texto en 
Inglés (pdf)
Articulo en XML
Referencias del artículo
Traducción automática
Permalink
INTRODUCTION
Anemia during pregnancy has been associated with increased maternal and perinatal mortality rates; increased risk of prematurity and low birth weight; and increased risk of anemia in the first months of the newborn’s life. Furthermore, women who suffered from anemia early in pregnancy, among other complications, are more likely to give birth to children with an increased risk of autism and developing attention deficit hyperactivity disorder (ADHD)1. According to the World Health Organization (WHO), the global prevalence of this problem among pregnant women was estimated at 41.8% between 1993 and 20052, and in Brazil, at 29.1%3.
Prior to 2016, WHO defined anemia during pregnancy as an Hb concentration below 110g/L at any time during pregnancy4. However, WHO recommendations on antenatal care and US Centers for Disease Control and Prevention guidelines recommended trimester-specific cutoffs for anemia (first and third trimester: <110g/L; second trimester: <105g/L)5.
However, some studies have performed analyses using multiple cutoff points and reported that only more extreme cutoff points were significantly associated with adverse birth outcomes6. In this sense, the work of Ohuma et al.6 was the pioneer in proposing that the diagnosis of anemia should be made based on hemoglobin values according to the gestational week – gestational week criterion. This prospective study analyzed data from the INTERGROWTH-21st Project carried out from 2009 to 2016, and sought, through secondary analysis, to verify the hemoglobin (Hb) percentiles in women who had uncomplicated pregnancies and who received optimal prenatal care, that is, who were in good health, nutrition, education, socioeconomic status and who started prenatal care early in pregnancy, that is, <14+0 weeks of gestation.
It is important to highlight that there is a strong association between socioeconomic issues and anemia, which always reveals a higher prevalence in lower economic levels and unfavorable socioeconomic conditions, such as being part of larger families and living in households with food insecurity7. The anemia situation is even worse for black pregnant women, who have a higher prevalence of anemia when compared to other women8. The probable cause of the high prevalence of anemia among black women may be associated with the lower socioeconomic level and high parity of this population9.
Therefore, this study aimed to analyze the factors associated with the presence of anemia in pregnant women identified using the gestational week criterion and the World Health Organization cutoff point.
METHODS
This is a cross-sectional study developed with women, after childbirth, who during conception were treated in public hospitals and hospitals affiliated with the SUS and who attended prenatal care in basic health units in the Metropolitan Region of Greater Vitória (RMGV) -ES, which is formed by the municipalities of Cariacica, Fundão, Guarapari, Serra, Viana, Vila Velha and Vitória.
Data were collected between April 2010 and February 2011 through interviews with 1,395 postpartum women and by consulting the SUS prenatal cards of former pregnant women10. As a criterion for compiling the data, it was necessary for the information provided by the women to agree with that recorded on the pregnant woman’s card. Pregnant women who lived in the Greater Vitória region and signed the Free and Informed Consent Form were included in the study, and those who did not have a pregnant woman’s card were excluded.
For data collection, ten interviewers were previously trained by researchers from the Federal University of Espírito Santo and a pilot study was conducted with 30 women after giving birth, who were subsequently not included in the main study. The pilot study took place in the maternity facilities, the same place where the main research was conducted.
During the data collection, a questionnaire was first administered to the participants, and then the records of interest for the research present in the pregnant women’s cards were collected. Among the relevant information, both from the interviews and from the cards, were sociodemographic data, health practices during pregnancy, dietary practices, knowledge about anemia in pregnancy and hemoglobin level.
The following sociodemographic variables were collected: age group, marital status, head of household, monthly family income, level of education, self-reported race/color, and number of members living in the household. To assess lifestyle, the following variables were collected: smoking and alcohol intake. The clinical variables collected were: presence of chronic diseases during pregnancy (pre-eclampsia) and presentation of any gestational complications. In addition, the following obstetric data were collected: type of delivery; birth weight; gestational age at birth; number of prenatal consultations; and guidance on healthy eating. Nutritional status was classified by the pre -gestational body mass index (BMI) collected from the pregnant woman’s card.
