Relationship among Habitual Tea Consumption ... - Wiley Online Library

Relationship among Habitual Tea Consumption, Percent Body Fat, and Body Fat Distribution Chih-Hsing Wu, Feng-Hwa Lu, Chin-Song Chang, Tsui-Chen Chang, Ru-Hsueh Wang, and Chih-Jen Chang

Abstract WU, CHIH-HSING, FENG-HWA LU, CHIN-SONG CHANG, TSUI-CHEN CHANG, RU-HSUEH WANG, AND CHIH-JEN CHANG. Relationship among habitual tea consumption, percent body fat, and body fat distribution. Obes Res. 2003;11:1088 –1095. Objective: To disclose the possible relationship between habitual tea consumption and changes in total body fat and fat distribution in humans. Research Methods and Procedures: A cross-sectional survey of 1210 epidemiologically sampled adults (569 men and 641 women) were enrolled in our study. Tea consumption and other lifestyle characteristics were obtained by structured questionnaires. Percent body fat (BF%) was measured using bioelectrical impedance analysis. Body fat distribution was assessed using waist-to-hip ratio (WHR). Results: Among the 1103 analyzed subjects, 473 adults (42.9%) consumed tea once or more per week for at least 6 months. The habitual tea drinkers were male-dominant, more frequently current smokers, and alcohol or coffee drinkers than the nonhabitual tea drinkers. Habitual tea drinkers for more than 10 years showed a 19.6% reduction in BF% and a 2.1% reduction in WHR compared with nonhabitual tea drinkers. The multiple stepwise regression models revealed that men, older age, higher BMI, and current smokers were positive factors for BF% and WHR. In contrast, longer duration of habitual tea consumption and higher total physical activity were negative factors for BF%. Longer duration of habitual tea consumption, higher socioeconomic status, and premenopausal status were negative factors for WHR.

Received for review November 22, 2002. Accepted in final form July 14, 2003. Department of Family Medicine, National Cheng Kung University Hospital, Tainan, Taiwan. Address correspondence to Chih-Jen Chang, Department of Family Medicine, National Cheng Kung University Hospital, 138 Sheng-Li Road, Tainan 70428, Taiwan. E-mail: [email protected] Copyright © 2003 NAASO

1088

OBESITY RESEARCH Vol. 11 No. 9 September 2003

Discussion: An inverse relationship may exist among habitual tea consumption, BF%, and body fat distribution, especially for subjects who have maintained the habit of tea consumption for more than 10 years. Key words: tea, percent body fat, bioelectrical impedance analysis, body fat distribution, waist-to-hip ratio

Introduction Recently, Dulloo et al. have demonstrated the positive effect of tea on thermogenesis in rats (1). Increasing fat oxidation and 24-hour energy expenditure in humans were observed in healthy young men who consumed green tea extract experimentally (2). Furthermore, the fact that oolong tea extract may enhance noradrenaline-induced lipolysis in adipose tissue and inhibit pancreatic lipase activity (3) seemed to suggest the possible antiobesity action of tea consumption (4). Interestingly, whether these experimental findings could be extrapolated to the change of percent body fat (BF%)1 in humans is still questionable. To the best of our knowledge, no human study has been reported. On the other hand, tea has also been proven to express a cardiovascular-protective and lipid-lowering effect (5,6). Several studies have shown a close relationship among general obesity, central (abdominal) obesity, and cardiovascular morbidity and mortality (7,8). Supposing there is a relationship between tea consumption and BF%, it would be valuable to know the possible effect of tea consumption on body fat distribution. However, this is also an unexplored issue. There are numerous methods of measuring BF% and body fat distribution (9,10). According to a recent consensus report (11), bioelectrical impedance analysis (BIA) is a feasible, reliable, and comparable method for measuring BF% by well-trained professionals. On the other hand, the World Health Organization recommended that both waist

1

Nonstandard abbreviations: BF%, percent body fat; BIA, bioelectrical impedance analysis; WHR, waist-to-hip ratio; ANCOVA, analysis of covariance.

