A1c Variability as an Independent Correlate of CVD in T2DM
A1c Variability as an Independent Correlate of CVD in T2DM
As previously reported, participants included in this analysis, i.e. those having multiple (3–5) HbA1c values, had a median (IQR) age and duration of diabetes at enrolment of 68 (61–74) and 14 (7–23) years, respectively, and a male/female ratio of 57/43. Likely due to longer disease duration, these subjects showed a worse CVD risk profile and a higher prevalence of any CVD event and were more frequently on glucose-, lipid- and BP-lowering drug treatment, than those excluded from the analysis due to unavailability of serial HbA1c measurements. HbA1c-MEAN of participants was 7.57% (6.86–8.38), HbA1c-SD was 0.46 (0.29–0.74) and adj-HbA1c-SD was 0.40 (0.25–0.65). Both variability measures, i.e. HbA1c-SD and adj-HbA1c-SD, were closely related to HbA1c-MEAN and between each other.
One or more major acute CVD events were adjudicated in 2,133 of the 8,290 patients analyzed (25.7%). Clinical characteristics of subjects with or without prior CVD event(s) are shown in Table 1 and Table 2. As compared with individuals without CVD, patients with prior event(s) were older, predominantly male, and had a longer diabetes duration, a worse metabolic control, a significantly higher rate of insulin treatment, and a slightly but significantly higher number of HbA1c measures (4.63±0.70 versus 4.48±0.78, P<0.0001). Concerning the other CVD risk factors, these patients were more frequently treated for dyslipidemia and/or hypertension and had higher triglycerides and lower HDL cholesterol, but also lower total, LDL and non-HDL cholesterol and (marginally) BP levels. Finally, subjects with prior CVD event(s) showed a higher prevalence of DR, albuminuria, reduced eGFR and CKD.
As previously reported, increasing HbA1c-MEAN and HbA1c-SD (or HbA1c-CV) were associated with longer and shorter diabetes duration, respectively, whereas both measures correlated with a more adverse CVD risk profile, more severe complications, and more intensive treatment, with a few differences.
Values for current HbA1c, HbA1c-MEAN, HbA1c-SD, HbA1c-CV and Adj-HbA1c-SD in subjects with or without any major acute CVD event are given in Table 1 (total) and 3 (by vascular bed, i.e. coronary, carotid and lower limb). As compared with men, women showed significantly higher values of current HbA1c (7.75±1.40 versus 7.58±1.32, P<0.0001), HbA1c-MEAN (7.80±1.24 versus 7.62±1.17, P<0.0001) and lower levels of HbA1c-SD (0.59±0.49 versus 0.60±0.47, P=0.049), HbA1c-CV (7.39±5.50 versus 7.65±5.38, P=0.001), and, non-significantly, Adj-HbA1c-SD (0.52±0.43 versus 0.52±0.40, P=0.060), as compared with men. When stratified by presence or absence of any CVD event, total and by vascular bed, data in men and women were almost identical to those from the whole cohort presented in Table 3 (Additional file 2: Tables S1-S2). Current HbA1c and HbA1c-MEAN were higher in patients with history of any CVD, either total or by vascular bed, whereas HbA1c-SD and Adj-HbA1c-SD were significantly increased in subjects with AMI, any lower limb vascular event and ulceration/gangrene and HbA1c-CV only in those with this latter event. Current HbA1c and HbA1c-MEAN values, but not measures of HbA1c variability, increased with the number of major acute CVD events and with the number of vascular beds affected (P<0.0001). However, while these values did not change significantly with increasing number of events (from 1 to ≥5), they were higher in subjects with 2 or 3 vascular beds affected than in those with involvement of only one site.
We used logistic models as multivariate regression analyses because of several potential confounding factors for the association between HbA1c variability and prevalence of macrovascular complications. Even after adjusting for CVD risk factors and complications, HbA1c-MEAN, but not HbA1c-SD (and independent of it in Models 2 and 3) was a significant correlate of any CVD (Table 4).
When CVD was examined by vascular bed, similar findings were observed in subjects with versus those without any coronary event (AMI and/or coronary revascularization) or any cerebrovascular event (stroke and/or carotid revascularization) in all 3 models (Table 5 and Table 6) and in those with versus without AMI in Model 3 only (Table 5). Conversely, none of these measures were associated with stroke (Table 6), whereas both correlated with any lower limb vascular event (ulceration/gangrene and/or lower limb revascularization; Hb-A1c-SD only in Model 2) and HbA1c-SD alone with ulceration/gangrene (Table 7).
In all models, expressing HbA1c variability as HbA1c-CV or adj-HbA1c-SD instead of HbA1c-SD did not change the results.
