1. Date of download: 11/12/2016 Copyright © 2016 American Medical Association. All rights reserved. From: Diet and Kidney Disease in High-Risk Individuals With Type 2 Diabetes Mellitus JAMA Intern Med. 2013;173(18):1682-1692. doi:10.1001/jamainternmed.2013.9051 Flowchart of Number of Participants and Outcomes at 5.5 Years of Follow-upGFR indicates glomerular filtration rate; ONTARGET, Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial; and UACR, urinary albumin-creatinine ratio. Figure Legend:

2. Date of download: 11/12/2016 Copyright © 2016 American Medical Association. All rights reserved. From: Diet and Kidney Disease in High-Risk Individuals With Type 2 Diabetes Mellitus JAMA Intern Med. 2013;173(18):1682-1692. doi:10.1001/jamainternmed.2013.9051 Single-Variable Multinomial Logit Models Adjusted With Known ConfoundersConfounders (at study entry) are age, duration of type 2 diabetes mellitus, albuminuria status, glomerular filtration rate, sex, Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial randomization arms, and urinary-albumin-creatinine ratio (UACR) to progression, which was defined as the difference between the participant-specific cutoff point of developing new microalbuminuria or macroalbuminuria and UACR at baseline on the log scale. Association of modified Alternate Healthy Eating Index (mAHEI) and 24-hour urinary sodium and relative odds (solid line) with 95% CI (shaded area) with chronic kidney disease (CKD) (A) or death (B) and respective histograms. The horizontal line on the top shows tertiles, and the numbers within each tertile give the percentage of participants experiencing the respective event. Figure Legend:

3. Date of download: 11/12/2016 Copyright © 2016 American Medical Association. All rights reserved. From: Diet and Kidney Disease in High-Risk Individuals With Type 2 Diabetes Mellitus JAMA Intern Med. 2013;173(18):1682-1692. doi:10.1001/jamainternmed.2013.9051 Forest Plot for Single-Variable Multinomial Logit Models Adjusted With Known ConfoundersFor confounders, see the legend to Figure 2. If not stated otherwise, food items are given in servings per week. Renal Outcome column gives odds ratios (ORs) comparing participants alive with chronic kidney disease (CKD) with participants alive but without CKD; Death column reports ORs comparing participants who died during follow-up with participants alive without CKD. For continuous independent variables, the ORs for the median of the second tertile (50.0th percentile [solid circle]) and the median of the third tertile (83.3rd percentile [open circle]) compared with the median of the first tertile (16.7th percentile) as reference are given. Independent variables in bold type have a significant association with CKD. The last column states the P value of inclusion of the respective variable. mAHEI indicates modified Alternate Healthy Eating Index. Figure Legend:

4. Date of download: 11/12/2016 Copyright © 2016 American Medical Association. All rights reserved. From: Diet and Kidney Disease in High-Risk Individuals With Type 2 Diabetes Mellitus JAMA Intern Med. 2013;173(18):1682-1692. doi:10.1001/jamainternmed.2013.9051 Multivariable Multinomial Logit Model Adjusted With Known ConfoundersAssociation of animal protein, fruits and fruit juices, and leafy green vegetables and relative odds with 95% CI with chronic kidney disease (CKD) (A) or death (B) and respective histograms. For confounders, see the legend to Figure 2. The horizontal line on top shows tertiles, and the numbers within each tertile give the percentage of participants experiencing the respective event. The remaining independent variables of the adjusted multivariable model are depicted in the Supplement (eFigure 3). Figure Legend:

5. Date of download: 11/12/2016 Copyright © 2016 American Medical Association. All rights reserved. From: Diet and Kidney Disease in High-Risk Individuals With Type 2 Diabetes Mellitus JAMA Intern Med. 2013;173(18):1682-1692. doi:10.1001/jamainternmed.2013.9051 Multivariable Multinomial Logit Model Adjusted With Known ConfoundersAssociation of alcohol, 24-hour urinary sodium, and 24- hour urinary potassium and relative odds with 95% CI with CKD (A) or death (B) and respective histograms. For confounders, see the legend to Figure 2. The horizontal line on top shows tertiles, and the numbers within each tertile give the percentage of participants experiencing the respective event. The remaining independent variables of the adjusted multivariable model are depicted in the Supplement (eFigure 3). Figure Legend:

6. Date of download: 11/12/2016 Copyright © 2016 American Medical Association. All rights reserved. From: Diet and Kidney Disease in High-Risk Individuals With Type 2 Diabetes Mellitus JAMA Intern Med. 2013;173(18):1682-1692. doi:10.1001/jamainternmed.2013.9051 Forest Plot for the Multivariable Multinomial Logit Model Adjusted With Known Confounders For confounders, see legend to Figure 2. If not stated otherwise, food items are given in servings per week. Renal outcome column gives odds ratios (ORs) comparing participants alive with chronic kidney disease (CKD) with participants alive but without CKD; column death reports ORs comparing participants who died during follow-up with participants alive without CKD. For continuous independent variables, the ORs for the median of the second tertile (50.0th percentile [solid circle]) and the median of the third tertile (83.3rd percentile [open circle]) compared with the median of the first tertile (16.7th percentile) as reference are given. Independent variables in bold type have a significant association with CKD. The last column states the P value of inclusion of the respective variable. Figure Legend: