American Journal of Respiratory and Critical Care Medicine

The use of oral corticosteroids is associated with an increased risk of fracture, but there is limited information on the relationship between corticosteroid dose, bone mineral density (BMD), and fracture. We examined this relationship in a community population (more than 50 years) taking oral corticosteroids for chronic lung disease. Details of corticosteroid use and lifestyle were obtained by questionnaire, general practice records, and patient interview. BMD was assessed at the lumbar spine and femur and vertebral fracture by morphometric X-ray absorptiometry. Of the 117 patients who participated (median age, 69), 48% were female. Fifty-eight percent had osteoporosis (a T score of less than −2.5), and 61% had a vertebral fracture. The presence of vertebral fracture was related to BMD at the femoral neck, with an odds ratio of 1.6 for a 1 SD reduction in BMD. The cumulative prednisolone dose ranged from 3.4 to 175 g and was strongly associated with vertebral fracture, with the odds ratio between the highest and lowest dose quartiles being 4.4 (95% confidence interval, 1.04, 18.8). The difference in femoral neck BMD between the same dose quartiles was only modest, however (0.5 SD; 95% confidence interval, 0.09, 0.94). In patients taking long-term oral corticosteroids for chronic lung disease, the relationship between vertebral fracture risk and BMD is similar to that seen in other populations. Cumulative prednisolone dose is strongly related to fracture risk, and this effect is independent of its more modest impact on BMD.

Oral corticosteroid treatment is associated with osteoporosis and an increased risk of fracture (14), although these relationships have never been clearly quantified. Studies of involutional osteoporosis suggest that fracture risk doubles with each standard deviation reduction in bone mineral density (BMD) (5). It is not known whether the same is true for corticosteroid-induced osteoporosis. There is some evidence that fractures associated with the use of corticosteroid occur at a higher BMD than those caused by involutional osteoporosis (6), although this has been disputed (7).

With the advent of effective treatments to prevent fracture in patients with osteoporosis, there is a need to quantify the impact of oral corticosteroids on BMD and to assess the risk of fracture associated with these changes. To do this, we have identified subjects from an established cohort of patients taking oral corticosteroids for chronic lung disease and have collected data on BMD and vertebral fracture.

Subjects were selected from an established cohort of patients, described elsewhere (8). Patients had been recruited from local general practices if they were 50 years or older and had taken continuous or frequent intermittent courses of oral corticosteroids for the last 6 months, at least, for asthma, chronic obstructive pulmonary disease, or fibrosing alveolitis. Patients taking frequent intermittent courses of oral corticosteroids were only included if the mean daily dose of prednisolone over the previous 6 months was 5 mg or more (or an equivalent dose for other corticosteroids). Data on lifetime cigarette consumption, alcohol intake, calcium intake, age at menopause, exercise determined from regular sporting activity between ages 15 and 25 years and at present, and height at age 25 years were collected at recruitment by questionnaire (8).

Members of the original cohort were invited to participate in this study unless they had other reasons for fracture, for example, prolonged immobility, early menopause, or taking of drugs that are known to affect BMD (anabolic steroids, sodium fluoride tablets, calcitonin, bisphosphonates [more than 1 month]), calcium supplements (more than 500 mg per day for more than 6 months), vitamin D (more than 400 IU/day), or hormone replacement therapy (currently, more than 3 months in the past 10 years, or more than 2 years ever). Subjects attended hospital for measurement of height (stadiometer), weight, BMD of the lumbar spine and femur, and morphometric analysis of the dorsal and lumbar spine. Nottingham City Hospital's ethics committee gave ethics approval; subjects provided written informed consent.

Details of dose and duration of oral corticosteroid therapy were obtained for each patient from a questionnaire and were corroborated against the general practice record without knowledge of the patients' BMD. Lifetime cumulative doses of prednisolone and inhaled corticosteroids were calculated, plus duration of therapy, as the total time that continuous or intermittent corticosteroids had been taken. The three inhaled corticosteroids used by patients (beclomethasone dipropionate, budesonide, and fluticasone propionate) were considered to be equipotent; no allowance was made for different formulations, inhaler, or spacer devices.

