Food Additives in Ultra-processed Packaged Foods: An Examination of US Household Grocery Store Purchases

Abstract

Background:

Food additives have been used mainly in the past century to perform specific functions in foods. Some types of food additives have been linked to adverse health outcomes, yet there is little research examining food additives in the US food supply.

Objective:

To examine the proportion of products purchased by US households containing four common technical food additives using time-specific food composition data and examine whether purchases have changed over time.

Participants/setting:

Nielsen Homescan Consumer Panels, 2001 and 2019.

Main outcome measures:

The proportion of packaged food products containing common types of food additives purchased by US households was determined overall and by food category.

Statistical analysis performed:

Differences examined using Students t-test, p-value of <0.01 considered significant.

Results:

Between 2001 and 2019, the proportion of food products purchased by US households that contained additives increased from 49.6% to 59.5% (p<0.001). The proportion of soft drinks purchased containing flavors decreased, with a subsequent increase in purchases containing non-nutritive sweeteners. Baby foods showed a 20% increase in the proportion of purchases containing additives and >15% increase in the proportion of purchases containing three or more additives.

Conclusions:

There is convincing evidence that US household purchases of common types of technical food additives are increasing. Despite some positive changes such as a decrease in the use of added colors in soft drinks, across most food categories an increase in purchases of all types of products containing additives was observed. In particular the finding that purchases of baby food products containing additives has increased substantially is crucial in informing future research in this area and warrants further investigation.

Introduction

Unhealthy diets are underpinned by the over-consumption of packaged food and beverage products. Over the last decade concern for ultra-processed foods (UPFs), which generally contain many additives added to enhance palatability, have been associated with increased bodyweight and other nutrition-related noncommunicable diseases and mortality.^1–15 A significant proportion of calories in the US diet derives from UPFs.^{16, 17} In most countries, be they high or low income countries, intake of UPFs is increasing.^18–20 Ingredients characteristic of UPFs include food additives (e.g., colors, flavors, flavor enhancers, emulsifiers, thickeners, and artificial sweeteners) designed to make the final product more palatable.²¹

Food additives have been used predominantly over the last century to perform specific functions in foods. In its broadest sense, a food additive is defined as any substance that is added to food. Legally, in the US, the term additive refers to “any substance the intended use of which results or may reasonably be expected to result -- directly or indirectly -- in its becoming a component or otherwise affecting the characteristics of any food.” Food additives are commonly used as coloring agents, flavoring agents, preservatives and sweeteners. While many additives are undoubtedly useful for increasing shelf life and food safety and have no negative health associations, there is evidence showing a number of food additives are linked to adverse health outcomes²² and that high consumption of UPFs (which in turn generally contain more additives compared to their less processed counterparts) potentially linked to addictive eating behavior.^23–26 For example, research has shown a potential link to higher intakes of artificial food colors with hyperactivity in children,²⁷ some emulsifiers with insulin resistance and weight gain,²⁸ flavors such as monosodium glutamate with a range of adverse health effects,²⁹ and a wide variety of food additives have been linked with negative changes to the gut microbiome.³⁰ In addition, there is some evidence from experimental research demonstrating that the interaction of different additives may lead to negative health effects, although research in this area is still in its infancy.³¹ On balance, technological advances have resulted in the introduction of food additives with no known negative health consequences to the US food supply that has allowed for longer shelf life, protection against microbial growth and numerous other benefits to the population.³²

There is little research examining sources and types of food additives consumed or purchased by the US population. Research globally is also lacking, with only a few studies from France,³³ Brazil³⁴ and Australia³⁵ examining the types of food additives in national food supplies. A small number of US-based studies exist that examine artificial color use in US foods and beverages.^{36, 37} One recent study has also examined the presence of sensory-related industrial additives in US food and beverage products.³⁸ Despite this it is yet unknown to what extent US households are purchasing packaged foods and beverages containing common food additives, and there has been no research examining whether there have been changes over time.

