1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Flavonoids are bioactive polyphenols found particularly in fruit and vegetables, but little is known about their role in bone health in humans. The aim of this observational study was to investigate whether dietary flavonoid intake was associated with bone mineral density (BMD) and bone resorption in a large group of perimenopausal Scottish women. Over 3000 women completed a food frequency questionnaire as part of an osteoporosis screening study. The diets were analyzed for flavonoid intake using a food composition database. BMD was measured at the femoral neck (FN) and lumbar spine (LS) by dual-energy X-ray absorptiometry (DXA). Free pyridinoline (PYD) and deoxypyridinoline (DPD) were measured by high-performance liquid chromatography (HPLC) in second early morning fasted urine samples. The mean flavonoid intake of the diet was 307 ±199 mg/d. The catechin family contributed the most to flavonoid intakes (55%), and the flavones the least (<1%). Associations were found between energy-adjusted total flavonoid intakes and BMD at the FN and LS (FN r = 0.054, LS r = 0.036, p ≤ .05). Annual percent change in BMD was associated with intakes of procyanidins and catechins (p ≤ .05), and flavanones were negatively associated with bone-resorption markers (PYD r = −0.049, DPD r = –0.057, p ≤ .001). These associations were still seen after adjusting for confounders. It is concluded that dietary flavonoid intakes are associated with BMD, supporting the evidence from animal and cellular studies. © 2011 American Society for Bone and Mineral Research.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

It has been estimated that one in two women and one in five men in the United Kingdom over the age of 50 years will have an osteoporotic fracture,1 and the prevalence of osteoporosis and related fractures is increasing.2 Poor nutrition is likely to play a role in the pathogenesis of the disease, and in the last decade, a number of population-based studies have demonstrated a beneficial effect of fruit and vegetables on indices of bone health.3–7

There could be several mechanisms by which fruit and vegetables act on bone, either via some pharmacologically active compound found within them or by the base buffering of excess dietary metabolic acids. A recent 2-year randomized, controlled trial found that supplementation with potassium citrate or three additional portions of fruit and vegetables a day did not reduce bone turnover or increase BMD, suggesting that any long-term benefits of fruit and vegetable intake on bone are not due to the provision of alkaline salts.8 Animal studies have shown that ovariectomy-induced bone resorption can be inhibited by onions and other vegetables commonly eaten by humans9 and that this effect is independent of their base-balancing properties,10 indicating that the protective effects of fruit and vegetables may be found in the nutrients they contain.

One family of nutrients, the flavonoids, which are found commonly in fruit, vegetables, tea, nuts, and seeds, has not been investigated with respect to bone health in humans. Human studies on dietary flavonoids and bone health have focused on tea because tea has been reported to protect against hip fracture.11, 12 Cellular models and animal studies have found associations between bone health and flavonoids13–15; therefore, the aim of this study was to examine whether dietary flavonoids (ie, flavonols, flavones, flavanones, catechins, and procyanidins) were associated with markers of bone health in a large cohort of Scottish women.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information


Subjects were recruited from a population-based screening program for osteoporotic fracture risk involving 5119 women aged 45 to 54 years who first visited our unit between 1990 and 1994.16, 17 They were selected randomly from the Community Health Index, which records the entire population registered with a general practitioner in Scotland. They all underwent bone densitometry and risk factor assessment by questionnaire. They were invited for further screening between 1997 and 1999, when 3883 returned. At the second visit, 3226 women completed a diet and physical activity questionnaire and had bone densitometry measurements, and the majority provided a second early morning fasted urine sample for analysis of bone-resorption markers (n = 2929). Written informed consent was obtained from all the women, and the Grampian Research Ethics Committee approved the study.

Assessment of bone health

BMD was measured at the femoral neck (FN) and lumbar spine (LS; L2–L4) by dual-energy X-ray absorptiometry (DXA; Norland XR26 and XR36; Cooper Surgical, Inc., Trumbull, CT, USA) at both visits. Most of the women were scanned using the XR26, but 7% of the women were scanned using the XR36. A comparison of 50 phantom measurements using both machines showed a small difference (1.258%) between the machines, so a correction factor was used to convert the XR36 values to equivalent XR26 values.18 The same trend in results was seen whether or not this correction factor was used. The percentage change in BMD from the first scan was calculated, and because the elapsed time between visits was different for each participant, this was standardized to annual percentage change in BMD. In our unit, the in vivo error for the LS is 0.56%, right hip 0.77%, and left hip 0.72%.

