Abstract
Results from animal studies have consistently suggested that lignans play a role in the regulation of in body weight, but evidence from human studies has been limited. We examined the associations between urinary excretion of enterolactone and enterodiol, the major intestinal microbial metabolites of dietary lignans, and 10-year prospective weight change using data from 2 well-characterized cohort studies of US women: the Nurses' Health Study (2000–2010) and Nurses' Health Study II (1997–2007). Urinary excretion levels of enterolactone and enterodiol were measured at baseline. Associations with prospective weight change were analyzed using a multivariable-adjusted linear mixed-effects model. We observed that women in the highest quartile of urinary excretion of total lignans had significantly lower baseline body mass indices (weight in kilograms divided by height in meters squared) (mean, 24.6, 95% confidence interval (CI): 23.9, 25.2) than did those in the lowest quartile (mean, 27.7, 95% CI: 27.0, 28.4; P for trend < 0.01). Compared with women in the lowest quartile of enterodiol excretion, those in the highest quartile gained 0.27 kg/year less weight (95% CI: 0.12, 0.41; P for trend < 0.01) during the 10-year follow-up. The association was borderline significant for enterolactone (for the fourth vs. first quartile, least square mean of weight change rate = −0.14 kg/year, 95% CI: −0.29, 0.00). Our data suggest that higher urinary excretion of lignan metabolites, especially enterodiol, is associated with modestly slower weight gain.
Keywords: enterodiol, enterolactone, lignan, weight change
Excessive and rapid weight gain is an important risk factor for various chronic diseases, including obesity, cardiovascular disease, type 2 diabetes mellitus, and certain cancers (1–4). Nationally representative survey data have shown that the prevalence of overweight and obesity has steadily increased over the past 3 decades, and more than 35% of US men and women were obese in 2009–2010 (5). Searching for effective dietary and lifestyle preventive approaches to mitigating excessive weight gain has become essential.
Higher consumption of certain plant-based foods, such as fruits, vegetables, whole grains, and nuts, predicts less weight gain (6), although the active ingredients in these foods that could account for the putative beneficial effects are largely unknown. Lignans are a group of bioactive, nonnutrient, noncaloric, phenolic plant compounds (7). Flax and sesame seeds are rich sources of lignans, although lignans are also found in lower concentrations in whole grains, nuts, fruits, vegetables, coffee, tea, and wine. Lariciresinol, pinoresinol, secoisolariciresinol, and matairesinol are the major lignan compounds in the human diet and are metabolized to enterolactone and enterodiol by gut microbiota in the colon (8). As phytoestrogens, lignan metabolites might be able to regulate adipogenesis through modulation of the estrogen receptor (ER) signaling pathway (9) or by inducing adiponectin mRNA expression (10). Although investigations using animal models have consistently suggested a possible role of lignans in the regulation of body weight (10, 11), evidence from human studies based either on biomarkers or lignan intakes is limited (12–14), probably because of the lack of reliable databases on lignans in food until recently.
To fill this gap in knowledge, we evaluated urinary excretion levels of lignan metabolites as objective biomarkers of lignan intake in relation to prospective weight change using data from 2 well-characterized cohort studies of US women: the Nurses' Health Study (NHS) and Nurses' Health Study II (NHSII). In addition, using existing data, we also explored the relations among several endocrine-disrupting chemicals (EDCs), including bisphenol A (BPA) and phthalate metabolites, lignans, and weight change in the present analysis. These chemicals were associated with faster weight gain (15), and lignans might interact with the EDCs by exerting counteracting mechanisms on weight change.
METHODS
Study population
The NHS was a cohort study that included 121,700 female registered nurses aged 30–55 years who were enrolled in 1976. The NHSII was a parallel cohort study established in 1989, and it consisted of 116,671 female registered nurses who were 25–42 years of age. The participants in both cohorts responded to a baseline questionnaire about their body weight, height, lifestyle practices, and medical histories. Participants were followed biennially using similar but expanded questionnaires to update the information.
Upon request, a total of 18,717 NHS participants aged 53–79 years and 29,611 NHSII participants aged 32–52 years provided blood and spot urine samples in 2000–2002 and 1996–2001, respectively. Urine samples were collected without preservative in a polypropylene container and returned to a central biorepository via overnight courier with an icepack. They were immediately processed upon arrival and aliquoted into polypropylene cryovials, which were then stored in the vapor phase of liquid nitrogen freezers at −130°C or less.
The present investigation was based on a prospective nested case-control study of type 2 diabetes mellitus in the 2 cohorts. Briefly, among participants who provided urine samples and were free of self-reported diabetes, cardiovascular disease, and cancer at sample collection, we prospectively documented and confirmed 1,111 cases with type 2 diabetes (in the NHS, n = 456; in the NHSII, n = 655) during a follow-up period of 2000–2010 in the NHS and 1997–2007 in the NHSII. One healthy control was randomly selected from the nondiabetic participants when a case was diagnosed. Cases and controls were matched for age at urine sample collection, month of sample collection, fasting status (≥8 hours or not), first morning urine (yes or no), race (white or other races), menopausal status, and hormone replacement therapy use (NHSII only). To avoid confounding by the association of diabetes with weight change, we conducted our analysis among the 1,111 control subjects from these studies. The study protocol was approved by the institutional review board of the Brigham and Women's Hospital and the Human Subjects Committee Review Board of the Harvard School of Public Health.