Habitual Tea Consumption, Percent Body Fat, and Body Fat Distribution, Wu et al.

circumference and waist-to-hip ratio (WHR) are useful methods in evaluating changes in body fat distribution (12). Therefore, it would be logical to use BIA and WHR to assess changes of BF% and body fat distribution, especially in an epidemiological survey. In this study, an epidemiologically sampled Chinese cohort who commonly consumed tea was assessed. The possible confounding lifestyle characteristics, including smoking habits (13,14), alcohol and coffee consumption (13), favored dietary habits, physical activity (13,15), diabetes history, menopausal status (16), and socioeconomic status (17) were queried simultaneously. BIA and WHR were used to measure the BF% and body fat distribution, respectively. From these results, the relationships among BF%, body fat distribution, and habitual tea consumption were evaluated.

Research Methods and Procedures Study Subjects The Tainan diabetes and related chronic disease survey was conducted for 1638 men and women sampled from Tainan, southern Taiwan, using the systemically stratified method during 1996 to 1997 (18). Five hundred sixty-nine male and 641 female subjects ⱖ20 years of age, who completed the second follow-up survey in 1998 to 2000, were enrolled from this report. Exclusion criteria for analyzing subjects were those who had severe renal, thyroid, or hepatic dysfunction, those who had received an operation with mechanical implantation or replacement, and those who suffered from major deformities or contracture of the extremities. Finally, a total of 1103 ambulatory subjects were analyzed in this report. None of the subjects were dieting or following any course of weight control during the study. Each sampled subject visited National Cheng Kung University Hospital, a tertiary medical center located in southern Taiwan, for examination and had their written consent approved by the research committee of this hospital. Questionnaires of Lifestyle Covariates Subjects were interviewed using structured questionnaires. Socioeconomic status was categorized by the modified Hollingshead index (including occupational and educational level) and reclassified into high (grades 3 to 5) and low (grades 1 and 2) status (19). Smoking was dichotomized into noncurrent smoker (never and ex-smokers) and current smoker (more than one cigarette daily) (13). Subjects who had drunk alcohol and coffee more than once per week regularly were recorded as habitual drinkers; otherwise, they were recorded as nonhabitual drinkers. Total physical activity, including leisure activity, occupational activity, and walking for exercise, was calculated as the metabolic equivalent-hour per week for all activities in the past year (20,21). The average frequency of main food intake during the past 1 year, including fatty meat, eggs, fried food, fast

food, smoked food, and seafood, was also recorded and dichotomized as favored (ⱖ3 meals per week) or nonfavored. Medical history including menopausal status and diabetes was checked accordingly. Tea Consumption The questionnaire of tea consumption was modified from the Mediterranean Osteoporosis study (22) and has been used in recent reports (23). The first question was “Have you drunk tea habitually once a week for at least 6 months?” Subjects who answered “yes” were coded as habitual tea drinkers and completed the following questions. 1) What kind of tea (green, black or oolong) was mostly consumed? 2) Do you regularly add milk or sugar to your tea? 3) How often do you drink tea each week? 4) How many times do you drink tea each day? 5) How much (milliliters) tea do you drink each time? 6) How many years have you been drinking tea in this way? Finally, the average amount of daily tea consumption (milliliters) was calculated [(days per week ⫻ times per days ⫻ volume of tea extracts each time)/7]. Total Body Fat and Fat Distribution Wearing light indoor clothes, each subject’s height and weight were measured, and BMI (kilograms per meters squared) was calculated. After overnight fasting, waist circumference was measured midway between the lateral lower rib margin and the superior anterior iliac crest. The subjects were in a standing position with a bare abdomen. Hip circumference was measured at the level of the bilateral great trochanters (12). WHR was calculated and represented the body fat distribution (12). BF% was measured using BIA (BC-Profile, Model 310: Biodynamics Corp.). Subjects were requested to empty the bladder before testing and were told to lie in a supine position in an air-conditioned room set at 26 °C to 28 °C. Two current-introducing and voltagesensing electrodes were placed at the standard positions of the right hand and foot, respectively (11). The agreement of BF% between BIA and DXA (Lunar Radiation Corp.) was performed for 145 men and 123 women. The regression model was as follows: DXA-BF% ⫽ ⫺ 3.76 ⫺ 0.0275 ⫻ Age ⫹ 6.648 ⫻ Sex (male⫽0, female⫽1) ⫹ 0.906⫻ BMI ⫹0.438 ⫻ BIA-BF% (R 2 ⫽ 0.952). The Bland-Altman analysis (24) showed a normal distribution with high agreement of BF% measurement between BIA and DXA in this study. Statistics Analysis Data were analyzed using the SPSSWIN software (SPSS, version 8.0; SPSS, Chicago, IL). ␹2 test, Student’s t test, and OBESITY RESEARCH Vol. 11 No. 9 September 2003

