糖心vlog

Key Points

Question听 How has dietary intake changed over the past 2 decades among pregnant and nonpregnant women in the US?

Findings听 In this cross-sectional analysis of 1999-2018 data from the National Health and Nutrition Examination Survey, pregnant and nonpregnant women of reproductive age reported consuming less carbohydrates, vitamin A, vitamin C, and iron. Intake of calcium, vitamin K, and magnesium increased over time.

Meaning听 This study suggests that decreases in vitamin A, vitamin C, and iron intake can compromise nutritional adequacy among women and could, in turn, affect maternal and fetal outcomes.

Importance听 Nutritional status before and during pregnancy is important for maternal health and fetal growth and development.

Objective听 To describe secular trends in nutrient intake from foods, beverages, and supplements among pregnant and nonpregnant women of reproductive age in the US.

Design, Setting, and Participants听 This was a secondary series of cross-sectional analyses of the 1999-2018 National Health and Nutrition Examination Survey (NHANES). Pregnant (n鈥�=鈥�1392) and nonpregnant (n鈥�=鈥�9737) women aged 20 to 44 years who provided at least 1 reliable dietary recall were included for analysis. These analyses were performed between February 2022 and July 2024.

Main Outcomes and Measures听 The primary outcomes included the mean usual intake of macronutrients and micronutrients, as well as the prevalence of inadequate intake of micronutrients.

Results听 This representative sample included 1392 pregnant women (mean [SE] age, 28.5 [0.3] years) and 9737 nonpregnant women (mean [SE] age, 32.2 [0.1] years). Among pregnant women, a weighted mean (SE) of 27.0% (1.8%) of women were in their first trimester, and 33.8% (2.2%) were in their second trimester. Mean (SE) carbohydrate intake decreased between 1999-2000 and 2013-2018 among pregnant women (306.9 [7.6] to 274.9 [5.7] g/d; 尾鈥�=鈥夆垝2.1 [0.4]; P鈥�<鈥�.001) and between 1999-2000 and 2017-2018 among nonpregnant women (251.9 [4.9] to 216.9 [3.3] g/d; 尾鈥�=鈥夆垝1.9 [0.4]; P鈥�=鈥�.002). Between 1999-2000 and 2013-2018, the proportion of pregnant women who consumed below the Estimated Average Requirement of vitamin A increased by 10.9 percentage points (pp) (95% CI, 5.2-16.7 pp), and the proportion of pregnant women who consumed below the Estimated Average Requirement of vitamin C increased by 8.9 pp (95% CI, 3.9-14.0 pp). Similarly, the proportion of nonpregnant women with inadequate intake of vitamin A, vitamin C, and iron increased by 19.9 pp (95% CI, 12.3-27.5 pp), 11.1 pp (95% CI, 4.5-17.7 pp), and 4.9 pp (95% CI, 1.7-8.2 pp), respectively, between 1999-2000 and 2017-2018. The mean (SE) calcium intake increased from 1120.6 (41.4) to 1308.7 (49.0) mg/d for pregnant women (尾鈥�=鈥�11.7 [4.3]; P鈥�=鈥�.03) and from 849.5 (19.8) to 981.2 (27.9) mg/d for nonpregnant women (尾鈥�=鈥�6.7 [2.6]; P鈥�=鈥�.03; 尾²鈥�=鈥夆垝1.3 [0.2]; P鈥�&濒迟;鈥�.001). Among pregnant women, the prevalence of inadequate intake decreased by 16.1 pp (95% CI, 8.3-23.9 pp) for magnesium (P鈥�<鈥�.001) and 33.2 pp (95% CI, 24.0-42.4 pp) for vitamin K (P鈥�<鈥�.001); among nonpregnant women, the proportion with inadequate intake decreased by 16.1 pp (95% CI, 10.4-21.7 pp) for calcium (P鈥�<鈥�.001), 15.5 pp (95% CI, 7.3-23.6 pp) for magnesium (P鈥�<鈥�.001), and 33.3 pp (23.5-43.0 pp) for vitamin K (P鈥�&濒迟;鈥�.001).

Conclusions and Relevance听 This cross-sectional study of pregnant and nonpregnant women of reproductive age found that vitamin A, vitamin C, and iron intake decreased over the past 2 decades, which may have substantial maternal and fetal health implications. By identifying these nutrient gaps and trends in inadequate intake in this at-risk population, scientific, health care, and regulatory communities may be better poised to adopt recommendations to improve nutrient intake.

