Serum Klotho and insulin resistance: Insights from a cross-sectional analysis

Abstract

The prevalence of diabetes has surged globally, posing significant health and economic burdens. Insulin resistance underlies the initiation and development of type 2 diabetes. Klotho is a crucial endogenous antiaging factor, associated with atherosclerotic cardiovascular diseases, cancer, neurological disorders, and renal diseases. It additionally has a function in controlling glucose metabolism and holds promise as a new therapeutic target for diabetes. However, its relationship with insulin resistance remains unclear. This study utilizes the National Health and Nutrition Examination Survey (NHANES) 2007 to 2016 data to investigate the relationship between serum Klotho concentrations and insulin resistance. In this observational study, information from the NHANES spanning 2007 to 2016 was employed. The sample consisted of 6371 participants. Weighted linear regression model and chi-square tests were utilized to assess differences in continuous and categorical variables, respectively, among groups categorized by Klotho quartiles. The relationship between Klotho and HOMA-IR (homeostatic model assessment of insulin resistance) was studied using multiple linear regression. Smooth curve fitting was used to analyze nonlinear relationships and the inflection point was determined through a 2-stage linear regression method. After adjusting for multiple confounding factors, serum Klotho levels were found to be positively correlated with insulin resistance [0.90 (0.68, 1.13)]. This correlation is nonlinear and exhibits a saturation effect, with the inflection point identified at 1.24 pg/µL. When Klotho levels are below 1.24 pg/µL, for every unit increase in Klotho, HOMA-IR increases by 1.30 units. Conversely, when Klotho levels exceed 1.24 pg/µL, there is no correlation between HOMA-IR and Klotho. Subgroup analysis reveals that the relationship between HOMA-IR and Klotho varies depending on diabetes and body mass index (BMI). This positive correlation was most prominent in the obese nondiabetic population. There is a positive correlation between serum Klotho and insulin resistance.

1. Introduction

The incidence of diabetes is swiftly rising across all global regions. The worldwide incidence of diabetes in people aged 20 to 79 in 2021 was approximately 10.5%, corresponding to 536.6 million individuals. It is anticipated that this figure will increase to 12.2% (equating to 783.2 million individuals) by 2045.^[1] Diabetes represents a considerable burden both to individuals and public health systems, in terms of the affected population size, diabetes-related complications, and the subsequent costs to national healthcare and social care systems. Consequently, there is an urgent need to actively identify new therapeutic targets for diabetes, reduce its incidence, and improve prognoses.

A core aspect of diabetes is insulin resistance. The onset of insulin resistance precedes the increase in blood glucose concentrations and is characterized by a reduced responsiveness of insulin target tissues to elevated physiological insulin levels. The liver and adipose tissue, as crucial sites for glucose-induced insulin signal transduction, are central to the mechanisms underlying insulin resistance.^[2] Insulin resistance is foundational in the development and advancement of type 2 diabetes (T2DM). We can better protect β-cell function, prevent diabetes, and slow its progression, while reducing diabetes-related complications, only by effectively alleviating insulin resistance. HOMA-IR (homeostatic model assessment of insulin resistance) is commonly used to assess insulin resistance and is the most widespread and convenient evaluation metric for this purpose.^[3]

The Klotho gene, discovered in recent years, is recognized as a gene that counteracts aging.^[4] α-Klotho, predominantly produced in the kidneys, undergoes cleavage by the a disintegrin and metalloprotease 10 and 17, and is then released into the bloodstream, cerebrospinal fluid, and urine, known as s-Klotho.^[5] Klotho is an essential endogenous antiaging factor associated with atherosclerotic cardiovascular diseases, cancer, neurological disorders, and kidney diseases.^[6–10] It also contributes to the regulation of glucose metabolism and holds promise as a new therapeutic target for diabetes.^[11,12]

Although numerous studies suggest a close relationship between Klotho and the progression of T2DM and insulin sensitivity impairment,^[13–15] the connection between Klotho and insulin resistance remains incompletely elucidated. Therefore, the objective was to explore the link between serum Klotho levels and insulin resistance, utilizing data from the National Health and Nutrition Examination Survey (NHANES) covering the years 2007 to 2016.

