Associations of atherogenic index of plasma related indices with subclinical carotid atherosclerosis Authors: Wenzhen Li, Hong Zhang, Shuo Yang, Dajie Chen, Lap Ah Tse Published: 2026 ABSTRACT AbstractBackgroundEvidence regarding the relationships between atherogenic index of plasma (AIP)-related indices and carotid atherosclerosis (CA) is limited.MethodsA total of 59123 participants from the China Stroke High-risk Population Screening and Intervention Program (CSHPSIP) in Hubei Province, China, from 2017 to 2020 were included in this cross-sectional study. Three indices were developed to assess AIP-related indices: AIP-body mass index (AIP-BMI), AIP-waist circumference (AIP-WC), and AIP-waist-height ratio (AIP -WHtR).ResultsPer 1-standard deviation (SD) increase in AIP-BMI (OR = 1.36, 95% CI = 1.31–1.41), AIP-WC (1.34, 1.29–1.38), and AIP-WHtR (1.33, 1.28–1.38) were associated with CIMT. Per 1-standard deviation (SD) increase in AIP-BMI (OR = 1.32, 95% CI = 1.28–1.37), AIP-WC (1.34, 1.29–1.38), and AIP-WHtR (1.33, 1.29–1.38) were associated with plaques. AIP-BMI, AIP-WC and AIP-WHtR were positively associated with stenosis when they were divided into interquartile range (IQR), compared with the Q1 group. The ROC analyses showed that AIP-BMI had the greatest diagnostic value for CA and CIMT, with the AUC of 0.833 (0.828, 0.838) and 0.817 (0.811, 0.824), respectively. AIP-BMI, AIP-WC and AIP-WHtR had the greatest diagnostic value for plaques, with the AUC of 0.842 (0.836, 0.848), 0.842 (0.836, 0.847), and 0.842 (0.836, 0.848), respectively. Meanwhile, all the four indices (AIP, AIP-BMI, AIP-WC and AIP-WHtR) showed the same diagnostic value for plaques, with the AUC of 0.871 (0.858, 0.884).ConclusionsAIP-related indices were positively associated with CA, carotid intima-media thickness (CIMT), plaque and stenosis. Future prospective studies are needed to confirm our findings and clarify their clinical utility.Keywords: Atherogenic index of plasma, carotid atherosclerosis, carotid intima-media thickness, plaque, stenosis Background An increasing number of studies have focused on the positive relationship between the atherogenic index of plasma (AIP) and cardiometabolic diseases including coronary artery disease [1], type 2 diabetes [2], hypertension [3], and mortality [4], This relationship may be attributed to the strong association of AIP with the development of atherosclerosis [5] and lipoprotein particle size [6]. AIP, expressed as a logarithmic transformation of the quotient of triglycerides (TG) divided by high-density lipoprotein cholesterol (HDL-C) [lg(TG/HDL-C)], has been proven to be a reliable and predictable index for coronary atherosclerotic disease [7,8]. A previous study has found that elevated AIP was associated with a higher prevalence of carotid atherosclerosis (CA) [9]. CA constitutes a pivotal pathological process underlying cardiometabolic diseases, especially for coronary atherosclerotic disease [10]. Early detection of CA among ostensibly healthy populations is crucial for the prevention of cardiometabolic diseases and reduction of mortality [11,12]. Recently, a study demonstrated that combining AIP with obesity-related indices such as body mass index (BMI), waist circumference (WC), and waist-to-height ratio (WHtR) can provide a more comprehensive assessment of stroke risk [13]. Nevertheless, it remains unclear whether similar associations are also observed with CA, and whether this combination of AIP with obesity-related indices is superior to using AIP alone. Therefore, this study aimed at examining the relationship of AIP-BMI, AIP-WC and AIP-WHtR with CA. The AIP-related indices developed in this study are calculated using routinely measured parameters, making them simple and convenient for community-based risk assessment, which may help support evidence-based risk stratification in clinical and public health practice. Methods Study design and participants Participants