Abstract
Background:
Both smoking and exposure to passive smoking have repeatedly been associated with increased multiple sclerosis (MS) risk, but have never before been studied together. We assessed the public health impact of these factors.
Methods:
In a Swedish population-based case-control study (2455 cases, 5336 controls), we calculated odds ratios of developing MS associated with different categories of tobacco smoke exposure, together with 95% confidence intervals, by using logistic regression. The excess proportion of cases attributable to smoking and passive smoking was calculated as a percentage.
Results:
Both smoking and exposure to passive smoking contribute to MS risk in a dose-dependent manner. At the population level, 20.4% of all cases were attributable to smoke exposure. Among subjects carrying the genetic risk factor HLA-DRB1*15 but lacking HLA-A*02, 41% of the MS cases were attributable to smoking.
Conclusions:
From a public health perspective, the impact of smoking and passive smoking on MS risk is considerable. Preventive measures in order to reduce tobacco smoke exposure are, therefore, essential. In particular, individuals with a history of MS in the family should be informed regarding the impact of smoking on the risk of MS, and the importance of preventing their children from being exposed to passive smoke.
Introduction
Smoking is one of the most established environmental risk factors associated with the onset of multiple sclerosis (MS). 1 Both duration and intensity of smoking seem to contribute independently to the risk of MS. Smoking only a few cigarettes a day for a longer period of time confers an increased MS risk. 2 In all previous studies on smoking and MS risk, smokers have been compared with never smokers, which is not an ideal comparison group, since passive smoking also increases the risk of MS, 3 implying that the influence from smoking on MS risk may have been underestimated in previous studies. Even though the prevalence of smoking in Sweden has decreased in the last decade, 11% of the Swedish population were regular smokers in 2013. Around 17% reported that they were involuntarily exposed to environmental tobacco smoke, and 13% of infants born in 2010 had at least one parent who smoked. 4 From a public health perspective it is of obvious interest to determine the proportion of MS cases that are attributable to smoking and exposure to passive smoking. Using the on-going Epidemiological Investigation of Multiple Sclerosis (EIMS), now comprising 2488 incident cases of MS and 5433 randomly selected population controls, we calculated the odds ratios of developing MS conferred by different combinations of smoking and exposure to passive smoking, compared with subjects without exposure to tobacco smoke. We further calculated the excess proportion (EP) of cases attributable to smoking and passive smoking to assess the public health impact of these factors. Since a gene–environment interaction exists between tobacco smoke exposure and HLA genotype with regard to MS risk,5,6 we also calculated the EP of cases attributed to smoking in the context of different HLA genotypes.
Methods
Study design and study subjects
This study was based on EIMS, a Swedish population-based case-control study on environmental and genetic risk factors for MS. The study group comprises the Swedish population aged 16–70 years. In Sweden there is equal access to publicly–sponsored health care for all residents, and patients with MS are cared for by neurologists. Cases were recruited via hospital-based and privately-run neurology units. In total, there were 40 reporting clinics, including all university hospitals in Sweden. All cases were examined and diagnosed by a neurologist located at the unit in which the case was entered. Cases that did not fulfill the McDonalds criteria 7 at the time of this report were excluded. Using the national population register, which is continuously updated, two controls were randomly selected for each case in connection to the inclusion of the case, matched by age, gender and residential area. If we were not able to make contact with a selected control, or if the control declined participation in the study, another control was selected. Ethical approval was obtained from the relevant ethics committee. More details on study design and methods are given elsewhere, 8 as well as in the supplemental web material.
Data collection and definition of smoking habits
Information regarding life-style factors and a variety of other exposures was collected using a standardized questionnaire given to the cases shortly after they had received their diagnosis, and sent by mail to the controls. Information on smoking was obtained by asking about current and previous smoking, including duration of smoking, average number of cigarettes smoked per day and type of cigarettes. Information on exposure to passive smoking was obtained by asking if the subjects had been daily exposed to environmental tobacco smoke at home or at work (Supplementary Table 1). During the study period April 2005 to December 2014, completed questionnaires were obtained from 2488 cases and 5433 controls, the response proportion being 92% for the cases and 67% for the controls.
For each case, the time of the initial appearance of MS symptoms was used as an estimate of the disease onset, and the year in which this occurred was defined as the index year. The corresponding controls were given the same index year. Tobacco smoking was considered prior to the end of the index year. Subjects who had smoked regularly before or at the index year were defined as ever smokers, and those who had never smoked before or at the index year were defined as never smokers. Subjects who had been exposed to passive smoking during at least one year before or during the index year were defined as exposed to passive smoking, whereas those who had never been exposed to passive smoking were defined as not exposed to tobacco smoke. Subjects were also categorized into groups based on the amount of smoking consumption (pack years), and the duration of exposure to passive smoking. Subjects who could not provide detailed information on both smoking habits and exposure to passive smoking were excluded, as were those who had been exposed to passive smoking for less than one year (33 cases and 97 controls). The study thus comprised 2455 cases and 5336 controls.
