Abstract
Background
Tracking the prevalence of imbalances for essential nutrients over time provides insight into nutritional status changes and highlights vulnerabilities of demographic subgroups. Consolidated information for a broad range of biomarkers in the US population is missing.
Objectives
We aimed to describe nutritional deficiency and excess prevalence trends for 12 biomarkers of water- and fat-soluble vitamins and iron status in the US population participating in the NHANES 1999–2023 after adjusting for demographic changes over time.
Methods
We estimated the prevalence of deficiency (vitamins A, B-6, B-12, C, D, and E, as well as iron), insufficiency [vitamins B-6, B-9 (folate), B-12, C, and D], and excess (vitamins A and D as well as iron). We assessed trends as overall change in prevalence using logistic regression (unadjusted and adjusted for age, sex, and race/Hispanic origin) and as linear trend (if overall change was significant).
Results
After adjusting for demographic changes, we observed small (typically <5 percentage points) temporal changes (significant linear trends) in the prevalence of deficiency and/or insufficiency for vitamin B-12 and methylmalonic acid (≥20 y), and for vitamin E and 25-hydroxyvitamin D (25OHD) (≥6 y). The prevalence of folate insufficiency (red blood cell folate (RBF) <748 nmol/L) and iron deficiency (ferritin ≤15 ng/mL) increased from 15.7% to 21.4% and from 11.6% to 17.4%, respectively, in women of reproductive age. The prevalence of excess vitamin D (25OHD >125 nmol/L) increased from <1% to 8.06% in persons ≥6 y, with the highest prevalence in adult subgroups (16.2% in supplement users, 13.2% in females, and 12.3% in non-Hispanic White persons).
Conclusions
The large increases in excess 25OHD concentrations and modest increases in low ferritin and RBF concentrations merit attention and highlight the importance of providing gold standard data to continuously monitor the nutritional status of the US population and of subpopulations.
Keywords: biomarker, trend, prevalence, fat-soluble vitamin, water-soluble vitamin, iron indicator
Introduction
The NHANES is a unique source for nationally representative data on the health and nutritional status of the US population in the form of biomarker measurements, information collected on dietary intake and supplement use, and health examinations and questionnaires [1]. The “Centers for Disease Control and Prevention (CDC) National Report on Biochemical Indicators of Diet and Nutrition in the US Population (CDC Nutrition Report)” is a published resource that presents descriptive data on nutritional biomarker concentrations measured as part of the continuous NHANES (1999 and forward) in a consistent format that allows for comparisons across demographic subgroups, biomarkers, and over time [2]. The “Second CDC Nutrition Report,” released in 2012, included data on concentrations for 58 biomarkers (NHANES 2003–2006). Deficiency and/or excess prevalence data for 12 of those biomarkers were also included.
The 2026 CDC Nutrition Report covers all publicly available data on concentrations for 131 nutritional biomarkers from NHANES 1999–2000 to the most recent survey cycle performed from August 2021 to August 2023 [3]. With >2700 tables and 500 figures, it is very large in size. Due to the complexities of presenting prevalence on nutritional deficiencies and excesses, such as multiple cutoff values to account for different stages of the imbalance and/or different population groups of interest, prevalence estimates were not included in the 2026 CDC Nutrition Report. Tracking prevalence over time provides critical insight into potential changes in nutritional status and highlights vulnerabilities of specific demographic subgroups. Articles focusing on deficiencies or excesses of some nutrients measured as part of NHANES, such as folate [4], vitamin C [5], vitamin D [6], and iron [7] have been published, but there are gaps for other nutrients. Furthermore, there are no consolidated analyses that integrate multiple nutrient imbalances for a broader collection of essential nutrients across decades of nationally representative biomarker data and that apply the same level of stratification across biomarkers.
The goal of this article is to describe the magnitude of and changes in the prevalence of nutritional deficiencies and excesses for 12 biomarkers of water- and fat-soluble vitamins and iron status in the US population participating in NHANES 1999–2023 after adjusting for demographic changes over time.
Methods
Participants and survey design
The CDC’s National Center for Health Statistics (NCHS) conducts NHANES, which has been a continuous survey since 1999. The survey collects cross-sectional data on the health and nutritional status of the civilian, noninstitutionalized US population in 2-y cycles as part of a stratified, multistage, probability sample designed to represent the US population based on self-reported age, sex, and race and Hispanic origin. In addition to conducting health interviews and examinations, the survey also collects biological specimens to measure various health status and exposure biomarkers. Certain population subgroups are oversampled to increase the precision of estimates [Supplementary Table 1 (footnote)]. Interview and examination response rates for each survey period are publicly available [8]. Participants aged ≥18 y consented, parents provided consent for those aged <18 y, and youth 7–17 y assented. The NCHS Ethics Review Board approved NHANES [9].
Biomarker laboratory methods and data adjustment
The CDC laboratory at the Division of Laboratory Sciences, National Center for Environmental Health, analyzed biological specimens for all 12 nutritional biomarkers presented in this article as part of the NHANES protocol (Supplementary Table 1). Assay performance was monitored using Westgard-type quality control multirules [10]. Detailed laboratory method information is provided on the NHANES website separately for each survey cycle [11]. Because of changes in laboratory methods over time, we followed the NHANES analytical guidelines and adjusted data to allow appropriate assessment of long-term trends in prevalence estimates [Supplementary Table 1 (footnotes)].
Study variables
Our analysis included the following biomarkers: red blood cell folate (RBF), serum pyridoxal-5'-phosphate (PLP), serum vitamin B-12 (B-12), plasma total homocysteine (HCY), plasma or serum methylmalonic acid (MMA), serum vitamin C (VIC; ascorbic acid), serum vitamin A (VIA; retinol), serum vitamin E (VIE; α-tocopherol), serum 25-hydroxyvitamin D (25OHD), serum ferritin (FER), serum soluble transferrin receptor (TFR), and serum body iron index (BII). We calculated BII as follows: BII (milligrams/kilograms) = −[log10 (TFR × 1000 / FER) – 2.8229] / 0.1207 [12], with the TFR concentration representing an adjusted concentration to make the Roche Tina-quant TFR concentrations equivalent to the original assay [13] used in the development of the BII model: original TFR = 1.5 × Roche TFR + 0.35 mg/L [14].
For most of these biomarkers, data were not available continuously over the course of the 24-y period during NHANES 1999–August 2023. Only RBF and 25OHD had data available throughout all or most NHANES cycles. We used generally accepted cutoff values (Table 1) [[15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29]] to assess prevalence estimates for nutrient deficiencies, insufficiencies (to include the prevalence of deficiency), and/or excesses when biomarker data were available for the respective cycle. For folate, we only assessed the prevalence of insufficiency indicative of risk for neural tube defects; we did not assess the prevalence of folate deficiency, as it is known to be <1% in the overall population and in subgroups as a result of mandatory folic acid fortification of enriched cereal-grain products, which has been in place for 3 decades [4].
TABLE 1.
Cutoff values used to assess the prevalence of abnormal concentrations for biomarkers of water- and fat-soluble vitamins and iron status.
| Biomarker (unit) | Deficiency (Reference) | Insufficiency1 (Reference) | Excess (Reference) |
|---|---|---|---|
| RBF (nmol/L) | N/A | WRA: <748 [15] | N/A |
| PLP (nmol/L) | All ages: <20 [16] | All ages: <30 [17] | N/A |
| B-12 (pmol/L) | All ages: <148 [18] | All ages: <221 [18] | N/A |
| MMA (nmol/L) | All ages: >376 [19] | All ages: >271 [19] | N/A |
| HCY (μmol/L) | All ages: >15 [20] | N/A | N/A |
| VIC (μmol/L) | All ages: <11.4 [21] | All ages: <23 [21] | N/A |
| VIA (μg/dL) | All ages: <20 [22,23] | N/A | All ages: >100 [24] |
| VIE (μg/dL) | All ages: <500 [25] | N/A | N/A |
| 25OHD (nmol/L) | All ages: <30 [26] | All ages: <50 [26] | All ages: >125 [26] |
| FER (ng/mL)2 | Ages >6 mo: ≤15 [27] | N/A | Ages >6 mo: >150 [28] |
| TFR (mg/L) | WRA: >5.33 [29] | N/A | N/A |
| Children 1–5 y: >6.0 [29] | N/A | N/A | |
| BII (mg/kg) | WRA: <0 [12] | N/A | N/A |
Abbreviations: 25OHD, 25-hydroxyvitamin D; B-12, vitamin B-12; BII, body iron index; FER, ferritin; HCY, total homocysteine; MMA, methylmalonic acid; PLP, pyridoxal-5'-phosphate; RBF, red blood cell folate; TFR, soluble transferrin receptor; VIA, vitamin A (retinol); VIE, vitamin E (α-tocopherol); VIC, vitamin C (ascorbic acid); WRA, women of reproductive age. N/A, not applicable (data available but not used)
Insufficiency cutoffs include the prevalence of deficiency.
