Introduction
In March 2020, the World Health Organization (WHO) declared the coronavirus disease (COVID-19) as a global pandemic(1). Subsequently, the United States (US) declared a national state of emergency, with state and local stay-at-home orders, closing schools and businesses. Economic implications including unemployment, food insecurity, financial instability and access to healthcare services worsened, deeply impacting low-income populations(Reference Wolfson and Leung2). Preventive measures became unavoidable, influencing lifestyle-related habits, food choices and diet. Through the monitoring of the pandemic's effect food spending, prices and sufficiency, the USDA (US Department of Agriculture), Economic Research Service (ERS) found expenditures at restaurants and other eating-out establishments fell from $79 billion in February 2020 to $42 billion in April 2020(3). Furthermore, COVID-19 illnesses led to a reduction of farm workers leading to decreased production of fresh vegetables and crops(3). Prior to the COVID-19 pandemic, in 2019, 10⋅5 % (13⋅7 million) of US households were food insecure, 13⋅6 % of which included children(4). As a result of the pandemic and the subsequent financial crisis, food insecurity rates increased. At the time, population experts had projected more than 42 million individuals may experience food insecurity, with approximately 13 million being children in 2021(4). Food insecurity has been recognised as a critical social determinant of health with important links to health and behavioural outcomes(Reference Drennen, Coleman and Ettinger de Cuba5).
Food insecurity is defined as the lack of or uncertain availability and limited access to nutritionally adequate food acquired in socially acceptable ways(Reference Andersen6). As determined by the USDA (US Department of Agriculture), food insecurity was associated with reduced food intake and disrupted eating patterns among households with children due to lost school meals support during the pandemic with shut down of schools(4,Reference Hales and Coleman-Jensen7) . Furthermore, food insecurity has been associated with consumption of lower quality food, poor nutrition and changes in dietary habits, specifically reduced fruit and vegetable (FV) consumption(Reference Araújo, Mendonça and Lopes Filho8). Additionally, purchasing packaged foods with longer shelf lives, rather than fresh food, led to lower consumption of the recommended portions of FV(Reference Richards and Rickard9). A recent cohort study showed disruption of dietary behaviours during the COVID-19 lockdown(Reference Deschasaux-Tanguy, Druesne-Pecollo and Esseddik10). Results of the present study indicated increased weight gain due to increased snacking and sedentary activity, and decreased fresh produce consumption(Reference Deschasaux-Tanguy, Druesne-Pecollo and Esseddik10). The study also examined shopping behaviours with 27 % of individuals reported buying less fresh produce and 26 % reported difficulty in accessing fresh produce(Reference Deschasaux-Tanguy, Druesne-Pecollo and Esseddik10).
Brighter Bites is a school-based nutrition and health promotion programme serving families with children in schools with 75 % of the student population eligible for free or reduced-price lunches. The programme uses a co-op model to distribute fresh FV to low-income families in elementary schools and early childcare centres in conjunction with nutrition education, recipes and recipe demonstrations(Reference Sharma, Markham and Chow11). Brighter Bites programming had been implemented during the school year in two 8-week sessions in the fall and spring, with additional summer programming for up to 8 weeks(Reference Sharma, Markham and Chow11). In March 2020, due to COVID-19-related school closures, Brighter Bites programming came to a halt.
Limited studies have assessed the impact of COVID-19-related environmental shifts on FV intake among food insecure families with children within the US. To understand the impact of the early phase of the pandemic, in April–June 2020, Brighter Bites conducted a rapid assessment of dietary behaviours, social determinants of health and COVID-19 concerns among participating families experiencing food insecurity. The purpose of the present study was to examine the cross-sectional relationship between both COVID-19-related concerns as well as FV shopping behaviours with self-reported changes in FV consumption among a sample of food insecure families with children enrolled in the Brighter Bites program.
Methods
Design
Cross-sectional data were collected during April–June 2020 as part of a rapid-response survey to understand social needs, COVID-19-related concerns and diet-related behaviours among families with children participating in Brighter Bites in the 2019–2020 school year at risk for food insecurity. Ethical approval for this study was obtained from the Institutional Review Board at the University of Texas Health Science Center (UTHSC) at Houston (IRB# HSC-SPH-15-0752).