The urinary bone-resorption markers free pyridinoline (fPYD) and free deoxypyridinoline (fDPD) were measured in second early morning fasting urine by reverse-phase high-performance liquid chromatography (HPLC) with fluorescence detection.19 Creatinine was measured in urine by standard automated techniques (Roche, Lewes, UK), and the results were expressed as fPYD/Cr and fDPD/Cr (nmol/mmol). The interassay coefficient of variation for both methods was 5.5%.

Assessment of diet and flavonoid intakes

Dietary intake was assessed using the same validated food frequency questionnaire (FFQ) that has been used in other studies of diet and bone health in Scotland.3, 5 The FFQ quantifies intakes of 98 foods in the present and past (up to 12 years and 20 to 30 years of age) and has been validated using biochemical markers of antioxidant status and 7-day weighed records.20, 21 Dietary flavonoid intakes of the foods were calculated using a nutrient database that was validated for use with an earlier version of the FFQ.22 When nutrient information was missing (eg, for cooked vegetables), the US Department of Agriculture (USDA) database was consulted.23 Dishes with foods that contained flavonoids (eg, vegetable soups) had their flavonoid content computed using a dietary analysis package to calculate the weight and nutrients lost during cooking. Flavonoid cooking losses were calculated based on data available for vitamin C.

Although the earlier version of the FFQ had been validated for use with flavonoids, we assessed the relative validity of our updated FFQ for use with flavonoids in a subset of the Aberdeen Prospective Osteoporosis Screening Study population (n = 218). The subjects completed 4-day estimated portion size food diaries and the FFQ 9 months apart as a measure of compliance in a fruit and vegetable intervention trial.8

Assessment of potential confounders

The women were weighed on a set of balance scales (Seca, Hamburg, Germany) calibrated to 0.5 kg while wearing light clothing and no shoes. Height was measured with a stadiometer (Holtain, Ltd., Crymych, United Kingdom) to 0.5 cm. Information on health, smoking status, menopausal status, and hormone-replacement therapy (HRT) was collected. Social deprivation category was assessed from postal codes.24 Physical activity levels (PALs) were obtained using the same questionnaire as used in the Scottish Heart Health Study.25 The PAL was calculated from the number of hours in the day spent doing light, moderate, or heavy activities and how many hours were spent in bed or resting. These questions were asked separately for working and nonworking days.

Statistical analysis

All statistical analyses were carried out using SPSS Version 17.0 (SPSS, Inc., Chicago, IL, USA). Descriptive statistics were determined for all the bone and confounding variables, and when the distribution was skewed, the natural logarithm was used to transform data to achieve normality. Dietary flavonoid intakes were adjusted for total energy intake using the residual method.26

For assessment of the relative validity of the FFQ, Pearson's correlation coefficients were calculated for the flavonoid variables in order to measure the degree of association between the two methods of dietary assessment, and Bland and Altman plots were examined to establish the relative precision of the FFQ compared with the food diaries.

Dummy variables were created for a combination of HRT use and menopausal status. There were five mutually exclusive categories for menopausal status: premenopausal, perimenopausal and postmenopausal (never used HRT), past HRT users, and finally, present HRT users (which were used as the reference category).

In order to study the relationship between bone health and diet, Pearson's correlation coefficients and multiple linear regressions were used to assess the relationship between the dietary and bone variables. The dummy variables and the independent predictors of age, height, weight, PAL, social deprivation category, and current smoking were entered in a stepwise manner.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Validation study

The mean age of the 218 postmenopausal women who completed both FFQ and 4-day food diary was 60.8 ± 2.2 years, and their mean energy intake was 8.05 ± 1.77 MJ/d, as calculated by the food diaries. The FFQ calculated greater intakes for flavonols, catechins, flavanones, and total flavonoids than the food diaries (Table 1). The only flavonoid group where the correlation between the two methods was not significant was the flavones (r = 0.021, p = .771). Flavonol intakes showed the strongest correlation between the two methods (r =  0.785, p < .001), and total flavonoids also were highly correlated (r = 0.757, p < .001). Bland Altman plots indicated reasonable precision between the FFQ and the food diaries for all flavonoid groups except the flavones.

