点击显示 收起
1 From the National Public Health Institute, Department of Epidemiology and Health Promotion, Helsinki, Finland; the Institute for Preventive Medicine, Nutrition, and Cancer, Folkhälsan Research Center, Helsinki, Finland; and the Department of Clinical Chemistry, University of Helsinki, Helsinki, Finland.
2 Supported by the Finnish Cultural Foundation and the Yrjö Jahnsson Foundation. 3 Address reprint requests to A Kilkkinen, National Public Heath Institute, Department of Epidemiology and Health Promotion, Mannerheimintie 166, FIN-00300 Helsinki, Finland. E-mail: annamari.kilkkinen{at}ktl.fi.
ABSTRACT
Background: The lignan enterolactone, which is produced by the intestinal microflora from dietary precursors, may protect against hormone-dependent cancers and cardiovascular diseases.
Objective: We examined the cross-sectional associations between the serum enterolactone concentration and variables related to diet and health in Finnish adults.
Design: Serum enterolactone was measured by using time-resolved fluoroimmunoassay in 2380 Finnish men and women aged 25–64 y who were participating in a cross-sectional national survey in 1997. Background information was collected with self-administered questionnaires and the diet was assessed with a food-frequency questionnaire.
Results: The median serum enterolactone concentration was 13.8 nmol/L (range: 0–95.6 nmol/L) in men and 16.6 nmol/L (range: 0–182.6 nmol/L) in women. Multiple regression analyses showed positive associations in men between the serum enterolactone concentration and constipation, consumption of whole-grain products, and intake of fruit and berries. In women, the serum enterolactone concentration was positively and independently associated with consumption of vegetables, subject age, and constipation and was negatively associated with smoking. Furthermore, female subjects of normal weight had significantly higher serum enterolactone concentrations than did their underweight or obese peers.
Conclusions: The serum enterolactone concentration varies widely in the population. Of the variables we examined, the most important determinants of the serum enterolactone concentration were consumption of lignan-containing foods and constipation; however, these appeared to explain only a small part of the variation. Therefore, the role of gut microflora in the metabolism of lignans might be very important. Further studies will also be needed to determine the bioavailability and absorption rate of lignans.
Key Words: Cross-sectional study • enterolactone • Finland • diet • lignans • phytoestrogens • cancer prevention • cardiovascular disease prevention • hormone-dependent cancers • intestinal microflora • gut microflora
INTRODUCTION
Lignans, a type of phytoestrogen, are naturally occurring diphenolic plant compounds. These compounds were shown to possess a broad spectrum of biological properties with the potential to reduce the risk of breast, prostate, and colon cancers and cardiovascular diseases (1–4). Some epidemiologic evidence also indicates that phytoestrogens may protect against breast cancer and cardiovascular diseases (5, 6).
Lignans are found in many plants. Flaxseed is the most abundant food source of lignans, whereas other seeds, nuts, whole grains, berries, fruit, and vegetables contain lesser amounts (7–9). After human consumption of plant lignans, gut microflora cause enzymatic metabolic conversions resulting in the formation of mammalian lignans, the main component of which is enterolactone (10–12). Many precursors of mammalian lignans may still be unknown (4).
The traditional Finnish diet seems to be richer in lignans than typical Western diets because of the relatively high intakes of whole-grain products, particularly rye bread, and berries. Therefore, we sought to define the range of serum enterolactone concentrations and to examine cross-sectional associations between serum enterolactone and variables related to diet and health in a Finnish adult population.
SUBJECTS AND METHODS
Subjects
The subjects of the present study were participants in a cross-sectional survey carried out in Finland during the early spring of 1997. Participants were aged 25–64 y and lived in 5 areas: Helsinki and Vantaa (the metropolitan area), the cities of Turku and Loimaa and some rural communities near Loimaa, and the provinces of North Karelia, Kuopio, and Oulu. An independent random sample (n = 10 000), stratified by 10-y age groups, regions, and sex, was drawn from the National Population Register. For the dietary survey subsample, 4000 of these subjects (40%) were chosen and 72% participated in the study. The participation rate was lower for men than for women and was lower among younger persons than among older persons (13). Subjects who had used antibiotics within the past 3 mo (n = 373) were excluded from this study. A total of 1168 men and 1212 women were included in the analysis. The survey was approved by the Ethics Committee of the National Public Health Institute. All subjects gave their written, informed consent to participate in this study.
