Sources of Evidence for Potassium and Sodium Intake Distributions
This appendix provides an overview of the sources of evidence the committee used for potassium and sodium intake distributions.1 These data were used to inform the Adequate Intake (AI) values for potassium (see Chapter 4), to inform the infant AI values for sodium (see Chapter 8), and to perform the third step of the Dietary Reference Intake (DRI) organizing framework (intake assessment; see Chapters 7 and 11). Three surveys provided estimates of usual intake of potassium and sodium from dietary sources: the National Health and Nutrition Examination Survey (NHANES), the Canadian Community Health Survey–Nutrition 2015 (CCHS Nutrition 2015), and the Feeding Infants and Toddlers Study (FITS) 2016 (Anater et al., 2018; CDC/NCHS, 2018; Statistics Canada, 2017). Estimates from NHANES 2009–2014 were computed for all DRI age, sex, and life-stage groups; however, the estimates excluded breastfed infants and children. The CCHS Nutrition 2015 did not include data on infants 0–12 months of age (Statistics Canada, 2017). Therefore, FITS 2016 data and additional published analyses of NHANES data were used to inform the intake estimates for infants (Ahluwalia et al., 2016b; Maalouf et al., 2015; Tian et al., 2013).
1 Intake distribution tables are available by request from the National Academies of Sciences, Engineering, and Medicine’s Public Access Records Office. For more information, email PARO@nas.edu.
NATIONAL HEALTH AND NUTRITION EXAMINATION SURVEY
NHANES is a representative survey of the noninstitutional civilian population of the United States. NHANES has been a continuous survey since 1999, surveying approximately 5,000 people per year from 15 counties in the United States, and releasing data in 2-year intervals. Because it is continuous, the data may be combined across years. NHANES is the primary source of monitoring of dietary intakes for the United States. Since 2002, when the two national dietary surveys in the United States—the U.S. Department of Agriculture (USDA) Continuing Survey of Food Intakes by Individuals and NHANES—were integrated, two 24-hour dietary recall interviews have been collected on participants using the USDA Automated Multiple-Pass Method (Ahluwalia et al., 2016a; Moshfegh et al., 2008; Raper et al., 2004); the first years of public data release were in 2003–2004. The first 24-hour dietary recall is conducted in the Mobile Examination Center by an interviewer (CDC, 2014), and the second is collected by telephone within 3–10 days (CDC, 2013). Interviews are conducted with a proxy for participants younger than 6 years of age, are conducted with a proxy with the participant present for participants 6–8 years of age, are conducted with the participant with a proxy present for participants 9–11 years of age, and are conducted independently for participants 12 years of age and older. Each interview is coded as being reliable, as assessed by the interviewer. Dietary data are coded and linked to the Food and Nutritient Database for Dietary Studies (FNDDS) using the Survey Net system (Raper et al., 2004); data summarized as nutrient intake per day are publicly available.
Intake Distributions for All DRI Age, Sex, and Life-Stage Groups
Distributions of usual intake of potassium and sodium from NHANES 2009–2014 were provided to the committee (CDC, unpublished data).2 The distributions included each of the DRI age, sex, and life-stage groups. The data were analyzed using the National Cancer Institute (NCI) method (Tooze et al., 2010). All individuals with a reliable 24-hour recall from the exam were included in the analytic sample, with the exception of infants and children who were breastfed and women whose pregnancy or lactation
2 Intake distribution tables included estimates for all participants, estimates stratified by race/ethnicity, and estimates stratified by hypertension status. For the hypertension-stratified intake distributions, the blood pressure status of children and adolescents was categorized using the 2017 American Academy of Pediatrics guidelines (Flynn et al., 2017); the blood pressure status of adults was categorized using the 2017 American College of Cardiology and the American Heart Association guidelines (Whelton et al., 2018). Participants who reported that a doctor or other health professional had ever told them that they had a stroke or heart attack (myocardial infarction) were excluded from the hypertension-stratified intake distributions. This footnote was added since the prepublication release.
status was uncertain. Pregnant and lactating women were only included in analyses of those specific categories. Usual intake estimates of potassium and sodium from NHANES 2009–2014 exclude salt added at the table and potassium and sodium intakes from supplements or medications. This information is, however, queried in NHANES.
