The association of visceral and subcutaneous fat areas with phenotypic age in non-elderly adults, mediated by HOMA-IR and HDL-C

Abstract

Background

Ageing results in diminished adaptability, as well as declines in physiological and psychological functions and resilience. The epigenetic clock ‘Phenotypic Age’ (PhenoAge) represents ‘preclinical ageing’. Phenotypic Age Acceleration (PhenoAgeAccel) is defined as the residual from a linear regression model predicting PhenoAge on the basis of chronological age. Abdominal subcutaneous adipose tissue, visceral adipose tissue, the Homeostasis Model Assessment of Insulin Resistance (HOMA-IR), and high-density lipoprotein cholesterol (HDL-C) have all been shown to correlate with ageing; however, the connections between these factors and PhenoAge are still insufficiently investigated.

Methods

Data for this study were sourced from the National Health and Nutrition Examination Survey (2015–2018), comprising 2580 participants. Complex survey designs were considered. To examine the association between body fat area and PhenoAgeAccel, logistic regression was applied. Additionally, subgroup analysis was used to identify variations in population characteristics. The dose‒response relationship between body fat area and PhenoAgeAccel was determined via restricted cubic spline analysis. Mediation and interaction analyses were further employed to investigate the roles of the HOMA-IR and HDL-C in this association.

Results

In nonelderly adults, the relationships between body fat area and PhenoAgeAccel differed chronological age. For abdominal subcutaneous fat area (SFA), this relationship was nonlinear in individuals aged 18–44 years and 45–59 years, with thresholds of 2.969 m² and 3.394 m², respectively. In contrast, a nonlinear relationship of visceral fat area (VFA) with PhenoAgeAccel was observed in individuals aged 18–44 years, while this relationship was linear in individuals aged 45–59 years, with thresholds of 0.769 m² and 1.220 m², respectively. Mediation effect analysis revealed that the HOMA-IR had a more pronounced mediation effect in individuals aged 18–44 years, accounting for 13.4% of the relationship between VFA and PhenoAgeAccel and 6.9% of the relationship between SFA and PhenoAgeAccel. Conversely, HDL-C had a greater mediating effect in individuals aged 45–59 years, accounting for 21.7% of the relationship between VFA and PhenoAgeAccel and 11.6% of the relationship between abdominal SFA and PhenoAgeAccel. HOMA-IR ≥ 2.73 or VFA > 0.925 m², as well as HOMA-IR ≥ 2.73 or abdominal SFA > 3.137 m², accelerated PhenoAge, whereas 1.60 < HDL-C ≤ 3.90 mmol/L combined with abdominal SFA ≤ 3.137 m² or VFA ≤ 0.925 m² decelerated PhenoAge.

Conclusion

In this study, the nonlinear relationships among abdominal SFA, VFA, and PhenoAgeAccel were elucidated, while characteristic thresholds across different age groups were identified. The results of this study emphasize the complex influence of fat distribution on the ageing process and refine the roles of HOMA-IR and HDL-C in various age cohorts. These findings provide a biological basis for future screening for accelerated ageing and appropriate intervention in high-risk populations and offer valuable insights for guiding personalized clinical interventions and health management strategies.

Background

With the extension of human lifespans, age-related health issues have become increasingly prominent, significantly adding to the global healthcare burden [1]. Aging is a complex and interconnected network of progressive functional deviations in harmful phenotypes, often driven by DNA damage, oxidative stress, and microbiota changes [2]. While chronological ageing progresses at a fixed rate and cannot be modified, biological ageing is variable and is shaped by genetic factors, environmental influences, and lifestyle choices, among other factors [3]. Compared with chronological age, biological age is a more accurate indicator of functional decline. Early determination of biological age and appropriate interventions are therefore critical for improving late-life health and reducing associated social burdens.

