Full text of DOI:10.1080/07315724.2002.10719242

This article was downloaded by: [University of Toronto Libraries]
On: 23 November 2014, At: 15:44
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of the American College of Nutrition
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/uacn20
Dietary Protein, Phosphorus and Potassium Are
Beneficial to Bone Mineral Density in Adult Men
Consuming Adequate Dietary Calcium
Susan J. Whiting PhDa, Jennifer L. Boyle MSca, Angela Thompson PhDa, Robert L. Mirwald
PhDa & Robert A. Faulkner PhDa
a College of Pharmacy and Nutrition and College of Kinesiology, University of
Saskatchewan, Saskatoon, Saskatchewan, CANADA
Published online: 19 Jun 2013.
To cite this article: Susan J. Whiting PhD, Jennifer L. Boyle MSc, Angela Thompson PhD, Robert L. Mirwald PhD &
Robert A. Faulkner PhD (2002) Dietary Protein, Phosphorus and Potassium Are Beneficial to Bone Mineral Density in
Adult Men Consuming Adequate Dietary Calcium, Journal of the American College of Nutrition, 21:5, 402-409, DOI:
10.1080/07315724.2002.10719242
To link to this article: http://dx.doi.org/10.1080/07315724.2002.10719242
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of
the Content. Any opinions and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied
upon and should be independently verified with primary sources of information. Taylor and Francis shall
not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other
liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions

Original Research
Dietary Protein, Phosphorus and Potassium Are
Beneficial to Bone Mineral Density in Adult Men
Consuming Adequate Dietary Calcium
Susan J. Whiting, PhD, Jennifer L. Boyle, MSc, Angela Thompson, PhD, Robert L. Mirwald, PhD, and
Robert A. Faulkner, PhD
College of Pharmacy and Nutrition and College of Kinesiology, University of Saskatchewan, Saskatoon, Saskatchewan, CANADA
Key words: bone mineral density, potassium, phosphorus, protein, calcium
Objective: The purpose of this study was to determine relationships of calcium (Ca), protein (Pr),
phosphorus (P) and potassium (K) to measures of bone mineral density in adult men.
Methods: Cross-sectional analysis of 57 men ages 39 to 42 years who were participants in an ongoing study.
Dietary assessment was conducted using the Block food frequency questionnaire (FFQ). BMD of total body
(TB), hip and lumbar spine (LS) were measured with dual X-ray absorptiometry (DXA).
Results: Ca, Pr, P and K, as well as lean body mass (LBM), showed significant correlation with BMD at the
total body, hip and lumbar spine. Stepwise forward regression selection method identified LBM, height and fat
mass as significant predictors of TB-BMD, LBM and height as significant predictors of hip BMD, and LBM as
a significant predictor of LS-BMD. As the nutrients tested correlated significantly with each other, only one
nutrient was entered into the regression model at a time to accommodate the potential for multicollinearity. In
regression analysis, adjusted for site-specific anthropometric variables and energy intake, K, Pr and P intake
accounted for significant (p Ͻ 0.05) prediction of TB-BMD and LS-BMD values by 7% to 13%. No bone-related
nutrient added significantly to the prediction of hip BMD. Ca intake was not significantly associated with BMD
at any site in the adjusted models.
Conclusions: Our analysis provides support that a moderate protein (1.2 g/kg) diet, plentiful in potassium
(Ͼ100 mmol/day) and phosphorus (1741 Ϯ 535 mg) is beneficial for maintaining bone mineral density in adult
men when Ca intake was adequate (1200 Ϯ 515 mg).
main area of research, it is noteworthy that diets adequate in
INTRODUCTION
calcium are often adequate in many other essential nutrients for
Failure to gain optimal bone density during adolescence and
young adulthood contributes to low bone density, leading to
osteoporosis later in life. Similarly, failure to reduce bone loss
throughout adulthood also contributes to low bone density.
Although efforts to prevent osteoporosis have focussed on
young people in their growing years, efforts could be made to
promote maintenance of bone mass by reducing bone loss
throughout adulthood. Although calcium intake has been a
good bone health, including vitamin D, potassium, magnesium,
phosphorus, and protein [1]. Determining dietary patterns is
important for making food-based recommendations.
Recent studies report a significant positive relationship be-
tween high protein intakes and bone density [2,3]; however,
others espouse the view that a high protein diet, by increasing
urinary calcium excretion, has a negative affect on bone [4,5].
On the other hand, there is growing evidence that low dietary
Address correspondence to: Susan J. Whiting, PhD, College of Pharmacy and Nutrition and College of Kinesiology, 110 Science Place, University of Saskatchewan,
Saskatoon SK, S7N 5C9, CANADA. E-mail: susan.whiting@usask.ca
Presented in abstract form at the American Society form Bone and Mineral Research annual meeting, Toronto, September 23, 2000.
Abbreviations: BMD ϭ bone mineral density, LS ϭ lumbar spine, TB ϭ total body, DXA ϭ dual X-ray absorptiometry, LBM ϭ lean body mass, FM ϭ fat mass, SD ϭ
standard deviation, HHHQ ϭ Health, Habits and History Questionnaire, FFQ ϭ food frequency questionnaire, SGDFS ϭ Saskatchewan Growth and Development
Followup Study, CV ϭ coefficient of variation.
