Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Maternal Hemoglobin Concentration During Pregnancy and Risk of Stillbirth

Maternal Hemoglobin Concentration During Pregnancy and Risk of Stillbirth Abstract Context High and low maternal hemoglobin concentrations during pregnancy have been reported to increase risk of small-for-gestational-age (SGA) birth, which is a predictor of stillbirth. The relationship between hemoglobin concentration during pregnancy and risk of stillbirth is unclear. Objective To study the associations among hemoglobin concentration at first measurement during antenatal care, change in hemoglobin concentration during pregnancy, and risk of stillbirth. Design, Setting, and Participants Population-based, matched case-control study of births from 1987 through 1996 in Sweden including 702 primiparous women with stillbirths occurring at 28 weeks' gestation or later and 702 primiparous women with live births. Main Outcome Measures Risk of stillbirth, classified as malformed or nonmalformed, antepartum or intrapartum, preterm or term, and SGA or non-SGA, compared by maternal hemoglobin concentration at first antenatal measurement and weekly changes in hemoglobin concentration during pregnancy, adjusted for maternal age, body mass index, height, smoking, socioeconomic status, and week of first hemoglobin measurement. Results In multivariate analyses, compared with women with hemoglobin concentrations of 126 to 135 g/L at first antenatal measurement, women with concentrations of 146 g/L or higher were at increased risk of stillbirth (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.0-3.3). This risk was slightly increased when the analysis was restricted to antepartum stillbirths without malformations (OR, 2.0; 95% CI, 1.1-3.8). When we further restricted the analyses to preterm and SGA antepartum nonmalformed stillbirths, the ORs increased to 2.7 (95% CI, 1.1-6.4) and 4.2 (95% CI, 1.3-13.9), respectively. Excluding women with preeclampsia and eclampsia further increased these risks. Average weekly change in hemoglobin concentration during early or late pregnancy was not significantly associated with risk of stillbirth, although a larger decrease in concentration tended to be protective. Anemia (hemoglobin concentration <110 g/L) was not significantly associated with risk of stillbirth in multivariate analyses (OR, 1.2; 95% CI, 0.5-2.7). Conclusions High hemoglobin concentration at first measurement during antenatal care appears to be associated with increased risk of stillbirth, especially preterm and SGA antepartum stillbirths. The relationship between maternal hematological parameters and pregnancy outcome has been a source of continuing controversy. In developed countries, not only maternal anemia1-3 but also high hemoglobin concentration during pregnancy4-9 has been reported to increase the risks of unfavorable outcomes such as small-for-gestational-age (SGA) birth, preterm birth, and perinatal death. The associations between hemoglobin concentration in early pregnancy, changes in hemoglobin concentration during pregnancy, and risk of stillbirth are not known.10 In this large population-based case-control study, we investigated the associations between the hemoglobin concentration at the first measurement during antenatal care, weekly change in hemoglobin concentration during early and late pregnancy, and risk of stillbirth. Because the causes of stillbirth are varied, we also investigated hemoglobin-related risks in different subgroups of stillbirths. Methods Description of Sample From the population-based Swedish Medical Birth Register,11 we obtained information on all single births to primiparous women living and giving birth within a geographically defined area in central Sweden from 1987 through 1996. Of 220,712 births, there were 725 stillbirths occurring at 28 weeks of gestation or later. For each case, we randomly selected 2 controls, matched by year and hospital of birth. Using the unique national registration number assigned to all Swedish residents, the standardized antenatal and obstetric records were retrieved from each of the 25 delivery archives and examined by one of us (O.S.), using a structured protocol. Of the 725 eligible cases, information on 702 (97%) was obtained: 10 cases were missing in the archives, 8 were incorrectly coded for delivery hospital, and 5 had an incorrect national registration number and were thereby impossible to trace. As for the control births, a second control birth was included only if the first selected control birth was not found. To obtain an equal number of case and control births (n = 702), we included another 25 controls: 24 in the first control group were missing in the archives and 1 had an incorrect national registration number (retrieval rate, 96%). In 4 case and 2 control births, the mother had not received antenatal care. In 12 cases and 4 controls, the antenatal records were incomplete. The number of antenatal visits ranged from 0 to 34; mean and median number of visits were 12.2 and 12, respectively. Information in antenatal records is prospectively recorded from the first to the last antenatal visit. At registration to antenatal care, we obtained information about maternal height, weight, occupation, and cigarette smoking status. During pregnancy, we repeatedly obtained information about blood pressure and proteinuria. From the obstetric and pediatric records, we recorded maternal age at delivery, gestational age at birth (or diagnosed stillbirth), weight at birth, and diagnosed malformations or chromosomal abnormalities. Gestational age was estimated using early second trimester ultrasound examination (generally before 19 gestational weeks) when available (94% of case and control births); otherwise, the last menstrual period was used. Body mass index (BMI) at registration to antenatal care was calculated as weight in kilograms divided by the square of height in meters. A BMI of less than 20.0 was categorized as lean; 20.0 through 24.9 as normal weight; 25.0 through 29.9 as overweight; and 30.0 or more as obese. Information on mother's occupation was classified according to the Swedish Socioeconomic Classification12 and grouped as blue-collar workers, low-level white-collar workers, intermediate and high-level white-collar workers, students, and others (54 were not in the labor force, 12 were self-employed, and 80 were not classifiable). Maternal smoking status was categorized as no daily smoking, 1 to 9 cigarettes per day, or 10 or more cigarettes per day. Preeclampsia and eclampsia were defined according to the criteria given by the National High Blood Pressure Education Program Working Group Report on High Blood Pressure in Pregnancy.13 Mild preeclampsia was defined as gestational hypertension (blood pressure ≥140/90 mm Hg in ≥2 readings ≥6 hours apart, occurring after 20 weeks of gestation) accompanied by mild or moderate proteinuria (≥2 urinary dipsticks with 1+ or 2+ or 300 mg to 3 g of protein in a 24-hour urine collection). Severe preeclampsia was defined