ABSTRACT
Aim
To establish clinically applicable reference intervals for complete blood count (CBC) and key biochemical parameters in a Japanese cohort of twin pregnancies and to compare these parameters with singleton pregnancies.
Methods
This retrospective cohort study included women with twin pregnancies who delivered at a Japanese perinatal center between 2013 and 2023. CBC parameters (white blood cell count [WBC], red blood cell count [RBC], hemoglobin [Hb], hematocrit [Ht], mean corpuscular volume [MCV], mean corpuscular hemoglobin [MCH], mean corpuscular hemoglobin concentration [MCHC], and platelet count [PLT]) and biochemical markers (plasma glucose, aspartate aminotransferase [AST], alanine aminotransferase [ALT], triglycerides [TG], total cholesterol [TC], and uric acid [UA]) were evaluated. Measurements were categorized into two gestational periods (< 20 weeks and ≥ 20 weeks). Reference intervals were the 2.5th and 97.5th percentiles with 95% confidence intervals estimated by bootstrap resampling. Twin and singleton pregnancies were compared using the Wilcoxon rank‐sum test.
Results
A total of 734 women were included, contributing more than 1300 CBC samples and nearly 1000 biochemical samples. The mean maternal age was 32.5 years and the mean gestational age at delivery was 36.3 weeks. Compared with < 20 weeks, RBC, Hb, and Ht levels were lower at ≥ 20 weeks, whereas TG and TC levels were higher. Most other parameters showed modest changes. Compared with singleton pregnancies, Hb levels were lower after 20 weeks, and lipid levels were higher in twin pregnancies.
Conclusions
Twin pregnancies demonstrate distinctive hematologic and metabolic patterns across gestation. These reference intervals may improve laboratory interpretation and clinical management.
Keywords: biochemical parameters, complete blood count, reference intervals, twin pregnancy
1. Introduction
In recent years, the incidence of twin pregnancies has increased worldwide [1]. In Japan, the twin birth rate has remained around 1%, but has risen to approximately 1.2%, largely due to the increasing use of assisted reproductive technology (ART) associated with delayed childbearing [1, 2]. Given these ongoing demographic and reproductive trends, the incidence of twin pregnancies is likely to remain stable or continue to rise.
Twin pregnancies are associated with a higher risk of obstetric complications and adverse perinatal outcomes, including hypertensive disorders of pregnancy (HDP), gestational diabetes mellitus (GDM), preterm birth, fetal growth restriction (FGR), and postpartum hemorrhage, compared with singleton pregnancies [3, 4, 5]. Consequently, twin gestations require closer surveillance and more careful clinical management.
Laboratory assessments, including complete blood count (CBC) and biochemical parameters, are essential components of prenatal care, and accurate reference intervals are critical for appropriate interpretation. Physiological adaptations in twin gestations—including greater plasma volume expansion and increased metabolic demand—may differ from those observed in singleton pregnancies [5]. Nevertheless, laboratory reference intervals derived from singleton pregnancies are often applied in clinical practice because large‐scale studies establishing clinically applicable reference intervals for twin pregnancies remain scarce.
The aim of this study was to establish clinically applicable reference intervals for hematological and biochemical parameters in a Japanese twin pregnancy cohort. These findings may contribute to improved interpretation of laboratory results and optimization of perinatal management in twin pregnancies.
2. Methods
2.1. Study Design
We conducted a retrospective cohort study using outpatient blood test data from pregnant women who delivered at Fukuda Hospital, a perinatal referral center in Kumamoto, Japan. The hospital is equipped with a 9‐bed maternal–fetal intensive care unit (MFICU) and a 21‐bed neonatal intensive care unit (NICU) and manages low‐ to moderate‐risk pregnancies, with approximately 4000 deliveries annually.
This study was approved by the Institutional Review Board of Kumamoto University (approval number 2906). Given the retrospective design and use of anonymized clinical data, the requirement for written informed consent was waived. Participation was ensured through an opt‐out approach in accordance with the Ethical Guidelines for Medical and Health Research Involving Human Subjects in Japan. Eligible individuals were informed of the study and provided the opportunity to decline participation. The data were used strictly within the scope of the approved research protocol, and any secondary use requires additional ethical approval.
