点击显示 收起
1 From the Elsie Widdowson Laboratory, MRC Human Nutrition Research, Cambridge, United Kingdom
1 From the Elsie Widdowson Laboratory, MRC Human Nutrition Research, Cambridge, United Kingdom.
2 Supported by Project Award N05044 from the Food Standards Agency.
3 Reprints not available. Address correspondence to CJ Prynne, Elsie Widdowson Laboratory, MRC Human Nutrition Research, Fulbourn Road, Cambridge, CB1 9NL, United Kingdom. E-mail: celia.greenberg{at}mrc-hnr.cam.ac.uk.
See corresponding editorial on page 1254.
ABSTRACT
Background: Evidence is increasing for positive effects of fruit and vegetable intakes on bone health. However, most of the studies to date were conducted in adults, and few reports included adolescents.
Objective: We explored the association between bone mineral status and fruit and vegetable intakes in adolescent boys and girls (aged 16–18 y), young women (aged 23–37 y), and older men and women (aged 60–83 y).
Design: Bone mineral measurements of the whole body, hip, and spine were made in all subjects by using dual-energy X-ray absorptiometry. Information on health and lifestyle and physical activity was obtained by questionnaire. Fruit, vegetable, and nutrient intakes were ascertained from 7-d food diaries.
Results: In adolescent boys and girls and older women, significant positive associations were observed between spine size-adjusted bone mineral content (SA-BMC) and fruit intake. In boys only, femoral neck SA-BMC was also significantly and positively associated with the intakes of both fruit and dietary vitamin C. No significant associations were found in the young women or older men, or between bone measurements and intake of vegetables alone (after adjustments) in any of the groups.
Conclusions: Higher fruit and vegetable intakes may have positive effects on bone mineral status in both younger and older age groups, especially at the spine and femoral neck. The specific mechanisms remain to be ascertained, but vitamin C, other fruit-specific antioxidants, and lifestyle may play a role.
Key Words: Fruit and vegetables • bone mineral content • bone mineral density • adolescents • adults
INTRODUCTION
Osteoporosis is a major public health problem in the United Kingdom and around the world. The deterioration in bone mass and microarchitecture associated with the disease leads to greater risk of fragility fracture, disability, and premature mortality. Diet is one of the modifiable risk factors for osteoporosis, and adequate dietary calcium intake and vitamin D status have long been recognized as factors in maintaining bone health (1). As long ago as 1968, it was suggested that a diet emphasizing fruit and vegetables and moderate amounts of milk should be considered as an adjunct to therapy for osteoporosis (2). Renewed interest in fruit and vegetable intakes came with the publication of several cross-sectional studies showing a positive link between bone mineral density (BMD) and fruit and vegetable consumption in adult men and women (3–9). However, studies involving postmenopausal women have not shown any association (9, 10). A 3-mo intervention study in 23–76-y-old men and women showed that consumption of a diet high in fruit and vegetables [ie, the Dietary Approaches to Stop Hypertension (DASH) diet] resulted in a significant reduction in bone turnover markers (3), which, if sustained for the long term, could result in bone mineral gains.
The mechanism whereby fruit and vegetables may affect bone is not clear and may be multifactorial. A higher dietary acid load (caused by greater consumption of acid-forming foods, such as cereals and meat, than of alkali-forming foods, such as fruit and vegetables) is believed to result in bone mineral dissolution and greater bone resorption, which results in the release of carbonate, citrate, calcium, sodium, and potassium (11). A diet rich in fruit and vegetables may result in a more alkaline environment, which has been shown to reduce urinary calcium excretion (12). Fruit and vegetables also are rich sources of antioxidant vitamins such as vitamin C and ß-carotene. These could act by combating oxidative stress, which has been shown to be negatively associated with BMD in adults (13). Vitamin C also has an essential regulatory role in osteoblast differentiation (14) and collagen formation and therefore may positively influence bone health (15–17). Vegetables are the principal source of vitamin K1 in the diet, and evidence is growing that vitamin K1 has a role in bone mineralization by acting as a cofactor in the -carboxylation of the bone protein osteocalcin (18–20). More emphasis is now being placed on the diet as a whole in relation to bone health rather than on isolated nutrients, but the evidence linking fruit and vegetables to bone health across the life cycle is not conclusive.