were from the China Stroke High-risk Population Screening and Intervention Program (CSHPSIP) in Hubei Province, China, from 2017 to 2020. The detailed study design has been described previously [14]. A total of 61638 subjects were surveyed during the study time. After excluding those with missing basic sociodemographic information (n = 982) and missing laboratory data (n = 1533), 59123 subjects were included in the final analysis. This study was based on the Stroke Screening and Prevention Project in Hubei Province, and conformed to the Declaration of Helsinki. The written informed consents were obtained from all the participants. Assessment of AIP-related indices Blood samples were collected to test fasting plasma glucose, triglyceride (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C). AIP-related indices include AIP-BMI, AIP-WC, and AIP-WHtR. They were calculated as follows [13]: 1) AIP = Log [TG (mg/dl)/HDL-C (mg/dL)]; 2) AIP-BMI = AIP × [weight (kg)/height2 (m2)]; 3) AIP-WC = AIP × WC (cm); 4) AIP-WHtR = AIP × [WC (cm)/height (cm)]. The combination of AIP and anthropometric measures (BMI, WC, WHtR) was based on biological plausibility: AIP characterizes atherogenic lipid profiles, whereas BMI, WC, and WHtR reflect central adiposity and adipose tissue dysfunction. Both pathways independently promote inflammation, endothelial injury, and atherosclerosis. The composite index serves as a simple integrated marker of combined lipid and metabolic risk. They were calculated using routinely measured parameters, making them simple and convenient for community-based risk assessment, which may help support evidence-based risk stratification in clinical and public health practice. Assessment of CA Carotid ultrasonography examination was performed by qualified ultrasound technologists using one of the ultrasound systems and all participants underwent standardized carotid ultrasound examinations [15]. Carotid intima-media thickness (CIMT) and plaque were used to assess CA, and CA was defined as increased intima-media thickness (IMT) or the presence of plaques. Increased IMT was defined as IMT ≥ 1.0 mm in either the left or right carotid artery. Plaque was defined as IMT ≥ 1.5 mm or focal narrowing of the vessel wall of > 50% relative to adjacent segments. Stenosis severity was classified into five categories, and ≥ 50% were defined as stenosis. Covariates Age, sex, education, smoking, drinking, physical activity and history of diseases including cerebrovascular disease, heart diseases, hypertension, dyslipidemia and diabetes were collected by face-to-face interviews. Smoking was defined as smoking at least one cigarette per day for more than half a year. Drinking was defined as regular heavy drinking (≥ 3 times/week). Blood pressure was calculated using the average of 3 measurements at 1-min intervals after 5 min of rest. Hypertension was defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, self-reported hypertension diagnosed by a physician, or use of antihypertensive medications. Statistical analysis Continuous variables were shown as the mean [standard deviation (SD)] and qualitative variables were presented as numbers with percentages. Logistic regression models were conducted to examine the association of AIP-related indices with CA, after adjusting for covariates (age, sex, education, smoking, drinking, physical activity, SBP, TC, LDL-C, history of diseases including cerebrovascular disease, heart diseases, hypertension, dyslipidemia and diabetes). A directed acyclic graph (DAG) was used to identify potential covariates (Figure S1). Restricted cubic spline (RCS) models with 4 knots were applied to explore the dose-response association between AIP-related indices and CA, with the minimum of AIP-related indices are the reference. Subgroup analyses, categorized by sex and age, were conducted to further validate the association between AIP-related indices and CA. Bootstrap internal