Risks and predictors of osteoporosis in patients with inflammatory bowel diseases in an Asian population: a nationwide population-based cohort study.

Risks and predictors of osteoporosis in patients with

inflammatory bowel diseases in an Asian population: a

nationwide population-based cohort study

M.-S. Tsai,

C.-L. Lin,

Y.-K. Tu,

P.-H. Lee,

C.-H. Kao

Introduction

Inflammatory bowel diseases (IBDs), including Crohn’s disease (CD) and ulcerative colitis (UC), are characterised by lifelong bowel inflammation and the consequences. Disease flare-ups require medical treatment and hospitalisation. It has been estimated that

approximately half of IBD patients show a substantial reduction in their bone mass during the disease course (1–5), either because of disease activities or medication, especially steroid. Conceptually, osteoporosis is a pathological condition characterised by

reduced density (mass/volume) of normally mineralised bone. The operational definition of osteoporosis

provided by World Health Organization is a bone mineral density 2.5 SD or more below the young adult mean value (T-score ≤ _2.5) (6). It is commonly believed that IBD can increase the risk of osteoporosis and disabling fracture (7–10), although it has been reported that IBD patients do not exhibit excessive risks of osteoporosis (11) or fracture (12,13).

Despite the epidemiological evidence of the association between IBD and osteoporosis, the IBD-specific

factors predicting osteoporosis and fracture development remain controversial, such as sex, steroid usage

and disease activity. Female IBD patients were

reported to have a higher risk of fracture (9) than male patients, whereas an animal study demonstrated more severe bone loss in male animals exhibiting colitis (14). Studies have shown that patients with CD have higher risks of fracture than do those with UC (8,9), whereas UC patients have higher bone turnover rates than CD patients do (3). The severity of IBD, assessed according to the number of symptoms, was shown to predict the risk of fracture (8).

In the present cohort study, we determined whether IBD is associated with increased risks of osteoporosis and pathological fracture in an Asian population by analysing the Taiwan National Health Insurance Research Database (NHIRD). We also

investigated the disease-specific predictors of osteoporosis and fracture in IBD patients. Methods

Data sources

The Taiwan National Health Insurance (NHI) programme was established in 1995 to provide affordable

healthcare to all residents of Taiwan. By end of 2009, the NHI programme covered nearly 99% of the 23.74 million residents and included 97% of all hospitals and clinics in Taiwan (14). The NHIRD consists of NHI programme reimbursement claims data. Details on the NHIRD have been published previously (15,16). Our study used the Longitudinal Health Insurance Database (LHID), a subset of the NHIRD. The LHID consists of historical claims data for one million patients randomly sampled from the entire insured population in 1996–2010. The National Health Research Institutes reported that there are no statistically significant differences in the distribution of sex, age or healthcare costs between cohorts in the LHID and the entire insured population.

The disease diagnoses were defined according to the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). This

study was approved by the Research Ethics Committee

of China Medical University Hospital.

Study participants

We identified patients aged 20 years and older with IBDs, namely UC (ICD-9-CM code 556) and CD (ICD-9-CM codes 555.0–555.2), who were newly diagnosed between 2000 and 2010, and refer to them as the IBD cohort. The date of IBD diagnosis was used as the index date to estimate the follow-up time. Patients who had been diagnosed at the baseline with osteoporosis (ICD-9-CM codes 733.0–

733.1) or for whom information on age or sex was missing were excluded from this study. For each IBD patient, four comparisons were randomly selected

from the pool of participants without IBD and osteoporosis at the baseline, frequency matched by the

year of index date, age (every 5-year span) and sex.

The subjects without IBD and osteoporosis were stratified by the index date of IBD cases. Non-IBD participants in each year stratum were further stratified by age in 5-year span: 20–24, 25–29, 30–34, 35–

39, 40–44, 45–49, 50–54, 55–59, 60–64 and 65+ years. Based on the specific age of each IBD case, four comparison subjects were randomly selected from non-IBD subjects with the appropriate age span and same sex for the non-IBD cohort. The flowchart of case selection in this study was presented in Figure 1.

Totally, this study included 3141 patients with IBD (including 1489 UC patients and 1652 CD patients); 12,564 subjects in the non-IBD cohort;

1652 (52.6%) male patients with IBD; and 6608 (52.6%) male patients without IBD. Among the 1489 UC patients, 80.7% was male. On the other hand, 73.9% was male patients in the 1652 CD patients (see also Table S1).