as gestational hypertension accompanied by severe proteinuria (≥2 urinary dipsticks with 3+ or ≥3 g of protein in a 24-hour urine collection), or gestational hypertension with a diastolic blood pressure of at least 110 mm Hg (in ≥2 readings ≥6 hours apart), regardless of proteinuria. Eclampsia was defined as seizures in a patient with preeclampsia that could not be attributed to other causes. Blood was taken for hemoglobin concentration estimation repeatedly during pregnancy. The mean gestational week of the first hemoglobin concentration measurement was 10.5 for cases and controls. There are no uniform criteria for categorization of hemoglobin measures during pregnancy. Based on a previous report, we divided hemoglobin concentration taken during antenatal care into 5 categories.9 A separate stratification was performed to address the question of anemia (<110 g/L) and risk of stillbirth.14 The average weekly change in hemoglobin concentration during pregnancy was estimated separately for early and late pregnancy periods using simple linear regression. The start of the early period was identified as the week that the first hemoglobin concentration measurement was taken, and the end of the early period was identified as the week of gestation closest to week 25, provided it was between week 17 and week 28. The regression line for the early period was fitted to all hemoglobin measurements during the period (the mean number of hemoglobin measurements was 2.9). The estimated weekly change in concentration was given by the slope of the fitted regression line. The method of estimating average change in hemoglobin concentration for the late period was similar; the start of the late period was identical to the end week of the early period, and the end week was defined as the week of the last concentration measurement in pregnancy (the mean number of hemoglobin measurements was 3.1). Average weekly change in hemoglobin concentration was not estimated if the difference between the start and end dates of the respective period was less than 3 weeks. Among cases, we only considered hemoglobin concentration measurements prior to stillbirth or prior to the birth week of the matched control. Among controls no hemoglobin concentration measurements were used after gestational week of stillbirth for the matched case. Average weekly change in hemoglobin concentration in early and late pregnancy was divided into 6 categories. Stillbirths were classified as malformed or nonmalformed, antepartum or intrapartum, preterm or term, and SGA or non-SGA. Malformations were only noted if they were lethal or potentially lethal.15 Preterm birth was defined as gestational age less than 37 completed weeks, and SGA was defined as birth weight more than 2 SDs below the mean birth weight for gestational age, according to the Swedish birth weight curve in common use.16 Statistical Methods We used standard statistical methods for the analysis of matched case-control studies, namely conditional logistic regression models estimated using SAS PROC PHREG.17 The data were individually matched, but some matched pairs had identical values for the matching variables. We therefore analyzed the data as if they were N:M matched, ie, each stratum contained N controls and M cases where neither N nor M were necessarily 1. All 702 controls were eligible for the analysis but were only included if they belonged to a stratum that contained at least 1 case. For each model, observations with missing values for explanatory variables were excluded from the analysis, although the corresponding matched case or control was included in the analysis if the stratum contained at least 1 case and 1 control with full covariate information. The number of observations excluded due to missing values was small (Table 1a). Odds ratios (ORs) with 95% confidence intervals (CIs) were used to estimate the relative risk. The study was approved by the research ethics committee at Karolinska Institutet, Stockholm, Sweden. The committee did not require informed consent because the information had been collected in hospital archives; however, no personal identification numbers were used in the databases. Results Low-(≤115 g/L) and high-≥146 g/L) hemoglobin concentrations at first antenatal measurement were associated with increased risk of stillbirth in the univariate analyses (Table 1). Twenty cases and 15 controls were anemic (<110 g/L), and the OR for stillbirth related to anemia was 1.5 (95% CI, 0.7-3.0). Compared with women with a moderate decrease in hemoglobin concentration during early or late pregnancy, women with the largest increase in hemoglobin concentration were at increased risk of stillbirth. The risk of stillbirth increased with advancing maternal age, increasing BMI, low maternal stature, smoking, low socioeconomic status, and severe preeclampsia or eclampsia. The relationship between hemoglobin concentration at first antenatal measurement and average change in hemoglobin concentration during pregnancy is shown in the Figure 1. In early pregnancy, mean hemoglobin concentrations decreased, with the exception of the group in the lowest concentration (≤115 g/L) category. Some of the results demonstrated in the Figure 1 can be explained by regression to the mean, that is, the tendency of observations that are extreme by chance to move closer to the mean when repeated. However, those women who began pregnancy with a high hemoglobin concentration continued to have higher concentrations throughout pregnancy and vice versa. The lowest hemoglobin concentrations were found close to the 30th week of gestation for all groups. In late pregnancy, there was a small increase in mean hemoglobin concentrations, with the exception of the group with the highest hemoglobin concentration at first antenatal measurement (≥146 g/L). We also wanted to investigate whether cases and controls within different first hemoglobin categories differed in the change of hemoglobin values during pregnancy. We therefore repeated the analysis in the Figure 1 separately for cases and controls, but the changes in hemoglobin values from first hemoglobin concentration were essentially the same (data available on request). Women with a high first hemoglobin measurement were more likely to have a high BMI, preeclampsia or eclampsia, whereas there was no clear association between first-hemoglobin measurement and maternal age, cigarette smoking, or occupational status (data available on request). Next, we performed conditional logistic regression analysis on the associations between first-hemoglobin concentration, early pregnancy change in hemoglobin values, and risk of stillbirth (Table 2). Compared with women with mid-range first hemoglobin concentrations (126-135 g/L), women with the highest hemoglobin concentrations were at increased risk of stillbirth (OR, 1.8). Anemia (<110 g/L) (16 cases and 13 controls) was not significantly associated with risk of stillbirth (OR, 1.2; 95% CI, 0.5-2.7). The subgroups of malformed and intrapartum stillbirths were too small