The hemoglobin level used to classify anemia was collected based on data available on the pregnant woman’s card and was measured as part of routine prenatal care. Information on healthy eating and the use of preventive or therapeutic supplements containing iron and folic acid was also collected from the pregnant women’s cards.
For both classification criteria, all hemoglobin data available on the pregnant woman’s card were evaluated. Therefore, pregnant women who presented any of the evaluations recorded below the established cutoff points were considered anemic.
The data were entered and analyzed in the Statistical program Package for Social Sciences (SPSS), version 21.0. Associations were tested through bivariate analyses with the chi- square test or Fisher’s exact test. Then, variables with a p-value of 5% were entered into a binary logistic regression model, using the “Enter” method, to identify independent determinants of anemia in pregnancy. In the final model, a probability value of less than 0.05 was considered statistically significant. Subsequently, for the purposes of this study, the power of the test was greater than 80% and the alpha error was less than 5%, to test the proportion of the gestational criterion in 604 and the WHO criterion in 990 women; for both situations, the minimum sample would be 120.
The study was authorized by the Research Ethics Committee of the Health Sciences Center of the Federal University of Espírito Santo (process nº. 3060797). Written informed consent was obtained from the respondents and they were assured that the information provided would be kept confidential and would be used only for study purposes.
RESULTS
Of the 1,395 postpartum women included in the initial study, 990 women were used in this study, as they had available hemoglobin information. It should be noted, however, that anemia classification according to the gestational week criterion was performed in only 604 women, with the others being excluded from this analysis because it was not possible to determine gestational age at the time of the hemoglobin test or because only blood count data prior to the 14th gestational week were available.
Among the 990 women with available hemoglobin data, the prevalence of anemia according to the WHO criteria was 29.6% (n=293). However, using the classification criterion by gestational week, among the 604 women whose gestational week and hemoglobin data were available, the prevalence found was only 4.6% (n=28), that is, there is a difference of 25 percentage points between the two diagnostic criteria.
When associating the anemia outcome with sociodemographic data, the WHO classification found a higher prevalence among women who did not live with a partner (35.5%), when compared to women who lived with a partner (27.8%, p=0.039). In the classification by gestational week, a borderline association was found with black women (10%), when compared to white women (4.9%) and mixed race women (3.7%, p=0.054). In addition, women with lower education (up to eight years) also had a higher prevalence of anemia (7.7%, p=0.001) (table 1).
Table 1 Prevalence of anemia according to WHO criteria and by gestational week according to sociodemographic data, RMGV – ES, 2010-2011
| WHO Criteria | Gestational week criteria | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Anemia | No anemia | p-value | Anemia | No anemia | p-value | |||||
| n | % | n | % | n | % | n | % | |||
