Zhang Yueyu, Chen Xinyi, Wang Yu, Tang Yi, Zhang Kangrui, Wu Juncang
Anhui Medical University, Hefei, 230032, Anhui, China.
Department of Neurology, The Second People's Hospital of Hefei, Hefei, 230011, Anhui, China.
Eur J Med Res. 2025 Jun 23;30(1):513. doi: 10.1186/s40001-025-02802-1.
Accumulating evidence indicates a potential link between insulin resistance (IR) and depression, although the bidirectional nature and underlying mechanisms of this association remain poorly understood. This study aims to systematically investigate the associations between multiple IR indices-specifically the Homeostatic Model Assessment of Insulin Resistance (HOMA-IR), Lipid Accumulation Product (LAP), and Triglyceride-Glucose indice (TyG)-and the prevalence of depression.
Data from 12,011 participants in the National Health and Nutrition Examination Survey (NHANES) were analyzed. IR was quantified using three indices: HOMA-IR, LAP, and TyG. Baseline demographic and clinical characteristics were compared between participants with and without depression following stratification by depression status. Weighted multivariate logistic regression models were employed to evaluate the associations between IR indices (categorized into quartiles) and depression. Nonlinear relationships were explored using threshold effect analysis, restricted cubic spline (RCS) models, and smooth curve fitting. Subgroup analyses were performed to assess heterogeneity by age, gender, poverty level, and comorbidities (e.g., cardiovascular disease, hypertension).
The depressed group (n = 971) exhibited significantly higher IR indices compared to the non-depressed group (n = 11,040). In the fully adjusted model (Model 3), both LAP (Q4 vs. Q1: OR = 1.569, 95% CI 1.234-1.998) and TyG (Q4 vs. Q1: OR = 1.497, 95% CI 1.182-1.896) were significantly associated with depression, whereas the association for HOMA-IR was attenuated (Q4 vs. Q1: OR = 1.310, p = 0.099). Threshold effect analysis revealed a nonlinear "inverted L-shaped" relationship between HOMA-IR, LAP, and depression, with effect modification observed at specific indice thresholds. Subgroup analyses demonstrated stronger associations in males (LAP: OR = 1.23, p < 0.01; TyG: OR = 1.31, p < 0.05), individuals with coronary heart disease (LAP: OR = 1.68, p < 0.001), and stroke survivors (LAP: OR = 1.42, p = 0.023 for interaction).
This study provides robust evidence of significant associations between IR indices (LAP and TyG) and depression, with a notable nonlinear "inverted L-shaped" relationship observed for LAP. Subgroup analyses highlighted stronger correlations in older adults (≥ 59 years), patients with coronary heart disease, stroke survivors, males, and individuals with hypertension. These findings enhance our understanding of the metabolic pathways underlying depression and emphasize the importance of integrating IR indices into mental health risk assessments. The results also offer a theoretical basis for personalized interventions targeting metabolic abnormalities in depression prevention and treatment.
越来越多的证据表明胰岛素抵抗（IR）与抑郁症之间存在潜在联系，尽管这种关联的双向性和潜在机制仍知之甚少。本研究旨在系统地调查多种IR指标（特别是胰岛素抵抗稳态模型评估（HOMA-IR）、脂质蓄积产物（LAP）和甘油三酯-葡萄糖指数（TyG））与抑郁症患病率之间的关联。
对国家健康与营养检查调查（NHANES）中12,011名参与者的数据进行了分析。使用三个指标对IR进行量化：HOMA-IR、LAP和TyG。在按抑郁状态分层后，比较了有抑郁症和无抑郁症参与者的基线人口统计学和临床特征。采用加权多变量逻辑回归模型评估IR指标（分为四分位数）与抑郁症之间的关联。使用阈值效应分析、受限立方样条（RCS）模型和平滑曲线拟合探索非线性关系。进行亚组分析以评估年龄、性别、贫困水平和合并症（如心血管疾病、高血压）的异质性。
与非抑郁组（n = 11,040）相比，抑郁组（n = 971）的IR指标显著更高。在完全调整模型（模型3）中，LAP（四分位数4 vs. 四分位数1：OR = 1.569，95% CI 1.234 - 1.998）和TyG（四分位数4 vs. 四分位数1：OR = 1.497，95% CI 1.182 - 1.896）均与抑郁症显著相关，而HOMA-IR的关联减弱（四分位数4 vs. 四分位数1：OR = 1.310，p = 0.099）。阈值效应分析揭示了HOMA-IR、LAP与抑郁症之间呈非线性“倒L形”关系，在特定指标阈值处观察到效应修饰。亚组分析表明，在男性（LAP：OR = 1.23，p < 0.01；TyG：OR = 1.31，p < 0.05）、冠心病患者（LAP：OR = 1.68，p < 0.001）和中风幸存者（LAP：交互作用p = 0.023）中关联更强。
本研究提供了有力证据，证明IR指标（LAP和TyG）与抑郁症之间存在显著关联，LAP呈现出明显的非线性“倒L形”关系。亚组分析突出了在老年人（≥59岁）、冠心病患者、中风幸存者、男性和高血压患者中更强的相关性。这些发现加深了我们对抑郁症潜在代谢途径的理解，并强调了将IR指标纳入心理健康风险评估的重要性。研究结果还为针对抑郁症预防和治疗中代谢异常的个性化干预提供了理论依据。
