Abstract

Background Evidence exists that normal gestational length varies with ethnicity. This UK-based study compares gestational length amongst a cohort of white European, Black and Asian women.

Methods The cohort comprised 122 415 nulliparous women with singleton live fetuses at the time of spontaneous labour, giving birth in the former North West Thames Health Region, London, UK.

Results The median gestational age at delivery was 39 weeks in Blacks and Asians and 40 weeks in white Europeans. Black women with normal body mass index (BMI) (18.5–24.9 kg/m2) had increased odds of preterm delivery (odds ratio [OR] = 1.33, 95% CI: 1.15, 1.56, adjusted for deprivation and BMI) compared with white Europeans. The OR of preterm delivery was also increased in Asians compared with white Europeans (OR = 1.45, 95% CI: 1.33, 1.56, adjusted for single unsupported status and smoking). Meconium stained amniotic fluid, which is a sign of fetal maturity, was statistically significantly more frequent in preterm Black and Asian infants and term Black infants compared with white European infants.

Conclusions This research suggests that normal gestational length is shorter in Black and Asian women compared with white European women and that fetal maturation may occur earlier.

The estimated date of delivery (EDD) is calculated clinically early in pregnancy, reflecting its social and medical importance. It is calculated using the date of the last menstrual period (LMP) by adding 280 days to the date of the first day of the LMP, giving a point estimate of 40 weeks for gestational length (including the 2 weeks before conception occurs). This method (sometimes known as Naegele's rule, although his method of adding 7 days and subtracting 3 months from the date of the LMP can give a date up to 3 days different to the 280-day method, because of the variation in the length of different months) has to be relied upon in areas without ultrasound access for dating. Even where ultrasound is available, the principle persists of using menstrual dates to determine the EDD, provided the date suggested by ultrasound measurement does not differ by more than 7 days. This is because the use of ultrasound as a dating technique requires the assumption that all fetal measurements are average for the gestation at which they are made, when they may truly be large or small for gestational age at that time. These methods of calculating the EDD are applied regardless of individual medical or demographic characteristics, and do not account for the fact that different babies mature at different rates.

Few women deliver on their calculated EDD and 5–10% of women deliver preterm.1,2 Several factors are known to affect the duration of pregnancy, including parity, socio-demographic characteristics, medical complications, previous preterm delivery, cigarette smoking, and maternal age.2,3

Obstetric outcomes may differ amongst ethnic groups when managed in the same setting. A British study found significant differences in duration and outcomes of labour when comparing white, Asian, and black women.4 Shorter gestational length has been observed in certain ethnic groups.1,5 Two studies have estimated an average gestational length 5 days shorter in black pregnancy.6,7 One study noted that differences were more strongly associated with the mother's rather than the father's race.8 Racial differences have also been observed in the rates of preterm (33–37 weeks) and very preterm (<33 weeks) birth in black compared with white women.9,10 A UK study explored the factors associated with preterm delivery in different ethnic groups and found that gestation was shorter in UK Africans and Afro-Caribbeans even after correction for socioeconomic risk factors.11

One hypothesis for shorter average gestational length amongst black infants is that earlier maturation of the feto-placental unit relates to the maternal pelvic size. A smaller pelvis benefits the mother in evolutionary terms in relation to posture and stability when running. However, a smaller pelvis is also associated with a higher incidence of both obstructed labour and maternal mortality. Indeed, Africans have been observed to have amongst the highest emergency caesarean section rates. In fetal terms it is advantageous for the fetus to have a large head because of the improved brain growth. Thus, this creates conflict in the maternal/fetal relationship. It therefore would be in the interest of the fetus to mature faster and deliver earlier to avoid the complications described.

It is well recognized that gross motor skills develop in black infants earlier than in their white counterparts.12 There is also evidence of earlier fetal maturation. The incidence of the fetal passage of meconium during labour is strongly related to gestational age, increasing from less than 5% at 34 weeks in white European women, to over 25% post EDD.13 Black infants are significantly more likely to pass meconium in utero at all gestational ages, indicating earlier maturation.14,15

Perinatal mortality rates also differ amongst ethnic groups. Black infants in the US experience overall higher mortality compared with white infants.16,17 In the UK, the highest perinatal mortality rates have also been seen in ethnic minorities.18,19 However, this oversimplifies the relationship between perinatal mortality and ethnicity. Black gestational age specific mortality has been observed to be lower than white infants amongst those born preterm.17,19 After 37 weeks, this pattern is reversed with higher perinatal mortality amongst black infants compared with white infants.16 These observations suggest black infants mature earlier compared with white infants hence their survival advantage if born preterm. By contrast, black infants born after 40 weeks gestation may be susceptible to complications of post maturity at earlier gestations than white infants.

The aim of this study was to compare gestational length amongst three ethnic groups in nulliparous women with singleton pregnancies and spontaneous labour.

Materials and Methods

The St Mary's Maternity Information system (SMMIS) database is a maternity database covering 18 of 20 hospitals in the former North West Thames Health Region since 1988. Each participating hospital collects data regarding maternal and neonatal factors on every patient throughout pregnancy. The collected data are sent annually to the Department of Epidemiology and Public Health at St Mary's Hospital. The main data set containing identifiers such as name and postcode is held on a Sun Workstation within the secure environment of the Department also housing the Department of Health Small Area Statistics Unit. To preserve confidentiality, all data used for this analysis were taken from a pseudo-anonymized data set from which all identifiers except the Oracle database number had been removed, thus complying with Section 60 of the Health and Social Care Act 2001. Analysis was performed on the non-attributable data set of 439 425 women who delivered between 1 January 1988 and 31 December 1998. Ethnicity data were self-reported. Women with missing ethnic group data or ethnic groups not in our study were excluded (51 402). Other exclusions were women with multiple pregnancies (4727), ante-partum stillbirths, or stillbirth of indeterminate timing, and induced or spontaneous abortions (964). Only nulliparous women (218 194 multiparous women excluded) and those who laboured spontaneously (excluded 41 723 women with induced or no labour) were included for analysis as previous preterm delivery is a risk factor for subsequent preterm delivery and those induced or who did not labour would not address the study aims.

The white European women were regarded as the reference (control) group. Black African and black Caribbean women were combined into one group, hereafter called Black. The second group comprises women from India, Pakistan, and Bangladesh, hereafter called Asian.

The main outcome measure was gestation at delivery (term birth). This is recorded in whole weeks on the SMMIS database and is derived from the EDD, which is calculated using a combination of LMP (available for 95% women), clinical examination, and ultrasound scan (available for 96% women). Term was defined as ≥37 completed weeks of pregnancy and preterm between 24 and 37 weeks. Length of neonatal stay in intensive care was compared as a binary variable of <1 day or ≥1 day.

Potential confounders were identified before analysis. These were marital status, single unsupported mother status, Carstairs deprivation score, maternal age at delivery, cigarette smoking, maternal height, body mass index (BMI) at booking, gestation at booking, history of diabetes mellitus, history of hypertension, any other ante-natal booking complications, and year of delivery.

Analysis strategy

The two ethnic groups (Black and Asian) under consideration were analysed separately and compared with the reference (white European) group. The distributions of ethnic groups in the cohort were calculated and differences in baseline characteristics were compared with the reference group using χ2 significance tests.20 The odds ratio (OR) and 95% CI for the association between the ethnic groups and the outcome of interest was calculated. Two-sided likelihood-based significance tests were deemed statistically significant if the associated P-value was ≤5%.21,22 The effect of the identified potential confounders on the unadjusted OR between ethnic group and gestational age at birth was analysed. Information from this analysis was used to develop a log linear logistic regression model. Covariates were added into the model according to their confounding effect and retained if the crude OR changed by over 10% after adjustment. Interaction terms were managed similarly. Statistical significance was assessed using the likelihood ratio test (LRT). Stata 7.0 software (StataCorp, Texas) was used for analysis.

Results

Characteristics of the cohort

There were 122 415 women remaining for analysis after the exclusions above. Of these, 98 370 were white European (80.4%), 7853 (6.4%) were Black, and 16 192 (13.2%) were Asian. Of white European women, 65% were married, compared with 42% of the Black population (χ2P < 0.001) and 95% of the Asians (χ2P < 0.001). The Carstairs‘ deprivation scores indicated greater deprivation amongst Black and Asian groups compared with the white European group (χ2P < 0.001 for both groups). Smoking was most common amongst white European women (23%) compared with 13% of Black women (χ2P < 0.001) and only 2% of Asian women (χ2P < 0.001). The majority of women had a normal BMI (18.5–24.9 kg/m2). Amongst Black women, 26% were overweight (BMI 25.0–29.9 kg/m2) and 9% were obese (BMI >30 kg/m2) (Table 1).

Table 1

The distribution of maternal characteristics by ethnic group

Maternal characteristics  N = 98 370   N = 7853   N = 16 192  
  White European
 
  Black
 
  Asian
 
 

 

