Introduction

Obesity is a multifactorial disease that is caused due to a combination of biological, social, genetic, behavioral, and environmental determinants [1]. Obesity, a feature of metabolic syndrome, is associated with chronic inflammation in obese subjects [2]. Obesity is recognized as a major disease that leads to the onset of many other chronic diseases, including cardiovascular disease (CVD), hypertension (HTN), and type 2 diabetes (T2D) [3]. The World Health Organization (WHO) reported that about 2 billion and 600 million adults worldwide were overweight and obese in 2014, respectively [4]. According to the World Obesity Atlas, around one billion adults were considered obese in 2020, and this number is expected to rise to approximately 1500 million by 2030 [5]. The prevalence of obesity and overweight was 22.7% and 59.3% in Iranian adults in 2016, respectively [6]. Furthermore, overweight and obesity prevalence was higher in women than men [7].

Increased BMI and obesity are strongly associated with changes in the physiological function of adipose tissue, leading to enhanced secretion of adipocytokines and inflammatory factors including leptin, interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), MCP-1, resistin [8], and hs-CRP [9]. As a result, obesity, particularly visceral obesity, is now considered a low-grade inflammatory disease [10,11,12]. Ghrelin is a hormone that exerts strong inhibitory effects on proinflammatory cytokines, such as IL-1β, IL-6, and TNF-α, following lipopolysaccharide (LPS)-induced inflammation. Consequently, low serum ghrelin levels have been observed in conditions with a positive energy balance, including obesity [13].

Polyphenols are a complex group of plant secondary metabolites and one of the most notable natural antioxidants widely distributed in plant-based foods and beverages, such as fruits, vegetables, grains, and tea [14]. The four main polyphenol classes are phenolic acids, flavonoids, stilbenes, and lignans, and epidemiological studies have suggested inverse associations between polyphenol intake and the risk of inflammatory and oxidative chronic diseases, including obesity. However, the existing evidence shows that the health effects of polyphenols are conflicting, and a paucity of studies have examined such effects [15].

Given the rising prevalence of obesity, and the lack of consistent evidence on the associations between inflammatory markers and dietary polyphenol intake, especially in obese and overweight/obese Iranian women, the present study aims to assess this association.

Method and materials

Study participants

The participants were overweight/obese Iranian women referred to Tehran health centers (Fig. 1). A random multistage sampling method was used to recruit the participants. The inclusion criteria were as follows: being female, being between the ages of 18 and 48, and having BMI ranging from 25 to 40 kg/m2. Participants with a history of HTN, CVD, diabetes, or any acute or chronic illnesses including thyroid disease, cancer, liver disease, or renal disease, smoking, taking medications for controlling blood sugar, blood pressure, blood lipids, weight, drinking alcohol, pregnancy, or lactation, following any specific diet, having weight fluctuations greater than 5% over the last 6 months, and having energy intake less than 800 or more than 4200 kcal per day were excluded [16].

Fig. 1
figure 1

Flowchart of study participants

Full size image

The study protocol has been approved by the Tehran University of Medical Sciences ethics committee (IR.TUMS.VCR.REC.1399.636). A consent form has been signed by each participant.

Anthropometric indices and body composition

With an accuracy of 0.1 kg, weight was measured using a digital scale (Seca, Hamburg, Germany), while participants wore light clothes and were without shoes. Using a body composition analyzer, the following measurements were made: fat-free mass (FFM), visceral fat area (VFA), body fat percentage (BFP), body fat mass (BFM), and BMI (InBody770 scanner; InBody, Seoul, Korea). A Seca 206 stadiometer (Hamburg, Germany) was used to measure the participants' height with an accuracy close to 0.2 cm. The WC and HC were measured with accuracy near 0.2 cm.

Evaluation of dietary intake and polyphenol consumption and its constituent parts

Participants' dietary consumption was evaluated using a semiquantitative Food Frequency Questionnaire (FFQ) with 147 food items [17]. The validity and reliability of the FFQ have been confirmed previously [18]. Participants were asked to provide information on portion size, regular cooking techniques, and types of oil. To convert portion amounts to grams, standard measures were used. Total polyphenol consumption was calculated using the Phenol-Explorer database (www.phenolexplorer.eu/) [16, 19]. To measure total polyphenol content and its constituent parts independently, the Folin–Ciocalteu test or the sum of four major classes (containing flavonoids, phenolic acids, stilbenes, lignans, and other polyphenols) was applied. To estimate nutrients and energy intake, Nutritionist IV software (version 7.0; N-Squared Computing, Salem, OR) was utilized [20].

