Background

Hyperlipidemia, characterized by elevated lipid levels, is a prevalent metabolic disorder with substantial implications for cardiovascular health and mortality1,2,3. Despite the extensive utilization of lipid-lowering medications, the prevalence of hyperlipidemia persists at alarming levels, highlighting the urgent for the exploration of novel therapeutic targets2,3.

Klotho, primarily represented by α-Klotho, alongside β-Klotho and γ-Klotho, is a transmembrane anti-aging protein4. α-Klotho is a type I single-pass transmembrane protein. Its extracellular domain can undergo cleavage by secretory enzymes to enter the bloodstream, urine and cerebrospinal fluid in soluble form5,6. Serum α-Klotho exerts its wide-ranging biological effects on various targets, even several distal organs, functioning as a hormone. The absence of α-Klotho may lead to aging-related symptoms, such as a shortened lifespan, vascular dysfunction, and cognitive impairment7. Conversely, elevated α-Klotho levels have demonstrated the potential to delay aging8. Deficiency or abnormal expression of α-Klotho has been observed in diseases linked to aging, chronic kidney disease (CKD), cardiovascular disease (CVD), diabetes, and osteoporosis. Serum α-Klotho contributes to antioxidant stress, anti-inflammatory responses, ion channel regulation, phosphate homeostasis, glucose metabolism improvement, and insulin/IGF-1 signaling pathways suppression. This indicates that the impact of α-Klotho on aging is probably linked to metabolic regulation. Soluble α-Klotho is negatively associated with body weight and waist circumference9,10, implying its involvement in energy balance regulation. Although α-Klotho inhibits insulin signaling, it can still enhance glucose-induced insulin secretion11 and is negatively linked to blood glucose levels12. Additionally, investigations have indicated that α-Klotho offers protection against adult cardiac metabolic diseases13,14, improving pressure-induced cardiac hypertrophy and remodeling15. Additionally, serum α-Klotho levels are negatively linked to the likelihood of developing CVD13. Klotho provides cardiac protection via inhibiting cardiac oxidative stress16 and attenuating cardiac fibrosis17. Furthermore, α-Klotho helps maintain vascular function by suppressing endothelial cell apoptosis and preventing vascular calcification.

Hyperlipidemia, a known risk factor for CVD, has been reported to be linked to α-Klotho, but findings are inconsistent. Previous investigations have indicated a link between soluble α-Klotho and hyperlipidemia in 40–79 aged people. Specifically, these studies have observed a negative association between α-Klotho and plasma triglyceride (TG) levels18,19,20. Interestingly, in another research, high-density lipoprotein (HDL) demonstrated a favorable association with female α-Klotho and a negative association with male α-Klotho21. Additionally, the negative link between TG and α-Klotho was only noted in females. Hence, the potential link between α-Klotho and hyperlipidemia may also exhibit sex difference, an aspect that has not been reported previously.

The present utilized representative data from the National Health and Nutrition Examination Survey (NHANES) to elucidate the link between serum α-Klotho levels and hyperlipidemia, particularly across different sexes. The primary objective was to evaluate whether the serum α-Klotho levels could be a predictive biomarker for hyperlipidemia.

Materials and methods

Study population and design

The data for this research were sourced from the NHANES, a comprehensive initiative to assess the nutritional and health status of the non-institutionalized civilian population. NHANES utilizes a thorough sampling design, employing a complex, multi-stage, and stratified probability method to assess the prevalence of prevalent health conditions and related risk factors. The NHANES protocol received approval from the Research Ethics Review Board of the National Center for Health Statistics. Prior to participation, written informed consent was diligently obtained from all enrolled participants. Detailed information regarding NHANES protocols and procedures can be obtained online at http://www.cdc.gov/nchs/NHANES.htm.

To increase sample size and mitigate potential sampling errors, this investigation combined data spanning NHANES cycles from 2013 to 201622. An initial cohort of 20,146 participants was enrolled in NHANES from 2013 to 2016. Rigorous application of exclusion and inclusion criteria meticulously narrowed down the study cohort to 1400 participants, a process systematically depicted in Fig. 1. Because α-Klotho concentration was assessed exclusively within the 40 to 79 age range, the study initially excluded individuals who were either under 40 years of age or above 79. Initially, participants with incomplete data on exposure and outcome variables were removed from the analysis. Subsequently, individuals lacking covariates data and those without recorded weights were also excluded to ensure the accuracy and reliability of the study’s findings. The weight selected for this study was wtsaf2yr.lipid, and utilizing sampling weights in all statistical analyses was paramount to ensuring the national representativeness of the research outcomes.