Assessment of body weight and covariates
Body weight was self-reported at baseline and updated every 2 years during follow-up. A validation study showed a correlation coefficient of 0.96 for the association between self-reported weight and measured weight among 184 NHS participants (16). The primary outcome of the present study was body weight assessed from the time of urine sample collection to the most recent follow-up cycle (2010 in the NHS; 2009 in the NHSII). The follow-up time in each study was 10 years. The secondary outcome was the baseline body mass index (BMI), which was calculated as weight in kilograms divided by height in meters squared.
Information on covariates, including age, smoking status, menopausal status, alcohol consumption, and physical activity level, was primarily based on information from the follow-up questionnaires that were administered in 2000 in the NHS or 1999 in the NHSII, when most of the samples were collected. Diet was assessed using a validated food frequency questionnaire that was administered every 4 years beginning in 1990 in the NHS or 1991 in the NHSII (17). Diet quality was assessed using the Alternative Healthy Eating Index, which measures adherence to a healthy diet pattern based on 11 dietary factors that include intakes of vegetables, fruits, whole grains, sugar-sweetened beverages and fruit juices, nuts and legumes, red/processed meat, trans fat, long-chain (n-3) fats, polyunsaturated fatty acid, sodium, and alcohol (18). We used cumulative average estimates of dietary covariates to reflect long-term dietary habits and reduce measurement errors (19).
Laboratory measurements
Urinary excretion of enterolactone and enterodiol were measured using electrospray ionization orbitrap liquid chromatography mass spectrometry (20). Urinary creatinine excretion was measured with a Roche-Cobas MiraPlus clinical chemistry autoanalyzer (Roche Diagnostics, Basel, Switzerland). A total of 224 blinded, split samples from 112 NHS and NHSII participants were interspersed in the assay plates. The coefficient of variation of the assays was calculated based on these quality-control samples. The average intra-assay coefficients of variation were 14.4% for enterodiol, 7.4% for enterolactone, and 5.4% for creatinine. Somehow, the average interassay coefficients of variation were 11.1% for enterodiol, 6.4% for enterolactone, and 3.5% for creatinine. The urinary excretions (nanomoles per liter) of enterolactone and enterodiol were divided by creatinine concentrations (grams per liter) to derive creatinine-adjusted values (nanomoles per gram of creatinine). A pilot study showed reasonable intraclass correlation coefficients (0.45 for enterodiol and 0.53 for enterolactone) between levels in 2 urine samples collected 1–2 years apart from 58 NHSII participants.
Statistical analyses
The baseline characteristics of the study population were presented according to the quartiles of enterolactone and enterodiol. Values for continuous variables with normal distribution were expressed as means and standard deviations, and those for variables with a skewed distribution were shown as medians and interquartile ranges. Values for categorical variables were presented as percentages. Generalized linear regression was used to assess the relationship between urinary excretion of lignan metabolites and BMI at baseline. Model 1 was the unadjusted crude model. Model 2 was adjusted for cohort origin (NHS or NHSII), age at baseline (years), cigarette smoking status (never, past, or current smoker), menopausal status (premenopausal, postmenopausal, or uncertain), alcohol consumption (grams per day, log-transformed), physical activity level (metabolic equivalent hours per week, log-transformed), Alternative Healthy Eating Index score, and total energy intake (kilocalories per day). We used the Akaike information criterion to perform model selection. After age at blood drawn, weight at baseline, and smoking status were forced in the model, inclusion of other covariates did not lead to significant changes in the Akaike information criterion statistic. P values for linear trend were obtained by modeling the median concentration of each quartile as a continuous variable in the regression models. To model the prospective rate of annual weight change by quartile of urinary lignan metabolites excretion, we used linear mixed-effects models with an unstructured correlation matrix. A product term between lignan metabolites and years since baseline was included in the model to represent the annual weight change rate. P values for linear trend were obtained by examining an interaction term between follow-up time and median concentration of the quartile in the mixed-effects models. We also repeated the analysis using BMI as the outcome to evaluate the association between baseline urinary excretion of lignans and the change in BMI.
A secondary analysis was conducted to evaluate the potential interaction between lignan metabolites and urinary excretion of several EDCs (BPA, phthalic acid, butyl phthalates, and monobenzyl phthalate (MBzP), in nanograms per milliliter) in relation to prospective weight change. These EDCs were associated with faster weight gain in a previous investigation in the same study participants (15). Enterolactone and enterodiol excretion were dichotomized according to the median value, and the associations of BPA, phthalic acid, and MBzP with prospective weight gain were examined using a linear mixed-effect model in the 2 strata of the lignan metabolites separately. P values for (3-way) interactions were calculated by including a product term of binary lignan metabolites excretion, continuous concentrations of the EDCs, and time since baseline.
