Critical data at the crossroads: the National Health and Nutrition Examination Survey faces growing challenges

Abstract

NHANES needs urgent attention to ensure its future, which is facing emerging challenges associated with data collection, stagnant funding that has undercut innovation, and the increased call for granular data for subpopulations and groups at risk. The concerns do not rest merely on securing more funding but focus on the need for a constructive review of the survey to explore new approaches and identify appropriate change. This white paper, developed under the auspices of the ASN’s Committee on Advocacy and Science Policy (CASP), is a call to the nutrition community to advocate for and support activities to prepare NHANES for future success in a changing nutrition world. Furthermore, because NHANES is much more than a nutrition survey and serves the needs of many in health fields and even commercial arenas, effective advocacy must be grounded in alliances among the survey’s diverse stakeholders so that the full range of expertise and interests can engage. This article highlights the complicated nature of the survey along with key overarching challenges to underscore the importance of a measured, thoughtful, comprehensive, and collaborative approach to considering the future of NHANES. Starting-point questions are identified for the purposes of focusing dialog, discussion forums, and research. In particular, the CASP calls for a National Academies of Sciences, Engineering, and Medicine study on NHANES to articulate an actionable framework for NHANES going forward. With a well-informed and integrated set of goals and recommendations that could be provided by such a study, a secure future for NHANES is more readily achievable.

Introduction

For 50 y, our society has benefited greatly from the analysis of data from NHANES. It is difficult to imagine a national survey with a greater effect across such a wide range of health and nutrition issues in public and private sectors or one that has been used by such a diverse set of stakeholders. The notable success of NHANES is due not only to the relevance of the survey content to its users’ needs but also to the excellence of the data collection methodologies, trustworthiness of the data, and reliable availability of data for analysis. However, new challenges are emerging along with increasing health and nutrition data needs. With an eye to the future and responding to current challenges and new technologies, stakeholders together with NHANES staff must now carefully consider the survey’s data collection goals, its methodologies, and needed changes.

Addressing the future of NHANES is urgent and timely. Funding limitations are inevitable owing to competing demands for federal funds. Response rates for virtually all federal surveys including NHANES have declined dramatically, requiring innovative ways to assure representative data. Nutrition-related interests are changing. Health disparities are a national focus and addressing the challenge requires data from NHANES that will be difficult to collect given limited sample size necessitated by the cost of data collection. Finally, modernization and “re-imagination” of the US public health data infrastructure is an emerging national priority. In short, NHANES will be expected to meet emerging needs within real-world constraints while taking on newer methodologies and potentially altered approaches for both how and what data are collected. It will take time and a range of expertise for these complicated challenges to be considered and for their solutions to be explored, tested and, as appropriate, implemented. Therefore, the work of evaluating and conceptualizing the future of NHANES should begin now. The nutrition community must engage and lead advocacy efforts for the survey. However, NHANES is much more than a nutrition survey, and effective advocacy will require a broad alliance of stakeholders including many outside the field of nutrition.

This white paper was developed under the auspices of the ASN Committee on Advocacy and Science Policy (CASP). The authors were asked to describe the valuable and unique characteristics of NHANES and, within that context, identify overarching and potentially disruptive challenges facing the survey. We propose a way forward based on collaborative dialogs and underscore the value and relevance of studies conducted by expert panels convened by the National Academies of Sciences, Engineering, and Medicine (NASEM). The audience for this article is the community of NHANES stakeholders—researchers, government officials, policy makers, public health officials, clinicians, commercial interests, pharmaceutical groups, and others—who must now be galvanized to support and advocate for NHANES in new, collaborative, and comprehensive ways. The advocacy rests on an understanding of the nature of the challenges. Although this article identifies the emerging challenges, it does not propose specific solutions. Rather, examples of the considerations that may come into play are offered to highlight the complicated interrelationships that exist and, in turn, the need for thoughtful and well-considered approaches to providing for the future of NHANES.

NHANES Foundation

Advocating for NHANES requires an understanding of its foundation, which in turn frames decisions about the survey’s future and the suitability of different approaches to meeting emerging challenges. For the purposes of this article, the foundation reflects the survey’s origin, mission and defining components, and impact.

Origin

The 1956 National Health Survey Act authorized federal agencies to collect statistics for a variety of health issues [1]. The Act created the US National Health Survey Program. It put in place a household interview survey (a version of which was already in the field), followed by a physical examination survey, and then a family of health provider surveys. The interview survey became the National Health Interview Survey, and the examination survey became the National Health Examination Survey (NHES). In 1960, the program merged with the National Office of Vital Statistics to form the National Center for Health Statistics (NCHS)—located within the US Department of Health, Education, and Welfare, where it served data needs throughout the department.

Initially, none of the program surveys collected information on dietary intake or nutritional status. In 1967, Congress directed the Department, and in turn NCHS, to conduct what was known as the 10-State Nutrition Survey [2], given interests originating from the War on Poverty [3]. The focus on hunger and malnutrition grew as the US Senate Select Committee on Nutrition and Human Needs became active in 1968 [4] and the 1969 White House Conference on Food, Nutrition, and Health flagged the need for increased health-related dietary data [5]. So, in 1971, nutritional assessment was added to NHES and the survey was renamed NHANES. Policy makers and researchers found it advantageous to link dietary data with the measures of physical health status. Table 1 summarizes the timeline for health examination surveys at NCHS.

TABLE 1.

Timeline for health examination surveys conducted by National Center for Health Statistics, 1959–2020

Survey	Dates	Ages
Prenutrition
NHES I National Health Examination Survey: focused on chronic disease	1959–19621	18–79 y
NHES II National Health Examination Survey: focused on child growth and development	1963–1965	6–11 y
NHES III National Health Examination Survey: focused on adolescent health	1966–1970	12–17 y
Nutrition added
10-State Nutrition Survey: focused on low-income populations in 10 states	1968–1970	All ages
NHANES I National Health and Nutrition Examination Survey: focused on US population	1971–1974	1–74 y
NHANES II National Health and Nutrition Examination Survey: added infants starting at 6 mo old	1976–1980	6 mo–74 y
NHANES III National Health and Nutrition Examination Survey: added infants starting at 2 mo; no upper age limit	1988–1994	2 mo and above
Continuous data collection
NHANES2	1999–2000	All ages
	2001–2002
	2003–2004
	2005–2006
	2007–2008
	2009–2010
	2011–2012
	2013–2014
	2015–2015
	2017–2018
	2017–Mar 2020

NHES, National Health Examination Survey; NHANES, National Health and Nutrition Examination Survey.