The principal aim of the current study was to investigate the associations between bone mineral status and actual fruit and vegetable intakes, as estimated from 7-d food diaries. This study used data collected previously from 3 different studies including 5 cohorts—adolescent boys and girls, young women, and older men and women aged 16–83 y. Bone mineral status was assessed by size-adjusted (SA) bone mineral content (BMC) and bone mineral density (BMD). These measures were related to fruit and vegetable consumption after adjustment for other dietary and lifestyle factors. The secondary aim of the project was to investigate whether vitamin C, specifically, was exerting an influence on bone mineral status.
SUBJECTS AND METHODS
Subjects
The adolescent subjects, 125 girls and 132 boys aged 16–18 y, had been recruited from local schools and colleges to take part in the Cambridge (United Kingdom) Bone Studies (21–23). The adult subjects also were recruited from the Cambridge area. The 120 young women were between 23 and 37 y old. They either were nulliparous or had not given birth in the past 2 y or lactated for 1 y and had been recruited for the Young Women's Pregnancy Study (an ongoing study within the research unit). Recruitment was through advertisements, from the volunteer lists maintained at MRC Human Nutrition Research, or by word of mouth. The older subjects, recruited for the Vitamins K and D Study (24), were 70 men and 73 women aged 60–83 y. They were recruited through advertisements placed in general practitioners' offices and by recruitment presentations at retirees' associations. Exclusion criteria for all subjects included any medical condition, a history of any eating disorder, and current medication known to interfere with bone and calcium metabolism [except the contraceptive pill and hormone replacement therapy (HRT)]. Seven percent of the girls and 46% of the young women were taking a contraceptive pill, and 19% of the older women were currently receiving HRT. No dietary advice was given to the subjects.
All volunteers and their parents or guardians (for the adolescents' study) gave written informed consent. Ethical approval for the adolescents' study was provided by the Ethics Committee of the MRC Dunn Nutrition Unit (now MRC Human Nutrition Research). Ethical approval for the young adult women's and the older men and women's studies was provided by the Cambridge Local Research Ethics Committee.
Anthropometric and bone mineral status measurements
In all subjects, height was measured to the nearest 0.1 cm by using a stadiometer (Doherty premier height measure; Sidhil Ltd, Halifax, United Kingdom). Weight was measured to the nearest 0.1 kg by using an electronic digital scale (Seca Scale; Todd Scales Ltd, Norwich, United Kingdom).
In adolescent girls and boys, BMC (g) and bone area [(BA) cm2] of the whole body, lumbar spine (L1–L4), and left hip (total, femoral neck, and greater trochanter) were measured by using dual-energy X-ray absorptiometry (DXA) in a Hologic QDR 1000/W scanner (Hologic Inc, Waltham, MA). DXA software (version 5.61; Hologic Inc) was used for whole-body measurements (enhanced analysis), and the performance mode (version 4.47P) was used for spine and hip measurements. In young women and older men and women, the BMC and BA of the whole body, lumbar spine, and hip were measured by using a Lunar MD DXA scanner (GE Lunar, Madison, WI (software 4.7d). Quality assurance was performed daily, and long-term instrument stability was assessed 2–3 times/wk.
