Preamble

Title: Association Between Lipid Levels and the Risk of Sarcopenic Obesity in Middle-aged and Elderly Chinese: A Cohort Study

Authors: XU Chunyan¹², HE Ling¹, GUO Canhui¹², LAI Hurong¹², LIAO Caifeng¹², TU Huaijun¹*

Affiliations:
1. Department of Geratology, the Second Affiliated Hospital, Jiangxi Medical College, Nanchang University, Nanchang 330006, China
2. Jiangxi Medical College, Nanchang University, Nanchang 330006, China

Corresponding author: TU Huaijun, Chief physician; E-mail: ndefy10061@ncu.edu.cn

Abstract

Background: With population aging intensifying, the incidence and impact of geriatric syndromes have garnered widespread attention, with sarcopenic obesity (SO) emerging as a key research focus. While lipid levels may represent an important influence on SO, current research has yet to reach consistent conclusions regarding the relationship between specific lipid components and SO risk.

Objective: To investigate the association between blood lipid levels and SO incidence risk among middle-aged and elderly Chinese adults.

Methods: Using data from the China Health and Retirement Longitudinal Study (CHARLS) 2011–2015, we conducted a cohort study of middle-aged and elderly individuals without SO at baseline in 2011. Baseline total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C) served as exposure variables, while SO occurrence in 2013 or 2015 constituted the outcome. SO was defined using dual criteria for both sarcopenia and obesity: sarcopenia was determined by appendicular skeletal muscle mass adjusted for weight (ASM/W), and obesity was defined as BMI ≥ 25 kg/m². Cox proportional hazards regression models were constructed to analyze the relationship between baseline lipid levels and SO risk, with potential non-linear associations examined using restricted cubic spline models.

Results: A total of 5,268 participants were included (median age 58 [52, 64] years), with 20,592 person-years of cumulative follow-up. We identified 382 incident SO cases, yielding a cumulative incidence of 7.25%—5.22% (128/2,451) in men and 9.02% (254/2,817) in women. In fully adjusted Cox models, compared with the lowest quartile (Q1), the highest TC quartile (Q4) showed significantly elevated SO risk (HR=1.35, 95%CI=1.00–1.82). For TG, Q2 (HR=1.55, 95%CI=1.08–2.21), Q3 (HR=2.07, 95%CI=1.48–2.90), and Q4 (HR=2.53, 95%CI=1.82–3.52) all demonstrated significantly increased risk. Similarly for LDL-C, Q2 (HR=1.38, 95%CI=1.02–1.88) and Q4 (HR=1.44, 95%CI=1.07–1.95) showed elevated risk. Conversely, higher HDL-C quartiles exhibited protective effects: Q2 (HR=0.75, 95%CI=0.58–0.96), Q3 (HR=0.54, 95%CI=0.41–0.71), and Q4 (HR=0.43, 95%CI=0.31–0.58). Restricted cubic spline analysis revealed an inverted L-shaped association between TG levels and SO risk (P-nonlinear<0.001), while TC (P-nonlinear=0.731), HDL-C (P-nonlinear=0.600), and LDL-C (P-nonlinear=0.400) showed linear relationships.

Conclusion: Elevated TG, TC, and LDL-C represent risk factors for SO in Chinese middle-aged and elderly populations, whereas HDL-C exerts a protective effect, with TG demonstrating an inverted L-shaped dose-response relationship. These findings underscore the importance of lipid management for SO prevention and treatment in this population.

Keywords: Middle aged; Aged; Sarcopenic obesity; Hyperlipidemias; Cholesterol, LDL; Cholesterol, HDL; Lipid levels; China Health and Retirement Longitudinal Study

1. Subjects and Methods

1.1 Study Subjects

Data were derived from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative longitudinal survey of community-dwelling adults aged 45 and older. The study is publicly available at https://charls.pku.edu.cn with assessments conducted every 2–3 years (five waves from 2011–2020). The CHARLS protocol was approved by the Biomedical Ethics Review Committee of Peking University (approval number: IRB00001052-11015), and all participants provided written informed consent.

To examine the longitudinal relationship between blood lipids and SO, we utilized data from the 2011, 2013, and 2015 waves, with 2011 as baseline and 2015 as the outcome assessment point. Data from 2018 and 2020 were excluded due to lack of physical examination information required for SO evaluation.