FER cutoffs stay the same after adjustment for inflammation.
Our covariates were age, sex (male and female), race and Hispanic origin categories [Mexican American (MA), non-Hispanic Black (NHB), and non-Hispanic White (NHW) for NHANES 1999–August 2023 and also non-Hispanic Asian (NHA) for NHANES 2011–August 2023; other race and Hispanic origin groups were included in overall estimates but not presented separately per NCHS analytic guidelines] [30], and supplement use (dichotomized as yes/no based on self-reported consumption of any dietary supplements in the past 30 d). Because biomarker data were not available for the same age groups across survey cycles, we defined consistent age-specific analytic populations: children ≤11 y, adolescents 12–19 y, and adults ≥20 y. For folate and iron indicators, we assessed prevalence estimates for women of reproductive age (WRA, 12–49 y) and for iron indicators also for children 1–5 y, as these groups are of particular public health interest.
Statistical analysis
We followed a data analysis plan approved by all authors before conducting the statistical analysis using SAS for Windows, version 9.4 (SAS Institute Inc.). Analyses accounted for the complex survey design of NHANES by applying Taylor series linearization, which calculates appropriate variance estimates by accounting for the stratified, multistage cluster sampling and incorporates survey weights. We included data from all available 2-y cycles from 1999–2000 to August 2021–August 2023, except for the 2019–2020 cycle, affected by the COVID-19 pandemic, during which field operations were suspended and the resulting data were not considered nationally representative [31]. We excluded pregnant and lactating women aged 20–44 y from our analytical sample using the NHANES-provided pregnancy and lactation information that was available for this specific age group across all cycles. Because FER is a positive acute-phase protein, we used R, a program for statistical computing (R Foundation for Statistical Computing) to calculate a specialized adjustment for FER concentrations (FERadj) using the Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA) approach, which considered each participants C-reactive protein (CRP) concentration to adjust for inflammation [[32], [33], [34]]. Biomarker data availability ranged from ∼250 children 1–5 y to ∼8000 participants ≥1 y of age per cycle (Table 2).
TABLE 2.
Sample sizes for biomarkers of water- and fat-soluble vitamins and iron status in the US population by age group, NHANES 1999–20231,2.
| Group | Biomarker | Survey cycle |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1999–2000 | 2001–2002 | 2003–2004 | 2005–2006 | 2007–2008 | 2009–2010 | 2011–2012 | 2013–2014 | 2015–2016 | 2017–2018 | August 2021–August 2023 | ||
| ≥1 y | PLP | — | — | — | 7966 | 8090 | 8343 | — | — | — | — | — |
| ≥6 y | VIC | — | — | 7041 | 6959 | — | — | — | — | — | 6659 | — |
| VIA | 6837 | 7615 | — | 6910 | — | — | — | — | — | 6727 | — | |
| VIE | 6791 | 7615 | — | 6910 | — | — | — | — | — | 6580 | — | |
| 25OHD | — | 7489 | 7160 | 7056 | 6336 | 7804 | 7008 | 7561 | 7202 | 6733 | 6988 | |
| ≥20 y | B-12 | 3889 | 4402 | 4249 | 4052 | — | — | 4804 | 5216 | — | — | — |
| MMA | 3924 | 4429 | 4111 | — | — | — | 4807 | 5213 | — | — | — | |
| HCY | 3925 | 4440 | 4270 | 4144 | — | — | — | — | — | — | — | |
| WRA | RBF | 1966 | 2192 | 1940 | 1976 | 1759 | 2042 | 1786 | 1967 | 1830 | 1608 | 1662 |
| FER | N/A | N/A | 1951 | 1999 | 1741 | 2037 | — | — | 1830 | 1595 | 1626 | |
| FERadj | N/A | N/A | 1951 | 1999 | 1740 | 2037 | — | — | 1810 | 1569 | 1618 | |
| TFR | — | — | 1906 | 2000 | 1739 | 2037 | — | — | 1798 | 1584 | 1626 | |
| BII | — | — | 1904 | 1998 | 1733 | 2037 | — | — | 1798 | 1584 | 1626 | |
| 1–5 y | FER | N/A | N/A | 796 | 686 | 670 | 778 | — | — | 751 | 602 | 262 |
| FERadj | N/A | N/A | 794 | 668 | N/A | N/A | — | — | 649 | 553 | 237 | |
| TFR | — | — | 696 | 679 | 667 | 778 | — | — | 738 | 577 | 261 | |
Abbreviations: 25OHD, 25-hydroxyvitamin D; B-12, vitamin B-12; BII, body iron index; FER, ferritin; FERadj, ferritin adjusted for inflammation (C-reactive protein only) using BRINDA equation; HCY, total homocysteine; MMA, methylmalonic acid; N/A, not applicable (data available but not used); PLP, pyridoxal-5'-phosphate; RBF, red blood cell folate; TFR, soluble transferrin receptor; VIA, vitamin A (retinol); VIC, vitamin C (ascorbic acid); VIE, vitamin E (α-tocopherol); WRA, women of reproductive age (females 12–49 y).
Pregnant and lactating women ages 20–44 y were excluded.
Data from 2019 to March 2020 cycle not included (field operations suspended in March 2020 due to COVID-19 pandemic; data not considered nationally representative).
We used 2-y mobile examination center (MEC) weights provided by NHANES to adjust for unequal probabilities of selection and nonresponse, with the following exceptions: for 2017–2018 a subsample weight was applied for RBF and for August 2021–August 2023 the phlebotomy weights were used to account for differences in nonresponse patterns identified in the phlebotomy component of the MEC [35]. Weighted prevalence estimates and 95% confidence intervals were calculated using the Clopper–Pearson method [36], with the Korn and Graubard adjustment applied to the degrees of freedom or sample size when appropriate, to account for the complex survey design [37]. Estimates were evaluated for reliability and reportability according to the NCHS Data Presentation Standards for Proportions [38]. Differences across survey cycles were assessed using both unadjusted and adjusted Wald F-tests reported from logistic regression models (with survey cycle modeled categorically) and adjusted for age, sex, and race and Hispanic origin to account for changes in population demographics over time. Vitamin D models were additionally adjusted for season to account for seasonal variation in 25OHD concentrations and the seasonal design of the NHANES data collection. Adjusted linear Wald F-tests were also conducted by modeling survey cycle as a continuous variable in the logistic regression models; results from these tests were not reported if the adjusted overall Wald F-test was not significant (P ≤ 0.05) or if <4 cycles were reported. The overall Wald F-tests were used as omnibus tests, whereas our interpretation focused on the linear Wald F-tests that captured directional changes over time.
Results
Trends in prevalences of abnormal water- and fat-soluble biomarker concentrations for the overall population
Data for RBF were available for all cycles. The prevalence of folate insufficiency (RBF <748 nmol/L) in WRA increased from 15.7% to 21.4% (significant linear trend) between 1999 and August 2023 (Table 3). Data for PLP for persons ≥1 y were available for 2005–2010 but linear trends were not assessed due to only 3 data points (Table 3). After adjusting for demographic changes, the prevalence of vitamin B-6 deficiency (PLP <20 nmol/L) increased significantly from 10.2% to 11.1% over these 3 cycles, whereas insufficiency (PLP <30 nmol/L; ∼25%) did not change. Data for B-12, MMA, and HCY were mostly available for 1999–2006 and 2011–2014 for persons ≥20 y (Table 3). The prevalence for vitamin B-12 deficiency (B-12 <148 pmol/L; <2%) did not change over time, whereas insufficiency (B-12 <221 pmol/L) increased from 7.26% to 10.7% (significant linear trend) between 1999 and 2014. MMA, a sensitive indicator of vitamin B-12 status, also increased from 6.47% to 9.42% (significant linear trend) between 1999 and 2014 at the lower cutoff (>271 nmol/L) corresponding to vitamin B-12 insufficiency, but did not change after adjusting for demographic changes at the higher cutoff (>376 nmol/L; ∼3%–4%) corresponding to vitamin B-12 deficiency. The prevalence of HCY >15 μmol/L, an indicator of folate, B-12, and/or B-6 vitamin deficiency, did not change over time (∼4%–5%). Data for VIC for persons ≥6 y were available for 2003–2006 and 2017–2018 but linear trends were not assessed due to only 3 data points (Table 3). Neither the prevalence of vitamin C deficiency (VIC <11.4 μmol/L; ∼5%–7%) nor insufficiency (VIC <23.0 μmol/L; ∼14%–17%) changed over time.
TABLE 3.
