Increased risk of cardiovascular events in
end-stage renal disease
patients with osteoporosis: a nationwide
population-based cohort
study
T.-M. Yu & C.-L. Lin & K.-H. Shu & Y.-L. Liu & C.-H. Chen &
S.-T. Huang & C.-H. Kao
Introduction
Bone disease in patients with chronic kidney disease (CKD) is complex and comprises a broad spectrum of metabolic bone disorders. As renal function declines, uremic bone disorders increase in patients with chronic renal failure.
Osteoporosis is related to the aging process and is a major public health problem in the elderly. It has been reported that over 55 % of patients with CKD stages 3–5 are older than 70 years and that this number is still growing [1]. Therefore, it is anticipated that the incidence of osteoporosis will substantially increase in patients with CKD, and it will become a prevalent but easily neglected disorder in patients with ESRD [2].
Atherosclerosis is a common degenerative disorder in the general population as well as in those with osteoporosis. Atherosclerosis and osteoporosis share several risk factors such as aging, diabetes mellitus, hypertension, smoking, and a sedentary lifestyle [3]. Osteoporosis-related fractures and atherosclerosis-related cardiovascular diseases such as coronary artery disease (CAD), stroke, and congestive heart failure
(CHF) are common in the general population, particularly in the elderly [4]. Previous studies have reported a causal link between osteoporosis and atherosclerosis, which suggests that one is predictive of the other [3]. For example, Kado et al.
reported that an increased 1.3-fold risk of CAD was associated with women over 65 years old with decreased bone mineral density (BMD) of the femoral neck [5]. Several similar or
identical biological and molecular pathways have been identified that are involved with cellular components in bone and
vessel walls [3, 4]. Most importantly, vascular calcification and attenuated bonemineralization have been reported to play an important role in this process, which suggests cross-talk between bone and vessels [3].
Vascular calcification is considered to be fundamental for all cardiovascular diseases and is a marker of atherosclerosis severity which has been found to be inversely correlated with BMD [6].
It is well known that cardiovascular complications are the leading cause of mortality in patients with ESRD, and arterial calcification has been suggested to substantially contribute to the cardiovascular diseases in ESRD which is beyond the established conventional risk factors [7]. Arterial calcification is estimated to occur in 80–85%of dialysis patients, and 60% of incident dialysis patients have been reported to have coronary artery or aortic calcification to a certain extent [8, 9]. In
addition to the influence of the aging process on the development of arterial calcification, arterial calcification in ESRD
has been reported to be largely associated with renal
osteodystrophy [7]. Accumulating evidence shows that arterial calcification in ESRD is closely associated with the condition of adynamic bone disease or low bone activity, which reveals a similar condition closer to osteoporosis–vascular calcification [10–12]. Nevertheless, a variety of histological bone changes related to renal osteodystrophy could be found in ESRD patients with osteoporosis. It is likely that fromhighturnover bone lesions caused by uncontrolled hyperparathyroidism
to the profound depression of bone remodeling activity in adynamic bone disease could be coexisted in ESRD
patients with osteoporosis. The bone changes seen in uremic osteoporosis and renal osteodystrophy in ESRD patients are different and not always parallel within the spectrum of renal osteodystrophy [13].
