It's a cohort study and they simply look at trends from a population of people. Tea is common in China and it could be that people who drink more tea drink less alcohol and/or sugar sweetened beverages.

That's the tricky part about cohorts. As the saying goes, correlation does not imply causation. People who eat healthier tend to exercise more and smoke less. All contribute to health but single studies can't tell us differing weights and effects.

Perhaps tea has some unique benefits or maybe it's a coincidence and representative of other lifestyle choices.