I feel like this warrants a longer, better supported post, but it seems like the meaning of "lonely" is extremely (overly) relative. Since a lot of studies hinge on self-reporting loneliness this has a huge impact on the data available to diagnose long term trends.

Two important things that get conflated into "loneliness" is social isolation (the quality/depth of social relationships) and "loneliness" (the quantity of social relationships).