Abstract
Aims
Numerous studies report positive associations between total carbohydrate (CHO) intake and incident metabolic syndrome (MetS), but few differentiate quality or type of CHO relative to MetS. We examined source of CHO intake, including added sugar (AS), AS-rich CHO foods, and sugar-sweetened beverages (SSBs) associated with incident MetS in adults enrolled in the Coronary Artery Risk Development in Young Adults (CARDIA) study.
Methods and results
Among 3154 Black American and White American women and men aged 18–30 years at baseline, dietary intake was assessed by diet history three times over 20 years. Sources of AS-rich CHO foods and beverages include sugar-rich refined grain products, candy, sugar products, and SSBs. Incident MetS was created according to standard criteria. Time-dependent Cox proportional hazards regression analysis evaluated the associations of incident MetS across quintiles of cumulative intakes of AS-rich CHO foods and beverages, AS, and SSBs adjusted for potential confounding factors over 30 years of follow-up. The associations of AS-rich CHO foods and beverages, AS, and SSB intakes with incident MetS were consistent. Compared with the lowest intake, the greatest intakes of AS-rich CHOs, AS, and SSBs were associated with 59% (Ptrend < 0.001), 44% (Ptrend = 0.01), and 34% (Ptrend = 0.03) higher risk of developing MetS, respectively. As expected, diet quality was lower across increasing quintiles of AS-rich CHO foods and beverages, AS, and SSBs (all Ptrend < 0.001).
Conclusion
Our study findings are consistent with an elevated risk of developing MetS with greater consumption of AS, AS-rich CHO foods, and SSBs, which support consuming fewer AS-rich CHO foods and SSBs.
Keywords: Carbohydrate foods and beverages, Added sugar, Sugar-sweetened beverages, Metabolic syndrome, Prospective cohort study
Introduction
Metabolic syndrome (MetS) is a cluster of three of five adverse cardiovascular disease (CVD) risk factors, including abdominal obesity, elevated blood pressure, elevated triglyceride, elevated glucose or diagnosis of diabetes, and low HDL cholesterol (HDL-C).1 Metabolic syndrome is also a known precursor to type 2 diabetes.1 Since 1999, the prevalence of MetS among US adults increased from 28.2 to 37.1% by 2018.2 To prevent and treat MetS, several dietary strategies, such as altering the type or quality of carbohydrate (CHO) and related food sources, have been proposed.3 In a meta-analysis of 18 observational studies conducted in several countries, adults consuming the highest amount of total CHO compared with the lowest had 25% greater risk of developing MetS, while the results across numerous studies were inconsistent for the associations between total CHO intake and MetS.4 The inconsistent results may be explained by the differing nutrient and compound composition of CHO foods and beverages, including grain products, vegetables, fruit, dairy products, and sugar-sweetened beverages (SSBs), as well as amount of added sugar (AS). Therefore, examining total CHO intake relative to a health outcome most likely masks the differential associations from the aforementioned sources of CHOs.5
Global CHO recommendations as well as the US 2020–25 Dietary Guidelines emphasize CHO quality, including intakes of fibre-rich foods but limiting AS or free sugar intake.6,7 In the USA, adults consume about 47% of energy (kcal) from CHOs.8 Notably, Americans typically consume over 25% of total energy from snack foods and SSBs according to National Health and Nutrition Examination Surveys 2015–16.6,9 Moreover, intakes of SSBs and sweetened coffee and tea contribute about 35% of total AS consumed by US adults,6 with the remaining AS intake from food sources, including sweet bakery products, ready-to-eat breakfast cereals, refined grain desserts, bread, rolls, tortillas, candy, and sugar products.6,10 Consumption of SSBs has been adversely related to health outcomes worldwide, including the development of obesity, abdominal obesity, and MetS in adults as well as obesity and MetS in adolescents.11–15
The objective of this study is to investigate the associations of AS and AS-rich CHO food and beverage intakes with the risk of developing MetS in adult Black American and White American women and men enrolled in the Coronary Artery Risk Development in Young Adults (CARDIA) study. We hypothesize that consuming AS and AS-rich CHO foods and beverages is positively associated with risk of incident MetS in CARDIA participants over 30 years of follow-up.
Methods
The institutional review board at each institution reviewed and approved the CARDIA study protocols. All study participants gave written consent prior to taking part in each exam.
Study population
Data from participants enrolled in the CARDIA study and followed for 30 years were used for this analysis. The CARDIA study is an ongoing prospective cohort study that initially sought to discern the role of a wide variety of risk factors for CVD among a diverse group of young Black American and White American adults 18–30 years old at baseline recruited from four field centres across the USA between 1985 and 1986. A total of 5115 participants were enrolled at baseline with nearly equal numbers of participants in terms of race, sex, education level, and age group (18–24 and 25–30 years old). Details about the study design and methods have been described previously.16 Nine follow-up exams have been conducted between 1987–88 and 2021–22.
Assessment of dietary intake
Trained and certified data collectors administered the validated CARDIA Diet History questionnaire to assess dietary intake at baseline (Y0) and follow-up exam Years 7 (Y7) and 20 (Y20).17 Participants were asked to report foods and beverages consumed in the past month, including brand name foods and beverages, amount consumed, and frequency of consumption.17 Reported foods and beverages were coded using the nutrient and food tables obtained from the Nutrition Data System for Research (NDSR), developed at the University of Minnesota Nutrition Coordinating Center. Added sugar–rich CHO foods and beverages include refined grain breads and rolls; quick breads (muffins, cornbread, and sweet breads); sweet bakery products (cakes, cookies, pies, bars, doughnuts, and pastries); ready-to-eat cereals; candy; sugar products (jams and jellies, honey, syrups, and sugars added to foods and beverages); and SSBs. Dairy desserts were excluded from the AS-rich CHO category due to potentially beneficial effects of dairy fatty acids on outcomes.18 Added sugar was defined, consistent with the Dietary Guidelines for Americans, as sugar, high-fructose corn syrup, and other caloric sweetener products added to foods during food and beverage preparation and commercial production and does not include naturally occurring sugars (intrinsic mono- and disaccharides in fruit, dairy products, and other foods).8 The Pearson correlations for AS with SSB and AS-rich CHO foods and beverages were 0.50 and 0.57, respectively (both P < 0.001), and SSB with AS-rich CHO foods and beverages was 0.19 (P < 0.001).
Assessment of clinical characteristics
Each MetS component was collected or measured according to standard CARDIA procedures at each examination.16 Anthropometrics, including height, weight, and waist circumference, were measured by trained staff with participants dressed in light clothing and no shoes. Height was measured using a wall-mounted stadiometer and recorded to the nearest 0.5 cm. Weight was measured using a beam balance scale and recorded to the nearest 0.2 kg. Body mass index (BMI) was calculated as weight in kilograms divided by height in square metre. While the participant was standing, waist circumference was measured two times using a cloth tape midway between the iliac crest and the bottom of the ribcage and the average recorded to the nearest 0.5 cm. In Y0–Y15, systolic and diastolic blood pressures were measured using a random zero sphygmomanometer, three times, while the participant was seated after a 5-min rest with 1-min intervals in between measurements. In Y20–Y30, blood pressure was similarly assessed using an Omron oscillometer, calibrated to random zero measurements at Y20. The average of the last two of three measurements was recorded.
A fasting blood sample was drawn and stored at −80°C for measurement of lipids and glucose. Blood lipids were assayed from stored samples at the Northwest Lipid Research Clinic Laboratory (Seattle, WA). Cholesterol and triglycerides were assayed by enzymatic methods. HDL cholesterol levels were measured after precipitation with dextran sulfate–Mg2+. Glucose was measured from stored samples by using hexokinase coupled to glucose-6-phosphate dehydrogenase. Glucose 2 h after a glucose drink was assayed at Y10, Y20, and Y25. Haemoglobin A1c was assayed from whole blood at Y20 and Y25 analysed nonenzymatically at the University of Minnesota using the Tosoh G7 Automated HPLC Analyzer. Diabetes was defined as fasting glucose ≥126 mg/dL or taking an antidiabetic drug. When measured, 2-h glucose ≥200 mg/dL or haemoglobin A1C ≥ 6.5% also classified a participant as having diabetes.