Statistical analysis
We calculated odds ratios of developing MS associated with different categories of tobacco smoke exposure as compared with the reference category never exposed to tobacco smoke (that is, subjects who reported they had never smoked or been exposed to passive smoking), together with 95% confidence intervals (95% CI), by using logistic regression models. Both matched and unmatched analyses of the data were performed. However, only the results from the unmatched analyses are presented in this report, since these were in close accordance with those from the matched analyses, but in general had higher precision in terms of more narrow confidence intervals.
All analyses were adjusted for age, gender, residential area, ancestry, snuff use and alcohol consumption habits. Assessment of ancestry was based on whether the subject was born in Sweden or not, and whether either of the subject’s parents had immigrated to Sweden. A subject who was born in Sweden, whose parents had not immigrated, was classified as Swedish. Snuff use was dichotomized into ever or never snuff users. Alcohol consumption habits was categorized based on alcohol consumption at the index year (yes, no or unknown). The results remained similar when the amount of alcohol consumption at the index year was considered. Other potential confounding factors taken into consideration were adolescent body mass index (BMI), ultra violet ray (UVR) exposure habits during the last five years, a history of infectious mononucleosis, educational level and socioeconomic status, but these factors had minor influences on the results of the study and were not adjusted for in the final analyses. BMI at age 20 was calculated by dividing self-reported weight in kilograms by self-reported height in metres squared. Based on three questions regarding exposure to UVR, where each answer alternative was given a number ranging from 1 (the lowest exposure) to 4 (the highest exposure), we constructed an index by adding the numbers together and thus acquired a value between 3 and 12. A history of infectious mononucleosis was dichotomized into yes or no. Educational level was categorized into no secondary education, post secondary education without university degree or university degree. The last occupation during the year before the index year was used as a marker for socioeconomic class, which was categorized into the following strata: 1, workers in goods production; 2, workers in service production; 3, employees at lower and intermediate levels; 4, employees at higher levels, executives, university graduates and 5; others such as pensioners, students and unemployed.
The trend test for a dose–response relationship regarding smoking and risk of MS was performed by using a continuous variable for cumulative dose of smoking, whereas the trend test regarding passive smoking and risk of MS was performed separately for ever and never smokers, by using a continuous variable for duration of exposure to passive smoking in a logistic regression model.
The excess proportion of cases attributable to smoking and passive smoking, respectively, was calculated in percent (EP%) 9 as an indicator of the impact of these exposures on the occurrence of MS in the population. All analyses were conducted using Statistical Analysis System (SAS) version 9.4.
Results
The mean duration between diagnosis and inclusion in the study was 11 months, and the mean duration between disease onset and diagnosis was 3.4 years. Characteristics of cases and controls among never smokers and ever smokers (further divided into more or less than 10 pack years) are presented in Table 1. From the table it can be seen that, among controls, smoking is correlated to passive smoking, snuff use and to a minor extent to alcohol.
Characteristics of cases and controls among never smokers and smokers (<10 pack years (p/y) and ⩾10 p/y, respectively).
data on alcohol consumption was missing for 382 cases and 1060 controls among never smokers, for 311 cases and 569 controls among smokers who had consumed less than 10 pack years and for 144 cases and 246 controls among smokers who had consumed 10 or more pack years.
When investigating the impact of smoking and exposure to passive smoking in different categories based on cumulative dose of smoking and duration of passive smoking, both exposures appeared to contribute to the risk of MS (Table 2). Duration of exposure to passive smoking was associated with MS risk in a dose-dependent manner among never smokers (p=0.006) and smoking was dose-dependently associated with increased risk of MS, regardless of exposure to passive smoking (p<0.0001). Those who reported smoking more than 10 pack years and exposure to passive smoking for more than 20 years had nearly three times higher risk of developing MS compared to those who reported no exposure to tobacco smoke (OR=2.7, 95% CI 2.0–3.8).
Odds ratio (OR) with 95% confidence interval (CI) of developing MS for subjects with different combinations of smoke exposure, compared with never-smokers who have never been exposed to passive smoking.
One pack year is the equivalent to smoking 20 cigarettes daily for one year.
number of exposed cases and controls.
adjusted for age, gender, residential area, and ancestry.
adjusted for age, gender, residential area, ancestry, snuff use, alcohol consumption habits and past/current smoking.
Our analyses were adjusted for a broad range of possible confounding factors including socioeconomic status, educational level, adolescent BMI and a history of infectious mononucleosis. However, these factors had minor influences on the results of the study and were not adjusted for in the final analyses.
Overall, 20.4% of all cases in the population were attributable to smoke exposure (active or passive). Among smokers, smoking was responsible for 33% of the MS cases, and among never smokers exposed to passive smoking 12% of the cases were attributable to passive smoking. In subjects with the genetic risk factors carriage of HLA-DRB1*15 and absence of HLA-A*02, 41% (23–55) of the MS cases were attributable to smoking (Table 3).
