British Journal of Radiology (2005) 78, 493-498
© 2005 British Institute of Radiology
doi: 10.1259/bjr/42380498
Differences in bone mineral density and geometry in men and women: the Newcastle Thousand Families Study at 50 years old
S P Tuck, BSc, MB, MRCP1,
M S Pearce, MSc, CStat, PhD2,
D J Rawlings, BSc, MPhil1,
F N Birrell, MA, MRCP, PhD3,
L Parker, PhD, FRCPCH, FFPHM2 and
R M Francis, MBChB, FRCP1
1 Department of Rheumatology, Musculoskeletal Unit, Freeman Hospital, Freeman Road, Newcastle upon Tyne NE7 7DN, 2 Sir James Spence Institute of Child Health, School of Clinical Medical Sciences, University of Newcastle, The Royal Victoria Infirmary, Queen Victoria Road, Newcastle upon Tyne NE1 4LP and 3 Department of Rheumatology, University of Newcastle upon Tyne, Medical School, Framlington Place, Newcastle upon Tyne NE2 4HH, UK
Correspondence: Dr R M Francis, Musculoskeletal Unit, Freeman Hospital, Freeman Road, Newcastle upon Tyne NE7 7DN, UK
 |
Abstract
|
|---|
In 1947 Sir James Spence initiated the Newcastle Thousand Families study, which recruited all 1142 children born in the city between May and June that year. At the age of 50 years, 832 survivors were traced and invited to attend for measurement of bone mineral density (BMD) by dual energy X-ray absorptiometry (DXA). The aim was to compare BMD measurements of men and women in this cohort, before and after adjustment for skeletal size. The femoral neck shaft angles (NSA) were also measured manually from the DXA scan printouts. A total of 171 men and 218 women agreed to participate. As expected men had greater bone mineral content and bone area at all sites (p<0.0001) and were taller and heavier (p<0.0001) than women. Men also had significantly higher BMD than women at all regions (p<0.0002), except at the femoral neck or lumbar spine. After correction for skeletal size and body weight, men had statistically significantly lower volumetric BMD at all sites. The measurement of NSA had good intra/interobserver errors and precision (coefficient of variations 0.79%, 1.2% and 1.2%). Men had significantly larger NSAs (mean 130°, range 121–138°) than women (mean 128°, range 119–137°). We conclude that there are gender differences in BMD, skeletal size and geometry in middle aged men and women, which together with the subsequent rate of bone loss, may influence fracture risk in later life.
|
Introduction
|
|---|
Osteoporotic fractures are a major health problem causing excess mortality and considerable morbidity, resulting in substantial cost to the health and social services. There is an inverse relationship between bone mineral density (BMD) and fracture risk, with 1 standard deviation (SD) decrease in BMD being associated with a two to threefold increase in fracture incidence [1]. The risk of fracture is not only determined by BMD, but also by skeletal geometry and morphometry. Femoral geometry, including neck shaft angle (NSA), is an independent predictor of fracture risk, with 1 standard deviation increase in femoral NSA resulting in a doubling of fracture risk in both men and women [2, 3].
Dual energy X-ray absorptiometry (DXA) generates two-dimensional or areal BMD (aBMD) measurements (g cm–2), which are influenced by the size of the skeleton. Throughout life males have larger bones than females, as a result of which men have a greater bone cross-sectional area and hence a higher aBMD than women [4, 5]. Melton et al found aBMD to be 15% higher in men than women, but when correction for skeletal size was made this difference fell to only 6% [4]. Indeed, the true volumetric BMD (vBMD) has been reported to be the same in healthy age-matched men and women [5].
The Newcastle Thousand Families cohort study was initiated in 1947 by Sir James Spence, when all 1142 children born in May and June that year to mothers resident in the city of Newcastle upon Tyne were recruited. The health, growth and development of this cohort were followed in great detail up to the age of 15 years [6]. More recently survivors were followed up when they reached 50 years of age. Results from this 50 year follow up have been published on the effects of early life and later determinants of adult disease [7], risk factors for cardiovascular disease [8], central metabolic syndrome [9] and implications of childhood obesity for adult health [10].