Osteoporosis is a systemic metabolic bone disease of diverse causes. It is characterized by impaired bone strength
caused by attenuated bone mineral density and compromised bone quality which can leave patients vulnerable to fragility fractures [14]. It has been suggested that osteoporosis/ osteopenia in uremia should not only be a component of mineral bone disorders, but also a surrogate marker to reflect the final status of overall disturbances of the mineralization pathway related to nutritional and inflammatory factors which are closely associated with vascular calcification in ESRD [15]. Evidence directly comparing the association between osteoporosis and cardiovascular events in patients with ESRD is limited. The aim of this study was to investigate the association between osteoporosis and cardiovascular events in
patients with ESRD in a nationwide cohort study. Methods
Data source
The National Health Insurance (NHI) program in Taiwan is a nationwide insurance program with approximately 99 % of the 23.74 million population enrolled (http://www.nhi.gov.tw/
english/index.aspx). The NHI Research Database (NHIRD) is
an encrypted secondary database derived from the NHI program, maintained and updated by the National Health Research
Institutes. The NHIRD provides researchers with
scrambled identification numbers associated with the relevant claims information, including the patient’s gender, date of birth, registry of medical services, and medication prescriptions. The Registry of Catastrophic Illnesses Patient Database
(RCIPD) is a sub-dataset of the NHIRD containing health claims data for the treatment of catastrophic illness, which consists of 30 categories of diseases that require long-term care. The insured cases with major diseases such as cancer or ESRD can then apply for a catastrophic illness certificate. This study was approved by the Ethics Review Board of China Medical University (CMU-REC-101-012). The diagnostic
codes used in the NHIRD are in the format of the International Classification of Disease, 9th Revision, Clinical Modification (ICD-9-CM). Subject selection
To be registered in the RCIPD, both inpatients and outpatients are required to obtain approval from the insurance authority. From the RCIPD, we identified the patients with the diagnosis of ESRD (ICD-9-CM code 585) as Electronic supplementary
material Figure.
ESRD in patients was defined as eGFR<15 ml/min with long-term dialysis.
Patients with newly diagnosed osteoporosis (including osteoporotic fractures) (ICD-9-CM code 733.0 and 733.1) as
well as an age of 20 years and older in the period from1998 to 2011 were selected as the osteoporosis cohort.
The date of diagnosis of osteoporosis was defined as the index date. Patients with any history of CAD (ICD-9-CM codes 410–414), CHF (ICD-9-(ICD-9-CM code 428), and stroke (ICD-9-CM codes 430–438) before the index date were excluded. The non-osteoporosis comparison cohort consisted of patients randomly selected from the rest of the patients with ESRD without a history of osteoporosis as follows. For each osteoporosis case, we randomly selected two comparison subjects frequency matched for gender, age (every 5-year span), year of diagnosis of ESRD, and year of index date, excluding those with a history of osteoporosis, CAD, CHF, and stroke. The patients with any cancer (ICD-9-CM 140–208) before the index date were also excluded from both cohorts. Definitions of end-point, comorbidities, and co-variables The endpoint of interest in this study was the development of cardiovascular events (including CAD, CHF, and stroke) or mortality. All subjects were followed from the index date to the date when the endpoint occurred, withdrawal from the NHI program, death, or the end of 2011, whichever came first. The comorbidities that were considered conventional risk factors before the index date included hypertension
(ICD-9-CM codes 401–405), diabetes (ICD-9-(ICD-9-CM code 250), hyperlipidemia CM code 272), mental disorders
(ICD-9-CM codes 290–319), hepatitis B infection (ICD-9-(ICD-9-CM codes V02.61, 070.20, 070.22, 070.30, 070.32), and hepatitis C infection (ICD-9-CM codes V02.62, 070.41, 070.44, 070.51,
070.54).
Statistical analysis
We compared the demographic data and comorbidities between the patients with ESRD with and without osteoporosis
using the chi-square test. The mean ages were compared using
the ttest. The gender- and age-specific incidence rates of cardiovascular events or mortality per 100 person-years of
follow-up for each cohort were calculated. The osteoporosisto-non-osteoporosis relative hazard ratios (HRs) and 95 %
confidence interval (CIs) were estimated by gender and age using a Cox proportional hazards regressions model. Multivariate Cox proportional hazards regressionwas used to assess
the risk of cardiovascular events or mortality associated with osteoporosis, adjusting for age and gender, and the
comorbidities of hypertension, diabetes, hyperlipidemia, mental disorders, hepatitis B infection, and hepatitis C infection.