Table 1. Results of the Validation Study Comparing FFQs and 4-Day Estimated Portion Size Food Diaries for Assessing Flavonoid Intake in 218 Postmenopausal Scottish Women
Flavonoid groupFFQ, mean ± SD (mg/d)Diary, mean ± SD (mg/d)ra (p)
  • a

    Pearson's correlation coefficient for energy-transformed flavonoid variable.

Flavonols39.3 ± 24.027.1 ± 17.40.785 (<.001)
Flavones0.3 ± 0.30.4 ± 1.00.021 (.771)
Procyanidins44.2 ± 30.147.9 ± 28.10.537 (<.001)
Catechins153.9 ± 124.2130.9 ± 97.30.664 (<.001)
Flavanones56.3 ± 46.636.6 ± 34.40.603 (<.001)
All flavonoids294.0 ± 183.8242.9 ± 146.10.757 (<.001)

Main study

The characteristics of the study population (n = 3226) are shown in Table 2. One-fifth of the women were obese (body mass index > 30 kg/m2), 38% were overweight, and 40% had a normal body mass index (BMI). Estimated dietary flavonoid intakes for the women are given in Supplemental Table S1. All subjects consumed some flavonols, and quercetin contributed to 64% of flavonol intake. Flavone intakes were very small, and they were absent in 3 women. Luteolin, which is found in honey, sweet peppers, and lettuce, contributed an average of 21% to flavone intakes. All subjects ate procyanidins, and a broad range of intakes was seen from less than 1 to more than 200 mg/d. The catechin family contributed the most to flavonoid intakes (55%), with every subject consuming some. Epigallocatechinogallate (EGCG) contributed the greatest percentage (25.2%) of the total catechins intake, with epicatechinogallate (ECG) contributing slightly less (24.7%). Both these catechins are found in high levels in black tea. The flavanones (eg, naringin and hesperidin) accounted for 15% of total flavonoid intake. Naringin and hesperidin were absent from the diet of fewer than 1% of the women. The mean intake of naringin was equivalent to that obtained from half a small grapefruit, and eating half an average size orange would give a similar mean intake of hesperidin.

Table 2. Subject Characteristics of the Cohort
  • a

    fPYD/Cr = free pyridinoline expressed relative to creatinine; fDPD/Cr free = deoxypyridoline expressed relative to creatinine; PAL = physical activity level; HRT = hormone-replacement therapy.

  • b

    Based on postcode classification, where 1 represents the most affluent and 6 represents the most deprived.

Bone mineral density (g/cm2)3226  
Lumbar spine 0.9950.166
Annual % change LS –0.641.21
Femoral neck 0.8300.119
Annual % change FN –0.771.06
Bone-resorption markers (nmol/mmol)a2929  
fPYD/Cr 19.26.8
fDPD/Cr 5.32.0
Weight (kg)323368.912.4
Height (cm)3233160.55.9
Age (years)323655.12.2
BMI (kg/m2)323126.74.6
Dietary flavonoid intake (mg/d)3226306.8198.9
Current smoker576  
HRT use and menopausal status   
 Past HRT user676  
 Present HRT user1183  
Social deprivation categoryb   

Tea was the main contributor of flavonoids to the diet (57%), and the women reported drinking an average of 596 ± 504 mL of tea per day. Fresh fruit and fruit juice contributed significant amounts of flavonoids to the diet (18% and 12%, respectively), and intakes of these were 214 ± 123 g/d (fresh fruit) and 105 ± 114 mL/d (juice). Vegetables provided less than 5% of flavonoids to the diet from a mean intake of 192 ± 88 g/d. Seventy-three percent of the women reported eating five 80-g portions of fruit and vegetables per day.

Femoral neck BMD was correlated with intakes of all flavonoid groups except flavones (Table 3). At the lumbar spine, flavonols, procyanidins, flavanones, and total flavonoids showed weaker associations than at the hip. Catechins and procyanidins were associated with the annual percentage change in BMD at both sites. Flavanones were the only flavonoid group to show a significant correlation with both bone-resorption markers. Total flavonoids also were associated with fDPD/Cr.

