In the fully adjusted models restricted to deprived tracts, the street tree count showed a marginally significant inverse association with the odds of preterm birth within a 100 m buffer ( Table 6 ) and fully significant inverse associations in the models for 250 m and 500 m buffers.
1
—
—