Dietary intakes of total polyphenol and its subclasses in relation to incidence of chronic kidney disease: a prospective population-based cohort study

Background: Chronic kidney disease (CKD) is nowadays a public health challenge worldwide. Accordingly, this study is aiming to evaluate the association between long-term intake of total polyphenol and its subclasses, and the incidence of CKD. Methods: A sample of 3021 Iranian Adults (47% men) were selected from the Tehran Lipid and Glucose Study population. The participants aged 20-79 years and had no diagnosis of CKD at baseline. Total polyphenol intake and its major subclasses including avonoids, phenolic acids, stilbenes, and lignans, was assessed by a validated and reliable food frequency questionnaire, and were categorized as avonoids, phenolic acids, stilbenes, and lignans. CKD should be dened by either reduction in eGFR or by morphological abnormalities of the kidneys or by abnormalities in the urinalysis persistent for 3 months. Since GFR is generally accepted as the best overall index of kidney function, in current study CKD was exclusively dened as eGFR <60 mL/min/1.73m 2 . The Modication of Diet in Renal Disease Study equation was used to calculate the estimated glomerular ltration rate (eGFR). Hazard ratio and 95% condence intervals of CKD by total polyphenols quartiles were assessed by Cox-regression analysis. Results: In this study, we documented 355 cases of diagnosed CKD over 11,058.464 person-years. The median [IQR] age of participants was 36 [27-46] years at baseline. After adjustment of the potential confounders, it was revealed that a moderate intake of lignans ( ≤ 6.8 mg) was negatively associated with the incidence of CKD, whereas there was no signicant association between the higher amounts of lignan intake and CKD. No signicant associations were observed between the consumption of total polyphenols and the incidence of CKD (HR: 0.97, 95% CI 0.67-1.40). Conclusions: Data of the current study suggest that in the case of lignan with protective properties, a moderate amount of lignan favorably reduced the incidence of CKD by approximately 32%, whereas higher amounts possessed a null effect.

confounders, it was revealed that a moderate intake of lignans (≤6.8 mg) was negatively associated with the incidence of CKD, whereas there was no signi cant association between the higher amounts of lignan intake and CKD. No signi cant associations were observed between the consumption of total polyphenols and the incidence of CKD (HR: 0.97, 95% CI 0.67-1.40).
Conclusions: Data of the current study suggest that in the case of lignan with protective properties, a moderate amount of lignan favorably reduced the incidence of CKD by approximately 32%, whereas higher amounts possessed a null effect.

Background
Chronic kidney disease (CKD), as a global public health challenge, manifests by loss of kidney function or structural kidney damage (1,2). Obesity, diabetes, hypertension and increased levels of in ammation and oxidative stress, as the underlying risk factors for cardiovascular disease (CVD), are also known as the most common chronic conditions associated with CKD (1,3). In this regard, habitual dietary intake is considered as an effective modi able factor in the etiology of CKD. Therefore, adopting a healthy and feasible lifestyle appears to be a powerful strategy in the prevention of CKD (4).
It is believed that adherence to speci c dietary patterns including the Mediterranean diet (5) and the Dietary Approaches to Stop Hypertension (DASH) (6, 7) has a favorable contribution in the renal function.
An increasing body of evidence has consensus that plant-based diets characterized by higher intake of fruits, vegetables, nuts, spices, herbs, cereals, legume, seeds, chocolate, and tea better prevent the occurrence of chronic conditions (8). These plant-based food items are well-known for their diverse content of phytochemicals (9). Thus far, polyphenols are referred to as the major constituent of human consumed micronutrients; with considerable antioxidant and biological activities, they are believed to bene cially affect various chronic disorders (10,11). Flavonoids, phenolic acids, stilbenes, and lignans are the four major subclasses of polyphenols (8). While several studies have investigated the effect of polyphenol-rich red grape juice (12), green tee, coca (13) and white wine (14) among patients with renal failure, there is an ongoing discussion about the CKD-preventive property of polyphenols in a regular diet.
With regards to the lack of data on the relationship between long-term intake and polyphenols with kidney function among adults, this study aimed to investigate the association of long-term consumption of total polyphenol and its subclasses, including avonoids, phenolic acids, stilbenes, and lignans with the incidence of CKD

