Estimated Reductions in Added Sugar Intake among US Children and Youth in Response to Sugar Reduction Targets

Abstract

Background

In 2021, the National Salt and Sugar Reduction Initiative (NSSRI) released voluntary sugar reduction targets for packaged foods and drinks in the United States.

Objective

The objectives of this study were to describe trends in added sugar intake from NSSRI foods and beverages among children and youth and estimate possible reductions if industry were to meet the targets.

Design

This study consisted of cross-sectional and trend analyses of demographic and 24-hour dietary recall data from eight survey cycles (2003–2004 to 2017–2018) of the National Health and Nutrition Examination Survey.

Participants/setting

The study sample included 23,248 children and youth (aged 2 to 19 years).

Main outcome measures

The main outcome measure was the percent of daily calories from added sugar for foods and beverages in NSSRI categories.

Statistical analyses performed

Foods and beverages reported by participants were mapped to one of the NSSRI’s categories or coded as a non-NSSRI item. Trends over time in added sugar intake were assessed using regression models. To assess possible reductions in added sugar intake if industry were to meet the targets, sales-weighted mean percent reductions for 2023 and 2026 targets were applied to NSSRI items in the 2017–2018 National Health and Nutrition Examination Suvey data. Results were examined overall and by demographic characteristics.

Results

From 2003–2004 to 2017–2018, added sugar intake from NSSRI foods and beverages declined, but consumption remained high. During 2017–2018, NSSRI categories accounted for 70% of US child and youth added sugar intake. If industry met the NSSRI targets, US children and youth would consume 7% (2023 targets) to 21% (2026 targets) less added sugar.

Conclusions

Although added sugar intake from NSSRI foods and drinks has declined over the past decade, added sugar intake from all sources remains high and consumption of added sugar from certain NSSRI categories has remained steady over time. If met, the NSSRI targets are expected to result in meaningful reductions in added sugar intake for US children and youth.

Added sugar intake among children and youth in the United States (US) is high, with 65% of those aged 2 to 19 years not meeting the 2020–2025 Dietary Guidelines for Americans’ recommendation to limit added sugar to <10% of total energy intake.^1,2 A large body of research links added sugar intake to adverse health outcomes, including weight gain, diabetes, cardiovascular disease risk factors, and dental caries.³ Reducing added sugar intake among children and youth is critical to achieving a healthier generation and reducing the nation’s burden of diet-related diseases.^2–4

Excessive intake of added sugar is driven by many factors, including the widespread availability of sweetened food and beverages, the high sugar content of these products, and their ubiquitous marketing.^5–7 Thus, a meaningful reduction in added sugar in the US population will likely require a suite of complementary, multilevel strategies, which could include government policy, consumer education and counter-marketing, and industry efforts to reduce added sugar in the food supply. In recent years, some progress has been made to implement sugar-reduction policies. For example, in 2016, the Food and Drug Administration published final rules on the Nutrition Facts label, which includes new information about added sugars.⁸ In the same year, the US Department of Agriculture published final rules on Nutrition Standards for Foods Sold in Schools, which prohibit the sale of sugary drinks and set sugar limits on snack foods.⁹ There has also been momentum in select states and municipalities, including sugar-sweetened beverage (SSB) excise taxes¹⁰ and healthy beverage ordinances requiring restaurants to offer only healthy beverages instead of SSBs with children’s meals.^11–13 These measures, together with growing media and public recognition of harms of SSBs in particular, have likely contributed to gradually declining added sugar consumption among US children and adults.^14,15 However, these declines are largely attributable to reductions in SSB intake (which still remains high), whereas decreases in added sugar from foods have been much smaller.^14,15

The National Salt and Sugar Reduction Initiative (NSSRI) is a partnership of more than 100 local, state, and national health groups convened by the New York City Department of Health and Mental Hygiene to encourage reductions in the sodium and sugar content of packaged foods. By creating changes at the level of the food supply, the initiative seeks to make it easier for all individuals to access healthier options, an upstream approach that may mitigate existing disparities in diet-related diseases.^16,17 Through analysis of national sales data, nutrition information, meetings with industry, and two public comment periods, the NSSRI developed voluntary sugar reduction targets for industry across 15 categories of packaged foods and drinks.¹⁷ The creation of sugar reduction targets was based on the moderate success of the National Salt Reduction Initiative,¹⁷ which itself was modeled on the United Kingdom’s approach to reducing sodium.¹⁸ Using data from highest-selling food and drink products in 2018 as the baseline, the 2023 NSSRI sugar reduction targets are a 10% reduction in sugar density (grams of sugar per 100 g food or per 100 mL beverage), whereas the 2026 targets are 40% for sugary drinks and 20% for other categories. Industry is encouraged to meet NSSRI targets; to do so, the mean sugar density of a company’s products must be at or below the target. Companies can influence mean sugar density by reformulating existing products to be lower in total sugar, increasing sales of lower total sugar products, and introducing new, lower total sugar products.¹⁹

The public health influence of the NSSRI relies on whether or not and to what extent the targets capture major sources of added sugar in the US diet. It is also of importance to document temporal trends in consumption of NSSRI foods and beverages before the initiative launch–understanding whether or not added sugar intake from NSSRI items is already changing will inform future evaluation efforts and can identify NSSRI foods and beverages for which added sugar intake is not already decreasing. It is particularly important to examine these gaps in knowledge about trends in sugary intake for US children and youth, given that taste preferences develop in childhood.^20,21 Ensuring healthier options may help shape dietary preferences, benefitting individuals throughout the life course.

Thus, the objectives of this study were to use nationally representative data to describe trends in added sugar intake from NSSRI foods and beverages among children and youth aged 2 to 19 years between 2003–2004 and 2017–2018, document to what extent food and drinks included in the NSSRI account for added sugar intake in the most recent years of data (2017–2018), and estimate possible reductions in added sugar intake in the case that industry had met the NSSRI sugar reduction targets in the most recent years of data (2017–2018). To understand potential effects that the initiative may have on diet-related disparities, differences in the results for these aims were examined by sociodemographic characteristics.

METHODS

Data and Study Population

This study pooled data from eight survey cycles (2003–2004 to 2017–2018) of the National Health and Nutrition Examination Surveys (NHANES), a repeated cross-sectional study released every 2 years and designed to represent the US noninstitutionalized population.²² The study sample consisted of participants aged 2 to 19 years with complete data on all covariates and a valid first 24-hour dietary recall. Although the NHANES administered two 24-hour dietary recalls, the analysis was limited to the first 24-hour recall to preserve sample size for subgroup analyses. Because this study analyzed de-identified publicly available data, it does not constitute human subjects research and institutional review board approval was not required.

Measures

Added Sugar Intake.