to be analyzed separately, and we assumed that these groups were not primarily associated with hemoglobin concentration during pregnancy. We therefore restricted the analysis to antepartum stillbirths without malformations. This restriction slightly increased the risk of stillbirth related to high first hemoglobin concentration (OR, 2.0). There was no significant interaction between first hemoglobin concentration and any of the covariates in the model. Women with the largest increase in hemoglobin concentration in early pregnancy were at increased risk of stillbirth, but change in hemoglobin concentration in early pregnancy was not statistically significant in any model. Hemoglobin concentration at first measurement during antenatal care and change in hemoglobin concentration in early pregnancy were also analyzed separately (after adjusting for the confounders included in Table 2) with no major differences in results. Antepartum stillbirths without malformations were also analyzed according to birth weight for gestational age (Table 3). For SGA stillbirths, a first hemoglobin measurement of 146 g/L or higher was associated with a more than 4-fold increase in risk, and when women with preeclampsia or eclampsia were excluded, this risk was substantially increased (OR, 15.1). We found no significant association between first hemoglobin level and non-SGA stillbirths (P = .19). Change in hemoglobin concentration during early pregnancy was not significantly associated with risks of SGA or non-SGA stillbirths (P = .18 and P = .13, respectively). We also stratified nonmalformed antepartum stillbirths by gestational age (Table 4). For preterm stillbirth, women with a first hemoglobin level of 146 g/L or higher had an almost 3-fold increase in risk, and this risk was further increased when women with preeclampsia or eclampsia were excluded. There was no significant association between term stillbirth and first hemoglobin concentration (P = .14), or change in hemoglobin during early pregnancy and risk of preterm or term stillbirth (P = .17 and P = .08, respectively). Finally, we introduced change in hemoglobin concentration during late pregnancy in 3 models of stillbirth (all, nonmalformed, and antepartum nonmalformed stillbirth). Change in hemoglobin concentration during early and late pregnancy was not statistically significant in any model, indicating that first hemoglobin concentration was a stronger predictor for stillbirth compared with changes in hemoglobin concentration during early or late pregnancy. Comment The results from this study suggest that a high hemoglobin concentration in early pregnancy is associated with an almost 2-fold increase in risk of stillbirth. The risk increase is even larger among specific stillbirth subgroups, such as antepartum nonmalformed preterm or SGA stillbirths. When cases and controls with preeclampsia or eclampsia were excluded, these risks were further increased. Change in hemoglobin concentration during early or late pregnancy did not significantly influence stillbirth risk. Because anemia (<110 g/L) in early pregnancy is rare in Sweden,18 our study did not have sufficient power to investigate the association between anemia and stillbirth. This population-based case-control study included more than 700 cases of stillbirth and 700 controls, for which 96% to 97% of the medical records were retrieved. Because exposure was registered prospectively during pregnancy, recall bias is unlikely. Possible confounders such as maternal age, BMI, smoking status and socioeconomic status were accounted for in the analyses. The relatively homogeneous population in Sweden, the standardized antenatal care, and the use of uniform records further minimized the possibility of confounding due to differences in sociodemographic factors or pregnancy management. We investigated primiparous women with singleton pregnancies, and the conclusions from this study can only be interpreted for this group. We were restricted to data found in the archive files and could not investigate other potential confounders, such as, nutritional intake, or physical activity. Furthermore, we could not study any additional hematological parameters besides hemoglobin concentration. Iron supplementation improves maternal hematological indexes during pregnancy.19 In Sweden the clinical guidelines recommend iron supplementation for pregnant women, unless hemoglobin concentration is high.18 Thus, women with high hemoglobin concentration at the first visit for antenatal care are less likely to receive iron supplementation compared with those with a lower hemoglobin concentration. A recent review of clinical trials states that routine iron supplementation "had no detectable effect on any substantive measures of either maternal or fetal outcome."20 In this study, we did not have information on iron supplementation and therefore could not investigate its possible confounding of the association between hemoglobin concentration and risk of stillbirth. A U-shaped association between lowest hemoglobin concentration during pregnancy and stillbirth rates has been reported previously,10 and 2 studies have reported increased rates of perinatal death with high- and low-hemoglobin concentration during pregnancy.4,6 However, change in hemoglobin concentration during pregnancy was not considered in these studies. In our investigation, we adjusted for week of first hemoglobin measurement and used no hemoglobin measurements after stillbirth of the case or delivery of the individually matched control. Thus, bias due to measurements of hemoglobin concentrations at different gestational weeks for cases and controls was minimized. Plasma volume expansion and lowered hemoglobin concentration are physiologic responses to pregnancy.21 Plasma volume expansion is considered important for fetal growth,22 and several studies have reported an increased incidence of low birth weight associated with a high maternal hemoglobin concentration.5-7,9 The mechanism by which expansion of the plasma volume enhances fetal growth is not known, but reduced blood viscosity may favor blood flow in the maternal intervillous space.9 High hemoglobin values are associated with placental infarction,4 and pregnancy hemodilution may, by preventing thrombosis in the uteroplacental circulation, promote fetal nourishment and growth.5 In our investigation, we used weekly change in hemoglobin concentration as an indirect measure of plasma volume expansion. However, weekly change in hemoglobin concentration was not significantly associated with stillbirth risk, and did not influence stillbirth risks related to first hemoglobin concentration. Moreover, although women with the highest initial hemoglobin value had the largest drop in hemoglobin concentration, the risk of stillbirth was almost exclusively confined to this group. Thus, high hemoglobin levels in early pregnancy may per se be deleterious for fetal survival. Small for gestational age is associated with a high hemoglobin concentration during pregnancy,6,9 and