The statistic can't be true because the study was conducted by a major milk manufacturer, they wouldn't want to have a negative outlook on their product since it could lead to a potential loss of market value.
Another reason why this statistic can't be trusted is that we don't know how many or groups of teens were surveyed. They could have gone to their faithful customers for all we know, and/or it could have been more people of one gender to the other or lack of diversity overall.
They could have tried to appeal to parents and teens to drink more milk and that it's healthy to make them buy more of it, lots of companies do that.