Metabolic syndrome
In this study, MetS was designated based on a harmonized definition proposed by the International Diabetes Federation Task Force on Epidemiology and Prevention, National Heart, Lung, and Blood Institute, and the American Heart Association specific to individuals of most ancestries living in the USA.1 Metabolic syndrome is characterized by the presence of at least three of five risk factors including (i) fasting plasma glucose ≥100 mg/dL or undergoing drug treatment for elevated glucose or diabetes, (ii) HDL-C < 40 mg/dL for males or <50 mg/dL for females, (iii) fasting triglycerides ≥150 mg/dL or undergoing drug treatment for elevated triglycerides, (iv) waist circumference >102 cm in males or >88 cm in females, and (v) systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg or undergoing drug treatment for hypertension.1
Other characteristics
Demographic information, education level, and smoking status were collected using self-report standardized questionnaires. Education status, a proxy for socioeconomic status, was defined as greater than high school graduation or less than or equal to high school. Smoking status was defined as a current smoker. Physical activity and leisure time were determined by a validated questionnaire administered by a trained technician.19
Statistical methods
For this study, participants who took part in the Year 30 (Y30) clinic exam were included in this analysis (n = 3358). However, we excluded those with prevalent MetS or diabetes at baseline (n = 60), bariatric surgery by the Y30 exam (n = 77), incomplete or implausible energy intake data [men: <800 or >8000 kcal/day, women: <600 or >6000 kcal/day (n = 46), and missing covariates (n = 21)]. As shown in Figure 1, the final study sample included n = 3154: 915 White American women, 865 Black American women, 769 White American men, and 605 Black American men. With a sample size of 3154, there was sufficient power (>95%) to detect significant associations between the exposures and incident MetS.
Figure 1.
Study population: flowchart of inclusion and exclusion criteria. MetS, metabolic syndrome.
SAS version 9.4 software (SAS Institute, Cary, NC) was used for all statistical analyses. Quintiles were created for each of AS-rich CHOs, AS, and SSBs at baseline. Baseline characteristics were reported as mean [standard error (SE)] or frequency (%) stratified across quintiles of baseline intake of AS-rich CHOs (sv/day), AS (g/day), and SSBs (sv/day) adjusted for age, sex, race, field centre, education, and energy intake. One serving of SSB is equivalent to eight fluid ounces. To account for dietary intake in the statistical models for the associations of AS-rich CHO foods, AS, and SSBs with MetS, three diet patterns were created. The Healthy Eating Index 2020 (HEI-2020) score, a measure of diet quality, was computed as the sum of servings from major food groups,20 but AS was omitted. Two western diet pattern scores were derived from principal component analysis: one without AS-rich CHO foods and beverages and another western diet pattern score without SSBs. To account for change in dietary habits over 20 years, average intake of AS-rich CHO foods and beverages (sv/day), SSBs (sv/day), and AS (g/day) for the three exposure measures was determined. The averaged exposure variables were created by computing the mean of each of AS, SSB, and AS-rich CHO foods and beverages at baseline, Y7, and Y20.
Multiple regression analysis evaluated the association of each MetS component at Y30 stratified across quintiles of averaged AS-rich CHO foods and beverages, AS, and SSBs adjusted for potential confounding factors. Time-dependent Cox proportional hazards regression analysis was used to evaluate the associations for each of AS-rich CHO foods and beverages, AS, and SSBs with risk of incident MetS by study Y30. Time at risk was defined as until first diagnosis or date of the last follow-up examination. Person-years were calculated. Hazard ratios were computed for Quintile 2 (Q2) through Quintile 5 (Q5; highest intake) of AS-rich CHOs, AS, and SSBs compared with Quintile 1 (Q1; lowest intake) as the referent group. The cumulative average of dietary intakes, which increases the precision of dietary exposure, was used to create our diet exposures of interest. The cumulative average for dietary intake was calculated based on participants’ follow-up times: (i) Y0 to predict incident MetS occurring before Y7; (ii) average Y0 and Y7 dietary intakes predicting incident MetS from Y7 to Y20; and (iii) average Y0, Y7, and Y20 predicting MetS from Y20 to Y30. Three models were developed: Model 1 was adjusted for baseline energy intake and demographic characteristics age, sex, race, education, and field centre; Model 2 adjusted for Model 1 plus baseline current smoking status, physical activity, and a diet pattern, as appropriate; and Model 3 was adjusted for Model 2 plus BMI and menopausal status. A linear trend across quintiles of AS-rich CHO foods and beverages, AS, or SSBs was tested with contrast statements using orthogonal polynomial coefficients. Effect modification of sex on the associations between the exposures (AS, SSBs, and AS-rich CHO foods and beverages) and incident MetS was tested; the interactions were not significant (Pinteraction ranged from 0.36 to 0.76).
Results
As shown in Table 1, those who consumed more daily servings of AS-rich CHOs at baseline were more likely to be younger, male, Black American, have less than a high school education, current smoker, and less likely to be physically active compared with those who consumed fewer AS-rich CHO foods and beverages per day. Physical and clinical measures at baseline were not associated with AS-rich CHO intake, although both fasting triglycerides and HDL-C were adversely associated with greater AS-rich CHO intake. Similar associations were observed for AS intake (Table 2). Participants with higher SSB intakes at baseline were more likely to be younger, male, Black American, have less than a high school education, and be less physically active compared with those with lower intakes of SSBs at baseline (see Supplementary material online, Table S1). Body mass index, waist circumference, systolic and diastolic blood pressure, and fasting triglycerides were positively associated with SSB intake, while HDL-C was inversely associated with SSB intake at baseline.
Table 1.
Baseline characteristics (mean ± SE) across quintiles of added sugar–rich carbohydrate food and beverage intake: CARDIA, n = 3154
| Characteristic | Quintiles of baseline AS-rich food and beverage intake (sv/day) | |||||
|---|---|---|---|---|---|---|
| Quintile (n = ) | 1 (n = 631) | 2 (n = 631) | 3 (n = 630) | 4 (n = 632) | 5 (n = 630) | P trend |
| AS-rich CHO, mean (SE), range | 2.9 (0.68) < 3.8 | 4.6 (0.46), 3.8–5.3 | 6.2 (0.49), 5.4–7.0 | 8.5 (0.89), 7.1–10.1 | 14.4 (4.27) > 10.1 | |
| Demographic characteristics | ||||||
| Age, years | 25.5 (0.15) | 25.3 (0.14) | 24.9 (0.14) | 25.1 (0.14) | 24.5 (0.17) | <0.001 |
| Female, % | 73.8 (1.95) | 62.1 (1.84) | 52.4 (1.77) | 48.7 (1.78) | 44.9 (2.16) | <0.001 |
| White, % | 67.6 (2.04) | 62.3 (1.91) | 53.5 (1.84) | 45.7 (1.85) | 37.9 (2.23) | <0.001 |
| >High school, % | 74.7 (1.95) | 70.5 (1.82) | 67.9 (1.75) | 63.1 (1.76) | 54.3 (2.13) | <0.001 |
| Lifestyle characteristics | ||||||
| Physical activity score | 489.9 (12.06) | 440.7 (11.24) | 418.7 (10.81) | 405.8 (10.90) | 374.1 (13.20) | <0.001 |
| Current smoking, % | 21.1 (1.86) | 20.5 (1.74) | 24.3 (1.67) | 26.5 (1.69) | 34.6 (2.05) | <0.001 |
| Current drinking, % | 89.3 (1.46) | 86.5 (1.36) | 87.7 (1.31) | 87.3 (1.32) | 84.6 (1.60) | 0.12 |
| Physical and clinical characteristics | ||||||
| BMI, kg/m2 | 23.9 (0.18) | 24.1 (0.17) | 24.0 (0.17) | 24.2 (0.17) | 23.9 (0.20) | 0.97 |
| Waist circ, cm | 75.8 (0.41) | 76.9 (0.38) | 76.9 (0.37) | 77.4 (0.37) | 76.6 (0.45) | 0.19 |
| Systolic BP, mmHg | 109.1 (0.42) | 110.2 (0.39) | 109.7 (0.38) | 109.9 (0.38) | 109.8 (0.46) | 0.47 |
| Diastolic BP, mmHg | 67.9 (0.39) | 68.5 (0.36) | 68.7 (0.35) | 68.5 (0.35) | 68.1 (0.43) | 0.71 |
| Glucose, mg/100 mL | 81.8 (0.38) | 81.6 (0.35) | 81.3 (0.34) | 81.5 (0.34) | 81.4 (0.41) | 0.45 |
| Triglycerides, mg/dL | 66.5 (1.71) | 67.3 (1.59) | 68.4 (1.53) | 71.5 (1.54) | 71.6 (1.87) | 0.03 |
| HDL-C, mg/dL | 56.2 (0.54) | 54.8 (0.51) | 53.8 (0.49) | 53.3 (0.49) | 52.0 (0.59) | <0.001 |
Adjusted for age, sex, race, education, field centre, and energy intake.
BMI, body mass index; waist circ, waist circumference; BP, blood pressure; HDL-C, HDL cholesterol.
Table 2.
