Bone mechanical properties after exercise training in young and old rats DIANE M. RAAB, EVERETT L. SMITH, THOMAS D. CRENSHAW, AND D. PAUL THOMAS Biodynamics and Biogerontology Laboratories and Department of Meat and Animal Sciences, University uf Wisconsin, Madison, Wisconsin 53705 RAAB, DIANE M., EVERETT L. SMITH, THOMAS D. CRENSHAW,AND D. PAU~T~OMAS.Bonemechanicalp~o~ertiesctfter exercise training in young and old rats. J. Appl. Physiol. 68(l): 130-134,1990.-The effects of a lO-wk training regimenon the mechanicalpropertiesof the femur and humeruswere evaluated in 2.5 and 25mo-old Fischer 344 female rats. The rats trained on a rodent treadmill 5 days/wk for 10 wk. Duration, grade, and speedincreaseduntil the rats maintained 1 h/day at 15% gradeand either 15m/min (old rats) or 36 m/min (young rats). Excised boneswere mechanically tested with a 3-point flexure test for mechanicalproperties of force, stress,and strain. Fatfree dry weight (FFW) and moment of inertia were also obtained. With aging, similar increaseswere observedin both the femur and humerus for FFW, moment of inertia, and force. Ultimate stresswasreducedin the senescentfemur while strain was elevated; a similar but nonsignificant trend was observed in the humerus.Irrespective of age,training increasedFFW in the femur and, to a lesserdegree, in the humerus. Breaking force was elevated for both bonesafter training. In young and old bones, the training-induced differences in bone massand force were similar, despite differences in training intensity. In the old trained rats, femur ultimate stresswas greater than that in control rat femurs and similar to that in young rat femurs. The results of the present study indicate that training effects were not limited by age. aging; morphometry; strength
RATS, KNOWN for growth
through adulthood, display signs of senescence with advanced age. The body weight of male Fischer 344 rats peaks at -75% of their maximal life span and decreases after -90% of their life span (20). Compared with young adult rats, the bones of old rats have equal or greater cortical area, width, and length (3, 5, 6, 11). Although they maintain normal Ca, P, and collagen concentrations (mg/g), the bones from old rats have a lower density (g/cm3) than those from young rats, indicating greater porosity with senescence (3). These changes in density may affect bone mechanical properties. Bone force increases with an increase in bone size, but stress (force per unit area of bone) has been found to decrease with age despite an increase in bone mass (5, 6) *Exercise training may stimulate bone formation and increase bone mass and strength (7, 16, 18, 19). Studies that have evaluated the effects of exercise on bone mechanical properties in rats have primarily used immature animals (7, 10, 16, 19). Such models are inappropriate 130
for inferences to adult skeletons, because there may be an interaction of growth and training. The senescent rat model may provide insights to help understand exercise effects in humans. Osteoporosis, diagnosed by fracture due to low bone mass and strength, is a common health risk for older women. By 90 yr of age, an estimated one-third of all women will experience an osteoporotic hip fracture (12). If exercise training can increase bone mass and the ability of senescent bone to withstand force, it may decrease the risk for fracture in older adults. Depending on animal species, age, length, and intensity of exercise training, various experiments have shown greater, lesser, or no differences in bone weight, density, size, and strength (1, 7,10,16, 18, 19). Training initiated with mature, 19- to 26-mo-old rats has resulted in increased bone weight, Ca, and hydroxyproline content with no increase in bone diameters or strength (3,11). It is unclear whether the lack of change in diameters or strength resulted from differences in methodology between studies, an interaction with growth, or a decrease in response to mechanical stimulation with age. Previous research has concentrated primarily on growing animals and has failed to consider the mechanical effects of training imposed on the mature skeleton. The current study was designed to examine the role of age on training-induced adaptation of weight bearing bone, specifically the humerus and femur. Bones from young adult and senescent rats, trained and untrained, were compared for differences in size and mechanical properties. METHODS
Fischer 344 virgin female rats, ages 2.5 (young, Y) and 25 mo (old, 0), were obtained from the National Institute on Aging (specific pathogen free) colony at Indianapolis, IN. Throughout the experiment the rats were maintained in a 20-22°C room with a 12-h light/dark cycle. The animals were individually housed in wire cages and allowed ad libitum access to food (Wayne Rodent Blox, Continental Grain) and water. The feed contained 1.46% Ca and 0.99% P. These levels exceed those recommended for growth by the National Research Council (13). Two weeks were allowed for acclimation before the initiation of the study. All rats walked on a rodent treadmill before assignment to treatment groups. Rats from each age group were randomly assigned to either a sedentary
0161-7567/90 $1.50Copyright0 1990the AmericanPhysiologicalSociety
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
BONE PROPERTIES
WITH
control (C) or training (T) group. The initial group consisted of 10 YC, 10 YT, 8 OC, and 9 OT. Training protocol. All rats from the same age group trained simultaneously on a lo-chamber rodent treadmill (Quinton model 42-15). An electric grid at the rear of the belt was used as the running stimulus. All rats trained 5 days/wk for 10 wk with speed and grade, and duration progressively increased. The old rats began training at 5 m/min and 0% grade for 10 min/day. Speed and duration were increased until, by the 4th wk, the OT rats ran at 15 m/min for 60 min/day. Thereafter the speed and duration were maintained, but grade was gradually increased to 15%. Young adult rats have a greater maximal oxygen uptake than old rats (9); therefore, the YT rats were exercised at a greater absolute intensity than the OT group. The young adult rats began training at 10 m/ min and 5% grade for 15 min/day. By the end of the 2nd wk, the YT rats ran at 15 m/min, 15% grade for 60 min/ day. Thereafter, the grade and duration were maintained but speed was increased 2-3 m/min each week. By the 10th wk, the YT rats were running at 36 m/min and 15% grade for 60 min/day. Throughout the experiment, control rats were placed on the stationary treadmill for 5 min/day, 5 days/wk. All rats were weighed twice each week. Weights were used as an indication of growth and health status. Data collection. After the lo-wk training program, the rats were anesthetized with chloral hydrate and killed by a 1 ml cardiac injection of saturated KCl. The right femur and humerus were removed, cleaned of all soft tissue, individually sealed in plastic bags, and frozen at -20°C. While in the sealed bags, the bones were thawed to room temperature on the day of mechanical testing. Freeze-thaw procedures do not alter the mechanical properties of bone (17). Immediately before the mechanical test, the bone length was measured and the midpoint marked. The mechanical properties of the femur and humerus were determined using a three-point bending test with an Instron Universal Testing Machine. In the threepoint bending test, the bone was rested on two support plates 11.48 mm apart. Force was applied at the midpoint of the bone by a plunger moving downward at a constant rate of 5.08 mm/min. The bone was simultaneously tested in compression on the concave surface and tension on the convex surface. The force withstood was dependent on the shape of the bone cross section, the amount and distribution of bone mass, and the distance between the support plates. Force was applied to the femur midshaft from the anterior to the posterior surface and to the humerus midshaft from the lateral to the medial surface. Orientation was established for maximum stability and to prevent rotation during the test. The amount of load withstood by the bone and the bending or deformation that occurred during each test was recorded on an X-Y plotter as a load-deformation curve (Fig. 1). The moment of inertia, a measurement of the area distribution about the neutral axis of the bone cross section, is dependent on the bone’s cross-sectional area and shape. The midshaft cross section of each bone was
EXERCISE
131
AND AGING
z 0 I
Deformat ion 1. Typical load-deformation curve generated during a 3-point bending test. y, Yield or bending point; U, ultimate or breaking point; horizontal axis, vertical deformation of the bone as plunger moved downward at a constant rate of 5.08 mm/min; vertical axis, load on bone or force with which bone resisted deformation. White area, elastic region where no permanent damage occurs to bone; hatched area, plastic region where bone is permanently damaged. FIG.