1089


Table 1. Comparison of demographic data between habitual and nonhabitual tea drinkers

Case no. (%) Female (cases %)* Premenopausal* Postmenopausal Age (year)† BMI (kg/m2)‡ Total physical activity (MET-hr/wk)* Socioeconomic status (cases %)* Low (levels 1 to 3) High (levels 4 and 5) Smoking status (cases %)* Noncurrent Current Alcohol drinking (cases %)* Never or nonhabitual Habitual (⭌1/wk) Coffee intake (cases %)* Never or nonhabitual Habitual (⭌1/wk) Favored food intake (⭌3 meals/wk)* Fatty meat (cases %) Eggs (cases %) Fried food (cases %) Fast food (cases %) Smoked food (cases %) Seafood (cases %) Diabetes history (cases %)* No Yes

Nonhabitual tea drinkers

Habitual tea drinkers

630 (57.1%) 384 (61.0%) 217 (56.6%) 167 (43.4%) 48.6 ⫾ 15.9 23.71 ⫾ 3.45 75.7 ⫾ 61.0

473 (42.9%) 186 (29.3%) 137 (73.7%) 49 (26.3%) 48.2 ⫾ 14.7 24.19 ⫾ 3.46 80.7 ⫾ 111.2

409 (64.9%) 221 (35.1%)

308 (65.1%) 165 (34.9%)

p Value ⬍0.001 ⬍0.001 ⬎0.2 ⬎0.2 ⬎0.2 ⬎0.2

⬍0.001 524 (83.2%) 106 (16.8%)

294 (62.2%) 179 (37.8%)

561 (89.0%) 69 (11.0%)

331 (70.0%) 142 (30.0%)

582 (92.4%) 48 (7.6%)

401 (84.8%) 72 (15.2%)

59 (9.4%) 206 (32.7%) 59 (9.4%) 24 (3.8%) 36 (5.7%) 70 (11.1%)

51 (10.8%) 156 (33.0%) 48 (10.1%) 12 (2.5%) 37 (7.8%) 55 (11.6%)

581 (92.2%) 49 (7.8%)

441 (93.2%) 32 (6.8%)

⬍0.001

⬍0.001

⬎0.2 ⬎0.2 ⬎0.2 ⬎0.2 0.179 ⬎0.2 ⬎0.2

* ␹2 test. † Student’s t-test, two-tailed. ‡ ANCOVA, adjusted with sex and age.

analysis of covariance (ANCOVA; age- and sex-adjusted) were used to analyze the differences of basic characteristics between nonhabitual and habitual tea drinkers. Subjects were categorized into four subgroups: nonhabitual, 1 to 5 years, 6 to 10 years, and ⬎10 years as habitual tea drinkers. The differences of BF% and body fat distribution among the four subgroups were analyzed by ANCOVA that was adjusted for age, BMI, sex, and all other lifestyle covariates. Finally, multiple stepwise regression models were applied to assess the independent effect of tea consumption on BF% and total fat distribution and expressed with standardized 1090


regression coefficients. Statistical significance was defined as p ⬍ 0.05 for two-tailed analysis.

Results Among the 1103 analyzed subjects, 473 (42.9%) were habitual tea drinkers. In Table 1, the results show that the habitual tea drinkers were male-dominant, more frequently current smokers, and habitual alcohol or coffee drinkers, but they had similar age, BMI, and total physical activity per week as the nonhabitual tea drinkers. The favored food


Table 2. Comparison of anthropometric parameters between habitual and nonhabitual tea drinkers in 1103 subjects

Case no. BMI (kg/m2) Total body fat (%) Waist circumference (cm) Hip circumference (cm) WHR

Habitual tea drinkers, duration of years


1 to 5 years

6 to 10 years

>10 years

620 25.72 ⫾ 0.43 25.35 ⫾ 0.81* 81.53 ⫾ 0.70† 93.54 ⫾ 0.69* 0.873 ⫾ 0.008†

200 26.09 ⫾ 0.46 25.22 ⫾ 0.88‡ 81.81 ⫾ 0.76§ 93.69 ⫾ 0.54§ 0.873 ⫾ 0.009¶

143 26.16 ⫾ 0.49 24.85 ⫾ 0.92** 81.35 ⫾ 0.81‡‡ 93.16 ⫾ 0.79‡‡ 0.871 ⫾ 0.009‡‡

140 25.62 ⫾ 0.49 20.39 ⫾ 0.92 79.86 ⫾ 0.80 91.88 ⫾ 0.79 0.855 ⫾ 0.009

Data were analyzed by ANCOVA and expressed as adjusted mean ⫾ SD by age, BMI, sex, smoking status, coffee and alcohol drinking habit, favored food intake, total physical activity, diabetes history, and socioeconomic status. Comparison between nonhabitual and ⬎10 years habitual tea drinkers groups: * p ⬍ 0.001, † p ⬍ 0.01. Comparison between 1 to 5 years and ⬎10 years habitual tea drinkers groups: ‡ p ⬍ 0.001, § p ⬍ 0.01, ¶ p ⬍ 0.05. Comparison between 6 to 10 years and ⬎10 years habitual tea drinkers groups: ** p ⬍ 0.001, ‡‡ p ⬍ 0.05.

intake was different between Western and Chinese subjects, but it was quite similar between habitual and nonhabitual tea drinkers. Among the habitual tea drinkers, 455 were green or oolong tea consumers and 18 were black tea consumers. The mean duration of tea consumption was ⬃10.26 years, and the average daily amount was 434.3 mL in habitual tea drinkers. In the dose-response analysis, subjects were categorized into four subgroups by the duration of tea consumption: nonhabitual (620 subjects), 1 to 5 years (200 subjects), 6 to 10 years (143 subjects), and ⬎10 years (140 subjects). After adjustment for age, BMI, total physical activity, favored

food intake, and major lifestyle covariates, the levels of BF%, waist circumference, hip circumference, and WHR were lowest among subjects who consumed tea habitually for more than 10 years compared with the other three subgroups (nonhabitual, 1 to 5 years, and 6 to 10 years habitual tea drinkers) in either total subject, male subject, or female subject groups (Tables 2 to 4). For total subjects, the multiple stepwise linear regression model revealed that the age, sex (male ⫽ 0, female ⫽ 1), BMI, and smoking status (noncurrent ⫽ 0, current ⫽ 1) were positive independent factors of BF%. The duration of tea consumption and total physical activity were negative

Table 3. Comparison of anthropometric parameters between habitual and nonhabitual tea drinkers in 533 male subjects




1 to 5 years

6 to 10 years

>10 years

239 25.29 ⫾ 0.57 20.01 ⫾ 0.98* 86.84 ⫾ 0.88 94.96 ⫾ 0.86† 0.920 ⫾ 0.010

87 26.21 ⫾ 0.65 20.69 ⫾ 1.12‡ 86.60 ⫾ 1.01 95.13 ⫾ 0.98§ 0.915 ⫾ 0.011

98 25.62 ⫾ 0.64 20.57 ⫾ 1.09¶ 86.68 ⫾ 0.98 94.41 ⫾ 0.95 0.922 ⫾ 0.011

109 25.23 ⫾ 0.62 16.81 ⫾ 1.05 86.03 ⫾ 0.95 93.40 ⫾ 0.92 0.915 ⫾ 0.011

Data were analyzed by ANCOVA and expressed as adjusted mean ⫾ SD by age, BMI, smoking status, coffee and alcohol drinking habit, favored food intake, total physical activity, diabetes history, and socioeconomic status. Comparison between nonhabitual and ⬎10 years habitual tea drinkers groups: * p ⬍ 0.001, † p ⬍ 0.05. Comparison between 1 to 5 years and ⬎10 years habitual tea drinkers groups: ‡ p ⬍ 0.001, § p ⬍ 0.05. Comparison between 6 to 10 years and ⬎10 years habitual tea drinkers groups: ¶ p ⬍ 0.001.


