SEARCH

SEARCH BY CITATION

Keywords:

  • corticosteroid;
  • epidemiology;
  • fracture;
  • inhaled;
  • topical

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

Objectives.  To study the fracture risk associated with the use of corticosteroids in any formulation and administration.

Design.  Case–control study.

Setting.  Community-based study in Denmark.

Subjects.  Cases were all subjects with any fracture sustained during the year 2000 (n = 124 655). For each case, three controls (n = 373 962) matched on age and gender were randomly drawn from the background population. Adjustments were made for concurrent diseases (lung diseases, rheumatic disorders), use of other drugs (inhaled bronchodilators), contacts to hospitals and general practitioners, and social variables.

Results.  An increased risk of any fracture, hip, spine, and forearm fractures was present with use of more than 2.5 mg of prednisolone or equivalent orally per day. For inhaled corticosteroids a limited increase in the risk of any fracture was present in users of more than 7.5 mg prednisolone equivalents per day (equivalent to 1875 μg of budesonide per day). However, no increase in the risk of hip, spine or forearm fractures was present in users of inhaled corticosteroids. For other topical corticosteroids (dermal, rectal, nasal, local application in the mouth, the eyes or the ears) no increase in fracture risk could be demonstrated even at high doses after adjustment for confounders.

Conclusions.  Ingestion of more than 2.5 mg of oral prednisolone equivalents per day is associated with an increase in fracture risk. No increase is associated with inhaled corticosteroids except at daily dosages above 7.5 mg of prednisolone equivalents. No increase in fracture risk is associated with other forms of topical corticosteroids.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

Corticosteroid use is detrimental to bone mass and metabolism through a number of mechanisms spanning from inhibition of sex steroid synthesis and secretion over impaired calcium absorption from the gut, and increased loss in the kidneys to effects on the osteoblasts and osteoclasts (bone turnover), the insulin-like growth factor system and collagen [1, 2]. This leads to an increased risk of fractures [3–6]. Indeed, a large proportion of fractures in users of corticosteroids (47% of all hip and 72% of all vertebral fractures) is attributable to the use of corticosteroids [7]. Corticosteroids are used widely in many conditions, meaning that a substantial number of people are exposed to these drugs [8]. Any fracture risk associated with the use of corticosteroids may thus have significant consequences on a population level.

Corticosteroids may be used in several formulations and preparations (oral, injected, dermal, rectal). An increased fracture risk has been shown for oral [3] and inhaled corticosteroids [4, 9], as well as for topical corticosteroids in general [10], whilst no increase has been present for nasal corticosteroids [6]. However, the disease for which the corticosteroids are given may per se contribute to fracture risk [11] (confounding by indication), and the more severe the disease is, the more it may increase fracture risk. At the same time the severity of the disease may also influence the amounts of corticosteroids needed to control the disease. A dose–response relationship between fracture risk and corticosteroid use may thus be misinterpreted as an effect of the drug instead of the disease. Recent data have suggested that this may be the case for inhaled corticosteroids [12].

Apart from lung diseases, many other diseases and conditions (rheumatic disorders, skin disorders, allergy, haematological disorders, inflammatory bowel disease, organ transplantation) may require treatment with corticosteroids administered systemically (orally or injected) or topically (dermal, inhaled, rectal, locally in joints, in the eyes, ears). These disorders may also per se alter fracture risk by various mechanisms.

The different forms of administration also represent differences in absorption and first-pass metabolism in the liver. After oral administration more than 80% of corticosteroids may undergo first pass metabolism limiting the systemic bioavailability [13].

Topically administered corticosteroids do not undergo first-pass metabolism, but the absorption may be very low. For rectally administered corticosteroids 3.1% are systemically available in healthy volunteers and 4.5% in patients [14] probably due to the inflammation in the rectum in patients with inflammatory bowel disease. For nasal corticosteroids a bioavailability of 0.4% has been reported [15], and for dermal preparations <0.3% are absorbed into systemic circulation [16]. For inhaled corticosteroids, systemic bioavailability is much higher (>15%) [17, 18]. However, healthy controls may absorb inhaled corticosteroids better than patients with lung diseases [19]. For inhaled corticosteroids, different methods of inhalation, different devices used for inhalation, and different operation of inhalation devices may all influence the amount of corticosteroids delivered in the lungs, and the amount going into the gastrointestinal tract. Furthermore, the absorption from the gastrointestinal tract and the degree of first-pass metabolism may vary.

For inhaled corticosteroids, a systemic effect on bone turnover has been demonstrated [20]. However, the effects on bone mineral density of inhaled corticosteroids are small [21] in contrast to oral corticosteroids where a rapid loss of bone [22], and a rapid increase in fracture risk is seen [3]. Inhaled corticosteroids may also have effects on growth and thus skeletal development and maturation in children, although this is under debate [18].

Study design

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

The study was designed as a case–control study. All subjects sustaining a fracture during the year 2000 in Denmark were included as cases (n = 124 655), and for each case three subjects of the same age (same birth year) and gender were randomly selected from the background population as controls (n = 373 962).

We examined the risk of fractures in subjects exposed to systemic, and a wide range of topical corticosteroids in a large community-based sample. We adjusted for co-morbidity and examined the effects of different routes of administration on dose–response relationships between corticosteroids and fracture risk at several important skeletal sites. We also adjusted for use of other drugs that could potentially interfere with fracture risk. We calculated the exposure to corticosteroids in such a way that the total exposure was comparable between oral and topical (dermal, inhaled) preparations. In this way it was possible to assess the effects of the corticosteroids per se in different preparations adjusted for the effects of the underlying disease and measures of the severity of the underlying diseases.

