Enrollment of participated women
During 2011-12, 112,425 women were recruited, among whom 2,002 (1.78%) multiple pregnancies, 1541(1.37%) fetal demise or abortion, 826(0.73%) fetal malformation, 39 (0.03%) fetal chromosome abnormality, 265 (0.24%) birth weight less than 800g and 4925 (4.38%) women with delivering age before 28 gestational weeks or after 43 gestational weeks, were excluded. After excluding women with missing information of birth weight (1192, 1.16%) and/or maternal height (12962, 12.61%), a total of 88,673 women delivering live birth were enrolled for final analysis (Fig. 1).
Baseline characteristics of participated women
Among enrolled 88,673 women, 2003 (2.26%) women gave birth to SGA infants and 4939 (15.04%) to LGA infants. As expected, SGA infants were more likely to born to mothers of short height (20.27%) compared to mothers of tall height (7.29%), while LGA infants were more likely to born to mothers of tall height (20.71%) than to mothers of short height (6.86%) (P < 0.05) (Table 1).
Table 1
Baseline characteristics of enrolled pregnant women.
| SGA N = 2003 | AGA N = 81731 | LGA N = 4939 |
| Short N = 406 | Average N = 1451 | Tall N = 146 | Short N = 8956 | Average N = 62607 | Tall N = 10168 | Short N = 339 | Average N = 3577 | Tall N = 1023 |
Maternal age | 27.71 ± 5.10 | 27.45 ± 5.20 | 27.45 ± 4.48 | 28.13 ± 4.99 | 28.05 ± 4.63 | 28.18 ± 4.23 | 29.48 ± 5.18* | 29.19 ± 4.75* | 29.24 ± 4.25* |
Paternal age | 29.84 ± 5.66 | 29.32 ± 5.58* | 29.39 ± 5.57 | 29.82 ± 5.30 | 29.89 ± 5.20 | 30.30 ± 5.12* | 30.97 ± 5.34* | 30.99 ± 5.38* | 31.17 ± 4.98* |
BMI, kg/m2 | | | | | | | | | |
< 18.5 | 39(15.79%) | 183(17.68%) | 28(25.00%) | 629(11.50%) | 5203(12.29%) | 1164(15.28%) | 4(1.79%) | 141(5.93%) | 70(9.27%) |
18.5–24.9 | 181(73.28%) | 760(73.43%) | 74(66.07%) | 4164(76.11%) | 32391(76.54%) | 5610(73.66%) | 164(73.21%) | 1673(70.32%) | 503(66.62%) |
≥ 25 | 27(10.93%) | 92(8.89%) | 10(8.93%) | 678(12.39%) | 4725(11.17%) | 842(11.06%) | 56(25.00%) | 565(23.75%) | 182(24.11%) |
Education level | | | | | | | | | |
< college | 246(60.59%) | 887(61.13%) | 76(52.05%) | 5038(56.25%) | 31723(50.67%) | 4219(41.49%) | 201(59.29%) | 1776(49.65%) | 407(39.78%) |
≥ college | 160(39.41%) | 564(38.87%) | 70(47.95%) | 3918(43.75%) | 30884(49.75%) | 5959(58.51%) | 138(40.71%) | 1801(50.35%) | 616(60.22%) |
Local resident, % | 261(64.29%) | 868(59.82%) | 95(65.07%) | 6479(72.34%) | 42410(67.74%) | 7328(72.07%) | 251(74.04%) | 2442(68.27%) | 758(74.10%) |
Region | | | | | | | | | |
Northwest | 9(2.22%) | 85(5.586%) | 6(4.11%) | 220(2.46%) | 1967(3.14%) | 302(2.97%) | 13(3.83%) | 169(4.72%) | 27(2.64%) |
North | 156(38.41%) | 842(58.03%) | 108(73.97%) | 4186(46.74%) | 38299(61.17%) | 6773(66.61%) | 204(60.18%) | 2362(66.03%) | 720(70.38%) |
South | 241(59.36%) | 524(36.11%) | 32(21.92%) | 4550(50.80%) | 22341(35.68%) | 3093(30.42%) | 122(35.99%) | 1046(29.24%) | 276(26.98%) |
Current smoker | 1(0.25%) | 5(0.34%) | 1(0.68%) | 10(0.11%) | 174(0.28%) | 43(0.42%) | 1(0.29%) | 13(0.36%) | 5(0.49%) |
Alcohol use | 1(0.25%) | 17(1.17%) | 1(0.68%) | 49(0.55%) | 638(1.02%) | 150(1.48%) | 2(0.59%) | 30(0.84%) | 19(1.86%) |
Family history of hypertension or diabetes | 1(0.25%) | 21(1.45%) | 5(3.42%) | 116(1.30%) | 1015(1.62%) | 220(2.16%) | 3(0.88%) | 81(2.26%) | 23(2.25%) |