Results
Patients' Characteristics
As previously reported, participants included in this analysis, i.e. those having multiple (3–5) HbA1c values, had a median (IQR) age and duration of diabetes at enrolment of 68 (61–74) and 14 (7–23) years, respectively, and a male/female ratio of 57/43. Likely due to longer disease duration, these subjects showed a worse CVD risk profile and a higher prevalence of any CVD event and were more frequently on glucose-, lipid- and BP-lowering drug treatment, than those excluded from the analysis due to unavailability of serial HbA1c measurements. HbA1c-MEAN of participants was 7.57% (6.86–8.38), HbA1c-SD was 0.46 (0.29–0.74) and adj-HbA1c-SD was 0.40 (0.25–0.65). Both variability measures, i.e. HbA1c-SD and adj-HbA1c-SD, were closely related to HbA1c-MEAN and between each other.
One or more major acute CVD events were adjudicated in 2,133 of the 8,290 patients analyzed (25.7%). Clinical characteristics of subjects with or without prior CVD event(s) are shown in Table 1 and Table 2. As compared with individuals without CVD, patients with prior event(s) were older, predominantly male, and had a longer diabetes duration, a worse metabolic control, a significantly higher rate of insulin treatment, and a slightly but significantly higher number of HbA1c measures (4.63±0.70 versus 4.48±0.78, P<0.0001). Concerning the other CVD risk factors, these patients were more frequently treated for dyslipidemia and/or hypertension and had higher triglycerides and lower HDL cholesterol, but also lower total, LDL and non-HDL cholesterol and (marginally) BP levels. Finally, subjects with prior CVD event(s) showed a higher prevalence of DR, albuminuria, reduced eGFR and CKD.
Average HbA1c and HbA1c Variability
As previously reported, increasing HbA1c-MEAN and HbA1c-SD (or HbA1c-CV) were associated with longer and shorter diabetes duration, respectively, whereas both measures correlated with a more adverse CVD risk profile, more severe complications, and more intensive treatment, with a few differences.
Values for current HbA1c, HbA1c-MEAN, HbA1c-SD, HbA1c-CV and Adj-HbA1c-SD in subjects with or without any major acute CVD event are given in Table 1 (total) and 3 (by vascular bed, i.e. coronary, carotid and lower limb). As compared with men, women showed significantly higher values of current HbA1c (7.75±1.40 versus 7.58±1.32, P<0.0001), HbA1c-MEAN (7.80±1.24 versus 7.62±1.17, P<0.0001) and lower levels of HbA1c-SD (0.59±0.49 versus 0.60±0.47, P=0.049), HbA1c-CV (7.39±5.50 versus 7.65±5.38, P=0.001), and, non-significantly, Adj-HbA1c-SD (0.52±0.43 versus 0.52±0.40, P=0.060), as compared with men. When stratified by presence or absence of any CVD event, total and by vascular bed, data in men and women were almost identical to those from the whole cohort presented in Table 3 (Additional file 2: Tables S1-S2). Current HbA1c and HbA1c-MEAN were higher in patients with history of any CVD, either total or by vascular bed, whereas HbA1c-SD and Adj-HbA1c-SD were significantly increased in subjects with AMI, any lower limb vascular event and ulceration/gangrene and HbA1c-CV only in those with this latter event. Current HbA1c and HbA1c-MEAN values, but not measures of HbA1c variability, increased with the number of major acute CVD events and with the number of vascular beds affected (P<0.0001). However, while these values did not change significantly with increasing number of events (from 1 to ≥5), they were higher in subjects with 2 or 3 vascular beds affected than in those with involvement of only one site.
Multiple Logistic Regression Models
We used logistic models as multivariate regression analyses because of several potential confounding factors for the association between HbA1c variability and prevalence of macrovascular complications. Even after adjusting for CVD risk factors and complications, HbA1c-MEAN, but not HbA1c-SD (and independent of it in Models 2 and 3) was a significant correlate of any CVD (Table 4).
When CVD was examined by vascular bed, similar findings were observed in subjects with versus those without any coronary event (AMI and/or coronary revascularization) or any cerebrovascular event (stroke and/or carotid revascularization) in all 3 models (Table 5 and Table 6) and in those with versus without AMI in Model 3 only (Table 5). Conversely, none of these measures were associated with stroke (Table 6), whereas both correlated with any lower limb vascular event (ulceration/gangrene and/or lower limb revascularization; Hb-A1c-SD only in Model 2) and HbA1c-SD alone with ulceration/gangrene (Table 7).
In all models, expressing HbA1c variability as HbA1c-CV or adj-HbA1c-SD instead of HbA1c-SD did not change the results.
Source...