Fig. 1

The probability of insomnia depending on hazardous alcohol use and levels of mental distress (N = 16 529). The moderator ‘mental distress’ was defined according to the cut-offs used when probing the interaction and identified as the regions of significance using the Johnson-Neyman technique. Using a non-centered mental distress variable these regions correspond to the following cut off values for mental distress, i.e. up to 1 (n = 5 494), from 1.1–1.20 (n = 4 594), from 1.21–1.60 (n = 4 265), from 1.61 up to 1.90 (n = 1 137) and 1.91 and above (n = 1 039)