Baseline characteristics (mean ± SE) across quintiles of added sugar intake: CARDIA, n = 3154
| Characteristic | Quintiles of baseline added sugar intake (g/day) | |||||
|---|---|---|---|---|---|---|
| Quintile (n = ) | 1 (n = 630) | 2 (n = 630) | 3 (n = 631) | 4 (n = 630) | 5 (n = 630) | P trend |
| Added sugar, mean (SE), range | 19.4 (5.9) < 30.2 | 39.5 (5.5), 30.3–49.1 | 60.0 (6.8), 49.2–71.6 | 90.7 (12.2), 71.7–114.3 | 178.2 (68.0) > 114.3 | |
| Demographic characteristics | ||||||
| Age, years | 25.5 (0.15) | 25.2 (0.14) | 24.9 (0.14) | 25.1 (0.14) | 24.6 (0.17) | 0.002 |
| Female, % | 61.6 (1.99) | 57.2 (1.86) | 52.1 (1.80) | 51.3 (1.82) | 59.7 (2.17) | 0.20 |
| White, % | 69.5 (2.02) | 61.8 (1.90) | 50.3 (1.84) | 42.4 (1.85) | 43.0 (2.21) | <0.001 |
| >High school, % | 70.7 (1.94) | 71.7 (1.81) | 68.6 (1.75) | 64.1 (1.78) | 55.3 (2.11) | <0.001 |
| Lifestyle characteristics | ||||||
| Physical activity score | 475.8 (12.01) | 427.5 (11.23) | 419.1 (10.86) | 423.3 (10.99) | 383.6 (13.10) | <0.001 |
| Current smoking, % | 26.2 (1.86) | 20.5 (1.74) | 24.1 (1.68) | 26.2 (1.70) | 30.1 (2.03) | 0.06 |
| Current drinking, % | 87.6 (1.45) | 87.9 (1.36) | 88.1 (1.31) | 88.7 (1.33) | 83.3 (1.58) | 0.17 |
| Physical and clinical characteristics | ||||||
| BMI, kg/m2 | 24.1 (0.18) | 24.0 (0.17) | 24.0 (0.17) | 23.8 (0.17) | 24.2 (0.20) | 0.88 |
| Waist circ, cm | 76.6 (0.41) | 76.5 (0.38) | 76.8 (0.37) | 76.5 (0.37) | 77.3 (0.44) | 0.40 |
| Systolic BP, mmHg | 109.4 (0.42) | 109.8 (0.39) | 109.8 (0.38) | 109.8 (0.38) | 109.9 (0.46) | 0.53 |
| Diastolic BP, mmHg | 67.9 (0.39) | 68.2 (0.36) | 68.0 (0.35) | 68.7 (0.35) | 68.8 (0.42) | 0.09 |
| Glucose, mg/100 mL | 81.7 (0.38) | 81.8 (0.35) | 81.6 (0.34) | 81.1 (0.34) | 81.4 (0.41) | 0.38 |
| Triglycerides, mg/dL | 67.6 (1.70) | 65.8 (1.59) | 70.5 (1.53) | 68.1 (1.55) | 73.3 (1.85) | 0.03 |
| HDL-C, mg/dL | 56.4 (0.54) | 55.2 (0.50) | 54.3 (0.49) | 52.8 (0.49) | 51.4 (0.59) | <0.001 |
Adjusted for age, sex, race, education, field centre, and energy intake.
BMI, body mass index; waist circ, waist circumference; BP, blood pressure; HDL-C, HDL cholesterol.
Food and beverage intakes were stratified across quintiles of AS-rich CHO (Table 3), AS (Table 4), and SSB (see Supplementary material online, Table S2) intakes. As shown in Tables 3 and 4, compared with study participants with lower AS-rich CHO or AS intake, respectively, those with higher intakes of AS-rich CHO or AS had lower HEI-2020 scores evident by their lower consumption of whole grain products, fruit and fruit juice, vegetables, legumes, nuts, poultry, fish and seafood, and higher consumption of red/processed meat and dairy products. Those in Q5 consumed significantly more candy, sugar products (honey, syrup, jam, etc.), and refined grain products and desserts compared with those in lower quintiles. Further, those who consumed more AS-rich CHOs and AS consumed more total CHOs, but similar amount of fibre across quintiles; as expected, the CHO:fibre ratio increased across quintiles of AS-rich CHOs and AS. A similar pattern of eating habits was observed across quintiles of SSB intake (see Supplementary material online, Table S2), including lower diet quality among those consuming more SSBs than fewer.
Table 3.
Dietary intake (mean ± SE) stratified across quintiles of intake of added sugar–rich carbohydrate foods and beverages: CARDIA, n = 3154
| Quintiles of averaged (Y0, Y7, Y20) AS-rich CHO food and beverage intake (sv/day) | ||||||
|---|---|---|---|---|---|---|
| Quintile, n = | 1 (n = 631) | 2 (n = 631) | 3 (n = 630) | 4 (n = 632) | 5 (n = 630) | P trend |
| AS-rich CHO, mean (SE), range | 2.9 (0.68) < 3.82 | 4.6 (0.46), 3.8–5.3 | 6.2 (0.49), 5.4–7.0 | 8.5 (0.89), 7.1–10.2 | 14.4 (4.27) > 10.2 | |
| HEI-2020 | 66.9 (0.38) | 64.5 (0.36) | 61.4 (0.34) | 58.6 (0.35) | 55.4 (0.42) | <0.001 |
| Daily nutrient intake | ||||||
| Energy, kcal | 1918 (31.9) | 2173 (30.6) | 2494 (30.2) | 2925 (30.4) | 3951 (32.4) | <0.001 |
| Total fat, g | 97.5 (1.29) | 102.0 (1.20) | 107.7 (1.15) | 115.2 (1.16) | 130.9 (1.41) | <0.001 |
| CHO, g | 264.6 (3.15) | 282.7 (2.94) | 301.7 (2.83) | 331.0 (2.85) | 401.7 (3.45) | <0.001 |
| Protein, g | 93.9 (1.08) | 94.8 (1.01) | 96.3 (0.97) | 101.2 (0.98) | 108.3 (1.18) | <0.001 |
| Added sugar, g | 47.0 (1.67) | 60.2 (1.55) | 72.7 (1.49) | 90.9 (1.51) | 140.3 (1.82) | <0.001 |
| Fibre, g | 21.0 (0.37) | 20.5 (0.34) | 20.4 (0.33) | 20.2 (0.33) | 20.5 (0.40) | 0.35 |
| CHO:fibre ratio | 12.9 (0.23) | 14.7 (0.21) | 16.3 (0.20) | 18.2 (0.20) | 21.0 (0.25) | <0.001 |
| Food group intake, (sv/day) | ||||||
| Candy | 0.1 (0.02) | 0.2 (0.02) | 0.3 (0.02) | 0.3 (0.02) | 0.5 (0.02) | <0.001 |
| Sugar products | 0.4 (0.07) | 0.8 (0.07) | 1.3 (0.07) | 1.9 (0.07) | 3.7 (0.08) | <0.001 |
| SSB | 0.5 (0.05) | 0.7 (0.05) | 1.0 (0.05) | 1.5 (0.05) | 2.5 (0.06) | <0.001 |
| RG desserts | 0.4 (0.03) | 0.5 (0.02) | 0.5 (0.02) | 0.6 (0.02) | 0.9 (0.03) | <0.001 |
| RG quick breads | 1.6 (0.06) | 2.0 (0.05) | 2.5 (0.05) | 3.0 (0.05) | 4.2 (0.06) | <0.001 |
| RG RTE cereal | 0.04 (0.01) | 0.05 (0.01) | 0.07 (0.01) | 0.08 (0.01) | 0.12 (0.01) | <0.001 |
| RG bread (loaf, rolls) | 0.5 (0.03) | 0.6 (0.02) | 0.7 (0.02) | 0.8 (0.02) | 1.1 (0.03) | <0.001 |
| Pasta/rice | 3.2 (0.07) | 3.8 (0.06) | 4.2 (0.06) | 4.8 (0.06) | 6.3 (0.07) | <0.001 |
| Whole grains | 1.9 (0.05) | 1.9 (0.05) | 1.8 (0.05) | 1.7 (0.05) | 1.6 (0.06) | <0.001 |
| Fruit, fruit juice | 2.6 (0.09) | 2.4 (0.08) | 2.2 (0.08) | 2.3 (0.08) | 1.9 (0.10) | <0.001 |
| Vegetables | 4.7 (0.10) | 4.3 (0.09) | 4.2 (0.09) | 4.2 (0.09) | 3.9 (0.11) | <0.001 |
| Legumes | 0.3 (0.01) | 0.2 (0.01) | 0.2 (0.01) | 0.2 (0.01) | 0.2 (0.01) | 0.06 |
| Nuts | 0.9 (0.05) | 0.9 (0.04) | 0.8 (0.04) | 0.8 (0.04) | 0.7 (0.05) | 0.008 |
| Red/process meat | 3.0 (0.08) | 3.1 (0.08) | 3.2 (0.08) | 3.5 (0.08) | 4.3 (0.09) | <0.001 |
| Poultry | 1.5 (0.05) | 1.4 (0.04) | 1.4 (0.04) | 1.3 (0.04) | 1.2 (0.05) | <0.001 |
| Eggs | 0.6 (0.02) | 0.6 (0.02) | 0.6 (0.02) | 0.6 (0.02) | 0.7 (0.02) | 0.05 |
| Fish/seafood | 1.2 (0.05) | 1.1 (0.04) | 1.0 (0.04) | 1.0 (0.04) | 0.9 (0.05) | <0.001 |
| Dairy prod | 2.7 (0.08) | 2.7 (0.07) | 2.8 (0.07) | 2.9 (0.07) | 3.1 (0.08) | <0.001 |
| Diet drinks | 0.7 (0.05) | 0.7 (0.04) | 0.7 (0.04) | 0.5 (0.04) | 0.4 (0.05) | <0.001 |
| Coffee/tea | 1.7 (0.10) | 1.6 (0.09) | 1.6 (0.09) | 2.0 (0.09) | 2.2 (0.11) | 0.001 |
| Alcohol | 1.0 (0.05) | 0.9 (0.05) | 0.8 (0.04) | 0.8 (0.04) | 0.7 (0.05) | 0.002 |
Adjusted for age, sex, race, education, field centre, and energy intake. AS-rich CHOs include candy, sugar products, SSB, RG desserts, RG quick breads, RG RTE cereal, and RG bread products.