Odds ratio (OR) with 95% confidence interval (CI) of developing MS for ever smokers, compared with never smokers, among subjects with different combinations of HLA-DRB1*15 and HLA-A*02, together with the proportion of cases attributable to smoking (Excess Fraction (EF)).
number of exposed cases and controls.
adjusted for age, gender, residential area, and ancestry, snuff use and alcohol consumption habits.
Discussion
Our study allowed us to quantify the influence from smoking and passive smoking on the risk of developing MS more accurately than previous studies. The increased risk due to these factors is dependent on both the amount of smoke exposure and genotype. Subjects who had smoked more than 10 pack years and had been exposed to long-term passive smoking had an almost three-fold increased risk of developing MS than those reporting no exposure to tobacco smoke.
At the population level, 20.4% of all cases could be attributable to smoke exposure (active or passive). Among smokers, smoking was estimated to be responsible for 33% of the MS cases. Among subjects with the genetic risk factors carriage of HLA-DRB1*15 and absence of HLA-A*02, 41% of the MS cases were attributable to smoking. Among never-smokers, around 5% of the MS cases were attributed to passive smoking. Both smoking and passive smoking are preventable factors, and our results illustrate what possibly could be gained by preventive effort against these factors.
Internationally Sweden has a low prevalence of smoking, and several preventive measures have been taken in order to reduce tobacco smoking and exposure to environmental tobacco smoke. However, the Swedish national goals of 2003 which aimed to decrease smoking and eliminate involuntary exposure to passive smoking have not been reached. 4 Even though comprehensive tobacco control policies that apply to the whole population appears to be the best way of reducing smoking, this paper is important since it may provide a rationale for specific counselling against smoke exposure for individuals where MS runs in the family.
It should be pointed out that the result regarding the excess proportion for smoke exposure at the population level is highly dependent on the exposure frequency. Thus, in other countries where smoking is more prevalent, the excess proportions may be higher than that observed in our study, which was based on data from Sweden, where the prevalence of smoking is low compared to many other countries.
The effect of smoking and passive smoking may be modified by other environmental and genetic factors in the development of MS.10–12 The precise effects of smoking and passive smoking on MS development may thus vary considerably in different contexts and in different populations. However, smoking has been found to be an essential risk factor for MS in almost all published studies.
The study was designed as a case-control study in which information regarding different exposures was collected retrospectively. In order to minimize recall bias, we primarily included incident cases of MS who had received the diagnosis within the past year. We made a great effort to obtain the information on smoking in an identical way for the cases and the controls. The questionnaire contained a wide range of questions regarding many potential environmental risk factors, and no section in the questionnaire was given prime focus.
The problem of selection bias was minimized by the population-based design, and even though there was a relatively high proportion of non-responders among the controls, this bias is probably modest because the prevalence of life style factors, such as smoking and alcohol consumption, among the controls was consistent with that of the general population of similar ages. 13 Furthermore, the distribution of socioeconomic status among controls was in line with that of the general population. 13 Some cases may have been unidentified in our study, for example those who were diagnosed in private clinics not participating in our studies. However, considering the structure of the public Swedish health care system, which provides equal free-of-charge access to medical services for all Swedish citizens, we believe almost all cases of MS are referred to neurological units. However, there is a risk that patients were not asked to participate in the study due to the attending neurologists’ lack of time.
All analyses were adjusted for age, gender, residential area, snuff use and alcohol habits. We further considered possible confounding from adolescent BMI, history of infectious mononucleosis, socio-economic status and educational level. However, these factors had minor influences on the results of the study and were not adjusted for in the final analyses.
In conclusion, the impact of exposure to tobacco smoke is substantial. Exposure to tobacco is a preventable risk factor, and preventive measures at a population level in order to reduce both smoking and exposure to passive smoking are essential. In particular, individuals with a history of MS in the family should be informed regarding the impact of smoking on the risk of MS, and the importance of preventing their children from being exposed to passive smoke.
Footnotes
Conflict of interest
Dr Hedström reports no disclosures. Dr Olsson received compensation for scientific advisory boards or lectures from Biogen and Genzyme, unrestricted MS research grants from Biogen, Bovartis, Genzyme, Allmiral and AstraZeneca, the Swedish Research Council (07488), EU fp7 Neurinox, Knut and Alice Wallenberg Foundation, Margareta af Ugglas Stiftelse, the Afa Foundation and the Swedish Brain Foundation. Dr Alfredsson receives research support from the Swedish Medical Research Council (K2013-69X-14973-10-4) and Swedish Council for Working life and Social Research (Dnr 2012-0325).
Funding
The study was supported by grants from the Swedish Medical Research Council; from the Swedish Research Council for Health, Working Life and Welfare, the Knut and Alice Wallenberg Foundation, the AFA foundation, the Swedish Brain Foundation and the Swedish Association for Persons with Neurological Disabilities.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
