Previous studies have suggested associations between the triglyceride glucose index (TyG) , body mass index (BMI) , and cognitive function. However, when exploring the association between TyG and cognitive function, these studies have tended to give less consideration to the potential influence of BMI, and the relationship among these three factors remains unclear. Examining the effect of BMI on the association between TyG and cognitive function is of great significance for developing precise interventions to prevent cognitive decline in older adults.

To explore the effect of BMI on the association between TyG and cognitive function in older adults, and to provide a reference for the prevention of cognitive decline in older adults.

From July to August 2019, the rural elderly residents aged 60 years and older were selected using the multi-stage cluster sampling method from 12 villages in 2 counties (districts) of Guizhou Province, and questionnaire surveys, general physical examinations, cognitive function assessments, and laboratory measurements were carried out among them. Cognitive function was evaluated using the Mini-Mental State Examination (MMSE) scale, and TyG was calculated using laboratory indices (triglyceride, fasting blood glucose) . Spearman rank correlation analysis and multiple linear regression model were used to investigate the association between TyG and cognitive function and the interaction effect between TyG levels and BMI subgroups on cognitive function in older adults.

The results of the interaction analysis of the impact of TyG and BMI on cognitive function showed that in model 1 without adjusting for covariates, the interaction term between TyG and low-BMI had an effect on MMSE scores of older adults (β=-0.515, 95%CI=-0.809--0.222, P<0.001) . In model 2 after adjusting for the relevant covariates, the interaction term between TyG and low-BMI still had an effect on MMSE scores in older adults (β=-0.351, 95%CI=-0.601--0.102, P<0.05) . Further stratification of body mass index followed by multiple linear regression analyses showed that in the under weight group, TyG was negatively associated with MMSE scores (r_s=-0.330, P<0.001) , higher TyG was associated with cognitive decline (β=-0.251, 95%CI=-0.421--0.081, P<0.05) , and taking the TyG index Q1 group as a reference, the risk of cognitive decline was significantly increased in both Q3 and Q4 groups (Q3: β=-0.453, 95%CI=-0.862--0.043, P<0.05; Q4: β=-0.503, 95%CI=-0.957--0.050, P<0.05) . TyG was not associated with MMSE in the normal weight group and the overweight and obese group.

The association between TyG and cognitive function is influenced by BMI, and an association between high TyG and cognitive decline has been found in the under weight older adults. Enhanced monitoring and control of insulin resistance in the under weight group may be helpful in preventing or delaying cognitive decline in older adults.

研究证据提示了三酰甘油葡萄糖指数（TyG）、BMI与认知功能均存在关联，但既往研究TyG与认知功能之间的关联时，较少考虑BMI在两者关系中的潜在影响，三者的关系仍不清楚，探究BMI对TyG与认知功能关联的影响对预防老年认知功能下降的精准干预具有重要意义。

探讨老年人BMI对TyG与认知功能关联的影响，为预防老年人认知功能下降提供参考依据。

于2019年7—8月采用多阶段整群抽样的方法，选取贵州省2个县（区）12个村的60岁及以上农村老年人作为研究对象，进行问卷调查、体格检查、认知功能检查和实验室检查。采用简易精神状态评价量表（MMSE）评估认知功能，利用实验室指标（三酰甘油、空腹血糖）计算TyG。采用Spearman秩相关分析及多重线性回归分析老年人TyG与认知功能的关联以及TyG与BMI分组的交互作用对认知功能的影响。

TyG与BMI对认知功能影响的交互作用分析结果显示：在未纳入混杂因素的模型1中，TyG与低BMI的交互项对老年人的MMSE评分存在影响（β=-0.515，95%CI=-0.809~-0.222，P<0.001）；在调整了相关协变量后的模型2中，TyG与低BMI的交互项对老年人MMSE评分仍存在影响（β=-0.351，95%CI=-0.601~-0.102，P<0.05）。进一步将BMI分层后进行多重线性回归分析，结果显示，在低体重组中，TyG与MMSE总分呈负相关（r_s=-0.330，P<0.001），较高的TyG与认知功能下降相关（β=-0.251，95%CI=-0.421~-0.081，P<0.05），以TyG的最低四分位数Q1组为参照，Q3、Q4组发生认知功能下降的风险升高（Q3：β=-0.453，95%CI=-0.862~-0.043，P<0.05；Q4：β=-0.503，95%CI=-0.957~-0.050，P<0.05）。在正常体重组和超重肥胖组没有发现TyG与MMSE的关联。

TyG与认知功能之间的关联受BMI的影响，在低体重老年人群中发现了高TyG与认知功能下降存在关联，加强对低体重人群中胰岛素抵抗的监测和控制，可能有助于预防或延缓老年人的认知功能下降。