Abstract
Rationale & Objective –
Ultra-processed foods (UPF) have become readily available in the global food supply in the past few decades. Several adverse health outcomes have been linked with higher consumption of UPF. However, the impact of UPF on chronic kidney disease (CKD) risk remains unknown.
Study Design –
Prospective cohort study.
Setting & Participants –
14,679 middle-aged adults without CKD at baseline in the Atherosclerosis Risk in Communities (ARIC) study.
Exposure –
UPF consumption (servings/d) calculated using dietary data collected via a food frequency questionnaire at visit 1 and visit 3.
Outcome –
Incident CKD defined as eGFR <60 mL/min/1.73 m2 accompanied by ≥25% eGFR decline, CKD-related hospitalization or death, or end-stage renal disease.
Analytical Approach –
Multivariable-adjusted Cox proportional hazards models were used to assess the association between UPF consumption and CKD. Restricted cubic splines were used to examine the shape of the association.
Results –
During a median follow-up of 24 years, there were 4,859 cases of incident CKD. The incidence rate for the highest quartile of UPF consumption was 16.5 per 1000 person-years (95% CI, 15.6–17.4) and 14.7 per 1000 person-years (95% CI, 13.9–15.5) for the lowest quartile of consumption. After adjusting for a range of confounders including lifestyle factors, demographic characteristics, and health behaviors, participants in the highest quartile of UPF consumption had a 24% higher risk (HR, 1.24; 95% CI, 1.15–1.35) of developing CKD compared to those in the lowest quartile. There was an approximately linear relationship observed between ultra-processed food intake and risk of CKD. By substituting one serving of ultra-processed foods with minimally processed foods, there was a 6% lower risk of CKD observed (HR, 0.94; 95% CI, 0.93–0.96; p < 0.001).
Limitations –
Self-reported data and residual confounding.
Conclusions –
Higher UPF consumption was independently associated with a higher risk of incident CKD in a general population.
Index words: ARIC study, diet and nutrition, epidemiology, kidney disease, ultra-processed foods, NOVA classification
Plain Language Summary
Ultra-processed foods are industrially processed foods and drinks that contain little to no intact foods and artificial additives and substances. The consumption of ultra-processed foods has been increasing around the world recently and has been linked with adverse health outcomes. In this study, we aimed to expand the evidence by investigating the relationship between ultra-processed foods consumption and risk of chronic kidney disease. We found that the higher consumption of ultra-processed foods is associated with higher risk of chronic kidney disease. Our results provided support to avoid ultra-processed foods, and further studies should explore the underlying mechanisms by which ultra-processed foods may be harmful to the kidneys.
INTRODUCTION
The consumption of ultra-processed foods is high and increasing in the U.S. and around the globe (1, 2). Ultra-processed foods are industrially processed food and drink products that contain little to no intact foods; they are mostly comprised of ingredients extracted from foods and contain non-culinary substances and artificial additives to enhance the shelf life and palatability of products (3). Ultra-processed foods contain a high amount of added sugar, refined carbohydrates, saturated and trans fats, and sodium, and contain a low amount of fiber, protein, and micronutrients. In addition to the poor nutritional composition of ultra-processed foods, contaminants generated during the process of physical or chemical alteration and packaging have a negative health impact (4).
Although ultra-processed food consumption has been linked to cardiovascular diseases, all-cause mortality, and cancers (5, 6, 7), less is known about the impact of ultra-processed food on kidney health. In a recent cross-sectional study conducted in Brazil, older adults on hemodialysis had worse diet quality and higher consumption of ultra-processed foods than older adults without chronic kidney disease (CKD) (8). In a cohort of Spanish older adults without CKD at baseline, ultra-processed food consumption was associated with a more than 50% higher risk of renal function decline (9). Given the high burden of kidney diseases and the poor nutritional value and high prevalence of ultra-processed foods in Western diets, there is a need for longitudinal evidence on the relationship between levels of food processing and risk of adverse kidney outcomes.
We aimed to assess the association between ultra-processed food consumption and incident CKD in a large, prospective study of generally healthy individuals without CKD at baseline.
METHODS
Study Population and Design
The Atherosclerosis Risk in Communities (ARIC) study is a prospective cohort of 15,792 black and white men and women aged 45–64 years. The study protocol at each participating site was approved by the institutional review board (IRB), and all participants provided written informed consent at each study visit (10). We treated study visit 1 (1987–1989) as baseline for the present analysis. After exclusions, the final analytic sample included 14,679 middle-aged U.S. adults (Figure 1).
Figure 1.
Flowchart for the selection of the analytic sample from the Atherosclerosis Risk in Communities (ARIC) Study. Number of participants excluded for each covariate: age (n = 0), body mass index (n = 11), physical activity score (n = 44), smoking status (n = 13), drinking status (n = 37), education level (n = 18), diabetes status (n = 119), hypertension status (n = 79), eGFR level (n = 27), serum cholesterol level (n = 104); Abbreviations: eGFR, estimated glomerular filtration rate; FFQ, food frequency questionnaire
Ultra-Processed Food Classification
Participants completed a 66-item modified semi-quantitative Willett food frequency questionnaire (FFQ) at baseline (1987–1989) and visit 3 (1993–1995) (11, 12). Food items were classified according to the NOVA classification system: 1) unprocessed or minimally processed foods, 2) processed culinary ingredients, 3) processed foods, and 4) ultra-processed foods (13). Ultra-processed food classification for food items on the ARIC study FFQ has been reported previously (14). Within each NOVA category, we calculated the total daily consumption for each participant (servings per day), adjusted for energy using the residual method (15), and then assigned participants to quartiles based on the ranked distribution of energy-adjusted frequency of ultra-processed food consumption.
Outcome Ascertainment
Incident CKD was defined by meeting at least one of the following four criteria over a total follow-up period of 32 years from the baseline visit (1987–1989) to the end of follow-up (December 31, 2018): (1) reduced kidney function (eGFR <60 mL/min/1.73 m2) accompanied by ≥25% eGFR decline at any follow-up study visit relative to baseline; (2) hospitalization involving CKD stage 3+ diagnosis defined by International Classification of Diseases (ICD) 9/10 code, identified through active surveillance of the ARIC cohort; (3) death involving CKD stage 3+ diagnosis defined by ICD 9/10 code, identified through linkage to the National Death Index; or (4) end-stage renal disease defined as dialysis or transplantation, identified by linkage to the US Renal Data System (USRDS) registry (16).