validation with 1000 resamples was used for internal validation of AUC, NRI, and IDI to reduce overfitting. Furthermore, the net reclassification improvement (NRI) and integrated discrimination improvement (IDI) indices were calculated by comparing models to assess incremental discriminative performance. Finally, receiver operating characteristic (ROC) curves and the areas under the curves (AUCs) were used to compare discriminative ability. Results Basic characteristics of participants are presented in Table 1. The mean age of the included participants was 60.0 ± 10.8 years old, and 55.7% were females. Among them, 9.9% (5833/59123) were diagnosed as CA. Compared with those without CA, participants with CA were more likely to be older, male, smoking, drinking, inactive physical activity, history of diseases, higher BMI, SBP, DBP, FPG, TG, TC, and LDL. Table 1.Characteristics of included participants (N = 59123) (p value was calculated by chi-square test/t-test. A p-value < 0.05 was regarded as statistically significant.). Overall (N = 59123)CA (N = 5833)No CA (N = 53290)p valueAge (years)60.0 ± 10.867.4 ± 9.559.2 ± 10.6<0.001Gender <0.001Male26197 (44.3)3082 (52.8)23115 (43.4) Female32926 (55.7)2751 (47.2)30175 (56.6) Education <0.001Primary school and below15747 (26.6)1701 (29.2)14046 (26.4) Junior middle school23238 (39.3)2276 (39.0)20962 (39.3) Senior high school13624 (23.0)1303 (22.3)12321 (23.1) College and above6514 (11.0)553 (9.5)5961 (11.2) Smoking <0.001Yes7590 (12.8)1793 (30.7)5797 (10.9) No51533 (87.2)4040 (69.3)47493 (89.1) Drinking <0.001Yes8862 (15.0)1147 (19.7)7715 (14.8) No50261 (85.0)4686 (80.3)44470 (85.2) Physical activity <0.001Yes38345 (64.9)3056 (52.4)35289 (66.2) No20778 (35.1)2777 (47.6)18001 (33.8) History of diseases Cerebrovascular disease <0.001Yes1678 (2.8)757 (13.0)921 (1.7) No57445 (97.2)5076 (87.0)52369 (98.3) Heart disease <0.001Yes1997 (3.4)553 (9.5)1444 (2.7) No57126 (96.6)5280 (90.5)51846 (97.3) Dyslipidemia <0.001Yes8471 (14.3)1565 (26.8)6906 (13.0) No50652 (85.7)4268 (73.2)46384 (87.0) Diabetes mellitus <0.001Yes8422 (14.2)1413 (24.2)7009 (13.2) No50701 (85.8)4420 (75.8)46281 (86.8) BMI, kg/m224.0 ± 3.325.3 ± 3.723.8 ± 2.9<0.001SBP, mmHg130.1 ± 16.7139.4 ± 18.8129.1 ± 16.2<0.001DBP, mmHg79.6 ± 10.282.1 ± 11.879.3 ± 10.0<0.001FPG, mmol/L5.6 ± 1.76.2 ± 2.25.5 ± 1.7<0.001TG, mmol/L1.7 ± 1.01.9 ± 1.11.6 ± 0.9<0.001TC, mmol/L4.7 ± 1.34.8 ± 1.44.7 ± 1.3<0.001LDL-C, mmol/L2.6 ± 0.92.7 ± 1.02.6 ± 0.9<0.001HDL-C, mmol/L1.4 ± 0.91.3 ± 0.41.4 ± 0.5<0.001CA: Carotid atherosclerosis; BMI: Body mass index; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; FPG: Fasting plasma glucose; TG: Triglyceride; TC: Total cholesterol; LDL-C: Low-density lipoprotein cholesterol; HDL-C: High-density lipoprotein cholesterol. Characteristics of included participants (N = 59123) (p value was calculated by chi-square test/t-test. A p-value < 0.05 was regarded as statistically significant.). CA: Carotid atherosclerosis; BMI: Body mass index; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; FPG: Fasting plasma glucose; TG: Triglyceride; TC: Total cholesterol; LDL-C: Low-density lipoprotein cholesterol; HDL-C: High-density lipoprotein cholesterol. Associations between AIP-related indices and CA The associations of AIP-related indices with CA, CIMT, plaque and stenosis are showed in Table 2. IR-related indices were positively associated with CA, carotid intima-media thickness (CIMT), plaque and stenosis, after adjusting for potential covariates. Per 1-standard deviation (SD) increase in AIP-BMI (OR = 1.39, 95% CI = 1.35–1.43), AIP-WC (1.39, 1.34–1.43), and AIP-WHtR (1.38, 1.34–1.42) were associated with CA. Per 1-standard deviation (SD) increase in AIP-BMI (OR = 1.36, 95% CI = 1.31–1.41), AIP-WC (1.34, 1.29–1.38), and AIP-WHtR (1.33, 1.28–1.38) were associated with CIMT. Per 1-standard deviation (SD) increase in AIP-BMI (OR = 1.32, 95% CI = 1.28–1.37), AIP-WC (1.34, 1.29–1.38), and