INTRODUCTION
Ossification of the posterior longitudinal ligament (OPLL) is a multifactorial condition characterized by abnormal bone formation, which results in the replacement of spinal ligamentous tissue13,15,16,18,20). The precise cause and mechanism of OPLL remain unestablished. The posterior longitudinal ligament (PLL) extends along the dorsal surface of the vertebral canal, spanning from the axis to the sacrum, where it functions to resist excessive spinal hyperflexion. Typically composed of well-aligned fibroblasts and lacking cartilage components, the PLL can undergo pathological ossification known as OPLL. In cases of OPLL, fibroblast-like cells transform into chondrocytes and osteoblasts, accompanied by an augmented vascularity attributed to the overexpression of bone morphogenetic protein (BMP). Consequently, this process leads to the replacement of the native ligament's fibrous structure with cartilaginous tissue, as well as the emergence of regions characterized by degeneration and ossification7,17,21). As OPLL develops, it leads to a narrowing of the spinal canal, thereby causing harm to the spinal cord6,19). Consequently, patients exhibit clinical manifestations of myelopathy, which significantly compromises their quality of life7,8).
To date, there has been a dearth of investigations exploring the direct link between congestive heart failure (CHF) and OPLL. Nevertheless, CHF has established associations with factors such as infection and atherosclerosis. In pursuit of advancing our understanding, our nationwide longitudinal study endeavors to illuminate the potential relationship between CHF and OPLL.
MATERIALS AND METHODS
1. Data Source
Within South Korea, there exists a form of universal health insurance encompassing vital medical expenses for all residents under a unified public framework known as the National Health Insurance Service (NHIS). The NHIS administers comprehensive health assessments annually or biennially for individuals above the age of 40 who are either full-time or temporary workers. Additionally, the NHIS compiles individual data, incorporating demographic attributes and medical examinations, within the National Health Information Database (NHID). For our study, we have duly secured the authorization to access NHID health examination data spanning from 2004 to 2015 through the assistance of the Institutional Review Board (approval no. 2020-01-011) at the Bundang Medical Center, CHA University11).
2. Establishment of the Study Cohort
A comprehensive national health examination was undergone by a total of 514,557 individuals, with their progress being closely monitored over a 12-year span until December 31, 2015. Individuals identified by the International Classification of Diseases, 10th Revision codes 'M48.8, M48.80, M48.81, M48.82, M48.83' were categorized as patients with OPLL (n=3,405). Among this group, 1,977 subjects were excluded due to the absence of computed tomography data. Within the remaining 1,428 subjects, individuals newly diagnosed with OPLL after January 1, 2003 were selected as the primary cohorts for our investigation, resulting in a total of 1,289 participants. Utilizing the 'Match IT' R package algorithm, a control group consisting of 6,445 subjects was meticulously chosen through a 1:5 age- and sex-stratified matching methodology (Fig. 1).
Throughout the duration of the study, the participants were under continuous observation commencing from the date of their initial CHF diagnosis. This surveillance persisted until either their mortality or the culmination of the designated observation period. The occurrence of CHF was meticulously analyzed, with meticulous consideration given to variables encompassing age, sex, income, and underlying medical conditions.
3. Statistical Analysis
Both the χ2 test and Student’s t-test were employed to compare the demographic variables between the OPLL group and the control group. To estimate the probability of CHF-free survival in each group, the Kaplan-Meier method was employed. The impact of OPLL on the occurrence of CHF was explored using multivariate analysis within the framework of the Cox proportional-hazards regression model (Table 1).
In detail, two refined models were developed. Model 1 was adjusted for age and sex, while Model 2 was adjusted for age, sex, income, and pertinent medical factors, including hypertension, diabetes, and dyslipidemia (Table 2). For subgroup analyses aimed at adjusting covariates, the Cox proportional-hazards regression model was employed, utilizing R software version 3.3.3 (The R Foundation for Statistical Computing, Vienna, Austria). This approach was chosen to ensure the meticulous examination of covariate effects within the context of the study.
RESULTS
1. Characteristics of the OPLL and Control Groups
For both the newly diagnosed OPLL group and the control group, a larger proportion of the subjects were female (51.67%) and the average age was 57.4 ± 9.65 years.
2. CHF in the OPLL and Control Groups
Regarding CHF, the experimental group exhibited a higher incidence rate (2.153) compared to the control group (0.874) per 1000 person-years. Moreover, we identified that the hazard ratio of the control group was 3.164 (95% confidence interval [CI], 1.867-5.360) in Model 1, while in Model 2, it was 3.355 (95% CI, 1.977-5.694) (Table 2). In Fig. 2, during the 12-year follow-up period for CHF, the cumulative incidence in the experimental group was observed to be higher compared to that in the control group.
