Abstract
OBJECTIVE
To examine differences in nutrient intake among U.S. adults with diagnosed diabetes, undiagnosed diabetes, prediabetes, and no diabetes, and how intake of nutrients has changed over time by diabetes status.
RESEARCH DESIGN AND METHODS
We conducted cross-sectional analyses of major macronutrient and micronutrient intakes using National Health and Nutrition Examination Survey (NHANES) 2011–2020 data. Energy-adjusted nutrient intakes across all year periods were compared across categories of diabetes status. Multiple linear regression models were used to examine the time trends in energy-adjusted nutrient intake by diabetes status after adjusting for age, sex, and race. All analyses were weighted to account for the complex survey design.
RESULTS
Compared with individuals without diabetes, individuals with diagnosed diabetes had greater average intakes of protein, total fat, cholesterol, calcium, and sodium, but lower total sugar, potassium, and magnesium intakes across all periods. Among individuals with diagnosed diabetes, carbohydrate, dietary fiber, magnesium, sodium, and potassium intake decreased from 2011 to 2020, whereas total fat and saturated fatty acid intakes increased. Adults with undiagnosed diabetes consumed less dietary fiber, more saturated fatty acids, and less magnesium than those without diabetes. The Healthy Eating Index was lower for people with prediabetes and undiagnosed and diagnosed diabetes than for those without diabetes (all P < 0.01).
CONCLUSIONS
In the 10-year study period, nutrient intakes did not improve in individuals with prediabetes, undiagnosed diabetes, or diagnosed diabetes. The lack of improvement in dietary quality, coupled with a high prevalence of diabetes in the U.S., warrants public health efforts to reverse these adverse trends.
Introduction
Diabetes is a prevalent health condition in the U.S., with more than 38.1 million adults affected, accounting for 14.7% of the population (1). Type 2 diabetes (T2D) and its associated macrovascular and microvascular complications can lead to higher mortality rate, a lower quality of life, and substantial health care burden (2,3). Lifestyle interventions including a healthy diet can improve glycemia and slow the progression of diabetes and related diabetes outcomes (4,5). The American Diabetes Association “Standards of Care in Diabetes” highlights the importance of consuming nutrient-dense, high-fiber, minimally processed foods and healthy fats as part of an overall healthy eating pattern (5).
Previous studies have shown that U.S. adults, regardless of diabetes status, generally fall short of meeting national dietary guidelines (6). Survey data from 1977 to 2018 suggested that dietary habits in the general population were gradually changing and this change varied by different population groups. Whereas higher-income individuals significantly improved their diet quality through an increased intake of fresh, home-prepared foods, younger and lower-income populations increasingly relied on fast food (7). Trends among individuals with self-reported diabetes have been similarly complex; research from 1988 to 2004 noted an increase in carbohydrate consumption, and subsequent data from 1999 to 2018 revealed a significant decline in the intake of vegetables, fruits, and refined grains, resulting in lower vitamin C and potassium levels (8,9). Despite these findings, it remains unclear whether dietary intakes among individuals with T2D have improved over time during the decade immediately preceding the pandemic. To address this gap, we examined trends in major macro- and micronutrient intakes among people with diagnosed diabetes, undiagnosed diabetes, and prediabetes, using data from a nationally representative sample.
Research Design and Methods
Methods
This study analyzed major macronutrient and micronutrient intakes by diabetes status using data from the National Health and Nutrition Examination Survey (NHANES) from the 2011 to March 2020 survey cycles (thus including data from before the COVID-19 pandemic) (10). NHANES is a national cross-sectional survey using a stratified, multistage, probability-cluster design to produce nationally representative estimates of the U.S. population. Data are collected from household interviews and include surveys of health history, dietary recall interview data, physical examination data, and laboratory test results (11). The U.S. Department of Agriculture’s (USDA) Five-Steps Automated Multiple Pass Method (AMPM) was used to collect 24-h recall information for each participant. Then, 64 nutrient components were estimated in relation to food intake by referring to the standards in the USDA Food and Nutrient Database for Dietary Studies. Venous blood samples were collected in the morning and transported to the central laboratory to measure fasting glucose (mg/dL) and insulin (μIU/mL) levels. All participants who met the 9-h fasting standard completed an oral glucose tolerance test (OGTT). A separate whole-blood sample was collected at any time during the day for hemoglobin A1c (HbA1c) measurement. Blood pressure was measured while the participants were in a seated position after 5 min of rest.
Study Population
During the 2011–March 2020 cycles, 23,790 adults who completed 24-h diet recall interviews participated in the examinations. We subsequently excluded 251 pregnant women and 835 participants with missing diabetes indicators. We also excluded 157 individuals with a higher likelihood of type 1 diabetes (defined as those who were diagnosed at <25 years of age.) In the final sample, 22,547 adults (representing 184,664,074 people in the population) were included in our analysis.