¹National Center for Health Statistics came into existence in 1960.

²In 2002, integrated with USDA’s Continuing Survey of Food Intakes by Individuals; the 2019–2020 cycle suspended in March 2020; 2017 to March 2020 prepandemic cycle; August 2021 to August 2023 currently in field.

A factor in the later development of NHANES is the collaborative activities with the US Department of Agriculture (USDA). USDA had been collecting data on household and individual food consumption since 1935 to inform the USDA food assistance programs and projections for agricultural commodities [6]. These USDA surveys were conducted periodically until 1985 when USDA began the Continuing Survey of Food Intakes by Individuals (CSFII). In 1990, the passage of the National Nutrition Monitoring and Related Research Act (NNMRRA) [7] stipulated that the 2 agencies harmonize dietary the data collection and analysis. Combining CSFII with NHANES made logistical, efficiency, and budgetary sense despite the surveys having different sampling plans and methodologies [8]. This harmonization required several years and resulted ultimately in NHANES collecting participants’ reports of food intake using an automated program developed and maintained by the USDA [9]. Responsibility for nutritional analysis of the NHANES food consumption reports was moved to the USDA Agricultural Research Service, which revamped the analysis to create What We Eat in America (WWEIA) [10]. The WWEIA is now the source of dietary intake data derived from NHANES participants. In addition, the passage of the NNMRRA further established the foundation for NHANES. The Act was driven by evidence of the emerging relationships among diet, malnutrition, and chronic disease risks coupled with Congressional interest in improved nutrition monitoring [11]. The legislation required regular reporting on the nutritional status of the United States and specified a 10-y plan to do so. Data collected from NHANES participants were a key component of the required nutrition monitoring reports. In turn, NNMRRA spurred the further refinement and application of NHANES along with the related USDA methodologies and databases. The legislation lapsed in 2000, and the NNMRRA was not reauthorized. However, much of the data continues to be collected and disseminated even in the absence of the required progress reports.

Mission and defining components

Overall, NHANES is a logistically complex survey involving in-depth household interviews and health examinations. It operates primarily out of mobile examination centers (MECs) that travel to selected sites to obtain a representative sample of the US population [12]. It collects data on health and nutrition status of a probability sample of noninstitutionalized adults and children and is intended to do the following: estimate the number and proportion of the population and subgroups with selected diseases and conditions and risk factors for them; monitor trends in the prevalence, awareness, treatment, and control of selected diseases and conditions; provide normative data; and monitor trends in health risks and environmental exposures. In this context, data on nutrition reflects a risk factor and an environmental exposure.

The overall design of NHANES reflects 3 considerations: 1) objectives of data collection, 2) specific data collected, and 3) survey platform design. The objectives of data collection focus on obtaining the representative sample and then on using objective measures of health and health-related characteristics that are standardized with known quality parameters and able to be disseminated in a way to allow statistical calculations. The NHANES mission has been consistently anchored to these objectives and changes to these objectives would notably alter the survey and the data application.

The data content needed for the different diseases, conditions, and risk factors of interest will determine the information obtained by the collection process. In contrast to the data collection objectives, determinations about data content have been relatively open to change and modification. Furthermore, decisions about the data content are interrelated with the survey platform. The platform produces the specified data while meeting the overall survey objectives. It includes the structure of data collection, such as the number and nature of the persons sampled or the use of an examination center compared with an in-house interview. It also includes the specific tools of data collection such as blood draw equipment and digital scales. Moreover, the platform works to ensure respondent burdens are realistic. The interface between the specification of the data content and the survey platform can be a pressure point when considering adjustments and modifications to the survey as may occur in the future.

Impact

Planning for the future of NHANES will be informed by the survey’s broad impact. Data from NHANES have driven an impressive array of dietary and nondietary uses and applications, examples of are summarized in Table 2. Much of this impact originates from the survey’s unique ability to combine information from personal interviews, physical examinations, diagnostic procedures, dietary intake, and laboratory tests. However, the data are also relevant to stakeholders who reflect large commercial interests but are not focused on health or nutrition but rather on the representativeness and standardization of the data.

TABLE 2.

Examples of dietary and non-dietary uses of National Health and Nutrition Examination Survey data

Federal Government Stakeholders
Developing the foundation for health policy and related decision-making
•Dietary Guidelines for Americans Scientific Report
•Dietary Reference Intakes
•Healthy People Objectives
•Department of Agriculture Food Plans
•Centers for Disease Control and Prevention child growth charts
Informing guidance, regulations, provisions, standards, educational programs, and reports
Department of Health and Human Services
•Food fortification guidelines, nutrition labeling rules, infant formula composition
•Tracking infectious disease
•Physical activity guidelines
•Cut points for cholesterol levels, blood pressure measures
Department of Agriculture
•School meals planning, Supplemental Nutrition Assistance Program
•Healthy Eating Index development
•Cost of healthy diet estimation
Environmental Protection Agency
•Exposure levels for heavy metals in foods
Department of Defense
•Garrison meal planning, military combat rations planning
Government Accountability Office
•Report: Obesity Drugs: Few Adults Used Prescription Drugs for Weight Loss
All Stakeholders1
Serving as a platform for research and scientific discovery
•Relationship between serum folate and neural tube defects
•Association of water intake and hydration status with risk of kidney stone formation
•Identification of biological markers of chronic musculoskeletal pain
Specifying trends and prevalence
•Prevalence of heart disease, diabetes, obesity among adults and children
•Trends in fruit and vegetable intake
•Levels of household food security
•Exposure to mercury from food intake
•Acuity of taste/smell
Informing clinical practice and decisions
•Guidelines for the diagnosis and management of food allergy
•Prediction of increasing cost of urolithiasis
Informing commercial decisions
•Agricultural commodity planning for input supply, marketing and prices
•Marketing of dietary supplements to groups at risk for deficiencies
•Determine sizing of jeans in the U.S. and airline seat size
Informing legal decisions
•Determination of alimony, child support, and foster care guidelines/payments
Informing outreach and educational programs
•National Dairy Council, National Osteoporosis Foundation, Food Allergy Research &Education
•Brochures developed by state health programs and health providers

¹May include government stakeholders.

