Why does smoking increase your risk of MS?

We don’t understand yet why people get MS. The accepted view is that there is an interaction between risk genes and environmental factors like EBV, smoking, and vitamin D status. Epidemiological studies have helped to identify these risk factors for developing MS, but it is unclear how these factors actually contribute to disease evolution and progression.

There is strong evidence that smoking increases your risk of developing MS. There is also strong evidence that in pwMS, smoking is bad for both disease activity and long-term disability progression. The effect of smoking is long-lived: it takes about 10 years after quitting for disease activity to return to the level of people who don’t smoke. As well as being harmful in terms of overall health, smoking increases your risk of Crohn’s disease and Rheumatoid Arthritis – other autoimmune diseases similar in some respects to MS.

Why smoking influences MS is unclear. One explanation is that the chemicals in cigarette smoking alter the tagging and folding of DNA in our cells, and thereby change the way that the genetic code is read out.

A brief interlude on this theme. For many years the central dogma of biology was that information flowed as a one-way street (excuse mixed metaphor) from our genetic code. Crick and Watson, the scientists credited with discovering the structure of DNA, suggested that the genetic material (DNA) tells the cell which proteins to make, but that this information does not flow the other way. Classically, DNA was thought to be used to generate messenger molecules (mRNA), which then formed the recipe for making the proteins required for cellular processes:

We now recognise that this is not the whole story. While the vast majority of our cells carry the same DNA, they do not all make the same proteins. In fact any one cell only reads out from a tiny fraction of its DNA at any one time. In addition, the parts of DNA that a cells reads out from changes dynamically from moment to moment.

How does this phenomenon take place? Well, although the backbone of DNA is a simple double-stranded helix, each molecule of DNA is folded into a complicated 3D structure in the cell. To read out from a particular portion of DNA, that portion needs to be accessible to the reading machinery. We now understand that cells have various ways of ‘tagging’ parts of the DNA molecule which can effectively turn genes on and off. These modifications essentially change the way the genetic code is read out without changing the DNA itself: they are therefore dubbed ‘epigenetic’ rather than genetic.
One of the best-understood epigenetic modifications is called methylation. A methyl group is a carbon with 3 hydrogen atoms attached. Addition of a methyl group to a part of DNA normally has the effect of switching off nearby genes. Studying DNA methylation is very interesting because it tells us how genes may be switched on and off in both health and disease.

Against this backdrop, a new Swedish study asked whether DNA methylation might be the link between smoking and MS. They looked at 2 cohorts from a huge case-control study of MS. First, they picked people who had a particular genetic signature known to interact with smoking: they were positive for HLA-DRB1*15:01 – the main risk gene – and negative for HLA-A*02 – the main protective gene. The second group were people with MS without any genetic criteria.

For each participant, they analysed the entire epigenome: the map of all methylated sites in the DNA. They then categorised people into three groups according to when they last smoked:
          Within 5 years
          Beyond 5 years
… and compared the epigenomes between these groups.

Interestingly, they found several positions in the genome at which the methylation status differed between never-smokers and recent smokers. There was no difference between never-smokers and ex-smokers who quite >5 years previously. Surprisingly, the effect of smoking on DNA methylation is greater in those with MS than those without. Whether this is due to synergistic effects of disease-modifying therapy or related to the underlying disease is not clear.

This study identifies a possible mechanistic link between smoking and the risk of developing MS. The authors prove that smoking leads to epigenetic changes in pwMS. It is therefore plausible that smoking has widespread effects on the expression of various genes which may facilitate the development of MS. The fact that these effects were only seen in the recent smokers supports the idea that smoking might promote MS through changes in gene expression, as this fits with the epidemiological data: the harms of smoking for MS dissipate 5 – 10 years after quitting. So, the timescale of the epigenetic modifications reported here is consistent with a causal link between smoking and MS.

While these data are very interesting, they do not really show how smoking exacerbates MS. Smoking influenced the methylation status of several genes, but only one of these was shown to be expressed differently as a downstream consequence. This gene was not obviously involved in the development of MS. It seems likely to me that smoking influences the expression of lots of genes; it also promotes lung inflammation, can introduce mutations into the underlying DNA, and promotes cardiovascular disease. The mechanism by which smoking influences MS is likely very complicated and is unlikely to be just due to a direct effect on gene expression. There are also important confounders like activity level, socio-economic status, and overall health which may influence smoking status and epigenetic modifications.