Dietary intakes
All subjects were asked to complete a prospective 7-d food diary; dietary records were obtained from 84% and 76% of the boys and girls, respectively; from 75% of the young women; and from 94% and 85% of the older men and women, respectively. Portion sizes were matched against food photographs, and quantities were described in household measures. The subjects were asked supplementary questions about their diet, such as the type of milk they usually drank, the type of fat used for cooking or spreading, whether they ate meat, and the type of water they drank. The diet records were coded by using the in-house software program, Diet In Data Out (DIDO; 25), and nutrient analysis was performed by using the in-house suite of software programs based on McCance and Widdowson's The Composition of Foods: the 4th (26) and 6th (27) editions for the adolescents and the older cohorts, respectively, and the supplements for cereal and cereal products (28) and milk and milk products (29). The 2 editions were used because dietary data were collected from the adolescents several years earlier than from the older cohorts. Intakes of calcium included calcium from the recorded intake of water, which was coded as Cambridge tap water unless otherwise specified. The analysis software allowed certain related items of foods to be grouped together to find the total weight of those related foods eaten per day. For the purposes of this investigation, total vegetable intake included green vegetables, root vegetables, pulses, salad vegetables including tomatoes, other vegetables, and mixed-vegetable dishes such as curries and stir-fry dishes, but it did not include potatoes. Total fruit intake included fresh fruit, cooked or canned fruit, dried fruit, nuts, and fruit juices. Only fruit and vegetables specifically reported were assessed; it was not possible to infer the contribution from composite dishes unless the contents were described by the subject.
Physical activity, health, and lifestyle
Habitual physical activity was ascertained in adolescent boys, young women, and older men and women by using the European Prospective Investigation into Cancer (EPIC) physical activity questionnaire (30). Activity levels in the adolescent girls were assessed by using a questionnaire adapted from the Allied Dunbar Fitness Survey UK, 1992 (31). For the purposes of this study, the measure of physical activity used was the total h/wk spent on recreational activities.
Other relevant information was collected from the subjects by using a general health and lifestyle questionnaire. Details of smoking habits and dietary supplement use were recorded for all subjects. Information was obtained on the use of the contraceptive pill use by the girls and young women and on the use of HRT by the older women. The girls were asked their age at menarche, and the older women were asked their age at menopause.
Statistical analysis
Data analysis was performed by using SPSS for MS WINDOWS software (version 10.0; SPSS Institute, Chicago, IL). Because the data for this project came from 3 separate studies, each of which used a different sampling method, all statistical analyses were stratified by sex and age group. Descriptive statistics (ie, means, SDs, medians, and interquartile ranges) were ascertained for all variables. To correct for skewed distributions and to examine proportional associations, all bone data and the weights of foods were converted to natural logarithms. Multiple linear regression analysis was carried out on log-transformed data and the associations between bone at each of the sites and intakes of fruit, vegetables, combined fruit and vegetables, intake and vitamin C were ascertained. When regression analysis was carried out with the use of log-transformed data and when the independent variable was continuous, the regression coefficient multiplied by 100 corresponded closely with the percentage of change in the dependent variable associated with each 100% change in the independent variable, as defined by (difference/mean) x 100 (32). Therefore, the results of the regression analysis are presented as percentage coefficients.
Univariate analysis was first carried out to examine the associations between BMC, BA, and BMD and fruit, vegetable, combined fruit and vegetable, and vitamin C intakes. Backward stepwise regression analysis was then carried out to ascertain the associations with BMC and BMD after adjustment for potential confounders [ie, age, total energy intake, calcium intake, vitamin K1 intake, supplement use, smoking, time spent per week in recreational activities, age at menarche (girls only), contraceptive pill use (girls and young women), age at menopause (older women only), and current HRT use (older women only)]. Because BMD may not adequately correct for bone and body size, models were also constructed in which BMC was adjusted for BA, height, weight, and the confounders listed above (SA-BMC model; 33). Results from the parsimonious models are presented as mean (±SE) percentage coefficients. In addition, for the older subjects only, to compare with an earlier publication (7), a further model was constructed in which BMD was the dependent variable, and height, body mass index (BMI; in kg/m2), age, total energy intake, supplement use, smoking, time spent in recreational activities, age at menopause, and HRT use were included as independent variables.