Respondents and recruitments
The rapid-response survey screener was electronically distributed in both English and Spanish versions via Formsite (Vroman Systems, Inc.) to participants who provided a telephone number and consent during enrolment in Brighter Bites. A total of 16 673 participating Brighter Bites families in Texas (4⋅0 % in Austin, 56⋅2 % in Houston and 25⋅7 % in Dallas), Southwest Florida (6⋅1 %) and Washington, D.C. (8⋅1 %) received a texted link from programme staff members. The survey was voluntarily completed by a parent or adult caregiver in the family and open for 3 months (April–June 2020) with a response rate of 10⋅7 % (1777 of 16 673). The survey is typically distributed at this time, at the end of the school year to assess the effect of the Brighter Bites program within the past Spring semester. This specific survey was distributed to report changes in activity due to the pandemic along with school closures. Response rates by region were 4⋅8 % in Austin, 12⋅5 % in Houston and 5⋅2 % in Dallas, Texas; 12⋅3 % in Southwest Florida and 6⋅9 % in Washington, D.C. Response rates may have been low due to school closures and changes in people's lives due to the pandemic, even though researchers followed up with a reminder link to complete the survey on a weekly basis for 3 months. The proposed response rate was expected to have been 70 % or higher, similar to previous assessment years(Reference Marshall, Markham and Ranjit12). Furthermore, Brighter Bites collected, de-identified and shared completed data with UTHSC for analysis as part of a data-sharing agreement.
Data collection and measures
The self-reported 30-item survey, pilot tested within the research team prior to full scale administration, took about 10 min to complete, and was administered electronically with accessibility on smartphones or computers. The survey was used to rapidly assess socio-demographic data, dietary and FV shopping behaviours, COVID-19-related concerns and social determinants of health described below.
Socio-demographic data
Respondent socio-demographic data included sex, age, relationship to child, respondent and child race/ethnicity, employment status, education level, and total number of adults and children in household. Respondents’ enrolment in government assistance programmes (Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), Supplemental Nutrition Assistance Program (SNAP), Double Dollars, and/or free or reduced-price National School Lunch Program (NSLP)) was also assessed. Participation in food assistance programmes was coded as participating in none, one, or two or more programmes.
Outcome: parental FV intake
As a result of the pandemic, parent's FV intake was measured using a created 3-item self-reported block screener based on a previously validated self-reported FV block screener(Reference Hunsberger, O'Malley and Block13). We were interested in knowing whether the overall frequency of FV intake had changed or stayed the same, rather than knowing changes in portion sizes or specific food items. The frequency of FV intake was assessed using the following question ‘Due to coronavirus, has your consumption of fruits and vegetables: increased, decreased or stayed the same?’ We were able to compare responses to this survey in Spring 2020 to the past survey in Fall 2019 before the peak of the pandemic. Moreover, questions within this survey are specifically geared to assess changes due to COVID-19 by tweaking the already validated questions to include changes due to COVID-19 and distributing these surveys during Spring 2020.
Exposure: FV shopping behaviour
FV shopping frequency and behaviour were characterised using items adapted from the National Cancer Institute's 2007 Food Attitudes and Behavior Survey(Reference Blanck, Thompson and Nebeling14). Both frequency (times/month or times/week) and type of store for FV purchase were assessed using the following questions and response options: ‘Due to coronavirus, currently how often do you buy or get fruits and vegetables and other groceries for the family from a large grocery store or super market?’ Five response options ranging from: ‘never’ to ‘2+ times per week’. ‘Due to coronavirus, currently how often do you buy or get fruits and vegetables and other groceries for the family from these locations?’ Response options: ‘A small local store or corner store (usually locally owned and do not sell gas), or a convenience store (such as 7–11 or mini market usually sell gas)’, ‘farmer's market/food co-op/farm stand’ and ‘food bank/food pantry, or other food distributions’. Survey items were modified by adding ‘Due to coronavirus’ to reflect the behavioural changes due to COVID-19 among an already food insecure population. Questions do not reflect shopping behaviours for items they need due to COVID-19, but rather their shopping behaviour for FV specifically during the pandemic.
Multiple responses could be selected for these questions. Shopping behaviour was categorised as the total number of times per month or per week an individual shopped for FV at a certain location. For every one time per month, frequency was assessed as a continuous variable and coded 0 (never), 0⋅5 (less than once a month), 1⋅5 (1–2 times per month), 4 (1 time per week) and 8 (2+ times per week). We tweaked the questionnaire and asked that the respondents answer these questions with the changes of their behaviours being due to the pandemic with no other modifications made to the specific questions. Moreover, we were able to compare responses during the pandemic in Spring 2020 to responses pre-pandemic in Fall 2019.
Exposure: variables for COVID-19-related concerns
The items assessing COVID-19-related concerns were adapted from previously validated survey instruments(15): ‘Due to coronavirus, are you concerned about any of the following in regards to you and your family? (Check all that apply)’ Multiple response options could be selected from: ‘Financial stability’, ‘My employment status will change in the near future’, ‘Availability of food’ and ‘Affordability of food’.
Confounders
Eating out frequency
As a result of the pandemic, the family's frequency of eating food from a restaurant was measured using a created 3-item self-reported block screener, by previously validated items for home mealtime environment(Reference Marshall, Markham and Ranjit12,Reference Ding, Sallis and Norman16) . We were interested in knowing whether the overall frequency of eating out had changed or stayed the same. The frequency of eating out was assessed using the following question, ‘During the past 7 days, how many times did your family: Eat food from any type of restaurant? This includes restaurants such as fast food, sit-down restaurants, buffet restaurants, taco shops, donut shops and pizza places’. Five response options ranged from ‘never’ to ‘7+ times per week’.
Food insecurity
The 2-item Hunger Vital Sign Screener assessed household food security status during the COVID-19 pandemic(Reference Hager, Quigg and Black17). We added the phrase ‘due to coronavirus’ to each of the 2-item screener to assess the risk of food insecurity that is due to the pandemic, rather than general risk of food insecurity. The following items were worded as such: ‘Due to coronavirus: You worried whether your food would run out before you got money to buy more’ and ‘The food you bought just didn't last and you didn't have money to get more’(Reference Hager, Quigg and Black17). Response options for both items included ‘Often True’, ‘Sometimes True’ or ‘Never True’. Household respondents who responded ‘Often True’ or ‘Sometimes True’ to either food security question was categorised as at risk for food insecurity. Households were categorised as food secure if respondents responded ‘Never True’ to both questions.
Statistical analysis
At baseline, the sample included 1777 adults participating in Brighter Bites, 92⋅2 % of whom reported experiencing food insecurity. Therefore, we only included those who were at risk for food insecurity in the analysis for this current study (n 1639). Since those who were food secure represented less than 10 % of the population, we did not include them in this analysis as a potential comparison group. All analyses were conducted using Stata version 16.0 (StataCorp LLC). Descriptive statistics and bivariate analyses, including Chi-square, examined differences in respondent characteristics. Means, standard deviations and frequency distributions were computed. Frequency distributions were stratified by a categorical FV intake. During this acute phase of the pandemic, multinomial logistic regression modelling was used to determine cross-sectional associations between FV shopping behaviours and changes in FV intake using Beta coefficients. Furthermore, multinomial logistic regression modelling was used to determine cross-sectional associations between variables for COVID-19-related concerns and changes in FV intake using relative risk ratios. The main dependent variable assessed was a self-reported change in FV intake due to the COVID-19 pandemic. The independent variables were (a) COVID-19-related concerns such as financial stability, change of employment status, food availability and food affordability, and (b) FV shopping behaviour. Confounding variables were chosen a priori using a directed acyclic graph and computing the measure of association both before and after adjusting(Reference Evans, Chaix and Lobbedez18). Moreover, confounding was present and included in the model if the difference between the two measures of association was at least 10 %(Reference Lee19). Models were adjusted for the following covariates: race/ethnicity, region, total number of adults and children in household, frequency of eating out at restaurants, food assistance criteria and FV shopping behaviours in only those who were food insecure based on the level of significance. The significance level used was P < 0⋅05.
Results
Descriptive statistics
Table 1 outlines socio-demographic characteristics of food insecure households (n 1639). Of these food insecure households, mean respondent age was 37 years old (±7⋅2); respondents were mostly Hispanic (84⋅1 %), predominantly located in Houston, Texas (71⋅4 %). Seventy-three percent reported their children utilised NSLP's free or reduced-price meals, 31 % received SNAP benefits and 25 % received WIC. Furthermore, 16 % of respondents reported not utilising any food assistance programmes, 48 % used only one programme and 37 % used two or more programmes. Food insecurity in households was positively associated with food assistance programme participation (P < 0⋅001).
Table 1. Descriptive table of survey/respondent characteristics with food insecurity bivariate relationship between food insecurity and covariates
WIC, Special Supplemental Nutrition Program for Women, Infants, and Children; SNAP, Supplemental Nutrition Assistance Program; sd, standard deviation.