AIP-WHtR (1.33, 1.29–1.38) were associated with plaques. AIP-BMI, AIP-WC and AIP-WHtR were positively associated with stenosis when they were divided into interquartile range (IQR), compared with the Q1 group. Table 2.Association of AIP-related indices with CA, CIMT, plaques and stenosis with logistic regression models (N = 59123) (p trend was assessed by including the median values of exposure within each category as a continuous variable in the model. A p-value < 0.05 was regarded as statistically significant.). Quartiles of index p trend Per 1-SD increaseQ1Q2Q3Q4CA AIP-BMI1.39 (1.35–1.43)Ref1.20 (1.09–1.32)1.54 (1.40–1.70)2.38 (2.17–2.60)<0.001AIP-WC1.39 (1.34–1.43)Ref1.16 (1.44–1.73)1.58 (1.44–1.73)2.26 (2.06–2.47)<0.001AIP-WHtR1.38 (1.34–1.42)Ref1.18 (1.06–1.30)1.57 (1.43–1.73)2.26 (2.06–2.47)<0.001CIMT AIP-BMI1.36 (1.31–1.41)Ref1.21 (1.07–1.37)1.67 (1.49–1.88)2.26 (2.02–2.53)<0.001AIP-WC1.34 (1.29–1.38)Ref1.23 (1.09–1.39)1.65 (1.47–1.85)2.14 (1.91–2.40)<0.001AIP-WHtR1.33 (1.28–1.38)Ref1.21 (1.07–1.37)1.68 (1.50–1.89)2.11 (1.89–2.37)<0.001Plaques AIP-BMI1.32 (1.28–1.37)Ref1.07 (0.96–1.20)1.26 (1.13–1.40)2.07 (1.87–2.29)<0.001AIP-WC1.34 (1.29–1.38)Ref1.01 (0.90–1.13)1.30 (1.17–1.45)2.00 (1.81–2.21)<0.001AIP-WHtR1.33 (1.29–1.38)Ref1.04 (0.93–1.17)1.27 (1.14–1.42)2.02 (1.83–2.23)<0.001Stenosis AIP-BMI1.08 (0.99–1.18)Ref1.53 (1.16–2.04)1.71 (1.31–2.25)1.52 (1.15–1.99)0.008AIP-WC1.07 (0.98–1.16)Ref1.53 (1.16–2.02)1.66 (1.27–2.17)1.34 (1.02–1.76)0.101AIP-WHtR1.06 (0.98–1.16)Ref1.58 (1.19–2.08)1.62 (1.24–2.12)1.38 (1.05–1.81)0.069Adjusted for age, sex, education, smoking, drinking, physical activity, SBP, TC, LDL-C, history of diseases including cerebrovascular disease, heart diseases, hypertension, dyslipidemia and diabetes.CA: Carotid atherosclerosis; CIMT: Carotid intima media thickness; AIP: atherogenic index of plasma; AIP-BMI: AIP-body mass index; AIP-WC: AIP-waist circumference; AIP-WHtR: AIP-waist to-height ratio. Association of AIP-related indices with CA, CIMT, plaques and stenosis with logistic regression models (N = 59123) (p trend was assessed by including the median values of exposure within each category as a continuous variable in the model. A p-value < 0.05 was regarded as statistically significant.). Adjusted for age, sex, education, smoking, drinking, physical activity, SBP, TC, LDL-C, history of diseases including cerebrovascular disease, heart diseases, hypertension, dyslipidemia and diabetes. CA: Carotid atherosclerosis; CIMT: Carotid intima media thickness; AIP: atherogenic index of plasma; AIP-BMI: AIP-body mass index; AIP-WC: AIP-waist circumference; AIP-WHtR: AIP-waist to-height ratio. The dose-response relationships of AIP-related indices with CA, CIMT, plaque and stenosis are presented in Figure S2. RCS results indicated that AIP-related indices were positively associated with CA, CIMT and plaque in a non-linearity way (Pnon-linearity < 0.05). Subgroup analysis by age and sex is shown in Figure S3-S5. The results were similar to those in the total population. The associations of AIP-related indices with CA, CIMT and plaque were statistically significant among all the subgroups; nevertheless, the associations of AIP-related indices with stenosis were significant among females and those aged ≥ 60, not among males and those aged below 60. Incremental discriminative values of AIP-related indices for CA Table 3 shows the performance for CA when AIP-related indices were added in the basic model. Among all these indicators, AIP-BMI (NRI = 0.034, p = 0.009; IDI = 0.009, p < 0.001), AIP-WC (NRI = 0.034, p = 0.009; IDI = 0.009, p < 0.001), and AIP-WHtR (NRI = 0.034, p = 0.007; IDI = 0.010, p < 0.001) showed statistically significant but modest incremental value for CA.; AIP-BMI exhibited modest incremental value for CIMT (NRI = 0.019, p < 0.008; IDI = 0.005, p < 0.001) and plaques (NRI = 0.025, p = 0.007; IDI = 0.008, p < 0.001); AIP-WC for stenosis (NRI = 0.004, p = 0.003). Table 