BMD was measured at the anterior–posterior lumbar spine (L2–L4) and left femur (neck, Ward's triangle, and trochanter) by dual energy X-ray absorptiometry (Lunar Expert; Lunar Corporation, Madison, WI), using the same scanner for all patients. Lumbar spine BMD was computed from vertebrae that were unaffected by fracture or osteoarthritis using regression analysis (9). Daily calibration checks remained stable throughout the study. Lateral dorsal and lumbar spine morphometry was assessed with the patient supine, using X-ray absorptiometry. Vertebral fractures were defined as described by McCloskey and colleagues (10, 11) using local reference data from 437 women aged 45 to 59 years. BMD was expressed in absolute values, as a T score and Z score (i.e., the difference in SDs from mean reference BMD values matched for race and sex from peak bone mass [T score] or matched for age [Z score]).

Analysis

The relationship of cumulative oral corticosteroid dose (in quartiles) to BMD at the lumbar spine and proximal femur was estimated by multiple linear regression (SPSS Inc., Chicago, IL). Potential confounding variables (body mass index, calcium intake, daily activity, exercise now, and at age 15–25, years since menopause, smoking, and inhaled corticosteroid use) were introduced in turn to a baseline model comprising age (in quintiles) and sex and oral corticosteroid use (quartiles) and were retained if they changed the estimated effect of corticosteroid dose by more than 10%. The effect of cumulative inhaled corticosteroid dose on BMD was modeled in a similar way adjusting for oral corticosteroid dose.

The relationship of oral corticosteroid dose to vertebral fracture (absent verses present or more than one fracture) was explored using multiple logistic regression and a similar modeling approach; that is, we allowed for age and sex and other variables found to alter the relationship between corticosteroid use and fracture. The relationship of vertebral fracture to BMD was determined using multiple logistic regression. Finally, we assessed how far the effect of prednisolone dose on BMD at each site contributed to the relationship between cumulative prednisolone dose and vertebral fracture by augmenting the logistic model with the BMD variable (quartiles).

Of the 169 subjects from the original cohort (8) who fulfilled the entry criteria for the study, 117 (69%) agreed to participate (median age, 69 years; 48% female); 29 were unwilling, and 23 failed to respond. Demographic details are shown in Table 1

TABLE 1. Demographic data for 117 patients with chronic lung disease taking oral corticosteroids



Female (n = 56)

Male (n = 61)
Patient Characteristics
Median (SD or Range)
Age, yr70 (8)69 (10)
Height, cm157 (6)173 (8)
Weight, kg67 (16)73 (15)
Body mass index, kg·m−226 (6)25 (6)
Never smokers, %4112
Current cigarette smokers, %12.515
Cigarettes smoked, pack years12.5 (0.5–75)18 (2–100)
Calcium intake, mg/d679 (263)658 (242)
Daily activity, %
Light6872
Moderate3228
Exercise now, %
Minimal6655
Walking regularly1812
Exercise at least once a wk1633
Exercise between 15–25 yr, %
Minimal2415
Walked regularly1815
Exercise at least once a wk5870
Inhaled corticosteroid use, %8990
Cumulative dose, g3.1 (0–16)3.3 (0–15)
Duration, yr
8.5 (0–23)
9 (0–25)
. Patients had been given oral corticosteroids for asthma (50%), chronic obstructive pulmonary disease (14%), asthma plus chronic obstructive pulmonary disease (32%), or fibrosing alveolitis (5%).

Ninety-two patients were taking an oral corticosteroid regularly, and 25 had had frequent intermittent courses. The median cumulative dose of prednisolone was 21.4 g (interquartile range, 12.1 to 47.3), median duration of years (interquartile range, 3.6 to 15.2), and median daily dose of 7.8 mg (interquartile range, 5.9 to 10.0). Most patients (90%) were also taking an inhaled corticosteroid, either beclomethasone dipropionate (50%), budesonide (20%), or fluticasone propionate (20%). The median cumulative dose of inhaled corticosteroid was 3.3 g, with a median duration of 9 years and a median daily dose of 917 μg.