Studies have shown that genomic instability is a hallmark of biological ageing, and methylation patterns at specific CpG sites in DNA are strongly linked to chronological age [4]. Levine et al. developed an epigenetic clock known as ‘Phenotypic Age’ (PhenoAge), which is based on DNA methylation (DNAm) and clinical biomarkers indicative of physiological dysregulation [5]. This clock substitutes chronological age with PhenoAge and serves as a highly robust predictor of transitions to subsequent stages in the ageing process, including morbidity and mortality. Additionally, PhenoAge can reflect ‘preclinical ageing’, potentially providing a better distinction of ageing among children, young adults, and exceptionally healthy individuals. Phenotypic Age Acceleration (PhenoAgeAccel) refers to the residual value obtained from a linear regression model that estimates phenotypic age using chronological age as a predictor. Positive PhenoAgeAccel indicates that the estimated age exceeds the expected chronological age, whereas negative PhenoAgeAccel reflects a slower-than-expected rate of ageing [5,6,7].

Mammals possess two main forms of adipose tissue: brown adipose tissue (BAT) and white adipose tissue (WAT). WAT functions mainly as an energy storage depot, storing free fatty acids, and can be divided into visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) depending on its location in the body [8]. Various pathological processes are associated with ageing, such as cellular senescence, low-grade systemic inflammation, immunosenescence, mitochondrial dysfunction, and shifts in body composition and hormone levels [9,10,11]. These changes alter the cellular structure, insulin sensitivity, and inflammatory state of adipose tissue. Consequently, inflammation originating from adipose tissue systematically promotes the accumulation of senescent cells, while reduced insulin sensitivity may lead to insulin resistance [12]. The Homeostasis Model Assessment of Insulin Resistance (HOMA-IR) is a technique employed to assess the interaction between fasting glucose and insulin levels, providing an estimate of the extent of insulin resistance [13]. An elevated HOMA-IR or insulin resistance is associated with ageing [14]. Additionally, dysfunction in adipose tissue is correlated with age-related alterations in blood lipid profiles [15]. High-density lipoprotein cholesterol (HDL-C) levels are negatively correlated with the subcutaneous fat area (SFA), visceral fat area (VFA), and the VFA/SFA ratio [16, 17]. HDL-C plays a role in reducing oxidative stress within plasma and cellular compartments and counteracts mitochondrial apoptosis, a function that is negatively associated with ageing [18]. However, the connections between these factors and PhenoAge are still insufficiently investigated.

This study aimed to clarify the connections between the abdominal SFA and PhenoAge, as well as between the VFA and PhenoAge, among individuals aged 18 to 59 years through dose‒response analysis. Furthermore, the mediating and interactive effects of HOMA-IR and HDL-C on these associations across different age subgroups were examined, thus providing a conceptual foundation for future research on aging acceleration and offering valuable insights for managing health.

Methods

Data source

The data for this study were sourced from the National Health and Nutrition Examination Survey (NHANES) database, designed to assess the health and nutritional conditions of both adults and children across the United States. This analysis incorporated data from two NHANES cycles, 2015–2016 and 2017–2018, encompassing a total of 19,225 participants. As shown in Fig. 1, after excluding individuals aged under 18 years; individuals aged over 60 years old; and individuals with incomplete data on PhenoAge, SFA, VFA, insulin, glucose, and other covariates, the final sample comprised 2580 participants.

Measurement of PhenoAge

In accordance with the literature [6, 19], PhenoAge was calculated on the basis of the following parameters: age, creatinine, albumin, glucose, and alkaline phosphatase levels; mean cell volume; lymphocyte percentage; white blood cell count; red cell distribution width; C-reactive protein level. Blood samples were obtained in accordance with standardized protocols at the Mobile Examination Centre (MEC). The formula for calculating PhenoAge is displayed in Fig. 2. PhenoAgeAccel is defined as the residual obtained from a linear regression model that predicts PhenoAge based on chronological age. Positive residuals indicate PhenoAge acceleration, while negative residuals indicate PhenoAge deceleration or stasis.

Assessment of covariates

Demographic information, such as gender (male/female), age, race (Mexican American/other Hispanic/non-Hispanic/other race), income status (at or above poverty/below poverty), education (college graduate or above/high school graduate or equivalent college/less than 12th grade), and marital status (married/other), was collected through interviews. The category ‘at or above poverty’ refers to individuals with a poverty income ratio (PIR) of 1 or higher.

