Peripheral arterial disease (PAD) is a prevalent disorder of the peripheral vasculature. Its hallmark pathology is atherosclerosis in the arteries of the lower extremities, resulting in arterial stenosis or occlusion and consequent haemodynamic compromise (1, 2). Clinically, PAD often presents as intermittent claudication, specifically, ischaemic muscular pain triggered by walking and relieved by rest. Sustained ischaemia may lead to tissue loss or even gangrene in the affected limb (3, 4). The global incidence of PAD is increasing, especially in older adults, with a prevalence exceeding 18% in individuals over 60 years of age (5). Despite the availability of numerous treatment strategies, the rate of PAD recurrence remains high (6, 7), posing a major challenge and placing a substantial economic burden on health care systems. Accordingly, the identification of modifiable risk factors for PAD is crucial for its effective prevention and treatment.

The development of PAD involves a complex interplay of genetic, environmental, dietary, lifestyle, and metabolic factors. Epidemiological studies have shown links between PAD risk and multiple factors, including sex, age, race, smoking status, alcohol consumption, and comorbid conditions such as hypertension and diabetes (8–12). Among these, dyslipidaemia plays a pivotal role in PAD progression by promoting atherosclerosis in the arteries of the lower extremities (13, 14). In particular, elevated levels of low-density lipoprotein cholesterol (LDL-C) and total cholesterol have been shown to contribute significantly to PAD development (14, 15), whereas higher high-density lipoprotein cholesterol (HDL-C) levels are associated with a lower PAD risk (16, 17). These divergent effects of LDL-C and HDL-C underscore the importance of cholesterol management in the maintenance of vascular health.

Compared with individual lipid measures, the non-HDL-C to HDL-C ratio (NHHR) has recently emerged as an innovative composite lipid index that may offer a better reflection of overall atherogenic risk. Unlike isolated measures such as LDL-C or total cholesterol (TC), the NHHR simultaneously captures both atherogenic (non-HDL-C) and protective (HDL-C) lipid components, thus integrating the balance between cholesterol deposition and clearance (18). Increasing evidence indicates that compared with traditional lipid parameters, the NHHR has superior ability to predict cardiovascular and cerebrovascular risk (19, 20). In addition, the NHHR offers several practical advantages over other emerging lipid indices. First, it is simple to calculate, requiring only TC and HDL-C values that are routinely available in clinical testing. Second, the NHHR is stable across fasting and non-fasting states, enhancing its utility in real-world clinical assessments (18). Third, its broad applicability has been demonstrated across various metabolic and vascular conditions—including cardiovascular disease (21), thyroid dysfunction (22), depression (23), kidney stones (24), suicidal ideation (25), osteoporosis (26), and even abdominal aortic aneurysm (AAA) (27).

Given that PAD is essentially an atherosclerotic disease, a higher NHHR—reflecting an imbalance between atherogenic and protective lipoproteins—may contribute to endothelial dysfunction, inflammation, and lipid deposition in peripheral arteries (28), thereby increasing the risk of PAD. Previous studies have indicated that in patients with PAD, the NHHR has a U-shaped relationship with major adverse cardiovascular events (29). These findings suggest that dynamic monitoring of the NHHR may aid in risk stratification and personalized lipid management for PAD patients. Furthermore, owing to variations in genetic, environmental, and lifestyle factors, the risk factors for and disease progression of PAD may differ among populations. As the incidence and clinical burden of PAD increase sharply with advancing age, individuals aged ≥50 years represent the population at highest risk and constitute the main demographic group undergoing vascular evaluation and surgical intervention (30, 31). Focusing on this group ensures clinical relevance and adequate event rates while minimizing the heterogeneity related to age-dependent metabolic variation. Therefore, the present study aimed to clarify the relationship between the NHHR and PAD in a Chinese inpatient population aged ≥50 years undergoing vascular surgery to inform the early clinical assessment and management of PAD.

The findings of this cross-sectional analysis of 3,532 inpatients aged ≥50 years who underwent vascular surgery indicated that the NHHR was inversely associated with the prevalence of PAD after controlling for confounders. Patients in the third NHHR quartile were found to have the lowest PAD risk. The relationship between the NHHR and PAD was nonlinear and J shaped, with an inflection of approximately NHHR = 2.75. NHHR values below this threshold were protective against PAD, whereas those above 2.75 were associated with a higher PAD risk. Moreover, this association was significantly modified by sex and was more pronounced among women. These results indicate that the risk of PAD in relation to the NHHR is affected by patient subgroup (e.g., patient sex), underscoring the need for personalized risk assessment and management strategies.

The present findings align with prior evidence linking dyslipidaemia to PAD risk in the lower extremities. For example, Bertrand et al. (17) reported that higher HDL-C levels were associated with lower PAD prevalence and incidence in patients with diabetes, whereas higher TC/HDL-C ratios and non-HDL-C levels were associated with a greater risk of PAD. Similarly, Liu et al. (14) observed in a community-based study that cumulative exposure to elevated LDL-C levels was an independent risk factor for PAD development, whereas non-HDL-C levels had no significant predictive value. In a large study of hypertensive adults, Ding et al. noted that several composite lipid indices (TC/HDL-C, TG/HDL-C, LDL-C/HDL-C, and non-HDL-C) were independently and positively correlated with PAD, with the TC/HDL-C and LDL-C/HDL-C ratios being the most strongly predictive (16). These studies confirm the importance of abnormal lipid profiles in PAD pathogenesis and emphasize the importance of aggressive lipid management for the prevention of PAD in at-risk patients.