Fig. 1
figure 1

Schematic illustration of the participant selection criteria.

Full size image

Quantification of serum α-Klotho level

The analyses were executed utilizing a commercial ELISA kit manufactured by IBL International (Japan). Before analysis, all samples were kept at − 80 °C. The assay sensitivity of the ELISA kit was found to be 6 pg/mL. Guidelines provided by NHANES were followed and discussed for further details on laboratory methodology, quality assurance, and monitoring.

Hyperlipidemia

The hyperlipidemia status of participants was assessed following the criteria outlined by the National Cholesterol Education Program. Hyperlipidemia was operationally characterized based on specific thresholds, including total cholesterol (TC) levels ≥ 200 mg/dL, TG levels ≥ 150 mg/dL, low-density lipoprotein (LDL) levels ≥ 130 mg/dL, or female HDL levels ≤ 50 mg/dL, and male HDL levels ≤ 40 mg/dL23. Participants who self-reported the use of lipid-lowering medications were also grouped under the hyperlipidemia category24,25. The presence of hyperlipidemia was established if any one of these five criteria was met, providing a comprehensive and clinically relevant approach to identifying individuals with lipid metabolism disorders.

In order to understand the association between α-Klotho and hyperlipidemia, we classified hyperlipidemia according to the levels of hyperlipidemia components. LDL levels ≥ 130 mg/dL were defined as hyper-LDL, female HDL levels ≤ 50 mg/dL and male HDL levels ≤ 40 mg/dL were defined as hypo-HDL, TG levels ≥ 150 mg/dL were defined as hyper_TG, TC levels ≥ 200 mg/dL were defined as hyper_TC.

Covariates

This investigation comprehensively examined various covariates including demographic, lifestyle-related and health-related factors, through interviews, laboratory tests and physical examinations. Demographic variables comprised age, sex, race, economic status, education level, and marital status, with race was categorized into four groups: Black, White, Mexican, or Other. Moreover, levels of education were stratified into three tiers: less than high school, high school, and college (encompassing individuals with a college degree or higher). Similarly, economic status was evaluated by the family monthly poverty level index, delineated as low (< 1.30), medium (1.30–3.50), and high income (> 3.50)26. Marital status was bifurcated into married and single categories. Participants in marital or cohabitating relationships were designated as "married," whereas those who were separated, widowed, unmarried, or divorced were grouped into the “single” category.

Body mass index (BMI), energy intake, metabolic equivalent (MET), alcohol status, and smoking status were all included in lifestyle-related factors. BMI was employed as a metric for body composition. During the body examination, BMI was computed as weight in kilograms divided by height in meters squared (kg/m2). Alcohol and smoking status were dichotomously classified as either “yes” or "no."

Health-related factors included the Charlson Comorbidity Index (CCI), a reliable tool for evaluating overall physical condition and predicting long-term mortality27. The 24-h dietary recall method was utilized at mobile examination centers to calculate the energy intake by the total nutrient intake on the first day (DR1TOT). Participants also completed a physical activity (PA) questionnaire that adhered to World Health Organization guidelines28. The questionnaire collected data on physical activity type, frequency, and duration over the past seven days. MET values were assigned to evaluate the intensity of PA based on World Health Organization guidelines21,29. The formula for calculating PA (MET-min/week) was: PA (MET-min/week) = MET × weekly frequency × duration of each PA. These values were summed to derive the total PA (MET) value30.

The diagnosis and risk factors for various health conditions were also considered. Hypertension was described as having a history of diagnosed hypertension, using antihypertensive drugs, or having an average systolic blood pressure > 140 mmHg and/or a diastolic blood pressure > 90 mmHg. Diabetes was identified by various criteria, including HbA1c ≥ 6.5%, fasting plasma glucose ≥ 7.0 mmol/L, random plasma glucose or two-hour plasma glucose (oral glucose tolerance test) ≥ 11.1 mmol/L. Additionally, a history of diagnosed diabetes and the use of antidiabetic medications or insulin were also considered for diagnosis31. Moreover, CVD diagnosis was based on self-reported instances of stroke, heart attack, coronary artery disease, heart failure, or angina32. CKD diagnosis followed the 2012 Kidney Disease Outcomes Quality Initiative guidelines, considering glomerular filtration rate and proteinuria classifications33. The estimated glomerular filtration rate (eGFR) was calculated utilizing the epidemiology collaboration formula to assess CKD. CKD was defined as eGFR < 60 ml/min per 1.73 m2 or urine albumin-to-creatinine ratio (UACR) ≥ 30 mg/g34. Additionally, isotope dilution liquid chromatography tandem mass spectrometry (ID-LC–MS/MS) was utilized to evaluate testosterone and estradiol levels in serum.

