The results of our controlled trial suggest that smoking is to be included in the growing list of known factors influencing the composition of the intestinal microbiota. We found evidence with two independent methodological approaches that smoking cessation induces a substantial modification of the intestinal microbial composition with no indication of any concomitant alteration in nutrition. A variety of different analyses of the high-throughput sequencing data indicated differences before and after smoking cessation, suggestive for profound and robust microbial shifts. On the phylum level strikingly even alterations regarding the major four phyla were identified in the individuals undergoing smoking cessation (, Figure S4
). On the level of OTUs quantitative changes (relative abundance of some bacterial lineages in the complex microbial mixture) are indicated by alterations in the weighted UniFrac mode, while the similar findings in the unweighted analyses in addition point to qualitative changes (i.e. occurrence of new OTUs after smoking cessation and disappearance of former OTUs prior to the intervention).
Furthermore, aside from pure microbial shifts an increase in microbial plurality is suggested by the increase in α-diversity after smoking cessation (, Figure S9
). However, due to the limited observation period after smoking cessation, it is not possible to draw any conclusions on a potential sustained long-term increase of α-diversity due to smoking cessation. A long-term effect towards an increased microbial richness rather appears doubtful, as the initial increase in diversity somewhat mitigated after 8 weeks in all our analyses of α-diversity. Also, there were no substantial differences in α-diversity between smokers and non-smokers in the control groups. Interestingly, in contrast to our findings, oropharyngeal microbial diversity was found to be increased in smokers compared to non-smokers 
. Thus, to clarify this issue further studies on the long-term effect of smoking on α-diversity of the intestinal microbiota are necessary.
We observed similar shifts in microbial composition on the phylum level after smoking cessation as described in obese versus lean humans 
and mice 
(a lower proportion of Bacteroidetes
and a higher proportion of Firmicutes
), while a concomitant average weight gain of 2.2 kg during the whole observational period occurred. However, neither a change in total daily calorie intake nor nutritional components was suggested owing to analyses of the food-frequency questionnaires. Thus, the common perception that a change in dietary habits is the key factor for weight gain after smoking cessation is challenged. Our findings may suggest a potential role of an altered composition of the intestinal microbiome and hence metabolome with subsequent modifications in gut microbiota's metabolic function in the pathogenesis of weight gain after smoking cessation.
The above mentioned disordered oropharyngeal and tracheal microbial communities in smokers 
appears to be intuitively plausible due to the direct contact of cigarette smoke with the respiratory epithelial cells and mucus harboring the local microbiota. Possible pathophysiologic links explaining the influence of smoking and smoking cessation on the intestinal microbial composition are much less evident. However, our study was not designed to investigate potential causative mechanisms. In addition, due to the study design it was not possible to investigate whether there are any loco-regional differences such as for instance between the terminal ileum and recto-sigmoid. This would be of interest with regard to the discrepancy of smoking between UC and CD.
Another limitation of our study is that despite the controlled study setup to largely rule out external influences such as for instance medication intake (e.g. antibiotics or supportive drugs for smoking cessation) or non-adherence to strict smoking cessation, we cannot completely ascertain lack of confounding effects such as exercise or subtle changes in dietary habits unrecognized by analysis of food-frequency protocols. However, the following considerations argue against a major contributing role of hidden dietary alterations. First, the assumption that an alteration of diet and increase of calorie intake would explain the observed weight gain is not based on robust evidence, as indications that increased calorie intake is the principal explanation for weight gain after smoking cessation are at best conflicting. In contrast, there are important studies suggesting alternative mechanisms for weight gain, such as a decreased metabolic resting rate 
or changes in metabolic properties in the adipose tissue, in specific an increase of the lipoprotein lipase activity 
. While a smaller study found no increase in calorie intake after smoking cessation in subjects gaining weight 
, an indicator for alternative mechanisms for the well-known effect of weight gain 
are the results of the multiple risk factor intervention trial 
; one of the largest studies ever performed in this field. This trial included more than 11000 men with an observational period of 6 years, thoroughly investigating weight gain and calorie intake in nonsmokers, quitters, recidivists, and continued smokers in a usual care group in comparison to a special intervention group with specific counseling. Interestingly, despite a reduced calorie intake and a generally healthier diet after counseling compared to baseline, and an even bigger reduction in calories compared to continuing smokers, men who ceased smoking gained weight. This was strikingly in contradiction to nonsmokers, continuing smokers, and even recidivists, who all lost weight. Second, although alterations in intestinal microbiota composition due to diet are well known, profound shifts seem to necessitate a longer time period as a recent human study found significant changes on the phylum level to occur predominantly after 12 weeks 
. In contrast no stable switches between enterotype groups occurred in the short-term, despite a controlled and randomized short-term feeding experiment using two extremely divergent diets (high-fat/low-fiber versus low-fat/high-fiber) 
. Third, a uniform shift in the phyla composition such as the one observed in our patients undergoing smoking cessation (Figure S4
) indeed might be assumed to occur in interventions with prespecified and standardized major changes of diet. But this would less likely to be expected in individually different and uncontrolled slight dietary changes undetected by food-frequency protocols.
Although we identified continuous microbial shifts 8 weeks after smoking cessation, this relatively short observational period does not permit conclusions about a long-term alteration. Nevertheless, the continuing increase in body mass observed in subsequent 4 years 
after the first year after smoking cessation (where the major part of the weight gain occurs) points to a potential sustained microbial alteration. Larger prospective long-term studies are needed to address this issue.
Although the role of cigarette smoking as one of the most important environmental risk factor in the pathogenesis of IBD is well established 
, the numerous epidemiological reports on smoking status and course of IBD are in sharp contrast with the very limited mechanistic explorations of the effect of smoking cigarettes on the gut and intestinal inflammation. The molecular and cellular mechanisms by which smoking interferes with the pathogenesis of CD and UC are only poorly understood. While several potential mechanisms, such as modulation of mucosal immune responses, alterations in intestinal cytokine and eicosanoid levels or modifications in gut permeability have been proposed, none of these hypotheses could hitherto offer a satisfying explanation 
. As the importance of the intestinal microbiota in the pathogenesis of IBD is well established, the profound shifts in the composition of the intestinal microbiota observed in our study may shed some light to a potential pathogenetic link of cigarette smoking, smoking cessation, and course of disease in IBD, including the discrepancy in CD and UC. However, as we did not investigate the intestinal microbial evolution during smoking cessation specifically in patients with IBD, this potential pathogenetic link largely remains a speculative deduction and our findings need to be confirmed in CD and UC patients. We nevertheless believe, that the results of our observational study, demonstrating intestinal microbial shifts after smoking cessation for the first time, point to a potential interaction of disease-modifying factors in IBD. This may give us a further piece of information with regard to an enormously complex interplay, where we clearly are just at the beginning of our understanding – the interplay of the intestinal microbiota, chronic inflammatory bowel disease, systemic metabolic interactions and environmental factors (such as smoking).
In conclusion, smoking cessation induces profound changes in intestinal microbial composition. Interestingly, these changes appear to occur in a comparable manner both microbiologically and clinically (weight gain) as those induced by transplantation of an ‘obese microbiota’ into lean mice 
. These alterations in conjunction with weight gain point to further evidence for the assumed role of the intestinal microbiota and its metabolic properties derived to the host in the pathogenesis of obesity and the metabolic syndrome. Our findings invite confirmatory larger and long-term studies as well as further research on the effect of environmental factors on intestinal microbial composition.