**4.1 Effects on structural (whole bone) properties**

Structural properties are calculated from original biomechanical testing raw data without any normalization. Related parameters are load (Nt), displacement (mm), energy (mini joule, mJ) and stiffness (Nt/mm) etc. As shown in previous studies, results of rodents' whole tissue biomechanical properties after a period of endurance running training were controversial. Some studies show that exercise groups were higher in load-withstanding capacity,51-53 while others revealed a higher energy absorption capacity.55 One possible explanation for these discrepancies may be differences in training protocol. Animals trained at a higher intensity tended to show higher bone strength (e.g. higher bending load or moment),51, 53 suggesting that higher mechanical loading generated by intensively running may benefit to bone strength. Moreover, the specific testing conditions would also affect testing results; for instance, Nordsletten and colleagues measured bone strength *in vivo*, at which time bone strength may be affected by muscle strength gains achieved through training.53

hormone in growing male rats may account for the significantly lower protein mass gain and BW gain. In female rats, however, estrogens would suppress body mass, food consumption and fat deposition.70, 71 Progesterone, on the other hand, has been verified to increase body fat and body mass.72, 73 Moreover, a previous study suggests that regular treadmill training results in extended periods of progesterone secretion, which was associated with significant weight gain.74 Women may respond to ERT with similar regulation in progesterone. However, the up-regulation of progesterone may be more pronounced in rats than in women, since female rats reveal no decrease in BW even under

An analysis of gender differences observed in the above animal studies reveals that ERT increases BMD and BMC in female rats but not in male rats. However, the physiological response (*e.g.* BW gain) of female rats to ERT differs from that in female humans. Given that animal studies are intended to clarify the mechanisms of biological phenomena in humans, female rats may not be a suitable model for investigating the effects of ERT on developing or young adult bone. On the other hand, the changes in BW and densitometric parameters associated with ERT in male rats were similar to those in humans, indicating that male rats

Aside from BMD and BMC, biomaterial related analysis will provide more valuable information to predict the capacity that bone tissue can withstand extra mechanical loading generated by daily physical activity or accidents (e.g. fall), and thus, prevent bone from

Generally, biomaterial properties of bone tissue can be analyzed at a structural level and a tissue level. Structure biomaterial properties are size-dependent, that is, tissues bigger in size tend to be stronger than smaller ones. Conversely, tissue-level properties are analyzed under size-independent conditions using mathematic methods (e.g. normalized tissue size by cross-sectional moment of inertia) or mechanical methods (e.g. a specimen with

Structural properties are calculated from original biomechanical testing raw data without any normalization. Related parameters are load (Nt), displacement (mm), energy (mini joule, mJ) and stiffness (Nt/mm) etc. As shown in previous studies, results of rodents' whole tissue biomechanical properties after a period of endurance running training were controversial. Some studies show that exercise groups were higher in load-withstanding capacity,51-53 while others revealed a higher energy absorption capacity.55 One possible explanation for these discrepancies may be differences in training protocol. Animals trained at a higher intensity tended to show higher bone strength (e.g. higher bending load or moment),51, 53 suggesting that higher mechanical loading generated by intensively running may benefit to bone strength. Moreover, the specific testing conditions would also affect testing results; for instance, Nordsletten and colleagues measured bone strength *in vivo*, at which time bone strength may be affected by muscle strength gains achieved through

are a suitable model for investigating the effects of endurance running.

**4. Effects of endurance training on bone biomaterial properties** 

vigorous ERT.60

loading-induced damage.

training.53

**3.2 Studies of male rats mimic human practice** 

consistent size is sectioned from a whole tissue).75

**4.1 Effects on structural (whole bone) properties** 

However, in considering the applicability of exercise, a training program with moderate exercise intensity would be expected to show a higher compliance and therefore be more appropriate for the general population. As shown in Table 2, animals trained at a relatively moderate intensity (20-24m/min), which corresponds to 70% V‧O2 max,76 also had lower body masses and slightly lower (~5% lower) total BMD (p = 0.04), but were not found to have enhanced structural bone strength.55 The authors' previous study used body mass as a covariate to equalize raw data, which then revealed a comprehensively stronger bone tissue either in structural or tissue-level biomaterial properties.52 With lower body mass, the data of the exercise group would be adjusted to a higher level, and the effects of ERT seemed to become "good" for animal bones. However, it would be a more relevant and natural study if animals were fed ad libitum and data were not adjusted

In aspects of biological efficiency, an athlete at her/his optimal physiological status will not necessarily be absolutely higher in every physiological parameter. Therefore, a smaller muscle mass or skeleton size seems to be a benefit, rather than a weakness, for a distance runner or an endurance athlete. With such a smaller bone size, moderate ERT rats did not show absolutely enhanced structural bending load values but, interestingly, they showed better energy absorption capacity in long bone tissue that ERT rats were found to have a four-fold increase in energy absorption after long bone tissue reached the yield-point (postyield energy).55 As mentioned in previous studies,41 post-yield behaviors are highly correlated to tissue-level changes (e.g. collagen fiber orientation).