Clearly, the US government makes extensive use of NHANES data in policy and regulatory decisions, and in fact, national surveys such as NHANES are required by the Office of Management and Budget to be of “practical utility to the government” [13]. Data from NHANES informs national nutrition policy decisions ranging from the development of the Dietary Guidelines for Americans [14] to folate fortification of foods. A nonnutrition example of government application of NHANES data is the regulation of lead in gasoline as informed by data from NHANES showing high levels of lead in blood in multiple population groups [15].

However, the effect of NHANES extends far beyond government and has found uses and applications in many fields and in commercial enterprises. Nutrition stakeholders will be familiar with the survey’s role in clinical and dietetics activities and the food industry’s applications of the data for food product development and reformulation. Nondietary applications include the use of data on the body size to inform decisions about clothing manufacture [16]. Growth charts for children 2 y and older are derived from the analysis of NHANES data [17]. Medical administrative decisions, as illustrated by a study designed to predict the increasing cost of urolithiasis [18], rely on NHANES. The data have even been used to adjust unmeasured confounders in other data sets [19]. Most notable is the extensive use of NHANES data for scientific discovery as performed by investigators within academia, the private sector, and government. The data not only underpin diverse nutrition research activities, such as study on the effect of increasing plant protein intake on protein quality and nutrient intake [20], but also inform considerable nonnutrition research. A 5-y scan of randomly selected peer-reviewed journals reveals thousands of publications citing NHANES in recent years and reveals impressively diverse applications and uses (Table 3). Furthermore, NHANES has been emulated by other countries, and there are countless international publications informed by NHANES.

TABLE 3.

Examples of research publications incorporating NHANES data: informal 5-y scan of randomly selected peer-reviewed journals

Medical and disease journals

Journal of the American Medical Association: J.W. McEvoy, N. Daya, F. Rahman, R.C. Hoogeveen, R.S. Blumenthal, A.M. Shah, et al., Association of isolated diastolic hypertension as defined by the 2017 ACC/AHA blood pressure guideline with incident cardiovascular outcomes, JAMA 323(4) (2020) 329–338.

Stroke: M.A. Kim-Tenser, B. Ovbiagele, D. Markovic, A. Towfighi, Prevalence and predictors of food insecurity among stroke survivors in the United States, Stroke 53(11) (2022) 3369.

Alcoholism: Clinical and Experimental Research: J.B Warner, K.H. Zirnheld, H. Hu, A. Floyd, M. Kong, C.J McClain, et al., Analysis of alcohol use, consumption of micronutrient and macronutrients, and liver health in the 2017-2018 National Health and Nutrition Examination Survey, Alcohol Clin. Exp. Res. 46(11) (2022) 2025.

Journal of the American Dental Association: G.T. Vu, B.B. Little, P.C. Lai, G.-L. Cheng, Tooth loss and uncontrolled diabetes among US adults, J. Am. Dent. Assoc. 153(6) (2022) 542.

Public health and nutrition journals

American Journal of Preventive Medicine: M. Zimmer, A.J. Moshfegh, J.A. Vernarelli, C.S. Barroso. Participation in the special supplemental nutrition program for women, infants, and children and dietary intake in children: associations with race and ethnicity, Am. J. Prev. Med. 62(4) (2022) 578–585.

Public Health Nutrition: Y.-J. Wang, Y. Du, G.-Q. Chen, Z.-Q. Cheng, X.-M. Liu, Y. Lian, Dose-response relationship between dietary inflammatory index and diabetic kidney disease in US adults, Public Helath Nutr. 26(3) (2022) 1–9.

Journal of Racial and Ethnic Health Disparities: S.O. Nwaobi, H.L. Richmond, D.A. Babatunde, F. Twum, A.K. Mallhi, Y. Wei, et al., Crossover trends in current cigarette smoking between racial and ethnic groups of US adolescents aged 12-19 years old, 1999-2018, J. Racial Ethn. Health Disparities (2022) 1–10.

Journal of Nutrition: K. Pachipala, V. Shankar, Z. Rezler, R. Vittal, S.H. Ali, M.S. Srinivasan, et al., Acculturation and associations with ultra-processed food consumption among Asian Americans: NHANES, 2011-2018, J. Nutr. 152(7) (2022) 1747.

American Journal of Public Health: D.T. Lau, P. Sosa, Disparate impact of the COVID-19 pandemic and health equity data gap, 112(10) (2022) 1404.

Journal of the Academy of Nutrition and Dietetics: A.Y. Rosinger, H.J. Bethancourt, S.L. Young, Tap water avoidance is associated with lower food security in the United States: Evidence from NHANES 2005-2018, J. Acad. Nutr. Diet. 21 (2022) S2212.

Journal of Nutrition: L.E. O'Connor, E.A. Wambogo, K.A. Herrick, R. Parsons, J. Reedy, A standardized assessment of processed red meat and processed poultry intake in the U.S. population ages > 2 years using NHANES, J. Nutr. 152(1) (2022) 190–199.

Environmental and toxicology journals

Journal of Exposure Science and Environmental Epidemiology: Z. Stanfield, R. Setzer, V. Hull, R.R. Sayre, K.K. Isaacs, J.F. Wambaugh, Bayesian inference of chemical exposures from NHANES urine biomonitoring data, J. Expo. Sci. Environ. Epidemiol. 32(6) (2022) 833–846.

Journal of Environmental Research: X. Guo, N. Li, H. Wang, W. Su, Q. Song, Q. Liang, et al., Combined exposure to multiple metals on cardiovascular disease in NHANES under five statistical models, Environ. Res. 215(Pt3) (2022) 114435.

Journal of Cleaner Production: C. Bassi, R. Maysels, R. Anex, Declining greenhouse gas emissions in the US diet (2003-2018): drivers and demographic trends, J. Clean. Prod. 351 (2022) 131465.