The food composition databases underpinning dietary surveys are often unable to identify the unique ingredients in individual food and beverage products that are purchased and consumed by US adults and children.³⁹ US consumers purchase >400,000 different packaged food and beverage products annually, and the specific products purchased are constantly changing due to product introduction and reformulation.^{40, 41} Hence, most studies to date that have examined food additives in the US have been limited by out-of-date product-specific food composition data, and no research has been undertaken examining changes that may have occurred over time, with the exception of a recent US study examining changes in purchases of products containing non-nutritive sweeteners (NNS) which showed purchases increased between 2001 and 2018.⁴²

The objective of this study was to examine the proportion of food and beverage products purchased by US households that contain four common types of technical food additives (colors, flavors, preservatives and NNS) using time-specific food composition data, and to examine whether changes have occurred over time. Food group sources of food additives were also examined.

Subjects and Methods

Study design and population

The Nielsen Homescan panel is an ongoing nationally representative longitudinal survey in which participating households record all Universal Product Code (UPC) purchases on a daily basis using a handheld scanner.^{43, 44} Households also report sociodemographic and household information and are sampled and weighted to be nationally representative. The Homescan dataset is used frequently by researchers to examine food consumption and purchasing patterns.^{42, 45} For this study Homescan data from 2001 and 2019 were used. These years were chosen as they represent the oldest and newest available datasets for which this analysis could be undertaken. By utilizing secondary deidentified Homescan data, this work was exempted from institutional review board review.

Linkage of UPC food products with Nutrition Facts Panel data and ingredient information

Each UPC captured in Nielsen Homescan 2001 and 2019 was matched with Nutrition Facts Panel data and ingredient information (when available) using commercial nutrition databases (Gladson, Label Insight, Product Launch Analytics, USDA National Nutrient Database for Standard Reference, and Mintel Global New Products Database) in a time-relevant manner. These commercial databases contain national brands and private label items at the barcode level, and data are updated regularly as new products enter the market. Further details regarding matching these commercial datasets have been published previously,⁴⁰ and involve matching and merging of data using UPCs, product descriptions, nutrient information and ingredients lists. A team of registered dietitians reviewed the Homescan purchases and assigned all products to major and minor food categories. Homescan does not include bulk items without barcodes, thus foods and beverages without a barcode or a Nutrition Facts Panel were excluded (e.g., unpackaged fresh fruits and vegetables, fresh meats, loose bread and bakery products without a barcode). It should be noted that these foods are much less likely to incorporate food additives than packaged food and beverage items and are consistent with the NOVA definition of minimally processed foods.

Identification of food additives

A list of industrialized food additives utilized in UPFs and their technical functions was compiled from FDA’s Substances Added to Food inventory, previously known as Everything Added to Foods in the United States (EAFUS).⁴⁶ Four common classes of technical food additives were selected to examine in this study (preservatives, NNS, color additives, and flavors). In brief (using the FDA’s Substances Added to Food inventory), colors were defined as substances authorized by a regulation in 21 CFR Part 73, 74, or 82. Flavors were defined as substances with a Flavour and Extracts Manufacturers Association (FEMA) or Joint FAO/WHO Expert Committee on Food Additives (JECFA) number, excluding those that are not listed as being a flavoring agent or flavor enhancer and excluding those in 182.10 category (herbs and spices). Preservatives were defined as substances identified as a “antimicrobial agent”, and NNS were defined as substances listed as having both a NNS function under the Substances Added to Foods inventory and that also appear in the FDA’s list of approved high-intensity sweeteners. Three additional additives (luo huan guo, sucralose and steviol glycosides) were also included in the definition of NNS as they appear in the FDA’s list of approved high-intensity sweeteners but not under the Substances Added to Food inventory as non-nutritive sweeteners. To remain consistent with existing literature on NNS, sugar alcohols were also included as NNS (sorbitol, xylitol, lactitol, mannitol, erythritol, and maltitol). All remaining substances were included as “other” additives. Table 1 lists the search terms used to define each class of food additives examined in this study.

Table 1:

Search terms used to identify color, flavor, preservative and non-nutritive sweetener additives in US packaged food and beverage products.