| Age | ||||||||||
| Under 18 years old | 38 | 36.9 | 65 | 63.1 | 0.105 | 3 | 4.9 | 58 | 95.1 | 0.228f |
| From 18 to 35 years old | 241 | 29.3 | 582 | 70.7 | 21 | 4.2 | 481 | 95.8 | ||
| Over 35 years old | 14 | 21.9 | 50 | 78.1 | 4 | 9.8 | 37 | 90.2 | ||
| Race/color | ||||||||||
| White | 37 | 28.0 | 95 | 72.0 | 0.661 | 4 | 4.9 | 77 | 95.1 | 0.054f |
| Black | 51 | 32.5 | 106 | 67.5 | 9 | 10.0 | 81 | 90.0 | ||
| Brown (brunette/mulatto) | 192 | 29.2 | 465 | 70.8 | 15 | 3.7 | 389 | 96.3 | ||
| Education | ||||||||||
| Up to 8 years old (elementary school) | 146 | 31.9 | 312 | 68.1 | 0.095 | 20 | 7.7 | 240 | 92.3 | <0.001f |
| 9 years or older | 141 | 27.0 | 381 | 73.0 | 6 | 1.8 | 333 | 98.2 | ||
| Marital status | ||||||||||
| Lives with a partner | 223 | 27.8 | 578 | 72.2 | 0.039 | 23 | 4.8 | 456 | 95.2 | 0.729 |
| Does not live with a partner | 66 | 35.5 | 120 | 64.5 | 5 | 4.1 | 11 8 | 95.9 | ||
| Head of the family | ||||||||||
| The woman herself | 38 | 34.9 | 71 | 65.1 | 0.103 | 6 | 9.4 | 58 | 90.6 | 0.124 |
| The companion | 174 | 27.3 | 463 | 72.7 | 17 | 4.6 | 356 | 95.4 | ||
| Others | 80 | 33.2 | 161 | 66.8 | 5 | 3.0 | 159 | 97.0 | ||
| Number of people in the household | ||||||||||
| Up to 4 people | 225 | 30.4 | 516 | 69.6 | 0.343 | 18 | 4.0 | 429 | 96.0 | 0.230f |
| 5 or more people | 68 | 27.2 | 182 | 72.8 | 10 | 6.4 | 147 | 93.6 | ||
| Family income | ||||||||||
| ≤ 1 salário-mínimo | 49 | 31.8 | 105 | 68.2 | 0.241 | 3 | 3.2 | 90 | 96.8 | 0.596f |
| >1 minimum wage | 188 | 27.1 | 505 | 72.9 | 21 | 4.9 | 406 | 95.1 | ||
| Income per person | ||||||||||
| Up to R$140.00 (poverty or extreme poverty) | 41 | 29.7 | 97 | 70.3 | 0.621 | 5 | 5.8 | 81 | 94.2 | 0.573 |
| Above R$ 140.00 | 196 | 27.6 | 513 | 72.4 | 19 | 4.4 | 415 | 95.6 | ||
f= Fisher’s test.
Table 3 shows the number of prenatal consultations as a determining factor for the occurrence of anemia (p-value <5%) according to WHO criteria and includes smoking during pregnancy as a factor associated with anemia in pregnant women in both methods used. Other variables, such as gestational hypertension, assumed a significance of less than 10%.
Table 3 Logistic regression between sociodemographic, clinical and obstetric factors and lifestyle habits and anemia according to the WHO criteria, RMGV – ES, 2010-2011
| Gross OR | Lower limit | Upper limit | p-value | Adjusted OR | Lower limit | Upper limit | p-value | |
|---|---|---|---|---|---|---|---|---|
| Marital status | ||||||||
| Lives with a partner | 1 | - | - | - | 1 | - | - | - |
| Does not live with a partner | 1,426 | 1,017 | 1,998 | 0.040 | 1,396 | 0.977 | 1,994 | 0.067 |
| Number of prenatal consultations | ||||||||
| Suitable (6 or more) | 1 | - | - | - | 1 | - | - | - |
| Inadequate (< 6) | 1,349 | 1,004 | 1,811 | 0.047 | 1,201 | 0.884 | 1,633 | 0.241 |
| Smoking during pregnancy | ||||||||
| No | 1 | - | - | - | 1 | - | - | - |
| Yes | 2,015 | 1,294 | 3,137 | 0.002 | 1,817 | 1,141 | 2,893 | 0.012 |
After logistic regression analysis, only smoking remained associated with the occurrence of anemia according to the WHO criteria. Women who reported being smokers had an 82% (95% CI: 1.14-2.89) higher chance of anemia when compared to those who did not smoke (table 3).
Regarding the classification of anemia according to gestational weeks, after adjusting the variables in the logistic regression analysis, education and smoking remained associated with the occurrence of the outcome (table 4).