* P < 0⋅05; P < 0⋅01; P < 0⋅001; or Significance based on 95 % CI.
Statistical test used: Chi-square.
Among individuals from our final sample, 41 % (n 672) reported decreased FV intake during the pandemic, 32 % (n 527) reported increased intake and 27 % (n 439) reported no change in intake. Moreover, 62 % of respondents reported never eating out at a restaurant during the pandemic, while 38 % reported eating food from a restaurant at least once a week. During the time of the pandemic, 16 % (n 256) reported either never shopping or shopping less than once a month at large grocery stores/supermarket. Among those who shopped at large grocery stores/supermarket, most respondents shopped either 1–2 times per month (36 %) or 1 time per week (34 %). The majority of respondents (64 %; n 1043) reported frequency of shopping as never or less than once a month at a small local store/corner store for FV. A smaller percentage of respondents (34 %) reported shopping less than once a month at a food bank/food pantry.
Respondents reported a high prevalence of concerns regarding financial stability (78 %), changes in employment status (41 %), food availability (70 %) and food affordability (53 %).
Relationship between FV shopping behaviours and changes in FV consumption among respondents from food insecure households
Among respondents whose FV intake decreased, shopping frequency at large grocery stores/supermarket was 2⋅3 [±2⋅0] times per month. However, among individuals whose FV intake stayed the same, frequency of shopping at large grocery stores/supermarket was 3⋅2 [±2⋅2] times per month. There was a positive association between frequency of FV shopping and changes in FV intake (Adjβ = −0⋅2, CI −0⋅3, −0⋅1; P < 0⋅001). Similarly, among respondents who reported increased FV intake during this time, frequency of shopping at a large grocery store/supermarket was 4⋅2 [±2⋅6] times per month (β = 0⋅2, CI 0⋅1, 0⋅2; P < 0⋅001).
The frequency of shopping at small local stores/corner stores was 1⋅1 [±1⋅7] times per month among those whose FV intake decreased. In contrast, among individuals whose FV intake stayed the same, the frequency of shopping at small local stores/corner stores was 1⋅3 [±2⋅1] times per month (β = −0⋅1, CI −0⋅2, −0⋅0; P = 0⋅006) (Table 2). There was no significant relationship between increased FV intake and FV shopping behaviour at small local stores/corner stores.
Table 2. Relationship between changes in FV intake and FV shopping behaviours, and concerns regarding various social determinants of health among those who were at risk for food insecurity (N 1639)
RR, relative risk ratio; CI, confidence interval; sd, standard deviation; REF, Reference.
* P < 0⋅05; P < 0⋅01; P < 0⋅001; or Significance based on 95 % CI.
Multinomial logistic regression.
Adjusted for: race/ethnicity, region, total number of adults living in home, total number of children living in home, frequency of eating out at restaurants, food assistant criterion, FV shopping behaviours (as continuous variables) in only those who are food insecure.
For every one time per month: 0 = Never, 0⋅5 = Less than once a month, 1⋅5 = 1–2 times per month, 4 = 1 time per week, 8 = 2+ times per week.
Among those whose FV intake increased, the frequency of shopping at a food bank/food pantry was 1⋅4 [±2⋅1] times per month, whereas among individuals whose FV intake stayed the same, frequency of shopping at a food bank/food pantry was 1⋅1 [±1⋅8] times per month (β = 0⋅1, CI 0⋅0, 0⋅1; P = 0⋅05), as shown in Table 2. There was no significant relationship between decreased FV intake and FV shopping behaviour at a food bank/food pantry.
There was also no significant relationship between changes in FV intake and frequency of shopping at a farmers market or food co-op, which represented the least frequented food venue across all FV intake groups.
Relationship between variables for COVID-19-related concerns and changes in FV consumption
Those reporting concerns about financial stability during COVID-19 had 40 % greater risk of decreased FV intake compared to those not concerned about financial stability (RR (relative risk ratio) 1⋅4; 95 % CI 1⋅0, 2⋅0; P = 0⋅03). Additionally, those reporting concerns about employment status changing during COVID-19 had 70 % greater risk of increased FV intake opposed to FV intake staying the same and compared to those not concerned about change of employment status (RR 1⋅7; 95 % CI 1⋅3, 2⋅3; P < 0⋅001) (Table 2).