2.2. Study Population
All women with twin pregnancies who delivered at Fukuda Hospital between January 1, 2013, and March 31, 2023, and who had at least one outpatient blood test (CBC or biochemical test) during pregnancy were eligible for inclusion. The following were excluded: women with singleton or higher‐order multiple pregnancies, those without blood testing at the hospital's outpatient clinic, samples with unknown gestational age at blood collection, and samples with missing or blank test results.
2.3. Data Collection
Maternal demographic and clinical characteristics were extracted from electronic medical records, including maternal age, height, pre‐pregnancy weight, body mass index (BMI), parity, gestational age at blood sampling, mode of delivery, gestational age at delivery, and the presence of GDM, HDP, and preterm delivery. Pre‐pregnancy weight and height were self‐reported at the first prenatal visit. Diagnoses of GDM and HDP were established according to the contemporaneous Japanese clinical practice guidelines at the time of delivery. These diagnoses were retained as recorded in the medical records and were not retrospectively reclassified.
2.4. Statistical Analysis
The following hematological and biochemical parameters were analyzed: white blood cell count (WBC), red blood cell count (RBC), hemoglobin (Hb), hematocrit (Ht), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), platelet count (PLT), random plasma glucose, aspartate aminotransferase (AST), alanine aminotransferase (ALT), triglycerides (TG), total cholesterol (TC), and uric acid (UA). Continuous variables such as maternal age, height, weight, and BMI were summarized as mean ± standard deviation (SD).
Hematological and biochemical parameters were categorized into two gestational periods (< 20 weeks and ≥ 20 weeks) based on the distribution of sampling time points in our dataset to ensure adequate sample size and statistical stability within each group, particularly in late gestation where the number of samples was limited. Laboratory measurements were obtained during routine outpatient prenatal care. Because laboratory testing was performed according to routine clinical practice, some women contributed laboratory measurements obtained at more than one gestational time point during pregnancy. All available outpatient laboratory measurements were categorized according to gestational age at sampling and included in the corresponding gestational period for analysis. Sampling time was summarized as median [interquartile range (IQR)] due to non‐normal distribution. Because most laboratory parameters showed skewed distributions, values were summarized as median [IQR], along with the 2.5th and 97.5th percentiles. These percentiles and their 95% confidence intervals (CIs) were estimated using a smoothed nonparametric bootstrap method with 5000 resamples. To approximate continuous distributions and avoid tied percentile estimates resulting from discrete laboratory values, small random noise (±0.05) was added during resampling. The percentile method was used to derive 95% CIs from the bootstrap distributions.
For contextual comparison, singleton pregnancy data for hematological parameters and biochemical parameters were obtained from our previously published cohort [6, 7]. These cohorts consisted of women with singleton pregnancies who delivered at the same institution during the same study period, and laboratory data were collected and analyzed using the same methodology as in the present study. These data were reclassified into two gestational periods (< 20 weeks and ≥ 20 weeks) to match the twin pregnancy categorization. Comparisons between singleton and twin pregnancies were performed using the Wilcoxon rank‐sum test. As a supplementary analysis, both singleton and twin pregnancy data were additionally categorized into three gestational age groups (< 20, 20–29, and ≥ 30 weeks) to further explore gestational trends.
Statistical analyses were performed using GraphPad Prism 10.6 (GraphPad Software, La Jolla, CA, USA), except for bootstrap resampling, which was conducted in R (version 4.5.1; R Foundation for Statistical Computing, Vienna, Austria) using the boot package. A two‐sided p‐value < 0.05 was considered statistically significant.
3. Results
3.1. Cohort Description
Between 2013 and 2023, a total of 37 245 births occurred at Fukuda Hospital after 22 weeks of gestation. Among them, 734 women with twin pregnancies who underwent at least one prenatal blood test at the hospital's outpatient clinic were included in the analysis (Figure 1). The total number of blood samples analyzed was as follows: WBC (N = 1397), RBC (N = 1398), Hb (N = 1400), Ht (N = 1398), MCV (N = 1308), MCHC (N = 1399), PLT (N = 1400), glucose (N = 837), AST (N = 970), ALT (N = 946), TG (N = 970), TC (N = 970), and UA (N = 970).
FIGURE 1.
