Whey Protein Supplementation and Higher Total Protein Intake Do Not Influence Bone Quantity in Overweight and Obese Adults Following a 36-Week Exercise and Diet Intervention^,,,

Abstract

Background: Controversy exists concerning the effects of higher total protein intake (TPro) on bone health, which may be associated with reduced bone mineral density (BMD). However, whey protein (WP) may induce bone formation because of its basic component, milk basic protein.

Objective: This study assessed the effects of WP supplementation, TPro, and change in TPro (postsupplementation − presupplementation) on BMD and bone mineral content (BMC; total body, lumbar spine, total femur, and femoral neck) in overweight and class I obese middle-aged adults following an exercise intervention.

Methods: This analysis used data from a double-blind, randomized, placebo-controlled 36-wk WP supplementation trial, wherein participants consumed a 1.7-MJ (400-kcal) supplement (0, 20, 40, or 60 g WP/d) along with their otherwise unrestricted diet while participating in a resistance and aerobic exercise intervention (3 d/wk). TPro was the summation of WP and habitual dietary intakes (4-d food record). Statistical analyses for WP were based on group and bone data [n = 186, 108 women; mean ± SD age: 49 ± 8 y; body mass index (BMI; in kg/m²): 30.1 ± 2.8], whereas TPro was based on dietary and bone data (n = 113, 70 women; age 50 ± 8 y; BMI 30.1 ± 2.9).

Results: WP supplementation, regardless of dose, did not influence BMD or BMC following the intervention. By using a multiple linear regression model, TPro (expressed as g/d or g · kg⁻¹ · d⁻¹) and change in TPro (expressed as g/d) were not associated with responses over time in total or regional BMD or BMC. By using a cluster analysis approach [<1.0 (n = 41), 1.0–1.2 (n = 28), and ≥1.2 g · kg⁻¹ · d⁻¹ (n = 44)], TPro was also not associated with responses in total or regional BMD or BMC over time.

Conclusion: WP supplementation and total dietary protein intake did not negatively or beneficially influence bone quantity in overweight and obese adults during a 9-mo exercise intervention. This trial was registered at clinicaltrials.gov as NCT00812409.

Introduction

An indispensable macronutrient, dietary protein and its amino acids serve as the functional components for total body structure, regulation, and locomotion. Structurally, protein constitutes 20–25% of total bone mass of which type 1 collagen fibers make up the majority (1). Sufficient dietary protein is essential for the formation and structural maintenance of bone because the body is unable to reutilize amino acids for collagen synthesis following bone turnover (1). As such, protein intakes below the current RDA, 0.8 g · kg⁻¹ · d⁻¹ (2), are shown to delay bone growth (3, 4), decrease bone mass (5, 6) and increase the risk of osteoporosis and skeletal fracture (7), particularly in older populations (8, 9). Although these effects of inadequate protein intake are well recognized, the effects of higher total protein intake (TPro)⁷ on bone (>0.8 g · kg⁻¹ · d⁻¹) are still actively debated and may be dependent on a number of factors, including quantity and source of dietary protein, calcium intake, energy balance (10), and physical activity (11).

The acid-ash hypothesis (12) theorizes that the consumption of excess dietary protein acidifies the environment surrounding bone, through the oxidation of sulfur-containing amino acids and phosphoproteins, leading to bone demineralization (8, 13), diet-induced hypercalciuria (14, 15), and a negative calcium balance (16–18). This hypothesis put forth by Sherman (12) nearly a century ago has long been proposed as the mechanism behind a detrimental effect of dietary protein on bone. Although increasing intakes of dietary protein are associated with increased urinary calcium excretion (19, 20), dual-stable calcium isotope studies have shown that higher TPro increases intestinal calcium absorption (21, 22). Therefore, the observed hypercalciuria effect of increasing protein intake is often explained by this enhanced calcium absorption, which subsequently does not alter total calcium balance (23). Despite dietary protein’s lack of effect on total calcium balance, prospective cohort studies have shown that higher TPro is associated with higher bone mineral density (BMD) (24–26). A 2009 meta-analysis of randomized controlled trials also supports this claim, showing a beneficial effect of higher TPro on spine BMD in adults (27). However, a consensus on the effect of dietary protein on bone is still lacking.

Conflicting outcomes reported in the literature may be partially explained by the predominate source of protein in the diet. Different sources of dietary protein contain varying amino acid compositions as well as additional micronutrient, bioactive, and nonnutritive components, which collectively may differentially influence bone health (10, 28–43). Whey protein (WP) is 1 of 2 proteins contained in dairy and is a commonly used protein supplement. The basic protein fraction of WP, known as milk basic protein (MBP), was shown to elicit an anabolic response in bone (44–49). Kininogen fragment 1.2 and cystatin C are considered the anabolic components of MBP, causing a stimulation of bone formation and a suppression of bone resorption (44, 46, 48, 50). In addition to MBP and its anabolic effects on bone, WP is also shown to increase serum insulin-like growth factor I (IGF-I) concentrations (50, 51), a hormone important in osteoblast differentiation (52). These inherent properties of WP highlight the differential effects of protein source on bone and suggest potential therapeutic effects.

Currently, no study to our knowledge has investigated the dose-response effects of WP supplementation or the effects of TPro and change in TPro in conjunction with exercise training on measurements of bone quantity in overweight and obese middle-aged adults. Therefore, a secondary analysis of data from a 36-wk exercise training and WP supplementation trial in middle-aged, overweight, and class I obese adults (53) was conducted to determine whether WP supplementation, TPro, or change in TPro elicits an anabolic or catabolic response on measurements of bone quantity. We hypothesized that WP supplementation would increase bone quantity in a dose-dependent manner, and higher TPro and positive changes in TPro would increase bone quantity.

Methods

Study design.

This investigation utilized data from a double-blind, placebo-controlled, community-based, randomized 36-wk intent-to-treat study conducted from 2007 to 2010 (53). After completing a baseline assessment, participants were randomly assigned to consume 1 of 4 WP supplements and participated in an exercise training program throughout the 36-wk intervention. Measurements were conducted at baseline (week 0) and postintervention (week 36).

Participants.

Overweight and class I obese adults were recruited to participate in this study. Study inclusion was based on the following criteria: 1) man or woman aged 35–65 y, 2) total body weight <136 kg, 3) BMI (in kg/m²) between 26 and 35, 4) fasting glucose <110 mg/dL, 5) blood pressure <160/100 mm Hg, 6) plasma total cholesterol <260 mg/dL, 7) LDL cholesterol <160 mg/dL, 8) triacylglycerol <400 mg/dL, 9) not currently or previously following a weight-loss diet or other special or unbalanced diet (in the past 6 mo), and 10) <2 h/wk of habitual resistance or aerobic exercise training in the past 6 mo. The Purdue University Biomedical Institutional Review Board approved the study protocol. Participants taking medications for elevated blood pressure, reduced HDL, or elevated triacylglycerol, including statins, were included in the study because their medication use did not affect the final analysis. Each participant signed an informed-consent form before enrollment and received a monetary stipend for participating in the study. Five hundred seventy-three potential participants were screened, 327 were randomly assigned (207 women; age 48 ± 8 y; BMI 30.4 ± 2.7), and 186 participants (108 women; age 49 ± 8 y; BMI 30.1 ± 2.8) completed the study. Figure 1 shows the flow of participants for WP and TPro assessments. Supplemental Table 1 shows the baseline characteristics of those who were randomly assigned (n = 327), those who completed (n = 186), and those who failed to complete (n = 141) the 36-wk intervention. Further details relating to recruitment, retention, and compliance were published previously (53).

FIGURE 1.

Consolidated Standards of Reporting Trials flow diagram. Flow of participants and study analysis. WP, whey protein.

WP supplementation and total dietary protein intake.

Participants were randomly assigned (Microsoft Excel 2010, randomization function) to 1 of 4 iso-energetic supplementation groups: 1) 0 g WP/d, placebo control; 2) 20 g WP/d; 3) 40 g WP/d; or 4) 60 g WP/d. A greater number of participants were assigned to the 0- and 20-g WP/d groups than to the 40- and 60-g WP/d groups [group assignment ratio of 16:16:5:5 participants in the 4 groups, respectively, based on sample size estimate for the primary outcome of change over time in lean body mass (53)]. Because of these limitations in group sample sizes, the 40-g (n = 21) and 60-g (n = 25) WP supplementation groups were combined to form one ≥40-g WP supplementation group for this secondary analysis (0 g WP/d: n = 68; 20 g WP/d: n = 72; ≥40 g WP/d: n = 46).