Measurement of Covariates
Information about age, sex, race, education level, smoking status, drinking status, physical activity, hypertension status, and diabetes status was collected at baseline through a validated questionnaire administered by trained interviewers (10). Total serum cholesterol level was measured using the enzymatic method with a single aqueous reagent (17). Body mass index (BMI) was calculated using measured weight in kilograms and height in meters taken during baseline visit. Blood pressure was measured three times and the mean of latter two was used for analysis. Hypertension was defined as using blood pressure lowering medication, systolic blood pressure ⩾140 mm Hg, or diastolic blood pressure ⩾90 mm Hg. Diabetes was defined as using diabetes medication, self-reported diagnosis of diabetes mellitus, fasting blood glucose ⩾126 mg/dL, or non-fasting blood glucose ⩾200 mg/dL. Stage 2+ CKD was defined as eGFR < 90 mL/min/1.73 m2. As a measure of overall diet quality, Alternative Heathy Eating Index-2010 (AHEI) score was constructed based on 11 dietary components with higher scores representing healthier diets (18).
Statistical Analyses
For participant and nutritional characteristics, descriptive statistics were reported and differences across quartiles of energy-adjusted ultra-processed food consumption were tested using χ2 tests and analysis of variance. For the primary analysis, we estimated hazard ratios (HR) and 95% CI for the association between quartiles of ultra-processed food intake and incident CKD using Cox proportional hazard regression models.
We incorporated dietary intake as a time-varying factor. Specifically, for participants who developed CKD or were censored between baseline and visit 3, we compared them with participants who had not yet had an event by that time. Visit 1 dietary intake data were used for all participants from visit 1 until visit 3, and the average of visit 1 and visit 3 values were used for those who developed CKD or were censored after visit 3.
We incrementally adjusted for potential confounders. Model 1 adjusted for demographic characteristics (age, sex), total energy intake, and race-center, which was combined to account for the non-uniform distribution of race groups across centers. Model 2 (main model) additionally adjusted for socioeconomic status (education level) and health behaviors (smoking, physical activity). Model 3 adjusted for all previous covariates with the addition of potentially mediating factors [BMI, hypertension status, diabetes status, kidney function (two linear spline terms with one knot at eGFR 90 mL/min/1.73 m2), total cholesterol level]. Model 4 adjusted for diet quality (AHEI) in addition to covariates in our main model (model 2) in order to evaluate whether the association between ultra-processed food and CKD was predominantly driven by diet quality or some other factor related to ultra-processed food. We calculated p-values for trend across quartiles using the median value of each quartile of ultra-processed food consumption. We estimated the association between each additional serving of ultra-processed food consumed (as a continuous variable) and the risk of incident CKD. We examined the association between ultra-processed food consumption and incident CKD using visit-based measures only (eGFR<60 mL/min/1.73 m2 and ⩾30% eGFR decline relative to baseline). To visually evaluate the shape of the association between ultra-processed food consumption and risk of CKD, we used a restricted cubic spline with 4 knots at the 5th, 35th, 65th, and 95th percentiles. We tested for effect modification by sex, race, BMI categories, diabetes status, and hypertension status using likelihood ratio tests and stratified the analysis by these subgroups.
As a secondary analysis, we investigated the relationship between the consumption of unprocessed or minimally processed foods and risk of incident CKD. We conducted a substitution analysis to estimate the impact of replacing ultra-processed foods with minimally processed foods in model 2 (19). We examined the association between intake of specific ultra-processed foods items (beverages; margarine; bakery goods; ultra-processed meats; cereals; fried foods; sugary snacks; hard liquor; ice cream) and CKD risk. Sensitivity analyses were performed by excluding CKD cases diagnosed within the first 5 years of follow-up to minimize the impact of potential reverse causation. The proportional hazards assumption was tested using Schoenfeld residuals, and we found that this assumption was not violated (p=0.67). A two-sided p-value <0.05 was considered statistically significant. All data analyses were performed using Stata version 16.0 (StataCorp, LLC).
RESULTS
Baseline Characteristics of Participants
Among the 14,679 participants included in the present study, 55.1% were female and 25.2% were black. The mean age of participants was 54.1 years and mean BMI was 27.6 kg/m2. More than two fifths of the participants had a higher than high school degree (44.4%) and never smoked tobacco (41.5%), and more than half currently drank alcohol (56.7%). The mean energy-adjusted ultra-processed food consumption was 8.4 servings per day in quartile 4 compared to 3.6 servings per day in quartile 1 (Table 1). Those participants in the highest quartile of ultra-processed food were younger, were more likely to be white, obese, and former smokers, and had lower levels of physical activity. Participants who consumed more ultra-processed food also had lower overall diet quality.
Table 1.
