I would assume you are correct. Judging by my personal observations I would assume that active cyclists are less likely to be smokers and/or heavy drinkers. I also wonder if the researchers recorded long term body weights and other habits/behaviors known to be catalysts if not causes of ill health of their sample subjects, and if the active cyclists had similar overall health profiles as the people in the non-active sample population.

I also assume that unhealthy people and people prone to known unhealthy habits like smoking, hard drug use and excessive drinking do not normally become active cyclists; or remain active cyclists for long.

Perhaps someone who has delved into this study has more info on the "active cyclist" sample and if it included cyclists who were not recreational or club cyclists, but rather people who are actively cycling because they no longer have the financial means to get around any other way, i.e people down on their luck, homeless and/or had their license revoked due to alcohol or drug abuse.