Measurements including body mass index (BMI), waist circumference, abdominal SFA, VFA, lipid levels, insulin levels, and blood glucose levels were obtained at the MEC by trained health technicians. Body fat areas were measured using dual-energy X-ray absorptiometry (DXA), which evaluates both VFA and SFA around the level of the fourth and fifth lumbar vertebrae (L4 and L5). Smoking status (yes/no) was determined on the basis of whether the participants had smoked at least 100 cigarettes in their life. Alcohol consumption (yes/no) was classified differently according to the study cycle: for the 2015–2016 cycle, alcohol consumption was defined as having consumed at least 12 alcoholic drinks in the past year or over a lifetime; for the 2017–2018 cycle, alcohol consumption was defined as having consumed any kind of alcohol and drinking more than 12 times. Having a history of cardiovascular disease was based on reports of congestive heart failure, coronary artery disease, angina, myocardial infarction, or stroke (MCQ160B-F). Hyperlipidaemia history was determined using responses to BPQ080, BPQ090D, and BPQ100D, whereas hypertension was determined using responses to BPQ020 and BPQ050A. Diabetes diagnosis was derived from responses to DIQ010, DIQ050, and DIQ070. Regular exercise was defined on the basis of responses to PAQ605, PAQ620, PAQ650, and PAQ665. The calculation of the HOMA-IR involved both blood glucose and insulin levels, as illustrated in Fig. 2.

Statistical analysis

In accordance with the survey methods and analytical guidelines outlined by NHANES, the NHANES sample was selected using a complex four-stage sampling design. For this study, sample weights were calculated using WTSAF2YR/2. Continuous variables are reported as the median ± standard deviation, with comparisons between groups performed using one-way analysis of variance (ANOVA). Categorical variables are expressed as percentages, and group differences were assessed using the chi-square test. To examine the association between body fat area and PhenoAgeAccel, logistic regression was applied. Model 1 was unadjusted, Model 2 included adjustments for gender, age, and race, while Model 3 built upon Model 2, incorporating adjustments for exercise, alcohol use, smoking status, marital status, education level, income, HbA1c, history of hypercholesterolemia, hypertension, diabetes, and cardiovascular disease. These variables encompass demographic, socioeconomic, lifestyle, and clinical factors, all of which have been shown in previous research to be intricately linked to metabolic health and ageing processes, potentially influencing adipose tissue function [13,14,15]. The restricted cubic spline (RCS) method was utilized for curve fitting and dose‒response analysis. Mediation analyses were conducted to investigate the relationships among HOMA-IR, HDL-C, body fat area, and PhenoAgeAccel. By analysing the combined effects of the association of HOMA-IR or HDL-C with body fat area on PhenoAgeAccel, further insights into their joint impact were obtained.

All statistical analyses were conducted using R software (version 4.2.2), Empower Stats software (version 4.2), and STATA software (version 18.0). P < 0.05 (two-tailed) was considered statistically significant.

Fig. 1

Flowchart

Fig. 2

Calculation formula

Results

Baseline characteristics

The baseline characteristics of a total of 2580 participants from the United States are presented in Table 1. Individuals exhibiting PhenoAgeAccel tended to be older; identify as Hispanic; and have elevated BMI, HbA1c, insulin, waist circumference, abdominal SFA, and VFA. Furthermore, individuals exhibiting PhenoAgeAccel demonstrated a higher prevalence of hypertension, cardiovascular diseases, and diabetes and were more likely to have a positive smoking status. Additionally, these individuals had lower HDL-C, education levels, and incomes (P < 0.05).

Relationships between body fat area and PhenoAgeAccel

The relationships between body fat area and PhenoAgeAccel are shown in Table 2. In this analysis, abdominal SFA and VFA were categorized on the basis of quartiles. A marked positive association of VFA with PhenoAgeAccel was observed after adjusting for variables including age, gender, race, exercise, alcohol consumption, smoking status, marital status, education level, income, HbA1c, and histories of hypercholesterolemia, hypertension, diabetes, and cardiovascular disease. Compared with that for quartile 1 (Q1; the lowest quartile), the odds ratio (OR) for Q4 (the highest quartile) was 8.339 (95% confidence interval (CI): 5.559 to 12.510), with P for trend < 0.001. For abdominal SFA, no statistically significant relationship was observed at the Q2 level. However, significant correlations were detected at the Q3 and Q4 levels. In Model 3, compared with Q1, the OR for Q4 was 7.756 (95% CI: 5.250 to 11.458), with P for trend < 0.001.