Statistical analysis

In this investigation, all analyses integrated sampling weights to ensure the representation of the non-institutionalized civilian population. Cycle weights were recalculated per NHANES guidelines to maintain accuracy. Continuous variables were depicted as geometric means with corresponding interquartile ranges (IQR), whereas categorical variables were represented as frequencies and percentages. α-Klotho levels were transformed with Log-base 2 due to the right-skewed distribution, facilitating an examination of the link between serum α-Klotho levels and the change in the risk of hyperlipidemia resulting from a doubling of these levels. Moreover, α-Klotho levels were stratified into tertiles (T1 to T3). Survey-weighted Wilcoxon tests were employed for continuous variables, while survey-weighted chi-square tests were utilized for categorical variables. Survey-weighted logistic regression analyses were employed to assess the link between α-Klotho levels and hyperlipidemia. This investigation involved adjustments for a comprehensive array of covariates. Model 1 adjusted for the lifestyle-related and health-related factors, such as age, sex, race, education level, economic status, marital status, BMI, energy intake, MET, alcohol status, smoking status, CCI, diabetes, hypertension, CVD and CKD. Model 2 further adjusted testosterone and estradiol based on Model 1. Logistic regression analyses were performed to determine the association between α-Klotho and hyper-LDL, hypo-HDL, hyper-TG or hyper_TC. Moreover, restricted cubic spline (RCS) analyses were utilized to investigate potential non-linear associations between males and females. Subsequently, based on the RCS findings, piecewise logistic regression analysis was executed. This approach facilitated a detailed examination of the link between α-Klotho levels and hyperlipidemia, enabling a thorough comprehension of potential curvilinear associations. Statistical analysis was executed utilizing R (v4.3.1), with significance established at a threshold of P < 0.05.

Results

Characteristics of the study population

The application of exclusion and inclusion criteria meticulously narrowed down the study cohort to 1400 participants, including 337 participants without hyperlipidemia and 1063 participants with hyperlipidemia. Table 1 presents a comprehensive summary the demographic and relevant characteristics of the participants, thoughtfully stratified by sex. It is crucial to acknowledge significant differences between the female and male groups regarding several parameters, such as economic status, marital status, energy intake, MET, alcohol status, smoking status, CKD, testosterone and estradiol. Notably, males exhibited considerably higher energy intake and MET than females. The proportion of males reporting drinking (92.45%) and smoking (53.71%) was significantly higher than that of females, who reported drinking (86.91%) and smoking (41.35%). Conversely, the prevalence of CKD was considerably elevated in females compared to males. Additionally, testosterone and estradiol levels were considerably increased in males compared to females.

Table 1 Characteristics of subjects from the National Health and Nutrition Examination Survey (2013–2016).
Full size table

The geometric mean of α-Klotho levels, a central aspect of the investigation, was calculated at 779.79 (range 756.56–803.74). No significant difference in α-Klotho levels was noted between the male and female groups (P > 0.05). Furthermore, α-Klotho levels were methodically stratified into three tertiles: T1 (273.10–629.97), T2 (692.97–892.87) and T3 (892.87–5038.30). The distribution of α-Klotho tertiles exhibited no considerable differences between male and female groups (Table 1). The variation in α-Klotho levels within each group was analyzed (Fig. 2). The α-Klotho levels in the hyperlipidemia group were considerably lower relative to those in non-hyperlipidemia group (P < 0.01) (Fig. 2A). Specifically, in females, the levels of α-Klotho of the hyperlipidemia group were considerably lower relative to the non-hyperlipidemia group (P < 0.01). No significant difference was observed in the male hyperlipidemia and non-hyperlipidemia groups (Fig. 2B). Furthermore, Table S1 presents the baseline characteristics by tertiles of α-Klotho. BMI and smoke status were significantly different in α-Klotho tertiles.

Fig. 2
figure 2

The α-Klotho levels in different groups. (A) The α-Klotho levels in non-hyperlipidemia and hyperlipidemia. (B) The α-Klotho levels of hyperlipidemia and non-hyperlipidemia in different sexes. ** P < 0.01.

Full size image

Association between α-Klotho levels and hyperlipidemia among study participants

Survey-weighted logistic regression analyses were undertaken to scrutinize the link between α-Klotho levels and hyperlipidemia in participants. The findings revealed a considerable reduction in the risk of hyperlipidemia with each doubling of α-Klotho levels (OR 0.68, 95% CI 0.47–0.98, P = 0.04) in crude model. However, the other models demonstrated no statistically significant differences (Table 2).

