Chinese Empirical Dietary Inflammatory Index Is Associated with High Risk of Systematic Inflammation and Aging in Rural China

doi:10.21203/rs.3.rs-2964068/v1

Background：China is rapidly transforming into an aging society and houses the largest elderly population in the world. Diet patterns increase inflammation and oxidative stress, which may provide effective intervention strategies for mitigating the health burden in aging societies. This study sought to develop a Chinese empirical dietary inflammatory index(CEDII) that assesses Chinese diet quality based on its inflammatory potential and examine whether the index was associated with an accelerated aging process.

Methods:A cross-sectional study conducted in the community of Beijing recruited 388 participants. Dietary intake was collected based on 24-h dietary recalls. The Chinese Diet Balance Index(DBI) and its derivatives, as well as the empirical dietary inflammatory pattern(EDIP) based on the western diet, were calculated to assess dietary quality in different aspects. Chinese empirical dietary inflammatory index(CEDII) was derived using reduced-rank regression(RRR) according to Chinese diet patterns. The population was categorized into tertiles of the CEDII. Leukocyte telomere length(LTL) and mitochondrial DNA copy number(mtDNAcn) were determined by polymerase chain reaction assay. Tumor necrosis factor α(TNFα) and interleukin-6(IL-6), 8-oxo-2'-deoxyguanosine (8-oxo-dG), superoxide dismutase (SOD) activities, and glutathione reductase (GR) were measured.

ResultsA total of 388 subjects were enrolled. High CEDII group was characterized by high intake of carbohydrates. This pattern predicted higher levels of TNFα, IL-6, shorter LTL and lower mtDNAcn. Neither DBI nor EDIP was correlated with biomarkers of inflammation or aging. High CEDII scores were associated with short LTL and decreased mtDNAcn using multivariate linear regression with adjustment for confounding variables. The stratified analysis demonstrated that the high CEDII group had an increased risk of short LTL in subjects with abnormal metabolic status.

Conclusions: This study developed a Chinese empirical dietary inflammatory index(CEDII). High CEDII scores were associated with an increased risk of aging. Compared with indexes related to DBI and EDIP, CEDII was much more effective in assessing inflammation and aging potential of diet patterns for Chinese residents.

the inflammatory potential of the diet

aging

inflammation

oxidative stress

The proportion of elders is steadily increasing worldwide due to significant public health and medical practice improvements. However, increasing lifespan is not necessarily mean being healthier. Dietary interventions have been identified as one of the primary strategies for aging-related diseases, yet often neglected in the past¹.

Chronic inflammation is widely held to play an essential role in aging and its related diseases², and is correlated with dietary patterns. A systematic review included 17 relevant studies has demonstrated that dietary components could drive systemic inflammation by altering the nonenzymatic glycation of proteins, nucleic acid³ and lipids. Existing evidence has indicated that dietary patterns rich in vegetables and fruit, fiber, or low-fat dairy have an inverse association with inflammation. In contrast, dietary patterns rich in saturated fatty acids, and micronutrients such as sodium and ultra-processed foods, are positively correlated with inflammation⁴. Therefore, we hypothesized that dietary patterns might be particularly relevant for aging by modulating systematic inflammation.

It is challenging to capture a comprehensive overview of the complex interconnection of dietary exposures when exploring the role of individual nutrients and foods in disease risk. Dietary patterns may be more informative for addressing the association between diet and disease^5,6. There are two dominant approaches to developing dietary patterns, the prior approach and the posterior approach. The priori approach is created according to the already available evidence regarding current dietary guidelines. In the present study, we adopted the Diet Balance Index (DBI) as the prior approach to assessing dietary quality of the study population based on the dietary guidelines for Chinese residents^7,8. However, the prior dietary index cannot accurately predict clinical outcomes. Reduced Rank Regression (RRR) is one of posterior approaches, created based on food groups by using data on selecting response variables related to the disease of interest. Researchers from the USA have developed an empirical dietary inflammatory pattern (EDIP) based on the western dietary pattern by RRR⁹. Chinese traditional diets are rich in vegetables and grain in comparison to the western diet. Therefore, EDIP may not apply to Chinese residents. The current study intended to develop Chinese empirical dietary inflammatory index(CEDII), a dietary index based on traditional Chinese food compounds by RRR, to assess diet inflammatory potential.

Recently, several studies have confirmed that EDIP plays an essential role in aging diseases, like cardiovascular diseases, diabetes, and metabolic syndrome^10,11. This indicates that diet inflammation potential may have an impact on aging progress. Thus, we aimed to investigate the association of different dietary indexes, especially CEDII, with inflammation and aging, in order to provide a new light on coping with aging.

Study participants

The participants were recruited from a Chinese rural community cohort from March 2014 to January 2015¹⁰. Overall, 599 individuals provided informed consent to participate. Trained researchers administered a questionnaire to obtain information on gender, age and medical history. All participants received a standardized 75 g oral glucose tolerance test (OGTT). Individuals were considered ineligible if they were positive for antibodies of type 1 diabetes and suffering from cardiovascular and cerebrovascular diseases or chronic kidney diseases, or taking steroids, antidiabetic or lipid-lowering agents in the past three months. Additional criteria for exclusion were subjects without assessment of LTL or mtDNAcn, or 24h food recalls, or with extreme calorie intake (< 500kcal/d or > 5000 kcal/d). A total of 388 participants were included in the final analysis(Supplementary Fig. 3). The ethics committee of Peking Union Medical College Hospital approved this cross-sectional study.

