Areal BMD was measured using routine methods. Mean values of BMD from narrow neck region are on the average about 14 % higher than the conventional neck region of interest (ROI) values on the same subjects, although age trends are similar in previous reports [15].

Average cortical thickness is an estimate of the mean cortical thickness assuming a circular (narrow neck or shaft) or elliptical (intertrochanter) annulus model of the cross section for use in the estimated buckling ratio. The model assumes that 60 %, 70 %, and 100 % of the measured bone mass is in the cortex for narrow neck, intertrochanter, and shaft, respectively. These methods have the possibility of overestimate of the cortical thickness in the cases having the increase of trabecular bone, especially in narrow neck and intertrochanter regions.

This is defined as the surface area of bone tissue in the cross section after excluding soft tissue (marrow) spaces. CSA is derived from the integral of the bone mass profile and is not equivalent to the total area within the periosteal envelope as widely misinterpreted. CSA is an index of resistance to forces directed along the long axis of the bone.

This is an index of resistance to bending forces and is calculated as CSMI/d _max where CSMI is the cross-sectional moment of inertia and d _max is the maximum distance from either bone edge to the centroid of the profile [13, 14]. The CSMI is derived from the integral of the bone mass profile across the bone weighted by the square of distance from the center of mass. Due to limitations of a 2D DXA image, the CSMI and section modulus derived by HSA are only relevant for bending in the plane of the DXA image.

Buckling ratio describes stable configurations of thin-walled tubes subjected to compressive loads and requires an estimate of the cortical thickness. The buckling ratio is computed as the ratio of d _max to the estimated mean cortical thickness. Buckling ratio is mainly presented in narrow neck and intertrochanter region, because this parameter in the shaft is not important in meaning.

In addition to these parameters, the HSA program measures neck-shaft angle and femoral neck length. The latter is defined as the distance from the center of the femoral head to the intersection of the neck and shaft axes.

Based on a preceding study, precision error (CV%) for HSA variables at the narrow neck, intertrochanter, and shaft ranges from 0.8 to 4.7 %, with an average of 2.2 % [16, 17].

The trends in aging into the proximal femur geometry in Japanese women were analyzed using a cross-sectional sample obtained from a clinical sample [17]. The data of BMD was acquired from outpatient clinic for the screening of osteoporosis on 409 Japanese women aged from 50 to 93 years. For the purpose of interpretation of apparent age trends, mean values for women in the 50–59 year age group were used as a reference standard value (100 %), and all expressed means relative to standard values in 5 year intervals (those over 80 were put in a single group). Sample sizes in each group ranged from 50 to 90 subjects.

Areal BMD in each femur site appears to decline in the same trends. BMD at narrow neck, intertrochanter, and shaft in the oldest group (80+ years) decreased by 19.5 %, 21.8 %, and 17.1 % compared to the standard value, respectively (Fig. 4.3). Measured on the same subjects, section modulus are quite discordant with those in areal BMD. Section modulus at narrow neck decreased age-dependent manner; the oldest group (80+ years) averages 17.0 % less than standard value. In contrast, section modulus at intertrochanter and shaft appears to decline more slowly before 75 years of age; after that, it significantly decreased by 15.6 % in the oldest group (80+ years) at intertrochanter. There are no significant differences between narrow neck and intertrochanter in the group of 75–79 years and 80+ years. On the other hand, section modulus at shaft appears to remain static by 95.4 % until 70–74 years of age and 90.8 % in the group of 80+ years. This was also true in the rate of expansion of subperiosteal diameter which partially explains the reduction in areal BMD and that provides a mechanical explanation for why the apparent decline in section modulus is smaller than the change in BMD.