NHANES also has economic impact. Data are lacking to allow direct quantification of the dollar value associated with the myriad of outcomes and activities that are informed by NHANES data. However, even casual observation underscores the economic value of NHANES. Federal regulations—which can be expected to reflect only a fraction of the economic impact associated with NHANES data to inform decisions—illustrate the point. An informal 2003 review using the Federal Register over a 5-y period found 8 final rules and 89 proposed rules referencing NHANES monitoring data that when combined suggested benefits of more than $100 million [11]. Recently, the Food and Drug Administration proposed a revised definition for the food labeling claim “healthy” and concluded that the use of the definition could reduce all-cause morbidity over time if consumers selected and consumed more healthful foods [21]. NHANES data played a role in the development of that definition and in the determination of the estimated economic impact. Another consideration is that diet and nutrition data collected from NHANES participants plays a role in the development and evaluation of USDA food programs that have annual costs in the billions of dollars.

NHANES Design and Implementation

Understanding the challenges facing NHANES requires a familiarity with the survey’s design and the nature of its current implementation. We focus on the key challenges of sample size, granular data, frequency of data release, response rate, emerging nutrition concerns, funding, and approaches to innovation, modernization, and efficiency. These challenges are briefly introduced in this section. Integration of the concerns for the purposes of stakeholder advocacy is addressed in the final section.

Survey design

Table 4 provides a general overview of the nature of NHANES. A full discussion of the survey’s many design features are beyond the scope of this article. However, methodologies, quality assurances, data processing and reporting, and the specific data collected have been modified over the years and are described in a myriad of publications and online resources [22]. The survey uses a 3-pronged approach to data collection including a household interview, physical examination in the MEC that is coupled with additional questionnaires, and postexamination interviews. The Survey Content Brochure [23] lists the measures included in the various survey cycles since 1999. Not all measures are included in every survey cycle. More than 75 questionnaire items and ∼150 laboratory measures have been included at various times over the years of data collection. More than 30 physical examination measures have been conducted, but only blood pressure, body measurements, and dual-energy X-ray absorptiometry have been included for each cycle since 1999.

TABLE 4.

Study characteristics and investigator access of National Health and Nutrition Examination Survey, 1999–2020¹

Study Characteristics
Objectives	NHANES is a program of studies designed to assess health and nutritional status of adults and children in the United States. It collects interviews and biospecimens and conducts physical examinations
Population	Target: Civilian, noninstitutionalized resident US population Sample: Households
Study design	Cross-sectional, population-based survey
Sampling	Multistage, stratified probability
Setting	National, 15 counties per year
Sample size	5000 examined per year
Recruitment method	Participants are determined by algorithm and recruitment is performed in-person
Compensation	Participants receive remuneration/travel expenses and a report of medical findings
Data collected	Interviews collect demographic, socioeconomic, dietary, and health-related information. The examination component consists of medical, dental, and physiological measurements and laboratory test results. Privacy is protected by the public law
Data collection methods	Health interviews are conducted in participants’ homes. Health measurements and additional health interviews are performed in specially designed and equipped mobile centers. Interviewers use notebook computers with electronic pens. Data from the mobile center is transmitted into data bases through such devices as digital scales and stadiometers. Touch-sensitive computer screens in private exam rooms let participants enter their own responses to certain sensitive questions
Funding	Congressional appropriation; other federal and related agencies
Investigator access
Data available	Nonidentifying individual-level health data including demographic, nutrition, biospecimen, data from examinations, and behavior information; restricted use data through Research Data Center
Access criteria	Data sets and user manuals are available for public download
Website	https://www.cdc.gov/nchs/nhanes/index.htm

¹Modified from the NHANES: U.S. Residents, US Department of Veterans Affairs, Cooperative Studies Program, 2020 [Internet]. Available from: https://www.vacsp.research.va.gov/CSPEC/Studies/INVESTD-R/Ntl-Health-and-Nutrition-Examination-Survey-US-Residents.asp

The sample size is a factor in the NHANES design and a key focus for stakeholders. The design must strike a balance between that needed to cover representative groups of interest and the limits on the sample size given the cost and current infrastructure that relies on MECs traveling to different locations to collect data [24]. In 1999, NHANES was restructured as a continuous survey, and data have been collected each y on a target sample of 5000 examined persons. Data are released every 2 y, providing approximately 10,000 examined persons. The continuous nature of the survey supports trend analysis, but at least 2 y of data, and often 4 or more years of data, are required for reliable statistical calculations. In addition, the interest in subpopulations means that the survey oversamples some subpopulations to increase the reliability of the data for those groups. Depending on the survey cycle, oversampling has included ethnicity (Hispanics, non-Hispanic Blacks, and Asians), low-income persons, adolescents, and older persons [12].

Survey design includes approaches to ensuring reasonable response rates, another key focus for stakeholders. NHANES has reported unweighted response rates for 1999–2020 for the interview component of the survey and the physical examination as summarized in Table 5. There has been a gradual decline over time, and a notable decline beginning in 2015. Response rates are an indirect indicator of the credibility of the survey. Low response rates undercut the ability to assume that the findings are representative of the population. A staff presentation on gaining participant cooperation while minimizing burden outlines some of the factors involved ranging from the time needed to complete the survey to perceived sensitivity related to answering survey questions [25]. Although sample weights are used in the data analysis of NHANES to correct for nonresponses [26], it is important to reduce nonresponse in the first place.

TABLE 5.

Unweighted response rate (%) for interview and examination survey components, NHANES 1999–2020

Interview/examination	Survey cycle
1999–2000	82/76
2001–2002	84/80
2003–2004	79/76
2005–2006	81/77
2007–2008	78/75
2009–2010	79/77
2011–2012	73/70
2013–2014	71/69
2015–2016	61/59
2017–2020	51/47

Source: https://wwwn.cdc.gov/nchs/nhanes/responserates.aspx.

Nutrition core

A “nutrition core” data set within NHANES does not officially exist. However, a subset of NHANES measures associated with food, diet, and nutrition interests are often informally referred to collectively as the “nutrition core.” The measures that commonly comprise the nutrition core are listed in Table 6, some of which are collected in every survey cycle, whereas others are collected less consistently. The dietary intake data are a central measure [27]. The data reflect 2, 24-h dietary recalls that are obtained from each participant. The first recall takes place in the MEC (through the 2020 cycle) during an in-person interview. The second recall is conducted through a telephone interview within 3–10 d after the first interview. Participants use models and drawings to estimate food amounts consumed. As noted earlier, NHANES provides the population sample and together with USDA coordinates and funds the collection of dietary data from these persons using a computerized method [9]. USDA also maintains an integrated data system on the nutrient content of foods, known as FoodData Central [28]. Intake data from WWEIA-NHANES are analyzed and updated by USDA for each NHANES cycle. The dietary intake data from WWEIA-NHANES can be found on both USDA and NHANES websites.