1091


Table 4. Comparison of anthropometric parameters between habitual and nonhabitual tea drinkers in 570 female subjects




1 to 5 years

6 to 10 years

>10 years

381 25.51 ⫾ 0.72 31.10 ⫾ 1.47* 76.05 ⫾ 1.25* 93.19 ⫾ 1.27† 0.810 ⫾ 0.015*

113 25.42 ⫾ 0.76 30.02 ⫾ 1.55‡ 76.52 ⫾ 1.31‡ 93.27 ⫾ 1.34§ 0.815 ⫾ 0.016‡

45 25.94 ⫾ 0.81 29.46 ⫾ 1.66¶ 75.89 ⫾ 1.41** 93.29 ⫾ 1.44†† 0.807 ⫾ 0.017**

31 25.83 ⫾ 0.87 21.78 ⫾ 1.79 72.21 ⫾ 1.51 90.69 ⫾ 1.54 0.765 ⫾ 0.018

Data were analyzed by ANCOVA and expressed as adjusted mean ⫾ SD by age, BMI, menopausal status, smoking status, coffee and alcohol drinking habit, favored food intake, total physical activity, diabetes history, and socioeconomic status. Comparison between nonhabitual and ⬎10 years habitual tea drinkers groups: * p ⬍ 0.001, † p ⬍ 0.05. Comparison between 1 to 5 years and ⬎10 years habitual tea drinkers groups: ‡ p ⬍ 0.001, § p ⬍ 0.05. Comparison between 6 to 10 years and ⬎10 years habitual tea drinkers groups: ¶ p ⬍ 0.001, ** p ⬍ 0.01, †† p ⬍ 0.05.

independent determinants of BF% (Table 5). On the other hand, age, BMI, smoking status, and menopausal status (pre- ⫽ 0, post- ⫽ 1) were positive independent factors for WHR, but the duration of years of tea consumption, sex, and socioeconomic status (grades 1 to 3 ⫽ 0, grades 4 and 5 ⫽ 1) were negative independent determinants for WHR. That is, the longer the duration of years of tea consumption, the lower the BF% and WHR. Others, such as the average daily amount of tea consumption, habit of alcohol or coffee

drinking, favored food intake, and history of diabetes, were all nonsignificant determinants for the changes of BF% and body fat distribution in the population we studied.

Discussion For total subjects, we found that habitual tea drinkers had a lower BF% and smaller WHR than the nonhabitual ones. In general, those who had maintained the habit of tea

Table 5. Multiple stepwise regression models of tea consumption, lifestyle associated factors, total BF%, and fat distribution in 1103 subjects

Independent variables

Model 1 (Adjusted R2 value ⴝ 0.495)

Model 2 (Adjusted R2 ⴝ 0.584)

Age (years) BMI (kg/m2) Sex (male ⫽ 0, female ⫽ 1) Menopausal status (pre- ⫽ 0, post- ⫽ 1) Total physical activity (MET-hour/wk) Duration of habitual tea consumption (years) Smoking status (noncurrent ⫽ 0, current ⫽ 1) Socioeconomic status (grades 1 to 3 ⫽ 0, grades 4 and 5 ⫽ 1)

0.302* 0.314* 0.571* — ⫺0.082* ⫺0.133* 0.055‡ —

0.309* 0.356* ⫺0.446* 0.070† — ⫺0.052‡ 0.083* ⫺0.051‡

Model 1 uses BF% as the dependent variable and model 2 uses WHR as the dependent variable. Only significant variables are shown with standardized regression coefficients. The other nonsignificant variables include average daily amount of tea consumption, alcohol or coffee drinking, major favored food intake, and diabetes history. * p ⬍ 0.001, † p ⬍ 0.01, ‡ p ⬍ 0.05.