Anthropometric measurements

Weight(Wt), Height(Ht), waist circumference(WC), hip circumference(HC) and blood pressure (BP) were obtained following the standard method as described previously¹². Body mass index(BMI) was calculated as weight in kilograms divided by the square of the height in meters(kg/m²).

Laboratory analysis

Venous blood from each subject who fasted overnight was collected. Fasting blood samples were drawn for lipid profile, oxidative stress and inflammation biomarkers, LTL and mtDNAcn. Total cholesterol (TC), triglyceride (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), uric acid(UA) and creatinine(Cr) were detected using an automated analyzer. HbA1c was assessed by high-performance liquid chromatography (HPLC, intra-assay coefficient of variation (CV) < 3%, inter-assay CV < 10%). Immediately after, participants underwent a standardized 75 g oral glucose tolerance test (OGTT). Blood samples for plasma glucose, insulin and C-peptide were determined at 4-time points(0, 30, 60 and 120minutes). Urine albumin-creatinine ratio(ACR) was assayed from morning urine samples. The estimated glomerular filtration rate (eGFR) was calculated according to Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) Eq. 1³.

Inflammation indicators, including serum tumor necrosis factor α (TNFα), interleukin 6 (IL-6), and oxidative stress indicators, including 8-oxo-2′-deoxyguanosine (8-oxo-dG), superoxide dismutase (SOD) activities and glutathione reductase (GR) were determined using Elisa kits (Cloud-Clone Corp, Houston, USA) by the Beijing Institute of Biotechnology. LTL and mtDNAcn were measured using quantitative polymerase chain reaction (qPCR). LTL was expressed as the ratio of repeat-copy number of the telomere (T) to a single-copy gene (S) copy number. MtDNAcn was expressed as the ratio of mtDNAcn to β-actin copy number.

Assessment of Insulin Sensitivity and β-Cell Function

The homeostatic model assessment of insulin resistance (HOMA-IR) was obtained for assessing fasting-based insulin resistance: HOMA-IR = fasting blood glucose (mmol/L) × fasting insulin (IU/mL)/22.5. Matsuda index(ISIM) used the following formula to index whole-body insulin sensitivity derived from OGTT¹⁴. ISIM=\(\frac{10000}{\sqrt{(Fasting Glucose\times Fasting Insulin)\times (Mean Glucose\times Mean Insulin)}}\).The hepatic insulin resistance index (HIRI) was calculated by the formula: HIRI= AUC_PG(0–30 min) × AUC_Insulin(0–30min)¹⁵. The homeostasis model assessment of insulin secretion (HOMA-β) was represented for basal β-cell function: HOMA-β = fasting insulin (IU/mL)/(fasting blood glucose (mmol/L) – 3.5). The estimate of first-phase insulin secretion was calculated using the formula of insulinogenic index 30 (IGI30)¹⁶. IGI30 = (Ins_0min – Ins_30min)/(PG_30min – FPG). The disposition index (DI) was used to evaluate integrated islet β-cell function that accounts for insulin sensitivity¹⁷. DI = (AUCInsulin_{(0–120 min)}/AUCPG_(0–120min)) × ISIM

Dietary assessment

During face-to-face interviews, trained interviewers estimated dietary data using 24h food recalls¹⁸. Certified dieticians analyzed energy and nutrient content of foods based on the China Food Composition Table (2004). Moreover, the energy intake ratios of macronutrients were extracted.

Assessment of Empirical Dietary Inflammatory Pattern and Diet Balance Index-16

Empirical dietary inflammatory pattern (EDIP) is devised to capture overall inflammatory potential of whole diets, as described in more detail previously^9,19. Briefly, 39 predefined food groups were added into reduced-rank regression(RRR) models followed by stepwise linear regression analyses to develop a dietary pattern most predictive of inflammatory indices (tumor necrosis factor-α receptor 2(TNFαR2), interleukin-6(IL-6) and C-reactive protein (CRP)). The weighted sum of food groups selected by the regression models^9,20 constitutes the EDIP. A higher score indicates more inflammatory potential of diets, and a lower score indicates anti-inflammatory potential of diets. Individuals were divided into 3 categories according to the EDIP tertile(low EDIP<-95.58,n = 129; -95.58 ≤ moderate EDIP < 65.24,n = 130; high EDIP ≥ 65.24,n = 129)

Dietary balance index-16(DBI-16) is developed to assess inadequate and excessive food intake among Chinese populations according to the Chinese Dietary Guidelines 2016²¹. DBI-16 comprises 8 food items, including cereals, vegetables and fruits, beans and dairy products, animal foods, empty energy foods, condiments, diet variety, and water. Due to a lack of water intake, the other 7 food items were enrolled in the current study. Based on the scores of each DBI-16, 2 indicators of diet quality were assessed. Diet quality distance(DQD)-DBI was defined as the sum of the absolute values of positive and negative scores and represented imbalanced food intake²². Individuals were divided into 3 categories according to the DQD-DBI tertile(low DQD-DBI < 45,n = 134; 45 ≤ moderate DQD-DBI < 51,n = 130; high DQD-DBI ≥ 51, n = 124).

