Examination of the relationship between different circulating progenitor or mature endothelial cells and the risk of AD/dementia by using Cox proportional hazards regression models after adjusting for covariates
Circulating cells | Number/mL (%)Median [Q1, Q3] | ADHR (95% CI) | P value | All-cause dementiaHR (95% CI) | P value |
---|---|---|---|---|---|
CD34+CD133+ | 2.61 × 104 [1.65 × 104, 4.39 × 104]0.031 [0.020, 0.049] | 0.64 (0.44, 0.94)n = 1,340 | 0.02* | 0.63 (0.45, 0.87)n = 1,362 | 0.006* |
CD34+CD133– | 2.13 × 104 [1.38 × 104, 3.19 × 104]0.025 [0.017, 0.038] | 0.79 (0.49, 1.27)n = 1,340 | 0.33 | 0.74 (0.50, 1.10)n = 1,362 | 0.14 |
CD34–CD133+ | 3.06 × 104 [2.33 × 104, 4.15 × 104]0.037 [0.028, 0.047] | 0.56 (0.28, 1.14)n = 1,340 | 0.11 | 0.81 (0.46, 1.43)n = 1,362 | 0.47 |
CD34–CD133– | 8.59 × 107 [6.91 × 107, 10.56 × 107]99.90 [99.90, 99.90] | 342.92 (0.62, 1.91 × 105)n = 1,340 | 0.07 | 62.05 (0.37, 1.03 × 104)n = 1,362 | 0.11 |
CD34+ | 6.36 × 104 [4.44 × 104, 9.51 × 104]0.076 [0.053, 0.110] | 0.62 (0.36, 1.04)n = 1,415 | 0.07 | 0.61 (0.39, 0.96)n = 1,436 | 0.03* |
CD34+/KDR+ | 3.38 × 106 [1.96 × 106, 5.71 × 106]3.96 [2.33, 6.58] | 0.91 (0.65, 1.28)n = 1,416 | 0.59 | 0.91 (0.68, 1.23)n = 1,437 | 0.55 |
CD31+/CD45– | 5.14 × 105 [2.93 × 105, 8.78 × 105]0.60 [0.36, 1.01] | 0.85 (0.62, 1.17)n = 1,431 | 0.32 | 0.85 (0.65, 1.11)n = 1,453 | 0.24 |
CD31+ | 1.73 × 107 [1.19 × 107, 2.42 × 107]20.24 [15.12, 26.00] | 0.86 (0.43, 1.70)n = 1,400 | 0.65 | 0.74 (0.42, 1.31)n = 1,421 | 0.30 |
CD31– | 3.11 × 107 [2.35 × 107, 4.02 × 107]39.39 [29.98, 47.93] | 1.46 (0.69, 3.08)n = 1,400 | 0.32 | 1.64 (0.85, 3.16)n = 1,421 | 0.14 |
CD31+DIM | 3.36 × 107 [2.38 × 107, 4.43 × 107]39.03 [31.99, 46.82] | 0.73 (0.32, 1.66)n = 1,400 | 0.45 | 0.73 (0.36, 1.48)n = 1,421 | 0.38 |
CD31+lymphoid | 6.57 × 107 [5.38 × 107, 8.22 × 107]79.76 [74.00, 84.88] | 1.91 (0.16, 22.65)n = 1,400 | 0.61 | 2.76 (0.32, 24.12)n = 1,421 | 0.36 |
The concentration (cell number/mL) and the proportion (%) of different circulating endothelial cell (EPCs) and EMCs are illustrated in the 2nd column. Cox proportional hazards regression models were used to study the relationship between the proportions (%) of different subtypes of EPCs and EMCs (log-transformed) and the risk of AD or all-cause dementia after adjusting for age, sex, years of education, APOE ε4, and vascular diseases. HR with 95% CI with P values is shown. * P value significant < 0.05; CI: confidence interval; HR: hazard ratio; Q: quartile