Association between lipoprotein(a) and peripheral arterial disease in coronary artery bypass grafting patients

Abstract Objective Peripheral arterial disease (PAD) affects a large population and is associated with various adverse clinical outcomes. Lipoprotein(a) has proatherogenic properties and is associated with PAD incidence and severity. The aim of this study is to explore the association between LP(a) and PAD in coronary artery bypass grafting (CABG) patients. Methods A total of 1001 patients were included and divided into two groups: low Lp(a) group [LP(a) < 30 mg/dL] and high Lp(a) group [LP(a) ≥ 30 mg/dL]. A comparison of PAD incidence diagnosed by ultrasound was made between the groups. Multivariate logistic regression was conducted to explore the risk factors for PAD. During the analysis, the influence of diabetes mellitus (DM) and gender on LP(a) serum level was taken into consideration. Results DM history (odds ratio [OR], 2.330, p = .000 for males; OR, 2.499, p = .002 for females) and age (OR, 1.101, p = .000 for males; OR, 1.071, p = .001 for females) were risk factors for PAD. LP(a) ≥ 30 mg/dL was a risk factor for PAD only in female patients (OR, 2.589, p = .003), while smoking history was a risk factor only in male patients (OR, 1.928, p = .000). LP(a) level was not associated with PAD severity in DM patients of both gender. As for female patients without DM, PAD was more severe in the high LP(a) group. Conclusions In CABG patients, DM history and age were risk factors for PAD. But a high level of LP(a) was a significant risk factor only in female patients. In addition, we are the first to propose a gender deviation in the correlation between LP(a) serum level and severity of PAD diagnosed by ultrasound.


| BACKGROUND
A systematic review estimated that 238 million people were living with peripheral arterial disease (PAD) in 2015 worldwide. 1 PAD affects a large population and has become an increasingly global problem. 2 A body of evidence indicated that PAD was associated with an increased risk of various adverse clinical outcomes, thus it should not be underestimated. 1 Traditionally, the term PAD refers to the disease of vascular beds all over except for the heart, which includes the carotid artery, renal artery, lower extremity artery, and so on. The most common cause of PAD is atherosclerosis followed by peripheral venous disease and lymphatic disease. 3 There exist different opinions on the definition of PAD in different academic groups. 4 For the purpose of our research, we referred to PAD as partial or complete atherosclerotic obstruction of the lower extremity artery. 2 Lipoprotein(a) [LP(a)] is a lipoprotein with a structure similar to lowdensity lipoprotein (LDL) synthesized in the liver. 5 LP(a) serum level is genetically determined. Although its serum level varies among individuals, it is relatively stable within a single individual. 6 Lp(a) has proinflammatory and proatherogenic properties. Due to these properties, Lp(a) serum level is associated with atherosclerotic cardiovascular diseases including PAD. 7 Previous studies have shown that Lp (a) is not only a significant independent risk factor for PAD but also associated with more severe forms of PAD in specific populations. 8,9 The association between LP(a) serum level and PAD in coronary artery bypass grafting (CABG) patients is unknown. We speculated that in this population, LP(a) may have a correlation with PAD. In this study, we aim to explore the risk factors for PAD in CABG patients and the correlation between LP(a) and severity of PAD. For all patients, we recorded preoperative baseline characteristics including age, sex, body mass index, DM history, hypertension history, smoking history, systolic blood pressure, diastolic blood pressure, and so on were all recorded. Preoperative biochemical blood tests including serum total cholesterol, triglycerides, highdensity lipoprotein (HDL) cholesterol, LDL cholesterol, Lp(a), and so on were all recorded. Each patient received a preoperative lower extremity artery sonographic examination. Generally, all routine examinations including lower extremity ultrasound were completed within 1 week after admission and then surgery was performed. Four arterial beds were imaged: femoral, popliteal, posterior tibia, and anterior tibia arteries. The femoral artery included common and superficial femoral arteries. All images were acquired with Philips, GE, or Hitachi ultrasound imaging systems by certified experienced ultrasound clinicians. Images were acquired using an L9-3 transducer.
For each artery bed, we define 40%-49% diameter stenosis as mild stenosis, 50%-69% diameter stenosis as moderate stenosis, and 70%-99% diameter stenosis as severe stenosis. Occlusion was grouped into severe stenosis. This study was approved by the Ethics Committee of Beijing Anzhen Hospital.

| Statistical analysis
Continuous variables were reported as mean ± standard deviation and categorical variables as numbers and percentages. Univariate comparisons between groups were performed using the χ 2 test for categorical variables and Student's t-test or Mann-Whitney rank-sum test for continuous variables, as appropriate. To explore the risk factors for PAD, a multivariate logistic regression analysis was performed. All analyses were performed using SPSS26.0 (SPSS).

| Basic characteristics
A total of 1001 patients were finally included. The mean age was 61.62 ± 8.52 years and 76.8% were male. In the high Lp(a) group, total cholesterol (TCHO) serum level (4.27 ± 1.11 vs. 4.02 ± 1.05, p = .001) and LDL serum level (2.69 ± 0.93 vs. 2.39 ± 0.87, p = .000) were significantly higher than the low LP(a) group. Triglyceride serum level was slightly higher in the low LP(a) group (1.53 ± 0.83 vs. 1.70 ± 1.28, p = .044). The proportion of male patients was similar between the two groups. The proportion of patients with DM history, hypertension (HTN) history, and smoking history were also of no significance between the two groups.
The results were shown in Table 1.