End-points

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

The study end-points were occurrence of any fracture (ICD10 codes: S02.0–S02.9, S07.0–S07.9, S12.0–S12.9, S22.0–S22.9, S32.0–S32.8, S42.0–S42.9, S52.0–S52.9, S62.0–S62.9, S72.0–S72.9, S82.0–S82.9, S92.0–S92.9) between 1 January 2000 and 31 December 2000. In Denmark almost all patients with fractures are managed in the hospital system (also including the emergency rooms) [23], even fractures sustained abroad are registered upon return for insurance reasons. The capture of fractures is thus very high.

Exposure variables

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

The primary exposure variable was use of corticosteroids as oral, injected, inhaled, topical in eye/ear, dermal, rectally, topical in the mouth or nasal. The different preparations were converted to milligrams of prednisolone equivalents using a conversion formula for anti-inflammatory potency [10] (see Appendix 1). The exposure was total number of milligrams of prednisolone equivalents bought at pharmacies from 1996 to the date of fracture amongst the cases or the date of censoring (a dummy date equivalent to the date of fracture amongst the controls). The daily dose was calculated as number of milligrams bought during the observation period (from 1996 to the date of fracture or equivalent) divided by the date of fracture (or equivalent) minus the first date of a prescription for the corticosteroid in question. The dose intervals were divided into ≥7.5 mg day−1 of prednisolone equivalents (the dose known to suppress endogenous cortisol secretion), 2.5–7.49 mg day−1 (2.5 mg day−1 being the lowest dose shown by other authors to be associated with fracture risk amongst users of oral corticosteroids [3]) and <2.5 mg day−1.

In order to address the issue of conversion of different preparations, formulations and administration forms, a separate analysis was performed, where the dose was expressed as cumulated defined daily dosages (DDD). In this way, the doses were not directly comparable between inhaled and oral corticosteroids, as 1 DDD for oral corticosteroids is very different from say 1 DDD for inhaled corticosteroids. However, DDD is a more general measure of the dose used not depending on conversion formulae.

The other exposure variables were occurrence of (i) diseases or conditions known to require treatment with corticosteroids, (ii) use of drugs other than corticosteroids known to be associated with fracture risk, (iii) number of contacts to the health service (hospitals, general practitioners or specialists) as a proxy variable for disease severity [24], and (iv) social variables [25].

The diseases and conditions included as confounders were: (i) obstructive lung diseases (asthma or chronic obstructive pulmonary disease – COPD) [4], (ii) rheumatic disorders (rheumatic arthritis, polymyalgia rheumatica) [26, 27], (iii) skin disorders often requiring treatment with corticosteroids (dermatitis, psoriasis), (iv) haematological disorders (leukaemia, lymphomas, multiple myeloma) [28], (v) Inflammatory bowel diseases (Crohn's disease and ulcerative colitis) [29], and (vi) organ transplantation (kidney, lung, heart, liver, pancreas). Other important disease confounders included (i) alcoholism and (ii) occurrence of a prior fracture or not [30].

The drugs associated with fracture risk were: (i) antiepileptic drugs [31], (ii) diuretics (separated into thiazides, loop, potassium sparing, and other types of diuretics) [32], (iii) sedatives, anxiolytics and hypnotics [33], (iv) neuroleptics [33], (v) antidepressants [34], and (vi) inhaled bronchodilators (beta-agonists and other types of bronchodilators such as anticholinergics) [4]. The exposure was ever use of the drug in question from 1996 to the date of fracture amongst the cases and the date of censoring amongst the controls.

The proxy variables for disease severity were: (i) number of bed days in hospital the year before the fractures and (ii) the number of contacts to general practitioner or practising specialist.

The social variables were: (i) working or not, (ii) income in the year of the fracture (dichotomised by average income), and (iii) living alone or together with another person.

Registers used

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

The information on fracture occurrence and occurrence of other diseases, prior fractures, alcoholism came from two registers: (i) The National Hospital Discharge Register [35] and (ii) The Psychiatric Central Register [36].

The National Hospital Discharge Register was founded in 1977 [35]. It covers all inpatient contacts from 1977 to 1994, and from 1995 also all outpatient visits to hospitals, outpatient clinics and emergency rooms [35]. Upon discharge, the doctor codes the reason for the contact using the ICD system. The code used is at the discretion of the individual doctor. The register has a nationwide coverage and an almost 100% capture of contacts [35]. In general the validity of registrations is high [37], especially for fractures, where a precision of 97% has been reported for fractures treated both on an inpatients basis and for fractures treated on an outpatient basis via emergency rooms (say a forearm fracture) [29]. The cases occurred only once in the analyses with the first occurrence of an incident fracture during the year 2000.

The Psychiatric Central Register was founded in 1968 and covers all in- and outpatient contacts to Danish mental hospitals [36]. It has a nationwide coverage and a high validity of diagnoses has been reported [23]. This register also uses the ICD system for coding contacts.

The National Health Service keeps a register of all contacts to general practitioners for reimbursement purposes. The register does not contain ICD codes for the contacts but codes for the nature of the contact (regular check-up visit, routine vaccination in children).

The number of bed days in the year 1999 was counted as the number of days the patient spent on an inpatient basis in any hospital in 1999. The number of contacts to general practitioner or specialist was counted as the total number of reimbursement codes issued by the general practitioner or specialist in the year 1999 for each patient.

