Study design and participants
We used anonymous secondary open data from NHANES for noninstitutionalized adults aged 20 years and older between 2017-2018 in the United States [22]. NHANES, conducted by the National Center for Health Statistics (NCHS), is a cross-sectional survey with a stratified, multi-stage probability sample design. NHANES collects 24-hour dietary intake recalls for two days using the interview-administered Automated Multiple Pass Method (AMPM) for a nationally representative sample over a two-year study period. The AMPM is a validated, multi-pass approach that aims to facilitate complete and accurate food recalls and reduce respondent burden [23, 24]. The first day of the dietary interview is conducted in-person in the mobile examination center (MEC) and the second day dietary interview was performed 3–10 days later over the phone. The questionnaires, data sets, and all related documents for each NHANES cycle are available on the NCHS website [25].
Demographic Data
For the first day of interviews, interviewees collected demographic information at each household, including gender and age. For age, we considered 10-year age groups (20-29, 30-39, 40-49, 50-59, 60-69, 70-79, and 80+).
Food And Sodium Intake Data
All foods and beverages reported in the interviews were assigned a food code using the Food and Nutrient Database for Dietary Studies (FNDDS) 2017-2018 edition. The food code converts consumed foods and beverages reported in the interviews into gram quantities and determines the corresponding nutrient (e.g. sodium) content. It should be noted that a previous study analyzing 24-hour urinary sodium data collected using AMPM suggests that AMPM is a valid means of determining sodium intake in adults [26].
The FNDDS provides an 8-digit food code to uniquely identify each food/beverage. The first digit in the food code identifies one of nine major food groups: (1) milk, (2) meat and fish, (3) eggs, (4) legumes, nuts and seeds, (5) grains, (6) fruits, (7) vegetables, (8) fats, oils and salad dressings, and (9) sugars, sweets and beverages. The second and subsequent digits of the food code indicate the specific sub-groups within the nine major food groups. In this study, all analyses were conducted at the subgroup level, but the estimation results are presented for the 10 major food groups, separating fish and meat.
In this study, the average intake of each food group and the corresponding sodium intake from the two-day dietary interview was calculated and analyzed as a daily value. Salt equivalent intake (g) was defined as sodium (mg) × 2.54/1,000. Appropriate sample weights were used to generate national, representative estimates [27].
Sodium reduction rate in various food products with the incorporation of umami substances
According to scientific literature, the incorporation of umami substances can reduce sodium in various food products, while maintaining their palatability. We obtained potential salt reduction rates for several food products from an extensive review by Tanaka et al. (2021) [21]. Based on previous studies and input from several food and nutrition experts (co-authors), we estimated salt reduction rates for umami substances by NHANES food subgroups as listed in Table 1.
Table 1
Sodium reduction assumptions due to incorporation of umami substances by FNDDS food code
Main group | Subgroup | FNDDS food code | Sodium reduction rate (%) | References | | | Umami substance |
Milk | Cheese | 14XXXXXX | 54–100% | Rodriques (2014)[45] | da Silva (2014)[46] | | MSG |
Meat | Sausage | 252XXXXX | 17–75% | Wooward (2003)[47] | Ichikawa Chemical Institution (1984)[48] | dos Santos (2014)[49] | MSG, CDG, Inosinate |
| Chicken broth | 283XXXXX, 285XXXXX | 11–38% | Chi (1992)[50] | Carter (2011)[51] | Wang (2019)[52] | MSG, CDG |
Fish | Salted fish | 26109170, 26109180 | 30–40% | Ichikawa Chemical Institution (1984) [48] | | | MSG, Inosinate |
Legume | Miso | 41601070 | 15–35% | Ishida (2011)[53] | Yamasa Corporation (2014)[54] | | MSG, Inosinate, Guanilate |
| Soy sause | 41420300 | 40–61% | Kao Corporation (2006)[55] | Ishida (2011)[53] | Kameda Seika Co., Ltd. (1997)[56] | MSG, Inosinate, Guanilate |
Grain | Snack | 540XXXXX, 543XXXXX, 544XXXXX | 51% | Buechler (2019)[57] | | | MSG, Inosinate, Guanilate |
Vegetable | Vegetable soup | 718XXXXX, 723XXXXX, 735XXXXX, 746XXXXX, 756XXXXX, 775XXXXX | 17–40% | Kremer (2009)[58] | Ball (2002)[59] | | Glutamates, CDG |
| Potate chips | 712XXXXX | 30% | Kongstad (2020)[60] | | | MSG |
| Salted vegetable | 755XXXXX | 55% | Tampei Pharmaceutical Co., Ltd. (1985)[61] | | | MSG |
Oil | Butter | 811XXXXX | 100% | de Souza (2014)[62] | | | MSG |
FNDDS: Food and Nutrient Database for Dietary Studies; MSG: monosodium glutamate; CDG: calcium diglutamate; X refers to all food codes for which the first one or two digits correspond. |
(Table 1 about here)
Estimating salt intake reduction with the incorporation of umami substances
As people in the United States may already consume a certain amount of low-sodium foods containing umami substances in their daily diets, we, therefore, set four hypothetical scenarios for the market share of low-sodium products: 0% (i.e. no low-sodium foods in the market); 30% 60%, 90%.
For each of the major food groups, at the population level, we calculated the amount of salt reduction possible for each of the four scenarios by gender. The salt reduction rate for each NHANES subgroup, expressed as an upper-lower interval in Table 1, represents the range of possible salt reduction rates estimated in the literature. The upper and lower limits were then used to calculate the maximum and minimum possible salt reduction for each subgroup at the individual level.
The following equations give the upper and lower limits of the j-th food subgroup-specific reduction in salt intake due to the incorporation of umami substances in the i-th individual;
$$Upper reduction in salt intake of the jth item under the kth scenario in the ith individual$$
$$Lower reduction in salt intake of the jth item under the kth scenario in the ith individual$$