Abstract
Background
Food insecurity, defined as limited or uncertain access to enough food, has been epidemiologically associated with cardiovascular diseases, mental health and cognitive decline. However, its association with healthy ageing remains obscure.
Objective
To examine the association between food security and healthy ageing using data from two US population-based surveys: the National Health Interview Survey (NHIS) and the National Health and Nutrition Examination Survey (NHANES).
Methods
In this cross-sectional study, a total of 42,147 and 3,020 participants aged 60 years and older were included from NHIS (2011–2018) and NHANES (2011–2014), respectively. Food security was assessed using the USDA Adult Food Security Survey Module, and healthy ageing was defined with the consideration of chronic diseases, cognitive function, physical limitations, mental and self-perceived health.
Results
In both cohorts, food security was significantly associated with each component of healthy ageing, even after controlling for demographic, lifestyles and metabolic comorbidities. Notably, food security was linked to substantially increased likelihoods of healthy ageing (NHIS: OR = 2.24, 95%CI: 1.98 to 2.54; NHANES: OR = 3.32, 95%CI: 1.59 to 6.94). The positive association between food security and healthy ageing was largely consistent across subgroups stratified by gender, race, lifestyle factors and obesity status.
Conclusion
Food security was associated with higher odds of healthy ageing, highlighting the public health relevance of food security in relation to multidimensional healthy ageing, especially in resource-constrained areas.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12937-026-01347-1.
Keywords: Food Security, Healthy Ageing, NHIS, NHANES
Highlights
Food security is notably associated with higher likelihoods of healthy ageing in both NHIS (OR=2.24, 95%CI: 1.98 to 2.54) and NHANES (OR=3.32, 95%CI: 1.59 to 6.94).
Food security is associated with higher odds of being free from chronic disease, cognitive impairment, and limitations.
The positive association between food security and healthy ageing is consistent across gender, races, lifestyles, and obesity status.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12937-026-01347-1.
Introduction
Driven by increased life expectancies, improved healthcare and declining birth rates, population ageing is experiencing an unprecedented increase, with the estimation that by 2050 the number of people aged 60 and over will be doubled [1]. Ageing is often accompanied by increased prevalences of chronic diseases and ageing-related conditions, imposing a major burden on all societies [2]. Although healthy ageing could potentially be promoted by following a healthy lifestyle [3], the adherence to a healthy diet may substantially be hindered by food insecurity [4].
Food insecurity, an important component of the social determinants of health, is defined as limited or uncertain access to adequate food in socially acceptable ways [5]. Approximately 6.8% of elderly US residents were reported to be confronted with food insecurity in 2020, with a disproportionately high proportion among low-income or self-identified Black populations [6]. Food insecurity had been epidemiologically linked to increased risk of cardiovascular diseases [7] and mental health disorders [8], all of which were established contributors to accelerated ageing. Despite the inverse association between food insecurity and physical and mental health [9], its role in healthy ageing, a multifaceted concept encompassing physical, mental, cognitive and social well-being [10], remains largely unknown. Addressing this gap is essential, as healthy ageing transcends isolated health domains and requires integrated public health strategies.
Therefore, in this study, we used data from two US major health surveys, the National Health Interview Survey (NHIS) and the National Health and Nutrition Examination Survey (NHANES), to examine the association of food security with healthy ageing and to evaluate the potential modifying effects by other social determinants.
Methods
Study population
NHIS and NHANES are both cross-sectional surveys, recruiting a representative sample of civilian, community-dwelling members of the US population, using a complex, stratified, multistage and probability-cluster design. The NHIS focuses on health status, healthcare access, and health-related behaviors through a combination of in-person interviews and questionnaires [11]. In contrast, NHANES provides a more comprehensive assessment by integrating questionnaire interviews, physical examinations, and laboratory measurements to evaluate health and nutritional status in greater detail [12].
Study protocols of these two cohorts were both approved by the ethics review board of both the National Center for Health Statistics and the Centers for Disease Control (CDC), and all participants provided informed written consent [13, 14].
For the NHIS, considering the unification of chronic disease variables, data from eight cycles from 2011 to 2018 were utilized, and 42,147 individuals aged over 60 years with comprehensive information on food security and healthy ageing were included in this study (Fig. 1 A).
Fig. 1.
Study flowchart
In parallel, for the NHANES, given the availability of information on cognitive function and chronic diseases, two cycles from 2011 to 2012 and 2013–2014 were utilized, and 3,020 participants aged over 60 years with comprehensive information on food security and healthy ageing were included in this study (Fig. 1B).
Assessment of food security
Food security in both cohorts was measured through the United States Department of Agriculture Adult Food Security Survey Module (AFSSM), a well-validated 10-item questionnaire for adults [15]. In the NHIS cohort, it reflects food security over the past 30 days [11], while in the NHANES cohort, it reflects food security over the past 12 months [12]. Affirmative responses to each item were summed up, with more affirmative responses indicating worse food security. We then defined the overall status as full food security (0), marginal food security (1 to 2), low food security (3 to 5), and very low food security (6 to 10) based on the total number of affirmative responses, according to the guide of U.S. AFSSM [15]. To maximize statistical power, we further dichotomized the results into either full food security (no affirmative responses) or food insecurity (≥ 1 affirmative response) [16, 17], as previously described. Food security, treated as a continuous variable (0–10) for both cohorts, was presented in Table S1 of the Supplementary Methods.
Assessment of healthy ageing
Given the multidimensional nature of healthy ageing [10], in both cohorts, it was defined as having no history of major chronic diseases, no impairment of cognitive function, no physical limitations, being in good mental health, and self-perceived health (Table 1), according to the definition proposed by the World Health Organization [18]. If a participant met all five of the above criteria, he or she would be considered as healthy ageing [19, 20]. In addition to such a simple dichotomized definition [19, 20], we also developed a continuous healthy ageing metric, calculated as the arithmetic mean (0-100) of five equally-weighted domains: chronic disease, cognition, physical limitation, mental health, and self-perceived health, which extended score range and therefore increased the sensitivity to interindividual differences [21]. Furthermore, given that currently healthy ageing is only an umbrella term with no unified definition [10], we applied another definition of healthy ageing based on the Rowe and Kahn model, a binary indicator requiring simultaneous absence of major chronic disease, cognitive impairment, disability, physical limitation, and restricted social engagement [22] to further examine the robustness of our findings. More details were provided in Table S2-S3 of Supplementary Methods.
