Professional Documents
Culture Documents
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/51173891
CITATIONS
READS
110
402
7 AUTHORS, INCLUDING:
Rafael Mikolajczyk
University of So Paulo
SEE PROFILE
SEE PROFILE
Rintaro Mori
SEE PROFILE
SEE PROFILE
Articles
Summary
Background Denition of small for gestational age in various populations worldwide remains a challenge. References
based on birthweight are decient for preterm births, those derived from ultrasound estimates might not be applicable
to all populations, and the individualised reference can be too complex to use in developing countries. Our aim was
to create a generic reference for fetal weight and birthweight that overcame these deciencies and could be readily
adapted to local populations.
Methods We used the fetal-weight reference developed by Hadlock and colleagues and the notion of proportionality
proposed by Gardosi and colleagues and made the weight reference easily adjustable according to the mean birthweight
at 40 weeks of gestation for any local population. For application and validation, we used data from 24 countries in Africa,
Latin America, and Asia that participated in the 200408 WHO Global Survey on Maternal and Perinatal Health
(237 025 births). We compared our reference with that of Hadlock and colleagues (non-customised) and with that of
Gardosi and colleagues (individualised). For every reference, the odds ratio (OR) of adverse perinatal outcomes (stillbirths,
neonatal deaths, referral to higher-level or special care unit, or Apgar score lower than 7 at 5 min) for infants who were
small for gestational age versus those who were not was estimated with multilevel logistic regression.
Findings OR of adverse outcomes for infants small for gestational age versus those not small for gestational age
was 159 (95% CI 153166) for the non-customised fetal-weight reference compared with 287 (273301) for our
country-specic reference, and 284 (271299) for the fully individualised reference.
Interpretation Our generic reference for fetal-weight and birthweight percentiles can be easily adapted to local
populations. It has a better ability to predict adverse perinatal outcomes than has the non-customised fetal-weight
reference, and is simpler to use than the individualised reference without loss of predictive ability.
Funding None.
Introduction
Infants who are small for gestational age, generally
dened as having birthweight below the tenth percentile
at a particular gestational week, have a higher risk of
various adverse outcomes in perinatal period1,2 and in the
long term.36 The fth or third percentiles are also used
sometimes to assess eects of more pronounced growth
restriction.7 However, denition of the status of small for
gestational age in various populations worldwide has
been a longstanding challenge.8 Although a birthweight
reference based on a mixed ethnic population of
California was proposed for international comparisons,9
whether this reference is suitable for countries with
dierent ethnic composition remains unclear.
Furthermore, birthweight references based on neonates
born at various gestational weeks have been used for
nearly half a century.10 Researchers have come to realise
that birthweight references are awed at early gestational
weeks because infants born before term are more likely
to be growth restricted. When the weight at the tenth
percentile is based on preterm neonates only, it is
substantially lower than that based on all unborn fetuses
and neonates at a particular gestational week.11 The
discrepancy could be large in early gestational weeks,
which could substantially underdiagnose small for
www.thelancet.com Vol 377 May 28, 2011
1855
Articles
Methods
Creation of the generic global reference
To create a generic global reference, we adapted the
following widely accepted, ultrasound-based reference
for estimated fetal weight proposed by Hadlock and
colleagues25
Fetal weight (g) = exp(0578 + 0332 GA 000354 GA)
GA refers to gestational age in exact weeks
(eg, 39 weeks + 5 days = 397 weeks). Hadlock and
Term births
Median
(3742 weeks, %) maternal
height*
(cm, IQR)
Median
maternal
weight*
(kg, IQR)
colleagues25 used ultrasound measurements between 10 and 41 weeks of gestation of 392 pregnant
women of the European Continental Ancestry Group
living in the USA to create this optimum growth
equation. They also showed that the statistical variation
of fetal weight in a given gestational week was a
constant fraction of the mean. On the basis of this
information, they provided fetal-weight percentiles for
each gestational week.
