Strong relationship between dyslipidemia and the ectopic ossification of the spinal ligaments

Study design

A retrospective cross-sectional study was conducted in accordance with the Declaration of Helsinki (1964), including subjects between April 2020 and May 2021. The study was approved by the ethical review board of the Hakodate Central Hospital and Hokkaido University Hospital, and the need for Obtaining patients’ informed consent was waived owing to the retrospective nature of the study and the deidentified data used.

patients

A database of 12,740 Japanese patients from a single institution was used. All subjects, with or without symptoms, underwent routine health examinations once per year or once every few years. Most of the subjects were community residents and facility staff, including physicians, nurses, nursing assistants, therapists, and clerks; the majority underwent blood tests at their discretion. The subjects underwent CT of the trunk at their discretion; they also selected the scan region (neck to chest, abdomen to pelvis, or neck to pelvis). Among the 1,002 subjects who underwent CT, 525 were selected for whom CT allowed assessment of the cervical spine to pelvis region and for whom blood test data were available. Finally, a total of 458 subjects (251 men, 207 women), aged 30 to 78 years, with OLF and/or OPLL and without spinal ligament ossification were included in this study (Fig. 1).

Figure 1

Flow chart of study participants. CT, computed tomography; OLF, ossification of the ligamentum flavum; OPLL, ossification of the posterior longitudinal ligament.

Demographics and comorbidities

Demographic data were obtained from all subjects using a questionnaire assessing height, weight, smoking history, the presence of comorbidities (hypertension, diabetes mellitus, dyslipidemia, ischemic heart disease, stroke, and renal disease), and the current and past use of medication for the comorbidities. The questionnaire also assessed the presence of weight gain from the age of 20 years and exercise routine of more than 30 min per day.

Serological assessment parameters

All serological assessments were performed using fasting blood samples. The assessed parameters included total cholesterol (T-Cho), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), glycated hemoglobin A1c (HbA1c), and blood glucose. The LDL-C/HDL-C ratio (L/H ratio) was calculated and used as one of the parameters to indicate the severity of dyslipidemia. The rationale is based on the suggestion that the L/H ratio is a better predictor of atherosclerosis progression and ischemic cardiovascular disease than LDL-C and HDL-C separately; an L/H ratio ≤ 2.0 significantly inhibits the progression of coronary artery plaque, while a ratio ≤ 1.5 further strengthens this effect27.

Diagnostic criteria for dyslipidemia by the Japan Atherosclerosis Society

For adults, the 2012 guidelines of the Japan Atherosclerosis Society define lipid abnormality as a TG concentration ≥ 150 mg/dL, LDL-C ≥ 140 mg/dL, and/or HDL-C < 40 mg/dL28.

In the present study, subjects with dyslipidemia were defined as those who met any of the above criteria or who were taking therapeutic drugs for dyslipidemia.

Assessment of the presence of OLF and OPLL

Axial CT images were used to evaluate the distribution of OLF and OPLL (cervical [C]thoracic [T]and/or lumbar [L]). CT was performed using an Aquilion One™/Genesis Edition system (Canon Medical Systems Inc., Tochigi, Japan). The presence of OLF and OPLL was determined according to previous reports (Fig. 2); OLF was essentially defined as ossification of the ligamentum flavum with a thickness of 3 mm or more. Ossification of less than 3 mm, which was clear on axial images, which was also considered OLF3,11, as was mushroom-shaped ossification localized in the center of the lamina3,11. OPLL was essentially defined as ossification of the posterior longitudinal ligament with a thickness of 2 mm or more on axial images11,29. All CT scans were assessed by two board-certified spine surgeons; Disagreements were resolved by consensus.

Figure 2

Criteria for the identification of OLF and OPLL on computed tomography. (a) OLF that is clearly visible but less than 3 mm in thickness (white arrowheads). (b) OLF with an apparent thickness of 3 mm or more (black arrowheads). (c) A mushroom-shaped OLF that is localized at the center of the laminae (white arrow). (d,e) OPLL with an apparent thickness of 2 mm or more (black arrow). (f) OPLL that is small but with a thickness of 2 mm or more (bold white arrow). OLF, ossification of the ligamentum flavum; OPLL, ossification of the posterior longitudinal ligament.

Grouping of subjects

All subjects were divided into the following four groups according to the type of concomitant spinal ligament ossification present: no ligament ossification (control; n = 230), OLF (n = 167), cervical OPLL (C-OPLL; n = 28), thoracic OPLL (T-OPLL; n = 33). Subjects with OPLL in the cervical spine exclusively—with or without concomitant OLF—were classified as the C-OPLL group, and subjects with OPLL in the thoracic spine—with or without concomitant cervical OPLL, lumbar OPLL (L-OPLL), or OLF —were classified as the T-OPLL group. Subjects with OLF without concomitant C-OPLL, T-OPLL, or L-OPLL were classified as the OLF group. This classification was based on the following rationale: (1) C-‍OPLL has a comparatively high probability of occurring concomitantly with OLF9; (2) compared to C-OPLL, T-OPLL has distinct features, such as morbid obesity, early onset of symptoms, and diffuse ligament ossification of the entire spine, including OLF, C-OPLL, and L-OPLL12,15,24 ; and (3) OLF occurs mostly in the thoracic spine and rarely in the cervical or lumbar spine29.

All subjects were also divided into the following two groups according to the presence or absence of concomitant dyslipidemia: DL(+) (n = 215), and DL(−) (n = 243).

Statistical analysis

The data were analyzed using BellCurve for Excel (version 3.10; Social Survey Research Information Co., Ltd., Tokyo, Japan). Results are presented as mean ± standard deviation. Four-group comparisons were evaluated using the Kruskal–Wallis and Fisher’s exact tests. Statistical significance was set at P < 0.05; however, in the four-group comparisons, it was set at P < 0.012 (0.05/4) with a Bonferroni correction considering multiple testing.

Univariate logistic regression analysis was performed for age, body mass index (BMI), sex, comorbidities, the presence of weight gain since the age of 20 years, T-Cho (mg/dL), TG (mg/dL), HDL- C (mg/dL), LDL-C (mg/dL), and the L/H ratio as risk factors for the prevalence of OLF, C-OPLL, and T-OPLL. For variables determined to be significant (P < 0.05) in the univariate logistic regression analysis, the relationship between factors affecting the presence of the three types of spinal ligament ossification (ie, OLF, C-OPLL, and T-OPLL) was evaluated using multivariate logistic regression analysis. Diabetes mellitus was analyzed as a candidate variable regardless of its significance in the univariate logistic regression analysis because it has been mentioned as a risk factor for C-OPLL in the literature. Statistical significance was set at P<0.016 (0.05/3) with a Bonferroni correction considering multiple testing of subjects with multiple concomitant types of ossification.

Cut-off values ​​of the parameters of lipid metabolism for the presence of OLF, C-OPLL, and T-OPLL were calculated from receiver operating characteristic (ROC) curves. In this analysis, only subjects who were not receiving medication were included, considering the effect of medications on serum parameter levels.