Study flowchart describing the selection of samples analyzed.
The mean maternal age was 32.5 ± 4.9 years, the mean pre‐pregnancy BMI was 21.5 ± 3.7 kg/m2 (Table 1). The mean gestational age at delivery was 36.3 ± 2.0 weeks, and 286 (39.0%) women delivered preterm. GDM and HDP were diagnosed in 138 (18.8%) and 66 (9.0%) women, respectively. Cesarean delivery was performed in 687 (93.6%) cases.
TABLE 1.
Clinical characteristics of women with twin pregnancies.
| Clinical demographics | Values | Valid cases, n |
|---|---|---|
| Twin pregnancies, n (%) | 734 (100) | 734 |
| Maternal age, years | 32.5 ± 4.9 | 734 |
| Height, cm | 158.6 ± 5.4 | 714 |
| Pre‐pregnancy weight, kg | 54.1 ± 10.1 | 715 |
| Pre‐pregnancy body mass index | 21.5 ± 3.7 | 714 |
| Primipara, n (%) | 339 (46.2) | 734 |
| Gestational age of delivery, weeks | 36.3 ± 2.0 | 734 |
| Preterm delivery, n (%) | 286 (39.0) | 734 |
| Gestational diabetes mellitus, n (%) | 138 (18.8) | 734 |
| Hypertensive disorders of pregnancy, n (%) | 66 (9.0) | 734 |
| Cesarean delivery, n (%) | 687 (93.6) | 734 |
Note: Data are presented as mean ± SD, or n (%), as appropriate. For variables with missing data, the “Valid cases” column indicates the number of cases with available data. Percentages are calculated based on available cases.
3.2. Reference Intervals
Figures 2 and 3 illustrate the distributions of hematological and biochemical parameters across the two gestational periods (< 20 weeks, ≥ 20 weeks). Tables 2 and 3 present the corresponding distributional statistics, expressed as median [IQR] and 2.5th and 97.5th percentiles with 95% CIs.
FIGURE 2.
Scatter plots of hematological parameters across gestational age. The number of samples (N) is shown in each panel.
FIGURE 3.
Scatter plots of biochemical parameters across gestational age. The number of samples (N) is shown in each panel.
TABLE 2.
Reference intervals of hematological parameters during twin pregnancy.
| Parameter | Pregnancy period | |
|---|---|---|
| < 20 weeks | ≥ 20 weeks | |
| Sampling time, weeks, median [IQR] | 12.0 [9.0–13.0] | 25.1 [24.4–26.7] |
| WBC (103/μL) | ||
| Median [IQR] | 8.3 [7.1–9.8] | 8.3 [7.3–9.8] |
| 2.5th (95% CI) | 4.7 (4.4–5.2) | 5.3 (4.9–5.6) |
| 97.5th (95% CI) | 13.0 (12.2–13.4) | 13.3 (12.7–13.9) |
| Samples, n | 651 | 746 |
| RBC (106/μL) | ||
| Median [IQR] | 4.04 [3.81–4.30] | 3.52 [3.30–3.73] |
| 2.5th (95% CI) | 3.32 (3.28–3.39) | 2.92 (2.89–2.96) |
| 97.5th (95% CI) | 4.77 (4.67–4.83) | 4.20 (4.13–4.23) |
| Samples, n | 652 | 746 |
| Hb (g/dL) | ||
| Median [IQR] | 12.1 [11.5–12.9] | 10.5 [9.9–11.2] |
| 2.5th (95% CI) | 9.9 (9.8–10.1) | 8.7 (8.5–8.9) |
| 97.5th (95% CI) | 14.1 (13.9–14.3) | 12.4 (12.2–12.5) |
| Samples, n | 652 | 748 |
| Ht (%) | ||
