Relation Between Hepatitis C Virus Exposure and Risk of Osteoporosis
A Nationwide Population-based Study
Chien-Hua Chen, MD, MPH, Cheng-Li Lin, MSc, and Chia-Hung Kao, MD
INTRODUCTION
A negative balance between the formation and resorption of bone mass can lead to the development of osteoporosis.
Moreover, osteoporosisis characterized by skeletal fragility
resulting from reduced bone mass and disrupted bone microarchitecture.
Osteoporosis increases the risk of fracture and thus
has been considered a major public health concern.1 Traditionally, the development of osteoporosis is related to several risk
factors, such as aging, immobility, hypertension, antihypertensive agents, hyperparathyroidism, menopause, diabetes mellitus, corticosteroid usage, low calcium intake, vitamin D
deficiency, and genetic vulnerability.2
Osteoporotic fractures mainly consist of vertebral fracture
and hip fracture. Vertebral fracture is the most common osteoporotic fracture, but only from one-third to one quarter of the
patients with vertebral fracture can be clinically identified.3–5 Furthermore, it is reported that asymptomatic vertebral fractures are associated with future hip fracture by threefold and
other nonvertebral fracture by twofold.6 Hip fracture is the
second common osteoporotic fracture and it may incur substantial healthcare costs resulted from disability. Furthermore,
the effect of hip fracture on mortality increase can adversely extend up to 10 years or more.3 The mortality following a hip fracture generally increases with the increment of age and is greater in men, but the sex difference declines after age 80.7 Although the rates of hip fracture have declined in the West, the rate is increasing in the developing world. It is estimated that
>50% of hip fractures worldwide will occur in Asia by 2050.8 The functional outcome and increased mortality of osteoporotic
fracture are heterogeneous and depend on age, activity of daily living prior to fracture, pre-fracture comorbidities, and the cognition.3,9,10
Hepatitis C virus infection (HCV) is a global health
problem estimated to affect 170 million people worldwide.11
Hepatitis C virus infection is a hepatotropic virus that mainly causes inflammation and fibrosis of the liver. It is reported that
_20% of HCV-infected patients will progress to liver cirrhosis.
12 The late hepatic sequelae include chronic hepatitis, cirrhosis, and even hepatocellular carcinoma. However, HCV can
also cause several extrahepatic manifestations, such as diabetes
mellitus, rheumatic disorders, lymphoproliferative disease, cardiovascular events, and cognitive impairment.13 Although the
role of osteoporosis as a sequence of cirrhosis or advanced liver disease has been thoroughly documented, the effect of HCV exposure on bone mineral density in the absence of advanced liver disease remains debated.14 Some scholars have proposed that chronic HCV infection without liver cirrhosis contributes to reduced bone mineral density, whereas other scholars have asserted the opposite.15–20
To assess the association between HCV exposure and subsequent development of osteoporosis, we conducted a nationwide population-based cohort study by analyzing data from a nationwide medical database, the National Health Insurance Research Database (NHIRD).
METHODS Data Source
The National Health Insurance (NHI) program, initiated on March 1, 1995, provides comprehensive coverage for the medical care of Taiwan residents. At the end of 2014, 23.75 million people (_99.9% of the population) were enrolled in the program.21 In cooperation with the Bureau of National Health Insurance (BNHI), the National Health Research Institutes established several data sets for public use, including the Longitudinal Health Insurance Database 2000 (LHID2000), a cohort data set comprising 1,000,000 randomly selected cases from the registry of NHI beneficiaries in 2000. To maintain
confidentiality, personal information, such as patient identification
numbers and sensitive personal data, are encrypted
in the database, and International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) codes are used for disease classification. This study was approved to fulfill the condition for exemption by the Institutional Review Board (IRB) of China Medical University (CMUH- 104-REC2-115). The IRB also specifically waived the consent requirement.
Participants
From 2000 to 2011, patients aged 20 years and older with diagnosed HCV infection (ICD-9-CM codes 070.41, 070.44, 070.51, and 070.54) identified from the LHID2000 comprised the HCV infection cohort. The date for HCV exposure coding was designated as the index date. Patients with a history of osteoporosis (ICD-9-CM codes 733.0 and 733.1) and hepatitis B virus (HBV) infection (ICD-9-CM codes 070.20, 070.22,
070.30, and 070.32) diagnosed before the index date, those with missing information, and those younger than 20 years were excluded. Using 1:m case-control studies is to increase the power and to control possible confounding. Based on the statistical efficiency does not gain much when m>4, we constructed a 1:4 matched cohort study. For each HCV case, 4 insurers with no history of HCVexposure, HBVinfection, and osteoporosis were assigned to a comparison cohort and frequency matched with the HCV exposure cohorts according to
age (every 5-year span), sex, and index year. Individuals were excluded from the comparison cohort using the same criteria used for the HCV exposure cohort.