cut 1 mm distal to the point of fracture. The cross section was photographed and enlarged N-fold. The AutoCAD program (Autodesk, Millvalley, CA) interfaced with a digitizing pad and was used to determine the centroid and moment of inertia about the axis perpendicular to the line of force application (femur: medial-lateral axis; humerus: anterior-posterior axis). Fat-free dry weight was determined after the bones were extracted in 100% ether for 5 days, allowed to dry at room temperature for 20 h, and subsequently dried at 80°C for 24 h. Computation of mechanical properties Mechani cal properties of th .e fe mur and humerus were calculated bY use of data from the load-deformation curves (Fig. 1) at both the yield and ultimate points (4). The yield point is the transition from the linear elastic phase (deformation is proportional to force) to the curvilinear or plastic phase (deformation increases more rapidly than force). The yield point occurs when the applied load or force creates permanent damage within the bone. If the load on the bone is removed before the yield point, the bone wil l return to its original shape and retain its original properties. The ultimate point is the point where load reaches a maximum and the bone fractures. Bending moment or the force (F) the bone withstood is calculated from the load required to break the bone distance between the two lower and the length support plates F = load x L/4 (kg* mm) Stress is the force withstood per unit area of bone, Stress is calculated from the force, the moment of inertia (I), and C, which is half the bone diameter measured parallel to the line of force application stress = F x C/I (kg/mm’) Strain is a measure of the bending (deformation) of the bone relative to its length. The terms C and L are defined above for stress and force strain = 12 x deformation x C/L The results were analyzed by a two-way analysis of
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
132
BONE PROPERTIES
WITH
variance procedure: the main effects were age and training. Results from each bone were analyzed separately. Differences were considered significant at P < 0.05. RESULTS
The results reported are based only on the rats that completed the IO-wk experiment (YC = 10, YT = 8, OC = 6, and OT = 8). During the study, two OC and one OT rats died and two YT rats were eliminated due to foot injuries that prevented further training. Age effects on growth. Young rats in both groups gained weight (P < 0.05) during the lo-wk study. Average weights at 2.5 mo were (mean t SE) 153.9 t 11.8 g (YC) and 153,6 t 12.2 g (YT) and at 5 mo were 199.0 t 4.5 g (YC) and 203.6 t 4.2 g (YT). In contrast, both senescent groups gradually lost body weight from 278.8 -I- 23.4 g (OC) and 281.8 t 22.5 g (OT) at 25 mo to 264.2 k 7.4 g (OC) and 259.5 t 8.0 g (OT) at 27.5 mo, Body weight was not affected by training in either age group. The old rats were larger and heavier than the young adult rats. Both the femur and humerus from old rats were longer and had greater fat-free dry weights and greater moments of inertia than those from young rats (Table 1). Although skeletal size increased with age, the fat-free dry weight of the bones remained a constant percentage of total body weight. The femurs averaged 0.21 and the humerus 0.10% of total body weight. Training effects on bone grcwth. Irrespective of age, the femurs from trained rats had a greater (P c 0.05) fat-free dry weight than those from control rats (Table 1). A similar trend was observed for the humerus, but the training-induced increase in weight was not significant. The relative weight of the femur to total body weight was greater (P C 0.05) in trained than in control rats within age groups, but there were no differences between age groups. Training did not affect the length or moment of inertia at the middiaphyseal site in either bone. Age effects on mechanical properties. The forces withstood by the femur and humerus were greater in the old than in the young rats at both the yield and ultimate 1. Physical characteristics and humerus of Fischer 344 rats
TABLE
Group
FFW,
YC
40928*
mg
FFW/BW,
%
of the femur Length,
mm
MI, mm4
Femur
UC
0.204~0.002 32.8O,tO.20*2.639~0.068" 422+11*t 0.207+0.002t 33.12kO.30" 2.800t0.100* 0.198~0.005 35.67kO.42 4.104kO.320 521k8
OT
574zk26t
0.222+0.012~
YG
188&3* 197*5* 248&4 257&B
0.095~0.002 26.00&0.15* 0.56&0.022* 0.097+0.001 26.00t_O.33*0.610&0.019* 0.094+0.003 27.83,tO.48 0.985&0.088 0*100zk0.004 28.~0~0.00 0.970~0.055
YT
36.OOkO.38
3.915~0.099
Humerus
YT
OC
OT Values are means t SE. YC, young control (n = 10); YT, young trained (n = 8); OC, old control (n = 6); OT, old trained (n. = 8). FFW/ BW, bone fat-free dry weight expressed as a percent of total body weight; MI, moment of inertia. * Significant age difference, young (YT + YC) vs. old (OT + OC), P < 0.05; t significant training difference, trained (YT + OT) vs. control (YC + OC), P < 0.05.