Biochemical parameters

After a 10–12-h fast, blood samples were collected, and serum was kept at − 80 °C. According to the manufacturer’s instructions, all tests were performed. Fasting plasma glucose was assessed using the glucose oxidase technique, and the affront level was determined using an enzyme-linked immunosorbent assay (ELISA) device (Human affront ELISA unit, DRG Pharmaceuticals, GmbH, Germany). Related packets were used to assess TG, Chole, LDL-c, and HDL-c (Pars Azemun, Iran). The Universal League of Clinical Chemistry and Research facility Medication standardization was used to test SGPT and SGOT. Gal-3 (R&D Systems, Minneapolis, MN), MCP-1 (Zell Bio GmbH, ULM, Germany), TGF-β and IL_1β (HUMAN TGF-BETA 1 and IL_1β Quantikine ELISA kit R&D System-USA), PA-I (Human PAI-1*96 T ELISA kit Crystal Company), serum leptin and ghrelin concentrations, and hs-CRP were measured using the ELISA method (Mediagnost, Reutlingen, Germany). For all tests, the variability between and within analyses was less than 12% and 10%, respectively [16].

Sociodemographic characteristics and physical activity

The sociodemographic characteristics including education (Illiterate, under diploma, diploma, bachelor and higher), occupation (Unemployed, employed), marital status (single, married), economic position (low, middle, high status), and supplement intake (yes, no), were collected using a questionnaire. To measure physical activity standards, the validated International Physical Activity Questionnaire (IPAQ) was translated to minutes per week using metabolic equivalents (MET-min/week) [21].

Statistical analyses

The sample size was computed according to the following formula: n = [(Z1-α + Z1-β) × (√1 − r2)/r) 2] which r = 0.27, β = 0.95, and α = 0.05; thus, 350 women were considered for the study population. The normality of quantitative dependent variables was checked using the Kolmogorov–Smirnov test (P value > 0.05) and assessment of the histogram curve. Furthermore, according to the central limit theorem, all dependent variables were considered to have a normal distribution [22, 23]. Categorical variables were reported as numbers and percentages, and quantitative variables were reported as means and standard deviation (SD). To compare the frequency of categorical variables and the mean difference of quantitative variables across polyphenol intake quartiles, chi-square (χ2) tests and one-way analysis of variance (ANOVA) were performed, respectively. Normality assumption as mentioned has been performed and the same variance was assessed using Levene’s test if there was no Welch test applied. Analysis of covariance (ANCOVA) was used to examine the mean difference of continuous variables over polyphenol intake quartiles, and the analysis was adjusted for potential confounders including age, BMI, physical activity, and energy intake. The covariates were identified based on the previous studies [16, 24, 25] and examining the associations between polyphenol intake and the variables. The variables that had a significant association with polyphenols were considered confounding variables (Table 1). BMI was considered a collinear variable for anthropometric and body composition measurements. All linear regression test assumptions were evaluated, including normality, normality of residual error, linearity, homoscedasticity, and collinearity. Linear regression analysis was used to examine the association between inflammatory and polyphenol intakes and their components. Bonferroni post hoc was applied to detect the significant mean difference. The adjusted model 1 was controlled for age, BMI, physical activity, total energy intake, supplement intake, economic status, and education. SPSS v.26 software (SPSS Inc., IL, USA) was used for statistical analysis. P value < 0.05 was considered significant, while 0.05, 0.06, and 0.07 were considered marginally significant.

Table 1 General characteristics of study participants over quartiles of polyphenols intake (n = 391)
Full size table

Result

Study population

A total of 391 participants were included in the analysis. The mean difference of age (P = 0.001) was statistically significant over polyphenols intake (ml/day) quartiles, and the mean difference of age (P = 0.063) according to polyphenols intake (mg/day) quartiles was marginally significant. The majority of participants were employed in quartile 1 (ml/day) (26.2%) and in quartile 2 (mg/day) (25.5%) of polyphenols intake. Most participants in quartile 3 (ml/day) (29%) and quartile 1 (mg/day) (29%) of polyphenol intake had high economic status (Table 1).

General characteristics of study participants over quartiles (mg/day)/(ml/day) of polyphenols intake

Table 1 shows the general characteristics of the study participants. The mean difference of PA after adjustment for age, BMI, and energy intake was statistically significant across polyphenols intake (ml/day) quartiles (P = 0.022). Also, the mean difference of PA according to polyphenols intake (mg/day) quartiles was significant and after adjustment (P = 0.004) remained significant (P < 0.05). According to Bonferroni's post hoc testing, this mean difference was higher in Q4. After adjustment for confounders, the mean difference of WHR (P = 0.054), and BF (%) (P = 0.060) was significant and marginally significant over polyphenols intake quartiles (mg/day), respectively, with a higher mean difference in Q1. The mean difference of LDL-c (P = 0.047) was significant over polyphenol intake quartiles (ml/day), after adjustment for confounders. Post hoc analysis showed a higher mean difference in Q2. The frequency of supplement consumption had a significant difference over polyphenol intake quartiles (mg/day) (P = 0.042), while after adjustment for confounders, the association was marginally significant (P = 0.051).

Dietary intakes across the polyphenol’s intake quartiles (mg/day)/ (ml/day)

The dietary intakes of participants over the polyphenol intake quartiles are presented in Table 2. The mean differences of whole grains, fruits, vegetables, legumes (P = 0.001), and sugar and sugar-sweetened beverages (SSB) consumption across the polyphenol intake quartiles (mg/day) were statistically significant after adjustment for confounders (P = 0.004). The mean difference of energy over the polyphenol intake quartiles (mg/day) was statistically significant (P = 0.046). The mean difference of linolenic acid (P = 0.042) and vitamin A consumption (P = 0.001) across polyphenols intake quartiles (mg/day) was statistically significant. Also, the mean difference of carbohydrates (P = 0.003), percentage of energy from protein (P = 0.003), percentage of energy from fat (P = 0.020), total fat (P = 0.018), saturated fatty acid (SFA) (P = 0.001), mono-unsaturated fatty acid (MUFA) (P = 0.020), vitamin E (P = 0.044), vitamin B5 (P = 0.042), magnesium (P = 0.011), selenium (P = 0.014), total fiber, β-carotene, vitamin C, folate, biotin, vitamin B6, and copper (P = 0.001) over polyphenols intake quartiles (mg/day) were statistically significant after adjustment for confounders. After adjustment for confounders, the mean difference of tea and coffee, caffeine, manganese (P = 0.001), vitamin E (P = 0.021), and vitamin B6 (P = 0.018) across polyphenol’s intake quartiles (ml/day) was statistically significant.

Table 2 Dietary intakes over quartiles of the polyphenol intake (n = 391)
Full size table

Association between inflammatory markers and polyphenol intakes (mg/day)/ (ml/day) over polyphenol intake quartiles

The association between inflammatory markers and polyphenol intakes (mg/day, ml/day) across quartiles of polyphenol intake in crude and adjusted models is presented in Table 3. In model 1, after controlling for potential confounders including age, BMI, energy intake, PA, educational status, income status, supplement consumption, and marital status, there was a marginally significant association between hs-CRP and polyphenol intakes (mg/day) in Q3 (P = 0.069). Also, in the crude model, there was a marginally significant association between PAI-1 and polyphenol intakes (mg/day) (P-trend = 0.068). After controlling for confounders, there was a marginally significant association between MCP-1 and polyphenol intakes (mg/day) in Q3 (P = 0.070).

Table 3 Association between inflammatory markers and polyphenol intakes (mg/day) and (ml/day) over quartiles of polyphenol intake (n = 391)
Full size table

The association between polyphenols intake components and inflammatory markers

The association between polyphenol intake components and inflammatory markers in the crude and adjusted models is presented in Table 4. Regarding flavonoids in model 1, there was a marginally negative significant association between flavonoids (mg/day) intake and hs-CRP (P = 0.001) and MCP-1 (P = 0.024), and also between lignans (mg/day) intake and MCP-1 (P = 0.017) and Gal-3 (P = 0.032). There was a negative significant relationship between IL-1β and other polyphenols (mg/day) intake (P = 0.014) and a positive significant relationship between TGF-β and phenolic acid (ml/day) intake (P = 0.014). Furthermore, there was a marginally negative significant association between IL-1β and flavonoids intake (mg/day) (P = 0.057), and between serum leptin and lignans (mg/day) intake (P = 0.061) in model 1. Moreover, a marginally negative significant association between hs-CRP and phenolic acid intake (mg/day) (P = 0.067) and between hs-CRP and stilbenes (mg/day) intake (P = 0.069) was found in model 1.

Table 4 Association between polyphenols intake and inflammatory markers (n = 391)
Full size table

There was a significant association between total flavonoids intake (mg/l) and hs-CRP (mg/l) in the crude model (P = 0.001) and after adjustment (P = 0.001), also between total flavonoids intake (mg/l) and MCP-1 (mg/l) in the crude model (P = 0.042) and after controlling covariates and confounding variables (P = 0.024) (Fig. 2).