| Incidence of PAD in CABG patients
The incidence of PAD was 47.0% (48.9% in male patients and 40.5% in female patients). The incidence of PAD had a significant difference between males and females (p = .025).

| Correlation between LP(a) serum level and PAD severity
The results were shown in Tables 2 and 3. As for DM patients, there was no significant correlation between LP(a) serum level YI ET AL. | 513 and severity of PAD for nearly all artery beds (anterior tibial artery, posterior tibial artery, and the entire lower extremity artery) in male and female patients. In male patients, incidences of PAD in some arterial beds were slightly higher in the low LP(a) group, although there might be no significant difference. In addition, the low LP(a) group had a higher incidence of femoral popliteal artery stenosis (34.4% vs. 20.3%, p = .027) and multiple vessel femoral popliteal artery stenosis (19.6% vs. 4.3%, p = .002).
In female patients, the incidence of PAD in the high and low LP(a) groups were of no significant difference.  T A B L E 2 Correlation between LP(a) and PAD severity in two groups (patients with DM).

| Risk factors for PAD
Our results showed that in CABG patients, risk factors for PAD had obvious gender differences. Surprisingly, LDL was not an independent risk factor in both male and female patients.

| DISCUSSION
In CABG patients, the incidence of PAD was 47.0%. In male patients, the incidence of PAD was 48.9%, and in female patients, it was 40.5%.
Our study showed that the incidence of PAD in CABG patients was quite high which was quite reasonable. Atherosclerosis is a systemic disease and affects multiple vascular beds. 11 PAD is always coexistent with coronary artery disease (CAD). 12 The significance of our study was that we provided the incidence of PAD in a CABG population and we used the ultrasonic method to evaluate PAD severity. In previous studies, PAD was always diagnosed using the ankle-brachial index (ABI), which is easier to perform but still has shortcomings. For instance, ABI can be falsely high in the presence of artery calcification, a condition often observed in patients with severe CAD such as CABG. 13 The ultrasonic method was cheap, safe, and had a satisfying accuracy as well. The sensitivities of ultrasound in detecting occlusion and stenosis have been reported to be 95% and 92% and specificities were 99% and 97%, respectively. 14 Due to these advantages, the application rate of lower extremity artery ultrasound in our hospital is quite high, and nearly all CABG patients underwent this examination. We propose that ultrasound might be a more suitable method for PAD at least in our country. PAD has been traditionally considered as affecting men more than women and some scholars ascribed this to hormonal differences. 15,16 But this point of view was controversial. Researchers also proposed that women had an equal or even higher prevalence of PAD diagnosed using ABI measurements. 3 influence of certain drugs. Since CABG patients were always with multiple comorbidities such as hyperlipidemia. The application of these drugs must affect LDL serum level, which we can clearly see from the basic characteristics that the mean serum level of LDL was 2.47 ± 0.899 mmol/L. Thus, the predictive value of LDL may have been concealed, while LP(a) level emerged as an important surrogate.
The association between elevated LP(a) serum level and PAD has already been proposed. 23 Prior studies have also reported an association between increased Lp(a) serum level and PAD severity. We speculate that LP(a) might be related to the prognosis or progression of male PAD.
As for female patients, we have to admit that the sample size of female patients with high LP(a) levels was small, thus the result might not be so convincing. But the incidence of severe stenosis of different segments of the lower extremity artery in the high LP(a) level group is much higher, thus we believe that this trend is true, and a still larger sample size is needed to obtain a more convincing result.
Although the correlation between LP(a) level and PAD severity in non-DM female patients needs to be put forward conservatively, we might still see the trend. Surprisingly, the trend was quite reasonable and there might be some possible explanations. LP(a) was recognized as an independent risk factor for CAD, stroke, PAD, and so on, but the situation might be somewhat more complex in women. 24,25 Although LP(a) serum level remains relatively stable in an individual's life, it may change in some specific circumstances. For instance, LP(a) serum level increases by about 25% after the onset of menopause. 26 In our study, most female patients were in postmenopausal status. have proposed that DM patients always have reduced Lp(a) levels. 28 We speculated that the existence of DM might have some influence on the function of LP(a), thus concealing the relation between LP(a) serum level and PAD. As our results had shown in male patients with DM incidence of PAD of several arterial beds was even higher in the low LP(a) group.
In short, our results confirmed previous arguments that LP(a) was more of a "conditional risk factor" and its correlation with PAD as well as other clinical conditions depends on an individual's background based on age, gender, and coexisting conditions such as DM. 29 As far as we know, we are the first to explore this correlation using the diagnostic method of ultrasound. There is another innovative part that we found, which is the gender deviation of this correlation in a reasonable, a larger sample size is still needed to obtain a more convincing result. Second, our study was only a cross-sectional study.
If we follow up for a period of time and observe the progression of PAD, we might find some impact of LP(a) on the prognosis of PAD. In addition, we could further study whether lowering LP(a) level improves the prognosis of PAD. Third, Lp(a) serum level is affected by certain lipid-lowering medications. In the CABG population, some patients were diagnosed with hyperlipidemia and took lipid-lowering drugs. The effect of these drugs on the serum level of Lp(a) could not be calculated.