EXERCISE
AND AGING
points (Table 2). When force was adjusted for the moment of inertia, the femurs from old rats withstood significantly less stress at the ultimate point. Strain at the ultimate point was significantly greater in old than in young femurs. There were no age differences in femur stress or strain at the yield point. Stress and strain followed similar patterns in the humerus, but the trends were not significant. Training effects on mechanical properties. Training increased the force the humerus was able to resist at both the yield and ultimate points. Force was similarly increased in the femur from trained rats but was significant only at the ultimate point. Combined across age groups, there were no training-induced differences in the material properties of stress and strain in either bone at the ultimate or yield points. A significant interaction between age and training was observed for femur ultimate stress. The old femurs showed a significant increase in stress with training, whereas no effect was detected in young femurs. No interactions were seen in the humerus for stress; however, the humerus from young rats tended to show greater training effects than the humerus from old rats. A significant interaction between age and training was observed for strain in the femur at the yield point (Table 2). Strain was greater in femurs from OT than from OC rats, whereas strain was lower in femurs from YT than YC rats. Although the interaction was not significant at the ultimate point or for the humerus at either point, bones from YT rats tended to have greater strain properties than those from YC rats. No consistent pattern was observed for a training effect on strain in the old bones. The only interactions between age and training were observed for yield strain and ultimate stress in the femur. Despite differences in training intensity, differences in bone mass and force properties due to weight-bearing exercise in old bones were similar to those in young bones. DISCUSSION
Age and bone mechanical properties. As would be expected from bones with a greater mass and moment of inertia, bones from the old rats withstood greater force at both the yield and ultimate points than those from young rats. In previous comparisons of femurs from 6and Wmo-old rats, there were no consistent increases in force with age although wet weights and cortical areas increased with age; yield force decreased in males (5), and force only increased at the ultimate point in females (6). In the present study, when the bones were corrected for differences in size, stress at the yield point was not different between age groups. Age differences were evident at the ultimate point where the old femurs withstood less stress and demonstrated greater strain than the young femurs. Kiebzak et al. (5,6) found decreased stress and no change in strain with age at both the yield and ultimate points. From these findings it can be concluded that increases in bone weight and size with age result in an ability to withstand greater force. The increase in
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
BONE PROPERTIES TABLE
WITH
EXERCISE
133
AND AGING
2. Mechanical properties of the femur and humerus at the yield and ultimate points in Fischer 344 rats Yield Group
Force, kg/mm
Stress, kg/mm2
Ultimate Strain
Force, kg/mm
Stress, kg/mm’
Strain
41.7t,l.l" 42.9&1.8*~ 45.7tl.l 53.1*3.4?
21.82&0.72*$ 21.56+0.64*$ 17.37+0.76$ 20.46+1.07$
0.059&0.002* 0.066t,O.O03* 0.0721kO.006 0.072kO.004
19.0t0.5* 21.6+1.0*? 24.5t0.8 25.6cO.7t
36.8721.75 40.57k2.37 33.43k4.02 34.221k2.65
0.043~0~002 0.049~0.003 0.053~0.004 0.049zk0.004
Femur YC
YT OC OT
24.0+0.7* 26.2kl.V 30.0~2.9 32.2t3.4
12.55t0.46 13.22t0.56 11.60tl.47 12.4Otl.15
0.022a0.001$
25.02kl.40 29.23tl.83 24.12t3.73 24.85zk2.16
0.0243-0.001 0.027t0.002 0.023~0.002 0.025kO.001
0.026&0.001$ 0.027+0.002$ 0.025,+0.001$ Humerus
YC 12.9AO.5" YT 15.4+0.6*=t OC 17.6tl.3 OT 18.5+O.?t Values are means & SE. See Table 1 < 0.05; t significant training difference,
footnote for description of groups. * Significant age difference, young (YT + UC) vs. old (OT + OC), P trained (YT + OT) vs. control (YC + OC), P < 0.05; $ interaction between age and training, P < 0.05.
force is not proportional to the increase in size due to a decrease in material strength (stress). This suggests that, given equivalent mass, an old bone would fracture with less force than a young bone. Training intensity. Change in bone mass is related to forces that act directly on the bone and alter the rate, magnitude, and direction of strain applied to the bone (8). Although no in vivo strain measurements were collected in the present study, strain in weight bearing bones should increase as running speed increases (15). The specific intensity levels and types of activity that will effectively stimulate bone responses in vivo have not yet been defined. In the present experiment, all rats exercised for the same duration and at a similar grade but the young animals trained at higher speeds and, therefore, at a greater absolute work intensity than the old rats. From estimates of maximal oxygen uptake (%2 max) for the two age groups (9), the young rats trained at a greater percent of their %2max than the older rats. By zueek four, the young animals were training at an estimated 75% of their predicted untrained VoZ m8Xwhereas the old rats never reached this relative intensity. Although the old rats trained at both a lower absolute and relative training intensity, the training-induced differences in bone mass and force were similar across age groups. Training level, monitored and discussed as a percent of aerobic capacity, although convenient, may not be relevant for the estimation of bone stimulation. In humans >60 yr of age, correlations between measured VOgrnax and spine bone mineral density (BMD) are weak in men (r = 0.41) and not significant in women (r = 0.04) (2). Tj02 maxis a systemic measure of an animal’s oxidative capacity, and although it may provide a good indication of the general level of activity and the relative amount of force regularly applied to bones, it is not predictive of local bone strain. Age differences in body weight, bone size, and normal activity patterns further confound estimation of training stimulation to the bones. From this study it appears that lower training speeds and relative intensities are required in old than in young rats to induce similar increases in bone mass and force. The observation that our training protocol did not influence body weight, bone length, or moment of inertia
might be anticipated in old rats, because the training was initiated after maturity. However, the lack of training response for these same parameters in the young rats, which increased body weight by 30% during the training period, was not expected. Matsuda et al.( 10) exercised 3wk-old rapidly growing roosters at 7O-80% Vozrnax and observed a decrease in bone length, width, and mechanical properties in trained compared with control roosters. Others have observed increases in moment of inertia, cross-sectional area, and bone volume after training in immature rats (1, 16) and pigs (19). The different responses to training may relate to the species, age of the animals, or training intensities. Stress and training. Training increased bone mass and force necessary for fracture. The increase in force for femurs from YT rats was proportional to the increase in size or moment of inertia. Stress, the ratio of force to moment of inertia, was unaffected by training in the young rats. This is consistent with previous findings (19). The femurs from OT rats, however, demonstrated significantly greater ultimate stress than those from the OC rats. Ultimate stress in OC femurs was significantly lower than that of young (training and control) femurs, whereas ultimate stress in OT femurs was similar to that of the young (training and control) femurs. In rodents, researchers have observed a decrease in bone density from 9 to 22 mo of age (3, 14). Training prevented the age-associated bone loss in mice (14) and increased bone density in old rats (3). In the present study, the femurs from old-trained rats had greater fat-free dry weight than control femurs with no increase in moment of inertia. This implies an increase in density with training, which would be consistent with the change in material strength (i.e., ultimate stress). In summary, the femur and humerus from both young and senescent trained rats had a greater resistance to fracture than bones from weight matched sedentary controls. The bones from old rats had significantly lower material strength than those from young rats. Training either maintained or increased ultimate stress in the old
femur at a level comparable with that of young femurs. The senescent rats displayed similar bone mass and force adaptations to training to the young adult rats, despite working at a lower absolute work intensity. Further study
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
134
BONE PROPERTIES
is necessary to investigate appropriate of exercise that beneficially influence and mechanical properties.