Ensuring adequate nutrition both before and during pregnancy is important for supporting maternal health, fostering fetal growth and development, and reducing the risk of chronic diseases in children later in life.^1-4 Energy requirements during pregnancy increase in the second and third trimesters due to uterine and fetal metabolic contributions and increased workload of the heart and lungs.⁵ During pregnancy, there is also an increased demand for multiple nutrients. These include protein to support maternal and fetal tissue development, iron for red blood cell production, folate to reduce the risk for neural tube defects, calcium to support fetal bone formation, and iodine to support increased thyroxine production for the fetus. Although calcium needs do not increase during pregnancy, adequate intake is imperative to support fetal bone formation.

Dietary patterns in the US have changed over recent decades, with overall dietary quality decreasing among adults older than 20 years of age,⁶ which is associated with the prevalence of inadequate nutrient intake and the inability to meet nutritional needs through diet alone.^7,8 According to a previous analysis of National Health and Nutrition Examination Survey (NHANES) data, adequate nutrient intake appears to be a concern in pregnant US women, with more than 40% of pregnant women not meeting the Estimated Average Requirement (EAR) for vitamin D, vitamin E, and magnesium, and most pregnant women also struggling to consume above the Adequate Intake (AI) for choline, vitamin K, and potassium, despite reporting greater supplementation frequency compared with nonpregnant counterparts.⁹ Although there is no established dietary requirement for preformed docosahexaenoic acid (DHA) intake, experts recommend that pregnant women consume at least 250 mg/d of DHA and eicosapentaenoic acid (EPA), with 200 mg/d from DHA.^10-13 However, 95% of women reported daily intakes below these thresholds.^13,14 Nonpregnant women of reproductive age are also at risk for inadequate intake of vitamins A, C, D, and E; calcium; and magnesium and have suboptimal vitamin K, choline, potassium, and dietary fiber intake.¹⁵

The most recent NHANES estimates of dietary intake among pregnant and nonpregnant women of reproductive age are available only up to the 2015-2016 cycle, and no comprehensive trend analysis has been performed; such analysis could help researchers understand changes in health behaviors over time and anticipate threats to nutritional status, especially in vulnerable subpopulations. Conducting frequent analyses of these data are therefore important to better understand the current prevalence of inadequate micronutrient intake and its shortfalls and trends, especially in vulnerable populations, such as pregnant and nonpregnant women of reproductive age, with significant public health implications. These data are also imperative to help inform policy, educational messages, and interventions to support women鈥檚 health before, during, and after pregnancy. Thus, the objective of this study is to describe trends in dietary intake among pregnant and nonpregnant women aged 20 to 44 years who participated in the 1999-2018 NHANES.

The NHANES program, overseen by the National Center for Health Statistics, consistently gathers cross-sectional health data from a nationally representative sample of the noninstitutionalized US population. Twenty years (10 cycles) of NHANES data (1999-2000, 2001-2002, 2003-2004, 2005-2006, 2007-2008, 2009-2010, 2011-2012, 2013-2014, 2015-2016, and 2017-2018) were combined for this analysis. Details of the survey design, protocol, and data collection are available online.¹⁶ The survey uses a complex, multistage probability sampling design in which study personnel collect data from in-home interviews followed by a visit to a mobile examination center. The National Center for Health Statistics research ethics review board reviewed and approved the NHANES protocol in this cross-sectional study, and trained NHANES personnel collected written informed consent for all participants or approved proxies. This analysis followed the Strengthening the Reporting of Observational Studies in Epidemiology () reporting guideline for cross-sectional studies.¹⁷

Inclusion criteria for this analysis were women aged 20 to 44 years who participated in the 1999-2018 NHANES and provided at least 1 reliable dietary recall. Pregnancy status was determined from a positive urine pregnancy test result or self-reported pregnancy status. Demographics included age, self-reported race and ethnicity (Hispanic, non-Hispanic Black, non-Hispanic White, and other [any races and ethnicity, including multiracial, not included in Hispanic, non-Hispanic Black, and non-Hispanic White]), educational attainment, and family income to poverty ratio. Races and ethnicity were assessed to describe the sample. Trimester status for pregnant respondents was estimated from the question: 鈥淲hat month of pregnancy are you in?鈥�

We categorized smoking status based on self-reported cigarette use. Respondents were classified into the categories of current, former, and never smokers. Adult food security classification was based on responses using the US Food Security Survey Module,¹⁸ and responses were grouped into the categories of full food security, marginal food security, low food security, and very low food security. Participation in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) was determined based on whether the respondent reported receiving WIC benefits within the past 12 months. Similarly, participation in the Food Stamps or Supplemental Nutrition Assistance Program (SNAP) was based on whether the respondents indicated they received benefits in the past 12 months.

