Sugar-sweetened beverages (SSB) contribute to a myriad of health conditions including dental caries(Reference Warren, Weber-Gasparoni and Marshall1,Reference Evans, Hayes and Palmer2) , obesity(Reference DeBoer, Scharf and Demmer3-Reference Cantoral, Tellez-Rojo and Ettinger5), type 2 diabetes(Reference Malik, Popkin and Bray6), fatty liver disease(Reference Nseir, Nassar and Assy7), hypertension(Reference Nguyen, Choi and Lustig8), heart disease(Reference Malik9), cancer(Reference Chazelas, Srour and Desmetz10) and death(Reference Collin, Judd and Safford11). The impact of SSB on health is evident both in adults and among children(Reference Evans, Hayes and Palmer2,Reference DeBoer, Scharf and Demmer3,Reference Cantoral, Tellez-Rojo and Ettinger5,Reference Nguyen, Choi and Lustig8,Reference Malik9) . Excess sugar consumption is a global concern as are efforts to reduce SSB intake through policy-level interventions(12-Reference Deren, Weghuber and Caroli14). The WHO recommends that no more than 5 % of total energy intake comes from added sugars(12). Both the American Academy of Pediatrics and the European Academy of Paediatrics recommend that children avoid SSB and endorse policies that restrict children’s access to SSB and promote water consumption(Reference Muth, Dietz and Magge13,Reference Deren, Weghuber and Caroli14) . Previous studies have demonstrated racial-ethnic disparities in the consumption of SSB among US children(Reference Taveras, Gillman and Kleinman15-Reference Elfassy, Adjoian and Lent20). These disparities likely contribute to higher rates of diseases among racial-ethnic minorities in the USA including type 2 diabetes(Reference Dabelea, Mayer-Davis and Saydah21) and fatty liver disease(Reference Botero, Hoy and Jimenez22). Understanding disparities in SSB consumption and tracking trends in consumption among different racial-ethnic groups is critical for the design and evaluation of policies to reduce SSB consumption.
California is home to one-eighth of the US child population, and nearly three-quarters of those children are racial-ethnic minorities, with 52 % identifying as Latino(23). Previous research using data from the California Health Interview Survey (CHIS), a telephone survey of California households, documented a downward trend in SSB consumption from 2003 to 2009 among California children aged 2–11 years, but also found persistently higher consumption among Latino children(Reference Beck, Patel and Madsen17). The decrease in SSB consumption coincided with statewide policies implemented in that time period, most notably passage of legislation to ban SSB sales in schools in 2003 and 2005(Reference Beck, Patel and Madsen17). Nonetheless, ongoing tracking and assessment of trends in child SSB consumption, with attention to racial-ethnic disparities, are important. From 2003 to 2012, CHIS determined child SSB consumption via a single question about consumption of soda, sweetened fruit drinks and sports drinks. Beginning in 2013, CHIS modified its approach to assessing child SSB consumption and began to ask parents two questions, one on soda and one on sweetened fruit drinks/sports drinks. In this report, we assess trends in consumption of soda, sweetened fruit drinks/sports drinks and any SSB among children in California aged 2–5 and 6–11 years from 2013 to 2016.
Methods
Data source
We analysed data on soda and sweetened fruit drink/sports drink intake from the 2013–2014 and 2015–2016 CHIS surveys. CHIS is a telephone survey of households in California and is the largest statewide health survey in the USA. The survey is designed to yield population-level estimates of surveyed health behaviours as well as estimates for major racial-ethnic groups. CHIS uses a dual frame, multi-stage sample design that includes both landlines and cell phones(24). There are separate surveys for children (0–11 years), adolescents (12–17 years) and adults (18 years or older). For children under age 12, an adult who is knowledgeable about the child’s health responds to the questions. Households are selected via a random digit dialing approach within pre-defined geographic strata. In households with multiple children under the age of 12 years, one child is randomly selected. CHIS was conducted every other year from 2001 to 2010. Since 2011, the survey has been conducted continuously over 2-year cycles. Public use data files are available for each year of the survey beginning in 2011. However, it is recommended to pool data from each 2-year cycle (i.e. 2013–2014, 2015–2016) to achieve statistically stable estimates for children’s outcomes.
Measure of outcome variables
To assess children’s soda intake parents were asked ‘Yesterday, how many glasses or cans of soda that contain sugar, such as Coke, did your child drink? Do not include diet soda’. To assess children’s sweetened fruit drink/sports drink intake parents were asked ‘Yesterday, how many glasses or cans of sweetened fruit drinks, sports, or energy drinks, did your child drink?’ To assess race-ethnicity, parents were first asked if their child is Latino or Hispanic. They were subsequently asked which of the following categories best described their child: Native Hawaiian, Other Pacific Islander, American Indian, Alaska Native, Asian, Black, African American or White. Based on parental responses, children are classified as Latino/Hispanic (Latino), non-Hispanic White (White), non-Hispanic Asian (Asian), non-Hispanic African-American (African-American), non-Hispanic mixed ethnicity or other ethnicity.