Outcome and Comorbidities
The primary endpoint in this study was defined as the
diagnosis of osteoporosis. Each participant was followed from the index date until the endpoint, withdrawal from the NHI program, or December 31, 2011. The baseline characteristics of participant comorbidities were also analyzed; the comorbidities included diabetes (ICD-9-CM code 250), hypertension (ICD-9- CM codes 401–405), hyperlipidemia (ICD-9-CM code 272), heart failure (ICD-9-CM code 428), stroke (ICD-9-CM codes 430–438), obesity (ICD-9-CM code 278), and cirrhosis (ICD-9-
CM codes 571.2, 571.5, and 571.6).
Statistical Analyses
The chi-square test and t test for categorical and continuous variables, respectively, were first used to compare the distributions of age, sex, and baseline comorbidities between the
HCV exposure and the comparison cohorts. The incidence densities of osteoporosis were estimated in person-years for the various risk factors. To estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) of osteoporosis, the univariate and multivariate Cox proportional hazards regression model was used. Multivariable models were simultaneously adjusted for age, sex, and the comorbidities of diabetes, hypertension, hyperlipidemia, heart failure, stroke, and cirrhosis. Kaplan–
Meier estimates were plotted to illustrate the cumulative incidence of osteoporosis, and the log-rank test was performed to
examine the difference between the HCV exposure and the comparison cohorts. All statistical analyses were performed using the SAS package (Version 9.4 for Windows; SAS institute, Inc, Cary, NC). A 2-sided p value<.05 was considered
statistically significant.
RESULTS
Of the 51,535 sampled patients, 41,228 and 10,307 were categorized as the comparison and HCV exposure cohorts, respectively (Table 1). Most patients were aged _50 years (61.7%), and 54.6% of the patients were women. The mean age was 54.1_15.3 years in the HCV exposure cohort and 53.7_15.5 years in the comparison cohort. Regarding baseline characteristics, the HCV exposure cohort exhibited a higher prevalence of diabetes, hypertension, hyperlipidemia, heart failure, stroke, obesity, and cirrhosis than did the comparison cohort. During the mean follow-up periods of 5.44 and 6.01 years for the HCV exposure and comparison cohorts, respectively, the Kaplan–Meier curve revealed that the cumulative
incidence of osteoporosis was higher in the HCV exposure cohort than in the comparison cohort (Figure 1, log-rank test P<0.001).
The overall incidence of osteoporosis in the HCV exposure
cohort was higher than that in the comparison cohort (8.27 vs 6.19
per 1,000 person-years; crude HR¼1.33, 95% CI¼1.20–1.47) (Table 2). After we adjusted for factors such as age, sex, and comorbidities, namely diabetes, hypertension, hyperlipidemia, heart failure, stroke, and cirrhosis, the risk of developing osteoporosis was significantly higher in theHCV exposure cohort than
in the comparison cohort (adjusted HR [aHR]¼1.35; 95%
CI¼1.21–1.51). Compared with patients aged _49 years, the risk of developing osteoporosis was 4.05-fold higher in those aged 50 to 64 years (95%CI¼3.845–4.760) and 8.82-fold higher in those aged 65 years or older (95% CI¼7.48–10.40). In the multivariate model, the risk for osteoporosis was 3.10-fold higher for women than for men (95% CI¼2.81–3.43) and was higher
for patients with the comorbidities of hypertension (aHR¼1.19, 95% CI¼1.07–1.31), hyperlipidemia (aHR¼1.17, 95% CI¼
1.05–1.29), and heart failure (aHR¼1.23, 95%CI¼1.02–1.49).
The incidence of osteoporosis increased with age, was
higher in women than in men, and increased with comorbidity in both cohorts (Table 3). The overall risk of osteoporosis related
to several variables including age, sex, and presence of comorbidities was compared in the HCV exposure cohort and the
comparison cohort. The risk of osteoporosis in patients exposed to HCV in all stratifications was higher than that in the
comparison cohorts. However, the risk of osteoporosis contributed by HCV decreased with age (aged _ 49: aHR¼1.79, 95%
CI¼1.32–2.43; aged 50–64: aHR¼1.36, 95% CI¼1.14–
1.62; aged _ 65: aHR¼1.23, 95% CI¼1.05–1.44) and the presence of comorbidity (no comorbidity: aHR¼1.54, 95%
CI¼1.26–1.89; comorbidity: aHR¼1.27, 95% CI¼1.12–
1.43).
