A population-based nested case–control study: the
use of
5-alpha-reductase inhibitors and the increased risk
of osteoporosis
diagnosis in patients with benign prostate
hyperplasia
Wen-Ling Lin*,†, Yow-Wen Hsieh*,†, Cheng-Li Lin‡, Fung-Chang Sung‡,§, Chieh-Hsi Wu†
and Chia-Hung Kao§,¶
Introduction
Osteoporosis is a disease of the bones that leads to an increased risk of fractures.1 In the treatment of benign prostatic hyperplasia,
finasteride decreases dihydrotestosterone (DHT), which is the potent androgen responsible for the development and enlargement of the prostate gland. Serum concentrations of DHT are reduced by 70% with daily dosing of finasteride 5 mg.2
A 70-years-old man taking finasteride (5 mg/days) for several
years developed progressive weakness of the arms and legs, combined with respiratory distress. Drug cessation resulted in a
gradual increase of muscle strength, and dyspneic symptoms improved over a 1-month period.3 Many of the physical and
behavioural changes that occur in men as they age are similar to those that occur in younger men with hypogonadism. These changes include decreased muscle mass, strength, bone mass and sexual function, and increased body fat, fatigue and depression. It is therefore reasonable to consider whether testosterone deficiency contributes to the adverse physical and behavioural
changes of ageing and whether these could be ameliorated by the administration of testosterone.4
disease in men 60 years and older, and patients may require medication to relieve urinary symptoms. Androgens play a critical role in prostatic growth and contribute to the pathogenesis of prostate disease.5,6 The 5-alpha-reductase inhibitors (5ARIs)
convert testosterone into the more potent dihydrotestosterone (DHT) and enhance androgenic signals in tissue.7 Finasteride
and dutasteride are two well-known 5ARIs that are used to treat BPH. Both agents result in similar volume reductions in prostate glands and the relief of urinary symptoms. In addition, both
agents achieve long-term results.8 Currently, no related research exists on the use of 5ARIs
and the increased risk of osteoporosis
in patients with BPH. One of the major secondary causes of osteoporosis in men is hypogonadism, which is observed in nearly 20% of men with symptomatic vertebral fractures and 50% of elderly men with hip fractures.9 Although the pathogenesis
of osteoporosis in men is multifactorial, testosterone is known to play a critical role in the maintenance of the male
skeleton. Testosterone replacement therapy improves bone density in men with hypogonadal osteoporosis, particularly if the
epiphyses remain open.10
Thus, it is reasonable to evaluate the association between the use of 5ARIs and osteoporosis diagnosis because the literature comprises only four comprehensive studies. We conducted this study using data from the National Health Insurance (NHI) system in Taiwan and analysed the risk of osteoporosis diagnosis
and the use of 5ARIs.
Materials
Data source
The National Health Insurance (NHI) programme, a universal health programme established in 1995, covers approximately 99% of the population (23_74 million people) and is contracted with 97% of the hospitals and clinics in Taiwan.11 In this study,
we used reimbursement claims data of the Longitudinal Health Insurance Database (LHID) established by the National Health Research Institute (NHRI) of the Department of Health. The database includes claims data for one million people covered by the NHI from 1996–2010. No significant difference exists in the distribution of gender and age between LHID and all beneficiaries.
We used the scrambled identification numbers to link data sets and safeguard the confidentiality of the insured population, avoiding ethical violations. In addition, this study was approved by the Institutional Review Board of China Medical University in Central Taiwan (CMU-REC-101-012).
Study sample
This study was conducted as a nested case–control study. First, we used the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) to identify newly diagnosed BPH patients (ICD-9-CM codes 600.xx; n = 46 997) as
the exposure cohort (Fig. 1). We excluded patients with a history of osteoporosis diagnosis before January 1, 2005. Each
patient was followed from 2005–2010. Patients with a diagnosis of osteoporosis (ICD-9-CM codes 733_0) became part of the case group, and patients without an osteoporosis diagnosis
became part of the nonosteoporosis diagnosis group. As a comparison group, we randomly selected four people from the nonosteoporosis diagnosis group and frequency matched them to
each patient of the case group based on age, BPH year, osteoporosis diagnosis index year and osteoporosis diagnosis index
month. In total, we included 1352 patients in the osteoporosis diagnosis case group and 5387 in the nonosteoporosis diagnosis control group.