Table 3. Pearson's Correlations for Energy-Adjusted Flavonoid Groups and Bone-Related Variables
Flavonoid groupaBone mineral densityb% Annual change in BMDResorption markersb
 FN (n =3230)LS (n = 3230)FN (n = 3230)LS (n = 3230)fPYD/Cr (n = 2933)fDPD/Cr (n = 2933)
  • FN = femoral neck; LS = lumbar spine; fPYD/Cr = free pyridinoline relative to creatinine; fDPD/Cr = free deoxy pyridinoline relative to creatinine.

  • a

    Adjusted for energy using the residuals method.

  • b

    BMD and bone-resorption marker variables transformed using the natural logarithm.

  • *

    p ≤ .05;

  • **

    p ≤ .01.

Total flavonoids0.054**0.036*0.0280.034–0.027–0.037*

Linear regression analysis (adjusting for the confounders of weight, PAL, height, smoking, social deprivation category, age, HRT use, and menopausal status) showed that the correlations (as seen in Table 3) still were significant with the exception of flavanones and LS BMD, and total flavonoids and fDPD/Cr. Associations between dietary flavonoids and BMD appeared to be stronger at the hip than at the spine. Both catechins and procyanidins were associated with annual percentage change in BMD at both sites, and only the flavanones showed any association with bone-resorption markers (Table 4). Menopausal status was the greatest predictor of fDPD/Cr (13.2%) and fPYD/Cr (data not shown). Energy-adjusted total flavonoids explained 0.3% of the model (p = .001) with a negative coefficient (unstandardized β) of –0.011 (95% confidence interval –0.018 to –0.003) showing that dietary flavanones were associated with decreased bone resorption.

Table 4. Results of Multiple Linear Regression Analysis to Identify Independent Predictors of FN BMD and fDPD/Cr for All Women
Predictorsa (n = 3230)Variation explained, %Unstandardized β95% Confidence interval for βp Value
  • FN BMD = femoral neck bone mineral density; fDPD/Cr  = free deoxypyridinoline relative to creatinine; HRT = hormone-replacement therapy

  • a

    Independent variables were selected from weight, height, smoking, physical activity levels, age, social deprivation category, HRT use, and menopausal status.

  • b

    To account for menopausal status and HRT use, dummy variables were used in the model with present HRT use as the comparison group.

 Intercept –0.551–0.731, –0.372<0.001
 Dummy variableb2.4
  Premenopausal0.0390.013, 0.064.003
  Perimenopausal0.0280.009, 0.047.004
  Postmenopausal–0.017–0.028, –0.006.002
  Past HRT user–0.033–0.046, –0.021<.001
 Weight (kg)13.60.0040.004, 0.005<.001
 Height (m)1.00.0020.001, 0.003<.001
 Age (years)0.7–0.006–0.008, –0.004<.001
 Energy adjusted flavonoids (mg)0.30.0090.003, 0.014.001
 Physical activity level0.10.0180.002, 0.033.028
 Intercept 2.0991.785, 2.413<.001
 Dummy variableb13.2
  Premenopausal0.1220.055, 0.190<.001
  Perimenopausal0.1820.134, 0.230<.001
  Postmenopausal0.2590.231, 0.287<.001
  Past HRT user0.2530.221, 0.284<.001
 Weight (kg)1.30.0040.003, 0.005<.001
 Height (m)1.0–0.006–0.008, –0.004<.001
 Energy-adjusted flavanones (mg)0.3–0.011–0.018, –0.003.005
 Current smoking0.20.0380.007, 0.069.015

The regression model for FN BMD showed that weight was the most important predictor, explaining 13.6% of the overall variation. Energy-adjusted total flavonoids explained 0.3% of the model (p = .001). Increasing age, past HRT use, and being postmenopausal had a negative effect on FN BMD, whereas dietary flavonoids appeared to protect the femoral neck, as did physical activity.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

To our knowledge, this is the first study to report on the potential influence of dietary flavonoids on bone health in a large cohort of women. We found a significant association between total flavonoid intake and BMD at the hip and lumbar spine. Dietary flavanones were shown to have a weak but significant negative correlation with bone-resorption markers, and catechins and procyanidins were associated with an annual change in BMD at both sites.