Table 4 Logistic regression between sociodemographic, clinical and obstetric factors and lifestyle habits and anemia according to the gestational week criterion, RMGV – ES, 2010-2011
| Gross OR | Lower limit | Upper limit | p-value | Adjusted OR | Lower limit | Upper limit | p-value | |
|---|---|---|---|---|---|---|---|---|
| Race/color | ||||||||
| White | 1 | - | - | - | 1 | - | - | - |
| Black | 2,139 | 0.632 | 7,233 | 0.221 | 1,079 | 0.265 | 4,392 | 0.916 |
| Brown (brunette/ mulatto) | 0.742 | 0.240 | 2,297 | 0.605 | 0.743 | 0.230 | 2,395 | 0.618 |
| Education | ||||||||
| 9 years or older | 1 | - | - | - | 1 | - | - | - |
| Up to 8 years old (elementary school) | 4.63 | 1.83 | 11,690 | 0.001 | 3,430 | 1,303 | 9,033 | 0.013 |
| Smoking during pregnancy | ||||||||
| No | 1 | - | - | - | 1 | - | - | - |
| Yes | 5,757 | 2,435 | 13,610 | 0.000 | 4,863 | 1,953 | 12,113 | 0.001 |
Women with up to eight years of schooling were 3.43 times (95% CI 1.30-9.03) more likely to have anemia than women with 9 or more years of schooling. Smokers were 4.86 times (95% CI 1.95-12.11) more likely to have anemia than non-smokers (table 4).
Table 2 Prevalence of anemia according to WHO criteria and by gestational week according to clinical and obstetric data and lifestyle habits, RMGV – ES, 2010-2011
| Second who criteria | gestational week criteria | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Anemia | No anemia | p-value | Anemia | No anemia | p-value | |||||
| n | % | n | % | n | % | n | % | |||
| Type of delivery | ||||||||||
| Normal | 191 | 31.1 | 424 | 68.9 | 0.184 | 19 | 5.0 | 363 | 95.0 | 0.612 |
| Caesarean section | 101 | 27.1 | 272 | 72.9 | 9 | 4.1 | 212 | 95.9 | ||
| Birth weight | ||||||||||
| Below 2500g | 27 | 29.7 | 64 | 70.3 | 0.982 | 3 | 5.7 | 50 | 94.3 | 0.728f |
| Greater than or equal to 2500g | 266 | 29.6 | 634 | 70.4 | 25 | 4.5 | 526 | 95.5 | ||
| Gestational weeks | ||||||||||
| Premature (<37 weeks) | 9 | 19.6 | 37 | 80.4 | 0.125 | 0 | 0.0 | 22 | 100.0 | 0.616f |
| Term (≥ 37 weeks) | 268 | 30.1 | 621 | 69.9 | 26 | 4.8 | 521 | 95.2 | ||
| Number of prenatal consultations | ||||||||||
| Adequado (≥ 6) | 93 | 25.7 | 269 | 74.3 | 0.046 | 6 | 2.5 | 237 | 97.5 | 0.060 |
| Inadequate (< 6) | 180 | 31.8 | 386 | 68.2 | 20 | 5.7 | 332 | 94.3 | ||
| Gestational hypertension | ||||||||||
| No | 223 | 29.7 | 529 | 70.3 | 0.074 | 21 | 4.7 | 426 | 95.3 | 1,000f |
| Yes | 12 | 19.0 | 51 | 81.0 | 1 | 2.9 | 34 | 97.1 | ||
| Guidance on healthy eating | ||||||||||
| No | 127 | 31.4 | 278 | 68.6 | 0.304 | 13 | 5.8 | 213 | 94.2 | 0.313 |
| Yes | 166 | 28.3 | 420 | 71.7 | 15 | 4.0 | 363 | 96.0 | ||
| Alcoholic beverages during pregnancy | ||||||||||
| No | 264 | 29.4 | 633 | 70.6 | 0.676 | 26 | 4.8 | 513 | 95.2 | 0.757f |
| Yes | 29 | 31.5 | 63 | 68.5 | 2 | 3.1 | 63 | 96.9 | ||
| Smoking during pregnancy | ||||||||||
| No | 238 | 27.5 | 627 | 72.5 | 0.002 * | 17 | 3.2 | 522 | 96.8 | 0.000f |
| Yes | 39 | 43.3 | 51 | 56.7 | 9 | 15.8 | 48 | 84.2 | ||
| Pre -gestational BMI | ||||||||||
| IMC ≤ 24,9Kg/m2 | 142 | 28.5 | 357 | 71.5 | 0.229 | 10 | 3.1 | 3 11 | 96.9 | 0.277f |
| BMI > 24.9kg/m2 | 50 | 24.0 | 158 | 76.0 | 7 | 5.5 | 120 | 94.5 | ||
f= Fisher’s test.