HSA in Japanese women showed that reduction in geometric strength as reflected by the section modulus was not dependent on decline in BMD. The section modulus is mainly influenced by changes in cortical dimensions. Mathematically, the section modulus is more strongly dependent on subperiosteal diameter than endocortical diameter, so that a greater bone loss at the endocortical surface can be compensated by a smaller increase in subperiosteal diameter. Subperiosteal expansion with age is thought to be a response to bone resorption in the endocortical surface for maintaining bone strength [19]. Indeed, subperiosteal diameters were positively correlated with age at all three regions in this study. Although cortices become progressively thinner through life, geometric strength at the femoral shaft is maintained.

Iki et al. showed the geographical differences in HSA parameters in four areas in Japan [20]. The areal differences were within the range of 10 % between the lowest area (Sanuki) and the highest area (Memuro) with an inverse association in buckling ratio.

The committee for osteoporosis treatment of the Japanese Orthopaedic Association had elucidated the current status of hip fracture incidence in Japan [21]. The number of patients with femoral neck fractures gradually increased with age from 60 to 75 years, and exceeded that with trochanteric fractures before 75 years. In contrast, the number of patients with intertrochanter fracture rapidly increases after 75 years of age, and these figures became inverted thereafter.

The results in a cross-sectional sample of Japanese women show that although BMD declines with age in all regions, the section modulus appears to decline at a slower rate than that of BMD. The changing pattern in the section modulus at narrow neck and intertrochanter in Japan might be consistent with the epidemiological evidence of hip fracture (neck and trochanteric fracture) rates in Japan (Fig. 4.3).

The previous reports have investigated gender differences in femoral neck geometry in an aging population and differences in pre- and postmenopausal hip geometry in white and black women [10, 15, 22, 23]. Note that there are some apparent differences in age trends between Japanese and US white women. Rates of bone loss (CSA decline) are actually greater in Japanese while rates of neck expansion (outer diameter) appear to be greater in US whites. The age trend in bending resistance is actually greater in Japanese women while the apparent increase in buckling ratio (toward greater instability) is greater in white women. These differences are reflected in geometric properties like the bone CSA and section modulus which define resistance to axial and bending loads, respectively.

Beck et al. analyzed structural parameters in non-Hispanic white men and women acquired in the Third National Health and Nutrition Examination Survey (NHANES III) [15]. The BMD decline with age in the narrow neck and shaft was similar to that in conventional regions. On the other hand, the section modulus at both the narrow neck and the shaft regions remains nearly constant until the fifth decade in females and then declined at a slower rate than BMD. In males, the narrow neck section modulus declined modestly until the fifth decade and then remained nearly constant whereas the shaft section modulus was static until the fifth decade and then increased steadily. The aging loss of BMD in the hip does not necessarily mean reduced mechanical strength.

Wang et al. measured the structural basis of racial and sex differences in femoral neck fragility in 829 healthy Chinese and 1181 healthy Caucasian subjects aged 18-93 years. For both races, women had a higher fracture risk in bending than men [22].

Longitudinal hip data in 4187 mostly white, elderly women were analyzed whether altered skeletal load should stimulate adaptive increases or decreases in hip geometry [27]. Subjects with increasing load (weight gain >5 % in 3.5 years) lost narrow neck BMD, but section modulus increased markedly. On the other hand, those with declining skeletal loads (weight lost >5 % in 3.5 years) showed the greatest loss of BMD and significant declines in section modulus (Fig. 4.4). These results indicated that mechanical homeostasis in the hip is evident in section modulus but not in BMD.

Takada et al. had reported the first presentation for the relations between the body composition (BMC, lean, fat) and HSA parameters in Japanese women [28]. One hundred eighty-three community-dwelling Japanese women (65.6 years of age) who had a checkup for osteoporosis were measured body composition in lower leg and hip geometry parameters by whole-body scan DXA and HSA, respectively. BMC and lean mass at narrow neck and intertrochanter were positively correlated with BMD, CSA, and section modulus (r = 0.425−0.874), and BMC was inversely correlated with buckling ratio (r = −0.494, −0.572). On the other hand, fat mass did not show significant correlation with CSA and section modulus (Table 4.4).