Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
1 23
1 23
Maternal-Fetal Medicine
Received: 7 December 2013 / Accepted: 7 February 2014 / Published online: 27 February 2014
Springer-Verlag Berlin Heidelberg 2014
Abstract
PurposeTo determine if correction of cut-offs of haemoglobin levels to define anaemia at high altitudes affects
rates of adverse perinatal outcomes.
Methods Data were obtained from 161,909 mothers
and newborns whose births occurred between 1,000 and
4,500m above sea level (masl). Anaemia was defined with
or without correction of haemoglobin (Hb) for altitude as
Hb <11g/dL. Correction of haemoglobin per altitude was
performed according to guidelines from the World Health
Organization. Rates of stillbirths and preterm births were
also calculated.
Results Stillbirth and preterm rates were significantly
reduced in cases of anaemia calculated after correction of
haemoglobin for altitude compared to values obtained without Hb correction. At high altitudes (3,0004,500 masl),
after Hb correction, the rate of stillbirths was reduced
from 37.7 to 18.3 per 1,000 live births (p<0.01); similarly, preterm birth rates were reduced from 13.1 to 8.76%
(p<0.01). The odds ratios for stillbirths and for preterm
births were also reduced after haemoglobin correction.
Conclusion At high altitude, correction of maternal haemoglobin should not be performed to assess the risks for
preterm birth and stillbirth. In fact, using low altitude Hb
cut-off is associated with predicting those at risk.
G.F.Gonzales(*) M.Gasco
Department ofBiological andPhysiological Sciences,
Universidad Peruana Cayetano Heredia, Honorio Delgado 430,
Lima 31, Peru
e-mail: gustavo.gonzales@upch.pe
G.F.Gonzales V.Tapia M.Gasco
Instituto de Investigaciones de la Altura, Universidad Peruana
Cayetano Heredia, Honorio Delgado 430, Lima 31, Peru
Introduction
Most populations living at high altitude (HA) show an
increase in haemoglobin (Hb) concentration as the result of
increased erythropoietic activity as a mechanism to compensate an effect of tissue hypoxia consequence of low
barometric pressure [1]. For such reason, populations at
high altitudes have the distribution curve of haemoglobin
shifted to the right in relation to that observed at low altitudes [2]. Because of this, the World Health Organization
(WHO) has suggested that the Hb cut-off values to define
anaemia at high altitudes should be shifted accordingly [3,
4]. In populations of mothers living between sea level and
1,000m above sea level (masl), the cut-off point of Hb to
define anaemia is 11g/dL. Using adjustments for altitude,
the Hb cut-off value to define anaemia increases as altitude
increases [3, 4]. These cut-off values can be observed in
Table1. For instance, in Cerro de Pasco at 4,340 masl, the
Hb cut-off value to define anaemia is 14.5g/dL. At low levels, this value of Hb is considered high and it is used as the
threshold to define erythrocytosis [2].
These adjustments to the Hb cut-off value to define
anaemia are based on a mathematical distribution of normality rather than the use of any clinical parameter. In
the highlands, the use of Hb correction to define anaemia
results in an increase of the rate of anaemia and this rate
is higher to the one observed when anaemia was defined
by iron deficiency [5]. In fact, in 800 Bolivian mothers
living at altitudes between 150 and 3,750 masl, the prevalence of anaemia was 26.6% after correcting Hb for altitude, whereas body iron measurements indicated that only
13
Table1Haemoglobin cut-offs to define maternal anaemia at different altitudes according to guidelines from WHO
Altitude (m)
Haemoglobin cut-off to
define maternal anaemia
(g/dL)
Subtracting value
to obtain Hb at
each altitude
<1,000
1,000
1,500
2,000
2,500
3,000
3,500
4,000
11
11.2
11.5
11.8
12.3
12.9
13.7
14.5
0
0.2
0.5
0.8
1.3
1.9
2.7
3.5
4,500
15.5
4.5
Materials andmethods
Study design
This is a retrospective cohort study based on data from the
Perinatal Information System in Peru that assess haemoglobin in pregnancy and perinatal outcomes.