Table 1.
Definition of healthy ageing
| Criterion | NHIS [11] | NHANES [12] | ||
|---|---|---|---|---|
| Evaluation method | Definition | Evaluation method | Definition | |
| No history of major chronic diseases | Participants were asked whether they have the following chronic diseases. For each item, participants were allowed to reply ‘yes’ or ‘no’. | No history of heart failure, coronary heart disease, angina pectoris, heart attack, stroke, emphysema or chronic bronchitis, liver condition, diabetes mellitus, kidney failure, cancer or malignancy. | Participants were asked whether they have the following chronic diseases. For each item, participants were allowed to reply ‘yes’ or ‘no’. | No history of heart failure, coronary heart disease, angina pectoris, heart attack, stroke, emphysema or chronic bronchitis, liver condition, diabetes mellitus, kidney failure, cancer or malignancy. |
| No impairment of cognitive function | Participants were asked whether they have the difficulty remembering or concentrating. The answer was coded as ‘no difficulty’, ‘some difficulty’, ‘a lot of difficulty’ and ‘cannot do at all/ unable to do’. | Participants with answers of no difficulty were defined as no impairment of cognitive function. | We used the lowest quartile of four assessments as the cutoff (CERAD W-L ≥ 16, CERAD D-R ≥ 4, AFT ≥ 12 and DSST ≥ 33). | Participants who performed in the top 75% on any or all assessments. |
| No physical limitations | These questions were phrased to assess the individual’s level of difficulty in performing the task without using any special equipment. The answer was coded as ‘no difficulty’, ‘some difficulty’, ‘a lot of difficulty’ and ‘cannot do at all/ unable to do’. | Impaired physical function was defined as those who reported difficulties in washing or dressing themselves, walking or climbing steps, using hands or fingers to pick up small objects, or doing household chores. Participants without the above limitations were defined as having good physical function. | These questions were phrased to assess the individual’s level of difficulty in performing the task without using any special equipment. The answer was coded as ‘no difficulty’, ‘some difficulty’, ‘much difficulty’ or ‘unable to do’ and ‘do not do this activity’. | Impaired physical function was defined as those who reported difficulties in dressing themselves, walking between rooms on the same floor, getting in and out of bed, using a fork, knife, drinking from a cup, managing money, doing household chores or preparing meals. Participants without the above limitations were defined as having good physical function. |
| Good mental health | Participants were asked how often do they feel depressed. The answer was coded as ‘daily’, ‘weekly’, ‘monthly’, ‘a few times a year’ and ‘never’. | Participants with answers of ‘a few times a year’ or ‘never’ were defined as good mental health. | PHQ-9 was used. For each item, participants were allowed to answer with the following frequencies: ‘not at all’, ‘several days’, ‘more than half the days’ and ‘nearly every day’. | PHQ-9 score ≤ 4. |
| Good self-perceived health | “Compared with 12 months ago, would you say your health is better, worse, or about the same?” | Participants with answers of better or the same were defined as having good self-perceived health. | “In general, would you say your health is: excellent, very good, good, fair, poor?” | Participants with answers of excellent, very good or good were defined as having good self-perceived health. |
CERAD Consortium to establish a registry for Alzheimer’s disease, CERAD W-L The CERAD word learning subtest, CERAD D-R The CERAD delayed recall subtest, AFT The animal fluency test, DSST The digit symbol substitution test, PHQ-9 Patient health questionnaire-9
Assessment of covariates
According to directed acyclic graph (Figure S1), a series of covariates were selected, including age, gender, self-reported race, education level, marital status, family income-poverty-ratio (PIR), smoking, alcohol drinking, sleep duration, physical activity and body mass index (BMI). Given the low missing rate for all covariates in both cohorts (NHIS: 0.14–8.64%; NHANES: 0.10–8.05%; footnotes of Table 2), we performed complete-case analysis without imputation [23]. More details were provided in Supplementary Methods.
Table 2.
Characteristics of participants from National Health Interview Survey (NHIS) and National Health and Nutrition Examination Survey (NHANES) stratified by the status of food security
| Characteristics | NHIS | NHANES | ||||
|---|---|---|---|---|---|---|
| Total (N = 42 147) |
Food insecurity (N = 5 186) |
Food security (N = 36 961) |
Total (N = 3 020) |
Food insecurity (N = 616) |
Food security (N = 2 404) |
|
| Age (years) | 70.22 (0.05) | 68.74 (0.13) | 70.40 (0.06) | 69.34 (0.20) | 67.87 (0.28) | 69.55 (0.21) |
| Male, n (%) | 17,919 (45.9) | 1898 (38.6) | 16,021 (46.8) | 1463 (46.3) | 275 (38.8) | 1188 (46.2) |
| Non-Hispanic White, n (%) | 34,552 (84.1) | 3450 (69.6) | 31,102 (85.9) | 1409 (78.7) | 167 (55.3) | 1242 (82.1) |
| Education level, n (%) † | ||||||
| Lower than high school | 6960 (15.0) | 1786 (34.3) | 5174 (12.7) | 804 (16.7) | 275 (39.5) | 529 (13.5) |
| High school or equivalent | 11,838 (27.7) | 1554 (30.5) | 10,284 (27.4) | 696 (22.2) | 156 (25.0) | 540 (21.9) |