Data collection
Data on health, socioeconomic factors, smoking, alcohol consumption, and dietary habits were collected with a self-administered questionnaire that subjects completed at the examination. Subjects also completed a food-frequency questionnaire (FFQ) regarding their dietary intake during the previous 12 mo; the FFQ included 38 food items. Consumption of whole-grain products was determined from the FFQ by summing the frequency of consumption (servings/mo) of rye bread, crisp bread, porridge, and cereals. Consumption of vegetables was calculated by summing the frequency of consumption of salad vegetables, root vegetables, legumes, and vegetable dishes. Consumption of fruit and berries was determined by summing the frequency of consumption of fruit, berries, fruit juices, and berry juices. For the analyses, subjects were divided into tertiles by frequency of consumption of whole-grain products, vegetables, and fruit and berries.
Alcohol consumption was calculated from the self-administered questionnaire, which covered the frequency of use and the average portion size for beer, wine, and spirits during the previous 12 mo. The following cutoffs for weekly ethanol intake were used for men: 0 g = none, 1–139 g = low, 140–280 g = moderate, and >280 g = heavy. The corresponding cutoffs for women were 0 g, 1–104 g, 105–190 g, and >190 g. The subjects were also divided into 3 categories of self-reported physical health (poor, satisfactory, and good) and were classified as nonsmokers, former smokers, or current smokers. Education categories were formed by dividing subjects born in the same year into tertiles on the basis of their total number of years of education.
Weight and height were measured at the examination. Body mass index (BMI) was computed as weight in kg divided by the square of height in m. The following cutoffs for BMI were used: <20.1 = underweight, 20.1–25.0 = normal weight, 25.1–30.0 = overweight, 30.1–35.0 = obese, and >35.0 = severely obese.
Analytic methods
At the examination, which occurred between 1100 and 1900 on a weekday, a venous blood sample was obtained from each subject after he or she had fasted for 4 h. The samples were fractionated and then frozen in aliquots. The serum aliquot used in this analysis was stored at -20°C for 1 y until it was analyzed for enterolactone in the spring of 1998.
The analyses were performed by using time-resolved fluoroimmunoassay (TR-FIA) (14). After this method was first published, it was improved and a more rapid modification was developed (15). The more rapid modification, which we used in this study, is as follows: 50-µL serum samples were incubated with 50 µL hydrolysis reagent containing sulfatase and ß-glucuronidase overnight at 37°C. After hydrolysis, 150 µL 0.5% bovine serum albumin–tris buffer (pH 7.8) was added to the sample to obtain the optimal pH and protein concentration for analysis. The analyses were performed on antirabbit antiserum-coated microtitration strips by using 20 µL sample. Enterolactone concentrations were measured with the Victor 1420 multilabel counter (Wallac Oy, Turku, Finland). Each batch was analyzed with 3 duplicate quality-control serum samples; the interassay CVs and the concentrations of the samples were 16.8% (4 nmol/L), 10.1% (14 nmol/L), and 13.1% (59 nmol/L).
The rapid modification of the method described above tends to yield 15–30% higher results than does the standard TR-FIA method with extraction (14). To be able to compare our results with those of previous studies, we developed an equation that adjusts the results from the rapid modification of the TR-FIA method to those expected with the standard TR-FIA method. We developed the equation by analyzing 92 samples by both methods. The final results of the present study were calculated by using this equation:
RESULTS
The characteristics of the subjects are shown in Table 1. The subjects were on average slightly overweight; 31% of men and 20% of women were smokers and 40% reported that they were in good physical health.
View this table:
TABLE 1.. Characteristics of the study population
The distributions of serum enterolactone concentrations in men and women are shown in Figures 1 and 2. In men, the mean serum enterolactone concentration was 17.3 nmol/L, with a median concentration of 13.8 nmol/L and a range of 0–95.6 nmol/L. The corresponding values for women were 20.5, 16.6, and 0–182.6 nmol/L, respectively.
FIGURE 1. . Serum enterolactone concentrations in men (n = 1168). The line was produced by Gaussian Kernel density estimates on the basis of logarithmic values transformed to the original scale; kernel width is 0.4.
FIGURE 2. . Serum enterolactone concentrations in women (n = 1212). The line was produced by Gaussian Kernel density estimates on the basis of logarithmic values transformed to the original scale; kernel width is 0.4.