Variables
Based on target nutrients in guidelines, we included the following major macronutrients and micronutrients in our analysis: protein, carbohydrate, total sugar, dietary fiber, total fat, saturated fatty acids, monounsaturated fatty acids, polyunsaturated fatty acids, cholesterol, calcium, magnesium, zinc, sodium, and potassium. Data used in our study on the intake of these nutrients were collected during the first day of the 24-h dietary recall interview (12). According to the analysis guide of NHANES, a single 24-h recall is sufficient for analyzing mean nutrient intakes (13). To assess overall food quality, we also calculated the Healthy Eating Index (HEI)-2015 score by running the Dietaryindex R package (14), which aligns with the 2015–2020 Dietary Guidelines for Americans (DGA). The HEI measures diet quality on a 0–100 scale by evaluating how an individual’s food choice aligns with the DGA; a higher score indicates better adherence to federal nutrition recommendations. Compared with the HEI-2010, the HEI-2015 introduced important changes, including added sugar, and replacing solid fats with saturated fatty acids. To capture trends that reflect these changes, we applied the HEI-2015 consistently across all years in the analysis.
Diabetes status was categorized as diagnosed diabetes, undiagnosed diabetes, prediabetes, or no diabetes. Diagnosed diabetes was defined as any self-reported diagnosis of diabetes by people older than 25 years and/or who used diabetes medications. Undiagnosed diabetes was defined as negative responses to all the questions mentioned above and meeting any of the following criteria: a fasting plasma glucose level ≥126 mg/dL, an HbA1c level ≥6.5%, or a post-OGTT plasma glucose level ≥200 mg/dL. Prediabetes was defined as meeting any of the following parameters: 126 mg/dL > fasting plasma glucose level ≥100 mg/dL, 6.5% > HbA1c level ≥5.7%, or 200 mg/dL > post-OGTT plasma glucose level ≥140 mg/dL. Nondiabetes was defined as a fasting plasma glucose level <100 mg/dL, an HbA1c level <5.7%, and a post-OGTT plasma glucose level <140 mg/dL (15).
Other participant characteristics of interest collected by NHANES included sex, age, race, education level, income, and BMI (16). The race and/or ethnicity of the participants were classified as non-Hispanic White American, non-Hispanic Black American, non-Hispanic Asian American, Mexican American, or other race or ethnicity. Education level was categorized as “less than 12th grade,” “high school graduate,” “some college,” or “college graduate.” We categorized income based on the U.S. Federal Poverty Guidelines, classifying it as either below three times the poverty level or at/above three times the poverty level.
Statistical Analysis
Sampling weights designed for survey analysis were applied in all analyses. The weighted means across the 10-year period and their corresponding 95% CIs for each demographic factor and energy-adjusted nutrient intake were calculated. The macronutrients were transformed into percentages of total caloric intake. Specifically, protein, carbohydrate, and total sugar intake were calculated by multiplying grams by 4 kcal/g, and total fat, saturated fatty acid, monounsaturated fatty acid, and polyunsaturated fatty acid intake were calculated by multiplying grams by 9 kcal/g. Other nutrients were standardized based on an intake of 2,000 kcal.
Analyses were based on average nutrient intake estimates across NHANES survey cycles. We compared the 10-year average energy-adjusted nutrient intakes by adults with prediabetes, undiagnosed diabetes, or diagnosed diabetes with those in people without diabetes. We used multiple linear regression models to assess the association between diabetes status and energy-adjusted nutrient intake. Each model included diabetes status, total energy intake, age, sex, and race and ethnicity as covariates. Separate models were conducted for each nutrient of interest. In the subsequent analysis of changes over time, we grouped the data into four periods: 2011–2012, 2013–2014, 2015–2016, and 2017–March 2020 (prepandemic). Multiple linear regression models were used to compare energy-adjusted nutrient intake over time; sex, age, and race and ethnicity were adjusted to minimize the impact caused by demographic shifts in estimating the trends. The significance of the trend was assessed by calculating the P value for the time variable in each model; a P value <0.05 indicated a statistical significance. All analyses were conducted using the “survey” package in R, version 4.3.1.
Results
Demographic and Clinical Characteristics
Our final data set comprised a total of 22,547 adults (representing 184,664,074 U.S. adults after weighting), including 10,619 individuals without diabetes (47.1%), 7,687 individuals with prediabetes (34.1%), 1,002 individuals with undiagnosed diabetes (4.4%), and 3,239 individuals with diagnosed diabetes (14.4%). Compared with individuals without diabetes, individuals with prediabetes or diabetes were more likely to be male, older, of a racial minority, have a lower education level, and have lower income. In addition, fasting insulin and fasting serum glucose levels and systolic and diastolic blood pressures were significantly greater in people with diabetes than in those without diabetes (P < 0.01). Among people with diagnosed diabetes, the average diabetes duration was 10.5 years; 27.7% of these individuals reported using insulin, and 71.5% reported taking oral diabetes medications (Table 1).
Table 1.