Table 1 Baseline characteristics according to age and phenotypic age

Table 2 Association of body fat area with phenoage acceleration

Subgroup analysis

Interaction tests revealed a differential relationship between VFA and PhenoAgeAccel across subgroups defined by chronological age, sex, and diabetes status (Fig. 3). In the chronological age subgroup analysis (P-interaction = 0.002), the ORs were 5.921 (95% CI: 4.364–8.034) for individuals aged 18–44 years and 2.934 (95% CI: 2.055–4.190) for those aged 45–59 years. In the sex subgroup (P-interaction = 0.026), the ORs were 5.331 (95% CI: 4.276–6.645) for males and 3.679 (95% CI: 2.618–5.170) for females. For the self-reported diabetes subgroup (P-interaction < 0.001), the ORs were 5.341 (95% CI: 2.338–12.199) for individuals with diabetes and 4.350 (95% CI: 3.350–5.648) for those without diabetes.

The relationship between abdominal SFA and PhenoAgeAccel varied across subgroups according to chronological age, alcohol consumption, and diabetes status. In the chronological age subgroup analysis (P-interaction = 0.035), the ORs were 1.783 (95% CI: 1.608–1.976) for individuals aged 18–44 years and 1.473 (95% CI: 1.284–1.690) for those aged 45–59 years. In the alcohol consumption subgroup (P-interaction = 0.004), the ORs were 1.920 (95% CI: 1.731–2.130) for individuals who consumed alcohol and 1.728 (95% CI: 1.481–2.015) for those who abstained. For the self-reported diabetes subgroup (P-interaction < 0.001), the ORs were 1.537 (95% CI: 1.269–1.863) for individuals with diabetes and 1.682 (95% CI: 1.550–1.825) for those without diabetes.

Fig. 3

Subgroup analysis. Each stratification was analyzed after making adjustments for age, gender, race, exercise, alcohol use, smoking status, marital status, education level, income, HbA1c, and histories of hypercholesterolemia, hypertension, diabetes, and cardiovascular disease. (except the stratification factor itself.) All analyses accounted for complex survey designs

RCS analysis

To evaluate the dose-response relationship between body fat area and PhenoAgeAccel, RCS analysis was performed, with chronological age used as a subgroup. In the population aged 18–59 years, both VFA and abdominal SFA exhibited nonlinear associations with PhenoAgeAccel, with thresholds of 0.925 m² and 3.137 m², respectively (Fig. 4). When analysed by age subgroup, RCS analysis revealed that for abdominal SFA and PhenoAgeAccel, nonlinear relationships were observedin individuals aged 18–44 years and 45–59 years, with corresponding thresholds of 2.969 m² and 3.394 m², respectively. In contrast, for VFA and PhenoAgeAccel, a nonlinear relationship was identified in individuals aged 18–44 years, with a threshold of 0.769 m², whereas a linear relationship was noted in those aged 45–59 years, with a threshold of 1.220 m².

Fig. 4

The dose-response relationship between body fat area and PhenoAgeAccel. Values corresponding to the vertical line indicate the body fat area when OR is equal to 1. (A). The dose-response relationship between VFA and PhenoAgeAccel stratified by chronological age. (B). The dose-response relationship between abdominal SFA and PhenoAgeAccel stratified by chronological age. Data were adjusted for factors including age, gender, race, physical activity, alcohol use, smoking status, marital status, education level, income, HbA1c, and histories of hypercholesterolemia, hypertension, diabetes, and cardiovascular disease. All analyses accounted for complex survey designs