intrauterine growth restriction is one of the main determinants of stillbirth.23,24 It is therefore noteworthy that the highest risks in our study were found for antepartum nonmalformed SGA stillbirths, suggesting that the risk of stillbirth related to high hemoglobin concentration measured during antenatal care is associated with impaired fetal growth. As fetal growth restriction is reported to be more important for preterm than term stillbirths,25 the hemoglobin-related risk increase in preterm stillbirths may have a similar mechanism. Hemoconcentration with hypovolemia is a feature of preeclampsia.6,26 The reduction in plasma volume is reported to be proportional to the severity of the condition,27 and the hematological changes occur before the onset of preeclampsia.28,29 Because preeclampsia and eclampsia may lie in the causal pathway between high hemoglobin concentration and stillbirth, we did not control for them in the multivariate analysis. However, when women with preeclampsia or eclampsia were excluded, the stillbirth-related risks due to a high hemoglobin concentration increased, suggesting that a high hemoglobin concentration primarily influences risk of stillbirth in nonpreeclamptic pregnancies. The exact mechanisms for the association between maternal hemoglobin concentration and stillbirth, as well as the possible effects of iron supplementation, require further studies. Women with a high hemoglobin concentration at first measurement during antenatal care are reported to be at increased risk of SGA births,6 and we found that a high first hemoglobin concentration increased the risk of stillbirth, especially antepartum-SGA stillbirths. We therefore suggest that pregnancies with high hemoglobin concentrations should be considered as high-risk pregnancies. To improve antenatal detection of fetal growth disturbances, it may be prudent to perform repeated ultrasound scannings on these pregnancies. References 1. Lieberman E, Ryan KJ, Monson RR, Schoenbaum SC. Association of maternal hematocrit with premature labor. Am J Obstet Gynecol.1988;159:107-114.Google Scholar 2. Klebanoff MA, Shiono PH, Selby JV, Trachtenberg AI, Graubard BI. Anemia and spontaneous preterm birth. Am J Obstet Gynecol.1991;164:59-63.Google Scholar 3. Scholl TO, Hediger ML, Fischer RL, Shearer JW. Anemia vs iron deficiency: increased risk of preterm delivery in a prospective study. Am J Clin Nutr.1992;55:985-988.Google Scholar 4. Naeye RL. Placental infarction leading to fetal or neonatal death: a prospective study. Obstet Gynecol.1977;50:583-588.Google Scholar 5. Sagen N, Nilsen ST, Kim HC, Bergsjo P, Koller O. Maternal hemoglobin concentration is closely related to birth weight in normal pregnancies. Acta Obstet Gynecol Scand.1984;63:245-248.Google Scholar 6. Murphy JF, O'Riordan J, Newcombe RG, Coles EC, Pearson JF. Relation of haemoglobin levels in first and second trimesters to outcome of pregnancy. Lancet.1986;1:992-995.Google Scholar 7. Huisman A, Aarnoudse JG. Increased 2nd trimester hemoglobin concentration in pregnancies later complicated by hypertension and growth retardation: early evidence of a reduced plasma volume. Acta Obstet Gynecol Scand.1986;65:605-608.Google Scholar 8. Knottnerus JA, Delgado LR, Knipschild PG, Essed GG, Smits F. Haematologic parameters and pregnancy outcome: a prospective cohort study in the third trimester. J Clin Epidemiol.1990;43:461-466.Google Scholar 9. Steer P, Alam MA, Wadsworth J, Welch A. Relation between maternal haemoglobin concentration and birth weight in different ethnic groups. BMJ.1995;310:489-491.Google Scholar 10. Garn SM, Ridella SA, Petzold AS, Falkner F. Maternal hematologic levels and pregnancy outcomes. Semin Perinatol.1981;5:155-162.Google Scholar 11. Cnattingius S, Ericson A, Gunnarskog J, Kallen B. A quality study of a medical birth registry. Scand J Soc Med.1990;18:143-148.Google Scholar 12. Swedish Socioeconomic Classification. Reports on Statistical Co-ordination.Stockholm, Sweden: Statistics Sweden; 1995.Google Scholar 13. National High Blood Pressure Education Program Working Group. Report on High Blood Pressure in Pregnancy. Am J Obstet Gynecol.1990;163:1691-1712.Google Scholar 14. Centers for Disease Control. Criteria for anemia in children and childbearing-aged women. MMWR Morb Mortal Wkly Rep.1989;38:400-404.Google Scholar 15. Winbo IG, Serenius FH, Dahlquist GG, Kallen BA. A computer-based method for cause of death classification in stillbirths and neonatal deaths. Int J Epidemiol.1997;26:1298-1306.Google Scholar 16. Marsal K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr.1996;85:843-848.Google Scholar 17. SAS Institute Inc. SAS/STAT Software: Changes and Enhancements Through Release 6.12. Cary, NC: SAS Institute Inc; 1997. 18. During Health Care Before and After Pregnancy [in Swedish]. Stockholm, Sweden: National Board of Health and Welfare; 1997. Report 1996:7. 19. US Preventive Services Task Force. Routine iron supplementation during pregnancy: policy statement. JAMA.1993;270:2846-2848.Google Scholar 20. Mahomed K. Iron supplementation in pregnancy [Cochrane Review on CD-ROM]. Oxford, England: Cochrane Library Update Software; 1999. Issue 4. 21. Whittaker PG, Macphail S, Lind T. Serial hematologic changes and pregnancy outcome. Obstet Gynecol.1996;88:33-39.Google Scholar 22. Steer PJ. Maternal hemoglobin concentration and birth weight. Am J Clin Nutr.2000;71(supp5):1285S-1287S.Google Scholar 23. Berkowitz GS, Papiernik E. Epidemiology of preterm birth. Epidemiol Rev.1993;15:414-443.Google Scholar 24. Cnattingius S, Haglund B, Kramer MS. Differences in late fetal death rates in association with determinants of small for gestational age fetuses: population based cohort study. BMJ.1998;316:1483-1487.Google Scholar 25. Gardosi J, Mul T, Mongelli M, Fagan D. Analysis of birthweight and gestational age in antepartum stillbirths. Br J Obstet Gynaecol.1998;105:524-530.Google Scholar 26. Sagen N, Koller O, Haram K. Haemoconcentration in severe pre-eclampsia. Br J Obstet Gynaecol.1982;89:802-805.Google Scholar 27. Chesley LC. Plasma and red cell volumes during pregnancy. Am J Obstet Gynecol.1972;112:440-450.Google Scholar 28. Blekta M, Hlavaty V, Trnkova M, Bendl J, Bendova L, Chytil M. Volume of whole blood and absolute amount of serum proteins in the early stage of late toxemia of pregnancy. Am J Obstet Gynecol.1970;106:10-13.Google Scholar 29. Gallery ED, Hunyor SN, Gyory AZ. Plasma volume contraction: a significant factor in both pregnancy-associated hypertension (pre-eclampsia) and chronic hypertension in pregnancy. Q J Med.1979;48:593-602.Google Scholar http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JAMA American Medical Association

Maternal Hemoglobin Concentration During Pregnancy and Risk of Stillbirth

Download PDF
7 pages

Loading next page...
 
/lp/american-medical-association/maternal-hemoglobin-concentration-during-pregnancy-and-risk-of-pTgiuMxDx5

References (41)

Publisher
American Medical Association
Copyright
Copyright © 2000 American Medical Association. All Rights Reserved.