3.Discriminative performance of models with AIP-related indices for CA, CIMT, plaques and stenosis (N = 59123) (p values were calculated by comparing models to assess whether adding AIP-related indices could improve the performance of CA. A p-value < 0.05 was regarded as statistically significant.). NRI p IDI p CA Basic modelRefRefRefRefBasic model + AIP0.0240.0080.006<0.001Basic model + AIP-BMI0.0340.0090.009<0.001Basic model + AIP-WC0.0340.0090.009<0.001Basic model + AIP-WHtR0.0340.0070.010<0.001CIMT Basic modelRefRefRefRefBasic model + AIP0.0120.0060.003<0.001Basic model + AIP-BMI0.0190.0080.005<0.001Basic model + AIP-WC0.0180.0080.004<0.001Basic model + AIP-WHtR0.0170.0070.004<0.001Plaques Basic modelRefRefRefRefBasic model + AIP0.0180.0060.004<0.001Basic model + AIP-BMI0.0250.0070.008<0.001Basic model + AIP-WC0.0240.0080.006<0.001Basic model + AIP-WHtR0.0240.0080.006<0.001Stenosis Basic modelRefRefRefRefBasic model + AIP0.0020.0020.0000.534Basic model + AIP-BMI0.0000.0050.0000.595Basic model + AIP-WC0.0040.0030.0000.502Basic model + AIP-WHtR0.0020.0020.0000.489Basic model: adjusted for age, sex, education, smoking, drinking, physical activity, SBP, TC, LDL-C, history of diseases including cerebrovascular disease, heart diseases, hypertension, dyslipidemia and diabetes.CA: Carotid atherosclerosis; CIMT: Carotid intima media thickness; AIP: atherogenic index of plasma; AIP-BMI: AIP-body mass index; AIP-WC: AIP-waist circumference; AIP-WHtR: AIP-waist to-height ratio; NRI: net reclassification improvement; IDI: integrated discrimination improvement. Discriminative performance of models with AIP-related indices for CA, CIMT, plaques and stenosis (N = 59123) (p values were calculated by comparing models to assess whether adding AIP-related indices could improve the performance of CA. A p-value < 0.05 was regarded as statistically significant.). Basic model: adjusted for age, sex, education, smoking, drinking, physical activity, SBP, TC, LDL-C, history of diseases including cerebrovascular disease, heart diseases, hypertension, dyslipidemia and diabetes. CA: Carotid atherosclerosis; CIMT: Carotid intima media thickness; AIP: atherogenic index of plasma; AIP-BMI: AIP-body mass index; AIP-WC: AIP-waist circumference; AIP-WHtR: AIP-waist to-height ratio; NRI: net reclassification improvement; IDI: integrated discrimination improvement. Table 4 and Figure 1 shows the discriminative performance of AIP-related indices for CA. The ROC analyses revealed that AIP-BMI had a relatively higher AUC for CA and CIMT, with the AUC of 0.833 (0.828, 0.838) and 0.817 (0.811, 0.824), respectively. AIP-BMI, AIP-WC and AIP-WHtR similar high AUC for plaques, with the AUC of 0.842 (0.836, 0.848), 0.842 (0.836, 0.847), and 0.842 (0.836, 0.848), respectively. Meanwhile, all the four indices (AIP, AIP-BMI, AIP-WC and AIP-WHtR) showed similar performance for stenosis, with the AUC of 0.871 (0.858, 0.884). Although AUC values were high, between-index differences were small and of uncertain clinical significance. Table 4.Discriminative performance of AIP-related indices for CA, CIMT, plaques and stenosis by receiver operating characteristic curves and the areas under the curves (N = 59123) (p values were calculated by comparative diagnostic performance of AIP-related indices. A p-value < 0.05 was regarded as statistically significant.). Sensitivity (%)Specificity (%)YoudenAUC (95% CI) p CA AIP81.0169.980.5100.830 (0.825, 0.835)RefAIP-BMI80.9270.560.5150.833 (0.828, 0.838)<0.001AIP-WC80.7770.450.5120.832 (0.827, 0.837)<0.001AIP-WHtR81.9169.310.5120.832 (0.827, 0.837)<0.001CIMT AIP77.8370.920.4880.814 (0.808, 0.821)RefAIP-BMI79.5169.530.4900.817 (0.811, 0.824)<0.001AIP-WC82.6166.420.4900.816 (0.811, 0.824)<0.001AIP-WHtR81.0867.970.4910.816 (0.810, 0.823)<0.001Plaques AIP80.5072.560.5310.840 (0.834, 0.846)RefAIP-BMI80.5572.750.5330.842 (0.836, 0.848)<0.001AIP-WC80.3872.940.5330.842 (0.836, 0.847)<0.001AIP-WHtR80.5572.720.5330.842 (0.836, 0.848)<0.001Stenosis AIP80.4778.980.5950.871 (0.858, 0.884)RefAIP-BMI80.2979.290.5960.871 (0.858, 0.884)0.021AIP-WC80.2979.300.5960.871 (0.858, 0.884)0.014AIP-WHtR80.2979.180.5950.871 (0.858, 0.884)0.013CA: Carotid atherosclerosis; CIMT: Carotid intima media thickness; AIP: atherogenic index of plasma; AIP-BMI: AIP-body mass index; AIP-WC: AIP-waist circumference; AIP-WHtR: AIP-waist to-height ratio; AUCs: areas under the curves. Discriminative performance of AIP-related indices for CA, CIMT, plaques and stenosis by receiver operating characteristic curves and the areas under the curves (N = 59123) (p values were calculated by comparative diagnostic performance of AIP-related indices. A p-value < 0.05 was regarded as statistically significant.). CA: Carotid atherosclerosis; CIMT: Carotid intima media thickness; AIP: atherogenic index of plasma; AIP-BMI: AIP-body mass index; AIP-WC: AIP-waist circumference; AIP-WHtR: AIP-waist to-height ratio; AUCs: areas under the curves. ROC curves of AIP-related indices for diagnosing (A) CA, (B) CIMT, (C) plaques and (D) stenosis (N = 59123). ROC: Receiver operating characteristic; AIP: Atherogenic index of plasma; CA: Carotid atherosclerosis; CIMT: Carotid intima media thickness; AIP-BMI: AIP-body mass index; AIP-WC: AIP-waist circumference; AIP-WHtR: AIP-waist to-height ratio. Sensitivity analyses were performed by further excluding TC and LDL-C in the covariates. The results were similar to those adjusting TC and LDL-C, including the association of AIP-related indices with CA, CIMT, plaques and stenosis (Table S1), discriminative performance of models with AIP-related indices for CA, CIMT, plaques and stenosis (Table S2), discriminative performance of AIP-related indices for CA, CIMT, plaques and stenosis (Table S3), and ROC curves of AIP-related indices for diagnosing CA, CIMT, plaques and stenosis (Figure S6). Discussion The present study demonstrated significant associations between AIP-related indices (i.e. AIP-BMI, AIP-WC, and AIP-WHtR) and subclinical CA. AIP-BMI showed relatively higher discriminative performance for CA and CIMT, while the three composite indices performed similarly for plaques. All four indices showed comparable performance for stenosis. Our findings revealed the positive association between AIP-related indices and CA. Results were robust when adjusting for covariates and in most subgroup analysis, which could promise the reliability of our results. Most previous studies just focused on AIP alone, and indicated that AIP was positively associated with several adverse outcomes, including hypertension [3], cardiorenal syndrome [16], stroke [17,18], cardiometabolic diseases [19,20], acute myocardial infarction [21], diabetes [22], etc. Our study extends these findings by evaluating composite markers that integrate lipid and obesity-related information. In the subgroup analysis, the positive association between AIP-related indices and stenosis were only significant among females and those aged over 60 years. Inconsistent findings regarding the relationship between AIP and stenosis were also reported in previous studies. For example, Huang et al. found that an elevated AIP was associated with a higher prevalence of CA, but not stenosis among 52380 subjects [9]. Zhao et al. showed that AIP were closely related to the degree of carotid plaque stenosis among 336 patients with acute ischemic stroke [23]. Garg et al. revealed that AIP was associated with a higher prevalence of carotid artery stenosis among a single center case-controlled study with 31 cases and 236 controls [24]. Guo et al. found no significant association between AIP and stenosis among 5314 participants [25]. These discrepancies may arise from differences in population characteristics, sample size, stenosis definition, covariates, and disease severity. With regard to the different results on age groups, it is well-known that age is an established risk factor