When significant associations with fruit intake were found in groups (adolescents and older women), analysis of variance was used to compare the mean intakes of energy and selected nutrients by thirds of fruit intake. Scheffe's test was performed to indicate significant differences between the thirds. For those vitamins with very skewed distribution, the analysis was carried out by using log-transformed data. Means and 95% CIs for each third are reported.
Results were not compared across the cohorts because the data were not collected at the same time and the assessments of bone mineral status were carried out with the use of different equipment. Each cohort had 80% power to detect correlations of 0.2 to 0.3 between bone mineral measurements and the consumption of fruit, vegetables, and combined fruit and vegetables.
To assess the likely validity of the reported energy intakes, the ratio of energy intake to the basal metabolic rate was calculated (34); a ratio of <1.1 was taken to indicate energy underreporters. Associations between bone mineral measurements were ascertained with and without underreporters.
RESULTS
Subject characteristics
General characteristics of subjects including age, weight, height, BMI, age at menarche, age at menopause, HRT use, contraceptive use, smoking, and hours of physical activity per week are shown in Table 1. Table 1 also shows the summary statistics for BMC, BA, and BMD of the whole body, spine, total hip, trochanter, and femoral neck for all subjects. It should be noted that the Lunar DXA gives systematically higher BMD values than does the Hologic DXA (35).
View this table:
Table 1. General characteristics and summary statistics for bone area and mineral measurements of whole body, spine, total hip, femoral neck, and trochanter in boys, girls, young women, and older men and women1
When the estimates of total EI were compared with the calculated BMR, it was found that 12% of the boys and 8% of the girls were potential underreporters (36). In the young women, 20% were found to have a ratio of energy intake to basal metabolic rate of <1.1; in older men and women, the proportion was 3% and 12%, respectively. Inclusion of the underreporters made no difference to the bone-diet associations.
Intake of fruit and vegetables
The arithmetic mean, median, and interquartile range for intakes (by wt) of total fruit (including nuts and juices), fruit juices (from the total fruit intake), vegetables (not including potatoes), and total fruit and vegetables consumed by each of the age groups are shown in Table 2. Intakes of fruit and vegetables increased with age and were greatest in the older women. Only the older subjects, who were >60 y old, consumed an average of more than the recommended 400 g/d. Fifty-four percent of the women and 49% of the men had individual mean intakes 400 g/d. The intake of nuts was very small—a mean of 4.7, 4.1, 5.6, and 2.7 g/d for adolescents, young women, old men, and old women, respectively.
View this table:
Table 2. Daily intakes of fruit and vegetables in adolescent boys, girls, young women, and older men and women1
Associations between bone measurements and fruit and vegetable intakes
The univariate associations between BMC, BA, and BMD and intakes of fruit, vegetables, and combined fruit and vegetables in boys and girls are shown in Table 3. In the boys, significant positive associations were found between BMC, BA, and BMD and fruit alone and combined fruit and vegetables at nearly all skeletal sites. For vegetables alone, significant associations were found only with whole-body and femoral neck BMC and whole-body BA. In the parsimonious multiple regression models, whole-body, total hip, femoral neck, and trochanter BMD and femoral neck and spine SA-BMC were significantly and positively associated with fruit and combined fruit and vegetable intakes, as shown in Table 4.