ISSN
0098-7484
eISSN
1538-3598
DOI
10.1001/jama.284.20.2611
Publisher site
See Article on Publisher Site

Abstract

Abstract Context High and low maternal hemoglobin concentrations during pregnancy have been reported to increase risk of small-for-gestational-age (SGA) birth, which is a predictor of stillbirth. The relationship between hemoglobin concentration during pregnancy and risk of stillbirth is unclear. Objective To study the associations among hemoglobin concentration at first measurement during antenatal care, change in hemoglobin concentration during pregnancy, and risk of stillbirth. Design, Setting, and Participants Population-based, matched case-control study of births from 1987 through 1996 in Sweden including 702 primiparous women with stillbirths occurring at 28 weeks' gestation or later and 702 primiparous women with live births. Main Outcome Measures Risk of stillbirth, classified as malformed or nonmalformed, antepartum or intrapartum, preterm or term, and SGA or non-SGA, compared by maternal hemoglobin concentration at first antenatal measurement and weekly changes in hemoglobin concentration during pregnancy, adjusted for maternal age, body mass index, height, smoking, socioeconomic status, and week of first hemoglobin measurement. Results In multivariate analyses, compared with women with hemoglobin concentrations of 126 to 135 g/L at first antenatal measurement, women with concentrations of 146 g/L or higher were at increased risk of stillbirth (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.0-3.3). This risk was slightly increased when the analysis was restricted to antepartum stillbirths without malformations (OR, 2.0; 95% CI, 1.1-3.8). When we further restricted the analyses to preterm and SGA antepartum nonmalformed stillbirths, the ORs increased to 2.7 (95% CI, 1.1-6.4) and 4.2 (95% CI, 1.3-13.9), respectively. Excluding women with preeclampsia and eclampsia further increased these risks. Average weekly change in hemoglobin concentration during early or late pregnancy was not significantly associated with risk of stillbirth, although a larger decrease in concentration tended to be protective. Anemia (hemoglobin concentration <110 g/L) was not significantly associated with risk of stillbirth in multivariate analyses (OR, 1.2; 95% CI, 0.5-2.7). Conclusions High hemoglobin concentration at first measurement during antenatal care appears to be associated with increased risk of stillbirth, especially preterm and SGA antepartum stillbirths. The relationship between maternal hematological parameters and pregnancy outcome has been a source of continuing controversy. In developed countries, not only maternal anemia1-3 but also high hemoglobin concentration during pregnancy4-9 has been reported to increase the risks of unfavorable outcomes such as small-for-gestational-age (SGA) birth, preterm birth, and perinatal death. The associations between hemoglobin concentration in early pregnancy, changes in hemoglobin concentration during pregnancy, and risk of stillbirth are not known.10 In this large population-based case-control study, we investigated the associations between the hemoglobin concentration at the first measurement during antenatal care, weekly change in hemoglobin concentration during early and late pregnancy, and risk of stillbirth. Because the causes of stillbirth are varied, we also investigated hemoglobin-related risks in different subgroups of stillbirths. Methods Description of Sample From the population-based Swedish Medical Birth Register,11 we obtained information on all single births to primiparous women living and giving birth within a geographically defined area in central Sweden from 1987 through 1996. Of 220,712 births, there were 725 stillbirths occurring at 28 weeks of gestation or later. For each case, we randomly selected 2 controls, matched by year and hospital of birth. Using the unique national registration number assigned to all Swedish residents, the standardized antenatal and obstetric records were retrieved from each of the 25 delivery archives and examined by one of us (O.S.), using a structured protocol. Of the 725 eligible cases, information on 702 (97%) was obtained: 10 cases were missing in the archives, 8 were incorrectly coded for delivery hospital, and 5 had an incorrect national registration number and were thereby impossible to trace. As for the control births, a second control birth was included only if the first selected control birth was not found. To obtain an equal number of case and control births (n = 702), we included another 25 controls: 24 in the first control group were missing in the archives and 1 had an incorrect national registration number (retrieval rate, 96%). In 4 case and 2 control births, the mother had not received antenatal care. In 12 cases and 4 controls, the antenatal records were incomplete. The number of antenatal visits ranged from 0 to 34; mean and median number of visits were 12.2 and 12, respectively. Information in antenatal records is prospectively recorded from the first to the last antenatal visit. At registration to antenatal care, we obtained information about maternal height, weight, occupation, and cigarette smoking status. During pregnancy, we repeatedly obtained information about blood pressure and proteinuria. From the obstetric and pediatric records, we recorded maternal age at delivery, gestational age at birth (or diagnosed stillbirth), weight at birth, and diagnosed malformations or chromosomal abnormalities. Gestational age was estimated using early second trimester ultrasound examination (generally before 19 gestational weeks) when available (94% of case and control births); otherwise, the last menstrual period was used. Body mass index (BMI) at registration to antenatal care was calculated as weight in kilograms divided by the square of height in meters. A BMI of less than 20.0 was categorized as lean; 20.0 through 24.9 as normal weight; 25.0 through 29.9 as overweight; and 30.0 or more as obese. Information on mother's occupation was classified according to the Swedish Socioeconomic Classification12 and grouped as blue-collar workers, low-level white-collar workers, intermediate and high-level white-collar workers, students, and others (54 were not in the labor force, 12 were self-employed, and 80 were not classifiable). Maternal smoking status was categorized as no daily smoking, 1 to 9 cigarettes per day, or 10 or more cigarettes per day. Preeclampsia and eclampsia were defined according to the criteria given by the National High Blood Pressure Education Program Working Group Report on High Blood Pressure in Pregnancy.13 Mild preeclampsia was defined as gestational hypertension (blood pressure ≥140/90 mm Hg in ≥2 readings ≥6 hours apart, occurring after 20 weeks of gestation) accompanied by mild or moderate proteinuria (≥2 urinary dipsticks with 1+ or 2+ or 300 mg to 3 g of protein in a 24-hour urine collection). Severe preeclampsia was defined as