Associations of obesity-related indices with mild cognitive impairment in adults 60 years and older with type 2 diabetes: a retrospective study

Study population

This retrospective study encompassed patients aged 60 years and older with T2D, who were admitted to Hebei General Hospital between January 2021 and November 2023 according to medical records. The diagnosis of T2D was established according to the criteria set forth by the American Diabetes Association (ElSayed et al., 2023). Exclusion criteria included: (1) patients diagnosed with type 1 diabetes mellitus; (2) patients lacking cognitive function assessments; (3) patients meeting the diagnostic criteria for dementia (Ismail et al., 2020); (4) patients experiencing cerebrovascular events, such as stroke, within three months prior to the study; (5) patients with other specific conditions that could influence cognitive function assessment, including anxiety, depression, brain injuries, or carbon monoxide poisoning. Ultimately, 587 elderly patients were included in our study. The authors did not have access to any information that could have identified individual participants either during or after the data collection process. The procedure for patient selection is illustrated in Fig. 1.

Ethical approval

This study followed the principles in the Declaration of Helsinki and was approved by the Ethical Committees of Hebei General Hospital (NO.2024-LW-109). The informed consent was waived because the study was a retrospective analysis and the data of all participants were anonymized.

Data collection

The characteristics of all patients were meticulously documented, encompassing age, body height (BH), body weight (BW), WC, hip circumference (HC), years of education, smoking and alcohol status, duration of T2D, anti-diabetic drugs, and history of stroke, hypertension, and coronary heart disease (CHD). Laboratory parameters including fasting plasma glucose (FPG), glycosylated hemoglobin (HbA1c), total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), blood urea nitrogen (BUN), serum creatinine (Scr), uric acid (UA), and estimated glomerular filtration rate (eGFR) were quantitatively analyzed through automated biochemical analyzer and glycosylated hemoglobin analyzer platforms in accordance with established clinical laboratory procedures.

Calculation of obesity-related indices

The obesity-related indices were calculated (Huang et al., 2022; Zhang et al., 2024; Ou et al., 2021) as follows.

1. Body mass index (BMI) = BW(kg)/BH²(m)

2. WHR = WC(cm)/HC(cm)

3. Waist-to-height ratio (WHtR) = WC(cm)/BH(cm)

4. Lipid accumulation product (LAP) = (WC(cm)–65)×TG(mmol/L) in males and

   LAP = (WC(cm)–58)×TG(mmol/L) in females

5. $\text{Body roundness index}\left(\mathrm{BRI}\right)=364.2-365.5\times \sqrt{1-{\left(\frac{\mathrm{WC}\left(\mathrm{m}\right)}{\pi \times \mathrm{BH}\left(\mathrm{m}\right)}\right)}^2}$

6. $\text{Conicity index}\left(\mathrm{CI}\right)=\frac{\mathrm{WC}\left(\mathrm{m}\right)}{0.109\times \sqrt{\frac{\mathrm{BW}\left(\mathrm{kg}\right)}{\mathrm{BH}\left(\mathrm{m}\right)}}}$

7. Visceral adiposity index (VAI) = $\left(\frac{\mathrm{WC}\left(\mathrm{cm}\right)}{39.68+\left(1.88\times \mathrm{BMI}\right)}\right)\times \left(\frac{\mathrm{TG}\left(\text{mmol}/\mathrm{L}\right)}{1.03}\right)\times \left(\frac{1.31}{\mathrm{HDL}-\mathrm{C}\left(\text{mmol}/\mathrm{L}\right)}\right)$ in males and

VAI = $\left(\frac{\mathrm{WC}\left(\mathrm{cm}\right)}{36.58+\left(1.89\times \mathrm{BMI}\right)}\right)\times \left(\frac{\mathrm{TG}\left(\text{mmol}/\mathrm{L}\right)}{0.81}\right)\times \left(\frac{152}{\mathrm{HDL}-\mathrm{C}\left(\text{mmol}/\mathrm{L}\right)}\right)$ in females

8. $\text{Body adiposity index}\left(\mathrm{BAI}\right)=\frac{\mathrm{HC}\left(\mathrm{cm}\right)}{\mathrm{BH}{\left(\mathrm{m}\right)}^{3/2}}-18$

9. $\text{Abdominal volume index}\left(\mathrm{AVI}\right)=\frac{2\times \mathrm{WC}{\left(\mathrm{cm}\right)}^2+0.7\times {\left(\mathrm{WC}\left(\mathrm{cm}\right)-\mathrm{HC}\left(\mathrm{cm}\right)\right)}^2}{1000}$

10. $\mathrm{A}\text{body shape index}\left(\text{ABSI}\right)=\frac{\mathrm{WC}\left(\mathrm{m}\right)}{\mathrm{BMI}{\left(\mathrm{kg}/{\mathrm{m}}^2\right)}^{2/3}\times \mathrm{BH}{\left(\mathrm{m}\right)}^{1/2}}$

11. TyG = ln [TG (mg/dl)×FPG (mg/dl)/2]

12. Cardiometabolic index (CMI) = TG/HDL-C×WHtR

Evaluation of cognitive function

The Montreal Cognitive Assessment (MoCA) was utilized to evaluate cognitive function in all participants. Cognitive impairment was evaluated using objective criteria based on MoCA scores: a score of ≤13 indicated impairment in individuals with no formal education, ≤19 for those with 1–6 years of education, and ≤24 for individuals with 7 or more years of education (Lu et al., 2011). According to the established criteria (Albert et al., 2011), patients enrolled in our study were categorized into either the normal cognitive group and MCI group.