As part of the 50 year follow up, participants of the Newcastle Thousand Families cohort study were invited to have BMD measured by DXA. The aim of the present study was to compare BMD measurements in this well-characterized cohort of 50-year-old men and women and to examine the effect of adjustment for skeletal size. We also investigated a method for the manual measurement of the femoral NSA using the DXA scan printouts, which enabled us to make gender comparisons of femoral geometry in this cohort.
|
Patients and methods
|
|---|
Design of the Thousand Families Study
Details of the original Newcastle Thousand Families study have been described elsewhere [6, 11–13], but will be outlined briefly. A total of 967 of these children were followed through until the end of the first year. In the 1990s, 832 study members were traced, either through the NHS Central Register or by contacting the study team in response to local, national and international media publicity. Between October 1996 and December 1998 self-completion questionnaires on health and lifestyle were sent out and members invited to attend for clinical examination and to take part in further studies [7–10]. The study has local ethics committee approval and all subjects gave their informed consent.
The clinical examination measured height, weight and a number of physiological risk factors for age-related disorders [8]. In addition participants were invited to attend for DXA scanning, which was performed using a Hologic QDR 2000 machine (Hologic Instruments, Waltham, MA). This had a daily calibration check using the Hologic spine phantom and had a coefficient of variation of 0.5% throughout the period of study. The lumbar spine (L1 to L4) and all regions of the hip (total hip, femoral neck, trochanteric and intertrochanteric) were measured. This generated data on area, bone mineral content (BMC) and BMD at these sites. In order to correct for size differences, an estimate of volumetric density was made (g cm–3); the bone mineral apparent density (BMAD). This was calculated using the formulae of Katzman et al [14] whereby:
Lumbar spine BMAD=BMC/area3/2
Femoral neck BMAD=BMC/area2
where the area at the lumbar spine or femoral neck was the projected bone area, as determined from the BMD scan. BMAD cannot be calculated for the remaining sites and so an adjustment was made by dividing the BMD by height in metres. This has previously been reported to be a useful method of reducing or eliminating the difference in BMD found between men and women owing to skeletal size [4].
Geometric measurements
No morphometric measurements were made during the BMD assessment. The DXA scanner used was of fan beam design and so magnification errors made it inappropriate to measure hip axis length or femoral neck width. However, the effect of magnification on femoral NSA was thought to be minimal. In an attempt to verify this, a cadaver femur was positioned in a tissue equivalent phantom at 2 cm intervals 0–12 cm above the DXA couch.
For both patient and phantom scans NSA was measured from the DXA scan printout using a method adapted from that of Faulkner et al [15, 16]. The femoral neck axis was identified on the printout by the DXA analysis software. A line was then drawn manually from the junction between the greater trochanter and the femoral neck down to a point in the middle of the shaft at the bottom of the scan (Figure 1
). The junction of these two lines gives the femoral NSA, which was measured with a long armed protractor with 0.5° intervals (a BIOMET Inc. goniometer; Biomet, Inc., Warsaw, IN). The intraobserver and interobserver errors were determined as the average coefficient of variation of a sample of 20 scans. The precision error (reproducibility) was calculated from a further 20 subjects who had had three successive scans. All printouts were generated from the same printer attached to the DXA workstation.