Development and Assessment of Chinese Empirical Dietary Inflammatory Index

Chinese Empirical Dietary Inflammatory Index(CEDII) was developed to obtain the overall inflammatory potential of diets in Chinese community residents. Food components were aggregated into 27 defined food groups based on dietary habits of the study population and the China Food composition 2009. These food groups were considered as predictor factors, and inflammatory indicators(IL-6 and tumor necrosis factorα(TNFα)) were considered as response factors. The PROC PLS procedure with the RRR method²³ option was performed to identified inflammatory dietary patterns, and developed two kinds of dietary pattern scores. The first dietary pattern score was retained for further analysis because it reflected the greatest amount of variation in food groups and the inflammatory potential of whole diets compared to the remaining dietary pattern scores. Then a stepwise liner regression analysis was adopted to derive the most essential food groups contributing to the first dietary pattern, with the first dietary pattern score as response variable, and food groups as predictor variables. The level of significance was set as p = 0.1 for entry into and retention in the model. The CEDII was defined as the sum of the intake of essential food groups weighted by regression coefficients according to the stepwise linear regression(Supplementary Table 1).

Statistical analysis

The characteristics were expressed as mean ± standard deviation with normal distribution and median (25th percentile, 75th percentile) with non-normal distribution. The RRR was performed through SAS statistical software (version 9.2; SAS Institute, Cary, NC, USA.) to derive the dietary patterns.

Differences between groups of each dietary index were adopted using one-way analysis of variance(ANOVA)( Bonferroni correction) or Kruskal-Wallis tests. Categorical variables were compared by liner-by-liner associated χ2 test. The correlations of each dietary pattern with aging, inflammation, oxidative stress, glucose and lipid metabolism biomarkers was analyzed with spearman analyses. Subsequently, partial correlations were applied to obtain the correlation between variables with adjustment of age, gender, BMI, WC, HC, BP, HbA1c, lipid profile, UA, eGFR, AST, ALT and daily energy intake. Multivariate regression analyses were performed to identify the contribution of CEDII to LTL after adjustment for the confounding variables. Multicollinearity was checked with a threshold of variance inflation factor(VIF) < 10. Cochran-Mantel-Haensel (CMH) test was used to examine the effect of CEDII on short-LTL and low-mtDNAcn after stratification by risk factors of metabolic diseases, including age(≤ 50 years or > 50 years), gender(male or female), TG(≤ 1.7mmol/L or > 1.7mmol/L), HDL-c(< 1.0mmol/l or ≥ 1.0mmol/L), and LDL-c(< 2.6mmol/L or ≥ 2.6mmol/L), and the presence or absence of overweight (≥ 24kg, according to Group of the Working Group on Obesity in China)²⁴, DM(FPG ≥ 7.0mmol/L, or 2hPG ≥ 11.1mmol/L, or HbA1c ≥ 6.5%, according to Chinese Diabetes Society 2020²⁵), Raised BP(SBP ≥ 130mmHg or DBP ≥ 80mmHg).

IBM SPSS Statistics software (version 26, IBMCorp., Armonk, NY, USA) and R (R studio version 1.3.959 and https://hiplot.com.cn/) were conducted for statistical analyses. Statistical tests were set at a 2-tailed P < 0.05.

Dietary patterns among Chinese residents in the current study

Supplementary Figure 1 showed the factor loadings of the dietary pattern based on food groups. Intakes of other vegetables, eggs, starch products, condiment were positively related to CEDII(loadings≥0.1). Intakes of coarse grain and oil and fats were negatively related to CEDII(loading≤-0.1)(Supplementary Figure 1 and Supplementary Figure 2).

Demographic Characteristics of the study population

388 subjects enrolled in the final analysis were classified into three categories based on the tertiles of CEDII levels(Supplementary Figure 3)(Low CEDII<-0.1124,n=126; -0.1124≤Moderate CEDII<0.0056,n=130; High CEDII≥0.0056,n=132). The demographic characteristics were shown in Table 1. Individuals in high CEDII group showed the highest level of HbA1c. There were no statistically significant differences in demographic characteristics between dietary patterns. However, the high CEDII group consumed more carbohydrates and fat, while the low CEDII group consumed more fat (Supplementary Table 2).

Biomarkers of inflammation, oxidative stress and aging among tertiles across different dietary indexes

In general, with increasing tertiles of CEDII, TNFα of participants tended to be elevated(p<0.001). Levels of IL-6 were the highest in participants with high CEDII. Overall, there were no significant differences in biomarkers of oxidative stress. Regarding aging indicators, individuals who consumed high CEDII dietary patterns had lower levels of LTL and mtDNAcn, than those who consumed low CEDII dietary patterns. (Table 2)

In contrast, indicators classified by EDIP or DQD-DBI did not appear to have any substantial differences, except for SOD activities. Participants with moderate EDIP had lower SOD activities than those in low EDIP and high EDIP groups (p=0.02). (Table 2)

Association of different dietary indexes and biomarkers of inflammation, oxidative stress and aging

CEDII scores were positively associated with TNFα(r=0.318, p<0.001), IL-6(r=0.138, p=0.012), SOD activities(r=0.129, p=0.020), LTL(r=-0.151, p=0.003), mtDNAcn(r=-0.175, p<0.001). There were no significant associations between CEDII and 8-oxo-dG, GR. Partial correlation analyses after adjusting for all confounders(including age, gender, BMI, WC, HC, SBP, DBP, HbA1c, TC, TG, HDL-c, UA, eGFR, AST, ALT, energy intake) showed CEDII still significantly correlated with TNFα(r=0.278，p<0.001), IL-6(r=0.106，p=0.044), LTL(r=-0.125, p=0.017), mtDNAcn(r=-0.127, p=0.016), but not SOD activities(r=0.067，p=0.207). Additionally, food varieties were inversely associated with TNFα(r=-0.110，p=0.031) and LTL(r=0.108, p=0.034). (Figure 1)