Journal of the American College of Nutrition, Vol. 21, No. 5, 402–409 (2002)
Published by the American College of Nutrition
402

Diet and Bone in Men
protein negatively affects bone metabolism [6,7]. Studies of
potassium intake indicate that fruit and vegetable intake may
decrease the adverse effect of dietary acid load on bone
[8,9,10]. While high phosphorus intakes may seriously alter
calcium metabolism [11], one study reported that protein, phos-
phorus and calcium intakes were positively correlated with
bone mineral density, suggesting a greater intake of each of
these, including phosphorus, was beneficial [3]. None of these
studies has looked at all three—protein, phosphorus and potas-
sium—at the same time.
Dietary Assessment
Usual dietary intake was assessed using the Health, Habits,
and History Questionnaire (HHHQ), a semi-quantitative ques-
tionnaire also known as the “Block” FFQ [15–17]. Although
mailed to subjects, there was opportunity to have questions
answered by a trained undergraduate nutrition student at the
time of bone scans. FFQs were analyzed using Dietsys Version
3.0, obtained from the National Cancer Institute. All FFQs were
coded and double-entered by the same researcher. One subject
had a low reported energy intake (below 1000 kcal) that re-
mained low even after repeated measurement; this subject was
excluded. Food group servings and amount of alcohol based on
medium-sized drinks consumed per day were provided by
Dietsys.
While many studies have examined dietary intakes of adult
women in relation to bone mineral mass, few have examined
men. The etiology of bone loss in adult men is multifactorial
and includes heredity, lifestyle such as smoking, alcohol con-
sumption and physical activity, hormonal function and different
nutritional factors [12]. Since a well-balanced diet may be an
important modifiable factor related to bone health [1], it is
important to determine the role of nutrition in maintenance of
bone mineral density in adult men. Therefore, we examined the
effect of dietary factors on bone mineral density in adult men in
a cross-sectional study in which dietary information was col-
lected using a food frequency questionnaire. Three outcome
variables were evaluated: total body bone mineral density (TB-
BMD), total hip bone mineral density (hip BMD), and lumbar
spine bone mineral density (LS-BMD).
Anthropometric Assessment
Body measurements were taken by a trained anthropom-
etrist according to published protocol [18] during the time the
bone scan was performed, in a hospital setting. Measurements
included height, weight and skinfolds of the triceps, subscap-
ular, biceps, pectoral, supraspinale, iliac crest, abdominal, front
thigh and medial calf skinfolds from the right side of the body.
Height was measured twice and recorded to the nearest 0.1 mm.
Weight was measured twice and recorded to the nearest 0.1 kg.
Skinfold measurements were also taken twice with Harpenden
calipers and recorded to the nearest 0.1 mm.
Bone Mineral Measurement
METHODS
Bone measures were obtained using dual-energy X-ray ab-
sorptiometry using the Hologic QDR 2000 in array mode at the
Department of Nuclear Medicine, Royal University Hospital,
Saskatoon. Subjects (n ϭ 57) wore t-shirts and loose fitting
shorts, with shoes and metal objects removed. BMD was as-
sessed at the lumbar spine, proximal femur and total body sites.
Bone mineral content and bone area were also determined. All
scans were analyzed by the same qualified individual. Short
term precision values (expressed as coefficient of variation,
CV%) were less than 1.1% at both the proximal femur and
lumbar spine, and 0.51% for the total body BMD [19].
Subjects
A cross-sectional analysis of all male subjects from the
Saskatchewan Growth and Development Follow-up Study
(SGDFS) who completed both the dietary assessment and bone
measurements was performed. Subjects recruited for the
SGDFS had previously participated in the Saskatchewan
Growth and Developmental Study between 1964 and 1973, a
longitudinal study of 131 boys [13]. The initial sample was
randomly stratified on a socioeconomic basis from elementary
schools in Saskatoon. Dietary intakes were not obtained. In
1998, the SGDFS began, with the purpose to examine the
relationship between physical activity and fitness in childhood/
adolescence and health habits, physical activity level and phys-
ical fitness in middle adulthood. Subjects were contacted
through media. Upon agreement to participate in the follow-up
study (SGDFS), the subjects were sent four questionnaires to
self-administer: 1) demographics, 2) milk history [14], 3) food
frequency and 4) health, habits and behaviors. Data on current
physical fitness level, physical activity level, bone mineral
density, health habits, anthropometric measures, dietary intake
and demographics were collected. Seventy men (53% of orig-
inal cohort) completed at least part of the follow-up study.
Other Questionnaires
Current activity levels were assessed using the Seven-Day
Physical Activity Recall [21], administered at the time subjects
had their bone scans. Each component of the activity question-
naire was multiplied by an appropriate weighting factor based
on an estimate of energy expenditure and summed to arrive at
an estimate of total kilocalories expended per day from activity
for each subject. Information on current and past smoking
habits was collected. The number of pack-years smoked (as-
suming 20 cigarettes per pack) was calculated for all past and
current smokers.
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
403

Diet and Bone in Men
Table 1. Characteristics of Subjects (n ϭ 57)
Statistical Analysis
All statistical analyses were performed with SPSS (Statistical
Package for the Social Science) Version 9.0 (SPSS Inc., Chicago).
