Type 2 diabetes is more predictable in women than men by multiple anthropometric and biochemical measures

Abstract

Men and women are sexually dimorphic but whether common anthropometric and biochemical parameters predict type 2 diabetes (T2D) in different ways has not been well studied. Here we recruit 1579 participants in Hainan Province, China, and group them by sex. We compared the prediction power of common parameters of T2D in two sexes by association, regression, and Receiver Operating Characteristic (ROC) analysis. HbA1c is associated with FPG stronger in women than in men and the regression coefficient is higher, consistent with higher prediction power for T2D. Age, waist circumference, BMI, systolic and diastolic blood pressure, triglyceride levels, total cholesterol, LDL, HDL, fasting insulin, and proinsulin levels all predict T2D better in women. Except for diastolic blood pressure, all parameters associate or tend to associate with FPG stronger in women than in men. Except for diastolic blood pressure and fasting proinsulin, all parameters associate or tend to associate with HbA1c stronger in women than in men. Except for fasting proinsulin and HDL, the regression coefficients of all parameters with FPG and HbA1c were higher in women than in men. Together, by the above anthropometric and biochemical measures, T2D is more readily predicted in women than men, suggesting the importance of sex-based subgroup analysis in T2D research.

Figure 1

HbA1c, Age, waist circumference and BMI predict T2D better in women than in men. (A) Comparisons of association and linear regression of HbAc1 to FPG in women and men. Spearman’s association coefficient (ρ) indicates a tighter association in women. Linear regression coefficient (B) indicates a stronger influence of HbA1c on FPG levels in women. P values, regression constant can be found in Tables S2 and S3. (B) HbA1c is a better T2D predictor in women. Receiver operating characteristic (ROC) was plotted for HbA1c and T2D and area under the curve (AUC) is indicated. (C) ROC-AUC analysis shows no sex-difference of FPG in predicting T2D. (D,E) Age has a tighter association (Spearman’s ρ) with and stronger influence (steeper slope) on FPG and HbA1c in women. (F) Age is a stronger T2D predictor for women (higher ROC-AUC values). (G,H) Waist circumference has a tighter association with and stronger influence on FPG and HbA1c in women. (I) Waist circumference is a stronger T2D predictor for women as judged by ROC-AUC. (J,K) BMI has a tighter association with and stronger influence on FPG and HbA1c in women. (L) BMI predicts T2D in women but not men as judged by ROC-AUC analysis. AUC area under the curve, ns not significant, *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2

Blood pressure predicts T2D better in women than in men. (A,B) Systolic blood pressure has a tighter association (Spearman’s ρ) with and stronger influence (steeper slope) on FPG and HbA1c in women. P values, regression constant can be found in Tables S2 and S3. (C) Systolic blood pressure is a stronger T2D predictor for women as judged by ROC-AUC analysis. (D) Diastolic blood pressure tends to have a tighter association with FPG and a stronger influence on FPG in women than in men. (E) Diastolic blood pressure has a tighter association with HbA1c in men but tends to have a stronger influence on women. (F) Diastolic blood pressure is a weak T2D predictor for women and has neglectable prediction power for men. AUC area under the curve, ns not significant, *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 3

Triglyceride, total cholesterol, and LDL predict T2D better in women than in men. (A,B) Triglyceride has a tighter association (Spearman’s ρ) with FPG and HbA1c in women and affects FPG and HbA1c stronger (steeper slope) in women. P values, regression constant can be found in Tables S2 and S3. (C) Triglyceride is a stronger T2D predictor for women as judged by ROC-AUC analysis. (D,E) Total cholesterol levels tend to have tighter associations with FPG and HbA1c in women and affect FPG and HbA1c levels stronger in women. (F) Total cholesterol is a stronger T2D predictor for women. (G,H) LDL levels associate with FPG and HbA1c stronger in women and tend to affect FPG and HbA1c levels stronger in women. (I) LDL levels predict T2D in women but not men as judged by ROC-AUC analysis. (J,K) HDL has a very weak, negative correlation with FPG and HbA1c and shows no sex-difference. (L) HDL predicts T2D better in women than in men. AUC area under the curve, ns not significant, *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 4

Fasting insulin and proinsulin levels predict T2D better in women than in men. (A,B) Fasting insulin levels associate with and affect FPG and HbA1c stronger in women than in men as judge from Spearman’s ρ and regression coefficient B. P values, regression constant can be found in Tables S2 and S3. (C) Fasting insulin is a stronger T2D predictor for women as judged by ROC-AUC values. (D) Fasting proinsulin levels associate with and affect FPG and HbA1c stronger in women than in men. (E) Fasting proinsulin levels are weakly associated with HbA1c in women. (F) Fasting proinsulin is a stronger T2D predictor for women as judged by ROC-AUC values. AUC area under the curve, ns not significant, **P < 0.01, ***P < 0.001.

Figure 5

Sex-difference in the cutoffs of the predicting parameters individually or in combination. (A) The cutoffs of 3 parameters in men and 11 parameters in women having significant (P < 0.05) prediction power (ROC-AUC values > 0.6). A comprehensive comparison between men and women on the cutoffs, ROC-AUC values, sensitivity, specificity, and Youden Index is shown in Table S4. (B) A combination of age, waist, BMI, systolic and diastolic blood pressure, triglyceride, LDL, and HDL in T2D prediction by ROC analysis. (C) A combination of fasting insulin and proinsulin and HbA1c with parameters in (C) for T2D prediction. AUC area under the curve, ***P < 0.001.