Those reporting concerns about food availability during COVID-19 had 50 % greater risk of increased FV intake opposed to FV intake staying the same and not being concerned for food availability (RR 1⋅5; 95 % CI 1⋅1, 2⋅0; P = 0⋅01). On the other hand, those reporting concerns about food availability during COVID-19 had 70 % greater risk of decreased FV intake opposed to FV intake staying the same and not being concerned for food availability (RR 1⋅7; 95 % CI 1⋅3, 2⋅2; P < 0⋅001).
Those reporting concerns about food affordability during COVID-19 had 40 % greater risk of increased FV intake opposed to FV intake staying the same and not being concerned for food affordability (RR 1⋅4; 95 % CI 1⋅1, 1⋅8; P = 0⋅02). On the other hand, those reporting concerns about food affordability during COVID-19 had 50 % greater risk of decreased FV intake as opposed to FV intake staying the same and not being concerned for food affordability (RR 1⋅5; 95 % CI 1⋅1, 2⋅0 P = 0⋅003). Both decreased and increased FV intake were significantly associated with concerns of food affordability during COVID-19.
Discussion
Our study adds to the growing body of literature to understand how the pandemic impacted dietary behaviours, specifically FV intake in low-income food insecure households. Overall, we found that among households at risk for food insecurity, there were changes in FV intake during this initial phase of the pandemic such that the intake increased or decreased across a majority of the participants, and these changes were associated with several social determinants of health. These results concur with other recent studies that have demonstrated increases in consumption of food products other than produce and changes of food intake patterns during COVID-19(Reference Górnicka, Drywień and Zielinska20). A previous study showed lower income was associated with higher boredom- or anxiety-related consumption of foods, disrupted mealtime schedules, decreased purchase of fresh produce and unhealthy changes in overall food consumption(Reference Deschasaux-Tanguy, Druesne-Pecollo and Esseddik10). Along with behavioural changes, emotional responses arose, which included fear of lack of food as well as feelings of stress and uncertainty. Consumers reduced FV shopping trips to limit the risk of exposure to COVID-19(Reference Hassen, Bilali and Allahyari21). A rise in emotional stress was positively associated with the overconsumption of processed meats and non-perishable items, including processed foods and savoury snacks(Reference Sadler, Thapaliya and Jansen22). Panic buying during the crisis included increased non-perishable foods with longer shelf lives and decreased fresh FV(Reference Baker, Farrokhnia and Meyer23). The prevalence of food insufficiency, meaning a household did not have enough food to eat within the past week, rose to 13⋅4 % as of December 2020(3). The impact of COVID-19 on the food supply chain and mobility restrictions to purchase fresh produce led to changes in food and dietary behaviours, unsold agriculture products and increased food loss and waste(Reference Hassen, Bilali and Allahyari21). Our study adds to this current body of sparse literature on how the initial phase of the pandemic impacted FV consumption behaviours among food insecure households with children.
Our study demonstrated that respondents who shopped for groceries at small local stores/corner stores reported significantly decreased FV intake; conversely, those who shopped at a food bank/food pantry had significant increases in FV intake. These results underscore the importance of alternative FV access avenues such as food pantries during times of natural disasters to maintain diet quality.
Latest research examined the effects of the pandemic such as COVID-19 fear, social conditions and influences on health and nutrition(Reference Aksoy, Kabadayi and Alan24). The pandemic affected individual's daily routines regarding where and how they shopped for produce. Also, fear of COVID-19 positively affected nutritional habits by consuming less unhealthy foods(Reference Aksoy, Kabadayi and Alan24). Specifically, individuals consumed more fruit, vitamins and yogurt during this time and less of fast food, snacks, desserts, cigarettes and packaged foods(Reference Aksoy, Kabadayi and Alan24). These results aligned with our study where 27 % of the respondents reported increasing their FV intake during that time, even while they reported being food insecure. The present study had a relatively low response rate, while typically the response rate is much higher in this same population, but the findings were sufficient to allow for a pivot in the programme to capture immediate needs of these families which were already utilising the programme prior to the pandemic.
Finally, our results highlight the vulnerability of food insecure households to food availability and affordability. These results concur with other recent studies that evaluated the prevalence of food insecurity associated with affordable, high quality foods and FV intake(Reference Horning, Alver and Porter25). Those who had access to affordable foods and cooked healthy foods were associated with lower odds of food insecurity overall(Reference Horning, Alver and Porter25). Individuals who were food insecure were less likely to consume fruits and vegetables(Reference Horning, Alver and Porter25). Similar to what was identified in our study, other studies demonstrated the impact of the food environment. Individuals using a small local corner store for food purchase may be less likely to find an assortment of healthy foods and thus more inclined to purchasing processed, unhealthy food items with longer shelf life(26). The relationship between food insecurity and FV consumption and shopping behaviours stipulates a need for further interventions to address food insecurity's effects on FV intake. These findings, along with those in our study highlight the disruption of both the food environment, and FV shopping behaviours during the initial phase of the pandemic which in turn impacted produce intake during this time.