Study population
13
67
Confounder variables
Potential confounders controlled for in the analysis were
age, body mass index (BMI), maternal education, antenatal
care, parity, preeclampsia, and urinary infection. These variables were correlated with exposure and with outcomes.
Quality control ofdata
compared with gestational age measured by physical examination and results (not shown) were also concordant.
Exposure variables
Altitude was defined in three categories, 1,0001,999 masl,
2,0002,999 masl and 3,0004,500 masl.
In some cultures, married women migrate back to their
original birthplace to deliver. In the present study, 0.59%
of women have their pregnancy in a place different to the
hospital of delivery. This group was excluded from the
analysis.
Anaemia was also defined in three categories according to the WHO classification [4]: (1) Mild anaemia,
when maternal haemoglobin values were between 9 and
<11g/dL, (2) moderate anaemia for haemoglobin values
between 7 and <9g/dL, and (3) severe anaemia for haemoglobin values below 7g/dL.
From the available data, newborns whose mothers were
diagnosed with mild, moderate or severe anaemia before
and after correction of haemoglobin cut-off for altitude were
selected for further analyses. Correction of Hb cut-off point
to define anaemia was defined according to altitude of residence using the guidelines from WHO showed in Table1.
13
Results
The rate of anaemia increased five times after Hb cut-off
was corrected to define maternal anaemia at high altitude (Fig.2). The rate of anaemia before Hb correction
decreased as altitude increased, whereas after Hb correction, the rate of anaemia was highest at the higher altitude
(3,000m) (Fig.2).
Table 2 shows the sociodemographic data according
to normal haemoglobin level or degree of anaemia (mild,
moderate or severe anaemia). Severe anaemia was associated with low and high maternal age, higher BMI, less education, less antenatal care, low and higher parity, higher
stillbirths and preterm births. Anaemia rates were decreased
as altitude of residence increases. Most of moderate/severe
anaemia was detected at third trimester of pregnancy.
The stillbirth rate in mothers with anaemia diagnosed
without correction of Hb was 28.0 per 1,000 live births,
a value significantly higher than that obtained when correction for altitude was performed, in which case the rate
was reduced to 18.2 per 1,000 live births (p<0.01). For
13
stillbirths, the ORs in anaemics defined before Hb correction were also significantly reduced from 1.69 (CI 1.48
1.94) to 1.17 (CI 1.071.27) in anaemics defined after Hb
correction for altitude (Fig.3a, b). The ORs for stillbirth
with variables as maternal age, BMI, maternal education,
number of visits of antenatal care, parity, trimester at Hb
measurement, preeclampsia, preterm births and altitude
were not different in the groups without or with Hb correction (data not shown).
The rate of preterm births in mothers with anaemia
without correction of Hb was 9.41%, which was reduced
to 8.24% (p<0.01) after Hb correction for altitude. For
preterm births, the OR in anaemics without Hb correction was 1.44 (CI 95% 1.341.55) and this value was
reduced to 1.20 (CI 95% 1.151.25) after Hb correction
for altitude (p<0.01) (Fig.3c, d). The ORs for variables
as maternal age, BMI, maternal education, number of visits of antenatal care, parity, trimester at Hb measurement,
preeclampsia, urinary infection and altitude were not different in the groups without or with Hb correction (data
not shown).
Results are similar if we break down the rates by different altitudes. Thus, the rate of stillbirths in mothers with
anaemia without Hb correction was higher than that after
Hb correction for altitude, both at 2,0002,999 masl (28.9
per 1,000 live births vs. 19.3 per 1,000 live births, p<0.01)
and at 3,0004,500 masl (37.7 per 1,000 live births vs. 18.3
per 1,000 live births, p<0.01) (Table3).
The rate of preterm births was also higher if anaemia
was diagnosed without Hb correction for altitude. This
was observed particularly at 2,0002,999 masl (9.20%
vs. 8.13%, p<0.01) and at 3,0004,500 masl (13.1 vs.
