Previous milk consumption is associated with greater bone density in young women

BACKGROUND

Dietary calcium and milk intakes at specific ages may influence bone mineral measures at specific sites during development of peak bone mass.

OBJECTIVE

Relations of previous milk intake and current calcium intake to current bone mineral measures were investigated in young women.

DESIGN

A food-frequency interview and recall of previous milk intake from early childhood to 12 y of age and during adolescence (13-19 y) were completed in a cross-sectional analysis in young women (age 18-31 y; n = 224). Three levels of previous milk intake were defined: 1) infrequently or never, 2) sometimes, and 3) at every or almost every meal. Total body (TB), femoral neck, radius (R), and spine (S) bone mineral density (BMD) and bone mineral content (BMC) were determined by using dual-energy X-ray absorptiometry.

RESULTS

Childhood and adolescent milk intakes were positively correlated (r = 0.66). Childhood and adolescent milk intakes correlated with current calcium intakes (r = 0.26 and 0.33, respectively). Adolescent milk intake correlated with RBMD (r = 0.16). When weight was controlled for, adolescent milk intake correlated with TBBMD (r = 0.16), TBBMC (r = 0.21), SBMC (r = 0.16), RBMD (r = 0.18), and RBMC (r = 0.15). Current calcium intakes correlated with SBMC (r = 0.17). Regression analyses supported these results.

CONCLUSIONS

Results were consistent with the hypothesis that higher milk intake during adolescence is associated with greater total body, spine, and radial bone mineral measures during development of peak bone mass, whereas current calcium intakes may influence SBMC. In addition, milk intake at a younger age may contribute to similar habits of milk intake later in life.

Calcium and oxalic acid kinetics differ in rats

Small molecular weight calcium salts, if absorbed intact, could provide a nutritional source of calcium in subjects with impaired absorption of calcium by the saturable pathway. An understanding of the mechanism of absorption of calcium oxalate (as a representative salt) may be important nutritionally and therapeutically. The aim of the present study was to develop models to study absorption, distribution and retention of calcium and oxalate in rats as a basis for studying calcium oxalate absorption. Labeled compounds (45Ca and [14C]-oxalic acid) were administered to separate groups of rats orally (n = 8-11) or intravenously (n = 3-5) and blood was sampled for up to 240 min. Data were analyzed using SAAM/CONSAM. Calcium kinetics were fitted by a model with three compartments in the body and one absorption pathway from the intestine. By contrast, oxalic acid kinetics were fitted by two pools in the body and two absorption pathways from the intestine. Calcium and oxalic acid, therefore, demonstrate different absorption and distribution kinetics in rats.

Rates of growth and loss of bone mineral in the spine and femoral neck in white females

This study characterizes the rates of growth and loss of bone mass as a function of age in white females. It combines longitudinal data from several studies of bone mass on healthy white female subjects ranging from age 6 to 90 years. Rates of change in bone area, bone mineral content (BMC) and bone mineral density (BMD) are estimated separately for the spine and the femoral neck of each individual using linear regression. The individual rates of change are then fitted as a nonparametric function of age using weighted moving averages, resulting in a curve of age-specific mean change as a function of age. When the curves of BMD were compared between the hip and the femoral neck, the cessation of bone growth and the onset of bone loss were found to occur at an earlier age at the hip than at the spine. No significant differences in the ages of maximum rates of growth or maximum loss were found between the two skeletal sites. This information will be useful for designing interventions to promote bone growth or retard bone loss.

Absorption of calcium oxalate does not require dissociation in rats

Calcium absorption is thought to occur only if calcium is in a soluble or dissociated form, although experimental evidence is lacking. The intestinal absorption of calcium oxalate, a small, neutral and virtually insoluble calcium salt, was elucidated in the whole body of awake rats. Suspensions of 45Ca ascorbate, 14C-oxalic acid and doubly labeled 45Ca-[14C]-oxalate were given by gavage to separate groups of rats. Following dosing, blood samples were drawn for up to 240 min through a previously inserted intravenous catheter. Serum was assayed for radioactive tracers, and data were then plotted as fraction of dose over time. Calcium absorption was 15% [with a loading of 0.3 mmol (15 mg) calcium], oxalic acid absorption was 22% and Ca-oxalate absorption was <2%. Appearance of 45Ca from calcium ascorbate and 14C from oxalic acid differed, whereas 45Ca and 14C from doubly labeled Ca-oxalate had identical serum appearance profiles. Therefore, we conclude that calcium oxalate was absorbed intact. Addition of excess, unlabeled calcium to the doubly-labeled calcium oxalate did not alter the relationship of the serum level of the two tracers, confirming absorption of calcium oxalate as the intact salt. Thus, calcium bound as a small, neutral, calcium salt such as calcium oxalate does not have to be dissociated prior to absorption. Possibly other small compounds would be similarly absorbed. These results alter our current understanding of calcium bioavailability from foods and therapeutic agents.