Baseline characteristics of participants according to quartiles of ultra-processed food consumption
| Quartile 1 (n=3,670) | Quartile 2 (n=3,670) | Quartile 3 (n=3,670) | Quartile 4 (n=3,669) | p-valueb | |
|---|---|---|---|---|---|
| Ultra-processed food intake, servings/d | 3.6 (1.0) | 5.2 (0.3) | 6.2 (0.3) | 8.4 (1.6) | <0.001 |
| Age, y | 54.3 (5.8) | 54.2 (5.8) | 54.1 (5.8) | 53.9 (5.7) | 0.009 |
| Female | 1977 (53.9%) | 2140 (58.3%) | 2054 (56.0%) | 1915 (52.2%) | <0.001 |
| Black | 1152 (31.4%) | 1115 (30.4%) | 936 (25.5%) | 494 (13.5%) | <0.001 |
| Education level | <0.001 | ||||
| Less than high school | 943 (25.7%) | 881 (24.0%) | 832 (22.7%) | 732 (20.0%) | |
| High school | 1073 (29.2%) | 1177 (32.1%) | 1230 (33.5%) | 1294 (35.3%) | |
| Higher than high school | 1654 (45.1%) | 1612 (43.9%) | 1608 (43.8%) | 1643 (44.8%) | |
| Body mass index category | 0.008 | ||||
| Normal weight, <25.0 kg/m2 | 1266 (34.5%) | 1269 (34.6%) | 1244 (33.9%) | 1139 (31.0%) | |
| Overweight, 25.0–<30.0 kg/m2 | 1460 (39.8%) | 1415 (38.6%) | 1447 (39.3%) | 1471 (40.1%) | |
| Obese, ≥30.0 kg/m2 | 944 (25.7%) | 986 (26.9%) | 982 (26.8%) | 1059 (28.9%) | |
| Smoking status | <0.001 | ||||
| Current smoker | 1063 (29.0%) | 926 (25.2%) | 903 (24.6%) | 933 (25.4%) | |
| Former smoker | 1125 (30.7%) | 1176 (32.0%) | 1160 (31.6%) | 1298 (35.4%) | |
| Never smoker | 1482 (40.4%) | 1568 (42.7%) | 1607 (43.8%) | 1438 (39.2%) | |
| Drinking status | <0.001 | ||||
| Current drinker | 2018 (55.0%) | 1972 (53.7%) | 2076 (56.6%) | 2250 (61.3%) | |
| Former drinker | 699 (19.0%) | 688 (18.7%) | 661 (18.0%) | 702 (19.1%) | |
| Never drinker | 953 (26.0%) | 1010 (27.5%) | 933 (25.4%) | 717 (19.5%) | |
| Physical activity score a | 2.5 (0.8) | 2.4 (0.8) | 2.4 (0.8) | 2.4 (0.8) | 0.04 |
| Diabetes | 321 (8.7%) | 340 (9.3%) | 333 (9.1%) | 380 (10.4%) | 0.1 |
| Hypertension | 1058 (28.8%) | 1090 (29.7%) | 1047 (28.5%) | 1002 (27.3%) | 0.2 |
| Total serum cholesterol, mg/dL | 215.1 (42.6) | 215.0 (41.8) | 215.6 (41.4) | 213.8 (41.6) | 0.3 |
| Diet quality (AHEI) score | 52.3 (10.8) | 47.9 (10.0) | 45.9 (9.7) | 45.5 (9.6) | <0.001 |
| Stage 2 + CKD, eGFR < 90 mL/min/1.73 m2 | 548 (14.9%) | 545 (14.9%) | 575 (15.7%) | 585 (15.9%) | 0.5 |
Baseline characteristics are reported as mean (standard deviation) for continuous variables and number (%) for categorical variables. Abbreviations: AHEI, alternative healthy eating index; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate.
Physical activity score for sport-related exercise during leisure time.
P-values were calculated from Pearson’s chi-squared test for categorical variables and analysis of variance for continuous variables.
Nutritional Characteristics of Ultra-Processed Food
Higher ultra-processed food consumption was associated with a lower intake of protein, cholesterol, fiber, and some micronutrients (e.g., niacin, vitamin A, vitamin B6, vitamin B12, calcium, phosphorous, magnesium, and potassium), and a higher intake of fat (total, saturated, monounsaturated, and polyunsaturated fat) (Table 2). Participants in the highest quartile of ultra-processed food consumption had a lower consumption of fruits and vegetables, and a higher consumption of sugar-sweetened beverages. The top food groups that contributed to the frequency of ultra-processed food consumption were: sugar-sweetened beverages (27%), margarine (18%), bakery goods (15%), and ultra-processed meats (11%).
Table 2.
Nutritional characteristics according to quartiles of ultra-processed food consumption
| Nutrient, mean (SD) | Quartile 1 (n= 3,670) | Quartile 2 (n=3,670) | Quartile 3 (n=3,670) | Quartile 4 (n=3,669) | p-valuea |
|---|---|---|---|---|---|
| Total energy, kcal/day | 1755.7 (611.5) | 1470.8 (536.9) | 1514.1 (559.5) | 1715.1 (639.8) | <0.001 |
| Macronutrient intake | |||||
| Protein, % of energy | 19.4 (4.3) | 18.6 (3.9) | 17.3 (3.8) | 16.3 (3.9) | <0.001 |
| Carbohydrate, % of energy | 48.5 (9.4) | 48.9 (9.0) | 49.3 (9.2) | 48.6 (9.9) | 0.002 |
| Total fat, % of energy | 31.6 (6.9) | 32.4 (6.4) | 33.0 (6.4) | 34.5 (6.9) | <0.001 |
| SFA, % of energy | 11.7 (3.2) | 11.8 (2.9) | 12.1 (2.8) | 12.4 (3.0) | <0.001 |
| MUFA, % of energy | 12.0 (3.1) | 12.4 (2.9) | 12.7 (2.8) | 13.3 (3.0) | <0.001 |
| PUFA, % of energy | 4.6 (1.3) | 4.8 (1.3) | 5.0 (1.3) | 5.6 (1.6) | <0.001 |
| Sugar, g/1000 kcal | 66.4 (23.9) | 68.8 (23.5) | 72.1 (26.8) | 71.2 (29.6) | <0.001 |
| Micronutrient intake | |||||
| Cholesterol, mg/1000 kcal | 167.4 (66.2) | 163.7 (62.1) | 154.7 (56.9) | 143.8 (53.3) | <0.001 |
| Folate, μg/1000 kcal | 148.7 (49.8) | 151.8 (56.0) | 146.5 (56.7) | 142.4 (61.8) | <0.001 |
| Niacin, mg/1000 kcal | 12.3 (3.1) | 12.2 (2.9) | 11.5 (2.9) | 11.0 (3.0) | <0.001 |
| Fiber, g/1000 kcal | 12.1 (4.7) | 11.4 (4.3) | 10.6 (3.7) | 9.9 (3.5) | <0.001 |
| Vitamin A, IU/1000 kcal | 6689.2 (5249.8) | 6310.5 (4299.4) | 5544.1 (3821.2) | 4854.5 (3781.4) | <0.001 |
| Vitamin B6, mg/1000 kcal | 1.2 (0.3) | 1.1 (0.3) | 1.1 (0.3) | 1.0 (0.3) | <0.001 |
| Vitamin B12, μg/1000 kcal | 5.2 (2.7) | 5.3 (3.0) | 4.8 (2.6) | 4.3 (2.3) | <0.001 |
| Vitamin C, mg/1000 kcal | 73.7 (40.5) | 78.8 (42.9) | 80.2 (45.2) | 77.9 (49.8) | <0.001 |
| Vitamin E, mg/1000 kcal | 3.2 (1.4) | 3.2 (1.8) | 3.0 (1.5) | 3.0 (1.8) | <0.001 |
| Sodium, mg/1000 kcal | 945.2 (216.9) | 933.8 (200.5) | 913.5 (202.9) | 917.1 (208.1) | <0.001 |
| Calcium, mg/1000 kcal | 452.8 (207.1) | 418.8 (182.3) | 394.1 (161.8) | 378.9 (157.3) | <0.001 |