The Danish Medicines Agency keeps a nationwide register of all drugs sold at pharmacies throughout the country from 1996 and onwards (The National Pharmacological Database run by the Danish Medicines Agency: http://www.dkma.dk). Any drugs bought are registered with ATC code, dosage sold, and date of sale for the period 1 January 1996 to 31 December 2000. As all sales are registered to the individual who redeemed the prescription, the capture and validity is high.

Information on income was obtained from the Tax authorities, and information on working status and marital status from the National Bureau of Statistics (Statistics Denmark).

It is possible to link these sources of information through the Central Person Register Number which is a unique registration code given to every inhabitant – to some degree similar to the American social security number – that allows registration on an individual basis.

The project was approved and controlled by the National Board of Health, the Danish Data Protection Agency, and the Directory board of the Psychiatric Central Register.

Statistical analyses

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

Mean and standard deviation were used as descriptive statistics. Crude and adjusted odds ratios (ORs), and 95% confidence intervals were calculated. A conditional logistic regression analysis was used to assess the association between any fracture and the exposure variable. Crude and multiple-adjusted ORs were calculated.

Analyses were performed for total dose of corticosteroids used (converted to mg prednisolone) and daily dose of prednisolone (mg day−1). A separate analysis was performed taking into account when the last prescription of prednisolone was reimbursed (<3 months vs. ≥3 months, <6 vs. ≥6 months, and <1 vs. ≥1 year ago). Separate analyses were also performed for total amount of prednisolone equivalents ingested (mg), along with a stratified analysis in patients with or without prior fractures.

Age- (<15, 15–49, ≥50 years, and <40, 40–64, and ≥65 years) and gender-stratified analyses were also performed.

Analyses were performed using stata 8.1 (STATA Corp., College Station, TX, USA) and spss 10.1.0 (SPSS Inc., Chicago IL, USA), both in the UNIX version.

Results

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

Table 1 displays baseline characteristics of the participants. The cases and controls were well matched. There were significant differences, as more cases were unmarried, unemployed, had previous fractures, and in general had used more drugs and had more concurrent diseases than the controls.

Table 1.  Characteristics of patients and controls: any fracture (values are number and % or mean ± SD)
VariableCases (n = 124 655)Controls (n = 373 962)P
  1. GP, general practitioner; SERM, selective oestrogen receptor modulators (raloxifene); HRT, hormonal replacement therapy (oestrogen compounds with or without progestogens) – women only.

Age (years)43.44 ± 27.3943.44 ± 27.39
Men48.248.2
Annual income (US$)25 974 ± 22 38527 955 ± 31 243<0.01
Married29.033.4<0.01
Working40.444.9<0.01
Previous fracture33.115.0<0.01
Number of bed days in hospital in 19999.7 ± 39.74.2 ± 20.3<0.01
Number of contacts to GP or specialists in 199923.9 ± 43.318.1 ± 31.4<0.01
Asthma3.32.3<0.01
COPD3.41.8<0.01
Rheumatic disorders2.31.5<0.01
Crohn's disease0.30.2<0.01
Ulcerative colitis0.40.3<0.01
Allergy1.21.0<0.01
Haematological disorders0.40.3<0.01
Skin disorders1.31.0<0.01
Inhaled beta-agonists19.216.3<0.01
Other inhaled bronchodilators1.40.9<0.01

Table 2 shows the exposure to corticosteroids. More cases than controls were exposed to corticosteroids, and in general the cases had used higher doses than the controls. Only limited doses of topical corticosteroids were used in the mouth, nose, and in the eyes or ears.

Table 2.  Exposure to corticosteroids
VariableCases (n = 124 655)Controls (n = 373 962)P
Use of corticosteroids (%)
 Ever use of any corticosteroid 54.3 50.7<0.01
 Topical in the eye or ear 12.7 11.7<0.01
 Topical on the skin 38.0 35.7<0.01
 Topical inhaled  9.3  7.5<0.01
 Systemic/topical injected  6.5  5.6<0.01
 Topical in the mouth  0.9  0.8<0.01
 Topical in the nose  8.2  8.00.01
 Systemic oral  7.0  4.8<0.01
 Topical in the rectum or colon  0.2  0.20.01
Cumulated prednisolone dose (mg) per user, median (range)
 Any corticosteroid 185 (1–193 630) 160 (1–150 000)<0.01
 Topical in the eye or ear  25 (10–3850)  25 (10–3892)0.62
 Topical on the skin 133 (13–130 208) 133 (13–150 000)<0.01
 Topical inhaled 640 (21–46 100) 560 (21–52 032)<0.01
 Systemic/topical injected  93 (25–21 875)  93 (25–60 107)0.01
 Topical in the mouth   1 (1–169)   1 (1–115)0.03
 Topical in the nose  80 (8–5952)  88 (8–14 239)0.60
 Systemic oral2000 (63–156 500)1000 (62–76 000)<0.01
 Topical in the rectum or colon 581 (56–13 125) 563 (56–23 100)0.99

Table 3 shows the risk of any fracture associated with use of different corticosteroid administration modes stratified by daily dose.

Table 3.  Dose-dependent risk of any fracture with use of corticosteroid (mg day−1 of prednisolone equivalents)
VariableCases/controlsCrude ORMultiple-adjusted ORa
  1. aMultiple adjusted for variables in this table and Table 4, and the confounders mentioned in footnote of Table 4.

