It seems to be based on time alone. From the article:
So that explains where the 1 year comes from. Of course, it should be considered a minimum, not "best practice." The longer you can stay RF, the safer you are - even adults are safer RF (good to know if you're flying Southwest or taking a train
).
I'm not sure where the "and 20 lb" part comes from. I'll see what I can find.
Turn-Around Time
There are no magical or visible signals to tell us, parents or pediatricians when the risk of facing forward in a crash is sufficiently low to warrant the change, and, when a parent drives around for months or years without a serious crash, the positive feedback that the system they have chosen "works" is very difficult to overcome. When in doubt, however, it's always better to keep the child facing rearward.
In the research and accident review(2) that I did a few years ago, the data seemed to break at about 12 months between severe consequences and more moderate consequences for the admittedly rare events of injury to young children facing forward that we were able to identify. One year old is also a nice benchmark, and the shift to that benchmark in the last few years has kept many kids in a safer environment longer and has probably saved some lives, some kids from paralysis and some parents from terrible grief.
So that explains where the 1 year comes from. Of course, it should be considered a minimum, not "best practice." The longer you can stay RF, the safer you are - even adults are safer RF (good to know if you're flying Southwest or taking a train

I'm not sure where the "and 20 lb" part comes from. I'll see what I can find.