Regression analyses for CEDII and ageing biomarkers

Multiple linear regression models including confounders were conducted to estimate the independent contributor of CEDII to ageing biomarkers. CEDII was firmly associated with LTL(β=-0.332，p=0.024) and mtDNAcn(β=-0.073，p=0.026) after adjusting for age, gender, BMI, WC, HC, SBP, DBP, energy intake, eGFR, UA, and indicators of oxidative stress, glucose and lipid metabolism, and liver function. (Table3)

Subgroup analyses of the associations of aging with CEDII across risk factors for aging

Participants were categorized into High CEDII(H_CEDII, individuals in high CEDII group) and Low and Moderate CEDII(LM_CEDII, individuals in moderate and low CEDII groups). The lower third of LTL of the study population was taken as the boundary of Short LTL group(LTL≤28.37) and Long LTL group(LTL>28.37). Taking the lower third of mtDNAcn as the boundary, Low mtDNAcn was defined as mtDNAcn≤82.25 and High mtDNAcn was defined as mtDNAcn＞82.25.

The risk for short LTL predicted by CEDII elevated in participants with overweight(OR 2.66, 95%CI: 1.56-4.6), raised BP(OR 3.15, 95%CI:1.65,6.01) or elevated TG(OR 3.27, 95%CI: 1.65,6.86). The inverse relationship between CEDII and LTL was similar when examined in different subgroups of age, gender, glucose, HDL-c and LDL-c. CEDII and mtDNAcn was negatively related when examined in different subgroups. (Figure 2)

The current study set out to develop a dietary index of dietary inflammatory potential based on food items driven from the traditional Chinese diet using RRR and stepwise linear regression, and identify the association between inflammatory potential of whole diets and aging. Low CEDII group was characterized by a higher intake of coarse grain and fat. In comparison, high CEDII group was characterized by a higher intake of carbohydrates consisting mainly of refined grains, higher consumption of meat and eggs and a lower intake of fat. This study showed that CEDII was robustly associated with inflammatory indicators but not all oxidative stress indicators. High inflammatory potential of habitual dietary patterns had increased risk of aging. Subgroup analyses highlighted that the risk of short LTL was higher in participants of high CEDII with overweight or raised BP or hypertriglyceridemia. However, we did not observe significant relationships between DQD-DBI, EDIP and indicators of inflammation, oxidative stress or aging.

Diet plays a prominent role in aging²⁶, although the underlying mechanisms remain undetermined. This study suggested that dietary inflammatory potential might be partially involved in aging. In contrast to those consuming low CEDII diet, individuals consuming high CEDII diet had shorter LTL and lower mtDNAcn. In addition, after adjusting for variable confounders, CEDII were firmly inversely related to aging indicators. Overall, these results indicate that CEDII represents a novel dietary inflammatory index, which could assess the association between dietary inflammation potential and aging based on Chinese diet. Simultaneous evaluations of associations of DQD-DBI and EDIP with aging were further conducted. No significant relationship between DQD-DBI and aging were found in the current study. This might due to that DQD-DBI is created by prior approach based on the Chinese Dietary Guidelines 2016²¹, which might lead to weak links to the clinical outcome. EDIP is a posterior dietary index based on whole diets’ inflammatory potential¹¹. It is developed according to Western dietary patterns distinguished by high intake of desserts, processed meat, and refined grains, which differs from the traditional Chinese dietary pattern characterized by high intake of carbohydrates. This may explain why EDIP was not correlated with biomarkers of inflammation and aging in the current study population. By contrast, CEDII was consistently associated with biomarkers of inflammation and aging after adjusting for potential confounding variables, indicating that CEDII might have greater abilities to predict the association of inflammatory potential of Chinese diet and aging than other dietary indexes.

The present study also found that a diet high in coarse grain and fat, and low in protein-to-carbohydrate ratio contributed to more favorable biological aging, which has been confirmed by both animal experiments and population-based cohorts previously.^{27,28 29}Advanced glycation products(AGEs) played an essential role in aging and were impacted by dietary patterns. A randomized controlled trial from Australia has demonstrated that a diet with high intake of refined grains, meats and processed meats might primarily be responsible for elevated AGEs³⁰. A previous research, including 15 428 adults aged 45–64 years in four American communities, have suggested diets high in carbohydrates and low in carbohydrates are both associated with increased mortality, with minimal risk observed at 50–55% carbohydrate consumption³¹. Hence the low CEDII group characterized by moderate carbohydrate intake with high coarse grain may delay the progression of aging.

In addition, fat intake was also associated with aging. High-fat diets(35–45% of energy from fat) were associated with an increased risk of chronic inflammation accompanied by aging³². However, the current study highlighted that the low CEDII group high in fat intake (39.75% of energy from fat) had a lower inflammation level and slower aging progress. The Prospective Urban Rural Epidemiology (PURE) study including individuals aged 35–70 years in 18 countries, in accordance with the present study, has shown that total fat intake(17.7%-30.5% of energy from fat) are associated with lower mortalities in countries with high carbohydrate intake(52% individuals with more than 60% of energy from carbohydrates³³) and disproportionately low intake(< 10%) of saturated fatty acids could be harmful. Thus, limiting carbohydrate intake and increasing fat intake might postpone aging, for subjects with high carbohydrate intakes.

By subgroups, statistically significant relationship between high CEDII and short LTL were stronger in participants with obesity, raised BP, or elevated TG. Previous studies have revealed that overweight/obese subjects consuming more sodium were more susceptible to hypertension than those with normal BMI. Moreover, subjects with hypertension intaking higher sodium levels are more likely to be overweight/obese than those without hypertension³⁴. It implies that high CEDII may have more prominent effects on aging among individuals with metabolic disorders than those without them.