Colors search terms	Flavors search terms	Preservatives search terms	Non-nutritive sweeteners search terms
acetic acid acetone alcohol sda-3a alcohol, denatured formula 23a aluminum hydroxide ammonium bicarbonate ammonium carbonate ammonium hydroxide ammonium phosphate, dibasic ammonium phosphate, monobasic ammonium sulfate ammonium sulfite annatto, extract (bixa orellana 1.) astaxanthin benzoin benzyl alcohol beta-apo-8’-carotenal beta-carotene butterfly pea flower extract butyl alcohol calcium carbonate calcium hydroxide canthaxanthin capsicum (capsicum spp.) capsicum, oleoresin (capsicum spp.) caramel carmine (coccus cacti 1.) carrot, oil (daucus carota 1.) castor oil (ricinus communis 1.) cetyl alcohol citric acid citrus red no. 2 cochineal extract (coccus cacti 1.) copals, manila corn endosperm oil corn steep liquor cottonseed flour, partially defatted, cooked, toasted cyclohexane damar gum (shorea dipterocarpaceae) dehydrated beets dextrose diethylene glycol distearate dioctyl sodium sulfosuccinate dipotassium phosphate dried algae meal edta, calcium disodium edta, disodium ethoxyquin ethyl acetate ethyl alcohol ethyl cellulose ethylene dichloride ethylene glycol distearate ethylene glycol monoethyl ether fd&c blue no. 1 fd&c blue no. 1, aluminum lake fd&c blue no. 2 fd&c blue no. 2, aluminum lake fd&c green no. 3 fd&c red no. 3 fd&c red no. 40 fd&c red no. 40, aluminum lake fd&c red no. 40, calcium lake fd&c yellow no. 5 fd&c yellow no. 5, aluminum lake fd&c yellow no. 5, calcium lake fd&c yellow no. 6 fd&c yellow no. 6, aluminum lake fd&c yellow no. 6, calcium lake ferric oxide ferrous gluconate ferrous lactate fruit juice grape color extract grape skin extract haematococcus algae meal hexane hydroxypropyl cellulose invert sugar iron oxide isobutyl alcohol isopropyl alcohol japan wax malt syrup (malt extract) methyl alcohol methylene chloride mica molasses (saccharum officinarum 1.) naphtha orange b paprika (capsicum annuum 1.) paprika oleoresin (capsicum annuum 1.) phaffia yeast phosphoric acid polyethylene glycol (m w 2009,500) polyglycerol esters of fatty acids polysorbate 60 polysorbate 65 polysorbate 80 polyvinyl acetate polyvinyl alcohol polyvinylpyrrolidone potassium bicarbonate potassium carbonate potassium hydroxide potassium phosphate, tribasic potassium sulfate potassium sulfite propylene glycol riboflavin rosin (pinus spp.) and rosin derivatives rosin, adduct with fumaric acid, pentaerythritol ester rosin, limed saffron (crocus sativus 1.) salts of fatty acids shellac, purified silicon dioxide sodium bicarbonate sodium carbonate sodium copper chlorophyllin sodium hydroxide sodium phosphate, dibasic sodium phosphate, monobasic sodium phosphate, tribasic sodium sulfate sodium sulfite sodium tripolyphosphate sorbitan monooleate sorbitan monostearate soy leghemoglobin spirulina extract sucrose sulfur dioxide sulfuric acid sulfurous acid tagetes meal & extract talc terpene resins, natural terpene resins, synthetic titanium dioxide tomato lycopene trichloroethylene turmeric (curcuma longa 1.) turmeric, oleoresin (curcuma longa 1.) ultramarine blue vegetable juice xanthophyll zinc oxide	All substances with a JECFA or FEMA number were included in the flavors search terms. Over 3,000 exist and hence too many to list in a single document. Please refer to: https://www.cfsanappsexternal.fda.gov/scripts/fdcc/?set=FoodSubstances	2,2-dibromo-3-nitrilopropionamide acetic acid acrolein ammonium persulfate ascorbic acid benzoic acid bisabolene butylated hydroxyanisole calcium benzoate calcium propionate carbon dioxide catalase from aspergillus niger catalase from bovine liver catalase from penicillium notatum chlorine chlorine dioxide chlorine solution, aqueous citric acid dehydroacetic acid dimethyl dicarbonate disodium cyanodithioimidocarbonate disodium ethylenebisdithiocarbamate egg white lysozyme erythorbic acid ethyl alcohol ethylene oxide--nlfg fumaric acid hydrogen peroxide lactic acid listeria-specific bacteriophage preparation l-malic acid methyl benzoate methyl esters of fatty acids (edible) methyl p-hydroxybenzoate nisin preparation nitrates, sodium & potassium nitrites, sodium & potassium o-phenylphenol ozone phosphoric acid potassium benzoate potassium bisulfite potassium metabisulfite potassium nitrate potassium n-methyldithiocarbamate potassium oleate potassium sorbate potassium stearate propionic acid propyl p-hydroxybenzoate propylene glycol quaternary ammonium chloride combination saccharin, sodium salt sodium acetate sodium benzoate sodium bisulfite sodium chloride sodium chlorite sodium dimethyldithiocarbamate sodium erythorbate sodium lactate sodium metabisulfite sodium nitrate sodium nitrite sodium phosphate, dibasic sodium phosphate, monobasic sodium propionate sodium sorbate sulfur dioxide	acesulfame potassium advantame aspartame brazzein enliten equal erythritol lactitol luo han guo maltitol mannitol monk fruit extract necta sweet neotame nutrasweet purevia saccharin sorbitol splenda stevia steviol glycosides sucralose sugar twin sunett sweet’n low sweet one truvia xylitol