Analysis
We used the survey function in Stata software (version 12) and the replicate weights provided by CHIS to obtain statewide estimates for each variable of interest. For our analysis, we dichotomised consumption of soda and sweetened fruit drinks/sports drinks on the day prior to the interview into any and none. We chose to dichotomise the primary outcome to facilitate comparisons to prior research(Reference Beck, Patel and Madsen17), and because the majority of children had not consumed an SSB on the day prior, rendering mean intake less meaningful. We also created a variable of any SSB consumption that was coded as positive if the child consumed any soda and/or any sweetened fruit drink/sports drink. We conducted separate analyses for children aged 2–5 years and children aged 6–11 years. For each age group, we determined the prevalence of consuming any soda, any sweetened fruit drinks/sports drink and any SSB among children in California in 2013–2014 and 2015–2016. We then determined consumption of any soda, any sweetened fruit drink/sports drink and any SSB among 2–5 and 6–11-year-olds in the four largest racial-ethnic groups: Latino, White, Asian and African-American. We used logistic regression to assess for differences in consumption among racial-ethnic groups and across survey years (2013–2014 v. 2015–2016).
Results
The 2013–2014 survey sample included 4901 children aged 2–11 years. In 2013–2014, 42 % of children were Latino, 38 % were White, 9 % were Asian, 3 % were African-American and 7 % were of another race/ethnicity or mixed race/ethnicity. The 2015–2016 survey sample included 3606 children. In 2015–2016, 47 % were Latino, 35 % were White, 8 % were Asian, 5 % were African-American and 5 % were of another race/ethnicity or mixed race/ethnicity. Consumption of any SSB was reported for 23 % of 2–5-year-olds in 2013–2014 and 22 % in 2015–2016 (Table 1). Latino 2–5-year-olds were more likely to consume sweetened fruit drinks/sports drinks relative to White children in 2013–2014 and were more likely to consume soda in 2015–2016. In both 2013–2014 and 2015–2016, Latino 2–5-year-olds were more likely to consume any SSB. There were no statistically significant differences in consumption for 2–5-year-olds from Asian or African American race/ethnicity relative to Whites. Comparing 2013–2014 to 2015–2016, there were no statistically significant differences in consumption of soda, sweetened fruit drinks/sports drinks or any SSB for 2–5- year-olds overall or for 2–5- year-olds in any of the four major racial-ethnic categories.
Table 1 Proportion of children in California aged 2–5 years who consumed any soda, any sweetened fruit drink/sports drink and any sugar-sweetened beverage (soda or sweetened fruit drink/sports drink) in the previous 24 h in 2013–2014 and 2015–2016 by major racial-ethnic category*
* Data from the California Health Interview Survey. All differences in consumption across years were non-significant. 2013–2014: Total n 1668, white n 631, Latino n 723, Asian n 141, African-American n 39. 2015–2016: Total n 1481, white n 475, Latino n 722, Asian n 96, African-American n 75.
† Consumption significantly higher than reference group (Whites) with P < 0·05.
Among 6–11- year-olds, 37 % consumed an SSB in 2013–2014 and 35 % consumed an SSB in 2015–2016 (Table 2). Among Latino 6–11- year-olds, soda consumption was higher than Whites in 2013–2014 and 2015–2016 and any SSB consumption was higher in 2013–2014. Among African-American 6–11- year-olds, soda consumption, any sweetened fruit drink/sports drink consumption and any SSB consumption were higher than Whites in 2013–2014. For 6–11-years-olds, there were no statistically significant differences in consumption for Asians relative to Whites. Differences in consumption of soda, sweetened fruit drinks/sports drinks or any SSB across years (2013–2014 to 2015–2016) were not statistically significant for 6–11- year old children overall or for any of the four major racial-ethnic categories.
Table 2 Proportion of children in California aged 6–11 years who consumed any soda, any sweetened fruit drink/sports drink and any sugar-sweetened beverage (soda or sweetened fruit drink/sports drink) in the previous 24 h in 2013–2014 and 2015–2016 by major racial-ethnic category*
* Data from the California Health Interview Survey (CHIS). All differences in consumption across years were non-significant. 2013–2014: Total n 3233, white n 1244, Latino n 1349, Asian n 302, African-American n 111. 2015–2016: Total n 2125, white n 796, Latino n 988, Asian, n 174, African-American n 97.
† Consumption significantly higher than reference group (Whites) with P < 0·05.
Discussion
We found no significant change in SSB consumption among children in California from 2013–2014 to 2015–2016. Child SSB consumption in 2013–2016 appears higher than in 2009 when 16 % of 2–5-year-olds and 33 % of 6–11-year-olds reported any SSB consumption(Reference Beck, Patel and Madsen17). However, this apparent increase may be due to changing from a single question on SSB consumption in 2003 to 2009 to separate questions about soda and sweetened fruit drinks/sports drinks in the 2013–2016 surveys. Regardless, there has certainly been no further decrease in SSB consumption among children in California relative to the last decade, suggesting that the impact of existing state-wide policies have reached their floor and additional measures are needed. Existing policies include legislation to ban sales of SSB in schools passed in 2003 and 2005, as well as a bill passed in 2010 that bans the provision of SSB to children in licensed childcare facilities in California(25). In 2019, several additional measures to discourage SSB consumption were introduced in the California legislature including a sugary beverage tax, product placement restrictions and health warning labels(Reference Mcgreevy26). Ultimately, none of these measures passed(Reference Mcgreevy26).