WITH
levels and modes bone metabolism
The authors thank L. Gosselin, C. Julin, K. McCormick, and D. Wanta for assistance with the care and training of the rats and M. Checovich for laboratory assistance. This work was supported by a grant to D. P. Thomas from the National Institute on Aging, National Institute of Biogerontology and by the University of Wisconsin Graduate School. Present address of D. M. Raab: Center for Hard Tissue Research, 601 N. 30th St., Suite 5740, Creighton Universtity, Omaha, NE 68131. Address for reprint requests: E. L. Smith, Biogerontology Laboratory, 504 North Walnut St., University of Wisconsin, Madison, WI 53705.
Received 10 April 1989; accepted in final form 28 August 1989.
EXERCISE
mice-physical properties of achilles tendons and long bones. Growth 8.
41: 123-137, 1977. LANYON, L. E. Functional
strain as a determinant for bone remodeling. Calcij. Tissue Int. 36: S56-S61, 1984. 9. MAZZEO, R. S., G. A. BROOKS, AND S. M. HORVATH. Effects of age on metabolic responses to endurance training in rats. J. Appl. Physiol. 57: 1369-1374, 1984. 10. MATSUDA, J. J., R. F. ZERNICKE, A. PEDRINI-MILLE, AND J. A. MAYNARD.
C. VAILAS, V. A. PEDRINI, A. Structural and mechanical adaptations of immature bone to strenuous exercise. J. Appl. Physiol. 60: 2028-2034, 1986. 11. MCDONALD, R., J. HEGENAUER, AND P. SALTMAN. Age-related differences in the bone mineralization pattern of rats following exercise. J. Gerontol. 41: 445-452, 1986. 12. MELTON, L. J., AND B. L. RIGGS. Epidemiology of age-related fractures. In: The Osteoporotic Syndrome: Detection, Preuention, and Treatment, edited by L. V. Avioli. New York: Grune & Stratton, 1983, p. 45-72. 13. NATIONAL
REFERENCES K. D., AND P. GRIMINGER. Long-term effects of activity and of calcium and phosphorus intake on bones and kidneys of female rats. J. Nutr. 113: 2111-2121, 1983. 2. BEVIER, W. C., R. A. WISWELL, G. PYKA, K. C. KIZAK, K. M. NEWHALL, AND R. MARCUS. Relationship of body composition , muscle strength, and aerobic capacity to bone mineral density in older men and women. J. Bone Min. Res. In press. 3. BEYER, R. E,, J. C. HUANG, ANT) G. B. WILSHIRE. The effects of endurance exercise on bone dimensions, collagen, and calcium in the aged male rat. Exp. GerontoL. 20: 315-323, 1985. 4. CRENSHAW, T. D., E. R. PEO, A. J, LEWIS, B. D. MOSER, AND D. OLSON. Influence of age, sex, and calcium and phosphorus levels on the mechanical properties of various bones in swine. J. Anim. 1. BAUER,
Sci. 52: 1319-1329,19&l. 5. KIEBZAK, G. M., R. SMITH, SACKT~R. Bone status of
RESEARCH
COUNCIL.
Nutrient
Requirements
of the Lab-
oratory Rut. Washington, DC: Natl. Acad. Sci, 1978, p. 7-37. 14. RINGE, J. D., AND E. STEINHAGEN-THIESSEN. Prevention of physiological age-dependent bone atrophy by controlled exercise in mice. Age Omaha 2: 44-47,19&5. 15. RUBIN, C. T., AND L. E. LANYON. Limb mechanics as a function of speed and gait: a study of functional strains in the radius and tibia of horse and dog. J. Exp. Biol. 101: 187-211, 1982. 16. SAVILLE, P. D,, AND M. P. WHYTE. Muscle and bone hypertrophy positive effect of running exercise in the rat. Clin. Orthop. Relat. Res. 65: 81-88, 1969. 17. SELDIN, E. D. A rheologic
model for cortical bone. A study of the physical properties of human femoral samples. Acta Orthop. Stand.