BMD and Vertebral Fracture

Mean BMD at the femoral neck, Ward's triangle, trochanter, and lumbar spine ranged from 0.6 to 1.02 g/cm2, and mean Z scores ranged from −0.03 to −0.6 (Table 2)

TABLE 2. Mean (sd) values for bone mineral density at each site expressed as absolute values and as a z score and t score*



BMD (th g/cm2)

Z Score

T Score
Site
Mean, SD
Femoral neck0.81, 0.16−0.4, 1.1−1.8, 1.2
Ward's triangle0.60, 0.14−0.6, 0.9−2.6, 1.0
Trochanter0.76, 0.2 −0.2, 1.3−0.95, 1.5
Lumbar spine
1.02, 0.18
−0.03, 1.5
−1.2, 1.8

*For definition, see text.

Excluding vertebrae with fracture or osteoarthritis.

Definition of abbreviation: BMD = bone mineral density.

. More women (64%) than men (46%) had a T score of less than −2.5 at any site, thereby fulfilling the current definition of osteoporosis (12) (Table 3)

TABLE 3. Prevalence of osteoporosis* and vertebral fracture by age and sex and the median prednisolone dose in each group


Age (yr)

Number of
 Patients
 (Men)

Median
 Prednisolone
 Dose (g)

Number with
 T Score <−2.5
 (Men)

Number with
 Vertebral Fracture
 (Men)
50–5917 (11)225 (2)9 (5)
60–6940 (21)2116 (8)21 (12)
70–7940 (19)2131 (13)26 (13)
80–86
15 (10)
32
12 (5)
12 (7)

*T score of less than −2.5.

. Age, sex, body mass index, current exercise, and number of years since menopause were related to BMD at all four sites in the univariate linear regression (Table 4)

TABLE 4. Effect of potential confounding factors on femoral neck bone mineral density*



Bone Mineral Density
Variable
Regression
 Coefficient
95% CI
Age, per quintile−0.03−0.054, −0.01
Sex, females versus males−0.13−0.18, −0.07
Body mass index, kg/m20.0110.006, 0.017
Calcium intake, mg/d0.0001−0.00004, 0.0002
Smoking, per pack year0.0002−0.001, 0.002
Time since menopause, yr−0.006−0.008, −0.004
Physical activity (Figure 1)
Daily activity, per category0.06−0.006, 0.12
Exercise now0.040.01, 0.08
Exercise at age 15–25 yr0.02−0.02, 0.06
Inhaled corticosteroids
Cumulative dose, per g−0.001−0.01, 0.008
Oral corticosteroids
Cumulative dose, per g−0.015−0.04, 0.015
Duration, per yr
−0.007
−0.03, 0.02

*Univariate linear regression.

Definition of abbreviation: CI = confidence interval.

.

A similar number of men and women (61%) had at least one vertebral fracture on morphometric assessment, and 39% had two or more fractures. For only 15% of these patients was a vertebral fracture recorded in the GP records or patient questionnaire. Having a vertebral fracture was related to BMD at the femoral neck such that a T-score reduction of one was associated with an odds ratio (OR) of 1.6 for vertebral fracture (95% confidence interval [CI], 1.03, 2.45). The findings at Ward's triangle were similar but not significant (OR, 1.54; 95% CI, 0.96, 2.48), whereas no relationship was seen at the trochanter (OR, 1.16; 95% CI, 0.83, 1.62) or lumbar spine (OR, 0.87; 95% CI, 0.68, 1.12). Vertebral fracture was also related to height loss. The mean ± SD reduction from self-reported height at the age 25 of years was 4.7 ± 3.7 cm for patients with a fracture compared with 3.1± 2.6 cm for those without (difference, 1.57; 95% CI, 0.25, 2.89).