The comorbidities and medications considered including medication of alpha-adrenergic blockers and comorbidities of hypertension CM codes 401 to 405), diabetes (ICD-9-CM code 250), hyperlipidaemia (ICD-9-(ICD-9-CM code 272), chronic obstructive pulmonary disease (COPD) (ICD-9-CM codes 490– 496), obesity ((ICD-9-CM code 278), bladder outlet obstruction and urinary obstruction (ICD9 codes 596_0 and 599_6, respectively), ureteral catheterization (ICD-9-CM procedural codes
59_8) and testosterone treatment. The average daily dose of finasteride and dutasteride were counted when patient finasteride
or dutasteride prescriptions prior to within 2 years of and at
index, and calculated the average exposure as total finasteride or dutasteride exposure (mg) divided by the period between within
2 years of and at index (per day). Finasteride or dutasteride average exposure is partitioned into two segments by median.
Medications were classified into three groups: none (control), dutasteride and finasteride.
Statistical analysis
We compared the distributions of demographic characteristics, medication use and comorbidity between the osteoporosis diagnosis and the nonosteoporosis diagnosis groups, using the chisquared test. We then calculated the odds ratio (OR: the ratio of
osteoporosis diagnosis risk of medication user to that for nonuser) and a 95% confidence interval (95% CI), using the multivariable logistic regression model. The multivariable models
were simultaneously adjusted for demographic characteristics, medication of alpha-adrenergic blockers, and comorbidities of hypertension, diabetes, hyperlipidaemia, COPD, obesity, bladder outlet obstruction and urinary obstruction, treatment of testosterone and ureteral catheterization. Models were also used for
estimating the risk between the dosages of dutasteride and finasteride for osteoporosis diagnosis.
Results
Table 1 shows a comparison of distributions of demographic characteristics, medication and baseline comorbidity between the osteoporosis diagnosis and the nonosteoporosis diagnosis groups. Among the 1352 patients with osteoporosis diagnosis, most were between 65 and 84 years of age (72_5%). The mean age was 74_0 _ 9_63 years for the osteoporosis diagnosis group and 73_8 _ 9_55 years for the nonosteoporosis diagnosis comparison group. The distributions of medication use between
patients with osteoporosis diagnosis (1_18% for dutasteride and 2_44% for finasteride) and patients with nonosteoporosis diagnosis (1_06% for dutasteride and 1_69% for finasteride) did not differ substantially between the two groups. Compared to
comparison group, medications were more common in the osteoporosis groups (P = 0_02). The osteoporosis diagnosis group was
more likely to have hypertension, diabetes mellitus, hyperlipidaemia and COPD compared with the nonosteoporosis diagnosis
group (all P-value <0_0001). The results of the multivariate logistic regression model of medication use associated with osteoporosis diagnosis risk in patients with BPH are shown in Table 2.
1_52-fold significant increase of osteoporosis diagnosis in patients with BPH using finasteride (95% CI, 1_01–2_30). The results of the medication dosage analysis are shown in Table 3. We observed that higher doses of finasteride are associated with a higher osteoporosis diagnosis risk (OR = 1_68; 95% CI, 1_01– 2_81), relative to the patients not using 5ARIs.
