Chronic kidney disease (CKD) is one of the major worldwide health concerns [1], which has a significant contribution to the global burden of disease by incrementing cardiovascular disease (CVD) and mortality risk [2, 3]. According to the international guidelines, the existence of structural and functional damage, albuminuria, and glomerular filtration rate (GFR) below 60 ml/min/1.73 m2 considered as the definition of CKD if these circumstances have been lasted for at least three months and have not been caused by other causes [4]. Based on the latest Global Burden of Disease report in 2017, the CKD prevalence worldwide was estimated at 9.1%. This prevalence increased by about 29.3% in all ages compared to 1990 [1]. Recent national data has also reported that the overall CKD prevalence in Iranian adults was more than 10% [4]. An unhealthy lifestyle and poor diet, alongside some multifactorial abnormalities such as hypertension, obesity, and diabetes, are readily treatable major risk factors for the causes of CKD [5, 6].
Low-grade systemic inflammation is of considerable importance in the pathophysiology of CKD [7], which can significantly increase the morbidity and mortality associated with chronic nephropathy [8]. Studies have shown that an unhealthy lifestyle is an important risk factor for elevated systemic inflammation [9]. Also, the individual relationship of main lifestyle determinants, including diet, physical activity, body weight, and smoking with inflammation has previously been well established and shown that each one is an important determinant of predicting the inflammation status, per se [10,11,12,13]. However, due to the interactions of these factors with each other, it makes sense that the simultaneous study of the combined role of these factors significantly improves the ability to predict the outcome.
Accordingly, in recent years, the diet inflammation score (DIS) and lifestyle inflammation score (LIS) has been introduced by Byrd et al. to evaluate the ability of diet and other lifestyle-related factors to aggravate or modulate systemic inflammation levels in the body [14]. Indeed, DIS and LIS determine the collective contributions of lifestyle and diet exposures to systemic inflammation. Another index recently developed by Na et al. in the Korean population is food-based dietary inflammatory potential (FBDI), which uses the hs-CRP biomarker as a response variable to determine the inflammatory effects of food groups [15]. Although, to the best of our knowledge, no study has yet assessed the association of DIS, LIS, and FBDI scores with CKD risk, the relationship of these lifestyle and diet inflammatory scores with CKD-related disorders such as type 2 diabetes (T2D) and metabolic syndrome (MetS) and other chronic diseases has been investigated previously [15,16,17,18,19,20]. Two studies showed that higher DIS and LIS scores are associated with an increased risk of T2D and MetS [17, 20]. Other studies have also linked DIS and LIS scores to CVD and cancers [16, 18, 19]. Limited studies have examined the performance of FBDI [15, 21]. Na et al. showed that an increased FBDI score was associated with a higher prevalence of MetS [15].
Although there are no data on the clear kidney-related effect of lifestyle and diet inflammatory score in subjects with healthy renal function, the observed association of these inflammatory indices with the risk of T2D, MetS, and CVD that are closely related to renal disease, suggests a possible link between these inflammatory indices and the CKD risk. Therefore, this study aimed to investigate the relationship between DIS, LIS, and FBDI scores and CKD risk among Iranian adults.
Materials and Methods
Study participants
The present study was performed in the framework of the Tehran Lipid and Glucose Study (TLGS), a population-based cohort study conducted to examine the chronic diseases risk factors among a representative urban population of Tehran, including 15 005 participants aged ≥ 3 years [19]. The first survey of TLGS was initiated in March 1999, and data collection conducted prospectively at 3-year intervals is ongoing. The baseline survey was a cross-sectional study performed from 1999 to 2001, and surveys II (2002–2005), III (2006–2008), IV (2009–2011), V (2012–2015), and VI (2015–2018) were prospective follow-up surveys. The details of the TLGS have been reported previously [22]. In the third survey of the TLGS (2006–08), of 12 523 participants, 3568 were randomly selected for dietary assessment. Also, in the fourth survey (2009–2011), of 12 523 participants, 7956 randomly selected subjects agreed to complete dietary assessment.
For the current study, participants aged ≥ 18 years, with complete nutritional information on the third examination of TLGS and the new entries participants in the fourth examination, which was 7761, were included. Individuals with cardiovascular accidents and myocardial infarction (n = 81), and prevalent cancer (n = 16) were excluded. Also, the pregnant and lactating women (n = 195), those with under- or over-reported dietary energy intakes (out of the range 800–4200 kcal/d) (n = 492), and individuals with CKD in the baseline (n = 692) were excluded; some of them may fell into more than one category. Of 6365 participants at baseline, who were followed-up to sixth survey of TLGS (2015-2018), 321 were lost to follow-up, and 6044 remained for final analysis (follow-up rate: 94.95%) (Fig. 1).