Relationship of Oral and Inhaled Corticosteroid Dose to BMD

Cumulative prednisolone dose was related to BMD at the femoral neck and Ward's triangle. There was little difference in BMD between the first and second quartile but a progressive reduction with the two higher quartiles (Figure 1)

. A similar pattern was seen at the trochanter and lumbar spine, but these were not statistically significant (p = 0.1 and 0.06, respectively). The difference in mean BMD between the highest and lowest prednisolone dose quartile was 0.082 g/cm2 (95% CI, 0.014, 0.15) at the femoral neck and 0.067 (95% CI, 0.005, 0.13) at Ward's triangle; this is equivalent to a difference of 0.5 SD at both sites.

The effect of cumulative inhaled corticosteroid dose followed a similar pattern but was of borderline significance at all four sites (p = 0.05 to 0.08); the difference in mean BMD between the highest (mean dose of 8 g) and lowest dose (0.5 g) quartiles varied between 0.021 and 0.057 g/cm2 at the four sites.

Relationship of Corticosteroid Dose to Vertebral Fracture and BMD

The presence of vertebral fracture increased with increasing quartiles of cumulative prednisolone dose, from 48% in patients in the lowest dose quartile to 76% in the highest quartile (Table 5)

TABLE 5. Effect of potential confounding factors on vertebral fracture*



Vertebral Fracture

Variable
Range per Category
 (Mean/Median)
Percentage
 with Fracture
Odds Ratio
95% CI
Age, per quintile
151–61, (57)521.0
262–67, (65)541.060.34, 3.4
368–72, (70)601.360.42, 4.4
473–76, (74)621.480.43, 5.0
577–88, (82)793.410.84, 13.8
Sex (m = 1, f = 2)611.010.47, 2.15
Body mass index quartiles, kg/m2
116–23, (21)751.0
223–26, (24)620.550.17, 1.7
326–29, (27)440.270.09, 0.84
429–48, (33)590.480.15, 1.53
Calcium intake, mg/d
1735–1068, (850)621.0
2550–773, (645)540.730.29, 1.82
380–518, (305)661.180.46, 3.01
Smoking, pack years
1None671.0
2⩽ 20, (10)650.920.34, 2.45
3> 20, (36)500.500.18, 1.42
Time since menopause, yr
10611.0
21–21, (14)520.680.27, 1.75
322–39, (27)681.340.5, 3.59
Physical activity (see Table 1)
Daily activity, per category
1611.0
2590.920.4, 2.13
Exercise now
1581.0
2591.030.35, 3.02
3681.520.6, 3.84
Exercise at age 15–25 yr
1571.0
2671.500.41, 5.54
3601.110.42, 2.98
Inhaled corticosteroids
Cumulative dose quartiles, g
10–1, (1)571.0
21–3, (2)611.160.4, 3.36
33–5, (4)611.160.4, 3.36
45–16, (7)641.350.46, 3.96
Oral corticosteroids
Cumulative dose quartiles, g
13–12, (7)481.0
212–21, (15)541.240.43, 3.58
321–45, (30)641.940.66, 5.71
447–175, (61)763.381.09, 10.55
Total duration quartiles, yr
10–4, (3)441.0
24–7, (6)672.500.88, 7.14
37–15, (10)561.600.53, 4.76
4
15–46, (23)
74
3.57
1.13, 11.25

*Univariate logistic regression.

Definition of abbreviation: CI = confidence interval.

. In the baseline age- and sex-adjusted logistic analysis, the OR for having a vertebral fracture between the first and fourth cumulative oral corticosteroid dose quartiles was 3.0. In the fully adjusted model, the OR for the highest relative to the lowest dose quartile was 4.4 (95% CI, 1.04, 18.8), as shown in Figure 2 , with a significant linear trend in odds across increasing quartiles (p trend = 0.03). The OR between the highest and lowest dose quartiles for two or more fractures was 6.2 (95% CI, 1.13, 34.1). The addition of the variable for femoral neck BMD to the model had no impact on these ORs, whether fitted as a continuous variable or as quartiles (OR = 4.5 and 4.9 versus 4.4). The same was true when BMD at other sites was added to the model in a similar way (OR ranged from 4.1 to 5.5).