Inhaled
 Never use 1.001.00
 <2.5 mg day−19206/23 0961.22 (1.19–1.25)1.01 (0.97–1.04)
 2.5–7.49 mg day−12034/44941.41 (1.34–1.48)0.99 (0.93–1.06)
 ≥7.5 mg day−1296/4961.82 (1.58–2.10)1.17 (1.00–1.38)
Oral
 Never use 1.001.00
 <2.5 mg day−14447/10 9381.25 (1.20–1.29)0.97 (0.93–1.01)
 2.5–7.49 mg day−12321/43731.66 (1.57–1.74)1.15 (1.09–1.22)
 ≥7.5 mg day−11906/27072.16 (2.04–2.29)1.59 (1.49–1.70)
Injected
 Never use 1.001.00
 <2.5 mg day−17992/20 8661.75 (1.71–1.80)1.04 (1.01–1.07)
 2.5–7.49 mg day−150/1271.22 (0.88–1.70)1.15 (0.82–1.61)
 ≥7.5 mg day−112/411.34 (0.70–2.54)0.84 (0.43–1.64)
Eye or ear
 Never use 1.001.00
 <2.5 mg day−115 752/43 6901.09 (1.07–1.11)0.97 (0.95–1.00)
 2.5–7.49 mg day−120/630.99 (0.60–1.63)0.84 (0.47–1.49)
 ≥7.5 mg day−111/301.11 (0.56–2.22)1.11 (0.53–2.30)
Skin
 Never use 1.001.00
 <2.5 mg day−145 845/129 5291.10 (1.09–1.12)1.00 (0.98–1.01)
 2.5–7.49 mg day−11140/31761.12 (1.04–1.19)0.91 (0.84–0.98)
 ≥7.5 mg day−1360/8381.34 (1.18–1.51)1.09 (0.95–1.25)
Mouth
 Never use 1.001.00
 <2.5 mg day−11112/29651.12 (1.05–1.21)0.95 (0.88–1.03)
 2.5–7.49 mg day−10/0
 ≥7.5 mg day−10/0
Nasal
 Never use 1.001.00
 <2.5 mg day−110 155/29 6111.03 (1.01–1.06)0.97 (0.94–1.00)
 2.5–7.49 mg day−141/1271.00 (0.78–1.43)1.01 (0.68–1.50)
 ≥7.5 mg day−110/251.20 (0.58–2.50)1.01 (0.42–2.42)
Rectal
 Never use 1.001.00
 <2.5 mg day−1242/5811.25 (1.08–1.45)1.09 (0.92–1.30)
 2.5–7.49 mg day−126/671.21 (0.77–1.90)0.89 (0.55–1.44)
 ≥7.5 mg day−12/260.23 (0.06–0.86)0.24 (0.06–1.03)

In the crude analysis an increased fracture risk was present for inhaled, skin, and for orally administered corticosteroids. Only a limited increase in relative fracture risk was present with low doses of eye/ear, injected, mouth, nasal and rectal corticosteroids.

Upon adjustment for multiple confounders, an increased risk only persisted for oral corticosteroids above 2.5 mg prednisolone equivalents per day and for inhaled corticosteroids at a dose above 7.5 mg prednisolone equivalents per day. In most cases the changes from crude to adjusted OR were limited.

Table 4 shows the effect of covariates on the risk of any fracture. In the crude analysis all diseases and conditions were associated with a significant increase in fracture risk.

Table 4.  Risk of any fracture: adjusted for the corticosteroids in Table 3. The reference category is no presence of the disease in question in the time period of interest
VariableCrude ORMultiple-adjusted ORa
  1. aMultiple adjusted for the variables in Table 3 and this table, plus adjusted for use of antiepileptic drugs, diuretics (loop diuretics, thiazide diuretics, potassium sparing diuretics, and other types of diuretics), anxiolytics and sedatives, neuroleptics, antidepressants, alcoholism number of bed days in 1999, number of contacts to GP or specialists in 1999, working status, income, and living with another person or not.

Asthma (yes/no)1.43 (1.38–1.48)1.08 (1.02–1.14)
COPD (yes/no)1.86 (1.79–1.94)1.18 (1.12–1.24)
Rheumatic disorders (yes/no)1.58 (1.51–1.66)1.11 (1.06–1.17)
Crohn's disease (yes/no)1.31 (1.15–1.49)0.98 (0.86–1.13)
Ulcerative colitis (yes/no)1.21 (1.09–1.34)0.93 (0.82–1.04)
Allergy (yes/no)1.22 (1.15–1.30)1.05 (0.98–1.12)
Haematological disorders (yes/no)1.46 (1.31–1.62)1.12 (1.00–1.25)
Skin disorders (yes/no)1.35 (1.27–1.43)1.07 (1.00–1.15)
Organ transplantation (yes/no)3.40 (2.61–4.44)2.12 (1.55–2.89)
Prior fracture (yes/no)2.69 (2.44–2.97)2.42 (2.38–2.46)
Inhaled bronchodilators other than beta-agonists (yes/no)1.54 (1.45–1.63)1.01 (0.95–1.08)
Inhaled beta-agonists (yes/no)1.22 (1.20–1.24)1.09 (1.06–1.11)

Upon multiple adjustments for confounders the risk estimates were attenuated, significant associations only persisting for asthma, chronic obstructive pulmonary disease, rheumatic disorders and organ transplant recipients.

Table 5 shows the multiple-adjusted fracture risk associated with corticosteroid use in the forearm, hip and spine. As for all fractures combined (Table 3), an increased fracture risk was present in the hip and spine with more than 2.5 mg prednisolone equivalents ingested per day. None of the topical preparations were associated with an increased fracture risk.

