To create a genetic model associated with hypothyroidism as a risk factor for preeclampsia.Design and method:
For 130 pregnant women with preeclampsia and 100 women with normal pregnancies we evaluated nine variables (MTHFR-C677T, M235T-AGT, A1166C-AGTR2, Ser447X-LPL, Asp287Glu-eNOS, D1-C/T, D2-X, VEGF-C936T, I/D-ACE), and a multiple logistic regression multivariate analysis model was constructed.Results:
The result evidenced that the multivariate analysis model was validated by a B value and a statistically significant p < 0.001. The MTHFR-C677T, D1-C/T and D2-Thr22Ala mutations showed significant p value levels of 0.003, 0.001 and < 0.001. The Ser447X-LPL and Asp287Glu-eNOS mutations showed marginal significant p value levels, and for the M235T-AGT, A2350C-AGTR2, C936T-VEGF and I/D-ACE genetic variations, a p-value of 0.211, 0.229, 0.993 and 0.243, respectively, was obtained. Based on these results, we built a prediction parameter called preeclampsia risk score, with the formula: Score = −1.974*C6772–1.479*D1–1.335*D2. The risk score was normalized by summing all values calculated with K = −4.78. Thus, the score ranged in the interval: [0–4.79], negative values being eliminated. The reliability of the new prediction score was assessed by a ROC analysis. The area under the curve of 0.701 was calculated (95%CI = 0.640–0.756), the newly built parameter being validated as a highly reliable predictor of preeclampsia by a significance level p < 0.001. Thus, parameter levels over 1.34 (cut-off value) presented a Se = 3.2 (0.9–8.0) and a Sp = 100 (97.2–100) for association with preeclampsia. The analysis revealed that score values below 1.34 could be classified as high risk, between 1.34 and 3.45 as medium risk, and beyond this value as low risk. To use the newly constructed score, an interface was created which allows users to introduce the position status (1 = homozygous mutation, 0 = normal or heterozygous mutation) for each of the three mutations included in the previous score.Conclusions:
Of all the variables included in the model, MTHFR-C677T, D1-C/T and D2-Thr22Ala were independent predictors of preeclampsia. The preeclampsia risk score is calculated automatically, using the created interface, as a continuous value and is interpreted semiquantitatively in accordance with the values set.