A particular strength of this study is the large cohort size and the inclusion of five different groups of flavonoids used to estimate intake. Most previous studies have focused on the flavonols and flavones, and many published intakes are based on the content of these two flavonoid groups only.27, 28 Our FFQ estimated that the mean intake of dietary flavonoids was 307 ± 199 mg/d in our population. This is greater than that found by Kyle,29 who used 4-day food diaries to estimate intakes in 81 Scottish men and women (median intake 132 mg/d), although the proportional contribution by individual flavonoids to total flavonoid intake was analogous with this study. The differences may be accounted for in the method of dietary assessment and the length of time over which Kyle's data were collected because longer periods of observation may be required to assess nutrient intakes accurately.30 Our study population had a healthy diet with greater fruit and vegetable intakes than those reported in the Scottish Health Survey of 2003, which also could explain the high estimated intakes of flavonoids. Flavonoid intakes in the United States calculated using the National Health and Nutrition Examination Survey (NHANES) 1999–2002 dietary records estimate an average of 190 mg/day, which is one-third lower than our estimation. This is not unexpected because the flavonoid density of the diet has been found to increase with age, income, female gender, and white race.31

The correlations of intakes between FFQ and 4-day food diaries were good (r = 0.76 for total flavonoids). Willett32 suggests that when FFQs are compared with multiple records of diet, correlation coefficients may reach 0.6 to 0.7, but values of 0.8 to 0.9 are unlikely. The validity study did not use biomarkers because correlations between these and dietary records of flavonoid intakes are not good (r = 0.35)33 owing to their bioavailability and metabolism in the gut.34–36

Limitations of this study include those associated with using an FFQ because FFQs are known to overestimate the true overall frequency of consumption of fruit and vegetables.37, 38 There are issues that affect the flavonoid content of foods, namely, analytical and environmental factors, processing, storage, and species differences, and these cannot be detected using a FFQ because it is a crude measure of diet. One particular concern in this study is the lack of information in the FFQ about the type of tea drunk and the length of infusion time because these factors are known to affect the concentration of flavonoids in tea.39 FFQs, despite inaccuracies, can be used to classify people into groups of intakes, and this ranking is the essential factor when studying diet and disease.40

A further limitation of this study is the flavonoid database, which used two sources of composition data. There were some differences between them, but care was taken to ensure that the same analytical techniques were used to measure flavonoids in foods. Adjusting for flavonoid losses during cooking was based on what happens to vitamin C because literature concerning flavonoid losses owing to cooking is limited. Vitamin C and flavonoids have similar chemical properties in that they are both hydrophilic and both leach from foods during cooking. Also, Fjelkner-Moidg and colleagues41 found close agreement between directly analyzed and calculated quercetin values in moussaka, giving validity to the method of estimation used for this study.

Urinary resorption markers show circadian rhythms, with a peak during the night and nadir in the afternoon; therefore, sampling time is important.42 This problem was addressed in this work, where the women fasted overnight and attended the clinic in the mornings.

Common sources of flavonoids in the human diet are fruit and vegetables, and previous work has found associations between better bone health and greater fruit and vegetables intakes.3, 5–7 Antioxidants specifically found in fruit and vegetables such as carotenoids may have a protective role,43 but many of these effects have not been studied extensively, and more research is required in this area. Flavonoids also are found in alcoholic drinks,44 moderate intakes of which have been associated with better bone health in some studies.5, 45

A universal source of flavonoids is tea, which is a rich combination of catechins and procyanidins. These were associated with annual percentage changes in BMD at both sites. A further confounder in relation to tea drinking is the Western custom of addition milk, thus increasing calcium intakes, but Hegarty and colleagues46 found that mean BMD did not vary between women who added milk to their tea and those who did not. In any case, in countries where the addition of milk to tea is uncommon, it has been found that tea drinking is still associated with greater BMD.47, 48 It has been suggested that the addition of milk to tea may alter the bioavailability of antioxidant polyphenols. However, a bioavailability study examining the effects of infusion time and addition of milk to tea found that while greater infusion time increased the antioxidant capacity and phenolic content of tea, adding milk had no effect on these parameters.39

In this study, femoral neck BMD was associated with all groups of flavonoids except flavones. This dietary factor explained a small amount of the variation (<1%); it is similar to that found for other key nutrients and greater than that explained by physical activity, a factor that has been shown to be important in the maintenance of bone mass.49