Physical activity assessment
The individual’s physical activity data were collected by a modifiable activity questionnaire (MAQ), previously modified and validated among Iranian adults [23]. Individuals were asked to report and identify the frequency and time spent on activities of light, moderate, hard, and very hard intensity, over the past year, based on a list of common activities of daily life; we reported the total physical activity of each participant as metabolic equivalent/hours per week (Met.h.wk).
Demographic and anthropometric assessment
Trained and expert interviewers used a standard questionnaire to collect study population data on socio-demographic characteristics, including age (years), sex, education level (high school and diploma, academic education), smoking habit, medical history, and medications at baseline. We used a standardized mercury sphygmomanometer with an accuracy of two mmHg to measure the systolic blood pressure (SBP) and diastolic blood pressure (DBP). All blood pressure measurement was performed for each participant twice on the right arm with a minimum interval of 30 s after a 15-min rest sitting on a chair; we considered the mean of the two measurements to be the participants' blood pressure. We measured participants' body weight using a digital scale (Seca 881, Germany) to the nearest 100 g while the participants were in light clothes and without shoes. Height was measured by a stadiometer in a standing position without shoes and recorded to the nearest 0.5 cm. BMI was computed as weight (kg) divided by the height squared (m2).
Biochemical measurements
The biochemical variables, including fasting blood glucose (FPG), 2-h blood glucose, and serum creatinine were measured in participants. Based on the standard protocol, participants' blood samples were taken after 12–14 h of overnight fasting in a sitting position and centrifuged within 30–45 min of collection. We performed all blood analyses at the TLGS research laboratory and used the Selectra 2 auto-analyzer (Vital Scientific, Spankeren, The Netherlands) to analyze the samples. FPG was determined using an enzymatic colorimetric method with glucose oxidase. Both inter-and intra-assay coefficient variations were 2.2% for FPG. For the oral glucose tolerance test, 82.5 g of glucose monohydrate solution (equivalent to 75 g anhydrous glucose) was administered orally to participants aged > 20 years. A second blood sample was taken 2-h after glucose ingestion. Serum creatinine concentration was measured based on the standard colorimetric Jaffe Kinetic reaction method. Both intra- and inter-assay CVs were < 3.1%. We performed all analyses using commercial kits (Pars Azmoon Inc., Tehran, Iran).
Definitions
Hypertension (HTN) was determined in the study population based on SBP/DBP ≥ 140/90 mm Hg for individuals aged < 60 years and SBP/DBP ≥ 150/90 mm Hg for those aged ≥ 60 years or using current antihypertensive medication [24]. The criteria of the American Diabetes Association (ADA) were used to determine T2D in participants according to the following criteria: FPG ≥ 126 mg/dl or 2-h post 75-g glucose load ≥ 200 mg/dl or current blood glucose-lowering medications [25]. The Epidemiology Collaboration (EPI) equation formula, as described by Levey et al. [26], was used to calculate eGFR in participants. We expressed the eGFR was in ml/min/1.73m2 of body surface area. CKD was defined based on participants’ eGFR levels using the national kidney foundation guidelines as follows: eGFR ≥ 60 ml/min/1.73m2 as not having CKD and eGFR < 60 ml/min/1.73m2 as having CKD.
Dietary intake assessment
A valid and reliable 168-item semi-quantitative food frequency questionnaire (FFQ) with standard serving sizes was used to determine the participant’s dietary intake data in the last year [27]. Expert nutritionists with at least five years’ experience in TLGS asked individuals to report the frequency of their intakes for each food item on a daily, weekly, monthly, or yearly basis; portion sizes of consumed foods, reported in household measures, were then converted to daily grams of food intake. Considering that the Iranian Food Composition Table (FCT) is incomplete and has limited data on the nutrient content of raw foods and beverages, we used the United States Department of Agriculture (USDA) FCT. However, the Iranian FCT was used for local food items not listed in the USDA FCT.
