People could just be systematically acting in a sub-optimal fashion, for instance

Ways this could be the case:

• they don't know about those other health-improving behaviours, or they misunderstand their effectiveness (or they believe they're effective but they lack the corresponding alief);
• they weigh their costs more heavily than you do (my grandma feels that if she couldn't eat as much as she does, she would feel deprived of one of the main pleasures of life, whereas that would hardly bother me);
• they correctly estimate their cost:effectiveness ratio, but they fail to adhere to them due to akrasia.

I'd guess that these effects are more relevant than signalling for a two-digit percentage of the people who buy organic food but don't do other health-benefitting behaviours.