Study population
This longitudinal study was performed within the framework of the ongoing community-based Tehran Lipid and Glucose Study (TLGS). The prospective TLGS was originally conducted to determine and prevent the risk factors for non-communicable diseases (NCDs) (15). In this respect, TLGS consists of 5 phases, the rst of which was a cross-sectional study and began in 1999. Four other subsequent phases Ultimately, a representative sample of 3021 individuals with 20-79 years of age and complete data, were recruited and followed to the fth phase. Based on the exclusion criteria, participants on pregnancy or lactating states, on energy consumption out of the prede ned limits (800 > x > 4200 kcal/d) or speci c diets or with any history of myocardial infarction, cerebral vascular accident, cancers and CKD at baseline were excluded. Hence, a total of 2054 individuals were enrolled for the nal follow-up analysis with 5.4 years of mean follow up time.
All participants were initially asked to provide written informed consent. The study protocol was also approved by the ethics committee research council of the Research Institute for Endocrine Science (RIES), Shahid Beheshti University of Medical Science, Tehran, Iran.

Dietary assessment
The habitual food intake was evaluated by a validated and reliable semi-quantitative food-frequency questionnaire (FFQ) at baseline (16,17). The individual consumption frequency of each food item was designated by trained and experienced dietitians on daily, weekly or monthly basis. The portion sizes were collected in household measures and converted to grams. The USDA Food Composition Table (FCT) was used to calculate and interpret the energy and nutrient content of each food item. The estimation of total polyphenol and subclasses' intake was based on the Phenol-Explorer database (www.phenolexplorer.eu/contents) (18).

Measurement of covariates
The physical activity level of each participant was assessed by the Modi able Activity Questionnaire which has previously been validated for the Iranian population (19). A metabolic equivalent (MET-h per week) was calculated according to a list of most common and daily routine activities.
Anthropometric measurements including weight and height were collected to the nearest 0·1 kg and 0·1 cm, respectively. The weight was recorded in light clothing via a SECA digital weighing scale (Seca 707; Seca Corporation; range 0·1-150 kg), and height was taken without shoes on. BMI was calculated as weight (kg) divided by square of height (m 2 ). Arterial blood pressure was measured manually, using a mercury sphygmomanometer with a suitable cuff size for each participant after a 15-min rest in the supine position.
Systolic (SBP) and Diastolic blood pressures were determined by the initial tapping and disappearance of Korotkoff sound, respectively. Blood pressure was measured twice and the average was considered as participant's nal blood pressure. Blood samples were taken from all participants at the TLGS research laboratory after 12 to 14 hours of overnight fasting.
Fasting plasma glucose (FPG) and 2-h plasma glucose (equivalent to 75 g anhydrous glucose; Cerestar EP) were measured by enzymatic colorimetric method using glucose oxidase, with both inter-and intraassay CV being < 2%. Serum creatinine was measured under the standard colorimetric Jaffe_Kinetic reaction method at baseline (2006-2008) and after 6 years of follow-up (2012-2015). Both Intra-and inter-assay CVs were below 3·1%; all analyses were performed using commercial kits (Pars Azmoon Inc.).
De nition CKD should be de ned by either reduction in eGFR or by morphological abnormalities of the kidneys or by abnormalities in the urinalysis persistent for 3 months. Since GFR is generally accepted as the best overall index of kidney function, in this study, the estimated GFR (eGFR) was expressed as ml/min/1.73m 2 of body surface area, using the Modi cation of Diet in Renal Disease (MDRD) equation (20). The MDRD equation is as follows: Patients were classi ed based on individual eGFR levels pertain to the national kidney foundation guidelines (21); given this, eGFR ≥ 60 ml/min/1.73m 2 are considered as non-CKD patients and eGFR < 60 ml/min/1.73m 2 are those diagnosed with CKD. Hypertension was primarily de ned as SBP/DBP ≥ 140/90 mm-Hg or current therapy for a de nite diagnosis of hypertension (22). Furthermore, diabetes was de ned in accordance with the criteria of the American Diabetes Association (ADA) as fasting plasma glucose ≥ 126 mg/dl, 2-h post 75-g glucose load ≥ 200 mg/dl or current therapy for a de nite diagnosis of diabetes (23).