| College and above | 23,143 (57.3) | 1801 (35.2) | 21,342 (60.0) | 1517 (61.1) | 184 (35.5) | 1333 (64.7) |
| Marital status, n (%) † | ||||||
| Never married | 3062 (5.5) | 532 (8.6) | 2530 (5.1) | 176 (4.5) | 46 (4.9) | 130 (4.4) |
| Widowed, divorced, or separated | 19,143 (32.7) | 3096 (48.6) | 16,047 (30.7) | 1113 (31.1) | 292 (50.1) | 821 (28.4) |
| Married or living with partner | 19,857 (61.8) | 1551 (42.8) | 18,306 (64.2) | 1727(64.4) | 278 (44.9) | 1449 (67.2) |
| Family income-poverty ratio level, n (%) † | ||||||
| Low to moderate | 22,143 (46.2) | 4569 (85.0) | 17,574 (41.4) | 1725 (49.0) | 535 (90.5) | 1190 (43.0) |
| High | 20,004 (53.8) | 617 (15.0) | 19,387 (58.6) | 1052 (51.0) | 43 (9.5) | 1009 (57.0) |
| Lifestyles | ||||||
| Never smoking, n (%) † | 21,953 (52.5) | 2409 (47.2) | 19,544 (53.1) | 1505 (50.0) | 261 (40.9) | 1244 (51.3) |
| No current drinker, n (%) † | 29,837 (69.4) | 4347 (83.9) | 25,490 (67.6) | 2175 (60.8) | 486 (79.1) | 1689 (65.1) |
| Moderate to vigorous physical activity, n (%) † | 20,160 (50.7) | 1708 (34.9) | 18,452 (52.6) | 782 (29.3) | 102 (15.6) | 680 (31.3) |
| Normal sleep duration 7–8 h, n (%) † | 24,983 (60.4) | 2405 (47.0) | 22,578 (62.1) | 1644 (60.2) | 288 (48.2) | 1356 (61.7) |
| Physical measurements | ||||||
| Body mass index (kg/m2) † | 27.77 (0.03) | 28.94 (0.11) | 27.63 (0.04) | 29.03 (0.20) | 30.47 (0.38) | 28.83 (0.21) |
| Components of healthy ageing | ||||||
| No history of major chronic diseases, n (%) | 18,696 (44.7) | 1810 (34.6) | 16,886 (46.0) | 1336 (44.7) | 249 (36.8) | 1087 (45.8) |
| No impairment of cognitive function, n (%) | 32,625 (78.4) | 3277 (64.0) | 29,348 (80.2) | 1475 (68.2) | 260 (49.8) | 1215 (70.8) |
| No physical limitations, n (%) | 25,202 (61.8) | 1813 (36.4) | 23,389 (65.0) | 2107 (74.4) | 331 (54.6) | 1776 (77.3) |
| Good mental health, n (%) | 36,426 (87.2) | 3630 (70.7) | 32,796 (89.2) | 2288 (78.8) | 364 (56.6) | 1924 (82.0) |
| Good self-perceived health, n (%) | 37,268 (88.7) | 4026 (77.2) | 33,242 (90.1) | 2142 (79.6) | 316 (55.6) | 1826 (83.0) |
NHIS National Health Interview Survey, NHANES National Health and Nutrition Examination Survey
All values represented are weighted means (SE) or unweighted counts (weighted %) for continuous and categorical variables, respectively. All estimates accounted for complex survey designs, and P values were calculated by analysis of variance for continuous variables and chi-square test for categorical variables
†These variables had missing values. In NHIS, the unweighted number (percentage) of participants with missing values was as follows: education level (n = 206, 0.49%), marital status (n = 85, 0.20%), never smoking (n = 57, 0.14%), no current drinker (n = 309, 0.73%), moderate to vigorous physical activity (n = 1213, 2.88%), normal nocturnal sleep 7–8 h (n = 259, 0.61%), body mass index (n = 3640, 8.64%). In NHANES, the unweighted number (percentage) of participants with missing values was as follows: education level (n = 3, 0.10%), marital status (n = 4, 0.13%), family income-poverty ratio level (n = 243, 8.05%), never smoking (n = 3, 0.10%), no current drinker (n = 8, 0.26%), normal nocturnal sleep 7–8 h (n = 7, 0.23%), body mass index (n = 56, 1.85%)
Statistical analysis
Both NHIS and NHANES recruited a representative sample of civilian, community-dwelling members of the US population, using a complex, stratified, multistage and probability-cluster design. Weighted analyses with standard errors were employed to generate nationally representative estimates and combine survey cycles according to the guidelines [24, 25]. Baseline characteristics were presented as weighted means (standard errors) or unweighted counts (weighted proportions) across the status of food security for continuous and categorical variables, respectively. Differences between groups were examined by analysis of variance adjusted for sampling weights and Rao-Scott χ2 test.
Multivariable-adjusted logistic regression models accounting for complex survey designs were used to estimate the odds ratios (ORs) and the corresponding 95%CI for the association between food security and the likelihood of healthy ageing. Model 1 included adjustments for age, gender, race, education level, marital status, and family income-poverty ratio level (PIR); Model 2 additionally adjusted for smoking, alcohol drinking, physical activity and sleep duration; and Model 3 further adjusted for BMI.
Moreover, to assess the potential modification effect, we further explored the association between food security and healthy ageing stratified by age group (60–64 years, and ≥ 65 years), gender (male and female), race (non-Hispanic White and others), education level (lower than high school, high school or equivalent, and college and above), PIR (≤ 3 and > 3), smoking status (never smoking, and former or current smoking), alcohol drinking (current, and no current drinker), physical activity (< 600 MET, and ≥ 600 MET per week), sleep duration (7–8 h/day, and < 7 or > 8 h/day), and BMI (< 25, and ≥ 25). Tests for interaction were performed by adding multiplicative terms in Model 3. Specifically, within the complex-sampling design, we fitted a survey-weighted quasi-binomial generalized linear model with interaction terms (e.g., food security × sex) to examine the multiplicative effect on healthy ageing with interaction significance assessed using the Wald χ² test.