gestational hypertension accompanied by severe proteinuria (≥2 urinary dipsticks with 3+ or ≥3 g of protein in a 24-hour urine collection), or gestational hypertension with a diastolic blood pressure of at least 110 mm Hg (in ≥2 readings ≥6 hours apart), regardless of proteinuria. Eclampsia was defined as seizures in a patient with preeclampsia that could not be attributed to other causes. Blood was taken for hemoglobin concentration estimation repeatedly during pregnancy. The mean gestational week of the first hemoglobin concentration measurement was 10.5 for cases and controls. There are no uniform criteria for categorization of hemoglobin measures during pregnancy. Based on a previous report, we divided hemoglobin concentration taken during antenatal care into 5 categories.9 A separate stratification was performed to address the question of anemia (<110 g/L) and risk of stillbirth.14 The average weekly change in hemoglobin concentration during pregnancy was estimated separately for early and late pregnancy periods using simple linear regression. The start of the early period was identified as the week that the first hemoglobin concentration measurement was taken, and the end of the early period was identified as the week of gestation closest to week 25, provided it was between week 17 and week 28. The regression line for the early period was fitted to all hemoglobin measurements during the period (the mean number of hemoglobin measurements was 2.9). The estimated weekly change in concentration was given by the slope of the fitted regression line. The method of estimating average change in hemoglobin concentration for the late period was similar; the start of the late period was identical to the end week of the early period, and the end week was defined as the week of the last concentration measurement in pregnancy (the mean number of hemoglobin measurements was 3.1). Average weekly change in hemoglobin concentration was not estimated if the difference between the start and end dates of the respective period was less than 3 weeks. Among cases, we only considered hemoglobin concentration measurements prior to stillbirth or prior to the birth week of the matched control. Among controls no hemoglobin concentration measurements were used after gestational week of stillbirth for the matched case. Average weekly change in hemoglobin concentration in early and late pregnancy was divided into 6 categories. Stillbirths were classified as malformed or nonmalformed, antepartum or intrapartum, preterm or term, and SGA or non-SGA. Malformations were only noted if they were lethal or potentially lethal.15 Preterm birth was defined as gestational age less than 37 completed weeks, and SGA was defined as birth weight more than 2 SDs below the mean birth weight for gestational age, according to the Swedish birth weight curve in common use.16 Statistical Methods We used standard statistical methods for the analysis of matched case-control studies, namely conditional logistic regression models estimated using SAS PROC PHREG.17 The data were individually matched, but some matched pairs had identical values for the matching variables. We therefore analyzed the data as if they were N:M matched, ie, each stratum contained N controls and M cases where neither N nor M were necessarily 1. All 702 controls were eligible for the analysis but were only included if they belonged to a stratum that contained at least 1 case. For each model, observations with missing values for explanatory variables were excluded from the analysis, although the corresponding matched case or control was included in the analysis if the stratum contained at least 1 case and 1 control with full covariate information. The number of observations excluded due to missing values was small (Table 1a). Odds ratios (ORs) with 95% confidence intervals (CIs) were used to estimate the relative risk. The study was approved by the research ethics committee at Karolinska Institutet, Stockholm, Sweden. The committee did not require informed consent because the information had been collected in hospital archives; however, no personal identification numbers were used in the databases. Results Low-(≤115 g/L) and high-≥146 g/L) hemoglobin concentrations at first antenatal measurement were associated with increased risk of stillbirth in the univariate analyses (Table 1). Twenty cases and 15 controls were anemic (<110 g/L), and the OR for stillbirth related to anemia was 1.5 (95% CI, 0.7-3.0). Compared with women with a moderate decrease in hemoglobin concentration during early or late pregnancy, women with the largest increase in hemoglobin concentration were at increased risk of stillbirth. The risk of stillbirth increased with advancing maternal age, increasing BMI, low maternal stature, smoking, low socioeconomic status, and severe preeclampsia or eclampsia. The relationship between hemoglobin concentration at first antenatal measurement and average change in hemoglobin concentration during pregnancy is shown in the Figure 1. In early pregnancy, mean hemoglobin concentrations decreased, with the exception of the group in the lowest concentration (≤115 g/L) category. Some of the results demonstrated in the Figure 1 can be explained by regression to the mean, that is, the tendency of observations that are extreme by chance to move closer to the mean when repeated. However, those women who began pregnancy with a high hemoglobin concentration continued to have higher concentrations throughout pregnancy and vice versa. The lowest hemoglobin concentrations were found close to the 30th week of gestation for all groups. In late pregnancy, there was a small increase in mean hemoglobin concentrations, with the exception of the group with the highest hemoglobin concentration at first antenatal measurement (≥146 g/L). We also wanted to investigate whether cases and controls within different first hemoglobin categories differed in the change of hemoglobin values during pregnancy. We therefore repeated the analysis in the Figure 1 separately for cases and controls, but the changes in hemoglobin values from first hemoglobin concentration were essentially the same (data available on request). Women with a high first hemoglobin measurement were more likely to have a high BMI, preeclampsia or eclampsia, whereas there was no clear association between first-hemoglobin measurement and maternal age, cigarette smoking, or occupational status (data available on request). Next, we performed conditional logistic regression analysis on the associations between first-hemoglobin concentration, early pregnancy change in hemoglobin values, and risk of stillbirth (Table 2). Compared with women with mid-range first hemoglobin concentrations (126-135 g/L), women with the highest hemoglobin concentrations were at increased risk of stillbirth (OR, 1.8). Anemia (<110 g/L) (16 cases and 13 controls) was not significantly associated with risk of stillbirth (OR, 1.2; 95% CI, 0.5-2.7). The