Survey respondents reported all foods and beverages consumed during the previous 24-hour period, specifying the type, quantity, and source of each intake occasion. This 24-hour dietary recall interview was conducted in-person by trained dietary interviewers on either a weekday or weekend day, utilizing the US Department of Agriculture Automated Multiple-Pass Method, a 5-step approach designed to enhance accurate food recall and reduce respondent burden.²³ Responses for participants aged 2 to 11 years were provided or assisted by a parent/guardian, whereas participants aged 12 years and older responded independently. All reported foods and beverage items are linked to the US Department of Agriculture Food and Nutrient Database for Dietary Studies (FNDDS) to obtain total calories and the Food Patterns Equivalents Database to obtain added sugar.^24,25

Given declines in reported energy intake over time in NHANES and concerns about measurement error,^26,27 the trends analyses were energy-adjusted by using percent of daily calories from added sugar as the primary outcome. This was calculated by dividing calories from added sugar by total calories for each participant. This mean ratio approach was chosen (in contrast to population ratio approach, which is calculated by summing the calories from added sugar for all participants and then dividing that by the sum of total calories for all participants) because the goal of the analysis was to estimate the reduction in added sugar resulting from the NSSRI sugar reduction targets at the individual vs population level. Total intake of grams of added sugar was also examined as a secondary outcome, energy-adjusted by including total calories as a continuous covariate in regression models.

Developing the NSSRI Categories.

The NSSRI categories were developed through analysis of national sales and nutrition information and considered added sugar contribution, opportunities and technical challenges for sugar reduction, and comments from industry. Briefly, the NSSRI includes 15 packaged food and beverage categories aggregated to form seven meta-categories (Table 1 available at www.jandonline.org).^19,28 In 2018, baseline sales-weighted mean (SWM) sugar density was calculated for each category using 2017 sales data from Nielsen and 2018 nutrition data from Label Insight and manufacturer websites.^29,30 SWM sugar density was calculated by dividing each product’s sugar content in grams by its weight in 100-g units (or volume in 100-mL units for liquids) and multiplying by the product’s percent unit sales in the category. Thus, more frequently purchased foods and beverages contribute more to the SWM sugar density than less frequently purchased foods and beverages.

Table 1.

National Salt and Sugar Reduction Initiative categories and subcategories^a

Packaged food category	Category description	Baseline 2018 sales-weighted meanb (g sugar per 100 g [or 100 mL])	Sales-weighted mean targets (g sugar per 100 g [or 100 mL])
			2023	2026
Drinks
Sugary drinks	Soda, sports drinks, fruit drinks, energy drinks, tea. Excludes 100% juice and drinks with milk or milk substitute as a first or second ingredient	8.9	8.0	5.3
Sweetened milk	Drinks containing milk as a first or second ingredient	6.0	5.4	4.8
Sweetened milk substitute	Flavored drinks containing milk substitute as a first or second ingredient	3.6	3.3	2.9
Grain-based desserts
Breakfast pastries	Donuts, cinnamon rolls, coffee cakes, Danishes, streusel, muffins, pies, and toaster pastries	27.2	24.5	21.7
Cakes	Cakes, cupcakes, brownies, and snack cakes	39.9	35.9	31.9
Cookies	Filled and unfilled cookies, sandwich cookies, and tea biscuits	35.6	32.1	28.5
Dry mixes	Dry mixes for cake, cookies, brownies, and muffins	49.4	44.5	39.6
Granola bars	Granola bars, cereal bars, breakfast bars, yogurt bars, and protein bars	27.1	24.4	21.7
Refrigerated and frozen desserts
Refrigerated and frozen desserts	Ice cream, frozen yogurt, gelato, ice pops, sherbet, sorbet, and premade pudding and gelatin. Excludes pudding and gelatin dry mixes	20.7	18.7	16.6
Candies
Sweet candies	Chewy, gummy, and hard candies. Includes caramels, nut rolls, and seasonal sweet candies	59.2	53.3	47.4
Chocolate candies	Chocolate bars, chocolate candies, filled bars, and seasonal chocolate	53.6	48.2	42.9
Breakfast cereals
Breakfast cereals	Ready-to-eat cereal, granola, and hot cereal	27.4	24.7	22.0
Condiments and toppings
Condiments	Ketchup, barbecue sauce, salad dressing, steak sauce, and Asian sauces	21.8	19.7	17.5
Dessert syrups and toppings	Chocolate and caramel syrups, fruit syrups and spreads, chocolate and hazelnut spreads, marshmallow topping, pancake syrup, and frosting. Excludes 100% maple syrup	54.3	48.9	43.4
Yogurt
Yogurt	Dairy and nondairy yogurt and yogurt drinks	6.5	5.9	5.2

^aAdapted with permission from the Revised Voluntary Sugar Reduction Targets from the National Salt and Sugar Reduction Initiative document published by the New York City Department of Health and Mental Hygiene on July 25, 2019.

^bCalculated in 2018 using 2017 sales data from Nielsen and 2018 nutrition data from Label Insight and manufacturer websites.^29,30

Applying the NSSRI Categories to NHANES Data.

The FNDDS codes corresponding to foods and beverages reported by NHANES participants were hand coded by one author as a non-NSSRI item or mapped to one of the NSSRI’s 15 food and beverage categories using added sugar amounts and item descriptions. This coding was checked in a nonindependent manner by two other authors and any discrepancies were discussed collaboratively as a team. Because the NSSRI sets targets for packaged items, the definition of NSSRI foods was restricted to those reported by participants to be acquired from stores (grocery, supermarket, and convenience stores) or vending machines. Foods in NSSRI categories acquired from other sources (eg, restaurants) were considered non-NSSRI foods. Sugary beverages were allowed to be obtained from any source.

Covariates.

To adjust for potential demographic shifts over time, analyses included the following covariates: age group (2 to 5 years, 6 to 11 years, and 12 to 19 years), sex (male or female), race/ethnicity (non-Hispanic White, non-Hispanic Black, Mexican American, other Hispanic, or other race/ethnicity), annual family income (<130% of the Federal Poverty Guidelines, ≥130% of the Federal Poverty Guidelines³¹), and parental educational attainment (less than college graduate, college graduate, or above). Other race/ethnicity included individuals reporting a race other than White or Black or individuals reporting multi-racial identity. Information on these covariates was collected from participants by trained NHANES interviewers via an at-home interview.

Analyses

Trends Analyses.