Table 5.  Risk of hip, forearm and spine fractures with use of corticosteroids (multiple-adjusted OR, 95% CI), adjusted for the covariates in Table 6 (mg day−1 of prednisolone equivalents) and the factors mentioned in footnote of Table 6
VariableForearm (20 035 cases /60 030 controls)Hip (10 530 cases/ 31 535 controls)Spine (3364 cases/ 10 079 controls)
Inhaled
 Never use1.001.001.00
 <2.5 mg day−10.99 (0.90–1.09)0.88 (0.78–1.00)1.03 (0.84–1.26)
 2.5–7.49 mg day−10.97 (0.82–1.16)0.79 (0.66–0.95)0.86 (0.62–1.19)
 ≥7.5 mg day−11.07 (0.70–1.64)1.13 (0.68–1.87)0.81 (0.36–1.82)
Oral
 Never use1.001.001.00
 <2.5 mg day−10.99 (0.89–1.09)0.97 (0.87–1.08)1.16 (0.95–1.41)
 2.5–7.49 mg day−11.14 (0.98–1.33)1.27 (1.11–1.44)1.54 (1.18–1.99)
 ≥7.5 mg day−11.19 (0.99–1.43)1.45 (1.25–1.69)2.08 (1.54–2.79)
Eye or ear
 Never use1.001.001.00
 <2.5 mg day−10.95 (0.89–1.01)0.89 (0.84–0.95)1.03 (0.91–1.16)
 2.5–7.49 mg day−10.57 (0.11–2.92)0.39 (0.03–4.31)1.73 (0.15–19.5)
 ≥7.5 mg day−11.10 (0.11–11.0)
Injected
 Never use1.001.001.00
 <2.5 mg day−11.04 (0.97–1.12)0.84 (0.77–0.91)1.05 (0.91–1.22)
 2.5–7.49 mg day−10.95 (0.37–2.43)0.78 (0.28–2.21)0.18 (0.02–1.47)
 ≥7.5 mg day−12.34 (0.38–14.2)
Skin
 Never use1.001.001.00
 <2.5 mg day−10.99 (0.95–1.04)0.90 (0.86–0.95)0.99 (0.90–1.08)
 2.5–7.49 mg day−10.87 (0.72–1.07)0.96 (0.80–1.15)1.13 (0.79–1.60)
 ≥7.5 mg day−11.15 (0.78–1.68)0.98 (0.69–1.40)0.81 (0.39–1.66)
Mouth
 Never use1.001.001.00
 <2.5 mg day−11.12 (0.92–1.36)0.93 (0.75–1.15)1.01 (0.67–1.51)
 2.5–7.49 mg day−1
 ≥7.5 mg day−1
Nasal
 Never use1.001.001.00
 <2.5 mg day−11.06 (0.99–1.14)0.78 (0.70–0.86)0.86 (0.52–1.11)
 2.5–7.49 mg day−10.69 (0.20–2.46)0.21 (0.03–1.61)2.67 (0.40–17.8)
 ≥7.5 mg day−14.12 (0.20–83.8)1.90 (0.12–30.8)
Rectal
 Never use1.001.001.00
 <2.5 mg day−11.18 (0.74–1.87)0.93 (0.55–1.57)1.59 (0.59–4.25)
 2.5–7.49 mg day−11.30 (0.40–4.28)0.74 (0.18–3.12)0.38 (0.03–5.17)
 ≥7.5 mg day−11.76 (0.15–21.2)

Table 6 shows the multiple-adjusted fracture risk associated with potential confounders for the corticosteroids in Table 5. A prior fracture and alcoholism were associated with an increased fracture risk at all skeletal sites whilst lung diseases, rheumatic disorders, Crohn's disease, haematological diseases and organ transplantation were associated with the risk of hip fractures. Only haematological diseases were associated with spine fractures.

Table 6.  Risk of forearm, hip and spine fracture for covariates (adjusted OR, 95% CI): adjusted for the corticosteroids in Table 5. The reference category is no presence of the disease in question in the time period of interest
VariableForearmHipSpine
  1. Adjusted for use of antiepileptic drugs, diuretics (loop diuretics, thiazide diuretics, potassium sparing diuretics, and other types of diuretics), anxiolytics and sedatives, neuroleptics, antidepressants, alcoholism, number of bed days in 1999, number of contacts to GP or specialists in 1999, working status, income, and living with another person or not.

Asthma (yes/no)1.18 (1.02–1.36)1.32 (1.11–1.56)0.84 (0.63–1.13)
COPD (yes/no)1.01 (0.89–1.15)1.63 (1.46–1.82)1.23 (0.98–1.55)
Rheumatic disorders (yes/no)0.88 (0.77–1.02)1.17 (1.04–1.31)1.15 (0.90–1.48)
Crohn's disease (yes/no)0.66 (0.42–1.03)1.57 (1.02–2.41)2.06 (0.96–4.39)
Ulcerative colitis (yes/no)0.84 (0.60–1.17)0.94 (0.65–1.37)0.64 (0.31–1.30)
Allergy (yes/no)0.99 (0.81–1.21)0.80 (0.60–1.06)0.87 (0.58–1.30)
Haematological disorders (yes/no)0.80 (0.58–1.10)1.31 (1.01–1.69)2.11 (1.36–3.28)
Skin disorders (yes/no)1.20 (0.98–1.46)1.18 (0.94–1.49)0.98 (0.66–1.44)
Organ transplantation (yes/no)1.58 (0.67–3.74)6.29 (1.21–32.7)0.93 (0.14–6.34)
Prior fracture (yes/no)1.92 (1.83–2.01)2.04 (1.94–2.15)2.36 (2.15–2.59)
Inhaled bronchodilators other than beta-agonists (yes/no)0.99 (0.83–1.19)1.04 (0.87–1.23)1.20 (0.87–1.65)
Inhaled beta-agonists (yes/no)0.99 (0.92–1.06)0.99 (0.91–1.07)1.11 (0.96–1.28)

A prior fracture was a significant risk factor for sustaining a fracture. Age and gender stratification did not change the estimates significantly.

