High-Dose Treatment With Vitamin A Analogues and Risk of Fractures

Objective  To study fracture risk associated with use of systemic vitamin A analogue (isotretinoin and acitretin) treatment.

Design  Case-control study.

Setting  Nationwide registry.

Participants  A total of 124 655 patients with fractures (cases) and 373 962 age- and sex-matched controls.

Main Outcome Measure  Incidence of fractures in patients with and without exposure to systemic vitamin A analogues. Confounder control was performed for social variables, contacts with hospitals and general practitioners, alcoholism, and a number of other variables known to potentially affect fracture risk, including use of systemic, intramuscular, and topical corticosteroids and antiepileptic drugs and comorbid conditions.

Results  No trend in risk of any fracture or of hip, forearm, or spine fractures was present with increasing doses or durations of treatment with vitamin A analogues. Subdividing vitamin A analogues into isotretinoin and acitretin did not change the results.

Conclusion  Risk of fracture is not associated with vitamin A analogue treatment.

High doses of vitamin A as dietary intake or supplements have been associated with adverse skeletal effects.¹ Most epidemiologic studies have failed to show a relationship between total retinol intake and risk of fractures^2-5 at retinol doses of greater than 1.5 mg⁶ and up to greater than 7.5 mg⁷ retinol equivalents per day; however, some studies have identified an increased risk of fractures even at these dosages.^8-10 Although most studies have not reported any detrimental effect of vitamin A intake on bone mineral density (BMD),⁷ some studies have reported such effects at high doses.⁹ Therefore, it is not clear whether very high doses of vitamin A analogues may have detrimental effects on the skeleton.¹

Patients may receive treatment with very high doses of vitamin A analogues such as isotretinoin (3cis-retinoic acid) and acitretin ([2E,4E,6E,8E]-9-[4-methoxy-2,3,6-trimethylphenyl]-3,7-dimethylnona-2,4,6,8-tetraenoic acid). Isotretinoin, acitretin, and analogues have been associated with bone changes such as hyperostosis,¹¹ hypercalcemia,¹² impaired markers of bone turnover,¹³ and decreased BMD.^14,15 However, single courses of isotretinoin for up to 20 weeks may not impair bone turnover and BMD.^16-18 It is not clear whether large doses of isotretinoin and equivalents or long-term treatment with these are associated with an increased risk of fractures.

We therefore conducted a large-scale, population-based, case-control study to assess the negative effects of pharmacologic treatment with vitamin A analogues.

This was a case-control study with data obtained from 2 nationwide registers in Denmark. All subjects sustaining a fracture during the year 2000 in Denmark were included as cases (n = 124 655), and for each case 3 subjects of the same age (same birth year) and sex were randomly selected from the background population as controls. Because not all cases could be matched with 3 controls, the total number of controls was 373 962, ie, 3 less than expected.

The study cases were occurrence of any fracture (International Statistical Classification of Diseases, 10th Revision [ICD-10] 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, and S92.0-S92.9) between January 1, 2000, and December 31, 2000. In Denmark almost all patients with fractures are treated in the hospital system (also including the emergency departments)¹⁹; even those with fractures sustained abroad are registered on return for insurance reasons. Thus, the capture of fractures is high.^20,21 The vertebral fractures included were clinical fractures collected from hospital records of patients referred to emergency departments or other departments and who had a vertebral fracture diagnosed.

Using the Civil Registration System, which has electronic records on all changes in vital status, including change of address and date of death for the entire Danish population since 1968, we randomly selected 3 female controls for each case, matched by year of birth. The controls were selected by means of the incidence-density sampling technique,²² ie, the controls had to be alive and at risk of fracture at the time the corresponding case was diagnosed. They were required to be individuals who did not have a fracture during the year 2000. Because not all cases could be matched with 3 controls, a total of 373 962 controls were included in the study, ie, 3 less than expected.

The information on fracture occurrence and occurrence of other diseases, previous fractures, and alcoholism came from 2 registers: (1) the National Hospital Discharge Register²⁰ and (2) the Psychiatric Central Register.²³ The National Hospital Discharge Register, founded in 1977, covers all inpatient contacts from 1977 through 1994 and, starting in 1995, also all outpatient visits to hospitals, outpatient clinics, and emergency departments.²⁰ On discharge of the patient, the physician codes the reason for the contact by using the ICD system. The code used is at the discretion of the individual physician. The register has nationwide coverage and an almost 100% capture of contacts.²⁰ In general, the validity of registrations is high,²¹ especially for fractures, for which a precision of 97% has been reported for fractures treated both on an inpatient basis and on an outpatient basis via emergency departments (eg, a forearm fracture).²⁴ The cases occurred only once in the analyses with the first occurrence of an incident fracture during 2000. The National Health Service keeps a register of all contacts with general practitioners for reimbursement purposes. The register does not contain ICD codes for the contacts but codes for the nature of the contact (eg, regular check-up visit or routine vaccination in children).

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

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

The primary exposure was use of systemic vitamin A analogues (isotretinoin and acitretin) and topical vitamin A analogues irrespective of indications. Topical vitamin A use was included as a confounder control. Topical vitamin A analogues were used only for skin diseases and not for skin aging.