1092



consumption for more than 10 years had a reduction of 19.6% of BF% and 2.1% of WHR compared with nonhabitual subjects. Therefore, long-term habitual tea consumption may influence fat metabolism. This effect may not only be expressed in experimental or animal energy-balancing theories (1– 4) but also may be demonstrated through clinical changes of BF% and fat distribution, which were first described in this study. Consistent with other reports, current smokers exhibited a greater degree of abdominal obesity, as represented by the higher WHR (13,14), than noncurrent smokers in our study. A pattern of steadily increasing BF% and central fat accumulation with age has been shown in our study and other studies (16,25). The impact of physical activity on adipose tissue metabolism (15), which may be reflected by the reduction of BF% (13), was also found in this study. On the other hand, lower socioeconomic status could represent a relatively higher depressive-stress reaction and cortisol arousal that may consequently result in centralization of body fat, as shown in Table 5 and in previous reports (17,19,26). Finally, the phenomenon of abdominal accumulation of body fat may be enhanced through the menopausal effect, as described in a recent report (16). Several studies (27–30) have used the average daily amount (or cups) of tea consumption as defining the level of tea consumption. We have also tried different cut-points (30,31) of daily tea consumption to classify study subjects into different subgroups, but no consistent results could be found. In our study, the multiple linear regression standardized coefficients of the daily tea consumption for BF% and WHR are around ⫺0.030 and ⫺0.009, respectively. Unfortunately, the effect of daily tea consumption on BF% or WHR was overwhelmed by the stronger effect of duration of tea drinking in the regression models as a whole. Moreover, as the daily amount of tea consumption may fluctuate more than the maintenance of a tea-consuming habit, the recall bias may be higher in the former. Therefore, it is plausible that duration of tea consumption was more appropriate than the average amount of daily tea consumption as an independent factor; this was found in a previous study (23) and in our report (Table 5). However, whether this unique finding could be a bias attributed to different drinking habits between Chinese and Western subjects needs to be investigated in the future. Tea extracts may enhance thermogenesis (1), decrease intestinal glucose (32) and lipid absorption (4), and increase 24-hour energy expenditure (2). They may also enhance noradrenaline-induced lipolysis in adipose tissue (4) or inhibit lipase activity (3). These results may explain the relationship between tea consumption and changes of BF%. On the other hand, regional fat cell lipolysis and accumulation were modulated by the counter-regulated actions of insulin and catecholamine (33). The catecholamine-induced lipolytic activity was found to be highest in visceral fat but lowest

in peripheral subcutaneous fat (33). Thus, the stimulated sympathetic nervous system activity is associated with variations in body fat and body fat distribution (34). Tea may promote the activity of catecholamine (4) and sympathetic tone (1), which in turn, contributes to the favorable changes of abdominal fat distribution observed in this study. BMI is the method commonly used as a surrogate measure for obesity; it is the degree of body fat accumulation (35). Compared with whites, an inconsistent relationship between BMI and BF% was demonstrated recently in Taiwanese subjects (36). There is evidence that different body sizes may accommodate a wide variation of BF%s and WHRs for similar BMI levels (37,38). While the increment of fat oxidation, 24-hour energy expenditure, and lipolysis in adipose tissue seem to suggest possible changes of body fat content in tea drinkers, BMI may be similar between tea and nontea drinkers (27). Therefore, the observed unsynchronized changes in BF%, WHR, and BMI are possible in this study. In this study, men had less BF% than women, but it was predominantly located around the abdomen. It has been demonstrated that sexual dimorphism of body fat distribution may be influenced by the plurimetabolic responses of fat cells to physiological regulation, such as sympathetic tone, insulin, catecholamine, estrogen levels, etc. (39,40). Tea may enhance insulin activity (41), and because it contains catecholamine and phytoestrogen, it may work in concert to express different changes of body fat distribution between men and women. In this study, compared with the women, the male subjects had a higher incidence of smoking (49.9% vs. 3.0%), had a higher incidence of alcohol intake (35.6% vs. 3.7%), had a higher BMI (24.5 vs. 23.3 kg/m2), and were older (49.4 vs. 47.5 years). These confounding factors may influence the changes of WHR (42), and even after adjustment, may outweigh the actual effect of tea consumption on the small changes of WHR in men (Table 3). Two potential limitations of this study should be considered. First, the habitual tea drinkers were classified using the questionnaire but not by direct measurement of tea catechin. However, an optimally designed dietary questionnaire has high validity with actual habits of food intake (43), and the habit of tea consumption is seen by the public to be healthy, so the amounts are likely to be accurate. Therefore, it is reasonable to assume that the level of tea consumption assessed by using questionnaires, which is similar to other epidemiological reports, was accurate (22,23). Second, recall bias was another possible limitation. However, most of the discrepancies that could be found in other studies were related to complicated behavior or distant past recall, such as 20 to 30 years ago (44). For the simple behavioral habit of tea consumption, the self-reported questionnaire used in this study should be sufficiently reliable. In conclusion, except for the well-known determinants, it is suggested that the habit of tea consumption may have an OBESITY RESEARCH Vol. 11 No. 9 September 2003

1093


inverse relationship with the changes of BF% and body fat distribution in both male and female adults. Habitual tea drinkers, especially those who have maintained the habit of tea consumption for more than 10 years, had a lower BF% and smaller WHR than nonhabitual tea drinkers. These findings indicate that tea could be used as a possible healthy weight loss beverage.