To examine changes in the percent of daily calories from added sugar for each NSSRI category and meta-category over time, linear regression models were used. In these models, the outcome variable was percent of daily calories from added sugar from each NSSRI category and meta-category and covariates were a categorical survey year term, age group, sex, race/ethnicity, annual family income, and parental educational attainment. To analyze the statistical significance of trends over time, models were also fit with a continuous survey year term. To assess potential nonlinearity in trends over time, quadratic and cubic year terms were also included as covariates, and a joint Wald test of the quadratic and cubic terms was performed. In the case that the test was statistically significant, the results from this model were reported. If not, no evidence of nonlinearity was concluded, and a model including only a linear term was fitted. From these models, the beta coefficient, 95% CI, and the P value for linear year term were reported. To account for multiple testing, a P value < 0.001 was considered statistically significant. In addition to examining changes over time, estimates of the average percent of daily calories from added sugar for each NSSRI category and meta-category for each survey year were also generated. To obtain estimates within subgroups, separate models were fitted within each subgroup, adjusting for all other covariates. Trends analyses with secondary outcomes (added sugar intake in grams and total quantity of foods or beverages in grams) were conducted in an analogous manner.

Estimated Reductions.

To estimate reductions in added sugar intake (in grams) if companies were to meet the 2023 and 2026 targets, the sample was restricted to the most recently available data (2017–2018). Next, following the approach of a previous study,³² the ratio of the target to baseline SWM sugar density was calculated for each NSSRI category. For example, the baseline SWM sugar density for the breakfast pastries categories was 27.2 g per 100 g, whereas the 2023 target was 24.5 g per 100 g; thus, the ratio of the 2023 target to baseline was 0.90 (ie, a 10% reduction). Generally, the 2023 targets reflect a 10% reduction in the SWM sugar density for both food and drinks, whereas the 2026 targets reflect a 40% reduction in the SWM sugar density for sugary drinks and a 20% reduction for other categories. Next, the applicable ratio of the target to the baseline SWM sugar density was multiplied by the amount of added sugar reported for that NSSRI category by each participant. For example, in the case that a participant reported consuming a total of 5 g sugar from the breakfast pastries category, their predicted intake under the 2023 targets would be 0.90 × 5 g = 4.5 g. Descriptive statistics (means and percent change) were used to summarize added sugar intake pre-NSSRI (2017–2018 data) and under 2023 and 2026 targets, overall and by subgroup.

Sensitivity Analyses

Two sensitivity analyses related to allowable food sources for NSSRI items were conducted. The first sensitivity analysis only allowed sugary beverages acquired from stores and vending machines to be considered an NSSRI item, taking a more conservative approach and assuming no reformulation of beverages from nonstore sources. The second sensitivity analysis placed no restrictions on food or beverage sources, taking a less conservative approach and assuming total reformulation of foods and beverages from all sources.

All analyses were conducted in 2020 using Stata version 14.2.³³ In addition, all analyses were weighted to account for the multistage, clustered probability sampling of the NHANES.

RESULTS

From a total of 29,712 eligible children and youth in NHANES aged 2 to 19 years, the final analytic sample included 23,248 participants with complete data on dietary intake and covariates. Participants were excluded in the case that they were missing any data on dietary intake (n = 4,038), family income (n = 2,330), and/or parental educational attainment (n = 1,273) (note: participants could be excluded for multiple reasons). Table 2 reports unweighted sample sizes and weighted proportions by demographic characteristics.

Table 2.

Sample characteristics of children and youth aged 2 to 19 years in the National Health and Nutrition Examination Survey 2003–2004 to 2017–2018 (n = 23,248)

Characteristic	n (%)a
Sex
Male	11,722 (51)
Female	11,526 (49)
Age (y)
2–5	5,522 (22)
6–11	7,528 (33)
12–19	10,198 (45)
Race/ethnicity
Non-Hispanic White	6,915 (57)
Non-Hispanic Black	6,256 (14)
Mexican American	5,750 (14)
Other Hispanic	1,922 (6)
Other race/ethnicityb	2,405 (8)
Annual family income
Income <130% of US federal poverty guidelines	10,311 (34)
Income ≥130% of US federal poverty guidelines	12,937 (66)
Parental education attainment
Less than college graduate	18,846 (74)
College graduate or above	4,402 (26)

^aReported as unweighted sample size (weighted proportion).

^bOther race/ethnicity included individuals reporting a race other than White or Black, including Asians or individuals reporting multiracial identity.

Between 2003–2004 and 2017–2018, added sugar intake from all NSSRI foods and beverages as a percent of daily calories declined (14.0% to 10.4%; P for trend < 0.001), driven primarily by a reduction in percent calories from added sugar in drinks (9.0% to 5.8%; P for trend < 0.001) (Figure 1 and Table 3). There was also a decrease over time in percent calories from added sugar for NSSRI foods (5.0% to 4.6%; P for trend = 0.001), although many of the individual food categories did not experience a significant decrease. Trends were similar when examining grams of added sugar from NSSRI items as the outcome (Table 4, available at www.jandonline.org). There was also a decline in the quantity (total grams) of NSSRI foods and drinks consumed by participants (Table 5, available at www.jandonline.org).

Figure 1.

Trends in added sugar intake as a percent of total energy intake^a from National Salt and Sugar Reduction Initiative (NSSRI) food, drinks, and nonincluded items, between 2003–2004 and 2017–2018 for children and youth in the National Health and Nutrition Examination Survey (NHANES). ^aOutcome was constructed as (grams of added sugar per day × 4)/(total calories per day) × 100. The margins command in Stata³³ was used to estimate the predicted percent of daily calories from added sugar from each category and meta-category for each survey year, when all other covariates were set to their mean values. ^bNSSRI = National Salt and Sugar Reduction Initiative.

Table 3.

Trends in added sugar intake as a percent of total energy intake^a (95% CI) from each National Salt and Sugar Reduction Initiative (NSSRI) category to meta-category to and all categories combined between 2003–2004 and 2017–2018 for children and youth in the National Health and Nutrition Examination Survey