We further conducted 2 sensitivity analyses to assess the robustness of our findings. We stratified the data by age and BMI at baseline separately to investigate whether the observed associations were consistent between young and older women (<53 years vs. ≥53 years) or between lean and heavier women (BMI <25 vs. BMI ≥25). Of note, using the cutoff of 53 years for age essentially separated the 2 cohorts. All P values were 2-sided. Data were analyzed using SAS software, version 9.3 (SAS Institute, Inc., Cary, North Carolina).
RESULTS
Tables 1 and 2 show the baseline characteristics of the participants according to quartiles of urinary excretion of enterolactone and enterodiol. On average, women with higher urinary excretion of enterolignan tended to be slightly older, were less likely to be current smokers, and were more likely to be postmenopausal. They also consumed more alcohol, had higher physical activity levels, and had better diet quality. Higher enterolactone and enterodiol excretion generally suggested higher consumption of coffee, red wine, whole grains, fruits, and vegetables.
Table 1.
Baseline Characteristics of Participants by Quartile of Urinary Enterolactone Excretion in the Nurses' Health Study (2000–2010) and Nurses' Health Study II (1997–2007)
| Characteristic | Quartile of Urinary Enterolactone Excretion |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 (n = 277) |
2 (n = 278) |
3 (n = 278) |
4 (n = 278) |
|||||||||
| Mean (SD) | % | Median (IQR) | Mean (SD) | % | Median (IQR) | Mean (SD) | % | Median (IQR) | Mean (SD) | % | Median (IQR) | |
| Age, years | 53.2 (11.2) | 52.8 (11.1) | 54.1 (11.1) | 54.8 (11.5) | ||||||||
| NHS participant | 39.7 | 37.8 | 43.5 | 43.2 | ||||||||
| White race | 97.8 | 95.3 | 98.2 | 98.2 | ||||||||
| Smoking status | ||||||||||||
| Nonsmoker | 60.3 | 59.4 | 56.8 | 56.5 | ||||||||
| Past smoker | 31.4 | 33.5 | 36.7 | 35.6 | ||||||||
| Current smoker | 8.3 | 6.8 | 6.5 | 7.6 | ||||||||
| Postmenopausal | 52.7 | 48.2 | 52.2 | 51.8 | ||||||||
| Alcohol consumption, g/day | 0.78 (0–3.65) | 1.35 (0–3.75) | 1.90 (0.16–6.64) | 2.61 (0.45–7.15) | ||||||||
| AHEI score | 48.1 (40.9–54.0) | 49.9 (43.6–56.0) | 50.4 (44.2–57.1) | 54.1 (46.7–60.7) | ||||||||
| Physical exercise, MET hours/week | 9.6 (2.9–18.6) | 11.9 (4.3–23.2) | 13.9 (6.4–26.0) | 14.9 (6.8–29.7) | ||||||||
| Coffee, servings/day | 1.57 (1.6) | 1.72 (1.56) | 2.14 (1.63) | 1.96 (1.47) | ||||||||
| Red wine, servings/day | 0.04 (0.12) | 0.05 (0.13) | 0.07 (0.16) | 0.09 (0.26) | ||||||||
| Whole grains, servings/day | 1.17 (0.94) | 1.34 (0.99) | 1.38 (0.99) | 1.47 (0.99) | ||||||||
| Vegetable, servings/day | 3.11 (1.78) | 3.31 (1.67) | 3.33 (1.85) | 3.66 (1.90) | ||||||||
| Fruit, servings/day | 1.20 (0.84) | 1.34 (0.84) | 1.28 (0.77) | 1.60 (0.99) | ||||||||
Abbreviations: AHEI, Alternative Healthy Eating Index; IQR, interquartile range; MET, metabolic equivalent; NHS, Nurses' Health Study; SD, standard deviation.
Table 2.