Suppl. 83: 20-43,1965. 18. STEINBERG, M. E., AND
J. TRUETA, Effects of activity on bone growth and development in the rat. Clin. Orthop. Relat. Res. 156: 52-60,19&l.
J. C. HOWE, AND B. senescent male rats: chemical, morphometric, and mechanical analysis. J. Bone Min. Res. 3: 37-45, 1988. 6. KIEBZAK, G. M., R. SMITH, J. C. HOWE, C. M. GUNDBERG, AND B. SACKTOR. Bone status of senescent female rats: chemical, morphometric, and biomechanical analysis. J. Bone Min. Res. 3: 439446,19&&. 7. KIISKINEN,
AND AGING
C. C. GUNDBERG,
A. Physical training
and connective tissue in young
Woo, S. L.-Y., S. C. KUEI, D. AMIEL, M. A. GOMEZ, W. C. HAYES, F. C. WHITE, AND W. H. AKESON. The effect of prolonged physical training on the properties of long bone: a study of Wolff’s law (Abstract). J. Bone Jt. Surg. Am. Vol. 63-A: 780-787, 1981. 20. Yu, B. P., E. 9. MASORO, I. MURATA, H. A. BERTRAND, AND F. T. LYND. Life span study of SPF Fischer 344 male rats fed ad libitum or restricted diets: longevity, growth, lean body mass and disease.
19.
J. Gerontol.
37: 130 -141,
1982.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
RATS, KNOWN for growth
through adulthood, display signs of senescence with advanced age. The body weight of male Fischer 344 rats peaks at -75% of their maximal life span and decreases after -90% of their life span (20). Compared with young adult rats, the bones of old rats have equal or greater cortical area, width, and length (3, 5, 6, 11). Although they maintain normal Ca, P, and collagen concentrations (mg/g), the bones from old rats have a lower density (g/cm3) than those from young rats, indicating greater porosity with senescence (3). These changes in density may affect bone mechanical properties. Bone force increases with an increase in bone size, but stress (force per unit area of bone) has been found to decrease with age despite an increase in bone mass (5, 6) *Exercise training may stimulate bone formation and increase bone mass and strength (7, 16, 18, 19). Studies that have evaluated the effects of exercise on bone mechanical properties in rats have primarily used immature animals (7, 10, 16, 19). Such models are inappropriate 130
for inferences to adult skeletons, because there may be an interaction of growth and training. The senescent rat model may provide insights to help understand exercise effects in humans. Osteoporosis, diagnosed by fracture due to low bone mass and strength, is a common health risk for older women. By 90 yr of age, an estimated one-third of all women will experience an osteoporotic hip fracture (12). If exercise training can increase bone mass and the ability of senescent bone to withstand force, it may decrease the risk for fracture in older adults. Depending on animal species, age, length, and intensity of exercise training, various experiments have shown greater, lesser, or no differences in bone weight, density, size, and strength (1, 7,10,16, 18, 19). Training initiated with mature, 19- to 26-mo-old rats has resulted in increased bone weight, Ca, and hydroxyproline content with no increase in bone diameters or strength (3,11). It is unclear whether the lack of change in diameters or strength resulted from differences in methodology between studies, an interaction with growth, or a decrease in response to mechanical stimulation with age. Previous research has concentrated primarily on growing animals and has failed to consider the mechanical effects of training imposed on the mature skeleton. The current study was designed to examine the role of age on training-induced adaptation of weight bearing bone, specifically the humerus and femur. Bones from young adult and senescent rats, trained and untrained, were compared for differences in size and mechanical properties. METHODS
Fischer 344 virgin female rats, ages 2.5 (young, Y) and 25 mo (old, 0), were obtained from the National Institute on Aging (specific pathogen free) colony at Indianapolis, IN. Throughout the experiment the rats were maintained in a 20-22°C room with a 12-h light/dark cycle. The animals were individually housed in wire cages and allowed ad libitum access to food (Wayne Rodent Blox, Continental Grain) and water. The feed contained 1.46% Ca and 0.99% P. These levels exceed those recommended for growth by the National Research Council (13). Two weeks were allowed for acclimation before the initiation of the study. All rats walked on a rodent treadmill before assignment to treatment groups. Rats from each age group were randomly assigned to either a sedentary
0161-7567/90 $1.50Copyright0 1990the AmericanPhysiologicalSociety
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
BONE PROPERTIES
WITH
control (C) or training (T) group. The initial group consisted of 10 YC, 10 YT, 8 OC, and 9 OT. Training protocol. All rats from the same age group trained simultaneously on a lo-chamber rodent treadmill (Quinton model 42-15). An electric grid at the rear of the belt was used as the running stimulus. All rats trained 5 days/wk for 10 wk with speed and grade, and duration progressively increased. The old rats began training at 5 m/min and 0% grade for 10 min/day. Speed and duration were increased until, by the 4th wk, the OT rats ran at 15 m/min for 60 min/day. Thereafter the speed and duration were maintained, but grade was gradually increased to 15%. Young adult rats have a greater maximal oxygen uptake than old rats (9); therefore, the YT rats were exercised at a greater absolute intensity than the OT group. The young adult rats began training at 10 m/ min and 5% grade for 15 min/day. By the end of the 2nd wk, the YT rats ran at 15 m/min, 15% grade for 60 min/ day. Thereafter, the grade and duration were maintained but speed was increased 2-3 m/min each week. By the 10th wk, the YT rats were running at 36 m/min and 15% grade for 60 min/day. Throughout the experiment, control rats were placed on the stationary treadmill for 5 min/day, 5 days/wk. All rats were weighed twice each week. Weights were used as an indication of growth and health status. Data collection. After the lo-wk training program, the rats were anesthetized with chloral hydrate and killed by a 1 ml cardiac injection of saturated KCl. The right femur and humerus were removed, cleaned of all soft tissue, individually sealed in plastic bags, and frozen at -20°C. While in the sealed bags, the bones were thawed to room temperature on the day of mechanical testing. Freeze-thaw procedures do not alter the mechanical properties of bone (17). Immediately before the mechanical test, the bone length was measured and the midpoint marked. The mechanical properties of the femur and humerus were determined using a three-point bending test with an Instron Universal Testing Machine. In the threepoint bending test, the bone was rested on two support plates 11.48 mm apart. Force was applied at the midpoint of the bone by a plunger moving downward at a constant rate of 5.08 mm/min. The bone was simultaneously tested in compression on the concave surface and tension on the convex surface. The force withstood was dependent on the shape of the bone cross section, the amount and distribution of bone mass, and the distance between the support plates. Force was applied to the femur midshaft from the anterior to the posterior surface and to the humerus midshaft from the lateral to the medial surface. Orientation was established for maximum stability and to prevent rotation during the test. The amount of load withstood by the bone and the bending or deformation that occurred during each test was recorded on an X-Y plotter as a load-deformation curve (Fig. 1). The moment of inertia, a measurement of the area distribution about the neutral axis of the bone cross section, is dependent on the bone’s cross-sectional area and shape. The midshaft cross section of each bone was
EXERCISE
131
AND AGING
z 0 I
Deformat ion 1. Typical load-deformation curve generated during a 3-point bending test. y, Yield or bending point; U, ultimate or breaking point; horizontal axis, vertical deformation of the bone as plunger moved downward at a constant rate of 5.08 mm/min; vertical axis, load on bone or force with which bone resisted deformation. White area, elastic region where no permanent damage occurs to bone; hatched area, plastic region where bone is permanently damaged. FIG.