The patients exposed to HCV exhibited a 1.38-fold (95%
CI¼1.24–1.55) higher risk of developing osteoporosis compared with the patients who were not exposed to HCV (Table 4).
The risk of osteoporotic fracture did not differ significantly between patients exposed to HCV and the comparison cohorts (aHR¼0.80, 95% CI¼0.44–1.45).
DISCUSSION
Consistent with the literature proposing that HCV seroprevalence peaks after age 55 and that women are predisposed
to HCV infection, our results revealed that most patients (61.7%) were aged _50 years and that 54.6% of the patients were women.22,23 The mean age in the HCV exposure cohort was 54.1_15.3 years. Compared with patients who were not exposed to HCV, patients who were exposed to HCV tended to have more comorbidities, including diabetes, hypertension, hyperlipidemia, heart failure, stroke, obesity, and cirrhosis.
Our results revealed that the HCV exposure cohort had more comorbidities; however, the risk of osteoporosis remained higher in the HCV exposure cohort after adjusting for age,
sex, and the comorbidities of diabetes, hypertension, hyperlipidemia, heart failure, stroke, and cirrhosis. The mechanism
affecting the pathophysiology that causes comorbidities in patients exposed to HCV may include HCV-associated insulin resistance and atherosclerosis.13 HCV may cause diabetes mellitus by directly interfering with insulin signaling and
inducing insulin resistance in hepatocytes; HCV-infected hepatocytes can secrete mediators that induce extrahepatic insulin
resistance, notably in the skeletal muscle. Hepatitis C virus may increase the risk of cardiovascular and cerebrovascular events through atherosclerosis induced by systemic inflammation, chronic endothelial injury, or direct infection of the arterial wall.24–26
Our study revealed that the osteoporosis incidence increased with age and was higher in women than in men.
The association between osteoporosis and aging has been confirmed in the literature, and the gene expression of the rennin–angiotensin system independent of hypertension in the skeletal tissue may contribute to osteoporosis.27,28 Estrogen deficiency and aging are the major etiologies of primary
osteoporosis. Estrogen can protect against osteoporosis by inhibiting osteoblast apoptosis and increasing osteoblast lifespan.
29,30 The decreasing rate of bone mineral density is swifter in the early postmenopausal period, which typically begins after age 50, and women aged 40 to 59 years have the highest risk of developing osteoporosis.23,30,31 Consistent with our result that women are predisposed to osteoporosis, the US Preventive
Services Task Force indicated that as many as 1 in 2 postmenopausal
women and 1 in 5 men are at risk for osteoporosisrelated fracture.32
Moreover, in the present study, osteoporosis was associated with hypertension, hyperlipidemia, and heart failure. Both osteoporosis and hypertension are common among the aging population and may share similar etiologies, such as low
calcium intake and levels, vitamin D and vitamin K deficiency, high salt consumption, imbalanced nitric oxide levels, and
antihypertensive agents that exert detrimental effects on the skeletal metabolism, strength, and density.33 Hyperlipidemia can reduce bone formation and promote bone loss by causing the products of lipid oxidation to accumulate in the subendothelial spaces of the vasculature and bone. Moreover, hyperlipidemia can induce secondary hyperparathyroidism, thereby
impairing bone regeneration and mechanical strength.34 Heart failure and osteoporosis share several risk factors, such as aging, smoking, and postmenopausal and antihypertensive agents;
heart failure can also accelerate bone loss by inducing hyperaldosteronism and secondary hyperparathyroidism.35 However,
the effect of obesity on bone metabolism is controversial.36 Obesity is traditionally regarded as a protective factor for osteoporosis, conferring a positive mechanical loading on bone formation. Nevertheless, evidence supports the deteriorating effect of obesity on osteoporosis. For example, both osteoblasts and adipocytes are derived from a common mesenchymal stem cell and agents inhibiting adipogenesis-stimulated osteoblast differentiation and vice versa. Furthermore, the reduced bone formation caused by aging is usually accompanied by adipogenesis in bone marrow cavities. Moreover, elevated oxidative
stress is common in people with obesity and osteoporosis.