Additional NHANES Analyses Specific to Infants
Although the distributions of usual intake of potassium and sodium described above included the DRI age categories of infants 0–6 and 7–12 months of age, the estimates excluded NHANES participants who were breastfed. The committee, therefore, sought additional analyses of usual potassium and sodium intake that included NHANES participants 0–12 months of age who were breastfed. For infants 7–12 months of age, the committee also sought additional analyses that provided estimates of potassium and sodium intakes from complementary foods. Key methodologies used in the identified publications are briefly described below:
- Tian et al. (2013) used the Iowa State University method (Nusser et al., 1996) to estimate the distribution of usual intakes of potassium and sodium using NHANES 2003–2010 data. Estimates were presented for three age groups (7–11 months, 1–3 years, and 4–5 years) and, as applicable, stratified by breastfeeding status. Among infants consuming breast milk, the volume of breast milk was assumed to be 600 mL/d in those fed only breast milk, and 600 mL/d minus the volume of infant formula plus other milk for infants who were not exclusively breastfed. Potassium and sodium concentrations in breast milk were assumed to be 177 mg/L and 531 mg/L, respectively, from the USDA National Nutrient Database for Standard Reference 25 values for 1,000 mL of mature human milk (USDA/ARS, 2018). Estimates were also presented for usual potassium and sodium intake from complementary foods, which was defined in the publication as foods and beverages other than breast milk, infant formula, and other milks (e.g., cow’s milk).
- Maalouf et al. (2015) examined food sources of sodium among NHANES 2003–2010 participants, birth to 24 months of age. The methodology for estimating sodium intake from breast milk was the same as in Tian et al. (2013). The publication provided estimated contribution of breast milk, infant formula, and cow’s milk to the sodium intake of infants 6–11.9 months, along with estimated total sodium intake per day. Based on this information, the committee estimated the contribution of complementary foods to total sodium intake among older infants.
- Ahluwalia et al. (2016b) used the NCI method to estimate the distribution of usual intakes of a range of nutrients using NHANES 2009–2012 data. Estimates of the distribution of usual potassium and sodium intakes were presented for two age groups (6–11 months, 12–23 months) and included participants who consumed breast milk. The methodology for estimating intake from breast milk was the same as in Tian et al. (2013). This analysis did not present estimates for the contribution of complementary foods to usual potassium and sodium intake.
CANADIAN COMMUNITY HEALTH SURVEY–NUTRITION 2015
The CCHS Nutrition 2015 was the second nationally representative nutrition survey of the 21st century of the people of Canada, with the prior survey conducted in 2004 (Health Canada, 2017; Statistics Canada, 2017). The CCHS Nutrition is a focused survey collected occasionally with the CCHS Annual Component Survey, which samples 65,000 people each year. The CCHS Nutrition 2015 included a sample of all private-living individuals in the 10 Canadian provinces 1 year of age and older, with more than 20,000 respondents. Computer-assisted interviews conducted primarily in participants’ homes were conducted during 2015 on all days of the week. Interviews for children 1–5 years of age were conducted with a parent or guardian, those 6–11 years of age included the participant and a parent or guardian, and those 12 years of age or older were interviewed independently. All participants completed an unannounced 24-hour recall using the USDA Automated Multiple-Pass Method at the first interview, and a random subset (approximately 7,600) were invited to complete a second interview by phone 3–10 days later on a different day of the week using a food model booklet. Nutrients were extracted from the Canadian Nutrient File Version 2015 (CNF, ref 16), a recipe file based on FNDDS 5.0 and 2011–2012 and modified for the Canadian food supply, and survey foods reported in the survey that were not in the CNF but had some nutrient information available.
Distributions of usual intake of potassium and sodium from CCHS Nutrition 2015 data were provided to the committee (Statistics Canada, unpublished).3 The distributions included each of the DRI age, sex, and life-stage group for individuals 1 year of age and older. The data were analyzed using the NCI method. Because the NCI method allows for esti-
3 Intake distribution tables included estimates for all participants and estimates stratified by hypertension status. For the hypertension-stratified intake distributions, blood pressure status was categorized based on the participant’s response to the question, “Do you have high blood pressure?” Participants who reported that a health professional had ever told them they had heart disease were excluded from the hypertension-stratified intake distributions. This footnote was added since the prepublication release.
mates to be made for subpopulations, data from the previous CCHS 2004 (cycle 2.2) were combined with the CCHS Nutrition 2015 data to increase the sample size and improve model precision, but estimates from the 2015 cycle were estimated separately using covariates.4 All individuals with a reliable 24-hour recall from the exam were initially included in the analytic sample. An outlier detection strategy was then applied, which identified observations where the differences between the first and second recall were large. The second recall was removed from the analysis, as it is generally considered less reliable than the first because of learning effects or the Hawthorne effect. Observations were removed if they were within ± 2, 2.5, or 3 standard deviations from the mean distribution of the difference of the first and second recall values, with the cutoff providing the greatest improvement in the within/between-person variance ratio chosen.5,6 Pregnant and lactating women were included in analyses of those specific categories, but were excluded from other analyses. Usual intake estimates of potassium and sodium from CCHS Nutrition 2015 excluded salt added at the table and potassium and sodium intakes from supplements or medications. This information, however, is queried in CCHS Nutrition 2015 (Statistics Canada, 2017).
