- Split View
-
Views
-
Cite
Cite
Marie Juul Ornstrup, Torben Harsløf, Thomas Nordstrøm Kjær, Bente Lomholt Langdahl, Steen Bønløkke Pedersen, Resveratrol Increases Bone Mineral Density and Bone Alkaline Phosphatase in Obese Men: A Randomized Placebo-Controlled Trial, The Journal of Clinical Endocrinology & Metabolism, Volume 99, Issue 12, December 2014, Pages 4720–4729, https://doi.org/10.1210/jc.2014-2799
- Share Icon Share
Metabolic syndrome (MetS) is associated with low-grade inflammation, which may harmfully affect bone. Resveratrol (RSV) possesses anti-inflammatory properties, and rodent studies suggest bone protective effects.
This study sought to evaluate effects of RSV treatment on bone in men with MetS.
The study was conducted at Aarhus University Hospital as a randomized, double-blinded, placebo-controlled trial assessing changes in bone turnover markers, bone mineral density (BMD), and geometry.
The study population comprised 74 middle-aged obese men with MetS recruited from the general community, of which 66 completed all visits. Mean age of participants was 49.3 ± 6.3 years and mean body mass index was 33.7 ± 3.6 kg/m2.
Oral treatment with 1.000 mg RSV (RSVhigh), 150mg RSV (RSVlow), or placebo daily for 16 weeks.
Prespecified primary endpoint was change in bone alkaline phosphatase (BAP).
BAP increased dose dependently with RSV (R = 0.471, P < .001), resulting in a significantly greater increase in BAP in the RSVhigh group compared with placebo at all time-points (week 4, 16.4 ± 4.2%, P < .001; week 8, 16.5 ± 4.1%, P < .001; week 16, 15.2 ± 3.7%, P < .001). Lumbar spine trabecular volumetric bone mineral density (LS vBMDtrab) also increased dose dependently with RSV (R = 0.268, P = .036), with a significant increase of 2.6 ± 1.3% in the RSVhigh group compared with placebo (P = .043). In addition, changes in BAP and LS vBMDtrab were positively correlated (R = 0.281, P = .027). No consistent changes were detected in bone density at the hip.
Our data suggest that high-dose RSV supplementation positively affects bone, primarily by stimulating formation or mineralization. Future studies of longer duration comprising populations at risk of osteoporosis are needed to confirm these results.
Obesity and metabolic syndrome (MetS) are major health problems worldwide, and due to several severe comorbidities, the health care costs related to these conditions are high (1, 2). MetS is associated with low-grade inflammation (3), and inflammation is a major cause of both local and systemic bone loss, caused by excessive bone resorption as well as impaired bone formation (4, 5). This imbalance in bone remodeling is at least in part mediated by cytokines activating osteoclasts and impairing osteoblast function (6), as seen in chronic inflammatory diseases such as inflammatory bowel disease (7) and rheumatoid arthritis (8). In the Swedish Osteoporotic Fractures in Men (MrOS) Study, elderly men with the highest tertile of high-sensitive C-reactive protein (hs-CRP) had an increased risk of vertebral fractures, and this association was independent of BMD (9).
Resveratrol (RSV) is a natural polyphenolic compound (3,5,4′-trihydroxy-trans-stilbene) present in nuts and fruits such as grapes. RSV has anti-inflammatory properties both in vitro and in vivo (10–13). Also, inflammation-independent effects of RSV have been described in relation to bone. RSV stimulates osteoblast differentiation (14–16), inhibits osteoclast activity (17, 18), and protects against bone loss in ovarectomized rats, immobilized rats, and aged mice (19–22). We previously analyzed the effect of RSV on biochemical markers of bone turnover in a randomized placebo-controlled trial designed to investigate potential effects of high-dose RSV supplementation on substrate metabolism, insulin sensitivity, and body composition (23). Obese men were randomly assigned to either placebo or 1.500 mg RSV daily for 4 weeks. We found a highly significant increase in bone alkaline phosphatase (BAP), without changes in other biochemical markers of bone turnover or calcium homeostasis (24).
To further explore the results from the initial trial we conducted a larger randomized placebo-controlled trial of longer duration and using two doses of RSV, to investigate the effects of RSV treatment on bone turnover markers, bone mass, and bone structure in obese men with MetS.
Materials and Methods
Study design and subjects
The study was a randomized, double-blinded, placebo-controlled trial (single center). Primary endpoint was changes in BAP. Secondary endpoints were changes in BMD assessed by quantitative computed tomography (CT) (QCT) and dual-energy x-ray absorptiometry (DXA), changes in bone geometry and microstructure assessed by QCT and high-resolution peripheral quantitative computed tomography (HRpQCT), and changes in other biochemical markers of bone turnover, as well as markers of calcium homeostasis. Subjects were randomly assigned (1:1:1) to treatment for 16 weeks with tablets containing 500 mg transresveratrol (RSVhigh)), 75 mg transresveratrol (RSVlow), or placebo twice daily. Tablets containing resveratrol were provided by Evolva, and placebo by Robinson Pharma. Randomization, blinding, packaging, and labeling were performed by the pharmacy at Aarhus University Hospital. In detail, the randomization was performed in blocks of six; initially RSVhigh and placebo 4:2, followed by RSVlow and placebo 4:2.
Participants were recruited through advertisements in local newspapers. We screened 123 men, enrolled and randomly assigned 76, of whom 66 completed the study (placebo, n = 24; RSVlow, n = 21; RSVhigh, n = 21) (Figure 1). Inclusion criteria were male sex, age 30–60 years, and presence of MetS. The International Diabetes Federation criteria for MetS (25) in men were used, according to which the following features should be present; Central obesity (Waist circumference ≥ 94 cm and/or body mass index [BMI] > 30 kg/m2) plus any two of the following: raised triglycerides (≥ 1.7 mmol/l), reduced high-density lipoprotein (≤ 1.03 mmol/l), raised blood pressure (systolic ≥ 130 mm Hg or diastolic ≥ 85 mm Hg), raised fasting plasma glucose (≥ 5.6 mmol/l). Exclusion criteria were other overt endocrine diseases, renal disease, hepatic disease, heart disease, malignant disease, anemia, alcohol abuse, and planned lifestyle changes.
The participants were instructed to maintain their body weight (BW) and lifestyle (including eating patterns, physical exercise etc.) and to abstain from any changes in intake of nutritional supplements (including calcium and vitamin D) during the study period. Information about adverse events was obtained at each visit, and the compliance was estimated based on tablet counting at each visit.
Ethical aspects
The protocol was approved by the Regional Committee on Health Research Ethics (M-20110111) and the Danish Data Protection Agency, and the study was conducted in agreement with the Declaration of Helsinki II. All participants were given oral and written information before written informed consent was obtained. According to the International Committee of Medical Journal Editors, the protocol was registered at ClinicalTrials.gov (NCT01412645) before recruitment was initiated.
General measurements
Standing height and weight were measured on a wall-mounted stadiometer with the participants lightly clothed. In addition, electrocardiogram, routine biochemistry, and physical examinations were performed at screening to investigate the presence of exclusion criteria.
Dual-energy x-ray absorptiometry
Areal bone mineral density (aBMD) and bone mineral content (BMC) at the lumbar spine (L1–L4), hip, and WB was measured by DXA using the same Hologic Discovery scanner at baseline and after 16 weeks of treatment. Coefficient of variation (CV) of repositioning is 1.5% for lumbar spine bone mineral density (LS aBMD) and 2.1% for femoral neck (26, 27).