Statistical analysis

Data were analysed using SAS 9.4. The volume of foods purchased overall (in grams), by food and beverage status and by food group was determined overall and for (1) all food additives; (2) colors; (3) flavors; (4) preservatives; and (5) NNS. Differences over time in the percent volume purchased overall and in each major and minor food category were examined through t-tests using household projection weights provided by Nielsen. Bonferroni multiple comparison adjustment was applied and a p-value of <0.001 was considered significant.

Results

Overall results

The number of products purchased overall between 2001 and 2019 increased by 17%, from 355,870 in 2001 to 414,629 in 2019 (Table 2). The mean number of additives in purchased food and beverage products overall increased from 3.7 in 2001 to 4.5 in 2019 (p<0.001). Between 2001 and 2019, the percent volume of products purchased by US households that contained additives increased from 49.6% to 59.5% (p<0.001; Table 3). Overall, there was a significant increase in the proportion of products purchased that contained one additive (4.9% to 7.1%; p<0.001), two additives (5.5% to 7.5%; p<0.001) and three or more additives (44.6% to 51.2%; p<0.001), with a subsequent decrease in the proportion of products containing zero additives (45.0% to 34.2%; p<0.001); Table 4.

Table 2:

Number of unique products purchased by US households and mean number of additives per product purchased, by food category (2001 and 2019).

Category	No. unique products purchased		Mean no. additives per product purchased
	2001	2019	2001	2019
Dairy products excl milk	21665	24431	3.7	3.4
Meat, poultry, fish and mixtures	20501	21535	2.7	3.8
Nuts and seeds	5483	7202	0.5	0.7
Grain products, no RTE desserts	45397	43187	5.8	7.4
RTE cereals and granola	3835	4417	5.3	4.7
Other grain products (e.g., pasta, bake mixes, RTE bread)	41562	38770	5.8	7.9
Fruits and vegetables	43512	50233	1.0	1.0
Fats and oils	4620	4993	2.9	1.5
Sauces and condiments	19883	23282	3.1	3.6
Sweets and snacks	93292	104779	5.0	6.1
RTE grain-based snacks	45065	53505	6.1	6.8
Spreads and toppings	8346	9497	3.5	3.4
Candy and gum	21220	24442	3.4	5.0
Sweeteners	2677	3001	1.1	1.8
Puddings and ice cream	15984	14334	5.4	7.5
Baby food	1050	1992	1.5	3.2
Other (e.g., baking supplies, spices)	17175	21853	1.7	1.6a
Mixed dishes and soups	22282	30751	9.3	9.6
Frozen entrees, appetizers and pizza	9226	13760	13.3	12.5
Other mixed dishes (e.g., soups, stews, RTE prepared dishes)	13056	16991	6.0	7.2
Beverages	61010	80391	2.3	2.9
Dairy beverages	9312	6895	0.4	1.1
Fruit and vegetables juices	13308	10734	2.4	3.6
Carbonated soft drinks	9317	10425	4.6	5.6
Other beverages (e.g., water, tea, coffee)	13703	21780	2.0	2.4
Alcohol	15370	30557	0.1	0.2
TOTAL	355870	414629	4.0	4.6

Note: RTE=Ready-to-Eat.