FEEDING INFANTS AND TODDLERS STUDY 2016
FITS 2016 is a cross-sectional study of the caregivers of infants and children younger than 4 years of age who live in the United States. Two previous FITS studies were conducted in 2002 and 2008 (Anater et al., 2018). Dietary intake data were collected using a 24-hour recall collected by telephone on all participants; 25 percent were invited to participate in a second 24-hour recall. The 24-hour recalls were collected using the Nutrition Data System for Research (NDSR) using certified dietary interviewers from the University of Minnesota Nutrition Coordinating Center. Participants were mailed a booklet to aid with the estimation of portion size. Although it is not a national probability sample, households were selected using stratified random sampling, and sampling weights that were calibrated to population totals for census divisions; the Special Supplemental Nutrition Program for
4 Three records missing the first day of recall were removed from CCHS 2004.
5 For potassium: 1- to 3-year-old ± 2.5 standard deviations was selected (n = 31 recalls removed); 31- to 50-year-old males ± 2 standard deviations was selected (n = 63 recalls removed); and 71-year-old and older males, ± 3 standard deviations was selected (n = 15 recalls removed).
6 For sodium: 19- to 30-year-old males, ± 2 standard deviations was selected (n = 39 recalls removed); 19- to 30-year-old females, ± 3 standard deviations was selected (n = 9 recalls removed); and 31- to 50-year-old females, ± 3 standard deviations was selected (n = 20 recalls removed).
Women, Infants, and Children (WIC) participation status; sex of child; race/ethnicity of child; and educational attainment of the caregiver. Nutrient data were analyzed using NDSR, which made updates to baby foods and infant formulas prior to the start of the study, and brand-name products were updated using user recipes during data collection. A total of 3,235 interviews were completed for the first 24-hour recall, including 600 infants aged 0–5.9 months, 902 infants aged 6–11.9 months, and 1,733 infants aged 1–47.9 months; 799 participants completed the second 24-hour recall.
Direct breastfeeding volumes were assessed using the methods of FITS 2008 (Ponza et al., 2004). Specifically, exclusively breastfed infants younger than 6 months were assumed to consume 780 mL of breast milk per day; for those who had both breast milk and formula, the volume of formula was subtracted from 780 mL to estimate daily breast milk consumption. For infants 6–11.9 months, the same method was used using 600 mL of breast milk per day. Expressed breast milk was quantified.
An analysis of potassium and sodium intake by food source was conducted for the FITS 2016 data from the first 24-hour recall. From these estimates, intakes from complementary foods, which include all food and beverage intakes other than baby milk (breast milk, infant formula, or toddler drinks) or other milk sources (e.g., cow’s milk), were estimated.
Ahluwalia, N., J. Dwyer, A. Terry, A. Moshfegh, and C. Johnson. 2016a. Update on NHANES dietary data: Focus on collection, release, analytical considerations, and uses to inform public policy. Advances in Nutrition 7(1):121-134.
Ahluwalia, N., K. A. Herrick, L. M. Rossen, D. Rhodes, B. Kit, A. Moshfegh, and K. W. Dodd. 2016b. Usual nutrient intakes of US infants and toddlers generally meet or exceed Dietary Reference Intakes: Findings from NHANES 2009-2012. American Journal of Clinical Nutrition 104(4):1167-1174.
Anater, A. S., D. J. Catellier, B. A. Levine, K. P. Krotki, E. F. Jacquier, A. L. Eldridge, K. E. Bronstein, L. J. Harnack, J. M. Lorenzana Peasley, and A. C. Lutes. 2018. The Feeding Infants and Toddlers Study (FITS) 2016: Study design and methods. Journal of Nutrition 148(9s):1516s-1524s.
CDC (Centers for Disease Control and Prevention). 2013. National Health and Nutrition Examination Survey. Phone follow-up dietary interviewer procedures manual. https://wwwn.cdc.gov/nchs/data/nhanes/2013-2014/manuals/Phone_Follow-up_Dietary_Interviewers_Manual.pdf (accessed October 23, 2018).
