Baseline clinical characteristics of the newly-diagnosed participants are shown in Table 1. Patients were adequately distributed for sex. Smoking habits were highly prevailing in males. Women were more obese and showed a worse lipid profile; however, LDL cholesterol was far from being at target (< 100 mg/dl in primary prevention according to international guidelines) in both sexes. 25.3% of the study group showed a scarce metabolic control (HbA1c > 8%), with no sex difference in fasting blood glucose or HbA1c. Renal function was similar in men and women; 13.2% of patients were already in a CKD stage ≥ 3 according to eGFR. Retinopathy was present only in seven patients, all males.
Table 1
Clinical and biochemical characteristics of the study cohort at baseline.
| All 289 | Men 158 (54.7%) | Women 131 (45.3%) | p value |
Age (yrs) | 63.3 ± 11.0 | 62.7 ± 12.0 | 64.0 ± 11.0 | ns |
Ethnic group (caucasian/black/Asian; %) | 95/2/3 | 94/3/3 | 97/1/2 | ns |
BMI (Kg/m2) | 30.5 ± 6.0 | 29.5 ± 5.3 | 31.7 ± 6.5 | 0.0026 |
Smoking habits (yes/no/ex; %) | 11/51/38 | 15/38/47 | 5/69/26 | < 0.001 |
SBP (mmHg) | 141.0 ± 19.7 | 140.0 ± 19.4 | 142.2 ± 20.3 | ns |
DBP (mmHg) | 83.0 ± 11.0 | 83.1 ± 11.0 | 83.0 ± 11.1 | ns |
Fasting glucose (mg/dL) | 158 ± 57 | 158 ± 60 | 157 ± 54 | ns |
Hb1Ac (%) | 7.6 ± 1.7 | 7.6 ± 1.64 | 7.7 ± 1.8 | ns |
Patients with HbA1c > 8% (n; %) | 73; 25.3 | 43; 27.2 | 30; 22.9 | ns |
Total cholesterol (mg/dL) | 202 ± 43 | 192 ± 41 | 214 ± 44 | < 0.0001 |
HDL-cholesterol (mg/dL) | 48 ± 16 | 43 ± 11 | 55 ± 18 | < 0.0001 |
LDL-cholesterol (mg/dL) | 122 ± 42 | 114 ± 42 | 131 ± 41 | 0.0022 |
Triglycerides (mg/dL) | 169 ± 106 | 179 ± 115 | 157 ± 92 | ns |
ALT (IU/L) | 32 ± 22 | 34 ± 25 | 30 ± 19 | ns |
AST (IU/L) | 24 ± 13 | 26 ± 13 | 23 ± 12 | ns |
γGT (IU/L) | 50 ± 59 | 57 ± 73 | 43 ± 36 | ns |
Serum creatinine (mg/dL) | 0.95 ± 0.41 | 1.1 ± 0.45 | 0.77 ± 0.16 | < 0.0001 |
eGFR CKD-EPI (ml/min/1.73 m2) | 79.2 ± 21.0 | 77.3 ± 23.0 | 81.5 ± 18.0 | ns |
eGFR < 60 ml/min/1.73 m2 (n; %) | 38; 13.2 | 23; 14.6 | 15; 11.5 | ns |
Uric acid (mg/dL) | 6.0 ± 1.8 | 6.3 ± 1.8 | 5.6 ± 1.5 | ns |
Diabetic retinopathy (n; %) | 7; 2.4 | 7; 4.4 | 0; 0 | 0.022 |
Previous CV disease (n; %) | 44; 15.2 | 37; 23.4 | 7; 5.3 | < 0.0001 |
Heart failure (n; %) | 19; 6.6 | 12; 7.6 | 7; 5.3 | ns |
Atrial fibrillation (n; %) | 23; 8.0 | 15; 9.5 | 8; 6.1 | ns |
Any cancer (n; %) | 50; 17.5% | 23; 14.6% | 27; 21.1% | ns |
On the basis of clinical records, a previous CV event was confirmed in 15% of the patients, with a significant prevalence in the male sex. Prevalence of heart failure and atrial fibrillation was 7% and 8%, respectively.
Table 2 shows the initial distribution of the pharmacologic treatments. At baseline visit, 54% of patients did not assume any treatment for diabetes. Metformin (mean dose 1000 mg/die) was already on board in one-third of the patients, while 17.3% of them received secretagogues, alone or in combination with metformin; in these latter, HbA1c was significantly higher vs those not receiving drugs (8.1 ± 2.0 vs 7.3 ± 1.3%, p = 0.012). Sixty-two percent of the patients received anti-hypertensive treatment; despite the insufficient control of lipid levels, only 30% of the patients were on statin treatment.
