic regression analyses were performed using MSSS (first tertile vs. third tertile) as the outcome. Results: The median MSSS was higher (3.2 vs. 2.3, p = 0.002) in ever- vs. never-smokers, but we did not find a difference between ever- and never-drinkers (2.7 vs. 2.8, p = ns). Ever-smokers were almost twice as likely to fall in the upper MSSS tertile than never-smokers. Ever-drinkers did not show a statistically significant association between alcohol intake and MS severity. The risk of falling in the worst MSSS tertile for smokers was 10.81 (2.0-58.48; p < 0.01) if they were never-drinkers, whereas it was only 1.65 (0.89-3.03, p = 0.11) if they were also drinkers. On the other side, the risk of falling in the worst MSSS tertile for drinkers did not change as much, whether they also were smokers (0.46; 0.13-1.65; p = 0.23) or not (1.49; 0.55-4.04, p = 0.43). Conclusions: Cigarette smoking, unlike alcohol consumption, is associated with MS severity. Alcohol consumption may attenuate the effect of smoking on disease severity, acting as an effect modifier. The biological background of this effect is unknown. The limitations of our study are mostly due to its cross-sectional design.