	2003–2004	2005–2006	2007–2008	2009–2010	2011–2012	2013–2014	2015–2016	2017–2018	P value for linear trendb	Beta coefficient (95% CI) for linear year term
	point estimate (95% CI)
All NSSRI categories combined	14.0 (13.1 to 15.0)	13.0 (12.4 to 13.6)	12.8 (12 to 13.6)	12.1 (11.4 to 12.8)	11.4 (11 to 11.8)	11.2 (10.4 to 12.0)	9.9 (9.4 to 10.4)	10.4 (9.7 to 11.1)	< 0.001	−0.28 (−0.33 to −0.22)
All NSSRI foods combined (no drinks)	5.0 (4.7 to 5.3)	5.1 (4.8 to 5.3)	4.9 (4.6 to 5.3)	4.6 (4.3 to 4.9)	4.2 (4.1 to 4.4)	4.7 (4.4 to 5)	4.3 (4 to 4.6)	4.6 (4.2 to 5.1)	0.001	−0.05 (−0.07 to −0.02)
Drinks	9.0 (8.2 to 9.8)	7.9 (7.3 to 8.5)	7.9 (7.1 to 8.6)	7.5 (6.9 to 8.1)	7.2 (6.8 to 7.6)	6.5 (5.7 to 7.3)	5.5 (5.1 to 5.9)	5.8 (5.3 to 6.3)	< 0.001	−0.23 (−0.28 to −0.18)
Sugary drinks	8.6 (7.8 to 9.4)	7.6 (7 to 8.1)	7.5 (6.8 to 8.3)	7 (6.4 to 7.7)	6.9 (6.4 to 7.3)	6.2 (5.4 to 7.1)	5.3 (4.9 to 5.7)	5.5 (5.1 to 6)	< 0.001	−0.22 (−0.27 to −0.17)
Sweetened milk	0.4 (0.3 to 0.4)	0.2 (0.2 to 0.3)	0.3 (0.2 to 0.4)	0.4 (0.3 to 0.5)	0.3 (0.2 to 0.4)	0.2 (0.2 to 0.3)	0.2 (0.2 to 0.3)	0.2 (0.1 to 0.3)	0.001	−0.01 (−0.02 to 0)
Sweetened milk substitute	0 (0 to 0)	0.1 (0 to 0.1)	0 (0 to 0.1)	0 (0 to 0.1)	0.1 (0 to 0.1)	0 (0 to 0.1)	0 (0 to 0.1)	0 (0 to 0.1)	0.42	0 (0 to 0)
Grain-based desserts and snack bars	1.4 (1.2 to 1.5)	1.6 (1.3 to 1.8)	1.5 (1.3 to 1.7)	1.5 (1.4 to 1.6)	1.6 (1.4 to 1.7)	1.5 (1.3 to 1.8)	1.6 (1.4 to 1.7)	1.5 (1.3 to 1.8)	0.335	0.01 (−0.01 to 0.02)
Breakfast pastries	0.4 (0.3 to 0.4)	0.5 (0.4 to 0.6)	0.5 (0.4 to 0.6)	0.5 (0.4 to 0.6)	0.5 (0.4 to 0.6)	0.4 (0.3 to 0.5)	0.5 (0.3 to 0.6)	0.4 (0.3 to 0.5)	0.46	0 (−0.1 to 0)
Cakes	0.5 (0.3 to 0.6)	0.5 (0.4 to 0.6)	0.5 (0.3 to 0.6)	0.4 (0.3 to 0.5)	0.4 (0.2 to 0.5)	0.4 (0.3 to 0.5)	0.4 (0.3 to 0.5)	0.5 (0.3 to 0.7)	0.523	0 (−0.01 to 0.01)
Cookies	0.4 (0.3 to 0.5)	0.5 (0.4 to 0.5)	0.4 (0.3 to 0.5)	0.5 (0.4 to 0.5)	0.6 (0.5 to 0.7)	0.5 (0.4 to 0.6)	0.5 (0.5 to 0.6)	0.5 (0.4 to 0.6)	0.012	0.01 (0 to 0.02)
Dry mixes	NAc	NA	NA	NA	NA	NA	NA	NA	NA	NA
Granola bars	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.2)	0.2 (0.1 to 0.2)	0.2 (0.1 to 0.2)	0.1 (0.1 to 0.2)	0.1 (0.1 to 0.2)	0.015	0 (0 to 0.01)
Refrigerated and frozen desserts	0.9 (0.7 to 1.2)	1 (0.8 to 1.2)	1.1 (0.8 to 1.3)	0.8 (0.7 to 0.9)	0.5 (0.4 to 0.6)	0.6 (0.5 to 0.7)	0.6 (0.5 to 0.7)	0.6 (0.5 to 0.8)	< 0.001	−0.04 (−0.05 to −0.02)
Candies	1 (0.8 to 1.2)	0.8 (0.7 to 0.9)	0.8 (0.7 to 0.9)	0.7 (0.6 to 0.8)	0.6 (0.5 to 0.7)	1 (0.8 to 1.1)	0.7 (0.6 to 0.8)	0.8 (0.6 to 0.9)	0.036	−0.01 (−0.02 to 0)
Sweet candies	0.6 (0.5 to 0.7)	0.5 (0.4 to 0.6)	0.5 (0.4 to 0.6)	0.4 (0.3 to 0.4)	0.3 (0.3 to 0.4)	0.7 (0.6 to 0.8)	0.4 (0.3 to 0.5)	0.6 (0.5 to 0.7)	0.699	0 (−0.01 to 0.01)
Chocolate candies	0.4 (0.3 to 0.5)	0.3 (0.3 to 0.4)	0.3 (0.3 to 0.4)	0.3 (0.2 to 0.4)	0.2 (0.2 to 0.3)	0.3 (0.2 to 0.5)	0.3 (0.2 to 0.3)	0.2 (0.1 to 0.3)	0.003	−0.01 (−0.02 to 0)
Breakfast cereals	1.1 (1 to 1.2)	1 (0.9 to 1.1)	0.9 (0.8 to 1)	1 (0.8 to 1.1)	0.9 (0.8 to 0.9)	0.9 (0.8 to 1)	1 (0.9 to 1.1)	1 (0.9 to 1.1)	0.084	−0.01 (−0.02 to 0)
Condiments and toppings	0.4 (0.3 to 0.5)	0.5 (0.4 to 0.6)	0.4 (0.3 to 0.6)	0.4 (0.3 to 0.5)	0.5 (0.3 to 0.7)	0.5 (0.4 to 0.5)	0.4 (0.3 to 0.6)	0.6 (0.4 to 0.7)	0.067	0.01 (0 to 0.02)
Condiments	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.2)	0.1 (0.1 to 0.1)	0.2 (0.1 to 0.2)	0.057	0 (0 to 0)
Dessert syrups and toppings	0.3 (0.2 to 0.4)	0.4 (0.3 to 0.4)	0.3 (0.2 to 0.4)	0.3 (0.2 to 0.4)	0.4 (0.2 to 0.6)	0.3 (0.3 to 0.4)	0.3 (0.2 to 0.5)	0.4 (0.3 to 0.6)	0.17	0.01 (0 to 0.02)
Yogurt	0.2 (0.1 to 0.3)	0.2 (0.2 to 0.3)	0.2 (0.1 to 0.2)	0.3 (0.2 to 0.3)	0.2 (0.2 to 0.3)	0.2 (0.2 to 0.3)	0.1 (0.1 to 0.1)	0.1 (0.1 to 0.2)	0.006	−0.01 (−0.01 to 0)
Non-NSSRI items	3.6 (3.4 to 3.9)	4.2 (3.9 to 4.5)	4.3 (4 to 4.5)	4.4 (4.2 to 4.7)	4.7 (4.3 to 5.1)	4 (3.9 to 4.2)	4.4 (4 to 4.9)	4.3 (4.1 to 4.6)	Nonlineard	NA

^aOutcome was constructed as (grams of added sugar per day × 4)/(totalcalories per day) × 100. The margins command in Stata³³ was used to estimate the predicted percent of daily calories from added sugar from each category and meta-category for each survey year to when all other covariates were set to their mean values. Any differences in estimates from Figure 1 are due to rounding.

^bTo assess the presence of nonlinearity in trends to linear to quadratic to and cubic year terms were included in the models. A joint Wald test of the significance of the quadratic and cubic terms was conducted. If these terms were significant to nonlinearity was concluded. If these terms were not significant to no evidence of nonlinearity was concluded and the results from a model that only included a linear year term were reported. To account for multiple testing toa P value < 0.001 was considered statistically significant.

^cNA = not applicable.

^dEvidence of a nonlinear trend in added sugar intake from non-NSSRI items over time to as indicated by statistically significant joint Wald test of the quadratic and cubic terms for survey year (P < 0.001).

Table 4.

Energy-adjusted^a trends in daily added sugar intake (in grams) (95% CI) from each National Salt and Sugar Reduction Initiative (NSSRI) category, meta-category, and all categories combined between 2003–2004 and 2017–2018 for children and youth in the National Health and Nutrition Examination Survey

	2003–2004	2005–2006	2007–2008	2009–2010	2011–2012	2013–2014	2015–2016	2017–2018	P value for linear trendb
	point estimate (95% CI)
All NSSRI categories combined	70.1 (65.6–74.7)	65.4 (61.1–69.8)	63.4 (59.1–67.8)	61.9 (58.6–65.2)	55.5 (54.1–57)	55.4 (52.4–58.4)	49.7 (47.2–52.2)	52.3 (48.9–55.7)	< 0.001
All NSSRI foods combined (not drinks)	24.8 (23–26.5)	24.6 (23.3–25.9)	24.2 (22–26.4)	23 (21.4–24.5)	20.5 (19.3–21.7)	22.6 (21.1–24.1)	21.4 (20–22.8)	23.5 (21.4–25.6)	0.005
Drinks	45.4 (41.7–49.1)	40.8 (36.6–45.1)	39.2 (35.7–42.8)	39 (36–41.9)	35 (33.2–36.8)	32.8 (29.6–36)	28.3 (26.2–30.4)	28.8 (26.8–30.8)	< 0.001
Sugary drinks	43.5 (39.8–47.2)	39.5 (35.3–43.6)	37.7 (34–41.3)	36.9 (33.7–40)	33.6 (31.6–35.5)	31.5 (28.2–34.7)	27.1 (25–29.1)	27.7 (25.9–29.5)	< 0.001
Sweetened milk	1.8 (1.3–2.3)	1.1 (0.9–1.4)	1.4 (1.1–1.7)	2 (1.5–2.5)	1.2 (0.7–1.7)	1.1 (0.9–1.3)	1 (0.8–1.2)	1 (0.6–1.3)	0.001
Sweetened milk substitute	0.1 (0–0.1)	0.2 (0–0.4)	0.2 (0.1–0.3)	0.2 (0.1–0.3)	0.2 (0.1–0.4)	0.2 (0–0.5)	0.2 (0.1–0.3)	0.2 (0.1–0.3)	0.194
Grain-based desserts and snack bars	6.8 (5.9–7.6)	7.8 (6.6–8.9)	7.6 (6.5–8.6)	7.6 (7–8.1)	8 (7.1–9)	8.1 (6.8–9.4)	8.1 (7.4–8.9)	8.2 (7–9.5)	0.049
Breakfast pastries	1.8 (1.5–2.2)	2.4 (1.8–2.9)	2.7 (2.2–3.3)	2.7 (2.2–3.2)	2.6 (2–3.1)	2.2 (1.7–2.6)	2.6 (1.8–3.3)	2.4 (1.8–2.9)	0.422
Cakes	2.5 (1.8–3.2)	2.8 (2–3.6)	2.5 (1.7–3.2)	1.9 (1.4–2.4)	1.9 (1.2–2.5)	2.5 (1.8–3.3)	2.1 (1.6–2.7)	2.8 (1.7–3.8)	0.78
Cookies	2 (1.5–2.4)	2.1 (1.8–2.5)	1.9 (1.4–2.4)	2.4 (2–2.7)	2.9 (2.4–3.4)	2.6 (2.1–3.1)	2.8 (2.3–3.3)	2.5 (2–2.9)	0.003
Dry mixes	NAc	NA	NA	NA	NA	NA	NA	NA	NA
Granola bars	0.5 (0.2–0.7)	0.4 (0.3–0.6)	0.5 (0.3–0.7)	0.5 (0.4–0.7)	0.7 (0.4–1)	0.8 (0.6–1)	0.7 (0.5–0.9)	0.6 (0.4–0.9)	0.015
Refrigerated and frozen desserts	4.5 (3.4–5.6)	4.7 (3.8–5.6)	5.5 (4.1–6.8)	4.1 (3.6–4.7)	2.4 (2–2.9)	3 (2.2–3.8)	2.8 (2.3–3.3)	3.2 (2.4–4.1)	< 0.001
Candies	5.1 (4.1–6)	4.1 (3.4–4.7)	4.1 (3.7–4.5)	3.5 (2.9–4.1)	2.7 (2.2–3.2)	4.6 (3.9–5.3)	3.4 (2.9–4)	3.9 (3.3–4.6)	0.032
Sweet candies	3 (2–3.9)	2.4 (1.8–3)	2.4 (1.9–2.9)	2 (1.6–2.4)	1.5 (1.1–1.8)	2.9 (2.4–3.5)	1.9 (1.6–2.2)	2.7 (2.1–3.2)	0.382
Chocolate candies	2.1 (1.6–2.6)	1.6 (1.5–1.8)	1.7 (1.4–2)	1.5 (1.2–1.8)	1.2 (0.9–1.5)	1.7 (0.8–2.5)	1.5 (1.1–2)	1.2 (0.9–1.6)	0.024
	5.6 (4.8–6.3)	4.7 (4.2–5.2)	4.1 (3.7–4.5)	4.4 (3.7–5.1)	3.9 (3.5–4.3)	3.7 (3.3–4.1)	4.2 (3.7–4.7)	4.7 (4.1–5.2)	Nonlineard
Breakfast cereals
Condiments and toppings	1.8 (1.3–2.3)	2.3 (1.7–2.8)	2.2 (1.6–2.8)	2.2 (1.9–2.6)	2.3 (1.3–3.4)	2.2 (1.9–2.6)	2.4 (1.8–3)	2.9 (2.3–3.6)	0.03
Condiments	0.5 (0.4–0.6)	0.5 (0.4–0.6)	0.6 (0.4–0.7)	0.5 (0.4–0.6)	0.4 (0.3–0.5)	0.5 (0.4–0.6)	0.6 (0.5–0.7)	0.7 (0.6–0.9)	0.035
Dessert syrups and toppings	1.4 (0.8–1.9)	1.8 (1.2–2.3)	1.6 (1.1–2.2)	1.7 (1.3–2.1)	1.9 (0.8–2.9)	1.8 (1.4–2.1)	1.8 (1.2–2.4)	2.2 (1.6–2.9)	0.079
Yogurt	1 (0.6–1.3)	1 (0.8–1.3)	0.7 (0.6–0.9)	1.1 (0.9–1.4)	1.1 (0.8–1.3)	1 (0.7–1.2)	0.5 (0.3–0.6)	0.6 (0.4–0.7)	0.002
Non-NSSRI items	17.6 (16.3–18.9)	20.9 (19.2–22.6)	21.3 (19.9–22.7)	21.4 (20.3–22.5)	23.7 (22–25.5)	20.4 (19.4–21.4)	23.1 (20.7–25.5)	22.1 (20.6–23.5)	< 0.001