Inhaled corticosteroid dose was not significantly related to vertebral fracture; the OR between the highest and lowest dose quartile for having a vertebral fracture was 1.4 (95% CI, 0.34 to 6.1) and was 2.8 for two or more fractures (0.53 to 14.5).

Our cross-sectional survey of older patients requiring oral corticosteroids for lung disease is the first to look in detail at the relationship between cumulative prednisolone dose, BMD, and vertebral fracture. Vertebral fracture was common in this population and related both to cumulative prednisolone dose and, inversely, to BMD. A high cumulative prednisolone dose was also associated with a lower BMD, although the size of this effect was small, and it did not appear to explain the large increase in risk of vertebral fracture associated with oral corticosteroids. These data suggest that BMD does not adequately capture the adverse effect of oral corticosteroids on fracture risk.

The patients included in this study were part of a larger cohort of patients who have been carefully characterized in terms of their underlying disease, comorbidity, and exposure to oral corticosteroids. Our initial study described the dose–response relationship between cumulative exposure to oral corticosteroids and a number of adverse events, including self-reported fracture, cataract, and bruising (8). In this study, we related corticosteroid use to objective measures of vertebral fracture and BMD. We excluded patients with other known risk factors for osteoporosis and fracture and those receiving treatment known to affect bone who were also those more likely to be taking higher doses of oral corticosteroids. Despite the fact that our target population was older patients with moderately severe chest disease, our response rate was good at almost 70%. Over half our subjects had a diagnosis of osteoporosis according to the World Health Organization criteria (12), and 61% had a vertebral fracture, although fracture had only been diagnosed clinically in 15% of the patients. The mean loss of height from self-reported height at the age of 25 was greater in patients with a vertebral fracture.

Despite excluding patients taking treatment for osteoporosis, the prevalence of vertebral fracture in our patients was five times higher than the figure of 12% seen in the large cross-sectional population based European Vertebral Osteoporosis Study. The men and women in European Vertebral Osteoporosis Study were on average 5 years younger (range, 50 to 79 years), and the response rate in European Vertebral Osteoporosis Study was lower at 54% (13). This marked difference in fracture prevalence is likely to relate to the specific characteristics of the population that we studied, in particular their chronic illness and use of corticosteroids. A difference in sensitivity of the methods used to define fracture seems unlikely (14, 15), as the use of morphometric X-ray absorptiometry to measure vertebral dimensions agrees well with radiographic morphometry (15) and the definition of fracture we used (10, 11) was similar to that used in European Vertebral Osteoporosis Study except that we used local reference data. Vertebral fracture was also related to loss of height. It seems likely therefore that our high prevalence for fracture reflects the true prevalence of fractures in older patients taking oral corticosteroids for respiratory disease.

An inverse association between the use of oral corticosteroids and BMD has been noted before (16), but we were able to characterize the dose–response nature of the relationship. Although only significant at the femoral neck and Ward's triangle, the relationship was nonlinear at all four sites, with no difference in BMD between the first two prednisolone dose quartiles and a progressive reduction thereafter. The reason for the nonlinear relationship is not apparent, but it remained when patients taking intermittent corticosteroids were removed and was present in patients with asthma and in those with chronic obstructive pulmonary disease. Longitudinal studies have suggested that bone loss may be more rapid soon after starting treatment (17) and then continues to decline more slowly. The magnitude of the effect of prednisolone between the highest and lowest dose quartiles was small, and it was not statistically significant at the lumbar spine. This may be because we excluded fractured vertebrae from our estimates of BMD, and thus, the values derived will be from “healthier” vertebrae. Lumbar spine BMD may be less reliable therefore in populations in which vertebral fracture is common.

