Data were obtained from the First Affiliated Hospital of Xinjiang Medical University. Between December 27, 2013, and September 30, 2020, a total of 12,904 symptomatic patients showing signs of CAD were continuously enrolled. The main symptoms of CAD are typical angina or atypical symptoms such as exertional dyspnea or episodes of chest pain at rest, which patients with CAD often experienced. All patients had undergone CTA with symptoms of suspected CAD as a first-line diagnostic imaging and had not received percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG). Patients with a history of PCI, CABG surgery, tumor, skin disease, immune disease, stroke, pulmonary embolism, infection status, and incomplete outcome data were excluded. The study was approved by the First Affiliated Hospital of Xinjiang Medical University institutional review board (K202106-02), and written informed consent was obtained from all the enrolled patients.

Those CTA were uniformly acquired by using multi-detector row computed tomography (CT) scanners consisting of 64-rows or greater (Somatom Definition or Somatom Definition Flash, Siemens Healthcare, Forchheim, Germany). CAC scores was calculated using Agatston’s scoring system. CAC scores were categorized into the following groups: 0–10, 11–100, and $\ge$101 [25, 26]. Coronary plaques were defined as lesions $>$1 mm${}^2$ within and/or adjacent to the coronary artery lumen, which were clearly distinguished from the vessel walls and the pericardium [27]. Plaques were classified according to phenotype: (1) noncalcified plaques (plaques having a lower density compared with the contrast-enhanced vessel lumen); (2) calcified plaques (high-density plaques); and (3) mixed plaques (noncalcified and calcified components within a single plaque). Plaques that exist in both calcified and noncalcified segments were classified as calcified plaques [27].

Prior to the vascular test, the patients had to stay in a reclined position for at least five minutes to maintain stable heart rate values. During coronary CTA acquisition, iodinated contrast (0.8 mL/kg) was delivered intravenously at a steady rate of 4–8 mL/s, with an intravascular contrast agent residence period of no less than 12 s followed by a 30–40 mL saline flush. Stenosis severity was categorized using the quantitative stenosis grading recommended by the Society of Cardiovascular Computed Tomography guidelines [28], it was classified as none (0% luminal stenosis), non-obstructive (1–49% luminal stenosis), or obstructive ($\ge$50% luminal stenosis).

The degree of coronary stenosis was classified as normal (no coronary stenosis), nonobstructive CAD (lesions 50%), and obstructive CAD (lesions $\ge$50%). Obstructive CAD was further subdivided into 1-, 2-, and 3-vessel obstructive CAD.

The following risk factors for coronary artery disease were considered: (1) current smoking status; (2) lipid status, i.e., total cholesterol levels $>$5.2 mmol/L prior to CTA; (3) triglyceride $>$1.69 mmol/L prior to CTA; (4) LDL $>$3.8 mmol/L prior to CTA; (5) high-density lipoprotein (HDL) 1.55 mmol/L prior to CTA; (6) systolic pressure $\ge$140 mmHg; (7) diastolic pressure $\ge$90 mmHg prior to CTA; (8) body mass index; (9) diabetes mellitus; (10) family history of CAD (defined as the presence of a first-degree relative with a CAD).

Clinical data were obtained through telephonic follow-ups. A total of 803 patients were lost at follow-up, after reviewing the medical records, these patients had no record of re-visits in our hospital. Participants were followed up every 6–12 months for endpoint events. All patients were tracked till April 30, 2021. All occurrences were documented to determine the vessel-related clinical events. All adverse events were assessed by two experienced cardiologists, and if there was any discrepancies, a third physician was consulted to ensure the data was reliable. The primary outcome was major adverse cardiac and cerebrovascular events (MACCE), which was defined as a composite outcome of nonfatal myocardial infarction, revascularization (PCI and CABG), stroke, and cardiovascular death. The secondary outcomes were congestive heart failure, cardiogenic shock, malignant arrhythmia, and all-cause mortality. Unless an unmistakable non-cardiovascular cause was identified, all deaths were classified as cardiovascular deaths. Diagnosis of myocardial infarction was based on the fourth universal definition of myocardial infarction [29].