Mediation effect analysis

As illustrated in Fig. 5, the HOMA-IR and HDL-C acted as partial mediators of the association between body fat area and PhenoAgeAccel. The mediating effect of the HOMA-IR accounted for 10.2% of the relationship between VFA and PhenoAgeAccel in participants aged 18–59 years (Table 3). When the data were stratified by chronological age, the mediating effect of HOMA-IR accounted for 13.4% of the relationship between VFA and PhenoAgeAccel in the 18–44 age group and 6.2% in the 45–59 age group. Additionally, the mediating effect of the HOMA-IR accounted for 4.7% of the relationship between abdominal SFA and PhenoAgeAccel. When the data were stratified by chronological age, the mediating effect of the HOMA-IR accounted for 6.9% and 3.5% of the relationship between abdominal SFA and PhenoAgeAccel for individuals aged 18–44 years and 45–59 years, respectively.

The mediating effect of HDL-C accounted for 15.3% of the relationship between VFA and PhenoAgeAccel in nonelderly adults (individuals aged 18–59 years) (Table 4). When the data were stratified by chronological age, the mediating effect of HDL-C accounted for 8.4% and 21.7% of the relationship between VFA and PhenoAgeAccel for individuals aged 18–44 years and 45–59 years, respectively. Additionally, HDL-C mediated 7.2% of the association between abdominal SFA and PhenoAgeAccel. Further analysis revealed that HDL-C did not significantly mediate the association between abdominal SFA and PhenoAgeAccel for those aged 18–44 years but did have a mediating effect in those aged 45–59 years, accounting for 11.6% of this relationship.

Fig. 5

Mediation effect analysis. (A). The mediation effect of HOMA-IR between body fat area and PhenoAgeAccel aged 18-59years. (B). The mediation effect of HDL-C between body fat area and PhenoAgeAccel aged 18-59years

Table 3 Association of body fat area with PhenoAgeAccel mediated by HOMA-IR

Table 4 Association of body fat area with PhenoAgeAccel mediated by HDL-C

Joint associations of body fat area combined with HOMA-IR or HDL-C with PhenoAgeAccel

In this study, the HOMA-IR threshold used for diagnosing insulin resistance was 2.73 [20]. As illustrated in Fig. 6, the relationships between body fat area combined with HOMA-IR and PhenoAgeAccel across chronological age subgroups were analysed. In individuals aged 18 to 59 years, a HOMA-IR ≥ 2.73, a VFA > 0.925 m², and an abdominal SFA > 3.137 m² were positively correlated with PhenoAgeAccel. The risk of developing PhenoAgeAccel was significantly greater when both HOMA-IR ≥ 2.73 and VFA > 0.925 m² were present (OR: 3.758, 95% CI: 2.918–4.840) or when both HOMA-IR ≥ 2.73 and abdominal SFA > 3.137 m² were present (OR: 4.395, 95% CI: 3.409–5.667) than when the HOMA-IR and body fat area were lower than these thresholds.

In this study, HDL-C levels were divided into four quartiles (Q1-Q4) for classification. As illustrated in Fig. 7, the relationships between body fat area combined with HDL-C and PhenoAgeAccel across different chronological age subgroups were analysed. In individuals aged 18 to 59 years, the combination of VFA > 0.925 m² and 0.16 ≤ HDL-C < 1.60 mmol/L, as well as the combination of abdominal SFA > 3.137 m² and 0.16 ≤ HDL-C < 1.60 mmol/L, were positively correlated with PhenoAgeAccel. Conversely, the combination of VFA ≤ 0.925 m² and 1.60 ≤ HDL-C < 3.90 mmol/L (OR: 0.576, 95% CI: 0.393–0.845), as well as the combination of abdominal SFA ≤ 3.137 m² and 1.60 ≤ HDL-C < 3.90 mmol/L (OR: 0.690, 95% CI: 0.482–0.988), demonstrated a negative relationship with PhenoAgeAccel.