Supplements were given in powder form and were manufactured and provided by Innovative Food Processors, Inc. (Faribault). The nutritional composition of the WP supplement and time of consumption were previously described (53). WP is a high-quality protein source (Protein Digestibility–Corrected Amino Acid Score = 1.00) (54), and Supplemental Table 2 describes the amino acid profile of each WP supplement. Participants were informed that the supplements provided 1.7 MJ/d (400 kcal/d) of energy but were not made aware of WP quantity or counseled to purposefully alter their usual eating behaviors. Four-day food records (3 weekdays and 1 weekend day) were completed at baseline, week 16, and postintervention (week 36) to calculate habitual daily energy and macronutrient intakes (Nutritionist Pro, version 1.3.36; First DataBank). Age- and sex-specific Schofield equations were used to estimate basal metabolic rates (55). Baseline and postintervention food records were considered valid only if they fell within the lower and upper 95% CIs established by Goldberg’s cutoff (56). Data from 113 of the 186 participants (0 g WP/d: n = 45; 20 WP g/d: n = 39; ≥40 WP g/d: n = 29) met this criterion, and their dietary data plus WP group assignment were used to calculate TPro.

To document differences in protein intake, 2 consecutive 24-h total urine collections were obtained at baseline and postintervention. Twenty-four–hour urine samples (weight divided by specific gravity; Digital Probe Refractometer; Misco Products Division) were aliquoted and stored at −20°C for subsequent 24-h urinary urea nitrogen (UUN) analyses (COBAS Integra 400; Roche Diagnostic Systems) to document and assess group-specific and individual TPro. All participants received a daily multivitamin/mineral tablet (Centrum; Pfizer Inc.) and calcium supplement (Citrical, 400 mg Ca/d; Bayer Healthcare, LLC).

Exercise training and testing.

All participants engaged in dynamic resistance and aerobic exercise training throughout the 36-wk intervention. Each participant was provided with a 36-wk gym membership for a local fitness facility located in Lafayette/West Lafayette, Indiana. Participants engaged in resistance exercise training 2 d/wk and aerobic exercise training 1 d/wk. The specific exercises and the progression toward the prescribed exercise training intensity for both resistance and aerobic exercises were previously described (53).

Body-composition measurements.

Body composition, BMD, and bone mineral content (BMC) were measured by DXA (GE LUNAR Prodigy with EnCORE software version 5.60) at baseline and postintervention. BMD and BMC were calculated for 4 skeletal sites [total body, lumbar spine (L1–L4), total femur, and femoral neck] and used for these analyses. CV was calculated from a sample of middle-aged adults (n = 13) and showed low levels of variability (total BMD, 0.603% CV; lumbar BMD, 0.678% CV; total femur, 0.512% CV; femur neck, 0.773% CV). Only participants with full DXA data were included in analyses.

Power calculations.

Given the nature of a secondary analysis, power calculations were not specifically conducted for the assessment of BMD and BMC. Original power calculations are published (20) and were done to determine the effects of this intervention on whole-body lean body mass. Retrospectively, it was calculated that a sample size of 64 participants in the placebo and WP supplement groups would allow for the detection of a 0.004 difference in femoral neck BMD at 80% power (2-sided significance) by using data from a previous WP-supplementation trial with BMD measurements at 9 mo (57).

Statistical analyses.

Because of differences in sample size populations (n = 186: 108 women or 58% women compared with n = 113: 70 women or 62% women), separate statistical analyses were conducted to assess the effects of WP supplementation (n = 186) and TPro or change in TPro (n = 113) on BMD and BMC at different skeletal sites (total body, lumbar spine, total femur, femoral neck). Both databases required the log transformation of lumbar spine, femoral neck, and total femur BMC data to comply with normality as assessed by the Shapiro-Wilk test (58). Outliers were detected in both databases and were excluded according to the outlier labeling rule (59).

