Table 4.
Multivariate analysis resultsa for prenatal and current exposures to pesticides as predictors of clinical outcomes and heart variability in Ecuadorian school children.
| Prenatal exposure |
Current exposure |
||||
|---|---|---|---|---|---|
| Anthropometrics and clinical results | None [mean (95% CI)b or ORc; n = 26] | Paternal only [β (95% CI)b or OR (95% CI)c; n = 23] | Maternal [β (95% CI)b or OR (95% CI)c; n = 35] | No [mean (95% CI)b or ORc; n = 59] | Yes [β (95% CI)b or OR (95% CI)c; n = 22] |
| Weight (kg)b | 22.5 (19.9 to 25.1) | −1.4 (−3.0 to 0.2)* | −0.7 (−2.1 to 0.8) | 21.8 (19.2 to 24.3) | 0.8 (−0.6 to 2.3) |
| Height (cm)b | 110.9 (105.6 to 116.3) | 1.0 (−2.4 to 4.5) | 2.1 (−0.9 to 5.2) | 111.8 (106.6 to 116.9) | 2.2 (−0.8 to 5.2) |
| Stuntedc | 1 | 1.04 (0.30 to 3.62) | 0.56 (0.18 to 1.76) | 1 | 0.52 (0.17 to 1.64) |
| BMIb | 18.0 (16.4 to 19.7) | −1.4 (−2.4 to −0.4)** | −1.1 (−2.0 to −0.2)** | 17.3 (15.6 to 18.9) | 0.0 (−1.0 to 1.0) |
| Hematocritb | 44.1 (41.7 to 46.6) | 1.2 (−0.4 to 2.7) | −0.3 (−1.4 to 1.3) | 43.9 (41.6 to 46.1) | 1.1 (−0.2 to 2.4) |
| Blood pressureb | |||||
| Systolic | 90.0 (83.6 to 96.4) | 1.2 (−2.9 to 5.3) | 3.6 (−0.1 to 7.2)** | 91.9 (85.7 to 98.0) | 0.4 (−3.1 to 3.9) |
| Diastolic | 55.2 (48.6 to 61.9) | 1.5 (−2.7 to 5.8) | 2.9 (−1.0 to 6.6) | 56.5 (50.2 to 62.8) | 1.0 (−2.6 to 4.7) |
| Cardiac parametersb | |||||
| Heart rate (per min) | 70.9 (60.0 to 81.8) | −2.5 (−9.5 to 4.6) | 1.7 (−4.5 to 7.9) | 70.0 (59.9 to 80.1) | 3.8 (−1.9 to 9.6) |
| Heart rate variability [CVRR (%)] | 6.7 (3.7 to 9.7) | 1.0 (−1.0 to 2.9) | 0.0 (−1.6 to 1.7) | 6.8 (4.0 to 9.6) | 0.5 (−1.1 to 2.1) |
Each row shows two multivariate models (linear or logistic regressions) controlled for the child’s sex, age, race, BMI, and stunting. The models with weight, height, stunted, and BMI as outcomes are adjusted only for child’s sex, age, and race. The nonexposed group was used as the reference category. Mutual adjustment for prenatal and current exposures did not affect the results.
Normally distributed outcome, linear regression models were used.
Because the distribution was different from normal, logistic regression models were used to calculate the odds of being stunted, that is, Z-score below −2.
p < 0.10;
p ≤ 0.05.