Is it necessary to create a constraint on a “secondary label” on Neo4j?

Question

I am parsing some CSV files to import that data on my Neo4j Enterprise single-instance database, and I set the NODE KEY type constraints on every node label so it is sped up as