8.76%, p<0.01) (Table4). Similarly, ORs for both stillbirths and preterm births were reduced in anaemics after
Hb correction, particularly for altitudes over 2,000 masl
(Tables 3, 4). For stillbirths, the ORs calculated for the
variables age, BMI, maternal education, antenatal care,
parity, preeclampsia, preterm birth were not modified in
the groups of anaemics with or without Hb correction. For
preterm births, the ORs calculated for the variables age,
BMI, maternal education, antenatal care, parity, preeclampsia, and urinary infection were not modified after Hb
correction.
Stillbirths rates increased from 14 per 1,000 in mothers with normal haemoglobin to 161 per 1,000 in women
with severe anaemia. These values are significantly reduced
after Hb correction for altitude. For instance, in severe
anaemia, the rate of stillbirths was reduced from 161 per
1,000 to 94.5 per 1,000 (p<0.01) (Fig.4 upper). A similar pattern was observed with preterm births. The preterm
birth rate increased with magnitude of anaemia and it was
reduced if Hb cut-off to define anaemia is corrected for altitude (Fig.4 bottom).
Variable
69
Normal (151,485)
n
Age (years)
<20
28,947
2034
104,206
<34
18,332
2
BMI (kg/m )
<19
13,884
2025
97,544
>25
40,057
Study
None/prim
118,820
Sec/sup
32,665
Antenatal care
<6
76,381
6
75,104
Parity
None
73,963
13
66,227
>3
11,295
Preeclampsia
4,693
Stillbirth
2,116
Preterm
10,614
Urinary inf
27,381
Haemoglobin measurement
1st trim.
35,317
2nd trim.
62,646
3rd trim.
53,522
Altitude (m)
1,0001,999
11,371
2,0002,999
59,375
>3,000
80,739
Severe anaemia
(149)
19.1
68.8
12.1
2,044
6,105
1,219
21.8
65.2
13.0
161
608
138
17.8
67.0
15.2
222
104
23
14.8
69.8
15.4
9.2
64.4
26.4
983
6,222
2,163
10.5
66.4
23.1
83
609
215
9.2
67.1
23.7
15
91
43
10.1
61.1
28.8
78.4
21.6
7,017
2,351
74.9
25.1
652
255
71.9
28.1
98
51
65.7
34.2
50.4
49.6
4,323
5,045
46.1
53.9
275
632
30.3
69.7
53
96
35.6
64.4
48.8
43.7
7.5
3.1
1.4
7.0
18.1
4,507
3,974
887
270
210
789
1,604
48.1
42.4
9.5
2.9
2.2
8.4
17.8
364
412
131
35
58
154
143
40.1
45.4
14.5
3.9
6.4
16.9
15.8
57
68
24
5
24
38
15
38.3
45.6
16.1
3.4
16.1
25.5
10.1
23.31
41.35
35.33
938
4,420
4,010
10.01
47.18
42.81
53
305
549
5.84
33.63
60.53
25
52
72
16.78
34.90
48.32
80.34
93.37
2,552
3,808
18.03
5.99
217
340
1.53
0.53
13
68
0.09
0.11
95.93
3,008
3.57
350
0.42
68
0.08
Discussion
In this large retrospective study on populations living
between 1,000 and 4,500 masl, we analyzed the rates of
stillbirths and preterm births associated with maternal anaemia. Anaemia was diagnosed when Hb cut-off to define
anaemia was uncorrected (Hb <11g/dL) and after correction for altitude [3, 4]. According to the results, stillbirth
and preterm birth rates were significantly reduced in those
cases of anaemia diagnosed after correction of haemoglobin for altitude. Since rates of anaemia increase after Hb
correction as observed in Fig.2 and anaemia is positively
correlated with stillbirths and prematurity, our results suggest that normal (non anaemic) women may have been
included as anaemic after correction of Hb for altitude. As
these cases correspond to women with low risk of stillbirths
13
30
25
20
