Increased risk of osteoporosis in
patients
with myasthenia gravis
A population-based cohort study
Myasthenia gravis (MG) is an acquired autoimmune disease and
corticosteroids are the standard
first-line therapy.
1,2Corticosteroid use is a well-documented risk factor for
osteoporosis,
3which
is also confirmed in corticosteroid-treated patients with MG.
4–6Prophylactic
medication for
osteoporosis is suggested for patients with MG at risk.
4In contrast,
absence of fracture risk
among patients with MG using oral glucocorticoids compared with
unexposed patients with
MG was recently reported using a national General Practice Research
Database in the United
Kingdom.
7Furthermore, the exact risk of osteoporosis for a
corticosteroid-naïve patient with
MG has not been comprehensively established among the small,
hospital-based studies reported.
We used a large cohort representing 99% of the population of Taiwan to
determine the risk of
osteoporosis in patients with MG.
METHODS Data source. The Taiwanese government initiated a universal single-payer health insurance program in 1995, which
comprises nearly 99% of the population of Taiwan. The National Health Insurance Research Database (NHIRD) contains the annual
original claims data for reimbursement and is maintained by the National Health Research Institutes (NHRI). To protect privacy, all
personal information was encrypted before release to researchers. The NHRI created scrambled and anonymous identification numbers
to link each patient’s information, including sex, birth date, and the registry of medical services. We used the registry for the Catastrophic Illness Patient Database (CIPD). The National Health
Insurance includes a catastrophic
illness program that exempts patients from copayments for the corresponding medical services, and the registry for the CIPD includes
major diseases such as cancers, type 1 diabetes, and MG. Disease was defined according to the ICD-9-CM. Standard protocol approvals, registrations, and patient
consents. This study was approved by the Ethics Review Board of China Medical University (CMU-REC-101-012).
Study population. We conducted a population-based retrospective cohort study from January 1, 2000 to December 31, 2011. The case cohort was constructed according to newly diagnosed MG (ICD-9-CM 358.0) during 2000–2005 from the CIPD, and the date of diagnosed MG was defined as the index date. We established a comparison cohort comprising patients without an MG diagnosis in the NHIRD who were 3-fold
frequency-matched by age (per 5 years) and sex and assigned the same index year. Patients in both cohorts with previous
osteoporosis before the index date were excluded from this study. The primary outcome was a new diagnosis of osteoporosis (including osteoporotic fractures) (ICD-9-CM 733.0 and 733.1) during the follow-up. Participants were followed from the index date to the date of osteoporosis diagnosis, the end date of the database (December 31, 2011), withdrawal from the insurance program, or death. Comorbidities were measured by the Charlson comorbidity index (CCI) score using the diagnoses recorded in the NHIRD before the index date.
Statistical analysis. The distributions of the comparison and MG cohorts were reported as the mean and SD for continuous variables and the number and percentage for categorical variables. To test the differences between these cohorts, we applied the t test for continuous variables and the x2 test for categorical variables.
We calculated the incidence rate of osteoporosis between the comparison and MG cohorts. We conducted Kaplan–Meier analysis to estimate the cumulative osteoporosis incidence curve
between these 2 cohorts. The log-rank test was performed to assess differences between these 2 incidence curves. We used multivariate Cox proportional hazards regression model to assess the effect of MG on the risk of osteoporosis, which was determined by the adjusted hazard ratio (HR) and the confidence
interval (CI).
We used SAS 9.3 software (SAS Institute, Cary, NC) to manage and analyze the data. The cumulative incidence curves were drawn using R software (R Foundation for Statistical Computing, Vienna, Austria). All p values were reported from 2-sided tests, and the level of statistical significance was set at 0.05.
RESULTS The study included a cohort containing 2,073 patients with MG and a cohort containing 6,219 individuals without MG with a similar average age (42 years) and sex ratio (58% women, table 1).
A total of 51% of the patients with MG used corticosteroids in the 6 months before the index date.
The proportion of CCI in the MG cohort was considerably higher than that of the comparison cohort
(p , 0.001).
Cumulative incidence curves of osteoporosis for the MG group and control group are shown in the figure. We observed a higher cumulative incidence of osteoporosis in patients with MG than in patients without MG (log-rank test, p , 0.001). The osteoporosis incidence rate was higher in the MG cohort
than in the comparison cohort (7.53 vs 4.08 per 1,000 person-years) (table 2). The MG cohort had a 1.96-fold (adjusted HR 5 1.96, 95% CI 1.57–
2.44) increased risk of developing osteoporosis compared with the comparison cohort after adjusting for
sex, age, and CCI score. The incidence of osteoporosis was higher than that of fracture. The MG cohort had HRs of 1.77 (95% CI 1.41–2.23) for osteoporosis events and 11.37 (95% CI 3.73–34.67) for fracture events compared with the comparison cohort. Sex-specific analysis indicated the incidence rate of osteoporosis in women and men with MG was 9.43 and 4.94 per 1,000 person-years, respectively—higher than that in the comparison cohort (5.15 and 2.57 per 1,000 person-years, respectively). In addition, women had 1.96-fold higher risk of osteoporosis (adjusted HR 5 1.96, 95% CI 5 1.51–2.54) and men had
1.94-fold higher risk of osteoporosis (adjusted HR 5 1.94, 95%CI 5 1.27–2.97) than the comparison cohort. We observed that patients with MG aged 30–44 years or with a CCI score of 2 or more had the greatest magnitude of osteoporosis risk compared with those without MG(HR 5 2.76, 95% CI 1.41–5.39; HR 5 3.32, CI 1.61–6.83, respectively).