At the lumbar spine, associations were found with flavonols, procyanidins, and total flavonoids. Weight was a strong predictor of both LS BMD and FN BMD. Menopausal status was a factor in both BMD models, showing that the presence or absence of estrogen plays a greater role in the more metabolically active lumbar spine than the femoral neck. The significant association between LS BMD and flavonols backs up the work of Horcajada-Molteni and colleagues,13 who suggested that rutin (a flavonol) and/or its metabolites inhibit ovariectomy-induced trabecular bone loss in rats by slowing down resorption and increasing formation.

Bone-resorption markers were negatively associated with intakes of flavanones, showing that flavanones, which accounted for about 15% of total flavonoid intakes, were associated with decreased bone resorption. The regression models for the bone markers showed different patterns from those for BMD because the dummy variable explained greater variations than weight, and smoking was seen to influence bone resorption. This work is supported by animal studies in which nonpharmacologic doses of citrus juice have been shown to exert a positive effect on decreasing bone resorption in male rats.50 A study by Chiba and colleagues14 found that hesperidin prevented bone loss in ovariectomized mice, suggesting that this flavanone may be useful in preventing disease arising from estrogen deficiency, although the molecular mechanisms behind this observation are still unclear. Hesperidin has a high antioxidant capacity, and it has been suggested that bone resorption is decreased by inhibition of osteoclastic superoxide availability.51 A further mechanism of action may be through estrogen receptors, to which flavanones have been found to have a high binding affinity.52 A further explanation may be that those with high intakes of hesperidin may have a greater ability to secrete the metabolites required for its absorption, thus increasing bioavailability.36

This cross-sectional study of over 3000 women has shown associations between greater intakes of dietary flavonoids and improved bone health, although causality cannot be inferred owing to the study design. To conclude, there are associations between flavonoid intakes and healthy bones, but these associations are not strong and should be considered only as indicators.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