TABLE 6.

Measures¹ associated with the nutrition core, NHANES 1971-2020

Administered Interview for Diet, Food, and Behavior Measures
•24-hour dietary recall2
•Dietary supplements2
•Food frequency
•Alcohol use
• Dietary behaviors, program participation, and food security o Dietary screener o Infant feeding practices2 o Complementary feeding o Iodized salt usage o Food program participation2 o U.S. Household Food Security Survey2,3 o Flexible Consumer Behavior Survey2,3 (nutrition knowledge)

Physical Examination and Clinical Measures
• Anthropometry o Weight2 o Height2 o Body circumference and skinfold
• Blood and urine4 o Nutrient biomarkers: folate, vitamin D, vitamin C, vitamin A/E/carotenoids, blood lipids, vitamin B12, iodine

Note: Measures listed are based on webinar “NHANES: Opportunities for Revitalization,” September 22, 2022 https://nutrition.org/strongnhanes-opportunities-for-revitalization-strong/.

Note: NHANES water intake, dual-energy x-ray absorptiometry, and physical activity measures may or may not be included within the nutrition core.

¹Inclusion in survey varies by survey cycle.

²Included in all survey cycles 1971-2020.

³Incorporated from USDA Economic Research Service.

⁴Collected in all survey cycles; specific biomarker analysis varies by cycle.

Funding

Funding for NCHS, and in turn for NHANES, comes from monies Congressionally appropriated and passed to NCHS through its parent agency, the CDC. The budget for NCHS is publicly available [29], but that for NHANES is not, despite the likelihood that stakeholders would find the information helpful. In any case, the NHANES budget would be complicated to present because it would reflect not only the funds allotted to it from NCHS but also “reimbursable” funds from other government groups. NHANES staff can work collaboratively with partners to design components of NHANES for a specific survey period. These components are supported with funds from the requesting partner, that is, reimbursable funds. For instance, the collection in 2011–2014 of normative reference data for the ability to taste and smell in the US population was funded by the National Institute on Deafness and Other Communications Disorders [30].

As shown in Figure 1, the NCHS budget has been stagnant for years. The nominal budget has remained virtually constant while real purchasing power has declined owing to inflation [31]. Of particular concern, shrinking resources have hampered innovation research for all NCHS surveys because the priority is to keep the surveys in the field. This has been especially true of NHANES because of its high cost per completed interview and examination. Furthermore, the distinction between basic data collection and targeted components is becoming increasingly blurred. Because NCHS funding continues to be flat-lined, NHANES has become increasingly dependent on the allocation of reimbursable funds to support what might be considered basic data collection. Moreover, reliance on reimbursable funds can be challenging to the survey planners. The monies vary from year to year and are subject to fluctuating priorities and availability of funds within government agencies. Such funding is not guaranteed, yet determinations about the data to collect must be made long in advance.

FIGURE 1.

The NCHS budget since 2009 shown in nominal and real (adjusted for inflation) dollars. Source: https://docs.google.com/spreadsheets/d/1_xt8oI2neZyTwaZvtyQOtujzuHnjemZPwPuYVsEELr0/edit#gid=131887528. “Real” values are expressed in FY09 dollars. Inflation rates based on the GDP deflator from the Bureau of Economic Analysis (https://www.bea.gov/taxonomy/term/796). Modified with permission from AmstatNews March 1, 2022 (https://magazine.amstat.org/blog/2022/03/01/nchs/).

The effect of COVID-19 pandemic

Many national surveys were affected by the COVID-19 pandemic, and NHANES was affected more than others given its focus on physical examinations and could not revert to telehealth data collection or online administration. In March 2020 collection for the 2019–2020 cycle was suspended [32]. Hence, the survey cycle could not be completed and the data collected to that point were not nationally representative. To address this, portions of the data from 2019 through March 2020 were combined with data from NHANES 2017–2018 to provide a “prepandemic” sample. The data files for NHANES 2017 to March 2020 were released in May 2021 [33].

In early 2021, NHANES staff proposed modifications to the 2021–2022 survey methods to allow data collection for the cycle to resume [32]. A general concern raised by NHANES staff was the possibility of reduced examined sample size and reduced response rate. Paulose-Ram et al. [32] highlighted proposed changes that focused on an adjusted sample design that eliminated oversampling for race, Hispanic ethnicity, and income and modifications to questionnaires and examination components, laboratory content, outreach materials, interview and examination mode, and operational procedures for household interviews. Physical modifications to the MEC were proposed and a switch to a telephone interview for the first day of the 24-h dietary recall to limit face-to-face contact. Data collection resumed in July 2021 and is now expected to continue through August 2023. Therefore, no data have been released after the pandemic. Information to clarify the survey changes in response to the pandemic and to address the implications for data users is pending [32].

NHANES data collection beyond 2023

Plans for “redesigning” NHANES and for putting in place a new 10-y contract for data collection services are in process. In 2021, NHANES solicited contract proposals for data collection services for the next 10 y of the NHANES survey [34]. In 2017, NHANES asked for ideas on sample design and data collection processes for the upcoming NHANES [35], and in 2022, the NHANES announced that it was moving ahead to its newest phase in the NHANES “redesign” and requested input on components of the redesign [36]. As part of a 2022 webinar [37], NHANES staff informally highlighted the following possible redesign options: oversampling for race, ethnicity, and possibly age groups; collecting a nationally representative sample each year rather than over 2 y; releasing data on an annual basis; and increasing the number of locations sampled. The webinar also suggested sample size will remain at 5000/y, and the survey will work to be more mobile by potentially shifting from connected trailers to smaller mobile units. The data collection services contract process had been delayed [38], but was completed in March 2023. Further announcements about the redesign or the impact of the COVID-19 pandemic have not been issued.

