Is there an association between inhaled corticosteroids and bone density in postmenopausal women?

Transcription

1 Is there an association between inhaled corticosteroids and bone density in postmenopausal women? Sölve Elmståhl, MD, PhD, a Henrik Ekström, MD, a Herbert Galvard, MD, a Olof Johnell, MD, PhD, b Maria Gerhardsson de Verdier, MD, PhD, c and Ensio Norjavaara, MD, PhD c,d Malmö, Lund, and Göteborg, Sweden Background: During the last decades, there has been increased concern about the association between oral corticosteroid (OC) therapy and osteoporosis. Objective: The question currently discussed is whether inhaled corticosteroids (ICs) in recommended doses have any clinically relevant effects on bone mineral density (BMD). Methods: We compared BMD in postmenopausal women exposed to corticosteroids only in inhaled form (IC group, n = 106) with that in women not exposed to corticosteroids (unexposed group, n = 674). BMD was also studied in 49 women exposed to OCs, intra-articular injections, or both in addition to ICs (OC group). The women were recruited from a population-based prospective cohort study. Methods: We used a dietary survey, bone density measurement of the forearm, and a health questionnaire including an interview about past and present medication use. Results: Mean BMD did not significantly differ between the IC group (0.434 g/cm 2 ) and the unexposed group (0.429 g/cm 2 ). The mean duration and dose of ICs was 8.2 ± 5.03 years and 853 µg daily. Within the IC group, BMD stratified for cumulative dose of IC, duration, or current dose greater than or less than 1000 µg did not differ. BMD in the OC group was lower than that in the IC group (0.408 vs g/cm 2 ). Conclusion: No difference in BMD was noted between the IC group and unexposed control subjects, nor was any doseresponse relationship observed between IC therapy and BMD. (J Allergy Clin Immunol 2003;111:91-6.) Key words: Bone mineral density, cohort study, inhaled corticosteroids, postmenopausal Treatment with systemic corticosteroids is known to increase the risk of osteoporotic fractures, 1 but the question is whether inhaled corticosteroids (ICs) in recommended doses have the same adverse effects on bone metabolism as oral medication. 2-4 The interest in this question has increased because ICs are used as first-line anti-inflammatory drugs for long periods, perhaps even for life, for some patients with asthma. The focus is on patients taking medium-to-high From a the Division of Geriatric Medicine, Department of Community Medicine, and b the Department of Orthopaedics, Lund University, Malmö; c AstraZeneca, Lund; and d Paediatric Growth Research Centre, Göteborg University, Göteborg. Received for publication April 12, 2002; revised September 9, 2002; accepted for publication September 25, Reprint requests: Sölve Elmståhl, MD, PhD, Department of Community Medicine, Malmö University Hospital, SE Malmö, Sweden Mosby, Inc. All rights reserved /2003 $ doi: /mai Abbreviations used BMD: Bone mineral density IC: Inhaled corticosteroid OC: Oral corticosteroid doses of ICs in high-risk populations, such as postmenopausal women. Studies on ICs and bone density in postmenopausal women have been difficult to interpret because of confounding by previous use of oral corticosteroids (OCs). 5-7 The evidence of an association between ICs and osteoporosis is still under debate, and several different results have been published. 