A p-value of Յ 0.05 was considered significant. The data were
initially explored for outliers and missing values. Descriptive
statistics were determined for all variables. The relation between
nondietary factors (age, weight, height, fat mass, lean body mass,
physical activity levels, alcohol consumption, coffee consumption
and pack-years smoked) and bone mass was examined using
Pearson correlation coefficients; then, forward stepwise multiple
regression was used to determine which factors significantly pre-
dicted BMD. Lean body mass, fat mass and height were identified
as significant predictors of total body BMD; lean body mass and
height were significant predictors of total hip BMD, and lean body
mass was identified as a significant predictor of lumbar spine
BMD. Thus, these factors were considered confounders and were
controlled in all subsequent nutritional analyses of BMD. Because
most nutrients correlate with energy intake, adjusting for energy
was necessary to assess the independent effect contributed by the
nutrient of interest [9,10]. Furthermore, controlling for energy
intake may account for differences that may have been due to body
size or activity levels, as well as correct for some of the measure-
ment error inherent in the FFQ [21]. Energy was controlled by
including it as a confounding variable within each regression
model. All distributions were checked for normalcy.
To better understand the potential of collinearity among
energy, calcium, potassium, protein and phosphorus intakes,
simple and partial (adjusted for identified non-dietary con-
founders) correlations among these variables were performed.
Because these analyses showed evidence of high collinearity
(r Ͼ 0.8) among most variables, it was not possible to assess
the independent effects of calcium, potassium, protein and
phosphorus on BMD in the same model. Intake for each nutri-
ent (calcium, potassium, protein or phosphorus) and energy
was entered into each BMD regression model of non-dietary
predictors/confounders by the forward stepwise regression pro-
cedure using a staged approach with only one nutrient entered
at a time [9,10]. Partial F tests [22] were calculated to deter-
mine if the addition of the nutrient to the model significantly
contributed to the prediction of BMD over and above that
achieved by the non-dietary predictors already in the model.
This method allowed the strongest independently predictive
factors of BMD to be identified.
Measurement
Age (years)
Weight (kg)
Height (cm)
Physical Activity (kcal/day)
Smoking (pack-yearsa)
Alcohol consumption
(drinks/day)
Coffee intake (cups/day)
BMI (kg/m2)
Body Fat (%)
Fat Mass (kg)
Lean Body Mass (kg)
Total Body BMD (g/cm2)
Hip BMD (g/cm2)
Lumbar BMD (g/cm2)
Mean Ϯ SD
Range
39.6 Ϯ 0.6
84.3 Ϯ 15.3
178 Ϯ 7
2260 Ϯ 1479
16.7 Ϯ 12.2
39–42
57.5–127.3
157–195
0–7092
0.1–57.5
0.8 Ϯ 0.2
2.7 Ϯ 2.4
0–8.4
0–10
26.5 Ϯ 4.2
23.7 Ϯ 7.8
20.9 Ϯ 10.3
60.4 Ϯ 6.8
1.15 Ϯ 0.09
1.01 Ϯ 0.14
1.03 Ϯ 0.13
19.1–39.3
9.3–42.3
5.3–53.8
49.3–76.7
0.94–1.43
0.61–1.38
0.77–1.36
BMD ϭ bone mineral density.
a average pack-years (#cigarettes per day/20 ϫ #years smoked) for former
and current smokers, n ϭ 32; value does not include 25 subjects who never
smoked.
of young adult non-Hispanic men, ages 20 to 29 years, measured
in NHANES III [23]. One subject had a hip BMD of 2.5 standard
deviations below the mean hip BMD of the young adult standard
reference group. The majority of the subjects, 82.5%, had normal
hip BMD values. Physical activity, alcohol consumption, coffee
consumption and smoking were considered possible confounders
in diet-bone relationships; however, no significant correlations
were found between these lifestyle factors and bone density mea-
surements (data not shown).
Mean (Ϯ SD) and 5th and 95th percentiles of nutrient intake
for energy and nutrients used in analyses and for servings/day of
the four food groups are presented in Table 2. Seventy-two per
cent of the subjects were below the average energy requirement
(2700 kcal/day) for adult men ages 25 to 49 [24]. Average protein
intake per kilogram body weight was 1.17 g/kg body weight, with
19% of subjects having an intake below the current Canadian
recommendation of 0.86 g of protein/kg body weight [24]. Aver-
age reported calcium intake was above the AI of 1000 mg/day
[25]. All subjects (100%) reported phosphorus intakes above the
1997 RDA of 700 mg/day [25]. The average intake for potassium
was above the U.S. 1989 RDA desirable intake of 90 mmol/day
[26]. Average servings of fruit and vegetables, milk products,
breads and cereals, and meat, fish and poultry were close to
recommended servings for Canadians (i.e., 5–10, 2–4, 5–12, 2–3,
respectively) considering Dietsys serving sizes were generally
larger than Health Canada sizes [27].