AS, added sugar; HEI, Healthy Eating Index; CHO, carbohydrate; RG, refined grain; RTE, ready-to-eat; SSB, sugar-sweetened beverages.
Table 4.
Dietary intake (mean ± SE) across quintiles of added sugar intake: CARDIA, n = 3154
| Quintiles of averaged (Y0, Y7, Y20) added sugar intake (g/day) | ||||||
|---|---|---|---|---|---|---|
| Quintile (n = ) | 1 (n = 630) | 2 (n = 631) | 3 (n = 631) | 4 (n = 631) | 5 (n = 631) | P trend |
| AS, mean (SE), range | 30.5 (8.04) < 41.5 | 49.8 (5.08), 41.5–58.1 | 68.2 (6.16), 58.2–79.2 | 94.7 (10.41), 79.3–115.1 | 167.7 (63.24) > 115.1 | |
| HEI-2020 | 65.0 (0.39) | 63.6 (0.36) | 62.1 (0.35) | 60.2 (0.36) | 56.0 (0.41) | <0.001 |
| Daily nutrient intake | ||||||
| Energy (kcal) | 2006 (33.5) | 2245 (32.0) | 2508 (31.7) | 2862 (32.0) | 3838 (33.4) | <0.001 |
| Total fat, g | 102.8 (1.29) | 104.7 (1.21) | 107.5 (1.18) | 113.3 (1.19) | 125.0 (1.37) | <0.001 |
| CHO, g | 257.1 (2.98) | 284.1 (2.80) | 303.3 (2.72) | 330.3 (2.75) | 406.5 (3.17) | <0.001 |
| Protein, g | 95.9 (1.07) | 96.9 (1.00) | 98.5 (0.98) | 99.5 (0.99) | 103.7 (1.14) | <0.001 |
| Fibre, g | 20.1 (0.36) | 20.6 (0.34) | 21.1 (0.33) | 20.6 (0.33) | 20.2 (0.38) | 0.89 |
| CHO: fibre ratio | 13.2 (0.22) | 14.5 (0.20) | 15.9 (0.20) | 17.7 (0.20) | 21.8 (0.23) | <0.001 |
| Food group intake, (sv/day) | ||||||
| AS-rich CHOs | 0.8 (0.04) | 1.4 (0.03) | 2.0 (0.03) | 2.6 (0.03) | 3.3 (0.04) | <0.001 |
| Candy | 0.1 (0.02) | 0.21 (0.02) | 0.3 (0.01) | 0.4 (0.01) | 0.6 (0.02) | <0.001 |
| Sugar products | 0.9 (0.08) | 1.1 (0.070) | 1.6 (0.07) | 1.8 (0.07) | 2.6 (0.08) | <0.001 |
| SSB | 0.3 (0.05) | 0.5 (0.04) | 0.9 (0.04) | 1.4 (0.04) | 3.1 (0.05) | <0.001 |
| RG desserts | 0.6 (0.03) | 0.6 (0.03) | 0.7 (0.03) | 0.8 (0.03) | 1.1 (0.03) | <0.001 |
| RG quick breads | 3.2 (0.07) | 3.3 (0.07) | 3.3 (0.06) | 3.4 (0.06) | 4.1 (0.07) | <0.001 |
| RG RTE cereals | 0.05 (0.01) | 0.05 (0.01) | 0.07 (0.01) | 0.08 (0.01) | 0.11 (0.01) | <0.001 |
| RG bread products | 0.8 (0.03) | 0.7 (0.02) | 0.7 (0.02) | 0.8 (0.02) | 0.7 (0.03) | 0.68 |
| Pasta/rice | 4.5 (0.07) | 4.5 (0.07) | 4.2 (0.07) | 4.4 (0.07) | 4.5 (0.07) | 0.70 |
| Whole grains | 2.1 (0.05) | 2.0 (0.05) | 1.8 (0.05) | 1.6 (0.05) | 1.4 (0.05) | <0.001 |
| Fruit, fruit juice | 2.4 (0.08) | 2.5 (0.08) | 2.4 (0.08) | 2.4 (0.08) | 1.9 (0.09) | <0.001 |
| Vegetables | 4.7 (0.09) | 4.5 (0.09) | 4.3 (0.09) | 4.2 (0.09) | 3.8 (0.09) | <0.001 |
| Legumes | 0.3 (0.01) | 0.2 (0.01) | 0.2 (0.01) | 0.2 (0.01) | 0.1 (0.01) | <0.001 |
| Nuts | 0.9 (0.05) | 1.0 (0.04) | 0.8 (0.04) | 0.8 (0.04) | 0.7 (0.05) | <0.001 |
| Red/process meat | 3.1 (0.08) | 3.3 (0.08) | 3.3 (0.08) | 3.6 (0.08) | 3.8 (0.08) | <0.001 |
| Poultry | 1.5 (0.05) | 1.4 (0.04) | 1.3 (0.04) | 1.2 (0.04) | 1.3 (0.05) | <0.001 |
| Eggs | 0.6 (0.02) | 0.6 (0.02) | 0.6 (0.02) | 0.5 (0.02) | 0.5 (0.02) | 0.02 |
| Fish/seafood | 1.1 (0.04) | 1.1 (0.04) | 1.0 (0.04) | 1.0 (0.04) | 0.9 (0.04) | <0.001 |
| Dairy prod | 3.0 (0.07) | 2.9 (0.07) | 2.8 (0.07) | 2.8 (0.07) | 2.6 (0.07) | <0.001 |
| Diet drinks | 0.9 (0.04) | 0.6 (0.04) | 0.6 (0.04) | 0.5 (0.04) | 0.5 (0.04) | <0.001 |
| Coffee, tea | 2.1 (0.09) | 1.9 (0.09) | 1.7 (0.0) | 1.8 (0.09) | 1.6 (0.09) | 0.001 |
| Alcohol | 0.9 (0.05) | 0.8 (0.04) | 0.8 (0.04) | 0.9 (0.04) | 0.8 (0.05) | 0.60 |
Adjusted for age, sex, race, education, field centre, and energy intake. AS-rich CHOs include candy, sugar products, SSB, RG desserts, RG quick breads, RG RTE cereal, and RG bread products.
AS, added sugar; HEI, Healthy Eating Index; CHO, carbohydrate; SSB, sugar-sweetened beverages; RG, refined grain; RTE, ready to eat.
The prevalence of MetS components by Y30 stratified across quintiles of averaged AS-rich CHOs, AS, and SSBs is shown in Table 5. Low HDL-C, but not abdominal obesity, elevated blood pressure, elevated glucose or diabetes, or elevated triglycerides, was significantly associated across increasing quintiles of AS-rich CHO food and beverage intake. Similar results were observed for AS intake. All MetS components, except elevated glucose/diabetes, were significantly associated across increasing quintiles of SSBs, although the % with elevated blood pressure was inverse to SSB intake.
Table 5.