Acknowledgments This study was supported by Grants NSC 87-2314-B006-081, NSC 88-2314-B-006-082, and NSC 91-2314-B006-077. We thank the colleagues of the Department of Family Medicine, NCKUH, for help in this epidemiological survey. References 1. Dulloo AG, Seydoux J, Giradier L, Chantre P, Vandermander J. Green tea and thermogenesis: interactions between catechin-polyphenols, caffeine and sympathetic activity. Int J Obes Relat Metab Disord. 2000;24:252– 8. 2. Dulloo AG, Duret C, Rohrer D, et al. Efficacy of a green tea extract rich in catechin polyphenols and caffeine in increasing 24-h energy expenditure and fat oxidation in humans. Am J Clin Nutr. 1999;70:1040 –5. 3. Han LK, Kimura Y, Kawashima MT, et al. Anti-obesity effects in rodents of dietary teasaponin, a lipase inhibitor. Int J Obes Relat Metab Disord. 2001;25:1459 – 64. 4. Han LK, Takaku T, Li J, Kimura Y, Okuda H. Antiobesity action of oolong tea. Int J Obes Relat Metab Disord. 1999;23:98 –105. 5. Vinson JA, Dabbagh YA. Effect of green and black tea supplementation on lipids, lipid oxidation and fibrinogen in the hamster: mechanism for the epidemiological benefits of tea drinking. FEBS Lett. 1998;433:44 – 6. 6. Imai K, Nakachi K. Cross sectional study of effects of drinking green tea on cardiovascular and liver disease. Br Med J. 1995;310:693– 6. 7. Folsom AR, Kushi LH, Anderson KE, et al. Association of general and abdominal obesity with multiple health outcomes in older women. The Iowa Women’s Health Study. Arch Intern Med. 2000;160:2117–28. 8. Abate N. Obesity as a risk factor for cardiovascular disease. Am J Med 1999;107:12S–13S. 9. Brodie DA. Techniques of measurement of body composition, Part I. Sports Med. 1988;5:11– 40. 10. Brodie DA. Techniques of measurement of body composition, Part II. Sports Med. 1988;5:74 –98. 11. Anonymous. Bioelectrical impedance analysis in body composition measurement: National Institutes of Health Technology Assessment Conference Statement. Am J Clin Nutr 1996; 64(suppl):524S–32S. 12. International Obesity Task Force. Defining the problem of overweight and obesity. In: Obesity: Preventing and Managing the Global Epidemic: Report of WHO Consultation on Obesity. Geneva, Switzerland: World Health Organization; 1998, pp. 7–17. 1094


13. Troisi RJ, Heinold JW, Vokonas PS, Weiss ST. Cigarette smoking, dietary intake, and physical activity: effects on body fat distribution—the Normative Aging Study. Am J Clin Nutr. 1991;53:1104 –11. 14. Shimokata H, Muller DC, Andres R. Studies in the distribution of body fat: III. Effects of cigarette smoking. JAMA. 1989;261:1169 –73. 15. Arner P. Impact of exercise on adipose tissue metabolism in humans. Int J Obes Relat Metab Disord. 1995;19(suppl 4): S18 –S21. 16. Chang CJ, Wu CH, Yao WJ, Yang YC, Wu JS, Lu FH. Relationships of age, menopause and central obesity on cardiovascular disease risk factors in Chinese women. Int J Obes Relat Metab Disord. 2000;24:1699 –704. 17. Rosmond R, Bjo¨rntorp P. Psychosocial and socio-economic factors in women and their relationship to obesity and regional body fat distribution. Int J Obes Relat Metab Disord. 1999; 23:138 – 45. 18. Lu FW, Yang YC, Wu JS, Wu CH, Chang CJ. A population-based study of the prevalence and associated factors of diabetes mellitus in southern Taiwan. Diabet Med. 1998;15: 564 –72. 19. Kraus J, Borhani NO, Franti CE. Socioeconomic status, ethnicity, and risk of coronary heart disease. Am J Epidemiol. 1980;111:407–14. 20. Kriska AM, Bennett PH. An epidemiological prospective of the relationship between physical activity and NIDDM: from activity assessment to intervention. Diabetes Metab Rev. 1992;4:355–72. 21. Paffenbarger RS, Blair SN, Lee IM, Hyde RT. Measurement of physical activity to assess health effects in free-living populations. Med Sci Sports Exerc. 1993;25:60 –70. 22. Dequeker J, Ranstam J, Valson J, Sigurgevisson B, Allander E, for the MEDOS Study Group. The Mediterranean Osteoporosis (MEDOS) study questionnaire. Clin Rheumatol 1991;10:54 – 69. 23. Wu CH, Yang YC, Yao WJ, Lu FH, Wu JS, Chang CJ. Epidemiological evidence of increased bone mineral density in habitual tea drinkers. Arch Intern Med. 2002;162:1001– 6. 24. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–10. 25. Going S, Williams D, Lohman T. Aging and body composition: biological changes and methodological issues. Exerc Sport Sci Rev. 1995;23:411–58. 26. Rosmond R, Lapidus L, Bjo¨rntorp P. The influence of occupational and social factors on obesity and body fat distribution in middle-aged men. Int J Obes Relat Metab Disord. 1996;20:599 – 607. 27. Tsubono Y, Takahashi T, Iwase Y, Iitoi Y, Akabane M, Tsugane S. Dietary differences with green tea intake among middle-aged Japanese men and women. Prev Med. 1997;26: 704 –10. 28. Arts ICW, Hollman PCH, Feskens EJM, Bueno de Mesquita HB, Kromhout D. Catechin intake and associated dietary and lifestyle factors in a representative sample of Dutch men and women. Eur J Clin Nutr. 2001;55:76 – 81. 29. Arts ICW, Hollman PCH, Feskens EJM, Bueno de Mesquita HB, Kromhout D. Catechin intake might explain the