In contrast to earlier studies conducted in general or diabetic populations, the present analysis focused specifically on an older Chinese inpatient cohort at high risk for PAD, evaluating the use of the NHHR, a novel composite lipid marker that integrates both atherogenic (non-HDL-C) and protective (HDL-C) cholesterol components, for a more holistic assessment of PAD risk. Notably, a clear threshold effect was identified in the NHHR–PAD relationship, and this relationship was modified by patient sex, findings that have not been reported previously. These results contribute to the literature by demonstrating that the NHHR is associated with PAD prevalence in an older vascular surgical patient population while also emphasizing the importance of considering patient sex when evaluating the NHHR for PAD risk stratification.

The NHHR may influence the prevalence of PAD through various mechanisms. Disturbances in glucose and lipid metabolism are important contributors to PAD pathogenesis. Insulin resistance can reduce the activity of lipoprotein lipase, impair catabolism of very-low-density lipoprotein (VLDL), and promote hepatic production of apoB100-containing lipoproteins, thereby increasing plasma triglyceride (TG) and remnant cholesterol (RC) levels (36). An elevated NHHR indicates an atherogenic, “lipotoxic” state characterized by endoplasmic reticulum stress, mitochondrial dysfunction, and oxidative stress, which can aggravate insulin resistance and pancreatic β-cell dysfunction (37, 38). Moreover, an increased TC: HDL-C ratio is associated with increased PAD risk, suggesting that imbalances in lipid ratios contribute to the formation of atherosclerotic plaques (14, 16). Conversely, extremely elevated HDL-C levels may reflect impaired reverse cholesterol transport or HDL dysfunction, potentially promoting cholesterol deposition in vascular tissues and contributing to atherogenesis (19, 39). This paradoxical association between very high HDL-C levels and adverse outcomes might partially explain the J-shaped relationship observed between the NHHR and PAD risk in the present study, suggesting that both very low and very high NHHR values could be linked to increased risk.

An elevated NHHR reflects a proinflammatory milieu. Small, dense LDL particles are readily oxidized to oxLDL, which is subsequently taken up by macrophages via scavenger receptors, resulting in the formation of foam cells. These foam cells secrete inflammatory cytokines (e.g., IL-6 and TNF-α), exacerbating vascular inflammation and atherosclerosis (36, 40). Clinically, higher levels of non-HDL-C are correlated with increased activity of proinflammatory macrophages, especially in obesity (41). Thus, an elevated NHHR likely perpetuates inflammation by increasing the levels of oxidized lipids and activating macrophage-driven inflammatory responses, accelerating PAD progression.

Endocrine factors exert a substantial influence on PAD risk partly through their regulation of lipid and glucose metabolism. For instance, insulin deficiency or resistance can downregulate hepatic LDL receptor expression and impair LDL clearance while simultaneously increasing hormone-sensitive lipase activity, leading to elevated free fatty acid levels and atherogenic dyslipidaemia (36). In addition, sex hormones may account for the sex-specific differences observed in the NHHR–PAD association. Previous studies have shown that higher HDL-C concentrations do not always confer cardiovascular benefit in women and, in some cases, are even associated with increased coronary risk (42). These paradoxical findings are attributed to HDL functional heterogeneity—in which HDL particles may become dysfunctional under oestrogen deficiency, oxidative stress, or chronic inflammation. In women, particularly those in the perimenopausal or postmenopausal stages, the withdrawal of oestrogen alters HDL composition and impairs ApoA1-mediated cholesterol efflux and antioxidative activity. Consequently, despite higher HDL-C levels, the overall functionality of HDL may decline, reducing its vascular protective effects (42). This decline in HDL functionality may explain why the association between the NHHR and PAD appears stronger in women. Therefore, future investigations should focus on the quality of HDL rather than solely its quantity, particularly in the context of menopause, to better understand the underlying mechanisms affecting cardiovascular risk in women.

Moreover, a high NHHR may also harm endothelial cells through oxidative mechanisms, thereby promoting the occurrence of PAD. Oxidized LDL impairs the function of endothelial nitric oxide synthase (eNOS), lowering nitric oxide availability and promoting vasoconstriction and platelet aggregation (36). OxLDL also triggers the apoptosis of endothelial cells, increases vascular permeability, and promotes monocyte adhesion and migration, accelerating plaque development (36, 43). Moreover, imbalances in lipid ratios (e.g., high LDL-C/HDL-C) can activate the NF-κB pathway in endothelial cells, exacerbating local inflammation (16). Thus, it is likely that an elevated NHHR contributes to PAD by inducing endothelial dysfunction and inflammation in arterial walls.