We used the Kaplan–Meier method with the log-rank test to compare the cumulative incidence curves of cardiovascular events or mortality between the two cohorts. The Kaplan– Meier survival curve was plotted using R software (version 2.14.1; R Development Core Team, Vienna, Austria). All
analyses were performed by SAS statistical software (version 9.2 for Windows; SAS institute, Inc., Cary, NC, USA). The
results were considered statistically significant when twotailed pvalues were less than 0.05.
Results
A total of 12,535 ESRD patients were enrolled in the cohort study, 4,153 (33.13 %) of whom had a diagnosis of osteoporosis. Table 1 shows the baseline demographic characteristics
and comorbidity status. There were more women than men (71 vs. 29 %). The average ages of the non-osteoporosis and osteoporosis groups were 58.5 (SD=14.2) and 59.8 (SD=
13.5) years, respectively. The duration of diagnosis with osteoporosis after ESRD was 4.37 (SD=3.59) years. The osteoporosis
group had more comorbidities than the nonosteoporosis group, including hypertension (70.6 vs.
66.4 %), hyperlipidemia (25.1 vs. 23.3 %), mental disorders (37.4 vs. 27.8 %), and hepatitis C infection (8.26 vs. 5.56 %).
Table 2 shows the incidence densities and HRs of cardiovascular events and mortality by gender and age. Overall, the
osteoporosis group had a higher incidence of CAD (4.10 vs. 3.00 per 100 person-years), CHF (1.96 vs. 1.53 per 100 person-years), stroke (2.90 vs. 2.47 per 100 person-years), and mortality (8.42 vs. 7.60 per 100 person-years) than the non- osteoporosis groups with crude HRs of 1.36 (95 % CI= 1.24–1.49), 1.28 (95% CI=1.13–1.46), 1.18 (95 % CI=1.06– 1.30), and 1.11 (95 % CI=1.05–1.17), respectively. After adjusting for age, gender, and related comorbidities of
hypertension, diabetes, hyperlipidemia,mental disorders, hepatitis B infection, and hepatitis C infection, the osteoporosis
group was associated with a significantly higher risk of CAD (HR=1.32, 95 % CI=1.20–1.45), which was significant in both genders (women, HR=1.35, 95 % CI=1.20–1.50; men, HR=1.27, 95 % CI=1.06–1.52) and age (≤49 years, HR= 1.41, 95 % CI=1.16–1.70; >49 years, HR=1.30, 95 % CI= 1.16–1.45). The osteoporosis group had a 1.26-fold (95%CI= 1.10–1.43) increased risk of developing CHF. Female gender (HR=1.31, 95 CI=1.13–1.52) and age (≤49 years, HR=1.43, 95 % CI=1.10–1.86; >49 years, HR=1.24, 95 % CI=1.07– 1.44) were associated with a significantly higher risk of CHF. The risk of stroke associated with osteoporosis had an adjusted overall HR of 1.13 (95%CI=1.02–1.26) and was higher in
the male patients (HR=1.31, 95 % CI=1.08–1.58) and in those older than 49 years of age (HR=1.16, 95 % CI=1.04– 1.30).
Compared with the non-osteoporosis group, the osteoporosis group was significantly associated with mortality (HR= 1.07, 95%CI=1.01–1.14), especially the male patients (HR= 1.26, 95 % CI=1.14–1.40) and age (≤49 years, HR=1.33, 95 % CI=1.13–1.57; >49 years, HR=1.10, 95 % CI=1.04– 1.17).
The results of multivariate Cox proportional hazard analysis for the risk of related variables contributing to cardiovascular events or mortality are shown in Table 3. There was a 1.32-fold increased risk of CAD (95 % CI= 1.21–1.45) in the patients with ESRD with osteoporosis.