 
N
 
%
 
N
 
%
 
Chi2 pa
 
N
 
%
 
Chi2 pa
 
Marital status Married 64 221 65.3 3329 42.4 <0.001 15 363 94.9 <0.001 
N = 122 415 Not married 34 149 34.71 4524 57.6  829 5.1  
Single unsupported mother No 76 704 78.0 4763 60.7 <0.001 15 664 96.7 <0.001 
 Yes 21 650 22.0 3083 39.3  528 3.3  
N = 122 392 Missing (23) (16)  (7)      
Carstairs deprivation score (quintiles)b 1 21 121 23.8 321 4.8 <0.001 1288 8.6 <0.001 
 2 21 554 24.3 665 10.0  1771 11.8  
N = 110 333 3 22 297 25.2 1228 18.5  3471 23.0  
 4 16 093 18.2 1904 28.7  4114 27.3  
 5 7570 8.5 2511 37.9  4425 29.4  
 Missing (12082) (9735)  (1224)   (1123)   
Mother's age at delivery (years) 12–20 10 014 10.2 1016 12.9 <0.001 1435 8.9 <0.001 
 20–29 59 221 60.2 5299 67.5  12 093 74.7  
N = 122 405 30–49 29 125 29.6 1538 19.6  2664 16.5  
Median 26, IQ range 23–30 Missing (10) (10)        
Cigarette smoker No 75 164 76.7 6805 87.0 <0.001 15 809 97.9 <0.001 
N = 121 971 Yes 22 844 23.3 1015 13.0  334 2.1  
 Missing (444) (362)  (33)   (49)   
Maternal height (cms) ≤157.5 14 843 17.2 1142 17.9 0.131 6893 49.6 <0.001 
Median 163, IQ range 158–168 >157.5 71 459 82.8 5224 82.1  6999 50.4  
 Missing (15 855) (8123)  (870)   (1473)   
N = 106 560          
Body mass index at bookingc Underweight 2433 2.9 223 3.6 <0.001 1333 10.0 < 0.001 
 Normal 58 270 69.9 3717 60.7  9381 70.3  
N = 102 825 Overweight 17 564 21.1 1620 26.4  2131 16.0  
Median 23, IQ range 21–25 Obese 5092 6.1 567 9.3  494 3.7  
 Missing (19 590) (15 011)  (1726)   (2853)   
Gestation at booking ≤20 weeks 80 445 90.5 5360 76.9 <0.001 12 646 83.9 <0.001 
N = 110 933 >20 weeks 8442 9.5 1611 23.1  2429 16.1  
Median 14, IQ range 12–16 Missing (11 482) (9483)  (882)   (1117)   
History of diabetes mellitus No 98 029 99.9 7817 99.9 0.700 16 140 99.9 0.565 
 Yes 87 0.1 0.1  12 0.1  
N = 122 093 Missing (322) (254)  (28)   (40)   
History of hypertension No 96 695 98.5 7693 98.4 0.18 16 077 99.5 <0.001 
N = 122 101 Yes 1432 1.5 129 1.7  75 0.5  
 Missing (314) (243)  (31)   (40)   
Booking complications No 79 665 84.4 5801 82.0 <0.001 12 145 87.1 <0.001 
N = 115 384 Yes 14 710 15.6 1270 18.0  1793 12.9  
 Missing (7031) (3995)  (782)   (2554)   
Year of delivery 1988–89 19 637 20.0 1393 17.7 <0.001 3212 19.8 <0.001 
N = 122 415 1990–91 20 467 20.8 1578 20.1  3325 20.5  
 1992–93 19 807 20.1 1786 22.7  3501 21.6  
 1994–95 15 592 15.9 1325 16.9  2546 15.7  
 1996–98 22 867 23.3 1771 22.6  3608 22.3  
Maternal characteristics  N = 98 370   N = 7853   N = 16 192  
  White European
 
  Black
 
  Asian
 
 

 