^aEnergy adjustment was conducted by including total calories as a continuous covariate in regression models. The margins command in Stata³³ was used to estimate the predicted outcome for each category and meta-category for each survey year, when all other covariates were set to their mean values. Negative predicted values were truncated at zero.

^bTo account for multiple testing, a P value < 0.001 was considered statistically significant. To assess the presence of nonlinearity in trends, linear, quadratic and cubic year terms were included in the models. A joint Wald test ofthe significance ofthe quadratic and cubic terms was conducted. If these terms were significant, nonlinearity was concluded. If these terms were not significant, no evidence ofnonlinearity was concluded and reported the results from a model that only included a linear year term.

^cNA = not applicable.

^dEvidence of a nonlinear trend in added sugar intake from breakfast cereals over time, as indicated by statistically significant joint Wald test ofthe quadratic and cubic terms for survey year (P = 0.0004).

Table 5.

Trends^a in mean daily intake (in grams) of food or drink (95% CI) from each National Salt and Sugar Reduction Initiative (NSSRI) meta-category and all categories combined between 2003–2004 and 2017–2018 for children and youth in the National Health and Nutrition Examination Survey

Category	2003–2004	2005–2006	2007–2008	2009–2010	2011–2012	2013–2014	2015–2016	2017–2018	P value for linear trendb
	point estimate (95% CI)
All NSSRI categories combined	693 (644.4–741.7)	630.5 (568–693.1)	576.5 (538.3–614.6)	560.2 (524.4–596)	514.3 (487.4, 541.1)	464.1 (429.9–498.3)	407.3 (380.4, −434.2)	431.8 (403.8–459.7)	< 0.001
All NSSRI foods combined	106.4 (97.8–115)	106 (99.5–112.5)	105.5 (97.7–113.2)	100 (91.9–108.1)	87.4 (82–92.9)	90.9 (84.7–97)	90.1 (84.1–96.2)	96.5 (87.1–105.8)	< 0.001
Drinks	586.6 (542.1–631.2)	524.5 (463.2–585.9)	471 (432.3–509.8)	460.2 (423.8–496.6)	426.8 (399.2–454.4)	373.2 (340–406.4)	317.2 (289.8–344.5)	335.3 (309.5–361.1)	< 0.001
Grain-based desserts and snack bars	25.7 (23.2–28.1)	26.1 (23, 29.2)	25.5 (22.4–28.6)	26.3 (23.5–29.1)	28.1 (25–31.2)	25.5 (21.3–29.6)	27 (23.7–30.3)	27.8 (23.4–32.2)	0.363
Refrigerated and frozen desserts	26.7 (21.5–31.8)	26.4 (22–30.8)	27.2 (22.3–32)	21.6 (18.5–24.6)	14.6 (12–17.2)	17.4 (12.9–21.9)	16.4 (13.5–19.2)	17.1 (12.8–21.5)	< 0.001
Candies	10.3 (8.5–12.1)	9.3 (8.1–10.5)	9.1 (7.9–10.2)	8.2 (7.3–9)	7.1 (5.8–8.4)	8.3 (7–9.7)	6.1 (5.1–7.2)	6.9 (5.7–8)	< 0.001
Breakfast cereals	21.2 (19.1–23.3)	20.2 (17.7–22.7)	21.4 (17.8–25)	18.9 (16.7–21.1)	17.2 (15.3–19)	16.9 (14.8–19.1)	17.9 (15.9–20)	18.2 (16.3–20.2)	0.001
Condiments and toppings	14.4 (9.5–19.2)	14.3 (10.5–18.1)	16 (12–20.1)	14.8 (11.3–18.2)	10.6 (7.2–13.9)	12.2 (9.2–15.2)	13.4 (9.4–17.4)	15.9 (12.5–19.3)	0.704
Yogurt	8.2 (5.9–10.5)	9.6 (7.2–12)	6.3 (4.9–7.8)	10.3 (7.7–12.9)	9.9 (7.9–11.8)	10.6 (8.1–13)	9.3 (6.6–12)	10.6 (7.8–13.3)	0.111
Non-NSSRI	337.6 (320.2–355)	281.8 (269.3–294.2)	280.2 (263.3–297)	289.7 (270.7–308.7)	337 (324.4–349.6)	296.4 (283.7–309)	314.5 (293.5–335.5)	308.7 (290.7–326.8)	Nonlinearc

^aThe margins command in Stata³³ was used to estimate the predicted outcome for each survey year, when all other covariates were set to their mean values.

^bTo accountfor multiple testing, a P value < 0.001 was considered statistically significant.To assessthe presence of nonlinearity in trends, linear, quadratic, and cubicyearterms were included in the models. A joint Wald testofthe significance ofthe quadratic and cubic terms was conducted. If these terms were significant, nonlinearity was concluded. If these terms were not significant, no evidence of nonlinearity was concluded and reported the results from a model that only included a linear year term.

^cEvidence of a nonlinear trend in grams of non-NSSRI items over time, as indicated by statistically significant joint Wald test of the quadratic and cubic terms for survey year (P < 0.001).

Between 2003–2004 and 2017–2018, added sugar as a percent of total calories from NSSRI items declined significantly across all age groups, most racial/ethnic groups, lower and higher income families, and those whose parents had both lower and higher educational attainment (Table 6). Across all years, 2- to 5-year olds had the lowest added sugar intake (as a percentage of total energy intake) from NSSRI categories, whereas 12- to 19-year olds had the highest. With respect to race/ethnicity, non-Hispanic White and Black participants had the highest intake of added sugar as a percentage of total energy intake from NSSRI categories across all years.