NHANES Future

The future of NHANES warrants thoughtful and deliberative consideration. The survey rests on a historically solid scientific base, and attention now must be focused on the challenges for the future. The issues and relevant questions should be outlined and openly considered. We have endeavored to further that process here. Because NHANES is complex and multilayered, many issues will ultimately need to be addressed. For now, a “starting point” set of questions can serve to focus the task and lay the groundwork for meaningful discussions and collaborations. These starting-point questions reflect urgent challenges. They of course cannot reflect all challenges and questions, and certainly not the detailed questions that will be inevitable, but they serve to provide a basis for stakeholder advocacy.

“Starting-point” questions

Box 1 provides starting-point questions which, by their nature, naturally rest within the context of limited funding and the inevitable interrelationships between data collection objectives, data content, and survey platform. The various questions are briefly discussed to clarify the nature of the concerns for NHANES stakeholders and to integrate the emerging challenges with the need for advocacy. The needed advocacy can take many forms ranging from informing and educating members of Congress to organizing discussion forums to supporting studies that elucidate and test solutions.

BOX 1. Starting-Point questions about the future of NHANES.

Mission and data collection objectives

• •Does the original reason for NHANES still stand, i.e., standardized, objective measures related to the health and nutritional status of a representative sample of the US population and subgroups?
• •What topics and design characteristics are essential to the mission of NHANES?

Data Content—Nutrition Core

• •Does the nutrition core align with current nutrition and dietary priorities?

Survey Platform

• •What are approaches to sampling subpopulations within the confines of a realistic sample size?
• •What are the sample design requirements that match the need for granular data on multiple subpopulations within the context of a platform that requires the collection of objective characteristics with a high level of standardization?
• •What strategies can increase more timely or frequent release of data?
• •How can adequate response rates be ensured?
• •What data collection methods will improve efficiency and take advantage of advances in collection modalities?
• •What strategies support and ensure dedicated and collaborative exploration of innovation and modernization as an ongoing part of the survey?

Alt-text: BOX 1

Mission and data collection objectives

Discussions about the future of NHANES should begin by revisiting the survey’s mission and the goals of standardization of measurement and representativeness of the sample. What are the aspects of NHANES that should be retained to be responsive to the data needs of the federal agencies and other stakeholders? NHANES could adopt adjustments and modifications that would address newer challenges and interests, but which could alter the survey as we now know it and its ability to meet current mission objectives. If standardization and representativeness remain central as part of any future survey, the question becomes how can needed and potentially significant changes be accommodated while still incorporating standardization and representativeness.

Standardization has required a certain level of control over the data collections process. Several hypothetical considerations help to illustrate the questions important to stakeholders. These include whether more efficiency could be realized by NHANES use of other available sources of information on health and nutrition such as interview surveys, health care records, and commercial data bases. In turn, questions arise as to whether these other sources present standardization and generalizability challenges and, if so, can they be overcome. Furthermore, the interest in standardization has led to the use of MECs. The need to physically move MECs reduces the speed with which data can be collected and limits collection to approximately 15 locations a year under the current budget. If the use of MECs is now seen as overly burdensome or too limiting for emerging data needs, how can the expectations for standardization be supported without the current MECs? Could smaller more mobile units be incorporated? Moreover, logistics and costs have dictated a clustered design with a limited sample size. If the sampling design needs to be reconfigured, how can that be accommodated while still providing representativeness and the needed granularity, and how does the reconfigured sample design affect the survey platform and methods? Urgent discussions such as these should be collaborative and include a wide range of stakeholders, including those with experience related to underserved communities.

Considerations about the mission and goals of NHANES may raise questions about its “match” with that of CDC, its current parent agency. The effect of the current location may be noteworthy as the future of NHANES is considered. NCHS and its survey systems including NHANES were transferred in 1987 to the CDC from the Office of the Assistant Secretary for Health in the Department of Health and Human Services (HHS) [39]. Statistical government agencies such as the NCHS tend to be located within a science-focused or research-focused unit and often report directly to the head of their government department. The NCHS, and therefore the NHANES, are located within CDC which has a programmatic mission, meaning that its focus is to develop and implement programs to conduct programs that improve health through state health departments. Although CDC conducts surveillance of diseases and medical procedures, these activities are not primarily statistical and are designed to support the CDC’s mission to improve health and prevent the spread of disease. NCHS has a role linked to supporting data needs throughout HHS, so its surveys may or may not be congruent with most of the information needs and interests of the CDC. Furthermore, the Director of NCHS does not report directly to either the Secretary of HHS or the Director of CDC. Stakeholders should not assume that because NHANES has successfully provided needed information for 50 y, it will readily continue to function at least in the same manner. A flat budget amid rising costs and data collection challenges coupled with a lack of investment should raise concerns for the HHS and other government departments and the research, professional, and corporate communities. It also raises questions as to whether its placement in the CDC has provided the support the survey requires. Issues such as these warrant stakeholder attention.

Nutrition core

Nutrition core measures (Table 6) reflect the data needs of policy makers, public health professionals, and researchers working within the field of nutrition and, of course, can be used by others to relate dietary data to a variety of health issues. The existing nutrition measures have been included in the survey in some form since at least the 1970s. For the most part, they have remained unchanged, whereas nutrition-related public policy concerns have evolved along with our understanding of nutrition science. There have been updates for Dietary Guidelines for Americans, objectives for Healthy People, and the Dietary Reference Intakes. US demographics have shifted since the nutrition core was put in place, obesity has become a major health problem, and dietary supplement use has increased. The 2022 White House Conference on Food, Nutrition, and Health recently outlined National Strategies for ending hunger and reducing nutrition-related disease [40]. Research studies and the advent of metabolomics have expanded the understanding of relationships between nutrients and other food substances and health or disease, and there is a growing interest in the effect of dietary patterns on health.

One issue raised about the nutrition core is the methodologies for dietary intake data collection. Work on the methods is ongoing [41], and technological innovations such as wearable devices and improved automation are rapidly emerging, for which feasibility and cost are factors to explore. However, aligning and updating the nutrition core with current interests should take a more expansive approach. A range of data needs can be anticipated. The aging of the US population could require more data pertaining to older Americans. A greater focus on cultural foods may become a priority and what measures could be developed to study dietary patterns. There would undoubtedly be a call for work related to identifying and incorporating nutritional biomarkers for nutritional status. Questions may be raised about the level of precision that is needed for measures that may inherently have notable variability such as food consumption reports.