RESULTS
Energy intake was significantly and highly correlated with
calcium, phosphorus, potassium and protein (r ϭ 0.56, 0.76, 0.80
and 0.84, respectively); therefore, we adjusted for energy in all
subsequent analyses. Correlations among dietary calcium, phos-
phorus, potassium, protein, adjusted for energy were strong, indi-
cating collinearity (Table 3). Lean body mass was significantly
correlated (p Ͻ 0.01) with all three BMD sites (r ϭ 0.37–0.45)
Mean values (Ϯ SD) for subject characteristics including life-
style factors, BMD, and anthropometric measurements are de-
scribed in Table 1. Forty-nine per cent of subjects were within a
BMI range of 20–26, 18% were marginally above, 28% were
overweight and the remaining 5% were underweight. The average
hip BMD of the subjects, ages 39 to 42 years, was similar to that
404
VOL. 21, NO. 5

Diet and Bone in Men
Table 2. Daily Intake of Bone-Related Nutrients of Male Subjects (n ϭ 57)
Nutrient/Food Group1
Mean Ϯ SD
5th
50th
2202
91.7
1050
1669
11.6
0.67
3572
3.5
95th
Energy (kcal)
Protein (g)
Ca (mg)
P (mg)
Ca:Pr (mg/g)
Ca:P (mg/mg)
K (mg)
Fruit & Vegetables (serving)
Milk Products (serving)
Breads and Cereals (serving)
Meat2 (serving)
2343 Ϯ 715
97 Ϯ 27
1370
61.1
508
1098
6.2
0.45
2283
3458
147.4
2112
2723
19.5
0.89
5578
8.2
1200 Ϯ 515
1741 Ϯ 535
12.1 Ϯ 3.9
0.66 Ϯ 0.16
3787 Ϯ 1093
4.2 Ϯ 2.1
3.4 Ϯ 2.5
3.1 Ϯ 1.4
2.1 Ϯ 0.7
1.7
0.8
1.0
1.0
2.7
2.9
2.1
10.2
5.8
3.8
1 Food Groups according to HHHQ [16–18].
2 Alternates such as eggs, peanut butter, beans and lentils not included.
Ca ϭ calcium, P ϭ phosphorus, Ca:Pr ϭ calcium:protein ratio, Ca:P ϭ calcium:phosphorus ratio, K ϭ potassium.
Table 3. Pearson Correlation Coefficients between Dietary
Calcium, Phosphorus, Potassium and Protein Adjusted for
Energy Intake1
correlated with all three BMD sites, while calcium intake was
significantly correlated with total body and spine BMD. These are
shown as unadjusted data in Table 4. When adjusted for energy
intake and appropriate anthropometric measures, all correlations
among protein, potassium and phosphorus and BMD measures
remained significant, while associations between calcium and
BMD measures became insignificant (adjusted values, Table 4).
Fruit and vegetable intake was significantly correlated with total
body BMD but became insignificant once adjusted for anthropo-
metric measures and energy intake. No other food group was
significantly associated with BMD at any site.
Calcium
1.00
Phosphorus
Potassium
Protein
Calcium
0.881
1.00
0.587
0.733
1.00
0.519
0.761
0.566
1.00
Phosphorus
Potassium
Protein
1 n ϭ 57; all Pearson correlation coefficients are significant, p Ͻ 0.001.
Calcium, potassium, phosphorus and protein intakes were
added individually to three BMD models—total body, hip and
spine. The complete display of regression models for total body
BMD only is shown in Table 5. Models containing a combi-
nation of two or more of calcium, potassium, phosphorus and
protein tended not to show independent effects because of the
high collinearity between dietary nutrients of interest (data not
shown). Generally, potassium, protein and phosphorus were
determined to significantly contribute to the prediction of TB
BMD and spine BMD, but not hip BMD, over and above the
models consisting of non-dietary predictors (Table 6). The
while weight approached significance (r ϭ 0.258, p ϭ 0.053) at
the total body BMD site. No significant correlations between age,
height, percent body fat, fat mass and bone mineral density mea-
sures were found. Lean body mass, height and fat mass were
selected to be included in the predictive model of total body BMD.
Lean body mass and height were significant predictors of total hip
BMD, and only LBM was significantly related to spine BMD.
These factors were considered confounders in the relationship
between calcium/bone-related nutrients and BMD and were con-
trolled for in all subsequent analyses.
Dietary protein, phosphorus and potassium were significantly
Table 4. Pearson Correlations Coefficients for Nutrient Intakes and Bone Mineral Density Measures Unadjusted and Adjusted for
Non-Dietary Factors and Energy Intake
Total body BMD
Hip BMD
Unadjusted
Spine BMD
Unadjusted
Adjusted1
Adjusted2
Unadjusted
Adjusted3
Calcium
0.285*
0.218
0.213
0.115
0.253
0.189
Phosphorus
Protein
Potassium
0.373**
0.358**
0.404**
0.297*
0.385**
0.383**
0.401**
0.187
0.301*
0.289*
0.346**
0.201
0.272*
0.322*
0.309*
0.122
0.296*
0.284*
0.325**
0.141
0.331*
0.419**
0.400**
0.129
Fruit & Vegetable Intake
n ϭ 57.
1 Controlled for lean body mass, fat mass, height, and energy intake.
2 Controlled for lean body mass, height, and energy intake.
3 Controlled for lean body mass, and energy intake.
Significance: * p Ͻ 0.05, ** p Ͻ 0.01.