| Iron, mg/1000 kcal | 7.1 (2.1) | 7.3 (2.4) | 7.1 (2.7) | 6.8 (2.6) | <0.001 |
| Phosphorus, mg/1000 kcal | 726.4 (166.0) | 691.7 (162.0) | 658.8 (162.3) | 655.4 (176.2) | <0.001 |
| Magnesium, mg/1000 kcal | 174.0 (42.6) | 166.9 (42.4) | 158.0 (41.2) | 151.0 (40.4) | <0.001 |
| Zinc, mg/1000 kcal | 6.8 (1.6) | 6.8 (1.7) | 6.7 (1.7) | 6.9 (2.1) | <0.001 |
| Potassium, mg/1000 kcal | 1784.3 (422.7) | 1732.2 (437.5) | 1648.9 (414.2) | 1572.1 (412.6) | <0.001 |
| Food consumption | |||||
| Fruit, servings/day | 1.9 (1.7) | 1.4 (1.1) | 1.3 (1.0) | 1.2 (1.1) | <0.001 |
| Vegetables, servings/day | 2.2 (1.4) | 1.7 (1.1) | 1.6 (1.0) | 1.5 (1.0) | <0.001 |
| Red meat, servings/day | 1.3 (1.0) | 1.2 (0.8) | 1.3 (0.8) | 1.4 (1.0) | <0.001 |
| Dairy, servings/day | 2.4 (1.9) | 1.7 (1.4) | 1.7 (1.3) | 1.9 (1.5) | <0.001 |
| Sugar-sweetened drinks, glass/day | 0.6 (0.7) | 0.7 (0.8) | 1.0 (1.0) | 1.9 (1.8) | <0.001 |
| Alcohol, g/day | 6.9 (15.2) | 5.1 (11.1) | 5.5 (11.3) | 6.9 (15.8) | <0.001 |
Nutritional characteristics are reported as mean (standard deviation). Abbreviations: IU, international units; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; SFA, saturated fatty acid.
P-values were calculated from analysis of variance for continuous variable.
Association between Ultra-Processed Food Consumption and Risk of CKD
During a median follow-up of 24 years, there were 4,859 cases (34.0%) of incident CKD. The incidence rate was 12% higher in the highest quartile of ultra-processed food consumption (16.5 per 1000 person-years, 95% CI, 15.6–17.4) compared to the lowest quartile (14.7 per 1000 person-years, 95% CI, 13.9–15.5) (Table 3). Higher intake of ultra-processed food consumption was associated with a higher risk of incident CKD, and the results were consistent across different models. In model 1, adjusting for age, sex, race-center, and total energy intake, those in the highest quartile of ultra-processed food intake had a 27% higher risk of incident CKD (HR for quartile 4 vs. quartile 1, 1.27; 95% CI, 1.17–1.37) compared to those in the lowest quartile of intake. Results were similar in the main model (model 2), which additionally adjusted for education level, smoking status, and physical activity (HR for quartile 4 vs. quartile 1, 1.24; 95% CI, 1.15–1.35). After further adjusting for potential mediators (model 3), the results were attenuated, but ultra-processed food consumption remained statistically significantly associated with incident CKD (HR for quartile 4 vs. quartile 1, 1.16; 95% CI, 1.07–1.26). Similarly, results were attenuated compared to model 2 after accounting for diet quality, but the association between ultra-processed food and incident CKD remained statistically significant (HR for quartile 4 vs. quartile 1, 1.19; 95% CI, 1.09–1.29). The results were generally consistent when using the visit-based definition (eGFR < 60 mL/min/1.73m2 and ≥30% eGFR decline) for incident CKD. The full model results are available in the supplementary materials (Table S1).
Table 3.
Risk of incident CKD associated with ultra-processed food consumption, analyzed according to quartiles and continuously
| Categorical analysis | Continuous analysis | ||||||
|---|---|---|---|---|---|---|---|
| Quartile 1 (n=3,670) | Quartile 2 (n=3,670) | Quartile 3 (n=3,670) | Quartile 4 (n=3,669) | p-value for trend | Per 1 additional serving of ultra-processed food per day | p-value | |
| CKD (composite definition) | |||||||
| Events observed (n, (%)) | 1,158 (31.6%) | 1,211 (33.0%) | 1,210 (33.0%) | 1,280 (34.9%) | |||
| Incidence rate per 1,000 person-years (95% CI) | 14.7 (13.9, 15.5) | 15.3 (14.5, 16.2) | 15.3 (14.4, 16.1) | 16.5 (15.6, 17.4) | |||
| Model 1a | 1 [reference] | 1.06 (0.98, 1.15) | 1.09 (1.01, 1.19) | 1.27 (1.17, 1.37) | <0.001 | 1.06 (1.04, 1.08) | <0.001 |
| Model 2b | 1 [reference] | 1.05 (0.96, 1.13) | 1.08 (0.99, 1.17) | 1.24 (1.15, 1.35) | <0.001 | 1.05 (1.04, 1.07) | <0.001 |
| Model 3c | 1 [reference] | 1.04 (0.96, 1.13) | 1.04 (0.96, 1.13) | 1.16 (1.07, 1.26) | <0.001 | 1.04 (1.02, 1.05) | <0.001 |
| Model 4d | 1 [reference] | 1.02 (0.94, 1.11) | 1.03 (0.95, 1.12) | 1.19 (1.09, 1.29) | <0.001 | 1.04 (1.03, 1.06) | <0.001 |
| CKD (visit-based definition) | |||||||
| Events observed (n, (%)) | 607 (16.5%) | 690 (18.8%) | 675 (18.4%) | 722 (19.7%) | |||
| Incidence rate per 1,000 person-years (95% CI) | 7.48 (6.91, 8.10) | 8.48 (7.87, 9.14) | 8.29 (7.69, 8.94) | 9.00 (8.36, 9.68) | |||
| Model 1a | 1 [reference] | 1.14 (1.02, 1.27) | 1.13 (1.01, 1.27) | 1.30 (1.16, 1.45) | <0.001 | 1.05 (1.03, 1.07) | <0.001 |
| Model 2b | 1 [reference] | 1.14 (1.02, 1.27) | 1.13 (1.01, 1.27) | 1.29 (1.16, 1.44) | <0.001 | 1.05 (1.03, 1.07) | <0.001 |
| Model 3c | 1 [reference] | 1.13 (1.01, 1.26) | 1.09 (0.98, 1.22) | 1.23 (1.10, 1.37) | <0.001 | 1.04 (1.02, 1.06) | <0.001 |
| Model 4d | 1 [reference] | 1.11 (0.99, 1.25) | 1.09 (0.97, 1.22) | 1.22 (1.09, 1.37) | 0.009 | 1.04 (1.01, 1.06) | 0.001 |
CKD (composite definition) was defined as estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m2 accompanied by ≥25% eGFR decline, International Classification of Diseases (ICD)-9/10 code related to CKD stage 3+ through Atherosclerosis Risk in Communities (ARIC) study surveillance or National Death Index, or end-stage renal disease identified by US Renal Data System registry. CKD (visit-based definition) was defined as eGFR<60 mL/min/1.73 m2 accompanied by ≥30% eGFR decline. Abbreviations: CKD, chronic kidney disease; CI, confidence interval.