Prevalence for nutrient deficiency, combined deficiency and insufficiency, and excess for water- and fat-soluble vitamin biomarkers in the US population by age group, NHANES 1999–20231,2.
| Group | Biomarker | Survey cycle (%) |
Trend3 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1999–2000 | 2001–2002 | 2003–2004 | 2005–2006 | 2007–2008 | 2009–2010 | 2011–2012 | 2013–2014 | 2015–2016 | 2017–2018 | Aug 2021–Aug 2023 | |||
| WRA | RBF <748 nmol/L | 15.7 (12.2–19.8) | 13.5 (10.9–16.6) | 19.8 (17.4–22.4) | 14.3 (11.8–17.1) | 19.7 (16.8–22.8) | 26.6 (23.4–30.0) | 22.8 (19.6–26.2) | 16.9 (13.6–20.5) | 16.3 (12.7–20.3) | 18.9 (15.8–22.4) | 21.4 (18.6–24.4) | HS; HS; S |
| ≥1 y | PLP <20 nmol/L | — | — | — | 10.2 (8.87–11.7) | 8.16 (6.72–9.79) | 11.1 (9.59–12.7) | — | — | — | — | — | S; S; — |
| PLP <30 nmol/L | — | — | — | 24.5 (22.4–26.7) | 21.8 (19.5–24.3) | 25.9 (23.7–28.3) | — | — | — | — | — | S; NS; — | |
| ≥6 y | VIC <11.4 μmol/L | — | — | 7.38 (5.46–9.71) | 4.87 (3.70–6.27) | — | — | — | — | — | 5.99 (4.30–8.09) | — | NS; NS; — |
| VIC <23.0 μmol/L | — | — | 16.2 (12.8–20.0) | 13.6 (11.6–15.9) | — | — | — | — | — | 17.0 (13.8–20.6) | — | NS; NS; — | |
| VIA <20 μg/dL | .129 (.059–.247) | .261 (.159–.404) | — | .285 (.138–.518) | — | — | — | — | — | .272 (.162–.428) | — | S; HS; NS | |
| VIA >100 μg/dL | 1.36 (1.01–1.80) | 2.09 (1.66–2.60) | — | 2.12 (1.59–2.78) | — | — | — | — | — | .751 (.455–1.17) | — | HS; HS; HS | |
| VIE <500 μg/dL | 1.05 (.658–1.60) | .442 (.306–.618) | — | .661 (.463–.913) | — | — | — | — | — | .141 (.055–.296) | — | HS; HS; HS | |
| 25OHD <30 nmol/L | — | 4.97 (3.69–6.53) | 6.97 (4.77–9.78) | 4.83 (3.52–6.45) | 5.97 (4.31–8.01) | 6.29 (4.70–8.20) | 5.13 (3.66–6.98) | 5.43 (3.94–7.27) | 4.57 (3.23–6.26) | 4.99 (3.94–6.22) | 5.76 (4.43–7.35) | NS; S; S | |
| 25OHD <50 nmol/L | — | 28.0 (24.6–31.6) | 28.6 (22.6–35.2) | 30.6 (25.9–35.6) | 24.7 (20.8–29.0) | 24.8 (20.9–29.0) | 24.2 (18.9–30.1) | 24.2 (20.5–28.3) | 23.2 (18.4–28.5) | 22.0 (18.2–26.2) | 21.6 (18.5–25.1) | S; HS; HS | |
| 25OHD >125 nmol/L | — | .779 (.498–1.16) | 1.40 (.781–2.31) | .827 (.554–1.18) | 2.28 (1.56–3.21) | 2.48 (1.73–3.43) | 4.07 (3.06–5.30) | 3.56 (2.77–4.48) | 4.58 (3.16–6.39) | 5.18 (4.08–6.47) | 8.06 (7.16–9.04) | HS; HS; HS | |
| ≥20 y | B-12 <148 pmol/L | 1.00 (.591–1.59) | 1.15 (.806–1.60) | 1.36 (1.01–1.79) | 1.51 (1.06–2.07) | — | — | 1.76 (1.28–2.35) | 1.82 (1.10–2.83) | — | — | — | NS; NS; — |
| B-12 <221 pmol/L | 7.26 (6.21–8.43) | 8.99 (7.81–10.0) | 7.82 (5.98–10.0) | 8.48 (6.79–10.4) | — | — | 10.9 (9.29–12.8) | 10.7 (9.55–11.8) | — | — | — | HS; HS; HS | |
| MMA >271 nmol/L | 6.47 (5.51–7.56) | 6.17 (5.28–7.16) | 7.60 (6.05–9.40) | — | — | — | 8.46 (6.88–10.3) | 9.42 (8.13–10.8) | — | — | — | HS; HS; HS | |
| MMA >376 nmol/L | 2.80 (2.05–3.73) | 2.83 (2.29–3.45) | 3.18 (2.52–3.95) | — | — | — | 4.07 (3.53–4.67) | 3.78 (3.06–4.61) | — | — | — | HS; NS; — | |
| HCY >15 μmol/L | 4.15 (3.21–5.27) | 4.47 (3.60–5.47) | 5.04 (3.92–6.37) | 3.71 (3.05–4.47) | — | — | — | — | — | — | — | NS; NS; — | |
Abbreviations: 25OHD, 25-hydroxyvitamin D; B-12, vitamin B-12; HCY, total homocysteine; HS, highly significant; MMA, methylmalonic acid; NS, not significant; N/A, not applicable (data available but not used); PLP, pyridoxal-5'-phosphate; RBF, red blood cell folate; S, significant; VIA, vitamin A (retinol); VIC, vitamin C (ascorbic acid); VIE, vitamin E (α-tocopherol); WRA, women of reproductive age (females 12–49 y).
Values are weighted prevalence estimates in percent (95% confidence interval). Estimates were evaluated for reliability and reportability according to the NCHS Data Presentation Standards for Proportions. Pregnant and lactating women ages 20–44 y were excluded. For sample sizes see Table 2.
Data from 2019–March 2020 cycle not included (field operations suspended in March 2020 due to COVID-19 pandemic; data not considered nationally representative).
Three trend tests are reported in the following order: (1) unadjusted Wald F-test for overall change using logistic regression; (2) adjusted Wald F-test for overall change controlling for age, sex, and race and Hispanic origin (with additional adjustment for season in VID models); and (3) adjusted linear trend with the same covariate adjustments. Tests were designed to detect any significant differences across groups, with the Wald F-tests serving as omnibus tests for overall change, and the linear trend test specifically assessing monotonic linear trends. Semiquantitative significance for each test is reported as follows: HS (highly significant), P < 0.01; S (significant), P = 0.01 to ≤ 0.05; NS (not significant), P > 0.05; linear trend test not reported if the adjusted overall change was not significant (P ≤ 0.05) or if <4 cycles are reported.
Data for the antioxidant vitamins A and E were mostly available for 1999–2006 and 2017–2018 for persons ≥6 y (Table 3). The prevalence of vitamin A (VIA <20 μg/dL) and vitamin E (VIE <500 μg/dL) deficiency were 1% or less, with the former increasing and the latter decreasing significantly. The prevalence of vitamin A excess (VIA >100 μg/dL) decreased from 1.36% to 0.751% (significant linear trend) between 1999 and 2018. The prevalence for risk of vitamin D insufficiency (25OHD <50 nmol/L) in persons ≥6 y decreased from 28.0% to 21.6%, whereas the prevalence for risk of excess (25OHD >125 nmol/L; “may be of concern”) and the prevalence for risk of deficiency (25OHD <30 nmol/L) increased from <1% to 8.06% and from 4.97% to 5.76% between 2001 and August 2023, respectively (significant linear trend).
In summary, we found significant linear trends for the following water- and fat-soluble vitamin biomarkers: RBF <748 nmol/L, B-12 <221 pmol/L, MMA >271 nmol/L, VIA >100 μg/dL, VIE <500 μg/dL, and 25OHD <30, <50, and >125 nmol/L (Table 4). The biomarkers showing no change over time were PLP <30 nmol/L, B-12 <148 pmol/L, MMA >376 nmol/L, HCY >15 μmol/L, and VIC <11.4 and <23.0 μmol/L.
TABLE 4.