cut 1 mm distal to the point of fracture. The cross section was photographed and enlarged N-fold. The AutoCAD program (Autodesk, Millvalley, CA) interfaced with a digitizing pad and was used to determine the centroid and moment of inertia about the axis perpendicular to the line of force application (femur: medial-lateral axis; humerus: anterior-posterior axis). Fat-free dry weight was determined after the bones were extracted in 100% ether for 5 days, allowed to dry at room temperature for 20 h, and subsequently dried at 80°C for 24 h. Computation of mechanical properties Mechani cal properties of th .e fe mur and humerus were calculated bY use of data from the load-deformation curves (Fig. 1) at both the yield and ultimate points (4). The yield point is the transition from the linear elastic phase (deformation is proportional to force) to the curvilinear or plastic phase (deformation increases more rapidly than force). The yield point occurs when the applied load or force creates permanent damage within the bone. If the load on the bone is removed before the yield point, the bone wil l return to its original shape and retain its original properties. The ultimate point is the point where load reaches a maximum and the bone fractures. Bending moment or the force (F) the bone withstood is calculated from the load required to break the bone distance between the two lower and the length support plates F = load x L/4 (kg* mm) Stress is the force withstood per unit area of bone, Stress is calculated from the force, the moment of inertia (I), and C, which is half the bone diameter measured parallel to the line of force application stress = F x C/I (kg/mm’) Strain is a measure of the bending (deformation) of the bone relative to its length. The terms C and L are defined above for stress and force strain = 12 x deformation x C/L The results were analyzed by a two-way analysis of
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
132
BONE PROPERTIES
WITH
variance procedure: the main effects were age and training. Results from each bone were analyzed separately. Differences were considered significant at P < 0.05. RESULTS
The results reported are based only on the rats that completed the IO-wk experiment (YC = 10, YT = 8, OC = 6, and OT = 8). During the study, two OC and one OT rats died and two YT rats were eliminated due to foot injuries that prevented further training. Age effects on growth. Young rats in both groups gained weight (P < 0.05) during the lo-wk study. Average weights at 2.5 mo were (mean t SE) 153.9 t 11.8 g (YC) and 153,6 t 12.2 g (YT) and at 5 mo were 199.0 t 4.5 g (YC) and 203.6 t 4.2 g (YT). In contrast, both senescent groups gradually lost body weight from 278.8 -I- 23.4 g (OC) and 281.8 t 22.5 g (OT) at 25 mo to 264.2 k 7.4 g (OC) and 259.5 t 8.0 g (OT) at 27.5 mo, Body weight was not affected by training in either age group. The old rats were larger and heavier than the young adult rats. Both the femur and humerus from old rats were longer and had greater fat-free dry weights and greater moments of inertia than those from young rats (Table 1). Although skeletal size increased with age, the fat-free dry weight of the bones remained a constant percentage of total body weight. The femurs averaged 0.21 and the humerus 0.10% of total body weight. Training effects on bone grcwth. Irrespective of age, the femurs from trained rats had a greater (P c 0.05) fat-free dry weight than those from control rats (Table 1). A similar trend was observed for the humerus, but the training-induced increase in weight was not significant. The relative weight of the femur to total body weight was greater (P C 0.05) in trained than in control rats within age groups, but there were no differences between age groups. Training did not affect the length or moment of inertia at the middiaphyseal site in either bone. Age effects on mechanical properties. The forces withstood by the femur and humerus were greater in the old than in the young rats at both the yield and ultimate 1. Physical characteristics and humerus of Fischer 344 rats
TABLE
Group
FFW,
YC
40928*
mg
FFW/BW,
%
of the femur Length,
mm
MI, mm4
Femur
UC
0.204~0.002 32.8O,tO.20*2.639~0.068" 422+11*t 0.207+0.002t 33.12kO.30" 2.800t0.100* 0.198~0.005 35.67kO.42 4.104kO.320 521k8
OT
574zk26t
0.222+0.012~
YG
188&3* 197*5* 248&4 257&B
0.095~0.002 26.00&0.15* 0.56&0.022* 0.097+0.001 26.00t_O.33*0.610&0.019* 0.094+0.003 27.83,tO.48 0.985&0.088 0*100zk0.004 28.~0~0.00 0.970~0.055
YT
36.OOkO.38
3.915~0.099
Humerus
YT
OC
OT Values are means t SE. YC, young control (n = 10); YT, young trained (n = 8); OC, old control (n = 6); OT, old trained (n. = 8). FFW/ BW, bone fat-free dry weight expressed as a percent of total body weight; MI, moment of inertia. * Significant age difference, young (YT + YC) vs. old (OT + OC), P < 0.05; t significant training difference, trained (YT + OT) vs. control (YC + OC), P < 0.05.