Furthermore, to test the robustness of our findings, several sensitivity analyses were performed using the NHANES dataset which supplied more detailed information on each component of healthy ageing. First, a continuous healthy ageing metric, which increased the sensitivity to interindividual differences [21], was employed. Second, another definition of healthy ageing based on Rowe and Kahn’s model [22] was used. Third, given the interconnected links between diet quality and food security, we additionally adjusted for Healthy Eating Index 2015, a well-established measurement aligning with the Dietary Guidelines for Americans [26]. In addition, to assess the robustness of our findings, several sensitivity analyses were conducted. In the NHIS cohort, where dietary data were unavailable, we applied a machine learning-based transfer learning framework [27] to predict HEI-2015 scores using NHANES-derived predictors (demographics, lifestyle factors, and nutritional covariates) through 10-fold cross-validated random forest with LASSO feature selection. The predicted continuous HEI-2015 values were subsequently incorporated as covariates in fully adjusted models. Furthermore, we performed a mediation analysis using the continuous HEI-2015 variable and the categorical complete lifestyle score variable as mediators to estimate its indirect effects using the coefficient product method [28]. Finally, a composite lifestyle score integrating five modifiable factors (smoking, alcohol, physical activity, sleep duration, and predicted HEI-2015) [29] was constructed and stratified into low (0–2), moderate (3), and high (4–5) categories for subgroup analysis.
All analyses were performed using R version 4.2.2 (The R Foundation for Statistical Computing, Vienna, Austria) and two-sided P values < 0.05 were considered statistically significant.
Results
Population characteristics
Table 2 showed basic characteristics of participants from NHIS and NHANES stratified by the status of food security. Among 42,147 participants in NHIS, the average age was 70.22 ± 0.05 years, with 45.9% male. Participants from NHANES (n = 3,020) had a slightly lower average age of 69.34 ± 0.20 years. Individuals with food security were more likely to be older, female, Non-Hispanic White, high education level, and having a better socioeconomic status, along with healthier lifestyle factors (Table 2). Of note, the weighted proportions of healthy ageing components, including no history of major chronic diseases, no impairment of cognitive function, no physical limitations, good mental health and good self-perceived health, were all markedly higher in subjects with food security in both cohorts (Table 2).
Association between food security and healthy ageing
Associations between food security and the likelihood of each healthy ageing component were shown in Fig. 2. Consistent with previous studies [9, 30–38], in both cohorts, food security was associated with a higher likelihood of each healthy ageing component, even after controlling for demographic, lifestyle factors and BMI. Specifically, in NHIS, participants with food security had a 52% higher likelihood of being free from chronic diseases (OR = 1.52, 95%CI: 1.39 to 1.68) and 100% higher likelihood of not having cognitive impairment (OR = 2.00, 95%CI: 1.81 to 2.21). Moreover, enjoying food security was associated with nearly three-fold increased probability of maintaining good physical function (OR = 2.41, 95%CI: 2.19 to 2.64), mental health (OR = 2.43, 95%CI: 2.18 to 2.72) and good self-perceived health (OR = 2.11, 95%CI: 1.88 to 2.37). In NHANES, the strength of association was even more pronounced, with a 68% and 69% higher likelihood of being free from chronic diseases (OR = 1.68, 95%CI: 1.14 to 2.49) or cognitive impairment (OR = 1.69, 95%CI: 1.11 to 2.57), respectively. Similarly, enjoying food security also nearly tripled the probability of maintaining good physical function (OR = 2.82, 95%CI: 1.86 to 4.27), mental health (OR = 2.52, 95%CI: 1.57 to 4.04), and good self-perceived health (OR = 2.24, 95%CI: 1.37 to 3.67).
Fig. 2.
Multivariable-adjusted odds ratios (95%CI) of food security with healthy ageing and its components. Model 1: adjustments for age, gender, race, marital status, education level, family income-poverty ratio level; Model 2: Model 1 plus adjustments for physical activity level, sleep duration, smoking status, alcohol consumption; Model 3: Model 2 plus adjustments for body mass index. For all these models, the reference was food insecurity. OR, odds ratio
Taken all these components into consideration, a total of 11,225 (weighted proportion: 28.00%) in NHIS, and 553 participants (weighted proportion: 25.00%) in NHANES could be defined as healthy ageing. After full adjustments, food security was found to be associated with a remarkably higher likelihood of being healthy ageing in NHIS (OR = 2.24, 95%CI: 1.98 to 2.54), and even more pronounced in NHANES (OR = 3.32, 95%CI: 1.59 to 6.94).
Subgroup and sensitivity analyses
In both the NHIS and NHANES, the positive association between food security and healthy ageing was not influenced by gender, race, education level, alcohol consumption, or BMI status (all P for interaction > 0.05, Fig. 3A, B). However, there were notable differences in the moderating effects within each cohort. In NHIS, the association between food security and healthy ageing was stronger among participants with mild physical activity levels (P for interaction = 0.014) and short or long sleep durations (P for interaction = 0.003), and it was particularly significant in those with low to moderate socioeconomic status (P for interaction < 0.001). In contrast, in NHANES, the association was more pronounced among the young-elderly aged 60 to 64 years (P for interaction = 0.012) and former or current smokers (P for interaction < 0.001). The magnitude of the association was also larger in individuals with higher socioeconomic status (P for interaction = 0.043).
Fig. 3.
Subgroup analysis of the association between food security and healthy ageing. A National Health Interview Survey (NHIS). B National Health and Nutrition Examination Survey (NHANES). Models were adjusted for age, gender, race, marital status, education level, family income-poverty ratio level, physical activity level, sleep duration, smoking status, alcohol consumption and body mass index, excepting for itself. For all models, the reference was food insecurity. All cases represented are unweighted counts (weighted %). OR, odds ratio
Furthermore, in sensitivity analyses using NHANES data, the strong positive association between food security and the possibility of being healthy ageing —initially observed in the primary analysis (OR = 3.32, 95%CI:1.59–6.94)—remained materially unchanged when a continuous healthy ageing score (β = 9.73; 95%CI 5.47–13.98) or an alternative definition of healthy ageing was employed (OR = 2.39; 95%CI 1.40–4.07) (Table 3). Moreover, despite the strong correlation between diet quality and food security and an established association between diet quality and healthy ageing, the link between food security remained significant even after the adjustment for HEI (OR = 3.12; 95% CI 1.49–6.55) (Table 3).
Table 3.
Sensitivity analyses of the association between food security and healthy ageing using data from NHANES
| N (%) a | Food insecurity | Food security | P value | ||
|---|---|---|---|---|---|
| Continuous healthy ageing score (N = 2,733 b) | |||||
| The history of major chronic diseases c | / | 1 (ref) | 1.39 (1.08–1.80) | 0.016 | |
| Cognitive functioning c | / | 1 (ref) | 1.57 (1.11–2.22) | 0.024 | |
| Physical limitations c | / | 1 (ref) | 2.66 (1.79–3.95) | < 0.001 | |
| Mental health c | / | 1 (ref) | 2.56 (1.59–4.13) | < 0.001 | |
| Self-perceived health c | / | 1 (ref) | 1.76 (1.08–2.87) | 0.036 | |
| Healthy ageing score d | / | 1 (ref) | 9.73 (5.47–13.98) | < 0.001 | |
| Rowe and Kahn’s model of healthy ageing (N = 2,391 e) | |||||
| No history of major chronic diseases f | 1090 (46.9) | 1 (ref) | 1.64 (1.01–2.67) | 0.047 | |
| No impairment of cognitive function f | 1839 (82.0) | 1 (ref) | 1.66 (0.99–2.79) | 0.055 | |
| No disability f | 2126 (91.4) | 1 (ref) | 2.39 (1.32–4.33) | < 0.001 | |
| No physical limitations f | 1994 (85.2) | 1 (ref) | 2.30 (1.24–4.27) | 0.011 | |
| Active engagement f | 2264 (95.8) | 1 (ref) | 3.18 (1.73–5.83) | < 0.001 | |
| Overall healthy ageing f | 737 (35.0) | 1 (ref) | 2.39 (1.40–4.07) | < 0.001 | |
| Additional adjustment for HEI-2015 (N = 2,803 g) | |||||
| No history of major chronic diseases h | 1230 (44.7) | 1 (ref) | 1.85 (1.26–2.73) | 0.003 | |
| No impairment of cognitive function h | 1510 (65.9) | 1 (ref) | 1.58 (1.00-2.47) | 0.048 | |
| No physical limitations h | 1977 (75.2) | 1 (ref) | 2.77 (1.82–4.21) | < 0.001 | |
| Good mental health h | 2138 (79.2) | 1 (ref) | 2.40 (1.41–4.08) | 0.003 | |
| Good self-perceived health h | 2003 (80.0) | 1 (ref) | 2.11 (1.30–3.41) | 0.004 | |
| Overall healthy ageing h | 494 (24.4) | 1 (ref) | 3.12 (1.49–6.55) | 0.005 | |
a N (%) represented unweighted counts (weighted %); For models using the continuous healthy ageing score as the outcome, N (%) was not reported
b Participants who answered “do not do this activity” in physical function (n = 287) were excluded
c Multivariate-adjusted ordinal logistic regression, with adjustments for age, gender, race, marital status, education level, family income-poverty ratio level, physical activity level, sleep duration, smoking status, alcohol consumption, body mass index, was used in this analysis
d Linear regression models were used to estimate the association of food insecurity with healthy ageing score and values were shown as β (95% confidence intervals)
e Participants with incomplete responses in cognitive tests (n = 52) or physical functions (n = 577) were excluded
f Multivariable-adjusted logistic regression, with adjustments for age, gender, race, marital status, education level, family income-poverty ratio level, physical activity level, sleep duration, smoking status, alcohol consumption, body mass index, was used in this analysis
g Participants without dietary data (n = 217) were excluded
h Multivariable-adjusted logistic regression, with adjustments for age, gender, race, marital status, education level, family income-poverty ratio level, physical activity level, sleep duration, smoking status, alcohol consumption, HEI-2015, body mass index, was used in this analysis
Besides, to address NHIS’s absence of dietary recall data, we applied a cross-survey transfer-learning framework (Table S4), using NHANES-derived random-forest predictions to impute continuous HEI-2015 for NHIS participants (Table S5). Subsequent causal-mediation analyses (Table S6-S7) and lifestyle-stratified models (Figure S2) confirmed the robustness of the food-security-healthy-ageing relationship, with only modest indirect pathways and a clear dose-response pattern across lifestyle strata.
Discussion
In these two US cohorts both from a nationally representative sample of elderly US population, we evaluated the proportion of healthy ageing using a combination form addressing several health domains and demonstrated that food security was associated with each healthy ageing component. Food security was associated with 2.24-fold (NHIS) and 3.32-fold (NHANES) higher odds of healthy ageing. This positive association extended across individual ageing components and persisted after adjustment for sociodemographic, lifestyle, and metabolic confounders (Fig. 2). Collectively, these findings indicate that food security is consistently associated with a more favorable multidimensional healthy-ageing profile in older adults.
Although food security consistently benefited various strata of gender, race, education, and BMI (P interaction > 0.05), cohort-specific variations were observed (Fig. 3). In NHIS, stronger effects were seen among physically inactive adults (P interaction = 0.014), those with short or long sleep durations (P interaction = 0.003) and smokers in NHANES (P interaction < 0.001), indicating increased vulnerability when food insecurity interacts with lifestyle risks. In NHANES, stronger associations were found among younger participants (60–64 years; P interaction = 0.012), suggesting heightened sensitivity during key health transitions. Socioeconomic gradients (PIR) diverged between cohorts: NHIS showed maximal protection in low-income groups (PIR ≤ 3; OR = 2.40, 95% CI: 2.06–2.78; P interaction < 0.001), while NHANES indicated greater effects in high-income strata (PIR > 3; OR = 16.30, 95% CI: 3.67–72.49; P interaction = 0.043). This discrepancy likely reflects limited statistical power in high-income and food-insecure subgroup (N = 43) from NHANES, and residual confounding by unmeasured stressors (e.g., wealth volatility in “high-income” groups). Nevertheless, the consistent directionality of effects, with point estimates > 1.0 across all subgroups, underscores the overarching relevance of food security to healthy ageing, regardless of modifying contexts. Furthermore, sensitivity analyses reinforced these findings, showing our findings highlight the importance of addressing food security as a key strategy to promote healthy ageing.