View larger version (95K):
[in this window]
[in a new window]
|
Figure 1. Measurement of the femoral neck shaft angle from the dual energy X-ray absorptiometry (DXA) scan printouts. The line AB is the hip axis marked by the scanner's software. A line is then drawn from C to D, in which C is the point at which the greater trochanter joins the femoral neck and D is the midpoint of the shaft at the bottom of the picture. E is the femoral neck shaft angle.
|
|
Analysis
Statistical analysis was performed using standard statistical software packages (Graphpad Prism, SanDiego, CA) and SPSS for Windows (SPSS, Chicago, IL). Descriptive statistics were obtained and the data tested for normality. Data, which were normally distributed, were then compared using Student's unpaired t-tests. Associations were assessed using Pearson's correlation. The derived variables BMD/height and BMAD were compared before and after adjustment for weight using regression analysis.
|
Results
|
|---|
Response rate
Of the 832 cohort members traced, 574 (69%) completed the questionnaire and 412 (50%) attended for clinical examination [8]. A total of 389 subjects (42%) comprising 171 men and 218 women underwent DXA scanning. This sample did not differ significantly from the remainder of the original cohort of 1142 individuals in terms of birth weight and social class at birth.
DXA scan results
The results of the DXA scans are shown in Table 1, which compares men and women. All subjects were scanned within a few months of their 50th birthdays resulting in a mean age of 50.1 years for both men and women. The BMD was significantly higher in men than women at the total hip, trochanteric and intertrochanteric regions, but not at the lumbar spine and femoral neck.
Men had a statistically significantly greater cross-sectional area and BMC at all sites measured and were significantly taller, reflecting their larger skeletal size (Table 1
). Men were also significantly heavier. The results of correction for skeletal size are shown in Table 2. Dividing the BMD by height resulted in the loss of the statistically significant difference previously seen between men and women at the total hip, trochanteric and intertrochanteric regions. However, at the femoral neck and lumbar spine, BMAD was significantly lower in the men.
View this table:
[in this window]
[in a new window]
|
Table 2. Comparison of volumetric bone mineral density (vBMD) (bone mineral apparent density (BMAD) and BMD/height in metres) by gender and site
|
|
Both height and weight were strongly correlated with BMD. For BMAD and BMD/height, there was no residual correlation with height, but the correlation with weight persisted. The data were therefore further adjusted to the mean weight for the entire study group (both men and women) of 74.05 kg. This resulted in men having statistically significantly lower adjusted bone mineral densities than women at all sites (Table 3).
View this table:
[in this window]
[in a new window]
|
Table 3. Comparison of volumetric bone mineral density (vBMD) (bone mineral apparent density (BMAD) and BMD/height in metres) by gender and site after adjustment to the mean weight of 74.05 kg
|
|
Comparison of the BMD measurements, with those of the currently available normal reference data from NHANES III and Hologic, show them to be similar to that expected once these have been standardized to age 50 years (Table 4). The BMD means and ranges are similar, indicating that the cohort is representative of the population as a whole.
View this table:
[in this window]
[in a new window]
|
Table 4. Comparison of mean bone mineral density (BMD) (g cm–2) and 95% confidence intervals (CI) between the Thousand Families study and the Hologic and NHANES III reference databases (standardized to age 50 years)
|
|
Femoral neck shaft angles
The method described gave an intraobserver error of 0.79%, interobserver error of 1.2% and precision of ±1.2%. The results are shown in Figure 2. Men had a mean femoral NSA of 130° (SD 3.3, range 121–138°). Women had a significantly (p<0.0001) smaller mean femoral NSA of 128° (SD 1.7, range 119–137°). Since the height of the femur neck above the scanning couch was not known for this group of subjects, these results take no account of possible image distortion caused by the shape of fan beam. Phantom measurements, however, indicate an apparent change in NSA of +0.27° degree per cm (Figure 3).