Demographic and medical characteristics of individuals from the NHANES 2011–March 2020 (prepandemic) by diabetes status
| No diabetes | Prediabetes | Undiagnosed diabetes | Diagnosed diabetes | P value | |
|---|---|---|---|---|---|
| No. of individuals (unweighted) | 10,619 | 7,687 | 1,002 | 3,239 | |
| Sex (%) | <0.001 | ||||
| Male | 46.2 (45.1–47.3) | 52.1 (50.7–53.6) | 52.1 (46.8–57.4) | 52.3 (49.5–55.0) | |
| Female | 53.8 (52.8–54.9) | 47.9 (46.5–49.3) | 47.9 (42.6–53.2) | 47.8 (45.0–50.5) | |
| Age (years) | 41.0 (40.2–41.7) | 51.8 (51.2–52.3) | 56.8 (55.6–58.0) | 61.0 (60.4–61.6) | <0.001 |
| Race and ethnicity (%) | <0.001 | ||||
| Mexican American | 8.2 (6.6–9.8) | 8.8 (7.0–10.6) | 10.6 (7.5–13.6) | 9.4 (7.0–11.9) | |
| Non-Hispanic White | 68.3 (65.1–71.6) | 64.7 (61.4–68.1) | 59.0 (52.5–65.5) | 61.2 (57.3–65.1) | |
| Non-Hispanic Black | 9.3 (7.7–11.0) | 11.9 (9.8–3.9) | 13.7 (10.6–16.8) | 14.1 (11.4–16.8) | |
| Asian American | 4.7 (3.8–5.5) | 5.0 (4.1–5.9) | 6.0 (4.5–7.4) | 5.0 (3.9–6.0) | |
| Other race and ethnicity | 9.5 (8.3–10.7) | 9.6 (8.5–10.7) | 10.8 (8.0–13.6) | 10.3 (8.6–12.0) | |
| Education level (%)* | <0.001 | ||||
| Less than 12th grade | 10.6 (9.3–12.0) | 15.3 (13.7–16.9) | 21.1 (18.0–24.2) | 20.0 (17.5–22.5) | |
| High school graduate | 21.1 (19.5–22.6) | 23.6 (21.9–25.4) | 25.8 (22.5–29.0) | 26.8 (24.0–29.5) | |
| Some college | 32.8 (31.0–34.6) | 30.9 (29.2–32.6) | 30.0 (25.4–34.6) | 32.2 (29.6–34.9) | |
| College graduate | 35.5 (32.7–38.3) | 30.2 (27.5–32.9) | 23.1 (18.6–27.6) | 21.0 (18.2–23.8) | |
| Income (%)† | <0.001 | ||||
| <3 times the poverty rate | 54.5 (51.6–57.3) | 56.5 (53.5–59.5) | 63.7 (57.7–69.6) | 62.6 (59.6–65.6) | |
| ≥3 times the poverty rate | 45.6 (42.7–48.4) | 43.5 (40.5–46.5) | 36.4 (30.4–42.3) | 37.5 (34.4–40.5) | |
| BMI (kg/m2) | 27.6 (27.4–27.8) | 30.2 (29.9–30.5) | 32.9 (32.2–33.5) | 33.3 (32.9–33.7) | <0.001 |
| Fasting glucose (mg/dL) | 91.6 (91.4–91.9) | 104.8 (104.4–105.3) | 137.4 (134.1–140.7) | 157.7 (153.2–162.2) | <0.001 |
| HbA1c (%) | 5.2 (5.2–5.2) | 5.7 (5.6–5.7) | 6.7 (6.6–6.8) | 7.3 (7.2–7.4) | <0.001 |
| Insulin (μIU/mL)‡ | 9.1 (8.7–9.4) | 13.5 (13.0–13.9) | NA | NA | <0.001 |
| Diabetes duration (years) | NA | NA | NA | 10.5 (10.1–11.0) | |
| Insulin use (%) | NA | NA | NA | 24.3 (22.4–26.2) | |
| Other glucose-lowering medication use (%) | NA | NA | NA | 76.0 (74.1–77.9) | |
| Systolic blood pressure (mmHg) | 118.3 (117.8–118.8) | 124.9 (124.3–125.6) | 130.4 (128.6–132.3) | 129.7 (128.6–130.8) | <0.001 |
| Diastolic blood pressure (mmHg) | 70.8 (70.3–71.2) | 72.5 (72.0–73.0) | 72.9 (71.5–74.3) | 70.6 (70.0–71.3) | <0.001 |
Values are reported as percentage or mean with 95% CI unless otherwise indicated. An ANOVA-Wald test was used for continuous variables, and a χ2 test was used for categorical variables. *Available for 90.96% of participants. †Available for 85.19% of participants. ‡Available for 47.15% of participants. NA, not applicable.