This work was funded by the UK Food Standards Agency Postgraduate Scholarship Scheme.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information
  • 1
    van Staa TP, Dennison EM, Leufkens HG, Cooper C. Epidemiology of fractures in England and Wales. Bone. 2001;29:517522.
  • 2
    Sambrook P, Cooper C. Osteoporosis. Lancet. 2006;367:20102018.
  • 3
    New SA, Bolton-Smith C, Grubb DA, Reid DM. Nutritional influences on bone mineral density: a cross-sectional study in premenopausal women. Am J Clin Nutr. 1997;65:18311839.
  • 4
    New SA, Robins SP, Campbell MK, et al. Dietary influences on bone mass and bone metabolism: further evidence of a positive link between fruit and vegetable consumption and bone health? Am J Clin Nutr. 2000;71:142151.
  • 5
    Macdonald HM, New SA, Golden MH, Campbell MK, Reid DM. Nutritional associations with bone loss during the menopausal transition: evidence of a beneficial effect of calcium, alcohol, and fruit and vegetable nutrients and of a detrimental effect of fatty acids. Am J Clin Nutr. 2004;79:155165.
  • 6
    Tucker KL, Hannan MT, Chen H, Cupples LA, Wilson PW, Kiel DP. Potassium, magnesium, and fruit and vegetable intakes are associated with greater bone mineral density in elderly men and women. Am J Clin Nutr. 1999;69:727736.
  • 7
    Prynne CJ, Mishra GD, O'Connell MA, et al. Fruit and vegetable intakes and bone mineral status: a cross sectional study in 5 age and sex cohorts. Am J Clin Nutr. 2006;83:14201428.
  • 8
    Macdonald HM, Black AJ, Aucott L, et al. Effect of potassium citrate supplementation or increased fruit and vegetable intake on bone metabolism in healthy postmenopausal women: a randomized controlled trial. Am J Clin Nutr. 2008;88:465474.
  • 9
    Muhlbauer RC, Li F. Effect of vegetables on bone metabolism. Nature. 1999;401:343344.
  • 10
    Muhlbauer RC, Lozano A, Reinli A. Onion and a mixture of vegetables, salads, and herbs affect bone resorption in the rat by a mechanism independent of their base excess. J Bone Miner Res. 2002;17:12301236.
  • 11
    Johnell O, Gullberg B, Kanis JA, et al., Risk factors for hip fracture in European women: the MEDOS Study. Mediterranean Osteoporosis Study. J Bone Miner Res. 1995;10:18021815.
  • 12
    Kanis J, Johnell O, Gullberg B, et al. Risk factors for hip fracture in men from southern Europe: the MEDOS study. Mediterranean Osteoporosis Study. Osteoporos Int. 1999;9:4554.
  • 13
    Horcajada-Molteni MN, Crespy V, Coxam V, Davicco MJ, Remesy C, Barlet JP. Rutin inhibits ovariectomy-induced osteopenia in rats. J Bone Miner Res. 2000;15:22512258.
  • 14
    Chiba H, Uehara M, Wu J, et al. Hesperidin, a citrus flavonoid, inhibits bone loss and decreases serum and hepatic lipids in ovariectomized mice. J Nutr. 2003;133:18921897.
  • 15
    Wattel A, Kamel S, Mentaverri R, et al. Potent inhibitory effect of naturally occurring flavonoids quercetin and kaempferol on in vitro osteoclastic bone resorption. Biochem Pharmacol. 2003;65:3542.
  • 16
    Garton MJ, Torgerson DJ, Donaldson C, Russell IT, Reid DM. Recruitment methods for screening programmes: trial of a new method within a regional osteoporosis study. BMJ. 1992;305:8284.
  • 17
    Torgerson DJ, Garton MJ, Donaldson C, Russell IT, Reid DM. Recruitment methods for screening programmes: trial of an improved method within a regional osteoporosis study. BMJ. 1993;307:99.
  • 18
    Macdonald HM, McGuigan FE, Fraser WD, New SA, Ralston SH, Reid DM. Methylenetetrahydrofolate reductase polymorphism interacts with riboflavin intake to influence bone mineral density. Bone. 2004;35:957964.
  • 19
    MacDonald AG, Birkinshaw G, Durham B, Bucknall RC, Fraser WD. Biochemical markers of bone turnover in seronegative spondylarthropathy: relationship to disease activity. Br J Rheumatol. 1997;36:5053.
  • 20
    Bolton-Smith C, Casey CE, Gey KF, Smith WC, Tunstall-Pedoe H. Antioxidant vitamin intakes assessed using a food-frequency questionnaire: correlation with biochemical status in smokers and non-smokers. Br J Nutr. 1991;65:337346.
  • 21
    New SA. An epidemiological investigation into the influence of nutritional factors on bone mineral density and bone metabolism. PhD Thesis, University of Aberdeen, Scotland, UK. 1995.
  • 22
    Kyle J, Masson LF, Grubb DA, Duthie GG, McNeill G. Estimating dietarty flavonoid intake: comparison of a semi-quantitative food frequency questionnaire with 4-day weighed diet records in a Scottish population, vol. 61 Proceedings of the Nutrition Society, pp 69A. 2002.
  • 23
    U.S. Department of Agriculture ARS 2004 USDA database for the flavonoid content of selected foods.
  • 24
    McLoone P. Carstairs scores for Scottish postcode sectors from the 1991 census, vol. PhD. University of Glasgow, Glasgow. 1994.