There were a few strengths worth mentioning in the current study. We identified traditional dietary patterns of North China related to inflammation using the RRR model based on the food groups. The inflammatory potential of whole diets was related to aging, which provided clues for optimizing interventions to slow aging. Additionally, we found prior indexes derived from DBI and EDIP derived from western dietary patterns did not present significant associations with inflammation and aging, which corroborated that CEDII was a more appropriate approach to assesses diet inflammatory potential quality and aging in Chinese residents. Finally, subgroup analyses according to different metabolic statuses were conducted, identifying the crucial role of diet management in individuals with abnormal metabolic status.

The present study had several limitations that merit consideration. First, our study could not make definitive causal inferences due to the cross-sectional study design. Second, as the participants were recruited from a rural community in Northern China, results may not be generalizable to other populations. Moreover, the prospect of false negatives might raise on account of the small sample size in subgroup analyses. Therefore, the results need to be validated in larger cohorts from different regions. Third, 24h food recalls may not be fully valid to represent for long-term food intake due to the unique Chinese dietary habits. Fourth, despite accounting for many crucial confounders, we could not completely rule out the other potential confounders, such as cooking styles, physical activity and smoking status. Thus, more covariates should be included in future studies.

In conclusion, our study has identified dietary patterns that assess diet quality from residents of North China based on inflammatory potential. Participants with high CEDII are more likely to have lower chronic inflammation and slower ageing progress. The relationship between CEDII and ageing acceleration is more closely in participants with abnormal metabolic status. These results may indicate that CEDII is a promising dietary measure to promote healthy aging in China. Further studies in large-scale validation from prospective cohorts are required to verify these findings.

Author Contributions: Drs. Lu Lyu and Huabing Zhang had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis

Concept and design: Huabing Zhang, Yuxiu Li

Acquisition, analysis, or interpretation of data: Lu Lyu, Wei Li, Huabing Zhang,

Drafting of the manuscript: Lu Lyu, Huabing Zhang, Wei Li, Yuxiu Li

Critical revision of the manuscript for important intellectual content: Lu Lyu, Jie Yu, Yiwen Liu, Shuli He, Mengya Qi, Na Yang, Liyun He, Jialu Wang,Fan Ping, Lingling Xu, Huabing Zhang, Wei Li, Yuxiu Li

Statistical analysis: Lu Lyu, Huabing Zhang, Wei Li, Yuxiu Li

Obtained funding: Huabing Zhang, Wei Li, Yuxiu Li

Supervision: Lyu Lu, Huabing Zhang, Wei Li, Yuxiu Li, Wei Li, Lingling Xu, Fan Ping

Conflict of Interest Disclosures: None of the authors had a conflict of interest.

Funding/Support: This study was supported by CAMS Innovation Fund for Medical Sciences(CIFMS)(2021-1-I2M-002), National High Level Hospital Clinical Research Funding(2022-PUMCH-B-015), Beijing Municipal Natural Science Foundation (No.M22014), the National Natural Science Foundation of China (No. 91846106).

Role of the Funder/Sponsor: The funder provide support for the ethical review fee and was involved in conduct of the study. The funder was not involved in collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See the Supplement.

Additional Contributions: We thank all the participants, physicians, investigators who participated in this study.