To our knowledge, this is the first population-based study
to assess the relation between HCV exposure and the incidences of osteoporosis and osteoporotic fracture.15–20 Our statistical analyses benefited from the use of a nationwide database and
the 12-year observation of participants selected from a representative cohort comprising 1,000,000 residents covered by the
NHI program. Hansen et al conducted a large-scale populationbased study to explore the association between HCV exposure
and all-site fracture, but omitted discussing osteoporosis development.
They concluded that the risk of fracture equally
increased in patients exposed to HCV (chronic or cleared infections), and the major determinants of fracture in such
patients are lifestyle-related factors, such as alcohol and drug abuse, which substantially increases fracture risk.20 By contrast, our epidemiological study demonstrated that HCV exposure increases the risk of osteoporosis and the detrimental effect of HCV on osteoporotic fracture was not obvious.
Several pathogenic mechanisms are involved in bone mineral density loss in patients exposed to HCV.37 First, fibronectin can infiltrate the bone matrix to enhance matrix mineralization, reducing its production by the liver. However, oncofetal fibronectin increases and can directly inhibit osteoblast function. Second, insulin-like growth factor 1, involved in
osteoblast differentiation and proliferation, is produced by the liver and is reduced in patients with chronic liver disease. Third, the receptor–activator ratio of nuclear factor kappa ligand and osteoprotegerin is higher in patients with chronic liver disease, enhancing bone resorption. Fourth, interleukin-6 is increased by chronic HCV infection and can activate osteoclasts to increase bone resorption. Fifth, hypogonadism in chronic liver disease can result in increased osteoclast activity. Finally, bilirubin elevation can inhibit in vitro osteoblast proliferation.38 By contrast, the association between HCV and osteoporosis may be due to shared risk factors since the prevalence of important osteoporosis risk factors was higher in the HCV exposed patients as compared to the patients without HCV exposure. However, it is reasonable to conclude that the increased risk of osteoporosis observed in HCV exposed patients was more likely to be due to the effect of their HCV status since the possible confounding effect of osteoporosis risk factors has been significantly minimized in our study. Moreover, the osteoporosis risk contributed by HCV decreased with the increment of age may be due to the absence or low prevalence of osteoporosis-associated comorbidities in the younger patients. The results of our subgroup analyses in which we excluded patients with important comorbidities at baseline
also confirmed the validity of our results. This finding, coupled with the results of the subgroup analyses, affirms the possible causal association between HCV and osteoporosis, and suggested that HCV may be a possible risk factor for osteoporosis.
Nevertheless, HCV may be less influential than other risk factors, such as hypertension, hyperlipidemia, and heart failure, in contributing to the development of osteoporosis.
The present study had some strength. First, the large-scale national database provided statistical benefits to our longitudinal study to evaluate the association between HCV and osteoporosis.
Second, the recruited subjects were a stable population and _99% of the residents in Taiwan have been covered by the NHI program. Nevertheless, our study had several limitations.
First, we could not ascertain the patients’ viremic status, such as the viral loads or the genotypes of the HCV. Nevertheless, all patients in our case cohort exhibited positive anti-HCV antibodies, which mean recent or past HCV exposure. Moreover,
the effect of antiviral therapy on the stage of liver fibrosis or the grade of liver inflammation could not be ascertained in our study. Therefore, we could not prove the ameliorating effect of antiviral therapy on bone mineral density through the arrest of liver cirrhosis even though the association between osteoporotic fracture and cirrhosis has been thoroughly documented. Second, osteoporosis-associated lifestyle factors could not be investigated in this study. However, potential osteoporosis-associated
comorbidities were confounded in our study. Third, validating the diagnosis of osteoporosis or osteoporotic fracture was difficult. However, to enhance the accuracy of diagnosis, we only recruited patients who received medical care for osteoporosis >3 separate visits. Moreover, the BNHI organizes regular audits performed by medical experts to ensure the accuracy of insurance claim codes in Taiwan. Finally,
we cannot clarify the temporal association between HCV exposure and osteoporosis since the date of HCV exposure
could not be ascertained.
In conclusion, this nationwide population-based cohort study concludes that HCV exposure increases the risk of developing subsequent osteoporosis, but no detrimental effect
on osteoporotic fracture was observed. Furthermore, HCV may be less influential than other risk factors, such as hypertension, hyperlipidemia, and heart failure, in contributing to the development of osteoporosis.