Prevalence of metabolic syndrome components by Year 30 across quintiles of added sugar–rich carbohydrate foods and beverages, added sugar, and sugar-sweetened beverage intakes: CARDIA, n = 3154
| Quintiles of averaged daily dietary intake | ||||||
|---|---|---|---|---|---|---|
| Quintile | 1 | 2 | 3 | 4 | 5 | P trend |
| Quintiles of AS-rich food and beverage intake (sv/day) | ||||||
| n | 631 | 631 | 630 | 632 | 630 | |
| AS-rich CHO, mean (SE), range | 2.9 (0.68) < 3.8 | 4.6 (0.46), 3.8–5.4 | 6.2 (0.49), 5.5–7.1 | 8.5 (0.89), 7.2–10.1 | 14.4 (4.27) > 10.1 | |
| Metabolic syndrome components, % (SE) | ||||||
| Abdominal obesity | 46.3 (2.21) | 49.1 (2.03) | 50.9 (1.94) | 54.3 (2.00) | 48.4 (2.43) | 0.27 |
| Elevated BP | 74.5 (1.95) | 72.9 (1.80) | 72.7 (1.72) | 73.4 (1.74) | 74.6 (2.14) | 0.91 |
| Elevated glucose or diabetes | 10.7 (1.39) | 11.4 (1.28) | 9.7 (1.22) | 9.3 (1.24) | 11.7 (1.52) | 0.99 |
| Elevated triglycerides | 14.9 (1.65) | 16.2 (1.52) | 13.4 (1.45) | 18.1 (1.47) | 16.4 (1.80) | 0.45 |
| Low HDL-C | 13.2 (1.74) | 18.1 (1.60) | 17.3 (1.53) | 18.4 (1.55) | 22.8 (1.91) | 0.004 |
| Quintiles of added sugar intake (g/day) | ||||||
| n | 630 | 631 | 631 | 631 | 631 | |
| AS, mean, range | 30.5 (8.04) < 41.5 | 49.8 (5.08), 41.5–58.1 | 68.2 (6.16), 58.2–79.2 | 94.7 (10.41), 79.3–115.1 | 167.6 (63.24) > 115.1 | |
| Metabolic syndrome components, % (SE) | ||||||
| Abdominal obesity | 49.7 (2.13) | 47.6 (2.00) | 49.9 (1.94) | 50.4 (1.96) | 51.3 (2.29) | 0.44 |
| Elevated BP | 73.1 (1.88) | 74.1 (1.77) | 73.8 (1.72) | 73.1 (1.73) | 73.9 (2.02) | 0.93 |
| Elevated glucose or diabetes | 11.0 (1.33) | 10.6 (1.25) | 10.8 (1.22) | 8.1 (1.23) | 12.2 (1.43) | 0.95 |
| Elevated triglycerides | 17.1 (1.59) | 14.3 (1.49) | 15.3 (1.45) | 17.2 (1.46) | 15.2 (1.70) | 0.88 |
| Low HDL-C | 13.8 (1.68) | 17.8 (1.58) | 17.4 (1.53) | 17.4 (1.54) | 23.3 (1.80) | 0.003 |
| Quintiles of SSB Intake (sv/day) | ||||||
| n | 631 | 628 | 633 | 631 | 631 | |
| SSB, mean (SE), range | 0.1 (0.06) < 0.2 | 0.4 (0.10), 0.2–0.5 | 0.8 (0.15), 0.6–1.07 | 1.5 (0.26), 1.08–1.9 | 3.5 (1.99) > 1.9 | |
| Metabolic syndrome components, % (SE) | ||||||
| Abdominal obesity | 41.8 (2.11) | 57.8 (1.98) | 52.5 (1.94) | 52.9 (1.98) | 53.9 (2.16) | <0.001 |
| Elevated BP | 77.3 (1.87) | 77.6 (1.75) | 72.8 (1.71) | 69.4 (1.75) | 71.1 (1.90) | 0.002 |
| Elevated glucose or diabetes | 10.2 (1.33) | 11.9 (1.25) | 9.7 (1.22) | 10.2 (1.24) | 10.8 (1.35) | 0.91 |
| Elevated triglycerides | 12.5 (1.58) | 15.2 (1.48) | 17.6 (1.44) | 16.2 (1.47) | 17.5 (1.60) | 0.05 |
| Low HDL-C | 13.5 (1.67) | 16.9 (1.57) | 19.0 (1.53) | 18.8 (1.56) | 21.6 (1.70) | 0.002 |
Adjusted for age, sex, race, education, field centre, energy intake, smoking status, status of drinking alcohol, physical activity, and the appropriate diet pattern score: AS-rich CHO was adjusted for the western diet pattern without AS-rich CHO foods and beverages; AS was adjusted for HEI without AS; and SSB was adjusted for the western diet pattern without SSB.
AS, added sugar; CHO, carbohydrate; BP, blood pressure; HDL-C, HDL cholesterol; SSB, sugar-sweetened beverages.
The risk of incident MetS by Y30 stratified across quintiles of AS-rich CHOs, AS, and SSBs intakes is shown in Table 6. After adjusting for demographic and lifestyle characteristics (Model 2), those in Q5 of AS-rich CHO consumption had a 59% higher risk of developing MetS compared with those in Q1 (P < 0.001). Likewise, higher intake of AS (Q5) was associated with a 44% higher risk of developing MetS compared with lower AS intake (Q1; P = 0.01). Finally, compared with no or rare intake of SSB (Q1), greater intake of SSB (Q5) was associated with a 34% higher risk of incident MetS by Y30 (P = 0.03). Further adjustment for BMI and menopause (Model 3) did not substantially change the risk estimates.
Table 6.
Risk (hazard ratio, 95% confidence interval) of incident metabolic syndrome by Year 30 across intakes of added sugar–rich carbohydrate foods and beverages, added sugar, and sugar-sweetened beverages: CARDIA, n = 3154
| Quintiles of averaged (Y0, Y7, Y20) daily dietary intake | ||||||
|---|---|---|---|---|---|---|
| Quintile | 1 | 2 | 3 | 4 | 5 | P trend |
| Quintiles of AS-rich food and beverage intake (sv/day) | ||||||
| AS-rich CHO, mean | 2.9 | 4.6 | 6.2 | 8.5 | 14.4 | |
| n | 631 | 631 | 630 | 632 | 630 | |
| MetS cases | 211 | 261 | 258 | 294 | 283 | |
| Model 1 | 1 | 1.42 (1.16, 1.75) | 1.54 (1.25, 1.89) | 2.03 (1.64, 2.51) | 2.53 (1.98, 3.24) | <0.001 |
| Model 2 | 1 | 1.23 (1.00, 1.51) | 1.13 (0.91, 1.40) | 1.49 (1.19, 1.87) | 1.59 (1.22, 2.06) | <0.001 |
| Model 3 | 1 | 1.24 (1.01, 1.53) | 1.17 (0.94, 1.45) | 1.49 (1.18, 1.87) | 1.51 (1.16, 1.97) | 0.001 |
| Quintiles of added sugar intake (g/day) | ||||||
| AS, mean | 30.5 | 49.8 | 68.2 | 94.7 | 167.6 | |
| n | 630 | 631 | 631 | 631 | 631 | |
| MetS cases | 236 | 241 | 270 | 256 | 304 | |
| Model 1 | 1 | 0.97 (0.79, 1.19) | 1.12 (0.92, 1.37) | 0.98 (0.79, 1.22) | 1.63 (1.30, 2.04) | <0.001 |
| Model 2 | 1 | 0.95 (0.77, 1.17) | 1.09 (0.78, 1.34) | 0.92 (0.74, 1.15) | 1.44 (1.14, 1.81) | 0.01 |
| Model 3 | 1 | 0.98 (0.80, 1.21) | 1.10 (0.90, 1.35) | 0.96 (0.77, 1.21) | 1.51 (1.20, 1.91) | 0.003 |
| Quintiles of sugar-sweetened beverage intake (sv/day) | ||||||
| SSB, mean | 0.07 | 0.37 | 0.79 | 1.45 | 3.49 | |
| n | 631 | 628 | 633 | 631 | 631 | |
| MetS cases | 213 | 238 | 274 | 277 | 305 | |
| Model 1 | 1 | 1.19 (0.95, 1.47) | 1.26 (1.02, 1.55) | 1.28 (1.03, 1.59) | 1.62 (1.30, 2.02) | <0.001 |
| Model 2 | 1 | 1.15 (0.93, 1.42) | 1.14 (0.93, 1.40) | 1.14 (0.92, 1.42) | 1.34 (1.06, 1.69) | 0.03 |
| Model 3 | 1 | 1.21 (0.98, 1.50) | 1.12 (0.91, 1.38) | 1.21 (0.97, 1.51) | 1.38 (1.09, 1.75) | 0.02 |
Model 1 adjusted for age, sex, race, field centre, education, and energy intake. Model 2 adjusted for Model 1 plus physical activity, current smoking, drinking status, and appropriate diet pattern score: AS-rich foods/beverages models were adjusted for the western diet pattern without the AS-rich products; AS was adjusted for HEI without AS; and SSB was adjusted for the western diet pattern without SSB. Model 3 adjusted for Model 2 plus BMI at baseline and menopausal status at Year 30.