| Median [IQR] | 35.0 [33.0–36.8] | 31.0 [29.2–32.7] |
| 2.5th (95% CI) | 29.0 (28.5–29.5) | 26.0 (25.3–26.6) |
| 97.5th (95% CI) | 40.2 (39.7–40.5) | 35.9 (35.6–36.4) |
| Samples, n | 651 | 747 |
| MCV (fL) | ||
| Median [IQR] | 86.6 [84.2–88.6] | 88.5 [85.4–91.6] |
| 2.5th (95% CI) | 76.1 (74.4–78.4) | 77.8 (76.6–79.2) |
| 97.5th (95% CI) | 93.2 (92.4–94.1) | 97.4 (96.5–98.4) |
| Samples, n | 623 | 685 |
| MCH (pg) | ||
| Median [IQR] | 30.3 [29.3–31.2] | 30.3 [28.7–31.6] |
| 2.5th (95% CI) | 25.3 (24.5–26.2) | 25.0 (24.4–25.6) |
| 97.5th (95% CI) | 33.0 (32.7–33.2) | 33.9 (33.4–34.3) |
| Samples, n | 652 | 747 |
| MCHC (g/dL) | ||
| Median [IQR] | 34.8 [34.2–35.5] | 34.1 [33.3–34.8] |
| 2.5th (95% CI) | 32.7 (32.6–33.0) | 31.6 (31.5–31.8) |
| 97.5th (95% CI) | 36.7 (36.6–36.9) | 36.1 (35.9–36.3) |
| Samples, n | 651 | 748 |
| PLT (103/μL) | ||
| Median [IQR] | 240 [211–272] | 227 [198–270] |
| 2.5th (95% CI) | 160 (151–164) | 144 (139–149) |
| 97.5th (95% CI) | 345 (335–355) | 355 (340–369) |
| Samples, n | 651 | 749 |
Note: Values are presented as median [Q1, Q3], 2.5th and 97.5th percentiles with 95% confidence intervals (CI), or n, as appropriate.
TABLE 3.
Reference intervals of biochemical parameters during twin pregnancy.
| Parameter | Pregnancy period | |
|---|---|---|
| < 20 weeks | ≥ 20 weeks | |
| Sampling time, weeks, median [IQR] | 9.1 [8.0–10.1] | 25.0 [24.4–25.7] |
| Plasma glucose (mg/dL) | ||
| Median [IQR] | 90 [84–98] | 92 [83–105] |
| 2.5th (95% CI) | 76 (71–77) | 73 (70–74) |
| 97.5th (95% CI) | 130 (123–144) | 133 (128–138) |
| Samples, n | 336 | 501 |
| Sampling time, weeks, median [IQR] | 13.0 [12.6–13.6] | 25.0 [24.4–25.7] |
| AST (U/L) | ||
| Median [IQR] | 15 [13–18] | 14 [13–17] |
| 2.5th (95% CI) | 11 (10–12) | 10 (10–11) |
| 97.5th (95% CI) | 40 (27–46) | 27 (24–29) |
| Samples, n | 316 | 654 |
| ALT (U/L) | ||
| Median [IQR] | 16 [12–21] | 13 [11–17] |
| 2.5th (95% CI) | 9 (7–9) | 7 (6–8) |
| 97.5th (95% CI) | 57 (44–75) | 30 (27–33) |
| Samples, n | 312 | 634 |
| TG (mg/dL) | ||
| Median [IQR] | 126 [101–165] | 216 [177–270] |
| 2.5th (95% CI) | 69 (64–74) | 115 (111–121) |
| 97.5th (95% CI) | 263 (232–327) | 446 (391–498) |
| Samples, n | 316 | 654 |
| TC (mg/dL) | ||
| Median [IQR] | 177 [163–199] | 257 [231–289] |
| 2.5th (95% CI) | 127 (120–137) | 187 (174–195) |
| 97.5th (95% CI) | 249 (235–255) | 367 (349–382) |
| Samples, n | 316 | 654 |
| UA (mg/dL) | ||
| Median [IQR] | 2.9 [2.5–3.3] | 3.2 [2.7–3.6] |
| 2.5th (95% CI) | 1.6 (1.4–1.8) | 1.8 (1.6–2.0) |
| 97.5th (95% CI) | 4.3 (4.0–4.4) | 4.9 (4.6–5.2) |
| Samples, n | 316 | 654 |
Note: Values are presented as median [Q1, Q3], 2.5th and 97.5th percentiles with 95% confidence intervals (CI) or n, as appropriate.
