It is well known that trabecular bone strength is determined not only by the amount of composite material (mineral, protein and water) but also the distribution of these materials (size, area, structural properties). A number of advantages, such as more abundant, thicker, well-connected, and plate-like trabeculae, confer a stronger trabecular bone compartment [27–30]. In the present study we demonstrated that diet-induced obese mice had a larger and stronger femoral metaphysis with more abundant and thicker trabecular bone. Voluntary wheel running decreased all of the measured cortical parameters, but increased trabecular bone mineral density and improved the 3D micro-structure.
Numerous data have shown that obesity is closely associated with dietary fat intake and sedentary life style [31, 32]. A close link between body mass and bone mass [17, 33] and increased risk for osteoporotic fracture due to low body and thus bone mass  has been reported. Our study also found a positive relationship between body mass and bone mass. Increased body mass requires stronger bone; this can be effectively realized through distributing bone mass further from the center of mass rather than dramatically increasing bone density. In our study, although total bone mineral density was decreased to some extent, the enlarged marrow cavity and increased total bone cross-sectional area resulted in a larger and stronger bone as indicated by the increased density-weighted moment of inertia (bending strength). This suggests that endosteal resorption and periosteal formation were enhanced in the obese mice. However, the effects of body mass on the skeleton remain controversial although well documented in obese subjects in previous studies [35, 36]. Some studies have shown that obese subjects have weaker bone to bear their over-weight body mass compared to normal counterparts in both humans and animals [37, 38]. In our study, after adjusting for body mass (unpublished data), no significant differences in bone traits between the obese and normal weight mice were found. This suggests increased bone strength through enlarged cross-sectional area thus distributing bone mass further from the centre of mass to adapt to the increase in body mass.
In order to further elucidate the effect of obesity and physical activity on bone, we separately estimated the trabecular and cortical bone compartments. It is well known that trabecular bone is the primary target for anabolic or catabolic factors and the most active bone site. The obese mice had larger trabecular area and higher trabecular bone mass than the control mice. In a recent report , the authors found increased SMI, decreased Conn.D, and similar Tb.Th in the proximal tibia in diet-induced obese mice. Similarly, we found increased SMI, but increased Conn.D and thicker trabeculae in the distal femur. These discrepancies could be explained by the differences in skeleton sites and our more accurate method of segmentation and higher resolution (2.8 μm), which could preserve natural structure and detect even tiny connections. The cortical parameters studied (cCSA, cBMD, cBMC, and cTh,) were not significantly influenced by diet-induced obesity, as also found in a previous study . However, some studies have shown positive effects of diet-induced obesity on cortical bone size , while negative effects on cortical bone mass  and size-independent mechanical properties  have also been suggested. These controversial results may be due to differences in animal age and the skeleton site measured as well as different measuring techniques.
Bone is a dynamic structure monitored by both intrinsic (body mass, hormone, cytokine, and other intrinsic factors) and extrinsic factors (environmental factors including physical activity, life style, etc.). On the one hand, the effect of diet-induced obesity on bone could be explained by alteration in the intrinsic mechanical loading environment caused by the increase in body mass. On the other hand, adipose tissue is regarded not just as a passive tissue for the storage of excess energy in the form of triglycerides, but also as an active endocrine organ secreting a variety of biologically active molecules, for example, leptin , resistin  and adiponectin . A cascade of events such as intense conversion of androgens into estrogens occurring in adipose tissue, alterations in other hormones or cytokines, and hyperinsulinemia may influence the bone microenvironment and increase bone mass . The elevated plasma leptin level in diet-induced obesity is a predictor of body mass accrual in different species [12, 45, 46]. The serum leptin level also regulates bone mass . However, the results of published studies on the effects of leptin level on bone are complex and controversial . Resistin is a controversial inflammatory-related factor , which also influences both osteoclast and osteoblast activity, resulting in increased bone remodelling . We found higher plasma leptin and resistin levels in obese mice, suggesting that plasma leptin and resistin levels have a positive effect on bone. However, their effects on bone mass and strength were the opposite, bone mass and strength showing a positive association with plasma leptin level and a negative association with resistin level (unpublished data). However, increased bone mass with increased body mass independent of leptin was also reported in a recent study . So far, the mechanism of interaction between bone metabolism and resistin remains unclear.