Summary of findings for water- and fat-soluble vitamin biomarkers, NHANES 1999–2023.
| Group | Biomarker and cutoff | Prevalence for overall population |
Prevalence by subgroup | ||||
|---|---|---|---|---|---|---|---|
| No overall change | No overall change after adjusting | Overall change after adjusting | Linear trend | Magnitude of overall change in prevalence | |||
| WRA | RBF <748 nmol/L | Yes | Yes | Increased from ∼15% to ∼20% | Mostly change after adjusting (except NHA) | ||
| ≥1 y | PLP <20 nmol/L | Yes | Not assessed | Fluctuated (10% to 14%) | Mostly no change (except females, NHB, S-users, S-nonusers) | ||
| PLP <30 nmol/L | Yes | Not assessed | No overall change | Mostly no change (except 12–19 y, 20+ y, females, S-users) | |||
| ≥6 y | VIC <11.4 μmol/L | Yes | Yes | No overall change | Mostly no change (except 6–11 y) | ||
| VIC <23.0 μmol/L | Yes | Yes | No overall change | Mostly no change (except NHB and S-users) | |||
| VIA <20 μg/dL | Yes | Prevalences <1% | Mostly no change (except ≥20 y, NHW, S-nonusers) | ||||
| VIA >100 μg/dL | Yes | Yes | Prevalences <1% | Mostly linear trend (except females, NHB) | |||
| VIE <500 μg/dL | Yes | Yes | Prevalences ∼1% | Differing trends | |||
| 25OHD <30 nmol/L | Yes | Yes | Prevalences ∼5% | Differing trends | |||
| 25OHD <50 nmol/L | Yes | Yes | Decreased from ∼30% to 22% | Differing trends | |||
| 25OHD >125 nmol/L | Yes | Yes | Increased from <1% to 8% | Mostly linear trend (except 6–11 y, 12–19 y, S-nonusers) | |||
| ≥20 y | B-12 <148 pmol/L | Yes | Yes | No overall change | Mostly no change (except 12-19 y, NHA, S-nonusers) | ||
| B-12 <221 pmol/L | Yes | Yes | Increased from ∼7% to ∼11% | Differing trends | |||
| MMA >271 nmol/L | Yes | Yes | Increased from ∼6% to ∼9% | Differing trends | |||
| MMA >376 nmol/L | Yes | No change | Differing trends | ||||
| HCY >15 μmol/L | Yes | Yes | No change | Mostly change after adjusting (except 12–19 y, females) | |||
Abbreviations: 25OHD, 25-hydroxyvitamin D; B-12, vitamin B-12; HCY, total homocysteine; MA, Mexican American; MMA, methylmalonic acid; NHA, non-Hispanic Asian; NHB, non-Hispanic Black; NHW, non-Hispanic White; PLP, pyridoxal-5'-phosphate; RBF, red blood cell folate; S-users, supplement users; S-nonusers, supplement nonusers; VIA, vitamin A (retinol); VIC, vitamin C (ascorbic acid); VIE, vitamin E (α-tocopherol).
Trends in the prevalence of abnormal iron status in WRA and children
Data for FER, TFR, and BII were available for 2003–2010 and 2015–August 2023 (Table 5). In WRA, the prevalence of iron deficiency (FER or FERadj ≤15 ng/mL) increased from 13.0% to 18.3% (significant linear trend) between 2003 and August 2023 for FERadj, whereas iron deficiency for other biomarkers (TFR >5.33 mg/L and BII <0 mg/kg) did not change over time. Iron excess (FER >150 ng/mL) decreased significantly over time (from 7.26% to 5.05%) but the linear trend was no longer significant after adjusting for inflammation. In children 1–5 y of age, none of the iron biomarkers changed significantly over time.
TABLE 5.
Prevalence for iron deficiency and excess for iron status indicators in US women of reproductive age 12–49 y and children 1–5 y, NHANES 1999–20231,2.
| Group | Biomarker | Survey cycle (%) |
Trend3 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1999–2000 | 2001–2002 | 2003–2004 | 2005–2006 | 2007–2008 | 2009–2010 | 2011–2012 | 2013–2014 | 2015–2016 | 2017–2018 | August 2021–August 2023 | |||
| WRA | FER ≤15 ng/mL | N/A | N/A | 11.6 (9.12–14.4) | 12.3 (10.9–13.8) | 15.0 (12.7–17.4) | 17.0 (14.7–19.6) | — | — | 18.0 (14.8–21.5) | 17.4 (15.0–19.9) | 17.4 (15.4–19.6) | HS; HS; HS |
| FERadj ≤15 ng/mL | N/A | N/A | 13.0 (10.4–16.0) | 12.9 (11.5–14.4) | 16.0 (13.7–18.6) | 17.0 (14.7–19.5) | — | — | 19.0 (15.9–22.4) | 18.5 (16.3–20.8) | 18.3 (16.4–20.3) | HS; HS; HS | |
| FER >150 ng/mL | N/A | N/A | 7.26 (5.81–8.93) | 5.31 (3.37–7.91) | 5.95 (4.73–7.36) | 5.34 (4.23–6.64) | — | — | 3.81 (2.89–4.91) | 4.18 (2.63–6.27) | 5.05 (4.01–6.28) | HS; S; S | |
| FERadj >150 ng/mL | N/A | N/A | 5.86 (4.55–7.40) | 4.56 (2.81–6.96) | 4.90 (3.75–6.28) | 4.11 (3.18–5.22) | — | — | 3.01 (2.15–4.09) | 3.35 (2.08–5.10) | 4.41 (3.34–5.69) | S; S; NS | |
| TFR >5.33 mg/L | — | — | 7.58 (5.77–9.75) | 8.63 (6.39–11.3) | 9.36 (7.48–11.5) | 7.23 (5.92–8.72) | — | — | 10.6 (7.45–14.6) | 8.03 (6.24–10.1) | 9.00 (7.50–10.7) | NS; NS; — | |
| BII <0 mg/kg | — | — | 7.96 (6.15–10.1) | 9.24 (8.00–10.6) | 10.4 (8.88–12.2) | 9.05 (7.66–10.6) | — | — | 12.3 (9.35–15.8) | 10.2 (8.10–12.6) | 10.9 (9.39–12.5) | NS; NS; — | |
| 1–5 y | FER ≤15 ng/mL | N/A | N/A | 11.8 (8.77–15.5) | 10.8 (7.46–15.0) | 13.1 (9.13–18.1) | 13.6 (10.7–16.9) | — | — | 11.5 (8.84–14.7) | 11.0 (8.60–13.8) | 10.2 (5.34–17.3) | NS; NS; — |
| FERadj ≤15 ng/mL | N/A | N/A | 12.4 (9.15–16.3) | 11.2 (7.73–15.6) | N/A | N/A | — | — | 11.8 (9.46–14.6) | 10.5 (8.02–13.5) | 8.00 (4.33–13.3) | NS; NS; — | |
| TFR >6.0 mg/L | — | — | 4.80 (2.73–7.76) | 5.25 (3.69–7.20) | 4.81 (3.05–7.17) | 4.61 (2.34–8.07) | — | — | 5.58 (3.73–7.98) | 3.47 (2.00–5.56) | 3.57 (1.30–7.70) | NS; NS; — | |
Abbreviations: BII, body iron index; FER, ferritin; FERadj, ferritin adjusted for inflammation (C-reactive protein only) using BRINDA equation; HS, highly significant; NS, not significant; N/A, not applicable (data available but not used); S, significant; TFR, soluble transferrin receptor; WRA, women of reproductive age.
Values are weighted prevalence estimates in percent (95% confidence interval). Estimates were evaluated for reliability and reportability according to the National Center for Health Statistics Data Presentation Standards for Proportions. Pregnant and lactating women ages 20–44 y were excluded. For sample sizes see Table 2.
Data from 2019–March 2020 cycle not included (field operations suspended in March 2020 due to COVID-19 pandemic; data not considered nationally representative).
Three trend tests are reported in the following order: (1) unadjusted Wald F-test for overall change using logistic regression; (2) adjusted Wald F-test for overall change controlling for age, sex, and race and Hispanic origin (with additional adjustment for season in VID models); and (3) adjusted linear trend with the same covariate adjustments. Tests were designed to detect any significant differences across groups, with the Wald F-tests serving as omnibus tests for overall change, and the linear trend test specifically assessing monotonic linear trends. Semiquantitative significance for each test is reported as follows: HS (highly significant), P < 0.01; S (significant), P = 0.01 to ≤ 0.05; NS (not significant), P > 0.05; linear trend test not reported if the adjusted overall change was not significant (P ≤ 0.05) or if <4 cycles are reported.
In summary, we found significant linear trends for the following iron indicators: FER and FERadj ≤15 ng/mL for WRA and FER >150 ng/mL for WRA (Table 6). The iron indicators showing no change over time were as follows: for WRA, TFR >5.33 mg/L and BII <0 mg/kg; for children, FER and FERadj ≤15 ng/mL, and TFR >6.0 mg/L.
TABLE 6.