3. Subgroup Analysis of the OPLL Incidence Rate
There are ten subgroups categorized by sex, age, diabetes mellitus, hypertension, and dyslipidemia (Table 3). Across all subgroup factors, the incidence rate of CHF exhibited a difference between the OPLL group and the control group. About the incidence of CHF, notable differences were evident among both males (95% CI, 2.948-6.174) and females (95% CI, 3.724-7.891). Moreover, significant differences were observed in the age <65 group (95% CI, 4.219-9.111), age ≥65 group (95% CI, 2.538-5.227), individuals without diabetes mellitus (95% CI, 3.722-6.627), individuals with diabetes mellitus (95% CI, 2.376-8.409), those without hypertension (95% CI, 3.591-7.888), and those with hypertension (95% CI, 3.298-6.694). Lastly, a notably distinct incidence of CHF was evident both in individuals without dyslipidemia (95% CI, 3.682-6.643) and those with dyslipidemia (95% CI, 2.425-7.769).
DISCUSSION
Our countrywide longitudinal research found that the OPLL group had a 3.164-fold greater incidence of CHF after controlling sex and age. The OPLL group also had a 3.355-fold greater incidence of CHF after controlling sex, age, wealth, hypertension, dyslipidemia, and diabetes mellitus. In addition, our study showed that the incidence of CHF was greater in the OPLL group than in the control group in sex, age, non-diabetes, diabetes, non-hypertension, hypertension, non-dyslipidemia, and dyslipidemia subgroups.
Until now, no studies have examined the direct connection between OPLL and CHF. Instead, several papers have shown that cytokines, signaling, and inflammation are associated with CHF. Yousefi et al.22) reported the transforming growth factor (TGF)-β and Wnt signaling pathways have been documented as pivotal regulators of myofibroblast biology within cardiac fibrosis, a process that contributes to the development of CHF. In previous studies related to cardiovascular disease and OPLL, there was a paper that studied the relationship between acute myocardial infarction (AMI) and OPLL. Kim et al.10) reported nationwide longitudinal cohort analysis between AMI and OPLL shows that patients with OPLL have an elevated risk of AMI, and the correlation is related to dysregulation of TGF-β.
Firstly, several members of the superfamily, including TGF-β1, TGF-β2, and TGF-β3 are associated with cardiac fibrosis following AMI and cardiac fibrosis is also closely related to CHF. For example, elevated expression levels of TGF-β1 due to mutation of TGF-β superfamily10), a prototypical fibrogenic cytokine, have been observed in cases of cardiac fibrosis, both in human subjects and experimental models22). Bai et al.1) reported rats were subjected to induced AMI through RNA manipulation by inhibiting the TGF-β1 gene. During the progression of OPLL, multiple genes, including TGF-β and BMP, play significant roles. The TGF-β/BMP signaling pathway orchestrates the intricate processes required for osteoblastogenesis, thereby facilitating the advancement of OPLL14,20). Consequently, this cytokine, TGF-β, could potentially elucidate a link between OPLL and CHF.
Secondly, the activation of the Wnt/β-catenin pathway by TGF-β has been extensively documented. TGF-β triggers an elevation in Akt phosphorylation via PI3K activation, leading to the inhibition of GSK3β (an enzyme crucial in β-catenin degradation). This sequence of events contributes to the augmentation of cardiac fibrosis22). Activation of Wnt/β-catenin signaling within osteocytes leads to enhanced bone formation and bone resorption, ultimately resulting in a positive balance that yields bone gain2). These mechanisms have the potential to contribute to the development of OPLL. Therefore, Wnt/β-catenin signaling is reasonable to explain the association between OPLL and CHF.
Thirdly, it has been noted that C-reactive protein (CRP) can serve as a robust indicator concerning vascular disease3,4,12). Habibi et al.5) reported the outcomes from intracranial vessel wall magnetic resonance imaging (MRI) revealed a robust connection between CRP and CHF (p<0.001). An MRI study provided compelling evidence linking CRP to the risk of CHF. Human genetic information indicates that CRP contributes causally to the development of CHF. Therefore, evaluating CRP levels could provide supplementary prognostic insights, enhancing overall risk assessment for patients with CHF. This genetic support underscores a potential causal connection between CRP and the onset of CHF.
Kawaguchi et al.9) reported a significant statistical disparity in CRP levels between patients with OPLL and those without OPLL was observed. The average serum CRP concentration was found to be 0.122 ± 0.141 mg/dL in the OPLL cohort, whereas it measured 0.086 ± 0.114 mg/dL in the comparison group (p=0.047). Therefore, CRP is reasonable to explain the association between OPLL and CHF.
Our longitudinal cohort study has several limitations. First of all, while our analysis accounted for variables such as sex, age, income, hypertension, diabetes, and dyslipidemia among participants, the study did not incorporate the influence of patients' lifestyle factors, such as smoking intensity, alcohol consumption, physical activity levels, and obesity status. These lifestyle elements could potentially exert a significant influence on the development of CHF.
At last, since no clear mechanism for OPLL has been elucidated, further studies will be needed to determine whether the signaling pathways discussed above are directly related to the incidence of OPLL. As a result, it is difficult to understand the exact relationship between OPLL and CHF through TGF-β, Wnt signaling, and CRP. Nevertheless, this is the first study attempting to find the connection between CHF and OPLL in Korean patients. Also, it is the largest nationwide longitudinal study indicating an increased risk of CHF in OPLL patients.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