 
N
 
%
 
N
 
%
 
Chi2 pa
 
N
 
%
 
Chi2 pa
 
Marital status Married 64 221 65.3 3329 42.4 <0.001 15 363 94.9 <0.001 
N = 122 415 Not married 34 149 34.71 4524 57.6  829 5.1  
Single unsupported mother No 76 704 78.0 4763 60.7 <0.001 15 664 96.7 <0.001 
 Yes 21 650 22.0 3083 39.3  528 3.3  
N = 122 392 Missing (23) (16)  (7)      
Carstairs deprivation score (quintiles)b 1 21 121 23.8 321 4.8 <0.001 1288 8.6 <0.001 
 2 21 554 24.3 665 10.0  1771 11.8  
N = 110 333 3 22 297 25.2 1228 18.5  3471 23.0  
 4 16 093 18.2 1904 28.7  4114 27.3  
 5 7570 8.5 2511 37.9  4425 29.4  
 Missing (12082) (9735)  (1224)   (1123)   
Mother's age at delivery (years) 12–20 10 014 10.2 1016 12.9 <0.001 1435 8.9 <0.001 
 20–29 59 221 60.2 5299 67.5  12 093 74.7  
N = 122 405 30–49 29 125 29.6 1538 19.6  2664 16.5  
Median 26, IQ range 23–30 Missing (10) (10)        
Cigarette smoker No 75 164 76.7 6805 87.0 <0.001 15 809 97.9 <0.001 
N = 121 971 Yes 22 844 23.3 1015 13.0  334 2.1  
 Missing (444) (362)  (33)   (49)   
Maternal height (cms) ≤157.5 14 843 17.2 1142 17.9 0.131 6893 49.6 <0.001 
Median 163, IQ range 158–168 >157.5 71 459 82.8 5224 82.1  6999 50.4  
 Missing (15 855) (8123)  (870)   (1473)   
N = 106 560          
Body mass index at bookingc Underweight 2433 2.9 223 3.6 <0.001 1333 10.0 < 0.001 
 Normal 58 270 69.9 3717 60.7  9381 70.3  
N = 102 825 Overweight 17 564 21.1 1620 26.4  2131 16.0  
Median 23, IQ range 21–25 Obese 5092 6.1 567 9.3  494 3.7  
 Missing (19 590) (15 011)  (1726)   (2853)   
Gestation at booking ≤20 weeks 80 445 90.5 5360 76.9 <0.001 12 646 83.9 <0.001 
N = 110 933 >20 weeks 8442 9.5 1611 23.1  2429 16.1  
Median 14, IQ range 12–16 Missing (11 482) (9483)  (882)   (1117)   
History of diabetes mellitus No 98 029 99.9 7817 99.9 0.700 16 140 99.9 0.565 
 Yes 87 0.1 0.1  12 0.1  
N = 122 093 Missing (322) (254)  (28)   (40)   
History of hypertension No 96 695 98.5 7693 98.4 0.18 16 077 99.5 <0.001 
N = 122 101 Yes 1432 1.5 129 1.7  75 0.5  
 Missing (314) (243)  (31)   (40)   
Booking complications No 79 665 84.4 5801 82.0 <0.001 12 145 87.1 <0.001 
N = 115 384 Yes 14 710 15.6 1270 18.0  1793 12.9  
 Missing (7031) (3995)  (782)   (2554)   
Year of delivery 1988–89 19 637 20.0 1393 17.7 <0.001 3212 19.8 <0.001 
N = 122 415 1990–91 20 467 20.8 1578 20.1  3325 20.5  
 1992–93 19 807 20.1 1786 22.7  3501 21.6  
 1994–95 15 592 15.9 1325 16.9  2546 15.7  
 1996–98 22 867 23.3 1771 22.6  3608 22.3  
a

Chi2 test comparing Black and Asian groups (exposures) separately with white Europeans (unexposed).

b

1 is least deprived, 5 is most deprived.

c

Underweight (<18.5 kg/m2 ) normal (18.5–24.9 kg/m2) overweight (25.0–29.9 kg/m2) obese (>30.0 kg/m2).

The median gestational age at delivery was 39 completed weeks for the Black and Asian groups but 40 completed weeks for the white European group. Figure 1 shows a left shift in the cumulative frequency curves for gestational age at birth for Black and Asian women compared with white European women, which is apparent from 28 weeks gestation. Of infants born to Black mothers, 0.90% had gestations less that 28 weeks, compared with 0.23% of infants born to white mothers. At 34 weeks 2.69% of infants born to Black mothers had been born compared with 1.26% of infants born to white mothers.

Figure 1

Cumulative frequency chart to show the gestational age at delivery by ethnic group

Figure 1

Cumulative frequency chart to show the gestational age at delivery by ethnic group

Risk of preterm delivery in the three ethnic groups

Table 2 presents the risk of preterm delivery overall by maternal age, in the three ethnic groups. Black women overall had the highest proportion of preterm births (7.6%) compared with white Europeans (5.1%) and Asian women (6.5%). In the youngest age group (<20 years) Black women experienced the highest proportion of preterm births.

Table 2

Proportion of preterm deliveries by ethnic group and maternal age

 White European
 
 Black
 
 Asian
 
 

 
N
 
%
 
N
 
%
 
N
 
%
 
All deliveries 4972 5.1 596 7.6 1049 6.5 
Maternal age (years)       
<20 738 14.8 103 17.3 134 12.8 
20/29 2789 56.1 365 61.2 722 68.8 
30 1445 29.1 128 21.5 193 18.4 
 White European
 
 Black
 
 Asian
 
 

 
N
 
%
 
N
 
%
 
N
 
%
 
All deliveries 4972 5.1 596 7.6 1049 6.5 
Maternal age (years)       
<20 738 14.8 103 17.3 134 12.8 
20/29 2789 56.1 365 61.2 722 68.8 
30 1445 29.1 128 21.5 193 18.4 