Nutrient Intake by Diabetes Status
After adjusting for age, sex, and race and ethnicity, individuals without diabetes had an average energy-adjusted intake of 15.6% from protein and 47.5% from carbohydrates. Total sugars contributed 20.4% of energy, and dietary fiber intake was 16.2 g/2,000 kcal on average. Total fat constituted 34.2% of energy, including 10.9% from saturated fatty acids, 12.0% from monounsaturated fatty acids, and 8.1% from polyunsaturated fatty acids. In a 2,000 kcal/day level, the average intakes of cholesterol, calcium, magnesium, and zinc were 276.6 mg, 883.7 mg, 289.6 mg, and 10.3 mg, respectively. Sodium intake was 3,304.9 mg, and potassium intake was 2,461.6 mg. The HEI score for this group was 51.1 (Table 2).
Table 2.
Energy-adjusted average dietary intake in nonpregnant adults from the NHANES 2011–March 2020 by diabetes status, adjusted for age, race, and sex
| No diabetes | Prediabetes | P value | Undiagnosed diabetes | P value* | Diagnosed diabetes | P value† | |
|---|---|---|---|---|---|---|---|
| No. of individuals (unweighted) | 10,619 | 7,687 | 1,002 | 3,229 | |||
| Energy (kcal/day)‡ | 2,195.3 (2,188.4–2,202.3) | 2,207.2 (2,200.2–2,214.2) | 0.59 | 2,165.2 (2,158.2–2172.2) | 0.54 | 20,56.1 (2,049.1–2,063.0) | <0.001 |
| Protein (%kcal) | 15.6 (15.5–15.6) | 15.6 (15.5–15.6) | 0.94 | 15.6 (15.6–15.6) | 0.90 | 16.1 (16.1–16.1) | 0.003 |
| Carbohydrates (%kcal) | 47.5 (47.5–47.5) | 48.0 (48.0–48.1) | 0.06 | 47.5 (47.5–47.5) | 0.90 | 46.5 (46.5–46.5) | 0.07 |
| Total sugars (%kcal) | 20.4 (20.4–20.4) | 20.9 (20.8–20.9) | 0.14 | 20.8 (20.8–20.8) | 0.400 | 18.8 (18.8–18.9) | 0.008 |
| Dietary fiber (g/2,000 kcal) | 16.2 (16.2–16.3) | 15.6 (15.5–15.6) | 0.01 | 14.9 (14.9–15.0) | 0.001 | 15.8 (15.8–15.9) | 0.36 |
| Total fat (%kcal) | 34.2 (34.2–34.2) | 34.9 (34.9–34.9) | 0.008 | 34.8 (34.7–34.8) | 0.27 | 36.7 (36.7–36.7) | <0.001 |
| Saturated fatty acids (%kcal) | 10.9 (10.9–10.9) | 11.3 (11.3–11.3) | 0.001 | 11.7 (11.6–11.7) | 0.001 | 11.8 (11.8–11.9) | <0.001 |
| Monounsaturated fatty acids (%kcal) | 12.0 (12.0–12.0) | 12.1 (12.1–12.1) | 0.19 | 12.1 (12.1–12.1) | 0.82 | 12.8 (12.8–12.8) | <0.001 |
| Polyunsaturated fatty acids (%kcal) | 8.1 (8.1–8.1) | 8.2 (8.2–8.2) | 0.28 | 7.7 (7.7–7.7) | 0.07 | 8.6 (8.6–8.6) | 0.006 |
| Cholesterol (mg/2,000 kcal) | 276.6 (276.3–276.9) | 286.5 (286.2–286.9) | 0.02 | 291.2 (290.9–291.5) | 0.10 | 319.5 (319.1–319.8) | <0.001 |
| Calcium (mg/2,000 kcal) | 883.7 (882.3–885.1) | 878.2 (876.8–879.6) | 0.59 | 931.4 (930.0–932.7) | 0.05 | 912.3 (911.0–913.7) | 0.008 |
| Magnesium (mg/2,000 kcal) | 289.6 (289.1–290.1) | 275.9 (275.4–276.4) | <0.001 | 263.6 (263.1–264.0) | <0.001 | 270.0 (269.5–270.4) | <0.001 |
| Zinc (mg/2,000 kcal) | 10.3 (10.3–10.4) | 10.3 (10.3–10.3) | 0.50 | 10.4 (10.4–10.4) | 0.76 | 10.5 (10.5–10.5) | 0.16 |
| Sodium (mg/2,000 kcal) | 3,304.9 (3,300.8–3,309.0) | 3,405.6 (3,401.5–3,409.7) | <0.001 | 3,463.6 (3,459.5–3,467.7) | 0.03 | 3,608.8 (3,604.7–3,612.9) | <0.001 |
| Potassium (mg/2,000 kcal) | 2,461.6 (2,457.3–2,465.9) | 2,395.6 (2,391.4–2,399.9) | 0.003 | 2,349.8 (2,345.5–2,354.1) | 0.04 | 2,379.9 (2,375.7–2,384.2) | 0.02 |
| HEI-2015 | 51.1 (51.1–51.2) | 48.9 (48.8–49.0) | <0.001 | 47.1 (47.1–47.2) | <0.001 | 47.9 (47.8–47.9) | <0.001 |
Values are reported as percentages or mean with 95% CI across 10 years unless otherwise indicated. All nutrient variables were adjusted for energy (2,000 kcal), age, race, or sex. Bold indicates a significant value. *P value indicates a difference between individuals without diabetes and individuals with undiagnosed diabetes. †P value indicates a difference between individuals without diabetes and individuals with diagnosed diabetes. ‡Average energy intake per day was adjusted for age, race, and sex.