Calculation of scores
DIS score were calculated based on the Byrd et al. study [14] based on 19 food components. High-sensitivity C-reactive protein and interleukins (interleukin-6, interleukin-8, and interleukin-10) were considered as response variables for the development of this inflammatory dietary index. According to each food item's effect on the inflammatory indicators levels, each item was assigned a specific weight, which could be positive or negative. DIS encompasses originally included leafy greens and cruciferous vegetables, tomatoes, apples and berries, deep yellow or orange vegetables and fruit, other fruits, and natural fruit juices, other vegetables, legumes, fish, poultry, red and organ meats, processed meats, added sugars, high-fat dairy, low-fat dairy and tea, nuts, other fats, refined grains, starchy vegetables, and supplements intake. However, supplement intake was excluded from the calculation of DIS in this study due to the lack of information in our dataset; then, we computed the overall score with 18 food groups. To compute the DIS score, each food item was multiplied by its specific weight (explained in Byrd et al. study) to determine the weighted values of each item. The weighted values were then standardized using the Z-score (to a mean of zero and SD of 1). Finally, all the items' standardized weighted values were summed to calculate the DIS score for participants [14].
LIS score was calculated using the physical activity, BMI, and smoking status data based on the Byrd et al. study [14]. Due to religious and legal restrictions in the Iranian population, alcohol is not consumed, or its consumption is not reported, so we did not consider alcohol consumption to calculate the LIS score. First, a dummy variable was created from each component and then multiplied for proposed regression coefficients [14]: physical activity was categorized into tertiles, and participants in the first, second, and third tertiles gave 0.0, -0.18, and -0.41, respectively. Participants were categorized into average weight (BMI < 25), overweight (25 ≤ BMI < 30), and obese (BMI ≥ 30); and then respectively received 0.0, 0.89, and 1.57 scores. Also, the proposed regression coefficients for smokers vs. non-smokers were 0.50 vs. 0.0, which were assigned. Finally, all the weighted values were summed to calculate the LIS score.
FBDI score was calculated according to the Na et al. study [15]. They used the Spearman correlation analysis and multiple regression for selecting food components and creating a formula to determine the FBDI score. In the Na et al. study, the spearman correlation analysis between log hs-CRP and the 51 food components was determined to select food components (ten dietary components) that were considered a significant correlation. Finally, FBDI was developed based on ten selected food items: mixed coffee and sweetened drinks, white rice, green vegetables, eggs, citrus, legumes, red fruits, beef, bread and wheat flour, and nuts. To calculate the FBDI score, the intakes of each of the mentioned food groups were multiplied by its specific applied value (weight). Finally, all ten weighted intake values were summed to form an overall FBDI score [15].
Statistical analyses
We used the Statistical Package for Social Sciences (Version 20.0; SPSS, Chicago, IL) to perform all statistical analyses. The Kolmogorov–Smirnov test and histogram chart were used to assess the normality of variables. Baseline characteristics of the participants are expressed as the mean ± SD or median (25–75 interquartile) for quantitative variables and percentages for qualitative variables. Individuals were classified according to FBDI and DIS quartiles cut-off points. Chi-square and linear regression were used to test for trends of categorical and continuous variables across quartiles of FBDI and DIS (as the median value in each quartile), respectively. Multivariable Cox regression models were used with CKD as the dependent variable and FBDI, DIS, and LIS as independent variables to estimate the risk of incident outcomes. We reported the hazard ratios (HRs) and 95% confidence intervals (CIs). The first quartile of each above-mentioned lifestyle and diet inflammatory score was considered the reference group. The multivariable model was adjusted for potential confounding factors, including age, sex, educational level, daily energy intake, hypertension, type 2 diabetes, BMI (adjusted for FBDI and DIS), smoking (adjusted for FBDI and DIS), physical activity (adjusted for FBDI and DIS), and baseline eGFR level. The proportional hazards assumption was checked using a log–log plot, and the assumption was satisfied (lines in the plots were parallel). P-values < 0.05 were considered to be statistically significant.
In the current study, as additional analysis, we assessed the combined role of dietary inflammatory scores (FBDI or DIS) along with lifestyle inflammatory score (LIS) in predicting the risk of CKD incident; In order to perform this analysis, the total of DIS Z-score or FBDI Z-score with LIS Z-score (determined by using the SPSS software) were summed for all participants, and two new inflammatory scores including DIS-LIS and FBDI-LIS were determined for them. Then, individuals were divided into quartiles based on their scores for DIS-LIS and FBDI-LIS. Finally, the risk of CKD incident was determined for participants across the quartiles of these new inflammatory scores (Fig. 2A-B).