Statistical analysis
In this study, the normality distribution of variables was assessed by the Kolmogorov-Smirnov test and Histogram chart. The participants' characteristics and nutritional states across quartiles of total polyphenols were presented by mean ± SD and median [IQR] for normal and skewed distribution, respectively; and by percentages for categorical variables. Linear regression and Chi-square tests were used to investigate the trend of continuous and categorical variables in proportion to total polyphenols quartiles, respectively. Hazard ratio (HR) and 95% con dence intervals (CI) of incident CKD were assessed by quartiles of total polyphenols by Cox regression analysis; the lowest quartile of dietary exposures considered as a reference group. Moreover, three models were speci ed for the rst being the crude, the second with adjustments for sex, age, physical activity and total calorie intake, and third additionally adjusted for BMI, diabetes and hypertension. The proportionality assumption underlying the Cox model was examined, and no evidence of violation was observed. All analyses were done using IBM SPSS version 16 (SPSS, Chicago, IL, USA) with the signi cance level set as P < 0.05 (two-tailed).

Results
In this study, the median [IQR] age of participants was 36 [27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44][45][46] years, with 47% being men. Within 11058.464 person-years of follow-up, 355 new cases of CKD were reported. Baseline characteristics and nutritional status of participants across quartiles of total polyphenols are stated in Table 1. The data had highlighted that across quartiles of total polyphenols, trends of age (P for trend= 0.015), BMI (P for trend= 0.001), energy (P for trend< 0.001), and fat (P for trend= 0.003) had increased, whereas, the trend of carbohydrates was mitigated (P for trend< 0.001). The HR for total polyphenol and its subclasses are described in Table 2. There were no signi cant associations between the consumption of total polyphenols and the incidence of CKD (HR: 0.97, 95% CI 0.67-1.40). Furthermore, the incidence of CKD was less signi cant among participants in the fourth quartile of avonoids and phenolic acids as compared to those in the rst quartile (HR: 1.07, 95% CI 0.74-1.55 and HR: 1.14, 95% CI 0.79-1.64, respectively). Across subclasses of polyphenols, lignans had a signi cant association with CKD in the multivariable-adjusted model. With that said, participants in the second and third quartile of lignans had a signi cantly reduced incident rate of CKD by 34% and 31%, respectively, comparing to those in the rst quartile. Also, participants in the rst and fourth quartiles of total dietary polyphenols were shown to be non-signi cantly associated with the incidence of CKD.
Compared to participants in the rst quartile of stilbenes, the HR was 0.64 (95% CI 0.48-0.87) for those in the last quartile of the unadjusted model and with a signi cant inverse but unstable trend that disappeared after adjusting for potential confounders. The association of total polyphenol and its subclasses with CKD did not substantially change when we further adjusted for fat, carbohydrate, whole grains, vegetables, fruits, and nuts.