Table 6.

Trends in added sugar intake as a percent of total energy intake^a (95% CI) from all National Salt and Sugar Reduction Initiative categories combined between 2003–2004 and 2017–2018 for children and youth in the National Health and Nutrition Examination Survey

	2003–2004	2005–2006	2007–2008	2009–2010	2011–2012	2013–2014	2015–2016	2017–2018	P value for linear trendb	Beta coefficient (95% CI) for linear year term
	point estimate (95% CI)
Age (y)
2–5	11.1 (10.1 to 12.2)	10.4 (9.7 to 11.2)	10.2 (9.6 to 10.8)	9.5 (9.1 to 10)	9.7 (9.1 to 10.4)	9.2 (8.4 to 10)	8.1 (7.4 to 8.9)	8.8 (7.2 to 10.4)	< 0.001	−0.18 (−0.27, −0.10)
6–11	13.2 (12.1 to 14.3)	11.6 (10.8 to 12.4)	12.6 (11.8 to 13.5)	11.3 (10.6 to 12)	11.2 (10.6 to 11.7)	10.4 (9.5 to 11.3)	9.5 (8.7 to 10.3)	10.3 (9.3 to 11.4)	< 0.001	−0.22 (−0.30, −0.15)
12–19	16.1 (14.8 to 17.4)	15.2 (14.4 to 16.1)	14.2 (12.9 to 15.6)	13.9 (12.5 to 15.3)	12.4 (11.6 to 13.3)	12.7 (11.4 to 14.1)	11 (10.2 to 11.9)	11.2 (10.1 to 12.4)	< 0.001	−0.36 (−0.45, −0.28)
Race/ethnicity
Non-Hispanic White	14.4 (13.2 to 15.6)	13.5 (12.4 to 14.5)	13.4 (12.3 to 14.4)	12.5 (11.5 to 13.5)	12.1 (11.5 to 12.7)	11.6 (10.1 to 13)	10.1 (9.4 to 10.8)	10.8 (9.6 to 11.9)	< 0.001	−0.29 (−0.37 to −0.20)
Non-Hispanic Black	15.1 (13.7 to 16.6)	13 (12.1 to 14)	13.5 (12.7 to 14.3)	12.6 (11.5 to 13.8)	11.8 (10.7 to 13)	12.4 (11.2 to 13.5)	11 (10.2 to 11.9)	11.6 (10.8 to 12.4)	< 0.001	−0.24 to (−0.33 to −0.15)
Mexican American	13.8 (12.7 to 14.9)	11.9 (10.8 to 13.1)	11.7 (10.8 to 12.7)	11.8 (10.4 to 13.3)	10.2 (9.1 to 11.3)	9.6 (9.1 to 10.2)	9.5 (8.7 to 10.4)	9.5 (8.2 to 10.7)	< 0.001	−0.30 (−0.38 to −0.21)
Other Hispanic	12.2 (9.5 to 15)	12.3 (10.3 to 14.3)	12.2 (9.8 to 14.6)	11.5 (10 to 13)	10.7 (9.2 to 12.2)	11.8 (10.7 to 13)	8.6 (7.4 to 9.7)	8.1 (7 to 9.2)	< 0.001	−0.32 (−0.46 to −0.17)
Other race/ethnicityc	11.5 (9 to 14)	12.3 (10.5 to 14.1)	9 (7.5 to 10.5)	9 (7.7 to 10.3)	8.8 (7.1 to 10.5)	8.6 (7.6 to 9.6)	7.9 (6.9 to 8.8)	9.3 (8.3 to 10.3)	0.001	−0.21 (−0.34 to −0.08)
Annual family income
Income <130% of	14.2 (13.1 to 15.2)	12.9 (12.2 to 13.6)	13.4 (12 to 14.8)	12.6 (11.8 to 13.3)	12.1 (11.3 to 12.9)	12 (10.6 to 13.3)	10.1 (9.4 to 10.9)	10.9 (9.8 to 12)	< 0.001	−0.25 (−0.33 to −0.16)
US FPG
Income ≥130% of US FPG	14 (13 to 15.1)	13 (12.3 to 13.7)	12.5 (11.5 to 13.4)	11.8 (10.8 to 12.8)	11.1 (10.4 to 11.7)	10.8 (10 to 11.5)	9.7 (9.2 to 10.3)	10.1 (9.3 to 11)	< 0.001	−0.30 (−0.36 to −0.23)
Parental educational attainment
Less than college graduate	14.7 (13.7 to 15.8)	13.4 (12.8 to 14.1)	13.3 (12.5 to 14.1)	12.6 (12 to 13.2)	12.2 (11.6 to 12.8)	11.7 (10.9 to 12.5)	10.3 (9.8 to 10.9)	11 (10.3 to 11.7)	< 0.001	−0.28 (−0.34 to −0.22)
College graduate or above	11.9 (10.5 to 13.3)	11.9 (10.4 to 13.3)	11.3 (10.4 to 12.3)	10.7 (9.3 to 12.1)	9.2 (8.5 to 10)	9.7 (8 to 11.4)	8.4 (7.7 to 9.1)	8.7 (7.3 to 10.2)	< 0.001	−0.27 (−0.38 to −0.17)

^aOutcome was constructed as (grams of added sugar per day × 4)/(totalcalories per day) × 100. The margins command in Stata³³ was used to estimate the predicted percent of daily calories from added sugar from each category and meta-category for each survey year, when all other covariates were setto their mean values. Negative predicted values were truncated at zero. To obtain trend estimates within subgroups, separate models were fitted within each subgroup, adjusting for all other covariates (eg, model fit among non-Hispanic White children, adjusting for age, income and parental educational attainment).

^bTo account for multiple testing, a P value < 0.001 was considered statistically significant. To assess the presence of nonlinearity in trends, linear, quadratic and cubic year terms were included in the models. A joint Wald test ofthe significance ofthe quadratic and cubic terms was conducted. Ifthese terms were significant, nonlinearity was concluded. If these terms were not significant, no evidence of nonlinearity was concluded and reported the results from a model that only included a linear year term.

^cOther race/ethnicity included individuals reporting a race other than White or Black, including Asians, or individuals reporting multiracial identity.

Whereas Tables 3 through 6 reports significant linear decreases for many outcomes, in some instances there was suggestion of an uptick in 2017–2018 compared with the decreasing trend observed in prior years.

In 2017–2018, mean overall added sugar intake among children and youth was 70.8 g, with the majority (49.5 g; 70.0%) of this added sugar intake estimated to come from foods and beverages covered under the NSSRI (Figure 2). In sensitivity analyses varying allowable food sources for NSSRI items, the proportion of added sugar intake comprised by NSSRI items varied from 58% (when only allowing sugary beverages acquired from stores and vending machines to be considered an NSSRI item, taking a more conservative approach and assuming no reformulation of beverages from nonstore sources) to 85% (when placing no restrictions on food or beverage sources, taking a less-conservative approach and assuming total reformulation of foods and beverages from all sources) (data not shown).