Table 3 shows incidence rates and the HR for osteoporosis risk stratified by the severity of MG. Compared with the non-MG cohort, the risk of developing osteoporosis increased from 1.52 (95%
CI 1.11–2.08) for patients with MG not using corticosteroids to 2.37 (95% CI 1.82–3.07) for those
using corticosteroids (p for trend , 0.001). The risk of developing osteoporosis increased from 1.91 (95% CI 1.50–2.42) for patients with MG not using corticosteroid treatment to 2.17 (95% CI 1.44–3.27) for
those receiving corticosteroid treatment, compared with the comparison cohort (p for trend , 0.001).
DISCUSSION This study demonstrated the risk of osteoporosis in the largest MG cohort to date. This nationwide population-based cohort study provides evidence that MG is associated with a high risk of osteoporosis, regardless of corticosteroid use, in 2,073 Taiwanese patients with MG. The higher risk of osteoporosis from MG is likely to be
multifactorial, including physical inactivity caused by muscle weakness, lack of outdoor activity and sunlight exposure caused by disability, and comorbidity with thyroid disorder exhibited by approximately 10% of patients with MG in Taiwan.8
The disease severity reflected by the need for intensive treatment including plasmapheresis, thymecotomy, ventilator, and radiotherapy was associated with a higher risk of osteoporosis compared with patients treated with medication alone.
Patients with autoimmune disease may have a high risk of osteoporosis and fracture.9,10 Our results
substantiate these findings. However, a recently published English study using a similar cohort approach
as our study reported an absence of fracture risk
among patients with MG treated with oral corticosteroids compared with corticosteroid-free patients
with MG and a lower risk compared with control patients treated with oral corticosteroids.7 A primary
difference we observed between these 2 cohort studies was the patients’ ages. The mean age of the Taiwanese MG cohort was 41 years, nearly 20 years younger than the English cohort (mean 5 60 years). Therefore, the incidence rate of osteoporotic fracture in the English control cohort was 1.7-fold times that of the Taiwanese MG cohort (12.6 vs 9.5) and 2.3-fold times that of the Taiwanese control cohort (12.6 vs 5.6). Based on our previous comparative studies between Chinese and Caucasian patients with MG, Chinese patients with MGtended to have more earlyonset and ocular cases and fewer severe cases.11
Differences in immunoglobulin allotypes were also observed among patients with MGfrom various countries,
12 suggesting a possible racial difference. Therefore,
Taiwanese patients with MG are younger than Caucasian patients with MG.
Aging is an inherent risk factor for osteoporosis.13
However, in this study, patients with MG older than 30 years developed an increased risk of osteoporosis, with the highest risk in the 30–44 year age group, compared with the control cohort. For younger
patients affected by osteoporotic fracture, the psychosocioeconomic burden should be much higher than
that of aging-related osteoporotic fracture. Therefore,
careful screening of patients with MG at risk for osteoporosis and early prophylactic intervention might
improve the quality of bone health and prevent avoidable
osteoporotic fractures.14 Our study was subject to the following limitations.
First, the NHIRD does not provide detailed information on patients such as their smoking habits, alcohol
consumption, body mass index, physical activity, socioeconomic status, and family history of systemic diseases, which are all major risk factors for developing osteoporosis. Second, the evidence derived from a cohort study is generally of lower methodologic quality than evidence from randomized trials because a cohort study design is subject to numerous biases related to adjustments for confounding factors. Despite our meticulous study design with adequate control of confounding factors, bias could have remained because of possible unmeasured or unknown
confounders. Third, although the data we obtained on MG and osteoporosis diagnoses were highly reliable, the diagnoses in the National Health Insurance
claims are primarily for administrative billing and do not undergo verification for scientific purposes. We were unable to contact patients directly to obtain additional information because of the anonymity ensured by the identification numbers. Finally, because no available detailed and individual data of the z scores were shown from dual x-ray absorptiometry (for hip, spine) and ultrasound densitometry (for
heel) in NHIRD, we could not present z scores in our study.
This population-based retrospective cohort study provides evidence that MG is associated with high risk of osteoporosis, regardless of corticosteroid use. Cautiously screening patients with MG at risk for osteoporosis and early prophylactic intervention might improve the quality of bone health and prevent avoidable osteoporotic fractures.