The other risk factors contributing to CAD included male
gender (HR=1.13, 95 % CI=1.02–1.25), increasing age (every one year 95 % CI=1.01–1.02), diabetes (HR=1.33,
95 % CI=1.18–1.50), hypertension (HR=1.27, 95 % CI= 1.14–1.41), hyperlipidemia (HR=1.19, 95 % CI=1.06– 1.32), and mental disorders (HR=1.14, 95 % CI=1.04– 1.26). The risk of developing CHF was 1.26-fold higher (95 % CI=1.10–1.43) in the osteoporosis group and was associated with increasing age (every year, 95 % CI= 1.01–1.02) and hypertension (HR=1.22, 95 % CI=1.06– 1.40).
Osteoporosis (HR=1.13, 95 % CI=1.02–1.26), male gender (HR=1.16, 95 % CI=1.04–1.29), older age (HR=1.04, 95 % CI=1.04–1.05), diabetes (HR=2.24, 95 % CI=1.99– 2.51), and hypertension (HR=1.35, 95 % CI=1.19–1.52)
were associated with an increased risk of stroke in the patients with ESRD. Osteoporosis (HR=1.07, 95 % CI=1.01–1.14), male gender (HR=1.20, 95 % CI=1.12–1.27), older age (HR=1.06, 95 % CI=1.06–1.07), diabetes (HR=2.04, 95 % CI=1.91–2.18), and hepatitis C infection (HR=1.26, 95 % CI=1.12–1.41) were associated with an increased risk of mortality.
We next explored the joint effect of osteoporosis and diabetes on cardiovascular events and mortality (Table 4). A
significantly higher risk was observed in the patients with both osteoporosis and diabetes (HR=1.84, 95 % CI, 1.58– 2.15 for CAD; HR=1.38, 95 % CI, 1.13–1.68 for CHF; HR= 1.71, 95 % CI, 1.44–2.03 for stroke; HR=1.72, 95 % CI, 1.33–2.22 for mortality) than in those without osteoporosis and diabetes.
Table 5 shows the incidence rates and HRs of the
different outcomes between the patients with either osteoporosis or pathological fractures. The patients with osteoporosis
were 1.33–, 1.26-, and 1.13-fold more likely to develop CAD (95 % CI=1.20–1.46), CHF (95 % CI= 1.11–1.44) and stroke (95 % CI=1.13, 95 % CI=1.02– 1.26), respectively, than those without osteoporosis or
1.72-fold higher risk of mortality
(95 % CI=1.45–2.04); however, this was not found in the patients with osteoporosis without fractures.
Figure 1a to d shows that, compared with the nonosteoporosis group, the osteoporosis group had significantly
higher cumulative incidence rates of CAD (p<0.001) (Fig. 1a), CHF (p<0.001) (Fig. 1b), stroke (Fig. 1c) (p= 0.002), and mortality (Fig. 1d) (p<0.001).
Discussion
To the best of our knowledge, this is the first study to report a causal link between osteoporosis and the clinical consequences of cardiovascular diseases in patients with ESRD.
The results showed that osteoporosis was significantly associated with the subsequent risk of cardiovascular events in
patients with ESRD in both gender and age.
The concept of a bone–vascular axis has recently gained
increasing attention, in that bone and vessels may share common pathophysiological pathways and interplay with each
other. Eventually, bone and vascular disorders result in bone
and vascular diseases such as osteoporosis and vascular calcification [16]. Osteoporosis can also be considered to be a
sign of compromised circulation with attenuated perfusion caused by arterial ischemia within the bone in patients with atherosclerosis [17]. Compared with the general population, a larger number of detrimental factors may aggravate the axis between bone and vessels in the setting of ESRD, including diabetes, dyslipidemia, uremic toxins, inflammation, oxidative stress, vitamin D deficiency, and mineral metabolism
disorders, and these may ultimately worsen vascular calcification and osteoporosis in patients with ESRD. Patients with
ESRD have been assumed to undergo a “perfect storm” of vascular calcification [16].