SPSS22 software (IBM Corp., Armonk, NY, USA), STATA version 16.0 software (Stata Corp, College Station, Texas, USA), and R programming language version 3.3.1 (R Foundation for Statistical Computing, Vienna, Austria) were used to perform statistical analysis. The baseline characteristics were outlined with median and interquartile ranges for continuous variables. and counts and percentages for categorical variables. Multivariable adjusted odds ratios (ORs) were used to assess the relationship between independent risk factors and CAC. The analyses were adjusted for age, sex, smoking, total cholesterol, triglyceride, LDL, HDL, systolic blood pressure, diastolic blood pressure, diabetes mellitus, and family history of CAD. In addition, we formed four groups based on the number of CAD risk factors (These risk factors have been enlisted and defined in the ‘CAD risk factors’ subsection). We analyzed the correlation between the number of risk factors and CAC utilizing unadjusted ORs and ORs that had been adjusted for age- and gender.

The log-rank test was used to evaluate the significance of the difference in survival for each age group and CAC group. We calculated the MACCE rate per 1000 person-years (with 95% CI) based on CAC groups with stratification by the number of risk factors and age groups. The MACCE rate per 1000 person-years (with 95% CI) based on stenosis with a luminal cross-sectional area was also assessed and was stratified by different age groups. The incidence rate ratio was calculated using the Poisson regression 1000 person-years approach to derive the relative risks. The nomogram was created using the RMS (Root Mean Square) package of R software, version 3.3.1 (R Foundation for Statistical Computing, Vienna, Austria). For MACCE prediction, ROC (Receiver Operating Characteristic) curve analysis was used to determine the optimal cutoff value of cardiovascular risk prediction model. Sensitivity, specificity, positive likelihood ratio (positive LR), negative likelihood ratio (negative LR), positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy were determined from the optimal threshold by the Youden index.

The study flow chart is presented in Fig. 1. A total of 7432 symptomatic patients were enrolled into the study. The baseline characteristics of those patients are shown in Table 1. The median age was 67 years (interquartile range [IQR]: 61–73 years). The number of patients with CAC 0–10 was 970 (13%, CAC 11–100 was 2331 (31%), and CAC $\ge$101 was 4131 (56%)). The median value and interquartile ranges of the CAC measurements for each group was 3 (IQR: 1.2–6), 42.4 (IQR: 23–66.7), and 379.6 (IQR: 197.4–827.1), respectively. 525 patients (7.1%) had 1 risk factor, 1275 patients (17.1%) had 2 risk factors, 1762 patients (23.7%) had 3 risk factors, and 3870 patients (52.1%) had 4 or more risk factors. In the total 7432 patients, 553 (7.4%) of whom had no CAD, 3889 (52.3%) nonobstructive CAD, and 2990 (40.3%) had obstructive CAD. The median duration of follow-up was 3.4 years (interquartile range: 1.8–5.4 years). A total of 478 patients (6.4%) had MACCE.

Fig. 1.

Flow chart of the study procedure. CTA, computed tomography angiography; PCI, percutaneous coronary intervention; CABG, coronary artery bypass graft; CAC, coronary artery calcium.

The percentage of CAC $\ge$101 increased with increasing age (Fig. 2A). The percentage of CAC $\ge$101 increased from the 55–65 to $\ge$75-year-old group from 46.5% to 68.2% (relative increase, 21.7%). By comparison, the proportion of CAC 0–10 significantly declined from 17.0% to 8.0% between the 55–65-year-old group and the $\ge$75-year-old group. Additionally, as the CAC score increased, the proportion of patients in the $\ge$75-year-old group increased from 12.9% to 25.8%, compared to the 55–65-year-old group (Fig. 2B).

Fig. 2.

Different age proportions in different CAC score groups in middle-aged and elderly patients ($\mathrm{\ge }$55 Years old). (A) CAC distribution across of age. (B) Age distribution across of CAC. CAC, coronary artery calcium.

Older age, LDL-C, and diabetes mellitus (DM) were the risk factors for the higher CAC scores, in comparison to the lower CAC scores (all ORs $>$1 and p values 0.05, Table 2). As the number of CAD risk factors increased, the increasing trend of ORs was observed in each CAC score group (Supplementary Table 1). The proportion of risk factor $\ge$4 in symptomatic patients aged over 55 years increased in correlation with the CAC scores (Fig. 3A) and the similar tendencies were observed among the different age subgroups (Fig. 3B–D).