Outcome definition

To measure the incidence of osteoporosis, the IBD

cohort and the non-IBD cohort were followed up

until osteoporosis was diagnosed or censored because of mortality, loss to follow-up, withdrawal from the insurance system, or December 31, 2010. Comorbidities,

including previous diabetes (ICD-9-CM code 250), hyperlipidaemia (ICD-9-CM code 272), hypertension

(ICD-9-CM codes 401–405), chronic kidney

diseases (CKDs) (ICD-9-CM code 585), stroke (ICD- 9-CM codes 430–438), chronic obstructive pulmonary disease (COPD) (ICD-9-CM codes 490–496)

and cancer (ICD-9-CM codes 140–208), were defined before the index date.

Statistical analysis

The proportionate distributions of demographic status and comorbidity of the IBD and non-IBD cohorts

were compared and tested using the v2 test for categorical variables, and the differences were tested using

the t-test for continuous variables. To assess the difference in the cumulative incidence rates of osteoporosis

between the two cohorts, a Kaplan–Meier

analysis and log-rank test were conducted. The sex-, age- and comorbidity-specific incidence of osteoporosis per 1000 person-year of follow-up for each cohort was calculated. The Poisson regression model was used to measure the incidence rate ratio (IRR) and 95% confidence interval (CI) of osteoporosis for the IBD cohort compared with the non-IBD cohort. The multivariable Cox proportional hazard model was used to estimate the hazard ratio (HR) of osteoporosis and the 95% CI of patients with IBD compared with

the people in the non-IBD cohort. The multivariableadjusted model simultaneously included age, sex and

comorbidities, namely diabetes, hyperlipidaemia,

hypertension, CKD, stroke, COPD and cancer, as covariates.

Further analysis was performed to verify the

impact of IBD severity on osteoporosis. All statistical analyses were performed using the SAS package (Version 9.3 for Windows; SAS Institute, Inc., Cary, NC).

Results

The IBD cohort comprised 3141 patients and the non-IBD cohort consisted of 12,564 persons in the non-IBD cohort. Table 1 shows the demographic data of the IBD and non-IBD cohorts. The male sex was slightly predominant in both cohorts, and 56.4%

of patients were younger than 50 years of age. The mean ages were 48.9 _ 16.5 years in the IBD cohort and 48.5 _ 16.7 years in the non-IBD cohort. The

IBD cohort exhibited a higher prevalence of comorbidities, namely diabetes, hypertension, hyperlipidaemia,

CKD, stroke, COPD and malignancies (Table 1).

The median follow-up period was 6.49 _ 3.09 years for the IBD cohort and 6.46 _ 3.08 years for the non-IBD cohort. During the follow-up period, the overall incidence of osteoporosis was 40% greater in the IBD cohort than in the non-IBD cohort (7.16 vs. 5.13 per 1000 person-year), with an adjusted hazard

ratio (aHR) of 1.32 (95% CI, 1.09–1.60) (Table 1). The cumulative incidence of osteoporosis was 1.63% greater in the IBD cohort than in the

non-IBD cohort (6.06% vs. 4.43%; p < 0.001) by the end of the follow-up period (Figure 2).

We also compared the incidence of osteoporosis in both cohorts grouped according to sex, age and comorbidity (Table 2). As expected, the sex-specific analyses showed that women had a greater incidence of osteoporosis than men in both cohorts (10.6 vs.

3.93 per 1000 person-year for the IBD cohort; 7.42 vs. 2.92 per 1000 person-year for the non-IBD

cohort). Further analysis showed that IBD is associated with a significantly higher osteoporosis risk in

women (aHR, 1.36; 95% CI, 1.09–1.70), but not in men (aHR, 1.28; 95% CI, 0.89–1.82). Age-stratification analysis indicated that the incidence of osteoporosis increased with age in both cohorts. We

observed the highest IRR and aHR in the patients

aged 50–64 years. Only in this age group, the IBD

cohort exhibited significantly higher risks of osteoporosis

than the non-IBD cohort did. The risk of osteoporosis was not elevated in the patients either

younger than 50 years or older than 65 years of age.

The comorbidity-specific analyses showed that

patients with comorbidities had a higher risk of osteoporosis than did those without comorbidities. However,

in patients without comorbidities, the incidence of osteoporosis was 1.60-fold greater in the IBD cohort than in the non-IBD cohort (3.38 vs. 2.11 per 1000 person-year), with an aHR of 1.81 (95% CI, 1.23–2.67). By contrast, IBD was not associated with an excessive risk of osteoporosis in patients with comorbidities.

Patients with CD had a higher risk of osteoporosis than did patients with UC (7.35 vs. 6.39 events per 1000 person-year) (Table 3). Moreover, patients with CD were 1.33-fold more likely to develop osteoporosis than were the people in the non-IBD cohort

(95% CI, 1.33–1.64). However, we did not observe an increased risk of osteoporosis in the UC patients compared with those in the non-IBD cohort (aHR, 1.28; 95% CI, 0.87–1.89) (Table 3).