of CA, which may contribute to a higher prevalence of CA [26,27]. As to sex, the prevalence of CA among males is usually higher than that among females, as males are more likely to smoking drinking and unhealthy lifestyles, which are related to CA [26,28]. However, studies also indicated that the male predominance in atherosclerosis declined after the age of 50 years, the CA prevalence being similar in elderly men and women, which may be related to an adverse effect of menopause on atherosclerosis [29]. And women in our study were just middle-aged and older adults. Compared with AIP alone, AIP-related indices showed modestly improved discriminative performance for CA, with the exception of stenosis. These findings were similar to the association between AIP-related indices and stroke [13]. Neeland et al. showed that the assessment of visceral fat is essential for assessing TG concentration, especially for BMI, WC and WHtR [30]. Besides, several studies revealed that incorporating obesity-related indicators, such as BMI, WC, and WHtR, provides a more comprehensive evaluation of cardiovascular diseases [31,32]. Thus, AIP-related indices may support more comprehensive cardiovascular risk assessment. Bootstrap validation was also conducted to address potential overestimation of model performance due to overfitting. The corrected estimates remained consistent, supporting the reliability of our findings. Although the present study found statistically significant improvements in discriminative ability with AIP-related indices, the absolute increases in NRI and IDI were modest, and the between-index differences in AUC were small. These findings suggest that the observed gains may not confer substantial advantages in routine clinical practice. Therefore, AIP-related indices should be regarded as convenient, low-cost screening tools for population-based risk stratification rather than standalone diagnostic or predictive biomarkers. Statistical significance should not be overinterpreted as clinical utility. The mechanism underlying the association between AIP-related indices and CA is multifactorial. AIP reflects atherogenic lipid profiles and anthropometric markers represent central obesity and adipose dysfunction; their combination integrates lipid and metabolic obesity phenotypes, which jointly contribute to atherogenesis. Although presented as a simple multiplicative combination, these composite indices provide a practical and accessible summary of combined cardiometabolic risk, which may be more informative than either component alone in community settings. First, AIP-related indices were related with the regulation of cholesterol, higher AIP-related indices mean more cholesterol in the inner layers of the atherosclerotic arteries, which may eventually lead to the formation of atherosclerosis [33], however, these simple multiplicative combinations require further biological validation. Second, it was also combined that obesity-related indices, dysfunctional adipose tissue, particularly visceral fat, could result in damage to arterial, and cardiac tissues, thereby promoting the occurrence of CA [34]. Third, AIP-related indices could promote the secretion of pro-inflammatory factors, resulting in inflammatory responses and oxidative stress, which drive endothelial dysfunction, plaque initiation, and progression [35] and are also considered in the progression of CA [36,37]. AIP-related indices may therefore serve as integrated markers of metabolic and inflammatory burden [38,39]. This study has several strengths. First, the large sample size of middle-aged and older Chinese adults enhanced statistical power and the reliability of the estimated associations. Second, we systematically constructed and comprehensively compared multiple AIP-related indices (AIP-BMI, AIP-WC, and AIP-WHtR) with detailed carotid atherosclerosis phenotypes including CIMT, plaque, and stenosis. Third, we applied a series of