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
405

Diet and Bone in Men
Table 5. Final Multiple Linear Regression Models of Total Body BMD with and without Addition of Bone Related Nutrients
Variable
Coefficient
SE
t
p
Model 11: TB-BMD ϭ Constant ϩ LBM ϩ Ht ϩ FM
Constant
LBM
1.373
0.011
Ϫ0.0046
Ϫ0.0024
0.265
0.002
0.002
0.001
5.18
5.12
Ϫ2.57
Ϫ2.11
Ͻ0.01
Ͻ0.01
0.01
Ht
FM
0.04
Adjusted R2 ϭ 0.296
Model 2: TB-BMD ϭ Constant ϩ LBM ϩ Ht ϩ FM ϩ Energy ϩ Ca2
Constant
LBM
Ht
FM
1.42674
0.01029
Ϫ0.00478
Ϫ0.00273
Ϫ0.00002
0.00004
0.26727
0.00215
0.00178
0.00115
0.00002
0.00002
5.34
4.78
Ϫ2.69
Ϫ2.37
Ϫ0.98
1.60
Ͻ0.01
Ͻ0.01
0.01
0.02
0.33
Energy
Ca
0.12
Adjusted R2 ϭ 0.303
Model 3: TB-BMD ϭ Constant ϩ LBM ϩ Ht ϩ FM ϩ Energy ϩ K3
Constant
LBM
Ht
FM
1.2860
0.0086
Ϫ0.0037
Ϫ0.0030
0.00006
0.00005
0.2509
0.0021
0.0017
0.0011
0.00002
0.00002
5.13
4.11
Ϫ2.57
Ϫ2.11
Ϫ2.57
3.13
Ͻ0.01
Ͻ0.01
0.01
0.04
0.01
Energy
K
0.01
Adjusted R2 ϭ 0.386
Model 4: TB-BMD ϭ Constant ϩ LBM ϩ Ht ϩ FM ϩ Energy ϩ Protein3
Constant
LBM
Ht
FM
1.36000
0.00941
Ϫ0.00439
Ϫ0.00298
Ϫ0.00006
0.00193
0.25142
0.00207
0.0017
0.00108
0.00002
0.00065
5.41
4.54
Ϫ2.61
Ϫ2.14
Ϫ2.52
2.96
Ͻ0.01
Ͻ0.01
0.01
0.04
0.02
Energy
Protein
Ͻ0.01
Adjusted R2 ϭ 0.376
Model 5: TB-BMD ϭ Constant ϩ LBM ϩ Ht ϩ FM ϩ Energy ϩ P3
Constant
LBM
Ht
FM
1.41683
0.00970
Ϫ0.00467
Ϫ0.00291
0.00005
0.25144
0.00205
0.00168
0.00109
0.00002
0.00003
5.63
4.73
Ϫ2.67
Ϫ2.77
Ϫ2.31
2.98
Ͻ0.01
Ͻ0.01
0.01
0.01
0.02
Energy
P
0.00008
Ͻ0.01
Adjusted R2 ϭ 0.377
n ϭ 57.
1 Model of non-nutrient predictors of BMD determined by forward stepwise regression (variables entered: age, LBM, FM, height, education level, physical activity, alcohol
consumption, coffee consumption, and pack years smoked).
2 Addition of nutrient to model is not significant (Partial F Test).
3 Addition of nutrient to model significant (Partial F Test), p Ͻ 0.05.
TB-BMD ϭ total body bone mineral density, LBM ϭ lean body mass, Ht ϭ height, FM ϭ fat mass, Energy ϭ energy intake, Ca ϭ calcium, K ϭ potassium, P ϭ
phosphorus.
addition of calcium to the models did not significantly contrib-
ute to the prediction of BMD at any site.
change in BMD associated with a 1000 mg increase in phos-
phorus intake was 6.9% in total body BMD and 11.7% in
lumbar spine BMD with every 1000 mg increase in phosphorus
intake. Approximately 8.1% of the variation in total body BMD
and 7.0% of the variation in spine BMD can be explained by
dietary phosphorus intake. Protein intake also made a signifi-
cant contribution to the prediction of TB BMD and spine BMD.
There was a 1.7% change in total body BMD and a 3.4%
change in lumbar spine BMD for every 10 g increase in dietary
protein. About 8.0% of the variation in total body BMD and
Potassium was significantly associated with total body and
spine BMD sites; there was a 4.3% increase in total body BMD
and a 7.3% change in lumbar spine BMD with every 1000 mg
increase in potassium intake. The differences in adjusted R2
values for the non-dietary model and the model containing
potassium indicate that about 9% of the variation in TB BMD
and 11.6% of the variation in spine BMD can be explained by
dietary potassium intake. Similarly for phosphorus intake, the
406
VOL. 21, NO. 5

Diet and Bone in Men
Table 6. Summary Table of Multiple Regression Models Showing R2 Values for Models with and without Addition of Bone
Related Nutrients
Model 1:
no nutrient
Model 2:
ϩ calcium
Model 3:
ϩ potassium
Model 4:
ϩ protein
Model 5:
ϩ phosphorus
Bone Site
R2
R2
0.303
0.169
0.131
R2
0.3864
0.238
0.2434
R2
R2
Total Body BMD
Hip BMD
Spine BMD
0.2961
0.1882
0.1273
0.3764
0.245
0.2564
0.3774
0.22
0.1974
1 R2 with lean body mass (LBM), fat mass (FM), and height in the model.
2 R2 with LBM, and height in the model.
3 R2 with LBM in the model.
4 Addition of nutrient and energy intake to model is significant (Partial F Test), p Ͻ 0.05.
12.9% of the variation in spine BMD can be explained by
protein intake.
was adequate may have impacted on the findings for the other
nutrients examined [34].
