4.2 MEASUREMENTS IN NORMAL AND OSTEOPENIC WOMEN
4.2.3. T Summary Points
4.2.4.5 Fixed versus, anatomical ROIs
The correlation coefficient of ultrasound parameters and BMD of the calcaneus was not significantly improved by matching the measurement site in the calcaneus. Also the presence o f cortical edge in the ultrasound scan path did not show a clear influence on results. Brooke-Wavell et al (1995) carried out a similar study and measured BUA and BMD at a fixed and an anatomical location. They found that the correlation between BUA and BMD in the calcaneus was significantly better at the fixed ROI
C hapter Four____________________________________ In-vivo Investigations with DXA and Ultrasound
(r=0.50 to r=0,74) compared to the anatomical ROI (r=0.74) with a p-value of <0.01. Better correlations were also observed between BUA and axial BMD at the fixed ROI. They used a CUBA system which does not have fixed transducers.
4.2,4,6 Patterns and rates o f bone loss
Spine and femoral neck BMD showed very similar changes with age with an overall annual decrease in postmenopausal women of 0.9 %. The rate o f loss in postmenopausal women was higher for the whole calcaneus at 1.2 %. Others who have compared rates of loss from cross-sectional data (Table 4.13) generally confirm the results from this study of a higher rate of loss in the calcaneus compared to axial sites. Ensrud et al (1995) report an annual rate of loss o f -1.43 % in the calcaneus of over 5000 women aged over 65 years compared to a loss rate of -0.51 % in the hip. Steiger et al (1992) found a higher rate of loss in the calcaneus than in spine or hip as did Vogel (1987) for women with and without vertebral fractures, and Yamada et al (1994) in postmenopausal women with the spine BMD measured by QCT. Houde et al (1994) found similarly high annual loss rates of over 3 % for both calcaneus and spine while Kotzki et al (1993) found no difference in rates o f for calcaneus and spine in 193 female controls. Their figures (-0.5 % for premenopausal women and -1.1 % for postmenopausal women) are in very close accord with the results from the current study for calcaneus BMD.
From peak bone mass to the eighth decade the largest decrease in BMD was observed for the calcaneus in this study. Steiger et al (1992) found that compared to women aged 65-69 years the bone density of women over age 85 years was 21.1 % lower in the calcaneus. Only Ward’s triangle showed a higher decrement (25.8 %). Truscott et al (1993) found very similar results to those in the current study and observed a decrease o f 27 % in spine and femoral neck BMD from the age of peak bone mass to 80 years in 329 pre and postmenopausal women aged 20-81 years.
The two major bone compartments, cortical and trabecular, behave differently according to age and sex. A large amount of data has been gathered on the BMC of cortical bone and it is generally accepted (Vogel et al 1988) that cortical bone loss
accelerates in women at menopause, and decreases 10-15 years later. The onset of trabecular bone loss in women is not generally agreed upon. Some studies show onset o f spinal bone loss well before menopause (Riggs et al 1981; Riggs et al 1986) while others found that it did not begin until the menopause (Cann et al 1980; Gordon and Genant 1985).
Hoshi et al (1993) measured the decreasing rate per year of BMD of the calcaneus measured by SXA. The rate increased from 0.07 % in the twenties to 2.5 % in the fifties before decreasing again. The vertebral BMD in Japanese women began to decrease at age 35 while that o f the calcaneus started decreasing 5 years earlier. These decreases seem to start earlier than generally reported for European and American women. In contrast Yamada et al (1993) found that, in women, loss of bone starts or is greatly increased at the time of menopause. Both calcaneal and lumbar spine BMD of the postmenopausal women aged 50-54 years were 14.3 % below the mean for young normal premenopausal women. However the pattern of bone loss was then different for the 2 sites as calcaneal BMD continued to decrease throughout the postmenopausal period to 34.9 % below young normal while spine BMD did not show further decrease in subjects more than 70 years old. The earlier decrease in calcaneal BMD observed by Hoshi et al (1993) and the continuation o f this decrease throughout the postmenopausal period reported by Yamada et al (1993) may be due to the fact that 30-40 % of the bone substance in the lumbar spine and 60-70 % of the bone substance in the femoral neck consists of cortical bone (Vogel 1987). The remodelling of cortical bone is known to be metabolically less active than that of trabecular bone (Tonna 1977). Therefore measuring an integral of cortical and trabecular bone may make DXA of the femoral neck or spine less sensitive to early disease changes that preferentially affect trabecular bone. However, the calcaneus is made up almost entirely of trabecular bone (Vogel 1987) and may provide a more sensitive measurement site for the assessment of bone mineral density by DXA.
As well as the onset of bone loss there are further controversies about the rate of bone loss in elderly women. From Figure 4.2 there is little evidence of an increase in the rate o f bone loss except possibly for the calcaneus. In contrast the graph for femoral
Chapter Four____________________________________ In-vivo Investigations with DXA and U ltrasound
neck BMD could be interpreted as a plateau from the age o f 56-60 years onwards. Ensrud et al (1995) measured the rates of change in bone density at the hip and calcaneus in a cohort of older women participating in the study of osteoporotic fractures. They found that the rate of bone loss increased significantly with age at the femoral neck and the calcaneus. The increase from age 67-69 years to over age 85 years was 0.35 % per year to 1.00 % per year for the femoral neck and 1.20 to 2.68 % per year for the calcaneus. The results of Ensrud et al (1995) disagree with results of previous cross-sectional studies (Riggs et al 1982; Hannan et al 1992; Mazess et al
1987) that have suggested that the rate of bone loss with ageing remains relatively constant. Cheng et al (1994) found that calcaneal BMD in 80 year old women was significantly lower compared with 75 year old women. The difference persisted even when controlled for body mass (p<0.0001), indicating that bone loss does not reach a plateau in elderly women.
While the rate o f bone loss with age at a particular bone site is important this must be related to the measurement precision at that site. For instance bone loss is often found to be highest at the Wards triangle region of the hip (Steiger et al 1992) but the measurement precision of this region is low. Fouilles et al (1993) showed that the ratio of rate of change per year in 85 healthy postmenopausal women to the precision of the technique was 1.6 for spine, 1.06 for femoral neck and 0.77 and 0.73 for Weirds triangle and trochanter respectively. The calcaneus with its higher rate o f bone loss and good measurement precision provides a suitable measurement site.