The development of obesity is associated with chronic inflammatory status, coinciding with significantly increased macrophage infiltration in adipose tissue and the expression of inflammatory cytokines, such as TNF-α, IL-6, monocyte chemotactic protein-1, and plasminogen activator inhibitor type-1 (PAI-1) . All of these inflammatory reactions are considered to be responsible for the majority of the obesity-related syndromes. Not surprisingly, this inflammatory status also influences bone metabolism through altering the micro-environment surrounding the bone cells. Over expression of PAI-1 increased bone strength and mineralization in an age- and gender-specific manner . Here, we found that a higher level of PAI-1 in obese mice correlated significantly with trabecular thickness, suggesting that PAI-1 had a positive effect on bone. In addition to these altered adipokins and inflammatory factors, we also detected higher levels of osteoprotegerin that is a bone resorption inhibitor . Together with the aforementioned factors, higher body weight may increase bone strength by shifting bone remodelling towards more active bone formation.
In order to investigate possible intervention methods, we also examined the effects of voluntary exercise. Here we found that voluntary wheel running was associated with a non-significant reduction in body mass with a concomitant improvement in glucose tolerance and insulin sensitivity and an increase in relative liver mass. Further studies are needed to find out whether the relative increase in liver mass in runners is associated with factors such as increased protein synthesis or glycogen storage. The minor reduction in body mass was not due to reduction in dietary intake. In fact, the runners consumed slightly more energy than their sedentary counterparts. Thus, the reduction in body mass was secondary to the increase in exercise-associated energy expenditure.
Physical activity has been shown to associate with bone mass and strength and to have positive effects on bone properties [51, 52]. In obese subjects, physical activity increased total, hip, and lumbar bone mineral content  and decreased plasma leptin level [54, 55]. More excitingly, long-term leisure time physical activity also showed positive effects on both cortical thickness and trabecular bone after controlling for the subjects' genetic background . The present study showed that voluntary exercise increased trabecular bone mineral density and improved the bone geometrical structure but led to a decrease in all of the cortical parameters. Previous animal studies have also shown positive effects of exercise on bone in different species at different ages with different types of exercises [56–59]. Most of these studies have focused on either cortical or trabecular bone mass or strength, with very few studies reporting micro-structural alteration induced by exercise. In C57BL/6J mice, from the age of six weeks onwards, trabecular volume and trabecular number are generally decreased  while trabecular thickness and trabecular space are increased up to an age of 24 weeks . Mori et al.  showed that intermittent voluntary climbing in eight-week-old C57BL/6J mice increased trabecular bone volume and reduced bone resorption, partially due to initial down-regulation of marrow osteoclastogenic cells and up-regulation of osteogenic cells, while further exercise desensitized them. In our study, the voluntary running lasted for 21 weeks, thus covering the entire growth period. We found that voluntary exercise tended to increase trabecular bone volume and decreased trabecular pattern factor and structure model index, shifting trabecular bone towards a stronger, more plate-like structure. However, in agreement with our previous report , we found that voluntary exercised animals under control diet had lower total BMC and cortical parameters. Similar findings were also reported in the rat tibia after intensive treadmill running , in the mice tibia after weight-bearing running during growth , and in 23-week-old female C57BL/6J mice after one month of voluntary wheel running . These effects might be explained by overactive modeling or remodeling of bone under continuous mechanical stimulation during growth [66–68]. However, histo-morphometric analyses  have suggested that decreased osteoblastic activity rather than a global adaptation of bone remodeling resulted in reduced longitudinal bone growth and bone loss in young rats under strenuous training. Another possible reason may be exercise-induced weight loss, which is accompanied by a reduced mechanical strain on the skeleton and decreased need of strong bones. Consequently, although numerous data indicate that, during growth, physical activity imposes its effect on bone more efficiently, the exercise programs or activities that will optimize bone structure and strength still remain unclear .