CDC. 2014. National Health and Nutrition Examination Survey. MEC in-person dietary interviewers procedures manual. https://wwwn.cdc.gov/nchs/data/nhanes/2013-2014/manuals/mec_in_person_dietary_procedures_manual_jan_2014.pdf (accessed October 23, 2018).
CDC/NCHS (National Center for Health Statistics). 2018. National Health and Nutrition Examination Survey. https://www.cdc.gov/nchs/nhanes/index.htm (accessed October 23, 2018).
Flynn, J. T., D. C. Kaelber, C. M. Baker-Smith, D. Blowey, A. E. Carroll, S. R. Daniels, S. D. de Ferranti, J. M. Dionne, B. Falkner, S. K. Flinn, S. S. Gidding, C. Goodwin, M. G. Leu, M. E. Powers, C. Rea, J. Samuels, M. Simasek, V. V. Thaker, E. M. Urbina, and Subcommittee on Screening Management of High Blood Pressure in Children. 2017. Clinical practice guideline for screening and management of high blood pressure in children and adolescents. Pediatrics 140(3).
Health Canada. 2017. Reference guide to understanding and using the data. 2015 Canadian Community Health Survey–Nutrition. Ottawa, Ontario: Health Canada.
Maalouf, J., M. E. Cogswell, K. Yuan, C. Martin, J. P. Gunn, P. Pehrsson, R. Merritt, and B. Bowman. 2015. Top sources of dietary sodium from birth to age 24 mo, United States, 2003–2010. The American Journal of Clinical Nutrition 101(5):1021-1028.
Moshfegh, A. J., D. G. Rhodes, D. J. Baer, T. Murayi, J. C. Clemens, W. V. Rumpler, D. R. Paul, R. S. Sebastian, K. J. Kuczynski, L. A. Ingwersen, R. C. Staples, and L. E. Cleveland. 2008. The US Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of energy intakes. American Journal of Clinical Nutrition 88(2):324-332.
Nusser, S. M., A. L. Carriquiry, K. W. Dodd, and W. A. Fuller. 1996. A semiparametric transformation approach to estimating usual daily intake distributions. Journal of the American Statistical Association 91(436):1440-1449.
Ponza, M., B. Devaney, P. Ziegler, K. Reidy, and C. Squatrito. 2004. Nutrient intakes and food choices of infants and toddlers participating in WIC. Journal of the American Dietetic Association 104(Suppl 1):s71-s79.
Raper, N., B. Perloff, L. Ingwersen, L. Steinfeldt, and J. Anand. 2004. An overview of USDA’s dietary intake data system. Journal of Food Composition and Analysis 17(3-4):545-555.
Statistics Canada. 2017. Canadian Community Health Survey–Nutrition (CCHS). http://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvey&SDDS=5049 (accessed October 23, 2018).
Tian, N., Z. Zhang, F. Loustalot, Q. Yang, and M. E. Cogswell. 2013. Sodium and potassium intakes among US infants and preschool children, 2003–2010. American Journal of Clinical Nutrition 98(4):1113-1122.
Tooze, J. A., V. Kipnis, D. W. Buckman, R. J. Carroll, L. S. Freedman, P. M. Guenther, S. M. Krebs-Smith, A. F. Subar, and K. W. Dodd. 2010. A mixed-effects model approach for estimating the distribution of usual intake of nutrients: The NCI method. Statistics in Medicine 29(27):2857-2868.
USDA/ARS (U.S. Department of Agriculture/Agricultural Research Service). 2018. USDA National Nutrient Database for Standard Reference. SR-Legacy. https://www.ars.usda.gov/northeast-area/beltsville-md-bhnrc/beltsville-human-nutrition-research-center/nutrientdata-laboratory/docs/usda-national-nutrient-database-for-standard-reference (accessed October 23, 2018).
Whelton, P. K., R. M. Carey, W. S. Aronow, D. E. Casey, Jr., K. J. Collins, C. Dennison Himmelfarb, S. M. DePalma, S. Gidding, K. A. Jamerson, D. W. Jones, E. J. MacLaughlin, P. Muntner, B. Ovbiagele, S. C. Smith, Jr., C. C. Spencer, R. S. Stafford, S. J. Taler, R. J. Thomas, K. A. Williams, Sr., J. D. Williamson, and J. T. Wright, Jr. 2018. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APHA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: A report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension 71(6):e13-e115.
This page intentionally left blank.