In addition to no overall decrease in SSB consumption, we found concerning racial-ethnic disparities. Our analysis is consistent with our previous study examining SSB intake in children in California from 2003 to 2009, which found elevated consumption of SSB among Latino children relative to White children(Reference Beck, Patel and Madsen17). Studies conducted in Massachusetts, Oregon and New York City have also found higher consumption of SSB among both Latino and African American preschoolers relative to White children(Reference Garnett, Rosenberg and Morris16,Reference Elfassy, Adjoian and Lent20,Reference Taveras, Gillman and Kleinman27) .
Racial-ethnic differences in SSB intake may be related to socioeconomic status. Lower income has been associated with higher SSB intake in prior studies(Reference Han and Powell28), and Latino and African American children in California are more likely than White children to live in poverty(29). Other factors that may contribute to racial-ethnic disparities in child SSB consumption include increased marketing of SSB to minorities(Reference Harris, Shehan and Gross30), lower awareness of the health effects of SSB(Reference Bogart, Cowgill and Sharma31) and mistrust of local water supplies(Reference Park, Ayala and Sharkey32–Reference Beck, Takayama and Halpern-Felsher34). Qualitative studies with low-income Latino parents provide insights about beliefs and cultural factors that contribute to child intake of SSB. Key findings include that low-income Latino parents typically recognise the negative health effects of soda, but often have misconceptions about other SSB, believing, for example, that beverages labelled as ‘all natural’ are healthy despite added sugar(Reference Scherzer, Barker and Pollick33,Reference Beck, Takayama and Halpern-Felsher34) . Among Latino immigrant families, high SSB consumption may reflect SSB consumption patterns in countries of origin; Mexico, the most common country of origin of California Latino immigrants(Reference Johnson and Sanchez35), has one of the highest rate of SSB consumption in the world(36).
There are a number of limitations to our study including the fact that SSB intake was determined via parental report and only asked about the day prior to the survey, which may not represent a child’s typical intake. In addition, the survey does not provide data on whether beverages were consumed at home, in school, or in another setting, information which is important for determining how to best intervene to reduce child SSB intake. Strengths of our study include the population-level design and the ability to demonstrate trends in intake over time.
Notwithstanding the limitations noted above, our study has important implications. It suggests that the reduction in SSB intake among California children seen from 2003 to 2009 has plateaued and that additional policy measures are needed to address child SSB consumption. Specific policy measures that have been recently endorsed by the American Academy of Pediatrics include increasing the price of sugary beverages through excise taxes, decreasing sugary drink marketing to children and ensuring access to credible nutrition information(Reference Muth, Dietz and Magge13). Sugary beverage taxes have been implemented at the local level in several US cities, including four cities in California. Evaluations of sugary beverage taxes in Berkeley (California)(Reference Falbe, Thompson and Becker37,Reference Lee, Falbe and Schillinger38) , Philadelphia(Reference Cawley, Frisvold and Hill39), Seattle(Reference Powell and Leider40) and Cook County, Illinois(Reference Powell, Leider and Leger41) have all demonstrated reductions in SSB intake to various degrees. Latin American countries have instituted multiple strategies to reduce SSB consumption(Reference Bergallo, Castagnari and Fernandez42). A suite of efforts to reduce added sugar intake was recently implemented in Chile including front of package warning labels, restrictions on marketing to children and a ban on sales of beverages and foods exceeding a specific threshold of sugar in schools(Reference Taillie, Reyes and Colchero43). A recent evaluation of this policy package found significant reductions in purchases of SSB(Reference Taillie, Reyes and Colchero43). SSB consumption in Mexico also declined following introduction of a beverage tax(Reference Colchero, Guerrero-Lopez and Molina44). Finally, given that parental knowledge(Reference Goodell, Pierce and Amico45,Reference SanGiovanni, Fallar and Green46) and attitudes(47) about SSB are associated with child beverage intake, statewide educational campaigns directed at parents are another important avenue to reduce SSB consumption among children in California. Such efforts should be attentive to disparities and ensure that materials and campaigns are culturally and linguistically appropriate to diverse groups of parents.
Acknowledgements
Acknowledgments: None. Financial support: This study was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (1K23HD080876-01A1) and the National Heart Lung and Blood Institute (1K01HL129087-01A1). Neither NICHD nor NHLBI had any role in the design, analysis or writing of this paper. Conflicts of interest: None. Authorship: Dr A.L.B. came up with the research question, conducted the analysis and drafted the manuscript. Dr S.M. contributed to the analysis and critically reviewed the manuscript. Dr A.I.P. contributed to the analysis and critically reviewed the manuscript. Dr A.F. contributed to the manuscript and critically reviewed the manuscript. Ethics of human subject participation: This study was collected using publicly available data files from the California Health Interview Survey.