In men, multiple regression analysis showed strong positive and independent associations between serum enterolactone concentrations and consumption of both whole-grain products and fruit and berries (Table 2). Furthermore, male subjects with constipation had a 23% higher median serum enterolactone concentration than did subjects who did not report constipation (P < 0.05). A significant positive association between years of education and serum enterolactone concentration was also observed, but the associations did not remain significant after adjustment for other variables. Similarly, nonsmokers and former smokers had serum enterolactone concentrations that were >26% higher than those of smokers, but the association did not remain significant after adjustment for other variables. Vegetable consumption was not an independent predictor of serum enterolactone concentration, and no significant relations were found between serum enterolactone and subject, age, BMI, self-reported physical health, or alcohol consumption. The final model for men, which included constipation, consumption of whole-grain products, and consumption of fruit and berries, explained only 2.68% of the variation in serum enterolactone concentration.
View this table:
TABLE 2.. Serum enterolactone concentrations according to subject characteristics and tertiles of intakes of foods rich in enterolactone precursors with summary of regression analysis in men (n = 1168)
In women, multiple regression analysis showed that BMI and smoking were the most significant predictors of serum enterolactone concentration (Table 3). Subjects of normal weight had serum enterolactone concentrations that were >21% higher than those of underweight or obese women. Nonsmokers and former smokers had serum enterolactone concentrations that were >28% higher than those of smokers. Furthermore, constipation, age, and consumption of vegetables were also positively and independently associated with serum enterolactone concentration. Consumption of whole-grain products and of fruit and berries, self-reported physical health, and education were also associated with serum enterolactone; however, the associations did not remain significant after adjustment for other variables. No association was observed between alcohol consumption and serum enterolactone concentration. In the final model for women, 5 variables (BMI, smoking, age, constipation, and consumption of vegetables) explained almost 14% of the variation in serum enterolactone concentration.
View this table:
TABLE 3.. Serum enterolactone concentrations according to subject characteristics and tertiles of intakes of foods rich in enterolactone precursors with summary of regression analysis in women (n = 1212)
DISCUSSION
In the present study, we defined the range of serum enterolactone concentrations and examined diet- and health-related determinants of serum enterolactone in a large population sample. Serum enterolactone concentrations varied greatly. Consumption of lignan-containing foods and constipation were the most important predictors of enterolactone status in both men and women.
The range of serum enterolactone concentrations was wide, from 0 to 183 nmol/L. We were surprised to find that high concentrations were measured in some women, but the distributions of the concentrations were greatly skewed to the right in both sexes and the concentration was <5 nmol/L in a large percentage of the samples. At these low concentrations, analytic variation is larger than that at higher concentrations (the overall CV of the method is 17%) and thus the actual differences between these samples with low enterolactone concentrations are difficult to detect. In contrast, the between-person variation of enterolactone concentration is so large (CV of 83%) that the analytic variation is probably of minor importance. The reliability coefficient of a single measurement of enterolactone appears to be moderately high (0.55), suggesting that serum measurements of this compound could be a useful tool in epidemiologic studies (16).
Previously, circulating concentrations of enterolactone were measured in only a few small studies. In our study, the average serum enterolactone concentrations in women were slightly lower than concentrations measured in American women (16) and were markedly lower than those measured previously in omnivorous and lactovegetarian Finnish women (17) and postmenopausal Israeli women (18). The serum enterolactone concentrations in our male subjects were slightly lower than the concentrations in the male controls in a nested case-control study of coronary heart disease in Finland (6) but were somewhat higher than concentrations in young Canadian men (19), Asian men (17), and British and Portuguese men (20). Because of the small sample sizes of previous studies, further investigations in representative populations are needed so that reliable comparisons can be made.
In our study, older subjects had significantly higher serum enterolactone concentrations than did younger subjects and women had higher serum enterolactone concentrations than did men. High consumption of whole-grain products, vegetables, and fruit and berries by women and older subjects (data not shown) can, in part, explain these differences. Furthermore, the prevalence of constipation, which was positively associated with serum enterolactone, is higher in women than in men and is higher in older subjects than in younger subjects (21, 22). Subjects with constipation may have slower intestinal motility and, as a result, the metabolism and absorption of lignans may be more complete than that which occurs in subjects without constipation. Moreover, subjects with constipation may increase their consumption of foods rich in fiber to prevent the symptoms of constipation. However, in our study, the consumption of lignan-containing foods was the same in subjects who did and did not report constipation (data not shown). Other sex- and age-related factors may also modulate the metabolism of lignans.
