Uhhh.... How on earth could pouring water in a hole increase cancer twenty miles away?
It doesn't. So what you do is look at the numbers in various ways - starting your comparative analysis from different years... Or looking only at this demographic (age, race, sex, etc.) or restricting your sources to this county or that county...
You keep looking and manipulating the numbers until you find a set/scheme that shows an uptick. Then you do something like instead of reporting a 2% rate vs a 1% rate you tout it as "100% INCREASE IN THE RISK!!!" You bury the fact that your error bars on the analysis far exceed the delta. You then build your whole "study" and analysis methodology to come to the conclusion you've already decided upon.
The anxiety caused by ignorance and stupidity raises stress hormone levels...