Strengths and limitations
A limitation of the present study was the use of self-reported dietary intake data and the relatively low response rate of 10⋅7 % from a sub-sample of participating families in Brighter Bites in the 2019–2020 school year. However, the intent of this rapid survey was to obtain information on the immediate needs of families during this time to allow organisations like Brighter Bites to pivot services, which was successfully done(Reference Haidar, Khoei and Alex27). Selection bias may have occurred due to the electronic nature of the survey including only those who could access this survey and a small proportion of families who participated. However, this was the most effective way to conduct the survey during the pandemic. These survey data provided meaningful insights into context and possible influences on dietary behaviour that have not been studied under the COVID-19 pandemic circumstances in five different regions in the US. However, sample size differed across regions with regards to the proportion of families enrolled in Brighter Bites. Additionally, this sample may not be generalisable to the population across the regions; those who responded may be individuals who needed the most help. In previous studies with a larger sample size, generalisability is high since the Brighter Bites population represents diverse, low-income groups from different regions in the US(Reference Crulcich28). However, the participants, primarily Hispanic who also qualified for school lunch assistance programme, reflect a sample of convenience, limited to the specific school-based programme, Brighter Bites, and in limited areas of the US; therefore, the study has limited power based on the sample size.
The survey used a well-developed two-item tool to assess and distinguish food insecurity through affirmative responses to at least one of the two questions from the HFSS (Household Food Security Survey)(Reference Hager, Quigg and Black17). However, the data were self-reported which could result in social desirability bias. Furthermore, modifying the validated questions in the survey to consider changes in FV intake due to the pandemic may have been interpreted in different ways, leading to measurement error. Finally, the cross-sectional nature of this study prevents true causal inferences from being made.
Implications for research and practice
These results contribute to the literature exploring how unexpected, large-scale pandemic events may impact FV consumption among food insecure families with children. To our knowledge, the present study examined possible influences on dietary behaviour that had not been studied under the COVID-19 pandemic circumstances. In addition, these findings enhance the current literature regarding the dynamic and complex influences on FV consumption. More effective core infrastructure investments are needed to build capacity to diminish the negative impact of natural disasters such as the COVID-19 pandemic on the population's health.
A global approach is needed to decrease food insecurity, a major concern of this pandemic, due to its negative impact on the population's overall health.
Acknowledgements
S. S. is on the Board of Directors of Brighter Bites non-profit organisation, the goal of which is to improve access to fresh fruits and vegetables and nutrition education among underserved communities. This is an unpaid, advisory board position. This work is also a publication of the USDA/ARS Children's Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX. This project was supported, in part, through federal funds from the US Department of Agriculture (USDA)/Agricultural Research Service under Cooperative Agreement no. 3092-51000-058-2S (J. M. D.). The contents of this publication do not necessarily reflect the views or policies of the USDA, nor does mention of trade names, commercial products or organisations imply endorsement by the US Government.
Formulating the research question, Designing and Carrying out the study: M. A., A. M., J. M. D., R.-J. C. and S. S.; Analysing the data: M. A.; Writing the article: M. A., A. M. and J. M. D.; Editing the article: M. A., A. M., J. M. D., R.-J. C., C. M. and S. S.
Support for this research paper was provided in part by Cooperative Agreement Number (5U48DP00498-05) funded by the Centers for Disease Control and Prevention's Division of Nutrition, Physical Activity, and Obesity (DNPAO) and Prevention Research Centers Program, which includes the Nutrition and Obesity Policy Research and Evaluation Network (NOPREN). The findings and conclusions in this product are those of the author(s) and do not necessarily represent the official position of the CDC or DHHS. NOPREN had no role in the design, analysis or writing of this article. Questions: contact NOPREN at nopren@ucsf.edu.
The authors declare no conflict of interest.
This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the Institutional Review Board at the University of Texas Health Science Center (UTHSC) at Houston (IRB# HSC-SPH-15-0752). Written or Verbal informed consent was obtained from all subjects/patients. Verbal consent was witnessed and formally recorded.
Open access