Discussion
This study investigated the association of total dietary polyphenol and its major subclasses with the incidence of CKD among adults in Tehran, Iran. While high intake of lignans up to 6.8 mg is negatively associated with the incidence of CKD independent of the potential cofounders, no similar associations were depicted with higher values (> 6.8 mg). Also, it was observed that the incidence of CKD in the highest quartile of stilbenes was 36% lower comparing to the lowest quartile in the unadjusted model. There were no signi cant ndings after controlling the confounding factors, which was speci cally dependent on the effect of potential confounders. Besides, no signi cant associations were detected between the consumption of total polyphenols and the incidence of CKD.
To the best of our knowledge, this is the rst study that evaluated the association of long-term consumption of total polyphenol and its major subclasses with the incidence of CKD. This study has implied that a moderate amount of lignan (≤6.8 mg) is CKD-protective, which can be attributed to the high content of antioxidants. Motivated by a piece of emerging evidence, reduced levels of oxidative stress, as a consequence of increased antioxidant defenses, correspondingly decreases the CVD risk factors (24). In this context, some epidemiological studies have emphasized that CVD and CKD share some common risk factors including low serum HDL cholesterol, hypertension, hypertriglyceridemia, and hyperglycemia (22,25). The cardiovascular risk factors tend to progressively increase as a result of renal function reduction, however, the proper management of the cardiovascular system, could decrease the risk of CVD and CKD manifestations (22,25). Therewith, data on the consumption of lignan remains inconsistent (26,27); it is believed that a moderate intake promotes bene cial health effects. However, high amounts of lignan could act as estrogen antagonists (28) or inhibit enzymes involved in the metabolism of sex hormones such as 5-a-reductase and 17b-hydroxysteroid dehydrogenase (23). In other words, a higher intake of lignan is believed to have null effects on CVD risk factors, for which as few as two mechanisms have been proposed for (23). In fact, with higher intake of lignan and the mechanisms mentioned earlier, the increased CVD risk factors can decrease the level of free estradiol and testosterone in women and men, respectively (23). Therefore, it seems as only a moderate amount of lignan (≤6.8 mg) may appear effective with an approximate 32% decrease of the CVD risk factors, and subsequently the risk of CKD.
Borges et al. performed a study to propose the CKD-reducing properties of green tee polyphenols by decreasing albuminuria among patients with diabetic nephropathy (4). The inhibition of in ammatory mediators (such as TNF-a) is the possible mechanism behind this investigation. Within this context, it was also indicated that total polyphenol intake had no association with a lower risk of CKD (4). The results of another study by Cynthia et al. contradicted the previous evidence, mostly attributed to the intervention of multifactorial treatments received by the diabetic participants along with the green tee polyphenols. In this sense, a meta-analysis endorsed the preliminary support of polyphenol-rich interventions in the improvement of cardiovascular risk factors among hemodialysis patients (29). Despite individual studies indicating signi cant improvements, pooled results depicted no effect for most outcomes excepting myeloperoxidase, diastolic blood pressure, and triglycerides (29). Only myeloperoxidase, a measure of oxidative stress, had a large pooled effect size (29). Also, the individualized polyphenol metabolism as a result of inter-individual gastrointestinal microbiota differences, brought diversity to the range of responses among the population (29). In this respect, no speci c associations were observed in the mentioned study, which could conceal the bene cial effect of total polyphenol consumption.
Concerning the current study, the potential limitations must be considered, as well. First, the consumption of total polyphenols and its subclasses were estimated by the Phenol-Explorer database (www.phenolexplorer.eu/contents) for the Iranian version not being available. Second, CKD is de ned as abnormalities of kidney structure or function, present for > 3 months, with implications for health, but in current study only reduction in eGFR was exclusively considered as CKD de nition. Third, the albumin-to-creatinine ratio (ACR) was not assessed; the ACR could have helped us interpret the results more accurately. Fourth, this study did not take into account a creatinine double-checking procedure, which could have possibly affected the ndings. Fifth, perhaps the sample size was small to defect signi cant associations. Finally, although the effect of potential confounders were adjusted, the impact of residual confounders could not be ruled out.

Conclusions
In conclusion, the result of this prospective study did not con rm any associations between total polyphenols and the incidence of CKD. However, it was nalized that despite the CKD-protective effect of moderate lignan intake (≤ 6.8 mg) by approximately 32% among adults of Tehran, higher amounts of lignan possess no signi cant effects. Therefore, there is no doubt that further research is required to differentiate the impact of low, moderate and high amounts of lignan intake on CKD incidence.