Baseline 1-factor ANOVA of participant demographics, anthropometrics, TPro, BMD, and BMC values revealed differences between men and women, pre- and postmenopausal women, and middle-aged (≤50 y) and older adults (≥51 y) such that the lowest measurements of bone quantity were observed in older (postmenopausal) women. Therefore age, sex, and menopausal status were included as covariates in subsequent analyses. To assess compliance to WP supplementation and protein cluster allotment, repeated-measures ANOVA was performed to assess UUN and blood urea nitrogen time-by-group effects.

Repeated-measures ANOVA was performed to determine the main effects of WP supplementation and time on BMD and BMC variables. By defining TPro as a continuous variable, multiple linear regression analyses were used to assess the effect of TPro (postintervention: expressed as g/d, g · kg body weight⁻¹ · d⁻¹, or % energy from protein) and change in TPro (postintervention − baseline; expressed as g/d, g · kg body weight⁻¹ · d⁻¹, or % energy from protein) on changes in BMD and BMC variables (postintervention − baseline). Considering that previous studies showed a potential threshold effect of 1.0 g dietary protein · kg⁻¹ · d⁻¹ on changes in lean soft tissue among middle-aged and older adults engaging in diet and exercise interventions (60), a cluster analysis approach was also used to determine the effects of TPro on BMD and BMC variables. Tertiles of TPro were created with <1.0 g · kg⁻¹ · d⁻¹ classified as normal protein intake (NP, n = 41), 1.0–1.2 g · kg⁻¹ · d⁻¹ as moderately high protein intake (MHP, n = 28), and >1.2 g · kg⁻¹ · d⁻¹ as high protein intake (HP, n = 44). By defining TPro as a categorical variable, repeated-measures ANOVAs were performed to determine the main effects of protein cluster and time on BMD and BMC variables. Statistical analyses were performed by using SPSS (IBM SPSS Statistics for Windows, version 22.0, 2013). Data are presented as means ± SDs with a P value of <0.05 considered statistically significant.

Results

Participant characteristics.

The baseline participant characteristics for both the WP and TPro assessments are presented in Table 1. Both data sets were composed of mostly overweight and obese older adult women with an average baseline protein intake above the current RDA (0.95 ± 0.31 g · kg⁻¹ · d⁻¹) (2). There were no differences among WP supplementation groups or TPro clusters at baseline for height, body mass, BMI, TPro, energy intake, or total-body BMD (Table 1). Compliance to WP supplementation and exercise training intervention are reported elsewhere (53); 87% of completers consumed >50% of WP supplements and 98% of participants who completed the study completed >70% of the exercise sessions. Additionally, both blood urea nitrogen (P = 0.002) and UUN (P = 0.019) increased over time with higher amounts of WP supplementation and TPro, and TPro did not differ between week 16 and week 36 of the intervention (P = 0.92).

TABLE 1.

WP supplementation and total protein intake cluster baseline subject characteristics¹