View this table:
Table 3. Results of univariate analysis between whole-body, spine, total hip, femoral peck, and trochanter bone mineral content (BMC), bone area (BA), and bone mineral density (BMD) and fruit, vegetable, and combined fruit and vegetable intakes in adolescent boys and girls1
View this table:
Table 4. Associations between bone mineral measurements at the whole-body, spine, and hip sites and fruit and vegetable intakes in adolescent boys and girls and fruit intake in older women from parsimonious regression models1
In the univariate analysis of the girls (Table 3), the only significant positive associations found were those between spine BMC and BMD and intakes of fruit alone and combined fruit and vegetables and that between whole-body BMD and combined fruit and vegetable intakes. In the parsimonious multiple regression models (Table 4), whole-body SA-BMC was significantly and positively associated with combined fruit and vegetable intakes, and spine SA-BMC was significantly and positively associated with fruit and combined fruit and vegetable intakes. Whole-body and spine BMD were significantly and positively associated with fruit intake, and the effect was greater when fruit and vegetables were combined
In young women and older men, no significant associations were found between any of the skeletal sites and fruit and vegetable intakes (when examined separately or in combination). When the association between BMD at various skeletal sites and total fruit and vegetable intakes was extended to include height and BMI in older men [to allow a direct comparison with a previously published study (7)], no significant associations were found.
In older women, significant univariate (negative) associations were found only between total hip and trochanter BA and fruit intake and between trochanter BA and combined fruit and vegetable intakes (results not shown). In the parsimonious regression model, a significant positive association was found between spine SA-BMC and fruit intake (Table 4). This showed that doubling the fruit intake would result in a 5% increment in BMC of the spine. Models similar to those used by Tucker et al (7) were constructed, and a significant (P 0.05) positive association between spine BMD and fruit intake was found (results not shown, but doubling the fruit intake would result in a 5.6% increment in BMD of the spine). Adjustment for energy underreporting did not change the results in any subject group.
Dietary characteristics by levels of fruit consumption
To further understand the associations found between SA-BMC and fruit intake in the adolescents and the older women, comparisons of energy and key nutrients were made between thirds of fruit intake. The results for adolescent boys and girls are shown in Table 5. In the adolescents and the older women, the highest third of fruit consumption was associated with significantly higher intakes of potassium, folic acid, and vitamin C than was the lowest third (results in older women not shown). In boys and girls, carotene intake was significantly greater in the highest third than in the lowest third of fruit intake. In boys only, intakes of calcium, vitamin E, and vitamin K1 were significantly higher in the highest third than in the lowest third of fruit intake, and there was a nonsignificant trend for protein intake and total weight-bearing activity to be higher (results not shown).
View this table:
Table 5. Mean (and 95% CI) daily nutrient intakes by thirds (T) of total fruit intake in boys and girls1
Bone mineral status and dietary vitamin C intake
The associations between dietary vitamin C and BMC, BA, and BMD in the adolescents and older women are shown in Table 6. In boys, significant positive associations were found between dietary vitamin C and BMC and BMD at all skeletal sites and between dietary vitamin C and whole-body BA. No significant univariate association was found in the girls, and a significant negative association with hip BA was found in the older women. In the boys, the parsimonious model showed significant associations with whole-body BMD, spine and femoral neck BMD, and femoral neck SA-BMC (Table 7). Adjustment for energy underreporting did not change the results in any subject group.
View this table:
Table 6. Results of univariate analysis between whole-body, spine, total hip, femoral neck, and trochanter bone mineral content (BMC), bone area (BA), and bone mineral density (BMD) and vitamin C intake in boys, girls, and older women1
View this table:
Table 7. Associations between bone mineral content (BMC), bone mineral density (BMD), and size-adjusted BMC (SA-BMC) at the whole-body, spine, and hip sites and vitamin C intakes in adolescent boys1
DISCUSSION
Fruit and vegetable intakes
The results of this study show that higher fruit and vegetable intakes may have a positive effect on bone mineral status in adolescent boys and girls and older women, especially at the spine and femoral neck. Higher intakes of fruit and vegetables were associated with other diet and lifestyle characteristics that may have an overall beneficial effect on bone health.
The World Health Organization recommended in 1990 that 5 portions of fruit and vegetables (400 g) be eaten per day to reduce the risk of chronic diseases, including cardiovascular disease and cancer (37). Only 25% of the adolescents, 38% of the younger women, and 50% of the older men and women in the current study had intakes 400 g/d. The figures for fruit and vegetable intakes indicate that some of the subjects did achieve intakes equaling or surpassing the recommendation, and, particularly in the boys, higher intakes were associated with better bone mineral status.