Neonatal sex (male) | 163(40.15%) | 497(34.25%) | 41(28.08%) | 4811(53.72%) | 33431(53.40%) | 5166(50.81%) | 264(77.88%) | 2798(78.22%) | 774(75.66%) |
Primipara | 319(78.57%) | 1170(80.63%) | 123(84.25%) | 6961(77.72%) | 50773(81.10%) | 8804(86.59%) | 211(62.24%) | 2634(73.64%) | 851(83.19%) |
Delivery week | 38.52 ± 2.30* | 38.54 ± 2.32* | 38.49 ± 2.49 | 38.92 ± 1.79* | 39.02 ± 1.66 | 39.09 ± 1.54* | 38.40 ± 2.68* | 38.90 ± 2.02* | 39.08 ± 1.64 |
Preterm birth | 60(14.78%) | 204(14.06%) | 21(14.38%) | 612(6.83%) | 3564(5.69%) | 463(4.55%) | 38(11.21%) | 187(5.23%) | 31(3.03%) |
GDM/DM | 15(3.69%) | 48(3.31%) | 9(6.16%) | 516(5.76%) | 2744(4.38%) | 490(4.82%) | 35(10.32%) | 333(9.31%) | 95(9.29%) |
*P < 0.05, compare with women of average height with AGA infants. |
BMI: body mass index; GDM: gestational diabetes mellitus; DM: diabetes mellitus; FGR: fetal growth restriction. |
Compared with mothers of average height with AGA infants, mothers of short height with LGA tended to be of elder maternal age (29.48 ± 5.18 vs 28.05 ± 4.63, P < 0.05), elder paternal age (30.97 ± 5.34 vs 29.89 ± 5.20, P < 0.05) and earlier delivery age (38.40 ± 2.68 vs 39.02 ± 1.66, P < 0.05), and higher overweight (BMI ≥ 25: 25.00% vs 11.17%, P < 0.05), preterm birth (11.21% vs 5.69%) and GDM/DM (10.32% vs 4.38%, P < 0.05) (Table 1).
Additionally, mothers of tall height with SGA infants tended to have low BMI level (BMI < 18.5: 25.00% vs 12.29%, P < 0.05), of Northwest (4.11% vs 3.14%, P < 0.05) or North (73.97% vs 61.17%, P < 0.05), complicated with family history of hypertension or diabetes (3.42% vs 1.62%, P < 0.05) and preterm birth (14.38% vs 5.69%, P < 0.05). There was no statistical significance of neonatal sex or proportion of primipara (P > 0.05) (Table 1).
Association of risk of preeclampsia with maternal height and infant’s birth weight
In this retrospective study, 2385 (2.69%) women were diagnosed as preeclampsia. Women of tall height were less likely to develop preeclampsia compared to those of average height (2.33% vs 2.70%, P < 0.05) (TableS3). The incidence of preeclampsia between women with short and average height was of no statistical significance (3.03% vs 2.70%, P > 0.05). When maternal height was shorter than 10 percentiles (154 cm), the odds for preeclampsia did not significantly alter with maternal height, but risk odds for preeclampsia increased with maternal height among women with 90 percentiles (167 cm) and 95 percentiles (169 cm) (Fig. 2, A). These trends remained at different maternal BMI levels (Figure S1, B), though odds for preeclampsia also increased with BMI in overweight women (Figure S1, A), indicating associations of height and altered odds for preeclampsia was unlikely to be confounding by maternal BMI. Furtherly, compared with women of average height, those with short height underwent higher risk of preeclampsia (3.03% vs 2.70%), despite of no statistical significance in unadjusted model (Table S3). This association was statistically significant when adjusted for maternal region (adjusted OR 1.26, 95%CI 1.11–1.42, Model1) (Table S3), and of no significance in other adjusted models. Additionally, women of tall height were associated with reduced risk odds for preeclampsia in both unadjusted (OR 0.86, 95%CI 0.75–0.98) and adjusted models (adjusted OR 0.79, 95%CI 0.68–0.93, Model2) (Table S3). In sensitivity analysis with excluding cases of FGR, this significant relationship remained (Table S3).