2. Methods

2.1. Study population

The NHANES database can be accessed via the National Center for Health Statistics website at the Centers for Disease Control and Prevention. The NHANES study from 2007 to 2016 encompassed a total of 50,588 participants. Among these, individuals who had their serum Klotho levels measured were considered for inclusion, amounting to 13,764 participants. Insulin users (n = 295), and those lacking complete data on fasting glucose (n = 7049) and insulin levels (n = 49) were excluded, resulting in 6371 participants in the study cohort. The NHANES 2007 to 2016 received ethical approval from its Institutional Review Board (Protocols #2005-06 and #2011-17), with all involved individuals providing informed consent. The selection procedure is depicted in Figure 1.

Figure 1.

Flowchart of participant selection.

The exposure variable is the concentration of serum Klotho in participants. Serum samples were gathered and preserved at −80°C to measure this concentration using an ELISA kit produced by IBL International, Japan. To ensure precision, all samples were tested in duplicate, and the average of the 2 measurements was taken as the final result. Due to the small effect size observed in our study, we converted the unit from the traditional pg/mL to pg/µL.

Our primary outcome variable is HOMA-IR, which represents insulin resistance. The calculation is made through the formula: HOMA-IR = Fasting Glucose (mmol/L) × Insulin (μU/mL)/ 22.5.^[16]

2.2. Covariates

From each NHANES cycle, a series of additional covariates was gathered. This encompassed age, creatinine (Cr), Hemoglobin A1c (HbA1c), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), total cholesterol, waist circumference (WC), and body mass index (BMI). Categorical variables included gender, race, smoking and alcohol use, diabetes, hypertension, and vigorous work activity. BMI is categorized as follows: <25 kg/m² is classified as normal weight, 25 to 30 kg/m² as overweight, and ≥ 30 kg/m² as indicating obesity. Race was categorized into Mexican American, other Hispanic, nonHispanic white, nonHispanic black, and additional racial groups. Individuals are classified as smokers if they have consumed more than 100 cigarettes in their lifetime, and as drinkers if they have consumed alcohol on more than 12 occasions per year throughout their life. Diabetes was defined based on any of the following criteria: Glycohemoglobin > 6.5%, or fasting plasma glucose level ≥ 7.0 mmol/L, or random blood glucose ≥ 11.1 mmol/L, or 2-hour oral glucose tolerance test plasma glucose ≥ 11.1 mmol/L, or participants who self-reported a diagnosis of diabetes. Hypertension was determined from questionnaires. Vigorous work activity was defined as tasks that are physically strenuous and lead to a significant increase in respiratory or heart rate, maintained for at least 10 consecutive minutes.

2.3. Statistical analysis

Klotho levels were stratified into quartiles, spanning from the lowest (Q1) to the highest (Q4). Continuous variables were reported as the mean ± standard deviation (SD), whereas categorical variables were denoted by their frequencies (expressed in percentages). Weighted linear regression model and chi-square tests were utilized to assess differences in continuous and categorical variables, respectively, among groups categorized by Klotho quartiles. The association between Klotho and HOMA-IR was examined using multivariable linear regression. In this analysis, Model 1 was unadjusted for covariates; Model 2 included adjustments for age, gender, and race; and Model 3 was further adjusted for all previously mentioned confounding factors. Further sensitivity analyses were conducted based on the Klotho quartiles, followed by subgroup analyses. nonlinear relationships were identified using a smoothing curve fit. Inflection points were identified through a 2-stage linear regression approach. A P value less than .05 was considered to signify statistical significance. A robust weighting strategy was adopted to mitigate potential biases. All analyses and graphical representations were conducted using R (version 4.2.0) and Empower Stats (version 4.0).

3. Results

3.1. Baseline characteristics

Taking into account the criteria for inclusion and exclusion, a total of 6371 adults were included in this study, with an average age of 56.20 ± 10.44 years. Of these, 47.02% were men and 52.98% were women. Additionally, 6.47% were Mexican American, 4.90% were Other Hispanic, 72.68% were nonHispanic White, 9.45% were nonHispanic Black, and 6.50% were from other races. The mean ± SD of HOMA-IR and Klotho were 3.49 ± 3.44 and 0.85 ± 0.30 pg/µL, respectively.

