Table 4.
Serum 25(OH)D categories | n* | No. cases | Risk per 100 births |
Unadjusted risk ratio |
95% CI | Adjusted risk ratio** |
95% CI |
---|---|---|---|---|---|---|---|
Preterm birth (<37 weeks) | 804 | 215 | 0.27 | ||||
<30 nmol/l | 92 | 21 | 0.23 | 0.82 | 0.54–1.25 | 1.06 | 0.63–1.78 |
30–49 nmol/l | 203 | 50 | 0.25 | 0.88 | 0.65–1.20 | 1.12 | 0.75–1.66 |
50–74 nmol/l | 251 | 72 | 0.29 | 1.03 | 0.78–1.36 | 1.18 | 0.84–1.65 |
≥75 nmol/l | 258 | 72 | 0.28 | Reference | Reference | ||
Indicated preterm birth (<37 weeks) | 712 | 123 | 0.17 | ||||
<50 nmol/l | 264 | 40 | 0.15 | 0.79 | 0.54–1.17 | 1.29 | 0.80–2.09 |
50–74 nmol/l | 218 | 39 | 0.18 | 0.94 | 0.63–1.38 | 1.24 | 0.79–1.95 |
≥75 nmol/l | 230 | 44 | 0.19 | Reference | Reference | ||
Spontaneous preterm birth (<37 weeks) | 681 | 92 | 0.14 | ||||
At <50 nmol/l | 255 | 31 | 0.12 | 0.93 | 0.58–1.50 | 0.97 | 0.54–1.73 |
50–74 nmol/l | 212 | 33 | 0.16 | 1.19 | 0.75–1.90 | 1.23 | 0.73–2.07 |
≥75 nmol/l | 214 | 28 | 0.13 | Reference | Reference | ||
Preterm birth (<35 weeks) | 805 | 89 | 0.11 | ||||
<50 nmol/l | 295 | 37 | 0.13 | 1.48 | 0.90–2.44 | 1.79 | 1.00–3.22 |
50–74 nmol/l | 251 | 30 | 0.12 | 1.41 | 0.83–2.37 | 1.57 | 0.89–2.76 |
≥75 nmol/l | 259 | 22 | 0.08 | Reference | Reference | ||
Indicated preterm birth (<35 weeks) | 757 | 41 | 0.05 | ||||
<50 nmol/l | 276 | 18 | 0.07 | 1.47 | 0.71–3.05 | 2.48 | 1.07–5.76 |
50–74 nmol/l | 233 | 12 | 0.05 | 1.16 | 0.52–2.58 | 1.51 | 0.65–3.47 |
≥75 nmol/l | 248 | 11 | 0.04 | Reference | Reference | ||
Spontaneous preterm birth (<35 weeks) | 764 | 48 | 0.06 | ||||
<50 nmol/l | 277 | 19 | 0.07 | 1.55 | 0.75–3.19 | 1.48 | 0.65–3.39 |
50–74 nmol/l | 239 | 18 | 0.08 | 1.70 | 0.82–3.52 | 1.61 | 0.74–3.52 |
≥75 nmol/l | 248 | 11 | 0.04 | Reference | Reference |
25(OH)D, 25-hydroxyvitamin D.
Total number is higher for preterm <35 weeks’ than preterm <37 weeks’ gestation because all births (including stillbirths) at ≥35 weeks’ gestation are included.
Mixed-effects Poisson regression model controlling for prepregnancy BMI, race, parity, marital status, season of blood draw, treatment group (aspirin versus placebo), and baseline risk group as fixed effects and study site as a random effect. Additional adjustment for C-reactive protein concentration, maternal age, maternal education or infant sex had no meaningful impact on results.