Summary of findings for iron status indicators, NHANES 1999–2023.
| Group | Biomarker and cutoff | Prevalence for overall population |
Prevalence by subgroup | ||||
|---|---|---|---|---|---|---|---|
| No overall change | No overall change after adjusting | Overall change after adjusting | Linear trend | Magnitude of overall change in prevalence | |||
| WRA | FER ≤15 ng/mL | Yes | Yes | Increased from ∼12% to ∼18% | Differing trends | ||
| FERadj ≤15 ng/mL | Yes | Yes | Increased from ∼13% to ∼19% | Differing trends | |||
| FER >150 ng/mL | Yes | Yes | Decreased from ∼7% to ∼5% | Differing trends | |||
| FERadj >150 ng/mL | Yes | Decreased from ∼6% to ∼4% | Differing trends | ||||
| TFR >5.33 mg/L | Yes | Yes | No change | Mostly no change (except 12–19 y, MA, NHA) | |||
| BII <0 mg/kg | Yes | Yes | Increased from ∼8% to ∼11% | Mostly no change after adjusting (except MA) | |||
| 1–5 y | FER ≤15 ng/mL | Yes | Yes | No change | No change throughout | ||
| FERadj ≤15 ng/mL | Yes | Yes | No change | Mostly no change (except S-nonusers) | |||
| TFR >6.0 mg/L | Yes | Yes | No change | No change throughout | |||
Abbreviations: BII, body iron index; FER, ferritin; FERadj, ferritin adjusted for inflammation (CRP only) using BRINDA equation; MA, Mexican American; NHA, non-Hispanic Asian; S-nonusers, supplement nonusers; TFR, soluble transferrin receptor; WRA, women of reproductive age (females 12–49 y).
Trends in prevalence by population subgroups
We generated heat maps to illustrate changes in prevalence over time by demographic subgroups and supplement use (FIGURE 1, FIGURE 2, FIGURE 3); corresponding data tables appear in Supplementary Tables 2–15. We also generated heat maps of P values from the 3 trend tests for a visual representation of the significance level by biomarker and population subgroup (Supplementary Figures 1–3). When reviewing the prevalence heat maps, numerous gaps in nutritional biomarker data across NHANES cycles are apparent. The prevalence of deficiency for B-12, VIA, and VIE were consistently low (typically <2%) and for FER among WRA consistently high (typically >14%) (Figure 1). Prevalences of deficiency increased with age and were generally the lowest in NHW persons compared with other race and Hispanic origin categories.
FIGURE 1.
Prevalence of nutrient deficiencies by survey cycle and demographic subgroup or supplement use in the US population, NHANES 1999–2023. For panels 1 and 4–6: data are shown for persons ≥1 y (PLP), ≥6 y (VIC, VIA, VIE, and 25OHD), ≥20 y (B-12, MMA, and HCY), or for WRA (FERadj, TFR, and BII). For panels 2–3 and 7–12: data are shown for persons ≥20 y (PLP, B-12, MMA, HCY, VIC, VIA, VIE, and 25OHD) or for WRA (FERadj, TFR, and BII). Empty white spaces mean that no data were available. 25OHD, 25-hydroxyvitamin D; B-12, vitamin B-12; BII, body iron index; FERadj, ferritin after adjusting for inflammation using BRINDA equation; HCY, total homocysteine; MMA, methylmalonic acid; N/A, not applicable (data available but not used); PLP, pyridoxal-5'-phosphate; TFR, soluble transferrin receptor; WRA, women of reproductive age (females 12–49 y); VIA, vitamin A (retinol); VIC, vitamin C (ascorbic acid); VIE, vitamin E (α-tocopherol).
FIGURE 2.
Prevalence of nutrient insufficiencies by survey cycle and demographic subgroup or supplement use in the US population, NHANES 1999–2023. For panels 1 and 4–6: data are shown for persons ≥1 y (PLP), ≥6 y (VIC and 25OHD), ≥20 y (B-12 and MMA), or for WRA (RBF). For panels 2–3 and 7–12: data are shown for persons ≥20 y (PLP, B-12, MMA, VIC, and 25OHD) or for WRA (RBF). Empty white spaces mean that no data were available. 25OHD, 25-hydroxyvitamin D; B-12, vitamin B-12; MMA, methylmalonic acid; N/A, not applicable (data available but not used); PLP, pyridoxal-5'-phosphate; RBF, red blood cell folate; WRA, women of reproductive age (females 12–49 y); VIC, vitamin C (ascorbic acid).
FIGURE 3.
Prevalence of nutrient excesses by survey cycle and demographic subgroup or supplement use in the US population, NHANES 1999–2023. For panels 1 and 4–6: data are shown for persons ≥6 y (VIA and 25OHD) or for WRA (FERadj). For panels 2–3 and 7–12: data are shown for persons ≥20 y (VIA and 25OHD) or for WRA (FERadj). Empty white spaces mean that no data were available. 25OHD, 25-hydroxyvitamin D; FERadj, ferritin after adjusting for inflammation using BRINDA equation; N/A, not applicable (data available but not used); WRA, women of reproductive age (females 12–49 y); VIA, vitamin A (retinol).
The prevalence of insufficiency was lowest for B-12 and MMA and generally >14% for other biomarkers with little differences across subgroups except for children aged 1–5 y (Figure 2). The prevalence of excess was low (<5%) for FER and VIA. A clear picture emerged for the prevalence of excess 25OHD with successive increases in prevalence over time, particularly noticeable for adult supplement users, females, and NHW and NHB persons (Figure 3).
Discussion
This comprehensive analysis investigated the magnitude of and changes in the prevalence of nutritional imbalances for 12 biomarkers of water- and fat-soluble vitamins and iron status assessed during NHANES 1999–2023 after adjusting for demographic changes over time. For most of the 25 biomarker-cutoff combinations we observed either no change over time (HCY, VIC, FER in children, and TFR in children and WRA), no change at 1 of the cutoff values (PLP <30 nmol/L, B-12 <148 pmol/L, and MMA >376 nmol/L), or a small change <5 percentage points (B-12 <221 pmol/L, PLP <20 nmol/L, MMA >271 nmol/L, VIA <20 or >100 μg/dL, VIE <500 μg/dL, 25OHD <30 nmol/L, FER >150 ng/mL in WRA, and BII <0 mg/kg in WRA) (TABLE 5, TABLE 6). Folate insufficiency (RBF <748 nmol/L) increased modestly (5–10 percentage points). We noted 3 larger changes: risk of vitamin D insufficiency decreased, whereas risk of excess increased; the prevalence of low FER in WRA increased. The extent of nutritional imbalances in adults or WRA across the 8 nutrients, using in each case the latest NHANES data available (Figure 4), varied from <1% for VIA and VIE deficiency (panel A) and for VIA excess (panel C), to 20%–30% for vitamin D, folate, and vitamin B-6 insufficiency (panel B).
FIGURE 4.
Prevalence of nutrient deficiencies (A), insufficiencies (B), or excesses (C) in descending order using the latest data available for US adults or women of reproductive age, NHANES 1999–2023. Error bars represent 95% confidence intervals. The y-axis label designates the biomarker, cutoff point, population group, and NHANES survey cycle. 25OHD, 25-hydroxyvitamin D; B-12, vitamin B-12; BII, body iron index; FERadj, ferritin after adjusting for inflammation using BRINDA equation; HCY, total homocysteine; MMA, methylmalonic acid; PLP, pyridoxal-5'-phosphate; RBF, red blood cell folate; TFR, soluble transferrin receptor; VIA, vitamin A (retinol); VIC, vitamin C (ascorbic acid); VIE, vitamin E (α-tocopherol).
The importance of adequate vitamin D status has received much scientific attention and media coverage over the last 2 decades, resulting in an exponential increase in peer-reviewed publications [39] and an 83-fold increase in the number of 25OHD blood tests among Medicare beneficiaries from 2000–2010 [40]. This was fueled by accumulating evidence from observational studies indicating that vitamin D status was inversely associated with nonskeletal health outcomes [41]. Large randomized clinical trials did not support the expected health benefits of vitamin D supplementation in vitamin D-replete adults regarding improved bone health, cancer or cardiovascular risk, the progression of prediabetes to type 2 diabetes, or overall mortality [42]. The jury is still out concerning benefits of vitamin D supplementation to the immune function. During active vitamin D treatment in the VITamin D and OmegA-3 TriaL (VITAL trial), 25OHD concentrations were associated with a temporary reduction in incident autoimmune disease [43] but the benefit was not sustained after trial termination [44]. Mendelian randomization studies have shown that individuals with genetically driven lower serum 25OHD concentrations have an increased risk of developing multiple sclerosis, either during adolescence or adulthood [42]. Large increases in vitamin D supplement use were noted from 1999–2014, 60-fold for ≥1000 IU/day (0.3%–18.2%) and 30-fold for ≥4000 IU/day (<0.1%–3.2%) [45]. Between NHANES 2007–2008 and August 2021–August 2023 there was essentially no change in the prevalence of 25OHD <30, 30–49, 50–125, and >125 nmol/L for persons not consuming vitamin D-containing supplements. However, the prevalence of 25OHD >125 nmol/L increased from 2.0% to 11.4% and from 4.7% to 30.6% for those reporting 1 and ≥2 vitamin D supplements, respectively [6]. Our findings support this trend, showing large increases in the prevalence of 25OHD >125 nmol/L in adult supplement users, females, and NHW and NHB persons. Both the Institute of Medicine and the European Food Safety Agency have cautioned about exceeding the upper limit for vitamin D intake of 4000 IU, as it may pose health risks [46]. Continued monitoring of vitamin D status in NHANES is therefore important. Despite the rightward shift of the 25OHD distribution, the prevalence of deficiency remained fairly constant over time (5%–6%), because few individuals were near this lower cutoff.