2-5,8-10 Whether there is an increased risk of osteoporosis for some of the different ICs on the market is also debated on the basis of the increased awareness of the difference in systemic effect of different ICs. 11 The aim of this study was to compare bone mineral density (BMD) in postmenopausal women exposed only to ICs (IC group) with that in women with combined oral and inhaled exposure to corticosteroids (OC group) or healthy control subjects. METHODS The study comprised 155 postmenopausal women treated with ICs and 674 control subjects. The women were recruited from a populationbased prospective cohort study on diet and health comprising 8301 women examined between 1994 and The examination included a questionnaire, bone density measurement, and assessment of the diet. Women born between 1923 and 1950 and resident in Malmö, a city in the southern part of Sweden, were selected from the Municipality Registry and invited by letter. Comparisons between study participants and the population cohort are presented in Table I. During 1999, a second invitation including a questionnaire and an interview with a nurse was sent to all 173 postmenopausal women who had reported use of ICs for at least 1 month at the examination from 1994 to Eighteen women were excluded, 10 declined to participate, 6 were deceased, 1 had moved abroad, and the medical record of 1 woman was not retrievable, leaving 155 (89.6%) subjects. A random sample of 1080 postmenopausal female nonusers was taken from the study cohort. Of the control subjects, 943 (87.3%) responded, and 269 of these were excluded, 233 because of previous oral or parenteral corticosteroid intake and 36 who were postmenopausal for less than 1 year, leaving 674 control subjects. The ethical committee at Lund University approved the study, and all subjects provided written consent. The participants completed a health questionnaire covering marital status, education, ethnicity, lifestyle factors, medication use, and medical history. Marital status was divided into 4 categories: married, unmarried, divorced, or widowed. The level of education 91

2 92 Elmståhl et al J ALLERGY CLIN IMMUNOL JANUARY 2003 TABLE I. Comparison between study participants and the female population cohort* Study participants, Population cohort, % (n = 829) % (n = 8103) Year of birth Marital status Unmarried Married Divorced Widowed Ethnicity Swedish Non-Swedish *One hundred ninety-eight women not postmenopausal for the past year were excluded. included 5 categories depending on the length of the education. Smoking habits was categorized into current smokers (including occasional and regular smoking), nonsmokers, and former smokers. Physical activity at work was divided into 3 categories: light, medium, and heavy. Assessment of physical activity during leisure time included a question on mean time spent on 17 predefined activities and an open-ended question. The Minnesota LTA questionnaire was used to create an activity index taking into account differences in intensity between activities. 12 The activity index comprising the added sum of activities and intensity factors was computed and categorized into 3 levels: light, medium, and heavy. A subjective health score from 1 (bad) to 7 (very good) was requested. Medical history included questions on previous, including a history of asthma, allergy, and chronic bronchitis. Questions were asked concerning medication use, and data on hormonal replacement therapy and thiazide associated with BMD are presented separately. No subject in the control or patient group was prescribed bisphosphonates. Further information regarding corticosteroid use was requested in the second questionnaire, which included detailed questions on all known prescribed corticosteroid drugs with oral, inhaled, nasal, intra-articular, and dermal application. Questions were asked on the number of daily doses, mean monthly use during the prescribed years, and year of start, as well as the eventual year of change or end of medication use. Photographs of the different pharmaceutical brands previously and currently available on the market were shown to facilitate recognition and identification. The same registered nurse interviewed all participants. Informed consent to retrieve medical records from the general practitioner was given in all cases, and IC intake was verified in all cases through medical records. Subjects who confirmed previous