Model 1 was adjusted for age, sex, race-center, and total energy intake.
Model 2 was adjusted for variables in model 1 plus education level, smoking status, and physical activity score.
Model 3 was adjusted for variables in both model 1 and 2 plus diabetes status, hypertension status, body mass index, serum cholesterol level, and kidney function (two linear spline terms with one knot at 90 ml/min/1.73 m2).
Model 4 was adjusted for variables in both model 1 and 2 plus Alternative Healthy Eating Index (AHEI) score.
Each additional serving of ultra-processed food consumed per day was significantly associated with 5% higher risk of incident CKD using the composite definition of CKD (model 2 HR, 1.05; 95% CI, 1.04–1.07; p<0.001) as well as the visit-based definition of CKD (model 2 HR, 1.05; 95% CI, 1.03–1.07; p<0.001) (Table 3).
There was an approximately linear dose-response relationship between ultra-processed food and the risk of incident CKD above the 25th percentile of 3.87 servings/day (Figure 2).
Figure 2.
Restricted cubic spline of the association between ultra-processed food consumption and risk of incident CKD. The grey histogram shows the distribution of ultra-processed food consumption. The black solid line represents the adjusted hazard ratios for incident CKD, modeled using restricted cubic splines with 4 knots at the 5th, 35th, 65th, and 95th percentiles. The reference level was set at the 25th percentile (3.87 servings/day). The black dashed lines represent 95% confidence intervals. The model was adjusted for age, sex, race-center, total energy intake, education level, smoking status, and physical activity score. Abbreviations: CI, confidence interval; CKD, chronic kidney disease; HR, hazard ratio.
Association between Minimally Processed Food Consumption and Risk of CKD
Replacing one serving per day of ultra-processed food with minimally processed food was significantly associated with 6% lower risk of incident CKD (HR, 0.94; 95% CI, 0.93–0.96; p < 0.001). Higher intake of unprocessed or minimally processed foods was associated with a lower risk of CKD (Table S2). In the main model, participants in the highest quartile of minimally processed foods intake had a 10% lower risk of incident CKD compared to those in the lowest quartile (model 2 HR for quartile 4 vs. quartile 1, 0.90; 95% CI, 0.83–0.98; p=0.003). The association was no longer significant after controlling for diet quality. For the visit-based CKD outcome, similar inverse associations were observed across all four models.
Association between Specific Ultra-Processed Foods and Risk of CKD
Higher consumption of sugar-sweetened beverages was associated with 22% higher risk of incident CKD (HR for quartile 4 vs. quartile 1, 1.22; 95% CI, 1.12–1.33) (Table 4). Ultra-processed meats were associated with 18% higher risk of incident CKD (HR for quartile 4 vs. quartile 1, 1.18; 95% CI, 1.08–1.29). No statistically significant association was observed for other ultra-processed food items.
Table 4.
Risk of incident CKD associated with consumption of specific ultra-processed food items or groups
| Ultra-processed food items or groups | HR (95% CI)a |
|---|---|
| Sugar-sweetened beveragesb | 1.22 (1.12, 1.33) |
| Margarine | 1.08 (0.99, 1.18) |
| Bakery goodsc | 0.92 (0.85, 1.01) |
| Ultra-processed meatsd | 1.18 (1.08, 1.29) |
| Cereals | 0.92 (0.84, 1,01) |
| Fried foodse | 1.00 (0.91, 1.10) |
| Sugary snacksf | 1.05 (0.96, 1.15) |
| Hard liquor | 0.97 (0.87, 1.08) |
| Ice cream | 0.95 (0.87, 1.04) |
Hazard ratios for incident CKD were calculated for highest consumption (quartile 4) versus lowest consumption (quartile 1) of the specific ultra-processed food items or groups. Models were adjusted for age, sex, race-center, total energy intake, education level, smoking status, and physical activity score.
Beverages included orange or grapefruit juice, low calorie soft drinks, regular soft drinks, fruitflavored punch or non-carbonated beverages (lemonade, Kool-Aid, Hawaiian Punch).
Bakery goods included ready-made pie, donuts, biscuits or cornbread; Danish pastry, sweet roll, coffee cake and croissant; cookies, cake or brownie.
Ultra-processed meats included hamburgers, hot dogs, processed meats (sausage, salami, bologna); beef, pork or lamb in dishes.
Fried foods included potato chips or corn chips, French fried potatoes, food fried away from home.
Sugary snacks included chocolate bars or pieces (Hershey’s, Plain M&M’s, Snickers, Reese’s), candy without chocolate.
Abbreviations: CI, confidence interval; CKD, chronic kidney disease; HR, hazard ratio. Bold font denotes statistically significant associations.