Comparing energy-adjusted average intake between adults with prediabetes and those without diabetes, adults with prediabetes had a lower intake of dietary fiber (14.92 vs. 16.21 g), magnesium (275.9 vs. 289.6 mg), and potassium (2,395.6 vs. 2,461.6 mg), and the HEI scores (48.9 vs. 51.1) were lower than for adults without diabetes. Conversely, the energy-adjusted mean intakes of total fat (34.9% vs. 34.2%), saturated fatty acids (11.3% vs. 10.9%), cholesterol (286.5 vs. 276.6 mg), and sodium (3,406.6 vs. 3,304.9 mg) were higher in adults with prediabetes (Table 2).
Comparing energy-adjusted average intake between adults with undiagnosed diabetes and those without diabetes, adults with undiagnosed diabetes consumed less dietary fiber (14.9 vs. 16.2 g/2,000 kcal), more saturated fatty acids (11.7% vs. 10.9%), less magnesium (263.6 vs. 289.6 mg), more sodium (3,463.6 vs. 3,304.9 mg), and less potassium (2,349.8 vs. 2,461.7 mg). The mean HEI scores of adults with undiagnosed diabetes, at 47.1, was the lowest among the four groups (Table 2).
Comparing energy-adjusted average intake between adults with diagnosed diabetes and those without diabetes, adults with diagnosed diabetes had higher protein (16.2% vs. 15.6%), total fat (36.7% vs. 34.2%), cholesterol (319.5 vs. 276.6 mg), calcium (912.3 vs. 883.7 mg), and sodium (3,608.8 vs. 3,304.9 mg) intakes and lower total sugar (18.8% vs. 20.4%), magnesium (269.9 vs. 289.6 mg), and potassium (2,380 vs. 2,461.7 mg) intakes. The average HEI score for individuals with diagnosed diabetes was 47.9, which was lower than that for individuals without diabetes (51.1). All these values were adjusted based on a 2,000 kcal/day dietary intake. All differences were statistically significant at P < 0.05 (Table 2).
Temporal Trends in Nutrients Intake by Diabetes Status
Among individuals without diabetes, the intake of carbohydrates and total sugar consumption decreased from 2011 to 2020, whereas upward trends were observed for the average intake of total fat, saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids (Fig. 1 and Supplementary Table 1). Additionally, the average calcium (902.05–851.7 mg) and potassium (2,519.1–2,352.7 mg) intake decreased from 2011 to 2020 (Fig. 2). These trends were also reflected in the decrease in the HEI score over the years (P < 0.05).
Figure 1.

Major macronutrient intake trend from 2011 to 2020 by diabetes status. *Statistical significance of trend over time (P < 0.05). All nutrient variables were adjusted for energy (2,000 kcal/day), age, race, or sex.
Figure 2.

Major micronutrient intake trend from 2011 to 2020 by diabetes status. *Statistical significance of trend over time (P < 0.05). All nutrient variables were adjusted for energy (2,000 kcal), age, race, or sex.
Individuals with prediabetes decreased their intake of carbohydrate and total sugars, but the intake of total fat, saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids increased (Fig. 1). The intake of cholesterol also increased, with a notable increase from 268.1 mg to 294.5 mg between 2011 and 2012 and between 2017 and 2020, respectively. Potassium intake decreased from 2,582.9 mg to 2,443.1 mg. These trends were reflected in the decline in the HEI score over the years (P < 0.05) (Fig. 2 and Supplementary Table 2).
In the undiagnosed diabetes population, the mean intake of total fat and saturated fatty acids gradually increased over the study period (Fig. 1). There were no significant changes in the intake of other nutrients (Fig. 2 and Supplementary Table 3).
Among individuals with diagnosed diabetes (Fig. 1), the average carbohydrate intake decreased from 46.81% to 44.25%, concomitant with an upward trend in total fat (36.59% to 38.84%) and saturated fatty acids consumption (11.6% to 12.3%). In addition, there was a downward trend in dietary fiber (from 18.29 to 15.36 g), magnesium (from 293.04 to 278.44 mg), sodium (from 3,615.28 to 3,326.31 mg), and potassium (from 2,724.89 to 2,411.61 mg) from 2011 to 2020 (Fig. 2 and Supplementary Table 4). All values were adjusted based on a 2,000 kcal/day dietary intake.