Fig. 3.

Interplay between age groups and increasing CAC score for the presence of risk factor burden in middle-aged and elderly patients ($\mathrm{\ge }$55 Years old). (A) Risk factors distribution in total population. (B) Risk factors distribution in 55–65-year-old group. (C) Risk factors distribution CAC in 65–75-year-old group. (D) Risk factors distribution in $\ge$75-year-old group. CAC, coronary artery calcium; RF, risk factors.

Plaque characteristics in relation to age appeared to vary considerably. For example, middle-aged and elderly women had 1.816 times (95% CI: 1.150–2.867) higher chances of having calcified plaques, while the inverse was true for diabetes mellitus (Supplementary Table 2). Overall, Patients aged $\ge$75 years accounted for 17% in the patients without CAD, 20% in the patients with nonobstructive CAD, and 23% in the patients with obstructive CAD (Fig. 4). The detection of non-obstructive CAD was similar among the three age groups (53.5% vs 51.9% vs 49.1%), while the percentage of obstruction CAD increased as the age increased (Fig. 5).

Fig. 4.

Age groups distribution across different type of coronary artery disease. (A) Age distribution of patients without coronary artery disease. (B) Age distribution of patients with non-obstructive coronary artery disease. (C) Age distribution of patients with obstructive coronary artery disease.

Fig. 5.

Age distribution of coronary vessels with obstructive coronary artery disease.

The cumulative incidence of survival decreases with increasing age and CAC scores (p 0.0001) (Fig. 6A,B). The risk of MACCE increased with increase of CAC score and the number of risk factors (Fig. 6C, Supplementary Table 3, Fig. 7). The risk of MACCE was significantly higher in patients with $\ge$4 risk factors and CAC $\ge$101 than in those with 1 risk factors and CAC 0–10. There were 1.71 times higher event rates in patients with CAC $\ge$101 and $\ge$4 risk factors than in patients with CAC $\ge$101 and 1 risk factor (p = 0.044) (Supplementary Table 4). As the number of CAD risk factors increased, the MACCE survival rate decreased significantly (Supplementary Fig. 1). As CAC scores increased, the risk of MACCE events rose, and the MACCE rate was highest in those aged $\ge$75 years and those CAC $\ge$101 (Fig. 6D and Supplementary Table 5). Similar trends were observed for the different types of CAD. For patients with non-obstructive CAD, the incidence of MACCE events gradually increases with age. As an illustration, in patients with non-obstructive CAD, the risk of MACCE in patients aged $\ge$75 years was 2.62 times (95% CI: 1.83–3.74; p 0.0001) higher than that in patients aged 55–65 years (Fig. 6E and Supplementary Table 6). A predictive MACCE-related prognostic nomogram was established using the results of the multivariate analysis to predict the 1-, 3-, and 5-year overall survival. As shown, this nomogram was able to assess several variables to predict a patient outcome including age, CAC and risk factor burdens (Fig. 8). Further using the Youden index test, we found that CAC combined with age and risk factors had the highest predictive power for MACCE (AUC = 0.614) (Fig. 9). The details of statistical results about each MACCE risk prediction model were showed in Supplementary Table 7.

Fig. 6.

Interplay of risk factors and CAC, age and CAC, age and different types of coronary artery disease for risk of MACCE in middle-aged and elderly patients ($\mathrm{\ge }$55 Years old). (A) Cumulative survival rates for different age groups. (B) Cumulative survival rates for different CAC groups. (C) Events rate (per 1000 person years) with increasing CAC score according to the burden of risk factors. (D) Events rate (per 1000 person years) with increasing CAC score according to different age groups. (E) Events rate (per 1000 person years) with different type of coronary artery disease according to different age groups. CAC, coronary artery calcium; MACCE, major adverse cardiac and cerebrovascular events; CAD, coronary artery disease.

Fig. 7.

Interplay between age group and coronary artery calcium in middle-aged and elderly symptomatic patients. CAC, coronary artery calcium; RF, risk factors.

Fig. 8.