Compared to those in the lowest quartile of Klotho levels, individuals in the highest quartile were more likely to be female and younger. They were also less likely to be smokers and drinkers, and less likely to be nonHispanic White. Moreover, they engaged in fewer high-intensity activities and had a smaller WC. Additionally, they exhibited higher levels of HbA1c, fasting glucose, insulin, and HDL-C, but lower levels of Cr and TG (Table 1).

Table 1.

Weighted characteristics of the study population based on Klotho quartiles.

	Klotho quartiles
	Q1 (N = 1593)	Q2 (N = 1592)	Q3 (N = 1593)	Q4 (N = 1593)	P value
Age (yr)	56.77 ± 10.91	56.03 ± 10.35	56.41 ± 10.23	55.58 ± 10.21	.0106
Gender (%)					<.0001
Men	50.63	50.40	45.69	40.92
Women	49.37	49.60	54.31	59.08
Race (%)					<.0001
Mexican American	6.08	6.22	6.58	7.06
Other Hispanic	4.73	4.25	5.31	5.34
NonHispanic White	74.44	75.30	73.66	66.92
NonHispanic Black	9.26	7.30	8.00	13.57
Other Race	5.49	6.92	6.45	7.11
Smokers (%)					<.0001
Yes	53.64	50.34	47.62	39.99
No	46.36	49.66	52.38	60.01
Drinkers (%)					<.0001
Yes	82.14	77.32	77.33	69.52
No	17.86	22.68	22.67	30.48
Hypertension (%)					.2726
Yes	43.57	41.57	40.22	41.18
No	56.43	58.43	59.78	58.82
Vigorous work activity (%)					.0001
Yes	22.23	19.89	17.14	15.88
No	77.77	80.11	82.86	84.12
Diabetes (%)					.7877
Yes	18.36	17.26	18.45	18.40
No	81.64	82.74	81.55	81.60
BMI	29.49 ± 6.30	29.40 ± 6.40	29.65 ± 6.78	29.08 ± 6.55	.1013
WC (cm)	102.28 ± 14.66	101.38 ± 15.00	101.92 ± 15.40	99.77 ± 15.70	<.0001
Cr (mmol/L)	81.77 ± 45.62	78.57 ± 43.95	75.72 ± 18.61	74.71 ± 27.30	<.0001
HDL-C (mmol/L)	1.44 ± 0.49	1.40 ± 0.43	1.42 ± 0.42	1.47 ± 0.45	.0002
TC (mmol/L)	5.22 ± 1.13	5.22 ± 1.05	5.20 ± 1.08	5.14 ± 1.06	.1467
LDL-C (mmol/L)	3.06 ± 0.93	3.14 ± 0.91	3.10 ± 0.93	3.06 ± 0.90	.0498
TG (mmol/L)	1.60 ± 1.68	1.51 ± 1.01	1.50 ± 1.11	1.36 ± 1.06	<.0001
HbA1c (%)	5.67 ± 0.72	5.68 ± 0.75	5.72 ± 0.82	5.83 ± 1.15	<.0001
Fasting glucose (mmol/L)	5.90 ± 1.21	5.92 ± 1.34	6.02 ± 1.56	6.19 ± 2.15	<.0001
Insulin (µU/mL)	11.56 ± 8.19	12.15 ± 9.36	12.69 ± 9.96	13.36 ± 11.51	<.0001
HOMA-IR	3.15 ± 2.64	3.33 ± 3.00	3.61 ± 3.68	3.90 ± 4.22	<.0001

BMI = body mass index, Cr = creatinine, HbA1c = hemoglobin A1c, HDL-C = high-density lipoprotein cholesterol, HOMA-IR = homeostatic model assessment of insulin resistance, LDL-C = low-density lipoprotein cholesterol, TC = total cholesterol, TG = triglyceride, WC = Waist circumference.

Table 2 shows a positive correlation between Klotho and HOMA-IR concentration in the nonadjusted model [0.81 (0.54, 1.09)], partially adjusted model [0.89 (0.61, 1.17)], and the fully adjusted model [0.90 (0.68, 1.13)].

Table 2.