Iron deficiency is the most common form of nutritional deficiency, with children and WRA being most vulnerable. Although the prevalence of low FER in WRA increased from ∼13% to ∼18% between 2003 and 2023, the prevalence of elevated TFR (∼9%) and low BII (∼11%) did not change. Iron deficiency represents a spectrum, starting with iron depletion where the amount of stored iron (measured by FER) is reduced, comparable with an insufficiency; it progresses to iron-deficient erythropoiesis where transport iron (measured by transferrin saturation) is diminished and a functional iron deficiency (measured by TFR) develops; it ends with iron-deficiency anemia where iron-containing functional components, including hemoglobin, are underproduced [47]. It was therefore not surprising that the prevalence of elevated TFR was only about half of that of low FER, both in WRA and in children. BII, a de facto ratio of FER to TFR, assesses the full range of iron status from positive balance with residual storage iron to negative balance representing functional iron deficiency. When compared with NHANES 1988–1994 data on iron deficiency in nonpregnant WRA using the ferritin model (2 biomarkers out of 3 with abnormal values for FER, transferrin saturation, and/or erythrocyte protoporphyrin) (9%–11%) [27], the prevalence of low BII in WRA between 2003 and 2023 (8%–11%) was comparable, suggesting no major changes in over 30 y.
Using pooled, multicountry BRINDA data to assess the impact of inflammation adjustment for FER (using CRP and α1-acid glycoprotein) on iron depletion prevalence, higher prevalences were reported in WRA and preschool children (7.5 and 25 percentage points, respectively) [48]; limiting the dataset to low- and medium-infection–burden countries including the United States and using CRP only, resulted in smaller increases in prevalence: 6.9 and 3.6 percentage points, respectively. We found an even smaller change (∼1 percentage point). A similarly small increase in prevalence was found in NHANES 2003–2006 after excluding persons with elevated CRP: 0.5 and 1.6 percentage points, respectively [32]. Given this minimal difference, discussions on the need for this adjustment in the US context may be warranted. Although our decision to use only CRP in the inflammation adjustment of FER was out of necessity (α1-acid glycoprotein data are only available starting with NHANES 2015–2016), recent findings showed that adjusting FER for 2 instead of 1 inflammation marker resulted in more overadjustment [49].
Our analysis included 5 biomarkers of 1-carbon metabolism to characterize folate and vitamin B-6 and B-12 status. The prevalence of folate insufficiency, representing the risk of neural tube defect-affected pregnancies among WRA at the population level increased from ∼15% to ∼21% between 1999 and 2023. Although this represented a linear trend, closer inspection showed cycle-to-cycle fluctuations (13.5% in 2001–2002, 26.6% in 2009–2010). About 2/3 of WRA in NHANES have ≥1 risk factor for birth defects (e.g., obesity, low RBF, smoking exposure, low food security, or diabetes) and the percentage increased from 65.3% (2007–2008) to 69.5% (2015–March 2020) [50]. This emphasizes the importance of continued NHANES monitoring to provide reliable data to policy makers on biomarker changes as well as on a broad range of risk factors affecting health of the US population and subpopulations.
Vitamins B-6 and B-12 have been identified as neglected micronutrients [51] and only few national nutrition surveys beyond NHANES have included these nutrients [52]. Demographic patterns of low PLP concentrations, such as higher prevalence with increasing age, in females compared with males, and in NHB compared with NHW and MA persons, aligned with demographic patterns of CRP (Supplementary Table 16) and with reports that inflammation may cause a functional vitamin B-6 deficiency [53]. Although there is uncertainty on the best cutoff values to assess vitamin B-6 and B-12 status [51,53], making it difficult to reach a conclusion on the burden of these nutrient inadequacies, there is agreement on using ≥2 biomarkers for vitamin B-12 status assessment when possible [54]. The prevalence of vitamin B-12 insufficiency (typically ∼10% for B-12 <221 pmol/L or MMA >271 nmol/L) showed race and Hispanic origin differences (higher in NHW compared with NHB and MA persons) that don’t seem to be explained by intake. According to NHANES 2013–2014 data from What We Eat in America, NHW persons had a higher daily mean vitamin B-12 dietary (5.07 μg from foods/beverages) and total (90.2 μg from foods/beverages and supplements) intake compared with both NHB (dietary: 4.76 μg; total: 36.2 μg) and Hispanic (dietary: 4.83 μg; total: 22.6 μg) persons [55]. Others have suggested that genetic rather than environmental factors are the main cause for this difference [56,57].
Our analysis has several strengths. We provide detailed information for 8 nutrients and 25 biomarker-cutoff combinations measured at some points over the course of 24 y of NHANES for the overall US population, 9 demographic subgroups, and separately for supplement users and nonusers. The use of heat maps allowed us to visualize this large volume of data into 3 key figures that provide a semiquantitative assessment of nutritional imbalances over time and across population subgroups. For readers interested in the underlying data, we also provide detailed supplementary tables with the corresponding information. To allow for accurate interpretation of trends we adjusted for demographic changes over time, and we excluded pregnant and lactating women to avoid an undue influence of metabolic changes at the tails of the biomarker distributions. The NHANES data are nationally representative and thus offer unique insights on the nutritional status of the US population that can be used as a reference for other populations. The statistical analysis is parsimonious and the initial omnibus test for overall change acted as a gatekeeper to control the false positive rate. We also limited the interpretation of the linear trend test to situations with >3 data points. Lastly, we employed high-quality laboratory methods to generate biomarker data.
Our analysis has some limitations. Although we performed data adjustments where necessary to account for method changes over time, these adjustments unlikely account for all differences in methods over time and this might be a particular issue when changes are small. Knowing the interest in prevalence data for children and adolescents, we included estimates for these groups, however, data are limited. To stay within the scope of our analysis, we restricted adjustments to demographic changes over time and to 2 analyte-specific factors that could confound the interpretation (inflammation for FER and season for 25OHD). However, other factors such as changes in dietary intakes or unmeasured confounders may impact trends over time. The decline in NHANES response rates over time has been documented [35]. Nonresponse bias assessments for 2017–2018 and August 2021–August 2023 showed that despite these declines and changes in sampling methods for August 2021–August 2023, no major sources of bias were found that were not addressed with weighting adjustments [35,58].
This article provides comprehensive information on the prevalence of imbalances for 12 biomarkers of water- and fat-soluble vitamins and iron status assessed in the US population during NHANES 1999–August 2023. Although there are numerous gaps in nutritional biomarker data across NHANES cycles, the rich history of this US survey allowed us to provide detailed information for 8 essential nutrients: vitamins A, B-6, B-9 (folate), B-12, C, D, and E, and iron. Continued monitoring of the nutritional status of Americans through NHANES is critical to provide objective, actionable data for public health assessments, which is essential for policy decisions and nutrition recommendations, and ultimately for health improvements.
Author contributions
The authors’ responsibilities were as follows – CMP, MRS: designed the research project; CMP, MRS, CDP: conducted the research; MRS: analyzed the data; CMP: wrote the paper. CMP: responsible for final content; and all authors read and approved the final manuscript.
Data availability
The NHANES data are publicly available at https://wwwn.cdc.gov/nchs/nhanes/default.aspx.
Declaration of generative AI and AI-assisted technologies in the writing process
No Generative AI and AI-assisted technologies were used in the writing process.
Funding
The authors reported no funding received for this study.
Conflict of interest
The authors report no conflicts of interest.
Acknowledgments
We acknowledge contributions from the following laboratory members: Madhulika Chaudhary-Webb, Huiping Chen, Cori Clifton, Christina Fischer, Kelly Hong, Gwendolyn Gabey, Shameem Jabbar, Donna L. LaVoie, Khin Maw, Shahzad Momin, Phuong Ngac, Neelima Paladugula, Jenny Pao, Sweta Patel, Elizabeth Pendergrast, Rosemary L. Schleicher, David Scully, Patrick Simon, Myat Win, Mary Xu.