OR for stillbirths
15
10
5
0
Without Hb correction
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Without Hb correction
With Hb correction
9.6
9.2
9.4
1.6
1.4
9
8.8
8.6
8.4
8.2
8
7.8
7.6
With Hb correction
Maternal Anaemia
Maternal Anemia
1.2
1
0.8
0.6
0.4
0.2
0
Without Hb correction
With Hb correction
Without Hb correction
Maternal Anemia
With Hb correction
Maternal Anaemia
Fig.3Stillbirth and preterm birth rates and their odds ratios with and
without Hb correction. a *p<0.001 with respect to the group without Hb correction; b *p<0.01. Logistic regression model adjusted
by age, BMI, maternal education, antenatal care, parity, trimester at
Hb measurement, preeclampsia, preterm birth and altitude of resi-
Table3Logistic regression for the relationship between anaemia with and without a correction factor and stillbirth rates for each altitude
(1,0004,500m)
Altitude (m)
ORc
ORa
CI 95%
ORc
ORa
CI 95%
1,0001,999
2,0002,999
41
122
14.7
28.9
1.41
2.26*
1.22
1.74
0.831.77
1.412.15
59
286
13.8
19.3
1.36
1.56*
1.21
1.24
0.861.70
1.071.44
3,0004,500
129
37.7
2.53*
1.74
1.432.12
591
18.3
1.25*
1.12
1.011.26
Logistic regression model adjusted for age, BMI, maternal education, antenatal care, parity, preeclampsia, preterm birth. Each row represents a
model for each altitude range comparing the population of anemics with stillbirth and non anaemic with stillbirths (OR=1.0)
ORc crude odds ratio, ORa adjusted odds ratio
Per 1,000 live births
* p<0.01
13
71
Table4Logistic regression for the relationship between anaemia with and without a correction factor and preterm birth rates for each altitude
(1,0004,500m)
Altitude (m)
1,0001,999
2,0002,999
141
388
3,0004,500
452
%
5.07
9.20
13.1
ORc
ORa
CI 95%
ORc
ORa
CI 95%
1.25*
1.46*
1.04
1.34
0.851.27
1.201.50
201
1,203
4.72
8.13
1.15
1.34*
0.96
1.22
0.811.15
1.131.31
1.79*
1.57
1.411.75
2,827
8.76
1.16*
1.18
1.121.24
Logistic regression model adjusted for age, BMI, maternal education, antenatal care, parity, preeclampsia, and urinary infection. Each row represents a model for each altitude range comparing the population of anemics with preterm births and non anaemic with pre term births (OR=1.0)
ORc crude odds ratio, ORa adjusted odds ratio
* p<0.01
180
Without Hb correction
160
With Hb correction
140
120
P<0.01
100
80
60
P<0.01
40
P<0.01
20
0
Normal Hb
30
Mild anaemia
Moderate anaemia
Severe anaemia
Without Hb correction
With Hb correction
25
P<0.01
20
P<0.01
15
P<0.01
10
5
0
Normal Hb
Mild anaemia
Moderate anaemia
Severe anaemia
13
Anaemia
Not
Mild
Moderate
Severe
Maternal age (years)
<20
2034
>34
BMI (kg/m2)
<19
1925
>25
None/low education
Secondary/superior
Antenatal care
<6
6
Parity
None
13
>3
Measure of Hb
1st trimester
2nd trimester
3rd trimester
Preeclampsia
Yes
Not
Preterm birth
<37weeks
37weeks
Altitude (m)
1,0001,999
2,0002,999
>3,000
Uncorrected haemoglobin
Corrected haemoglobin
OREE
OREE
95% CI
1.00
1.44
2.69
6.70
0.11
0.41
1.79
1.23
1.99
3.97
1.67
3.63
11.30
1.00
1.04
1.54
3.79
0.05
0.14
0.52
0.95
1.29
2.89
1.15
1.83
4.96
0.73
1.0
1.21
0.05
1.0
0.08
0.64
0.83
0.73
0.05
0.64
0.83
1.06
1.37
1.21
0.08
1.07
1.37
0.77
1.0
0.97
1.77
1
0.06
1.0
0.05
0.09
0.66
0.91
