I forget the details, but a different study in recent years found that decaf and caffeinated did have different links. IIRC it was measuring both cancer rates and heart disease, and found positive correlation with both reduced for caffeinated coffee but only one for decaffeinated. And years before that, a study that found positive correlations with caffeinated coffee, but not straight caffeine, and not decaffeinated coffee, which indicated some sort of interaction with the coffee compounds and the caffeine in particular (very fuzzy on this one).