The important finding is
After adjustment for caffeinated and decaffeinated coffee intake amounts, sleep hours, and other confounders, the morning-type pattern, rather than the all-day-type pattern, was significantly associated with lower risks of all-cause (hazard ratio: .84; 95% confidential interval: .74–.95) and cardiovascular disease-specific (hazard ratio: .69; 95% confidential interval: .55–.87) mortality as compared with non-coffee drinking.
So people who drank coffee ONLY in the morning had significant advantages compared to the non-coffee drinking population. All-day drinkers did not have significant advantages.