Many of the nutrition core measures will remain relevant, but it is time to consider the measures in a comprehensive way to ensure that they align with the current data needs. The starting point is to conduct a review identifying the new landscape of emerging nutritional data needs and to specifically target what data are missing and what data may now be less relevant. Changes would need to be considered in light of the effect on the survey platform and whether they comport with the goal of providing representative and standardized measures. In addition, if new methods were adopted for measures for which time series are important, bridging studies would be needed to retain the ability to track trends. These are only a few examples of the emerging challenges.

Survey platform: sample size

On its face, the question of sample size seems straight forward. To obtain representative data on smaller subgroups within a population, a sufficiently large sample of individuals is needed. However, the sample design is limited and centers on not only ensuring national representation but also keeping response rates as high as possible without exceeding the current survey budget [24]. Although the oversampling described earlier increases the reliability and precision for the targeted subgroups, the platform’s reliance on a 2-y sample of ∼10,000 examined individuals is insufficient to provide granular data for several at-risk populations and cannot provide estimates for small populations such as Indigenous Americans and breast-feeding women. Even when it is possible to provide point estimates, sample sizes are insufficient to determine whether the change has occurred or whether groups differ from each other. Some have pointed to the special survey conducted in the 1980s known as Hispanic HANES as an approach worth exploring. Hispanics in the United States were selected as a subpopulation to be surveyed owing to their number and projected growth and the lack of available health status data for them [42]. The survey demonstrated the ability of NHANES to collect representative data on an at-risk subpopulation. However, although the outcomes were much needed and widely used, the experience exposed difficulties owing to the considerable length of time that regular data collection was put on hold.

Nonetheless, as the recent White House Conference on Hunger, Nutrition, and Health [43] made clear, focusing on health disparities in the context of food insecurity and diet-related disease is a national priority and an ever-increasing focus within health fields in general. The pending update to the Dietary Guidelines for Americans will also be seeking such granular data and information on health disparities. To address such disparities, data from NHANES will be needed to establish baselines, articulate the nature of the health disparities, track progress, and conduct evaluations. Therefore, NHANES faces the future challenge of providing increasingly granular data based on ethnicity, race, income status, and even geographical locations; however, the sample size needed is likely beyond the capacity of the current infrastructure and budget.

These are difficult challenges and finding solutions will take time. Creative approaches will be needed for both consideration within stakeholder forums and supporting a NASEM study to examine solutions for NHANES. For instance, one “thought experiment” question is whether a survey platform based on—for example—5000 individuals surveyed every year with the goal of collecting data for every measure in the survey could be changed to an approach where data are collected—for example—for 15,000 persons in a year with the goal of collecting data only for the heavily used measures of blood, urine, weight, and height. Only a subsample of 5000 individuals would receive all components of the survey. Although this subsample would not be sufficient to provide estimates for some populations, the ability to address some characteristics of interest for all subpopulations would be expanded. If the total sample could be increased to 20,000 persons, then estimates for some would increase in stability and would be possible for others. However, changes such as these are not resource neutral, have various pros and cons, and would require research ahead of implementation. Finally, there are clear public health and research needs for real-time data release or at least for data released more frequently than every 2 y. The obvious solution of increasing the sample size in the current design to meet this concern is unrealistic, without a very large increase in the budget for NHANES. Questions and complicated decisions such as these are only illustrations of topics that could surface.

Survey platform: response rate

As pointed out earlier, NHANES has been experiencing a steady decline in response rates (Table 5). In the face of challenges presented by declining response rates, an initial focus has been to use a weighting factor to correct for nonresponse bias. In this context, Fakhouri et al. [26] reported on nonresponse for the 2017–2018 NHANES. They characterized the nature of nonresponders and concluded that enhanced weighting adjustments compared with previous survey cycles were needed, particulary targeting education and income. These adjustments still reflect standard adjustments but are more complicated than for earlier surveys given the higher nonresponse. Although perhaps not true for the 2017–2018 data, such adjustments as a general rule can lead to increased variability in weights and less reliability of estimates that can affect the ability to make comparisons between groups and across time. In any case, as a general matter, the decline in response rates calls into question the credibility of the data and its utility and is a chief concern for the future of NHANES. Some professional journals specify a minimum response rate for publication of results, usually ∼60%–80% [44]. As a collateral damage for the NHANES, lower response rates increase the cost for each completed interview because sampling must continue until enough individuals are recruited [45].

In addition to standard approaches to weighting data, efforts to explore characteristics that may be unique to, or strongly associated, with refusal or failure to take part in NHANES are likely a current focus of NHANES staff and are much needed. For instance, it may be possible to engage nongovernment or community partners to help encourage participation in the survey, but these approaches would require further study. Respondent burden as a component of survey design was discussed earlier, and a related question is whether there are ways to better determine what motivates people to participate. A 2021 report [46] examined the willingness of selected individuals to participate in surveys that involve physical measures and biomarkers. As outlined by the study authors, people with unmet health needs or chronic conditions were more likely to agree to take part in a health survey that includes collection of biomarkers. However, participation was more about the salience of the specific health issues than the goal of receiving additional health care. That is, if the survey were to be labeled as a general health survey, those with no existing health conditions would be less likely to respond than those with any existing health condition. Furthermore, if the survey were labeled as focused on a particular issue (e.g., asthma), then those with related conditions—asthma, allergies, or other respiratory issues—would be more likely to respond than someone with heart disease or diabetes. Apparently, how the survey is “advertised” to potential participants may impact response, but focusing on some areas will reduce the propensity to respond for those with other concerns.

The COVID-19 pandemic likely lowered the response rate. The 2021–2023 survey is not completed, and response rates have not yet been reported. In any case, NHANES response rates must receive increased attention and more targeted study. The issue is not only urgent but also fundamental to the viability of the survey. Some solutions may be found in collaborating with other government survey developers to examine factors influencing nonresponse, but the unique characteristics of NHANES add special challenges. Response rates can be counted among the most important challenges for the future of NHANES.

Survey platform: data collection innovation, efficiency, and modernization

Stagnant funding over time has limited the ability of NHANES to benefit from a robust and anticipatory innovation program. Staff have diligently worked to meet the needs for change and adjustment as they arise, but a survey such as NHANES needs a dedicated innovation program. A complicated set of issues arise when a long view of innovation and modernization is taken.

