Edited by: Uma Tiwari, Technological University Dublin, Ireland
Reviewed by: Suzanne E. Judd, University of Alabama at Birmingham, United States; Mirta Crovetto, Universidad de Playa Ancha, Chile
This article was submitted to Nutrition and Sustainable Diets, a section of the journal Frontiers in Nutrition
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
To inform dietary interventions, it is important to understand antecedents of recommended (henceforth: healthy) dietary behaviors, beyond dietary beliefs and self-efficacy. We used the validated “Eating Identity Type Inventory” to assess the extent to which participants identified as healthy eaters, meat eaters, emotional eaters or picky eaters. We examined correlations between participants' race/ethnicity and other socio-demographic characteristics and affinity with these eating identities, how affinity with these eating identities correlated with self-reports of dietary beliefs, self-efficacy, dietary behaviors and Body Mass Index (BMI), and how well affinity with these eating identities predicted self-reported dietary behaviors and BMI, as compared to self-reported dietary beliefs and self-efficacy.
In an online survey, a diverse sample of 340 Los Angeles County adults reported eating identities, dietary beliefs, and self-efficacy, dietary behaviors and BMI.
Pearson correlations revealed that identifying more as a healthy eater was positively associated with self-reports of being non-Hispanic White, non-Hispanic mixed race, older, and college-educated, while identifying more as a meat eater was positively associated with self-reports of being non-Hispanic Black, younger, and male (α = 0.05). Pearson correlations also showed that healthy eaters had more accurate dietary beliefs and self-efficacy, and emotional eaters had lower self-efficacy (α = 0.05). In linear regressions, identifying more as a healthy eater was associated with self-reporting healthier dietary behaviors and lower BMI, and identifying more as a meat eater and emotional eater was associated with reporting less healthy dietary behaviors and higher BMI, even after accounting for correlations with socio-demographics, dietary beliefs, and self-efficacy (α = 0.05).
Our findings highlight the importance of eating identities in understanding dietary behaviors and outcomes, with implications for dietary interventions.
The Dietary Guidelines for Americans recommends “healthy eating behaviors,” including consuming more fruit and vegetables, and limiting intake of fast food, sweets, sugary beverages and fried food (
Although the diets of most American adults fail to meet recommendations for healthy eating, there are notable differences between racial/ethnic groups (
According to theories of health behavior change (
A growing body of literature has found that it is also important to consider the extent to which people identify with the identity of a healthy eater (
Recently, researchers have recognized that four other eating identities, in addition to a healthy-eater identity, are linked to dietary behaviors (
While research on eating identities shows promise for predicting dietary behaviors, most research on eating identities has focused on the healthy eating identity only (
(1) How do participants' race/ethnicity and other socio-demographic characteristics correlate with their affinity with these eating identities?
(2) How is participants' affinity with these eating identities correlated with dietary beliefs and self-efficacy as well as self-reported dietary behaviors and BMI?
(3) How well does participants' affinity with these eating identities predict self-reported dietary behaviors and BMI, as compared to dietary beliefs and self-efficacy?
Participants aged 18 and older were recruited from Los Angeles County (henceforth LA county) through two sampling strategies. First, participants were recruited through the LA county sample of the Understanding American Study, which was recruited from randomly selected LA county addresses (
Of the 574 invited participants, 410 (71%) completed the online survey between May 3rd, 2017 and August 31st, 2017. We removed 66 individuals who had missing data on our study variables, and 4 who had biologically implausible values in reported height (<1 meter or <2.5 meters) (
Participants completed the validated Eating Identity Type Inventory (
Descriptive statistics.