Footnotes
The findings and conclusions in this report are those of the authors and do not necessarily represent the official views or positions of the Centers for Disease Control and Prevention, the Food and Drug Administration, NIH, or the Department of Health and Human Services
Supplementary data to this article can be found online at https://doi.org/10.1016/j.cdnut.2026.107689.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
References
- 1.U.S. Centers for Disease Control and Prevention, National Center for Health Statistics, About NHANES [cited 17 November, 2025], https://www.cdc.gov/nchs/nhanes/about/index.html.
- 2.U.S. Centers for Disease Control and Prevention, National Center for Environmental Health, Second National Report on Biochemical Indicators of Diet and Nutrition in the U.S. Population 2012 [cited 17 November, 2025], https://www.cdc.gov/nutrition-report/php/resources-2012/index.html.
- 3.U.S. Centers for Disease Control and Prevention . 2026. National Center for Environmental Health, National Report on Biochemical Indicators of Diet and Nutrition in the U.S. Population.http://www.cdc.gov/nutritionreport [cited 17 November, 2025] [Google Scholar]
- 4.Pfeiffer C.M., Sternberg M.R., Zhang M., Fazili Z., Storandt R.J., Crider K.S., et al. Folate status in the US population 20 years after the introduction of folic acid fortification. Am. J. Clin. Nutr. 2019;110(5):1088–1097. doi: 10.1093/ajcn/nqz184. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Powers C.D., Sternberg M.R., Patel S.B., Pfeiffer C.M., Storandt R.J., Schleicher R.L. Serum vitamin C status in US adults: NHANES 2017-2018. J. Appl. Lab. Med. 2023;8(2):272–284. doi: 10.1093/jalm/jfac093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Couch C.A., Williams A.M., Stierman B., Storandt R., Leachman J., Mishara S., et al. Trends in vitamin D status in the United States, 2007–2023: a cross-sectional analysis of data from the National Health and Nutrition Examination Survey. Am. J. Clin. Nutr. 2026;123 doi: 10.1016/j.ajcnut.2025.09.046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Jefferds M.E.D., Mei Z., Addo Y., Hamner H.C., Perrine C.G., Flores-Ayala R., et al. Iron deficiency in the United States: limitations and guidelines, data, and monitoring of disparities. Am. J. Pub. Health. 2022;112(S8):S826–835. doi: 10.2105/AJPH.2022.306998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.U.S. Centers for Disease Control and Prevention, National Center for Health Statistics, NHANES response rates and population totals [cited 17 November, 2025], https://wwwn.cdc.gov/nchs/nhanes/responserates.aspx.
- 9.U.S. Centers for Disease Control and Prevention, National Center for Health Statistics, Ethics Review Board Approval [cited 9 January, 2026], https://www.cdc.gov/nchs/nhanes/about/erb.html.
- 10.Caudill S.P., Schleicher R.L., Pirkle J.L. Multi-rule quality control for the age-related eye disease study. Stat. Med. 2008;27:4094–4106. doi: 10.1002/sim.3222. [DOI] [PubMed] [Google Scholar]
- 11.U.S. Centers for Disease Control and Prevention, National Center for Health Statistics, NHANES questionnaires, datasets, and related documentation [cited 17 November, 2025], https://wwwn.cdc.gov/nchs/nhanes/default.aspx.
- 12.Cook J.D., Flowers C.H., Skikne B.S. The quantitative assessment of body iron. Blood. 2003;101:3359–3364. doi: 10.1182/blood-2002-10-3071. [DOI] [PubMed] [Google Scholar]
- 13.Flowers C.H., Skikne B.S., Covell A.M., Cook J.D. The clinical measurement of serum transferrin receptor. J. Lab. Clin. Med. 1989;114:368–377. [PubMed] [Google Scholar]
- 14.Pfeiffer C.M., Cook J.D., Mei Z., Cogswell M.E., Looker A.C., Lacher D.A. Evaluation of an automated soluble transferrin receptor (sTfR) assay on the Roche Hitachi analyzer and its comparison to two ELISA assays. Clin. Chim. Acta. 2007;382:112–116. doi: 10.1016/j.cca.2012.05.022. [DOI] [PubMed] [Google Scholar]
- 15.Pfeiffer C.M., Sternberg M.R., Hamner H.C., Crider K.S., Lacher D.A., Rogers L.M., et al. Applying inappropriate cutpoints leads to misinterpretation of folate status in the US population. Am. J. Clin. Nutr. 2016;104:1607–1615. doi: 10.3945/ajcn.116.138529. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Institute of Medicine, Food and Nutrition Board . National Academy Press; Washington, D.C.: 1998. Dietary reference intakes: thiamin, riboflavin, niacin, vitamin B6, folate, vitamin B12, pantothenic acid, biotin, and choline. [PubMed] [Google Scholar]
- 17.da Silva V.R., Russel K.A., Gregory III J.F. In: Present Knowledge in Nutrition. Erdman Jr. J.W., Macdonald I.A., Zeisel S.H., editors. International Life Sciences Institute (ILSI); 2012. Vitamin B6; pp. 225–237. [DOI] [Google Scholar]
- 18.Carmel R. Biomarkers of cobalamin (vitamin B-12) status in the epidemiologic setting: a critical overview of context, applications, and performance characteristics of cobalamin, methylmalonic acid, and holotranscobalamin II. Am. J. Clin. Nutr. 2011;94:348S–358S. doi: 10.3945/ajcn.111.013441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Allen R.H., Stabler S.P., Savage D.G., Lindenbaum J. Diagnosis of cobalamin deficiency I: Usefulness of serum methylmalonic acid and total homocysteine concentrations. Am. J. Hematol. 1990;34:90–98. doi: 10.1002/ajh.2830340204. [DOI] [PubMed] [Google Scholar]
- 20.Committee members NACB LMPG, Myers G.L., Christenson R.H., Cushman M., Ballantyne C.M., Cooper G.R., et al. National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines: emerging biomarkers for primary prevention of cardiovascular disease. Clin. Chem. 2009;55:378–384. doi: 10.1373/clinchem.2008.115899. [DOI] [PubMed] [Google Scholar]
- 21.Gibson R.S. 2nd edn. Oxford University Press; New York: 2005. Principles of nutritional assessment. [Google Scholar]
- 22.Institute of Medicine, Food and Nutrition Board . National Academy Press; Washington, D.C.: 2001. Dietary reference intakes: vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, manganese, molybdenum, nickel, silicon, vanadium, and zinc. [PubMed] [Google Scholar]
- 23.World Health Organization . World Health Organization; Geneva: 2011. Serum retinol concentrations for determining the prevalence of vitamin A deficiency in populations. Vitamin and mineral nutrition information system.https://www.who.int/publications/i/item/WHO-NMH-NHD-MNM-11.3 WHO/NMH/NHD/MNM.11.3) [cited 17 November, 2025] [Google Scholar]
- 24.Bendich A., Langseth L. Safety of vitamin A. Am. J. Clin. Nutr. 1989;49:358–371. doi: 10.1093/ajcn/49.2.358. [DOI] [PubMed] [Google Scholar]
- 25.Johnson L.E. In: Vitamin deficiency, dependency, and toxicity. Merck Manual of Diagnosis and Therapy. Beers MH, editor. Merck & Co., Inc.; Rahway, NJ: 2025. Vitamin E Deficiency.http://www.merck.com/mmpe/sec01/ch004/ch004l.html [cited 17 November, 2025] [Google Scholar]
- 26.Institute of Medicine, Food and Nutrition Board . National Academies Press; Washington, D.C.: 2011. Dietary reference intakes for calcium and vitamin D. [PubMed] [Google Scholar]
- 27.U.S. Centers for Disease Control and Prevention Recommendations to prevent and control iron deficiency in the United States, Morb. Mort. Wkly. Rep. 1998;47(RR-3):1–29. https://www.cdc.gov/mmwr/preview/mmwrhtml/00051880.htm [cited 17 November, 2025] [PubMed] [Google Scholar]
- 28.World Health Organization . WHO Guideline on Ferritin Concentrations to Assess Iron Status in Individuals and Populations; 2020. https://iris.who.int/bitstream/handle/10665/331505/9789240000124-eng.pdf?sequence=1&isAllowed=y [cited 17 November, 2025] [PubMed] [Google Scholar]
- 29.Mei Z., Pfeiffer C.M., Looker A.C., Flores-Ayala R.C., Lacher D.A., Mirel L.B., et al. Serum soluble transferrin receptor concentrations in US preschool children and non-pregnant women of childbearing age from the National Health and Nutrition Examination Survey 2003-2010. Clin. Chim. Acta. 2012;413:1479–1484. doi: 10.1016/j.cca.2012.05.022. [DOI] [PubMed] [Google Scholar]
- 30.Johnson C.L., Paulose-Ram R., Ogden C.L., Carroll M.D., Kruszon-Moran D., Dohrmann S.M., et al. National Health and Nutrition Examination Survey: Analytic guidelines, 1999-2010, National Center for Health Statistics, Vital. Health. Stat. 2013;2(161) https://wwwn.cdc.gov/nchs/data/nhanes/analyticguidelines/99-10-analytic-guidelines.pdf [cited 11 January, 2026] [PubMed] [Google Scholar]
- 31.Paulose-Ram R., Graber J.E., Woodwell D., Ahluwalia N. The National Health and Nutrition Examination Survey (NHANES), 2021–2022: adapting data collection in a COVID-19 environment. Am. J. Public. Health. 2021;111(12):2149–2156. doi: 10.2105/AJPH.2021.306517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Namaste S.M.L., Rohner F., Huang J., Bhushan N.L., Flores-Ayala R., Kupka R., et al. Adjusting ferritin concentrations for inflammation: Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA) project. Am. J. Clin. Nutr. 2017;106(Suppl):359S–371S. doi: 10.3945/ajcn.116.141762. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Luo H., Geng J., Zeiler M., Nieckula E., Sandalinas F., Williams A., et al. A practical guide to adjust micronutrient biomarkers for inflammation using the BRINDA method. J. Nutr. 2023;153:1265–1272. doi: 10.1016/j.tjnut.2023.02.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.R package version 0.1.5 [cited 17 November, 2025], https://www.brinda-nutrition.org/r-package.