A separate analysis taking into account the last time when a prescription for prednisolone was reimbursed did not change the estimates significantly when using the average daily dose of prednisolone, the total dose (mg or DDD).

Shifting the analyses to cumulated dose of prednisolone failed to show an association except at very high doses of above 2000 mg for oral prednisolone equivalents and 4000 mg for injected corticosteroids. No associations could be demonstrated for the topical preparations. Shifting the analysis to cumulated DDD did not change the results significantly.

Discussion

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

In this large-scale community-based case–control study we found an increased fracture risk with use of oral corticosteroids at dosages higher than 2.5 mg prednisolone equivalents per day. No systematic increase in overall fracture risk could be demonstrated for topical preparations except for a borderline significant increase of limited magnitude with inhaled corticosteroids above a dosage of 7.5 mg prednisolone equivalents per day (equivalent to 1875 μg of budesonide or beclomethasone) after adjustment for confounders.

The advantages of the study are that it is community based and that the precision of information is high as selection and information bias are limited.

Limitations of the study

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

The major weakness is the presence of confounders for whom it is not possible to adjust for such as smoking and body mass index [38]. Furthermore, differences in absorption of inhaled and other topical corticosteroids may influence the amount of corticosteroids actually delivered into the systemic circulation.

In this study a general conversion formula was used to express the exposure as prednisolone equivalents. However, this general formula may not be applicable for topical preparations. It is based on anti-inflammatory properties and this may not apply to the potency of effects on bone metabolism.

Another major weakness is that not all diagnoses can be controlled for. Only subjects in contact with a hospital (on an inpatient or an outpatient basis) have a diagnosis of say COPD registered, as the general practitioners do not register their contacts by ICD codes. Only more severe cases of lung diseases come to the attention of the hospital system, and these conditions are thus under-diagnosed by the register leading to an overestimation of the risk. However, due to the presence of COPD amongst the controls, the risk estimates associated with the disease in Tables 4 and 6 may be underestimated.

The study relies on register data collected for other purposes. Not all fractures may be registered, and not all drugs ingested may be accounted for in the prescription database. The validity of other diagnoses (COPD, asthma, haematological disorders) is not known, and many cases of skin disorders, and eye and ear diseases are managed outside the hospital system and thus not registered. This will tend to overestimate the effects of skin disorders and eye and ear diseases, as probably only the more severe cases come in contact with the hospital system. However, despite this no significant associations were found for these conditions.

The register data are incomplete measures of disease severity and co-morbidity, and this may lead to residual confounding. The association between fracture risk and inhaled corticosteroids may thus be the result of residual confounding.

The register of prescriptions measures doses bought. However, the patients may actually not take all of the drugs purchased, and this may lead to an overestimation of the doses to which the patients were exposed. However, once a drug is bought it is likely that it will be used because the prescription is nonrefundable. Furthermore, corticosteroids administered whilst the patients were in hospital were not registered – however the amounts administered in this way are small. Patients with chronic conditions are more likely to be hospitalized and receive corticosteroids. During hospital stay, more corticosteroids will be administered intravenously or orally, for e.g. an exacerbation of COPD or asthma. This will underestimate the total dose of prednisolone, and thus tend to underestimate the ORs, as patients registered as not having had corticosteroids may have received them.

The number of fractures, especially of the spine, is a minimum. The register has nationwide coverage and the capture rate is high. In Denmark almost all fractures are managed via the hospital system (either on an outpatient basis via the emergency rooms or on an inpatient basis), and thus entered into the database. Almost all hip fractures are managed on an inpatient basis and are thus registered. Many Colles’ fractures are managed on an outpatient basis; however, they are still registered via the emergency rooms. Only a minor proportion of spine fractures is symptomatic [39], and thus never come to the attention of the medical system. A number of smaller fractures, for e.g. the fingers and ankles, may also never come to the attention of the health care system.

Effects of corticosteroids

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

The risk estimates for use of topical corticosteroids were attenuated upon adjustment for use of oral corticosteroids (Table 3) for inhaled, injected, and for low doses of rectal corticosteroids, whilst only small changes occurred for other topical preparations. This was due to the fact that use of systemic (oral) steroid was most prominent in these disease categories, whilst other topical preparations were used in conditions that did not need systemic corticosteroids as often or not at all.

After further adjustment for type of disease requiring corticosteroid treatment, the risk estimates were further attenuated for inhaled and oral corticosteroids whilst they did not change much for other preparations (Table 3).

Topical corticosteroids may often be used intermittently (e.g. on the skin, in the mouth or in the eyes or years). Previous studies with oral corticosteroids have shown a marked increase in fracture risk within the first months of use, and a similar decrease within the first few months of discontinuation [3]. Analyses of average dosages may thus not completely capture short-term changes in fracture risk. However, no permanent major increases in fracture risk seemed associated with the use of topical corticosteroid preparations.

The absence of larger effects of the time interval since last prescription of prednisolone may be due to the rapid increases and decreases in risk when starting and discontinuing prednisolone [3], i.e. these changes did not influence the long-term results.

Effects of underlying disease

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

A somewhat larger increase in fracture risk was seen in patients with COPD than with asthma except for fractures of the forearm (Tables 4 and 6). The reason for this may be systemic inflammation, smoke burden, limited physical activity, diet and lifestyle with less sun exposure, and thereby lower vitamin D. Furthermore, COPD is a disease of subjects of a higher age than asthma.