In sensitivity analysis, we revised the AS-rich CHO food group to include dairy desserts and then examined the association between the revised AS-rich food and beverage group and incident MetS. Results for the association between the revised AS food and beverage group and MetS did not substantially change (from Table 6) as shown for Model 2: Q1: 1.0; Q2: 1.22 (0.99, 1.51); Q3: 1.13 (0.91, 1.40); Q4: 1.53 (1.22, 1.91); and Q5: 1.59 (1.22, 2.06); Ptrend < 0.001.
Discussion
Consistent with our study hypothesis, higher intakes of AS-rich CHOs, AS, and SSBs were significant predictors of incident MetS over 30 years of follow-up. Individuals who consumed greater amounts of AS-rich CHO foods, AS, and SSBs had lower HEI-2020 scores and lower intake of nutrient dense foods compared with those who consumed fewer AS-rich CHO foods and beverages. Intakes of AS-rich CHOs, AS, and SSBs at baseline were greater among younger participants, Black American participants, and current smokers. While AS-rich CHO intake was greater among men, intakes of AS and SSBs were greater among women. Greater AS-rich CHO foods, AS, and SSB intakes at baseline were associated with being less physically active. Furthermore, the three exposures significantly predicted the prevalence of low HDL-C at follow-up Y30. Only high SSB intake was associated with prevalent abdominal obesity, elevated blood pressure, and triglycerides.
Although total CHO intake has been associated with higher risk of MetS in previously published studies,4 not all CHOs are associated with worse health outcomes.3,5 Differentiating the type or quality of CHO as well as identifying the food and beverage source is important to better understand their health benefits or risks.3,5 Few studies have examined the association of total AS intake or consumption of AS-rich CHO foods and beverages with MetS, except for SSBs that only account for about 35% of total AS intake.6 We observed, during 30 years of follow-up in the current study, higher risk of incident MetS with greater intakes of AS-rich CHOs, AS, and SSBs. In the China Health and Nutrition Survey, investigators reported a 9% increased risk of incident MetS in women who consumed 5–20 g of free sugar each day compared with lower free sugar intake, while this association was not significant in men.21 But, interestingly, adults consuming >20 g of free sugar per day were not at higher risk of developing MetS21; this finding may potentially be explained by the small cell size in the highest free sugar group. In a community-based cross-sectional study of over 7000 Korean men and women, men consuming >20% of energy from total AS had 33% higher odds of MetS than men consuming less; however, this association was not observed in women.22 Examining food sources of fructose in a meta-analysis of 13 prospective studies showed SSB intake positively associated with MetS, while fructose-containing yogurt was inversely associated with incident MetS.23 Although yogurt may contain AS, i.e. fructose as an ingredient, other ingredients in yogurt such as dairy fatty acids and protein promote health benefits.24 A study conducted in US adults revealed a 15% lower odds of MetS among chocolate candy consumers compared with nonconsumers; however, the statistical models were not adjusted for lifestyle factors or dietary intake that, most likely, would attenuate the association.25
The evidence from prospective observational studies has been inconsistent for the SSB-incident MetS association.26–31 Notably, in our current study with 30 years of follow-up, we observed a 34% higher risk of incident MetS among adults consuming two or more daily servings of SSBs. Consistent with our findings, one recent meta-analysis of seven prospective studies reported an increased risk of incident MetS with greater SSB consumption,32 while another meta-analysis of four prospective studies reported a null association.33 The difference in these findings between the two meta-analyses is likely due to the inclusion of different studies and study populations as well as the analytical methods. Not unexpected, results from cross-sectional studies have also been inconsistent, but meta-analyses of these studies resulted in a significant and positive association between SSB intake and prevalent MetS.32–34
Added sugar and AS-rich CHO foods and beverages influence the development of some MetS components. Meta-analysis of 39 randomized clinical trials designed to test the effect of AS on lipids and blood pressure demonstrated increased triglycerides, total and LDL cholesterol, and systolic and diastolic blood pressure with higher intake of AS compared with lower.35 Further, numerous studies have reported weight gain and development of obesity among consumers of SSB.12 Results have generally been consistent among published studies of AS and SSB intakes promoting adverse lipid concentrations,12,28,29,31,32,35 although not all study results are consistent for abdominal obesity, low HDL-C, elevated triglycerides, elevated blood pressure, and elevated glucose.36–39 Differing results may be due to variation in study design (cross-sectional, prospective, or clinical study design), study population (healthy vs. patient population and age group), different diet assessment instruments and definitions of exposures and outcomes (such as prevalent or incident outcomes), and statistical methods. In the current prospective study, the prevalence of low HDL-C was associated with higher consumption of AS-rich CHOs, AS, and SSB, but only SSB intake was significantly associated with prevalent abdominal obesity and triglyceride levels. However, elevated blood pressure was inversely related to SSB intake in this cohort of adults, which was unexpected. In another prospective study, higher waist circumference and lower HDL-C were observed among those consuming >10% of energy from AS compared with those consuming <10%.33 Notably, <10% of AS from total energy intake or limited free sugar intake are general country guidelines.6,7
Finally, excessive consumption of these AS-rich CHO foods and beverages may result in a higher risk for MetS and its components due to overall poor diet quality,40 as opposed to healthy diet patterns, including a Mediterranean, dietary approach to stop hypertension (DASH), Mediterranean-DASH intervention for neurodegenerative delay, or other plant-based diet pattern, that protects cardiometabolic health.41,42 High AS intake in this study and others was related to poor diet quality, as represented by a lower HEI2015 score, characterized by low intakes of whole grain products, fruit, and vegetables, thus resulting in lower intakes of fibre and micronutrients (i.e. iron, zinc, magnesium, folate, and other vitamins and minerals).40,43 Further evidence of the importance of CHO composition is shown in a study about CHO to fibre ratio, especially if the CHO:fibre ratio is high (high CHO and low fibre intake).44 High CHO:fibre ratio, commonly linked with high AS intake, has been associated with higher risk of MetS and insulin resistance compared with lower CHO:fibre ratio (lower CHO and high fibre intake).44 This type of low-quality diet pattern characterizes a ‘western diet’ pattern and is known to promote gut dysbiosis, a condition that disrupts intestinal homeostasis resulting in metabolic disorders, obesity, dyslipidaemia, inflammation, and insulin resistance.45–48 This could explain why, in our study, we found higher intake of AS-rich CHOs consisting of CHOs low in fibre (high CHO:fibre ratio) that was significantly associated with MetS even after adjusting for BMI (Table 6, Model 3).
Limitations and strengths
A limitation of this study is the use of self-reported dietary intake, generally known for measurement error, but the CARDIA diet history is validated and more detailed and comprehensive than a food frequency questionnaire.17,49 Furthermore, all dietary interviewers were trained and certified to administer the CARDIA diet history.17 Another limitation may be under-reporting AS-rich food intake, which may lead to misclassification. However, results from a clinical study demonstrated under-reporting of snack foods in all study participants;50 which may potentially result in attenuation of the risk estimate. Finally, residual confounding may be present even though statistical models were adjusted for numerous potential factors that may confound the associations between dietary intake and MetS.
Strengths of this study include its prospective study design including data from 30 years of follow-up, the large sample of a biracial population, including men and women enrolled in the study as young adults, and three repeated measures of dietary intake. The CARDIA diet history uses NDSR, a diet data entry software program that includes brand name information in the database allowing specific and more accurate estimation of nutrient intake, including AS intake using the same definition as the dietary guidelines.6 Further, total AS intake was quantified in our study, whereas many other studies assessing dietary intake by food frequency questionnaire are only able to quantify SSB consumption, which accounts for about one-third of total AS consumption.6 Our findings emphasize that the majority of remaining AS intake comes from other food sources [i.e. refined grain products and snack foods, i.e. sweet bakery products, ready-to-eat cereal, candy, and sugar products (sugar, honey, syrup, jams, etc.)].6,9,10 Finally, targeting type or quality of CHOs,5 those that are easily metabolized, i.e. intakes of AS-rich CHO foods and beverages, as opposed to total CHO intake, associated with higher risk of MetS is novel.
Study implications
Although the main functions of AS in various forms in food products include preservation, fermentation, text, flavour, and colour,51 we do not foresee food manufacturers changing their recipes anytime soon, such as reducing AS in food products. Consequently, the responsibility for selecting healthy foods and beverages remains with the consumer. In the USA and other countries, governments require food labelling, such as placing the amount of AS for a serving of food or beverage product on the nutrition facts panel. In many countries, warning labels or front-of-pack nutrition labels are appearing on food packages, providing a clear message about the healthfulness of a food or beverage product.52 Given the availability of nutrition facts panels, front-of-pack labels, or warning labels on packaging, consumer education interventions have shown positive results for consumer understanding and using nutrition labels when purchasing groceries.53 Perhaps with more knowledge about the nutrient composition of foods and beverages, consumers may select healthier food and beverage products in the marketplace.