Gardosi and colleagues13 created the individualised
reference by using the Hadlock formula and transforming
it into a so-called proportionality function (individual
weight expressed as a percentage of the expected weight
based on Hadlocks formula) and adjusting it for ethnic
group, parity, sex of the infant, and maternal height and
weight. Once the mean birthweight and standard
deviation (SD) of birthweight at 40 weeks is identied,
the mean and SD of fetal weight and birthweight for all
gestational weeks are xed; and so are the weight
Coecient of
Mean (SD)
Birthweight
reported only in birthweight at variation (%)
full 100s g* (%) 40 weeks of
gestation* (g)
Adverse
perinatal
events (%)
Deaths (%)
SGA
according to
Hadlock24 (%)
Africa
Algeria
14 775
95%
162 (159165)
72 (6580)
77%
3511 (467)
133%
40%
20%
14%
Angola
3358
59%
160 (155165)
62 (5669)
69%
3202 (463)
145%
96%
24%
19%
Congo
8354
84%
156 (152161)
59 (5565)
49%
3091 (440)
142%
57%
28%
35%
Kenya
2658
93%
158 (154163)
65 (6073)
83%
3176 (448)
141%
52%
23%
32%
Niger
7985
96%
160 (158165)
65 (5873)
51%
3103 (429)
138%
46%
31%
42%
Nigeria
7567
88%
160 (156165)
73 (6483)
97%
3298 (498)
151%
101%
39%
22%
Uganda
10 457
91%
159 (155164)
64 (5872)
95%
3336 (456)
137%
41%
24%
16%
Latin America
Argentina
7074
91%
159 (155163)
72 (6580)
26%
3494 (428)
122%
28%
09%
15%
Brazil
4847
92%
1585 (154163)
68 (6175)
10%
3331 (439)
132%
60%
15%
25%
Cuba
12 489
95%
159 (155163)
68 (6276)
34%
3374 (446)
132%
22%
11%
22%
Ecuador
11 397
93%
154 (150159)
65 (5972)
53%
3222 (422)
131%
54%
13%
31%
Mexico
19 394
91%
156 (152160)
69 (6276)
47%
3288 (432)
131%
25%
10%
24%
Nicaragua
5576
93%
155 (151158)
66 (6073)
55%
3214 (414)
129%
22%
13%
24%
Paraguay
3060
91%
158 (155161)
71 (6579)
37%
3506 (441)
126%
41%
15%
14%
13 692
92%
154 (150158)
66 (6173)
27%
3456 (422)
122%
53%
14%
15%
47%
Peru
Asia
Cambodia
5362
93%
155 (151158)
58 (5464)
91%
3126 (402)
129%
66%
19%
China
14 286
95%
159 (156162)
66 (6172)
64%
3410 (411)
121%
12%
03%
21%
India
23 960
78%
152 (150155)
55 (5059)
78%
2790 (396)
142%
88%
41%
60%
Japan
3204
95%
158 (154162)
62 (5767)
2%
3160 (357)
113%
44%
01%
44%
Nepal
8268
88%
150 (148153)
58 (5263)
87%
3016 (448)
149%
66%
21%
50%
Philippines
10 533
92%
155 (152158)
60 (5468)
48%
3052 (408)
134%
37%
16%
51%
Sri Lanka
14 708
93%
154 (150158)
60 (5467)
30%
3079 (399)
130%
18%
05%
47%
Thailand
9334
90%
156 (152160)
66 (6073)
14%
3237 (412)
127%
30%
04%
30%
Vietnam
13 167
95%
155 (152159)
60 (5565)
93%
3255 (385)
118%
11%
02%
36%
*Sample restricted to women giving birth at 40 weeks of gestation. Gestational age is measured in completed weeks. Coecient of variation is SD divided by mean birthweight times 100%. Adverse perinatal
events include any of: fresh stillbirth, dead within or after 24 h of birth, alive on 7th day postpartum but referred to a higher level or special care unit, or Apgar score lower than 7 at 5 min. Deaths include: fresh
stillbirth and dead within or after 24 h of birth.
Table 1: Characteristics of the study population in the 200408 WHO Global Survey on Maternal and Perinatal Health
1856
Articles
Birthweight (g)
4000
3000
2000
0
37
38 39 40
Gestational weeks
Local reference mean birthweight
41
37
38 39 40 41
Gestational weeks
Local reference 10th and 90th percentile
37
38 39 40 41
0
37
38 39 40
Gestational weeks
Gestational weeks
Empirical mean birthweight
Empirical 10th and 90th percentile
41
Figure 1: Comparison between recorded birthweight and birthweight estimated from the generic reference in the four countries with lowest fraction of
rounded birthweight
Argentina (A), Brazil (B), Japan (C), and Thailand (D) were the four countries with the lowest fraction of rounded birthweight in table 1.