Risk of term delivery in Black women compared with white European women

Amongst Black women 8% delivered preterm compared with 5% of white European women. The unadjusted odds of preterm delivery in all Black women was significantly higher than in white European women (OR = 1.54, 95% CI: 1.41, 1.69). The unadjusted odds were also calculated for the two sub-groups of the Black group. Black African women had significantly increased odds of preterm delivery (OR = 1.41, 95% CI: 1.23, 1.59) as did black Caribbean women (OR = 1.69, 95% CI: 1.52, 1.92). Further analysis was performed using the combined Black groups. The baseline analysis identified Carstairs' deprivation score and gestation at booking as confounding variables but marital status, cigarette smoking, maternal age, maternal height, booking complications, year of delivery, and single unsupported mother status were not found to be confounders. BMI was an effect modifier (test for homogeneity P < 0.001). Carstairs' deprivation score and gestational age at booking were incorporated into a logistic model and the risk of term delivery by ethnic group presented according to different BMI strata. This demonstrated increased odds of Black women with normal BMI of preterm delivery (OR = 1.33, 95% CI: 1.15, 1.56, P < 0.001) compared with white European women (Table 3).

Table 3

The estimated odds (OR) ratios for the association between Black ethnicity and preterm delivery

  Total women = 106223
 
      
  N = 98 370  N = 7853     
  White European
 
 Black
 
    

 

 
N
 
%
 
N
 
%
 
ORa
 
95% CI
 
P
 
Unadjusted        
Pre-term delivery No 93 398 95.0 7257 92.4    
 Yes 4972 5.1 596 7.6 1.54 1.41–1.69 <0.001 
(LRT P < 0.001)         
Stratified by BMI and adjusted for Carstairs deprivation score and gestational age at booking:         
BMIb Underweight 1950 2.9 179 3.8 0.61 0.28–1.33 0.22 
 Normal 47696 69.7 2816 60.0 1.33 1.15–1.56 <0.001 
N = 73064 Overweight 14517 21.2 1253 26.7 0.87 0.65–1.15 0.32 
 Obese 4211 6.2 442 9.4 0.99 0.67–1.49 0.94 
(LRT P value ≤0.001)         
  Total women = 106223
 
      
  N = 98 370  N = 7853     
  White European
 
 Black
 
    

 

 
N
 
%
 
N
 
%
 
ORa
 
95% CI
 
P
 
Unadjusted        
Pre-term delivery No 93 398 95.0 7257 92.4    
 Yes 4972 5.1 596 7.6 1.54 1.41–1.69 <0.001 
(LRT P < 0.001)         
Stratified by BMI and adjusted for Carstairs deprivation score and gestational age at booking:         
BMIb Underweight 1950 2.9 179 3.8 0.61 0.28–1.33 0.22 
 Normal 47696 69.7 2816 60.0 1.33 1.15–1.56 <0.001 
N = 73064 Overweight 14517 21.2 1253 26.7 0.87 0.65–1.15 0.32 
 Obese 4211 6.2 442 9.4 0.99 0.67–1.49 0.94 
(LRT P value ≤0.001)         
a

Odds ratio of having preterm delivery in Black compared with white European women.

b

Underweight (<18.5 kg/m2) normal (18.5–24.9 kg/m2) overweight (25.0–29.9 kg/m2) obese (>30.0 kg/m2).

Risk of term delivery in Asian women compared with white European women

Amongst Asian women, 7% delivered preterm. The unadjusted odds of preterm delivery in Asian women were significantly higher compared with white European women (OR = 1.30, 95% CI: 1.21, 1.39). Marital status, maternal age, hypertension at booking, other ante-natal booking complications (considered as a binary variable), maternal height, BMI at booking, and year of delivery were found not to confound the relationship between ethnicity and preterm delivery. Cigarette smoking and single unsupported mother status were identified as effect modifiers. The data are thus presented according to smoking and single unsupported mother strata. The association between single unsupported mother status and gestational age at delivery remained significant (LRT P = 0.04); in general all results remained statistically significant in the various smoking groups (Table 4).

Table 4

The estimated odds ratios (OR) for the association between Asian ethnic group and preterm delivery

  Total women = 114 562
 
      
  N = 98 370  N = 16 192     
  White European
 
 Asian
 
    

 

 
N
 
%
 
N
 
%
 
ORa
 
95% CI
 
P
 
Unadjusted:         
Preterm delivery No 93 398 95.0 15 143 93.5    
 Yes 4972 5.1 1049 6.5 1.30 1.21–1.39 <0.001 
(LRT P < 0.001)         
Stratifying by smoking and single unsupported mother status:         
Non-smoker:         
Single unsupported mother No 62 803 83.6 15 370 97.2 1.45 1.33–1.56 <0.001 
N = 90 964 Yes 12 352 16.4 439 2.8 1.92 1.41–2.63 <0.001 
Smoker:         
Single unsupported mother No 13 644 59.7 247 74.0 1.92 1.26–2.86 0.002 
N = 23 172 Yes 9194 40.3 87 26.1 2.33 1.28–4.17 0.005 
  Total women = 114 562
 
      
  N = 98 370  N = 16 192     
  White European
 
 Asian
 
    

 