View larger version (20K):
[in this window]
[in a new window]
|
Figure 3. Measurement, from the dual energy X-ray absorptiometry (DXA) scan printout, of neck shaft angle for a single cadaver femur. The femur had been positioned, with appropriate internal rotation, at a range of distances above the DXA scanning couch to characterize image distortion caused by the fan beam.
|
|
|
Discussion
|
|---|
This paper presents the results of areal and estimated volumetric BMD and femoral geometry measurements in the Newcastle Thousand Families Study at age 50 years. Although previous studies have made gender comparisons of BMD and skeletal size [4, 5], these have included subjects with a broad age range. Our study utilizes a well-characterized cohort of men and women at the age of 50 years, before significant involutional bone loss has occurred or the confounding effect of degenerative changes on spine bone density measurements becomes a major problem.
Men had greater bone mineral content and area, indicating a larger skeletal size, consistent with them being significantly taller and heavier than women. Consistent with this finding, BMD was greater at all sites in men than women except for the femoral neck and lumbar spine. This is broadly consistent with other studies comparing men and women, which have found men to have significantly higher BMD at all sites [4, 17]. Correction for body size, either by using height or by calculation of BMAD, resulted in significantly lower size-adjusted BMD at the lumbar spine and femoral neck in men. After adjusting for weight, there was statistically significantly lower BMD in men at all sites. The contrasting effects of adjustment at different skeletal sites suggests that size cannot be wholly represented by a simple scaling factor. Hence, it would not be anticipated that all data might demonstrate an identical response to the adjustment for body size.
Other studies have demonstrated that correction for skeletal size resulted in a reduction in the difference in BMD between the sexes [4, 17, 18]. Duan et al found the volumetric BMD (vBMD) at the lumbar spine to be similar in young men and women aged 18–43 years, but in those aged over 60 years, vBMD was greater in men than in women [17]. Our results may be different because the subjects are all the same age and the women are perimenopausal. It may therefore represent an age group for which BMD is transitional, before menopausal bone loss results in lower vBMD in women than men. Long-term follow up of this cohort may answer this question. An alternative explanation is that this is a cohort effect. The group was born in 1947 at a time when there was post war rationing and it is possible that this and other unknown factors operating at that time could have affected the results.
Our method of femoral NSA measurement proved both reliable and precise. The precision is similar to the 0.5% and 1% quoted by Qureshi and Faulkner, respectively, [15, 16] and the intraobserver error of 1.5% and precision of 1.7% quoted by Gomez Alonso et al [2]. Furthermore, the mean values and ranges are similar to those reported in other studies [2, 3, 15, 16, 19, 20]. Using this method we were able to demonstrate that men had a larger NSA than women.
However, despite being statistically significant, this difference in NSA was small. The fan beam could have distorted the image and our phantom measurements showed an apparent change in NSA of +0.27 degree per cm from the scanning couch. Anthropometric data relating to a normal population of seated individuals are available [21]. Assuming that these translate directly to subjects lying supine on a DXA scanner then the median height above the couch for a male neck of femur is 12.0 cm (SD 0.5 cm). For a female neck of femur the median is 13.0 cm (SD 0.5 cm). Given the magnitude and direction of the measured influence of couch height on neck shaft angle, correcting patient NSA data for any such effects would have resulted in small individual changes in NSA which would have increased, rather than decreased, the significance of the above gender differences. Hence, manual assessment of NSA from fan beam DXA printouts is reliable and can be performed without additional software.
We found men to have a significantly larger NSA than women, consistent with that reported by Gomez Alonso et al [2]. This larger femoral NSA might be anticipated to predispose to an increased hip fracture risk in later life [2, 19], yet men have a lower risk than women. This is likely to be related to slower bone loss with advancing age in men, but may also be due to other gender differences in skeletal geometry. Both Karlsson et al [19] and Gomez Alonso et al [2] have demonstrated that men have larger femoral neck widths than women. Looker et al found men to have greater periosteal and cortical widths than women at all sites in the hip, even after correction for body size [22]. Furthermore, the gender differences in cortical bone size increased with age [22], suggesting that these structural differences may contribute to the lower fracture risk in men.