The association between Klotho and HOMA-IR.

	Model 1 β (95%) P value	Model 2 β (95%) P value	Model 3 β (95%) P value
Klotho	0.81 (0.54, 1.09) <.0001	0.89 (0.61, 1.17) <.0001	0.90 (0.68, 1.13) <.0001
Klotho quartiles
Q1	Ref	Ref	Ref
Q2	0.18 (−0.05, 0.42) .1323	0.20 (−0.03, 0.44) .0895	0.21 (0.03, 0.39) .0258
Q3	0.45 (0.22, 0.69) .0002	0.48 (0.25, 0.72) <.0001	0.38 (0.20, 0.56) <.0001
Q4	0.75 (0.51, 0.99) <.0001	0.80 (0.56, 1.05) <.0001	0.81 (0.62, 1.00) <.0001
P for trend	<.0001	<.0001	<.0001

The analysis models were defined as: Model 1: No adjustments. Model 2: Adjusted for age, gender, and race. Model 3: Adjusted for age, gender, race, smoking and alcohol use, vigorous work activity, diabetes, hypertension, BMI, WC, LDL-C, HDL-C, TG, HbA1c, Cr.

BMI = body mass index, Cr = creatinine, HbA1c = hemoglobin A1c, HDL-C = high-density lipoprotein cholesterol, HOMA-IR = homeostatic model assessment of insulin resistance, LDL-C = low-density lipoprotein cholesterol, TG = triglyceride, WC = waist circumference.

Compared to the first quartile (Q1), the HOMA-IR in Q4 demonstrated increases of 0.81 in Model 3. Notably, the P for trend in model 3 was less than 0.05, indicating a consistent and significant trend across the quartiles.

As shown in Figure 2, there was a positive correlation between Klotho levels and HOMA-IR in most of the subgroups evaluated. The association remained consistent across subgroups stratified by gender, age, smoking habits, alcohol consumption, hypertension, and the intensity of work activities. However, notable disparities in this association were noted among subgroups delineated by diabetes and BMI, indicated by significant P values for interaction. The obese nondiabetic population exhibited the most pronounced positive correlation.

Figure 2.

Subgroup analysis of the correlation between Klotho and HOMA-IR. In addition to stratification variables, each subgroup analysis was adjusted for age, gender, race, smoking and alcohol use, vigorous work activity, diabetes, hypertension, BMI, WC, LDL-C, HDL-C, TG, HbA1c, Cr. BMI = body mass index, Cr = creatinine, HbA1c = hemoglobin A1c, HDL-C = high-density lipoprotein cholesterol, HOMA-IR = homeostatic model assessment of insulin resistance, LDL-C = low-density lipoprotein cholesterol, TG = triglyceride, WC = waist circumference.

Subsequently, a smooth curve fitting and segmented regression analysis were conducted to examine the nonlinear relationship between Klotho levels and HOMA-IR concentration. Following comprehensive adjustment, the smooth curve revealed a nonlinear association between Klotho levels and HOMA-IR (Fig. 3). The segmented regression analysis indicated that the inflection point value of Klotho was 1.24 pg/µL. In our study, for Klotho levels below 1.24 pg/µL, there was a marked association where an increment of 1 pg/µL in Klotho corresponded to a significant elevation of 1.30 in HOMA-IR (P < .05). Conversely, when Klotho levels were above 1.24 pg/µL, this relationship was not statistically significant (P > .05). The log likelihood ratio test confirmed that these associations differed between different Klotho thresholds with a significant value of less than 0.0001 (Table 3). An inverted U-shaped relationship between Klotho and HOMA-IR was observed in the diabetic population after stratification by diabetes status (Fig. 4).

Figure 3.

The association between Klotho and HOMA-IR. The solid red line represents the smooth curve fit between variables. Blue bands represent the 95% confidence interval from the fit. All models were adjusted for age, gender, race, smoking and alcohol use, vigorous work activity, diabetes, hypertension, BMI, WC, LDL-C, HDL-C, TG, HbA1c, Cr. BMI = body mass index, Cr = creatinine, HbA1c = hemoglobin A1c, HDL-C = high-density lipoprotein cholesterol, HOMA-IR = homeostatic model assessment of insulin resistance, LDL-C = low-density lipoprotein cholesterol, TG = triglyceride, WC = waist circumference.