Non-Hispanic white | 40.0% |
Non-Hispanic black | 8.2% |
Hispanic | 37.9% |
Asian | 8.8% |
Non-Hispanic mixed race | 5.0% |
Age | M = 44.38 (SD = 15.43) |
Male | 37.3% |
College education | 41.2% |
Recruited through birth records (vs. addresses) | 32.4% |
Healthy eater (1 = strongly disagree; 5 = strongly agree; α = 0.89) | M = 3.51 (SD = 0.9) |
Meat eater (1 = strongly disagree; 5 = strongly agree; α = 0.79) | M = 3.13 (SD = 1.23) |
Emotional eater (1 = strongly disagree; 5 = strongly agree; α = 0.84) | M = 2.60 (SD = 1.18) |
Beliefs in healthy diet (1 = not all important; 5 extremely important; α = 0.76) | M = 4.15 (SD = 0.53) |
Dietary self-efficacy (1 = I know I cannot; 3 = I know I can; α = 0.94) | M = 2.51 (SD = 0.42) |
Daily fruit and vegetable intake (open-ended) | M = 3.93 (SD = 2.46) |
Daily water intake (open-ended) | M = 5.53 (SD = 3.53) |
Weekly fast-food frequency (open-ended) | M = 1.33 (SD = 1.25) |
Sugary-drinks intake (neve |
M = 1.20 (SD = 1.08) |
Never | 31.3% |
Less frequently than once a day | 35.6% |
Once a day | 14.7% |
More than once a day | 18.2% |
Sweets intake (neve |
M = 1.65 (SD = 0.88) |
Never | 5.0% |
Less frequently than once a day | 47.4% |
Once a day | 25.0% |
More than once a day | 22.6% |
Fried-food intake (neve |
M = 1.09 (SD = 0.60) |
Never | 8.5% |
Less frequently than once a day | 80.0% |
Once a day | 5.9% |
More than once a day | 5.6% |
Overall BMI | M = 27.70 (SD = 6.22) |
Underweight (<18.5) | 2.9% |
Normal (>18.5 <24.99) | 35.6% |
Overweight (>24.99 and <30) | 32.7% |
Obese (>30) | 28.8% |
Assessments of dietary beliefs were adapted from the Low Income Diet and Nutrition Survey (
Dietary self-efficacy was assessed through the validated Self-efficacy and Eating Habits Survey (
We asked participants about the following dietary behaviors: 1) daily fruit and vegetable intake 2) daily water intake 3) weekly fast-food frequency 4) sugary-drinks intake 5) sweets intake 6) fried-food intake. Dietary guidelines (
Participants were asked to answer open-ended questions about their height (e.g., What is your height?) and their weight (e.g., How much do you approximately weigh?). They could answer in pounds or kilograms for weight and feet and inches or meters and centimeter for height. Responses were transformed to meters and kilograms and BMI was calculated as kilograms divided by squared meters (kg/m2).
All analyses were conducted using SPSS version 25. The one exception was the Confirmatory Factor Analysis on the Eating Identity Type Inventory, which was computed in R (
Our first preliminary analyses focused on comparing invited individuals who were included in the analyses with those who were not. We conducted a
Our second preliminary analyses examined the internal consistency of the Eating Identity Type Inventory, including Cronbach's α, inter-item correlations, and Confirmatory Factor Analysis. We retained subscales with Cronbach's alpha ≥.7 and inter-item correlations ≥.3 and acceptable model fit in the Confirmatory Factor Analysis including CFI ≥0.95, TLI ≥0.90, RMSEA ≤ 0.10, and SRMR ≤.08 (
Research question 1 asked about the race/ethnicity and other socio-demographic characteristics of individuals who identified more as a healthy eater, meat eater, emotional eater, or picky eater. To answer Research Question 1, we computed Pearson correlations between participants' reported affinity with these eating identities and their race/ethnicity, age, gender, and education. Additionally, we conducted linear regressions that modeled reported affinity with eating identities as a function of race/ethnicity and other socio- demographic characteristics.
Research question 2 asked how affinity with eating identities was correlated to dietary beliefs, self-efficacy, self-reported dietary behaviors, and BMI. To answer Research Question 2, we computed Pearson's r correlations between eating identities and dietary beliefs, dietary self-efficacy, self-reported dietary behaviors, and BMI.