Statistical analysis

Continuous variables adhering to a normal distribution were presented as mean (standard deviation) and analyzed using an independent samples t-test. For continuous variables that did not follow a normal distribution, the median (interquartile range) was utilized for description, with statistical significance evaluated via the Mann–Whitney U test. Categorical variables were summarized as frequency (percentage) and examined using chi-square tests. Logistic regression models were utilized to examine the association between obesity-related indices and MCI in elderly patients with T2D. Tests for trend were conducted to evaluate the dose-dependence relationships between obesity-related indices and MCI. Spearman correlation analysis was performed to assess the correlations between the MoCA scores and obesity-related indices. To further investigate potential non-linear relationships, restricted cubic splines were employed. The performance and predictive capability of obesity-related indices in identifying MCI among elderly patients with T2D were assessed using receiver operating characteristic (ROC) curves and the area under the curve (AUC). Statistical analyses were performed using SPSS version 26 (IBM Corporation, Armonk, NY) and the R programming environment (version 4.2.3, the R Foundation for Statistical Computing, Vienna, Austria). P < 0.05 was considered statistically significant.

Participants characteristics

A total of 587 patients with T2D (368 in the normal cognitive group and 219 in the MCI group) were included in our study. Table 1 shows the characteristics of patients in both groups. Patients with MCI were significantly older and had lower education levels than those in the normal cognitive group (P < 0.05). Patients in MCI group had lower frequencies of glucagon-like peptide-1 receptor (GLP-1R) agonists and sodium-dependent glucose transporters 2 (SGLT-2) inhibitors usage than those in the normal cognitive group (P < 0.05). In addition, compared with normal cognitive group, the MCI group presented higher FPG, HbA1c and TG, but lower HDL-C (P < 0.05).

Correlations between obesity-related indices and MoCA scores

As illustrated in Fig. 2, the TyG index (r = −0.22, P < 0.001), LAP (r = −0.20, P < 0.001), CMI (r = −0.19, P < 0.001), and VAI (r = −0.19, P < 0.001) exhibited significantly stronger negative correlations with MoCA scores compared to other obesity-related indices such as AVI, WHR, WHtR, BRI, and CI. Notably, There were no significant correlations between BAI, BMI, ABSI, and MoCA scores (P > 0.05).

Associations between obesity-related indices and MCI

Compared to T2D patients with normal cognition, those with MCI exhibited significantly higher values in BMI, WHR, WHtR, LAP, BRI, VAI, CI, AVI, TyG index, and CMI (P < 0.05) (Table 2). The results of logistic regression analyses are shown in Table 3. In the unadjusted model (Model 1), obesity-related indices, excluding WHR, BAI, and ABSI, were associated with MCI in elderly T2D patients (P < 0.05). After adjusting for age, education, hypertension, stroke, anti-diabetic drugs, and HbA1c (Model 3), elevated BMI, WHR, WHtR, LAP, BRI, CI, VAI, AVI, TyG index, and CMI were significantly associated with an increased risk of MCI in elderly T2D patients (all P < 0.05).

The levels of obesity-related indices, excluding CI, BAI, and ABSI, were positively associated with the risk of MCI in elderly T2D patients (Table 4). Relative to the lowest quartile, the odds ratios (ORs) for the highest quartile were as follows: BMI had an OR of 1.983 (95% confidence interval (CI) [1.187–3.312], P for trend = 0.004); WHR had an OR of 1.519 (95% CI [0.918–2.513], P for trend = 0.037); WHtR had an OR of 1.903 (95% CI [1.140–3.177], P for trend = 0.006); LAP had an OR of 5.743 (95% CI [3.282–10.050], P for trend < 0.001); BRI had an OR of 1.939 (95% CI [1.159–3.245], P for trend = 0.006); VAI had an OR of 5.456 (95% CI [3.125–9.443], P for trend < 0.001); AVI had an OR of 2.734 (95% CI [1.623–4.605], P for trend < 0.001); TyG index had an OR of 7.440 (95% CI [4.139–13.373], P for trend < 0.001); and CMI had an OR of 7.355 (95% CI [4.134–13.085], P for trend < 0.001).

After adjusting for age, education, hypertension, stroke, anti-diabetic drugs (GLP-1R agonists and SGLT-2 inhibitors), and HbA1c, there was a linear trend between BMI (P for non-linearity = 0.637), WHR (P for non-linearity = 0.430), WHtR (P for non-linearity = 0.452), BRI (P for non-linearity = 0.252), AVI (P for non-linearity = 0.944), and TyG index (P for non-linearity = 0.514) and the risk of MCI in elderly patients with T2D (Figs. 3A–3F). Figures 3G, 3H, and 3I show that there was a nonlinear association (S shaped relation) between LAP, VAI or CMI and risk of MCI in elderly patients with T2D (all P for non-linearity < 0.001). The risk of MCI increased when the level of LAP exceeded 36.11 after adjusting for age, education, hypertension, stroke, anti-diabetic drugs (GLP-1R agonists and SGLT-2 inhibitors) and HbA1c (Fig. 3G). The level of VAI (more than 1.91, Fig. 3H) and CMI (more than 0.65, Fig. 3I) also showed similar nonlinear patterns.

Figure 4 illustrates the the ROC curves and AUC values of the twelve obesity-related indices for identifying MCI among elderly patients with T2D. The AUC values for the various indices were as follows: CMI achieved the highest value of 0.682, with VAI at 0.679, TyG index at 0.673, LAP at 0.669, AVI at 0.580, WHtR and BRI both at 0.575, BMI at 0.560, CI at 0.556, WHR at 0.554, BAI at 0.547, and ABSI at 0.536. Detailed ROC curve statistics are in Table S1.