subgroups of malformed and intrapartum stillbirths were too small to be analyzed separately, and we assumed that these groups were not primarily associated with hemoglobin concentration during pregnancy. We therefore restricted the analysis to antepartum stillbirths without malformations. This restriction slightly increased the risk of stillbirth related to high first hemoglobin concentration (OR, 2.0). There was no significant interaction between first hemoglobin concentration and any of the covariates in the model. Women with the largest increase in hemoglobin concentration in early pregnancy were at increased risk of stillbirth, but change in hemoglobin concentration in early pregnancy was not statistically significant in any model. Hemoglobin concentration at first measurement during antenatal care and change in hemoglobin concentration in early pregnancy were also analyzed separately (after adjusting for the confounders included in Table 2) with no major differences in results. Antepartum stillbirths without malformations were also analyzed according to birth weight for gestational age (Table 3). For SGA stillbirths, a first hemoglobin measurement of 146 g/L or higher was associated with a more than 4-fold increase in risk, and when women with preeclampsia or eclampsia were excluded, this risk was substantially increased (OR, 15.1). We found no significant association between first hemoglobin level and non-SGA stillbirths (P = .19). Change in hemoglobin concentration during early pregnancy was not significantly associated with risks of SGA or non-SGA stillbirths (P = .18 and P = .13, respectively). We also stratified nonmalformed antepartum stillbirths by gestational age (Table 4). For preterm stillbirth, women with a first hemoglobin level of 146 g/L or higher had an almost 3-fold increase in risk, and this risk was further increased when women with preeclampsia or eclampsia were excluded. There was no significant association between term stillbirth and first hemoglobin concentration (P = .14), or change in hemoglobin during early pregnancy and risk of preterm or term stillbirth (P = .17 and P = .08, respectively). Finally, we introduced change in hemoglobin concentration during late pregnancy in 3 models of stillbirth (all, nonmalformed, and antepartum nonmalformed stillbirth). Change in hemoglobin concentration during early and late pregnancy was not statistically significant in any model, indicating that first hemoglobin concentration was a stronger predictor for stillbirth compared with changes in hemoglobin concentration during early or late pregnancy. Comment The results from this study suggest that a high hemoglobin concentration in early pregnancy is associated with an almost 2-fold increase in risk of stillbirth. The risk increase is even larger among specific stillbirth subgroups, such as antepartum nonmalformed preterm or SGA stillbirths. When cases and controls with preeclampsia or eclampsia were excluded, these risks were further increased. Change in hemoglobin concentration during early or late pregnancy did not significantly influence stillbirth risk. Because anemia (<110 g/L) in early pregnancy is rare in Sweden,18 our study did not have sufficient power to investigate the association between anemia and stillbirth. This population-based case-control study included more than 700 cases of stillbirth and 700 controls, for which 96% to 97% of the medical records were retrieved. Because exposure was registered prospectively during pregnancy, recall bias is unlikely. Possible confounders such as maternal age, BMI, smoking status and socioeconomic status were accounted for in the analyses. The relatively homogeneous population in Sweden, the standardized antenatal care, and the use of uniform records further minimized the possibility of confounding due to differences in sociodemographic factors or pregnancy management. We investigated primiparous women with singleton pregnancies, and the conclusions from this study can only be interpreted for this group. We were restricted to data found in the archive files and could not investigate other potential confounders, such as, nutritional intake, or physical activity. Furthermore, we could not study any additional hematological parameters besides hemoglobin concentration. Iron supplementation improves maternal hematological indexes during pregnancy.19 In Sweden the clinical guidelines recommend iron supplementation for pregnant women, unless hemoglobin concentration is high.18 Thus, women with high hemoglobin concentration at the first visit for antenatal care are less likely to receive iron supplementation compared with those with a lower hemoglobin concentration. A recent review of clinical trials states that routine iron supplementation "had no detectable effect on any substantive measures of either maternal or fetal outcome."20 In this study, we did not have information on iron supplementation and therefore could not investigate its possible confounding of the association between hemoglobin concentration and risk of stillbirth. A U-shaped association between lowest hemoglobin concentration during pregnancy and stillbirth rates has been reported previously,10 and 2 studies have reported increased rates of perinatal death with high- and low-hemoglobin concentration during pregnancy.4,6 However, change in hemoglobin concentration during pregnancy was not considered in these studies. In our investigation, we adjusted for week of first hemoglobin measurement and used no hemoglobin measurements after stillbirth of the case or delivery of the individually matched control. Thus, bias due to measurements of hemoglobin concentrations at different gestational weeks for cases and controls was minimized. Plasma volume expansion and lowered hemoglobin concentration are physiologic responses to pregnancy.21 Plasma volume expansion is considered important for fetal growth,22 and several studies have reported an increased incidence of low birth weight associated with a high maternal hemoglobin concentration.5-7,9 The mechanism by which expansion of the plasma volume enhances fetal growth is not known, but reduced blood viscosity may favor blood flow in the maternal intervillous space.9 High hemoglobin values are associated with placental infarction,4 and pregnancy hemodilution may, by preventing thrombosis in the uteroplacental circulation, promote fetal nourishment and growth.5 In our investigation, we used weekly change in hemoglobin concentration as an indirect measure of plasma volume expansion. However, weekly change in hemoglobin concentration was not significantly associated with stillbirth risk, and did not influence stillbirth risks related to first hemoglobin concentration. Moreover, although women with the highest initial hemoglobin value had the largest drop in hemoglobin concentration, the risk of stillbirth was almost exclusively confined to this group. Thus, high hemoglobin levels in early pregnancy may per