The bottom line is that this study shows one possible mechanism – epigenetic modification – through which smoking might influence the risk of MS.

Cigarette smoking is an established environmental risk factor for Multiple Sclerosis (MS), a chronic inflammatory and neurodegenerative disease, although a mechanistic basis remains largely unknown. We aimed at investigating how smoking affects blood DNA methylation in MS patients, by assaying genome-wide DNA methylation and comparing smokers, former smokers and never smokers in two Swedish cohorts, differing for known MS risk factors. Smoking affects DNA methylation genome-wide significantly, an exposure-response relationship exists and the time since smoking cessation affects methylation levels. The results also show that the changes were larger in the cohort bearing the major genetic risk factors for MS (female sex and HLA risk haplotypes). Furthermore, CpG sites mapping to genes with known genetic or functional role in the disease are differentially methylated by smoking. Modeling of the methylation levels for a CpG site in the AHRR gene indicates that MS modifies the effect of smoking on methylation changes, by significantly interacting with the effect of smoking load. Alongside, we report that the gene expression of AHRR increased in MS patients after smoking. Our results suggest that epigenetic modifications may reveal the link between a modifiable risk factor and the pathogenetic mechanisms.

About the author


  • Smoking among the general population is c.20% compared with c.60-70% in the 1930s. If smoking was a casual factor, MS would have been going through the roof in the 1930s and would be at much lower levels today (the opposite is true). I am the most anti-smoking person in the world (never smoked) yet got an RRMS diagnosis in my mid-30s. Bad mono in my mid- teens is the reason I got MS. Please don't waste any more time on research relating to smoking – lower number of smokers and higher levels of MS show it can't be a causal factor.

    • Good point about WWII saw huge amounts of stress and high levels of smoking. But lower levels of MS relapses and lower levels of MS.

      Now, as modern life marches on, MS rates are increasing (as are other autoimmune conditions).

    • Good point – if smoking were the only risk factor than you would expect rates of MS to be closely related to rates of smoking. Because it's one of many contributing factors the relationship isn't straigtforward

    • Between 1939 and 1945 a significant portion of the UK population in the age group 18 to 30 spent a lot of time outdoors and much further South. Also a lot never got old enough to find out if they had MS.

    • Be careful not to mix up incidence, with total numbers of people with a disease. Incidence is the number of new cases per year per fixed proportion of population. This can remain steady while the number of people with the disease increases, as medicine stops them dying. I cannot find any papers comparing incidence in the 1930s and now. Smoking is obviously not the only risk factor, but it is one.

  • Great post thanks Ben and so well explained I understood every word 🙂

    From someone who knows even less about genetics than she does about immunology 😉

  • So then it makes sense with other environmental factors too? There's nothing that smoking can do that other e.g. artificial sweeteners, chemically altered fats, can't do either.

    Ultimately these are toxins in the body that increases inflmmatory markers, no?

    Btw I love my Orange Fanta.

    • Smoking is particularly nasty because there are loads of different toxic chemicals in smoking, but you're right that epigenetic modifications are probably caused by lots of environmental toxins we're exposed to

  • Carbon monoxide affects the immune system and smoking increases the level of carbon monoxide in the blood. Another possible mechanism?

  • Is it still thought to be the case that exposure to cigarette smoke confers no additional MS risk to the risk already posed by EBV in infected individuals? I believe there was a post on this blog about this, and how it is therefore thought that smoking and EBV somehow act on the same pathway in terms of MS risk?

  • I did quite a bit of passive smoking in pubs and nightclubs before the UK smoking ban. I would dance away the hours in nightclubs most weekends and come out the club with my hair and clothes stinking of smoke, my throat would be sore.

  • Could it not also be that smoking adds to susceptibility? I remember a Spanish doctor was commenting on why lung cancer rates had been consistently higher in the UK than in Spain, where at the time just about EVERYBODY smoked. He said that the deadly combination was smoking while living in industrial areas.

    • Yes for sure smoking is only one risk factor of many – the increased risk of getting MS if you smoke vs if you don't is still quite small.

By DrBenJ



Recent Posts

Recent Comments