Additionally, genetic susceptibility may modulate the influence of dyslipidaemia on the risk of PAD. For example, the apolipoprotein Eε4 allele is linked to increased susceptibility to PAD development, likely because of alterations in lipoprotein metabolism and increased inflammation (44). Similarly, mutations in genes associated with familial hypercholesterolemia (e.g., LDLR and PCSK9) result in lifelong elevated levels of LDL-C, which accelerate atherosclerosis and PAD development (14). These genetic factors may partly explain interindividual variability in PAD risk despite the presence of similar lipid profiles, suggesting a role for genetic screening in high-risk patients.

The findings of the present study have significant clinical implications for assessing the risk of PAD. The observed nonlinear (J-shaped) relationship between the NHHR and PAD suggests the need for a nuanced and personalized approach. Specifically, NHHR values above approximately 2.75 were associated with an increasing risk of PAD, while values well below this threshold were also linked to a higher PAD risk. This finding indicates that while a low non-HDL-C/high HDL-C profile is generally viewed as favourable, excessively low NHHR values may paradoxically be detrimental to the health of peripheral arteries.

Clinicians should be mindful of this threshold effect, recognizing that patients with either very low or very high NHHR levels may both require closer monitoring, rather than assuming that “lower is always better” in terms of the NHHR. Regular monitoring of the NHHR may provide useful information for clinicians assessing PAD risk in patients undergoing vascular surgery. Identifying patients at the extremes of the NHHR spectrum may facilitate early preventive interventions—such as tailored lipid-modifying therapy and optimization of metabolic profiles—aimed at mitigating PAD risk and promoting vascular health. Additionally, sex interaction effects in the context of NHHR and PAD risk were explored in this study. Preliminary analyses suggest that the relationship between the NHHR and PAD may differ by sex, highlighting the need for further investigations to replicate these findings and confirm their clinical relevance.

A key strength of the present study is its use of three analytic models (unadjusted, partially adjusted, and fully adjusted) to assess the robustness of the association between the NHHR and PAD, thereby minimizing potential confounding bias. Additionally, unlike the ABI, which requires specialized equipment and training to measure, the NHHR can be readily determined from routine bloodwork. These findings suggest that the NHHR may be a potentially useful marker for identifying patients who warrant further evaluation for PAD risk, but validation through prospective studies is needed to confirm its clinical utility.

Nevertheless, this study has several limitations. First, its cross-sectional design limits causal inference, and it cannot be determined whether an elevated NHHR causes PAD or is merely associated with it. Prospective studies are needed to clarify causality. Second, 2,391 patients were excluded primarily because of incomplete data, along with a small number of patients who had recent acute illnesses or significant health issues. While multiple imputation analyses supported the robustness of our findings, the potential influence of missing data and these excluded conditions on the results cannot be completely ruled out. Third, patients receiving lipid-modifying medications were excluded. This exclusion was primarily due to our inability to obtain detailed medication data on the use of drugs that affect lipid metabolism. Since the use of such medications can substantially influence the outcomes of lipid profiles and related analyses, we opted to exclude these patients to maintain the integrity of our results. While this improved internal validity by isolating the effect of the NHHR, it may have reduced external validity, as many real-world patients are on such therapies. Excluding patients receiving lipid-modifying therapy may limit the applicability of our results to real-world patients who are on standard lipid management. Therefore, future studies should consider a more comprehensive patient population, such as one including medication history, to increase the generalizability of the findings. Fourth, the sample size (n = 3,532) and single-centre nature may limit the generalizability of the results and the statistical power for subgroup analyses. Additionally, the data were collected from inpatients aged ≥50 years who underwent vascular surgery at a single tertiary hospital in China, which may introduce clustering effects related to the performing surgeon, surgical techniques, and institutional practices. Larger multicentre studies involving more diverse populations are warranted to confirm these findings. Fifth, while we have discussed the association between the NHHR and PAD, the possibility of reverse causation cannot be completely ruled out; specifically, the presence of PAD may influence lipid metabolism and subsequently affect the NHHR. Addressing these issues in future research will provide clearer insights into the relationship between the NHHR and PAD, helping to differentiate the causal pathways and interactions involved. Sixth, although we adjusted for several important covariates, unmeasured confounding factors such as body mass index, renal function, sex hormones, inflammatory markers, dietary habits, physical activity, genetic predisposition, and potential misclassification of PAD may still influence the observed associations. Therefore, future studies that encompass a broad range of lifestyle factors, biological markers, and genetic data are essential for further understanding the role of the NHHR in PAD. Finally, while a nonlinear association between the NHHR and PAD, as well as a potential sex-related interaction, was identified, these findings are exploratory in nature. The underlying biological mechanisms require further investigation, highlighting the need for additional mechanistic and prospective studies.

This study revealed an association between the NHHR and the presence of peripheral artery disease, with evidence suggesting a nonlinear pattern and potential sex-specific differences. Importantly, owing to the retrospective cross-sectional design, these findings do not support causal inference or strong predictive claims. These results indicate that this biomarker may help identify individuals who could benefit from further clinical evaluation rather than serve as a definitive screening or predictive tool. Future prospective studies are necessary to clarify the temporal relationship and to determine its clinical value in risk assessment and decision-making.