Habitual tea consumption and 5-year incident metabolic syndrome among older adults: a community-based cohort study

Metabolic syndrome (MetS) is a global health concern associated with cardiovascular diseases, common cancers, increased mortality and great socieconomic burden, particularly in aged populations. The prevalence of MetS has increased over recent decades, reaching alarming rates worldwide [1, 2]. In 2016, the World Health Organization estimated that one-fifth of the world’s adult population will develop MetS during their lifespan and the impact on their wellbeing and the health systems is tremendous. Thus, understanding its modifiable risk factors is crucial for formulating population-based intervention strategies for this condition.

Tea, the dried leaves of the plant Camellia sinensis, is a popular beverage worldwide. Recent evidence has suggested that tea consumption was associated with the risk of MetS but the effect seemed to be controversial. While some studies concluded that tea consumption is protective for MetS [3, 4], others reported non-significant associations [5, 6]. Furthermore, one study found that frequent consumption of tea increased the risk of MetS in rural China [7].

China is the most populous country all over the world and may have the greatest burden of MetS [8]. It is estimated that one-third of the Chinese adults are affected by MetS [9]. Meanwhile, tea, especially green tea, is very popular among middle-aged to elderly Chinese adults. Suzhou is one of the most important green tea producing areas in China, where tea drinking is a common lifestyle habit among general populations. In addition, the existing conflicting associations of tea and MetS are likely due to the lack of longitudinal study design. The purpose of this study is to explore the prospective associations between tea consumption and 5-year incidence of MetS in a community-based cohort study on older adults in Suzhou of China. The findings might be important for formulating non-pharmacological strategies for supplementing the management of MetS and its associated complications, especially in areas where tea drinking is a widely accepted cultural practice.

Among the 4579 participants with complete data in the baseline survey, we excluded 183 (4.0%) participants who had died prior to the 5-year follow-up examination and 526 (11.5%) participants who were lost to follow-up or deceased. The mean age of the participants at follow-up examinations was 70.7 ± 5.7 years (range: 65.0-76.4 years). Among the 3870 participants who successfully attended the follow-up examinations, 779 with MetS at baseline and 86 who had missing information for MetS features in the follow-up period were excluded as well, leaving 3005 participants available for the current prospective analyses ultimately. In the remained 3005 participants free of MetS at baseline survey, there were 1077 habitual and 1928 non-habitual tea drinkers, respectively. Table 1 compares the baseline characteristics between habitual and non-habitual tea drinkers. In general, habitual tea drinkers were more likely to be men (P < 0.001), current smokers (P < 0.001), alcohol drinkers (P < 0.001), and better educated P < 0.001). They also tended to live with spouse (P < 0.001), have higher income (P < 0.001), and have more physical activity (P < 0.001).

The results of sensitivity analysis are presented in Fig. 1. We selected age, gender, initial BMI, educational level, monthly income and marriage status as confounding factor to be controlled. Prospective associations between tea consumption and the incident MetS were further examined and the results are shown in Table 2. After adjusting for baseline demographics such as initial BMI, educational level, monthly income and marriage status (Model 2), habitual drinkers had a 31% increased risk of developing MetS during the 5-year follow-up period compared with non-habitual drinkers (OR = 1.31, 95% CI: 1.01-1.71; P = 0.04). However, the significant association was not observed either in men (OR = 1.38, 95%CI: 0.98-1.95; P = 0.06) or women (OR = 1.17, 95%CI: 0.76-1.79; P = 0.48). Correspondingly, 5-year cumulative incidence of MetS was found to be higher in those who drank tea more than 5 times per week (OR = 1.38, 95% CI: 1.05-1.82; P = 0.02) as compared with non-habitual drinkers. This relationship still persisted in men (OR = 1.43, 95%CI: 1.00-2.01; P = 0.05) but was not significant in women (OR = 1.28, 95%CI: 0.79-2.08; P = 0.32) when stratified by gender. The association between green tea consumption and incident MetS was non-significant in all study participants (OR = 1.27, 95% CI: 0.97-1.67; P = 0.09), men (OR = 1.26, 95% CI: 0.91-1.76; P = 0.17) and women (OR = 1.22, 95% CI: 0.75-1.99; P = 0.42). Participants who drank tea for 16 to 30 years were more likely to develop MetS than non-habitual drinkers (OR = 1.53, 95%CI: 1.02-2.28; P = 0.04). The results were significant in neither men (OR = 1.39, 95%CI: 0.85-2.28; P = 0.19) nor women (OR = 1.79, 95%CI: 0.87-3.71; P = 0.11).

Figure 2 illustrates the prospective associations between tea consumption and individual components of MetS. No significant relationship was observed between the tea consumption at baseline and incident high BP, high BMI, diabetes mellitus or low HDL-C levels. The presence of tea consumption at baseline was related to high TG in both Model 1 (OR = 1.34, 95% CI: 1.03-1.73; P = 0.03) and Model 2 (OR = 1.32, 95% CI: 1.02-1.72; P = 0.04) among the participants, and the significant association still exist in men after stratified by gender (OR = 1.47, 95% CI: 1.03-2.10; P = 0.03).

Considering that tea consumption was much more frequent in men than women, we performed a gender-stratified analysis and the results are depicted in Fig. 1. Gender could be a possible influencing factor on the associations between habitual tea consumption and individual components of MetS. We found that low HDL-C levels were related to tea consumption in men (OR = 2.03, 95% CI: 1.11-3.69; P = 0.02). However, in women, tea consumption was positively associated with high BMI and elevated BP. Additionally, tea consumption was related with the presence of diabetes mellitus among women in Model 2 (OR = 2.02, 95% CI: 1.07-3.83; P = 0.03). The interaction effects between tea consumption and other variables were not observed.