Conclusions
Our results support our study hypothesis that greater intakes of AS-rich CHO foods and beverages and AS among Black American and White American men and women were associated with higher risk of developing MetS. These results support the US 2020–25 Dietary Guidelines as well as World Health Organization (WHO) Nutrition Guidelines for CHO to limit daily intake of AS and free sugars, respectively.6,7 Further research of CHO composition and quality is warranted to better understand the complexities between type of CHO and cardiometabolic risk factors.
Supplementary Material
Acknowledgements
We thank the participants, staff, and investigators of the CARDIA study for their dedication and highly valued contributions.
Contributor Information
Rae K Goins, Division of Epidemiology and Community Health, University of Minnesota School of Public Health, 300 West Bank Office Building 1300 S. 2nd St., Minneapolis, MN, USA.
Lyn M Steffen, Division of Epidemiology and Community Health, University of Minnesota School of Public Health, 300 West Bank Office Building 1300 S. 2nd St., Minneapolis, MN, USA.
So-Yun Yi, Division of Epidemiology and Community Health, University of Minnesota School of Public Health, 300 West Bank Office Building 1300 S. 2nd St., Minneapolis, MN, USA.
Xia Zhou, Division of Epidemiology and Community Health, University of Minnesota School of Public Health, 300 West Bank Office Building 1300 S. 2nd St., Minneapolis, MN, USA.
Linda Van Horn, Department of Preventive Medicine, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA.
James M Shikany, Division of Preventive Medicine, University of Alabama at Birmingham School of Medicine, Birmingham, AL, USA.
James G Terry, Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, USA.
David R Jacobs, Division of Epidemiology and Community Health, University of Minnesota School of Public Health, 300 West Bank Office Building 1300 S. 2nd St., Minneapolis, MN, USA.
Supplementary material
Supplementary material is available at European Journal of Preventive Cardiology.
Author contributions
L.M.S., R.K.G., S.-Y.Y., L.V.H., J.M.S., and D.R.J. designed and conducted the research. L.M.S., R.K.G., S.-Y.Y., and J.G.T. drafted the manuscript. X.Z., L.M.S., and S.-Y.Y. analysed the data. L.M.S., R.K.G., S.-Y.Y., X.Z., L.V.H., J.M.S., J.G.T., and D.R.J. had primary responsibility for final content. All authors read and approved the final manuscript.
Funding
The Coronary Artery Risk Development in Young Adults (CARDIA) study is supported by the National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute (NHLBI) in collaboration with the University of Alabama at Birmingham (HHSN268201800005I and HHSN268201800007I), Northwestern University (HHSN268201800003I), University of Minnesota (HHSN268201800006I), and Kaiser Foundation Research Institute (HHSN268201800004I). This work was also supported by the National Institutes of Health (NIH)/NHLBI grants R21 HL135300 and R01 HL150053 to L.M.S.
Ethics approval and consent to participate
The institutional review board at each institution reviewed and approved the study protocols. All study participants gave written consent prior to taking part in each exam.
Data availability
Data described in the manuscript, data collection forms, and code book will be made available upon request pending application to the CARDIA Coordinating Center (https://www.cardia.dopm.uab.edu/).
References
- 1. Alberti KGMM, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Donato KA, et al. Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; And International Association for the Study of Obesity. Circulation 2009;120:1640–1645. [DOI] [PubMed] [Google Scholar]
- 2. Yang C, Jia X, Wang Y, Fan J, Zhao C, Yang Y, et al. Trends and influence factors in the prevalence, intervention, and control of metabolic syndrome among US adults, 1999–2018. BMC Geriatr 2022;22:979. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Castro-Barquero S, Ruiz-León AM, Sierra-Pérez M, Estruch R, Casas R. Dietary strategies for metabolic syndrome: a comprehensive review. Nutrients 2020;12:2983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Liu YS, Wu QJ, Xia Y, Zhang JY, Jiang YT, Chang Q, et al. Carbohydrate intake and risk of metabolic syndrome: a dose-response meta-analysis of observational studies. Nutr Metab Cardiovasc Dis 2019;29:1288–1298. [DOI] [PubMed] [Google Scholar]
- 5. Campos V, Tappy L, Bally L, Sievenpiper JL, Lê KA. Importance of carbohydrate quality: what does it mean and how to measure it? J Nutr 2022;152:1200–1206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. U.S. Department of Agriculture and U.S . Department of Health and Human Services. Dietary Guidelines for Americans, 2020–2025. 9th Edition. December 2020. Available at DietaryGuidelines.gov. Accessed 10 Dec 2022.
- 7. Carbohydrate intake for adults and children . WHO Guideline. Geneva: World Health Organization; 2023. Licence: CC BY-NC-SA 3.0 IGO. [PubMed] [Google Scholar]
- 8. National Center for Health Statistics, National Health and Nutrition Examination Survey, and U.S . Department of Agriculture, Agriculture Research Service. Beltsville Human Nutrition Research Center, Food Surveys Research Group, What We Eat in America. Table McrNutr. Mean macronutrient intake among adults aged 20 and over, by sex and age: United States, selected years 1988–1994 through 2015–2018. www.cdc.gov/nchs/data/hus/2020-2021/McrNutr.pdf Accessed 03/10/2023.
- 9. Center for Disease Control and Prevention, National Center for Health Statistics; National Health and Examination and Nutrition Survey 2015–2016 . NCHS Fact Sheet: NCHS Nutrition Data, January 2019. https://www.cdc.gov/nchs/about/factsheets/factsheet_nutrition.htm#. Accessed 03/10/2023.
- 10. Ricciuto L, Fulgoni VL III, Gaine PC, Scott MO, DiFrancesco L. Sources of added sugars intake among the U.S. population: analysis by selected sociodemographic factors using the National Health and Nutrition Examination Survey 2011–18. Front Nutr 2021;8:687643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Cho Y-A, Choi J-H. Association between carbohydrate intake and the prevalence of metabolic syndrome in Korean women. Nutrients 2021;13:3098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Malik VS, Hu FB. Sugar-sweetened beverages and cardiometabolic health: an update of the evidence. Nutrients 2019;11:1840. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Semnani-Azad Z, Khan TA, Blanco Mejia SB, de Souza RJ, Leiter LA, Kendall CWC, et al. Association of major food sources of fructose-containing sugars with incident metabolic syndrome: a systematic review and meta-analysis. JAMA Netw Open 2020;3:e209993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Rodríguez LA, Madsen KA, Cotterman C, Lustig RH. Added sugar intake and metabolic syndrome in US adolescents: cross-sectional analysis of the National Health and Nutrition Examination Survey 2005–2012. Publ Hlth Nutr 2016;19:2424–2434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Malik VS, Hu FB. The role of sugar-sweetened beverages in the global epidemics of obesity and chronic diseases. Nat Rev Endocrinol 2022;18:205–218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Friedman GD, Cutter GR, Donahue RP, Hughes GH, Hulley SB, Jacobs DR, et al. CARDIA: study design, recruitment, and some characteristics of the examined subjects. J Clin Epidemiol 1988;41:1105–1116. [DOI] [PubMed] [Google Scholar]
- 17. McDonald A, Van Horn L, Slattery M, Hilner J, Bragg C, Caan B, et al. The CARDIA dietary history: development, implementation, and evaluation. J Am Diet Assoc 1991;91:1104–1112. [PubMed] [Google Scholar]
- 18. Steffen LM, Kroenke CH, Yu X, Pereira MA, Slattery ML, Van Horn L, et al. Associations of plant food, dairy product, and meat intakes with 15-y incidence of elevated blood pressure in young black and white adults: the Coronary Artery Risk Development in Young Adults (CARDIA) study. Am J Clin Nutr 2005;82:1169–1177. [DOI] [PubMed] [Google Scholar]