Advantages and limitations of this study
This study was more rigorous than were previous studies that reported a positive association between fruit and vegetable intakes and bone health. In previous studies, dietary information was collected with the use of food-frequency questionnaires and retrospective reporting of fruit and vegetable intakes in earlier life (4, 7, 38). The current study is, to our knowledge, the only study in which the weights of fruit and vegetables, estimated separately from a prospective diet record, have been related to bone mineral status. In addition, this was the only study in which the analysis was carried out with the use of SA-BMC as a measure of bone mineral status. It is considered that the use of BMD does not adequately correct for bone and body size, particularly in adolescent subjects (32).
The main limitation of the current study was that the subjects were not nationally representative; they were generally affluent and well educated, and, as volunteers, all were well motivated. The mean intakes of fruit and vegetables of each age group or sex in this study were greater than those in age-matched, nationally representative samples (39–41). In addition, the number in subjects in each of the cohorts was low, albeit large enough to address the question posed.
Fruit and vegetable intakes and bone health in adolescents
Adolescence is a crucial period of bone growth during which maximization of peak bone mass may reduce the risk of osteoporosis later in life (42). However, there have not been many studies of this nature involving young people. McGartland et al (8) reported a higher heel BMD by using a Lunar PIXI bone densitometer in 12- and 15-y-old girls, but not in boys, who had a higher fruit intake. However, this association disappeared in the older girls when adjustment was made for height, weight, pubertal status, activity level, social class, alcohol, smoking, and use of a nutritional supplements. In the current study, after similar adjustments were made, the results showed that doubling the fruit intake of the girls (aged 16-18 y) could increase spine BMD by 2.3%. In the current study, compared with the study by McGartland et al, the boys and girls were older, different bone sites were examined, and the subjects' combined consumption of fruit and vegetables was considerably higher, 92 and 74 g/d, for boys and girls, respectively. Data contrary to those of McGartland et al but in closer agreement with the current study were obtained from a longitudinal study of 85 Canadian boys and 67 girls aged 8–20 y, which showed that intakes of fruit and vegetables had a significant effect on whole-body BMC in boys but not in girls (43). In that study, the increment in whole-body BMC was 22 g for a twofold increase in fruit and vegetable consumption, whereas, in the current study, doubling the mean fruit and vegetable intakes of the boys would result in a much greater increment (240 g) in the mean whole-body BMC.
Fruit and vegetable intakes and bone health in adults
Of the groups included in the current study, the adolescent boys probably had the most metabolically active bone, as indicated by the significantly higher concentrations of bone turnover markers (44) and by the fact that they were still growing. The young women appeared to be the most stable; they had the lowest concentration of bone turnover markers among the female subjects, whereas the older women may have been in a state of bone flux because of their postmenopausal status. No positive association between bone and fruit and vegetable intakes was found in the young women, and no significant difference was found between vegetarians and nonvegetarians. This finding contrasts with a study by New et al (4) in premenopausal women aged 45–49 y in whom lumbar spine BMD was greater in the higher quartiles of intakes of nutrients found in fruit and vegetables—potassium, magnesium, and vitamin C. However, our findings in boys, girls, and women aged >60 y are similar to those of New et al and suggest that the skeleton may be most susceptible to variations in diet during the growth and mineral consolidation of adolescence and again later in life, when bone loss is greater.
Tucker et al (7) examined dietary patterns in men and women aged 69–83 y and identified a group of high fruit and vegetable consumers. Men in that group (mean vitamin C intake: 305 mg/d) had greater hip BMD (including femoral neck, trochanter, and Ward's area) than did men who consumed less fruit and vegetables. In the current study, using the same model as Tucker et al, we found no association between BMD at any site and fruit and vegetable consumption in the older men, whose mean intake of vitamin C, 99 mg/d, was very much lower. Although we did find a positive association between fruit consumption and spine BMD in the older women. we could not find a positive association with vitamin C intake.