0.78
0.06
0.66
0.91
0.88
1.61
1.07
1.95
0.97
1.77
1.00
0.05
0.09
0.88
1.61
1.07
1.95
1.86
1.0
0.10
1.0
1.68
2.07
1.87
0.10
1.68
2.07
1.03
1.0
1.38
0.05
1.0
0.10
0.93
1.15
1.03
0.05
0.93
1.15
1.19
1.59
1.36
0.10
1.18
1.57
1.0
1.16
0.91
1.0
0.09
0.06
1.02
0.80
1.31
1.03
1.16
0.91
0.07
0.06
1.02
0.79
1.31
1.03
0.79
1.0
0.08
1.0
0.65
0.96
0.79
0.08
0.65
0.96
14.71
1.0
0.66
13.46
16.06
14.65
0.66
13.41
16.01
1.0
1.01
1.0
0.10
0.84
1.20
0.94
0.08
0.79
1.12
1.10
0.10
0.92
1.31
0.98
0.09
0.83
1.16
13
CI 95%
73
Table6Relationship between degree of anaemia (with and without correction factor) and preterm birth (%) in Peruvian populations living at
altitudes between 1,000 and 4,500m
Degree of anaemia
Normal
Mild
Moderate
Severe
ORc
ORa
CI 95%
ORc
ORa
CI 95%
10,614
789
154
7.01
8.42
16.9
1.0
1.22*
2.71*
1.0
1.24
2.28
1.14 1.34
1.90 2.73
7,740
3,346
637
6.67
7.63
11.09
1.0
1.15*
1.74*
1.0
1.07
1.47
1.03 1.12
1.35 1.61
25.5
4.54*
4.07
2.77 5.99
172
20.3
3.56*
2.60
2.18 3.10
38
Logistic regression model adjusted for age, BMI, maternal education, antenatal care, parity, preeclampsia, altitude and urinary infection
ORc crude odds ratio, ORa adjusted odds ratio
* p<0.01
resulting in various haemoglobin values. However, the pattern of haemoglobin values varied harmoniously according to altitude, increasing haemoglobin values as altitude
increased (not shown), thus revealing that measurement
variation was probably small.
There is an underlying assumption that all stillbirths take
place in hospitals, when many take place at home. Hence,
a considerable number of stillbirths may have gone unrecorded. This is not a problem in the present study since we
compared rates of stillbirths at hospitals in situation when
Hb was corrected for altitude or not to define anaemia.
Another limitation could be that the study was retrospective. However, as we are comparing rates of adverse perinatal outcomes when anaemia was defined after correcting
or not Hb cut-off levels for altitudes, we can overcome this
limitation.
In conclusion, at high altitude, correction of maternal
haemoglobin should not be performed to assess the risks
for preterm birth and stillbirth. In fact, using low altitude
Hb cut-offs is associated with predicting those at risk.
AcknowledgmentsThe authors appreciate the collaboration of
the Ministry of Health and their representatives from different public hospitals in Peru. We acknowledge technical assistance from Dr.
Carlos Carrillo, MPH Leopoldo Bejarano and Lic. Vanessa Vasquez.
The study was supported by a grant from the UNDP/UNFPA/WHO/
WORLD BANK Special Programme of Research, Development
and Research Training in Human Reproduction at the World Health
Organization. The funding sources had no involvement in the design,
analysis or conclusions of this study.
Conflict of interestNone.
References
1. Gonzales GF (2011) Hemoglobin and testosterone: importance
on high altitude acclimatization and adaptation. Rev Peru Med
Exp Salud Publica 28:92100 Spanish
2. Gonzales GF, Steenland K, Tapia V (2009) Maternal hemoglobin
level and fetal outcome at low and high altitudes. Am J Physiol
Regul Integr Comp Physiol 297:R1477R1485
13
13