Fig. 6

The joint association of body fat area and HOMA-IR with PhenoAgeAccel. All models were adjusted for factors including age, gender, race, physical activity, alcohol use, smoking status, marital status, education level, income, HbA1c, and histories of hypercholesterolemia, hypertension, diabetes, and cardiovascular disease

Fig. 7

The joint association of body fat area and HDL-C with PhenoAgeAccel. Q1: 0.16 ≤ HDL-C < 1.09mmol/L; Q2: 1.09 ≤ HDL-C < 1.32mmol/L; Q3: 1.32 ≤ HDL-C < 1.60mmol/L; Q4: 1.60 ≤ HDL-C < 3.90mmol/L. All models were adjusted for factors including age, gender, race, physical activity, alcohol use, smoking status, marital status, education level, income, HbA1c, and histories of hypercholesterolemia, hypertension, diabetes, and cardiovascular disease

Discussion

This study analyzed the dose-response relationships between abdominal SFA and PhenoAge, as well as between VFA and PhenoAge, providing new insights into the factors contributing to accelerated aging. The analyses revealed a nonlinear relationship between abdominal SFA and PhenoAgeAccel in individuals aged 18 to 59 years, with a threshold of 3.137 m². Further analysis by age subgroup revealed that this nonlinear relationship persisted in individuals aged 18 to 44 years (threshold: 2.969 m²), whereas it shifted to a linear relationship in those aged 45 to 59 years (threshold: 3.394 m²). Similarly, a nonlinear relationship between VFA and PhenoAgeAccel was observed, with an overall threshold of 0.925 m²; within the 18-44-year age group, the threshold was 0.769 m², and the threshold increased to 1.220 m² in the 45-59-year age group. To further investigate the factors influencing the relationship between body fat area and PhenoAgeAccel, mediation and interaction analyses were conducted. These analyses revealed that HOMA-IR and HDL-C partially mediated these associations. Additionally, elevated body fat area or insulin resistance was positively correlated with PhenoAgeAccel, whereas a lower body fat area combined with high HDL-C levels was negatively associated with PhenoAgeAccel. Taken together, these findings indicate that, in nonelderly populations, insulin resistance and dyslipidaemia may act as key mediating mechanisms linking adiposity to ageing acceleration. Furthermore, age-specific management of body fat area, the HOMA-IR, and HDL-C may serve an essential function in mitigating the ageing process.

Most research findings indicate that VAT contributes to the acceleration of ageing, whereas SAT is associated with a deceleration of the ageing process. VAT influences the ageing process mainly through immune inflammation mediated by macrophages, dendritic cells, B cells, and T cells [21]. Additionally, the hypertrophy, hyperplasia, and fibrosis of VAT can also have substantial impacts on the ageing process [22, 23]. As individuals age, adipose tissue is progressively redistributed to the abdomen and visceral organs, resulting in increased VAT and reduced SAT. This shift increases the VAT/SAT ratio and is associated with impaired adipogenesis, altered cytokine and hormone secretion, and disruptions in lipid storage and metabolism [4, 24,25,26,27]. Reduced SAT function leads to imbalances in adipokine secretion and increased low-grade chronic inflammation, contributing to metabolic disorders metabolic disorders, including diabetes and cardiovascular diseases. Chronic, sterile inflammation resulting from these metabolic interactions further accelerates ageing [12, 28]. Additionally, SAT influences ageing in specific tissues. For example, reduced facial SAT thickness and altered adipocyte size correlate with accelerated facial ageing [29]. In a prior dose-response study, a positive correlation was found between the visceral adiposity index (VAI) and PhenoAgeAccel when the VAI value was below 10.543, but this association became nonsignificant when the VAI exceeded this threshold [30]. These results indicate that there may be a critical threshold beyond which excess fat tissue contributes to accelerated ageing. This study focused on evaluating abdominal SFA and VFA as key variables. An analysis of the dose‒response relationships between body fat area and PhenoAgeAccel revealed that these relationships are modulated by specific thresholds across age groups. In individuals aged 18 to 59 years, the threshold for VFA was 0.925 m², and for abdominal SFA, the threshold was 3.137 m². Further age-based subgroup analyses revealed distinct thresholds: for individuals aged 18 to 44 years and those aged 45 to 59 years, a lower body fat area showed a negative correlation with PhenoAgeAccel, suggesting that both VFA and SFA may act as protective factors against ageing. Conversely, when the body fat area exceeded these thresholds, positive correlations with PhenoAgeAccel emerged, indicating that elevated VFA and abdominal SFA may be risk factors for accelerated ageing.