se be deleterious for fetal survival. Small for gestational age is associated with a high hemoglobin concentration during pregnancy,6,9 and intrauterine growth restriction is one of the main determinants of stillbirth.23,24 It is therefore noteworthy that the highest risks in our study were found for antepartum nonmalformed SGA stillbirths, suggesting that the risk of stillbirth related to high hemoglobin concentration measured during antenatal care is associated with impaired fetal growth. As fetal growth restriction is reported to be more important for preterm than term stillbirths,25 the hemoglobin-related risk increase in preterm stillbirths may have a similar mechanism. Hemoconcentration with hypovolemia is a feature of preeclampsia.6,26 The reduction in plasma volume is reported to be proportional to the severity of the condition,27 and the hematological changes occur before the onset of preeclampsia.28,29 Because preeclampsia and eclampsia may lie in the causal pathway between high hemoglobin concentration and stillbirth, we did not control for them in the multivariate analysis. However, when women with preeclampsia or eclampsia were excluded, the stillbirth-related risks due to a high hemoglobin concentration increased, suggesting that a high hemoglobin concentration primarily influences risk of stillbirth in nonpreeclamptic pregnancies. The exact mechanisms for the association between maternal hemoglobin concentration and stillbirth, as well as the possible effects of iron supplementation, require further studies. Women with a high hemoglobin concentration at first measurement during antenatal care are reported to be at increased risk of SGA births,6 and we found that a high first hemoglobin concentration increased the risk of stillbirth, especially antepartum-SGA stillbirths. We therefore suggest that pregnancies with high hemoglobin concentrations should be considered as high-risk pregnancies. To improve antenatal detection of fetal growth disturbances, it may be prudent to perform repeated ultrasound scannings on these pregnancies. References 1. Lieberman E, Ryan KJ, Monson RR, Schoenbaum SC. Association of maternal hematocrit with premature labor. Am J Obstet Gynecol.1988;159:107-114.Google Scholar 2. Klebanoff MA, Shiono PH, Selby JV, Trachtenberg AI, Graubard BI. Anemia and spontaneous preterm birth. Am J Obstet Gynecol.1991;164:59-63.Google Scholar 3. Scholl TO, Hediger ML, Fischer RL, Shearer JW. Anemia vs iron deficiency: increased risk of preterm delivery in a prospective study. Am J Clin Nutr.1992;55:985-988.Google Scholar 4. Naeye RL. Placental infarction leading to fetal or neonatal death: a prospective study. Obstet Gynecol.1977;50:583-588.Google Scholar 5. Sagen N, Nilsen ST, Kim HC, Bergsjo P, Koller O. Maternal hemoglobin concentration is closely related to birth weight in normal pregnancies. Acta Obstet Gynecol Scand.1984;63:245-248.Google Scholar 6. Murphy JF, O'Riordan J, Newcombe RG, Coles EC, Pearson JF. Relation of haemoglobin levels in first and second trimesters to outcome of pregnancy. Lancet.1986;1:992-995.Google Scholar 7. Huisman A, Aarnoudse JG. Increased 2nd trimester hemoglobin concentration in pregnancies later complicated by hypertension and growth retardation: early evidence of a reduced plasma volume. Acta Obstet Gynecol Scand.1986;65:605-608.Google Scholar 8. Knottnerus JA, Delgado LR, Knipschild PG, Essed GG, Smits F. Haematologic parameters and pregnancy outcome: a prospective cohort study in the third trimester. J Clin Epidemiol.1990;43:461-466.Google Scholar 9. Steer P, Alam MA, Wadsworth J, Welch A. Relation between maternal haemoglobin concentration and birth weight in different ethnic groups. BMJ.1995;310:489-491.Google Scholar 10. Garn SM, Ridella SA, Petzold AS, Falkner F. Maternal hematologic levels and pregnancy outcomes. Semin Perinatol.1981;5:155-162.Google Scholar 11. Cnattingius S, Ericson A, Gunnarskog J, Kallen B. A quality study of a medical birth registry. Scand J Soc Med.1990;18:143-148.Google Scholar 12. Swedish Socioeconomic Classification. Reports on Statistical Co-ordination.Stockholm, Sweden: Statistics Sweden; 1995.Google Scholar 13. National High Blood Pressure Education Program Working Group. Report on High Blood Pressure in Pregnancy. Am J Obstet Gynecol.1990;163:1691-1712.Google Scholar 14. Centers for Disease Control. Criteria for anemia in children and childbearing-aged women. MMWR Morb Mortal Wkly Rep.1989;38:400-404.Google Scholar 15. Winbo IG, Serenius FH, Dahlquist GG, Kallen BA. A computer-based method for cause of death classification in stillbirths and neonatal deaths. Int J Epidemiol.1997;26:1298-1306.Google Scholar 16. Marsal K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr.1996;85:843-848.Google Scholar 17. SAS Institute Inc. SAS/STAT Software: Changes and Enhancements Through Release 6.12. Cary, NC: SAS Institute Inc; 1997. 18. During Health Care Before and After Pregnancy [in Swedish]. Stockholm, Sweden: National Board of Health and Welfare; 1997. Report 1996:7. 19. US Preventive Services Task Force. Routine iron supplementation during pregnancy: policy statement. JAMA.1993;270:2846-2848.Google Scholar 20. Mahomed K. Iron supplementation in pregnancy [Cochrane Review on CD-ROM]. Oxford, England: Cochrane Library Update Software; 1999. Issue 4. 21. Whittaker PG, Macphail S, Lind T. Serial hematologic changes and pregnancy outcome. Obstet Gynecol.1996;88:33-39.Google Scholar 22. Steer PJ. Maternal hemoglobin concentration and birth weight. Am J Clin Nutr.2000;71(supp5):1285S-1287S.Google Scholar 23. Berkowitz GS, Papiernik E. Epidemiology of preterm birth. Epidemiol Rev.1993;15:414-443.Google Scholar 24. Cnattingius S, Haglund B, Kramer MS. Differences in late fetal death rates in association with determinants of small for gestational age fetuses: population based cohort study. BMJ.1998;316:1483-1487.Google Scholar 25. Gardosi J, Mul T, Mongelli M, Fagan D. Analysis of birthweight and gestational age in antepartum stillbirths. Br J Obstet Gynaecol.1998;105:524-530.Google Scholar 26. Sagen N, Koller O, Haram K. Haemoconcentration in severe pre-eclampsia. Br J Obstet Gynaecol.1982;89:802-805.Google Scholar 27. Chesley LC. Plasma and red cell volumes during pregnancy. Am J Obstet Gynecol.1972;112:440-450.Google Scholar 28. Blekta M, Hlavaty V, Trnkova M, Bendl J, Bendova L, Chytil M. Volume of whole blood and absolute amount of serum proteins in the early stage of late toxemia of pregnancy. Am J Obstet Gynecol.1970;106:10-13.Google Scholar 29. Gallery ED, Hunyor SN, Gyory AZ. Plasma volume contraction: a significant factor in both pregnancy-associated hypertension (pre-eclampsia) and chronic hypertension in pregnancy. Q J Med.1979;48:593-602.Google Scholar

Journal

JAMA – American Medical Association

Published: Nov 22, 2000

Keywords: pregnancy,hemoglobin,fetal death,prenatal care,mothers,congenital abnormality,hemoglobin measurement,anemia

There are no references for this article.