Table 4.
Effect of T2D on each neuropsychological test in fully adjusted stepwise linear regression analysis: un-standardized betas (B) ± standard error (SE) and level of significance (p).
| Independent Variables: | B ± SE | p |
|---|---|---|
| FCSRT ITR (immediate total recall) * | 1.09 ± 0.41 | 0.01 |
| FCSRT DTR (Delayed Total Recall) * | 0.70 ± 0.26 | 0.01 |
| FCSRT ISC (Index of Sensitivity of Cueing) * | 0.06 ± 0.02 | 0.01 |
| Prose Memory Test ^ | 3.00 ± 0.63 | 0.00 |
| Rey–Osterrieth Complex Figure B (delayed recall) ^ | 5.49 ± 1.32 | 0.00 |
| Fluency for semantic categories ^ | 3.99 ± 1.32 | 0.00 |
| Animal fluency ^ | 2.75 ± 1.03 | 0.01 |
| Noun Naming (CAGI) * | 1.29 ± 0.48 | 0.01 |
Coefficients were estimated with ordinary least squares linear regression for normally distributed variables (*) and with quantile (median) linear regression for non-normally distributed variables (^). Each model is adjusted for gender, age, education, BMI, and clinical complications of diabetes (hypertension, neuropathy, nephropathy, chronic renal failure, retinopathy, supra-aortic trunks arterial disease, obliterative arteriopathy of the lower limbs, and myocardial ischemia).