RBC, Hb, and Ht levels were lower at ≥ 20 weeks compared with < 20 weeks. PLT levels also showed a modest decrease in median values with advancing gestation. For example, the median Hb levels decreased from 12.1 g/dL at < 20 weeks to 10.5 g/dL at ≥ 20 weeks, and the 2.5th percentile declined from 9.9 to 8.7 g/dL. In contrast, median WBC, MCH, and MCHC levels showed minimal differences between gestational periods, although slight upward shifts were observed in the lower and upper percentile limits of WBC levels. MCV levels increased modestly at ≥ 20 weeks.
When applying the anemia thresholds proposed by the Royal College of Obstetricians and Gynecologists (RCOG) [8], which were established for singleton pregnancies (Hb < 11.0 g/dL in early pregnancy and < 10.5 g/dL thereafter), 6.9% (45/652) of women before 20 weeks and 47.2% (353/748) at ≥ 20 weeks met the criteria for anemia in this twin cohort.
Glucose, AST, and ALT levels showed minimal changes between gestational periods. In contrast, TG and TC levels increased substantially at ≥ 20 weeks, with median TG levels rising from 126 mg/dL at < 20 weeks to 216 mg/dL at ≥ 20 weeks, and median TC levels from 177 to 231 mg/dL. UA levels also increased modestly at ≥ 20 weeks. Compared with CBC measurements, glucose measurements (n = 336 and 501 for < 20 and ≥ 20 weeks, respectively) were less frequent because plasma glucose was routinely assessed only at the first prenatal visit and again at 24–28 weeks, whereas other biochemical parameters (AST, ALT, TG, TC, and UA) were primarily measured at 12–16 and 24–28 weeks according to the institutional testing protocol.
3.3. Comparison With Singleton Pregnancies
For contextual interpretation, hematological and biochemical parameters in twin pregnancies were compared with those from our previously published singleton cohort (Table 4). The median sampling time at ≥ 20 weeks differed between cohorts (35.7 weeks in singletons vs. 25.1 weeks in twins).
TABLE 4.
Comparison of hematological and biochemical parameters between singleton and twin pregnancies (< 20 vs. ≥ 20 weeks).
| Pregnancy period | Singleton pregnancy | Twin pregnancy | p | |
|---|---|---|---|---|
| Sampling time, weeks, median [IQR] | < 20 weeks | 13.10 [12.60, 13.70] | 12.00 [9.00, 13.00] | |
| ≥ 20 weeks | 35.70 [25.30, 36.40] | 25.10 [24.40, 26.67] | ||
| WBC (103/μL) | < 20 weeks | 8.10 [6.90, 9.34] | 8.30 [7.07, 9.80] | 0.002 |
| n = 18 533 | n = 651 | |||
| ≥ 20 weeks | 8.27 [7.10, 9.60] | 8.34 [7.30, 9.80] | 0.058 | |
| n = 51 221 | n = 746 | |||
| RBC (106/μL) | < 20 weeks | 4.01 [3.79, 4.23] | 4.04 [3.81, 4.30] | 0.020 |
| n = 18 526 | n = 652 | |||
| ≥ 20 weeks | 3.71 [3.50, 3.92] | 3.52 [3.30, 3.73] | < 0.001 | |
| n = 51 216 | n = 746 | |||
| Hb (g/dL) | < 20 weeks | 12.10 [11.50, 12.70] | 12.10 [11.50, 12.85] | 0.253 |
| n = 18 526 | n = 652 | |||
| ≥ 20 weeks | 11.20 [10.60, 11.70] | 10.50 [9.90, 11.20] | < 0.001 | |
| n = 51 221 | n = 748 | |||
| Ht (%) | < 20 weeks | 34.60 [32.90, 36.30] | 35.00 [33.00, 36.80] | 0.039 |