DALYs GBD, Collaborators H. Global, regional, and national disability-adjusted life-years (DALYs) for 359 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392(10159):1859–922.
Kuilman T, Michaloglou C, Vredeveld LC, et al. Oncogene-induced senescence relayed by an interleukin-dependent inflammatory network. Cell. 2008;133(6):1019–31.
Van Puyvelde K, Mets T, Njemini R, Beyer I, Bautmans I. Effect of advanced glycation end product intake on inflammation and aging: a systematic review. Nutr Rev. 2014;72(10):638–50.
Bujtor M, Turner AI, Torres SJ, Esteban-Gonzalo L, Pariante CM, Borsini A. Associations of Dietary Intake on Biological Markers of Inflammation in Children and Adolescents: A Systematic Review. Nutrients 2021;13(2).
Jacques PF, Tucker KL. Are dietary patterns useful for understanding the role of diet in chronic disease? Am J Clin Nutr. 2001;73(1):1–2.
Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol. 2002;13(1):3–9.
Zang J, Yu H, Zhu Z et al. Does the Dietary Pattern of Shanghai Residents Change across Seasons and Area of Residence: Assessing Dietary Quality Using the Chinese Diet Balance Index (DBI). Nutrients 2017;9(3).
Li H, Li D, Wang X, et al. The Role of Dietary Patterns and Dietary Quality on Body Composition of Adolescents in Chinese College. Nutrients. 2022;14:21.
Tabung FK, Smith-Warner SA, Chavarro JE, et al. Development and Validation of an Empirical Dietary Inflammatory Index. J Nutr. 2016;146(8):1560–70.
Shakeri Z, Mirmiran P, Khalili-Moghadam S, Hosseini-Esfahani F, Ataie-Jafari A, Azizi F. Empirical dietary inflammatory pattern and risk of metabolic syndrome and its components: Tehran Lipid and Glucose Study. Diabetol Metab Syndr. 2019;11:16.
Li J, Lee DH, Hu J, et al. Dietary Inflammatory Potential and Risk of Cardiovascular Disease Among Men and Women in the U.S. J Am Coll Cardiol. 2020;76(19):2181–93.
Zhou MC, Zhu L, Cui X, et al. Reduced peripheral blood mtDNA content is associated with impaired glucose-stimulated islet beta cell function in a Chinese population with different degrees of glucose tolerance. Diabetes Metab Res Rev. 2016;32(7):768–74.
Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16(1):31–41.
Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care. 1999;22(9):1462–70.
Abdul-Ghani MA, Matsuda M, Balas B, DeFronzo RA. Muscle and liver insulin resistance indexes derived from the oral glucose tolerance test. Diabetes Care. 2007;30(1):89–94.
Utzschneider KM, Prigeon RL, Faulenbach MV, et al. Oral disposition index predicts the development of future diabetes above and beyond fasting and 2-h glucose levels. Diabetes Care. 2009;32(2):335–41.
Santos JL, Yevenes I, Cataldo LR, et al. Development and assessment of the disposition index based on the oral glucose tolerance test in subjects with different glycaemic status. J Physiol Biochem. 2016;72(2):121–31.
Yu J, Liu H, He S, et al. Dietary Magnesium Intake and Leukocyte Telomere Attrition in Adults: The Regulatory Role of Serum Tumor Necrosis Factor alpha. Mediators Inflamm. 2020;2020:7610436.
Romanos-Nanclares A, Tabung FK, Sinnott JA, et al. Inflammatory and insulinemic dietary patterns and risk of endometrial cancer among US women. J Natl Cancer Inst. 2023;115(3):311–21.
Jin Q, Shi N, Aroke D, et al. Insulinemic and Inflammatory Dietary Patterns Show Enhanced Predictive Potential for Type 2 Diabetes Risk in Postmenopausal Women. Diabetes Care. 2021;44(3):707–14.
Wang SS, Lay S, Yu HN, Shen SR. Dietary Guidelines for Chinese Residents (2016): comments and comparisons. J Zhejiang Univ Sci B. 2016;17(9):649–56.
Wang Y, Su X, Chen Y et al. Unfavorable Dietary Quality Contributes to Elevated Risk of Ischemic Stroke among Residents in Southwest China: Based on the Chinese Diet Balance Index 2016 (DBI-16). Nutrients 2022;14(3).
Hoffmann K, Schulze MB, Schienkiewitz A, Nothlings U, Boeing H. Application of a new statistical method to derive dietary patterns in nutritional epidemiology. Am J Epidemiol. 2004;159(10):935–44.
Zhou BF, Cooperative Meta-Analysis Group of the Working Group on Obesity in C. Predictive values of body mass index and waist circumference for risk factors of certain related diseases in Chinese adults–study on optimal cut-off points of body mass index and waist circumference in Chinese adults. Biomed Environ Sci. 2002;15(1):83–96.
Chinese Elderly Type 2, Diabetes P, Treatment of Clinical Guidelines, Writing G, Geriatric E, et al. [Clinical guidelines for prevention and treatment of type 2 diabetes mellitus in the elderly in China (2022 edition)]. Zhonghua nei ke za zhi. 2022;61(1):12–50.
Roberts SB, Silver RE, Das SK, et al. Healthy Aging-Nutrition Matters: Start Early and Screen Often. Adv Nutr. 2021;12(4):1438–48.
Wahl D, Solon-Biet SM, Wang QP, et al. Comparing the Effects of Low-Protein and High-Carbohydrate Diets and Caloric Restriction on Brain Aging in Mice. Cell Rep. 2018;25(8):2234–2243e2236.
Capurso C. Whole-Grain Intake in the Mediterranean Diet and a Low Protein to Carbohydrates Ratio Can Help to Reduce Mortality from Cardiovascular Disease, Slow Down the Progression of Aging, and to Improve Lifespan: A Review. Nutrients. 2021;13:8.
Gkotzamanis V, Panagiotakos D, Yannakoulia M, et al. Trajectories of healthy aging and their association with the Mediterranean diet: The HELIAD Study. Maturitas. 2022;159:33–9.
Kim Y, Keogh JB, Deo P, Clifton PM. Differential Effects of Dietary Patterns on Advanced Glycation end Products: A Randomized Crossover Study. Nutrients 2020;12(6).
Seidelmann SB, Claggett B, Cheng S, et al. Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis. Lancet Public Health. 2018;3(9):e419–28.
Zhang Y, Chen L, Hu M, et al. Dietary type 2 resistant starch improves systemic inflammation and intestinal permeability by modulating microbiota and metabolites in aged mice on high-fat diet. Aging. 2020;12(10):9173–87.
Dehghan M, Mente A, Zhang X, et al. Associations of fats and carbohydrate intake with cardiovascular disease and mortality in 18 countries from five continents (PURE): a prospective cohort study. Lancet. 2017;390(10107):2050–62.
Stern N, Buch A, Goldsmith R, et al. The role of caloric intake in the association of high salt intake with high blood pressure. Sci Rep. 2021;11(1):15803.