During the mobile examination center visit, a trained dietary interviewer administered a 24-hour dietary recall using the automated multiple pass method.^19,20 Beginning in 2003, the NHANES collected a second 24-hour dietary recall 3 to 11 days later over the telephone after the mobile examination center visit using the automated multiple pass method. Dietary supplement use was assessed via a questionnaire at the mobile examination center visit, in which participants identified their vitamin, mineral, herbal, and other supplement use over the past 30 days. For the 1999-2006 survey cycles, values for daily supplement intake during the past 30 days were estimated from the available product information and ingredient databases provided by the NHANES.^21,22 Nutrients included for analysis were the following macronutrients: carbohydrates, proteins, total fat, fatty acids, and dietary fiber. The micronutrients included for analyses were calcium, choline, copper, folate, iron, magnesium, niacin, phosphorus, potassium, riboflavin, selenium, sodium, thiamin, vitamin B₁₂, vitamin B₆, vitamin A, vitamin C, vitamin D, vitamin E, vitamin K, and zinc. Food and Nutrient Database for Dietary Studies nutrient values from subsequent databases were matched to food codes for nutrients not reported in earlier cycles (eg, vitamin D, choline, and vitamin K).

All analyses were performed between February 2022 and July 2024, using SAS software, version 9.4 (SAS Institute Inc). Descriptive statistics for participant characteristics were generated using the correct sampling weights and accounted for the NHANES survey design.²³ Participant characteristics were compared between pregnant and nonpregnant women using independent samples t tests for continuous variables and Rao-Scott 蠂² tests for categorical variables. No statistical comparisons for dietary intake were made between pregnant and nonpregnant women.

Usual nutrient intake was assessed using the National Cancer Institute (NCI) method, which reduces the effect of intraindividual variation to provide reliable estimates for usual nutrient intake and proportions of those consuming above or below the Dietary Reference Intakes.^24,25 Moreover, the NCI method allows for the adjustment of relevant covariates.

The Simulating Intake of Micronutrients for Policy Learning and Engagement (SIMPLE) and SIMPLE-Iron macros, which were developed based on the NCI method and connect multiple NCI macros, were used to estimate usual intake.²⁶ The proportion of the population of women who consumed below the EAR or above the AI were estimated using the cut-point approach for all nutrients except for iron for nonpregnant women. The full probability model was applied to iron intake for nonpregnant women, and a bioavailability adjustment of 18% was applied. All nutrients were treated as consumed daily except for iron, DHA, and EPA, which were estimated using the NCI 2-part model.²⁷ The covariates included in the NCI modeling approach were whether the recall was collected on the weekend (Friday-Sunday) or weekday, the sequence of the recall, dietary supplement use (yes or no), pregnancy status (yes or no), and survey cycle. Variance estimates were generated using balanced repeated half replicates (Fay adjustment factor鈥�=鈥�0.3).

The NHANES oversampled pregnant women during the 1999-2006 cycles and discontinued this from 2007 to 2018. As such, the resulting sample sizes are smaller per cycle. Therefore, we pooled survey cycles 2007 to 2012 and 2013 to 2018 for our analysis of pregnant women鈥檚 dietary intake to provide more reliable estimates.

Trends in nutrient intake from 1999 to 2018 were assessed using linear regression separately for pregnant and nonpregnant women. Survey cycle was the independent variable, and the dependent variables included the mean usual nutrient intake, the percentage of intake below the Acceptable Macronutrient Distribution Range (AMDR), the percentage of intake above the AMDR, the percentage of intake below the EAR, and the percentage of intake above the AI.²⁸ Linear and quadratic trends were assessed for unpooled estimates.²⁹ Changes in the proportion of women who reported micronutrient intake below the EAR or above the AI between 1999-2000 and 2013-2018 for pregnant women and between 1999-2000 and 2017-2018 for nonpregnant women were also calculated, and the corresponding 95% CIs were generated. Statistical significance was considered as P鈥�<鈥�.05, and all tests for significance were 2-sided.