DISCUSSION
Smoking was the factor most strongly linked to the observed results, however, it is important to mention social vulnerability (low level of education), which also contributes to the development of this complication during pregnancy when the outcome was assessed according to the gestational week criterion. It was observed that the prevalence of anemia found in the study differed substantially between the two classification methods (29.6% according to the WHO criterion and 4.6% according to the gestational weeks).
Globally, it is estimated that 40% of pregnant women (95% CI: 36.4–44.7%) have anemia, with a higher prevalence in Southeast Asia (58.2%) and a lower prevalence in the Americas (25.5%)11. In Brazil, other studies that also used the WHO classification criteria recorded higher prevalences of anemia than that found in the Greater Vitória Metropolitan Region12,13. Ferreira et al.13 found a prevalence of 50% in the semiarid region of the state of Alagoas13, while Bresani et al.12, in a study carried out in Recife, Pernambuco, found that 56.6% of pregnant women had anemia12. However, in Bahia, in a more recent study, the prevalence was 18.9% in a study involving 328 pregnant women treated at urban health units in Vitória da Conquista14.
This study is the first in Brazil to classify anemia according to the cutoff points proposed by Ohuma et al.6, so the comparison of results using this method becomes limited. This method considered the specific hemoglobin percentiles proposed for each gestational age based on data from 3,502 healthy and well-nourished women from eight countries participating in the Fetal Growth Longitudinal Study (FGLS), whose healthy babies were followed up until 2 years of age. Through the findings of these authors, it was possible to define, for the first time, normative hemoglobin trajectories to establish specific gestational age distributions compatible with normal functional results, such as fetal growth, neonatal morbidity and child growth and development up to 2 years of age6.
In the present study, the use of cutoff points by gestational week differentiated the prevalence of anemia compared to the WHO classification method, which clearly has serious implications for public health. This is because, once the use of cutoff points by gestational week is implemented, the prevalence of low Hb levels will be substantially lower, allowing for more targeted treatment with reduced treatment costs.
The differences in prevalence rates found can be attributed to two main factors. The first is related to the WHO cutoff point, which derives from a statistical approach based mainly on aggregated data from four European studies with very small sample sizes, while the new proposed hemoglobin distributions are compatible with the levels of functional health outcomes observed in a large sample of healthy pregnant women from countries on several continents. The second factor highlights that the WHO classification takes into account only a fixed cutoff point, which is an implausible biological concept, while the other criterion considers Hb trajectories according to each gestational age6. Considering the scientific advances made since the understanding of hemoglobin biology, a review of the criteria proposed by the WHO may be necessary.
The cutoff points proposed by Ohuma et al.6 are much lower than the current WHO parameters. Thus, moving from a more sensitive parameter to a more specific one may generate false negatives for anemia, since some cases previously characterized as anemia would no longer be included in the classification. With underdiagnosis, many cases of anemia would not be identified, and there would not be enough early interventions for prevention and treatment, especially in the most vulnerable populations. It is known that low maternal hemoglobin levels are associated with unfavorable birth outcomes (low birth weight, premature birth, small-for-gestationalage newborns – SGA, stillbirth, and perinatal and neonatal mortality) and adverse maternal outcomes (postpartum hemorrhage, preeclampsia, and blood transfusion)15.
During pregnancy, low hemoglobin levels may occur in response to several pregnancy factors, such as hormonal changes, increased total blood volume, weight gain, and increased fetal size. All of these factors have a physiological impact on the entire system of the pregnant woman, as the blood is diluted over the months16. Gestational anemia is therefore related to the physiology of pregnancy, however social factors may also influence its occurrence, as corroborated by the data from this study.