Fig. 2
figure 2figure 2

The association between polyphenol intake and its components with inflammatory factors (A-N). A: The association between total polyphenol intake (mg/d) and hs-CRP (mg/l), P = 0.046, adjusted P = 0.069. B: The association between total polyphenol intake (mg/d) and MCP-1 (mg/l), P = 0.061, adjusted P = 0.070. C: The association between total polyphenol intake (mg/d) and Gal-3 (mg/l), P = 0.518, adjusted P = 0.049. D: The association between total flavonoids intake (mg/d) and hs-CRP (mg/d), P = 0.001, adjusted P = 0.001. E: The association between total flavonoids intake (mg/d) and IL-1 β (mg/l), P = 0.664, adjusted P = 0.057. F: The association between total flavonoids intake (mg/d) and MCP-1 (mg/l), P = 0.042, adjusted P = 0.024. G: The association between other polyphenols intake (mg/d) and IL-1 β (mg/l), P = 0.610, adjusted P = 0.725. H: The association between other polyphenols (ml/d) and TGF-β (mg/l), P = 0.149, adjusted P = 0.008. I: The association between total phenolic acids polyphenols intake (mg/d) and hs-CRP (mg/l), P = 0.024, adjusted P = 0.067. J: The association between total phenolic acids (mg/l) and TGF-β (mg/l), P = 0.990, adjusted P = 0.578. K: The association between total lignans (mg/d) and leptin (ng/ml), P = 0.494, adjusted P = 0.061. L: The association between total lignans (mg/d) and MCP-1 (mg/l), P = 0.042, adjusted P = 0.017. M: The association between total lignans (mg/d) and Gal-3 (mg/l), P = 0.064, adjusted P = 0.032. N: The association between total stilbenes (mg/d) and Gal-3 (mg/l), P = 0.134, adjusted P = 0.187. Gal-3 Galectin-3, hs-CRP high-sensitivity C-reactive protein, IL-1β interleukin-1 beta, MCP-1 monocyte chemoattractant protein-1, PAI-1 plasminogen activator inhibitor-1, TGF-β Transforming growth factor beta

Full size image

A significant positive association between other polyphenols (mg/l) and TGF-β (mg/l) in the adjusted model (P = 0.008) was observed. A significant negative association between total lignans (mg/d) and MCP-1 (mg/l) in both the crude model (P = 0.042) and adjusted model was found (P = 0.017), while between total lignans (mg/d) and Gal-3 (mg/l) a marginally negative significant association in the crude model (P = 0.064), and a statistically negative significant association in the adjusted model (P = 0.032) was observed.

Discussion

The current findings showed a novel association between polyphenol intake and inflammatory markers in overweight/obese Iranian women. There was a significant negative association between flavonoids (mg/day) and hs-CRP, IL-1b, MCP-1, lignan (mg/day), and MCP-1, Gal-3, and serum leptin. Also, there was a significant negative association between phenolic acid (mg/day) (ml/day) and hs-CRP, stilbenes (mg/day), and hs-CRP. Furthermore, a significant positive association between phenolic acid (ml/day) and TGF- β was observed.

Given that the prevalence of obesity has increased in Iran from two million in 1980 to 11 million in 2015 [26], our findings are of importance regarding major public health and suggest that higher polyphenol intake might be an effective strategy in management of obesity and obesity-related diseases such as inflammation, especially in overweight/obese Iranian women. It should be mentioned that women typically have a higher adherence to healthy dietary patterns than men [27].

In line with our study, Hsieh et al. study in 2021 showed a negative association between flavonoid intake and CRP levels in Taiwanese [28]. By trapping the chain-initiating radicals at the membrane interface, flavonoids may reduce oxidative stress in the phospholipid bilayer. Inhibition of cytokine gene expression and production has also been demonstrated for flavonoids [28, 29]. By preventing nuclear factor kappa B (NF-κB) from being activated and by inhibiting the binding with genes, flavonoids are hypothesized to prevent the production of CRP [30,31,32]. The European Prospective Investigation into Cancer and Nutrition cohort (EPIC) in 2020 on the general population from 10 European countries demonstrated that higher plasma concentration of polyphenols is associated with lower odds of hs-CRP. Previous studies have reported that a diet with a higher intake of bioactive polyphenol compounds could be an effective strategy to prevent or modulate inflammation [33]. A systematic review and meta-analysis of 17 RCTs with 736 subjects reported that resveratrol (as a polyphenol) significantly reduced hs-CRP and TNF-α levels but had no significant effect on IL-6 levels [34].