Subgroup Analyses
The magnitude of the association for results stratified by sex, race, BMI, diabetes, and hypertension were generally similar to the results for the overall study population (Figure 3). There was no statistical evidence of effect modification by sex (p = 0.9), race (p = 0.08), BMI (p = 0.6), diabetes status (p = 0.08), or hypertension status (p = 0.6).
Figure 3.
Risk of incident CKD associated with ultra-processed food consumption according to subgroups of the study population. HRs and 95% CI were presented for quartile 4 versus quartile 1 of ultra-processed food consumption. The model was adjusted for age, sex, race-center, total energy intake, education level, smoking status, and physical activity score.
Abbreviations: BMI, body mass index; CI, confidence interval; CKD, chronic kidney disease; HR, hazard ratio.
Sensitivity Analysis
Results were similar to the main findings when CKD cases that occurred during the first 5 years of follow-up were excluded.
DISCUSSION
In this analysis of a community-based biracial cohort of 14,679 middle-aged adults over 3 decades, higher consumption of ultra-processed foods was associated with higher risk of incident CKD. The association remained significant after adjusting for a range of confounders including sociodemographic characteristics and health behaviors, and was independent of potentially mediating health conditions and diet quality. There was an approximately linear relationship found between each additional serving of ultra-processed food consumed and risk of CKD above the 25th percentile of 3.87 servings per day. The results were similar within population subgroups by sex, race, BMI categories, hypertension status, and diabetes status. Higher consumption of minimally processed foods was associated with lower risk of incident CKD.
Our study revealed a high level of ultra-processed food consumption (mean of 8.4 servings/day in quartile 4), which is in line with previous studies reported from U.S. populations (20, 21, 22). Our results on nutritional characteristics were mostly consistent with previous findings suggesting that higher level of ultra-processed food consumption is associated with an overall poor diet quality. In contrast to previously mentioned studies, we found inverse relationships between ultra-processed food consumption and cholesterol intake, and there was no significant difference in carbohydrate intake across quartiles of ultra-processed food consumption. Such discrepancies between studies may be due to differences in capturing dietary and nutritional information, and differences in methods for classifying of ultra-processed foods. Our study advances the understanding of the impact of ultra-processed food on kidney outcomes by incorporating clinically diagnosed CKD outcomes and improving the generalizability of the results by using a large biracial sample of middle-aged healthy adults (8, 9). This study is, to our knowledge, the first to assess the prospective relationship between ultra-processed food consumption and incident CKD in the general population.
Several potential mechanisms may explain the observed association between ultra-processed food consumption and CKD. High intake of sugar, specifically in the form of sugar-sweetened beverages, has been linked with an elevated risk of ESRD in a general population and an elevated risk of CKD in a community-based cohort of black Americans (23). Higher consumption of sugar-sweetened beverages is also found to be associated with development of type 2 diabetes in a meta-analysis (24). We also found that sugar-sweetened beverages, as a specific ultra-processed food category, contributed the most to the frequency of ultra-processed food consumption in our study sample, and was associated with an increased risk of CKD. Low fiber content is another nutritional characteristic of ultra-processed foods and was observed in the current study. Several studies have reported that high dietary fiber is associated with a lower inflammatory response and lower risk of CKD (25, 26). Dietary intake of fiber is also believed to play an important role in achieving adequate gut microbiota composition and thus mitigating CKD risk factors such as obesity, diabetes, and dyslipidemia (27, 28). We additionally found that participants who consumed higher levels of ultra-processed foods had a lower diet quality, as measured by AHEI. In a previous analysis conducted in the ARIC study, higher adherence to healthy dietary patterns, including AHEI, was associated with lower risk of CKD (29). Higher AHEI score has been linked to lower concentrations of inflammatory and endothelial dysfunction biomarkers, which are precursors to major CKD risk factors such as diabetes and cardiovascular disease (30).
Food additives are a common component of ultra-processed foods that may contribute to the observed association between ultra-processed food and CKD. Inorganic phosphate is a type of biochemical compound that is often added to ultra-processed foods such as processed meats, baked goods, and soft drinks in order to enhance flavor, alter texture, and prolong shelf life (31). This form of phosphate can be easily absorbed due to its high bioavailability, and elevated phosphate has been consistently found to be independently associated with eGFR decline, CKD progression, and CKD-related mortality (32, 33, 34). High levels of phosphate can increase fibroblast growth factor-23 which is also associated with incident kidney disease (35). Other additives such as emulsifiers and non-caloric sweeteners can also increase the risk of cardiometabolic diseases and type 2 diabetes specifically which are risk factors for CKD (36, 37). Food processing also impacts nutrient availability and the microbiome (38). Ultra-processed foods can increase the risk of CKD through the introduction of neo-formed contaminants during food processing. Advanced glycation end products (AGEs) are chemical compounds generated by heating foods that can increase gut barrier permeability and in turn lead to local inflammation and CKD in rodent models (39). AGEs provide flavor to baked, grilled, or roasted foods, and they have been artificially added to ultra-processed foods to increase flavor and palatability. We found the association between ultra-processed food and incident CKD was not entirely mediated by diet quality, suggesting that other factors such as food additives, emulsifiers, non-caloric sweeteners, and neo-formed contaminants may have played a role.
Our study has several strengths including the prospective study design which allowed us to establish temporality between ultra-processed food consumption and incident CKD; long-term follow-up (median of 24 years) for the enumeration of a sufficient number of incident CKD cases; a large sample size and racially diverse cohort, which provided external validity and allowed us to explore effect modification by race; time-varying exposure (two repeated dietary assessments) to better fit the assumption of our statistical model; and comprehensive ascertainment of incident CKD through surveillance to capture events among participants who may have missed study follow-up.
There are several limitations to this study. First, although the FFQ was validated and trained interviewers and visual guidance were used to minimize reporting errors, self-reported dietary data is prone to measurement error and recall bias. Second, our FFQ was not specifically designed to collect ultra-processed food consumption data and, as such, detailed information such as preparation method and additives were not available, which may have led to misclassification. Third, reverse causality is a concern given the observational study design. To minimize this issue, we excluded participants with prevalent CKD at baseline in the main analysis and excluded CKD cases that were identified in the first 5 years of follow-up as a sensitivity analysis. Fourth, residual confounding due to unmeasured or imprecisely measured covariates is likely due to the nature of observational studies. However, a wide range of demographic, clinical, and lifestyle variables were rigorously obtained in person by trained staff and our analyses adjusted for these factors. Fifth, the NOVA classification system has been criticized for simplifying processing levels (40). Nonetheless, the use of NOVA classification in epidemiologic studies has been useful for building an evidence base on the adverse health outcomes associated with ultra-processed food consumption. The NOVA classification system is also widely used in studies of ultra-processed food and health outcomes, which allows for comparison with other studies.
