Green Tea Consumption May Be Associated with Cardiovascular Disease Risk and Nonalcoholic Fatty Liver Disease in Type 2 Diabetics: A Cross-Sectional Study in Southeast China

Introduction

Cardiovascular disease (CVD) represents a major adverse outcome for type 2 diabetes mellitus (T2DM), with T2DM patients having a two- to four fold increased risk of CVD. ¹ Therefore, reducing CVD risk factors is crucial for prevention and/or delay of CVD among patients with T2DM, especially in some developing countries, such as China, where the burdens of both CVD and T2DM and are continuously increasing. ^{2 –4}

Mounting evidence supports the importance of healthy dietary patterns in the prevention of CVD risk. ⁵ Tea, which is made from the Camellia sinensis plant, has been one of the most popular beverages around the world for centuries, especially in many Asian countries, including China. Tea contains hundreds of chemical compounds, and some of these (e.g., polyphenols and catechins) are thought to be beneficial in terms of weight management and prevention of CVD, T2DM, and cancer. ⁶ However, cumulative epidemiologic evidence does not support a clear association between tea consumption and the risk of CVD. ^7,8 In a cohort of middle-aged Japanese, plasma tea catechins were even found to be positively associated with the risk of CVD. ⁹

The effects of tea consumption on cardiovascular risk factors have been evaluated in individuals with or at risk for T2DM. In a meta-analysis including seven randomized controlled trials (RCTs), ¹⁰ consumption of green tea or green tea extracts neither reduced glycemic traits nor improved insulin sensitivity in individuals at risk of T2DM. In another meta-analysis of RCT, ¹¹ similar neutral effects of green tea or green tea extracts on insulin resistance and glycemic control were observed in T2DM patients. However, in the third meta-analysis, ¹² tea or tea extracts supplementation effectively controlled fasting plasma glucose (FPG) and reduced waist circumference (WC) among T2DM patients, without effects on the other metabolic traits. Collectively, the potential role of tea consumption in modulating CVD risk factors among T2DM patients remains uncertain. Therefore, we used the data from a community-based cross-sectional study in Suzhou City to examine the associations between green tea consumption and numerous CVD risk factors, including obesity, hypertension, dyslipidemia, hyperuricemia, metabolic syndrome (MetS), and nonalcoholic fatty liver disease (NAFLD), among patients with T2DM.

Materials and Methods

Results

The mean age of the 1013 patients included in the analysis was 53.47 ± 11.84 years. A total of 400 participants (39.49%) reported drinking green tea regularly. Among the 400 participants who drank green tea frequently, 84.50% were men and 15.50% were women. Sex-specific characteristics of participants according to the frequencies of green tea consumption are given in Table 1. In men, green tea drinkers were younger, had a higher education level, and were more likely to smoke and drink alcohol. In women, green tea drinking was associated with younger age and higher levels of education and income.

Since green tea drinking was common in men, regular green tea drinkers were further categorized into once per day and more than or equal to twice per day. Adjusted ORs (95% CI) of CVD risk factors associated with tea consumption in men and women are reported in Table 2. In men, after full adjustment for potential confounders, green tea consumption was found to be associated with increased risks of general obesity and NAFLD. As compared with rarely or no green tea drinking, ORs of general obesity (P-trend = .021) and NAFLD (P-trend = .007) were increased gradually with the frequency of green tea consumption. No association was found between green tea consumption and other CVD risk factors in men.

Female participants were divided into two groups, including nontea drinkers and tea drinkers owing to the small number of participants reporting regular tea drinking. We did not find any association between green tea drinking and CVD risk factors in women (Table 2).

We performed an additional sensitivity analysis to examine the associations between green tea consumption and CVD risk factors, according to participants' self-awareness of their T2DM status. Among men subjects who were not self-aware of T2DM (n = 256), the positive association between green tea drinking and general obesity disappeared. For NAFLD, the positive association was weaker in participants who were not self-aware of their T2DM status than in those who were self-aware (Table 3). Although higher green tea consumption was also significantly associated with higher risks of central obesity and MetS among participants who were aware of their T2DM status, these associations should be treated cautiously due to potential reverse causation. In women, no association was found in either group.