In the present study, individuals in the higher quartiles of polyphenol intake consumed more whole grains, legumes, fruits and vegetables. The existing evidence showed that the consumption of polyphenol-rich foods such as fruits, vegetables, dark chocolate, tea, and coffee has modulated low-grade inflammation [35,36,37,38]. By interacting with proteins involved in gene expression and cell communication, polyphenols suppress the transcription factors that promote inflammation and protect from a number of chronic diseases that are triggered by inflammation [39]. Also, individuals in the higher quartiles of polyphenol intake had lower BF (%) and WHR. In accordance with our study, another cross-sectional study in 2022 reported that individuals in the higher tertiles of polyphenol intake had lower WHR and waist-to-height ratio (WHtR) [16]. Rosli et al. indicated that polyphenol intake was associated with lower neck circumference and obesity [40]. Cellular studies showed that dietary polyphenols play a role in adiposity reduction through suppressing adipocyte viability and preadipocyte proliferation, reducing adipocyte differentiation and triglyceride accumulation, stimulating lipolysis and fatty acid -oxidation, and decreasing inflammation [41]. In our previous study conducted by Aali et al. 2022 a significant negative association between stilbenes intake and BMI, lignan intake and BMI, polyphenol intake and WHR, and a marginally negative significant association between total polyphenol intake and WHtR was found [16]. According to our results, a marginally negative significant association between serum leptin and lignans (mg/day) was observed. Based on the previous studies, polyphenol intake may affect leptin [42,43,44]. One mechanism by which lignans can affect leptin is that they have capacity to inhibit protein tyrosine phosphatase 1B (PTP1B) [45, 46] that is a negative regulator of leptin [47]. In terms of ghrelin, no relationship with polyphenol intake was observed. However in other studies, polyphenols consumption had an effect on ghrel [48, 49].

Several studies have indicated that obesity causes inflammation [50,51,52]. Obesity-induced inflammation involves multiple organs, including adipose, liver, pancreas, heart, skeletal muscle, and the brain [50]. Dietary interventions using natural bioactive food compounds are promising treatments for obesity and metabolic diseases with limited side effects [53]. The current studies have reported that bioactive compounds play a role as anti-inflammatory agents and antioxidants through increasing thermogenesis and energy expenditure, reducing oxidative stress, which results in weight loss and the reduction in metabolic disorders [53, 54]. It has been shown that polyphenol compounds can inhibit the NF-κB signaling pathway [55]. NF-κB regulates cell proliferation, apoptosis, morphogenesis, and differentiation in addition to promoting the production of inflammatory cytokines, chemokines, and adhesion molecules [56]. Animal studies suggest that usual intake of polyphenols significantly affect obesity by decreasing fat mass, body weight, and triglycerides and increasing energy expenditure, fat utilization, and modulating glucose hemostasis [57, 58]. The studies that examined associations between polyphenol intake and inflammatory markers are limited and showed inconsistent results which could be due to the different study designs, different participants' characteristics (gender, age, ethnicity), and the chemical type of the dietary polyphenols used [59]. No specific mechanism has been found for the increasing effect of other polyphenols (ml/d) and even Phenolic acid.

This study has several strengths. To the best of our knowledge, this is the first study that examined associations between polyphenol intake and inflammatory markers in overweight/ obese Iranian women. Furthermore, a comprehensive and validated semiquantitative FFQ was used for analyzing dietary intakes. Anthropometric indices and body composition outcomes were assessed by the same person each time to improve the accuracy of the measurements [17].

There are limitations that need to be acknowledged. Given that this is a cross-sectional study, causality cannot be established. Despite using a validated FFQ, dietary intake measurement errors cannot be avoided. Given this study included only women, the results are not generalizable to the Iranian population. Furthermore, due to the small sample size, our purpose of reaching an association between polyphenol intake and inflammatory markers was limited. Finally, although all the analyses were adjusted for potential confounders, residual confounding may still exist.

Conclusion

In conclusion, there was a negative association between flavonoids (mg/day) and hs-CRP, IL-1b, MCP-1, lignan (mg/day) and MCP-1, Gal-3, leptin, and between phenolic acid (mg/day) and hs-CRP, phenolic acid (ml/day), stilbenes (mg/day) and hs-CRP. Also, a significant positive association between phenolic acid (ml/day) and other polyphenol intakes (mg/d), and polyphenol intake (ml/d) and TGF-B was found. The present study suggests that higher consumption of polyphenols could be effective in controlling obesity and obesity-related diseases and inflammation. Future studies with larger sample sizes including both genders are needed for comparison with our findings. Furthermore, experimental studies are needed to elucidate the exact molecular mechanism of the mentioned association.