Table 2 Logistic regression analyses of the association between α-Klotho levels and hyperlipidemia in participants.
Full size table

Association of α-Klotho levels with hyperlipidemia in females and males

The participants were stratified by sex and subjected to analysis by logistic regression model. Table 3 illustrates the link between α-Klotho levels and hyperlipidemia within the female participant subgroup. Even after adjusting for all covariates, a significant negative association remained between Log2 α-Klotho and hyperlipidemia. In model 2, each doubling of α-Klotho levels corresponded to a significant reduction in the risk of hyperlipidemia (OR 0.45, 95% CI 0.24–0.82, P = 0.02). As α-Klotho tertiles increased among females, there was a gradual and significant decrease in the risk of hyperlipidemia. In contrast, logistic regression analysis conducted on males (as detailed in Table 4) demonstrated a positive trend between α-Klotho levels and hyperlipidemia (P > 0.05). In model 2, each doubling of α-Klotho levels corresponded to an increased risk of hyperlipidemia (OR 1.14, 95% CI 0.63–2.06, P = 0.63) (Table 4).

Table 3 Logistic regression analyses of the association between α-Klotho levels and hyperlipidemia in females.
Full size table
Table 4 Logistic regression analyses of the association between α-Klotho levels and hyperlipidemia in males.
Full size table

The results of logistic regression analyses of the association between α-Klotho levels and hyperlipidemia components were shown in Table 5. The association between α-Klotho and all the components were inverse in both males and females, but only the link between α-Klotho and hyper_TC was statistically significant in females.

Table 5 Logistic regression analyses of the association between α-Klotho levels and hyperlipidemia components in females and males.
Full size table

The analyses of nonlinear relationship between α-Klotho levels and hyperlipidemia

RCS analysis examined the non-linear correlation between α-Klotho levels and hyperlipidemia. The outcomes revealed no non-linear correlation between α-Klotho levels and hyperlipemia in males (Fig. 3B), while a non-linear association was observed in females (Fig. 3A). In females, the inflection point of α-Klotho was determined to be 1106.87 pg/mL. Subsequently, piecewise logistic regression analysis was conducted (Table 6). The findings revealed a substantial reduction in the risk of hyperlipidemia for every doubling of α-Klotho levels in model 2 (OR 0.22, 95% CI 0.09–0.54, P < 0.01). However, when α-Klotho levels exceeded 1106.87 pg/mL, the association was no longer significant in any models (P > 0.05).

Fig. 3
figure 3

The RCS analyses were conducted using 3 nodes to illustrates the correlation between α-Klotho and hyperlipidemia based on sex-based variations. (A) Female; (B) male. The red line represents odd ratio (OR), and the red transparent area depicts 95% CI. ORs outcomes are adjusted based on Model 2.

Full size image
Table 6 The piecewise logistic regression analyses of the association between α-Klotho levels and hyperlipidemia in females.
Full size table

Discussion

This study represents the pioneering sex-based investigation of unveiling a novel relationship between serum α-Klotho levels and hyperlipidemia. Specifically, in females, α-Klotho levels displayed a considerable negative association with hyperlipidemia, characterized by a saturation effect. A consistent positive trend was observed in males between α-Klotho levels and hyperlipidemia. Consequently, α-Klotho emerges as a potential predictive biomarker for the risk of hyperlipidemia in the females.

Previous research has highlighted an inverse relationship between serum α-Klotho levels and metabolic syndrome, emphasizing its potential to attenuate risk factors, for CVD, such as TG and TC levels20,21. Additionally, there is a positive association between plasma α-Klotho concentration and HDL levels, and a negative link with TG levels and fibroblast growth factor 23 concentration13,35. The research revealed that the association between α-Klotho and the various components of hyperlipidemia exhibits sex-specific variations, which in turn may account for the observed sex disparities in the association between α-Klotho and the hyperlipidemia risk. The underlying mechanisms driving this association remain unclear and necessitate further exploration.

