Elsevier

Archives of Gerontology and Geriatrics

Volume 57, Issue 2, September–October 2013, Pages 226-233
Archives of Gerontology and Geriatrics

Response of bone mineral density, inflammatory cytokines, and biochemical bone markers to a 32-week combined loading exercise programme in older men and women

Abstract

This study examines the effects of 32 weeks of exercise training on balance, lower-extremity muscle strength, bone mineral density (BMD) and serum levels of bone metabolism and inflammatory markers in older adults. Forty-seven healthy older adults (women = 24, men = 23; mean age 68.2 years) participated in a exercise intervention (60 min/session) that included resistance exercise training (2 days/week) at 75–80% of maximum plus a multicomponent weight-bearing impact exercise training (1 day/week). Outcome measures included lumbar spine and proximal femoral BMD, dynamic balance, muscle strength, serum levels of bone metabolism markers [osteocalcin (OC), C-terminal telopeptide of Type I collagen (CTX), osteoprotegerin (OPG) and receptor activator of nuclear factor kappa B ligand (RANKL)] and serum levels of inflammatory markers [high sensitive (hs)-C-reactive protein (CRP), interleukin (IL)-6, tumor necrosis factor (TNF)-α, and interferon (IFN)-γ]. Potential confounding variables included body composition, dietary intake (using 4-day diet records), and accelerometer-based physical activity. After 32 weeks, both men and women increased dynamic balance (6.4%), muscle strength (11.0%) and trochanter (0.7%), intertrochanter (0.7%), total hip (0.6%), and lumbar spine BMD (1.7%), while OC, CTX, OPG and RANKL levels remained unchanged. In addition, hs-CRP and IFN-γ levels were decreased, while TNF-α levels were unchanged, and a decrease in IL-6 levels was only observed in men. These findings suggest that our combined impact protocol reduces inflammation and increases BMD, balance, and lower-extremity muscle strength, despite having little effect on bone metabolism markers. This reinforces the role of exercise to counteract the age-related inflammation, and the muscle strength, balance and BMD reduction.

Abbreviations

ANOVA
analysis of variance
B-ALP
bone alkaline phosphatase
BMD
bone mineral density
OC
osteocalcin
CTX
C-terminal telopeptide of Type I collagen
OPG
osteoprotegerin
RANKL
receptor activator of nuclear factor kappa B ligand
hs-CRP
high sensitive C-reactive protein
IL-6
interleukin-6
TNF-α
tumor necrosis factor-alpha
IFN-γ
interferon-gamma
PA
physical activity
BMI
body mass index
MVPA
moderate to vigorous physical activity
ES
effect size
RCT
randomized controlled trial
1RM
one-repetition maximum

Keywords

Bone mass
Elderly
Resistance exercise
Weight-bearing exercise
Inflammation
Biomarkers

1. Introduction

Aging is linked to a reduced amount of bone tissue, which consequently motivates bones to become weaker, commonly leading to osteoporosis (Kostenuik & Shalhoub, 2001). This is a common, serious, and disabling condition due to the inherent association with low-energy trauma or fragility fractures, and with also severe economic consequences (Cummings & Melton, 2002). Increasing evidence has suggested a central role for falls as the strongest single risk factor for a fracture (Peeters, van Schoor, & Lips, 2009). Accordingly, strategies targeting the prevention of bone fractures in the elderly should focus on reducing the risk of falls and maintaining or improving bone health. Mixed loading exercise programs combining impact activity with high-magnitude exercise as resistance training and odd-impact protocols appear effective against aging-induced bone weakness (Marques et al., 2012, Martyn-St James and Carroll, 2009). Despite the importance of bone density in the elderly and their attenuated bone response to physical forces (exercise) (Lanyon & Skerry, 2001), most studies have focused on postmenopausal women.

In addition, experimental animal studies have recently implicated inflammation in the pathogenesis of osteoporosis (Nanes, 2003). The effect is primarily driven on the differentiation and activity of the bone-resorbing cell, the osteoclast; and it is established that pro-inflammatory cytokines suppress osteoprotegerin (OPG) expression while simultaneously enhancing receptor activator of nuclear factor kappa B ligand (RANKL) expression (Schett, 2011). Evidence exists to support a relationship between regular exercise and improvements in systemic low-grade inflammation (Gleeson et al., 2011), even in old age (Ogawa, Sanada, Machida, Okutsu, & Suzuki, 2011), along with in vitro and in vivo experiments (Saunders et al., 2006) suggesting that mechanical stimulation can inhibit osteoclast formation and activity by increasing OPG/RANKL ratio. Nevertheless, to date there are no reports documenting whether changes in bone-related inflammatory cytokines are associated with alterations in BMD in both older men and women after long-term exercise training.