OC use for more than 1 month, injections of corticosteroids, or both were categorized separately. For calculation of cumulative dose of ICs, see below. Permission to retrieve medical records was also requested from 46 (5%) of the control subjects, and 44 women agreed. Agreement between records and questionnaire was found in 43 (98%) of 44 control subjects. One control subject had a previous history of OC intake, and this prevalence would correspond to 2.3%, or 15 of all 674 control subjects. The diet assessment method used was a modified diet history method combining a 7-day menu book and a quantitative food frequency questionnaire for the past year. Studies of the validity and reproducibility of the method, with an 18-day weighed dietary record as a reference method, have shown correlations in the order of 0.5 to 0.7 for nutrients and 0.5 to 0.8 for foods. 13,14 BMD was measured with an x-ray absorptiometry technique. Scans were made with the nondominant arm placed in a water container. BMD in grams per square centimeter was automatically calculated from the distal forearm at a point at which the distance between the radius and ulna was 8 mm. Body composition was measured with a bioelectrical impedance analyzer (BIA-109, RJL Systems, Inc). The analysis relies on the principle of differences in electrical conductivity between the 2 compartments, total body water, and body fat. Validity and reliability studies have shown good correlation between this method and measurements of total body water, lean mass, and body fat, with correlation coefficients of between 0.79 and ,16 Systolic and diastolic blood pressure was measured with the subject in the supine position. Weight was measured without shoes and clothes to the nearest 0.2 kg. Differences between patients and control subjects were tested by using the Student t test unless stated otherwise. Association between duration of exposure and BMD adjusted for age was tested by means of a linear regression model. A cumulative dose was calculated by multiplying the dose and duration for the specific ICs. For patients with different doses, the added cumulative doses for the specific periods have been calculated. The exposed women were dichotomized according to duration of exposure to ICs. A 1-tailed t test was used for power calculation, assuming a risk with increased exposure. Z scores, which adjust for age, were calculated to standardize BMD. ANOVA was used to compare z scores. RESULTS The mean duration of use of ICs among the women was 8.2 ± 5.03 years (range, years; median, 6.3 years). Mean annual duration of inhalation varied from 8.5 to 12 months for the different IC brands. The mean daily dose for all women was 853 µg, and 23% had a daily intake of greater than 1000 µg. Budesonide was taken by 136 women, beclomethasone by 13 women, and fluticasone by 6 women. Forty-nine of the 155 women also reported previous oral intake or intra-articular injections of corticosteroids. Mean values of BMD did not differ significantly between the 155 women exposed to ICs (0.426 g/cm 2 ) and the 674 control subjects (0.429 g/cm 2,Table II). A similar mean BMD was noted among all women who used ICs for less than 6.3 years (0.427 g/cm 2 ) compared with the BMD of those who used ICs for greater than or equal to 6.3 years (0.426 g/cm 2 ; 6.3 years was chosen as the cutoff point for the duration of IC use because it is the sample median). No association was seen between BMD adjusted for age and annual duration of IC use among the 155 women or for the women exposed only to ICs (partial correlation, 0.02). BMD values are given separately for women exposed to corticosteroids only in the inhaled form (IC group) and women with both inhaled and oral/intra-articular intake of corticosteroids (OC group, Table III). BMD stratified for cumulative dose by tertiles, duration in years, or current dose of greater than or less than 1000 µg did not differ within either the IC group or the OC group. A numerically higher age was noted in the IC subgroup receiving a higher daily dose. Mean BMD in the OC group (n = 49) was lower than that in the IC group (n = 106, vs g/cm 2, P =.044; Table II). The groups did not differ regarding mean age (58.1 and 59.0 years, respectively).

