Table 3.

Regression coefficients (β) and 95% confidence intervals (95% CI) resulting from linear regression analyses with network distance to closest food retailers as exposure measure and the two outcomes: percentage of consumption in grams from ultra-processed food (UPFs) and percentage of consumption in kilocalories from UPFs (n = 8104)

Closest food retailers within a range of	Percentage of consumption from UPFs in
	Grams	Kilocalories
	β (95% CI)	β (95% CI)
Fast-food restaurant
> 1500 m	Ref.a	Ref.a
1001–1500 m	− 0.3 (− 1.4; 0.8)	− 1.4 (− 2.8; − 0.0)
501–1000 m	− 0.5 (− 1.4; 0.4)	− 0.6 (− 1.7; 0.5)
0–500 m	− 0.3 (− 1.2; 0.6)	− 0.8 (− 1.9; 0.4)
Convenience stores
> 1500 m	Ref.	Ref.
1001–1500 m	0.2 (− 0.5; 0.8)	− 0.1 (− 0.8; 0.7)
501–1000 m	− 0.4 (− 0.9; 0.2)	− 0.8 (− 1.5; − 0.0)
0–500 m	− 0.4 (− 1.1; 0.2)	− 1.1 (− 2.0; − 0.3)
Restaurants
> 1500 m	Ref.a	Ref.a
1001–1500 m	− 0.1 (− 1.0; 0.7)	− 0.5 (− 1.6; 0.6)
501–1000 m	− 0.8 (− 1.5; 0.0)	− 0.6 (− 1.5; 0.4)
0–500 m	− 1.6 (− 2.4; − 0.8)	− 1.6 (− 2.6; − 0.6)
Supermarket
> 1500 m	Ref.	Ref.
1001–1500 m	− 1.2 (− 2.2; − 0.3)	− 1.1 (− 2.3; 0.1)
501–1000 m	− 1.8 (− 2.7; − 0.9)	− 1.8 (− 2.8; − 0.7)
0–500 m	− 2.1 (− 3.0; − 1.2)	− 2.2 (− 3.3; − 1.1)
Candy stores and cafés
> 1500 m	Ref.a	Ref.a
1001–1500 m	0.1 (− 0.5; 0.7)	0.4 (− 0.3; 1.2)
501–1000 m	− 0.3 (− 0.8; 0.3)	− 0.3 (− 1.0; 0.3)
0–500 m	− 1.2 (− 1.9; − 0.6)	− 1.3 (− 2.1; − 0.4)

Coefficients were adjusted for age, sex, region of residency, educational attainment, urbanisation, marital status, total kilocalorie intake, and proximity to local food shops. Effect modification was tested in separate models by adding an interaction term between urbanisation and each of the proximity measure

^aIndicates significant effect modification by urbanisation levels (p < 0.05)