^ano significant change between 2001 and 2019 (p>0.001).

Table 3:

Percent volume purchased among US households (weighted means) of foods and beverages containing additives (N=34,925 households in 2001; N=61,467 households in 2019).

	Additives overall (%)		Colors (%)		Flavors (%)		Preservatives (%)		Non-nutritive sweeteners (%)
	2001	2019	2001	2019	2001	2019	2001	2019	2001	2019
Dairy products excluding milk	57.0	61.4	29.9	40.1	47.7	48.2a	29.9	22.3	4.1	7.0
Meat, poultry, fish and mixtures	39.8	67.5	17.6	29.3	26.6	37.6	34.0	48.1	1.0	1.5
Nuts and seeds	7.6	16.1	5.5	10.4	6.2	10.3	0.1	3.7	0.0	0.4
Grain products, no RTE desserts	68.7	85.5	66.3	79.9	39.5	54.9	34.9	49.8	0.4	2.2
RTE cereals and granola	88.4	94.3	87.9	88.0a	46.5	42.4	27.7	17.9	1.4	0.7
Other grain products	65.1	84.1	62.4	78.7	38.5	57.2	36.7	55.4	0.2	2.5
Fruits and vegetables	19.7	27.5	12.1	19.0	9.5	14.8	11.4	17.0	0.1	0.5
Fats and oils	43.8	40.9	30.1	15.6	38.9	17.8	38.8	18.5	0.0	0.0
Sauces and condiments	73.9	83.8	47.7	55.7	37.0	40.2	36.2	47.9	1.0	1.5
Sweets and snacks	59.7	77.6	44.4	59.0	39.7	51.6	25.1	36.1	2.7	5.4
RTE grain-based snacks	64.7	79.3	59.2	70.2	41.4	51.6	31.2	40.6	2.6	3.5
Spreads and toppings	60.5	75.7	52.8	60.8	28.8	32.4	40.2	43.0	1.1	1.7
Candy and gum	75.4	89.3	31.8	40.8	36.0	34.2	22.8	27.4	3.4	8.9
Sweeteners	20.6	32.5	16.5	23.7	14.0	18.2	13.1	18.2	5.2	11.9
Puddings and ice cream	72.7	93.7	41.3	62.2	60.0	84.7	22.7	40.5	3.4	9.0
Baby food	50.0	72.5	37.0	48.3	38.9	57.8	32.7	47.4	0.0	3.3
Mixed dishes and soups	67.2	81.2	56.5	63.6	60.9	62.9	38.7	50.0	0.5	1.0
Frozen entrees, appetizers and pizza	64.9	85.8	61.4	75.0	60.4	70.7	52.1	58.2	0.5	0.4a
Other mixed dishes	67.9	77.1	52.9	54.8	60.6	56.7	29.7	43.9	0.6	1.4
Beverages	48.6	56.0	41.8	46.5	40.4	36.3	44.2	40.7	12.1	16.5
Dairy beverages	6.9	22.5	3.9	15.6	6.5	15.0	1.6	4.6	0.0	1.2
Fruit and vegetable juices	68.0	90.9	36.1	56.4	44.2	55.8	53.3	70.4	2.5	21.1
Carbonated soft drinks	90.2	95.1	89.5	86.6	80.3	75.4	89.3	87.1	28.0	32.7
Other beverages	23.9	43.5	21.7	38.7	17.1	18.3a	18.4	21.7	3.3	11.5
Alcohol	4.9	9.9	2.7	7.5	4.0	5.6	4.7	8.6	0.0	0.7
TOTAL	49.6	59.5	40.7	47.3	37.5	38.7	35.9	37.5	7.1	9.9

Note: RTE=Ready to Eat. Volume is in grams or mls as provided on product packaging.