| n = 18 534 | n = 651 | |||
| ≥ 20 weeks | 32.70 [31.10, 34.40] | 31.00 [29.20, 32.65] | < 0.001 | |
| n = 51 224 | n = 747 | |||
| MCV (fL) | < 20 weeks | 86.70 [84.30, 89.00] | 86.60 [84.20, 88.60] | 0.320 |
| n = 17 279 | n = 623 | |||
| ≥ 20 weeks | 88.40 [85.20, 91.40] | 88.50 [85.40, 91.60] | 0.425 | |
| n = 47 948 | n = 685 | |||
| MCH (pg) | < 20 weeks | 30.40 [29.40, 31.30] | 30.30 [29.30, 31.20] | 0.115 |
| n = 18 532 | n = 652 | |||
| ≥ 20 weeks | 30.30 [28.80, 31.60] | 30.30 [28.70, 31.60] | 0.941 | |
| n = 51 224 | n = 747 | |||
| MCHC (g/dL) | < 20 weeks | 34.90 [34.30, 35.60] | 34.80 [34.20, 35.50] | 0.055 |
| n = 18 529 | n = 651 | |||
| ≥ 20 weeks | 34.10 [33.30, 34.80] | 34.10 [33.30, 34.80] | 0.775 | |
| n = 51 219 | n = 748 | |||
| PLT (103/μL) | < 20 weeks | 23.00 [20.10, 26.40] | 24.00 [21.20, 27.20] | < 0.001 |
| n = 18 535 | n = 651 | |||
| ≥ 20 weeks | 22.50 [19.30, 26.00] | 22.70 [19.80, 27.00] | 0.009 | |
| n = 51 212 | n = 749 | |||
| Sampling time, weeks, median [IQR] | < 20 weeks | 9.60 [8.70, 10.70] | 9.10 [8.00, 10.10] | |
| ≥ 20 weeks | 25.60 [24.70, 33.40] | 25.00 [24.40, 25.70] | ||
| Plasma glucose (mg/dL) | < 20 weeks | 90.00 [84.00, 98.00] | 90.00 [84.00, 98.00] | 0.112 |
| n = 19 543 | n = 336 | |||
| ≥ 20 weeks | 90.00 [82.00, 102.00] | 92.00 [83.00, 105.00] | 0.034 | |
| n = 25 078 | n = 501 | |||
| Sampling time, weeks, median [IQR] | < 20 weeks | 13.10 [12.60, 13.70] | 13.00 [12.60, 13.60] | |
| ≥ 20 weeks | 25.60 [24.90, 33.40] | 25.00 [24.40, 25.70] | ||
| AST (U/L) | < 20 weeks | 14.00 [13.00, 16.00] | 15.00 [13.00, 18.00] | < 0.001 |
| n = 18 513 | n = 316 | |||
| ≥ 20 weeks | 15.00 [13.00, 17.00] | 14.00 [13.00, 17.00] | 0.650 | |
| n = 29 826 | n = 654 | |||
| ALT (U/L) | < 20 weeks | 14.00 [11.00, 17.00] | 16.00 [12.00, 21.00] | < 0.001 |
| n = 17 967 | n = 312 | |||
| ≥ 20 weeks | 13.00 [11.00, 17.00] | 13.00 [11.00, 17.00] | 0.467 | |
| n = 28 673 | n = 634 | |||
| TG (mg/dL) | < 20 weeks | 110.00 [86.00, 143.00] | 126.00 [101.50, 164.25] | < 0.001 |
| n = 18 515 | n = 316 | |||
| ≥ 20 weeks | 190.00 [147.00, 248.00] | 216.00 [177.00, 268.75] | < 0.001 | |
| n = 29 826 | n = 654 | |||
| TC (mg/dL) | < 20 weeks | 174.00 [157.00, 194.00] | 177.00 [163.00, 199.00] | 0.018 |
| n = 18 517 | n = 316 | |||
| ≥ 20 weeks | 240.00 [215.00, 269.00] | 257.00 [231.00, 288.75] | < 0.001 | |
| n = 29 823 | n = 654 | |||
| UA (mg/dL) | < 20 weeks | 2.90 [2.50, 3.40] | 2.90 [2.50, 3.30] | 0.334 |
| n = 18 497 | n = 316 | |||
| ≥ 20 weeks | 3.30 [2.80, 3.90] | 3.20 [2.73, 3.60] | < 0.001 | |
| n = 29 801 | n = 654 |
Note: Values are presented as median [Q1, Q3].
Among hematological parameters, WBC levels were largely comparable between singleton and twin pregnancies at both gestational periods. In contrast, RBC, Hb, and Ht levels were significantly lower in twin pregnancies at ≥ 20 weeks (all p < 0.001). PLT counts differed only modestly between singleton and twin pregnancies and varied according to gestational age.