Dietary calcium, protein, and phosphorus are related to bone mineral density and content in young women

BACKGROUND

Dietary factors have been implicated in modifying bone health, although the results remain controversial, particularly in young women.

OBJECTIVE

The objective of the study was to determine relations of selected dietary factors and anthropometric measurements to bone mineral density (BMD) of the spine, femoral neck, trochanter, Ward's triangle, radius, and total body and the bone mineral content (BMC) of the spine, radius, and total body.

DESIGN

The study was a cross-sectional analysis of 215 women aged 18-31 y.

RESULTS

Weight, height, and lean mass were correlated with bone mineral measures at every site (r = 0.17-0.78). Postmenarcheal age (years since onset of menses) was positively correlated with total-body BMD and BMC, radius BMD and BMC, and spine BMC, and negatively correlated with Ward's triangle BMD. Radius BMD was correlated with protein, calcium, and phosphorus intakes, and spine BMD and BMC were correlated with energy, protein, calcium, and phosphorus intakes. These correlations remained significant when postmenarcheal age, lean mass, and fat mass were controlled. A pattern emerged in multiple regression analyses that showed a complex relation among calcium, protein or phosphorus, and the calcium-protein or calcium-phosphorus ratio and spine or total-body BMC and BMD. All 3 variables (calcium, protein or phosphorus, and calcium-protein or calcium-phosphorus ratio) were required in the model for significance.

CONCLUSIONS

Anthropometric measures were predictors of bone mass. A single ratio of calcium to phosphorus or protein did not optimize bone mass across the range of calcium intakes.

Use of calcium tracers and biomarkers to determine calcium kinetics and bone turnover

The purpose of this article is to review approaches to measuring calcium metabolism and bone turnover, which can be used to assess bone loss during spaceflight and to evaluate the effectiveness of interventions aimed at reducing bone loss. Isotopic tracers of calcium can be used to determine absorption, excretion, retention, body pool sizes, and turnover rates. Because 99% of the body's calcium is in the bone, calcium tracer kinetics can provide an accurate assessment of bone turnover rates. Biomarkers can be used to assess qualitative changes in bone formation and bone resorption, but they do not predict bone density or quantitate net calcium retention. The usefulness of biomarkers has been validated against calcium tracer kinetics and bone histomorphometry. Calcium kinetics and biomarkers of bone turnover can assess changes in calcium metabolism and bone turnover long before changes in bone density can be detected.

Aluminum (26AI) metabolism in rats

Because of the lack of a suitable isotope and a sensitive technique of analysis, aluminum has been studied indirectly using analogs such as 67Ga (t1/2 = 78 hr). Recently, with the development of accelerator mass spectrometry (AMS), it has become possible to use the artificially produced radionuclide of aluminum, aluminum 26 (26AI), (t1/2 = 7.16 x 10(5) years). AMS is used for measuring long-lived and stable isotopes with the sensitivity of an attomole (10(-17) mol). To study aluminum metabolism, 26AlCl3 was administered to rats intraperitoneally (ip) by injection and orally by gavage (n = 3/group). Blood was collected periodically. On Day 8 following perfusion, blood, liver, kidney, femur, brain, and spleen were collected and analyzed for 26AI. Of all the tissues studied, 26AI accumulation was greatest in the bone. 26AI accumulated in tissues as: bone > spleen > kidney approximately liver > brain, but absorption was low (0.97% of dose). AMS offers great potential in AI research as it is the only technique available for tracer aluminum study.

