Skip to main content
. 2020 Aug 13;8:319. doi: 10.3389/fpubh.2020.00319

Table 5.

Multiple linear regression for Symphysis-fundal height (SFH) Z-scores, with (A) mean arterial pressure (MAP) and hemoglobin (grams/dL) and (B) diastolic blood pressure (DBP) and hematocrit (%).

A. SFH Z-scorea,c Non-standardized Coef. ± SE P 95% CI Standardized β-coef. Overall model
History of displacement −1.01 ± 0.31 0.001 −1.72, −0.48 −0.36 P = 0.0010 Adj. R2 = 0.262
MAP, centileb 0.49 ± 0.20 0.017 0.09, 0.89 0.30
Hemoglobin, grams/dL 0.36 ± 0.21 0.096 −0.07, 0.79 0.21
Taking iron supplements (0 = no, 1 = yes) 0.54 ± 0.29 0.073 −0.05, 1.39 0.22
Constant −3.40 ± 0.62 <0.0001 −4.65, −2.14
B. SFH Z-scorea, c Non-standardized Coef. ± SE P 95% CI Standardized β-coef. Overall model
History of displacement −1.09 ± 0.31 0.001 −1.72, −0.48 −0.41 P < 0.0001 Adj. R2 = 0.379
Taking iron supplements (0 = no, 1 = yes) 0.59 ± 0.28 0.041 0.02, 1.15 0.24
Hematocrit, centile 0.47 ± 0.18 0.014 0.10, 0.84 0.29
Diastolic blood pressure, centileb 0.91 ± 0.25 0.001 0.39, 1.43 0.41
Constant −4.37 ± 0.65 <0.0001 −5.68, −3.06
a

Variables that were included but were not retained in the final models (P > 0.10): # sleep hours (score), eating nuts/avocado at least once/week (0 = no, 1 = yes), history of hypertensive disorders of pregnancy (0 = no, 1 = yes).

Model A: n = 52. VIF = 1.02. Condition number: 10.63.

Model B: n = 50. VIF = 1.03. Condition number: 12.41.

b

25th and 75th centiles of MAP, DBP, hemoglobin and hematocrit were calculated from our population and used instead of continuous variables in order correct for non-linear associations: MAP: <73, between 73 and 83 and >83 mmHg, DBP: <60, between 60 and 70 and >70 mmHg; hemoglobin: <12, ≥12 and ≤13.7, >13.7 grams/dL; hematocrit: <35.7, ≥35.7 and ≤41.2, >41.2%.

c

After 50 bootstrap replications of linear regression models we observed that percentile- and bias-corrected confidence intervals were similar (Table S3).