Inclusion criteria: Participants in the 2011 wave with complete SO-related indicators.

Exclusion criteria: (1) Age <45 years or missing data on age, lipids, sex, or medical history; (2) SO diagnosis at baseline; (3) Missing SO information during follow-up (2013 and 2015).

The final analytic sample comprised 5,268 participants [FIGURE:1].

1.2 Measurements

1.2.1 Exposure Variables

Baseline levels of four lipid parameters served as exposure variables: total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C). Fasting venous blood samples were collected by trained medical staff after overnight fasting, centrifuged, and transported to the Capital Medical University laboratory in Beijing for enzymatic colorimetric assay. Participants were stratified into quartile groups based on baseline lipid levels:

• TC: Q1 <167.39 mg/dL, Q2 167.39–190.59 mg/dL, Q3 190.59–214.56 mg/dL, Q4 ≥214.56 mg/dL
• TG: Q1 <73.46 mg/dL, Q2 73.46–102.66 mg/dL, Q3 102.66–148.67 mg/dL, Q4 ≥148.67 mg/dL
• HDL-C: Q1 <40.97 mg/dL, Q2 40.97–50.25 mg/dL, Q3 50.25–60.69 mg/dL, Q4 ≥60.69 mg/dL
• LDL-C: Q1 <93.55 mg/dL, Q2 93.55–114.04 mg/dL, Q3 114.04–136.37 mg/dL, Q4 ≥136.37 mg/dL

1.2.2 Outcome Variable

The outcome was incident SO, defined using combined criteria for obesity and sarcopenia. Obesity was defined as BMI ≥25 kg/m² according to Asia-Pacific standards [8]. For sarcopenia, we used appendicular skeletal muscle mass adjusted for weight (ASM/W) as the metric [9], calculated as: sarcopenia index = (ASM/W) × 100%.

ASM was estimated using a validated anthropometric equation for Chinese populations that demonstrates good agreement with dual-energy X-ray absorptiometry (DXA) measurements [10-11]:

ASM = 0.193 × weight (kg) + 0.107 × height (cm) – 4.157 × sex – 0.037 × age (years) – 2.631

Height and weight were measured using Seca TM213 stadiometers and Omron TM HN-286 scales, respectively (male=1, female=2). Following Janssen et al. [9], sarcopenia was defined as a sarcopenia index below the gender-specific mean minus one standard deviation of a young reference group (aged 18–39 years). As CHARLS lacks young participants, we applied reference values from Kim et al. [12] derived from 4,918 Asian adults aged 20–39 years, with cutoffs of 30.18% for men and 23.75% for women.

1.2.3 Covariates

Covariates included sociodemographic characteristics (sex, age, education, marital status, residence type), lifestyle factors (smoking and alcohol use history), chronic disease history (hypertension, diabetes, lipid-lowering medication use), and C-reactive protein (CRP) levels.

1.3 Statistical Analysis

Data were analyzed using SPSS 27.0 and R 4.2.2. Normally distributed continuous variables were presented as mean ± standard deviation, while non-normally distributed variables were expressed as median (P25, P75). Group comparisons employed independent t-tests or Mann-Whitney U tests for continuous variables and chi-square tests for categorical variables.

Cox proportional hazards regression was used to examine associations between baseline lipid levels and SO risk across three models: Model 1 (unadjusted), Model 2 (adjusted for age, sex, urban/rural residence, education, and marital status), and Model 3 (additionally adjusted for smoking, alcohol use, disease history, medication history, and CRP). Lipid levels were analyzed as categorical variables (quartile groups), with Q1 as the reference. Trend tests were performed across quartiles. Results are presented as hazard ratios (HR) with 95% confidence intervals (CI). Restricted cubic spline analysis was conducted for significant indicators to explore potential non-linear associations. Two-sided tests were used with statistical significance set at P<0.05.