Previous studies have found that adopting a healthy lifestyle, which is easily accessible and affordable, plays an essential role in promoting healthy ageing [3]. Following a healthy diet or active engagements in physical activity has been demonstrated to be effective in preventing or delaying the progression of chronic diseases [39, 40]. For example, a Spanish cohort of 10,063 participants found that a healthier lifestyle was associated with a lower risk of incident depression [41]. Likewise, a prospective cohort study conducted in China found that a healthy lifestyle was linked to slower memory decline, even in the presence of apolipoprotein E genotype, suggesting that the health- protective role of lifestyle may offset genetic risk [42]. However, adherence to healthy diets or other healthy lifestyle factors may be affected by food security [43–45], which was defined as all people, at all times, having physical, social, and economic access to sufficient, safe, and nutritious food that meets their food preferences and dietary needs for an active and healthy life [8]. Food security has been reported to be associated with chronic diseases, such as diabetes [46], cardiovascular disease [47], and asthma [48]. Although food insecurity has a disproportional distribution among different ethnic groups, excessive cardiovascular disease risk induced by food insecurity was found in the general population [49]. Furthermore, a longitudinal cohort showed that economic food insecurity, rather than proximity to unhealthy food options, was associated with risk of incident cardiovascular diseases, independent of socioeconomic factors, traditional cardiovascular risk factors, diet quality and other health behaviors [50].
Although several studies have explored the association between food security and ageing-related diseases, these studies were limited to one or several dimensions of ageing without a comprehensive evaluation of healthy ageing. For example, participants with very low food security had elevated risks of depression compared with food-secure adults [8]. A meta-analysis of 64 studies highlighted a dose-responsive relationship between food insecurity and poor mental health [51]. Consistent with our findings, a retrospective cohort study found that food insecurity was prevalent and associated with a decline in executive function [9]. The higher probability of developing sarcopenia, reflected by low grip strength, in individuals with food insecurity [16] also supported our findings in the domain of physical limitation. However, healthy ageing is a multidimensional process of developing and maintaining the functional ability which enables wellbeing in older age [52]. In this regard, using a more comprehensive definition and several sensitivity analyses using different definitions for healthy ageing, our findings extended previous observations by providing evidence that food security was closely linked with overall healthy ageing.
To the best of our knowledge, this is the first study focusing on the association between food security and comprehensive assessments of healthy ageing. Adopting validated metrics to define food security as well as healthy ageing, using a large and nationally representative population sample broadens the generalizability of our findings. Critically, our study is the first to examine the association between food security and composite healthy ageing-defined as the concurrent absence of deficits across all five WHO domains. While prior work focused on isolated components, we showed that food security conferred a 2.24-3.32-fold higher likelihood of achieving all criteria simultaneously. This underscores the strong association between food security and preserved multidimensional functional ability, a core tenet of healthy ageing [49]. Our findings thus shift the paradigm from targeting single domains to addressing food insecurity as a foundational strategy for holistic ageing well-being.
However, we also acknowledge that there are several limitations. First, given the cross-sectional nature of our study design, causal inference is precluded. Nevertheless, the observed associations remained largely consistent across multiple sensitivity analyses and after comprehensive adjustment for a wide range of potential confounders. Second, minor differences existed between cohorts in the operational definitions of healthy ageing components and in the recall period for food security assessment (30 days in NHIS vs. 12 months in NHANES). Despite these methodological variations, the associations remained consistently positive and significant, underscoring the robustness of our findings. Third, due to the sample size, food security was divided as a more commonly used dichotomous variable but not the four-categorical variable recommended by FSSM [16, 17]. Further research is needed to explore the relationship between more different levels of food security and healthy ageing. Fourth, due to the utilization of self-reported measures for disease status, physical function, and self-perceived health in our analysis, misclassification and recall biases were inevitable. Nevertheless, such non-differential misclassification was more likely to attenuate the findings towards null, thereby underestimating the observed associations. Fifth, although we applied complete-case analysis, missingness in covariates may introduce bias if the data are not missing at random. Nevertheless, the consistency of observed associations across two independent cohorts with distinct proportions of missingness lends support to the overall stability of the findings.
These nationally representative data indicate that mitigating food insecurity may serve as an important upstream strategy to preserve the holistic functional capacity of older adults; public-health initiatives should therefore prioritize enhancing access to nutrition assistance and community-based supports as a core pillar of healthy-ageing policy.
Conclusions
In conclusion, our study provides strong evidence of the association between food security and healthy ageing, highlighting food insecurity as a potentially important factor for consideration in strategies aimed at improving well-being among older adults.
Supplementary Information
Acknowledgements
None.
Abbreviations
- NHIS
National Health Interview Survey
- NHANES
National Health and Nutrition Examination Survey
- CDC
Centers for Disease Control
- PIR
Family income-poverty-ratio
- BMI
Body mass index
- ORs
Odds ratios
- CERAD
Consortium to establish a registry for Alzheimer’s disease
- CERAD W-L
The CERAD word learning subtest
- CERAD D-R
The CERAD delayed recall subtest
- AFT
The animal fluency test
- DSST
The digit symbol substitution test
- PHQ-9
Patient health questionnaire-9
Authors’ contributions
QZ and JY performed data analysis and wrote the manuscript; YW, SL, JZ, WC, WZ, ZW, FL, LZ, YG, and LL helped with data collection, and data analysis; MX, XZ and YL critically revised the manuscript; YL and XZ provided financial support; YL conceived, designed and supervised the study. All authors have read and approved the final manuscript.
Funding
This study was supported by Joint research center for the high-quality development of primary public health (440306241173100400106), and Fundamental Research Funds for the Central Universities, Sun Yat-sen University (22hytd03).
Data availability
The dataset supporting the conclusions of this article is available on the NHIS website (https://www.cdc.gov/nchs/nhis/documentation/index.html) , and NHANES website (https://wwwn.cdc.gov/nchs/nhanes/default.aspx).
Declarations
Ethics approval and consent to participate
This study adhered to the ethical standards of the Declaration of Helsinki and was granted approval by the Institutional Review Board at the National Center for Health Statistics (NCHS). Written informed consent was obtained from all participants involved in the research.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Qi Zhang, Jialu Yang and Yong Wang contributed equally to this work.