Prior research has explored the link between hyperlipidemia and Klotho. Investigations on hyperlipidemic rats with non-insulin-dependent diabetes36 and hypercholesterolemic uremic atherosclerotic ApoE knockout mice37, observed a considerable decrease in Klotho expression. The expression of Klotho was reduced by acute inflammation and oxidative stress38,39. In ApoE knockout hyperlipidemic mice, oxidized-LDL (ox-LDL) production is heightened, triggering inflammation, intensifying oxidative stress, and inducing cell apoptosis. Ox-LDL reduces Klotho expression by activating ERK and NF-kB40. Inflammatory cytokines TWEAK and TNF-α downregulate Klotho in renal tubular cells via an NF-kB-dependent mechanism41,42. The investigations suggest that inflammation negatively impacts Klotho expression, whereas Klotho exhibits anti-oxidative stress and anti-inflammatory properties43,44. Klotho improves oxidative stress triggered by ox-LDL by regulating the PI3K/Akt/eNOS pathway45, and inhibits the secretion of IL-6 and TNF-α44. The inhibition of lipid accumulation in the liver and adipose tissue by α-Klotho is achieved by suppressing the PI3K/AKT signaling pathway. It contributes to improve lipid homeostasis in type 2 diabetes46. These mechanisms elucidate the considerable negative association between Klotho and the risk of hyperlipidemia.

This study also revealed considerable sex-specific variations in the relationship between α-Klotho levels and hyperlipidemia. Although the levels of α-Klotho in females were considerably higher in comparison to males, this disparity did not attain statistical significance. Notably, lower α-Klotho levels were observed in females with hyperlipidemia compared to females without this condition. However, there was no significant variation in α-Klotho levels between males with hyperlipidemia and those without this condition. It suggests that there may be a sex-specific variation in the regulatory mechanism between hyperlipidemia and α-Klotho. In this study, no significant association was detected between α-Klotho and hyperlipidemia in males, a phenomenon that remains unexplained by current studies. Prior research indicates that inflammatory responses differ by sex, with females resolving inflammation more rapidly than males47. Additionally, a lower systemic inflammatory burden has been reported in females48. Egr1 and PPARγ have been found to directly upregulate the expression of α-Klotho49,50. Studies suggest that inflammation generally upregulates Egr151, potentially representing a molecular mechanism regulating α-Klotho. Furthermore, PPARγ expression correlates with inflammation induced by hyperlipidemia52. The mechanism may involve the conversion of LDL into oxLDL, which is the endogenous PPARγ ligand, thereby activating PPARγ expression53. Testosterone can increase LDL oxidation54, which suggests a potential pathway for activating PPARγ expression in males. Studies have shown a positive association between testosterone and α-Klotho in males55, in contrast to the absence of such a relationship in females56. Current research suggests that the regulatory mechanisms of α-Klotho expression may vary between males and females, with males potentially being more susceptible to the upregulation of α-Klotho through inflammatory stimulation under hyperlipidemic conditions, which could explain the absence of a significant association between α-Klotho and hyperlipidemia in males. Although the molecular mechanisms are not fully elucidated, sex-based differences in α-Klotho expression have been substantiated by various studies. The influence of Life’s Essential 8 score on serum α-Klotho levels appears to be more pronounced in females57. The negative association between energy-dense diet and α-Klotho levels among sedentary women aged 45–65 years, a relationship not observed in males58. Moreover, serum α-Klotho levels exhibited a negative association with the depression in females, contrasting with a positive association found in males59. In brief, this study offers avenues for future biological explorations into α-Klotho.

This study underscores the pivotal role of α-Klotho in hyperlipidemia risk assessment among community-dwelling females, with profound implications for health management. By quantifying serum α-Klotho levels, healthcare professionals can delineate individuals at elevated risk, facilitating the formulation of tailored health strategies. Women with diminished α-Klotho levels should be encouraged to embrace a preventative lifestyle, characterized by a balanced diet, consistent physical activity, and the eschewing of detrimental habits, to mitigate hyperlipidemia risk. For those identified as high-risk, prioritization of medical resources and services is imperative to ensure prompt medical engagement and intervention. This approach not only enhances individual health outcomes but also optimizes the allocation of healthcare resources, thereby achieving a more judicious distribution of medical assets60.

While this study represents the pioneering effort to unravel the independent protective role of α-Klotho in hyperlipidemia, it is imperative to acknowledge certain limitations. Firstly, the detection of α-Klotho in NHANES is limited to individuals aged 40–79, excluding data on children and young adults. This age constraint emphasizes the need for careful consideration when extrapolating the results to a broader demographic range. Secondly, it’s crucial to recognize that α-Klotho levels can fluctuate due to circadian rhythms and temporal variations, which may introduce measurement variability. Lastly, the observational nature of the study prohibits making causal conclusions about the relationship between α-Klotho levels and hyperlipidemia. It underscores the importance of further interventional studies to elucidate causality.

The investigation insights into the sex-based relationship between serum α-Klotho levels and the risk of hyperlipidemia. Significantly, the results reveal a considerable negative relationship between serum α-Klotho levels and hyperlipidemia only in females. Serum α-Klotho levels could be a valuable biomarker for assessing the risk of hyperlipidemia in females.