This nationally representative sample (n鈥�=鈥�11鈥�129) of 1392 pregnant women (mean [SE] age, 28.5 [0.3] years) and 9737 nonpregnant women (mean [SE] age, 32.2 [0.1] years) was included in our analysis. Figure 1 illustrates how the final analytical samples were generated. Pregnant women were younger compared with nonpregnant women and had greater representation from Hispanic and non-Hispanic Black women (Table 1). Most women had at least some college education and income above 185% of the federal poverty level. Dietary supplement use was significantly greater among pregnant women compared with nonpregnant women (weighted mean [SE], 78.0% [1.7%] vs 46.4% [0.8%]; P鈥�&濒迟;鈥�.001). Dietary supplement use among pregnant women has been trending down slightly, while the proportion of nonpregnant women who report supplement use has been increasing over the past decade (eFigure in Supplement 1 and eTable 1 in Supplement 2).

Trends in usual macronutrient intake for pregnant and nonpregnant women are provided in Table 2. Dietary carbohydrate intake decreased from a mean (SE) of 55.2% (0.6%) of energy per day in 1999 and 2000 to 50.7% (0.5%) of energy per day from 2013 to 2018 for pregnant women (mean [SE] 尾鈥�=鈥夆垝0.2 [0.1]; P鈥�=鈥�.005) and from 52.2% (0.5%) of energy per day in 1999 and 2000 to 46.9% (0.5%) of energy per day in 2017 and 2018 for nonpregnant women (尾鈥�=鈥夆垝0.3 [0.1]; P鈥�=鈥�.007). Despite decreases in dietary carbohydrate intake, dietary fiber intake increased from 16.9 (0.7) g/d to 18.7 (0.5) g/d for pregnant women (尾鈥�=鈥�0.1 [0.03]; P鈥�=鈥�.009) and from 13.5 (0.5) g/d to 14.9 (0.4) g/d for nonpregnant women (尾鈥�=鈥�0.1 [0.03]; P鈥�=鈥�.004). Although dietary fat intake remained stable for pregnant women (尾鈥�=鈥�0.1 [0.2]; P鈥�=鈥�.48) and nonpregnant women (尾鈥�=鈥�0.2 [0.2]; P鈥�=鈥�.28) from 1999 to 2018, energy contribution from dietary fat increased exponentially from a mean (SE) of 32.1% (0.5%) of energy per day to 35.0% (0.4%) of energy per day for pregnant women (尾²鈥�=鈥�0.02 [0.01]; P鈥�=鈥�.04) and from 32.0% (0.5%) of energy per day to 36.5% (0.5%) of energy per day for nonpregnant women (尾²鈥�=鈥�0.03 [0.009]; P鈥�=鈥�.02). Moreover, the change in the proportion of women whose dietary fat intake was above the AMDR increased by 23.2 percentage points (pp) (尾²鈥�=鈥�0.2 [0.1]; P鈥�=鈥�.04) for pregnant women and 33.4 pp (尾²鈥�=鈥�0.2 [0.1]; P鈥�=鈥�.02) for nonpregnant women. eTable 2 in Supplement 2 provides the usual macronutrient intake estimates for all survey cycles included in this analysis.

Pregnant and nonpregnant women鈥檚 usual micronutrient intake from foods and supplements are provided in Table 3. Calcium intake increased from a mean (SE) of 1120.6 (41.4) mg/d to 1308.7 (49.0) mg/d for pregnant women (mean [SE] 尾鈥�=鈥�11.7 [4.3]; P鈥�=鈥�.03) from 1999 to 2018, yet this increase was attenuated over time (mean [SE] 尾²鈥�=鈥夆垝2.1 [0.6]; P鈥�=鈥�.01). Likewise, nonpregnant women鈥檚 mean (SE) calcium intake increased from 849.5 (19.8) mg/d to 981.2 (27.9) mg/d (尾鈥�=鈥�6.7 [2.6]; P鈥�=鈥�.03; 尾²鈥�=鈥夆垝1.3 [0.2]; P鈥�<鈥�.001) from 1999 to 2018. The mean (SE) dietary iron intake from foods and supplements decreased from 52.8 (4.6) mg/d in 1999 and 2000 to 29.5 (1.8) mg/d from 2013 to 2018 among pregnant women (尾鈥�=鈥夆垝0.9 [0.2]; P鈥�=鈥�.001) and from 19.9 (0.8) mg/d in 1999 and 2000 to 16.9 (0.8) mg/d in 2017 and 2018 for nonpregnant women (尾鈥�=鈥夆垝0.2 [0.02]; P鈥�&濒迟;鈥�.001). eTable 3 in Supplement 2 provides the usual micronutrient intake estimates for all survey cycles included in this analysis. Micronutrient intake from foods and beverages alone is provided in eTable 4 in Supplement 2.