Contributor Information
Xiaotao Zhou, Email: cybersoul@126.com.
Yan Liu, Email: liuyan215@mail.sysu.edu.cn.
References
- 1.Cho SJ, Stout-Delgado HW. Aging and Lung Disease. Annu Rev Physiol. 2020;82:433–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Nkuipou-Kenfack E, Koeck T, Mischak H, Pich A, Schanstra JP, Zürbig P, et al. Proteome analysis in the assessment of ageing. Ageing Res Rev. 2014;18:74–85. [DOI] [PubMed] [Google Scholar]
- 3.Zhou Y-F, Song X-Y, Pan X-F, Feng L, Luo N, Yuan J-M, et al. Association Between Combined Lifestyle Factors and Healthy Ageing in Chinese Adults: The Singapore Chinese Health Study. Journals Gerontol Ser Biol Sci Med Sci. 2021;76:1796–805. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jun S, Cowan AE, Dodd KW, Tooze JA, Gahche JJ, Eicher-Miller HA, et al. Association of food insecurity with dietary intakes and nutritional biomarkers among US children, National Health and Nutrition Examination Survey (NHANES) 2011–2016. Am J Clin Nutr. 2021;114:1059–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Core indicators of. nutritional state for difficult-to-sample populations. J Nutr. 1990;120(Suppl 11):1559–600. [DOI] [PubMed] [Google Scholar]
- 6.America) FAF. The state of senior hunger in America in 2020: an annual report %U. https://www.feedingamerica.org/research/senior-hunger-research/senior.
- 7.Brandt EJ, Chang T, Leung C, Ayanian JZ, Nallamothu BK. Food Insecurity Among Individuals With Cardiovascular Disease and Cardiometabolic Risk Factors Across Race and Ethnicity in 1999–2018. JAMA Cardiol. 2022;7:1218–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Reeder N, Tolar-Peterson T, Bailey RH, Cheng W-H, Evans MW. Food insecurity and depression among US adults: NHANES 2005–2016. Nutrients. 2022;14(15):3081. [DOI] [PMC free article] [PubMed]
- 9.Kim B, Samuel LJ, Thorpe RJ, Crews DC, Szanton SL. Food Insecurity and Cognitive Trajectories in Community-Dwelling Medicare Beneficiaries 65 Years and Older. JAMA Netw Open. 2023;6:132–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zanjari N, Sharifian Sani M, Chavoshi MH, Rafiey H, Mohammadi Shahboulaghi F. Successful aging as a multidimensional concept: An integrative review. Med J Islam Repub Iran. 2017;31:100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.(CDC/NCHS) CfDCaPNCfHS. About the National Health Interview Survey %U https://www.cdc.gov/nchs/nhis/about/?CDC_AAref_Val=https://www.cdc.gov/nchs/nhis/about_nhis.htm.
- 12.(CDC/NCHS) CfDCaPNCfHS. About the National Health and Nutrition Examination Survey %U https://www.cdc.gov/nchs/nhanes/about/?CDC_AAref_Val=https://www.cdc.gov/nchs/nhanes/about_nhanes.htm.
- 13.(CDC/NCHS) CfDCaPNCfHS. National Center for Health Statistics-How do I know this is a legitimate survey? %U https://www.cdc.gov/nchs/nhis/new_nhis/aboutus.htm.
- 14.(CDC/NCHS) CfDCaPNCfHS. Ethics Review Board Approval of National Health and Nutrition Examination Survey %U https://www.cdc.gov/nchs/nhanes/about/erb.html.
- 15.(USDA) USDOA. U.S. ADULT FOOD SECURITY SURVEY MODULE. %U https://www.ers.usda.gov/sites/default/files/_laserfiche/DataFiles/50764/26623_ad2012.pdf?v=3672.4.
- 16.Lynch DH, Petersen CL, Van Dongen MJ, Spangler HB, Berkowitz SA, Batsis JA. Association between food insecurity and probable sarcopenia: Data from the 2011–2014 National Health and nutrition examination survey. Clin Nutr. 2022;41:1861–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bundy JD, Mills KT, He H, LaVeist TA, Ferdinand KC, Chen J, et al. Social determinants of health and premature death among adults in the USA from 1999 to 2018: a national cohort study. Lancet Public health. 2023;8:e422–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.(WHO) WHO. Development of a proposal for a Decade of Healthy Ageing 2020–2030%U https://apps.who.int/gb/ebwha/pdf_files/EB146/B146_23-en.pdf.
- 19.Zhou Y-F, Lai JS, Chong MF-F, Tong EH, Neelakantan N, Pan A, et al. Association between changes in diet quality from mid-life to late-life and healthy ageing: the Singapore Chinese health study. Age Ageing. 2022;51(10):afac232. [DOI] [PubMed]
- 20.Sun Q, Townsend MK, Okereke OI, Rimm EB, Hu FB, Stampfer MJ, et al. Alcohol consumption at midlife and successful ageing in women: a prospective cohort analysis in the nurses’ health study. PLoS Med. 2011;8:e1001090. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhang J, Zhao A. Dietary Diversity and Healthy Aging: A Prospective Study. Nutrients. 2021;13:1787. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rowe JW, Kahn RL. Success Aging Gerontologist. 1997;37:433–40. [DOI] [PubMed] [Google Scholar]
- 23.Bennett DA. How can I deal with missing data in my study? Aust N Z J Public Health. 2001;25:464–9. [PubMed] [Google Scholar]
- 24.(CDC/NCHS) CfDCaPNCfHS. NHANES Survey Methods and Analytic Guidelines %U https://wwwn.cdc.gov/Nchs/Nhanes/AnalyticGuidelines.aspx.
- 25.(CDC/NCHS) CfDCaPNCfHS. National Health Interview Survey 1997–2018%U https://archive.cdc.gov/www_cdc_gov/nchs/nhis/1997-2018.htm.