Baseline Characteristics of Participants by Quartile of Urinary Enterodiol Excretion in the Nurses' Health Study (2000–2010) and Nurses' Health Study II (1997–2007)
| Characteristic | Quartile of Urinary Enterodiol Excretion |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 (n = 277) |
2 (n = 278) |
3 (n = 278) |
4 (n =278) |
|||||||||
| Mean (SD) | % | Median (IQR) | Mean (SD) | % | Median (IQR) | Mean (SD) | % | Median (IQR) | Mean (SD) | % | Median (IQR) | |
| Age, years | 51.0 (10.4) | 52.2 (11.3) | 55.2 (11.3) | 56.5 (11.0) | ||||||||
| NHS participant | 30.7 | 34.9 | 46.4 | 52.2 | ||||||||
| White race | 97.8 | 96.8 | 98.6 | 96.4 | ||||||||
| Smoking status | ||||||||||||
| Nonsmoker | 63.5 | 58.6 | 58.3 | 52.5 | ||||||||
| Past smoker | 27.8 | 34.2 | 34.5 | 40.7 | ||||||||
| Current smoker | 8.3 | 7.2 | 7.2 | 6.5 | ||||||||
| Postmenopausal | 41.2 | 43.5 | 57.2 | 63.0 | ||||||||
| Alcohol consumption, g/day | 0.9 (0–3.05) | 1.4 (0–4.78) | 1.8 (0.15–5.91) | 2.14 (0.38–7.70) | ||||||||
| AHEI score | 48.0 (40.3–53.5) | 50.6 (43.1–56.6) | 50.5 (45.2–57.6) | 52.9 (46.4–60.9) | ||||||||
| Physical exercise, MET hours/week | 10.9 (3.6–20.5) | 12.6 (5.1–25.9) | 12.7 (4.6–23.5) | 14.9 (6.8–26.2) | ||||||||
| Coffee, servings/day | 1.67 (1.61) | 1.89 (1.62) | 1.90 (1.57) | 1.93 (1.52) | ||||||||
| Red wine, servings/day | 0.04 (0.12) | 0.05 (0.13) | 0.07 (0.19) | 0.09 (0.24) | ||||||||
| Whole grains, servings/day | 1.18 (0.87) | 1.34 (0.98) | 1.36 (0.97) | 1.48 (1.09) | ||||||||
| Vegetable, servings/day | 3.10 (1.76) | 3.13 (1.55) | 3.42 (1.75) | 3.75 (2.09) | ||||||||
| Fruit, servings/day | 1.14 (0.75) | 1.31 (0.84) | 1.47 (0.97) | 1.50 (0.88) | ||||||||
Abbreviations: AHEI, Alternative Healthy Eating Index; IQR, interquartile range; MET, metabolic equivalent; NHS, Nurses' Health Study; SD, standard deviation.
Associations between urinary excretion of lignan metabolites and baseline BMI are shown in Table 3. For both enterolignans, women with the highest urinary excretion levels had the lowest baseline BMI. In the multivariable-adjusted models, participants in the highest quartile of enterolactone excretion had an average BMI of 24.6 (95% confidence interval (CI): 23.9, 25.3), whereas women in the lowest quartile had an average BMI of 27.5 (95% CI: 26.8, 28.2; P for trend < 0.01). The corresponding figures were 25.5 (95% CI: 24.8, 26.2) and 26.6 (95% CI: 25.9, 27.3; P for trend =0.03) for enterodiol. Women in the highest quartile of total excretion of urinary lignan metabolites also had the lowest baseline BMI.
Table 3.
Associations Between Urinary Excretion of Lignan Metabolites and Baseline Body Mass Index in the Nurses' Health Study (2000–2010) and Nurses' Health Study II (1997–2007)
| Lignan and Quartile | Median Creatinine Concentration, nmol/g creatinine | Model 1a |
Model 2b |
||
|---|---|---|---|---|---|
| BMIc | 95% CI | BMIc | 95% CI | ||
| Enterolactone | |||||
| 1 | 234.8 | 27.7 | 27.1, 28.3 | 27.5 | 26.8, 28.2 |
| 2 | 1,229.8 | 26.3 | 25.7, 26.9 | 26.3 | 25.6, 27.0 |
| 3 | 2,679.3 | 25.7 | 25.1, 26.3 | 25.9 | 25.3, 26.6 |
| 4 | 5,649.5 | 24.4 | 23.8, 25.0 | 24.6 | 23.9, 25.3 |
| P for trend | <0.01 | <0.01 | |||
| Enterodiol | |||||
| 1 | 16.4 | 26.7 | 26.1, 27.3 | 26.6 | 25.9, 27.3 |
| 2 | 56.0 | 26.0 | 25.4, 26.6 | 26.0 | 25.3, 26.7 |
| 3 | 114.6 | 26.0 | 25.4, 26.6 | 26.0 | 25.4, 26.7 |
| 4 | 342.8 | 25.4 | 24.8, 26.0 | 25.5 | 24.8, 26.2 |
| P for trend | 0.01 | 0.03 | |||
| Total lignans | |||||
| 1 | 312.4 | 27.8 | 27.2, 28.4 | 27.7 | 27.0, 28.4 |
| 2 | 1,361.0 | 26.3 | 25.7, 26.8 | 26.3 | 25.6, 27.0 |
| 3 | 2,831.6 | 25.6 | 25.0, 26.2 | 25.8 | 25.1, 26.5 |
| 4 | 5,872.1 | 24.4 | 23.8, 25.0 | 24.6 | 23.9, 25.2 |
| P for trend | <0.01 | <0.01 | |||
Abbreviations: BMI, body mass index; CI, confidence interval.
a Unadjusted.
b Adjusted for age, cohort origin, smoking status, physical activity level, menopausal status, alcohol consumption, Alternative Healthy Eating Index score, and total energy intake.
c Least square means of baseline BMI (weight in kilograms divided by height in meters squared).
