There’s always the chance that an age in any record is incorrect. If I’m looking at a census enumeration for someone who is a little bit older and their age ends in a 0 or a 5, I look to see how common it is for others who are a little bit older to have ages that end in a 0 or a 5.
If the ages for “those of a certain age” are entered accurately, one-fifth of them should end in a 0 or a 5. If it is significantly more than that there is the possibility that the ages have been approximated.
Just a possibility. Really determining that the ages for “those of a certain” were approximated requires a little more analysis than looking at the ages on a few pages.