Vitamin A intake decreased from a mean (SE) of 1233.6 (78.9) 渭g/d to 802.0 (29.6) 渭g/d among nonpregnant women (mean [SE] 尾鈥�=鈥夆垝12.4 [3.0]; P鈥�=鈥�.003), yet this decrease flattened in the latter half of the time frame (mean [SE] 尾²鈥�=鈥�1.4 [0.4]; P鈥�=鈥�.009) (Table 3). Pregnant women鈥檚 vitamin C intake decreased linearly from 1999 to 2018 (mean [SE] 尾鈥�=鈥夆垝4.6 [0.7]; P鈥�&濒迟;鈥�.001). Nonpregnant women鈥檚 mean (SE) vitamin C intake decreased from 184.6 (15.4) mg/d in 1999 and 2000 to 118.6 (5.9) mg/d in 2009 and 2010, followed by an increase to 156.5 (28.4) mg/d in 2017 and 2018 (mean [SE] 尾鈥�=鈥夆垝3.5 [1.0]; P鈥�=鈥�.007; mean [SE] 尾²鈥�=鈥�0.5 [0.1]; P鈥�=鈥�.005).

Although pregnant women reported a decrease in mean [SE] folate intake from 1518.4 (113.8) mg/d to 1373.3 (77.3) mg/d (mean [SE] 尾鈥�=鈥夆垝3.1 [6.3]; P鈥�=鈥�.64) and mean [SE] vitamin A intake from 1675.5 (126.0) 渭g/d to 1380.7 (76.9) 渭g/d (mean [SE] 尾鈥�=鈥夆垝5.6 [9.3]; P鈥�=鈥�.56), these trends were not statistically significant (Table 3). Sodium, potassium, and combined DHA and EPA intake for pregnant and nonpregnant women also remained consistent across the past 2 decades.

Figure 2 displays changes in the prevalence of inadequate micronutrient intake among pregnant women between 1999-2000 and 2013-2018 and nonpregnant women between 1999-2000 and 2017-2018. The largest increases in the prevalence of inadequate intake among both groups were for vitamin A and vitamin C, and the largest decreases were for magnesium and vitamin K. The prevalence of inadequate intake from 1999 to 2018 among pregnant women increased by 10.9 pp (95% CI, 5.2-16.7 pp) for vitamin A (P鈥�<鈥�.001) and 8.9 pp (95% CI, 3.9-14.0 pp) for vitamin C (P鈥�&濒迟;鈥�.001). In contrast, the prevalence of inadequate intake decreased by 9.2 pp (95% CI, 0.7-17.7 pp) for DHA and EPA (P鈥�=鈥�.03), 16.1 pp (95% CI, 8.3-23.9 pp) for magnesium (P鈥�<鈥�.001), and 33.2 pp (95% CI, 24.0-42.4 pp) for vitamin K (P鈥�<鈥�.001) among pregnant women. The prevalence of inadequate intake among nonpregnant women increased by 4.2 pp (95% CI, 0.8-7.6 pp) for copper (P鈥�=鈥�.02), 19.9 pp (95% CI, 12.3-27.5 pp) for vitamin A (P鈥�<鈥�.001), 11.1 pp (95% CI, 4.5-17.7 pp) for vitamin C (P鈥�=鈥�.001), and 4.9 pp (95% CI, 1.7-8.2 pp) for iron (P鈥�=鈥�.003). The prevalence of inadequate intake among nonpregnant women decreased by 16.1 pp (95% CI, 10.4-21.7 pp) for calcium (P鈥�<鈥�.001), 15.5 pp (95% CI, 7.3-23.6 pp) for magnesium (P鈥�<鈥�.001), 1.9 pp (95% CI, 0.6-3.4 pp) for niacin (P鈥�=鈥�.006), 6.8 pp (95% CI, 1.9-11.7 pp) for vitamin B₆ (P鈥�=鈥�.006), and 33.3 pp (95% CI, 23.5-43.0 pp) for vitamin K (P鈥�&濒迟;鈥�.001).