2. Results

2.1 Baseline Characteristics

The final sample included 5,268 participants with a median age of 58 (52, 64) years and 20,592 person-years of cumulative follow-up. We identified 382 incident SO cases, yielding a cumulative incidence of 7.25%. The incidence rate was 5.22% (128/2,451) in men and 9.02% (254/2,817) in women. By age group, incidence was 4.85% (87/1,795) for ages 45–<55 years, 7.55% (163/2,158) for 55–<65 years, 10.22% (106/1,037) for 65–<75 years, and 9.35% (26/278) for ≥75 years.

Compared with those who remained SO-free, participants who developed SO were older and had higher baseline weight, BMI, CRP, TC, TG, and LDL-C, but lower height, sarcopenia index, and HDL-C. The SO group also had higher proportions of women, urban residents, individuals with lower education levels, non-smokers, non-drinkers, hypertensive patients, and those using lipid-lowering medications (all P<0.05). No significant differences were observed in marital status or diabetes history [TABLE:1].

2.2 Cox Proportional Hazards Regression Analysis

Using incident SO during follow-up as the outcome variable (yes=1, no=0) and baseline lipid quartiles as predictors (with Q1 as reference), Cox regression analysis revealed significant associations between all four lipid parameters and SO risk after adjusting for sociodemographics, health behaviors, and chronic diseases (all P<0.05).

Specifically, compared with TC Q1, the Q4 group showed significantly elevated SO risk (HR=1.35, 95%CI=1.00–1.82). For TG, Q2 (HR=1.55, 95%CI=1.08–2.21), Q3 (HR=2.07, 95%CI=1.48–2.90), and Q4 (HR=2.53, 95%CI=1.82–3.52) all demonstrated significantly increased risk. For LDL-C, Q2 (HR=1.38, 95%CI=1.02–1.88) and Q4 (HR=1.44, 95%CI=1.07–1.95) showed elevated risk. Conversely, higher HDL-C quartiles were protective: Q2 (HR=0.75, 95%CI=0.58–0.96), Q3 (HR=0.54, 95%CI=0.41–0.71), and Q4 (HR=0.43, 95%CI=0.31–0.58). Significant dose-response trends were observed for all lipids (P-trend<0.05) [TABLE:2].

2.3 Dose-Response Relationship Between Lipid Levels and SO Risk

Restricted cubic spline models adjusting for multiple confounders revealed a non-linear, inverted L-shaped association between baseline TG levels and SO risk (P-nonlinear<0.001). SO risk increased sharply with rising TG levels but plateaued when TG exceeded 116.67 mg/dL. In contrast, TC (P-nonlinear=0.731), HDL-C (P-nonlinear=0.600), and LDL-C (P-nonlinear=0.400) showed linear associations with SO risk [FIGURE:2].

3. Discussion

This nationally representative longitudinal study of Chinese middle-aged and older adults demonstrates that TC, TG, and LDL-C are positively associated with SO risk, while HDL-C exhibits an inverse relationship. Notably, TG shows an inverted L-shaped dose-response pattern.

Our findings indicate that, compared with the lowest quartile, elevated levels of TC (HR=1.35, 95%CI=1.00–1.82), TG (HR=2.53, 95%CI=1.82–3.52), and LDL-C (HR=1.44, 95%CI=1.07–1.95) in the highest quartiles constitute risk factors for SO, whereas HDL-C (HR=0.43, 95%CI=0.31–0.58) is protective. The non-linear relationship between TG and SO risk represents a novel contribution to understanding the lipid-SO association.

Current evidence on lipid abnormalities and SO remains limited and largely cross-sectional [6-7,13-14]. The Korea National Health and Nutrition Examination Survey (KNHANES) of 3,483 adults aged ≥65 found that the SO group had higher dyslipidemia risk (OR=2.82, 95%CI=1.76–4.51) than those with sarcopenia or obesity alone, with high TC and TG and low HDL-C independently associated with SO in men [6]. Lu et al. [13] analyzed body composition in 600 Chinese adults and found high TG and low HDL-C correlated with increased SO risk. A cross-sectional study of 4,500 Chinese adults aged ≥50 reported TC (OR=1.35, 95%CI=1.12–1.63) and LDL-C (OR=1.45, 95%CI=1.15–1.83) as SO risk factors, but found no association for TG or HDL-C [7]. Another observational study of 14,926 individuals aged 35–74 reported that only TC and HDL-C correlated positively with SO risk in univariate analysis, but these associations disappeared after adjustment [14]—findings inconsistent with our results. These discrepancies may stem from different SO diagnostic criteria. Notably, Baek et al. [6] and Lu et al. [13] used weight-adjusted muscle mass (ASM/W) and defined obesity as BMI ≥25 kg/m², matching our definition. In contrast, Liu et al. [7] and Yin et al. [14] used height-squared adjustment (ASM/Ht²) and defined obesity by body fat percentage. Furthermore, our cohort design enabled examination of longitudinal associations between baseline lipids and incident SO, revealing the novel non-linear TG relationship. Variations in adjusted covariates across studies may also contribute to divergent findings.