In this large prospective cohort of middle-aged adults, higher ultra-processed food consumption was associated with a higher risk of incident CKD. This association was independent of CKD risk factors, was not entirely explained by potential mediating health conditions and diet quality, and was consistent across subgroups of the study population by sex, race, BMI, diabetes status, and hypertension status. Given the rise of ultra-processed foods in the global food supply, our study provides further support to avoid ultra-processed foods and to replace ultra-processed foods with minimally processed or unprocessed foods. These findings should be confirmed in other settings and populations, and further studies should explore the specific mechanisms by which ultra-processed foods may be harmful to the kidneys.
Supplementary Material
Table S1. Full model hazard ratios and 95% confidence interval for risk of incident CKD with ultra-processed food consumption, analyzed according to quartiles.
Table S2. Risk of incident CKD associated with unprocessed and minimally processed foods, analyzed according to quartiles and continuously.
Acknowledgements:
We thank the staff and participants of the Atherosclerosis Risk in Communities study for their important contributions.
Support:
The Atherosclerosis Risk in Communities Study was carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts (HHSN268201700001I, HHSN268201700002I, HHSN268201700003I, HHSN268201700004I, HHSN268201700005I). CMR was supported by grants from the National Institute of Diabetes and Digestive and Kidney Diseases (K01 DK107782, R03 DK128386) and the National Heart, Lung, and Blood Institute (R01 HL153178). Some of the data reported here were supplied by the United States Renal Data System (USRDS). The study funders had no role in the design, implementation, analysis, or data interpretation for the present study.
Footnotes
Financial Disclosure: The authors declare that they have no relevant financial interests.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Publisher's Disclaimer: Disclaimer: The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy or interpretation of the US government.
REFERENCES
- 1.Vandevijvere S, Jaacks LM, Monteiro CA, et al. Global trends in ultraprocessed food and drink product sales and their association with adult body mass index trajectories. Obesity Reviews. 2019;20(S2):10–19. doi: 10.1111/obr.12860 [DOI] [PubMed] [Google Scholar]
- 2.Wang L, Martínez Steele E, Du M, et al. Trends in Consumption of Ultraprocessed Foods Among US Youths Aged 2–19 Years, 1999–2018. JAMA. 2021;326(6):519–530. doi: 10.1001/jama.2021.10238 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Gibney MJ. Ultra-Processed Foods: Definitions and Policy Issues. Current Developments in Nutrition. 2019;3(2):nzy077. doi: 10.1093/cdn/nzy077 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Miclotte L, Van de Wiele T. Food processing, gut microbiota and the globesity problem. Crit Rev Food Sci Nutr. 2020;60(11):1769–1782. doi: 10.1080/10408398.2019.1596878 [DOI] [PubMed] [Google Scholar]
- 5.Srour B, Fezeu LK, Kesse-Guyot E, et al. Ultra-processed food intake and risk of cardiovascular disease: prospective cohort study (NutriNet-Santé). BMJ. 2019;365:l1451. doi: 10.1136/bmj.l1451 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rico-Campà A, Martínez-González MA, Alvarez-Alvarez I, et al. Association between consumption of ultra-processed foods and all cause mortality: SUN prospective cohort study. BMJ. 2019;365:l1949. doi: 10.1136/bmj.l1949 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Fiolet T, Srour B, Sellem L, et al. Consumption of ultra-processed foods and cancer risk: results from NutriNet-Santé prospective cohort. BMJ. 2018;360:k322. doi: 10.1136/bmj.k322 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Martins AM, Bello Moreira AS, Canella DS, et al. Elderly patients on hemodialysis have worse dietary quality and higher consumption of ultraprocessed food than elderly without chronic kidney disease. Nutrition. 2017;41:73–79. doi: 10.1016/j.nut.2017.03.013 [DOI] [PubMed] [Google Scholar]
- 9.Rey-García J, Donat-Vargas C, Sandoval-Insausti H, et al. Ultra-Processed Food Consumption is Associated with Renal Function Decline in Older Adults: A Prospective Cohort Study. Nutrients. 2021;13(2):428. doi: 10.3390/nu13020428 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.The ARIC investigators. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. Am J Epidemiol. 1989;129(4):687–702. [PubMed] [Google Scholar]
- 11.Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122(1):51–65. doi: 10.1093/oxfordjournals.aje.a114086 [DOI] [PubMed] [Google Scholar]
- 12.Stevens J, Metcalf PA, Dennis BH, et al. Reliability of a food frequency questionnaire by ethnicity, gender, age and education. Nutrition Research. 1996;16(5):735–745. doi: 10.1016/0271-5317(96)00064-4 [DOI] [Google Scholar]
- 13.Monteiro CA, Cannon G, Moubarac J-C, et al. The UN Decade of Nutrition, the NOVA food classification and the trouble with ultra-processing. Public Health Nutr. 2018;21:5–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Du S, Kim H, Rebholz CM. Higher Ultra-Processed Food Consumption Is Associated with Increased Risk of Incident Coronary Artery Disease in the Atherosclerosis Risk in Communities Study. J Nutr. 2021;151(12):3746–3754. doi: 10.1093/jn/nxab285 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr. 1997;65:1220S–1228S; discussion 1229S-1231S. [DOI] [PubMed] [Google Scholar]
- 16.Grams ME, Rebholz CM, McMahon B, et al. Identification of incident CKD stage 3 in research studies. Am J Kidney Dis. 2014;64(2):214–221. doi: 10.1053/j.ajkd.2014.02.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Allain CC, Poon LS, Chan CS, Richmond W, Fu PC. Enzymatic determination of total serum cholesterol. Clin Chem. 1974;20(4):470–475. [PubMed] [Google Scholar]