Our analysis of this nationally representative sample of US women shows that usual dietary intake among pregnant and nonpregnant women has changed considerably over the past 2 decades. Carbohydrate intake decreased for all women of reproductive age, while protein and fat intake were unchanged. Calcium, magnesium, and vitamin K intake increased, while iron and vitamin C intake decreased for all women.

Despite reductions in carbohydrate intake, women did not appear to change their intake of fat or protein between 1999 and 2018. Although women are not substituting carbohydrate intake with fat or protein, the energy contribution from total fat increased considerably. Consequently, the proportion of pregnant and nonpregnant women whose dietary fat intake exceeded the AMDR more than doubled. Although it is recommended to restrict total fat intake to less than 35% of energy, large prospective cohort studies have found no association between total fat intake and gestational diabetes^30,31 or hypertensive disorders.³²

The decrease in carbohydrate intake reported in this analysis is consistent with other reports that dietary carbohydrate intake has been decreasing in North America since 1999, which appears to be associated largely with decreases in intake of added sugars and flour and cereal products.³³ Although public policy guidelines largely favor limiting added sugar intake,³⁴ reducing intake from fortified flour and cereals is discouraged because this can lead to decreases in folate, iron, magnesium, vitamin A, and vitamin C intake if not replaced with other sources.^35,36 Our analysis indicated decreases in folate, vitamin C, and iron intake, but not magnesium and vitamin A, from foods and beverages alone (eTable 4 in Supplement 2).

The rates of spina bifida have decreased an estimated 28% after the mandatory folic acid fortification of enriched grains.³⁷ Similarly, women aged 12 to 49 years reported improvements in folate intake and red blood cell folate from 2007 to 2016.³⁸ In the present analysis, pregnant women鈥檚 total folate intake has remained consistent over the past 2 decades, while nonpregnant women have decreased their overall consumption of folate. This small reduction in intake over the past 2 decades appears to be associated with reductions in intake from foods and beverages (eTable 4 in Supplement 2).

The 2005 iteration of the US Dietary Guidelines for Americans included vitamin C as a nutrient of concern due to its low intake among adults.³⁹ Consumption of vitamin C鈥搑ich foods to enhance the absorption of nonheme iron is a specific recommendation for pregnant and nonpregnant women of reproductive age. Although this special consideration for all women of reproductive age remained in subsequent iterations of the guidelines,^40-42 vitamin C is no longer considered a nutrient of public concern despite continuing decreases in intake, especially among these subpopulations. The present analysis demonstrated that the risk for inadequate intake of vitamin C among pregnant women has increased 3-fold between 1999-2000 and 2017-2018 and by 40% among nonpregnant women. The decreases in intake appear to be largely associated with decreased consumption from foods and beverages alone (eTable 4 in Supplement 2) rather than from changes in supplement use habits. Although vitamin C deficiency and related pregnancy complications are rare, a diet low in vitamin C may negatively affect iron status.

Iron intake prior to and during pregnancy is important for fetal and maternal health outcomes. Modest increases in the prevalence of inadequate iron intake among pregnant and nonpregnant women from 1999 to 2018 were observed in this analysis. Intake of iron from foods and beverages decreased, which may be partially associated with decreased carbohydrate intake from fortified sources; however, this requires further inquiry. In addition, supplemental iron intake has decreased over the past 2 decades in these populations, which could be due to mild adverse effects associated with iron supplementation, including constipation, nausea, and gastric upset.⁴³ However small, the effect that inadequate intake has on iron status may be exacerbated in these populations by the concurrent decrease in vitamin C intake.

Calcium intake during pregnancy is important to support fetal skeletal growth; however, needs during pregnancy are not greater than during nonpregnancy due to increased maternal calcium absorption.⁴⁴ Our results showed that calcium intake increased for both pregnant and nonpregnant women of reproductive age, resulting in a decrease in the proportion at risk for inadequate intake in both groups. However, nearly two-thirds of nonpregnant women reported inadequate calcium intake.