The underlying mechanisms linking lipids to SO remain incompletely understood. Research suggests SO development is mediated by aging-related structural and functional changes in adipose and muscle tissues that interact reciprocally. Adipose tissue inflammation triggers ectopic fat distribution, particularly infiltration into skeletal muscle, impairing muscle strength and function and culminating in SO [15]. Lipid accumulation in myocytes disrupts mitochondrial function, promotes excessive reactive oxygen species generation, and impairs fatty acid β-oxidation, collectively inducing lipotoxicity, insulin resistance, and local inflammatory responses [16]. Inflammation, in turn, exacerbates muscle fat deposition and insulin resistance [17]. As muscle is a primary target of insulin, muscle mass reduction further worsens insulin resistance [18]. These processes create a vicious cycle of chronic inflammation, insulin resistance, and hyperlipidemia that promotes SO development. Thus, dyslipidemia may critically drive SO pathogenesis through multiple pathways including enhanced fat deposition, inflammation, and insulin resistance.

This study has several limitations. First, skeletal muscle mass was estimated using a validated Chinese-specific anthropometric equation rather than direct measurement via DXA or bioelectrical impedance analysis, though the equation has demonstrated good agreement with DXA in prior research. Second, we relied on a single baseline lipid measurement, precluding assessment of dynamic lipid changes over time. Finally, despite adjusting for numerous confounders, unmeasured factors such as nutritional status and physical activity levels may remain. We could not incorporate physical activity due to substantial missing data, and CHARLS does not collect nutrition data.

In conclusion, this cohort study reveals that TG, TC, and LDL-C are risk factors for SO incidence in Chinese middle-aged and elderly populations, while HDL-C is protective, with TG exhibiting an inverted L-shaped dose-response relationship. These findings emphasize the importance of lipid management for SO prevention and provide potential biomarkers and therapeutic targets.

Author Contributions: XU Chunyan conceptualized the study, curated data, performed statistical analysis, and wrote the original draft. HE Ling curated data, provided statistical design input, and assisted with manuscript revision. GUO Canhui and LAI Hurong curated data and contributed to feasibility analysis. LIAO Caifeng curated data. TU Huaijun supervised quality control and review and assumed overall responsibility for the manuscript. All authors approved the final version.

Conflict of Interest: The authors declare no conflicts of interest.

Funding: Jiangxi Provincial Natural Science Foundation Youth Project (20224BAB216014); Nanchang University Second Affiliated Hospital Internal Funding Project (2023efyB01)

References

[1] JI T, TANG Z, LI Y, et al. Current status of diagnostic criteria for sarcopenic obesity in the elderly[J]. Practical Geriatrics, 2021, 35(1): 74-76.

[2] SILVEIRA E A, DA SILVA FILHO R R, SPEXOTO M C B, et al. The role of sarcopenic obesity in cancer and cardiovascular disease: a synthesis of the evidence on pathophysiological aspects and clinical implications[J]. Int J Mol Sci, 2021, 22(9): 4339. DOI: 10.3390/ijms22094339.

[3] FÁBREGA-CUADROS R, CRUZ-DÍAZ D, MARTÍNEZ-AMAT A, et al. Associations of sleep and depression with obesity and sarcopenia in middle-aged and older adults[J]. Maturitas, 2020, 142: 1-7. DOI: 10.1016/j.maturitas.2020.06.019.

[4] HIRANI V, NAGANATHAN V, BLYTH F, et al. Longitudinal associations between body composition, sarcopenic obesity and outcomes of frailty, disability, institutionalisation and mortality in community-dwelling older men: The Concord Health and Ageing in Men Project[J]. Age Ageing, 2017, 46(3): 413-420. DOI: 10.1093/ageing/afw214.