3 J ALLERGY CLIN IMMUNOL VOLUME 111, NUMBER 1 Elmståhl et al 93 TABLE II. Values for BMD, medication, anthropometric, and obstetric variables between postmenopausal women exposed only to ICs (IC group) and women with a history of exposure to both OCs and ICs (OC group) Control subjects Total patients IC group OC group (n = 674) (n = 155) (n = 106) (n = 49) BMD (g/cm 2 ) ± ± ± ± Age (y) 57.6 ± ± ± ± 9.5 Weight (kg) 68.8 ± ± ± ± 12.9 Height (m) 1.64 ± ± ± ± 0.06 Lean weight (kg) 46.7 ± ± ± ± 5.7 Fat weight (kg) 21.2 ± ± ± ± 7.0 Body fat (%) 30.3 ± ± ± ± 5.1 BMI (kg/m 2 ) ± ± ± ± 4.93 Butt (cm) 98.0 ± ± ± ± 9.61 Waist (cm) 78.0 ± ± ± ± 13.8 Weight at age 20 y (kg) 55.8 ± ± ± ± 7.3 Diastolic blood pressure (mm Hg) 83.4 ± ± ± ± 8.7 Systolic blood pressure (mm Hg) ± ± ± ± 18.7 Thiazide treated (%) Subjective health (score 1-7) 5.4 ± ± ± ± 1.3 Estrogen replacement therapy (y) 10.3 ± 7.7 (49%) 10.9 ± 8.4 (49%) 10.6 ± 8.6 (48%) 11.6 ± 7.7 (51%) No. of children 1.8 ± ± ± ± 1.6 Values are given as means ± SD. Body composition was missing for 2 control subjects. Treatment with estrogen replacement therapy was reported by 328 control subjects, 51 patients in the IC group, and 25 patients in the OC group. BMI, Body mass index. TABLE III. BMD in postmenopausal women exposed only to ICs (IC group) and women with a previous history of exposure to OCs and ICs (OC group) IC group OC group n BMD (g/cm 2 ) n BMD (g/cm 2 ) P value adjusted for age Cumulative dose by tertiles <3996 mg/<4000 mg (IC/OC) ± ± Age 55.0 ± ± 9.5 Z score BMD 0.36 ± ± mg/ mg (IC/OC) ± ± Age 61.5 ± ± 9.1 Z score BMD 0.04 ± ± >4800 mg/>8000 mg (IC/OC) ± ± Age 58.2 ± ± 9.2 Z score BMD 0.13 ± ± Duration of exposure <6.3 y ± ± Age 60.1 ± ± 9.7 Z score BMD 0.04 ± ± >6.3 y ± ± Age 56.7 ± ± 8.7 Z score BMD 0.21 ± ± Current daily dose <1000 µg ± ± Age 57.3 ± ± 9.5 Z score BMD 0.16 ± ± µg ± ± Age 61.5 ± ± 9.7 Z score BMD 0.25 ± ± Values are given as means ± SDs, z scores, and comparisons of BMD z scores between patients in the IC and OC groups (ANOVA). BMD, expressed as z score and adjusted for age, was lower in the OC group than in the IC groups for women with long duration of exposure, current daily dose of less than 1000 µg, and cumulative dose within the lowest tertile (Table III). Budesonide was prescribed to 136 women, 91 in the IC group and 45 in the OC group, and corresponding BMD values were g/cm 2 and g/cm 2, respectively (P =.72). The z scores are shown in Fig 1. As expected, higher frequencies of allergy, asthma, and respiratory symptoms were reported by patients than by control subjects; 90% of all patients reported lifetime respiratory symptoms of dyspnea, and 61% reported a prolonged period of coughing compared with 7% and 13%, respectively, among the control subjects. Lower prevalences of inflammatory bowel disease, intermittent claudication, and diabetes were noted among control subjects without IC or OC exposure. However, as anticipated, these prevalences between patients and control subjects did not differ when analyses also included the 233 control subjects excluded because of previous corticosteroid intake.