Quantification of biochemical markers of bone turnover by kinetic measures of bone formation and resorption in young healthy females

The quantification of biochemical markers of bone formation and resorption with kinetic measures of bone turnover is an essential step in their validation. Some biochemical markers have been validated by quantification against formation and resorption rates measured by calcium kinetics in adults with bone disease. However, none has been validated in healthy individuals who are undergoing skeletal growth and bone consolidation. Therefore, we have measured biochemical markers of bone formation (serum osteocalcin [OC], bone-specific alkaline phosphatase [BAP], and total alkaline phosphatase [ALP]) and resorption (serum tartrate resistant acid phosphatase [TRAP], urinary cross-linked N teleopeptides of type I collagen/creatinine [NTx/Cr], and hydroxyproline/creatinine [OHP/Cr]) in healthy females aged 11-32 years (n = 31) after an overnight fast to determine their relationship with bone formation (Vo+) and bone resorption (Vo-) as measured by calcium kinetics and balance. All biochemical markers were highly intercorrelated (r > 0.6, p < 0.001) as were Vo+ and Vo- (r = 0.91, p < 0.001). Highly significant correlations were present between bone formation measured by calcium kinetics (Vo+) and serum levels of bone biochemical markers (OC, r = 0.82, p = 0.001; ALP, r = 0.92, p = 0.001; and BAP, r = 0.90, p = 0.001) and between bone resorption measured by calcium kinetics (Vo-) and fasting serum levels and urine creatinine ratios of biochemical markers (TRAP, r = 0.77, p < 0.001; OHP/Cr, r = 0.79, p < 0.001; and NTx/Cr, r = 0.70, p < 0.001). Thus, biochemical markers of bone formation and resorption can be used to predict calcium kinetic rates during skeletal growth and the early years of formation of peak bone mass, ages at which strategies to build peak bone mass are important. Biochemical markers of formation and resorption are equally useful in predicting either the bone formation rate or the resorption rate.

Calcium retention in relation to calcium intake and postmenarcheal age in adolescent females

Achievement of maximal calcium retention during adolescence may influence the magnitude of peak bone mass and subsequently lower the risk of osteoporosis. Calcium retention is generally considered to reach a plateau at a certain calcium intake. To test this hypothesis, calcium balance was measured in 35 females with a mean (+/-SD) age of 12.7 +/- 1.2 y (range: 12-15 y) who consumed from 841 +/- 153 to 2173 +/- 149 mg Ca/d. Subjects ate a basal diet that included a fortified beverage containing different amounts of calcium citrate malate. Twenty-one subjects were studied at two dietary calcium intakes with use of a crossover design. Results from a previous study in 14 subjects who were studied at only one calcium intake were included in the data analysis. Calcium retention was modeled as a nonlinear function of calcium intake that included a parameter representing mean maximal retention. Mean maximal calcium retention was 473 mg/d (95% CI: 245, 701 mg Ca/d). At higher postmenarcheal ages, maximal calcium retention was lower but the intake required to achieve this was not affected. Calcium intake explained 79% and 6%, respectively, of the variation in fecal and urinary calcium excretion. Intake of 1200 mg Ca/d, the recommended dietary allowance for calcium published in 1989, resulted in a mean calcium retention that was 57% of the maximal value (95% CI: 25%, 89%). Intake of 1300 mg Ca/d was the smallest intake that allowed some adolescent females to achieve 100% of maximal calcium retention (95% CI: 26%, 100%). These data support the idea that calcium retention plateaus at a certain calcium intake although it continues to increase at intakes > 2 g/d.

Calcium retention estimated from indicators of skeletal status in adolescent girls and young women

To determine clinically useful predictors of calcium retention during postpubertal growth, calcium balance, bio-chemical markers of bone turnover, and anthropometric variables were determined in 14 girls aged 11-14 y and in 11 young women aged 19-30 y. Subjects participated in a 3-wk calcium-balance study with a calcium intake of 1332 mg/d. Biochemical markers of bone turnover (serum osteocalcin, total alkaline phosphatase, bone alkaline phosphatase, tartrate-resistant acid phosphatase, and urinary cross-linked N-teleopeptides of type I collagen and hydroxyproline as the creatinine ratios) were measured in fasting samples. Total-body bone mineral density and total-body calcium content were significantly higher in adults than in adolescents (1.17 compared with 1.05 g/cm2 and 1019 compared with 791 g, respectively). At the observed retention of 326 mg/d, adolescents would require 2 y to reach the total bone calcium of the young adults. All biomarkers of bone turnover were strikingly higher in adolescents than in adults and were strongly correlated with calcium retention. A multiple-regression model using a biochemical marker of bone turnover (serum osteocalcin) and postmenarcheal age (a measure of sexual maturation) described 75% of the variability in calcium retention.

