This study was a single-center retrospective study. The subjects were divided into the following two groups. The dissection group: patients who were consecutively included from January 2019 to April 2023, presenting with symptoms such as chest and back pain and diagnosed with acute or subacute Aortic Dissection (AD) by full aortic CT angiography in our hospital. A total of 99 cases were included, among which 78 were in the acute phase (onset ≤ 14 days) and 21 were in the subacute phase (onset 15 - 90 days); according to the Stanford classification, 42 were type A and 57 were type B. The control group: patients who underwent full abdominal CT plain scan due to non-specific abdominal pain, routine physical examination or other non-cardiovascular emergencies during the same period were selected based on the age distribution (matched within ±5 years) and gender of the dissection group. To ensure that the control group had no aortic disease, two radiologists jointly reviewed their abdominal CT plain scan images (window width and window level suitable for observing large blood vessels) and reviewed their medical records to exclude any individuals with signs of aortic lesions such as aortic widening, calcification, ulceration, intramural hematoma or dissection. Inclusion criteria: 1) Age ≥ 18 years; 2) Image quality met the requirements for QCT measurement. Exclusion criteria: 1) Severe motion or metal artifacts in CT images; 2) Missing clinical data (such as smoking history, blood pressure, and blood lipids); 3) Previous history of aortic surgery; 4) Coexisting malignant tumors or long-term chemotherapy or glucocorticoid treatment. Ultimately, 99 patients were included in the dissection group and 85 subjects in the control group.

The following data were collected through the electronic medical record system: 1) Demographic characteristics: gender, age; 2) Lifestyle: smoking history (defined as cumulative smoking ≥ 100 cigarettes), drinking history (defined as regular drinking with an ethanol intake ≥ 30 g/d for more than one year); 3) Medical history: hypertension (based on medical history or multiple blood pressure measurements ≥ 140/90 mmHg after admission), diabetes (based on medical history or fasting blood glucose ≥ 7.0 mmol/L or glycated hemoglobin ≥ 6.5%); 4) First fasting lipid profile upon admission: total cholesterol, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol.

2.3.1. CT Scanning

The Siemens SOMATOM Definition dual-source 64-slice CT machine was used. The parameters for total aortic CT angiography: tube voltage 120 kV, automatic tube current modulation (reference mAs: 200 - 350), detector collimation 128 × 0.6 mm, pitch 0.8, rotation time 0.5 s, reconstruction layer thickness 1.5 mm, and slice spacing 1.5 mm. All images were reconstructed using the standard vascular reconstruction algorithm to ensure the comparability of CT values.

2.3.2. QCT Measurement

The plain scan DICOM images were imported into the QCT (Mind-ways Company) measurement and analysis software. Abdominal fat measurement: Two radiologists with over 10 years of abdominal imaging diagnostic experience and standardized QCT training independently operated. At the L2/3 intervertebral disc center level [7], the Total Abdominal Fat (TAT) area was manually delineated along the abdominal wall contour using the tissue composition analysis module. The software automatically identified and segmented the VAT area (with the abdominal wall muscles and spine as the inner boundary), and the SAT area was obtained by subtracting the VAT area from the TAT area. The results of the two measurements were recorded and averaged. Liver fat measurement: At the level of the right branch of the portal vein, circular Regions Of Interest (ROIs) with an area of approximately 300 mm² were placed in the right anterior lobe, right posterior lobe, and left medial lobe of the liver, avoiding large bile ducts, intrahepatic vessels, and liver fissures. The average value of the three ROI measurements was taken as Fat%-QCT. Consistency test: 30 images were randomly selected, and the two radiologists repeated the measurements after a two-week interval. The Intraclass Correlation Coefficient (ICC) of abdominal VAT area, SAT area, and Fat%-QCT was calculated to assess the intra- and inter-observer consistency. The results showed that the ICC of all indicators was greater than 0.95, indicating excellent measurement consistency.

SPSS 25.0 software was used. For the measurement data that followed a normal distribution, they were expressed as x ¯ ± s, and the comparison between groups was conducted using the independent sample t-test; for the non-normal distribution data, they were represented as M (P25, P75), and the comparison between groups was performed using the Mann-Whitney U test. For the count data, they were expressed as the number of cases (percentage), and the comparison between groups was conducted using the χ² test or Fisher’s exact test. Spearman’s rank correlation was initially used to preliminarily analyze the correlation between each factor and AD (assigned as 1 = yes, 0 = no). Variables from the univariate analysis with P < 0.1 and variables with clear clinical significance (such as age, gender) were included in the multivariate binary Logistic regression to analyze the independent related factors of AD occurrence. In the Logistic regression model, categorical variables such as smoking history, drinking history, and hypertension were all regarded as the reference category as “no”. The results were expressed as Odds Ratio (OR) and its 95% Confidence Interval (CI). Before conducting the regression analysis, collinearity diagnosis was performed on continuous independent variables (SAT area, VAT area, Fat%-QCT), and the Variance Inflation Factor (VIF) was all less than 5, indicating the absence of severe multicollinearity. As P value < 0.05 was considered statistically significant.

There was no statistically significant difference in age and gender composition between the dissection group and the control group (P > 0.05), and the matching was successful. The proportions of smoking history, drinking history, and hypertension in the dissection group were significantly higher than those in the control group (all P < 0.001). In terms of lipid indicators, there was no statistically significant difference in total cholesterol, triglycerides, high-density and low-density lipoprotein cholesterol levels between the two groups (all P > 0.05). In terms of imaging parameters, the Fat%-QCT, VAT area, and SAT area in the dissection group were significantly higher than those in the control group (P = 0.012, <0.001, 0.002). See Table 1.

Table 1. Comparison of baseline data and fat parameters between the dissection group and the control group.

Note: a represents χ² test or Fisher’s exact test; b represents Mann-Whitney U test.

Spearman rank correlation analysis revealed that smoking history, drinking history, hypertension, Fat%-QCT, SAT area, and VAT area were all positively correlated with the occurrence of AD (correlation coefficient R values ranged from 0.185 to 0.427, all P < 0.05), among which hypertension had the strongest correlation (R = 0.427, P < 0.001). See Table 2.

Table 2. Spearman correlation analysis of various factors with the occurrence of aortic dissection.

The variables with significant results from the univariate analysis (smoking history, drinking history, hypertension, Fat%-QCT, VAT area, SAT area) and age and gender were included in the multivariate Logistic regression model. The collinearity diagnosis revealed that the Variance Inflation Factors (VIF) for all continuous variables were all less than 2, indicating that there were no serious multicollinearity issues in the model. The results showed that, after controlling for other factors, smoking history, drinking history, hypertension and SAT area were independent risk factors for AD. See Table 3.

Table 3. Multivariate logistic regression analysis of risk factors for aortic dissection.