4 94 Elmståhl et al J ALLERGY CLIN IMMUNOL JANUARY 2003 FIG 1. BMD of the forearm expressed as z score (BMD z score) for postmenopausal women exposed only to inhaled budesonide (IC group) and women exposed to OCs and inhaled budesonide (OC group) and control subjects. Body weight, lean weight, and reported weight at 20 years were similar between groups (Table II). No differences were noted between patients and control subjects for known risk factors that could influence BMD, such as smoking habits, alcohol intake, physical activity, medication with diuretics, and dietary variables, such as vitamin D, calcium, and zinc (Tables II and IV). Similar distributions between groups were noted for sociodemographic variables, such as marital status, education, and ethnicity. A nonsignificant but numerically lower BMD value was noted in the IC group with a daily dose of greater than 1000 µg compared with that of women with an intake of less than 1000 µg. When analysis was stratified for physical activity (light activity vs others), the BMD among women with light activity was g/cm 2 compared with g/cm 2 among those with medium to heavy physical activity. Corresponding values in women with a daily IC intake of less than 1000 µg were g/cm 2 and g/cm 2,respectively. The proportion of women with low physical activity was 42% among women with a daily dose of greater than 1000 µg compared with 26% among women with a daily intake of less than 1000 µg. DISCUSSION This population-based study was designed to investigate the associations between IC therapy and BMD. There were no differences in BMD values between women exposed to ICs and control subjects, nor was there any dose-response relationship between IC therapy and BMD. The population was restricted to postmenopausal women, a group that are at high risk of osteoporotic fractures. The annual loss of bone increases with greater variability in bone density during the perimenopausal period and up to 5 years after menopause. On the other hand, the gonadal steroids might counteract glucocorticoid effects on bone. Thus the optimal study design to investigate a possible effect of ICs on BMD would be a prospective study on a population with small annual changes in bone density with and without gonadal steroids. Studying a postmenopausal population makes it possible to investigate whether use of ICs during the premenopausal period, the perimenopausal period, or both affects bone density. The analyses of ICs and BMD by duration does not indicate that there are any differences between different windows of exposure (Table III). The present study had 80% power to detect a difference of 0.03 g/cm 2 of BMD for comparison of exposure within the IC group, a power sufficient to detect changes that might be of clinical relevance. Contrary to other methods, SXA of the forearm might be assessed in large population-based studies and seems to be less influenced by weight and physical activity. Both dual-energy x-ray absorptiometry and single-photon absorptiometry have an accuracy error of 3% to 5% and a precision error of 1%, and BMD measurements at appendicular sites seem to be as effective as measurements of the central skeleton in identifying patients with low bone mass. 17 Furthermore, forearm BMD is a good predictor of fractures at any site in female 18 and male 19 subjects. Oral glucocorticoid treatment has been shown to decrease BMD of trabecular bone (spine) more rapidly and severely than cortical bone BMD. 20,21 Because the distal site of the forearm consists mainly of cortical bone (75%), the glucocorticoid effects on bone in the distal forearm require a longer follow-up time. 17 In the present study the duration of use of inhaled glucocorticoids varied from 1.5 to 26 years. As expected, the results showed that women treated with OCs had lower BMD than women exposed only to ICs, a finding in accordance with those of previous studies that demonstrated an effect of OCs primarily on trabecular bone. 22