Table 1 The demographic characteristics according to tertiles of CEDII

	Low CEDII (n=126)	Moderate CEDII (n=130)	High CEDII (n=132)	Total (n=388)	p-value
Gender，N（%）					0.591
Female	83(65.87%)	80(61.54%)	90(68.18%)	253(65.21%)
Male	43(34.13%)	50(38.46%)	42(31.82%)	135(34.79%)
Age（year）	53.05±10.59	51.08±11.29	51.69±10.5	51.93±10.8	0.330
BMI（kg/m2)	26.25±3.92	25.98±3.76	25.73±3.61	25.98±3.76	0.547
WC（cm）	88±10.38	87.01±9.39	86.29±9.2	87.08±9.66	0.373
HC（cm）	92.85±10.25	90.85±11.95	91.44±9.46	91.70±10.61	0.309
SBP（mmHg）	126(114,140)	128(118,140)	126(114,136)	126(114,136.75)	0.417
DBP（mmHg）	76(70,84.5)	76(70,82)	76(70,82.5)	76(70,82)	0.767
HbA1c（%）	5.5(5.2,5.9)^c	5.5(5.3,5.9)^c	5.7(5.4,6.2)^a,b	5.6(5.3,6)	0.005^**
FPG(mmol/L)	6(5.43,6.6)	5.86(5.45,6.52)	5.91(5.5,6.71)	5.91(5.45,6.58)	0.571
2hPG(mmol/L)	7.63(5.71,9.83)	7.02(5.43,8.91)	7.35(6.2,10.66)	7.35(5.8,9.96)	0.222
lg(HOMA-IR)	0.42±0.27	0.42±0.29	0.43±0.26	0.43±0.27	0.906
sqrt(ISIM)	7.84±3.01	7.73±2.83	7.58±2.69	7.71±2.84	0.760
HIRI	63.83(40.31,111.70)	67.83(50.82,117.50)	66.14(46.41,113.79)	65.69(45.77,112.84)	0.659
sqrt(HOMA-β)	1.98±0.65	1.98±0.7	1.96±0.74	1.97±0.7	0.959
DI	6.12(3.25,9.35)	6.79(3.76,10.07)	5.99(3.29,9.16)	6.25(3.33,9.45)	0.177
IGI30	11.02(4.56,20.02)	12.53(5.15,30.45)	10.31(4.13,23.71)	11.26(4.39,25.10)	0.687
lgTG(mmol/L)	0.19±0.28	0.17±0.26	0.16±0.25	0.17±0.26	0.547
TC(mmol/L)	5.43±1.09	5.51±1.06	5.52±1.05	5.49±1.07	0.780
sqrtHDL-c(mmol/L)	1.12±0.12	1.14±0.11	1.13±0.11	1.13±0.11	0.547
LDL-c(mmol/L)	2.77±0.73	2.87±0.72	2.89±0.7	2.85±0.72	0.376
eGFR(ml/min/1.73m2)	96.12(83.25,104.84)	94.17(82.24,105.36)	97.76(83.64,104.96)	96.64(83.22,105.08)	0.967
ACR(mg/mmolCr)	1.27(0.8,2.14)	1.32(0.67,2.76)	1.01(0.59,2.14)	1.14(0.67,2.2975)	0.095
UA(ummol/L)	283.77±85.54	295.39±82.99	290.83±76.89	290.02±81.75	0.526

BMI, body mass index; WC, waist circumference; HC, hip circumference SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucose; 2hPG, 2h postprandial plasma glucose; HOMA-IR, homeostasis model assessment for insulin resistance; ISIM： Matsuda index；HIRI, hepatic insulin resistance index; HOMA-β, homeostasis model assessment for β cell function; DI, disposition index of insulin; IGI30, insulinogenic index 30; TC, cholesterol; TG, triglyceride; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; ACR, urine albumin-creatinine ratio; UA, uric acid;

^*: p＜0.05, ^** p<0.01; ^a, a significant difference with the low CEDII; ^b, a significant difference with the moderate CEDII; ^c, a significant difference with the high CEDII;

Differences among the three clusters were performed by either ANOVA, Kruskal-Wallis tests or liner χ² test

Table 2 Indicators of inflammation, oxidative stress and aging across tertiles of different dietary indexes

	Groups			p值
	Low	Moderate	High	p值
CEDII
sqrtTNFa(fmol/ml)	4.3±1.07^b,c	4.74±1.05^a,c	5.16±1.02^a,b	<0.001^**
lgIL-6(pg/ml)	0.47±0.32^c	0.48±0.3^c	0.57±0.26^a,b	0.006^**
lg8-oxo-dG(pg/ml)	1.5±0.34	1.53±0.34	1.53±0.37	0.761
SOD activities(U/ml)	57.77(50.56,67.12)	59.03(50.92,68.33)	62.72(53.07,70.31)	0.117
GR(U/L)	6.43(4.82,9.65)	6.43(3.22,9.65)	8.04(4.82,9.65)	0.681
LTL	28.87±0.75^c	28.73±0.78	28.59±0.94^a	0.023^*
lgmtDNAcn	2.02±0.21^c	2±0.16^c	1.95±0.17^a,b	0.003^**
EDIP
sqrtTNFa(fmol/ml)	4.87±1.07	4.66±1.05	4.68±1.17	0.257
lgIL-6(pg/ml)	0.53±0.3	0.47±0.3	0.53±0.29	0.161
lg8-oxo-dG (pg/ml)	1.52±0.35	1.52±0.37	1.52±0.34	0.993
SOD activities (U/ml)	61.7(52.48,71.52)^b	57.68(48,63.48)^a,c	61.6(51.62,70.64)^b	0.021^*
GR(U/L)	6.43(4.82,9.65)	6.43(3.22,9.65)	8.04(4.82,9.65)	0.462
LTL	28.73±0.82	28.78±0.85	28.67±0.84	0.547
lgmtDNAcn	1.98±0.16	1.98±0.22	2±0.18	0.613
DQD-DBI
sqrtTNFa(fmol/ml)	4.77±1.02	4.8±1.05	4.63±1.24	0.428
lgIL-6(pg/ml)	0.48±0.29	0.54±0.3	0.5±0.32	0.262
lg8-oxo-dG(pg/ml)	1.49±0.37	1.51±0.35	1.57±0.32	0.17
SOD activities (U/ml)	60.53(51.07,70.73)	59.11(51,67.57)	60.83(51.57,70.16)	0.468
GR(U/L)	6.43(4.82,9.65)	6.43(3.22,9.65)	8.04(4.82,9.65)	0.822
LTL	28.75±0.88	28.68±0.77	28.76±0.85	0.711
lgmtDNAcn	1.99±0.16	2±0.18	1.98±0.22	0.663

