DISCUSSION
Our investigation of the spatial variations in mental
substance-related disorders in Malmo¨, Sweden, led us to
compare a spatial analytical perspective with the usual
multilevel approach. We found that the notion of space
used during analysis conditioned the information we could
obtain on the contextual dimension of mental disorders.
Building a notion of space into statistical models, we were
able to show not only important neighborhood variations
in substance-related disorders but also that the spatial scale
of variations was much larger than that of administrative