Although there is no established consumption requirement for long-chain unsaturated fatty acids, an amount above 250 mg/d is generally recommended for all adults.^13,34 Several expert scientific organizations also recommend that pregnant and lactating women supplement with at least 200 mg/d of DHA alone to improve maternal and infant serum DHA levels.^10-12 This is especially important during the last trimester of pregnancy, as DHA is accreted in the fetal brain at a considerable rate, thus supporting neural, cognitive, and retinal development of the infant. Higher-dose prenatal DHA supplementation when compared with no omega-3 long-chain polyunsaturated fatty acid supplementation has also been found to reduce early preterm delivery among women with low serum DHA status.⁴⁵ In this study, we showed that the mean DHA and EPA intake remained relatively stable among both groups, while the proportion of pregnant women who consumed at least 250 mg of combined DHA and EPA increased significantly. Intake was largely associated with supplemental DHA and EPA rather than foods and beverages (eTable 4 in Supplement 2). Consistent with previous studies, these analyses showed that most women (>90%) do not consume the recommended 250 mg/d of combined DHA and EPA, which also suggests that most women are not meeting the recommendation for 200 mg/d of DHA alone.¹⁴

A strength of this study is that it is a comprehensive analysis of trends in macronutrient and micronutrient intake among a nationally representative sample of US pregnant women. However, this study is not without limitations. First, sample sizes within survey cycles for pregnant women were small, which could lead to unstable variance estimates. Dietary recalls are prone to random error, which can lead to large variances; to address this possibility, the NCI method was used to minimize the effect of random error on variance estimates. In addition, this study did not assess serum biomarkers, which would have allowed for estimates of deficiency. Furthermore, the present study uses NHANES data collected until 2018; however, since these analyses were completed, the 2019 to March 2020 prepandemic datasets were released. Due to the incomplete sample size, the Centers for Disease Control and Prevention combined the prepandemic cycle with the 2017-2018 cycle for release.

Adequate preconception nutrition and nutrient status is critical for optimizing maternal health and fetal development.⁴⁶ Suboptimal intake of certain nutrients, including folate, riboflavin, vitamin B₆, DHA, and inositol, during preconception and pregnancy have been associated with increased risk for pregnancy complications and infant health outcomes.⁴⁷ Such adverse outcomes are associated with inadequate nutrient intake and include neural tube defects,⁴⁸ impaired brain development,^49,50 small size for gestational age,⁵¹ and preterm delivery,⁵² which can have significant effects on public health as well as increased health care costs and burdens.

For these reasons, among others, it is necessary to understand how women鈥檚 dietary patterns change over time among those who are pregnant and who may become pregnant, to better improve the nutritional status for these populations. By identifying these nutrient gaps and trends in inadequate intake in this at-risk population, scientific, health care, and regulatory communities may be better poised to adopt recommendations to improve nutrient intake. This cross-sectional study demonstrates that nutrient intake among women of reproductive age has changed considerably over the past 2 decades, with corresponding changes in the risk of inadequate intake for specific nutrients.

Accepted for Publication: July 30, 2024.

Published: October 10, 2024. doi:10.1001/jamanetworkopen.2024.38460

Open Access: This is an open access article distributed under the terms of the CC-BY-NC-ND License. 漏 2024 Miketinas D et al. 糖心vlog Open.

Corresponding Author: Derek Miketinas, PhD, RD, Department of Nutrition and Food Sciences, Texas Woman鈥檚 University, 6700 Fannin St, Houston, TX 77030 (dmiketinas@twu.edu).

Author Contributions: Dr Miketinas had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Miketinas, Luo, Firth, Bailey, Brink.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Miketinas, Firth, Bailey, Bender, Gross, Brink.

Critical review of the manuscript for important intellectual content: Miketinas, Luo, Firth, Bailey, Gross, Brink.

Statistical analysis: Miketinas, Luo, Firth.

Obtained funding: Miketinas, Brink.

Administrative, technical, or material support: Miketinas, Luo, Bailey, Brink.

Supervision: Miketinas, Luo, Bender, Brink.

Conflict of Interest Disclosures: Dr Miketinas reported receiving grants from Mead Johnson outside the submitted work and was employed by Texas Woman's Universit. Dr Luo reported serving as a consultant for Mead Johnson during the conduct of the study and was employed by Emory University. Drs Firth, Bailey, Gross, and Brink and Ms Bender were employed at Reckitt/Mead Johnson. No other disclosures were reported.

Funding/Support: This research was funded by Reckitt/Mead Johnson.

Role of the Funder/Sponsor: Reckitt/Mead Johnson had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Authors who were employees of Reckitt/Mead Johnson participated in each of these activities.

Data Sharing Statement: See Supplement 3.