Table 3.

Threshold effect analysis of Klotho on HOMA-IR using a linear regression model.

HOMA-IR	Adjusted β (95%) P value
Klotho
Inflection point	1.24
Klotho < 1.24	1.30 (0.98, 1.61) <.0001
Klotho > 1.24	0.11 (−0.38, 0.61) .6550
Log likelihood ratio	<0.0001

All models were adjusted for age, gender, race, smoking and alcohol use, vigorous work activity, diabetes, hypertension, BMI; WC, LDL-C, HDL-C, TG, HbA1c, Cr.

BMI = body mass index, Cr = creatinine, HbA1c = hemoglobin A1c, HDL-C = high-density lipoprotein cholesterol, HOMA-IR = homeostatic model assessment of insulin resistance, LDL-C = low-density lipoprotein cholesterol, TG = triglyceride, WC = waist circumference.

Figure 4.

Association between Klotho and HOMA-IR stratified by diabetes. All models were adjusted for age, gender, race, smoking and alcohol use, vigorous work activity, hypertension, BMI, WC, LDL-C, HDL-C, TG, HbA1c, Cr. BMI = body mass index, Cr = creatinine, HbA1c = hemoglobin A1c, HDL-C = high-density lipoprotein cholesterol, HOMA-IR = homeostatic model assessment of insulin resistance, LDL-C = low-density lipoprotein cholesterol, TG = triglyceride, WC = waist circumference.

4. Discussion

Studies have shown that serum Klotho levels are positively correlated with insulin resistance after adjusting for a variety of confounding factors [0.90 (0.68, 1.13)]. There was a saturation effect in this positive correlation. Below 1.24 pg/µL Klotho levels, each unit increase in Klotho corresponds to a 1.30 unit increase in HOMA-IR, whereas above 1.24 pg/µL, no significant correlation exists between HOMA-IR and Klotho. Subgroup analysis reveals variations in the relationship between HOMA-IR and Klotho among populations stratified by diabetes presence and BMI.

Klotho significantly influences the onset and advancement of T2DM. Insulin resistance is key to the development and progression of T2DM, but the link between Klotho and insulin resistance is still contested in current research. Some research suggests an inverse relationship between Klotho concentrations and insulin resistance.^[15,17] For example, in healthy school-aged girls, serum α-Klotho protein levels were found to be negatively correlated with C-reactive protein levels, the HOMA-IR index, visceral fat levels, body fat percentage, waist size, and BMI.^[18] In sedentary middle-aged men (i.e., 40–65 years old), plasma s-Klotho levels were negatively correlated with cardiac metabolic risk and insulin resistance, independent of their actual age, cardiopulmonary health, physical activity levels, and dietary intake. Such a correlation was not found in young individuals (i.e., 18–25 years old).^[19] Conversely, some studies align with our research, suggesting a positive correlation between Klotho and insulin resistance. Silencing Klotho can reverse the decline in cell activity induced by high glucose levels and alleviate insulin resistance.^[20] In diabetic mice, Klotho treatment could elevate plasma insulin levels, but insulin sensitivity was not improved.^[21] Another study showed that after knocking out Klotho, insulin production in mice decreased, but insulin sensitivity increased, leading to hypoglycemia.^[22] In contrast to the studies mentioned above, Lorenzi O and colleagues question the direct involvement of Klotho in insulin signaling. They propose that a decrease in Klotho expression is not a consistent characteristic observed across rodent models of insulin resistance.^[23]