Differences in calcium kinetics between adolescent girls and young women

Rapid bone accretion occurs throughout childhood but peaks during adolescence. The achievement of optimal peak bone mass, which can protect against osteoporosis later in life, is greatly dependent on rates of bone accretion. To identify differences in calcium metabolism during rapid vs. slower bone accretion, calcium kinetics were compared in 14 healthy girls aged 11-14 yr and 11 women aged 19-31 yr. Calcium kinetics were measured while subjects were undergoing a calcium balance study in a camp simulating a free-living environment. After 7 days on a diet containing 1,330 mg Ca/day, two stable isotopes were administered (44Ca orally and 42Ca intravenously), and blood samples and all urine and feces were collected for 14 days. Samples were analyzed for total calcium by atomic absorption spectrophotometry and for enrichment of 42Ca and 44Ca by fast atom bombardment mass spectrometry. Data from serum, urine, and feces were analyzed using the Simulation, Analysis and Modeling (SAAM) software. Data were fitted by a three-compartment model; the first pool was the same size in girls and women (1.6 g) but the second and third pools were larger in girls (2.85 vs. 1.66 g and 12 vs. 5 g). Compared with the women, girls absorbed more calcium from the diet (38 vs. 22% or 494 vs. 283 mg/day), excreted less calcium in urine (100 vs. 203 mg/day), deposited more calcium in bone (1,459 vs. 501 mg/day), and resorbed more calcium from the skeleton (1,177 vs. 542 mg/day), whereas endogenous calcium excretion did not differ between girls and women (112 vs. 121 mg/day). Girls retained more calcium than women (282 vs. -41 mg/day) through increased absorption, lower urine excretion, and higher bone turnover.

Calcium bioavailability from bovine milk and dairy products in premenopausal women using intrinsic and extrinsic labeling techniques

Stable isotopes were used to compare calcium fractional absorption from intrinsically and extrinsically labeled bovine milk as well as intrinsically labeled dairy product and cheese analogue. Healthy Caucasian women were fed a controlled diet for 4 d during the follicular phase of their menstrual cycle. With breakfast on the third day, participants ingested milk containing 44Ca (intrinsic) and 42CaCl2 (extrinsic) or dairy products containing 44Ca. Total feces were collected for 2 d prior to and 10 d after isotope ingestion. Polyethylene glycol was administered to monitor completeness of fecal collections. Total calcium was determined by atomic absorption spectrophotometry, and isotopic abundance was determined by high resolution fast atom bombardment mass spectrometry. Fractional absorption was determined as the difference between the administered isotopic dose and the quantity of 44Ca or 42Ca excreted in feces. The fractional absorption of calcium from milk was not affected by the method of labeling, lactose content, fermentation or the chemical form of calcium in dairy products or cheese analogue.

Previous physical activity relates to bone mineral measures in young women

Exercise may increase accretion of bone, potentially reducing the risk of osteoporosis. Previous physical activity was assessed in 204 minimally active young women (18-31 yr). Bone mineral content (BMC) and bone mineral density (BMD) for the total body, femoral neck, and spine were assessed by a dual x-ray absorptiometer, and the radius by a single photon absorptiometer. Self-reported occupation and leisure activity for the 5 yr before enrollment in the study, as well as high school and college sports participation, were assigned energy expenditure (EE) values. From this information, EE variables were created as follows: 1) occupation EE + leisure EE + high school sport and/or college sport EE if within prior 5 yr (5-yr EE); 2) occupation EE + leisure EE (occupation + leisure EE); and 3) high school sport EE (high school EE). These variables were correlated with bone mineral measures and significant results follow (P < 0.05). Five-year EE and occupation + leisure EE correlated with all measures of bone health (r from 0.13 to 0.39). High school EE correlated with total body BMD (r = 0.25) and BMC (r = 0.28), femoral neck BMD (r = 0.28), radius BMC (r = 0.20), as well as spine BMD (r = 0.20) and BMC (r = 0.27). When weight was controlled, 5-yr EE and occupation + leisure EE remained correlated with all BMC measures (r from 0.14 to 0.22). When controlled for weight, high school EE remained associated with femoral neck BMD (r = 0.24), total body BMD (r = 0.20) and BMC (r = 0.26), and spine BMC (r = 0.17). To partially control for selection bias, data were also controlled for total body BMD. Five-year EE and occupation + leisure EE remained positively correlated with all measures of BMC. High school EE remained correlated both with femoral neck BMD and total body BMC. In multiple regression analyses, 5-yr EE or occupation + leisure EE were significant predictors of all measures of bone health, except femoral neck BMD. High school EE was a significant predictor for total body BMD and BMC, femoral neck BMD, and spine BMC.