Quantitative CT
Volumetric bone mineral density (vBMD) at lumbar vertebra 2 (LS vBMD) and the proximal femur was measured using a Philips Brilliance 40 multidetector helical CT scanner. For this purpose, CT scans were acquired from the distal endplate of L1 to the proximal endplate of L3, and from acetabulum directly above the femoral head to 2 cm below the lesser trochanter, with 3-mm slice thickness and spacing. The scans were performed at 120 kV and 50 mAs/slice (LS) or 125 mAs/slice (hip), rotation time, 1 second; field of view, 360 mm; and collimation, 40 × 0.625. QCT Pro (version 4.2.3, Mindways Software) was used to determine vBMD, in conjunction with a solid-state CT calibration phantom (Mindways phantom QA 3325), which was scanned simultaneously with the patient. MJ Ornstrup analyzed all patient QCT data, blinded to treatment allocation, with a reanalysis precision of 0.6 ± 0.6% for LS vBMDtrab, and 0.6 ± 0.7% for total hip integral volumetric bone mineral density (vBMDintegral).
L2 was initially rotated manually in 3D, followed by automatic positioning of the volume of interest (VOI) by QCTPro, in the anterior trabecular portion of the vertebral body. VOI was manually adjusted in case of inaccurate positioning. At the nondominant hip, proximal femur was rotated in 3D, followed by automatic positioning of VOI at different hip sites. We determined vBMDintegral in standard regions (total hip (TH), femoral neck (FN), trochanter (TR), and intertrochanteric (IT)), along with separate trabecular estimates (vBMDtrab). A fixed threshold for cortical separation of 0.350 g/cm3 was used (28, 29). CV of repositioning is 1.3% for LS vBMDtrab and 1.8% for TH vBMDintegral (30, 31).
High-resolution peripheral QCT
Assessment of geometry and microarchitecture of the nondominant distal radius and tibia (or in case of a previous fracture, the nonfractured limb) were obtained using HRpQCT (XtremeCT, Scanco Medical). To immobilize the arm and leg, a standard carbon fiber cast was used. A scout view was used to define the measurement region, using an offset from the endplate of radius and tibia of 9.5 mm and 22.5 mm, respectively. At each skeletal site, 110 slices were obtained, providing a 3D representation of approximately 9 mm in the axial direction. After each scan, image quality was assessed, and a rescan was conducted if necessary. Analyses were performed solely by MJ Ornstrup. CV of repositioning for radius and tibia density measures are 0.5–1.5% and of the structural parameters 1–5% (32, 33). The following parameters were measured or calculated: Trabecular bone volume fraction (BV/TV) is derived from the trabecular volume density, assuming a mineral density of fully mineralized bone of 1200 mg hydroxyapatite per cm3. Trabecular number (Tb.N) is a direct measure. Trabecular thickness (Tb.Th) and trabecular space (Tb.Sp) are calculated and from Tb.N and BV/TV, using standard stereologic relations and assuming plate model geometry (34).
Biochemistry
Blood and urine samples were collected between 0730 and 1100 hours after an overnight fast. Routine biochemistry (creatinine, sodium, potassium, ionized calcium, alanine transaminase [ALT], bilirubin, total alkaline phosphatase [AP], and hemoglobin) was analyzed continuously throughout the study. Blood and urine for the analysis of biochemical markers of calcium homeostasis (PTH, vitamin D, phosphate, magnesium, and urinary excretion of calcium and bone turnover were frozen at −80°C and −20°C, respectively, until the time of analysis. Samples were analyzed in a single batch to reduce analytical variation. Bone turnover markers were collected at all visits: at randomization (wk 0), after 4 (wk 4) and 8 weeks (wk 8) of treatment, and at the end of the study (wk 16). We analyzed BAP, osteoprotegerin (OPG), intact and N-terminal midfragment osteocalcin, and procollagen I N-terminal propeptide (P1NP) as markers of bone formation. As markers of bone resorption we analyzed C-terminal telopeptide of type 1 collagen (CTx), and cross-linked N-terminal telopeptide of type 1 collagen (NTx). All analyses, except OPG, were performed by standard laboratory methods at the Department of Clinical Biochemistry (Aarhus University Hospital, Aarhus, Denmark) (Supplemental Table 1). OPG was measured with ELISA OPG kit (Biomedica) and CV was 4.2%.
Statistics
Normality of data was checked by Q–Q plots, and equal variance between groups was assessed by Levene's test for equal variances. Baseline comparisons were assessed by ANOVA, or ANOVA on ranks if data were not normally distributed. Baseline results are presented as mean ± SD or median with interquartile (25%; 75%) range. To evaluate possible dose-dependent responses to RSV treatment, linear regression analysis was performed and dependence between two variables was evaluated by Pearson Product Moment Correlation. Absolute changes from baseline were calculated for each time point, and differences between study groups were assessed by unpaired Student t test and within groups by paired t test. Changes are presented as mean ± SEM. The primary endpoint was changes in BAP. To detect a treatment difference in BAP of 5 U/L at a two-sided 0.05 significance level with a power of 0.80, 24 participants should be included in each group, assuming a SD of 6U/L.
Results
Baseline characteristics
A total of 74 middle-age obese men with MetS were included in this study, of whom 66 completed all study visits. Average age was 49.3 ± 6.3 years and average BMI was 33.7 ± 3.6 kg/m2 at baseline. Baseline characteristics of the participants in the three intervention groups are comparable (Table 1).
Characteristic . | Placebo (n = 26) . | RSVlow (n = 23) . | RSVhigh (n = 25) . | P Value . |
---|---|---|---|---|
Age, y | 48.2 ± 6.4 | 48.9 ± 6.5 | 50.9 ± 5.9 | .31 |
Body mass index, kg/m2 | 34.3 ± 3.8 | 33.4 ± 4.0 | 33.3 ± 3.0 | .59 |
Smoking (yes/no) | (3/23) | (2/21) | (3/22) | N/A |
Biochemistry | ||||
AP, U/L | 76.0 ± 16.5 | 68.5 ± 15.3 | 69.9 ± 16.1 | .22 |
ALT, U/L | 39.8 ± 16.6 | 41.1 ± 13.6 | 44.8 ± 17.0 | .51 |
U-calcium, mmol/L | 3.4 ± 2.2 | 4.0 ± 3.0 | 3.2 ± 1.9 | .48 |
Plasma Ca2+, mmol/L | 1.22 ± 0.03 | 1.21 ± 0.03 | 1.22 ± 0.03 | .39 |
PTH, pmol/L | 5.39 (4.28; 7.04) | 5.06 (3.98; 6.39) | 4.91 (4.17; 6.07) | .62 |
25-hydroxy vit D, nmol/L | 36.4 (28.8; 65.0) | 48.2 (31.3; 79.2) | 45.3 (30.7; 74.2) | .41 |
BAP, U/L | 29.9 (23.9; 36.4) | 27.5 (24.1; 31.9) | 26.9 (21.6; 38.1) | .76 |
Osteocalcin, μg/L | 22.2 (17.6; 24.7) | 21.1 (17.0; 26.1) | 18.6 (17.1; 22.7) | .55 |
P1NP, μg/L | 48.7 ± 14.1 | 45.2 ± 14.2 | 42.6 ± 12.3 | .28 |
OPG, pmol/L | 4.83 ± 1.32 | 4.45 ± 1.42 | 4.90 ± 1.74 | .56 |
CTx, ng/ml | 0.299 (0.25; 0.45) | 0.308 (0.23; 0.40) | 0.292 (0.25; 0.42) | .94 |
NTx, nmol/L | 18.2 (16.9; 21.9) | 17.8 (15.0; 20.8) | 19.2 (15.0; 23.2) | .58 |
DXA, g/cm2 | ||||
Spine aBMD | 1.057 ± 0.096 | 1.095 ± 0.148 | 1.047 ± 0.139 | .41 |
Total hip aBMDa | 1.095 ± 0.106 | 1.084 ± 0.122 | 1.048 ± 0.125 | .34 |
WB aBMD | 1.176 ± 0.081 | 1.182 ± 0.088 | 1.155 ± 0.089 | .52 |
QCT, g/cmc | ||||
Spine vBMDtrab | 0.149 ± 0.033 | 0.147 ± 0.034 | 0.134 ± 0.029 | .19 |
TH vBMDintegrala | 0.323 ± 0.052 | 0.321 ± 0.038 | 0.308 ± 0.049 | .47 |
FN vBMDintegrala | 0.300 ± 0.042 | 0.303 ± 0.048 | 0.295 ± 0.048 | .84 |
TR vBMDintegrala | 0.246 ± 0.040 | 0.239 ± 0.030 | 0.238 ± 0.042 | .72 |
IT vBMDintegrala | 0.378 ± 0.063 | 0.379 ± 0.045 | 0.355 ± 0.058 | .24 |
HRpQCT | ||||
Ultra-distal radius | ||||
BV/TV, (1) | 0.156 ± 0.028 | 0.158 ± 0.032 | 0.151 ± 0.029 | .68 |
Tb.N, mm−1 | 2.194 ± 0.190 | 2.177 ± 0.260 | 2.185 ± 0.227 | .97 |
Tb.Th, mm | 0.072 ± 0.015 | 0.073 ± 0.012 | 0.069 ± 0.010 | .55 |
Tb.Sp, mm | 0.39 (0.36; 0.41) | 0.39 (0.34; 0.41) | 0.40 (0.36; 0.41) | .87 |
Ultra-distal tibiab | ||||
BV/TV, (1) | 0.163 ± 0.022 | 0.157 ± 0.027 | 0.163 ± 0.025 | .66 |
Tb.N, mm−1 | 2.164 ± 0.228 | 2.063 ± 0.282 | 2.103 ± 0.266 | .43 |
Tb.Th, mm | 0.076 ± 0.009 | 0.076 ± 0.008 | 0.078 ± 0.008 | .68 |
Tb.Sp, mm | 0.39 (0.35; 0.43) | 0.40 (0.37; 0.45) | 0.40 (0.35; 0.45) | .55 |
Characteristic . | Placebo (n = 26) . | RSVlow (n = 23) . | RSVhigh (n = 25) . | P Value . |
---|---|---|---|---|
Age, y | 48.2 ± 6.4 | 48.9 ± 6.5 | 50.9 ± 5.9 | .31 |
Body mass index, kg/m2 | 34.3 ± 3.8 | 33.4 ± 4.0 | 33.3 ± 3.0 | .59 |
Smoking (yes/no) | (3/23) | (2/21) | (3/22) | N/A |
Biochemistry | ||||
AP, U/L | 76.0 ± 16.5 | 68.5 ± 15.3 | 69.9 ± 16.1 | .22 |
ALT, U/L | 39.8 ± 16.6 | 41.1 ± 13.6 | 44.8 ± 17.0 | .51 |
U-calcium, mmol/L | 3.4 ± 2.2 | 4.0 ± 3.0 | 3.2 ± 1.9 | .48 |
Plasma Ca2+, mmol/L | 1.22 ± 0.03 | 1.21 ± 0.03 | 1.22 ± 0.03 | .39 |
PTH, pmol/L | 5.39 (4.28; 7.04) | 5.06 (3.98; 6.39) | 4.91 (4.17; 6.07) | .62 |
25-hydroxy vit D, nmol/L | 36.4 (28.8; 65.0) | 48.2 (31.3; 79.2) | 45.3 (30.7; 74.2) | .41 |
BAP, U/L | 29.9 (23.9; 36.4) | 27.5 (24.1; 31.9) | 26.9 (21.6; 38.1) | .76 |
Osteocalcin, μg/L | 22.2 (17.6; 24.7) | 21.1 (17.0; 26.1) | 18.6 (17.1; 22.7) | .55 |
P1NP, μg/L | 48.7 ± 14.1 | 45.2 ± 14.2 | 42.6 ± 12.3 | .28 |
OPG, pmol/L | 4.83 ± 1.32 | 4.45 ± 1.42 | 4.90 ± 1.74 | .56 |
CTx, ng/ml | 0.299 (0.25; 0.45) | 0.308 (0.23; 0.40) | 0.292 (0.25; 0.42) | .94 |
NTx, nmol/L | 18.2 (16.9; 21.9) | 17.8 (15.0; 20.8) | 19.2 (15.0; 23.2) | .58 |
DXA, g/cm2 | ||||
Spine aBMD | 1.057 ± 0.096 | 1.095 ± 0.148 | 1.047 ± 0.139 | .41 |
Total hip aBMDa | 1.095 ± 0.106 | 1.084 ± 0.122 | 1.048 ± 0.125 | .34 |
WB aBMD | 1.176 ± 0.081 | 1.182 ± 0.088 | 1.155 ± 0.089 | .52 |
QCT, g/cmc | ||||
Spine vBMDtrab | 0.149 ± 0.033 | 0.147 ± 0.034 | 0.134 ± 0.029 | .19 |
TH vBMDintegrala | 0.323 ± 0.052 | 0.321 ± 0.038 | 0.308 ± 0.049 | .47 |
FN vBMDintegrala | 0.300 ± 0.042 | 0.303 ± 0.048 | 0.295 ± 0.048 | .84 |
TR vBMDintegrala | 0.246 ± 0.040 | 0.239 ± 0.030 | 0.238 ± 0.042 | .72 |
IT vBMDintegrala | 0.378 ± 0.063 | 0.379 ± 0.045 | 0.355 ± 0.058 | .24 |
HRpQCT | ||||
Ultra-distal radius | ||||
BV/TV, (1) | 0.156 ± 0.028 | 0.158 ± 0.032 | 0.151 ± 0.029 | .68 |
Tb.N, mm−1 | 2.194 ± 0.190 | 2.177 ± 0.260 | 2.185 ± 0.227 | .97 |
Tb.Th, mm | 0.072 ± 0.015 | 0.073 ± 0.012 | 0.069 ± 0.010 | .55 |
Tb.Sp, mm | 0.39 (0.36; 0.41) | 0.39 (0.34; 0.41) | 0.40 (0.36; 0.41) | .87 |
Ultra-distal tibiab | ||||
BV/TV, (1) | 0.163 ± 0.022 | 0.157 ± 0.027 | 0.163 ± 0.025 | .66 |
Tb.N, mm−1 | 2.164 ± 0.228 | 2.063 ± 0.282 | 2.103 ± 0.266 | .43 |
Tb.Th, mm | 0.076 ± 0.009 | 0.076 ± 0.008 | 0.078 ± 0.008 | .68 |
Tb.Sp, mm | 0.39 (0.35; 0.43) | 0.40 (0.37; 0.45) | 0.40 (0.35; 0.45) | .55 |
Data are expressed as mean ± sd or median with interquartile range (25%; 75%).
The P values were calculated by ANOVA or ANOVA on Ranks where appropriate.
Data available on 25/26 (placebo).
Data available on 24/26 (placebo), 21/23 (RSVlow), and 24/25 (RSVhigh).
Characteristic . | Placebo (n = 26) . | RSVlow (n = 23) . | RSVhigh (n = 25) . | P Value . |
---|---|---|---|---|
Age, y | 48.2 ± 6.4 | 48.9 ± 6.5 | 50.9 ± 5.9 | .31 |
Body mass index, kg/m2 | 34.3 ± 3.8 | 33.4 ± 4.0 | 33.3 ± 3.0 | .59 |
Smoking (yes/no) | (3/23) | (2/21) | (3/22) | N/A |
Biochemistry | ||||
AP, U/L | 76.0 ± 16.5 | 68.5 ± 15.3 | 69.9 ± 16.1 | .22 |
ALT, U/L | 39.8 ± 16.6 | 41.1 ± 13.6 | 44.8 ± 17.0 | .51 |
U-calcium, mmol/L | 3.4 ± 2.2 | 4.0 ± 3.0 | 3.2 ± 1.9 | .48 |
Plasma Ca2+, mmol/L | 1.22 ± 0.03 | 1.21 ± 0.03 | 1.22 ± 0.03 | .39 |
PTH, pmol/L | 5.39 (4.28; 7.04) | 5.06 (3.98; 6.39) | 4.91 (4.17; 6.07) | .62 |
25-hydroxy vit D, nmol/L | 36.4 (28.8; 65.0) | 48.2 (31.3; 79.2) | 45.3 (30.7; 74.2) | .41 |
BAP, U/L | 29.9 (23.9; 36.4) | 27.5 (24.1; 31.9) | 26.9 (21.6; 38.1) | .76 |
Osteocalcin, μg/L | 22.2 (17.6; 24.7) | 21.1 (17.0; 26.1) | 18.6 (17.1; 22.7) | .55 |
P1NP, μg/L | 48.7 ± 14.1 | 45.2 ± 14.2 | 42.6 ± 12.3 | .28 |
OPG, pmol/L | 4.83 ± 1.32 | 4.45 ± 1.42 | 4.90 ± 1.74 | .56 |
CTx, ng/ml | 0.299 (0.25; 0.45) | 0.308 (0.23; 0.40) | 0.292 (0.25; 0.42) | .94 |
NTx, nmol/L | 18.2 (16.9; 21.9) | 17.8 (15.0; 20.8) | 19.2 (15.0; 23.2) | .58 |
DXA, g/cm2 | ||||
Spine aBMD | 1.057 ± 0.096 | 1.095 ± 0.148 | 1.047 ± 0.139 | .41 |
Total hip aBMDa | 1.095 ± 0.106 | 1.084 ± 0.122 | 1.048 ± 0.125 | .34 |
WB aBMD | 1.176 ± 0.081 | 1.182 ± 0.088 | 1.155 ± 0.089 | .52 |
QCT, g/cmc | ||||
Spine vBMDtrab | 0.149 ± 0.033 | 0.147 ± 0.034 | 0.134 ± 0.029 | .19 |
TH vBMDintegrala | 0.323 ± 0.052 | 0.321 ± 0.038 | 0.308 ± 0.049 | .47 |
FN vBMDintegrala | 0.300 ± 0.042 | 0.303 ± 0.048 | 0.295 ± 0.048 | .84 |
TR vBMDintegrala | 0.246 ± 0.040 | 0.239 ± 0.030 | 0.238 ± 0.042 | .72 |
IT vBMDintegrala | 0.378 ± 0.063 | 0.379 ± 0.045 | 0.355 ± 0.058 | .24 |
HRpQCT | ||||
Ultra-distal radius | ||||
BV/TV, (1) | 0.156 ± 0.028 | 0.158 ± 0.032 | 0.151 ± 0.029 | .68 |
Tb.N, mm−1 | 2.194 ± 0.190 | 2.177 ± 0.260 | 2.185 ± 0.227 | .97 |
Tb.Th, mm | 0.072 ± 0.015 | 0.073 ± 0.012 | 0.069 ± 0.010 | .55 |
Tb.Sp, mm | 0.39 (0.36; 0.41) | 0.39 (0.34; 0.41) | 0.40 (0.36; 0.41) | .87 |
Ultra-distal tibiab | ||||
BV/TV, (1) | 0.163 ± 0.022 | 0.157 ± 0.027 | 0.163 ± 0.025 | .66 |
Tb.N, mm−1 | 2.164 ± 0.228 | 2.063 ± 0.282 | 2.103 ± 0.266 | .43 |
Tb.Th, mm | 0.076 ± 0.009 | 0.076 ± 0.008 | 0.078 ± 0.008 | .68 |
Tb.Sp, mm | 0.39 (0.35; 0.43) | 0.40 (0.37; 0.45) | 0.40 (0.35; 0.45) | .55 |
Characteristic . | Placebo (n = 26) . | RSVlow (n = 23) . | RSVhigh (n = 25) . | P Value . |
---|---|---|---|---|
Age, y | 48.2 ± 6.4 | 48.9 ± 6.5 | 50.9 ± 5.9 | .31 |
Body mass index, kg/m2 | 34.3 ± 3.8 | 33.4 ± 4.0 | 33.3 ± 3.0 | .59 |
Smoking (yes/no) | (3/23) | (2/21) | (3/22) | N/A |
Biochemistry | ||||
AP, U/L | 76.0 ± 16.5 | 68.5 ± 15.3 | 69.9 ± 16.1 | .22 |
ALT, U/L | 39.8 ± 16.6 | 41.1 ± 13.6 | 44.8 ± 17.0 | .51 |
U-calcium, mmol/L | 3.4 ± 2.2 | 4.0 ± 3.0 | 3.2 ± 1.9 | .48 |
Plasma Ca2+, mmol/L | 1.22 ± 0.03 | 1.21 ± 0.03 | 1.22 ± 0.03 | .39 |
PTH, pmol/L | 5.39 (4.28; 7.04) | 5.06 (3.98; 6.39) | 4.91 (4.17; 6.07) | .62 |
25-hydroxy vit D, nmol/L | 36.4 (28.8; 65.0) | 48.2 (31.3; 79.2) | 45.3 (30.7; 74.2) | .41 |
BAP, U/L | 29.9 (23.9; 36.4) | 27.5 (24.1; 31.9) | 26.9 (21.6; 38.1) | .76 |
Osteocalcin, μg/L | 22.2 (17.6; 24.7) | 21.1 (17.0; 26.1) | 18.6 (17.1; 22.7) | .55 |
P1NP, μg/L | 48.7 ± 14.1 | 45.2 ± 14.2 | 42.6 ± 12.3 | .28 |
OPG, pmol/L | 4.83 ± 1.32 | 4.45 ± 1.42 | 4.90 ± 1.74 | .56 |
CTx, ng/ml | 0.299 (0.25; 0.45) | 0.308 (0.23; 0.40) | 0.292 (0.25; 0.42) | .94 |
NTx, nmol/L | 18.2 (16.9; 21.9) | 17.8 (15.0; 20.8) | 19.2 (15.0; 23.2) | .58 |
DXA, g/cm2 | ||||
Spine aBMD | 1.057 ± 0.096 | 1.095 ± 0.148 | 1.047 ± 0.139 | .41 |
Total hip aBMDa | 1.095 ± 0.106 | 1.084 ± 0.122 | 1.048 ± 0.125 | .34 |
WB aBMD | 1.176 ± 0.081 | 1.182 ± 0.088 | 1.155 ± 0.089 | .52 |
QCT, g/cmc | ||||
Spine vBMDtrab | 0.149 ± 0.033 | 0.147 ± 0.034 | 0.134 ± 0.029 | .19 |
TH vBMDintegrala | 0.323 ± 0.052 | 0.321 ± 0.038 | 0.308 ± 0.049 | .47 |
FN vBMDintegrala | 0.300 ± 0.042 | 0.303 ± 0.048 | 0.295 ± 0.048 | .84 |
TR vBMDintegrala | 0.246 ± 0.040 | 0.239 ± 0.030 | 0.238 ± 0.042 | .72 |
IT vBMDintegrala | 0.378 ± 0.063 | 0.379 ± 0.045 | 0.355 ± 0.058 | .24 |
HRpQCT | ||||
Ultra-distal radius | ||||
BV/TV, (1) | 0.156 ± 0.028 | 0.158 ± 0.032 | 0.151 ± 0.029 | .68 |
Tb.N, mm−1 | 2.194 ± 0.190 | 2.177 ± 0.260 | 2.185 ± 0.227 | .97 |
Tb.Th, mm | 0.072 ± 0.015 | 0.073 ± 0.012 | 0.069 ± 0.010 | .55 |
Tb.Sp, mm | 0.39 (0.36; 0.41) | 0.39 (0.34; 0.41) | 0.40 (0.36; 0.41) | .87 |
Ultra-distal tibiab | ||||
BV/TV, (1) | 0.163 ± 0.022 | 0.157 ± 0.027 | 0.163 ± 0.025 | .66 |
Tb.N, mm−1 | 2.164 ± 0.228 | 2.063 ± 0.282 | 2.103 ± 0.266 | .43 |
Tb.Th, mm | 0.076 ± 0.009 | 0.076 ± 0.008 | 0.078 ± 0.008 | .68 |
Tb.Sp, mm | 0.39 (0.35; 0.43) | 0.40 (0.37; 0.45) | 0.40 (0.35; 0.45) | .55 |
Data are expressed as mean ± sd or median with interquartile range (25%; 75%).
The P values were calculated by ANOVA or ANOVA on Ranks where appropriate.
Data available on 25/26 (placebo).
Data available on 24/26 (placebo), 21/23 (RSVlow), and 24/25 (RSVhigh).
Effect of resveratrol on bone mass and density
LS vBMDtrab increased +2.6 ± 1.3% in the RSVhigh group after 16 weeks of treatment compared with placebo (P = .043) and +2.6 ± 0.9% compared with baseline (P = .009). LS vBMDtrab in the RSVlow group increased nonsignificantly compared with placebo (+1.0 ± 1.1%, P = .305) (Figure 2). A linear regression analysis suggested a dose-dependent increase in LS vBMDtrab with increasing RSV dose (R = 0.268, P = .036; and R = 0.273, P = .047 after adjusting for changes in vitamin D during the study). In addition, both aBMD and BMC at the spine increased significantly within the RSVhigh group (P = .017 and P = .014, respectively). However, the changes were not different from the placebo group (Table 2).
. | Placebo (n = 24) . | RSVlow (n = 21) . | RSVhigh (n = 21) . | P Value RSVlow vs PLC . | P Value RSVhigh vs PLC . |
---|---|---|---|---|---|
DXA | |||||
L1–L4 aBMD | 0.91 ± 0.57 | 0.62 ± 0.53 | 1.02 ± 0.38a | .70 | .88 |
L1–L4 BMC | 0.95 ± 0.84 | 1.20 ± 0.63 | 1.68 ± 0.61a | .97 | .63 |
Total hip aBMD | 0.58 ± 0.31 | 0.52 ± 0.30 | 0.30 ± 0.36 | .94 | .50 |
Total hip BMC | 0.87 ± 0.61 | 0.96 ± 0.69 | 0.40 ± 0.60 | .75 | .58 |
WB aBMD | 0.44 ± 0.40 | 0.23 ± 0.29 | −0.16 ± 0.30 | .62 | .20 |
WB BMC | 0.45 ± 0.41 | 1.05 ± 0.29b | −0.02 ± 0.26 | .28 | .29 |
QCT | |||||
Spinec | |||||
L2 vBMDtrab | −0.04 ± 0.89 | 0.93 ± 0.62 | 2.55 ± 0.88b | .31 | .043 |
Total hipd | |||||
vBMDintegral | 0.99 ± 0.44 | 0.35 ± 0.47 | 0.70 ± 0.61 | .29 | .62 |
vBMDtrab | 0.40 ± 0.32 | −0.42 ± 0.20 | 0.58 ± 0.35 | .039 | .68 |
Femoral neckd | |||||
vBMDintegral | 0.92 ± 0.71 | 0.25 ± 0.48 | −0.77 ± 0.61 | .43 | .08 |
vBMDtrab | 0.23 ± 0.45 | −1.12 ± 0.60 | 1.55 ± 0.77 | .06 | .14 |
Trochantericd | |||||
vBMDintegral | 0.53 ± 0.39 | 0.36 ± 0.29 | 1.14 ± 0.38b | .74 | .34 |
vBMDtrab | 0.50 ± 0.28 | 0.05 ± 0.23 | 0.75 ± 0.42 | .17 | .71 |
Intertrochantericd | |||||
vBMDintegral | 0.87 ± 0.55 | 0.43 ± 0.67 | 1.32 ± 0.92 | .59 | .86 |
vBMDtrab | 0.29 ± 0.38 | −0.57 ± 0.35 | −0.04 ± 0.49 | .11 | .70 |
HRpQCT | |||||
Ultra-distal radius | |||||
BV/TV | −0.13 ± 0.44 | 0.17 ± 0.31 | −0.47 ± 0.31 | .27 | .88 |
Tb.N. | 1.54 ± 1.93 | 1.84 ± 2.16 | −0.10 ± 1.81 | .92 | .53 |
Tb.Th. | −0.86 ± 1.86 | −0.83 ± 1.92 | 0.23 ± 1.66 | .89 | .51 |
Tb.Sp. | −0.68 ± 1.87 | −0.96 ± 2.10 | 0.85 ± 1.89 | .98 | .57 |
Ultra-distal tibiae | |||||
BV/TV | −0.43 ± 0.29 | −0.31 ± 0.27 | −0.27 ± 0.35 | .70 | .84 |
Tb.N. | −0.97 ± 1.32 | −0.68 ± 0.77 | −1.76 ± 1.91 | .82 | .77 |
Tb.Th. | 0.83 ± 1.38 | 0.72 ± 0.78 | 2.15 ± 2.20 | .95 | .67 |
Tb.Sp. | 1.46 ± 1.33 | 0.83 ± 0.75 | 2.56 ± 2.06 | .77 | .58 |
. | Placebo (n = 24) . | RSVlow (n = 21) . | RSVhigh (n = 21) . | P Value RSVlow vs PLC . | P Value RSVhigh vs PLC . |
---|---|---|---|---|---|
DXA | |||||
L1–L4 aBMD | 0.91 ± 0.57 | 0.62 ± 0.53 | 1.02 ± 0.38a | .70 | .88 |
L1–L4 BMC | 0.95 ± 0.84 | 1.20 ± 0.63 | 1.68 ± 0.61a | .97 | .63 |
Total hip aBMD | 0.58 ± 0.31 | 0.52 ± 0.30 | 0.30 ± 0.36 | .94 | .50 |
Total hip BMC | 0.87 ± 0.61 | 0.96 ± 0.69 | 0.40 ± 0.60 | .75 | .58 |
WB aBMD | 0.44 ± 0.40 | 0.23 ± 0.29 | −0.16 ± 0.30 | .62 | .20 |
WB BMC | 0.45 ± 0.41 | 1.05 ± 0.29b | −0.02 ± 0.26 | .28 | .29 |
QCT | |||||
Spinec | |||||
L2 vBMDtrab | −0.04 ± 0.89 | 0.93 ± 0.62 | 2.55 ± 0.88b | .31 | .043 |
Total hipd | |||||
vBMDintegral | 0.99 ± 0.44 | 0.35 ± 0.47 | 0.70 ± 0.61 | .29 | .62 |
vBMDtrab | 0.40 ± 0.32 | −0.42 ± 0.20 | 0.58 ± 0.35 | .039 | .68 |
Femoral neckd | |||||
vBMDintegral | 0.92 ± 0.71 | 0.25 ± 0.48 | −0.77 ± 0.61 | .43 | .08 |
vBMDtrab | 0.23 ± 0.45 | −1.12 ± 0.60 | 1.55 ± 0.77 | .06 | .14 |
Trochantericd | |||||
vBMDintegral | 0.53 ± 0.39 | 0.36 ± 0.29 | 1.14 ± 0.38b | .74 | .34 |
vBMDtrab | 0.50 ± 0.28 | 0.05 ± 0.23 | 0.75 ± 0.42 | .17 | .71 |
Intertrochantericd | |||||
vBMDintegral | 0.87 ± 0.55 | 0.43 ± 0.67 | 1.32 ± 0.92 | .59 | .86 |
vBMDtrab | 0.29 ± 0.38 | −0.57 ± 0.35 | −0.04 ± 0.49 | .11 | .70 |
HRpQCT | |||||
Ultra-distal radius | |||||
BV/TV | −0.13 ± 0.44 | 0.17 ± 0.31 | −0.47 ± 0.31 | .27 | .88 |
Tb.N. | 1.54 ± 1.93 | 1.84 ± 2.16 | −0.10 ± 1.81 | .92 | .53 |
Tb.Th. | −0.86 ± 1.86 | −0.83 ± 1.92 | 0.23 ± 1.66 | .89 | .51 |
Tb.Sp. | −0.68 ± 1.87 | −0.96 ± 2.10 | 0.85 ± 1.89 | .98 | .57 |
Ultra-distal tibiae | |||||
BV/TV | −0.43 ± 0.29 | −0.31 ± 0.27 | −0.27 ± 0.35 | .70 | .84 |
Tb.N. | −0.97 ± 1.32 | −0.68 ± 0.77 | −1.76 ± 1.91 | .82 | .77 |
Tb.Th. | 0.83 ± 1.38 | 0.72 ± 0.78 | 2.15 ± 2.20 | .95 | .67 |
Tb.Sp. | 1.46 ± 1.33 | 0.83 ± 0.75 | 2.56 ± 2.06 | .77 | .58 |
Abbreviation: PLC, placebo.
Data are expressed as percentage change from baseline (mean ± sem).
The P values for between-group differences were calculated by unpaired Student t test. Significant results are marked in bold.
Changes from baseline within groups were calculated by paired t test (superscript notation if significant within-group changes:
P < .05 and
P < .01).
Data available on 22/24 (PLC), 20/21 (RSVlow), and 20/21 (RSVhigh).
Data available on 21/24 (PLC), 21/21 (RSVlow), and 19/21 (RSVhigh).
Data available on 22/24 (PLC), 19/21 (RSVlow), and 19/21 (RSVhigh).
. | Placebo (n = 24) . | RSVlow (n = 21) . | RSVhigh (n = 21) . | P Value RSVlow vs PLC . | P Value RSVhigh vs PLC . |
---|---|---|---|---|---|
DXA | |||||
L1–L4 aBMD | 0.91 ± 0.57 | 0.62 ± 0.53 | 1.02 ± 0.38a | .70 | .88 |
L1–L4 BMC | 0.95 ± 0.84 | 1.20 ± 0.63 | 1.68 ± 0.61a | .97 | .63 |
Total hip aBMD | 0.58 ± 0.31 | 0.52 ± 0.30 | 0.30 ± 0.36 | .94 | .50 |
Total hip BMC | 0.87 ± 0.61 | 0.96 ± 0.69 | 0.40 ± 0.60 | .75 | .58 |
WB aBMD | 0.44 ± 0.40 | 0.23 ± 0.29 | −0.16 ± 0.30 | .62 | .20 |
WB BMC | 0.45 ± 0.41 | 1.05 ± 0.29b | −0.02 ± 0.26 | .28 | .29 |
QCT | |||||
Spinec | |||||
L2 vBMDtrab | −0.04 ± 0.89 | 0.93 ± 0.62 | 2.55 ± 0.88b | .31 | .043 |
Total hipd | |||||
vBMDintegral | 0.99 ± 0.44 | 0.35 ± 0.47 | 0.70 ± 0.61 | .29 | .62 |
vBMDtrab | 0.40 ± 0.32 | −0.42 ± 0.20 | 0.58 ± 0.35 | .039 | .68 |
Femoral neckd | |||||
vBMDintegral | 0.92 ± 0.71 | 0.25 ± 0.48 | −0.77 ± 0.61 | .43 | .08 |
vBMDtrab | 0.23 ± 0.45 | −1.12 ± 0.60 | 1.55 ± 0.77 | .06 | .14 |
Trochantericd | |||||
vBMDintegral | 0.53 ± 0.39 | 0.36 ± 0.29 | 1.14 ± 0.38b | .74 | .34 |
vBMDtrab | 0.50 ± 0.28 | 0.05 ± 0.23 | 0.75 ± 0.42 | .17 | .71 |
Intertrochantericd | |||||
vBMDintegral | 0.87 ± 0.55 | 0.43 ± 0.67 | 1.32 ± 0.92 | .59 | .86 |
vBMDtrab | 0.29 ± 0.38 | −0.57 ± 0.35 | −0.04 ± 0.49 | .11 | .70 |
HRpQCT | |||||
Ultra-distal radius | |||||
BV/TV | −0.13 ± 0.44 | 0.17 ± 0.31 | −0.47 ± 0.31 | .27 | .88 |
Tb.N. | 1.54 ± 1.93 | 1.84 ± 2.16 | −0.10 ± 1.81 | .92 | .53 |
Tb.Th. | −0.86 ± 1.86 | −0.83 ± 1.92 | 0.23 ± 1.66 | .89 | .51 |
Tb.Sp. | −0.68 ± 1.87 | −0.96 ± 2.10 | 0.85 ± 1.89 | .98 | .57 |
Ultra-distal tibiae | |||||
BV/TV | −0.43 ± 0.29 | −0.31 ± 0.27 | −0.27 ± 0.35 | .70 | .84 |
Tb.N. | −0.97 ± 1.32 | −0.68 ± 0.77 | −1.76 ± 1.91 | .82 | .77 |
Tb.Th. | 0.83 ± 1.38 | 0.72 ± 0.78 | 2.15 ± 2.20 | .95 | .67 |
Tb.Sp. | 1.46 ± 1.33 | 0.83 ± 0.75 | 2.56 ± 2.06 | .77 | .58 |
. | Placebo (n = 24) . | RSVlow (n = 21) . | RSVhigh (n = 21) . | P Value RSVlow vs PLC . | P Value RSVhigh vs PLC . |
---|---|---|---|---|---|
DXA | |||||
L1–L4 aBMD | 0.91 ± 0.57 | 0.62 ± 0.53 | 1.02 ± 0.38a | .70 | .88 |
L1–L4 BMC | 0.95 ± 0.84 | 1.20 ± 0.63 | 1.68 ± 0.61a | .97 | .63 |
Total hip aBMD | 0.58 ± 0.31 | 0.52 ± 0.30 | 0.30 ± 0.36 | .94 | .50 |
Total hip BMC | 0.87 ± 0.61 | 0.96 ± 0.69 | 0.40 ± 0.60 | .75 | .58 |
WB aBMD | 0.44 ± 0.40 | 0.23 ± 0.29 | −0.16 ± 0.30 | .62 | .20 |
WB BMC | 0.45 ± 0.41 | 1.05 ± 0.29b | −0.02 ± 0.26 | .28 | .29 |
QCT | |||||
Spinec | |||||
L2 vBMDtrab | −0.04 ± 0.89 | 0.93 ± 0.62 | 2.55 ± 0.88b | .31 | .043 |
Total hipd | |||||
vBMDintegral | 0.99 ± 0.44 | 0.35 ± 0.47 | 0.70 ± 0.61 | .29 | .62 |
vBMDtrab | 0.40 ± 0.32 | −0.42 ± 0.20 | 0.58 ± 0.35 | .039 | .68 |
Femoral neckd | |||||
vBMDintegral | 0.92 ± 0.71 | 0.25 ± 0.48 | −0.77 ± 0.61 | .43 | .08 |
vBMDtrab | 0.23 ± 0.45 | −1.12 ± 0.60 | 1.55 ± 0.77 | .06 | .14 |
Trochantericd | |||||
vBMDintegral | 0.53 ± 0.39 | 0.36 ± 0.29 | 1.14 ± 0.38b | .74 | .34 |
vBMDtrab | 0.50 ± 0.28 | 0.05 ± 0.23 | 0.75 ± 0.42 | .17 | .71 |
Intertrochantericd | |||||
vBMDintegral | 0.87 ± 0.55 | 0.43 ± 0.67 | 1.32 ± 0.92 | .59 | .86 |
vBMDtrab | 0.29 ± 0.38 | −0.57 ± 0.35 | −0.04 ± 0.49 | .11 | .70 |
HRpQCT | |||||
Ultra-distal radius | |||||
BV/TV | −0.13 ± 0.44 | 0.17 ± 0.31 | −0.47 ± 0.31 | .27 | .88 |
Tb.N. | 1.54 ± 1.93 | 1.84 ± 2.16 | −0.10 ± 1.81 | .92 | .53 |
Tb.Th. | −0.86 ± 1.86 | −0.83 ± 1.92 | 0.23 ± 1.66 | .89 | .51 |
Tb.Sp. | −0.68 ± 1.87 | −0.96 ± 2.10 | 0.85 ± 1.89 | .98 | .57 |
Ultra-distal tibiae | |||||
BV/TV | −0.43 ± 0.29 | −0.31 ± 0.27 | −0.27 ± 0.35 | .70 | .84 |
Tb.N. | −0.97 ± 1.32 | −0.68 ± 0.77 | −1.76 ± 1.91 | .82 | .77 |
Tb.Th. | 0.83 ± 1.38 | 0.72 ± 0.78 | 2.15 ± 2.20 | .95 | .67 |
Tb.Sp. | 1.46 ± 1.33 | 0.83 ± 0.75 | 2.56 ± 2.06 | .77 | .58 |
Abbreviation: PLC, placebo.
Data are expressed as percentage change from baseline (mean ± sem).
The P values for between-group differences were calculated by unpaired Student t test. Significant results are marked in bold.
Changes from baseline within groups were calculated by paired t test (superscript notation if significant within-group changes:
P < .05 and
P < .01).
Data available on 22/24 (PLC), 20/21 (RSVlow), and 20/21 (RSVhigh).
Data available on 21/24 (PLC), 21/21 (RSVlow), and 19/21 (RSVhigh).
Data available on 22/24 (PLC), 19/21 (RSVlow), and 19/21 (RSVhigh).
Neither aBMD nor BMC at total hip changed during the 16 weeks of treatment (Table 2). Results from QCT at the hip were ambiguous. TR vBMDintegral increased significantly from baseline within the RSVhigh group (P = .008), and likewise did FN vBMDtrab (P = .053), without reaching significance when compared with placebo. There were no dose-dependent effects at the hip. Actually, low-dose RSV had negative effects compared with placebo at TH vBMDtrab (P = .035), but was not significantly different from baseline (P = .063). Likewise, vBMDtrab at FN was borderline significantly lower in the RSVlow group compared with placebo (P = .064) (Table 2).
Changes in WB aBMD and WB BMC did not differ between groups; however, WB BMC increased significantly within the RSVlow group (P = .002) (Table 2).
HRpQCT-derived estimates of microarchitecture at distal radius and distal tibia did not change significantly between or within groups (Table 2), and neither did strength estimates from finite element modeling (data not shown).
Effect of resveratrol on biochemical markers of bone turnover
We found a highly significant dose-dependent increase in BAP (R = 0.471, P < .001; and R = 0.471, P < .001 after adjusting for changes in vitamin D during the study). At all time-points the RSVhigh group had significantly greater increase in BAP from baseline compared with the placebo group (wk 4: 16.4 ± 4.2%, P < .001; wk 8: 16.5 ± 4.1%, P < .001; wk 16: 15.2 ± 3.7%, P < .001). The changes seen in the RSVlow group followed the same pattern but were not significantly different from placebo (wk 4: 5.3 ± 3.2%, P = .20; wk 8: 4.7 ± 2.8%, P = .12; wk 16: 5.2 ± 3.5%, P = .14) (Figure 3A). Within the entire population changes in BAP and LS vBMD were positively correlated (R = 0.281, P = .027).
Other markers of bone formation did not change consistently. Changes in P1NP were significantly different in the RSVhigh group compared with placebo at week 8 (+8.0 ± 4.5%, P = .049) (Figure 3B). Osteocalcin increased in the RSVhigh group, but not significantly different from the placebo group (wk 4: 6.2 ± 4.2%, P = .10; wk 8: 6.9 ± 4.9%, P = .12; wk 16: 3.7 ± 4.6%, P = .42) (Figure 3C). Changes in OPG and the two bone resorption markers CTx and S-NTx were comparable between groups at all time points (Supplemental Table 2 and Figure 3D).
25-hydroxy vitamin D increased similarly throughout the study in the placebo group and RSVhigh group. However, changes in the RSVlow group were different from changes in the placebo group at all time points (P < .05) (Figure 4A). PTH levels decreased more in the placebo group compared with RSVhigh group at week 4 and week 8 (Figure 4B), whereas ionized calcium was decreased at week 8 in the RSVhigh group compared with placebo (Figure 4C). We found no differences in urinary excretion of calcium, ALT, bilirubin, or creatinine between groups (Supplemental Table 2). AP increased significantly in the RSVlow and RSVhigh group compared with placebo (Figure 4D).
Compliance and tolerability
Compliance rates were 97% (94; 99)%, 93% (87; 98)%, and 96% (93; 99)% in the placebo group, the RSVlow group, and the RSVhigh group, respectively. Generally, treatments were well tolerated and there were no serious adverse events. Complaints of the gastrointestinal tract were the most common; (7/25) in RSVhigh, (3/23) in RSVlow, and (4/26) in the placebo group. The gastrointestinal complaints were mild, and primarily in relation to increased frequency and/or softer stools, especially the first 3–4 weeks of treatment. Other adverse effects were rare and unlikely to be related to resveratrol treatment. However, one trial subject from the high-dose RSV treated group developed a transient pruritic skin rash after 1 month of treatment, which was resolved 14 days after stopping the treatment.
Discussion
In this randomized, placebo-controlled, single-center study of the effects on RSV on bone, we found a significant dose-dependent increase in the trabecular vBMD at the spine, reaching +2.6% in the group receiving high-dose RSV supplementation. The bone formation marker BAP increased by 16% after 4 weeks in the high-dose RSV group compared with placebo and remained elevated throughout the study. Furthermore, the changes in LS vBMDtrab and BAP were significantly correlated, supporting a causal relationship. This increase in LS vBMD is impressive considering the short intervention period of 16 weeks, and the age and phenotype of the trial subjects. For comparison, McClung et al (35) reported increases in trabecular vBMD at the spine, in postmenopausal osteoporotic women treated for 24 weeks with either 20 μg teriparatide or 10 mg alendronate daily, of 12.2% and 5.1%, respectively. Although the effect of RSV was inferior to these recognized antiosteoporotic drugs, we believe that a 2.6% increase over a shorter intervention period in a nonosteoporotic population of obese men makes it worth the effort to further investigate the antiosteoporotic potential of RSV.
DXA-derived aBMD and BMC at the spine increased significantly within the RSVhigh group, however not significantly different from placebo. Although aBMD by DXA is an integrated measurement of the trabecular and cortical bone, vBMD measured by QCT only comprises trabecular bone at the spine. Using vBMD measures offers advantages, especially in our population of obese men. The lumbar vertebrae consist primarily of trabeculae, enclosed by a very thin layer of cortex. The trabecular compartment is the most metabolic active compartment and therefore reveals the earliest and most dramatic changes (30). The precision error of DXA measures increases with increasing BMI, due to increased tissue thickness and fat inhomogeneity (36). The QCT technique reduces noise from excess soft tissue to a minimum, and QCT is less susceptible to disparities in patient positioning from one scan to the next, because postscan rotation of the vertebrae and hip is possible within the analysis software. QCT-derived vBMD may therefore be a better estimate of bone strength and quality in this category of patients (37, 38). No consistent changes were detected in bone density at the hip.
The increase seen in BAP is a confirmation of the findings of our previous study on short-term effects of resveratrol (24). That study lasted only 4 weeks and in the present study we have shown that the initial increase in BAP is maintained for at least 16 weeks. Also, AP increased in the RSV groups compared with placebo, but to a lesser extent. The AP measurement includes all iso-enzymes of alkaline phosphatase. In adults, AP primarily represents bone and liver iso-enzymes, with an intestinal fraction of less than 10%. Cross-reactivity between bone and liver iso-enzymes is modest (around 5%). The differences in AP changes between groups are probably driven by differences in BAP.
None of the markers of bone resorption changed differently between groups during the study, whereas the formation marker osteocalcin followed a pattern similar (nonsignificant) to the changes seen in BAP, and P1NP was significantly increased in the RSVhigh group at wk 8.
Baseline vitamin D levels were low and comparable between groups. Changes in vitamin D during the 16 weeks of intervention were modest; however, significantly different between groups, as levels increased in the RSVhigh group and the placebo group, but not in the RSVlow group. There was a tendency toward more participants being included in the late fall/early winter in the RSVlow group and late winter/early spring in the other two groups. We speculate if time of year at inclusion could explain the different changes in vitamin D levels, because a nadir of vitamin D is present in late winter (39). Because differences in vitamin D could have potentially affected the changes found in vBMD and BAP, we included changes in vitamin D in a multiple linear regression analysis. However, increasing doses of RSV still predicted an increasing LS vBMDtrab and increasing BAP independently of changes in vitamin D. The increase in serum levels of vitamin D and the accompanying decrease in serum levels of PTH in the placebo group is probably the explanation to why bone resorption and formation decline during the course of the study. Although we did see similar changes in serum levels of vitamin D in the RSVhigh group, PTH surprisingly remained at baseline level. The increase in bone formation and/or mineralization could explain the temporary decrease in serum levels of ionized calcium, due to calcium deposition in the bone and this decrease in calcium would tend to stimulate PTH secretion and thereby counteract the expected decrease in PTH with increasing vitamin D. Surprisingly, the higher serum levels of PTH in the RSVhigh group did not seem to induce osteoclast activity, as bone resorption markers were comparable to placebo. Thus, our data suggest that high-dose RSV supplementation affects bone primarily by stimulating formation or mineralization; however, we speculate that RSV may inhibit osteoclast activity, as shown in cell culture studies (17, 18), but the expected decrease in bone resorption is counteracted by the higher serum levels of PTH. Regardless, RSV seems to uncouple bone formation and resorption, possibly due to direct effects on both osteoblasts and osteoclasts. To gain better insight, future studies should include double-labeled bone biopsies useful to determine for example mineral apposition rate and to quantify osteoclasts by tartrate-resistant acid phosphatase (TRAP) staining. We measured urine calcium levels to ensure that calcium was not simply excreted, and thereby causing the temporary decrease in ionized calcium in the RSVhigh group. There were no differences between groups, which support calcium deposition in the bone as a possible explanation, and ties in well with the increase in LS vBMDtrab.
Changes in level of low-grade inflammation could also have an effect on bone turnover in these men with MetS. However, no effect on inflammation markers after 16 weeks of RSV treatment could be demonstrated (40). Therefore, the positive effect of RSV on bone is not explained by anti-inflammatory effects. Our previous study supports this claim given that we found increasing BAP after 4 weeks of RSV treatment with no simultaneous reduction in inflammatory markers (23, 24).
The strengths of this study include the design; a randomized, placebo-controlled, double-blind, single-center study with a high adherence rate, and that the effect of RSV on bone have been examined using several methods, including DXA, QCT, HRpQCT, and biochemical markers of bone turnover. The study also has limitations. The short intervention period prevents us from drawing any conclusions about the long-term effect of RSV on bone and calcium metabolism. The different changes in vitamin D makes the interpretation of some of the findings more difficult; however, the changes were very modest and regression analyses demonstrated that the differences found in BAP and vBMD in response to RSV treatment were not explained by changes in vitamin D. Finally, recognizing that effects of RSV are not mediated through reduced inflammation, the chosen study population is not the best suited to investigate effects on bone, and generalizability is limited.
In conclusion, we have investigated the effects of resveratrol on bone, and found dose-dependent and correlated increases in bone alkaline phosphatase and lumbar spine volumetric BMD after only 16 weeks of treatment. The effects of RSV were not mediated by changes in inflammation, and we speculate that RSV stimulates bone formation or mineralization directly. Future studies of longer duration comprising populations at risk of osteoporosis are needed to confirm these positive effects of resveratrol.
Acknowledgments
We thank laboratory technicians Lotte Sørensen, Liselotte Stenkjær, Pia Hornbek, Lisbeth Flyvbjerg, and Tove Stenum for their excellent technical assistance. We thank the Danish Council for Strategic Research for funding; and Evolva, Switzerland for providing us with the RSV tablets.
This study was registered in ClinicalTrials.gov as trial number NCT01412645.
This work was supported by the Long-term Investigation of Resveratrol on Management of Metabolic syndrome, Osteoporosis and Inflammation, and Identification of plant derived anti-inflammatory compounds (LIRMOI) research center (www.LIRMOI.com), which is supported by the Danish Council for Strategic Research (Grant 10-0934999). The HRpQCT scanner was supported by The Toyota Foundation, Karen Elise Jensen Foundation, AP Møller Maersk Foundation, and the Danish Osteoporosis Association.
Disclosure Summary: The authors have nothing to disclose.
Abbreviations
- aBMD
areal bone mineral density
- ALT
alanine transaminase
- AP
alkaline phosphatase
- BAP
bone alkaline phosphatase
- BMC
bone mineral content
- BMD
bone mineral density
- BMI
body mass index
- BV/TV
bone volume/tissue volume
- CT
computed tomography
- CTx
C-terminal telopeptide of type 1 collagen
- CV
coefficient of variation
- DXA
dual-energy x-ray absorptiometry
- FN
femoral neck
- hs-CRP
high-sensitive C-reactive protein
- HRpQCT
high-resolution peripheral quantitative computed tomography
- IT
intertrochanteric
- LS
lumbar spine
- MetS
metabolic syndrome
- NTx
N-terminal telopeptide of type 1 collagen
- OPG
osteoprotegerin
- P1NP
procollagen I N-terminal propeptide
- QCT
quantitative computed tomography
- RSV
resveratrol
- RSVhigh
1.000 mg resveratrol
- RSVlow
150 mg resveratrol
- Tb.Th
trabecular thickness
- Tb.N
trabecular number
- Tb.Sp
trabecular space
- TH
total hip
- TR
trochanter
- vBMD
volumetric bone mineral density
- vBMDintegral
integral volumetric bone mineral density
- vBMDtrab
trabecular volumetric bone mineral density
- VOI
volume of interest
- WB
whole body.
References