For biochemical parameters, glucose levels were largely comparable between singleton and twin pregnancies across gestation. AST and ALT levels were slightly higher in twin pregnancies at < 20 weeks, whereas no significant differences between singleton and twin pregnancies were observed at ≥ 20 weeks. In contrast, TG and TC levels were consistently higher in twin pregnancies and demonstrated more pronounced elevations at ≥ 20 weeks. UA levels were slightly lower in twin pregnancies at ≥ 20 weeks, although the absolute difference between groups was small.
In supplementary analyses using three gestational categories (< 20, 20–29, and ≥ 30 weeks), the results for the 20–29 and ≥ 30 week groups are presented in Table S1. In the 20–29 week group, RBC, Hb, and Ht levels were lower in twin pregnancies than in singleton pregnancies, whereas TG and TC levels were higher. Similar patterns were observed in the ≥ 30 week group. Sampling times for blood testing were comparable between singleton and twin pregnancies in the 20–29 week group, whereas a difference in sampling timing remained in the ≥ 30 week group. Overall, the findings were broadly consistent with the main two‐category analysis.
4. Discussion
In this study, we established clinically applicable reference intervals for complete blood count and biochemical parameters in one of the largest reported cohorts of Japanese twin pregnancies. Twin gestations were characterized by a marked decline in Hb and Ht after 20 weeks of gestation, as well as substantial increases in TG and TC. Compared with singleton pregnancies, differences were particularly evident for Hb and lipid parameters, suggesting distinctive hematologic and metabolic adaptations in twin gestations.
The decline in Hb after mid‐gestation warrants particular attention because it directly affects the clinical interpretation of anemia in twin pregnancies. The progressive reduction in Hb observed in our cohort is consistent with previous reports describing similar trends in twin gestations [9, 10, 11]. Maternal plasma volume increases progressively during pregnancy, rising by approximately 20% by 20 weeks and reaching nearly a 50% increase by late gestation [12, 13]. In twin pregnancies, plasma volume expansion may be further augmented due to greater placental mass and the metabolic demands of two fetuses. When applying the RCOG criteria [8], nearly half of women after 20 weeks met the definition of anemia. Previous studies suggest that a substantial proportion of anemia in twin pregnancies represents dilutional rather than iron‐deficiency anemia, particularly when Hb levels range between 10.0 and 10.5 g/dL [10, 14], and such mild anemia has not consistently been associated with adverse perinatal outcomes [15]. Furthermore, Dera‐Szymanowska et al. demonstrated no significant differences in maternal iron status, including ferritin concentrations, between singleton and twin pregnancies [16]. These findings are consistent with our observation that hemoglobin concentrations were lower, whereas changes in MCV were comparable between singleton and twin pregnancies, supporting physiological hemodilution rather than more severe iron deficiency as the primary mechanism. Together, these findings suggest that reduced Hb levels in twin pregnancies should be interpreted in the context of physiological hemodilution rather than automatically considered pathological anemia.
WBC counts, PLT counts, and several biochemical markers frequently used in the clinical evaluation of preterm labor and HDP showed relatively limited differences between singleton and twin pregnancies. WBC counts were largely comparable between singleton and twin pregnancies in the main analysis, although additional gestational stratification suggested lower WBC counts in twins in late gestation, consistent with previous observations [9]. PLT showed a modest decrease after 20 weeks, consistent with previous reports [9, 11], which may reflect gestational thrombocytopenia associated with plasma volume expansion. Similarly, AST, ALT, and serum UA demonstrated minimal variation across gestation between singleton and twin pregnancies. Although some previous studies have suggested higher uric acid levels in twin pregnancies, differences between singleton and twin pregnancies were small in our cohort [17, 18]. Plasma glucose levels were comparable between singleton and twin pregnancies before 20 weeks of gestation, whereas slightly higher values were observed in twin pregnancies after 20 weeks, although the magnitude of the difference was small. These findings suggest that laboratory markers commonly used for the evaluation of inflammation, HDP, and GDM may follow broadly similar trajectories in singleton and twin pregnancies, and that interpretation of these indices in clinical practice may not require substantial adjustment for plurality.