Research question 3 asked how well participants' affinity with eating identities predicts self-reported dietary behaviors and BMI, as compared to dietary beliefs and self-efficacy. To answer Research Question 3, we compared linear regressions models that predicted self-reported dietary behaviors (daily fruit and vegetable intake, daily water intake, weekly fast-food frequency, sugary-drinks intake, sweets intake, and fried-foods intake) from affinity with eating identities (Model 1 in
There were no significant differences between the 340 invitees who were included in the data analyses presented here and the 410 who were not, in terms of the percent identifying as non-Hispanic Black, Asian, non-Hispanic Mixed Race, or their mean age (all
For three of the four eating identities, items had sufficient internal consistency, seen in Cronbach's alpha being at least 0.70. Specifically, Cronbach's alpha was α = 0.89 for identifying as a healthy eater, α = 0.79 for identifying as a meat eater, and α = 0.84 for identifying as an emotional eater. Internal consistency was not sufficient for the picky-eater identity (α = 0.37). As in previous research (
This decision was supported by Confirmatory Factor Analyses, which was unable to fit all 11 items on four factors representing the healthy eater, meat eater, emotional eater, and picky eater identities. After removing the picky-eater subscale, a Confirmatory Factor Analysis still showed a relatively poor fit of the remaining 8 items on three factors representing the healthy eater, meat eater and emotional eater, seen in CFI ≥0.95, TLI ≥0.90, RMSEA ≤ 0.10 (48), CFI, and SRMR ≤.08 (χ2 (24) = 152.29,
Next, we computed Pearson correlations between mean scores for each eating identity we analyzed. Identifying more as a healthy eater was negatively correlated identifying more as a meat eater (
Identifying more as a healthy eater was correlated with self-reports of being non-Hispanic White (
Identifying more as a meat eater was correlated with self-reports of being non-Hispanic Black (
The degree to which participants identified as an emotional eater was positively correlated with self-reports of being non-Hispanic mixed race (
Identifying more as a healthy eater was correlated with self-reports of eating more fruit and vegetables (
Identifying more as a meat eater was associated with self-reporting more frequent fast-food dining (
Identifying more as an emotional eater was related to more self-reported consumption of sugary drinks (
Most of the relationships of eating identities with self-reported dietary behaviors and BMI that were reported for Research Question 2 still held in linear regressions that controlled for dietary self-efficacy, dietary beliefs, and socio-demographics (
Linear regressions predicting self-reported daily fruit and vegetable intake.
Adj. R2 = 0.11 F(11,339) = 4.67 |
Adj. R2 = 0.01 F(10,339) = 1.52 p = 0.13 | Adj. R2 = 0.11 F(13,339) = 4.12 |
||||
---|---|---|---|---|---|---|
Healthy eater | ||||||
Meat eater | −0.07 | (−0.36, 0.08) | −0.07 | (−0.37, 0.09) | ||
Emotional eater | 0.04 | (−0.14, 0.29) | 0.03 | (−0.15, 0.29) | ||
Dietary beliefs |
−0.03 | (−0.65, 0.35) | −0.08 | (−0.84, 0.13) | ||
Dietary self–efficacy |
0.07 | (−0.23, 1.05) | 0.00 | (−0.62, 0.67) | ||
(−0.84, 0.13) | ||||||
Demographics | ||||||
Non–Hispanic white | 0.00 | (−1.21, 1.24) | −0.11 | (−1.81, 0.73) | 0.00 | (−1.22, 1.24) |
Non–Hispanic black | 0.04 | (−1.11, 1.79) | −0.03 | (−1.77, 1.28) | 0.03 | (−1.15, 1.77) |
Hispanic | −0.02 | (−1.41, 1.17) | −0.17 | (−2.18, 0.48) | −0.02 | (−1.40, 1.19) |
Asian | 0.00 | (−1.38, 1.41) | −0.04 | (−1.81, 1.13) | 0.01 | (−1.35, 1.46) |
Age | −0.05 | (−0.03, 0.01) | 0.01 | (−0.02, 0.02) | −0.05 | (−0.03, 0.01) |
Male | ||||||
College education | 0.04 | (−0.37, 0.81) | 0.04 | (−0.40, 0.84) | 0.03 | (−0.43, 0.76) |
Recruited through birth | ||||||
records (vs. addresses) | 0.07 | (−0.30, 1.07) | 0.12 | (−0.09, 1.34) | 0.08 | (−0.28, 1.09) |
Linear regressions predicting self–reported daily water intake.