The study has a number of strengths and weaknesses. It is a large study with 389 subjects, all at age 50 years. Of the 1142 study members recruited at birth in 1947, 34% were included in the current study. This inclusion rate, 50 years after the study began, is higher than that reported in other studies (for example, 23% of a hospital based cohort were included in a study of fetal growth and adult hyperinsulinaemia [23] and less than 5% of a Hertfordshire birth cohort were included in a study of impaired glucose tolerance [24]).
There was no difference between the participants and the remainder of the original cohort in terms of birth weight and social class at birth. In addition, inclusion of study members no longer resident in the study region (18% of those attending a health check were resident outside of the Northern Region of England) increased how well the studied represented the population studied. Hence the study members are representative of all those born in Newcastle upon Tyne in May and June 1947.
At the time of scanning no morphometric measures were undertaken at the lumbar spine, so neither the degree of deformity nor direct measures of vertebral size were possible. Also, femoral neck geometry was not considered, which restricted analysis to neck shaft angles alone. Therefore we were unable to examine all aspects of size, structure and geometry, which may differ between men and women and affect their fracture risk.
|
Conclusions
|
|---|
In this paper we have presented the results of BMD measurements in the Newcastle Thousand Families Study. We have demonstrated that differences in areal BMD measurements between men and women are reversed following adjustment for skeletal size: vBMD is lower in men than in women at the age of 50 years. However, this may represent an age group for which BMD is transitional, before menopausal bone loss results in lower vBMD in women than men.
We have confirmed that this manual method of femoral NSA measurement is both reliable and precise and can be used retrospectively. Using this method we have shown men to have larger femoral NSAs than women, despite their lower fracture risk.
The present study did not examine the relationship between skeletal size, BMD and geometry on fracture incidence. However, post-menopausal bone loss may contribute to the increased risk of fracture in older women, when the vBMD becomes lower than men. Future studies of the pathogenesis of fracture should examine the interaction between skeletal size, BMD and skeletal geometry. Further work on the Newcastle Thousand Families study will explore the contribution of birth weight, growth, development and socioeconomic factors to adult BMD and skeletal geometry.
|
Acknowledgments
|
|---|
We would like to thank Miss Emma Scott, physics student at the University of Newcastle-upon-Tyne, for assistance and involvement in the phantom tests and the subsequent analysis of this data. We also thank the previous research teams involved in the Thousand Family Study and the study members for taking part in the present investigation. We are grateful to the Wellcome Trust, the Minnie Henderson Trust, the Sir John Knott Trust and the Special Trustees of the Newcastle Hospitals for supporting this work.
Received for publication December 12, 2003.
Revision received November 25, 2004.
Accepted for publication January 5, 2005.
|
References
|
|---|
- Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996;312:223–4.[Abstract/Free Full Text]
- Gomez Alonso CG, Diez Curiel M, Hawkins Carranza F, Perez Cano R, Diez Perez A. Femoral bone density, neck-shaft angle and mean femoral neck width as predictors of hip fracture in men and women. Osteoporosis Int 2000;11:714–20.[CrossRef][Medline]
- Gnudi S, Ripamonti C, Lisi L, Fini M, Giardino R, Giavaresi G. Proximal femur geometry to detect and distinguish femoral neck fractures from trochanteric fractures in postmenopausal women. Osteoporosis Int 2002;13:69–73.[CrossRef][Medline]
- Melton LJ, Khosla S, Achenbach SJ, O'Connor MK, O'Fallon WM, Riggs BL. Effects of body size and skeletal site on the estimated prevalence of osteoporosis in women and men. Osteoporosis Int 2000;11:977–83.[CrossRef][Medline]
- Seeman E. Sexual dimorphism in skeletal size, density and strength. J Clin Endocrinol Metab 2001;86:4576–84.[Free Full Text]
- Miller FJW, Court SDM, Knox EG, Brandon S. The School years in Newcastle upon Tyne 1952–1962. London: Oxford University Press, 1974.