Basic and clinical studies have established the relevance of biochemical markers of bone metabolism, showing an early response following treatment compared with BMD; and was proved to be useful for monitoring therapeutic response and efficacy on individual patients (Garnero, 2008). A combination of markers has been used to evaluate the rate of bone remodeling, including measuring predominantly osteoblastic or osteoclastic enzyme activities or assaying bone matrix components in blood and/or urine (Garnero, 2008). Currently, there are very limited data that have addressed the influence of long-term exercise (>12 weeks) on those biomarkers on older adults (Bemben et al., 2010, Vincent and Braith, 2002).

Therefore, the aim of the this study was to analyze several bone turnover and inflammatory biomarkers that may be associated with increased BMD after combined loading training in older adults. In addition, the alterations in balance and lower-extremity muscle strength as key factors associated with fall risk were also evaluated. We hypothesized that exercise would improve the bone-related inflammatory cytokines and bone turnover markers. Favorable alterations on BMD, muscle strength and balance were also expected.

2. Materials and methods

2.1. Subjects and experimental design

Subjects were recruited through advertisements in Porto area newspapers for participation in this university-based study. A total of 55 Caucasian older adults (29 women and 26 men) volunteered to participate in the study. The eligible subject pool was restricted to older adults with the following characteristics: aged 60–85 years, community-dwelling status, not engaged in regular exercise training in the preceding year, lack of use of bone-acting drugs and nutritional supplements known to affect bone metabolism (such as vitamin D and calcium) within the previous year, and lack of and significant sensory/cognitive impairment or medical conditions that contraindicated exercise participation. On the initial screening visit, all participants received a complete explanation of the purpose, risks, and procedures of the investigation and, after signing a written consent form, medical history and current medications of the subjects were documented. Participants were instructed to continue their daily routines and to refrain from changing their physical activity levels during the course of the experiment.

The baseline characteristics of the participants are given in Table 1. The study was carried out in full compliance with the Helsinki Declaration, and all methods and procedures were approved by the institutional review board.

Table 1. Baseline characteristics of the sample.

VariableWomen (n = 24)Men (n = 23)p value
Age (years)68.2 ± 5.768.2 ± 5.20.876
Education (years)7.9 ± 4.58.4 ± 3.60.373
Number of routine medications1.9 ± 1.82.2 ± 1.60.459
History of cigarette smoking (n/%)1/4.22/8.30.525

Anthropometry and body composition
 Weight (kg)64.2 ± 10.283.0 ± 11.7<0.001
 BMI (kg/m2)28.6 ± 4.129.2 ± 3.40.134
 Lean mass (kg)38.4 ± 4.854.9 ± 5.9<0.001
 Fat mass (%)37.8 ± 5.827.6 ± 27.2<0.001

Diet
 Energy intake (kcal/day)1444.6 ± 345.41618.3 ± 496.50.169
 Protein intake (g/day)68.7 ± 14.671.5 ± 19.40.578
 Calcium intake (mg/d)643.7 ± 337.9658.3 ± 253.30.868
 Phosphorus intake (mg/day)988.7 ± 299.61049.1 ± 319.20.507
 Vitamin D intake (μg/day)1.7 ± 1.91.5 ± 1.20.619
 Coffee intake (mL/day)62.4 ± 57.098.2 ± 48.50.025

Daily physical activity
 MVPA (min/day)80.4 ± 30.691.9 ± 31.50.294
 Daily counts per minute377.6 ± 123.6418.1 ± 139.10.383
 Daily step count9255.7 ± 3195.210,629.1 ± 9668.40.614

Bone mineral density
 Lumbar spine (g/cm2)0.848 ± 0.1211.039 ± 0.173<0.001
 Femoral neck (g/cm2)0.687 ± 0.1080.817 ± 0.0990.001
 Lumbar spine (T-score)−1.8 ± 1.2−0.4 ± 1.60.011
 Femoral neck (T-score)−1.5 ± 1.0−0.7 ± 0.90.014

BMI, body mass index; MVPA, moderate to vigorous physical activity.

2.2. Exercise protocol

The 32-week combined loading training involved odd-impact loading training performed once a week (Wednesdays) and high-magnitude joint reaction force loading through resistance training performed twice a week on separate days (Mondays and Fridays). Each session lasted approximately 60 min, and three physical education instructors specialized in PA for older adults, and supervised by the researchers led all sessions at the University of Porto – Faculty of Sport facilities.

The odd-impact training was designed to load bones with intermittent and multidirectional compressive forces, introducing atypical and novel stress on the bone, and to improve neuromuscular function. Each training session included six different components:

I)

A 10-min light stretching and warm-up exercise;

II)

15 min of weight-bearing activities, consisting of stepping exercise at a speed of 120–125 beats per minute using a 15-cm-high bench, bounding exercises, and heel-drops performed on a hard surface – a heel-drop consists of raising the body weight onto the toes and then letting it drop to the floor, keeping the knees locked and hips extended;

III)

Muscular endurance exercises performed concentrically and eccentrically for about 10 min, involving squats while wearing weight vests, hip flexors, extensors, and abductors; knee flexors and extensors and upper body exercises performed using elastic bands and dumbbells;

IV)

10 min of balance training with static and dynamic exercises (e.g. walking in a straight line, walking heel to toe, using additional resources such as ropes, sticks, balls, and balloons);

V)

10-min games where movements included directional elements that the body is not normally accustomed to, and agility training aimed at challenge hand-eye coordination, foot-eye coordination, dynamic balance, standing and leaning balance;

VI)

5-min of stretching.

For weight-bearing and strength exercises, the repetitions were increased from 8 to 15 and the number of sets increased to 3.

Resistance exercise training sessions involved the following: a standardized warm-up period (8–10 min) on a bicycle ergometer (Bike-Max; Tectrix, Irvine, CA) and/or rowing ergometer (Concept II, Morrisville, VR) at low intensity (50–60 rpm) and some stretching exercises; specific resistance training period (30–40 min) that included leg press and leg extension, seated leg curl, hip abduction, double chest press, lateral raise, overhead press, and abdominal machine; and a cool-down period (5–10 min) that included walking and stretching exercises. Subjects exercised on variable resistance machines (Nautilus Sports/Medical Industries, Independence, VA). To minimize fatigue, the exercises for the upper/lower body were performed in a non-consecutive way, with approximately 2 min rest in-between. Training intensity was gradually increased during the first four weeks. Participants underwent a 2-week familiarization period with the equipment and the exercises. The intensity of the training stimulus was initially set at 60% of one-repetition maximum (1RM), as determined at week 2, with a work range of three sets of 12–15 repetitions. Subjects then progressed from 75% to 80% of 1RM at a work range of six to eight repetitions (3 sets) and remained at this level until the end of the program. Training was continuously monitored by heart rate monitors (Polar Vantage XL, Polar Electro Inc., Port Washington, NY) and ratings of perceived exertion (Borg's 10-point psychometric scale) (Borg, Hassmen, & Lagerstrom, 1987). 1RM tests were performed every two weeks for the first month and then every four weeks until the end of the program. Between these tests, the load was increased for those subjects who were able to easily complete 12 or more repetitions for both sets.

The three research assistants were responsible for warm-up, cool down, and stretching exercises; the monitoring of correct lifting form; the appropriate amount of exercise and rest intervals; the maintenance of daily exercise logs; and the progression of the exercises. During training sessions, subjects were also encouraged to exercise with a training partner to provide additional motivation.

Exercise compliance was defined as the number of exercise sessions reported divided by the number of maximum exercise sessions possible.

2.3. Measurements

Participants were tested prior to the beginning of training (last week of September 2009) and after eight months of training (first week of June 2010).

2.3.1. Blood sampling and analysis

Fasting venous blood samples were drawn between 8 a.m. and 10 a.m. always on Mondays to ensure at least 2 days without exercise training. Serum samples were clotted at room temperature for 90 min after which they were centrifuged for 10 min at 1000 × g. Samples were aliquoted and stored at −80 °C until analysis.

Serum N-MID Osteocalcin (OC) was measured using an electrochemiluminescence immunoassay on a Cobas E module analyzer (Roche Diagnostics, Penzberg, Germany). For serum C-terminal telopeptide of Type I collagen (CTX) the serum CrossLaps enzyme-linked immunosorbent assay kit (Immunodiagnostic Systems Ltd, Boldon, UK) was used. Serum concentrations of IL-6, TNF-α, and IFN-γ (Human 3-plex Cytokine panel), serum OPG (Human bone panel 1A), and serum RANKL (Human RANKL Single Plex) were measured using Milliplextm map kits (Millipore, St. Charles, MO) in a Luminex® 200TM analyzer (Luminex Corporation, Austin, TX). Raw data (mean fluorescence intensity, MFI) were analyzed using ISTM 2.3 software (Luminex Corporation, Austin, TX). All these measurements were performed according to the manufacturers’ protocols. Standards and samples were measured in duplicate and assay methods had coefficients of variation-values (intra- and inter-assay) <9%.

Capillary blood samples were collected on the same occasion from the earlobe using a 50 μl lithium heparin-coated capillary tube and immediately assayed using the Cholestech LDX® Analyzer (Cholestch Corporation – Hayward, CA, USA) for determination of hs-CRP.

2.3.2. Bone mineral density measurements

BMD was measured using dual-energy X-ray absorptiometry (DXA; QDR 4500A, Hologic, Bedford, MA) at the lumbar spine (L1–L4) and proximal femur on the non-dominant side using standard protocols. To minimize interobserver variation, the same blinded technician carried out all analyses. Bone phantoms were scanned daily, and coefficients of variation were verified before and during the experimental period to ensure assessment reliability, as previously described (Marques et al., 2011a).

2.3.3. Balance and muscle strength measures

Dynamic balance was assessed using the 8-foot Up and Go test (Rikli & Jones, 1999a). Before starting the test, participants remained seated and rested for 5 min. The score corresponded to the shortest time to rise from a seated position, walk 2.44 m (8 feet), turn, and return to the seated position, measured to the nearest 1/10th's. Two attempts were allowed, with 1-min rest in-between, and the best performance was used for analysis.

Lower-extremity muscle strength was measured using the 30 s Chair Stand test (Rikli & Jones, 1999a). Participants were asked to sit in a 43-cm-high chair with arms crossed at the wrists and held against the chest. Participants completed as many “stand ups” as possible during 30 s. The score was the total number of stands executed correctly during 30 s. Both tests include normative performance standards established according to participants’ age and gender (Rikli & Jones, 1999b).

2.3.4. Potential confounding variables

Total body scans were taken using the same DXA instrument. Scans were analyzed for total lean mass (kg), and body fat mass (%). Body mass and height were recorded using a digital medical scale (Seca GmbH, model 708, Germany) and a height rod (Seca 220). Body mass index (BMI) was calculated using the standard formula: mass (kg)/height2 (m).

The Actigraph GT1M accelerometer (Manufacturing Technology, Fort Walton Beach, FL) was used as an objective measure of daily PA, as described previously (Marques et al., 2011b). For both test periods (pre- and post-training), four subjects had only two valid days. Those 4 participants were contacted and agreed to wear the accelerometer again for seven days (one week later than the rest of the group). The average daily moderate to vigorous PA (MVPA), number of steps, and daily activity counts per minute were analyzed.

Nutritional status was assessed using 4-day diet records over three weekdays and one weekend day. To ensure standardization of the dietary records, a dietician gave individual instruction to the subjects concerning how to fill out the diet records and assess food-serving sizes. Diet records were analyzed using Food Processor Plus® (ESHA Research, Salem, OR), which uses the table of food components from the US Department of Agriculture. Some traditional Portuguese dishes were added based on the table of Portuguese food composition. Total caloric, protein, calcium, phosphorous, vitamin D, and caffeine intakes were compared between men and women.

A baseline self-administered questionnaire to assess the impact of present and past lifestyle choices and clinical status was completed by interview to avoid misinterpretation of items and/or skipping of questions. Questionnaire included information regarding education, marital status, fall and fracture history, medical history, current medical conditions, medication use, as well as current and past smoking.

2.4. Statistical analysis

All statistical analyses were performed using SPSS Statistics (version 18; SPSS, Inc., Chicago, IL) for Windows with a significance level of 5%. Data were checked for normal distribution using the Shapiro–Wilk test, and the means ± standard deviations were calculated. Primary outcomes were changes from baseline in response to training in balance, muscle strength, BMD and serum bone metabolism markers and inflammatory cytokines. Secondary outcomes included 32-week changes from baseline in dietary intake, daily MVPA, and body composition (BMI, fat mass percentage, and lean mass). The results were analyzed on an intention-to-treat basis, and missing data due to lack of follow-up (the method assumed data were missing at random) was replaced using the process of multiple imputations. This method has been adapted to the analysis of longitudinal data (Mazumdar, Liu, Houck, & Reynolds, 1999). Between-group comparisons of continuous variables were performed using independent t-tests. Pearson's correlation coefficients were used to analyze the association between BMD of the femoral neck, total hip, and lumbar spine, body composition, bone biomarkers, and inflammatory markers. Correlation analysis (adjusted for gender) was also used to determine relationships among the variables with significant change over time.

The delta percentage was calculated with the standard formula: % change = [(post training score − baseline score)/baseline score] × 100, and the effect size for within-subjects (pre-test mean – post-test mean/pre-test standard deviation), was also calculated. An ES of 0.2 or less is considered small, an ES around 0.5 is moderate, and an ES of 0.8 or greater is large (Thomas, Nelson, & Silverman, 2005).

A two-way (group and time) factorial ANOVA, with repeated measures on one factor (time), was performed for main effects and time by group (males vs. females) interactions for each dependent variable. Main effects were considered when interactions were not significant. When significant interactions were found, Bonferroni post hoc tests were used to determine significant differences amongst mean values.

A power analysis based on a formulation of 75% power, an effect size of 0.5 for overall muscle strength, balance and BMD from previous studies, and a significance level of 5% deemed that a sample of 23 per group was sufficient to address the research questions.

3. Results

3.1. Subjects

Eight subjects did not meet selection criteria due to use of medication known to affect bone metabolism (n = 3), use of hormone replacement therapy (n = 2), and current involvement in water-based activities (n = 3). Forty of the original 47 subjects (women = 24, men = 23) who underwent the initial assessment completed the study (women = 20, men = 20). Dropout rates were similar amongst men and women. Three participants dropped out because of surgery, two participants dropped out due to medical issues unrelated to the intervention, and two other subjects because of personal reasons. One hundred percent compliance to the exercise sessions was set at 96 training sessions. Excluding dropouts, mean compliance to exercise sessions was 82.6% (60.0–100%). There were no exercise- or assessment-related (pre- and post-training) adverse events. In comparison to individuals who completed the trial, those who failed to provide follow-up data had lower daily MVPA level and lower performance in balance test. No significant differences in the remaining baseline measurements were found, including age, body weight, and inflammatory or bone-related variables. Our analysis included all subjects as results were analyzed on an intention-to-treat basis, and missing data due to lack of follow-up (the method assumed data were missing at random) were replaced using the process of multiple imputation. Yet, using the per-protocol analysis (n = 40), the results for all outcomes were similar, in direction and statistical significance, to intention-to-treat analysis.

Demographics and descriptive parameters of all groups are listed in Table 1. Of all the participants, 45% were overweight, almost half of them had hypertension, women never smoked, and only a small proportion of participants (6%) had a history of cigarette smoking (2 men and one women). On average, participants obtained 78 min of MVPA per day. Compared with men, women had significantly higher fat mass percentage (p < 0.001), lower weight and lean mass (p < 0.001), lower caffeine intake (p = 0.025), and significantly less BMD and T-score values.

3.2. Bone mineral density, balance, and muscle strength responses

There were no significant interactions between group and time on all measurements of BMD, balance and muscle strength (Table 2). A significant main effect of time for all balance, muscle strength, and bone variables was observed, excepting for femoral neck BMD. Accordingly, both men and women improved lower-extremity strength, the time to perform the balance test, and improved BMD at several bone sites, including lumbar spine, total hip, trochanter, and intertrochanteric region. However, the magnitude of the effect observed on muscle strength and balance was moderate in women, and low in men. In addition, effect sizes for BMD sites were low (<0.2). A significant main effect of group was also observed for all variables, excepting for muscle strength; thus, as expected, women had lower BMD at both time-points, and significant lower performance on the up and go test at baseline and after training.

Table 2. Pre- and post-training values, and effect sizes (ES) for proximal femur and lumbar spine BMD, dynamic balance and lower-extremity muscle strength.

VariableWomenESMenESp (Group)p (Time)p (Interaction)
Pre-trainingPost-trainingPre-trainingPost-training
Femoral neck (g/cm2)a0.715 ± 0.1190.705 ± 0.1040.080.822 ± 0.1130.821 ± 0.1150.010.0020.2410.406
Troch (g/cm2)a0.640 ± 0.0810.648 ± 0.0800.100.773 ± 0.1120.774 ± 0.1140.01<0.0010.0320.070
Inter (g/cm2)a1.031 ± 0.1421.041 ± 0.1390.071.169 ± 0.1651.172 ± 0.1630.020.0050.0380.276
Total hip (g/cm2)a0.864 ± 0.1080.872 ± 0.1110.081.004 ± 0.1401.006 ± 0.1380.010.0010.0210.129
Lumbar spine (g/cm2)a0.877 ± 0.1220.896 ± 0.1290.151.051 ± 0.1611.065 ± 0.1720.08<0.001<0.0010.427
8 ft Up and go (s)a5.33 ± 1.044.74 ± 0.540.564.55 ± 0.744.35 ± 0.510.280.004<0.0010.059
30 s Chair stand (rep)18.54 ± 4.4819.59 ± 3.510.4318.65 ± 4.2719.59 ± 3.110.220.7090.0040.304
a

Significant difference between groups at baseline and post-training, p < 0.05. 8 ft, 8-foot; BMD, bone mineral density; Troch, trochanter; Inter, intertrochanteric region; Rep, number of repetitions.

3.3. Bone markers and inflammatory cytokines responses to training

The 32-week exercise training did not significantly change in both bone turnover markers (OC and CTX) and in OPG, RANKL and their ratios (Table 3). OPG was significantly greater (p < 0.05) in the female group compared to the male group at baseline and post-training.

Table 3. Serum bone-related and pro-inflammatory markers responses to 32 weeks of exercise training.

VariableWomenESMenESp (Group)p (Time)p (Interaction)
Pre-trainingPost-trainingPre-trainingPost-training
OC (ng/mL)14.82 ± 3.6415.43 ± 4.12−0.1714.08 ± 2.8713.75 ± 2.800.120.1950.7170.225
CTX (ng/mL)0.38 ± 0.140.38 ± 0.150.010.37 ± 0.120.36 ± 0.120.120.7220.5710.667
OC/CTX42.91 ± 15.3543.83 ± 12.55−0.0639.45 ± 7.9240.67 ± 10.88−0.150.3090.4500.915
OPG (pg/mL)a514.61 ± 117.57503.45 ± 118.850.09432.80 ± 126.10423.81 ± 107.640.070.0180.2940.909
RANKL (pg/mL)29.64 ± 13.8226.53 ± 14.640.2327.20 ± 9.7226.53 ± 9.890.070.7220.0900.269
OPG/RANKL30.38 ± 39.6031.49 ± 36.85−0.0317.64 ± 6.8319.66 ± 12.32−0.300.1220.5570.865
IL-6 (pg/mL)1.18 ± 0.811.11 ± 0.910.091.62 ± 1.250.97 ± 0.84b0.520.5380.0420.013
TNF-α (pg/mL)7.30 ± 2.467.28 ± 2.160.017.39 ± 2.057.99 ± 2.52−0.300.5380.1640.149
IFN-γ (pg/mL)0.75 ± 0.510.48 ± 0.360.540.63 ± 0.430.44 ± 0.210.450.3930.0020.393
hs-CRP (mg/L)3.06 ± 2.272.54 ± 1.810.232.53 ± 1.951.63 ± 1.010.460.1330.0060.428
a

Significant difference between groups at baseline and post-training, p < 0.05.

b

Significantly different from baseline, p < 0.05. ES, effect size; OC, osteocalcina; CTX, C-terminal telopeptide of Type I collagen; RANKL, receptor activator of nuclear factor kappa B ligand; OPG, osteoprotegerin; IL, interleukine; TNF, tumor necrosis factor; IFN, interferon; hs-CRP, high sensitive-C reactive protein.

Inflammatory markers were not significantly different between groups at baseline. There was a significant treatment effect (decrease) for IL-6, IFN-γ, and hs-CRP (Table 3). A significant group x time interaction was observed only for IL-6, thus a different response between males and females was evident over time, supported by the significant decrease observed in the male group (effect size = 0.52 vs. effect size = 0.09). There was no significant interaction or main effects of group and time in TNF-α.

3.4. Change correlations

Adjusting to gender, there were no significant correlations between change values of inflammatory markers (IL-6, IFN-γ, and hs-CRP) and change in total hip, trochanter, intertrochanter and lumbar spine BMD.

3.5. Confounding variables

Total energy intake was similar amongst men and women at baseline and during the period of intervention. Energy intake was 1 530 ± 430 kcal/day at baseline and 1 452 ± 463 at 32-week. Dietary protein, phosphorus, caffeine, calcium, and vitamin D intake measured with a 4-day dietary record, remained unchanged after 32 weeks of intervention.

In total, pre- and post-training data from all participants were included in the analysis (47 files with seven valid days, 7 files with 6 valid days, and 11 files with 5 valid days). No significant changes in MVPA level were observed. There was no significant interactive (p = 0.210) or main effect of group (p = 0.102) and time (p = 0.326) on MVPA changes.

No significant interaction occurred for body composition in response to exercise intervention (Table 2). There was a main effect of group on lean mass (p < 0.001) and fat mass percentage (p < 0.001). Thus, women had significant lower lean mass and greater fat mass percentage on both time-points. However, the effect sizes for all body composition variables were low (<0.2) and changes in body composition were non-significant. Finally, no significant interactions were observed between the covariates fat mass and MVPA with all the main outcomes.

4. Discussion

Regular exercise training is consistently linked with a varied range of health-related benefits, including improvements in low-grade inflammation, BMD and bone metabolism. In fact, increased inflammation seems to be associated with a reduction in BMD, as key cytokines are associated with critical factors for bone remodeling (Schett, 2011). However, few comprehensive efforts have been made to characterize the relationship between changes in BMD, bone metabolism, and inflammatory response after long-term exercise training in older adults. We found that the 32-week resistance exercise combined with weight-bearing exercise training increased BMD and favorably modulated inflammatory markers in older adults. No changes were observed in OPG, RANKL, and OC and CTX (bone formation and resorption markers, respectively) after exercise. Finally, balance and lower body strength, as markers of fall risk factors, significantly improved after exercise training.

Evidence regarding exercise effects on bone mass and metabolism has been mostly based on women and single-stimuli exercise interventions, thus fewer studies have focus on elderly (age > 60y), particularly men (Marques et al., 2012). Human studies examining the effects of long-term exercise interventions in older men have shown positive results on both femoral neck and lumbar spine BMD (Bemben and Bemben, 2010, Kukuljan et al., 2011). All previous studies were based on evidence from 6 to 18 months resistance exercise training, which is consistent with our intervention. Our data also revealed that the relative improvements in BMD with training were similar for both men and women, supporting the findings of Bemben and Bemben (2010). We examined the effect of a combined training protocol, which included 1-day/week odd-impact exercise (such as aerobic or step classes, bounding exercises, agility exercises and games, and impact activities) with 2 days/week of resistance exercises, with site-specific exercises that applied mechanical loads to the skeleton. The positive results mediated by this type of exercise intervention are in agreement with results from a recent meta-analysis (Marques et al., 2012). However, as results were based on a disproportionate emphasis toward women, evidence to support the efficacy of exercise training on bone health in older men is warranted.

Data from exercise training and bone turnover markers are however inconsistent with some studies showing positive effects of exercise in attenuating bone turnover in aging adults (Karabulut et al., 2011). Others showed no effects on both bone formation and resorption markers (Yarasheski, Campbell, & Kohrt, 1997). RANKL, a member of TNF superfamily, is the key osteoclastogenic cytokine, because osteoclast formation requires its presence or its priming of precursor cells. OPG is a decoy receptor for RANKL and can block RANKL/RANK interactions, thus protecting bones from excessive bone resorption (Kostenuik & Shalhoub, 2001). Therefore, an up-regulation of OPG and a down-regulation of RANKL would inhibit osteoclast formation, and therefore prevent bone loss. Our previous work also demonstrated no changes in serum OPG and RANKL levels and their ratio after 8 months of exercise training (Marques et al., 2011b). Similarly, Esen et al. (2009) reported no significant changes in OPG levels after a 10-week walking program in middle-aged men, and a study using human cell lines showed that mechanical stimulation had not affected RANKL (Saunders et al., 2006). Our findings did not support the hypotheses that mechanical load may induce down-regulation of RANKL (Esen et al., 2009, Lau et al., 2010). Findings of the present study regarding changes in both bone turnover markers did not reach statistical significant level, in agreement with data reported by Ryan, Treuth, Hunter, and Elahi (1998) and Bemben et al. (2010). However, the ratio of OC to CTX increased 5% and the ratio of OPG to RANKL increased 15%, which may suggest a positive bone metabolism change. Previous exercise-based studies on bone metabolism were mostly short-term, lasting commonly 16 weeks, as bone marker responses to training are more rapid than BMD responses (Harris et al., 1993). Karabulut et al. (2011) found a significant increase in bone specific alkaline phosphatase (B-ALP) and B-ALP to CTX ratio after 6 weeks of resistance exercise in older men. Therefore, other results may have been observed if more serial measurements would have been taken over the training period, rather than only at baseline and after 32 weeks.

Although regular exercise training results in lower circulating levels of pro-inflammatory cytokines such as IL-6 and TNF (Gleeson et al., 2011), the association between inflammation and bone metabolism after long-term exercise is less clear. The mechanisms by which pro-inflammatory cytokine mediate bone damage have been postulated (Schett, 2011). Briefly, TNF exerts its effect on osteoclastogenesis by acting directly on osteoclast precursors, as well as indirectly, by upregulating the production of macrophage colony-stimulating factor and RANKL on mesenchymal cells (Lam et al., 2000). IL-6 can upregulate RANKL and thus indirectly support osteoclast formation via the interaction with mesenchymal cells (Udagawa et al., 1995). IFN-γ is also an important, yet controversial, osteoclastogenic-regulating factor (Takayanagi, Sato, Takaoka, & Taniguchi, 2005). In vitro, IFN-γ has a marked suppressor effect on osteoclastogenesis by inhibiting RANK signaling (Takayanagi et al., 2002, Takayanagi et al., 2000). However, the role of IFN-γ in vivo is more complex because it was shown to either decrease osteoclastic bone resorption, leading to an improvement of bone mass (Duque et al., 2011, Xu et al., 2009) or to increase osteoclastic bone resorption, leading to a decrease in bone mass (Gao et al., 2007). Therefore, a decrease in both IL-6 and TNF-α probably has some protective effect on bone loss by inhibiting osteoclast formation, while increased IFN-γ levels could support skeletal integrity.

Previous cross-sectional studies have linked high CRP levels with lower BMD (de Pablo, Cooper, & Buckley, 2012), higher levels of bone turnover markers (Kim et al., 2007), and greater risk of fracture (Pasco et al., 2006). Others have also associated elevated levels of pro-inflammatory cytokines with increased risk of bone loss (Pfeilschifter et al., 2002, Schett, 2011). The circulating levels of IFN-γ and hs-CRP decreased in both men and women, and IL-6 significantly decreased only in men in response to the training program (cf. Table 3). Of note, the observed decrease in inflammation was independent of weight loss, as no significant reductions on body weight or fat mass percentage occurred after the 32-week exercise intervention. The effect of exercise on the inflammatory profile was not due to the reduction of fat mass but to the exercise-induced muscle work per se. Indeed, most evidence concerning the exercise effect on inflammation had consistently focused on obese and/or type 2 diabetic subjects (Jorge et al., 2011, Silverman et al., 2009). In addition, as no significant changes were observed in lean mass, the exercise intervention was effective in maintaing and therefore counteract the common age-related decline in muscle mass. In accordance with our results, some previous studies involving elderly subjects found that exercise lead to a significant decrease in IL-6 (Nicklas et al., 2008, Prestes et al., 2009) and CRP (Martins et al., 2010, Ogawa et al., 2011), and no significant changes on TNF-α (Ogawa et al., 2011, Prestes et al., 2009). In addition, exercise-associated changes in IFN-γ have been poorly studied, and results from prospective training studies are controversial, as the IFN-γ production have been reduced (Golzari, Shabkhiz, Soudi, Kordi, & Hashemi, 2010) or did not change (Touvra et al., 2011) following a 8-week exercise training program in young patients. Yet, in our study, exercise induced a reduction in IFN-γ, which may inhibit is protective role on osteoclast differentiation and bone loss, but no associations were found between change in IFN-γ and bone biomarkers or BMD in response to our exercise program.

Finally, previous studies reported that exercise training is associated with balance and muscle strength improvements in healthy older adults (Marques et al., 2011a, Marques et al., 2011b); reinforcing the notion that exercise training has the potential to reduce fall risk in elderly people. In the present study, both older men and women significantly increased lower-extremity muscle strength and dynamic balance, which is in line with the prevailing evidence that exercise protocols that include a specific strengthening and balance component are the most effective exercise interventions for fall prevention compared with other modes of training (Sherrington et al., 2008).

The study does not have a control group so the within group differences reported over 32 weeks may be due to seasonal change, test familiarization and/or lifestyle change. However, relevant lifestyle-related variables (such as clinical and nutritional status, body composition, and daily MVPA) were measured as potential confounding variables. As no significant changes were detected, those variables probably did not account for the observed outcomes results. Due to the nature of the employed testing methodology, no familiarization is possible for physical activity, BMD and serum measurements. It is possible that the failure to detect changes in bone turnover marker and the modest improvement of BMD at several sites might be due to the aging process, and including a control group would possibly clarify this issue. Using imaging techniques to assess bone changes, exercise interventions with longer duration (more than 1 year) are clearly advised, however biomarkers can effectively detect changes within a smaller time period (after 3–6 months). Although a longer exercise interventions would be convenient, it would increase dropout rates, due to participants’ difficulties in managing both summer or vacation period and continuing attending to training sessions. While we did not conduct a randomized controlled trial (RCT), the improvements in femoral neck (0.6%) and lumbar spine BMD (1.7%) demonstrated in this study are consistent with RCTs (Bemben and Bemben, 2010, Marques et al., 2011a, Marques et al., 2011b). It should be noted that controlled trials may overestimate treatment effects compared to RCTs (Wolff, van Croonenborg, Kemper, Kostense, & Twisk, 1999), which is not perceptible in the present study. Although DXA is the method most commonly used to measure areal BMD (g/cm2) because of its speed, precision, low radiation exposure and availability of reference data (Watts, 2004), it cannot yield direct measures of bone strength.

This study adds to the current literature by focusing on the relationship between a structured exercise program and bone changes, including BMD measurements and bone metabolism markers, and inflammatory cytokines in this particular population. Another unique aspect of our study was the inclusion of both men and women, which allowed gender comparisons of the responses to training. Moreover, this study has considered the possible influence of critical confounding variables such as body composition, daily MVPA levels objectively measured by accelerometers as well as dietary intake.

In conclusion, our results support the beneficial role of long-term exercise training on lumbar spine and proximal femur BMD and inflammatory markers in elderly subjects. The exercise-induced anti-inflammatory effect seems to be dissociated from any effect on weight loss or lean mass. Moreover, we found no indication of an exercise-induced effect on biochemical markers of bone turnover. We have also demonstrated that exercise training elicits significant gains in muscle strength and balance. The relationship between exercise and changes in BMD as well as bone metabolism markers, and inflammatory response needs to be further explored given the public health importance of bone fragility and susceptibility to falls with aging.

Conflict of interest statement

None.

Acknowledgments

The authors thank Gustavo Silva for his kind support in biochemical assays and Andreia Pizarro for carrying out BMD measurements by DXA. This research was funded by individual grants SFRH/BD/36319/2007 and SFRH/BSAB/1025/2010 from Portuguese Foundation of Science and Technology.

References

View Abstract
Copyright © 2013 Elsevier Ireland Ltd. All rights reserved.