Adj. R2 = 0.10 F(11,339) = 4.25 |
Adj. R2 = 0.07 F(10,339) = 3.39 |
Adj. R2 = 0.10 F(13,339) = 3.81 |
||||
---|---|---|---|---|---|---|
Healthy eater | ||||||
Meat eater | 0.11 | (−0.01, 0.63) | 0.10 | (−0.03, 0.62) | ||
Emotional eater | 0.03 | (−0.22, 0.41) | 0.03 | (−0.24, 0.40) | ||
Dietary beliefs |
0.10 | (−0.02, 1.38) | 0.08 | (−0.15, 1.24) | ||
Dietary self–efficacy |
−0.03 | (−1.17, 0.62) | −0.03 | (−1.23, 0.64) | ||
Demographics | ||||||
Non–Hispanic white | −0.15 | (−2.87, 0.67) | −0.21 | (−3.28, 0.28) | −0.14 | (−2.82, 0.73) |
Non–Hispanic black | −0.13 | (−3.74, 0.45) | −0.14 | (−3.90, 0.35) | −0.12 | (−3.63, 0.59) |
Hispanic | −0.17 | (−3.12, 0.60) | −0.17 | (−3.09, 0.65) | ||
Asian | −0.06 | (−2.77, 1.27) | −0.08 | (−3.09, 1.03) | −0.06 | (−2.78, 1.28) |
Age | −0.13 | (−0.06, 0.00) | −0.13 | (−0.06, 0.00) | ||
Male | ||||||
College education | 0.09 | (−0.24, 1.46) | 0.12 | (−0.04, 1.70) | 0.10 | (−0.15, 1.56) |
Recruited through birth records (vs. addresses) | −0.04 | (−1.27, 0.70) | −0.02 | (−1.13, 0.86) | −0.04 | (−1.32, 0.65) |
Linear regressions predicting self–reported weekly fast–food frequency.
Adj. R2 = 0.20 F(11,339) = 8.72 |
Adj. R2 = 0.09 F(10,339) = 4.31 |
Adj. R2 = 0.20 F(13,339) = 7.56 |
||||
---|---|---|---|---|---|---|
Healthy eater | ||||||
Meat eater | ||||||
Emotional eater | 0.04 | (−0.06, 0.15) | 0.05 | (−0.05, 0.16) | ||
Dietary beliefs |
−0.05 | (−0.36, 0.13) | 0.00 | (−0.24, 0.22) | ||
Dietary self–efficacy |
−0.03 | (−0.40, 0.22) | 0.08 | (−0.07, 0.55) | ||
Demographics | ||||||
Non–Hispanic white | −0.01 | (−0.60, 0.57) | 0.06 | (−0.48, 0.76) | −0.02 | (−0.65, 0.53) |
Non–Hispanic black | 0.04 | (−0.51, 0.88) | 0.09 | (−0.32, 1.17) | 0.03 | (−0.58, 0.82) |
Hispanic | 0.19 | (−0.14, 1.10) | 0.16 | (−0.20, 1.05) | ||
Asian | −0.01 | (−0.73, 0.61) | 0.01 | (−0.66, 0.78) | −0.03 | (−0.80, 0.55) |
Age | −0.07 | (−0.02, 0.00) | −0.07 | (−0.02, 0.00) | ||
Male | 0.04 | (−0.16, 0.39) | 0.06 | (−0.14, 0.43) | 0.05 | (−0.15, 0.40) |
College education | ||||||
Recruited through birth records (vs. addresses) | −0.09 | (−0.58, 0.08) | −0.12 | (−0.67, 0.03) | −0.09 | (−0.57, 0.09) |
When controlling for race, age, gender, and education, identifying less as a healthy eater (β = −0.18;
Linear regressions predicting self–reported sugary–drinks intake.
Adj. R2 = 0.26 F(11,339) = 11.86 |
Adj. R2 = 0.22 F(10,339) = 10.71 |
Adj. R2 = 0.26 F(13,339) = 10.26 |
||||
---|---|---|---|---|---|---|
Healthy eater | ||||||
Meat eater | 0.11 | (0.00, 0.18) | ||||
Emotional eater | 0.09 | (−0.01, 0.17) | ||||
Dietary beliefs |
−0.05 | (−0.30, 0.09) | −0.02 | (−0.24, 0.14) | ||
Dietary self–efficacy |
−0.08 | (−0.45, 0.06) | ||||
Demographics | ||||||
Non–Hispanic white | −0.05 | (−0.60, 0.37) | 0.00 | (−0.49, 0.49) | −0.03 | (−0.56, 0.41) |
Non–Hispanic black | −0.01 | (−0.63, 0.52) | 0.03 | (−0.46, 0.72) | 0.00 | (−0.59, 0.57) |
Hispanic | 0.07 | (−0.36, 0.66) | 0.14 | (−0.21, 0.82) | 0.09 | (−0.31, 0.72) |
Asian | 0.00 | (−0.57, 0.54) | 0.03 | (−0.45, 0.70) | 0.01 | (−0.52, 0.60) |
Age | ||||||
Male | 0.01 | (−0.21, 0.24) | 0.00 | (−0.23, 0.23) | 0.00 | (−0.22, 0.23) |
College education | ||||||
Recruited through birth records (vs. addresses) | 0.07 | (−0.11, 0.43) | 0.05 | (−0.15, 0.40) | 0.07 | (−0.12, 0.43) |
Identifying less as a healthy eater (β = −0.21;
Linear regressions predicting self–reported sweets intake.