- Lamont DW, Parker L, Cohen MA, White M, Bennett SMA, Unwin NC, Craft AW, et al. Early life and later determinants of adult disease: a 50 year follow-up study of the Newcastle thousand families cohort. Public Health 1998;112:85–93.[Medline]
- Lamont DW, Parker L, White M, Unwin N, Bennett SMA, Cohen M, et al. Risk of cardiovascular disease measured by carotid intima-media thickness at age 49-51: a life course study. BMJ 2000;320:273–8.[Abstract/Free Full Text]
- Parker L, Lamont DW, Unwin N, Pearce MS, Bennett SMA, Dickinson HO, et al. A lifecourse study of risk for hyperinsulinaemia, dyslipidaemia and obesity (the central metabolic syndrome) at age 49-51. Diabetic Med 2003;20:406–15.[CrossRef][Medline]
- Wright CM, Parker L, Lamont DW, Craft AW. Implications of childhood obesity for adult health: findings from the Newcastle thousand families cohort study. BMJ 2001;323:1280–4.[Abstract/Free Full Text]
- Spence J, Walton WS, Miller FJW, Court SD. A Thousand Families in Newcastle upon Tyne. London: Oxford University Press, 1954.
- Miller FJW, Court SDM, Walton WS, Knox EG. Growing up in Newcastle upon Tyne. London: Oxford University Press, 1960.
- Kolvin I, Miller FJW, Fells H. Continuities of deprivation: The Newcastle Thousand Families Study. Aldershot: Gower, 1990.
- Katzman DK, Bachrach LK, Carter DR, Marcus R. Clinical and anthropometric correlates of bone mineral acquisition in healthy adolescent girls. J Clin Endocrinol Metab 1991;73:1332–9.[Abstract]
- Faulkner KG, Cummings SR, Black D, Palermo L, Gluer CC, Genant HK. Simple Measurement of femoral geometry predicts hip fracture: the study of osteoporotic fractures. J Bone Miner Res 1993;8:1211–7.[Medline]
- Qureshi AM, McGuigan FEA, Seymoor DG, Hutchinson JD, Reid DM, Ralston SH. Association between COLIA1 Spi alleles and femoral neck geometry. Calci Tissue Int 2001;69:67–72.
- Duan Y, Turner CH, Bonn-Taeck K, Seeman E. Sexual dimorphism in vertebral fragility is more the result of gender differences in age-related bone gain than bone loss. J Bone Miner Res 2001;16:2267–75.[CrossRef][Medline]
- Duan Y, Seeman E, Turner CH. The biomechanical basis of vertebral fragility in men and women. J Bone Miner Res 2001;16:2276–83.[CrossRef][Medline]
- Karlsson KM, Sernbo I, Obrant KJ, Redlund-Johnell I, Johnell O. Femoral neck geometry and radiographic signs of osteoporosis as predictors of hip fracture. Bone 1996;18:327–30.[Medline]
- Nelson DA, Barondess DA, Hendrix SC, Beck TJ. Cross-sectional geometry, bone strength and bone mass in the proximal femur in black and white postmenopausal women. J Bone Miner Res 2000;15:1992–7.[CrossRef][Medline]
- Pheasant S. Bodyspace: anthropometry, ergonomics and design. London: Taylor & Francis, 1988.
- Looker AC, Beck TJ, Orwoll ES. Does body size account for gender differences in femur bone density and geometry? J Bone Miner Res 2001;16:1291–9.[CrossRef][Medline]
- Martyn CN, Hales CN, Barker DJP, Jesperson S. Fetal growth and hyperinsulinaemia in adult life. Diabet Med 1998;15:688–94.[CrossRef][Medline]
- Hales CN, Barker DJP, Clark PMS, Jesperson S. Fetal and infant growth and impaired glucose tolerance at age 64. Br Med J 1991;303:1019–22.
Top
Abstract
Introduction