^aNon-significant difference between 2001 and 2019 (p>0.001)

Table 4:

Percent volume of food and beverage products purchased by US households with zero to 3+ additives; by subcategory (N=34,925 households in 2001; N=61,467 households in 2019).

	Zero additives		One additive		Two additives		Three additives
	2001	2019	2001	2019	2001	2019	2001	2019
Dairy products excluding milk	44.5	41.2	6.5	9.4	4.5	4.9	44.5	44.5
Meat, poultry, fish and mixtures	55.5	27.3	4.8	13.3	2.5	5.6	37.3	53.9
Nuts and seeds	91.5	84.4	1.7	3.5	1.0	3.5	5.8	8.6
Grain products, no RTE desserts	28.8	11.1	6.2	7.1	6.0	8.2	59.0	73.6
RTE cereals and granola	11.3	6.2	1.4	4.9	4.7	12.7	82.5	76.2
Other grain products	32.6	12.0	7.3	7.5	6.2	7.4	53.9	73.1
Fruits and vegetables	76.6	74.3	5.1	4.8	5.6	4.9	12.7	16.0
Fats and oils	47.6	54.7	6.4	16.0	1.5	8.2	44.4	21.1
Sauces and condiments	25.0	11.6	13.2	14.1	8.3	13.3	53.5	61.1
Sweets and snacks	33.3	17.4	4.0	6.1	7.7	10.9	55.0	65.6
RTE grain-based snacks	33.8	20.2	2.6	5.0	2.6	6.6	61.0	68.2
Spreads and toppings	39.0	23.3	8.4	17.2	24.4	17.3	28.2	42.2
Candy and gum	27.5	7.5	3.6	4.8	15.8	22.8	53.1	64.9
Sweeteners	70.1	58.5	13.4	16.6	1.4	3.3	15.1	21.5
Puddings and ice cream	25.6	6.8	4.5	5.0	8.8	9.4	61.0	78.8
Baby food	50.2	29.5	4.7	9.3	16.3	13.0	28.7	48.2
Mixed dishes and soups	33.4	20.1	2.1	5.6	4.6	5.7	59.9	68.5
Frozen entrees, appetizers and pizza	31.2	11.3	1.5	3.6	1.2	5.8	66.1	79.4
Other mixed dishes	35.1	26.9	2.5	7.2	7.2	5.7	55.2	60.2
Beverages	52.4	42.1	3.4	4.9	4.2	5.9	40.0	47.0
Dairy beverages	91.2	75.6	0.4	0.7	1.9	2.3	6.6	21.4
Fruit and vegetable juices	30.5	8.7	12.8	12.3	11.8	15.4	44.9	63.6
Carbonated soft drinks	10.7	3.6	0.8	6.1	2.1	0.6	31.3	34.6
Other beverages	65.3	52.7	1.0	3.9	2.4	8.9	31.3	34.6
Alcohol	97.4	95.6	0.2	0.4	0.1	0.3	2.4	3.7
Total	45.0	34.2	4.9	7.1	5.5	7.5	44.6	51.2

Note: RTE=Ready to Eat. Volume is in grams or mls as provided on product packaging.

^aNon-significant difference between 2001 and 2019 (p>0.001)

The overall increase in additives was driven primarily by food purchases (50.9% to 63.7%; p<0.001). There were significant increases in volume purchases of food products containing colors (39.1% to 47.9%; p<0.001), flavors (34.2% to 41.7%; p<0.001), preservatives (26.7% to 34.5%; p<0.001) and NNS (1.2% to 2.8%; p<0.001).

Smaller significant change was observed overall for the proportion of beverages containing additives (48.6 to 56.0%; p<0.001); Figure 1. Interestingly, although beverage products had a significant increase in volume purchases of products containing colors (41.8% to 46.5%; p<0.001) and NNS (12.1% to 16.5%; p<0.001), a significant decrease was observed in beverages containing flavors (40.4% to 36.3%; p<0.001) and preservatives (44.2% to 40.7%; p<0.001) (Figure 1).