- 35.U.S. Centers for Disease Control and Prevention, National Center for Health Statistics, Brief overview of sample design, nonresponse bias assessment, and analytic guidelines for NHANES August 2021–August 2023 [cited 17 November, 2025], https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/overviewbrief.aspx?Cycle=2021-2023.
- 36.Clopper C.J., Pearson E.S. The use of confidence or fiducial limits illustrated in the case of the binomial distribution. Biometrika. 1934;26(4):404–413. doi: 10.2307/2331986. [DOI] [Google Scholar]
- 37.Korn E.L., Graubard B.I. Confidence intervals for proportions with small expected number of positive counts estimated from survey data. Surv. Methodol. 1998;4(2):193–201. [Google Scholar]
- 38.Parker J.D., Talih M., Malec D.J., Beresovsky V., Carroll M., Gonzalez J.F., et al. National Center for Health Statistics Data Presentation Standards for Proportions, National Center for Health Statistics, Vital. Health. Stat. 2017;2(175) https://www.cdc.gov/nchs/data/series/sr_02/sr02_175.pdf [cited 17 November, 2025] [PubMed] [Google Scholar]
- 39.Cashman K.D., Kiely M. Contributions of nutrition science to the vitamin D field – clarity or confusion? J. Steroid. Biochem. Mol. Biol. 2019;187:34–41. doi: 10.1016/j.jsbmb.2018.10.020. [DOI] [PubMed] [Google Scholar]
- 40.Shahangian S., Alspach T.D., Astles J.R., Yesupriya A., Dettwyler W.K. Trends in laboratory tests volumes for Medicare Part B reimbursements 2000–2010. Arch. Pathol. Lab. Med. 2014;138:189–203. doi: 10.5858/arpa.2013-0149-oa. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Holick M.F. Vitamin D deficiency. N. Engl. J. Med. 2007;357:266–281. doi: 10.1056/nejmra070553. [DOI] [PubMed] [Google Scholar]
- 42.Bouillon R., Manousaki D., Rosen C., Trajanoska K., Rivadeneira F., Richards J.B. The health effects of vitamin D supplementation: evidence from human studies. Nature. Rev. Endocrinol. 2022;18:96–110. doi: 10.1038/s41574-021-00593-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Hahn J., Cook N.R., Alexander E.K., Friedman S., Walter J., Bubes V., et al. Vitamin D and marine omega 3 fatty acid supplementation and incident autoimmune disease: VITAL randomized controlled trial. Br Med J. 2022;376 doi: 10.1136/bmj-2021-066452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Costenbader K.H., Cook N.R., Lee I.M., Hahn J., Walter J., Bubes V, et al. Vitamin D and marine n-3 fatty acids for autoimmune disease prevention: outcomes two years after completion of a double-blind, placebo-controlled trial. Arthritis Rheumatol. 2024;76(6):973–983. doi: 10.1002/art.42811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Rooney M.R., Harnack L., Michos E.D., Ogilvie R.P., Sempos C.T., Lutsey P.L. Trends in use of high-dose vitamin D supplements exceeding 1000 or 4000 international units daily, 1999-2014. J. Am. Med. Assoc. 2017;317:2448–2450. doi: 10.1001/jama.2017.4392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Bouillon R. Safety of high-dose vitamin D supplementation. J. Clin. Endocrinol. Metab. 2020;105:1290–1291. doi: 10.1210/clinem/dgz282. [DOI] [PubMed] [Google Scholar]
- 47.Lynch S., Pfeiffer C.M., Georgieff M., Brittenham G., Fairweather-Tait S., Hurrell R.F., et al. Biomarkers of Nutrition for Development (BOND) - Iron review. J. Nutr. 2018;148:1001S–1067S. doi: 10.1093/jn/nxx036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Suchdev P.S., Williams A.M., Mei Z., Flores-Ayala R., Pasricha S.R., Rogers L.M., et al. Assessment of iron status in settings of inflammation: challenges and potential approaches. Am. J. Clin. Nutr. 2017;106(Suppl):1626S–1633S. doi: 10.3945/ajcn.117.155937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Ko Y.A., Suchdev P.S., Geng J., Luo H., Young M.F., Williams A.W. Evaluation of inflammation adjustment methods to assess iron deficiency using longitudinal data from norovirus human challenge trials, PLOS. Glob. Public. Health. 2024;4(12) doi: 10.1371/journal.pgph.0003964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Wang A., Zauche L.H., Crider K.S., Mai C.T., Qi Y.P., Yeung L.F., et al. Trends and prevalence of modifiable risk factors for birth defects among U.S. women of reproductive age: National Health and Nutrition Examination Survey (NHANES) 2007 to March 2020. Am. J. Prev. Med. 2025;69(6) doi: 10.1016/j.amepre.2025.107947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Brown K.H., Hess S.Y., Moore S.E., Combs G.F., Cashman K., McNulty H., et al. Neglected micronutrients – considering a broader set of vitamins and minerals in public health nutrition programs worldwide: a narrative review. Am. J. Clin. Nutr. 2025;122:680–694. doi: 10.1016/j.ajcnut.2025.06.030. [DOI] [PubMed] [Google Scholar]
- 52.World Health Organization, Vitamin and mineral nutrition information system [cited 17 November, 2025], https://www.who.int/teams/nutrition-and-food-safety/databases/vitamin-and-mineral-nutrition-information-system..
- 53.Sakakeeny L., Roubenoff R., Obin M., Fontes J.D., Benjamin E.J., Bujanover Y., et al. Plasma pyridoxal-5-phosphate is inversely associated with systemic markers of inflammation in a population of U.S. adults. J. Nutr. 2012;142:1280–1285. doi: 10.3945/jn.111.153056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Allen L.H., Miller J.W., de Groot L., Rosenberg I.H., Smith A.D., Refsum H., et al. Biomarkers of Nutrition for Development (BOND): Vitamin B-12 review. J. Nutr. 2018;148:1995S–2027S. doi: 10.1093/jn/nxy201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.U.S. Department of Agriculture, Agricultural Research Service, What we eat in America [cited 17 November, 2025], https://www.ars.usda.gov/northeast-group/docs/wweianhanes-overview/.
- 56.Saxena S., Carmel R. Racial differences in vitamin B12 levels in the United States. Am. J. Clin. Pathol. 1987;88(1):95–97. doi: 10.1093/ajcp/88.1.95. [DOI] [PubMed] [Google Scholar]
- 57.O’Logbon J., Crook M., Steed D., Harrington D.J., Sobczynska-Malefora A. Ethnicity influences total serum vitamin B12 concentrations: a study of Black, Asian, and White patients in a primary care setting. J. Clin. Pathol. 2022;75:598–604. doi: 10.1136/jclinpath-2021-207519. [DOI] [PubMed] [Google Scholar]
- 58.Fakhouri T.H.I., Martin C.B., Chen T.C., Akinbami L.J., Ogden C.L., Paulose-Ram R., et al. An investigation of nonresponse bias and survey location variability in the 2017–2018 National Health and Nutrition Examination Survey. National Center for Health Statistics. Vital Health Stat. 2020;2(185):1–36. https://www.cdc.gov/nchs/data/series/sr_02/s02-185-508.pdf [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 NHANES data are publicly available at https://wwwn.cdc.gov/nchs/nhanes/default.aspx.