Wheat bran abolishes the inverse relationship between calcium load size and absorption fraction in women

Fractional calcium absorption from varying intakes of calcium carbonate co-ingested with wheat bran, as well as alone, was measured in a randomized crossover study in healthy adult women. The calcium carbonate was intrinsically labeled with 45Ca. Absorption from the carbonate, ingested without bran, showed the expected inverse relationship to the logarithm of ingested load size (slope = -0.1199; not substantially different from the value previously reported for milk). At 0.5 mmol calcium load, fractional absorption averaged 0.769 +/- 0.134, whereas at 12.5 mmol load it averaged 0.378 +/- 0.069. In contrast, fractional absorption from calcium carbonate co-ingested with 40 g of a cereal product containing 16 g wheat bran, across a calcium load range from 0.5 to 15.5 mmol, was essentially constant (mean for all loads: 0.230 +/- 0.069). Thus, the calcium-binding capacity of the bran cereal altered the usual inverse relationship between calcium load and fractional absorption. In vitro calcium binding to the bran cereal was linear over a wide range of calcium levels. This suggests that binding of calcium to one or more components of the bran cereal is sufficient to explain the reduced absorption demonstrated in vivo.

Effects of long-term moderate exercise on iron status in young women

The impact of long-term (6-month) moderate exercise on the iron status of previously sedentary women was determined by randomly assigning 62 college-age women into one of the following four groups: 1) 50 mg.d-1 iron supplement, low iron diet (N = 16); 2) Placebo, free choice diet (N = 13); 3) Meat supplement to achieve 15 mg.d-1 iron intake (N = 13); and 4) Control, free choice diet (N = 20). All groups except the Control group exercised 3 d.wk-1 at 60%-75% of their heart rate reserve. VO2max was measured at baseline and week 24. Blood was sampled at baseline and every 4 wk thereafter for 24 wk to measure iron status and to elucidate the causes for alterations in iron status. Subjects had depleted iron stores throughout the study as indicated by their serum ferritin levels (< 15 ng.ml-1). Serum iron, total iron binding capacity and transferrin saturation were not compromised with exercise. Mean hemoglobin level in the Placebo/Ex group was significantly (P < 0.05) lower than the 50 Fe/Ex and the Meat/Ex groups by week 24. However, changes in serum albumin, haptoglobin, and erythropoietin data from the study cannot explain these changes.

Peak bone mass in young women

Increasing peak bone mineral density (BMD) or content (BMC) in young women may help to reduce the incidence of osteoporosis. Identifying the age when peak bone content or density is attained is essential to develop strategies aimed at optimizing peak BMD and BMC. Total body bone mineral density (TBBMD) and content (TBBMC) were measured by a dual X-ray absorptiometer in healthy females (n = 247, aged 11-32 years). TBBMD and TBBMC were modeled separately as a nonlinear function of age. By age 22.1 +/- 2.5 years, 99% of peak BMD is attained, and by age 26.2 +/- 3.7 years, 99% of peak BMC is attained. Nonlinear relationships between weight and TBBMD or TBBMC were also modeled. In this model, the influence of several parameters, including age, weight, and height, on BMC and BMD were simultaneously assessed. A model with age and weight described the best fit for TBBMD, whereas age, weight, and height described the best fit for total body TBBMC.

Magnesium supplementation and osteoporosis

Among other things, magnesium regulates active calcium transport. As a result, there has been a growing interest in the role of magnesium (Mg) in bone metabolism. A group of menopausal women were given magnesium hydroxide to assess the effects of magnesium on bone density. At the end of the 2-year study, magnesium therapy appears to have prevented fractures and resulted in a significant increase in bone density.

Differences in calcium metabolism between adolescent and adult females

A 3-wk metabolic study measured calcium balance in 14 white adolescent girls and 11 young adult women. Subjects were housed in a sorority to simulate a free-living environment. A 6-d menu cycle consisted of foods typically eaten by teenagers and averaged 1332 mg Ca/d. Adolescents had a significantly higher calcium balance of 326 +/- 107 mg/d (mean +/- SD) than adults, who averaged 73 +/- 104 mg/d (P < 0.001). No adult > age 21 y was in positive calcium balance. Adolescents had lower urinary calcium excretion values (P < 0.001), lower fecal calcium excretion (P < 0.01), and greater net absorption (P < 0.001) than adults. Calcium balance was negatively correlated with years postmenarche (r = -0.788, P = 0.0001) and height (r = -0.650, P = 0.001). Net calcium absorption was positively correlated with parathyroid hormone concentrations (r = 0.537, P = < 0.01). Thus, the growth demands of adolescents are met by more efficient net absorption and retention of calcium compared with young adults.

