Maternal vitamin D status and small-for-gestational-age offspring in women at high risk for preeclampsia

Abstract

Objective: To examine the association between second-trimester maternal serum 25-hydroxyvitamin D concentrations and risk of small for gestational age (SGA) in singleton live births.

Methods: We assayed serum samples at 12-26 weeks of gestation for 25-hydroxyvitamin D in a sample of participants in a multicenter clinical trial of low-dose aspirin for the prevention of preeclampsia in high-risk women (n=792). Multivariable log-binomial regression models were used to assess the association between 25-hydroxyvitamin D and risk of SGA (birth weight less than the 10 percentile for gestational age) after adjustment for confounders including maternal prepregnancy obesity, race, treatment allocation, and risk group.

Results: Thirteen percent of neonates were SGA at birth. Mean (standard deviation) 25-hydroxyvitamin D concentrations were lower in women who delivered SGA (57.9 [29.9] nmol/L) compared with non-SGA neonates (64.8 [29.3] nmol/L, P=.028). In adjusted models, 25-hydroxyvitamin D concentrations of 50-74 nmol/L and 75 nmol/L or greater compared with less than 30 nmol/L were associated with 43% (95% confidence interval [CI] 0.33-0.99) and 54% (95% CI 0.24-0.87) reductions in risk of SGA, respectively. Race and maternal obesity each modified this association. White women with 25-hydroxyvitamin D 50 nmol/L or greater compared with less than 50 nmol/L had a 68% reduction in SGA risk (adjusted risk ratio 0.32, 95% CI 0.17-0.63) and nonobese women with 25-hydroxyvitamin D 50 nmol/L or greater compared with less than 50 nmol/L had a 50% reduction in SGA risk (adjusted risk ratio 0.50, 95% CI 0.31-0.82). There was no association between 25-hydroxyvitamin D and risk of SGA in black or obese mothers.

Conclusion: Maternal vitamin D status in the second trimester is associated with risk of SGA among all women and in the subgroups of white and nonobese women.

Level of evidence: II.

Figures

Figure 1

Adjusted risk (solid line) of small-for-gestational-age birth by maternal serum 25-hydroxyvitamin D (25(OH)D) at 12–26 weeks of gestation in 792 women at high risk for preeclampsia (P=0.02). Gray shading is 95% confidence intervals. 25(OH)D was modeled by restricted cubic splines with three knots (28.8, 60.0, and 103.3) at percentiles 10%, 50%, and 90% in a log-binomial model adjusted for latitude, obesity status, race, treatment group, and risk group. 25(OH)D values less than 15 nmol/L (n=9) and more than 120 nmol/L (n=34) were omitted from the figure.

Figure 2

Adjusted risk (solid line) of small-for-gestational-age birth by maternal serum 25-hydroxyvitamin D (25(OH)D) at 12 – 26 weeks of gestation in (A) 298 white women (P less than 0.01, includes American Indian/Alaskan, n=2), (B) 494 black women (P=0.09), (C) 444 non-obese women (P=0.01, body mass index (BMI) less than 30 kg/m2), and (D) 348 obese women (P=0.39; BMI greater than 30 kg/m2) all at high risk for preeclampsia. Gray shading is 95% confidence intervals. 25(OH)D was modeled by restricted cubic splines with three knots (28.8, 60.0, and 103.3) at percentiles 10%, 50%, and 90% in a log-binomial model adjusted for latitude, obesity status, race, treatment group, and risk group. 25(OH)D values less than 15 nmol/L (n=9) and greater than 120 nmol/L (n=34) were omitted from the figure.

Figure 2

Adjusted risk (solid line) of small-for-gestational-age birth by maternal serum 25-hydroxyvitamin D (25(OH)D) at 12 – 26 weeks of gestation in (A) 298 white women (P less than 0.01, includes American Indian/Alaskan, n=2), (B) 494 black women (P=0.09), (C) 444 non-obese women (P=0.01, body mass index (BMI) less than 30 kg/m2), and (D) 348 obese women (P=0.39; BMI greater than 30 kg/m2) all at high risk for preeclampsia. Gray shading is 95% confidence intervals. 25(OH)D was modeled by restricted cubic splines with three knots (28.8, 60.0, and 103.3) at percentiles 10%, 50%, and 90% in a log-binomial model adjusted for latitude, obesity status, race, treatment group, and risk group. 25(OH)D values less than 15 nmol/L (n=9) and greater than 120 nmol/L (n=34) were omitted from the figure.

Figure 2

Adjusted risk (solid line) of small-for-gestational-age birth by maternal serum 25-hydroxyvitamin D (25(OH)D) at 12 – 26 weeks of gestation in (A) 298 white women (P less than 0.01, includes American Indian/Alaskan, n=2), (B) 494 black women (P=0.09), (C) 444 non-obese women (P=0.01, body mass index (BMI) less than 30 kg/m2), and (D) 348 obese women (P=0.39; BMI greater than 30 kg/m2) all at high risk for preeclampsia. Gray shading is 95% confidence intervals. 25(OH)D was modeled by restricted cubic splines with three knots (28.8, 60.0, and 103.3) at percentiles 10%, 50%, and 90% in a log-binomial model adjusted for latitude, obesity status, race, treatment group, and risk group. 25(OH)D values less than 15 nmol/L (n=9) and greater than 120 nmol/L (n=34) were omitted from the figure.

Figure 2

Adjusted risk (solid line) of small-for-gestational-age birth by maternal serum 25-hydroxyvitamin D (25(OH)D) at 12 – 26 weeks of gestation in (A) 298 white women (P less than 0.01, includes American Indian/Alaskan, n=2), (B) 494 black women (P=0.09), (C) 444 non-obese women (P=0.01, body mass index (BMI) less than 30 kg/m2), and (D) 348 obese women (P=0.39; BMI greater than 30 kg/m2) all at high risk for preeclampsia. Gray shading is 95% confidence intervals. 25(OH)D was modeled by restricted cubic splines with three knots (28.8, 60.0, and 103.3) at percentiles 10%, 50%, and 90% in a log-binomial model adjusted for latitude, obesity status, race, treatment group, and risk group. 25(OH)D values less than 15 nmol/L (n=9) and greater than 120 nmol/L (n=34) were omitted from the figure.