Table 2
Pharmacologic treatment of the study cohort at baseline
| All 289 | Men 158 (54.7%) | Women 131 (45.3%) | p value |
No treatment (n; %) | 155; 53.6 | 81; 51.3 | 74; 56.5 | ns |
Metformin (n; %) | 84; 29.1 | 48; 30.4 | 36; 27.5 | ns |
Secretagogues (± metformin; n; %) | 50; 17.3 | 29; 18.4 | 21; 16.0 | ns |
Anti-hypertensive treatment (n; %) | 179; 61.9 | 103; 65.2 | 76; 58.0 | ns |
Statins (n; %) | 88; 30.4 | 50; 31.6 | 38; 29.0 | ns |
Follow up Mean duration of the follow up was 51.2 months (median: 49 months); during such observation, we registered 36 deaths (12.5% of the study population). The mortality rate in this cohort was 29.2‰/year. Table 3 shows the baseline phenotype of patients alive or deceased at the index date. Patients encountering exitus within four years from the diagnosis of T2D were older, had a remarkably lower eGFR, lower uric acid levels, and a higher prevalence of atrial fibrillation. No significant differences emerged in the distribution of the treatments for T2D and comorbidities, as well as in the mean HbA1c, even though in the group of deceased patients there was a significant greater percentage of patients with impaired glycaemic control (HbA1c > 8%) at baseline. After multivariate analysis, age at diagnosis remained as the only basal predictor of all-cause mortality.
Table 3
Baseline phenotype and treatments of patients deceased or alive over the follow up
| Deceased 36 (12.8%) | Alive 253 (87.2%) | p value |
Age (yrs) | 72.8 ± 9.0 | 61.9 ± 11.3 | < 0.0001 |
Sex (n; %) | F 14 (38.9%) M 22 (61.1%) | F 117 (46.2%) M 136 (53.8%) | ns |
BMI (Kg/m2) | 30.0 ± 5.8 | 30.6 ± 6.0 | ns |
SBP (mmHg) | 145.6 ± 19.8 | 140.3 ± 19.7 | ns |
DBP (mmHg) | 84.8 ± 11.3 | 82.8 ± 11.0 | ns |
Fasting glucose (mg/dL) | 156 ± 50 | 158 ± 59 | ns |
Hb1Ac (%) | 7.9 ± 1.3 | 7.6 ± 1.8 | ns |
Patients with HbA1c > 8% (n; %) | 15; 45.5 | 58; 25.5 | 0.0174 |
Total cholesterol (mg/dL) | 196 ± 40 | 203 ± 44 | ns |
HDL-cholesterol (mg/dL) | 51 ± 14 | 48 ± 16 | ns |
LDL-cholesterol (mg/dL) | 122 ± 53 | 122 ± 41 | ns |
Triglycerides (mg/dL) | 136 ± 75 | 173 ± 108 | ns |
ALT (IU/L) | 25 ± 12 | 33 ± 23 | ns |
AST (IU/L) | 21 ± 7 | 25 ± 13 | ns |
γGT (IU/L) | 44 ± 28 | 51 ± 62 | ns |
Serum creatinine (mg/dL) | 1.06 ± 0.6 | 0.93 ± 0.37 | ns |
eGFR CKD-EPI (ml/min/1.73 m2) | 70.6 ± 19.2 | 81.0 ± 20.0 | 0.0256 |
eGFR < 60 ml/min/1.73 m2 (n; %) | 8; 22.2 | 30; 11.9 | ns |
Uric acid (mg/dL) | 4.9 ± 1.1 | 6.2 ± 0.2 | 0.0291 |
Diabetic retinopathy (n; %) | 1; 2.8% | 6; 2.4 | ns |
Previous CV disease (n; %) | 6; 16.7 | 38; 15.0 | ns |
Heart failure (n; %) | 4; 11.1 | 15; 5.9 | ns |
Atrial fibrillation (n; %) | 6; 16.7 | 17; 6.7 | 0.0067 |
Any cancer (n; %) | 10; 27.8 | 40; 15.8 | ns |
No treatment (n; %) | 16; 44.4 | 139; 54.9 | ns |
Metformin (n; %) | 12; 33.3 | 72; 28.5 | ns |
Secretagogues (n; %) | 8; 22.2 | 42; 16.6 | ns |
Anti-hypertensive treatment (n; %) | 23; 63.9 | 156; 61.6 | ns |
Statins (n; %) | 14; 38.9 | 74; 29.2 | ns |
Role of the renal phenotype Glomerular function strongly influences the prognosis in T2D; even though eGFR was not an independent predictor of all cause-mortality in this cohort of patients with short T2D duration, we explored its behaviour over time by building up the slopes on the basis of serum creatinine measured every six months, and relating them to all-cause mortality. Data are shown in Fig. 1a; interestingly, ΔeGFR at 52 months was significantly higher in deceased individuals (-6.9 ± 3.8 vs + 2.5 ± 0.1 ml/min/1.73 m2, p < 0.001); moreover, eGFR slope in these subjects was markedly unstable, with ample oscillations over the observation period (SD: 6.3 [4.8–12.3] vs 5.6 [3.5–8.5] ml/min/1.73 m2); the slope of alive individuals remained virtually stable. In the multivariate analysis, ΔeGFR emerged as significant determinant of mortality (p = 0.0453) also when adjusted for confounders (sex, age, BMI, blood pressure, glycaemia, HbA1c, uric acid). We also checked the impact of metabolic control over time, but the mean HbA1c value registered during the whole follow up (mean 6.3 determinations/patient) did not differently marked deceased vs surviving individuals (Fig. 1b).