EXERCISE
AND AGING
points (Table 2). When force was adjusted for the moment of inertia, the femurs from old rats withstood significantly less stress at the ultimate point. Strain at the ultimate point was significantly greater in old than in young femurs. There were no age differences in femur stress or strain at the yield point. Stress and strain followed similar patterns in the humerus, but the trends were not significant. Training effects on mechanical properties. Training increased the force the humerus was able to resist at both the yield and ultimate points. Force was similarly increased in the femur from trained rats but was significant only at the ultimate point. Combined across age groups, there were no training-induced differences in the material properties of stress and strain in either bone at the ultimate or yield points. A significant interaction between age and training was observed for femur ultimate stress. The old femurs showed a significant increase in stress with training, whereas no effect was detected in young femurs. No interactions were seen in the humerus for stress; however, the humerus from young rats tended to show greater training effects than the humerus from old rats. A significant interaction between age and training was observed for strain in the femur at the yield point (Table 2). Strain was greater in femurs from OT than from OC rats, whereas strain was lower in femurs from YT than YC rats. Although the interaction was not significant at the ultimate point or for the humerus at either point, bones from YT rats tended to have greater strain properties than those from YC rats. No consistent pattern was observed for a training effect on strain in the old bones. The only interactions between age and training were observed for yield strain and ultimate stress in the femur. Despite differences in training intensity, differences in bone mass and force properties due to weight-bearing exercise in old bones were similar to those in young bones. DISCUSSION
Age and bone mechanical properties. As would be expected from bones with a greater mass and moment of inertia, bones from the old rats withstood greater force at both the yield and ultimate points than those from young rats. In previous comparisons of femurs from 6and Wmo-old rats, there were no consistent increases in force with age although wet weights and cortical areas increased with age; yield force decreased in males (5), and force only increased at the ultimate point in females (6). In the present study, when the bones were corrected for differences in size, stress at the yield point was not different between age groups. Age differences were evident at the ultimate point where the old femurs withstood less stress and demonstrated greater strain than the young femurs. Kiebzak et al. (5,6) found decreased stress and no change in strain with age at both the yield and ultimate points. From these findings it can be concluded that increases in bone weight and size with age result in an ability to withstand greater force. The increase in
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
BONE PROPERTIES TABLE
WITH
EXERCISE
133
AND AGING
2. Mechanical properties of the femur and humerus at the yield and ultimate points in Fischer 344 rats Yield Group
Force, kg/mm
Stress, kg/mm2
Ultimate Strain
Force, kg/mm
Stress, kg/mm’
Strain
41.7t,l.l" 42.9&1.8*~ 45.7tl.l 53.1*3.4?
21.82&0.72*$ 21.56+0.64*$ 17.37+0.76$ 20.46+1.07$
0.059&0.002* 0.066t,O.O03* 0.0721kO.006 0.072kO.004
19.0t0.5* 21.6+1.0*? 24.5t0.8 25.6cO.7t
36.8721.75 40.57k2.37 33.43k4.02 34.221k2.65
0.043~0~002 0.049~0.003 0.053~0.004 0.049zk0.004
Femur YC
YT OC OT
24.0+0.7* 26.2kl.V 30.0~2.9 32.2t3.4
12.55t0.46 13.22t0.56 11.60tl.47 12.4Otl.15
0.022a0.001$
25.02kl.40 29.23tl.83 24.12t3.73 24.85zk2.16
0.0243-0.001 0.027t0.002 0.023~0.002 0.025kO.001
0.026&0.001$ 0.027+0.002$ 0.025,+0.001$ Humerus
YC 12.9AO.5" YT 15.4+0.6*=t OC 17.6tl.3 OT 18.5+O.?t Values are means & SE. See Table 1 < 0.05; t significant training difference,
footnote for description of groups. * Significant age difference, young (YT + UC) vs. old (OT + OC), P trained (YT + OT) vs. control (YC + OC), P < 0.05; $ interaction between age and training, P < 0.05.
force is not proportional to the increase in size due to a decrease in material strength (stress). This suggests that, given equivalent mass, an old bone would fracture with less force than a young bone. Training intensity. Change in bone mass is related to forces that act directly on the bone and alter the rate, magnitude, and direction of strain applied to the bone (8). Although no in vivo strain measurements were collected in the present study, strain in weight bearing bones should increase as running speed increases (15). The specific intensity levels and types of activity that will effectively stimulate bone responses in vivo have not yet been defined. In the present experiment, all rats exercised for the same duration and at a similar grade but the young animals trained at higher speeds and, therefore, at a greater absolute work intensity than the old rats. From estimates of maximal oxygen uptake (%2 max) for the two age groups (9), the young rats trained at a greater percent of their %2max than the older rats. By zueek four, the young animals were training at an estimated 75% of their predicted untrained VoZ m8Xwhereas the old rats never reached this relative intensity. Although the old rats trained at both a lower absolute and relative training intensity, the training-induced differences in bone mass and force were similar across age groups. Training level, monitored and discussed as a percent of aerobic capacity, although convenient, may not be relevant for the estimation of bone stimulation. In humans >60 yr of age, correlations between measured VOgrnax and spine bone mineral density (BMD) are weak in men (r = 0.41) and not significant in women (r = 0.04) (2). Tj02 maxis a systemic measure of an animal’s oxidative capacity, and although it may provide a good indication of the general level of activity and the relative amount of force regularly applied to bones, it is not predictive of local bone strain. Age differences in body weight, bone size, and normal activity patterns further confound estimation of training stimulation to the bones. From this study it appears that lower training speeds and relative intensities are required in old than in young rats to induce similar increases in bone mass and force. The observation that our training protocol did not influence body weight, bone length, or moment of inertia
might be anticipated in old rats, because the training was initiated after maturity. However, the lack of training response for these same parameters in the young rats, which increased body weight by 30% during the training period, was not expected. Matsuda et al.( 10) exercised 3wk-old rapidly growing roosters at 7O-80% Vozrnax and observed a decrease in bone length, width, and mechanical properties in trained compared with control roosters. Others have observed increases in moment of inertia, cross-sectional area, and bone volume after training in immature rats (1, 16) and pigs (19). The different responses to training may relate to the species, age of the animals, or training intensities. Stress and training. Training increased bone mass and force necessary for fracture. The increase in force for femurs from YT rats was proportional to the increase in size or moment of inertia. Stress, the ratio of force to moment of inertia, was unaffected by training in the young rats. This is consistent with previous findings (19). The femurs from OT rats, however, demonstrated significantly greater ultimate stress than those from the OC rats. Ultimate stress in OC femurs was significantly lower than that of young (training and control) femurs, whereas ultimate stress in OT femurs was similar to that of the young (training and control) femurs. In rodents, researchers have observed a decrease in bone density from 9 to 22 mo of age (3, 14). Training prevented the age-associated bone loss in mice (14) and increased bone density in old rats (3). In the present study, the femurs from old-trained rats had greater fat-free dry weight than control femurs with no increase in moment of inertia. This implies an increase in density with training, which would be consistent with the change in material strength (i.e., ultimate stress). In summary, the femur and humerus from both young and senescent trained rats had a greater resistance to fracture than bones from weight matched sedentary controls. The bones from old rats had significantly lower material strength than those from young rats. Training either maintained or increased ultimate stress in the old
femur at a level comparable with that of young femurs. The senescent rats displayed similar bone mass and force adaptations to training to the young adult rats, despite working at a lower absolute work intensity. Further study
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
134
BONE PROPERTIES
is necessary to investigate appropriate of exercise that beneficially influence and mechanical properties.