The nomogram was used to calculate predicted MACCE risk for symptomatic patients $\mathrm{\ge }$55 years old. The term “Points” represents the score of each variable under different values, “Total Points” means the total score of the collection after the sum of the corresponding individual fractions of all variables. To use the nomogram, first draw a vertical line to the top points row to assign points for each variable; then, add the points from each variable together and drop a vertical line from the total points row to obtain the 1-year survival, 3-year survival, 5-year survival, and median survival time (in years). CAC, coronary artery calcium.

Fig. 9.

The predictive effects of age, CAC, risk factors, CAC combined with age and risk factors on MACCE were analyzed. CAC, coronary artery calcium; MACCE, major adverse cardiac and cerebrovascular events.

Currently, the existing evidence on the relationship between risk factors and CAC in middle-aged and elderly patients is limited. A prospective study on a large multi-ethnic cohort of individuals aged 45–84 years who had no pre-existing clinical cardiovascular disease revealed that those with CAC = 0 had a low 10-year risk of event rate, but individuals with CAC $\ge$100 had a higher risk of event rate which was consistently above 7.5% [18]. A cohort study of 12,441 Korean patients with an average age of 52 years demonstrated that individuals with diabetes mellitus experienced greater CAC progression than those without diabetes mellitus, and diabetes had an incremental effect on CAC progression [30]. Previous research has suggested that an increase in traditional CAD risk factors is associated with an increase in CAC [7]. Across the spectrum of risk factor burden, a greater CAC score is strongly associated with an increased risk of long-term all-cause mortality, as well as a larger proportion of death caused by CVD and CAD [16, 31, 32, 33]. Our research adds to the preceding study, verifying the association between traditional risk factors and CAC scores, demonstrating the potential clinical application of risk factors. In addition, our research revealed that the middle-aged and elderly patients with four or more risk factors had 1.87 times higher odds of having CAC $\ge$101 compared to those with one risk factor. Notably, it is worth mentioning that 1 in 2 middle-aged and elderly adults with CAC $\ge$101 had $\ge$4 risk factors. Combing the results, it became evident that the risk factors are critical in predicting the probability of developing coronary calcification in middle-aged and elderly adults.

Stenosis of a vessel blocking the bloodstream leads to CAD, resulting in underperfusion of the heart region due to blocked vessel. Distinct types of CAD are caused by diverse degrees of coronary obstruction. However, obstructive CAD is associated with higher myocardial infarction rates [34, 35, 36]. Our research showed that when the number of CAD risk factors increased, the rate of obstructive CAD also increased. Specifically, we found that the rate of obstructive CAD was 5.7% among those with CAC 0–10 and 1 risk factor, while it was 16.5% among those with CAC 0–10 and 4 or more risk factors. Clinicians should alert those with a lower CAC score but higher risk factors to their possible health hazards. The MESA study indicated that cardiovascular events significantly increased with increasing CAC in middle-aged and elderly individuals [16]. Our study showed that the incidence of MACCE was lower in the group with CAC 0–10, but the incidence of MACCE was higher in the group with CAC 0–10 and $\ge$4 risk factors than in the group with CAC 0–10 and 1 risk factor, which indicated that risk factor burden can increase the risk of MACCE. Previous studies have shown that increased CAC was associated with a higher risk of future cardiovascular events in asymptomatic patients [37, 38]. In our study, the number of patients in CAC 0–10, CAC 11–100, and CAC $\ge$101 group was 970 (13%), 2331 (31.4%), and 4131 (55.6%), respectively. At a mean follow-up of 3.4 years period, the incidence of MACCE was 3.9%, 5.2%, and 7.7%, respectively. We found that combining CAC with age and risk factors improves the predictive value of MACCE events. The AUC ranged from 0.59 to 0.64 indicating that this model’s predictive ability for MACCE was moderate. These questions raised by this study warrant further investigation of a better risk score model.

Although we adjusted for gender in our analyses, potential unadjusted residual confounding factors may still exist. All the patients in our study were symptomatic, which may reduce the generalizability of the results to asymptomatic patients. When assessing symptomatic patients with CTA, it cannot exclude a degree of bias, as those deemed to be high-risk (e.g., aged above 75 years) may be more likely to be referred for invasive angiography. Therefore, this may reduce the generalizability of the results to the elderly and higher-risk groups. Age, risk factors, and CAC score are closely related and may affect each other with an increased severity of coronary artery disease. However, our study was based on real world data which is a major strength of this study. In addition, we obtained information on the CAC score and CT angiography results.