Dietary potassium was determined a significant predictor of
TB BMD when LBM, FM, height and energy intake were
controlled and a significant predictor of spine BMD when LBM
and energy intake were controlled. Others have reported that
greater potassium intakes, from fruit and vegetables, were
associated with increased BMD at hip sites in elderly men and
women [9] and with spine BMD in premenopausal women [8].
The beneficial effects of dietary potassium on bone health is
well established in the literature [35–38]. Experimentally, sup-
plemental KHCO3 has been shown to be hypocalciuric in
subjects consuming moderate protein diets [37], to reduce
high-protein induced hypercalciuria [38] and reduce the rate of
bone resorption [36] by improving calcium balance. Epidemi-
ological research has also been successful in showing a positive
relation between dietary potassium and BMD [8–10]. In the
final linear regression models for TB BMD and spine BMD,
potassium was able to explain 9% of the variation in TB BMD
and 12% of the variation in spine in our male subjects. These
subjects had a diet with good variety, which was relatively high
in fruits and vegetables. With an increase in fruit and vegeta-
bles in the diet, there is also an increase in a variety of other
nutrients, such as magnesium, vitamin C and vitamin K, that
may also have an effect on bone health [39]. Magnesium was
not directly investigated in this study; however, it was posi-
tively correlated with TB BMD and spine BMD in a crude
analysis (data not shown). As both potassium and magnesium
have been reported to be positively related to BMD in studies
that indicated that increased fruit and vegetable intake was
positively related to BMD [8–10], our findings on potassium
may also apply to magnesium.
DISCUSSION
The primary focus of this study was to examine the effect of
calcium and bone-related nutrients (potassium, protein and
phosphorus) on bone mineral density in male subjects ages 39
to 42 years. The average daily intake of calcium, as well as
potassium, protein and phosphorus, exceeded recommended
intakes in our subjects. A diet high in calcium intake appears to
be a marker of a diet which is high in most or all other essential
nutrients [28]. Indeed, this group of middle-aged men could be
described as an above average group of individuals in terms of
eating a well-balance diet, at an ideal body weight for height,
and having good bone density.
Many factors, such as age, weight, height, BMI, physical
activity, are reported to play a role in the degree of bone density
obtained in men [29]. Age was not a confounder because our
group had a narrow age range, 39 to 42 years. Our stepwise
procedure identified lean body mass, fat mass and height as
significant predictors of TB BMD. Lean body mass and height
were significant predictors of hip BMD, and lean body mass
was a significant predictor of spine BMD. Nutrient intakes
were adjusted for total energy intake. No significant relation-
ships between the nutrients of interest (calcium, phosphorus,
protein, potassium) and BMD were found at the three sites
when all four were added simultaneously; therefore; each nu-
trient was entered individually into the models, as others have
done in situations of collinearity [9,10,31].
The evidence of a positive relationship between dietary
calcium and bone density is well accepted from clinical ran-
domized control trials [25] and increasingly from observational
studies [32]. Our study did not find an association between
calcium and bone mineral density at any of the three sites, once
anthropometric factors were controlled for, a finding consistent
with calcium’s being a threshold nutrient [33]. Our subjects’
mean intake of calcium exceeded the Adequate Intake level of
1000 mg [25]. As discussed below, the fact that calcium intake
Protein was a significant predictor of both TB BMD (p Ͻ
0.01) and spine BMD (p Ͻ 0.01), but not hip BMD, when
appropriate anthropometric variables and energy intake were
controlled. The average protein intake in this group of men was
96 g/day or 1.17 g/kg body weight, a level considered “mod-
erate” [40–42]. Until recently the focus has been on the neg-
ative effects of excess protein in the diet, yet low dietary
intakes of protein (below the 1989 RDA of 0.8 g/kg body
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
407

Diet and Bone in Men
weight) negatively affect bone [41,43]. Kerstetter and col-
leagues [42] suggested that calcium metabolism is normal
when subjects consumed a well-balanced diet containing mod-
erate amounts of protein (ϳ1.0 g/kg body weight/day). Animal
protein and vegetable protein may have differential effects on
calcium balance [5]; however, not all studies are in agreement
[6,34]. Recently, a study examining the effect of dietary protein
on BMD in elderly men and women was able to show low
calcium absorption with high protein intakes when calcium
intakes were low and higher BMD with high protein when
calcium intakes were high [34]. Together, this study and our
data suggest that protein exerts a positive effect on BMD when
calcium intake is adequate.
REFERENCES
1. Miller GD, Groziak SM, Dirienzo D: Age considerations in nutri-
ent needs for bone. J Am Coll Nutr 15:553–555, 1996.
2. Cooper C, Atkinson EJ, Hensrud DD, Wahner HW, O’Fallon WM,
Riggs BL, Melton LJ: Dietary protein intake and bone mass in
women. Calcif Tissue Int 58:320–325, 1996.
3. Teegarden D, Lyle RM, McCabe GP, Proulx WR, Michon K,
Knight AP, Johnston CC, Weaver C: Dietary calcium, protein, and
phosphorus are related to bone mineral density and content in
young women. Am J Clin Nutr 68:749–754, 1998.
