Demographic characteristics of study participants

The present study recruited 2,092 participants with an average age of 73.21 ± 6.711 years. Among them, 971 were males and 1,121 were females. Among them, 1,746 had a history of T2DM and 346 had no history of T2DM. There was no significant difference in the history of T2DM between males and females. Differences in hypertension (P < 0.001) and gender (P < 0.001) were observed between groups with and without T2DM. In contrast, no statistically significant difference was found in age, smoking, alcohol consumption, or history of coronary heart diseases between groups with and without T2DM. A statistically significant difference was also found in body weight (P < 0.001), waist circumference (P < 0.001), neck circumference (P < 0.001), BMI (P < 0.001) and systolic blood pressure (P < 0.001) between groups with or without a history of T2DM and between groups of different genders. A significant difference was observed in laboratory tests, such as FPG (P < 0.001), TG (P < 0.001) and LAP (P < 0.001), between groups with and without a history of T2DM. The gender difference was also seen in these tests (FPG: P < 0.001; TG: P < 0.001; LAP: P < 0.001)(shown in Tables 1 and 2).

Screening of factors related to type 2 diabetes and gender differences of related factors

Spearman correlation analysis showed that body weight (P < 0.001), neck circumference (P < 0.001), systolic blood pressure (P < 0.001), FPG (P < 0.001), LAP (P < 0.001) and BMI (P < 0.001) were all significantly associated with T2DM. Detailed results are shown in Table 3. Stepwise regression analysis demonstrated that FPG (P < 0.01) and NC (P = 0.019) were associated with T2DM in the study population. However, only FPG (P < 0.01) and NC (P < 0.001) were significantly associated with the history of T2DM in the male population. In contrast, only FPG (P < 0.01) was significantly associated with the history of T2DM in the female population (Table 4).

Analysis of related indicators and risk factors for T2DM

FPG (P < 0.001) and LAP(P < 0.001) were significantly associated with the risk of T2DM before and after adjusting for age, sex, smoking, alcohol consumption, and other factors. This relationship between FPG (P < 0.001) and the risk of T2DM remained after adjusting associated indices of lipids and blood pressure, whereas no correlation was found between the risk of T2DM and NC (P > 0.05), nor between T2DM and LAP (P > 0.05).

With Q1 as the reference, before and after adjusting these parameters for age, gender, smoking, alcohol consumption and other factors, NC (before: OR = 2.17; after: OR = 2.5), FPG (before: OR = 1.94; after: OR = 1.94), LAP (before: OR = 1.97; after: OR = 1.97) were significantly associated with the risk of T2DM for patients with their LAP in Q4 (P < 0.001). After adjusting for lipid and blood pressure associated indices, LAP (P > 0.05) was not significantly associated with the risk of T2DM, but NC (P < 0.001) and FPG (P < 0.001) were.

For males, NC (P < 0.001), FPG (P < 0.001) and LAP (P < 0.001) were significantly associated with the risk of T2DM before and after adjusting for age, sex, smoking, alcohol consumption, and other factors. FPG (P < 0.001) was significantly related to the risk of T2DM after adjusting lipids and blood pressure-related indices, but NC and LAP (P > 0.05) were not. With Q1 as the reference, before and after adjusting these parameters for age, gender, smoking, alcohol consumption and other factors, NC (before: OR = 3.24; after: OR = 2.5), FPG (before: OR = 1.94; after: OR = 1.94), LAP (before: OR = 1.97; after: OR = 1.97) were significantly associated with the risk of T2DM for patients with their LAP in Q4 (P < 0.001). After adjusting for lipid and blood pressure associated indices, LAP (P > 0.05) and NC (P > 0.05) were not significantly associated with the risk of T2DM, but FPG (P < 0.001) was.

For females, NC (P < 0.001), FPG (P < 0.001), and LAP (P < 0.001) were significantly associated with the risk of T2DM before adjusting for age, sex, smoking, alcohol use, and only FPG (P < 0.001) and LAP (P < 0.001) were associated with the risk of T2DM after adjusting for these factors.

When Q1 was set as the reference, NC, FPG, and LAP before [NC (OR = 3.97), FPG (OR = 22.53), LAP (OR = 2.06)] and after [NC (OR = 4.04), FPG (OR = 22.8), LAP (OR = 2.08)] adjusting for age, sex, smoking, alcohol consumption and other factors were significantly associated with the risk of T2DM for patients with their LAP in Q4 (P < 0.001). NC (P < 0.05) and FPG (P < 0.001) were significantly associated with the risk of T2DM after adjusting for lipid and blood pressure-associated indices, whereas LAP (P > 0.05) was not. These suggest that there is a stronger correlation between the above factors and the risk of T2DM when the value of LAP is higher in either the overall population or the male or female population. In addition, LAP was influenced by several confounding factors (Table 5).

Confounding factor analysis of NC, FPG, and LAP

Logistic regression analysis was conducted to test possible factors that impact NC, FPG, and LAP, aiming to find confounding factors associated with LAP. Univariate analysis showed that NC [OR = 1.11(95%CI 1.07–1.15) P < 0.001], FPG [OR = 2.2 (95%CI 2.01–2.42) P < 0.001], LAP [OR = 1.01 (95%CI 1.007–1.012) P < 0.001] showed statistically significant difference. Multivariate Logistic regression analysis showed that FPG significantly impacted LAP of 3 groups [OR = 1.001 (95%CI 0.998–1.005) (overall), OR = 0.998 (95%CI 0.993–1.004)(male), OR = 1.003 (95%CI 0.99–1.008)(female)] after taking NC, FPG, LAP, BMI and TG into consideration. In the male population, FPG significantly influenced NC [OR = 1.05 (95%CI 0.97–1.13) P > 0.05], whereas NC influenced LAP [OR = 1.003 (95%CI 0.99–1.008) P > 0.05] in the female population (Table 6).

Prediction of T2DM using ROC curves of related factors

For the general population, FPG had the largest AUC (0.839, 0.812–0.866) and it showed the best predicting effect at the concentration of 6.4 mmol/L. LAP had the second largest AUC (0.613, 0.581–0.645) and the best predicting effect at 33.8 mmol/L. NC had an AUC of 0.6 (0.569 − 0.532) and the best prediction effect at 35.75 cm. For males, FPG had the best predicting effect (AUC = 0.835, 0.795–0.874) at 6.49 mmol/L, followed by LAP (AUC = 0.604, 0.577-652) at 37.95 mmol/L and NC (AUC = 0.597, 0.594–0.644) at 36.9 cm. For females, FPG was also the best predictor (AUC = 0.842, 0.805–0.878) at 6.42 mmol/L, followed by LAP (AUC = 0.617, 0.574–0.660) at 60.2 mmol/L (Table 7; Fig. 1).