In the diabetic population, Klotho shows a complex nonlinear relationship with HOMA-IR. The divergent roles of Klotho in Insulin Resistance may depend on the specific mechanisms involved. The insulin resistance induced by Klotho may differ from pathological insulin resistance associated with diabetes. Some theories suggest that Klotho-induced mild insulin resistance could slow down aging. The mechanism may involve reducing the supply of intracellular glucose stimulated by insulin, thereby preventing intracellular lipid overload and lipotoxicity.^[24] Klotho can inhibit the insulin- and insulin-like growth factor 1-induced autophosphorylation of the insulin receptor and insulin-like growth factor 1 receptor. It interferes with intracellular signaling mediators like Protein Kinase B (Akt), Glycogen Synthase Kinase 3β, and Phosphofructokinase by affecting their phosphorylation status. This weakens intracellular signal transduction, prevents the translocation of Glucose Transporter Type 4, and triggers insulin resistance in adipocytes.^[25,26] Furthermore, Klotho plays a significant role in the internal regulation of insulin activity as a negative feedback mechanism: insulin promotes the shedding of Klotho, elevating serum Klotho protein levels, thereby hindering peripheral insulin signaling and blocking extended insulin activity.^[27] On the other hand, some studies have found that in T2DM mice models, Klotho could improve insulin sensitivity and hepatic glucose homeostasis. The mechanism involves Klotho targeting the phosphoinositide 3-kinase/AKT/mammalian target of rapamycin complex 1 signaling pathway ‘s IGF1R interaction, thereby upregulating Peroxisome Proliferator-Activated Receptor α in T2DM, which in turn improves insulin sensitivity and hepatic glucose and lipid homeostasis.^[28] Another reason for Klotho potential role in enhancing insulin resistance may be its antiinflammatory and antioxidative stress effects. Klotho shows a negative correlation with inflammation and oxidative stress,^[8,29,30] can downregulate the levels of inflammatory cytokines^[31,32], and can reduce Reactive Oxygen Species (ROS) through the expression of antioxidant proteins, as well as inhibit ROS-related oxidative stress signaling pathways.^[33] These physiological mechanisms could contribute to resolving issues related to insulin sensitivity.^[34,35] Given these intricate interactions and mechanisms, a comprehensive understanding of Klotho role in insulin resistance requires further investigation.

Subgroup analysis revealed that BMI modulated the relationship between Klotho levels and insulin resistance. This positive correlation is most prominent in obese individuals. The complex mechanistic relationships among Klotho, obesity, and insulin resistance are yet to be completely understood. Insulin resistance is often triggered by disproportionate expansion of white adipose tissue and abnormal recruitment of adipocyte precursor cells. In vitro experiments have indicated that Klotho promotes adipocyte differentiation, potentially playing a role in adipocyte maturation and systemic glucose metabolism. Mice lacking the Klotho gene displayed resistance to obesity induced by a high-fat diet. Moreover, knocking out the Kl gene in leptin-deficient Lepob/ob mice resulted in reduced obesity, enhanced insulin sensitivity, and subsequently, decreased blood glucose levels.^[36] These results highlight the multifaceted role of Klotho in metabolic regulation and highlight the importance of further investigations to understand its potential as a therapeutic target for metabolic disorders.

5. Study strengths and limitations

This study presents several strengths, including its substantial sample size which ensures robust statistical analyses and reliable results. Additionally, we employed weighting to mitigate potential biases, further enhancing the study validity. Detailed subgroup analyses were conducted to explore the relationship between Klotho and insulin resistance in different populations. Through statistical adjustments, multiple potential confounders were controlled for, enhancing the credibility of our conclusions. However, the study has inherent limitations. Due to its cross-sectional design, causality cannot be established, underscoring the need for future prospective studies for further clarity. While we utilized HOMA-IR as an indicator of insulin resistance, and it is widely accepted, it remains a surrogate marker and might not wholly represent an individual state of insulin resistance. Despite controlling for multiple confounders, potential unmeasured or overlooked confounding factors might still exist. The sample, though large, might be subject to certain biases and may not be wholly representative of the broader population. Variability in laboratory techniques and equipment may have introduced inconsistencies in the measurements of serum Klotho levels.

6. Conclusion

In cross-sectional studies, a positive correlation between serum Klotho levels and insulin resistance was evident, which was particularly prominent in obese nondiabetics. These findings underscore the critical role of Klotho in glucose metabolism and its potential implications for diabetes management. To firmly establish causality and delineate the underlying mechanisms, future research necessitates large-scale, well-designed prospective studies.

Author contributions

Data curation: Xiaoyan Hu, Shanshan Wu.

Writing – original draft: Laisha Yan.

Writing – review & editing: Laisha Yan, Shunying Zhao.