Conclusions
In this nationally representative sample spanning 10 years, we found that adults with diabetes had lower overall diet quality than those without diabetes, and that the diet quality of these adults did not improve over time. In people with diabetes, we saw a decline in carbohydrate and fiber intake, particularly fiber-rich carbohydrates, alongside an increase in total fat, especially saturated fat. Consumption of important micronutrients, such as potassium and magnesium, also decreased, which contributed to lack of improvement in the HEI score. These results build upon historical trends: whereas earlier data showed increased carbohydrate intake (1988–2004) (8) and higher levels of sodium, calcium, and cholesterol (1988–2012) (17), more recent analysis (1999–2018) (9) confirms a persistent pattern of increasing intake of saturated fats and declining potassium that led to HEI scores declining.
People with prediabetes also had lower overall diet quality compared with those without diabetes, with no discernible improvement across survey cycles. Our trend analysis showed that among those with prediabetes, reductions in total sugar and higher intakes of unsaturated fat indicated positive shifts. However, there were also negative changes, including increased saturated fat and cholesterol intake.
Although sodium intake has shown some improvement among people with diagnosed diabetes, it remains above recommended levels (i.e., limited to 2,300 mg per 2,000 kcal) and was well above levels recently documented to lower blood pressure in individuals with T2D (18). No decline in sodium intake was seen among those without diabetes, with prediabetes, or with undiagnosed diabetes. These findings highlight persistent challenges in reducing sodium intake, which may be exacerbated by certain dietary patterns that are high in ultra-processed foods and low in fruits, vegetables, or whole grains (19). In addition, previous studies showed individuals with diabetes consume more processed foods, which is likely leading to the decreasing intake of potassium, magnesium, and fiber (20–22).
Comparing dietary trends with that of individuals without diabetes revealed several pronounced nutrient shifts that are unique to the individuals with diabetes. Whereas total carbohydrate and fiber intake declined in both groups, the fiber decrease was more pronounced among individuals with diabetes. Furthermore, individuals with diabetes did not have the improvements seen among the general population regarding increased intake of mono- and polyunsaturated fats and decreased cholesterol. Notably, magnesium and sodium levels decreased specifically among those with diabetes, a trend not observed in the population without diabetes. These differences suggest that patients with diabetes may be adopting restrictive dietary patterns that successfully reduce sugar and sodium but inadvertently lead to a greater loss of essential fiber and minerals. These findings highlight the need for customized public health outreach, and future recommendations must move beyond simple macronutrient distribution and focus on high-quality food patterns to ensure patients with diabetes meet critical requirements.
The average dietary intake of U.S. adults showed limited adherence to general nutrition guidelines, a pattern that was consistent across all diabetes statuses. The DGA, issued by the USDA, provides general nutrient standards for all Americans. The optimal recommendations, based on a 2,000 kcal/day diet, for fiber was more than 18 g, saturated fat less than 10%, cholesterol less than 300 mg, calcium more than 1,000 mg, magnesium more than 310 mg, zinc more than 8 mg, sodium less than 2,300 mg, and potassium more than 2,600 mg (6). Our analysis showed the overall average intake of dietary fiber, saturated fatty acids, calcium, magnesium, sodium, and potassium across all four groups (nondiabetes, prediabetes, undiagnosed diabetes, and diagnosed diabetes) did not achieve any of these recommended levels. Among individuals with diagnosed diabetes, levels of saturated fatty acids were 20% above the recommended limits, and dietary fiber intake was around 40% below the recommendation. Calcium intake and magnesium intake were both approximately 10% lower than recommendations. Additionally, sodium intake was about 60% higher than suggested, whereas potassium intake was 10% lower. These patterns resulted in lower HEI scores for individuals with prediabetes (48.9), undiagnosed diabetes (47.1), and diagnosed diabetes (47.9) compared with those without diabetes (51.13).
In light of the widening gaps in key nutrient intake, the effectiveness of future public health strategies should depend on moving beyond generic advice to provide practical, evidence-based recommendations on food sources rich in the necessary nutrients. It is critical to enhance dietary fiber sources such as vegetables, whole grains, and legumes (e.g., soybeans, peas, lentils, chickpeas). Low potassium could be corrected by increasing consumption of fresh vegetables and fruits, especially beans, potatoes (with skin), prunes, raisins, acorn squash, and spinach (23–25). Likewise, magnesium intake can be improved through the consumption of more fruits, green vegetables, legumes, nuts, and seeds such as pumpkin and sesame seeds (26). The main sources of dietary cholesterol and saturated fat that can be reduced are from high-fat meats, eggs, butter, and full-fat dairy products. Dietary Approaches to Stop Hypertension (DASH) and Mediterranean-style diets are associated with high fiber, high potassium, low fat, and low cholesterol consumption and have been shown to improve blood pressure and glycemia among people with diabetes (18,27).
