Methodological point on mediation analysis

Recently, King et al.1 assessed the effect of disability on mental health through bullying in adolescents using a causal mediation analysis method developed by Valeri and VanderWeele2 in which exposure–mediator interactions are allowed. In the latter method, one can only handle time-fixed exposure and mediator, and, for this reason, the authors only used the exposure in wave 5 and mediator (and outcome) in wave 6