Our findings showed that, after adjusting for confounders including age, gender, diabetes mellitus, hypertension, and hyperlipidemia, osteoporosis was associated with a 1.32-fold increase risk of CAD in the patients with ESRD. A similar trend was also observed for CHF and stroke. In addition, the ESRD patients with osteoporosis were associated with more
underlying diseases including hypertension, hyperlipidemia, mental disorders, and hepatitis C virus infection compared with the patients without osteoporosis. It is noteworthy that our data disclosed that osteoporosis was significantly correlated with cardiovascular events and mortality in the ESRD
patients who were younger (≤49 years). In addition, diabetes mellitus has been reported to be an important risk factor contributing to cardiovascular events in patients with ESRD and also to be associated with osteoporosis in the general population. The effect of osteoporosis and diabetes on the
patients with ESRD was determined the current study, and our results showed that the patients who had both were at a
higher risk of cardiovascular events. In contrast to the findings of osteoporosis in the general population, our results disclose more distinct association between osteoporosis and cardiovascular diseases in ESRD patients and suggest that factors other
than the traditional risk factors may be involved in uremic osteoporosis.
Uremic osteoporosis should be considered a systemic disorder adversely affected by uremic toxins through disabling bone material properties and bone elasticity, and this cannot be simply explained as the consequences of mineral bone disorder [18]. Previous study
has shown that uremic osteoporosis is frequently associated with some unfavorable conditions such as low
body weight, general health status, dialysis adequacy, and nutritional status in patients with ESRD [19]. Matsubara et al. compared the BMD of 277 incident dialysis patients and found that both male and female patients with a lower BMD (<1.09 g/cm [2]) were associated with protein-energy wasting and CVD and suggested that osteoporosis may be a surrogate marker of factors which could adversely affect mortality [15]. Moreover, in a study on the relationship between vascular calcification and BMD in patients with CKD,
Toussaint et al. reported a strong inverse correlation between BMD and vascular calcification [20]. These
and extra-osseous vascular calcification in patients with ESRD, which is relevant to our findings.
In the present study, the more distinct findings in our cohort may be because our patients with osteoporosis had a BMD T score of −2.5 or lower, meaning a worse osteoporosis status.
The measurement of BMD could be falsely increasing in patients with ESRD because of the confounding effect of coexisting metastatic calcifications which was commonly found in ESRD with renal osteodystrophy [4, 11].
The gold standard for the determination of renal
osteodystrophy is bone biopsy; however, this is not easily accessible. Instead, non-invasive measurements of wholebody BMD through dual energy X-ray absorptiometry are
easier to obtain in daily practice.
The strengths of our study are its population-based design, generalizability of findings, and use of
population-based data and NHIRD records with a very large sample size, including study and control cohorts. In addition, NHIRD covers a highly representative sample of Taiwan’s general population because the reimbursement policy is universal and operated by a single-buyer,
the government in Taiwan. All insurance claims should be scrutinized by medical reimbursement specialists and peer review. Therefore, the diagnoses of osteoporosis and ESRD in this study were highly reliable. However, several limitations inherent to our database should be mentioned. First, the diagnoses recorded in the NHIRD primarily serve the purpose of administrative billing and have not undergone verification for scientific purposes. Second, the database does not contain information regarding the daily lives or behaviors of the patients, including information on smoking habits, alcohol consumption, body mass index, socioeconomic status, and family history. Third,
there were lacking of personal laboratory, pathologic, and imaging data in NHIRD; therefore, we could not
evaluate the diseases’ severities for ESRD and osteoporosis. Finally, despite our meticulous study design in which
the confounding factors were adequately controlled, data derived from a retrospective cohort study are generally of lower statistical quality than those derived from randomized trials because of potential biases.
In conclusion, osteoporosis is significantly correlated with cardiovascular diseases in ESRD patients.When encountering patients with ESRD and osteoporosis, physicians should be alert to the cardiovascular risk factors in incident dialysis
patients to prevent the subsequent occurrence of these adverse events.