Figure 1:

Search terms used to identify color, flavor, preservative, and nonnutritive sweetener additives in US packaged food and beverage products. ^aJECFA = Food and Agriculture Organization/World Health Organization Expert Committee on Food Additives. ^bFEMA = Flavor and Extract Manufacturers Association of the United States. ^cRefer to: https://www.cfsanappsexternal.fda.gov/scripts/fdcc/?set=FoodSubstances.

Food category results

Frozen entrees, appetizers and pizza was the subcategory with the highest mean number of additives in both 2001 and 2019 (11.1 and 10.5, respectively), followed by Carbonated soft drinks (7.4 and 8.0). Between 2001 and 2019 the majority of categories and subcategories showed an increase in the volume of purchases containing all types of additives (Table 3). Figure 2 and Table 4 show changes in the mean number of additives in each food category and subcategory. Baby food showed the largest increase in the proportion of purchases containing three or more additives (28.7% to 48.2%; p<0.001), followed by Meat, poultry, fish and mixtures (37.3% to 53.9%; p<0.001) and Grain products (59.0% to 73.6%; p<0.001). Only one category (Fats & Oils) showed an increase in purchases of products containing zero additives (47.6% to 54.7%; p<0.001); Figure 2. More than half of all food categories had a larger proportion of purchases of products containing three or more additives compared to zero additives.

Figure 2:

Changes in the proportion of US household food and beverage purchases between 2001 and 2019 containing additives. University of North Carolina calculation based in part on data reported by NielsenIQ through its Homescan Services for all food categories including beverages for 2018 across the U.S. market.⁴⁴. Authors’ calculations based in part on data reported by NielsenIQ through its Homescan Services for all food categories including beverages for 2018 across the U.S. market. The conclusions drawn from the data are those of University of North Carolina and do not reflect the views of NielsenIQ. NielsenIQ is not responsible for and had no role in, and was not involved in, analyzing and preparing the results reported herein.

There was a significant decrease in purchases of Carbonated soft drinks containing flavors between 2001 and 2019 (80.3% to 75.4%; p<0.001) (Table 3). Despite this significant decrease, in 2019 more than 75% of Carbonated soft drinks purchases still contained added colors and more than 95% contained at least one additive. Of all beverage subcategories, Carbonated soft drinks had the highest mean number of additives in both survey years; 7.4 in 2001 and 8.0 in 2019 (Table 2). Almost every subcategory under Sweets & snacks and Beverages had an increase in the mean number of additives in purchased products, as well as an increase in the proportion of purchases containing NNS (Table 2 and Table 3).

Discussion

UPFs represent a growing proportion of the US diet, and consumption of industrial food additives has been identified as a growing concern alongside consumption of UPFs.^{38, 47} To the knowledge of the authors, this is the first study to examine the extent to which US consumers are purchasing food and beverage products containing industrial food additives, with the exception of one recent study examining changes in purchases of packaged foods containing NNS showing increases between 2001 and 2018.⁴² Results showed not only an increase between 2001 and 2019 in the proportion of food and beverage purchases containing additives⁴⁸ but also an increase in those containing three or more additives. Importantly along with these findings there was a more than 10% decrease in the proportion of products purchased that contained zero additives.

The overall results align for the most part with existing research. For example, a recent study examining sensory-related industrial additives in US food and beverage products found that most contained at least one additive, and more than a third had three or more additives.³⁸ The same study also found that more than 80% of sweets, non-alcoholic beverages, and RTE cereals contained additives, and in the present study it was observed that products purchased from these three categories had the highest mean number of additives in both 2001 and 2019. Another study conducted in France found that 54% of packaged foods and beverages contained at least one additive, with beverages and sweets having the highest number of additives.³³ A recent Australian study found that 65% of packaged food and beverage products contained additives.³⁵ Estimates from the afore mentioned studies are slightly higher overall than what was observed in the present study, with our observation that in 2019 just over 50% of all products purchased by US households contained additives.