rigorous statistical methods, including logistic regression, restricted cubic splines, ROC analysis, NRI, IDI, and bootstrap internal validation, to ensure robust evaluation of the associations and discriminative performance. Fourth, the studied indices are derived from routine laboratory and anthropometric measurements, supporting their potential use in community-based risk assessment. The limitations should also be acknowledged. First, the cross-sectional design only permits identification of associations, not causal inference, and that AIP-related indices may reflect underlying metabolic disturbances rather than independent causal factors. Second, the study population was derived from a regional health screening program in China, which may limit generalizability to other ethnicities, regions, and healthcare settings. Third, residual confounding from unmeasured factors including detailed dietary intake, specific medication use (especially statins), and socioeconomic status, which were not fully captured in the current dataset. And the predictive performance of the models was reported without internal validation, which may raise potential overfitting and overestimation of model performance. Fourth, although statistically significant, NRI/IDI improvements and OR effect sizes were modest, and the clinical meaningfulness required further validation. Fifth, the potential mechanism needs to be confirmed in future studies. Besides, this study did not directly compare AIP-related indices with well-established cardiovascular risk scores such as the Framingham risk score or ASCVD risk score due to data limitation. Therefore, the incremental clinical value of these indices relative to conventional risk models remains unclear. Last but not least, subgroup findings should be interpreted cautiously due to multiple comparisons. Besides, associations with stenosis were inconsistent, which may reflect differences between early and advanced carotid lesions. Conclusion AIP-related indices were positively associated with subclinical CA, CIMT, plaque and stenosis. Although statistically significant, the effect sizes and improvements in reclassification were modest. These indices may serve as accessible surrogate markers for risk assessment. Future prospective studies are needed to confirm our findings and clarify their clinical utility. Supplementary Material Supplementary materials.docx Acknowledgements The contributions of all the participants, staffs of Stroke Screening and Prevention Project in Hubei Province, and all members of study team are greatly acknowledged. WZL performed the statistical analysis and wrote the manuscript. LAT, DJC and WZL conceived and designed the study, and interpreted data. DJC, HZ, SY critically revised the manuscript and approved the final version. LAT and DJC are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Funding Statement The study was supported by National Natural Science Foundation of China (42507573). The funder did not play any role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; nor in the preparation, review, or approval of the manuscript. Ethics approval and consent to participate The study was performed according to the guidelines of the Helsinki Declaration. This study was based on the Stroke Screening and Prevention Project in Hubei Province. The written informed consents were obtained from all the participants. The study protocol was approved by the Ethics Committee of Xuan Wu Hospital, Capital Medical University (Beijing, China) ([2015]024). Consent for publication Not applicable. Disclosure statement No potential conflict of interest was reported by the author(s). Data availability statement The data are available with reasonable request from the corresponding author. References Associated Data Supplementary Materials Supplementary materials.docx Data Availability Statement The data are available with reasonable request from the corresponding author.