1857
Articles
Percentage of SGA
353%
64%
36%
159 (153166)
(2)
106%
109%
39%
287 (273301)
0679
0699
(3)
107%
109%
39%
289 (275303)
0698
(4)
105%
110%
39%
288 (274302)
0698
(5)
105%
109%
39%
285 (272299)
0698
(6)
106%
109%
39%
284 (271299)
0698
Perinatal mortality
(1)
353%
25%
13%
177 (165189)
0711
(2)
106%
49%
14%
363 (339390)
0737
(3)
107%
48%
14%
365 (340391)
0737
(4)
105%
49%
14%
361 (337388)
0737
(5)
105%
49%
14%
359 (335385)
0737
(6)
106%
49%
14%
365 (340392)
0738
SGA=small for gestational age. OR=odds ratio. AUC=area under the curve. *References used for SGA classication were
dened as 10th percentile of the following denitions: in (1) original fetal growth formula developed by Hadlock and
colleagues.25 Because the global survey24 reported gestational age in completed weeks, 05 weeks were added to each
gestational age. In (2)(6) the birthweight at 40 weeks of gestation is predicted by an increasing number of variables:
(2) country, (3) country+sex of the infant, (4) country+sex of the infant+maternal height+maternal weight,
(5) country+sex of the infant+maternal height+maternal weight+maternal weight squared, (6) country+sex of the
infant+maternal height+maternal weight+maternal weight-squared+parity (reference 6 is the individualised reference
by Gardosi et al12). ORs for infants with SGA compared with those without SGA were obtained by means of random
eects logistic regression with country as random eect. Estimates for the random eects are not presented. AUC was
estimated from a model with country included as a xed eect, which was nearly the same as the random eects model
used for estimation of ORs. Adverse perinatal events include any of the following conditions: fresh stillbirth, dead
within or after 24 h of birth, alive on 7th day postpartum but referred to a higher-level or special-care unit, or Apgar
score lower than 7 at 5 min. Deaths include: fresh stillbirths and deaths within or after 24 h of birth.
Table 2: Comparison of references (16) for classication* of infants SGA with respect to the ability to
predict adverse perinatal outcomes
97th
95th
24
820
786
25
957
918
26
1110
1064
25th
10th
90th
75th
Mean
5th
3rd
1st
768
741
695
644
593
547
520
502
897
865
812
752
692
639
607
586
547
1040
1003
941
872
803
741
703
679
634
468
27
1278
1225
1198
1155
1083
1004
924
853
810
782
730
28
1461
1401
1369
1320
1238
1147
1057
975
926
894
834
29
1658
1590
1554
1498
1405
1302
1199
1106
1051
1015
947
30
1869
1792
1751
1689
1584
1468
1352
1247
1184
1144
1067
31
2091
2005
1960
1890
1773
1643
1513
1395
1325
1280
1194
32
2324
2228
2178
2100
1970
1825
1681
1551
1473
1422
1327
33
2564
2459
2403
2317
2173
2014
1854
1711
1625
1569
1464
34
2809
2694
2632
2538
2381
2206
2032
1874
1780
1719
1604
35
3056
2930
2864
2761
2590
2400
2210
2039
1937
1870
1745
36
3301
3165
3093
2983
2798
2593
2387
2203
2092
2020
1885
37
3540
3395
3318
3199
3001
2781
2561
2362
2244
2167
2021
38
3770
3615
3533
3407
3196
2961
2727
2516
2390
2308
2153
39
3987
3823
3736
3603
3380
3132
2884
2660
2527
2440
2276
40
4186
4014
3923
3783
3549
3288
3028
2794
2653
2562
2390
41
4365
4185
4090
3944
3700
3428
3157
2913
2766
2671
2492
Figure 2: Selected percentiles of fetal weight and birthweight for a population with the mean birthweight at
40 weeks of gestation of 3288 g in Mexico
Data obtained from the calculation sheet for Microsoft Oce Excel software (webappendix B). *Gestational age
in completed weeks. Standard deviation of 132% of the corresponding mean weight was used for calculation
of percentiles.
1858
Results
Articles
Discussion
We showed that although the more complex, individualised reference improved prediction of adverse
events, a simpler generic reference adjusted for country
(or geographical area) or ethnic origin could achieve the
same eect (panel). Admittedly, the comparison with a
fully customised reference was rather hampered by the
fact that only maternal weight from late pregnancy was
available. The use of weight measured before pregnancy
or in early pregnancy as requested by Gardosi and
colleagues13 might somewhat improve the performance
of the customised reference. Nevertheless, the absence of
measurements before pregnancy is a clinical reality and
reliance on self-report instead could again oset the gain
of including maternal weight in the reference.
www.thelancet.com Vol 377 May 28, 2011
1859
Articles
Articles
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
1861