4. Freskanich D, Willett WC, Stampfer MJ, Colditz GA: Protein
consumption and bone fractures in women. Am J Epidemiol 143:
472–479, 1996.
Phosphorus intake, similar to potassium and to protein, was
also a significant predictor of both TB BMD and spine BMD,
but not hip BMD. Another study found a similar relation
between phosphorus and BMD, using the same FFQ to gather
dietary intake data [3]. There has been concern about high
phosphorus intakes as current intakes in the U.S. are very high
while at the same time calcium intakes are very low [11].
Furthermore, an individual’s actual phosphorus intake may be
higher than estimated due to inaccurate nutrient composition
tables for phosphorus values of processed foods [44]. To com-
pound the issue, an FFQ does not differentiation between the
use of processed foods (high in phosphates) or unprocessed
foods. For example, there is only one descriptor for soft drinks,
which can be phosphate-containing cola or not (and the phos-
phorus analysis of this food was 0 mg). Therefore, the useful-
ness of an FFQ for determining phosphorus in diet-disease
relationships is limited. Protein and phosphorus intakes were
highly correlated, even when adjusted for energy intake (r ϭ
0.761, p Ͻ 0.001). Generally, foods high in protein are also
high in phosphorus, so that dietary phosphorus intakes mea-
sured by the FFQ are most likely reflecting protein intake and
are not likely to be a true reflection of phosphorus in the diet.
Our study suggests that for adult men, a dietary pattern of
adequate calcium intake, moderate protein, and generous po-
tassium contributes to higher BMDs than a pattern where
calcium is adequate but protein and potassium are low. Because
of the collinearity in intake among these nutrients it is difficult
to separate individual effects. Nevertheless, the former dietary
pattern appears to be beneficial to BMD maintenance in men
and can be achieved by incorporating fruit, vegetables and
dairy products into the diet.
5. Sellmeyer DE, Stone KL, Sebastian A, Cummings SR, for the
Study of Osteoporotic Fractures Research Group: A high ratio of
dietary animal to vegetable protein increases the rate of bone loss
and the risk of fracture in postmenopausal women. Am J Clin Nutr
73:118–122, 2001.
6. Hannan MT, Tucker KL, Dawson-Hughes B, Cupples LA, Felson
DT, Kiel DP: Effect of dietary protein on bone loss in elderly men
and women. The Framingham Study. J Bone Miner Res 15:2504–
2512, 2000.
7. Kerstetter JE, Svastisalee CM, Caseria DM, Mitnick, ME, Insogna
KL: A threshold for low-protein-diet-induced elevations in para-
thyroid hormone. Am J Clin Nutr 72:168–173, 2000.
8. New SA, Bolton-Smith C, Grubb DA, Reid DM: Nutritional in-
fluences on bone mineral density: a cross-sectional study in pre-
menopausal women. Am J Clin Nutr 65:1831–1839, 1997.
9. Tucker KL, Hannan MT, Chen H, Cupples LA, Wilson PWF, 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 69:727–736, 1999.
10. New SA, Robins SP, Campbell MK, Martin JC, Garton MJ,
Bolton-Smith C, Grubb DA, Lee SJ, Reid DM: 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 71:142–151, 2000.
11. Calvo MS, Park YK: Changing phosphorus content of the U.S.
diet: potential for adverse effects on bone. J Nutr 126:1168S–
1180S, 1996.
12. Elmstahl S, Gullberg B, Janzon L, Johnell O, Elmstahl B: In-
creased incidence of fractures in middle-aged and elderly men with
low intakes of phosphorus and zinc. Osteoporos Int 8:333–340,
1998.
13. Mirwald RL: The Saskatchewan Growth and Development Study.
In Ostyn M, Beunen G, Simons J (eds): “Kinanthropometry II,”
International Series on Sport Sciences, vol. 9. Baltimore: Univer-
sity Park Press, 1978.
14. Sandler RB, Slemenda CW, Laporte RE, Cauley JA, Schramm
MM, Barresi ML, Kriska AM: Postmenopausal bone mineral den-
sity and milk consumption in childhood and adolescence. Am J
Clin Nutr 42:270–274, 1985.
ACKNOWLEDGMENT
This work was supported by a grant from the Canadian
Fitness and Lifestyle Research Institute (#961R001) to RLM
and RAF. Preliminary dietary interviews and dietary assess-
ments were conducted by Carolyn Medernach and Michelle
Mackenzie.
15. Block G, Woods M, Potosky A, Clifford C: Validation of a
self-administered diet history questionnaire using multiple diet
records. J Clin Epidemiol 43:1327–1335, 1990.
16. Block G, Thompson FE, Hartman AM, Larkin FA, Guire KE:
408
VOL. 21, NO. 5

Diet and Bone in Men
Comparison of two dietary questionnaires validated against mul-
tiple dietary records collected during a 1-year period. J Am Diet
Assoc 92:686–693, 1992.
Dequeker J, Dilsen G, Gennari C, Vaz AL, Lyritis G, Mazzuoli G,
Miravet L, Passeri M, Perez Cano R, Rapado A, Ribot C: Risk
factors for hip fractures in men from Southern Europe: The ME-
DOS study. Osteoporosis Int 9, 45–54, 1999.