- 26.Krebs-Smith SM, Pannucci TE, Subar AF, Kirkpatrick SI, Lerman JL, Tooze JA, et al. Update of the Healthy Eating Index: HEI-2015. J Acad Nutr Dietetics. 2018;118:1591–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Bastani H. Predicting with Proxies: Transfer Learning in High Dimension. Manage Sci. 2021;67:2964–84. [Google Scholar]
- 28.Valeri L, Vanderweele TJ. Mediation analysis allowing for exposure-mediator interactions and causal interpretation: theoretical assumptions and implementation with SAS and SPSS macros. Psychol Methods. 2013;18:137–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zhang YB, Chen C, Pan XF, Guo J, Li Y, Franco OH, et al. Associations of healthy lifestyle and socioeconomic status with mortality and incident cardiovascular disease: two prospective cohort studies. BMJ. 2021;373:n604. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Seligman HK, Schillinger D. Hunger and Socioeconomic Disparities in Chronic Disease. N Engl J Med. 2010;363:6–9. [DOI] [PubMed] [Google Scholar]
- 31.Laraia BA. Food Insecurity and Chronic Disease123. Adv Nutr. 2013;4:203–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Myers CA, Mire EF, Katzmarzyk PT. Trends in Adiposity and Food Insecurity Among US Adults. JAMA Netw Open. 2020;3:e2012767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Na M, Dou N, Ji N, Xie D, Huang J, Tucker KL, et al. Food Insecurity and Cognitive Function in Middle to Older Adulthood: A Systematic Review. Adv Nutr. 2020;11:667–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Lee JS, Frongillo EA Jr. Factors Associated With Food Insecurity Among U.S. Elderly Persons: Importance of Functional Impairments. Journals Gerontology: Ser B. 2001;56:S94–9. [DOI] [PubMed] [Google Scholar]
- 35.Awuviry-Newton K, Amoah D, Tavener M, Afram AA, Dintrans PV, Byles J et al. Food Insecurity and Functional Disability Among Older Adults in Ghana: The Role of Sex and Physical Activity. Journal of the American Medical Directors Association. 2022;23:1432.e1431-1432e1437. [DOI] [PubMed]
- 36.Lee YS, Kim TH. Household food insecurity and breakfast skipping: Their association with depressive symptoms. Psychiatry Res. 2019;271:83–8. [DOI] [PubMed] [Google Scholar]
- 37.Smith L, Il Shin J, McDermott D, Jacob L, Barnett Y, López-Sánchez GF, et al. Association between food insecurity and depression among older adults from low- and middle-income countries. Depress Anxiety. 2021;38:439–46. [DOI] [PubMed] [Google Scholar]
- 38.Lee JS, Frongillo EA. Nutritional and health consequences are associated with food insecurity among U.S. elderly persons. J Nutr. 2001;131:1503–9. [DOI] [PubMed] [Google Scholar]
- 39.Volpp KG, Berkowitz SA, Sharma SV, Anderson CAM, Brewer LC, Elkind MSV, et al. Food Is Medicine: A Presidential Advisory From the American Heart Association. Circulation. 2023;148:1417–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Isath A, Koziol KJ, Martinez MW, Garber CE, Martinez MN, Emery MS, et al. Exercise and cardiovascular health: A state-of-the-art review. Prog Cardiovasc Dis. 2023;79:44–52. [DOI] [PubMed] [Google Scholar]
- 41.Pano O, Sayón-Orea C, Hershey MS, de la Fernández-Lázaro OV, Bes-Rastrollo C. The risk of incident depression when assessed with the Lifestyle and Well-Being Index. Public Health. 2023;220:165–71. [DOI] [PubMed] [Google Scholar]
- 42.Jia J, Zhao T, Liu Z, Liang Y, Li F, Li Y, et al. Association between healthy lifestyle and memory decline in older adults: 10 year, population based, prospective cohort study. BMJ. 2023;380:e072691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Liu J, Rehm CD, Onopa J, Mozaffarian D. Trends in Diet Quality Among Youth in the United States, 1999–2016. JAMA. 2020;323:1161–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Isaura ER, Chen YC, Su HY, Yang SH. The relationship between food security status and sleep disturbance among adults: a cross-sectional study in an Indonesian population. Nutrients. 2020;12(11):3411. [DOI] [PMC free article] [PubMed]
- 45.Taylor Cunningham A, Carson Weinstein L, Stefancic A, Silverio A, Cabassa LJ. The association between food insecurity and physical activity in adults with serious mental illness living in supportive housing. Prev Med Rep. 2022;30:102008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Gucciardi E, Vahabi M, Norris N, Del Monte JP, Farnum C. The Intersection between Food Insecurity and Diabetes: A Review. Curr Nutr Rep. 2014;3:324–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Liu Y, Eicher-Miller HA. Food Insecurity and Cardiovascular Disease Risk. Curr Atheroscler Rep. 2021;23:24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Stupplebeen DA. Housing and Food Insecurity and Chronic Disease Among Three Racial Groups in Hawaiʻi. Prev Chronic Dis. 2019;16:E13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Vercammen KA, Moran AJ, McClain AC, Thorndike AN, Fulay AP, Rimm EB. Food Security and 10-Year Cardiovascular Disease Risk Among U.S. Adults. Am J Prev Med. 2019;56:689–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Zierath R, Claggett B, Hall ME, Correa A, Barber S, Gao Y, et al. Measures of Food Inadequacy and Cardiovascular Disease Risk in Black Individuals in the US From the Jackson Heart Study. JAMA Netw Open. 2023;6:e2252055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Trudell JP, Burnet ML, Ziegler BR, Luginaah I. The impact of food insecurity on mental health in Africa: A systematic review. Soc Sci Med. 2021;278:113953. [DOI] [PubMed] [Google Scholar]
- 52.Michel J-P, Leonardi M, Martin M, Prina M. WHO’s report for the decade of healthy ageing 2021-30 sets the stage for globally comparable data on healthy ageing. Lancet Healthy Longev. 2021;2:e121–2. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The dataset supporting the conclusions of this article is available on the NHIS website (https://www.cdc.gov/nchs/nhis/documentation/index.html) , and NHANES website (https://wwwn.cdc.gov/nchs/nhanes/default.aspx).