Adj. R2 = 0.08 F(11,339) = 3.69 |
Adj. R2 = 0.03 F(10,339) = 2.21 |
Adj. R2 = 0.08 F(13,339) = 3.21 |
||||
---|---|---|---|---|---|---|
Healthy eater | ||||||
Meat eater | 0.01 | (−0.07, 0.09) | 0.00 | (−0.08, 0.08) | ||
Emotional eater | ||||||
Dietary beliefs |
−0.06 | (−0.29, 0.07) | −0.03 | (−0.23; 0.12) | ||
Dietary self–efficacy |
−0.05 | (−0.35, 0.12) | ||||
Demographics | ||||||
Non–Hispanic white | ||||||
Non–Hispanic black | 0.06 | (−0.35, 0.71) | 0.09 | (−0.24, 0.84) | 0.07 | (−0.32, 0.74) |
Hispanic | 0.13 | (−0.24, 0.70) | 0.21 | (−0.08, 0.86) | 0.15 | (−0.21, 0.74) |
Asian | 0.09 | (−0.23, 0.79) | 0.12 | (−0.14, 0.91) | 0.10 | (−0.20, 0.83) |
Age | −0.05 | (−0.01, 0.00) | −0.08 | (−0.01, 0.00) | −0.05 | (−0.01, 0.00) |
Male | 0.01 | (−0.19, 0.22) | −0.01 | (−0.23, 0.19) | 0.01 | (−0.20, 0.22) |
College education | −0.01 | (−0.24, 0.19) | −0.02 | (−0.26, 0.18) | −0.02 | (−0.24, 0.19) |
Recruited through birth records (vs. addresses) | 0.13 | (−0.01, 0.49) |
In linear regressions predicting self-reported fried food consumption, identifying less as a healthy eater (β = −0.14;
Linear regressions predicting self–reported fried–food intake.
Adj. R2 = 0.11 F(11,339) = 4.92 |
Adj. R2 = 0.09 F(10,339) = 4.27 |
Adj. R2 = 0.12 F(13,339) = 4.42 |
||||
---|---|---|---|---|---|---|
Healthy eater | ||||||
Meat eater | 0.07 | (−0.02, 0.09) | 0.05 | (−0.03, 0.08) | ||
Emotional eater | ||||||
Dietary beliefs |
−0.04 | (−0.16, 0.07) | −0.01 | (−0.13, 0.10) | ||
Dietary self–efficacy |
−0.10 | (−0.29, 0.02) | ||||
Demographics | ||||||
Non–Hispanic white | 0.00 | (−0.30, 0.29) | 0.03 | (−0.26, 0.34) | 0.02 | (−0.28, 0.32) |
Non–Hispanic black | 0.02 | (−0.30, 0.40) | 0.06 | (−0.23, 0.49) | 0.04 | (−0.27, 0.44) |
Hispanic | 0.04 | (−0.26, 0.37) | 0.10 | (−0.19, 0.44) | 0.07 | (−0.23, 0.40) |
Asian | −0.02 | (−0.38, 0.30) | 0.01 | (−0.32, 0.37) | 0.00 | (−0.35,0.34) |
Age | ||||||
Male | 0.01 | (−0.13, 0.15) | 0.00 | (−0.14, 0.13) | 0.00 | (−0.13, 0.14) |
College education | −0.04 | (−0.20, 0.09) | −0.04 | (−0.20, 0.09) | −0.04 | (−0.19, 0.09) |
Recruited through birth records (vs. addresses) | 0.04 | (−0.11, 0.22) | 0.03 | (−0.13, 0.21) | 0.04 | (−0.12, 0.22) |
Finally, higher BMI (
Linear regressions predicting Body Mass Index.