5 J ALLERGY CLIN IMMUNOL VOLUME 111, NUMBER 1 Elmståhl et al 95 TABLE IV. Comparison of sociodemographic and lifestyle variables between postmenopausal women exposed only to ICs (IC group) and women with a history of exposure to both OCs and ICs (OC group) Control subjects Total patients IC group OC group (n = 674) (n = 155) (n = 106) (n = 49) Marital status (%) Married 61.8 (n = 417) 53.5 (n = 83) 58.5 (n = 62) 42.9 (n = 21) Unmarried 8.6 (n = 58) 8.4 (n = 13) 7.5 (n = 8) 10.2 (n = 5) Divorced 18.1 (n = 122) 25.8 (n = 40) 20.8 (n = 22) 36.7 (n = 18) Widowed 11.5 (n = 77) 12.3 (n = 19) 13.2 (n = 14) 10.2 (n = 5) Ethnicity (%) Swedish 89.0 (n = 601) 86.5 (n = 134) 87.7 (n = 93) 83.7 (n = 41) Education (%) Elementary school not completed 0.6 (n = 4) 0.0 (n = 0) 0.0 (n = 0) 0.0 (n = 0) Elementary school 64.8 (n = 437) 67.7 (n = 105) 58.5 (n = 62) 67.3 (n = 33) Secondary school 7.3 (n = 49) 1.2 (n = 19) 13.2 (n = 14) 10.2 (n = 5) 1 y extra or university degree 27.0 (n = 182) 20.0 (n = 31) 18.9 (n = 20) 22.5 (n = 11) Smoking habits (%) Current smoker 24.2 (n = 163) 23.2 (n = 36) 22.7 (n = 24) 24.5 (n = 12) Former smoker 27.0 (n = 182) 35.5 (n = 55) 34.9 (n = 37) 36.7 (n = 18) Nonsmoker 48.7 (n = 329) 41.3 (n = 64) 42.5 (n = 45) 38.8 (n = 19) Smoking duration (y) 23.0 (n = 345) 25.2 (n = 91) 25.2 (n = 61) 25.2 (n = 30) Alcohol habits (%) Teetotaler 10.4 (n = 70) 12.3 (n = 19) 11.3 (n = 12) 14.3 (n = 7) Intake last year 12.0 (n = 81) 12.9 (n = 20) 13.2 (n = 14) 12.2 (n = 6) Intake last month 77.6 (n = 522) 74.8 (n = 116) 75.5 (n = 80) 73.5 (n = 36) Mean alcohol consumption (g) 8.2 (n = 673) 9.3 (n = 155) 9.1 (n = 106) 9.7 (n = 49) Coffee (g) (n = 662) (n = 152) (n = 105) (n = 47) Tea (g) (n = 662) (n = 152) (n = 105) (n = 47) Vitamin D (µg) 6.28 (n = 662) 6.17 (n = 152) 6.01 (n = 105) 6.54 (n = 47) Calcium (mg) 1098 (n = 662) 1072 (n = 152) 1052 (n = 105) 1116 (n = 47) Zinc (mg) 9.9 (n = 662) 9.9 (n = 152) 9.8 (n = 105) 9.9 (n = 47) Physical activity at work (%) Light 56.7 (n = 383) 54.2 (n = 84) 51.9 (n = 55) 59.2 (n = 29) Medium 28.7 (n = 194) 30.3 (n = 47) 33.0 (n = 35) 24.5 (n = 12) Heavy 12.0 (n = 81) 12.9 (n = 20) 12.2 (n = 13) 14.3 (n = 7) Physical activity at leisure time Light 30.8 (n = 208) 34.2 (n = 53) 29.2 (n = 31) 44.9 (n = 22) Medium 33.3 (n = 225) 27.7 (n = 43) 32.1 (n = 34) 18.4 (n = 9) Heavy 29.9 (n = 202) 31.6 (n = 49) 32.1 (n = 34) 30.6 (n = 15) Information retrieved from the questionnaire and from the medical records were discussed during an interview with each woman to avoid misclassification of corticosteroid use. Because ICs were introduced to the Swedish market in the first half of the 1980s, the mean duration of use of ICs of 8.2 years among these women would be expected, as well as the mean annual duration of inhalation of 8.5 to 12 months, which probably reflects that the women have fewer symptoms during the summer period. The main findings in the study could be limited by the relatively short duration of exposure of the study group, but no difference in BMD was noted between the 2 windows of exposure. Important confounders, such as smoking, physical activity, body composition, dietary habits, alcohol consumption, medication, and ethnicity, did not differ between groups exposed and unexposed to ICs. Information on cumulative exposure to smoking and estrogen were lacking in this study, but this would also introduce possible recall bias of the included confounders. Any selection bias seems unlikely because of the high response rate of between 87% and 90% and the fact that the investigated cohort had demographic characteristics similar to those of the background population with respect to distribution of age, marital status, and ethnicity. Previous studies have provided somewhat conflicting results regarding the association between ICs and BMD. 23,24 Studies on high-dose IC and bone loss have not fully corrected for previous OC intake. Results similar to those seen in our study were reported in a study of postmenopausal women taking ICs for 4 years, showing no difference in lumbar BMD between asthmatic and nonasthmatic subjects. 