Table 3.
Unadjusted and adjusted estimates of the associations between walkshed food environment characteristics, covariates and diet quality (C-HEI score)
| Unadjusted+ | Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | Model 6 | |
|---|---|---|---|---|---|---|---|
| Food destination presence a | Food destination densityb | Food destination diversityc | Supermarket/ grocery store presenced | Convenience/multi-product store presencee | Restaurant presencef | ||
| β (95 % CI) | β (95 % CI) | β (95 % CI) | β (95 % CI) | β (95 % CI) | β (95 % CI) | β (95 % CI) | |
| Intercept (β0) | 58.15 (50.75, 65.54) | 57.56 (50.22, 64.90) | 57.58 (50.20, 64.96) | 57.49 (50.08, 64.91) | 57.67 (50.27, 65.05) | 57.90 (50.51, 65.29) | |
| Unadjusted+ β (95 % CI) | −0.98 (−3.23, 1.26) | 0.05 (−0.01, 0.10) | 0.04 (−0.09, 0.17) | −1.05 (−4.72, 2.63) | 0.35 (−2.40, 3.10) | −0.62 (3.00, 7.80) | |
| Adjusted~ β (95 % CI) | −1.32 (−3.58, 0.94) | 0.06 (0.01, 0.12)* | 0.06 (−0.08, 0.19) | −0.86 (−4.50, 2.77) | 0.44 (−2.37, 3.25) | −0.71 (−3.11, 1.68) | |
| Covariates | |||||||
| Sex^ | |||||||
| Women | 4.77 (2.76, 6.78)# | 4.65 (2.59, 6.71)* | 4.84 (2.78, 6.90)* | 4.73 (2.66, 6.79)* | 4.65 (2.59, 6.71)* | 4.69 (2.63, 6.78)* | 4.66 (2.60, 6.72)* |
| Age^ | |||||||
| 40–59 years | −1.35 (−4.03, 1.32) | −1.66 (−4.28, 0.96) | −1.32 (−3.91, 1.27) | −1.34 (−3,95, 1.27) | −1.46 (−4.07, 1.32) | −1.37 (−4.06, 1.25) | −1.54 (−4.15, 1.07) |
| ≥ 60 years | −1.49 (−4.28, 1.30) | −0.37 (−3.60, 2.85) | −0.26 (−3.46, 2.94) | −0.11 (−3.32, 3.10) | −0.13 (−3.34, 3.09) | −0.11 (−3.35, 3.14) | −0.26 (−3.48, 2.96) |
| Race^ | |||||||
| All other races | 1.05 (−3.01, 5.11) | 2.96 (−1.04, 6.96) | 3.30 (−0.68, 7.23) | 3.26 (−0.74, 7.26) | 3.09 (−0.91, 7.08) | 3.19 (−0.81,7.19) | 3.05 (−7.06, 0.95) |
| Marital status^ | |||||||
| Others | −3.00 (−5.41, −0.57) | −1.34 (−3.90, 1.21) | −1.50 (−4.06, 1.05) | −1.41 (−3.97, 1.16) | −1.23 (−3.80, 1.35) | −1.34 (−3.90,1.22) | −1.31 (−3.89, 1.25) |
| Dependents at home^ | |||||||
| ≥ 1 dependent | 2.29 (0.30, 4.28) | 1.90 (−0.27, 4.07) | 2.11 (−0.06, 4.27) | 2.03 (−0.14, 4.21) | 1.99 (−0.18, 4.16) | 2.00 (−0.18, 4.20) | 1.93 (−0.24, 4.11) |
| Highest education completed^ | |||||||
| College/trade/diploma | 3.38 (−0.30, 7.05) | 1.81 (0.03, 7.15)* | 3.31 (−0.24, 6.85) | 3.39 (−0.17, 6.98) | 3.50 (−0.07, 7.08) | 3.47 (−0.09, 7.03) | 3.55 (−0.01, 7.11) |
| University | 4.29 (1.24, 7.34) | 1.58 (0.16, 6.36)* | 2.69 (−0.41, 5.80) | 2.89 (−0.23, 6.01) | 3.14 (0.04, 6.24)* | 3.01 (−0.09, 6.12) | 3.19 (0.08, 6.31)* |
| Total gross household income^ | |||||||
| $60 000–$119 999 | 0.42 (−3.36, 4.21) | −0.36 (−4.14, 3.42) | −0.50 (−4.26, 3.28) | −0.37 (−4.15, 3.40) | −0.26 (−4.05, 3.53) | −0.37 (−4.16, 3.42) | −0.34 (−0.41, 3.44) |
| $ ≥ 120 000 | 2.63 (−1.07, 6.33) | 1.26 (−2.74, 3.42) | 1.02 (−2.98, 5.01) | 1.18 (−2.83, 5.19) | 1.47 (−2.56, 5.51) | 1.22 (−2.79, 5.23) | 1.27 (−2.73, 5.28) |
| Refused | 0.44 (−3.91, 4.73) | −1.07 (−2.73, 5.27) | −1.75 (−6.04, 2.56) | −1.36 (−5.65, 2.94) | −0.97 (−5.28, 3.33) | −1.25 (−5.54, 3.04) | −1.16 (−5.45, 3.12) |
| Smoking status in past 12 months^ | |||||||
| Non-smoker | 7.86 (2.52,13.20)# | 8.90 (3.76, 14.04)* | 9.14 (4.02, 14.26)* | 9.10 (3.96, 14.23)* | 9.15 (4.00, 14.29)* | 9.14 (3.98, 14.31)* | 9.02 (3.89, 14.21) |
| Car available for personal use^ | |||||||
| Never or sometimes | −0.45 (−4.00, 3.10) | −1.60 (−5.01, 1.82) | −1.27 (−4.68, 2.15) | −1.42 (−4.84, 1.99) | −1.46 (−4.87, 1.96) | 1.47 (1.96, 4.87) | 1.50 (−1.91, 4.91) |
| Dog ownership in past 12 months^ | |||||||