Smoking was the only factor that remained strongly associated with anemia, regardless of the classification criteria used. The effects of smoking during pregnancy are universally known. Smoking can cause deficiency in the absorption of vitamin B12, since the hydrocyanic acid contained in cigarettes reduces its levels. Vitamin B12 deficiency, in turn, causes a drop in hemoglobin, which is associated with premature birth, reduced erythropoiesis and leukopoiesis, leading to anemia17. This demonstrates the importance of health promotion actions that clarify to pregnant women the importance of not smoking during pregnancy.
The results of this study also demonstrate that populations with lower levels of education are more vulnerable to developing anemia, which confirms the findings of other studies18,19. These individuals may possibly have impaired knowledge about healthy eating and dietary strategies necessary to prevent anemia20, in addition to having greater difficulty understanding the need for preventive vitamin and mineral supplementation at this stage of life. Furthermore, although it was not a factor that remained associated with anemia in this study, individuals with lower levels of education tend to have lower incomes, which may also hinder access to a balanced diet, the acquisition of foods rich in iron (especially meat products, considering their high added value) or iron and folic acid supplementation, in the event of their unavailability in the Unified Health System.
Reducing maternal anemia is widely recognized as central to the health of women and children. Current WHO global targets call for a 50% reduction in anemia among women of reproductive age by 202521.
It is important to highlight that health services play a very important role in the prevention or early diagnosis of this complication, since prenatal consultations allow the monitoring of the pregnant woman’s biochemical parameters, with the aim of identifying nutritional deficiencies early and establishing appropriate interventions22. It is important that nutritional guidance be provided throughout the prenatal period or even in the period prior to pregnancy, since this monitoring plays a fundamental role in the obstetric outcome, in addition to being an important ally in the adoption of healthy eating habits with the aim of improving the consumption of essential nutrients for pregnancy, such as iron, acting preventively and therapeutically against anemia23. It is also possible to mention the importance of prenatal care in providing guidance on the adoption of healthy habits, including the importance of not smoking, since smoking was an important risk factor found in the analyses.
Another challenge for the health service is early identification and continuity of care for this more vulnerable population, given that mothers with little education are those who start prenatal care later and have fewer appointments. Low maternal education levels can lead to difficulties in understanding18, even though having adequate prenatal care, in itself, is not a determining factor in the worsening of anemia. Therefore, it is essential to actively seek out these pregnant women for quality prenatal care and to control any complications that may arise24.
Considering that there are still important gaps in the understanding of the association between maternal hemoglobin concentration and maternal and child health15, the results of this study provide evidence for a critical evaluation of the current hemoglobin cutoff points proposed by the WHO and adopted by the Ministry of Health to define anemia in pregnancy. Given the importance of maternal anemia worldwide, it is essential to reach a consensus on the definition of this condition. Revision of the criteria currently used may be necessary to avoid overestimation of the prevalence of anemia and consequent unnecessary treatment and allocation of health resources. Further studies, especially prospective cohorts, would be valuable for a better assessment of the implications that revision of the currently adopted cutoff points could entail6.
Among the limitations of this study, we can mention the loss of sample due to lack of information on the pregnant woman’s card regarding hemoglobin data and the gestational week for each available test. In addition, the criterion for classifying anemia according to the gestational week can only be applied to women over the 14th week of gestation, which also implied sample loss. The hemoglobin levels used for this study were collected from the pregnant woman’s card, therefore they were not performed in a standardized manner by the same laboratory and may have been influenced by the methodology used.
The small number of individuals in some categories may have compromised the quality of the analyses. It should also be noted that the findings cannot be inferred for all pregnant women, since the study population consisted only of SUS users during prenatal care and childbirth. Finally, since this is a cross-sectional study, there is no temporal relationship between the factors that explain the development of anemia.
CONCLUSION
The study showed that factors such as maternal education and smoking increased the chance of anemia in pregnant women, depending on the classification criteria used. The diagnostic criteria recommended by the WHO are more sensitive and end up classifying a greater number of women as anemic. Further research is needed to use the criterion of the week of pregnancy to evaluate its efficacy and effectiveness as a classification criterion and as a predictor of harm to maternal and fetal health.
