Table 3. Baseline characteristics of the population associated with conversion to new onset diabetes and with prediabetes reversion, survival approach—Multivariate Cox analysis.
Multivariate Cox analysis | ||||
---|---|---|---|---|
Before selection | After stepwise selection† | |||
HR [95% CI] | p-value | HR [95% CI] | p-value | |
Study of conversion to new onset diabetes (135 events/383) |
||||
Age (+1 SD) | 0.93 [0.77; 1.12] | 0.43 | - | - |
Sex (Women/men) | 0.90 [0.60; 1.34] | 0.59 | - | - |
Diabetes risk score (+1 SD) | 1.35 [1.08; 1.68]** | 0.0081 | 1.31 [1.07; 1.60]** | 0.0089 |
Hypertension (y/n) | 1.00 [0.67; 1.47] | 0.98 | - | - |
Statin therapy (y/n) | 0.92 [0.61; 1.40] | 0.71 | - | - |
Fasting plasma glucose (+1 SD) | 1.47 [1.23; 1.75]*** | <0.0001 | 1.46 [1.23; 1.74]*** | <0.0001 |
HbA1c (+1 SD) | 1.54 [1.26; 1.88]*** | <0.0001 | 1.52 [1.25; 1.84]*** | <0.0001 |
Fatty Liver Index (+1 SD) | 1.28 [1.03; 1.58]* | 0.024 | 1.31 [1.07; 1.61]** | 0.008 |
Before selection | After stepwise selection† | |||
Study of prediabetes reversion (265 events/383) | HR [95% CI] | p-value | HR [95% CI] | p-value |
Age (+1 SD) | 0.93 [0.81; 1.06] | 0.29 | - | - |
Sex (Women/men) | 1.22 [0.91; 1.65] | 0.19 | - | - |
Diabetes risk score (+1 SD) | 0.88 [0.76; 1.03] | 0.11 | - | - |
Hypertension (y/n) | 1.13 [0.85; 1.51] | 0.40 | - | - |
Statin therapy (y/n) | 1.02 [0.74; 1.41] | 0.89 | - | - |
Fasting plasma glucose (+1 SD) | 0.75 [0.66; 0.86]*** | <0.0001 | 0.75 [0.66; 0.86]*** | <0.0001 |
HbA1c (+1 SD) | 0.79 [0.68; 0.90]*** | 0.00063 | 0.79 [0.69; 0.90]*** | 0.0003 |
Fatty Liver Index (+1 SD) | 0.89 [0.77; 1.03] | 0.11 | 0.85 [0.75; 0.96]** | 0.0077 |
†Considered candidates for multivariate adjusted Cox model were defined as follows: baseline fasting plasma glucose and Fatty liver index were “forced” in the model; age, sex, diabetes risk score, hypertension, statin therapy, HbA1c were candidates. Multivariate Fractional Polynomial approach was also considered, but no polynomial transformation was proposed for a threshold selection value = 0.20
* p <0.05;
** p < 0.01;
*** p < 0.001