Practice of reporting social characteristics when describing representativeness of epidemiological cohort studies – A rationale for an intersectional perspective