- 19. Jacobs DR, Hahn LP, Haskell WL, Pirie P, Sidney S. Validity and reliability of short physical activity history: CARDIA and the Minnesota Heart Health Program. J Cardiopulm Rehabil 1989;9:448–459. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Shams-White MM, Pannucci TE, Lerman JL, Herrick KA, Zimmer M, Meyers Mathieu K, et al. Healthy Eating Index-2020: review and update process to reflect the Dietary Guidelines for Americans,2020–2025. J Acad Nutr Diet 2023;123:1280–1288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Pan F, Wang Z, Wang H, Su C, Zhang J, Du W, et al. Association between free sugars intake and risk of metabolic syndrome in Chinese adults: results from the China Health and Nutrition Survey, 2000–2018. Nutrients 2022;14:5385. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Seo EH, Kim H, Kwon O. Association between total sugar intake and metabolic syndrome in middle-aged Korean men and women. Nutrients 2019;11:2042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Semnani-Azad Z, Khan TA, Blanco Mejia S, de Souza RJ, Leiter LA, Kendall CWC, et al. Association of major food sources of fructose-containing sugars with incident metabolic syndrome: a systematic review and meta-analysis. JAMA Netw Open 2020;3:e209993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Wang H, Livingston KA, Fox CS, Meigs JB, Jacques PF. Yogurt consumption is associated with better diet quality and metabolic profile in American men and women. Nutr Res 2013;33:18–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. O'Neil CE, Fulgoni VL, Nicklas TA. Candy consumption was not associated with body weight measures, risk factors for cardiovascular disease, or metabolic syndrome in US adults: NHANES 1999–2004. Nutr Res 2011;31:122–130. [DOI] [PubMed] [Google Scholar]
- 26. Kang Y, Kim J. Soft drink consumption is associated with increased incidence of the metabolic syndrome only in women. Br J Nutr 2017;117:315–324. [DOI] [PubMed] [Google Scholar]
- 27. Lutsey PL, Steffen LM, Stevens J. Dietary intake and the development of the metabolic syndrome: the Atherosclerosis Risk in Communities study. Circulation 2008;117:754–761. [DOI] [PubMed] [Google Scholar]
- 28. Duffey KJ, Gordon-Larsen P, Steffen LM, Jacobs DR, Popkin BM. Drinking caloric beverages increases the risk of adverse cardiometabolic outcomes in the Coronary Artery Risk Development in Young Adults (CARDIA) study. Am J Clin Nutr 2010;92:954–959. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Barrio-Lopez MT, Martinez-Gonzalez MA, Fernandez-Montero A, Beunza JJ, Zazpe I, Bes-Rastrollo M. Prospective study of changes in sugar-sweetened beverage consumption and the incidence of the metabolic syndrome and its components: the SUN cohort. Br J Nutr 2013;110:1722–1731. [DOI] [PubMed] [Google Scholar]
- 30. Ferreira-Pêgo C, Babio N, Bes-Rastrollo M, Corella D, Estruch R, Ros E, et al. Frequent consumption of sugar- and artificially-sweetened beverages and natural and bottled fruit juices is associated with an increased risk of metabolic syndrome in a Mediterranean population at high cardiovascular disease risk. J Nutr 2016;146:1528–1536. [DOI] [PubMed] [Google Scholar]
- 31. Dhingra R, Sullivan L, Jacques PF, Wang TJ, Fox CS, Meigs JB, et al. Soft drink consumption and risk of developing cardiometabolic risk factors and the metabolic syndrome in middle-aged adults in the community. Circulation 2007;116:480–488. [DOI] [PubMed] [Google Scholar]
- 32. Muñoz-Cabrejas A, Guallar-Castillón P, Laclaustra M, Sandoval-Insausti H, Moreno-Franco B. Association between sugar-sweetened beverage consumption and the risk of the metabolic syndrome: a systematic review and meta-analysis. Nutrients 2023;15:430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Narain A, Kwok CS, Mamas MA. Soft drink intake and the risk of metabolic syndrome: a systematic review and meta-analysis. Int J Clin Pract 2017;71:e12927. [DOI] [PubMed] [Google Scholar]
- 34. Malik VS, Popkin BM, Bray GA, Després JP, Willett WC, Hu FB. Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes: a meta-analysis. Diabetes Care 2010;33:2477–2483. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Te Morenga LA, Howatson AJ, Jones RM, Mann J. Dietary sugars and cardiometabolic risk: systematic review and meta-analyses of randomized controlled trials of the effects on blood pressure and lipids. Am J Clin Nutr 2014;100:65–79. [DOI] [PubMed] [Google Scholar]
- 36. Welsh JA, Sharma A, Abramson JL, Vaccarino V, Gillespie C, Vos MB. Caloric sweetener consumption and dyslipidemia among US adults. JAMA 2010;303:1490–1497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Vorster HH, Kruger A, Wentzel-Viljoen E, Kruger HS, Margetts BM. Added sugar intake in South Africa: findings from the Adult Prospective Urban and Rural Epidemiology cohort study. Am J Clin Nutr 2014;99:1479–1486. [DOI] [PubMed] [Google Scholar]
- 38. Siqueira JH, Pereira TSS, Moreira AD, Diniz MFHS, Velasquez-Melendez G, Fonseca MJM, et al. Consumption of sugar-sweetened soft drinks and risk of metabolic syndrome and its components: results of the ELSA-Brasil study (2008–2010 and 2012–2014). J Endocrinol Invest 2023;46:159–171. [DOI] [PubMed] [Google Scholar]
- 39. Dunnigan MG, Fyfe T, McKiddie MT, Crosbie SM. The effects of isocaloric exchange of dietary starch and sucrose on glucose tolerance, plasma insulin and serum lipids in man. Clin Sci 1970;38:1–9. [DOI] [PubMed] [Google Scholar]
- 40. Louie JCY, Tapsell LC. Association between intake of total vs added sugar on diet quality: a systematic review. Nutr Rev 2015;73:837–857. [DOI] [PubMed] [Google Scholar]
- 41. Rosato V, Temple NJ, La Vecchia C, Castellan G, Tavani A, Guercio V. Mediterranean diet and cardiovascular disease: a systematic review and meta-analysis of observational studies. Eur J Nutr 2017;58:173–191. [DOI] [PubMed] [Google Scholar]
- 42. AkbarZ FS, Ouagueni A, Alalwani J, Sukik A, Al-Jayyousi GF, Bassil M, et al. Associations of the MIND diet with cardiometabolic diseases and their risk factors: a systematic review. Diabetes Metab Syndr Obes 2023;16:3353–3371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. O'Neil CE, Nicklas TA, Zanovec M, Cho S. Whole-grain consumption is associated with diet quality and nutrient intake in adults: the National Health and Nutrition Examination Survey, 1999–2004. J Am Diet Assoc 2010;110:1461–1468. [DOI] [PubMed] [Google Scholar]
- 44. Bernardino C, McLellan KC, Corrente JE, Burini RC. Carbohydrate intake and dietary fiber ratio as diagnostic tool for diet quality and prognostics of insulin resistance in Brazilian free-living adults. Int J Food Sci Nutr Diet 2016;5:319–324. [Google Scholar]
- 45. Arnone D, Chabot C, Heba AC, Kökten T, Caron B, Hansmannel F, et al. Sugars and gastrointestinal health. Clin Gastroenterol Hepatol 2022;20:1912–1924. e7. [DOI] [PubMed] [Google Scholar]
- 46. Thomas MS, Blesso CN, Calle MC, Chun OK, Puglisi M, Fernandez ML. Dietary influences on gut microbiota with a focus on metabolic syndrome. Metab Syndr Relat Disord 2022;20:429–439. [DOI] [PubMed] [Google Scholar]
- 47. Garcia K, Gerreira G, Reis F, Viana S. Impact of dietary sugars on gut microbiota and metabolic health. Diabetology 2022;3:549–560. [Google Scholar]
- 48. Santos-Marcos JA, Perez-Jimenez F, Camargo A. The role of diet and intestinal microbiota in the development of metabolic syndrome. J Nutr Biochem 2019;70:1–27. [DOI] [PubMed] [Google Scholar]
- 49. Liu K, Slattery M, Jacobs D, Cutter G, McDonald A, Van Horn L, et al. A study of the reliability and comparative validity of the CARDIA dietary history. Ethn Dis 1994;4:15–27. [PubMed] [Google Scholar]
- 50. Poppitt SD, Swann D, Black AE, Prentice AM. Assessment of selective under-reporting of food intake by both obese and non-obese women in a metabolic facility. Int J Obes 1998;22:303–311. [DOI] [PubMed] [Google Scholar]
- 51. Clemens RA, Jones JM, Kern M, Lee SY, Mayhew EJ, Slavin JL, et al. Functionality of sugars in foods and health. Compr Rev Food Sci Food Saf 2016;15:433–470. [DOI] [PubMed] [Google Scholar]
- 52. Kanter R, Vanderlee L, Vandevijvere S. Front-of-package nutrition labelling policy: global progress and future directions. Public Health Nutr 2018;21:1399–1408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Moore SG, Donnelly JK, Jones S, Cade JE. Effect of educational interventions on understanding and use of nutrition labels: a systematic review. Nutrients 2018;10:1432. [DOI] [PMC free article] [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
Data described in the manuscript, data collection forms, and code book will be made available upon request pending application to the CARDIA Coordinating Center (https://www.cardia.dopm.uab.edu/).