BMI was an independent predictor of the serum enterolactone concentration in women, but not in men. Although self-reported consumption of lignan-containing foods was almost the same in normal-weight women as in their underweight and obese counterparts (data not shown), it is possible that obese individuals overestimated their consumption of these healthy foods, as was shown in a previous study (23). Thus, the lower serum enterolactone concentrations in overweight individuals than in their normal-weight counterparts may be partly explained by recall bias. Furthermore, enterolactone was shown to pass through the preadipocyte cell membrane (2, 24, 25), which means that in obese subjects enterolactone concentrations may be diluted by rapid transport into the cells, resulting in lower serum concentrations.
Our findings indicate that smokers had lower serum enterolactone concentrations than did nonsmokers. As far as we know, the association between smoking and phytoestrogens has not been studied before. However, serum estrogen concentrations during oral hormone administration appear to be lower in postmenopausal smokers than in nonsmokers, whereas no differences are observed in women not receiving oral hormones (26, 27). The fact that smoking alters hormone concentrations in postmenopausal women receiving oral hormones, but not in women receiving percutaneous hormone administration, implies that the effects of smoking on sex hormones occur during absorption or during liver metabolism of hormones. The structural similarity of enterolactone to mammalian estrogens may enable smoking to also affect the metabolism of enterolactone. Furthermore, cigarette smoke increases oxidative stress (28). Like all phenolic compounds, enterolactone is likely to have antioxidant properties (4) and thus it might be used by the body to eliminate free radicals, leading to lower serum enterolactone concentrations in smokers than in nonsmokers. In addition, eating habits of smokers were shown to be less healthy than those of nonsmokers (29–31); in our study, smokers consumed less whole-grain products, vegetables, and fruit and berries than did nonsmokers (data not shown).
A relatively strong association existed between the serum enterolactone concentration and consumption of lignan-containing foods, even though the FFQ we used did not focus on assessing the intake of lignans. Furthermore, the FFQ assessed diet during the previous 12 mo, whereas enterolactone probably has a relatively short half-life in the body (32). This strongly supports the importance of regular consumption of lignan-containing foods to maintain high serum enterolactone concentrations. In previous studies, urinary lignan excretion was correlated with intakes of total fiber and fiber from berries, fruit, grains, legumes, and vegetables (1, 33) and with consumption of fruit (33) and pulses and beans (34). Consumption of vegetables was associated with urinary lignans in Japanese men and women (34) but not in American adults (33). However, a controlled dietary intervention study indicated that increasing the intake of vegetables and fruit increases enterolactone excretion in a dose-dependant manner (35).
It was shown previously that enterolactone excretion differs among groups consuming different diets, such as macrobiotic, vegetarian, and omnivorous diets (1, 36). Our data indicate that even among individuals consuming a Western diet, serum enterolactone can reflect differences in dietary patterns. However, the between-person variability in serum enterolactone concentration in our study was high and the factors we studied, including diet, did not explain all of this variability. This suggests that interindividual differences in the intestinal environment may play an important role in determining the serum enterolactone concentration. These differences in intestinal environment may include the composition of gut microflora, the ability to metabolize lignans, and the rate of lignan absorption. Previously, the importance of microflora in the metabolism of lignans was illustrated by the observation that use of antibiotics almost completely eliminates the formation of enterolactone from plant precursors in the gut (11, 36). None of the subjects investigated in our study had taken antibiotics during the past 3 mo. However, we do not know how long it takes to normalize gut microflora and enable metabolism of lignans after antibiotic administration. Other factors that may affect metabolism of lignans have not yet been examined in any detail.
In conclusion, an exceptionally wide range of serum enterolactone concentrations was found. Constipation and consumption of whole-grain products and fruit and berries appear to be the most important determinants of serum enterolactone in men. In women, the main determinants appear to be BMI, smoking, age, constipation, and consumption of vegetables. However, these variables explain only a small part of the variation in serum enterolactone concentration. Our results suggest that the role of the gut environment in the metabolism of lignans may be very important and should be investigated in future studies.
REFERENCES