Lack of improvement in diet quality may result from socioeconomic barriers, specifically food insecurity, policy instability, and the lack of affordable fresh produce. For instance, the Supplemental Nutrition Assistance Program (SNAP) is the largest food assistance program in the U.S., and participation is associated with improved glucose control. However, current legislation plans to reduce SNAP funding by approximately $186 billion over 10 years (28). These policy shifts may further worsen adherence to dietary guidelines by people with diabetes and those who need specific food accommodations. Furthermore, people with low income and those in minoritized racial and ethnic groups in the U.S. often reside in disinvested communities with limited availability of healthy foods (“food deserts”) and an overabundance of unhealthy foods (“food swamps”) (29). Beyond this, the limited clarity of public health recommendations, as well as a lack of specific knowledge about preparing nutrient-dense meals, have hindered significant improvement in healthy eating (30,31).
Our study has several strengths. Using a nationally representative sample, results from this study are generalizable. To categorize individuals with diagnosed and undiagnosed diabetes, both laboratory tests and questionnaire data were used to minimize selection bias. Notably, unlike previous studies that often looked at these groups in isolation, this study is the first, to our knowledge, to comprehensively compare trends in nutrient intake patterns across individuals of differing diabetes status. Nonetheless, there are limitations to our study. The dietary trend analyses using 24-h dietary recall data may not accurately capture long-term dietary behavior. A single 24-h dietary recall interview was used to estimate nutrient intake; however, given the substantial sample size of the NHANES, 1-day dietary recall is reliable enough to reflect the groups’ dietary intake level (12,13). Additionally, the NHANES data set does not distinguish between type 1 and type 2 diabetes, although more than 90% of the adults with diabetes have T2D (32). In addition, this analysis did not include data regarding specific food sources. Without distinguishing this, the factors driving the observed variations in total nutrient intake remain speculative. The AMPM food categories and USDA Food and Nutrient Database for Dietary Studies were used as models to calculate and estimate nutrient components. This approach may not capture all noncategory foods that do not fit into specific What We Eat in America (WWEIA) food categories, and there was difficulty accurately calculating nutrients (e.g., sodium, calcium) with increasing intake of processed foods, leading to potential discrepancies in nutrient intake.
In conclusion, the quality of nutrient intakes by people with diagnosed diabetes or prediabetes declined from 2011 to 2020. Because nutrition is crucial for both preventing and managing T2D, future public health efforts should focus on encouraging the adoption of healthy dietary patterns characterized by high-fiber, low-cholesterol foods and those rich in magnesium and potassium.
This article contains supplementary material online at https://doi.org/10.2337/figshare.31843873.
Article Information
Duality of Interest
No potential conflicts of interest relevant to this article were reported.
Author Contributions
X.L. and H.-C.Y. designed the research and wrote the manuscript. X.L. conducted research, analyzed data, and performed statistical analysis. S.J.P., S.O., L.J.A., and H.-C.Y. provided essential information and materials. S.J.P., S.O., and L.J.A. provided critical revision. All authors read and approved the final manuscript. X.L. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Supporting information
References
- 1. Centers for Disease Control and Prevention . National Diabetes Statistics Report. 23 July 2024. Accessed 28 December 2024. Available from https://www.cdc.gov/diabetes/php/data-research/index.html
- 2. Parker ED, Lin J, Mahoney T, et al. Economic costs of diabetes in the U.S. in 2022. Diabetes Care 2024;47:26–43 [DOI] [PubMed] [Google Scholar]
- 3. Mulnier HE, Seaman HE, Raleigh VS, Soedamah-Muthu SS, Colhoun HM, Lawrenson RA. Mortality in people with type 2 diabetes in the UK. Diabet Med 2006;23:516–521 [DOI] [PubMed] [Google Scholar]
- 4. Galaviz KI, Weber MB, Suvada K, et al. Interventions for reversing prediabetes: a systematic review and meta-analysis. Am J Prev Med 2022;62:614–625 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. American Diabetes Association Professional Practice Committee . 5. Facilitating positive health behaviors and well-being to improve health outcomes: Standards of Care in Diabetes—2025. Diabetes Care 2024;48:S86–S127 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. U.S. Department of Agriculture, U.S. Department of Health and Human Services . Dietary Guidelines for Americans, 2020–2025. Accessed 20 October 2023. Available from DietaryGuidelines.gov. [Google Scholar]
- 7. Lin BH, Guthrie J, Smith T, Eds. Dietary Quality by Food Source and Demographics in the United States, 1977–2018. Economic Research Service, U.S. Department of Agriculture, 2023 [Google Scholar]
- 8. Oza-Frank R, Cheng YJ, Narayan KMV, Gregg EW. Trends in nutrient intake among adults with diabetes in the United States: 1988-2004. J Am Diet Assoc 2009;109:1173–1178 [DOI] [PubMed] [Google Scholar]