 
N
 
%
 
N
 
%
 
ORa
 
95% CI
 
P
 
Unadjusted:         
Preterm delivery No 93 398 95.0 15 143 93.5    
 Yes 4972 5.1 1049 6.5 1.30 1.21–1.39 <0.001 
(LRT P < 0.001)         
Stratifying by smoking and single unsupported mother status:         
Non-smoker:         
Single unsupported mother No 62 803 83.6 15 370 97.2 1.45 1.33–1.56 <0.001 
N = 90 964 Yes 12 352 16.4 439 2.8 1.92 1.41–2.63 <0.001 
Smoker:         
Single unsupported mother No 13 644 59.7 247 74.0 1.92 1.26–2.86 0.002 
N = 23 172 Yes 9194 40.3 87 26.1 2.33 1.28–4.17 0.005 
a

Odds ratio of having preterm delivery in Asian compared with white European women.

Neonatal maturity

Length of neonatal stay in a special care unit varied amongst ethnic groups. Overall, all (term and preterm) Black and Asian neonates had a significantly shorter inpatient stay compared with white European infants (Black OR = 1.31, 95% CI: 1.06, 1.62, P = 0.014, Asian OR = 1.19, 95% CI: 1.02, 1.38, P = 0.023). (Table 5). Black infants were at increased odds of passing meconium both preterm (OR = 1.55, 95% CI: 1.14, 2.17) and at term (OR = 1.52, 95% CI: 1.44, 1.60) compared with white European infants. The Asian preterm infants were also at increased odds of passing meconium compared with white European infants (OR = 1.48, 95% CI: 1.13, 1.93).

Table 5

The estimated odds ratios of meconium stained amniotic fluid at birth by ethnic group and time of delivery

Preterm infants Total women = 6610
 
         
  N = 4968  N = 594    N = 1048    
  White European
 
 Black
 
   Asian
 
   

 

 
N
 
%
 
N
 
%
 
ORa
 
95% CI
 
N
 
%
 
ORb
 
95% CI
 
Meconium stained amniotic fluid No 4725 95.1 550 92.6   974 92.9   
 Yes 243 4.9 44 7.4 1.55 1.14–2.17 74 7.1 1.48 1.13–1.93 
Preterm infants Total women = 6610
 
         
  N = 4968  N = 594    N = 1048    
  White European
 
 Black
 
   Asian
 
   

 

 
N
 
%
 
N
 
%
 
ORa
 
95% CI
 
N
 
%
 
ORb
 
95% CI
 
Meconium stained amniotic fluid No 4725 95.1 550 92.6   974 92.9   
 Yes 243 4.9 44 7.4 1.55 1.14–2.17 74 7.1 1.48 1.13–1.93 
Term infants Total women = 115 778
 
         
  N = 93 384  N = 7256    N = 15 138    
  White European
 
 Black
 
   Asian
 
   

 

 
N
 
%
 
N
 
%
 
ORa
 
95% CI
 
N
 
%
 
ORb
 
95% CI
 
Meconium stained amniotic fluid No 74258 79.5 5217 71.9   11986 79.2   
 Yes 19126 20.5 2039 28.1 1.52 1.44–1.60 3152 20.8 1.02 0.98–1.07 
Term infants Total women = 115 778
 
         
  N = 93 384  N = 7256    N = 15 138    
  White European
 
 Black
 
   Asian
 
   

 

 
N
 
%
 
N
 
%
 
ORa
 
95% CI
 
N
 
%
 
ORb
 
95% CI
 
Meconium stained amniotic fluid No 74258 79.5 5217 71.9   11986 79.2   
 Yes 19126 20.5 2039 28.1 1.52 1.44–1.60 3152 20.8 1.02 0.98–1.07 
a

Odds of having meconium stained amniotic fluid in Black infants compared with odds in white European infants.

b

Odds of having meconium stained amniotic fluid in Asian infants compared with odds in white European infants.

Conclusions

For nulliparous women delivering single infants after spontaneous onset of labour the median gestational age at delivery was 39 completed weeks in the Black and Asian groups and 40 completed weeks in the white European group. This divergence in gestational length was apparent from 28 weeks. Black ethnicity was associated with increased odds of preterm delivery. Amongst Black women with a normal BMI, the odds of delivering a preterm baby was 33% higher than that of white European women. Asian ethnicity was also associated with increased odds of preterm delivery. Asian women who were both smokers and single unsupported mothers had the highest odds of preterm delivery compared with white European women.