Compared to women delivering AGA infants, those women delivering LGA and SGA were accompanied with higher proportion of preeclampsia (2.98% vs 2.34%, P < 0.05 and 16.08% vs 2.34%, P < 0.05, respectively) (Table S4). When neonatal weight was less than − 1.96 Z-score or higher than + 1.96 Z-score, risk odds for preeclampsia increased with infant’s birth weight (Fig. 2B). Women delivering SGA infants were associated with increased risk of preeclampsia in both unadjusted (unadjusted OR 7.98, 95%CI 7.02–9.07) and adjusted models (adjusted OR 8.58, 95%CI 7.30-10.07, Model2) (Table S4). In contrast, though mothers with LGA infants seemed to have increased odds of preeclampsia in unadjusted model (unadjusted OR 1.28, 95%CI 1.08–1.52) and in the model of adjusting for maternal region (adjusted OR 1.21, 95%CI 1.02–1.44, Model1), this association was of no statistical significance after adjusting for others variates (adjusted OR 0.98, 95%CI 0.79–1.21, Model2) (Table S4).
The association of neonatal weight for gestational age by stratums of maternal height levels and preeclampsia was further analyzed. Those women of tall height with SGA infants underwent the highest occurrence of preeclampsia (19.18%) (Table 2). Compared with women of average height delivering AGA infants, those women of all heights delivering SGA infants were at increased odds for preeclampsia, especially those women of tall height (adjusted OR 10.00, 95%CI 6.06–16.48, model 2) (Table 2). In contrast, proportion of preeclampsia did not show statistical significance between short women delivering LGA infants and those of average height delivering AGA infants (3.24% vs 2.35%, P > 0.05). Compared with those of average height delivering AGA infants, neither short women nor tall women delivering LGA infants were related to altered risk odds for preeclampsia, except for women of average height delivering LGA infants (adjusted OR 1.25, 95%CI 1.03–1.53, Model 1) (Table 2).
Table 2
Risk of preeclampsia by different maternal height levels and neonatal weight for gestational age.
| Preeclampsia | Unadjusted model | Model 1‡ | Model 2§ |
SGA | | | | |
Short | 12.32(50/406) | 5.83(4.32–7.87) | 6.90(5.09–9.35) | 5.93(3.84–9.16) |
Average | 16.82(244/1451) | 8.39(7.24–9.72) | 8.35(7.20–9.69) | 9.16(7.64–10.98) |
Tall | 19.18% (28/146) | 9.85(6.50-14.92) | 9.19(6.05–13.96) | 10.00(6.06–16.48) |
AGA | | | | |
Short | 2.60%(233/8956) | 1.11(0.96–1.28) | 1.22(1.06–1.40) | 1.16(0.97–1.39) |
Average | 2.35%(1473/62607) | Ref | Ref | Ref |
Tall | 2.07%(210/10168) | 0.88(0.76–1.01) | 0.85(0.74–0.98) | 0.81(0.67–0.96) |
LGA | | | | |
Short | 3.24%(11/339) | 1.39(0.76–2.54) | 1.39(0.76–2.54) | 0.78(0.32–1.91) |
Average | 3.08%(110/3577) | 1.32(1.08–1.60) | 1.25(1.03–1.53) | 1.08(0.85–1.38) |
Tall | 2.54%(26/1023) | 1.08(0.73–1.60) | 1.04(0.70–1.54) | 0.71(0.42–1.19) |
*Compared with women of average height delivering infant of AGA. Risks were presented as OR (95%CI). |
‡Adjusted for maternal region. |
§Adjusted for multivariate and variates of maternal region, maternal age and maternal body-mass index, maternal GDM/DM, and family history of hypertension or diabetes were retained in the model irrespective of significance. |