	WP supplementation, g WP/d (n = 186)			Total protein intake, g · kg−1 · d−1 (n = 113)
	0	20	≥40	NP, <1	MHP, 1–1.2	HP, >1.2
Sample size, n	68	72	46	41	28	44
Women, n (%)	38 (56)	41 (57)	29 (63)	23 (56)	19 (68)	28 (64)
Postmenopausal, n (%)	20 (29)	14 (19)	12 (26)	13 (32)	8 (29)	10 (23)
Age, y	50 ± 7	48 ± 8	49 ± 8	52 ± 7	50 ± 9	48 ± 8
Body mass, kg	87.5 ± 12.6	88.5 ± 12.6	88.5 ± 13.4	90.2 ± 11.7	89.6 ± 11.7	83.7 ± 11.3
Height, m	1.71 ± 0.1	1.70 ± 0.1	1.70 ± 0.1	1.71 ± 0.1	1.72 ± 0.1	1.68 ± 0.1
BMI, kg/m2	29.7 ± 2.8	30.4 ± 2.7	30.3 ± 2.9	30.6 ± 2.8	30.0 ± 3.1	29.5 ± 2.9
Lean mass, kg	50.6 ± 10.9	49.7 ± 9.8	49.9 ± 10.9	50.6 ± 9.9	50.2 ± 9.2	48.2 ± 9.7
Fat mass, kg	34.0 ± 6.0	35.3 ± 5.8	35.1 ± 8.1	35.8 ± 7.8	36.7 ± 8.0	32.7 ± 6.0
Total body BMD, g/cm2	1.24 ± 0.13	1.22 ± 0.11	1.22 ± 0.09	1.24 ± 0.13	1.21 ± 0.14	1.20 ± 0.08
Lumbar spine BMD, g/cm2	1.25 ± 0.17	1.24 ± 0.15	1.24 ± 0.16	1.26 ± 0.18	1.24 ± 0.17	1.20 ± 0.15
Total femur BMD, g/cm2	1.07 ± 0.13	1.05 ± 0.13	1.07 ± 0.12	1.07 ± 0.14	1.05 ± 0.13	1.05 ± 0.13
Femoral neck BMD, g/cm2	1.02 ± 0.12	1.00 ± 0.11	1.01 ± 0.11	1.01 ± 0.12	1.00 ± 0.15	1.00 ± 0.12
Protein intake, g/d	87.0 ± 29.0	80.5 ± 24.7	81.2 ± 31.0	86.9 ± 19.7	89.9 ± 21.2	90.7 ± 24.8
Protein intake, g · kg−1 · d−1	1.01 ± 0.35	0.92 ± 0.27	0.92 ± 0.32	0.98 ± 0.24	1.01 ± 0.23	1.08 ± 0.26
Energy from protein, %	16.0 ± 4.9	15.7 ± 4.1	15.8 ± 4.7	15.7 ± 2.7	16.2 ± 3.2	16.6 ± 3.4
Total energy intake, MJ/d	9.35 ± 2.34	8.76 ± 2.35	8.76 ± 2.75	9.37 ± 1.96	9.40 ± 1.89	9.26 ± 1.50

¹Values are means ± SDs unless otherwise indicated. BMD, bone mineral density; HP, high protein; MHP, moderately high protein; NP, normal protein; WP, whey protein.

Changes in soft tissue among completers over the 36-wk intervention.

Among the participants who completed the study (n = 186), total body mass did not change over time. Lean body mass increased by 0.95 ± 1.33 kg (P < 0.001), and fat mass decreased by 0.89 ± 3.08 kg (P < 0.001). WP did not affect changes in soft tissue over the 36-wk intervention; however, TPro was negatively associated with changes in total body mass (P = 0.016) and fat mass (P = 0.021).

Effect of WP supplementation and TPro cluster on bone quantity over the 36-wk intervention.

Changes in BMD and BMC measurements in each WP and TPro group following the 36-wk intervention are presented in Table 2. Baseline and postintervention BMD and BMC values for the WP and TPro sample size populations are also presented in Supplemental Table 3. Both WP supplementation and TPro cluster, regardless of dose or TPro, did not affect measurements of BMD or BMC over the 36-wk intervention (Table 2, Supplemental Table 4). Independent of WP supplementation and TPro, decreases in lumbar spine BMD in the WP analysis (P = 0.049, n = 186) and increases in total femur BMD in the TPro analysis (P = 0.017, n = 113) were observed over time.

TABLE 2.

Total population and group-specific changes in BMD and BMC measurements following a 36-wk exercise and WP supplementation intervention in overweight and obese adults¹