SMMIS covers 80% of the population within a geographical area as not all hospitals in the region participate in the SMMIS collaboration. The effect of their exclusion on the results can not be quantified, making the introduction of selection bias possible. In the UK, few women have private obstetric care and this enhances the representativeness of this cohort to the general population. SMMIS is useful for studies considering ethnicity as it is based in urban areas, which reflect the ethnic diversity of the UK. By maintaining a uniform system of data collection SMMIS provides high quality data for research, which has been validated by other studies.23,24

Classification of racial or ethnic groupings is problematic and reflected by the inconsistent approach of other studies addressing such issues. SMMIS ethnicity classification is based on Department of Health guidelines, which require self-reporting. The ’missing‘ group formed 3.3% of the total. This may be missing due to non-collection of data or alternatively represent individuals unwilling to assign themselves to an ethnic group. If this is confined to a particular ethnic group this could bias results; however, given the small size of this group, it is unlikely to have a significant influence.

Calculation of gestational age in the hospitals contributing to SMMIS is by a combination of methods as described previously. Inaccuracies are inherently associated with estimation of EDD using LMP alone, but are reduced by using additional ultrasound scan information.25 EDD is liable to further error the later the first ultrasound scan. Data were unavailable for this study regarding timing of first scan and could be a source of bias in estimating EDD if ethnic groups book for antenatal care at different times. A recent study in England and Wales found that women from ethnic minorities initiated antenatal care later than white British women.26

The results of this research are consistent with other studies and indicate that Black ethnicity is associated with decreased gestational length.1,6,10,11 There is a paucity of data exploring this research question with British Asian women. One study found minimal effect of maternal characteristics on length of gestation but it was carried out in an area with less than the UK average of ethnic minorities and did not define what was meant by the term Asian.27 Aveyard et al.11 found an increased risk of preterm delivery for Afro-Caribbean women but not for African women. In the former group, they found that half of the excess risk was associated with marital status and deprivation. Henderson & Kay5 found a decrease in ’Negro‘ compared with white pregnancy duration but only included women of low socioeconomic class and used LMP for dating.

The finding that Black infants had increased odds of 1.5 of meconium stained amniotic fluid is comparable to the few published studies in this area.15 It provides epidemiological evidence concerning fetal maturity to support the hypothesis of earlier maturation. Sedaghatian et al.14 found that meconium stained amniotic fluid varied with ethnicity and was highest in ‘East African blacks’. However, the ethnic grouping in this study is flawed as it is not based on standard classification, with African and Asian groups subdivided largely on the basis of skin colour.

If there is a difference in gestational length by ethnic group this could potentially influence a few areas of clinical practice, although further data regarding infant morbidity and mortality by ethnicity would be needed first. For example, steroid injections are administered to mothers before 34 weeks to decrease the risk of respiratory distress syndrome, if delivery is needed. There is evidence to support the safety of steroids although repeated doses may increase adverse maternal outcomes.28 If Black infants mature earlier, it may be possible to halt steroid treatment earlier without compromising survival.

In ethnic groups with shorter gestation it may be appropriate to utilize different definitions of term, for instance reducing cut-off points by one week. For example, women who deliver preterm are regarded as high risk in subsequent pregnancies, which in turn influences clinical management and may limit patient choice. Black women who deliver at 36 weeks may not warrant inclusion in this category. Elective caesarean section is typically performed at 39 weeks gestation. If this standard is applied to all ethnic groups Black and Asian women may be at higher risk of requiring emergency caesarean section with the extra complications this confers. Thus, consideration could be given to slightly earlier delivery in some groups. There are advantageous social implications of increased accuracy of EDD prediction such as facilitating better maternity leave planning and child care arrangements.

The presented results suggest that gestational length may vary by ethnicity. However, it is possible that these results have arisen from confounding by other variables, such as environmental factors. The number of available variables in the dataset permits limited further exploration of this point, although there is little evidence that adjustment for socioeconomic deprivation, BMI, maternal height, cigarette smoking, or marital status has any major effect on our results.

Summary

This research provides evidence that the length of human gestation differs amongst ethnic groups in a heterogeneous maternity population. Black and Asian women may have shorter pregnancy duration compared with white European women. The discrepancy in gestational length could be associated with earlier fetal maturation. To predict accurately EDD maternal factors may need consideration. Recognition that gestational length varies by ethnicity could potentially modify principles of obstetric practice.

KEY MESSAGES

  • Obstetric outcomes are recognized to vary by ethnicity.

  • Gestational length appears to vary by ethnicity.

  • Asian and Black women have shorter gestational lengths compared with white European women.

  • This may be related to earlier fetal maturation as evidenced by passage of meconium.