Disease severity, assessed according to hospitalisation for IBD, was related to the risks of osteoporosis

and pathological fracture (Table 4). Hospitalisation was associated with a significantly increased risk of osteoporosis with pathologic fracture (aHR, 17.1;

95% CI, 5.78–50.9) and without fracture (aHR, 4.46;

95% CI, 2.74–7.27).

Discussion

Based on a thorough review of relevant research, this study is the first to address the long-term risks and predictors of osteoporosis and pathological fracture in patients with IBD in an Asian population. This

study demonstrates that the long-term risk of osteoporosis is significantly elevated in patients with IBD

and identifies several predictors, namely the female

sex, an age between 50 and 65 years, an absence of

comorbidities, CD and hospitalisation for IBD.

Among these predictors, hospitalisation was associated with the highest risks of both osteoporosis

(aHR, 4.46; 95% CI, 2.74–7.27) and pathological fracture (aHR, 17.1; 95% CI, 5.78–50.9). Our results implied that disease severity, indicated by admission, significantly correlated with the development of osteoporosis and fracture. The disease-specific predictors

might help clinicians in differentiating between IBD patients with a low risk for bone disease from those with a high risk.

The exact mechanisms predisposing patients with IBD to osteoporosis are multifactorial, and the factors include vitamin D deficiency, systemic inflammation, malnutrition, the use of oral corticosteroids,

and decreased physical activity (10,17). However, it is difficult to determine the relative contribution of each factor, particularly for patient-related variables and corticosteroid use. Several studies have shown that steroid use is significantly related to reductions in bone mineral density (BMD) (4,5,18,19) and a risk of fracture (7,20) in IBD patients. By contrast, other studies have not supported the association between steroid use and reduced BMD (21,22). It has also been shown that risks of osteoporosis and fracture are not related to steroid use in IBD patients (8,23). In this study, hospitalisation significantly predicted osteoporosis and pathological fracture in IBD

patients. Because of the close relationship between

disease activity and corticosteroid use, distinguishing the impact of steroid use from the consequences of

chronic inflammation on bone health in IBD patients is difficult. Moreover, the hazardous effects of steroids on bone metabolism might be confounded by

their capability to reduce IBD activity.

Our results indicated that female, but not male,

IBD patients exhibited greater risks of osteoporosis

compared with those in the control cohort (Table 2).

The observed sex-related difference was consistent with that observed in several previous studies. A 6- year follow-up study reported a negative correlation between BMD and relapse of UC in women, but not in men (24). A population-based case–control study revealed that the 10-year probability of hip fracture was higher in women (7%) than in men (2.8%) aged 65 years with severe IBD (8). A recent study analysing a nationwide inpatient sample also concluded

that female IBD patients are associated with a higher risk of hospitalisation for fracture than male IBD patients are (25). The differences in risk related to sex might provide etiologic clues, indicating that further study is warranted.

It is unclear whether the risk of fracture differs

between patients with CD and those with UC. A previous population-based study showed similar

increases in the risk of fracture between patients with CD and those with UC (26). However, some studies have reported that the risks of osteoporosis and facture increased in patients with CD but not in those

with UC (9,11,27), which is in agreement with the findings of this study (Table 3). One possible explanation is the differing disease severity and systemic

inflammation between CD and UC. CD is often associated with a marked systemic inflammatory

response, which requires intensive corticosteroid therapy, whereas the inflammation of UC is limited to the colonic mucosa, and systemic inflammation is usually relatively not severe. No difference was observed in the risk of fracture between CD and UC patients after disease severity and steroid use were controlled (8).

This study has some limitations. First, the NHIRD does not provide detailed patient information on critical confounding variables such as smoking habits, alcohol consumption, body mass index (BMI),

physical activity, socioeconomic status, family history,

and medication use. In particular, the lack of information regarding smoking habits and BMI can bias the results (28). Second, we could not review charts and verify the diagnoses of IBD, osteoporosis, and fracture. Moreover, it is possible that we missed some patients with obscure fractures that can be identified only by using radiographic studies and cause little or no symptoms. By analysing clinically diagnosed fractures and not performing screening radiography, we might have underestimated the total number of radiographic fractures in both the IBD and non-IBD cohorts. However, these factors are not expected to have severely biased the results of this study, mainly because of the easy accessibility and

high coverage of the universal health insurance programme in Taiwan.

In summary, we determined that patients with

IBD have increased risks for osteoporosis and pathological fracture. Female sex, middle age, hospital

admission, and CD are predictive of developing osteoporosis and pathological fractures. Our findings are

helpful in ranking the osteoporotic and fracture risks

in IBD patients.