Table 4 presents the results of prospective annual weight change rate by quartiles of lignan metabolite excretion. Urinary enterodiol excretion levels in the highest quartiles of were associated with a significantly slower annual rate of weight gain. Women in the highest quartile of enterodiol (median, 342.8 nmol/g creatinine) gained 0.27 kg/year less (95% CI: −0.41, −0.12; P for trend = 0.003) than did those in the lowest quartile (median, 16.4 nmol/g creatinine) during 10-year follow-up. The associations were borderline significant for enterolactone excretion (least square mean of weight change rate = −0.14 kg/year, 95% CI: −0.29, 0.00; P for trend = 0.12) and total lignans excretion (least square mean of weight change rate = −0.16 kg/year, 95% CI: −0.30, −0.02; P for trend = 0.09). In a sensitivity analysis in which we restricted the analysis to women who provided first morning urine samples, we observed slightly stronger associations. For example, in the multivariable model, the annual weight change rate among women in the highest quartile of total lignans was 0.20 kg less (95% CI: −0.36, −0.04; P for trend = 0.05) than that among women in the lowest quartile. The results for the association between baseline urinary lignan excretion and prospective BMI change were consistent with the results for weight change (Web Table 1, available at http://aje.oxfordjournals.org/). Women in the highest quartile of the total urinary excretion of lignans had a BMI increase that was 0.06 per year less (95% CI: −0.11, −0.01; P for trend 0.06) than that in women in the lowest quartile.
Table 4.
Prospective Annual Weight Change Rate by Quartile of Lignan Metabolite Excretion in the Nurses' Health Study (2000–2010) and Nurses' Health Study II (1997–2007)
| Lignan and Quartile | Median Creatinine Concentration, nmol/g creatinine | Model 1a |
Model 2b |
||
|---|---|---|---|---|---|
| Weight Change Ratec | 95% CI | Weight Change Ratec | 95% CI | ||
| Enterolactone | |||||
| 1 | 234.8 | 0.00 | Referent | 0.00 | Referent |
| 2 | 1,229.8 | −0.11 | −0.25, 0.03 | −0.13 | −0.27, 0.02 |
| 3 | 2,679.3 | −0.12 | −0.27, 0.02 | −0.14 | −0.28, 0.00 |
| 4 | 5,649.5 | −0.14 | −0.29, 0.00 | −0.14 | −0.29, 0.00 |
| P for trend | 0.09 | 0.12 | |||
| Enterodiol | |||||
| 1 | 16.4 | 0.00 | Referent | 0.00 | Referent |
| 2 | 56.0 | −0.16 | −0.30, −0.01 | −0.17 | −0.31, −0.02 |
| 3 | 114.6 | −0.17 | −0.32, −0.03 | −0.17 | −0.32, −0.03 |
| 4 | 342.8 | −0.26 | −0.40, −0.12 | −0.27 | −0.41, −0.12 |
| P for trend | <0.01 | <0.01 | |||
| Total lignans | |||||
| 1 | 312.4 | 0.00 | Referent | 0.00 | Referent |
| 2 | 1,361.0 | −0.13 | −0.27, 0.01 | −0.14 | −0.29, 0.00 |
| 3 | 2,831.6 | −0.11 | −0.25, 0.03 | −0.12 | −0.27, 0.02 |
| 4 | 5,872.1 | −0.16 | −0.31, −0.02 | −0.16 | −0.30, −0.02 |
| P for trend | 0.06 | 0.09 | |||
Abbreviation: CI, confidence interval.
a Unadjusted.
b Adjusted for age, cohort origin, smoking status, physical activity level, menopausal status, alcohol consumption, Alternative Healthy Eating Index score, total energy intake, and baseline weight.
c Least square means of weight change rate (in kilograms per year).
In the secondary analysis, the positive association between EDCs and prospective weight gain was more apparent among women with lower level of lignan metabolite excretion, especially for BPA and phthalic acid, although none of the interaction tests reached significance (Figure 1). The P values for interaction were 0.28, 0.25, 0.90, and 0.41 for BPA, phthalic acid, butyl phthalates, and MBzP, respectively. Among women with lower excretion of total lignans, having a BPA level in the highest quartile was significantly associated with a 0.31 kg/year (95% CI: 0.07, 0.56; P for trend = 0.03) higher weight gain compared with the lowest quartile, whereas the association was greatly attenuated among women with high excretion of total lignans (0.13 kg/year, 95% CI: −0.08, 0.34; P for trend = 0.35). A similar pattern of associations was observed for phthalic acid and MBzP but not for butyl phthalates.
Figure 1.
Prospective annual weight change rate by quartile of endocrine-disrupting chemical according to total lignan metabolite excretion in the Nurses' Health Study (2000–2010) and Nurses' Health Study II (1997–2007). A) Bisphenol A (P for interaction = 0.28); B) phthalic acid (P for interaction = 0.25); C) butyl phthalates (P for interaction = 0.90); D) monobenzyl phthalate (P for interaction = 0.41). The results were adjusted for age, cohort origin, smoking status, physical activity level, menopausal status, alcohol consumption, Alternative Healthy Eating Index score, total energy intake, and baseline weight.