Effect of psyllium on absorption of co-ingested calcium

OBJECTIVE

To determine whether psyllium fiber (in the form of Metamucil) interferes with absorption of calcium ingested at the same meal.

DESIGN

Three-way, randomized, cross-over study in 15 healthy postmenopausal women, with calcium-fortified orange juice as the calcium source in all three meals. The test load of calcium was 219 mg (approximately 5.5 mmol). One test meal contained Metamucil, providing 3.4 g psyllium fiber; one had an equivalent amount of neutral cellulose; the third had no added fiber.

METHODS

The calcium-fortified orange juice was extrinsically labeled with 45Ca; calcium absorption fraction was calculated from the specific radioactivity of serum calcium at 5 hours after tracer ingestion.

RESULTS

Fractional absorption of calcium from orange juice without added fiber averaged 0.341 +/- 0.059; in the presence of psyllium fiber, 0.317 +/- 0.067; and in the presence of neutral fiber, 0.354 +/- 0.083. While the absorption with cellulose was slightly higher than in the absence of fiber, and the absorption with psyllium was slightly lower, neither difference was statistically significant. However, the difference between added psyllium and cellulose was statistically significant (P < .05).

CONCLUSION

Psyllium in the form of Metamucil makes little practical difference to availability of co-ingested calcium when ingested at typical therapeutic doses.

Age related calcium requirements due to changes in absorption and utilization

An attempt was made to estimate average calcium requirements for various age groups by gender. The factorial method that sums calcium accretion and endogenous losses modified by fractional absorption was used during growth and calcium intakes to achieve calcium balance was used for adults. These estimates exceed the current RDA for females aged 6-16 y and older than 30 y and for males aged 6-30 y. They also exceed current calcium intakes in the U.S. during adolescence for both sexes and for females above the age of 11 y. A number of factors influence calcium requirements with age. Requirements for calcium accretion is dominant until peak daily calcium retention during the adolescent growth spurt is achieved. The rate of calcium accretion with age varies widely, depending on the method of estimation. Urinary calcium losses increase until young adulthood. Contributions of endogenous fecal calcium and sweat calcium to obligatory calcium losses are largely unknown during growth. After growth ceases, calcium requirements increase with menopause, unless estrogen-treated, and with age as calcium absorption efficiency decreases. Data required to estimate calcium requirements during growth are insufficient to precisely define dietary requirements with age. Determining calcium intakes which optimize calcium retention, and therefore, bone mass as has been estimated by Matkovic and Heaney (1992) may be a sound approach. The requirements estimated by this approach will be difficult to achieve given current dietary intakes in the U.S.

Geometric structure of the femoral neck measured using dual-energy x-ray absorptiometry

An algorithm was developed to estimate the strength of the femoral neck from data generated by the dual-energy x-ray absorptiometry (DXA). This algorithm considers shape of the proximal femur as well as cross-sectional moment of inertia (CSMI) in the estimate. Proximal femora (10) from cadavers of white adults and an aluminum step wedge were scanned with the Lunar DPX to validate the calculation of CSMI. After scanning, each femoral neck was sectioned at its narrowest portion for direct measurement of CSMI. Three healthy young women were scanned five times each to evaluate the reproducibility of geometric measurements using DXA. There was a strong linear association between the CSMI measured directly and using DXA in both cadaver bones (r2 = 0.96) and the aluminum step wedge (r2 = 0.99). The coefficient of variation for CSMI from repeated measurements using DXA was less than 3%. This indicates that it is possible to estimate reproducibly the bending rigidity of bone from DXA measurements. The data from 306 normal subjects were analyzed to investigate geometric changes in the femoral neck with age. Although there was no strong correlation between CSMI and age in normal subjects of either sex, safety factor (SF, an index of strength of the femoral neck during walking) and fall index (FI, an index of the strength of the femoral neck during a fall) decrease with age in both sexes. We observed an alteration of the geometric structure of the femoral neck with age that may increase the stress on the femoral neck and decrease SF and FI.