Based on another set of “thought experiments,” we can consider that at one end of the spectrum there is the goal of enhanced use of newer technologies such as wearable devices and additional remote measurement or increased use of computerized self-report. Such changes have potential downsides and benefits and must be carefully studied before being implemented. Funds would clearly be needed to investigate these approaches. At the other end of the spectrum, the goal might be structural changes to the survey platform such as relying on different venues to obtain data from physical examinations if standardization could be maintained or, perhaps, collaborating with other relevant surveys to off-load some of the data collection. NHANES has experience in partnering with local communities to obtain measurements that perhaps could be explored as additions to a national survey [47]. There is also the possibility of reducing the comprehensiveness of the survey while still meeting the NHANES mission given the delay in data release. In any case, the newer challenges signal that the future of NHANES may be different than before. The implications must be explored and will include tests of new equipment, crossover studies between current and proposed methods, exploration of incentives strategies, and more. Stakeholders need to advocate for and support these efforts.

Clearly, the question of increased efficiency, innovation, and modernization brings up considerations unique to NHANES. However, exploration of these challenges must be made against the backdrop of changes anticipated for government health surveys in general. The COVID-19 pandemic exposed weaknesses in US public health data collection and reporting processes, particularly related to state reporting of infectious disease [48]. The pandemic did not cause a breakdown as much as it revealed long-standing weaknesses and, importantly, weaknesses that relate to more than infectious disease. The concerns include inadequate and inconsistent definitions of data across different government jurisdictions, ambiguous timing in reporting, data gaps, problems in accessing data, and the changing interpretations of outcomes [48]. Recent efforts among public health officials and others in addition to proposed legislation targeting improvements and coordination among health data systems [49,50] have transformed public health data modernization into a national priority [51]. In 2020, the CDC launched the Data Modernization Initiative (DMI) [52] to address these types of concerns. It reflects a national effort to create modern, integrated, and real-time public health data and surveillance. The focus of DMI is not only the timeliness and quality of data but how also data activities and systems can be better coordinated at the federal, state, and local levels. The stated goal is to move from siloed and brittle public health data systems to connected, resilient adaptable, and sustainable “response-ready” systems. These interests may indirectly affect NHANES in a positive manner and could point to ways NHANES could fill the gaps inherent in state-based surveillance systems.

A way forward

Specifying the solutions for the challenges facing NHANES and, in turn, the survey’s future will be a process. The goal is to recognize the key questions and seek exploration of solutions. The CASP has come together with others to consider ways in which these tasks can be accomplished. To better inform the ASN membership and other stakeholders, the CASP organized a webinar in September 2022 focused on the revitalization of NHANES [53]. The CASP plans to seek partnerships with NHANES staff and stakeholders to pursue suggestions offered during the webinar. These include building alliances, information gathering, and securing support for studies.

Building alliances and partnerships among the survey’s stakeholders is a critical aspect of advocacy for NHANES. Support from government agencies, Congress, and other authorities is essential if the future of the survey is to be secured. The broad base of NHANES stakeholders must be enlisted to inform policy makers about the uniqueness and usefulness of NHANES data to health care services, research, environmental monitoring, personal disease prevention and treatment, and economic interests. Advocacy coalitions are needed and should include all stakeholders and private and voluntary sectors such as major foundations who as yet have not engaged.

Professional research societies and related organizations that rely on NHANES not only have a stake in the future of NHANES but also have an obligation to work with NHANES staff to assist in clarifying the issues. Efforts to sponsor workshops, forums, and conferences focused on the future of NHANES will be necessary. Defining questions and encouraging targeted research projects is an important way forward.

Finally, the single most effective action at this time would be an overall and integrated examination of NHANES in the context of its future. Without a well-informed and defined set of goals and recommendations, there cannot be meaningful progress. Such a study would need to not only carry the gravitas of diverse experts with experience and knowledge related to health surveys and to health and nutrition data collection but also provide an independent and consensus-driven report. The NASEM offers a strategic platform for such work because it has the capability to convene independent consensus panels and to draw on world renowned experts and practitioners. Its organization includes units that focus on relevant topics, including a Health and Science Division, the National Committee on Health Statistics, and a Food and Nutrition Board. NASEM studies are initiated and funded by federal agencies through contracts that specify the project and often detail the questions to be asked. The scope of work for the NASEM study could be designed around addressing questions such as those in Box 1 and, per standard NASEM protocol, would include engagement with NHANES and stakeholders to ensure comprehensiveness and an understanding of background considerations. NASEM studies and workshops are not new to NHANES. For instance, the NASEM is conducting work to determine whether, when, and how to return genetic results to NHANES survey participants [54]. Moreover, other government organizations that collect data, for example, the United States Census Bureau [55] and the Bureau of Justice Statistics in the Department of Justice [56], have made use of NASEM studies.

In summary, NHANES is at crossroads. Its many past successes are impressive, given the ambition of its goals and the complex nature of the data it has been designed to collect. However, the effect of years of inflation on the survey’s stagnant budget has undercut activities to meet the future, and the potentially game-changing nature of newer challenges cannot be avoided. A NASEM study to set the stage for the future of NHANES—that is, to provide an actionable framework—is a critical and prudent step forward. Certainly, maintaining the status quo and failing to adapt to emerging challenges cannot be an option for a survey vital to the nation’s health.

We wish to thank those who voluntarily reviewed this manuscript and offered suggestions including Dr. Regan Bailey (Texas A & M University), Dr. Victor Fulgoni, III (Nutrition Impact, LLC), and Dr. Shiriki Kumanyika (Drexel University). A draft of the manuscript was shared with NHANES staff and with staff at the Agricultural Research Service, USDA. We are indebted to Dr. Jaime Gahche (National Institutes of Health) for providing detailed technical information, and we thank Dr. Sie Sadohora (University of Minnesota) for formatting assistance. CLT: was responsible for initial drafts of the manuscript and for coordinating its development; and all authors: contributed to the writing of the text, figures and tables and read and approved the final manuscript. The authors report no conflicts of interest.

Funding

The authors reported no funding received for this study.

Author disclosures

The members of ASN’s Committee on Advocacy and Science Policy (CASP) requested this manuscript on NHANES, and some members of the CASP served as authors for this manuscript. The entire Committee reviewed and approved this manuscript. This article did not undergo peer review with The American Journal of Clinical Nutrition but was reviewed by ASN members.