The other exposure variables were (1) use of drugs known to be associated with fracture risk (eg, corticosteroids and antiepileptic drugs); (2) number of contacts with the health service (hospitals, general practitioners, or specialists) as a proxy variable for disease severity²⁵ and the Charlson index, which is an index of 19 comorbid conditions²⁶; and (3) social variables.²⁷

The corticosteroids variable included all forms of corticosteroids, ie, oral, intramuscular, and topical. These factors were chosen because they were known to potentially affect fracture risk and were regarded as important potential confounders in a setting in which many variables besides the main factor may influence the risk of fractures (confounding by indication). The variables were entered into the statistical analysis and analyses for interaction were performed. Other important disease confounders included (1) alcoholism²⁸ and (2) occurrence of a previous fracture.²⁹

The social variables were as follows: (1) employment status, (2) income in the year of the fracture (dichotomized by average income), and (3) living situation (alone or with another person). These factors were included because previous studies have indicated that living in a relationship rather than living alone may be associated with a decreased risk of fractures; having a job may be associated with fewer fractures than being retired or out of a job; and income in some settings may be a predictor of fracture risk.²⁷ Information on body weight or body mass index was not available.

Mean and standard deviation were used as descriptive statistics. Crude and adjusted odds ratios 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 multiply adjusted odds ratios were calculated. Analyses were performed with Stata 8.2 (StataCorp, College Station, Texas) and SPSS 14.0 (SPSS Inc, Chicago, Illinois), both in the UNIX version. Stratified analyses by sex were performed to include the mandatory use of oral contraceptives in fertile women who were using these vitamin A analogues systemically.

The patients and controls were well matched for age and sex (Table 1). People experiencing fractures were more likely than controls to have comorbid conditions, have a diagnosis of alcoholism, have had previous fractures, and have taken antiepileptic drugs or any form of glucocorticoid.

None of the forms of administration of systemic and topical vitamin A analogues was associated with any change in fracture risk at any skeletal site (Table 2). There was no trend with increasing dose (Table 3) or treatment duration (Table 4) of systemic vitamin A analogues for fracture risk at any of the skeletal sites. Subdividing the vitamin A analogues into isotretinoin and acitretin did not influence the risk of fractures (Table 5), nor did analysis by cumulative dose of each drug (Table 6) except for a non–dose-dependent deviation for 5251 to 8750 mg of isotretinoin for hip fractures.

Neither acne nor psoriasis, indications for systemic treatment with vitamin A analogues, influenced the risk of fractures.

Adjustment for oral contraceptive use did not significantly change any of the risk estimates in women. There was no statistical interaction between fracture risk, use of vitamin A analogues, and use of corticosteroids either as any form (oral, dermal, or injected) or limited to oral corticosteroids. Therefore, analyzing oral corticosteroids alone instead of any form of corticosteroid did not change the results.

In this large-scale, population-based, case-control study, no trend in risk of fractures was associated with dose or duration of treatment with systemic vitamin A analogues. Even very large daily doses of 14 mg of vitamin A analogues (equivalent to 14 000 μg of retinol equivalents per day) were not associated with an increased risk of fractures. It thus seems that vitamin A analogues are safe in terms of fractures even at very high doses. Even though some studies have reported a decreased BMD with high doses of vitamin A as retinol in dietary intake or as supplements,^14,15 the decrease may not have been of such magnitude that it altered bone biomechanical competence.

Leachman et al¹⁴ reported an increase of 1.1% (an absolute increase of 0.1 in z score) in the lumbar spine and a decrease of 1.3% in the femoral neck (an absolute decrease of 0.1 in z score) after 6 months. These changes are very small and only cover a short time span. A z score change of 0.1 would equal a change in overall relative risk of fractures of only 1.6^0.1 = 1.05 by means of the estimates of Marshall et al,³⁰ ie, a very limited change. This is in keeping with our finding of no major change in risk of fractures. The study by DiGiovanna et al¹⁵ reported that isotretinoin was not associated with a decrease in BMD of the spine, hip, or forearm but rather a small trend toward an increase compared with normal controls. In contrast, etretinate (with the metabolite acitretin) was associated with a trend toward a decrease for the femoral neck and radius but not for the spine.¹⁵ It did not change the results to subdivide the exposure into isotretinoin and acitretin. This speaks against any significant detrimental effect of acitretin over isotretinoin.

The main advantages of this study are the large study sample and the uniform nature of the data. The main limitations are lack of information on sun exposure, use of UV radiation for treatment, smoking habits, and body mass index. In conclusion, no detrimental effect on fracture risk seems present with use of vitamin A analogue treatment.

Correspondence: Peter Vestergaard, MD, PhD, DrMedSc, The Osteoporosis Clinic, Aarhus Amtssygehus, Tage Hansens Gade 2, DK-8000 Aarhus C, Denmark (p-vest@post4.tele.dk).

Accepted for Publication: August 17, 2009.

Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Vestergaard, Rejnmark, and Mosekilde. Acquisition of data: Vestergaard. Analysis and interpretation of data: Vestergaard, Rejnmark, and Mosekilde. Drafting of the manuscript: Vestergaard. Critical revision of the manuscript for important intellectual content: Vestergaard, Rejnmark, and Mosekilde. Statistical analysis: Vestergaard. Obtained funding: Vestergaard. Administrative, technical, and material support: Vestergaard, Rejnmark, and Mosekilde. Study supervision: Rejnmark and Mosekilde.

Financial Disclosure: Dr Vestergaard has received research grants from Servier and Bayer Pharmaceuticals and travel grants from Eli Lilly and Company, Novartis, and Sanofi-Aventis.

Funding/Support: This study was supported by grant 22-04-0495 from the Danish Medical Research Council.

Role of the Sponsor: The sponsor had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; or in the preparation, review, or approval of the manuscript.

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