With US dietary guidelines recommending lowered consumption of processed foods and a higher consumption of less processed foods, it’s concerning that there has been an increase in purchases of products containing additives since 2001. Of specific concern, baby foods (which include products such as baby formula, purees and infant cereals) showed a 20% increase in the proportion of purchases containing additives, a 15% increase in the proportion of purchases containing three or more additives between 2001 and 2019, as well as an increase in colors, flavors and NNS. To the knowledge of the authors there is currently no research examining the extent to which food additives are used in baby foods in the US, however with research suggesting a negative association between some food additives and adverse health effects in young children,²² as well as the findings in the present study are cause for concern. It is important to note that certain additives may, for this category in particular, provide certain benefits (such as for preservative effects), it is unlikely that the large increases seen for added colors (>10% increase) and flavors (>20% increase) are providing nutritional and health benefits and warrants further investigation.

Results showed that purchases of food and beverage products containing added colors have increased between 2001 and 2019. Added colors are a widely studied food additive in relation to adverse behavioral effects,^{27, 36, 37} and may be cause for concern, particularly in beverages that are marketed to, and consumed by, children. A US-based study from 2016 examined 810 food and beverage products from a single grocery store that were marketed to children and found that 43% contained at least one food color.³⁶ In 2019 more than 50% of carbonated soft drink purchases still contained added colors and more than 95% still contained at least one additive. Soft drinks had on average four additives in 2019 and in fact every beverage subcategory had an increase in the mean number of additives purchased by US households between 2001 and 2019.

This study found an increase in total food products purchased containing NNS from 12.1% to 16.5%, and in carbonated soft drinks specifically from 28.0 to 32.7%. This fits with existing research which has demonstrated the increasing use of NNS in beverages available in the US. This increase is shown to be occurring in tandem with a reduction in purchases of beverages containing caloric sweetener (sugar) and an increase in purchases of beverages containing both caloric sweetener and NNS.⁴² It will be important to examine whether the move towards products containing both caloric sweetener and NNS (and in some cases, color additives as well) have any effect on population health. Despite negative health having been demonstrated for a number of specific additives, both colors and NNS in particular have been negatively associated with changes to gut microbiota,³⁰ and little is known about whether such effects are further affected by the combination of multiple additive types in a product. Clinical research in this area is still in its infancy, with few studies examining the interactive health effects of different types of food additives.

A strength of this study was the use of nationally representative food purchase data. An important point to raise is that purchase data cannot be equated with intake due to unknown intra-household distribution, plus foods acquired through other sources.

This study is not without limitations. One challenge in the use of Homescan data is that estimates of household purchases might not be comparable with per capita intake from dietary intake surveys such as NHANES. For example, in a given household all purchases of products containing additives could be consumed by a single member of the household, rather than being consumed by all household members. Homescan also does not capture purchases from fast-food chains and other restaurants, which would undoubtedly have resulted in an underestimate of purchased products containing additives in the present study. Another potential limitation is that each ingredient was not further examined to determine if each listed ingredient contained additives. In a study on added sugar, Ng and colleagues developed a technique to use a commercial database that included the sub-ingredients of each ingredient to estimate the amount of added sugar in each US food product.⁴⁹ In that study they found added sugar in many of these ingredients above and beyond the added sugar terms that were indicated in the ingredients list. This indicates that the present study could potentially have underestimated the number of additives purchased by US households.

Given the importance of understanding what components make up UPFs in the US diet, this study found convincing evidence that US household purchases of packaged foods containing food additives is increasing. When an increase of about 10% of the volume of food and beverage products purchased from retail outlets contain additives, this is a topic of concern to the nutrition profession. Despite some positive changes such as a decrease in the use of added flavors in carbonated soft drinks, across most food categories an increase in purchases of all types of additives was observed. In particular our finding that purchases of baby food products containing additives has increased substantially is crucial in informing future research in this area and warrants further investigation.

Research Snapshot.

Research Question:

Has the prevalence and volume purchased of packaged food and beverages containing food additives with specific technical effects changed between 2001 and 2019 in the USA?

Key Findings:

The proportion of food products purchased by US households that contained additives increased from 49.6% in 2001 to 59.5% in 2019. The proportion of carbonated soft drinks purchased containing flavors decreased, with a subsequent increase in purchases of these products containing non-nutritive sweeteners. Baby foods showed a 22% increase in the proportion of purchases containing additives and >15% increase in the proportion of purchases containing three or more additives. Baby food also had the largest percent increase in purchases containing additives between 2001 and 2019 out of all major food categories examined.