Mediation and interaction analyses suggest that the HOMA-IR, a critical marker of insulin resistance, mediates the relationship between body fat area and PhenoAgeAccel. Elevated body fat area and insulin resistance were both positively correlated with PhenoAgeAccel, consistent with previous research findings [14]. Changes in the metabolism of fatty acids in VAT and the expression of adipogenic genes are associated with markers of insulin resistance. The hypertrophy of abdominal subcutaneous adipocytes may trigger lipid efflux, leading to ectopic fat deposition, and the expansion of VAT, all of which exacerbate systemic insulin resistance [31]. Notably, insulin resistance is closely associated with ageing. Research has demonstrated that enhanced insulin sensitivity plays a crucial role in promoting longevity, while heightened insulin resistance is linked to a higher risk of functional decline and frailty, particularly among older adults without diabetes [32, 33]. Moreover, the onset of insulin resistance during early to mid-adulthood may act as a significant predictor of long-term risks related to metabolic disorders, as well as psychological and neurobehavioural conditions [34]. Age-related variations in insulin sensitivity may explain the differences in the mediation effects of HOMA-IR across different age subgroups. As individuals age, structural and functional changes in adipose tissue reduce insulin responsiveness [35]. Consequently, in young adults, increases in body fat area have a more pronounced impact on the HOMA-IR, establishing a stronger link to the ageing process.

Another key finding from this study is that HDL-C mediates the relationship between body fat area and PhenoAgeAccel. Specifically, a combination of lower fat area and higher HDL-C levels is negatively correlated with accelerated aging. Additionally, the results suggest that the mediating effect of HDL-C differs across age groups. Although research suggests that age does not significantly impact HDL-C levels, older individuals generally have higher HDL-C levels than younger individuals do [15]. This observance may explain why different mediating effects of HDL-C were observed in this study population. Adipose tissue can influence HDL-C metabolism, as cholesterol efflux from adipose tissue is essential for the creation of fully mature and functional HDL-C particles [31]. Inflammation in adipose tissue stimulates the secretion of cytokines like TNF-α, CRP, IL-1β, and IL-6, while suppressing cholesterol transport proteins such as ABCA1, thereby impairing HDL-C maturation [36]. Dyslipidaemia plays a crucial role in the progression of cellular senescence and atherosclerosis, both of which are closely linked to lifespan [37]. The effects of atherosclerosis and ageing on HDL-C are thought to be primarily determined by the functional properties of HDL-C rather than its cholesterol content [38]. HDL-C may delay the ageing process by directly modulating senescence signals or influencing the survival factor Klotho [18]. However, dysfunctional HDL-C, as well as the pro-inflammatory and pro-oxidative particles within it, are linked to an increased risk of acute coronary syndrome [37]. Further studies are necessary to completely clarify HDL-C’s role in the aging process.

Strengths and limitations of the research

There are several strengths of this study. First, weighted logistic regression and RCS analysis were employed to evaluate the dose‒response relationships between body fat area and PhenoAgeAccel, ensuring that all the data accounted for complex survey designs, which increased the precision of the results. Furthermore, subgroup analyses were conducted to explore how body fat area, the HOMA-IR, and HDL-C influence PhenoAgeAccel across different age groups, thus increasing the relevance and applicability of the findings. Nonetheless, this study has several limitations. The data were sourced from the NHANES database, and some data are missing from this database. Additionally, the analysis was limited to abdominal adipose tissue, and future research is needed to explore the impact of adipose tissue from various anatomical locations. Finally, as this study used a cross-sectional design, prospective studies are necessary to validate these findings and offer a more comprehensive understanding of the causal relationships between the variables.

Conclusions

In conclusion, this study revealed the nonlinear relationships among abdominal SFA, VFA, and PhenoAgeAccel while identifying characteristic thresholds across different age groups. This study highlights the intricate impact of fat distribution on the ageing process and refine the roles of HOMA-IR and HDL-C in various age cohorts. Furthermore, these findings provide a biological basis for future screening for accelerated ageing and appropriate intervention in high-risk populations and offer valuable insights for guiding personalized clinical interventions and health management strategies.