17. Hartman AM, Block G, Chan W, Williams J, McAdams M, Banks
WL, Robins A: Reproducibility of a self-administered diet history
questionnaire administered three times over three different sea-
sons. Nutr Cancer 25:305–315, 1996.
31. Elmstahl S, Gullberg B: Bias in diet assessment methods—
consequences of collinearity and measurement errors on power and
observed relative risks. Int J Epidemiol 26:1071–1079, 1997.
32. Heaney RP: Calcium, dairy products and osteoporosis. J Am Coll
Nutr 19:83S–99S, 2000.
18. Ross WD and Marfell-Jones MJ: Kinanthropometry. In MacDou-
gall JD, Wenger HA, Green JJ (eds): “Physiological Testing of the
High-Performance Athlete.” Champaign: Human Kinetics, pp
223–308, 1990.
33. Matkovic V, Heaney RP: Calcium balance during human growth:
evidence for threshold behavior. Am J Clin Nutr 55:992–996,
1992.
19. McKay H, Bailey DA, Wilkinson A, Houston CS: Familial com-
parison of bone mineral density at the proximal femur and lumbar
spine. Bone Miner 24:95–107, 1994.
34. Dawson-Hughes B, Harris SS: Calcium intake influences the as-
sociation of protein intake with rates of bone loss in elderly men
and women. Am J Clin Nutr 75:773–779, 2002.
20. Sallis JF, Haskell WL, Wood PD, Fortmann SP, Rogers T, Blair
SN, Paffenbarger Jr RS: Physical activity assessment methodology
in the Five-City Project. Am J Epidemiol 121:91–106, 1985.
21. Willett WC, Stampfer MJ: Total energy intake: implications for
epidemiologic analyses. Am J Epidemiol 124:17–27, 1986.
22. Kleinbaum DG, Kupper LL, Muller KE: “Applied Regression
Analysis and Other Multivariate Methods,” 2nd ed. Belmont, CA:
Duxbury Press, pp 206–217, 1988.
35. Lemann J: Relationship between urinary calcium and net acid
excretion as determined by dietary protein and potassium: a re-
view. Nephron 81:18–25, 1999.
36. Morris RC, Schmidlin O, Tanaka M, Forman A, Frassetto L,
Sebastian A: Differing effects of supplemental KCl and KHCO3:
Pathophysiological and clinical implications. Semin Nephrol 19:
487–493, 1999.
37. Lemann J, Pleuss JA, Gray RW: Potassium causes calcium reten-
tion in healthy adults. J Nutr 123:1623–1626, 1993.
23. Looker AC, Orwoll ES, Johnston CC, Lindsay RL, Wahner HW,
Dunn WL, Calvo MS, Harris TB, Heyse SP: Prevalence of low
femoral bone density in older U.S. Adults from NHANES III.
J Bone Miner Res 12:1761–1768, 1997.
38. Green TJ, Whiting SJ: Potassium bicarbonate reduces high protein-
induced hypercalciuria in adult men. Nutr Res 7:991–1002, 1994.
39. Ilich JZ, Kerstetter JE: Nutrition in bone health revisited: a story
beyond calcium. J Am Coll Nutr 19:715–737, 2000.
40. Remer T, Manz F: Estimation of the renal net acid excretion by
adults consuming diets containing variable amounts of protein.
Am J Clin Nutr 59:1356–1361, 1994.
24. Health and Welfare Canada: “Nutrition Recommendations: The
Report of the Scientific Review Committee.” Ottawa: Supply and
Services Canada, 1990.
25. Institute of Medicine, Food and Nutrition Board: “Dietary Refer-
ence Intakes: Calcium, Magnesium, Phosphorus, Vitamin D, and
Fluoride.” Washington, DC: National Academy Press, 1997.
26. National Research Council: “Recommended Dietary Allowances,”
10th ed. Washington DC: National Academy Press, 1989.
27. Health and Welfare Canada: “Canada’s Food Guide to Healthy
Eating.” Ottawa: Minister of Supply and Services, 1992.
28. Barger-Lux MJ, Heaney RP, Packard PT, Lappe JM, Recker RR:
Nutritional correlates of low calcium intake. Clin Appl Nutr 2:39–
44, 1992.
41. Kerstetter JE, Caseria DM, Mitnick ME, Ellison AF, Gay LF,
Liskov TAP, Carpenter TO, Insogna KL: Increased circulating
concentrations of parathyroid hormone in healthy, young women
consuming a protein-restricted diet. Am J Clin Nutr 66:1188–1196,
1997.
42. Kerstetter JE, O’Brien KO, Insogna KL: Dietary protein affects
intestinal calcium absorption. Am J Clin Nutr 68:859–865, 1998.
43. Kerstetter JE, Looker AC, Insogna KL: Low dietary protein and
low bone density. Calcif Tissue Int 66:313, 2000.
29. Nordstrom P, Nordstrom R, Lorentzon R: Correlation of bone
density to strength and physical activity in young men with a low
or moderate level of physical activity. Calcif Tissue Int 60:332–
337, 1997.
44. Calvo MS: Dietary phosphorus, calcium metabolism and bone. J
Nutr 123:1627–1633, 1993.
30. Kanis J, Johnell O, Gullberg B, Allander E, Elffors L, Ranstam J,
Received September 24, 2001; revision accepted May 24, 2002.
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
409