25 Association between doses of ICs of greater than 1 mg and lower values of BMD has been found in a study of 43 postmenopausal women by Toogood et al. 4 However, the same study noted a contradictory association between higher BMD and cumulative IC dose, and therefore different mechanisms of effect of IC therapy on BMD were hypothesized. Another alternative explanation to those findings could be differential misclassification. Recent studies on premenopausal women have reported associations between cumulative dose of ICs and BMD in the lumbar spine, total hip, and trochanter of the femur It is difficult to exclude the effect of the disease per se on BMD and whether the reports on lower BMD in women with higher inhaled doses is the result of less physical activity among more severely diseased women. Our findings indicated that low physical activity seemed to be more common in women

6 96 Elmståhl et al J ALLERGY CLIN IMMUNOL JANUARY 2003 with current high daily doses of ICs. The Cochrane Database of Systematic Reviews has recently published an article on ICs and bone metabolism. 29 The pooled results from the 3 studies showed no significant effect of inhaled steroids on BMD in patients with asthma or chronic obstructive pulmonary disease, as measured by using dual-energy x-ray absorptiometry at the lumbar spine and femoral neck. Two studies that collected fracture data prospectively showed no significant effect of ICs on vertebral fractures at conventional therapeutic doses. 30,32 A large retrospective study on the General Practice Research Database in the United Kingdom also investigated the use of ICs and the risk of fracture. 33 The study comprised 170,818 IC users, 108,786 bronchodilator users, and 170,818 control subjects. The findings suggested that low-dose IC therapy was not associated with an increased risk of fracture; however, patients with chronic respiratory who use inhaled therapy appeared to be at increased risk when compared with a population control group. The study sample in this study was small, and numerically lower values of BMD were noted among women with the highest cumulative dose and daily dose, which might suggest a trend with increasing exposure. However, the power of this study allows for an identification of a BMD within a group difference of 0.03 g/cm 2 with a 0.80 power and a significance level of.05 for cumulative dose and.04 for current daily dose, and no significant differences were noted within the IC groups. Corresponding power for the IC group patient and control comparison was 0.05 g/cm 2. Women with a higher current daily dose had numerically higher age and comprised a higher proportion of subjects with low physical activity (42% vs 26%) compared with those with a lower current daily dose, and this could explain, in part, the noted differences in BMD. This cross-sectional study will not allow for interpretation of confounding effects, and longitudinal studies are required to confirm or reject the main findings. In conclusion, moderate doses of ICs seem to carry less risk than traditional OC therapy provided that the dose is kept at the lowest daily dose sufficient to maintain optimum control of the disease. REFERENCES 1. Van Staa TP, Leeufkens HGM, Abenhaim B, Zhang B, Cooper C. Use of oral corticosteroids and risk of fractures. J Bone Miner Res 2000;15: Saag KG, Emkey R, Schnitzer TJ, et al. Bone loss and glucocorticoid therapy in patients with respiratory disease. Thorax 1999;54(suppl 2): Maxwell DL. Adverse effects of inhaled corticosteroids. Biomed Parmacother 1990;44: Toogood JH, Baskerville JC, Markov AE, et al. Bone mineral density and the risk of fracture in patients receiving long-term inhaled steroid therapy for asthma. J Allergy Clin Immunol 1995;96: Toogood JH. Side effects of inhaled corticosteroids. J Allergy Clin Immunol 1998;102: Jackson LD, Ploygenis D, McIvor RA, Worington I. Comparative efficacy and safety of inhaled corticosteroids in asthma. Can J Clin Pharmacol 1999;6: Cave A, Arlett P, Lee E. Inhaled and nasal corticosteroids: factors affecting the risk of systemic adverse effects. Pharmacotherapy 1999;83: Meeran K, Hattersley A, Burrin J, Shiner R, Ibbertson K. Oral and inhaled corticosteroids