| Non-owner | −1.70 (−3.80, 0.40) | −2.45 (−4.50,-0.41)* | −2.25 (−4.29, −0.20)* | −2.36 (−4.40, −0.32)* | −2.47 (−4.52 -0.43)* | −2.41 (−4.45, −0.36) | −2.41 (−4.45,-0.36)* |
| Self-reported mental health^ | |||||||
| Very Good | −1.17 (−3.64, 1.29) | −3.50 (−6.04, −0.95)* | −3.15 (−5.67, −0.62)* | −3.20 (−5.73, −0.66)* | −3.24 (−5.78, −0.70)* | −3.25 (−5.80, −0.69)* | −3.37 (−5.91, 0.83)* |
| Excellent | 0.76 (−1.92, 3.43) | −1.18 (−4.17, 1.80) | −0.91 (−3.87, 2.06) | −0.97 (−3.95, 2.01) | −0.86 (−3.86, 2.14) | −0.99 (−4.00, 2.00) | −1.08 (−4.07, 1.89) |
| Self-reported physical health^ | |||||||
| Very Good | 3.62 (1.39, 5.83) | 3.59 (1.24, 5.93)* | 3.44 (1.10, 5.78)* | 3.52 (1.17, 5.86)* | 3.44 (1.10, 5.80)* | 3.52 (1.17, 5.86)* | 3.54 (1.20, 5.89)* |
| Excellent | 3.51 (0.80, 6.22) | 2.23 (−0.88, 5.34) | 1.90 (−1.20, 5.00) | 2.07 (−1.03, 5.19) | 2.14 (−0.97, 5.24) | 2.12 (−0.98, 5.24) | 2.21 (−0.90, 5.33) |
| Sitting hours/day | −0.32 (−0.57, −0.68) | −0.14 (−0.41, 0.12) | −0.13 (−0.39, 0.14) | −0.14 (−0.40, 0.13) | −0.14 (−0.40, 0.12) | −0.13 (−0.40, 0.13) | −1.08 (−4.07, 1.89) |
| Hours in neighbourhood/ typical week | 0.05 (0.02, 0.08)# | 0.06 (0.03, 0.09)* | 0.06 (0.02, 0.09)* | 0.06 (0.02, 0.09)* | 0.06 (0.02, 0.09)* | 0.06 (0.02, 0.09)* | 0.06 (0.02, 0.09)* |
| Walkshed-level socioeconomic deprivation IndexŦ | −0.24 (−0.59, 0.11) | −0.08 (−0.46, 0.30) | −0.24 (−0.61, 0.14) | −0.19 (−0.57, 0.19) | −0.15 (−0.52, 0.23) | −0.16 (−0.53, 0.22) | −0.11 (−0.48, 0.27) |
| R2 | 0.18 | 0.19 | 0.18 | 0.18 | 0.18 | 0.18 | |
+Unadjusted estimates for the association between walkshed food environment measure (presence, density, diversity, presence of supermarket/grocery store, presence of convenience/multiproduct store, presence of restaurant) and C-HEI score are presented in the row with the “unadjusted” heading. Unadjusted estimates of the association between each covariate and C-HEI scores are presented in the first column of the table
~Adjusted estimates control for all covariates (sex, age, race, marital status, dependents at home, level of education, total gross household income, smoking status, car availability for personal use, dog ownership, self-reported physical health, self-reported mental health, number of hours spent sitting per day, number of hours spent in the neighbourhood during a typical week and walkshed level socioeconomic deprivation). These covariates represent socio-demographic, socioeconomic, health behaviours, and neighbourhood characteristics previously noted to be associated with diet quality. Fully-adjusted estimates are intended to isolate the effects of the walkshed socioeconomic status and walkshed food environment on diet quality
aModel 1 used linear regression to estimate the association between the presence of any food destination within the 400 m walkshed and C-HEI score. Presence was defined as at least one food destination present within the 400 m walkshed. The intention of this model was to determine if having a food destination within a 400 m walkshed of home address, regardless of type or count, was associated with diet quality (C-HEI score). All covariate estimates in the Model 1 column are fully-adjusted
bModel 2 used linear regression to estimate the associations between the density of food destinations within the 400 m walkshed and C-HEI score. Density was a continuous variable and defined as the total count of food destinations (all types) per walkshed area (km2). The intention of this model was to determine if the number of food destinations, regardless of type, within a 400 m walkshed of home address was associated with diet quality (C-HEI score). All covariate estimates in the Model 2 column are fully-adjusted
