Here's a bit from James Hohmann's Washington Post article about the study:
Shannon M. Monnat, an assistant professor of rural sociology and demography, created a data set with numbers from 3,106 counties. She found this trend to be true nationally but especially so in two regions: In the industrial Midwest, which is how academics refer to the Rust Belt, Trump ran ahead of Romney by an average of 16.7 percent in the quarter of counties with the highest mortality, compared to 8.1 percent in the lowest quartile. In New England, Trump did worse than Romney by an average of 3.1 percent in the lowest mortality counties but better than the former Massachusetts governor by an average of 10 percent in the highest mortality counties.
I saw this firsthand on the campaign trail all year, in countless interviews with folks who were down in the dumps and struggling to get ahead (or, quite frankly, just get by). Many supported Barack Obama eight years ago because they were desperate for hope and change. They’re still desperate, and now they’re hopeful Trump can bring the change they’re looking for.