[5] DONINI L M, BUSETTO L, BAUER J M, et al. Critical appraisal of definitions and diagnostic criteria for sarcopenic obesity based on a systematic review[J]. Clin Nutr, 2020, 39(8): 2368-2388. DOI: 10.1016/j.clnu.2019.11.024.

[6] BAEK S J, NAM G E, HAN K D, et al. Sarcopenia and sarcopenic obesity and their association with dyslipidemia in Korean elderly men: the 2008-2010 Korea National Health and Nutrition Examination Survey[J]. J Endocrinol Invest, 2014, 37(3): 247-260. DOI: 10.1007/s40618-013-0011-3.

[7] LIU X, HAO Q, YUE J, et al. Sarcopenia, obesity and sarcopenia obesity in comparison: prevalence, metabolic profile, and key differences: results from WCHAT study[J]. J Nutr Health Aging, 2020, 24(4): 429-437. DOI: 10.1007/s12603-020-1332-5.

[8] Obesity: preventing and managing the global epidemic. Report of a WHO consultation[J]. World Health Organ Tech Rep Ser, 2000, 894: i-xii, 1-253.

[9] JANSSEN I, HEYMSFIELD S B, ROSS R. Low relative skeletal muscle mass (sarcopenia) in older persons is associated with functional impairment and physical disability[J]. J Am Geriatr Soc, 2002, 50(5): 889-896. DOI: 10.1046/j.1532-5415.2002.50216.x.

[10] YANG M, HU X Y, WANG H Z, et al. Sarcopenia predicts readmission and mortality in elderly patients in acute care wards: a prospective study[J]. J Cachexia Sarcopenia Muscle, 2017, 8(2): 251-258. DOI: 10.1002/jcsm.12163.

[11] WEN X, WANG M, JIANG C M, et al. Anthropometric equation for estimation of appendicular skeletal muscle mass in Chinese adults[J]. Asia Pac J Clin Nutr, 2011, 20(4): 551-556.

[12] KIM B, KIM G, LEE Y, et al. Triglyceride-glucose index as a potential indicator of sarcopenic obesity in older people[J]. Nutrients, 2023, 15(3): 555. DOI: 10.3390/nu15030555.

[13] LU C W, YANG K C, CHANG H H, et al. Sarcopenic obesity is closely associated with metabolic syndrome[J]. Obes Res Clin Pract, 2013, 7(4): e301-7. DOI: 10.1016/j.orcp.2012.02.003.

[14] YIN T, ZHANG J X, WANG F X, et al. The association between sarcopenic obesity and hypertension, diabetes, and abnormal lipid metabolism in Chinese adults[J]. Diabetes Metab Syndr Obes, 2021, 14: 1963-1973. DOI: 10.2147/DMSO.S308387.

[15] BATSIS J A, VILLAREAL D T. Sarcopenic obesity in older adults: aetiology, epidemiology and treatment strategies[J]. Nat Rev Endocrinol, 2018, 14(9): 513-537. DOI: 10.1038/s41574-018-0062-9.

[16] ZHAI X Y, HAN Y Y, ZHOU Y. Research progress on pathogenesis, diagnosis and treatment of sarcopenic obesity[J]. Electronic Journal of Metabolism and Nutrition of Cancer, 2023, 10(5): 579-584. DOI: 10.16689/j.cnki.cn11-9349/r.2023.05.002.

[17] SCHRAGER M A, METTER E J, SIMONSICK E, et al. Sarcopenic obesity and inflammation in the InCHIANTI study[J]. J Appl Physiol (1985), 2007, 102(3): 919-925. DOI: 10.1152/japplphysiol.00627.2006.

[18] PETERSEN K F, DUFOUR S, SAVAGE D B, et al. The role of skeletal muscle insulin resistance in the pathogenesis of the metabolic syndrome[J]. Proc Natl Acad Sci USA, 2007, 104(31): 12587-12594. DOI: 10.1073/pnas.0705408104.

(Received: January 10, 2025; Revised: June 15, 2025)
(Editor: ZHAO Yuecui)