TNFα, tumor necrosis factor-α; IL-6, interleukin-6; 8-oxo-dG, 8-oxo-2′-deoxyguanosine; SOD activity, superoxide dismutase activity; GR, glutathione reductase; LTL, leukocyte telomere length; mtDNAcn, copy number of mitochondrial DNA

^*: p＜0.05, ^** p<0.01; ^a, a significant difference with the low dietary index; ^b, a significant difference with the moderate dietary index; ^c, a significant difference with the high dietary index;

Table 3 Multiple regression analysis evaluating aging biomarkers as dependent variables

	LTL				lgmtDNAcn
	Unstandardized-β	Standardized β	95%CI	p-value	Unstandardized-β	Standardized β	95%CI	p-value
CEDII	-0.332	-0.12	(-0.62,-0.044)	0.024^*	-0.073	-0.12	(-0.137,-0.009)	0.026^*
gender	0.218	0.125	(-0.007,0.443)	0.057	-0.024	-0.063	(-0.075,0.026)	0.341
age	-0.001	-0.006	(-0.01,0.009)	0.92	-0.001	-0.052	(-0.003,0.001)	0.419
BMI	0.019	0.088	(-0.012,0.051)	0.225	0.003	0.066	(-0.004,0.01)	0.367
WC	0.018	0.207	(-0.001,0.037)	0.062	0.002	0.081	(-0.003,0.006)	0.466
HC	-0.010	-0.124	(-0.025,0.006)	0.219	-0.003	-0.155	(-0.006,0.001)	0.126
SBP	2.180E-04	0.005	(-0.005,0.005)	0.933	-2.25E-04	-0.023	(-0.001,0.001)	0.695
DBP	-0.003	-0.037	(-0.013,0.006)	0.528	-4.46E-04	-0.024	(-0.003,0.002)	0.680
lg8-oxo-dG	-0.031	-0.013	(-0.274,0.212)	0.800	-0.017	-0.032	(-0.071,0.038)	0.548
SOD activities	-0.005	-0.101	(-0.01,4.15E-04)	0.072	-0.001	-0.143	(-0.003,3.28E-04)	0.012^*
GR	-0.007	-0.027	(-0.033,0.02)	0.611	-0.001	-0.013	(-0.007,0.005)	0.804
lgHOMAIR	0.181	0.058	(-0.371,0.733)	0.519	-0.066	-0.096	(-0.189,0.058)	0.298
sqrtHMOAβ	-0.136	-0.103	(-0.362,0.09)	0.237	0.026	0.089	(-0.025,0.076)	0.315
HbA1c	-0.040	-0.046	(-0.169,0.089)	0.546	0.002	0.009	(-0.027,0.031)	0.902
Energy intake	-6.147E-05	-0.048	(-2.00E-04,7.80E-05)	0.387	1.03E-05	0.037	(-2.10E-05,4.10E-05)	0.517
eGFREPI	-0.001	-0.017	(-0.006,0.005)	0.766	0.001	0.133	(0,0.003)	0.023^*
lgTG	-0.064	-0.019	(-0.508,0.379)	0.775	-0.007	-0.009	(-0.106,0.093)	0.897
sqrtHDL-c	-0.023	-0.003	(-0.939,0.893)	0.961	-0.141	-0.086	(-0.347,0.064)	0.177
LDL-c	0.037	0.032	(-0.097,0.171)	0.591	0.01	0.038	(-0.02,0.04)	0.525
UA	-0.003	-0.251	(-0.004,-0.001)	<0.001^**	-2.77E-04	-0.122	(-0.001,1.10E-05)	0.059
ALT	0.004	0.076	(-0.006,0.013)	0.449	-0.001	-0.141	(-0.004,0.001)	0.164
AST	-0.010	-0.095	(-0.031,0.01)	0.323	0.003	0.119	(-0.002,0.008)	0.222

BMI, body mass index; WC, waist circumference; HC, hip circumference SBP, systolic blood pressure; DBP, diastolic blood pressure; 8-oxo-dG, 8-oxo-2′-deoxyguanosine; SOD activity, superoxide dismutase activity; GR, glutathione reductase; FPG, fasting plasma glucose; 2hPG, 2h postprandial plasma glucose; HOMA-IR, homeostasis model assessment for insulin resistance; HOMA-β, homeostasis model assessment for β cell function; TG, triglyceride; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; UA, uric acid;

^*: p＜0.05, ^** p<0.01

No competing interests reported.

Supplementarymaterials.docx

Chinese Empirical Dietary Inflammatory Index Is Associated with High Risk of Systematic Inflammation and Aging in Rural China

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Study participants

Anthropometric measurements

Laboratory analysis

Assessment of Insulin Sensitivity and β-Cell Function

Dietary assessment

Assessment of Empirical Dietary Inflammatory Pattern and Diet Balance Index-16

Development and Assessment of Chinese Empirical Dietary Inflammatory Index

Statistical analysis

Results

Discussion

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1