Furthermore, sensitivity analyses according to age, diet quality, smoking status, and alcohol drinking were conducted to assess the robustness of the findings (Table 4). The association of green tea consumption with NAFLD did not change appreciably among participants aged 52 years or older, with lower dietary quality, nonsmokers, and nonalcohol drinkers. However, the association disappeared among participants who were <52 years old, with higher dietary quality, smokers, and alcohol drinkers.

In addition, analyses among participants who were not self-aware of their dyslipidemia status, hypertension status, and not obese were performed. Subjects who were self-aware of dyslipidemia/hypertension were defined as “have dyslipidemia/hypertension” and “know their serum lipids/blood pressure status,” the others were thought to be unaware of their dyslipidemia/hypertension status. As given in Supplementary Table S1, the association between tea drinking and NAFLD was significantly attenuated in subjects who were not self-aware of hypertension, indicating the possible potential reverse causation.

Discussion

In the present community-based cross-sectional study of T2DM patients, regular green tea drinking was found to be associated with an increased risk of NAFLD among male participants aged 52 years or older, and in those with lower dietary quality, but the association should be treated cautiously due to potential reverse causation caused by hypertensive patients. We did not find any association between regular green tea drinking and CVD risk factors in women.

Although we found a positive association between green tea drinking and general obesity in men, this association was not reliable in stratified analyses. Since tea consumption has been suggested to have some beneficial effects on body weight and cardiometabolic risk, ^{6,16 –19} the participants of this study who are self-aware of their T2DM status may have changed their dietary habits, such as drinking more tea. Therefore, potential reverse causation could be the reason why green tea drinking was associated with obesity in men.

Interestingly, green tea drinking was associated with an increased risk of NAFLD among nonsmokers and nondrinkers, while not in smokers and drinkers. Furthermore, a positive association between green tea consumption and NAFLD was found among participants aged ≥52 years, whereas no association was observed among participants aged <52 years (Table 4). Thus, it is of interest to investigate whether smoking and drinking differed by age. Our analysis indicated subjects who smoked and drank were younger than their nonsmoking and nonalcohol drinking counterparts (smokers: 51.69 ± 10.29, nonsmokers: 55.04 ± 12.50, P = .002; alcohol drinkers: 50.14 ± 10.37, nonalcohol drinkers: 54.72 ± 11.65, P < .001). Furthermore, we divided the ages into five groups according to the quintile. For those aged ≤55 years old, no association was found between green tea consumption and risk of NAFLD, whereas for those aged ≥56 years old, ORs of NAFLD were gradually increased with age (Supplementary Table S2). Therefore, the results of smoking- and alcohol drinking-stratified analyses may be explained by age differences.

Tea has had a major influence on the development of Chinese culture, and Chinese traditional culture is closely connected with Chinese tea. Although a large number of studies have demonstrated the potential cardiometabolic benefits of consuming green tea and its functional components, ^{6,16 –19} evidence on the potential detrimental role of green tea consumption on cardiometabolic health is scarce. In a cross-sectional study of Iranian adults, green tea drinking was also associated with a higher prevalence of diabetes. ²⁰ Furthermore, the potentially negative influence of drinking tea on diabetes was also suggested by findings from two large cohort studies conducted in Shanghai, a city that is close to Suzhou, in which participants of this study were residing. Findings from the Shanghai Women's Health Study and the Shanghai Men's Health Study suggested that current green tea drinkers had an increased risk of T2DM compared with noncurrent drinkers, with apparent dose–response relationships for duration and amount of tea drinking. ²¹