Predictor variables | Adj. R2 = 0.29 F(11,339) = 13.47 |
Adj. R2 = 0.14 F(10,339) = 6.61 |
Adj. R2 = 0.31 F(13,339) = 12.63 |
|||
β | [95% Cl] | β | [95% Cl] | β | [95% Cl] | |
Healthy eater | ||||||
Meat eater | ||||||
Emotional eater | ||||||
Dietary beliefs |
0.10 | (−0.03, 2.33) | ||||
Dietary self–efficacy |
−0.10 | (−3.00, 0.01) | 0.04 | (−0.85, 2.02) | ||
Demographics | ||||||
Non–Hispanic white | 0.00 | (−2.70, 2.82) | 0.03 | (−2.60, 3.40) | −0.01 | (−2.80, 2.66) |
Non–Hispanic black | 0.09 | (−1.13, 5.41) | 0.15 | (−0.31, 6.86) | 0.09 | (−1.13, 5.37) |
Hispanic | ||||||
Asian | −0.12 | (−5.75, 0.54) | −0.10 | (−5.67, 1.26) | −0.14 | (−6.12, 0.14) |
Age | ||||||
Male | 0.07 | (−0.33, 2.23) | 0.07 | (−0.55, 2.23) | 0.08 | (−0.27, 2.26) |
College education | ||||||
Recruited through birth records (vs. addresses) | 0.00 | (−1.59, 1.49) | −0.03 | (−2.07, 1.30) | −0.01 | (−1.66, 1.38) |
Overall, these regressions suggest that eating identities were the stronger predictors of self-reported dietary behaviors, as compared to dietary beliefs and self-efficacy. Neither models with dietary beliefs and self-efficacy alone nor full models that included eating identities together with dietary beliefs and self-efficacy had larger predictive power (Adjusted R2) than models with eating identities alone (
The Dietary Guidelines for Americans suggest that for most people, eating healthy can help in fighting obesity and chronic disease risk (
First, due to our diverse Los Angeles County country sample, we were able to report how eating identities varied by race/ethnicity and other socio-demographic differences. Specifically, identifying as a healthy eater was seen more among individuals who self-reported being non-Hispanic White and non-Hispanic Mixed Race, and less among Hispanic participants. Moreover, identifying more as a meat eater was more common among non-Hispanic Black participants. We also found that identifying more as a healthy eater was associated with being older and college-educated, while identifying as a meat eater was associated with being younger and male. These results can potentially help to tailor interventions to specific socio-demographic needs. As noted by the Dietary Guidelines for Americans, healthy dietary behaviors can be individualized to address personal preferences and cultural traditions (
Our second finding was that eating identities were associated with dietary beliefs, dietary self-efficacy, and self-reported dietary behaviors. Previous research had already shown that identifying more as a healthy eater was positively associated with accurate dietary beliefs (
Our third main finding was that eating identities were better predictors of self-reported dietary behaviors and BMI as compared to dietary self-efficacy and dietary beliefs, while taking into account socio-demographic characteristics. Our findings have a potential implication for interventions that aim to promote healthy diets. Indeed, interventions may be more effective if they address people's eating identities, including those related to healthy eating, meat eating, and emotional eating.
The study has four main limitations. First, all dependent variables were self-reported, which may undermine their validity. Self-reported dietary behaviors may be more valid when using diary methods (
Despite these limitations, our study has implications for dietary interventions. Specifically, health interventions that aim to improve dietary behaviors could benefit from focusing on encouraging people to identify more as a healthy eater, less as a meat eater, and less as an emotional eater. Text-message reminders have been found to increase identifying as a healthy eater while decreasing identifying as a meat eater–which helped to promote healthier dietary behaviors (
The data underlying this article are publicly available:
This study involved human participants. It was reviewed and approved by University of Southern California's Institutional Review Board. The participants provided their written informed consent to participate in this study.
PS, WB, LA, and TG: conceptualization and methodology. PS: formal analysis and writing—original draft. WB: supervision. WB and PS: writing—review and editing. LA, WB, and TG: funding acquisition. All authors have approved the final manuscript.
Data collection was funded by the LA Department of Public Health and First Five LA as part of their Choose Health LA program, as well as the Dornsife Center for Economic and Social Research. Wändi Bruine de Bruin was partially funded by the University of Southern California's Schaeffer Center for Health Policy and Economics.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
We thank Michael Sobolev for commenting on an earlier draft of the paper.
The Supplementary Material for this article can be found online at: