HomeLarge Type Edition
HOME ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
PubMed
Right arrow PubMed Citation
The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 57:M203-M208 (2002)
© 2002 The Gerontological Society of America

Maintenance of BMD in Older Male Runners Is Independent of Changes in Training Volume or V·O2peak

Robert A. Wiswella, Steven A. Hawkinsb, Hans C. Dreyera and S. Victoria Jaquea

a Department of Biokinesiology and Physical Therapy, University of Southern California, Los Angeles
b Department of Kinesiology and Nutritional Science, California State University, Los Angeles

Robert A. Wiswell, Department of Biokinesiology and Physical Therapy, University of Southern California, 1540 East Alcazar Street, CHP 155, Los Angeles, CA 90033 E-mail: wiswell{at}hsc.usc.edu.

Accepted November 8, 2001


   Abstract
Top
 Abstract
Methods
 Results
 Discussion
 References
 
Background. The results of several recent research studies have questioned the benefit of chronic running in reducing the risk of developing osteoporosis. These negative conclusions are based upon the findings that runners, in general, are not different from controls with regard to their bone mineral density (BMD). This has led one to speculate that the value of weight-bearing exercise might be related to its ability to maintain bone mass or, perhaps, decrease the expected rates of loss.

Methods. We examined a group of 54 male master athletes ranging in age from 40 to 80 years longitudinally over a 5- to 7-year intervening period. Physiological parameters included V·O2peak, body composition, bone density, and running performance. Medical histories and training records were obtained via questionnaire.

Results. Over the average 4.6 years between tests one and two, significant increases in body weight and lean body mass were observed. Aerobic fitness declined, as did weekly mileage and 5K and 10K performance times. Bone mineral density was lower in the whole body but not in the hip or the spine. Finally, we report no significant relationship between the change in training volume, change in body weight or lean mass, and change in aerobic capacity with changes in BMD.

Conclusions. Hip and spine BMD are maintained over a 4- to 5-year period in master runners. Furthermore, changes in bone density or content are not sensitive to moderate changes in training volumes. We conclude that bone density can be maintained by running in older active men. These findings suggest that if a minimal threshold of mileage is required, the level is certainly below the average mileage of master runners.

STUDIES investigating the influence of running mileage and/or marathon running on bone mineral density (BMD) in young and middle-aged runners report little differences between runners and age-matched controls (1)(2). MacDougall and colleagues reported that mileage (from 5–70+ miles per week) had little influence on BMD even though the tibia and fibula cross-sectional area, when normalized to body weight, tended to be greater as mileage increased (3). Goodpaster and colleagues reported in a group of men 42–73 years of age that distance runners did not have significantly different hip or spine BMD than individuals who ran short distances or not at all (4). From these and other findings, it has been suggested that weight-bearing exercise (e.g., running) will not lead to increases in BMD or bone mineral content (BMC) but may help to retard the rate of loss in skeletal mass associated with aging. This lack of weight-bearing exercise effect on increasing BMD in runners has been attributed to (i) decreased body mass (5)(6), (ii) inadequate loading (7), and (iii) decreased testosterone and estrogen associated with chronic training (2).

To address the issue of direct effect of running on bone maintenance, one would need to follow subjects over several years and relate changes in bone to changes in training volume and training intensity. It is only by a longitudinal comparison of active subjects that one could determine the extent of the benefit or detriment associated with chronic running. Therefore, to test the hypothesis that BMD is maintained in runners who train at high running intensities or volume, we examined 54 male runners who are participating in a 20-year longitudinal study started in 1986 (5). It was our specific purpose to relate changes in training volume (miles run per week and days per week of exercise) and V·O2peak to changes in whole body, spine, and hip BMD (by dual-energy x-ray absorptiometry) over a 4- to 5-year period.


   Methods
Top
 Abstract
 Methods
Results
 Discussion
 References
 
Design
This study evaluated 54 male master runners, ranging in age from 39 to 80 years. All subjects were actively training and competing in their respective athletic events at the time of the baseline (T1) measurement. All are currently participants in a 20-year longitudinal study at the University of Southern California (USC). The USC Institutional Review Board approved the protocol, and all subjects provided written informed consent upon entry into the study. (For a detailed description of methods, see reference 5.)

Body Composition Assessment
Residual lung volume was assessed utilizing the oxygen dilution technique (12), and body composition was determined via hydrodensitometry utilizing the Brozek equation. Three-compartment body composition was measured by dual energy x-ray absorptiometry (DEXA) simultaneously with the whole body scan using a Hologic QDR 1500 (Hologic, Inc., Bedford, MA).

Fitness Assessment
Oxygen uptake at fatigue (V·O2peak) was determined using a continuous, graded exercise test on a motor-driven treadmill. The test began at 2.5 mph and 0% grade and increased by 0.5 mph and 2%, respectively, every 2 minutes during exercise. Exercise continued until the subject terminated the test at a point of subjective exhaustion. All metabolic parameters were measured using a SensorMedics 2900 metabolic measurement cart (SensorMedics Corp., Yorba Linda, CA) interfaced with a Marquette electronic MAX-1 exercise electrocardiogram (EKG) system (Marquette Electronics, Inc., Milwaukee, WI). The flow meter and gas analyzers were calibrated daily. The EKG was monitored continuously throughout exercise and during 5 minutes of recovery.

Bone Mineral Density Assessment
Bone mineral density and bone mineral content were measured using the Hologic QDR 1500 DEXA (v. 7.1). Normalized values (T- and Z-scores) were generated from a gender-matched, and in the case of hip scores ethnicity-matched, control data set provided as part of the DEXA software. For the purpose of discussion in this article, we will use the World Health Organization (WHO) guidelines in defining clinical osteopenia (T-score < -1.0 and > -2.5) and osteoporosis (T-score < -2.5) (8). We recognize that these values are based upon relative risk of osteoporosis in postmenopausal women; however, we believe that these guidelines can be used for comparative purposes in a constructive way. The sample size of those with whole body scans was only 36 in that several subjects were evaluated in the lab prior to the university's acceptance of the whole body scan as a research tool.

Performance
Subjects self-reported training and performance data via questionnaire. Subjects were asked to report years of training, distance/time trained per week, days trained per week, best performances (5K, 10K, marathon) for each year they had been competing, and best performances within 2 months of the testing date. Subject responses were confirmed by oral interview on the day of testing.

Statistical Methods
All statistics were performed using the SPSS (v. 10.1) statistical software (SPSS Inc., Chicago, IL). Paired sample t tests were utilized to examine mean T1 and T2 differences. Linear regression was used to provide the magnitude and direction of the relationships among variables. When relating bone density to training and performance variables, partial order correlation was used, controlling for change in body weight or lean body mass (p was considered significant at the .05 level).


   Results
Top
 Abstract
 Methods
 Results
Discussion
 References
 
Table 1 presents the descriptive values for the subjects at T1 and T2. Test–retest reliabilities are provided, as well, and paired t tests are presented to indicate if significant time differences occurred. For most variables, even with 4.6 years intervening, there is excellent reproducibility. We recognize that bivariate correlations do not address the issues of precision and that without a measure of concordance, one might overlook data bias. To address the issue of bias, we ran a regression analysis on the T1 and T2 data to ensure that the data approximated a slope of 1.0 and an intercept of 0.0. When slope differences were observed, they were reported in the specific results sections.


View this table:
[in this window]
[in a new window]
  		Table 1. Subject Description at Baseline and Present (N = 54; means ± SD)

 
Subjects significantly increased body weight (1.4%) and body fat (1.0%) with a concomitant increase in the derived body mass index (BMI) (2.6%). Significant decreases were also observed for V·O2peak expressed in absolute or relative oxygen uptake.

A summary of the differences in training volume and running performance is presented in Table 2 . The disparity between the actual number of subjects (N = 54) and the change in running performance (n = 39) relates to the fact that several runners actually changed their competitive event over the intervening years. The results indicate that as these male runners get older, they decrease both mileage and the number of days per week they train. In this sample of older athletes, running performance significantly declined with age in the 5K and 10K events. Marathon times did not differ.


View this table:
[in this window]
[in a new window]
  		Table 2. Training Histories and Running Times (N = 54; means ± SD)

 
Thirty-seven percent of the runners in this study had hip or spine T-score values below -1.0, suggesting low BMD. Thirteen runners (24%) had T-scores for the spine between -1.0 and -2.5; and four subjects (7.4%) had spine T-scores below -2.5. Ten athletes (18%) had hip T-scores below -1.0; none were below -2.5. Interestingly, only three subjects demonstrated low hip T-scores without similarly low spine values, while nine subjects had low spine values without reduced hip scores. In order to identify factors that would discriminate between those with differences in BMD, we reclassified the sample into three subgroups as to whether they had T-scores greater than -1.0 [G1 (n = 35)], between -1.0 and -2.49 [G2 (15)], or less than -2.5 [G3 (4)]. Using a general linear model in the multivariate analysis of SPSS, significant differences were observed between G3 and G1 for lean body mass (LBM) (G1 = 61.5 ± 6.9; G3 = 54.21 ± 1.46 kg), and a significant difference was observed in the direction and magnitude of change in spine BMD between G1 and G2 with G3 (G1 = 0.006 ± 0.03 gm/cm2; G2 = 0.005 ± 0.03; G3 = -0.004 ± 0.01). Furthermore, while the differences were not statistically significant, real number differences were observed; G3 had lower body mass (9.1%) (G1 = 74.5 ± 9.3 kg; G3 = 67.8 ± 6.8) and higher V· O2peak ml · kg-1 · min-1 (14%) (G1 = 48.3 ± 9.9; G3 = 55.1 ± 15.2). These numbers suggest that those individuals who have the lowest bone density are likely to be lighter and more aerobically fit. No differences were observed for mileage, days per week of training, age, or percentage of body fat among these groups.

The longitudinal changes in BMD and BMC are presented in Table 3 . No significant change in whole body BMD or BMC was observed in these subjects. A significant increase in spine BMC and BMD and an increase in hip BMC occurred. The adjusted Z-score increased significantly for several sites—total spine (p < .016), total hip (p < .002), femoral neck (p < .006), and trochanter (p < .025). The increase in Z-score indicates that these runners exhibit a slower decline from peak bone mass than expected for individuals of similar age but not necessarily similar activity patterns. As reported, the direction of bone change over time for the hip in these athletes is toward bone maintenance not loss.


View this table:
[in this window]
[in a new window]
  		Table 3. Bone Density and Content (N = 54)

 
The left three panels in Fig. 1 present graphically the relationship between age and BMD in this sample. The data are present for both the T1 and T2 visit. The T1 WB (whole body) BMD and hip BMD declined significantly with age. The magnitude of the decline when expressed by the R2 (coefficient of determination) indicated a very small explained variance (5–10%) between age and these variables. T2 BMD values were not significantly related to age. The three graphs in the right panel of Fig. 1 provide evidence for the lack of association between changes in training volume and any of the BMD parameters. A dotted line has been placed in these graphs to separate the group into quadrants. The importance of this procedure is to demonstrate the number of subjects who contradict that any association exists in these parameters. That is, individuals in either (+,-) quadrant speak against the postulation of a significant relationship between the variables. Further, no significant relationship was observed between change in mileage and change in BMD.



View larger version (36K):
[in this window]
[in a new window]
  		Figure 1. The relationship between bone mineral density (BMD) and age is presented in the figures in the left panel. (•) represents the T1 values while ({circ}) represents the T2 values. Regression equations are presented only when a significant r value was observed. The relationships between changes in BMD and changes in miles run per week are presented in the right panel graphs. NS = not significant.

 
Fig. 2 presents graphically the percentage changes in a number of select variables. It can be observed that significant declines in aerobic power and mileage do not relate to significant changes in hip or spine BMD. The 12% reduction in mileage was not related to changes in BMD or BMC, V·O2peak, nor running performance in the 5K or marathon events. However, a significant relationship between change in 10K performance was related to changes in spine BMD (r = .530; p = .042). This suggests that greater increases in 10K time would indicate a greater decline in spine BMD.



View larger version (16K):
[in this window]
[in a new window]
  		Figure 2. Percentage differences (M ± SEM) in bone and fitness variables. (T1 value is represented as the 0 line on the Y-axis.) WB = whole body; BMD = bone mineral density.

 
To address the issue of the confounding variables, we performed partial order correlations to account for the influence of change in weight, LBM, V·O2peak expressed in absolute or relative units, change in age, and change in mileage on the relationship between select variables and the changes in BMD over the time period. No significant relationship between change in mileage and change in bone density was observed when changes in age, weight, or LBM were accounted for in the analysis.


   Discussion
Top
 Abstract
 Methods
 Results
 Discussion
References
 
The most remarkable findings of this study were that (i) changes in BMD and BMC are not sensitive to moderate changes in training volume or aerobic capacity, (ii) hip and spine bone density are maintained over 4 to 5 years resulting in improved age-adjusted Z-scores, and (iii) lean mass and/or changes in lean mass may be the most statistically important determinant of one's likelihood of being diagnosed with osteoporosis. Furthermore, evidence for a site-specific protection of the hip over the spine was observed in this sample of older male runners.

MacDougall and coworkers (3) reported no differences in trunk, spine, pelvis, or thigh BMD regardless of the running mileage in a group of 53 runners ranging in age from 20 to 45 years old. In fact, the BMD in these sites was not significantly different from a sedentary control group of similar age. Their data on this younger sample of runners are similar to our findings, that is, mileage is not a significant predictor of BMD. The mileage in their study was between 5 and 10 miles per week in the low group to 60–75+ in the high mileage group and was quite similar to our subjects, who ranged from 2 to 90 miles per week. Our findings extend their conclusion that mileage is not a significant predictor of BMD to the contention that changes in mileage over a number of years are not predictive of changes in BMD over time.

Karlsson and colleagues (9), studied 273 sedentary men over a 3-year period and reported, in sedentary men older than 50 years old, a 0.1% year loss in whole body BMD, a 1.5% decrease per year in femoral neck BMD, and a significant 0.45% per year gain in L2–L4 BMD. Unlike their findings, we report a significant greater loss in whole body BMD (0.27% per year) while neither hip nor spine BMD changed significantly. Our active men did increase spine density by 0.06% per year, but the difference was not significant and the real number value is lower than the 0.45% reported by Karlsson's group (9). It is likely that the increased spine density in both studies is not truly a reflection of changes in bone quality but perhaps a result of excessive compression and degeneration of the vertebrae associated with normal aging. The fact that the spine and hip BMD are maintained, while the WB BMD declines, suggests a site-specific influence of distance running.

The relatively high number of runners (37%) who were less than -1.0 standard deviations below peak BMD estimates for the spine versus the lower percentage at the hip (18%) was puzzling. One explanation of this finding is that the chronic impact of running accelerates bone loss at the spine. This was not the case in this study in that spine (L2–L4) BMD actually increased over the 3.6 years of the study. Second is that running does nothing to the prevalence of osteopenia at the spine but reduces it at the hip as a direct effect of specific hip loading during running. A final explanation may be that the prevalence of spine osteopenia is normally higher than at the hip. Kelsey reported fracture prevalence for the hip of 6% and spine of 5% in men, refuting the premise that differences in our hip/spine osteopenia rates are linked to fracture prevalence (10).

It is difficult to suggest any influence of running on fracture incidence until more longitudinal studies have been conducted. In dogs, Puustjarvi and colleagues (11) reported that chronic running led to a decrease in BMD versus sedentary dogs; however, collagen fibrils became reorganized into a more parallel manner in the runners, which the authors suggested accounted for maintenance of strength properties in bone. In other words, the benefit of exercise in this example was to change the architecture, not the bone mineral density or content. If this is true in humans as well, one could postulate that selective remodeling within specific bone sites more than compensates for the lack of distance running effect on BMD. Obviously, this is only conjecture, but several studies on the value of exercise on reduction of fracture risk may give credence to such speculation.

In conclusion, these data are presented as a longitudinal comparison of physiological variables in a group of older master runners. These data suggest that BMD can be maintained as a result of continuous training in male runners. They also suggest that changes in training patterns (either increases or decreases in days per week of training or miles per week of running) have little influence on the ability to maintain skeletal mass. It is obvious that future studies should increase the sample size and should include control subjects who either do not run or run a limited amount (<2–3 miles/week).


   Acknowledgments
 
This research was supported by the Wadt Memorial Research Fund and the Pickford Foundation. Currently, Steve Hawkins is in the Department of Kinesiology and Nutritional Science at California State University, Los Angeles.

Received October 29, 2001


   References
Top
 Abstract
 Methods
 Results
 Discussion
 References
 

  1. Morley JE, 2000. Editorial: The aging athlete. J Gerontol Med Sci. 55A:M627-M629. [Free Full Text]
  2. Aloia JF, Cohn SH, Babu T, Abesamis C, Kalici N, Ellis K, 1982. Skeletal mass and body composition in marathon runners. Metabolism. 27: (12) 1793-1796.
  3. MacDougall JD, Webber CE, Martin J, et al. 1992. Relationship among running mileage, bone density, and serum testosterone in male runners. J Appl Physiol. 73: (3) 1165-1170. [Abstract/Free Full Text]
  4. Goodpaster BH, Costill DL, Trappe SW, Hughes GM, 1996. The relationship of sustained exercise training and bone mineral density in aging male runners. Scand J Med Sci Sports. 6:216-221. [Medline]
  5. Wiswell RA, Hawkins SA, Jaque SV, et al. 2001. Relationship between physiological loss, performance decrement, and age in master athletes. J Gerontol Med Sci. 56A:M618-M626. [Abstract/Free Full Text]
  6. Ravaglia G, Forti P, Maioli F, et al. 2000. Body composition, sex steroids, IGF-1, and bone mineral status in aging men. J Gerontol Med Sci. 55A:M516-M521. [Abstract/Free Full Text]
  7. Frost HM, 1997. Why do marathon runners have less bone than weight lifters? A vital-biomechanical view and explanation. Bone. 20: (3) 183-189. [Medline]
  8. World Health Organization. Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis. WHO Technical Report Series. Geneva: WHO; 1994;843:1–129.
  9. Karlsson MK, Obrant KJ, Nilsson BE, Johnell O, 2000. Changes in bone mineral, lean body mass and fat content as measured by dual energy x-ray absorptiometry: a longitudinal study. Calcif Tissue Int. 66: (2) 97-99. [Medline]
  10. Kelsey JL, Browner WS, Seeley DG, Nevill MC, Cummings SR, 1992. Risk factors for fractures of the distal forearm and proximal humerus. The study of osteoporotic fractures research group. Am J Epidemiol. 135: (5) 447-489.
  11. Puustjarvi K, Nieminen J, Rasanen T, et al. 1999. Do more highly organized collagen fibrils increase bone mechanical strength in loss of mineral density after one-year running training?. J Bone Miner Res. 14: (3) 321-329. [Medline]



This article has been cited by other articles:


Home page
 Journals of Gerontology Series A: Biological Sciences and Medical SciencesHome page
M. R. Deschenes, J. A. Carter, E. N. Matney, M. B. Potter, and M. H. Wilson
Aged Men Experience Disturbances in Recovery Following Submaximal Exercise
J. Gerontol. A Biol. Sci. Med. Sci., January 1, 2006; 61(1): 63 - 71.
[Abstract] [Full Text] [PDF]

Home page
 Journals of Gerontology Series A: Biological Sciences and Medical SciencesHome page
J. E. Morley
The Need for a Men's Health Initiative
J. Gerontol. A Biol. Sci. Med. Sci., July 1, 2003; 58(7): M614 - 617.
[Full Text] [PDF]

Home page
 Postgrad. Med. J.Home page
J A Todd and R J Robinson
Osteoporosis and exercise
Postgrad. Med. J., June 1, 2003; 79(932): 320 - 323.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
PubMed
Right arrow PubMed Citation

HOME ARCHIVE SEARCH TABLE OF CONTENTS