For example, it is often said that cancer is the second leading cause of death in most modern countries. But it is easy to make it first: Divide up the “cardiovascular disease” category that is typically the one item above it in the list into stroke, artery disease, MI, etc. Do that right and the arbitrary category “cancer” rises to the top of the list of arbitrary categories. If you want to move suicide to the top of the list it will take a lot more work (and some of the division will probably seem pretty silly, like dividing the other based on the exact anatomical point where something happened), but it can be done. Moving something to “second leading killer” is much easier, just create a category that includes most everything else and chop up what remains. Making something fifth is easy. And I trust that at this point it goes without saying that something being fifth rather than sixth means nothing at all.
The standard categorizations clearly have some misleading arbitrariness to them. Infectious diseases are typically divided up into many specific infections which end up relatively low on the list, while various cancers or heart diseases (many of which have infectious origins) are bundled together. If you want to get “infectious disease” above “heart disease” I suspect you could pull it off by choosing the right population (world, not American) and by using an aggressive enough definition of infectious disease (staring with the obvious like malaria, adding in various cancers, all deaths of the elderly who have pneumonia or any infection weakening them, all infant deaths, etc.).
Just because the “position on the list” statements are largely meaningless does not mean that direct explicit comparisons are not useful. It is meaningful to observe “heart disease kills more women than breast cancer.” It is just the “most” (and even more so, “second most”) that is arbitrary.
As soon as you add qualifiers and orthogonal categories the claims become even more arbitrary. Anti-smoking activists are fond of observing that smoking is “the leading preventable cause of death”. (They often try to mislead people about their opportunity to employ harm reduction by saying “tobacco” instead of “smoking”, but that is another story.) What is this “preventable” of which they speak? Undoubtedly a time will come when most every currently deadly infectious disease can be cured as well as most cancers and even cellular transcription errors. They are all perfectly preventable, it is just that no one knows how to prevent them yet – just like smoking. I suspect that most of the people who repeat that claim do not even engage their brains to think about what they are saying when they type “preventable”. If they do, they probably are thinking something like “if I was made king right now, I could eliminate smoking, but I could not eliminate pneumonia”. That kind of makes sense, though it is obviously wrong – just ask the king of Bhutan.
You might have noticed another complication here, the orthogonality bit (meaning categories that run, in some sense, perpendicular to each other). Smoking and pneumonia are not really the same type of category, even though both are preventable causes of death. You can meaningfully say “smoking causes more deaths than pneumonia” if you want, but if they are somehow entries in the same list, the superlative game starts to look even less sensible.
Of course, if anti-smoking activists said what they really mean – that among all the choices people make, the choice to smoke causes more deaths than any other – then the comparisons start to make a little more sense. “Choices people make” (behaviors) is well defined, while “preventable” is not. But the activists work hard to avoid saying this, going so far as to make up a concept like “preventable” and thus make their claim fall somewhere between meaningless and flat-out wrong, because their mythology and activist approach requires that smoking be considered a disease not a choice people make. This example is more obvious than most, but it points out that most of the “second most deadly” type claims are motivated by activist goals not an effort to aid understanding.
Coming back to the original point, smoking is only superlative among choices people make because the other implicit entries on the list are chopped up fine enough. If we include the category “choice of what to eat”, I am pretty sure it would edge out smoking now, with its combination of overweight, unhealthy foods, and not getting the best combination of micronutrients (if only we knew what that was). If that did not beat out smoking, expanding the category to include infant feeding, and the world’s starvation and contaminated food would certainly do so. We might think that the fact that the “smoking is the leading…” claim is meaningless as typically phrased, and that it serves no apparent purpose and is pointless waste of ink anyway, when added to the fact that it is not true if the other entries on the list are chosen correctly, might discourage people from writing it. After all, almost-meaningless arguably-incorrect euphemistic claims are the leading preventable cause of over-zealous activists being ridiculed – so long as you chop up the other reasons into small enough bits.