Table 4.

Coefficients and 95 % CI as derived from generalized estimating equation–linear regression analyses indicating the associations of absolute measures of exposure to food retailers with dietary patterns among adults in neighbourhoods from five urban regions in Europe, February–September 2014. The SPOTLIGHT Project (n 4942)

	Healthy dietary pattern* †			Less healthy dietary pattern† ‡
	Coefficient	95 % CI	P value	Coefficient	95 % CI	P value
Tertiles of density of healthier food retailers per km2 §
Lowest	1			1
Medium	−48·2	−233·6, 137·3	0·611	13·7	−98·1, 125·5	0·810
Highest	−10·5	−178·8, 157·8	0·903	−56·6	−163·0, 49·9	0·297
Tertiles of density of less healthy food retailers per km2 ║
Lowest	1			1
Medium	−78·1	−244·9, 88·8	0·359	−61·3	−176·3, 53·7	0·296
Highest	−23·3	−181·3, 122·7	0·706	−129·6	−224·3, −34·8	0·007
Tertiles of spatial access score for healthier food retailers§
Lowest	1			1
Medium	−77·2	−183·1, 28·7	0·153	−18·2	−115·5, 79·2	0·714
Highest	−99·7	−222·9, 23·5	0·113	−34·4	−133·8, 65·0	0·498
Tertiles of spatial access score for less healthy food retailers║
Lowest	1			1
Medium	−68·0	−164·9, 28·9	0·169	−6·67	−90·9, 77·6	0·876
Highest	−100·0	−224·3, 24·2	0·114	−9·00	−117·4, 99·4	0·871

Dietary patterns were obtained from principal component analysis. All models were adjusted for age, sex, educational attainment, household composition, urban region and self-selection variables.

^*Healthy dietary pattern is composed of fruits, vegetables and fish.

^†Scores for healthy and less healthy dietary patterns were multiplied by 1000.

^‡Less healthy dietary pattern is composed of fast foods, sweets and sweetened beverages.

^§Healthier food retailers: supermarkets and local shops;

^║Less healthy food retailers: fast-food restaurants, cafés/bars and convenience/liquor stores.