Figure 2.

Percentage of daily added sugar intake by National Salt and Sugar Reduction Initiative (NSSRI) meta-category, for children aged 2 to 19 years in the National Health and Nutrition Examination Survey 2017–2018.

Of the 21.3 g (30.0%) of added sugar intake not covered by NSSRI categories in 2017–2018, about half was contributed by items that would otherwise have been eligible for NSSRI, but were obtained from sources that are not covered by the NSSRI (eg, restaurants, cafeterias, and gifts from another household), whereas the remainder came from a variety of food and beverages not covered by the NSSRI such as dips/spreads/sauces, mixed dishes (eg, pasta with tomato sauce), and salty snacks.

Assuming no substitution, daily added sugar intake for US children and youth would have been 7% lower in the case that the 2023 NSSRI targets had been met and would have been 21% lower in the case that the 2026 targets had been met. Estimated reductions were comparable across population subgroups, although these differences were not tested statistically (Figure 3).

Figure 3.

Daily added sugar intake^a in grams in 2017–2018 and intake if 2023 and 2026 National Salt and Sugar Reduction Initiative (NSSRI) targets were met, overall and by sociodemographic characteristics.^b Panel A: By age. Panel B: By race/ethnicity. Panel C: By annual family income and parental educational attainment. ^aAdded sugar intake refers to sum of both NSSRI and non-NSSRI items. ^bTo obtain estimates within subgroups, separate models were fitted within each subgroup, adjusting for all other covariates (eg, model fit among non-Hispanic White children, adjusting for age, income, and parental educational attainment). Other race/ethnicity included individuals reporting a race other than White or Black, including Asians or individuals reporting multiracial identity.

DISCUSSION

During 2017–2018, packaged foods and drinks covered under the NSSRI accounted for 70% of added sugar intake among children and youth in the United States, with drinks comprising the largest proportion at 38%. Although added sugar intake from NSSRI foods and drinks has declined over the past decade, added sugar intake from all sources remains high at about 71 g per day (equivalent to roughly 17 tea-spoons), which is almost triple the 25 g/day limit recommended by the American Heart Association for this age group.³⁴ In addition, consumption of certain NSSRI categories has remained steady over time (or, for some categories, declined but not by a meaningful amount) and there is some suggestion of an uptick in intake in the most recent year of data (2017–2018) compared with prior years. Although many factors contribute to these trends, including widespread availability and promotion of sugary foods and beverages, the findings of this study indicate that reducing sugar in the food supply could play a role in reducing added sugar intake among children and youth. In the case that industry were meeting the NSSRI targets during 2017–2018, children and youth would have consumed 7% (2023 targets) to 21% (2026 targets) less added sugar. Estimated reductions were similar across demographic characteristics, suggesting the NSSRI categories capture key sources of added sugar intake among children and youth from a variety of racial/ethnic and socioeconomic groups.

Global evidence indicates that target setting initiatives like the NSSRI can work to promote public health goals through industry reformulation of the food supply. Following implementation of government-led voluntary sugar targets in England, a 3.0% reduction in the SWM sugar content of food was observed in the first three years after implementation, with greater progress for some food categories (eg, 13.3% reduction for cereals).³⁵ Lessons can also be learned from a larger body of research evaluating global sodium reduction initiatives. More than 50 countries have established national sodium content targets for products,³⁶ with important reductions in the SWM sodium content of products and population-level dietary sodium intake observed in many countries.³⁷ For example, in the United States, a 7% reduction in the SWM sodium content of top-selling packaged foods was observed in the 5-year period during implementation of the National Salt Reduction Initiative’s voluntary sodium reduction targets.³⁸ In countries where substantial reductions in dietary sodium intake have been observed, strong support from central government, as well as multipronged efforts encompassing public education campaigns and other complementary strategies have been keys to success.^39,40 In addition, evidence suggests that mandatory targets may have a greater effect than voluntary targets.⁴¹

Although this analysis focused on reductions in added sugar intake, it is important to consider other shifts in industry behavior and population dietary intake that may result from the NSSRI. First, industry might replace sugars with ingredients such as nonnutritive sweeteners, which is problematic in light of evidence that exposure to nonnutritive sweeteners during childhood may influence future taste preferences and have implications for long-term health.⁴² Second, companies might acquire existing lower-sugar brands to meet the targets, which would change the composition of their product portfolio but would not affect the composition of the food supply. Third, consumers may make product substitutions away from reformulated products in favor of higher sugar items. Alternatively, consumers may be more likely to add sweeteners to products at the point of consumption (eg, adding sugar to unsweetened iced tea). Fourth, because this analysis suggests that US children and youth are consuming a decreasing quantity of NSSRI foods and beverages over time, reducing the added sugar content of these items might not have as strong of an effect as anticipated. The New York City Department of Health and Mental Hygiene has the ability to monitor these potential changes over time by rebuilding their database to track ingredients and sugar content for the years before and after the initiative.⁴³

There are limitations of this work. The analysis did not account for possible substitution that may take place during the initiative and assumed homogeneity in the influence of NSSRI across subgroups. The analysis also did not forecast the influence of the NSSRI targets on added sugar intake in 2023 and 2026, but instead used current population estimates to assess what added sugar intake could have looked like had industry already met the targets. In addition, the NSSRI targets are for total sugar, not added sugar. However, for many categories (eg, sugary drinks), added sugar and total sugar are equivalent, and in other categories (eg, sweetened milk), targets used an adjustment factor to account for sugars that are naturally occurring. Because the 24-hour dietary recall for children younger than age 12 years was completed or assisted by primary caregivers, added sugar intake may be underestimated in the case that children consume items without their caregiver’s knowledge.

This study also has many strengths. The analysis used eight survey cycles of nationally representative data, conducted extensive mapping of FNDDS food codes to NSSRI categories, and included several sensitivity analyses varying analytic assumptions.

CONCLUSIONS

Added sugar intake from packaged food and drinks among children and youth in the United States is high. By meeting NSSRI sugar reduction targets, industry could contribute to reducing child and youth consumption of added sugars.

RESEARCH SNAPSHOT.

Research Question:

In 2021, the National Salt and Sugar Reduction Initiative released voluntary sugar reduction targets for packaged foods and drinks in the United States. How much would added sugar intake among US children and youth be expected to change if industry were to meet the National Salt and Sugar Reduction Initiative targets?

Key Findings:

If industry currently met the targets, US children and youth would consume 7% (2023 targets) to 21% (2026 targets) less added sugar.