We found a progressive dose–response relationship between cumulative prednisolone dose and vertebral fracture with an OR of 4.4 between the highest and lowest dose quartiles. The findings were similar when duration of oral corticosteroid therapy was substituted for dose. The prevalence of vertebral fracture was high in all prednisolone dose quartiles, including the lowest, where 48% of the patients had a vertebral fracture. Because cumulative prednisolone dose was determined at study recruitment rather than at the time of fracture, our values for oral corticosteroid exposure will overestimate the dose at the time of fracture. The dose–response relationship with vertebral fracture may therefore underestimate the true relationship.

In our study, a T-score reduction of one at the femoral neck was associated with a 1.6-fold increase in vertebral fracture, and this value is broadly similar to that noted in studies of involutional osteoporosis (5). The effect of prednisolone on fracture risk was higher than expected from the prednisolone-related reduction in BMD, however. Subjects in the highest cumulative prednisolone dose quartile had a 4.4-fold increase in vertebral fracture compared with the lowest dose quartile but only a small (0.5 SD) reduction in BMD. This finding is in keeping with previous smaller studies (3, 4, 6) in which oral corticosteroids had little or no effect on BMD and has led to the suggestion that the main effect of corticosteroids relates to a disruption of bone architecture or collagen matrix rather than a reduction in bone mineralization (3, 6). Alternatively, corticosteroids might increase the risk of fracture through other mechanisms such as proximal myopathy (8).

Most patients in the study were taking an inhaled corticosteroid and often in high doses. Cumulative inhaled corticosteroid dose was associated with an independent reduction in BMD, which was of borderline statistical significance. The magnitude of the effect was in keeping with our previous study in young patients with asthma (18). The relationship between inhaled corticosteroid dose and vertebral fracture was not, however, significant.

Thus, in a population of older patients taking long-term oral corticosteroids for chronic chest disease, low BMD was a risk factor for fracture, and the magnitude of this relationship was similar to that seen in patients with involutional osteoporosis. The use of oral corticosteroids was associated with a large dose-dependent increase in vertebral fracture, which did not appear to be due to a reduction in BMD. When assessing the risk of fracture in such patients, cumulative oral corticosteroid dose is a strong risk factor independent of that of BMD.

The authors thank the following general practitioners for their help, as well as the practice managers and computer staff members associated with their practices: A. Hutton; J. S. McCracken; N. Quershi; J. Bilkhu; C. A. Brown; F. Coutts; I. W. L. McCulloch; J. D. Spencer; P. D. Sprackling; A. Khalique; K. G. Bratt; N. Gorbutt; M. E. Carr; G. Mansford; B. S. Mehat; N. A. Silcock; H. C. Barkataki; J. Ioannou; R. Kime; I. C. McNulty; T. Venables; S. Karim; O. P. Sharma; Y. U. S. Rao; K. F. Winterbottom; W. Holmes; G. D. R. Martin; E. Toms; N. V. Gould; S. J. Kingdon; H. M. Earwicker; J. A. Rudin; I. A. Dunn; R. H. Smith; P. J. Carberry; A. K. Tangri; M. Hepden; T. P. Connery; B. L. Parsons; G. J. Cox; A. L. Bridgewater; A. J. Avery; A. T. Harrison; A. J. Marsh; S. E. Annesley; D. Jenkinson; F. Badrashi; D. J. Mile; V. G. Golshetti; P. Smith; M. D. Barrett; P. Rose; R. R. Sheikh; G Stein; J. F. S. Eborall; J. E. Selwyn; V. A. Pollard; R. Nam; P. J. Enoch; C. Anderson; N. Foster; A. M. Cowe; and J. S. Ashcroft.