Discussion
Approximately 5% of the serum testosterone produced in men undergoes 5-alpha-reduction to form the more potent androgen DHT.12 During embryogenesis, DHT affects the development of
the male external genitalia. In adults, DHT affects the prostate and hair follicles.13 Two 5ARI isoenzymes have been observed in
humans.14 Type 2 5ARI is observed primarily in the prostate
and is inhibited by finasteride, which reduces the serum level of DHT by approximately 70%.15 Previous studies have shown that
the long-term administration of finasteride is not associated with bone mineral density, lipoprotein concentrations or overall
health.16,17 Finasteride has been used successfully in many men for the treatment of BPH 18
and male pattern baldness.19 Dutasteride
inhibits Type 1 5-alpha-reductase (located in the skin,
gut, liver and other tissues) and Type 2 5-alpha-reductase, causing a 95% decrease in serum DHT levels.20 Dutasteride is an
inhibitor of both 5-alpha-reductase enzymes and may be more potent than finasteride.21 It is less clear whether the greater
DHT suppression resulting from inhibition of both isozymes of 5-alpha-reductase by dutasteride has important effects on other androgen responsive tissues such as bone. Because low serum concentrations of androgens are associated with an increased risk of osteoporosis and fracture, therapy that decreases DHT has the potential to decrease BMD and increase fracture risk.22
In men with BPH, finasteride or dutasteride administration causes a marked decrease in DHT production, which leads to a 25% reduction in prostate volume, an improvement of the symptoms associated with BPH, and a significant reduction in the risk of acute urinary retention and the need for surgical
intervention.23–25 It is unclear whether the greater DHT suppression
resulting from the inhibition of both 5-alpha-reductase isozymes with dutasteride affects additional androgen responsive
tissues, such as bones, lipoprotein metabolism, haemoglobin, prostate-specific antigen (PSA) and sexual functions. Therapies that reduce DHT have the potential to reduce bone mineral
density and increase fracture risk because low serum concentrations of androgens are associated with an increased risk of osteoporosis and fractures.22 Amory et al. indicated that the
profound suppression of circulating serum DHT induced with finasteride and dutasteride in 1 year does not adversely affect bones, serum lipoproteins or haemoglobin.26 Circulating DHT
does not appear to play a clinically significant role in modulating bone mass, hematopoiesis or lipid metabolism in healthy
men. A case–control study found no positive association between use of finasteride and hip fracture and actually found some evidence of lower risk of fracture with finasteride use.27
However, we demonstrated a 1_52-fold significant increase of osteoporosis diagnosis in patients with BPH using finasteride (95% CI, 1_01–2_30). Furthermore, the results of the medication dosage analysis are shown in Table 3. We observed that higher
doses of finasteride are associated with an increased risk of osteoporosis diagnosis (OR = 1_68; 95% CI, 1_01–2_81) relative to
the patients not using 5ARIs. The combined use of alpha blockers and 5ARI is more effective in treating BPH.28
The strengths of this study include its use of populationbased data and the use of records from the NHIRD, rather than from self-reported drug use. In addition, Taiwan launched a
national health insurance (NHI) in 1995, operated by a singlebuyer, the government. All insurance claims should be scrutinized
by medical reimbursement specialists and peer review. The diagnoses of osteoporosis and BPH were based on the ICD-9 code determined by qualified clinical physicians under strict audit in the reimbursement process. Therefore, the diagnoses and codes for osteoporosis and BPH should be accurate
and reliable. However, misclassification of cases based on ICD9 codes alone without other more definitive criteria such as fractures, bone mineral density and/or osteoporosis medication (e.g.
bisphosphonate) use is possible. In addition, certain limitations should be mentioned. First, critical data are not included in the NHIRD, such as detailed demographic information, including
dietary factors, smoking habits, alcohol consumption, body mass index, socioeconomic status and family history of systemic diseases. These are major risk factors for osteoporosis diagnosis and may be indirectly associated with 5ARI use. However, because the NHIRD covers a highly representative sample of the general population in Taiwan and the reimbursement policy is universal, it is unlikely that these factors affected the prescription of 5ARIs. Second, the evidence derived from a nested
case–control study typically has a lower quality than that from randomized trials because a nested case–control study design is subject to numerous biases related to the adjustments for confounding variables. Despite our meticulous study design, which
featured adequate control of confounding factors, a critical limitation is that bias can remain if unmeasured or unknown confounders are present. Third, the diagnoses recorded in NHI
claims primarily serve the purpose of administrative billing and do not undergo verification for scientific purposes. We were unable to contact patients directly to inquire as to their use of
5ARIs because the anonymity of all beneficiaries listed in the NHIRD is assured. Furthermore, our analysis was unable to
consider prescriptions for 5ARIs issued before 1996. This omission could have caused an underestimation of the cumulative
dosage and may have weakened the observed association. However, the data for the prescription of the 5ARIs and the BPH
diagnoses were reliable.
Conclusion
This population-based nested case–control study suggests that the use of finasteride can increase osteoporosis diagnosis risk among patients with BPH. The effects were more prominent in patients using higher doses of finasteride.