An increased hip fracture risk was present in patients with Crohn's disease but not in patients with ulcerative colitis in accordance with previous studies [29]. The absence of an increase in fracture risk in patients with ulcerative colitis may thus be linked to the fact that more patients with this disease are managed by rectal corticosteroids than in Crohn's disease.

Haematological disorders were associated with hip and, in particular, spine fractures, the latter perhaps due to multiple myeloma in the spine.

Skin disorders and allergy did not contribute to fracture risk, probably because the inflammation was too limited to give rise to systemic perturbations in bone active inflammatory cytokines as seen in Crohn's disease [40]. The production of cytokines inducing increased bone resorption and decreased bone formation [40] is probably also the reason why rheumatic disorders in general were associated with an increase in the risk of any fracture and the risk of hip fractures [26, 27].

A prior fracture was a significant risk factor for sustaining a fracture. This could point at an effect of either a decreased BMD or an increased risk of sustaining traumas, e.g. through falls.

In conclusion, use of daily doses of oral prednisolone equivalents >2.5 mg is associated with an increase in overall fracture risk and in the risk of fractures of the hip and spine. No increase was associated with inhaled corticosteroids except at daily dosages above 7.5 mg of prednisolone (equivalent to 1875 μg of budesonide or beclomethasone). No increase in fracture risk was associated with corticosteroids used as injections or topically on the skin, in the nose, locally in the mouth, in the eyes or ears or rectally.

Acknowledgements

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

The National Bureau of Statistics (Statistics Denmark) is acknowledged for their help, without which this project would not have been possible.

Financial support was provided by the Danish Medical Research Council (grant number 22-04-0495).