References

1
Shiono PH, Klebanoff MA. Ethnic differences in preterm and very preterm delivery.
Am J Public Health
 
1986
;
76
:
1317
–21.
2
Peacock JL, Bland JM, Anderson HR. Preterm delivery: effects of socioeconomic factors, psychological stress, smoking, alcohol, and caffeine.
BMJ
 
1995
;
311
:
531
–35.
3
Berkowitz GS, Blackmore-Prince C, Lapinski RH, Savitz DA. Rick factors for preterm birth subtypes.
Epidemiology
 
1998
;
9
:
279
–85.
4
Tuck SM, Cardozo LD, Studd JW, Gibb DM, Cooper DJ. Obstetric characteristics in different racial groups.
Br J Obstet Gynaecol
 
1983
;
90
:
892
–97.
5
Henderson M, Kay J. Differences in duration of pregnancy.
Arch Environ Health
 
1967
;
14
:
905
–11.
6
Omigbodun AO, Adewuyi A. Duration of human singleton pregnancies in Ibadan, Nigeria.
J Natl Med Assoc
 
1997
;
89
:
617
–21.
7
Papiernik E, Alexander GR, Paneth N. Racial differences in pregnancy duration and its implications for perinatal care.
Med Hypotheses
 
1990
;
33
:
181
–86.
8
Migone A, Emanuel I, Mueller B, Daling J, Little RE. Gestational duration and birthweight in white, black and mixed-race babies.
Paediatr Perinat Epidemiol
 
1991
;
4
:
378
–91.
9
Adams MM, Read JA, Rawlings JS, Harlass FB, Sarno AP, Rhodes PH. Preterm delivery among black and white enlisted women in the United Sates Army.
Obstet Gynecol
 .
1993
;
81
:
65
–71.
10
Zhang J, Savitz DA. Preterm birth subtypes among blacks and whites.
Epidemiology
 
1992
;
3
:
428
–33.
11
Aveyard P, Cheng KK, Manaseki S, Gardosi J. The risk of preterm delivery in women from different ethnic groups.
Br J Obstet Gynaecol
 
2002
;
109
:
894
–99.
12
Iloeje SO, Obiekwe VU, Kaine WN. Gross motor development of Nigerian children.
Ann Trop Paediatr
 
1991
;
11
:
33
–39.
13
Steer PJ, Danielian PJ. Fetal distress in labor. In: James DK, Steer PJ, Weiner CP, Gonik B (eds). High Risk Pregnancy—Management Options. London: WB Saunders,
1999
, pp. 1121–49.
14
Sedaghatian MR, Othman L, Hossain MM, Vidyasagar D. Risk of meconium-stained amniotic fluid in different ethnic groups.
J Perinatol
 
2000
;
20
:
257
–61.
15
Alexander GR, Hulsey TC, Robillard PY, De Caunes F, Papiernik E. Determinants of meconium-stained amniotic fluid in term pregnancies.
J Perinatol
 
1994
;
14
:
259
–63.
16
Sappenfield WM, Buehler JW, Binkin NJ, Hogue CJ, Strauss LT, Smith JC. Differences in neonatal and postneonatal mortality by race, birth weight and gestational age.
Public Health Rep
 
1987
;
102
:
182
–92.
17
Mittendorf R, Williams MA, Kennedy JLJ, Berry RE, Herschel M, Aronson MP. A hypothesis to explain paradoxical racial differences in neonatal mortality.
Am J Prev Med
 
1993
;
9
:
327
–30.
18
Raleigh VS, Botting B, Balarajan R. Perinatal and postneonatal mortality in England and Wales among immigrants from the Indian subcontinent.
Indian J Pediatr
 
1990
;
57
:
551
–62.
19
Lyon AJ, Clarkson P, Jeffrey I, West GA. Effect of ethnic origin of mother on fetal outcome.
Arch Dis Child
 
1994
;
71
:F69.
20
Mood AM, Graybill FA, Boes DC. Introduction to the Theory of Statistics. 3rd Edn. New York: McGraw Hill,
1974
.
21
Cox DR, Hinkley DV. Theoretical Statistics. London: Chapman and Hall,
1974
.
22
McCullagh P, Nelder JA. Generalized Linear Models. 2nd Edn. London: Chapman and Hall,
1989
.
23
Cleary R, Beard RW, Coles J et al. The quality of routinely collected maternity data.
Br J Obstet Gynaecol
 
1994
;
101
:
1042
–47.
24
Bugg GJ, Atwal GS, Maresh M. Grand multiparae in a modern setting.
Br J Obstet Gynaecol
 
2002
;
109
:
249
–53.
25
Todros T, Ronco G, Lombardo D, Gagliardi L. The length of pregnancy: an echographic reappraisal.
J Clin Ultrasound
 
1991
;
19
:
11
–14.
26
Kupek E, Petrou S, Vause S, Maresh M. Clinical, provider and sociodemographic predictors of late initiation of antenatal care in England and Wales.
Br J Obstet Gynaecol
 
2002
;
109
:
265
–73.
27
Mongelli M, Opatola B. Duration and variability of normal pregnancy. Implications for clinical practice.
J Reprod Med
 
1995
;
40
:
645
–48.
28
Crowley P. Prophylactic corticosteroids for preterm birth (Cochrane Review). In: The Cochrane Library. Oxford: Update Software.
2002
.