I'm sure there will be lots who disagree, but this is one of a few findings that seem linked to sleep/excess exposure artificial light to me. If you drink coffee, that very likely means that you drink it in the morning; and if you slug down coffee in the morning you're much less likely to binge watch anything on Netflix 'till 3AM, 'cause you'll be too exhausted by then. So you'll get your requisite darkness (and melatonin.) (The research really shows that long (predictably timed) stretches of real darkness is important for health, not sleep per se.)
Now, back in the day, a few decades ago, coffee was clearly unhealthy, because most people used it not only to wake up, but to stay up well into the evening, very frequently. But it became obvious this wasn't a healthy use (I suspect employers were not keen on it.) So the habitual drinking of coffee in the late evening went from being what everybody did to being what almost nobody did (or admitted to, anyway.)
Find a few people who drink coffee only or mostly in the evening - there must be some - and redo this study. I bet you get quite different results.
I don't discount the antioxidants in coffee, but I doubt that's what's primarily driving the OP study findings.
Now, back in the day, a few decades ago, coffee was clearly unhealthy, because most people used it not only to wake up, but to stay up well into the evening, very frequently. But it became obvious this wasn't a healthy use (I suspect employers were not keen on it.) So the habitual drinking of coffee in the late evening went from being what everybody did to being what almost nobody did (or admitted to, anyway.)
Find a few people who drink coffee only or mostly in the evening - there must be some - and redo this study. I bet you get quite different results.
I don't discount the antioxidants in coffee, but I doubt that's what's primarily driving the OP study findings.