WITH
levels and modes bone metabolism
The authors thank L. Gosselin, C. Julin, K. McCormick, and D. Wanta for assistance with the care and training of the rats and M. Checovich for laboratory assistance. This work was supported by a grant to D. P. Thomas from the National Institute on Aging, National Institute of Biogerontology and by the University of Wisconsin Graduate School. Present address of D. M. Raab: Center for Hard Tissue Research, 601 N. 30th St., Suite 5740, Creighton Universtity, Omaha, NE 68131. Address for reprint requests: E. L. Smith, Biogerontology Laboratory, 504 North Walnut St., University of Wisconsin, Madison, WI 53705.
Received 10 April 1989; accepted in final form 28 August 1989.
EXERCISE
mice-physical properties of achilles tendons and long bones. Growth 8.
41: 123-137, 1977. LANYON, L. E. Functional
strain as a determinant for bone remodeling. Calcij. Tissue Int. 36: S56-S61, 1984. 9. MAZZEO, R. S., G. A. BROOKS, AND S. M. HORVATH. Effects of age on metabolic responses to endurance training in rats. J. Appl. Physiol. 57: 1369-1374, 1984. 10. MATSUDA, J. J., R. F. ZERNICKE, A. PEDRINI-MILLE, AND J. A. MAYNARD.
C. VAILAS, V. A. PEDRINI, A. Structural and mechanical adaptations of immature bone to strenuous exercise. J. Appl. Physiol. 60: 2028-2034, 1986. 11. MCDONALD, R., J. HEGENAUER, AND P. SALTMAN. Age-related differences in the bone mineralization pattern of rats following exercise. J. Gerontol. 41: 445-452, 1986. 12. MELTON, L. J., AND B. L. RIGGS. Epidemiology of age-related fractures. In: The Osteoporotic Syndrome: Detection, Preuention, and Treatment, edited by L. V. Avioli. New York: Grune & Stratton, 1983, p. 45-72. 13. NATIONAL
REFERENCES K. D., AND P. GRIMINGER. Long-term effects of activity and of calcium and phosphorus intake on bones and kidneys of female rats. J. Nutr. 113: 2111-2121, 1983. 2. BEVIER, W. C., R. A. WISWELL, G. PYKA, K. C. KIZAK, K. M. NEWHALL, AND R. MARCUS. Relationship of body composition , muscle strength, and aerobic capacity to bone mineral density in older men and women. J. Bone Min. Res. In press. 3. BEYER, R. E,, J. C. HUANG, ANT) G. B. WILSHIRE. The effects of endurance exercise on bone dimensions, collagen, and calcium in the aged male rat. Exp. GerontoL. 20: 315-323, 1985. 4. CRENSHAW, T. D., E. R. PEO, A. J, LEWIS, B. D. MOSER, AND D. OLSON. Influence of age, sex, and calcium and phosphorus levels on the mechanical properties of various bones in swine. J. Anim. 1. BAUER,
Sci. 52: 1319-1329,19&l. 5. KIEBZAK, G. M., R. SMITH, SACKT~R. Bone status of
RESEARCH
COUNCIL.
Nutrient
Requirements
of the Lab-
oratory Rut. Washington, DC: Natl. Acad. Sci, 1978, p. 7-37. 14. RINGE, J. D., AND E. STEINHAGEN-THIESSEN. Prevention of physiological age-dependent bone atrophy by controlled exercise in mice. Age Omaha 2: 44-47,19&5. 15. RUBIN, C. T., AND L. E. LANYON. Limb mechanics as a function of speed and gait: a study of functional strains in the radius and tibia of horse and dog. J. Exp. Biol. 101: 187-211, 1982. 16. SAVILLE, P. D,, AND M. P. WHYTE. Muscle and bone hypertrophy positive effect of running exercise in the rat. Clin. Orthop. Relat. Res. 65: 81-88, 1969. 17. SELDIN, E. D. A rheologic
model for cortical bone. A study of the physical properties of human femoral samples. Acta Orthop. Stand.
Suppl. 83: 20-43,1965. 18. STEINBERG, M. E., AND
J. TRUETA, Effects of activity on bone growth and development in the rat. Clin. Orthop. Relat. Res. 156: 52-60,19&l.
J. C. HOWE, AND B. senescent male rats: chemical, morphometric, and mechanical analysis. J. Bone Min. Res. 3: 37-45, 1988. 6. KIEBZAK, G. M., R. SMITH, J. C. HOWE, C. M. GUNDBERG, AND B. SACKTOR. Bone status of senescent female rats: chemical, morphometric, and biomechanical analysis. J. Bone Min. Res. 3: 439446,19&&. 7. KIISKINEN,
AND AGING
C. C. GUNDBERG,
A. Physical training
and connective tissue in young
Woo, S. L.-Y., S. C. KUEI, D. AMIEL, M. A. GOMEZ, W. C. HAYES, F. C. WHITE, AND W. H. AKESON. The effect of prolonged physical training on the properties of long bone: a study of Wolff’s law (Abstract). J. Bone Jt. Surg. Am. Vol. 63-A: 780-787, 1981. 20. Yu, B. P., E. 9. MASORO, I. MURATA, H. A. BERTRAND, AND F. T. LYND. Life span study of SPF Fischer 344 male rats fed ad libitum or restricted diets: longevity, growth, lean body mass and disease.
19.
J. Gerontol.
37: 130 -141,
1982.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 11, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