Phenotype and incident CV events Table 4 compares the baseline phenotype of people developing or not an acute CV event over the follow up; seven missed patients were excluded from this analysis. We registered 18 new CV events; 50% of them were first episodes, and 50% recurrent events. Only 3 events occurred in patients deceased during the follow-up, while the remaining were surely non-fatal events. Clinical characteristics marking such events were the male sex, the previous personal history of CV disease and retinopathy and, interestingly, a relative delay in the time of starting the antidiabetic treatment; in detail, macrovascular events occurred in a significantly higher percentage in patients not pharmacologically treated since the very beginning of the disease, or in those receiving sulphonylureas. As expected, multivariate analysis identified previous CV events as the only independent predictor of future CV events in such study cohort.
Table 4
Baseline phenotype of patients developing or not developing acute CV events over the follow up (seven missing patients excluded from the analysis)
| New CV events 18 (6.4%) | No new CV events 264 (93.6%) | p value |
Age (yrs) | 66.6 ± 8,0 | 63.0 ± 11,8 | ns |
Sex (n; %) | F 3; 16.6% M 15; 83.2% | F 122; 46.2% M 142; 53.8% | 0.0146 |
BMI (Kg/m2) | 28.7 ± 5.0 | 30.5 ± 5.9 | ns |
SBP (mmHg) | 142.7 ± 21.2 | 140.5 ± 19.7 | ns |
DBP (mmHg) | 81.4 ± 10.4 | 83.0 ± 11.0 | ns |
Fasting glucose (mg/dL) | 138 ± 51 | 159 ± 58 | ns |
Hb1Ac (%) | 7.25 ± 1.0 | 7.68 ± 1.75 | ns |
Total cholesterol (mg/dL) | 185 ± 29 | 203 ± 43 | ns |
HDL-cholesterol (mg/dL) | 43 ± 11 | 48 ± 14 | ns |
LDL-cholesterol (mg/dL) | 114 ± 27 | 123 ± 43 | ns |
Triglycerides (mg/dL) | 140 ± 49 | 170 ± 108 | ns |
ALT (IU/L) | 25 ± 9 | 32 ± 23 | ns |
AST (IU/L) | 23 ± 4 | 25 ± 13 | ns |
γGT (IU/L) | 41 ± 25 | 50 ± 60 | ns |
Serum creatinine (mg/dL) | 1.07 ± 0.3 | 0.94 ± 0.42 | ns |
eGFR CKD-EPI (ml/min/1.73 m2) | 71.8 ± 17.3 | 80.0 ± 21.2 | ns |
eGFR < 60 ml/min/1.73 m2 (n; %) | 3; 16.7 | 34; 12.9 | ns |
Uric acid (mg/dL) | 6.6 ± 1.2 | 5.9 ± 1.8 | ns |
Previous CV disease (n; %) | 9; 50.0 | 33; 12.5 | < 0.0001 |
Heart failure (n; %) | 3; 16.7 | 15; 5.7 | ns |
Atrial fibrillation (n; %) | 1; 5.6 | 22; 8.3 | ns |
Any cancer (n; %) | 5; 27.8 | 45; 17.0 | ns |
Diabetic retinopathy (n; %) | 3; 16.7 | 4; 1.5 | < 0.0002 |
No treatment (n; %) | 8; 44.4 | 143; 54.2 | 0.0400 |
Metformin (n; %) | 3; 16.7 | 79; 29.9 |
Secretagogues (n; %) | 7; 38.9 | 42; 15.9 |
Anti-hypertensive treatment (n; %) | 13; 72.2 | 160; 60.1 | ns |
Statins (n; %) | 7; 41.1% | 76; 29.0% | ns |
Table 5 shows the prevalence of macrovascular complications and retinopathy at baseline and at the end of the observation period, according to baseline treatment for T2D. Early use of metformin was coupled, even in this small cohort, to less incidence of CV events over the follow up. Of note, no retinopathy was present at diagnosis, neither developed at follow up (n = 13), in patients starting an early treatment with metformin.
Table 5
Patients with CVD, retinopathy and CKD stage ≥ 3 at baseline and at the end of the follow up according to treatment of T2D at baseline (seven missing patients excluded from the analysis)
| All | No treatment 148 (53.6%) | Metformin 84 (29.1%) | Secretagogues 50 (17.3%) | p value |
Baseline CV disease (n; %) | 44 | 26; 17.6 | 9; 10.7 | 9; 18 | ns |
Follow up CV disease (n; %) | 18 | 8; 5.4 | 3; 3.6 | 7; 14.0 | 0.04 |
Baseline any retinopathy (n; %) | 7 | 3; 2.0 | 0; 0 | 4; 8.0 | 0.015 |
Follow up any retinopathy (n; %) | 13 | 7; 4.7 | 0; 0 | 6; 12 | 0.007 |
Baseline CKD stage ≥ 3 | 38 | 21; 14.2 | 9; 10.7 | 8; 16.0 | ns |
Follow up CKD stage ≥ 3 | 17 | 9; 6.1 | 2; 2.4 | 6; 12 | ns |