For those of you who want to do your own analysis, assuming their data are authentic, take a look at their Table I:
http://ascopubs.org/na101/home/literatum/publisher/asco/jour...
Out of those 60 non-smokers with lung cancer apparently just 20 were men (out of 14,208 study participants who were men and non-smokers => 0.14%)
http://ascopubs.org/na101/home/literatum/publisher/asco/jour...
-----
Curiously this also means non-smoking women have 1.28x higher chance to get lung cancer than non-smoking men (unrelated to vitamins).
-----
For never-smokers vs smokers men risk ratios are then (again unrelated to vitamins):
- current smoker: 31.7x higher risk
- recent smoker (stopped < 10 years ago): 21.3x higher risk
- past smoker (stopped > 10 years ago): 8.9x higher risk
-----
Take home message: stop smoking!
Precautionary principle: if you take B vitamin supplements, you can continue (for slight decrease of cancer risk), just make sure you aren't taking mega-doses (especially if you still smoke), but really - stop smoking - the sooner you do less cumulative harm you get.
Researchers themselves excluded never-smokers from more detailed stratified analysis because of their small number.
There were only 60 cases of lung cancer in never-smoker category (out of 36,381 people => 0.16%).
In comparison there were 748 cases of lung cancer in smoker categories (former + recent + current) (out of 40,737 people => 1.84%).
So there is 11.5x higher risk of lung cancer simply by smoking (including people who stopped smoking).
For comparison the highest hazard ratio from this study was 3.71x for a category of current smoker taking >55ug/d B12 vs current smoker who is non-user of B12.
-------
Also curiously, the worse risks seem to be associated with people who stopped taking vitamins than people who currently use vitamins (B6: 1.97x vs 1.38x, B9: 1.65x vs 1.05x, B12: 2.58x vs 1.19x - individual supplement use status former vs current).
Plus smaller doses of B6/B9/B12 shown in this study to pretty much universally lower lung cancer risk by a bit (hazard ratios of 0.8-0.9x ranges; one noticeable outlier >600ug/d B9 in recent smokers halving risk of cancer).