In the sensitivity analyses that were stratified by age or BMI, the linear trends were less apparent within certain strata, although the highest quartiles of enterolignan excretion were generally associated with less weight gain in these underpowered analyses (Web Tables 2 and 3).
DISCUSSION
In the present analysis, we found that a higher baseline excretion of urinary enterolignan was significantly associated with a lower baseline BMI and a modestly lower weight gain during a 10-year follow-up period. These associations were independent of established and potential predictors of weight change.
Evidence from several cross-sectional studies showed that higher lignan consumption was associated with lower overall fat mass (12, 13, 21), as well as with less abdominal fat mass (14). These observations were in line with evidence generated from rodent models, in which supplementation of flaxseed lignans exerted beneficial effects on reducing body weight and fat accumulation, improving lipid profile, and lowering blood pressure (10, 11). In the present study, we consistently observed a significantly lower baseline BMI among women with higher urinary enterolignan excretion. Moreover, our results showed that women with higher lignan excretion had less weight gain after we controlled for baseline body weight. Overall, these findings support the notion that increased lignan consumption might potentially lead to less weight gain.
Although the exact underlying biological mechanisms are still not clear, several lines of evidence from basic science research suggest that the estrogenic effects of the enterolignans might contribute to the regulation of body weight. Lignans are structurally similar to 17β-estradiol and can bind to ERs and modulate downstream signaling (22, 23). It has been demonstrated in animal studies that the effects of 17β-estradiol on body weight regulation are predominately through binding to ER α (24). Estrogens and ERs are involved in the regulation of energy metabolism and utilization by directly or indirectly stimulating the expression/activity of the key enzymes/proteins involved in glycolysis (25–29). Moreover, it has been suggested in animal studies that there is a signaling cross-talk between ERs and peroxisome proliferator-activated receptor γ in the regulation of estrogen-responsive genes and peroxisome proliferator–responsive genes (30). Lignans preferably bind to ER α, which leads to subsequent ER-mediated gene expression both in vivo and in vitro (31). Enterolignans are also capable of inducing adiponectin mRNA expression, leading to higher levels of plasma adiponectin, which is known to promote β-oxidation in the skeletal muscle (32).
In our previous study, we observed that higher urinary excretions of BPA and certain phthalate metabolites (phthalic acid, MBzP, and butyl phthalates) were significantly associated with modestly greater weight gain (15). In the present study, the aforementioned positive associations were somewhat attenuated among women with higher urinary enterolignan excretion, suggesting a potential antagonistic process between enterolignans and EDCs. BPA is reported to preferably bind to ER β (33). It is likely that enterolignans and BPA might counteract each other through binding to the 2 different ERs, although more evidence is needed to substantiate this notion. Nevertheless, a similar antagonistic interaction between BPA and isoflavones, another group of polyphenols with estrogenic effects, on genome methylation was observed in animal experiments (34). In addition, it has been demonstrated that the expression of either ER α or β lowers both basal and stimulated peroxisome proliferator-activated receptor γ–mediated receptor activity (35). Many EDCs, including phthalate monoesters, have been identified as peroxisome proliferator-activated receptor γ agonists and thus are capable of promoting adipocyte maturation (36–38). Higher expression levels of ERs induced by enterolignans might be able to attenuate this obesity development process.
The strengths of the current study include the prospective study design, the relatively long follow-up period, the use of objective markers to assess lignan exposure, and the adjustment for a wide range of potential confounders. There are also several limitations to discuss. First, because the half-lives of enterolignans are relatively short, a single measurement might not be able to capture long-term levels. Multiple assessments during the follow-up period are preferable but were not feasible in the present study. Second, the sample size was suboptimal to evaluate the interaction between enterolignans and EDCs in relation to weight gain. Third, the generalizability might be limited to women of European ancestry. Future studies in men and people of other ethnicities are warranted. Finally, despite the adjustment for a wide array of covariates, we cannot rule out the role of residual confounding in the associations of interest.
In conclusion, our data suggest that higher urinary enterolignan excretion is associated with modestly less weight gain. Further studies, preferably with larger sample sizes and multiple measurements of lignan metabolites, are needed to confirm our findings.
Supplementary Material
ACKNOWLEDGMENTS
Author affiliations: Department of Nutrition, Harvard School of Public Health, Boston, Massachusetts (Yang Hu, Frank B. Hu, Qi Sun); Department of Epidemiology, Fielding School of Public Health, University of California, Los Angeles, Los Angeles, California (Yan Song); Department of Food Science and Human Nutrition, College of Tropical Agriculture and Human Resources, University of Hawaii at Manoa, Honolulu, Hawaii (Adrian A. Franke); Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts (Frank B. Hu); Channing Division of Network Medicine, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts (Frank B. Hu, Qi Sun); and Saw Swee Hock School of Public Health and Yong Loo Lin School of Medicine, National University of Singapore, National University Health System, Singapore, Singapore (Rob M. van Dam).