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix
  • 1
    Canalis E. Mechanisms of glucocorticoid action in bone: implications to glucocorticoid-induced osteoporosis. J Clin Endocrinol Metab 1996; 81: 34416.
  • 2
    Patschan D, Loddenkemper K, Buttgereit F. Molecular mechanisms of glucocorticoid-induced osteoporosis. Bone 2001; 29: 498505.
  • 3
    van Staa TP, Leufkens HGM, Abenhaim L, Zhang B, Cooper C. Use of oral corticosteroids and risk of fractures. J Bone Mineral Res 2000; 15: 9931000.
  • 4
    van Staa TP, Leufkens HGM, Cooper C. Use of inhaled corticosteroids and risk of fractures. J Bone Mineral Res 2001; 16: 5818.
  • 5
    van Staa TP, Leufkens HGM, Cooper C. The epidemiology of corticosteroid-induced osteoporosis: a meta-analysis. Osteoporos Int 2002; 13: 77787.
  • 6
    Suissa S, Baltzan M, Kremer R, Ernst P. Inhaled and nasal corticosteroid use and the risk of fracture. Am J Respir Crit Care Med 2004; 169: 838.
  • 7
    van Staa TP, Abenhaim L, Cooper C, Zhang B, Leufkens HGM. Public health impact of adverse bone effects of oral corticosteroids. Br J Clin Pharmacol 2001; 51: 6017.
  • 8
    van Staa TP, Leufkens HGM, Abenhaim L, Begaud B, Zhang B, Cooper C. Use of oral corticosteroids in the United Kingdom. QJM 2000; 93: 10511.
  • 9
    Hubbard RB, Smith CJP, Smeeth L, Harrison TW, Tattersfield AE. Inhaled corticosteroids and hip fracture: a population-based case–control study. Am J Respir Crit Care Med 2002; 166: 15636.
  • 10
    Vestergaard P, Olsen ML, Johnsen SP, Rejnmark L, Sørensen HT, Mosekilde L. Corticosteroid use and risk of hip fracture: a population-based case–control study in Denmark. J Intern Med 2003; 254: 48693.
  • 11
    van Staa TP, Cooper C, Brusse LS, Leufkens H, Javaid MK, Arden NK. Inflammatory bowel disease and the risk of fracture. Gastroenterology 2003; 125: 15917.
  • 12
    van Staa TP, Bishop N, Leufkens HG, Cooper C. Are inhaled corticosteroids associated with an increased risk of fracture in children? Osteoporos Int 2004; 15: 78591.
  • 13
    Edsbacker S, Andersson T. Pharmacokinetics of budesonide (Entocort EC) capsules for Crohn's disease. Clin Pharmacokinet 2004; 43: 80321.
  • 14
    Tromm A, Mollmann H, Barth J et al. Pharmacokinetics and rectal bioavailability of hydrocortisone acetate after single and multiple administration in healthy subjects and patients. J Clin Pharmacol 2001; 41: 53641.
  • 15
    Daley-Yates PT, Kunka RL, Yin Y, Andrews SM, Callejas S, Ng C. Bioavailability of fluticasone propionate and mometasone furoate aqueous nasal sprays. Eur J Clin Pharmacol 2004; 60: 2658.
  • 16
    Mizuchi A, Miyachi Y, Tamaki K, Kukita A. Percutaneous absorption of betamethasone 17-benzoate measured by radioimmunoassay. J Invest Dermatol 1976; 67: 27982.
  • 17
    Dahlstrom K, Thorsson L, Larsson P, Nikander K. Systemic availability and lung deposition of budesonide via three different nebulizers in adults. Ann Allergy Asthma Immunol 2003; 90: 22632.
  • 18
    Pedersen S. Do inhaled corticosteroids inhibit growth in children? Am J Respir Crit Care Med 2001; 164: 52135.
  • 19
    Singh SD, Whale C, Houghton N, Daley-Yates P, Kirby SM, Woodcock AA. Pharmacokinetics and systemic effects of inhaled fluticasone propionate in chronic obstructive pulmonary disease. Br J Clin Pharmacol 2003; 55: 37581.
  • 20
    Pouw EM, Prummel MF, Oosting H, Roos CM, Endert E. Beclomethasone inhalation decreases serum osteocalcin concentrations. BMJ 1991; 302: 6278.
  • 21
    Wong CA, Walsh LJ, Smith CJ et al. Inhaled corticosteroid use and bone-mineral density in patients with asthma. Lancet 2000; 355: 1399403.
  • 22
    Laan RF, van Riel PL, van de Putte LB, van Erning LJ, Van't Hof MA, Lemmens JA. Low-dose prednisone induces rapid reversible axial bone loss in patients with rheumatoid arthritis. A randomised controlled study. Ann Intern Med 1993; 119: 9638.
  • 23
    Vestergaard P, Emborg C, Støving RK, Hagen C, Mosekilde L, Brixen K. Fractures in patients with anorexia nervosa, bulimia nervosa, and other eating disorders – a nation-wide register study. Int J Eating Disord 2002; 32: 3018.
  • 24
    Vestergaard P, Rejnmark L, Mosekilde L. Fracture risk associated with use of anti-epileptic drugs. Epilepsia 2004; 45: 13307.
  • 25
    Jones S, Johansen A, Brennan J, Butler J, Lyons RA. The effect of socioeconomic deprivation on fracture incidence in the United Kingdom. Osteoporos Int 2004; 15: 5204.
  • 26
    Cooper C, Carbone L, Michet CJ, Atkinson EJ, O'Fallon WM, Melton LJ III. Fracture risk in patients with ankylosing spondylitis: a population-based study. J Rheumatol 1994; 21: 187782.
  • 27
    Cooper C, Coupland C, Mitchell M. Rheumatoid arthritis, corticosteroid therapy and hip fracture. Ann Rheum Dis 1995; 54: 4952.
  • 28
    Abildgaard N, Brixen K, Kristensen JE, Vejlgaard T, Charles P, Nielsen JL. Assessment of bone involvement in patients with multiple myeloma using bone densitometry. Eur J Haematol 1996; 57: 3706.
  • 29
    Vestergaard P, Mosekilde L. Fracture risk in patients with celiac disease, Crohn's disease, and ulcerative colitis: a nation-wide follow-up study in 16,416 patients in Denmark. Am J Epidemiol 2002; 156: 110.
  • 30
    Klotzbuecher CM, Ross PD, Landsman PB, Abbott TA III, Berger M. Patients with prior fractures have an increased risk of future fractures: a summary of the literature and statistical synthesis. J Bone Miner Res 2000; 15: 72139.
  • 31
    Vestergaard P, Tigaran S, Rejnmark L, Tigaran C, Dam M, Mosekilde L. Fracture risk is increased in epilepsy. Acta Neurol Scand 1999; 99: 26975.
  • 32
    Jones G, Nguyen T, Sambrook PN, Eisman JA. Thiazide diuretics and fractures: can meta-analysis help? J Bone Miner Res 1995; 10: 10611.
  • 33
    Ray WA, Griffin MR, Schaffner W, Baugh DK, Melton LJ Jr. Psychotropic drug use and the risk of hip fracture. N Engl J Med 1987; 316: 3639.
  • 34
    Ensrud KE, Blackwell T, Mangione CM et al. Central nervous system active medications and risk for fractures in older women. Arch Intern Med 2003; 163: 94957.
  • 35
    Andersen TF, Madsen M, Jørgensen J, Mellemkjær L, Olsen JH. The Danish National Hospital Register. Dan Med Bull 1999; 46: 2638.
  • 36
    Munk-Jørgensen P, Mortensen PB. The Danish Psychiatric Central Register. Dan Med Bull 1997; 44: 824.
  • 37
    Mosbech J, Jørgensen J, Madsen M, Rostgaard K, Thornberg K, Poulsen TD. The Danish National Patient Register: evaluation of data quality. Ugeskr Læger 1995; 157: 37415.
  • 38
    Vestergaard P, Mosekilde L. Fracture risk associated with smoking – a meta-analysis. J Intern Med 2003; 254: 57283.
  • 39
    Black DM, Cummings SR, Kerpf DB et al. Randomised trial of the effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet 1996; 348: 153541.
  • 40
    Sylvester FA, Wyzga N, Hyams JS, Gronowicz GA. Effect of Crohn's disease on bone metabolism in vitro: a role for interleukin-6. J Bone Miner Res 2002; 17: 695702.

Appendix

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Subjects and methods
  5. Study design
  6. End-points
  7. Exposure variables
  8. Registers used
  9. Statistical analyses
  10. Results
  11. Discussion
  12. Limitations of the study
  13. Effects of corticosteroids
  14. Effects of underlying disease
  15. Conflict of interest statement
  16. Acknowledgements
  17. References
  18. Appendix

Appendix 1

Equipotency of different corticosteroids (anti-inflammatory potency).

CorticosteroidEquipotent dose (mg)
Aclomethasone4
Beclomethasone1.25
Bethamethasone0.75
Budesonide1.25
Clobetasole0.75
Clobetason4
Desoxymethasone0.75
Dexamethasone0.75
Difluorcortenolone0.625
Flumethasone0.75
Fluocinoide0.625
Fluocinolone0.625
Fluormetholone0.75
Fluticasone0.625
Hydrocortisone20
Hydrocortisone butyrate4
Methylprednisolone4
Momethasone0.625
Prednisolone5
Prednisone5
Rimexolone5
Triamcinolone4