- 9. Sun H, Weaver CM. Trends in diet quality and increasing inadequacies of micronutrients vitamin C, vitamin B12, iron and potassium in US type 2 diabetic adults. Nutrients 2023;15:1980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. National Center for Health Statistics, Centers for Disease Control and Prevention. National Health and Nutrition Examination Survey Homepage . 24 September 2024. Accessed 14 October 2024. Available from https://www.cdc.gov/nchs/nhanes/index.htm
- 11. National Center for Health Statistics, Centers for Disease Control and Prevention. About the National Health and Nutrition Examination Survey . 31 May 2023. Accessed 24 October 2023. Available from https://www.cdc.gov/nchs/nhanes/about_nhanes.htm
- 12. McClure ST, Schlechter H, Oh S, et al. Dietary intake of adults with and without diabetes: results from NHANES 2013–2016. BMJ Open Diabetes Res Care 2020;8:e001681. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. National Center for Health Statistics, Centers for Disease Control and Prevention. NHANES dietary analyses . Accessed 4 March 2026. Available from https://wwwn.cdc.gov/nchs/nhanes/tutorials/dietaryanalyses.aspx
- 14. Zhan JJ, Hodge RA, Dunlop AL, et al. Dietaryindex: a user-friendly and versatile R package for standardizing dietary pattern analysis in epidemiological and clinical studies. Am J Med Sci 2024;120:1165–1174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. American Diabetes Association . 2. Classification and diagnosis of diabetes: Standards of Care in Diabetes—2023. Diabetes Care 2023;46(Suppl. 1):S19–S40 [DOI] [PMC free article] [PubMed]
- 16. Walker RJ, Strom Williams J, Egede LE. Influence of race, ethnicity and social determinants of health on diabetes outcomes. Am J Med Sci 2016;351:366–373 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Casagrande SS, Cowie CC. Trends in dietary intake among adults with type 2 diabetes: NHANES 1988–2012. J Hum Nutr Diet 2017;30:479–489 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Pilla SJ, Yeh H-C, Mitchell CM, et al.; DASH4D Collaborative Research Group . Dietary patterns, sodium reduction, and blood pressure in type 2 diabetes: the DASH4D randomized clinical trial. JAMA Intern Med 2025;185:937–946 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Murphy MM, Scrafford CG, Barraj LM, et al. Potassium chloride-based replacers: modeling effects on sodium and potassium intakes of the US population with cross-sectional data from NHANES 2015–2016 and 2009-2010. Am J Clin Nutr 2021;114:220–230 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Brandão-Lima PN, Carvalho GBd, Santos RKF, et al. Intakes of zinc, potassium, calcium, and magnesium of individuals with type 2 diabetes mellitus and the relationship with glycemic control. Nutrients 2018;10:1948. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Baqar S, Michalopoulos A, Jerums G, Ekinci EI. Dietary sodium and potassium intake in people with diabetes: are guidelines being met? Nutr Diabetes 2020;10:23–29 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Barbagallo M, Dominguez LJ. Magnesium and type 2 diabetes. World J Diabetes 2015;6:1152–1157 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Ley SH, Hamdy O, Mohan V, Hu FB. Prevention and management of type 2 diabetes: dietary components and nutritional strategies. Lancet 2014;383:1999–2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. D’Elia L, Masulli M, Cappuccio FP, Zarrella AF, Strazzullo P, Galletti F. Dietary potassium intake and risk of diabetes: a systematic review and meta-analysis of prospective studies. Nutrients 2022;14:4785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Chatterjee R, Slentz C, Davenport CA, et al. Effects of potassium supplements on glucose metabolism in African Americans with prediabetes: a pilot trial. Am J Clin Nutr 2017;106:1431–1438 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Schiefermeier-Mach N, Egg S, Erler J, et al. Electrolyte intake and major food sources of sodium, potassium, calcium and magnesium among a population in Western Austria. Nutrients 2020;12:1956. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Davis C, Bryan J, Hodgson J, Murphy K. Definition of the Mediterranean diet: a literature review. Nutrients 2015;7:9139–9153 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Aussenberg RA., Supplemental Nutrition Assistance Program (SNAP) and related nutrition programs in P.L. 119-21: an overview . 15 August 2025. Accessed 11 February 2026. Available from https://www.congress.gov/crs-product/R48552
- 29. Karpyn AE, Riser D, Tracy T, Wang R, Shen YE. The changing landscape of food deserts. UNSCN Nutr 2019;44: 46–53 [PMC free article] [PubMed] [Google Scholar]
- 30. Booth AO, Lowis C, Dean M, Hunter SJ, McKinley MC. Diet and physical activity in the self-management of type 2 diabetes: barriers and facilitators identified by patients and health professionals. Prim Health Care Res Dev 2013;14: 293–306 [DOI] [PubMed] [Google Scholar]
- 31. Denniss E, Lindberg R, McNaughton SA. Quality and accuracy of online nutrition-related information: a systematic review of content analysis studies. Public Health Nutr 2023;26: 1345–1357 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Centers for Disease Control and Prevention . Methods for the National Diabetes Statistics Report. 22 May 2024. Accessed 14 August 2025. Available from https://www.cdc.gov/diabetes/php/data-research/methods.html
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