  • 25
    Tunstall-Pedoe H, Smith WC, Crombie IK, Tavendale R. Coronary risk factor and lifestyle variation across Scotland: results from the Scottish Heart Health Study. Scott Med J. 1989;34:556560.
  • 26
    Willett W, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol. 1986;124:1727.
  • 27
    Hertog MG, Kromhout D, Aravanis C, et al. Flavonoid intake and long-term risk of coronary heart disease and cancer in the seven countries study. Arch Intern Med. 1995;155:381386.
  • 28
    de Vries JH, Janssen PL, Hollman PC, van Staveren WA, Katan MB. Consumption of quercetin and kaempferol in free-living subjects eating a variety of diets. Cancer Lett. 1997;114:141144.
  • 29
    Kyle JA. Flavonoids in Health and Disease [PhD thesis]. Aberdeen, Scotland: University of Aberdeen; 2006.
  • 30
    Bingham SA, Gill C, Welch A, et al. Comparison of dietary assessment methods in nutritional epidemiology: weighed records v. 24h recalls, food-frequency questionnaires and estimated-diet records. Br J Nutr. 1994;72:619643.
  • 31
    Chun OK, Chung SJ, Song WO. Estimated dietary flavonoid intake and major food sources of U.S. adults. J Nutr. 2007;137:12441252.
  • 32
    Willett WC. Invited commentary: comparison of food frequency questionnaires. Am J Epidemiol. 1998;148:11571159 discussion 1162-1155.
  • 33
    Nielsen SE, Freese R, Kleemola P, Mutanen M. Flavonoids in human urine as biomarkers for intake of fruits and vegetables. Cancer Epidemiol Biomarkers Prev. 2002;11:459466.
  • 34
    Williamson G, Manach C. Bioavailability and bioefficacy of polyphenols in humans. II. Review of 93 intervention studies. Am J Clin Nutr. 2005;81 (1 Suppl): 243S255S.
  • 35
    Manach C, Scalbert A, Morand C, Remesy C, Jimenez L. Polyphenols: food sources and bioavailability. Am J Clin Nutr. 2004;79:727747.
  • 36
    Silberberg M, Morand C, Mathevon T, et al. The bioavailability of polyphenols is highly governed by the capacity of the intestine and of the liver to secrete conjugated metabolites. Eur J Nutr. 2006;45:8896.
  • 37
    Haraldsdottir J. Minimizing error in the field: quality control in dietary surveys. Eur J Clin Nutr. 1993;47 (Suppl 2): S1924.
  • 38
    Amanatidis S, Mackerras D, Simpson JM. Comparison of two frequency questionnaires for quantifying fruit and vegetable intake. Public Health Nutr. 2001;4:233239.
  • 39
    Kyle JA, Morrice PC, McNeill G, Duthie GG. Effects of infusion time and addition of milk on content and absorption of polyphenols from black tea. J Agric Food Chem. 2007;55:48894894.
  • 40
    Block G. A review of validations of dietary assessment methods. Am J Epidemiol. 1982;115:492505.
  • 41
    Fjelkner-Modig S, Johansson K, Ewald C, Akesson B, Sjoholm I. Flavonoids in vegetables - the influence of processing Proceedings of the 2nd International Conference on Natural Antioxidants and Anticarcinagens in Nutrition, Health and Disease. 1998.
  • 42
    Delmas PD. Biochemical markers of bone turnover. J Bone Miner Res. 1993;8 (Suppl 2): S549555.
  • 43
    Sahni S, Hannan MT, Blumberg J, Cupples LA, Kiel DP, Tucker KL. Protective effect of total carotenoid and lycopene intake on the risk of hip fracture: a 17-year follow-up from the Framingham Osteoporosis Study. J Bone Miner Res. 2009;24:10861094.
  • 44
    Sun B, Spranger I, Yang J, et al. Red wine phenolic complexes and their in vitro antioxidant activity. J Agric Food Chem. 2009;57:86238627.
  • 45
    Tucker KL, Jugdaohsingh R, Powell JJ, et al. Effects of beer, wine, and liquor intakes on bone mineral density in older men and women. Am J Clin Nutr. 2009;89:11881196.
  • 46
    Hegarty VM, May HM, Khaw KT. Tea drinking and bone mineral density in older women. Am J Clin Nutr. 2000;71:10031007.
  • 47
    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;162:10011006.
  • 48
    Chen Z, Pettinger MB, Ritenbaugh C, et al. Habitual tea consumption and risk of osteoporosis: a prospective study in the women's health initiative observational cohort. Am J Epidemiol. 2003;158:772781.
  • 49
    Hingorjo MR, Syed S, Qureshi MA. Role of exercise in osteoporosis prevention--current concepts. J Pak Med Assoc. 2008;58:7881.
  • 50
    Deyhim F, Garica K, Lopez E, et al. Citrus juice modulates bone strength in male senescent rat model of osteoporosis. Nutrition. 2006;22:559563.
  • 51
    Key LLJ, Wolf WC, Gundberg CM, Ries WL. Superoxide and bone resorption. Bone. 1994;15:431436.
  • 52
    Kuiper GG, Lemmen JG, Carlsson B, et al. Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology. 1998;139:42524263.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

jbmr_285_sm_SuppTabS1.doc37KSupplementary Table S1

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.