Why "linked to" doesn't mean "caused by"

"Coffee linked to longer life." "Screen time linked to poor sleep." These headlines are everywhere, and they share a quiet trap: linked to means the two things tend to show up together, not that one causes the other. An association is a starting point for a question, not an answer to it.

The reason is that a third factor can be driving both. People who drink a lot of coffee might also, on average, be wealthier, more social, or more active, and any of those could be the real reason they live longer. That hidden third factor is called a confounder, and it is the usual culprit behind a correlation that looks like cause. Ice cream sales and drowning both rise in summer, but ice cream does not cause drowning. Hot weather drives both.

This is why the phrasing of a headline matters so much. Words like linked to, associated with, and tied to are honest signals that researchers found a pattern, not a cause. The strongest way to actually show cause is a randomized controlled trial, where people are randomly assigned to one group or another so the groups are otherwise similar. Most "linked to" studies are not that kind of study.

So when you read that something is linked to a good or bad outcome, resist the leap to "so it causes it." Ask whether it was an experiment that assigned people to groups, or just an observation that two things traveled together. If it is the second kind, treat it as a lead worth watching, not a fact to act on.