Current evidence remains limited as to how consumption of tea may affect CVD risk factors among T2DM. Two RCTs demonstrated that drinking green tea has no effect on glucose homeostasis, lipid metabolism, inflammatory response, or oxidative stress in T2DM patients. ^22,23 Evidence from some observational studies suggested inverse associations between tea drinking and some CVD risk factors, such as obesity, total body fat, and WC. ^18,19 In the study conducted by Micek et al., ⁸ tea consumption was not associated with MetS in either genders and, although tea consumption was inversely associated with WC in the whole population, the relationship disappeared in men. In another RCT, green tea extract effectively increased leptin and reduced LDL-C in overweight and obese women after 6 weeks of treatment, but no remarkable change was found for BMI, WC, or other CVD biomarkers. ²⁴ Our study adds to the literature by evaluating green tea consumption and various risk factors for CVD in patients with T2DM in a large cohort of community-dwelling patients.

NAFLD is a condition characterized by lipid infiltration and accumulation in liver cells. Tan et al. ²⁵ found that green tea polyphenols ameliorate NAFLD by upregulating AMPK activation in Zucker fatty rats fed with a high-fat diet. Current evidence from animal studies generally indicates that tea and its components may prevent steatosis and its progression to NAFLD. ²⁶ However, hepatotoxicity induced by tea or tea extract is also reported. An in vitro study in hepatocytes found that a high dose of epigallocatechin-3-gallate (EGCG) from green tea reduces mitochondrial membrane potential, damages mitochondrial outer membranes, and uncouples oxidative phosphorylation. ²⁷ The cytotoxicity of major tea phenolics toward isolated rat hepatocytes can be arranged as follows: EGCG > propyl gallate > epicatechin-3-gallate > gallic acid epigallocatechin. ²⁸ In an animal study, a high-cholesterol diet for 6 weeks resulted in hepatic steatosis and liver damage. In addition, dietary supplementation with green tea extract increased inflammatory and oxidative stresses. ²⁹ Similarly, the administration of paracetamol or green tea extract in rats resulted in hepatotoxicity, and these changes were more pronounced in the combined intervention. ³⁰ More importantly, Mazzanti et al. summarized the publications between 2008 and 2015, and found that at least 19 reports indicated that the use green tea supplement was associated with adverse hepatic reactions in humans. ³¹

The possibility that tea consumption may also serve as a vehicle for toxic substances that may have potential adverse effects on NAFLD or other metabolic conditions also need to be considered, as other researchers also acknowledged the positive association of tea consumption with T2DM could be attributed to pesticide residues found in tea. ^21,32 Huang et al. ³³ detected that 67% of green tea samples contained some pesticide residues, and most of them contained more than five pesticides. Moreover, the levels of 11 pesticide residues in 18 samples were greater than the values permitted based on international standards. ³³ A meta-analysis including 22 studies found that pesticide exposure increased diabetes incidence by 58%. ³⁴ The presence of heavy metals in tea leaves is another factor to be considered. For example, Zhang et al. detected 10 heavy metals in young and mature tea leaves and attributed the risk on adults mainly to Mn and Al. ³⁵

Our study had several limitations. First, the cross-sectional nature of the study did not allow us to establish any temporal relationship between green tea consumption and cardiovascular risk factors in T2DM patients. Second, the potential reverse causation should be considered as T2DM patients with CVD risk factors may have been drinking more tea, for example, due to the acknowledgment of potential cardiometabolic benefits of tea drinking, ⁶ or due to T2DM-induced thirst. ²¹ Nevertheless, the observed findings did not appear to be fully explained by reverse causation as the awareness rate of NAFLD should be lower than those of other CVD risk factors, and the positive association was not changed among those newly diagnosed T2DM. Third, the amount of tea consumption was not collected, which leads to no quantitative analysis in this study. Fourth, although we were able to perform stratified analyses, the limited sample size in each subgroup limits the reliability of the findings. Fifth, the relatively low proportion of tea drinkers among women not only does not allow stratification by the frequency of green tea drinking but also limits the statistical power for the detection of an association. Finally, our findings may not be generalized to the general populations as we included only T2DM patients.