Researchers who studied older adults’ internet use over time expected to find what most people assume: that the heaviest scrollers in the evening would turn out to be the loneliest. A 2025 study in Scientific Reports, using longitudinal survey data on adults over 60, found close to the opposite. Higher social internet use predicted lower loneliness later on, not higher, and the effect held up even after accounting for how many people were in a person’s regular contact network.
What the data actually showed
Jeroen Janssen and colleagues tracked older adults’ internet use, the size of their contact networks, and their loneliness across multiple survey waves. They expected social internet use to reduce loneliness mainly by helping people maintain more contacts, and built their analysis around that expectation. The mediation did not show up. Instead, the researchers found something closer to substitution at the margins: as in-person contact naturally decreases with age, for reasons ranging from mobility limits to friends moving away or dying, social internet use fills part of that gap rather than adding a new one on top of it. The number of people someone stays in contact with holds roughly steady across the transition. The channel through which that contact happens shifts.
That distinction matters for evening scrolling specifically. An older adult browsing photos, messages, or updates from family late in the day is not, on this evidence, showing a symptom of isolation. They are more often maintaining the same web of contact they always had, through a channel that has become more available to them than the alternative, at an hour when calling feels like more of an imposition than it once did.
Why scrolling and calling are not interchangeable
None of this means scrolling and a phone call do the same job. Research on older adults and communication has repeatedly found that direct contact, in person or by phone, carries a more reliable loneliness benefit than passive browsing does. AARP’s 2026 survey of technology use among adults 50 and older found that nine in ten now use social media regularly, and that texting has overtaken calling as the single most common way this group stays in touch. Scrolling fills the space a call used to occupy at a lower cost in effort and a lower risk of an answer nobody wants, silence, a bad connection, a conversation neither person has the energy for that night, rather than replacing the call because it serves the relationship better.
A 2009 study on what researchers call social surrogacy helps explain why the lower-effort option gets chosen so often. Jaye Derrick and colleagues found that engaging with familiar, favored media can measurably reduce loneliness in the moment and buffer against rejection, even though it involves no real exchange with another person. The original research focused on television rather than social feeds, but the same mechanism plausibly extends to an evening spent scrolling through people one already knows: it delivers a real, if partial, sense of connection without requiring either person to be available, articulate, or in the right mood at the same time.
The call that does not happen
What the research does not resolve is why the call gets skipped in favor of the scroll on a given night. Calling someone directly carries a small but real social cost that scrolling does not: the risk of interrupting them, of asking for their attention at a moment they cannot easily spare, of a conversation that reveals more distance than either person wants to feel right then. Scrolling carries none of that risk. It can be started and stopped without anyone else knowing it happened, and it asks nothing of the person on the other end of the feed. For someone who is not sure their call would land well tonight, or who has simply had a full-enough day already, that difference is what tips the choice toward the scroll, not a lack of people to call.
What this means for anyone writing to or about an older audience
Content aimed at older readers often carries an unspoken assumption that heavy phone or tablet use signals isolation, and that the fix is getting that reader to call someone instead. The research suggests a more accurate frame: for a meaningful share of older adults, the scrolling is already doing real relational work, just through a channel that asks less of both parties involved. Writing that treats every older reader’s screen time as a problem to be solved misreads what the data shows. Writing that takes seriously why the lower-cost option got chosen, and what would need to be true for a direct call to feel safer or easier by comparison, has more to say that is actually new.
That gap does not close through exhortation. Telling someone to call more, framed as a moral upgrade over scrolling, ignores the reason the scroll won out in the first place: it was already available, already low-risk, and already doing part of the job. The more useful question, for anyone covering this territory, is what would make the direct alternative feel as easy to reach for on a given night as the feed already is.