This study was supported by the National Institutes of Health (grants U54CA155626, P30 DK46200, CA87969, CA49449, DK58845, DK58785, DK082486, CA50385, CA67262, and CA71789). Q.S. is supported by career development award R00HL098459 from the National Heart, Lung, and Blood Institute. Y.S. was a recipient of Burroughs Wellcome Fund Inter-school Training Program in Metabolic Diseases at the University of California, Los Angeles.
We thank Dr. Xiangnan Li for her skillful performance of the liquid chromatography–mass spectrometry assays.
The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Conflict of interest: none declared.
REFERENCES
- 1.Stettler N, Kumanyika SK, Katz SH, et al. Rapid weight gain during infancy and obesity in young adulthood in a cohort of African Americans. Am J Clin Nutr. 2003;776:1374–1378. [DOI] [PubMed] [Google Scholar]
- 2.Huang Z, Hankinson SE, Colditz GA, et al. Dual effects of weight and weight gain on breast cancer risk. JAMA. 1997;27817:1407–1411. [PubMed] [Google Scholar]
- 3.Chan JM, Rimm EB, Colditz GA, et al. Obesity, fat distribution, and weight gain as risk factors for clinical diabetes in men. Diabetes Care. 1994;179:961–969. [DOI] [PubMed] [Google Scholar]
- 4.Willett WC, Manson JE, Stampfer MJ, et al. Weight, weight change, and coronary heart disease in women. Risk within the ‘normal’ weight range. JAMA. 1995;2736:461–465. [DOI] [PubMed] [Google Scholar]
- 5.Ogden C, Carroll M, Kit B, et al. Prevalence of Obesity in the United States, 2009–2010. Hyattsville, MD: National Center for Health Statistics; 2012. [Google Scholar]
- 6.Mozaffarian D, Hao T, Rimm EB, et al. Changes in diet and lifestyle and long-term weight gain in women and men. N Engl J Med. 2011;36425:2392–2404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Peterson J, Dwyer J, Adlercreutz H, et al. Dietary lignans: physiology and potential for cardiovascular disease risk reduction. Nutr Rev. 2010;6810:571–603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Landete JM. Plant and mammalian lignans: a review of source, intake, metabolism, intestinal bacteria and health. Food Res Int. 2012;461:410–424. [Google Scholar]
- 9.Chen JQ, Brown TR, Russo J. Regulation of energy metabolism pathways by estrogens and estrogenic chemicals and potential implications in obesity associated with increased exposure to endocrine disruptors. Biochim Biophys Acta. 2009;17937:1128–1143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fukumitsu S, Aida K, Ueno N, et al. Flaxseed lignan attenuates high-fat diet-induced fat accumulation and induces adiponectin expression in mice. Br J Nutr. 2008;1003:669–676. [DOI] [PubMed] [Google Scholar]
- 11.Park JB, Velasquez MT. Potential effects of lignan-enriched flaxseed powder on bodyweight, visceral fat, lipid profile, and blood pressure in rats. Fitoterapia. 2012;835:941–946. [DOI] [PubMed] [Google Scholar]
- 12.Kilkkinen A, Stumpf K, Pietinen P, et al. Determinants of serum enterolactone concentration. Am J Clin Nutr. 2001;736:1094–1100. [DOI] [PubMed] [Google Scholar]
- 13.Morisset A-S, Lemieux S, Veilleux A, et al. Impact of a lignan-rich diet on adiposity and insulin sensitivity in post-menopausal women. Br J Nutr. 2009;1022:195–200. [DOI] [PubMed] [Google Scholar]
- 14.De Kleijn MJ, van der Schouw YT, Wilson PW, et al. Dietary intake of phytoestrogens is associated with a favorable metabolic cardiovascular risk profile in postmenopausal U.S. women: the Framingham Study. J Nutr. 2002;1322:276–282. [DOI] [PubMed] [Google Scholar]
- 15.Song Y, Hauser R, Hu FB, et al. Urinary concentrations of bisphenol A and phthalate metabolites and weight change: a prospective investigation in US women. Int J Obes (Lond). 2014;3812:1532–1537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Willett W, Stampfer MJ, Bain C, et al. Cigarette smoking, relative weight, and menopause. Am J Epidemiol. 1983;1176:651–658. [DOI] [PubMed] [Google Scholar]
- 17.Salvini S, Hunter DJ, Sampson L, et al. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol. 1989;184:858–867. [DOI] [PubMed] [Google Scholar]
- 18.Chiuve SE, Fung TT, Rimm EB, et al. Alternative dietary indices both strongly predict risk of chronic disease. J Nutr. 2012;1426:1009–1018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hu FB, Stampfer MJ, Rimm E, et al. Dietary fat and coronary heart disease: a comparison of approaches for adjusting for total energy intake and modeling repeated dietary measurements. Am J Epidemiol. 1999;1496:531–540. [DOI] [PubMed] [Google Scholar]