- 18.Chiuve SE, Fung TT, Rimm EB, et al. Alternative dietary indices both strongly predict risk of chronic disease. J Nutr. 2012;142(6):1009–1018. doi: 10.3945/jn.111.157222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Faerch K, Lau C, Tetens I, et al. A statistical approach based on substitution of macronutrients provides additional information to models analyzing single dietary factors in relation to type 2 diabetes in danish adults: the Inter99 study. J Nutr. 2005;135(5):1177–1182. doi: 10.1093/jn/135.5.1177 [DOI] [PubMed] [Google Scholar]
- 20.Juul F, Vaidean G, Lin Y, et al. Ultra-Processed Foods and Incident Cardiovascular Disease in the Framingham Offspring Study. J Am Coll Cardiol. 2021;77(12):1520–1531. doi: 10.1016/j.jacc.2021.01.047 [DOI] [PubMed] [Google Scholar]
- 21.Zhong G-C, Gu H-T, Peng Y, et al. Association of ultra-processed food consumption with cardiovascular mortality in the US population: long-term results from a large prospective multicenter study. Int J Behav Nutr Phys Act. 2021;18(1):21. doi: 10.1186/s12966-021-01081-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kim H, Hu EA, Rebholz CM. Ultra-processed food intake and mortality in the USA: results from the Third National Health and Nutrition Examination Survey (NHANES III, 1988–1994). Public Health Nutr. 2019;22(10):1777–1785. doi: 10.1017/S1368980018003890 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Rebholz CM, Young BA, Katz R, et al. Patterns of Beverages Consumed and Risk of Incident Kidney Disease. Clin J Am Soc Nephrol. 2019;14(1):49–56. doi: 10.2215/CJN.06380518 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Malik VS, Popkin BM, Bray GA, Després J-P, Willett WC, Hu FB. Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes: a meta-analysis. Diabetes Care. 2010;33(11):2477–2483. doi: 10.2337/dc10-1079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Xu H, Huang X, Risérus U, et al. Dietary fiber, kidney function, inflammation, and mortality risk. Clin J Am Soc Nephrol. 2014;9(12):2104–2110. doi: 10.2215/CJN.02260314 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mirmiran P, Yuzbashian E, Asghari G, Sarverzadeh S, Azizi F. Dietary fibre intake in relation to the risk of incident chronic kidney disease. Br J Nutr. 2018;119(5):479–485. doi: 10.1017/S0007114517003671 [DOI] [PubMed] [Google Scholar]
- 27.Cronin P, Joyce SA, O’Toole PW, O’Connor EM. Dietary Fibre Modulates the Gut Microbiota. Nutrients. 2021;13(5):1655. doi: 10.3390/nu13051655 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Zhao L, Zhang F, Ding X, et al. Gut bacteria selectively promoted by dietary fibers alleviate type 2 diabetes. Science. 2018;359(6380):1151–1156. doi: 10.1126/science.aao5774 [DOI] [PubMed] [Google Scholar]
- 29.Hu EA, Steffen LM, Grams ME, et al. Dietary patterns and risk of incident chronic kidney disease: the Atherosclerosis Risk in Communities study. Am J Clin Nutr. 2019;110(3):713–721. doi: 10.1093/ajcn/nqz146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Fung TT, McCullough ML, Newby PK, et al. Diet-quality scores and plasma concentrations of markers of inflammation and endothelial dysfunction. Am J Clin Nutr. 2005;82(1):163–173. doi: 10.1093/ajcn.82.1.163 [DOI] [PubMed] [Google Scholar]
- 31.Houston J, Isakova T, Wolf M. Chapter 20 - Phosphate Metabolism and Fibroblast Growth Factor 23 in Chronic Kidney Disease. In: Kopple JD, Massry SG, Kalantar-Zadeh K, eds. Nutritional Management of Renal Disease. Academic Press; 2013:285–308. doi: 10.1016/B978-0-12-391934-2.00020-5 [DOI] [Google Scholar]
- 32.Vervloet MG, Sezer S, Massy ZA, Johansson L, Cozzolino M, Fouque D. The role of phosphate in kidney disease. Nat Rev Nephrol. 2017;13(1):27–38. doi: 10.1038/nrneph.2016.164 [DOI] [PubMed] [Google Scholar]
- 33.Kestenbaum B, Sampson JN, Rudser KD, et al. Serum phosphate levels and mortality risk among people with chronic kidney disease. J Am Soc Nephrol. 2005;16(2):520–528. doi: 10.1681/ASN.2004070602 [DOI] [PubMed] [Google Scholar]
- 34.O’Seaghdha CM, Hwang S-J, Muntner P, Melamed ML, Fox CS. Serum phosphorus predicts incident chronic kidney disease and end-stage renal disease. Nephrol Dial Transplant. 2011;26(9):2885–2890. doi: 10.1093/ndt/gfq808 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Rebholz CM, Grams ME, Coresh J, et al. Serum Fibroblast Growth Factor-23 Is Associated with Incident Kidney Disease. J Am Soc Nephrol. 2015;26(1):192–200. doi: 10.1681/ASN.2014020218 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Suez J, Korem T, Zeevi D, et al. Artificial sweeteners induce glucose intolerance by altering the gut microbiota. Nature. 2014;514(7521):181–186. doi: 10.1038/nature13793 [DOI] [PubMed] [Google Scholar]
- 37.Chassaing B, Koren O, Goodrich JK, et al. Dietary emulsifiers impact the mouse gut microbiota promoting colitis and metabolic syndrome. Nature. 2015;519(7541):92–96. doi: 10.1038/nature14232 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Zinöcker MK, Lindseth IA. The Western Diet-Microbiome-Host Interaction and Its Role in Metabolic Disease. Nutrients. 2018;10(3):E365. doi: 10.3390/nu10030365 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Snelson M, Tan SM, Clarke RE, et al. Processed foods drive intestinal barrier permeability and microvascular diseases. Sci Adv. 2021;7(14):eabe4841. doi: 10.1126/sciadv.abe4841 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Gibney MJ, Forde CG, Mullally D, et al. Ultra-processed foods in human health: a critical appraisal. Am J Clin Nutr 2017;106:717–24. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Full model hazard ratios and 95% confidence interval for risk of incident CKD with ultra-processed food consumption, analyzed according to quartiles.
Table S2. Risk of incident CKD associated with unprocessed and minimally processed foods, analyzed according to quartiles and continuously.