Vitamin C intake and bone health
In the current study, vitamin C was positively associated with bone mineral status at several sites in boys only. In 371 girls and boys aged 11–17 y whom Gunnes and Lehmann (16) studied, results indicated that an increase of 10 mg/d in vitamin C would result in an increment of 0.001 g/cm2 in distal radius BMD. The current study showed that, for the same increase in vitamin C intake, the boys achieved a greater increment of 0.003 (whole-body) to 0.005 (femoral neck) g/cm2. However, there were no significant results in the current study in girls, most of whom were older than the girls in the study of Gunnes and Lehmann. Nor were there any significant results in any of the older groups in the current study. This finding contrasts with previously published studies, in which positive associations were shown between BMD and dietary vitamin C in older women (4, 9, 15, 17, 45) and older men (15). The mechanism whereby the association between dietary vitamin C and bone mineral status in the boys was affected could be the well-defined role of dietary vitamin C in bone formation.
Bone health and the whole diet
In the current study, a greater intake of fruit and vegetables may have been an indicator of a better diet as a whole, and thus the whole diet could have contributed to the positive associations seen. It is not surprising that intakes of nutrients such as vitamin C, carotene, and potassium, which could be considered as markers for fruit, were higher in the high fruit consumers than in the low fruit consumers. Intakes of protein, calcium, and vitamin E were also higher in the boys who were high fruit consumers than in those who were low fruit consumers, which may indicate that the former group follows a better diet overall. However, calcium was not found to be a significant determinant of bone measurements in the boys or in any other age group. When considering the effect of fruit and vegetables on bone health, it should be remembered that these foods are part of the whole diet and may also represent a healthier lifestyle.
Conclusions
The results of this study show that higher fruit and vegetable intakes may have positive effects on bone mineral status in adolescents and older women, especially at the spine (girls and older women) and femoral neck (boys). The specific mechanisms remain to be determined, but vitamin C and other fruit-specific antioxidants may play a role (M O'Connell, G Mishra, G Muniz, et al, unpublished observations, 2005; M O'Connell, G Mishra, A Prentice, F Ginty, unpublished observations, 2005), and this is an area that merits further investigation. In a separate study of the same subjects (44), no significant associations were found between SA-BMC and 2 different measures of renal net acid excretion estimated from the diet. So the associations found between fruit and vegetable intakes and bone mineral status in these subjects did not appear to be related to a more alkaline environment (44). A recommendation to increase fruit and vegetable intakes, particularly in young people, would be expedient, because indications are that intakes in this age group are generally very low, and the results in the current study show that a considerable enhancement of BMC could be achieved. Future investigations could include intervention studies, particularly in younger groups. The fact that we were unable to show any associations between fruit and vegetable intakes and bone mineral status in young women and older men should not be interpreted as meaning that there is no advantage to be gained by an increase in intakes in these groups. Bone mineral status is only one aspect of health in which greater fruit and vegetable intakes may have long-term beneficial consequences.
ACKNOWLEDGMENTS
We gratefully acknowledge the Food Standards Agency of the United Kingdom for funding this study and the Medical Research Council of the United Kingdom for continued support. We are grateful to the participants of the Cambridge Bone Studies (adolescent boys and girls), the Young Women's Pregnancy Study (the young women), and the Vitamins K and D Study (the older men and women) and for the participants' permission to use their samples and data. We also thank the following people for their invaluable contributions: S Levitt for data entry; A Gent for dietary coding; and S Nigdikar and L McKenna for bone metabolism marker analysis.
GDM, FG and MAO were responsible for the conception and design of the study. AP, LY, and MAL collected the data. GDM, GM, and CJP analyzed the data, and AP provided significant advice on the analysis of the bone-related variables. CJP wrote the manuscript, and FG edited the manuscript. None of the authors had a financial or personal conflict of interest.
REFERENCES