- 20.Franke AA, Custer LJ, Wilkens LR, et al. Liquid chromatographic-photodiode array mass spectrometric analysis of dietary phytoestrogens from human urine and blood. J Chromatogr B Analyt Technol Biomed Life Sci. 2002;777(1-2):45–59. [DOI] [PubMed] [Google Scholar]
- 21.Milder IE, Feskens EJ, Arts IC, et al. Intake of the plant lignans secoisolariciresinol, matairesinol, lariciresinol, and pinoresinol in Dutch men and women. J Nutr. 2005;1355:1202–1207. [DOI] [PubMed] [Google Scholar]
- 22.Low Y-L, Taylor JI, Grace PB, et al. Phytoestrogen exposure correlation with plasma estradiol in postmenopausal women in European Prospective Investigation of Cancer and Nutrition-Norfolk may involve diet-gene interactions. Cancer Epidemiol Biomarkers Prev. 2005;141:213–220. [PubMed] [Google Scholar]
- 23.Touillaud MS, Pillow PC, Jakovljevic J, et al. Effect of dietary intake of phytoestrogens on estrogen receptor status in premenopausal women with breast cancer. Nutr Cancer. 2005;512:162–169. [DOI] [PubMed] [Google Scholar]
- 24.Hart-Unger S, Korach KS. Estrogens and obesity: is it all in our heads? Cell Metab. 2011;144:435–436. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Yan L, Ge H, Li H, et al. Gender-specific proteomic alterations in glycolytic and mitochondrial pathways in aging monkey hearts. J Mol Cell Cardiol. 2004;375:921–929. [DOI] [PubMed] [Google Scholar]
- 26.Pastorelli R, Carpi D, Airoldi L, et al. Proteome analysis for the identification of in vivo estrogen-regulated proteins in bone. Proteomics. 2005;518:4936–4945. [DOI] [PubMed] [Google Scholar]
- 27.Reiss NA. Ontogeny and estrogen responsiveness of creatine kinase and glycolytic enzymes in brain and uterus of rat. Neurosci Lett. 1988;842:197–202. [DOI] [PubMed] [Google Scholar]
- 28.Kostanyan A, Nazaryan K. Rat brain glycolysis regulation by estradiol-17 beta. Biochim Biophys Acta. 1992;11333:301–306. [DOI] [PubMed] [Google Scholar]
- 29.Bogacka I, Roane DS, Xi X, et al. Expression levels of genes likely involved in glucose-sensing in the obese Zucker rat brain. Nutr Neurosci. 2004;72:67–74. [DOI] [PubMed] [Google Scholar]
- 30.Houston KD, Copland JA, Broaddus RR, et al. Inhibition of proliferation and estrogen receptor signaling by peroxisome proliferator-activated receptor gamma ligands in uterine leiomyoma. Cancer Res. 2003;636:1221–1227. [PubMed] [Google Scholar]
- 31.Penttinen P, Jaehrling J, Damdimopoulos AE, et al. Diet-derived polyphenol metabolite enterolactone is a tissue-specific estrogen receptor activator. Endocrinology. 2007;14810:4875–4886. [DOI] [PubMed] [Google Scholar]
- 32.Dyck DJ, Heigenhauser GJ, Bruce CR. The role of adipokines as regulators of skeletal muscle fatty acid metabolism and insulin sensitivity. Acta Physiol (Oxf). 2006;1861:5–16. [DOI] [PubMed] [Google Scholar]
- 33.Soriano S, Alonso-Magdalena P, García-Arévalo M, et al. Rapid insulinotropic action of low doses of bisphenol-A on mouse and human islets of Langerhans: role of estrogen receptor β. PLoS One. 2012;72:e31109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Dolinoy DC, Huang D, Jirtle RL. Maternal nutrient supplementation counteracts bisphenol A-induced DNA hypomethylation in early development. Proc Natl Acad Sci U S A. 2007;10432:13056–13061. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Wang X, Kilgore MW. Signal cross-talk between estrogen receptor alpha and beta and the peroxisome proliferator-activated receptor gamma1 in MDA-MB-231 and MCF-7 breast cancer cells. Mol Cell Endocrinol. 2002;194(1-2):123–133. [DOI] [PubMed] [Google Scholar]
- 36.Hurst CH, Waxman DJ. Activation of PPARalpha and PPARgamma by environmental phthalate monoesters. Toxicol Sci. 2003;742:297–308. [DOI] [PubMed] [Google Scholar]
- 37.Feige JN, Gelman L, Rossi D, et al. The endocrine disruptor monoethyl-hexyl-phthalate is a selective peroxisome proliferator-activated receptor gamma modulator that promotes adipogenesis. J Biol Chem. 2007;28226:19152–19166. [DOI] [PubMed] [Google Scholar]
- 38.Desvergne B, Feige JN, Casals-Casas C. PPAR-mediated activity of phthalates: a link to the obesity epidemic? Mol Cell Endocrinol. 2009;304(1-2):43–48. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.