1. Lieberman P, Patterson R, Kunske R. Complications of long-term steroid therapy for asthma. J Allergy Clin Immunol 1972;49:329–336.
2. Adinoff AD, Hollister JR. Steroid induced fractures and bone loss in patients with asthma. N Engl J Med 1983;309:265–268.
3. Verstraeten A, Dequeker J. Vertebral and peripheral bone mineral content and fracture incidence in postmenopausal patients with rheumatoid arthritis: effect of low dose corticosteroids. Ann Rheum Dis 1986; 45:852–857.
4. Peel NFA, Moore DJ, Barrington NA, Bax DE, Eastell R. Risk of vertebral fracture and relationship to bone mineral density in treated rheumatoid arthritis. Ann Rheum Dis 1995;54:801–806.
5. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1995;312:1254–1259.
6. Luengo M, Picado C, Del Rio L, Guañabens N, Montserrat JM, Setoain J. Vertebral fractures in steroid dependant asthma and involutional osteoporosis: a comparative study. Thorax 1991;46:803–806.
7. Selby PL, Halsey JP, Adams KRH, Klimiuk P, Knight SM, Pal B, Stewart IM, Swinson DR. Corticosteroids do not alter the threshold for vertebral fracture. J Bone Miner Res 2000;15:952–956.
8. Walsh LJ, Wong CA, Oborne J, Cooper S, Lewis SA, Pringle M, Hubbard R, Tattersfield AE. Adverse effects of oral corticosteroids in relation to dose in patients with lung disease. Thorax 2001;56:279–284.
9. Altman DG. Practical statistics for medical research. London: Chapman and Hall; 1996. p. 299–303.
10. McCloskey EV, Spector TD, Eyres KS, Fern ED, O'Rourke N, Vasikaran S, Kanis JA. The assessment of vertebral deformity: a method for use in population studies and clinical trials. Osteoporos Int 1993;3:138–147.
11. McCloskey EV, Kanis JA. Assessing vertebral deformities [letter]. Osteoporos Int 1994;4:117–119.
12. Kanis JA, Melton J III, Christiansen Claus, Johnston CC, Khaltaev N. The diagnosis of osteoporosis. J Bone Miner Res 1994;9:1137–1141.
13. O'Neill TW, Felsenberg D, Varlow J, Varlow J, Cooper C, Kanis JA, Silman AJ. The prevalence of vertebral deformity in European men and women: the European Vertebral Osteoporosis Study. J Bone Miner Res 1996;11:1010–1018.
14. Black DM, Palermo L, Nevitt MC, Genant HK, Epstein R, San Valentin R, Cummings SR. Comparison of methods for defining prevalent vertebral deformities: the study of osteoporotic fractures. J Bone Miner Res 1995;10:890–902.
15. Lang T, Takada M, Gee R, Wu C, Li J, Hayashi-Clark C, Schoen S, March V, Genant HK. A preliminary evaluation of the Lunar Expert-XL for bone densitometry and vertebral morphometry. J Bone Miner Res 1997;12:136–143.
16. Reid IR, Heap SW. Determinants of vertebral mineral density in patients receiving long-term glucocorticoid therapy. Arch Intern Med 1990;150: 2545–2548.
17. Sambrook P, Birmingham J, Kempler S, Kelly P, Eberl S, Pocock N, Yeates M, Eisman J. Corticosteroid effects on proximal femur bone loss. J Bone Miner Res 1990;5:1211–1216.
18. Wong CA, Walsh LJ, Smith CJP, Wisniewski AF, Lewis SA, Hubbard R, Cawte S, Green DJ, Pringle M, Tattersfield AE. Inhaled corticosteroid use and bone mineral density in patients with asthma. Lancet 2000; 355:1399–1403.
Correspondence and requests for reprints should be addressed to Dr. Lesley J. Walsh, M.D., M.R.C.P., Consultant Physician, Sandwell and West Birmingham Hospitals, NHS Trust, Lyndon, West Bromwich, West Midlands, B71 4HJ, UK. E-mail:

Related

No related items
American Journal of Respiratory and Critical Care Medicine
166
5

Click to see any corrections or updates and to confirm this is the authentic version of record