reduce bone formation as shown by plasma osteocalcin levels. Am J Respir Crit Care 1995;151: Storms WW. Clinical adverse effects of inhaled corticosteroids: results of a questionnaire survey of asthma specialists. Ann Allergy Asthma Immunol 1998;80: Boyd G. Effect of inhaled corticosteroids on bone. Respir Med 1994;88: Lipworth B. Systematic adverse effects of inhaled corticosteroid therapy. Arch Intern Med 1999;159: Taylor HL, Jacobs D, Schuker B, Knudsen J, Leon A, Debacker G. A questionnaire for the assessment of leisure time physical activities. J Chron Dis 1978;31: Elmståhl S, Riboli E, Lindgärde F, Gullberg B, Saracci R. The Malmö Food Study. The relative validity of a modified diet history method and an extensive food frequency questionnaire for measuring food intake. Eur J Clin Nutr 1996;50: Elmståhl S, Gullberg B, Riboli E, Saracci R, Lindgärde F. The reproducibility of a novel diet history method and an extensive food frequency questionnaire. The Malmö Food Study. Eur J Clin Nutr 1996;50: Lukaski HC, Johnson PE, Bolonchuk W, Lykken GE. Assessment of fat free mass using bioelectrical impedance measurements of the human body. Am J Clin Nutr 1985;41: Steen B, Boseaus I, Elmståhl S, Galvard H, Isaksson B, Robertsson E. Body composition in the elderly estimated with an electrical impedance method. Comprehensive Gerontology 1986;1A: Augat P, Fuerst T, Genant HK. Quantitative bone mineral assessment at the forearm: a review. Osteoporos Int 1998;8: Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996;312: Melton LJ, Atkinson EJ, O Connor MK, O Fallon WM, Riggs BL. Bone density and fracture risk in men. J Bone Miner Res 1998;13: Lane NE, Lukert B. The science and therapy of glucocorticoid-induced bone loss. Endocrinol Metab Clin North Am 1998;27: Van Staa TP, Leufkens HG, Abenhaim L, Zhang B, Cooper C. Use of oral corticosteroids and risk of fractures. J Bone Miner Res 2000;15: Ebeling P, Erbas B, Hopper J, Wark J, Rubinfeld A. Bone mineral density and bone turnover in asthmatics treated with long-term inhaled or oral glucocorticosteroids. J Bone Miner Res 1998;3: Wong CA, Subakumar G, Casey PM. Effects of asthma and asthma therapies on bone mineral density. Curr Opin Pulm Med 2002;8: Bonay M, Bancal C, Crestani B. Benefits and risks of inhaled corticosteroids in chronic obstructive pulmonary disease. Drug Safety 2002;25: Luengo M, del Rio L, Ponsand F, Picado C. Bone mineral density in asthmatic patients treated with inhaled corticosteroids: a case-control study. Eur Respir J 1997;10: Ip M, Lam K, Yam L, Kung A, Ng M. Decreased bone mineral density in premenopausal asthma patients receiving long-term inhaled steroids. Chest 1994;105: Wong CA, Walsh LJ, Smith CJ, et al. inhaled corticosteroid use and bonemineral density in patients with asthma. Lancet 2000;355: Israel E, Banerjee TR, Fitzmaurice GM, Kotlov TV, Lahive K, Leboff MS. Effects of inhaled glucocorticoids on bone density in premenopausal women. N Engl J Med 2001;345: Jones A, Fay JK, Burr M, Stone M, Hood K, Roberts G. Inhaled corticosteroid effects on bone metabolism in asthma and mild chronic obstructive pulmonary disease. Cochrane Database Syst Rev 2002;(1): CD Pauwels RA, Lofdahl C-G, Laitinen LA, et al. Long-term treatment with inhaled budesonide in persons with mild chronic obstructive pulmonary disease who continue smoking. N Engl J Med 1999;340: Li JTC, Ford LB, Chervinsky P, et al. Fluticasone propionate powder and lack of clinically significant effects on hypothalmic-pituitary-adrenal axis and bone mineral density over 2 years in adults with mild asthma. J Allergy Clin Immunol 1999;103: Tattersfield AE, Town GI, Johnell O, et al. Bone mineral density in subjects with mild asthma randomised to inhaled corticosteroids or non-corticosteroid treatment for two years. Thorax 2001;56: Van Staa TP, Leufkens HGM, Cooper C. Use of inhaled corticosteroids and risk of fractures. J Bone Miner Res 2001;16:581-8.