30.

31.

32.

33. 34.

35.

36.

inverse relation between tea consumption and ischemic heart disease: the Zutphen elderly study. Am J Clin Nutr. 2001;74: 227–32. Geleijnse JM, Launer LJ, van der Kuip DAM, Hofman A, Witteman JCM. Inverse association of tea and flavonoid intakes with incident myocardial infarction: the Rotterdam study. Am J Clin Nutr. 2002;75:880 – 6. Hegraty VM, May HM, Khaw KT. Tea drinking and bone mineral density in older women. Am J Clin Nutr. 2000;71: 1003–7. Kobayashi Y, Suzuki M, Satsu H, et al. Green tea polyphenols inhibit the sodium-dependent glucose transporter of intestinal epithelial cells by a competitive mechanism. J Agric Food Chem. 2000;48:5618 –23. Arner P. Regulation of lipolysis in fat cells. Diabetes Rev. 1996;4:450 – 63. Poehlman ET, Gardner AW, Goran MI, et al. Sympathetic nervous system activity, body fatness, and body fat distribution in younger and older males. J Appl Physiol. 1995;78:802– 6. World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Report of a WHO Consultation on Obesity. Geneva, Switzerland: World Health Organization; 1998. Chang CJ, Wu CH, Chang CS, et al. Low body mass index but high percent body fat in Taiwanese—implications of obesity cut-offs. Int J Obes Relat Metab Disord. 2003;27:253–9.

37. Lohman TG. Advances in Body Composition Assessment, Current Issues in Exercise Science Series. Champaign, IL: Human Kinetics Publishers; 1992. 38. Slosman DO, Casez JP, Pichard C, et al. Assessment of whole-body composition with dual energy X-ray absorptiometry. Radiology. 1992;185:593– 8. 39. Jensen MD. Gender differences in regional fatty acid metabolism before and after meal ingestion. J Clin Invest. 1995;96: 2297–303. 40. Poehlman ET, Toth MJ, Ades PA, Calles-Escandon J. Gender differences in resting metabolic rate and noradrenaline kinetics in old individuals. Eur J Clin Invest. 1997;27:23– 8. 41. Anderson RA, Polansky MM. Tea enhances insulin activity. J Agric Food Chem. 2002;50:7182– 6. 42. Slattery ML, McDonald A, Bild DE, et al. Associations of body fat and its distribution with dietary intake, physical activity, alcohol, and smoking in blacks and whites. Am J Clin Nutr. 1992;55:943–9. 43. Nes M, van Staveren W, Zajkas G, Inelmen E, Moreiras O. Validity of the dietary history method in elderly subjects. Eur J Clin Nutr. 1991;45;97–104. 44. Falkner KL, Trevisan M, McCann SE. Reliability of recall of physical activity in the distant past. Am J Epidemiol. 1999; 150:195–205.


1095