The comma in the first sentence marks the ger-ppl clause as a non-restrictive supplement, a new predication outside the predication in the main clause. The clause does not 'modify' anything: it comments on the main clause, equivalent to saying
... one in five people have a sleep disorder of some kind, and having a sleep disorder affects their ability to get enough sleep.
If there were no comma the ger-ppl clause would be understood to modify some kind. In this case it would taken as restrictive: the sentence would not be talking about all sleep disorders but only those which affect the ability to get enough sleep. This is a different matter: it leaves open the possibility that some people have a sleep disorder of a kind which does not affect the ability to get enough sleep.
Likewise the absence of a comma in the second sentence marks the ger-ppl clause as a restrictive modifier on people: the quantity which is said to be increasing is not the number of all people but of the number of [people with malnutrition].