cModel 3 used linear regression to estimate the associations between the diversity of food destinations within the 400 m walkshed and C-HEI score. Diversity was defined by an index variable [minimum = 0, maximum = 9] capturing the variety of food destination types available within 400 m from home address. The nine food destination types were: fast-food restaurants, cafés, carry-out restaurants, full-service restaurants, supermarkets, grocery stores, convenience stores, multiproduct stores selling groceries (e.g. pharmacies), and single product specialty stores (e.g.. butchers, fruit and vegetable stands, and bakeries). Presence of a food destination type was defined as at least one destination within the 400 m walkshed. The intention of this model was to determine if greater diversity in food purchase opportunity within a 400 m walkshed of home address was associated with diet quality (C-HEI score). All covariate estimates in the Model 3 column are fully-adjusted
d Model 4 used linear regression to estimate the associations between the presence of a supermarket or grocery store within the 400 m walkshed and C-HEI score. Presence was defined as at least one supermarket or grocery store within the 400 m walkshed. The intention of this model was to determine if the presence of a food destination assumed to offer opportunity to purchase a variety of food types (e.g., fresh produce, lean proteins, dairy, whole grains) was associated with diet quality (C-HEI score). All covariate estimates in the Model 4 column are fully-adjusted
eModel 5 used linear regression to estimate the associations between the presence of a convenience or multiproduct store within the 400 m walkshed and C-HEI score. Presence was defined as at least one convenience or multiproduct store within the 400 m walkshed. The intention of this model was to determine if the presence of a food destination assumed to offer opportunity for limited variety of food purchase (e.g., primarily packaged and high fat, high sugar convenience foods) was associated with diet quality (C-HEI score). All covariate estimates in the Model 5 column are fully-adjusted
fModel 6 used linear regression to estimate the associations between the presence of a restaurant within the 400 m walkshed and C-HEI score. Presence was defined as at least one restaurant within the 400 m walkshed. The intention of this model was to determine if the presence of a restaurant assumed to offer limited variety of opportunity for food purchase (e.g., prepared dishes often high in sodium and fat) was associated with diet quality (C-HEI score). All covariate estimates in the Model 6 column are fully-adjusted
^Reference groups: Model1 = 0 food destinations within 400 m street network; Model 4 = 0 supermarket or grocery stores within 400 m street network, 0 convenience or multiproduct stores within 400 m street network, 0 restaurants within 400 m street network; Sex = men; Age = 21–39 years; Race = Caucasian; Marital status = married or living with partner; Dependents at home = no dependents at home; Highest education = high school diploma or less; Total gross income = ≤$59 000; Smoking status = non-smoker; Car available for personal use = always have a car available for personal use; Dog ownership = owner; Self-reported mental health = poor/fair/good; Self-reported physical health = poor/fair/good
*Statistically significant at alpha = 0.05
#Statistically significant at alpha = 0.003 (Bonferonni adjustment)