	WP supplementation, g WP/d				Total protein intake, g · kg−1 · d−1
	Total (N = 186)	0 (n = 68)	20 (n = 72)	≥40 (n = 46)	Total (n = 113)	NP, <1 (n = 41)	MHP, 1–1.2 (n = 28)	HP, >1.2 (n = 44)
BMD, g/cm2 × 102
Total body	−0.08 ± 1.22	0.11 ± 1.21	−0.16 ± 1.23	−0.23 ± 1.22	0.01 ± 1.11	0.18 ± 0.85	<0.01 ± 1.24	−0.14 ± 1.24
Lumbar spine	−0.49 ± 0.342	−0.07 ± 3.32	−0.74 ± 3.32	−0.68 ± 3.35	−0.47 ± 3.50	−0.05 ± 4.33	−0.49 ± 3.38	−0.85 ± 2.65
Total femur	0.14 ± 1.13	0.23 ± 1.14	0.15 ± 1.08	−0.01 ± 1.19	0.29 ± 1.122	0.46 ± 1.05	0.06 ± 1.36	0.26 ± 1.01
Femoral neck	−0.15 ± 1.47	−0.14 ± 1.54	−0.17 ± 1.43	−0.15 ± 1.45	0.01 ± 1.49	−0.01 ± 1.65	−0.03 ± 1.43	0.05 ± 1.41
BMC, g
Total body	−1.66 ± 29.8	−1.27 ± 24.0	1.79 ± 31.0	−7.62 ± 34.9	−0.70 ± 30.8	5.43 ± 23.1	−7.61 ± 32.3	−2.02 ± 35.3
Lumbar spine	0.06 ± 3.41	0.47 ± 4.62	−0.26 ± 2.31	−0.04 ± 2.75	0.28 ± 3.98	1.13 ± 5.70	0.05 ± 3.33	−0.36 ± 1.71
Total femur	0.03 ± 0.50	0.01 ± 0.65	0.07 ± 0.35	−0.02 ± 0.44	0.06 ± 0.56	0.08 ± 0.76	0.04 ± 0.47	0.06 ± 0.36
Femoral neck	<0.01 ± 0.09	−0.01 ± 0.09	0.01 ± 0.09	−0.01 ± 0.07	0.01 ± 0.09	0.02 ± 0.11	<0.01 ± 0.08	<0.01 ± 0.07

¹Values are means ± SDs; change in values = postintervention − baseline. BMC, bone mineral content; BMD, bone mineral density; HP, high protein; MHP, moderately high protein; NP, normal protein; WP, whey protein.

²Significant main effect of time: repeated-measures ANOVA controlling age, sex, and menopausal status; P < 0.05.

TPro and change in TPro on changes in bone quantity.

Neither TPro nor change in TPro influenced changes in bone quantity measurements over the 36-wk intervention (Supplemental Figures 1 and 2, Table 3). Similar results occurred regardless of protein intake classification (expressed as g/d, % energy from protein, or g · kg⁻¹ · d⁻¹).

TABLE 3.

The effect of total protein intake or change in total protein intake on changes in bone quantity following a 36-wk exercise and whey protein supplementation intervention in overweight and obese adults (n = 113)¹

	Total protein intake		Change in total protein intake
	P	β (95% CI ✕ 102)	P	β (95% CI ✕ 102)
Change in BMD, g/cm2
Total body	0.70	−0.04 (−0.008 to 0.006)	0.74	−0.03 (−0.006 to 0.004)
Lumbar spine	0.54	−0.06 (−0.028 to 0.015)	0.07	−0.17 (−0.028 to 0.001)
Total femur	0.98	0.01 (−0.007 to 0.007)	0.96	−0.01 (−0.005 to 0.005)
Femoral neck	0.12	0.15 (−0.002 to 0.0164)	0.24	0.11 (−0.003 to 0.01)
Change in BMC, g
Total body	0.44	−0.07 (−26.2 to 11.5)	0.87	0.02 (−21.1 to 14.3)
Lumbar spine	0.64	−0.05 (−3.1 to 1.9)	0.31	−0.10 (−2.6 to 0.8)
Total femur	0.50	−0.06 (−0.5 to 0.2)	0.15	−0.13 (−0.4 to 0.1)
Femoral neck	0.55	0.06 (−0.038 to 0.1)	0.52	0.06 (−0.026 to 0.05)

¹Values are multiple linear regression controlling for age, sex, and menopausal status; change in protein intake = postintervention − baseline. Similar results were seen if total protein intake was categorized as percentage of total energy or grams per kilogram per day. BMC, bone mineral content; BMD, bone mineral density.

Discussion

Dietary protein is an essential component in supporting skeletal growth, development, and adaptation. However, the quantity of dietary protein required to maximize these skeletal benefits without negatively influencing bone metabolism is still an area of debate. The purpose of this secondary analysis was to assess the dose-response effects of WP supplementation and the effects of TPro and change in TPro on changes in bone quantity (BMD and BMC) following a 36-wk diet and exercise intervention. Contrary to our hypotheses, increasing intakes of WP, higher TPro, and a greater change in TPro did not influence measurements of bone quantity in overweight and class I obese middle-aged adults following a 9-mo exercise and diet intervention. Additional retrospective analyses using the UUN-to–urinary creatinine ratio and change in this ratio as surrogate markers of TPro also saw a similar lack of effect on changes in bone quantity. The WP results from the current study are in agreement with 2 previous WP supplementation trials (51, 57), while the TPro results are inconsistent with results from prospective cohort studies (24–26) and a meta-analysis of randomized controlled trials showing a beneficial effect with increased TPro on bone quantity (27).

Both Zhu et al. (51) and Kerstetter et al. (57) assessed the effects of WP supplementation on bone quantity and, similar to the current study, found neither a deleterious nor a beneficial effect. Over a 2-y period, Zhu and colleagues (51) provided 30 g WP isolate/d before breakfast to 179 overweight older women (74.3 ± 2.7 y, BMI ∼27) and assessed changes in bone quantity through DXA and quantitative computed tomography. Although Zhu and colleagues (51) noted increases in IGF-I concentrations with WP supplementation and decreases in total hip and femoral neck BMD among all participants over time, there were no effects of WP supplementation on bone quantity. In a similar experimental design and study population, Kerstetter and colleagues (57) gave 45 g WP isolate/d for 18 mo to 121 overweight older men and women (∼70 y, BMI ∼26) and assessed changes in bone quantity through DXA and quantitative computed tomography. Akin to the results from Zhu and colleagues (51), increases in IGF-I concentrations were observed with WP supplementation along with decreases in total hip and femoral neck BMD plus increases in lumbar spine BMD over time (57). However, similar to the results for Zhu and colleagues (51), WP supplementation did not differentially influence bone quantity in comparison with placebo. The present study further supports and expands on this lack of effect in a different population, with a different experimental design, and with varying doses of both WP and TPro.

In contrast to the 2 previous studies (51, 57), the present study’s population was composed of middle-age overweight and obese adults who were on average 50 y of age. With evidence showing an increase in IGF-I concentrations (50, 51, 57), intestinal calcium absorption (21, 22), and the retention or improvement in BMD (10, 28–43), increasing WP supplementation and/or TPro during this period of rapid bone loss could differentially influence bone metabolism and provide a more measurable change in bone quantity in comparison with older adults in whome the rate of bone loss has decreased (61). Furthermore, in contrast to the 2 previous studies (51, 57), higher concentrations of fat mass in the current study’s population provide a metabolically different environment, which not only influences bone metabolism via increased concentrations of leptin (62) and inflammatory cytokines (63) but potentially the effects of WP supplementation and/or TPro on bone quantity. Collectively, these age- and obesity-related differences provide a novel environment to test the study’s hypotheses. However, despite these differences, results from this study mirror the 2 previous studies and their null effect.

This lack of an effect on bone quantity was even present following 9 mo of exercise training, in which participants engaged in 2 d resistance training and 1 d aerobic training/wk. Given the mechanical stimulation of exercise training and previous studies showing a differential effect on bone quantity with exercise training alone (64), the combination of WP supplementation and/or increased TPro with exercise training would theoretically provide an alternative result. Yet, even in the presence of exercise training and increasing amounts of WP supplementation (20–60 g/d), the current study’s results are similar to the 2 previous studies, further supporting a neutral effect of WP and/or TPro on bone quantity. In contrast, prospective (65, 66) and animal studies (67) have cited beneficial effects of WP or TPro on bone quantity. These discrepancies in the literature may partially be explained by the composition of the WP supplement and the length of the study intervention.

The WP supplement used in the present study was composed of an 80% WP concentrate. A study by Mullins and Sinning (68) investigating the short-term effects of WP supplementation and resistance exercise on bone turnover markers used a similar supplement. The authors revealed that neither resistance training nor WP supplementation evoked a skeletal response in young, healthy females. The results of the present study support these findings over a longer duration. In contrast, the majority of remaining studies investigating the effects of WP on bone quantity have supplemented participants with the basic fraction of whey, MBP (44–46, 48, 49, 69). MBP and its bioactive components, kininogen fragment 1.2 and cystatin C, were previously shown to elicit several beneficial effects on bone including increased bone formation and the suppression of bone resorption (46, 48, 50). MBP is a component of WP that is formed following the coagulation of dairy milk in the cheese-making process. To isolate MBP, WP must undergo fractionation followed by several rounds of processing to refine its composition and remove any acidic milk proteins (49). The 80% WP concentrate used in the current study displays a slightly different amino acid profile compared with the MBP. Not only does the WP contain the acidic (low isoelectric point) WP compounds, but it also has a lower concentration of lysine and arginine than MBP, which are both shown to promote osteoblast activity (70). Therefore, discrepancies between previously conducted studies with MBP and the current study’s 80% WP concentrate may be explained by differences in amino acid and bioactive compositions.

The bone remodeling transient period is an important consideration for any study designed to isolate the effects of diet and exercising training on measurements of bone quantity, particularly when considering the appropriate length of the study intervention. Under normal circumstances, bone remodeling is a complex series of tightly regulated sequential steps characterized by bone resorption, formation, and then mineralization. Measuring bone quantity at different time points within this perpetual cycle of bone adaptation will therefore inherently display different results. This underlines the importance of allowing a full remodeling cycle to occur before taking a measurement, particularly when designating the appropriate length of an intervention with an exercise component. The bone remodeling transient period is a temporary alteration in the bone remodeling cycle and the balance between bone resorption and bone formation (71), which can be initiated by exercise training (72). Eventually, this bone remodeling transient period will stabilize and develop a new remodeling cycle, which will then be maintained for the entire duration of the study (71, 72). Given that this transient period is believed to persist for ∼10 mo (72), to isolate the effects of diet and/or exercise training on bone quantity it is believed that study durations must extend >10 mo. Although the current study was a randomized controlled trial, in terms of determination of a true bone quantity response it was relatively short at only 9 mo in length. Techniques such as dual-stable calcium isotope kinetics and the regular assessment of bone formation and resorption biomarkers would provide beneficial and dynamic information regarding bone metabolism in these relatively short-term interventions. Although such techniques would provide evidence for future clinical changes in bone quantity, taking into account the bone remodeling transient period and the gradual pace by which changes in bone quantity actually occur, the need for longer-term trials is reemphasized.

Strengths of the current study include the large sample size; the randomized, double-blind, placebo controlled design; and the assessment of both total body and regional BMD and BMC skeletal sites. To our knowledge, the current study was also the first study to assess changes in bone quantity at various doses of WP supplementation and TPro during an exercise intervention. There are admittedly some limitations to the study. As a secondary analysis, results from the current study do not provide evidence of causality. The 9-mo WP supplementation and exercise training intervention are also within the bone remodeling transient period and therefore may not allow the intervention’s full effect to completely manifest. However, it is important to note the negligible, albeit statistically significant, changes in bone quantity observed over the 9-mo intervention. Lumbar spine BMD decreased over time, whereas total femur BMD increased over time. The decrease in lumbar spine BMD is on average only 10–20% of the proposed annual decrease in lumbar spine BMD (73, 74), whereas the 0.2% increase in total femur BMD is within the DXA imaging error rate (75). Therefore, such small alterations in bone quantity bring into question the clinical significance of these changes and whether the current study was originally powered to detect such a small change. The study’s exercise training component could explain this tempered effect and the observed alterations in bone quantity over the 36-wk intervention. In addition, the direct effect of exercise training on bone quantity via mechanical stimulation could have influenced or minimized the effect of WP or TPro on changes in bone quantity. Nutritional factors, such as dietary protein, are thought to elicit an indirect (e.g., via hormonal factors) or generalized effect on bone quantity (76), which in theory could be minimized by a more direct effect as with exercise training. However, despite these claims, the study by Josse et al. (77) further highlights the potential complimentary or synergistic effects of dietary protein and exercise training on bone quantity because overweight and obese middle-aged adults in the higher TPro group showed greater improvements in markers of bone metabolism than those in the normal TPro group during a diet- and exercise-induced weight-loss intervention.

In conclusion, neither WP supplementation nor higher TPro increased measurements of bone quantity in overweight and class I obese middle-aged adults following a 36-wk exercise and diet intervention. Alhough WP supplementation and increased TPro during exercise training may elicit beneficial effects on body composition, particularly soft tissue, it appears to have no measurable effect on bone quantity in overweight and class I obese middle-aged adults.