In this 5-year prospective study on older Chinese adults, increased risk of developing MetS was observed in those who drank tea more than 5 times per week as compared with non-habitual drinkers. A dose-response relationship was also observed with increased frequency of green tea consumption being associated with a higher risk of MetS. In addition, the association of tea consumption and individual components of MetS varied between men and women. These findings added novel knowledge to the current literatures regarding the controversial effect of tea consumption on cardiovascular and metabolic health among the elderly.

The main finding of our study was different from most studies, which reported beneficial effects of tea consumption for MetS and some of its individual components [3, 4, 19]. A systematic review and meta-analysis summarized the findings of 6 observational studies and demonstrated that tea consumption were associated with a reduced risk of MetS [20]. Meanwhile, null associations were also reported in some other studies [5, 6, 21]. Weak relation of tea with MetS pointed towards the potential importance of composition of polyphenols and the types of tea consumed, and these observations needs to be confirmed in well-designed cohort studies. Based on the results of our analysis, there is, however, an increased risk of developing MetS in older adults who drank tea more frequently. Particularly, we found that habitual drinkers who drank tea more than 5 times per week were more susceptible to MetS. In addition, the relationship remained significantly different among men. In our study, women were less likely to be habitual tea drinkers compared with men, which reduced the sample size of women and may mask the association between tea consumption and MetS. Moreover, gender-stratified analysis may weaken the statistical power as the sample size was reduced by nearly half, making the effect of tea consumption on MetS insignificant either in men or women. Previous results indicated the deleterious effects of tea consumption, which was independently associated with an increased risk of kidney stones in Northern Chinese [22]. Concentrated herbal extracts from green tea may increase the burden of liver metabolism and may not be free of adverse effects under certain circumstances [23]. Additionally, about 80-85% of oxalate in diets of Chinese individuals was derived from tea and coffee [24], as a result, heavy tea-drinking habits will increase the excretion of urinary oxalic acid, which may lead to kidney stone formation, and eventually increase the burden of circulation. Similar results were not found when stratified by tea drinking duration. We considered that the information which relied on the subjective judgment and recall of the participants may lead to misclassification and information bias, thus may have distorted the effect estimates of tea consumption on MetS. In addition, perennial tea drinking habits may change due to some influential factors, such as health education and disease. What’s more, the information about the preparation method of tea (brewing or boiling) and the container of tea drinking was not available, so the serving size of tea may not be described accurately. Moreover, we found that habitual drinkers were more likely to be alcohol consumers, as shown in Table 1. The baseline characteristics of the 709 participants who did not enter the analysis were not different from those who entered the analysis. In rural China, having tea after taking alcohol is thought to be able treat effects of drunkenness or hangover. By leading the toxic acetaldehyde converted from alcohol to the kidney along with tea absorption before it decomposes, tea consumption after drinking could damage the kidney and finally disrupt normal blood flow [25]. Further studies are warranted to validate these hypotheses.

With regards to the association between tea consumption and individual components of MetS, we observed a gender difference. Low HDL-C levels were related to tea consumption in men while tea consumption was positively associated with high BMI, elevated BP and the presence of diabetes mellitus in women. The mechanisms underlying this phenomenon might be attributed to different lifestyles between men and women. For instance, women are significantly less physically active than men and obese women are most functionally impaired, and had lower perceived behavioral control toward physical activity [26]. Apart from this, one survey also found that men were more likely to smoke and drink alcohol [27]. Physical activity, smoking and alcohol intake are well-established risk factors for MetS and thus may modify the tea-MetS association between men and women.

The public health implication of our findings needs to be discussed. Tea consumption is traditionally considered to be a promising non-pharmacological strategy for supplementing the management of hypertension, obesity or diabetes, especially in places where tea drinking is a widely accepted cultural practice. In the era of rapid population aging and high prevalence of chronic diseases, tea drinking is regarded as an inexpensive and applicable dietary practice, which may be clinically relevant and show its public health importance. However, our study indicated that these beneficial effects need carefully re-consideration, especially when the drinkers are restricted to older adults. Some other aspects such as frequency of drinking, way of preparations and time for drinking should also be taken into consideration. The findings might be important for formulating non-pharmacological strategies for supplementing the management of MetS.

Our study is a community-based cohort study with a large sample size and a reasonable follow-up rate. Therefore, we provided a more cogent evidence than case-control or cross-sectional studies. Several limitations should also be acknowledged. First, the included participants were restricted to older Chinese adults who lived in the eastern part of China with proportionately high prevalence and incidence of MetS and extensive extrapolation of the findings to other populations needs further clarification. Second, tea consumption was self-reported and thus may result in recall biases, especially in older adults with cognitive decline. Finally, although we had controlled for a wide range of confounders, the possibility of unmeasured and residual confounding such as chronic diseases or medication treatment, the type of diet and dietary patterns, and whether malnutrition was involved that might have contributed to the risk of MetS, which might distort the associations observed in this study. Diet in Jiangnan region is healthier than that in other regions. On the one hand, residents are used to cooking with squeezed rapeseed oils, which is rich in Omega-6 and contributes to reducing total mortality [28‐30]. On the other hand, due to the sufficient rainfall and fertile plains, Jiangnan residents often eat abundant aquatic products and fresh vegetables and fruits [31]. More studies are required to confirm the findings from this current study.

In conclusion, increased risk of developing MetS was observed in high-frequent tea drinkers among older adults. Further well-designed cohort studies with more precise measurement of tea consumption and long-term follow-up duration are warranted to replicate our findings and the mechanisms underlying the adverse effect of tea consumption needs to be elucidated.