TG and TC increased progressively with advancing gestation and were higher in twin pregnancies than in singleton pregnancies, consistent with previous studies [11, 19, 20]. During pregnancy, increasing insulin resistance and enhanced lipolysis promote hepatic lipid synthesis, leading to substantial elevations in circulating lipid concentrations [21]. These metabolic adaptations may be further accentuated in twin pregnancies to meet the nutritional demands of two fetuses. Such enhanced lipid metabolism may represent an important physiological mechanism supporting fetal growth in twin gestations.
Overall trends in our cohort were broadly consistent with prior studies, although some differences in absolute reference limits were observed. International differences in laboratory reference intervals have been widely documented and may reflect variations in ethnicity, genetic background, nutritional status, and healthcare environments [22, 23]. These population‐specific factors may also influence laboratory parameters in twin pregnancies.
Several limitations should be acknowledged. First, this was a retrospective single‐center study based on outpatient laboratory data. Because the mean gestational age at delivery in this cohort was approximately 36 weeks, some women delivered before the routine outpatient blood test typically performed around 36 weeks of gestation. In addition, although preoperative laboratory tests were routinely performed before cesarean delivery, these data were not included in the database used for this analysis. Consequently, laboratory measurements immediately preceding delivery were not fully captured. Second, the cohort included all twin pregnancies encountered in routine clinical practice rather than strictly uncomplicated pregnancies. While this approach may introduce the influence of obstetric complications, it also enhances clinical applicability by reflecting real‐world twin populations. Likewise, all available outpatient laboratory measurements obtained during routine antenatal care were included according to gestational age at sampling rather than limiting the analysis to a single observation per participant. Although this approach resulted in some women contributing laboratory measurements at more than one gestational time point, it better reflects routine outpatient practice and is consistent with the objective of establishing clinically applicable reference intervals. Third, although supplementary analyses were conducted to better align gestational age categories between singleton and twin pregnancies, sampling times were not perfectly matched between cohorts, particularly after 30 weeks of gestation. Furthermore, although more granular gestational categorization may provide additional insight into physiological changes, we prioritized statistical robustness in estimating reference intervals.
A major strength of this study is the relatively large sample size compared with most previous reports of twin pregnancies. Earlier studies examining hematologic changes in twin gestations have typically included fewer than 100 cases [9, 17, 18]. A recent large international study reported laboratory data from more than 2000 twin pregnancies; however, several commonly used parameters, including WBC, MCH, MCHC, AST, UA, and glucose, were not evaluated in that analysis [11]. The present study therefore complements existing evidence by providing reference intervals for a broader range of hematologic and biochemical markers in twin pregnancies.
In conclusion, we established clinically applicable reference intervals for hematological and biochemical parameters in a Japanese twin pregnancy cohort. Twin pregnancies demonstrated characteristic changes across gestation, particularly reductions in hemoglobin after mid‐pregnancy. These findings may facilitate more appropriate interpretation of laboratory values and support optimized clinical management of twin pregnancies.
Author Contributions
Masaru Kobayashi: writing – review and editing. Risa Shimokawa: conceptualization, writing – original draft, formal analysis, methodology, writing – review and editing. Akihito Sagara: writing – review and editing. Saori Yoshimura: writing – review and editing. Yutaka Iwagoi: writing – review and editing. Munekage Yamaguchi: writing – review and editing. Yasuhiro Yamamoto: writing – review and editing. Shoichi Kawakami: formal analysis, writing – review and editing. Eiji Kondoh: conceptualization, writing – review and editing, methodology, formal analysis, supervision, writing – original draft. Fumitaka Saito: writing – review and editing.
Funding
The authors have nothing to report.
Ethics Statement
Ethical approval was obtained from the institutional review board of Kumamoto University (approval number 2906). Informed consent was obtained through an opt‐out mechanism, in accordance with the Ethical Guidelines for Medical and Health Research Involving Human Subjects in Japan, whereby participants were provided with information regarding the study and given the opportunity to decline participation.
Consent
Written informed consent was waived due to the retrospective nature of the study, and consent was obtained through an opt‐out mechanism. No patient identifiable data are included in this article.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Comparison of hematological and biochemical parameters between singleton and twin pregnancies in the 20–29 and ≥ 30 weeks gestational age groups.
Acknowledgments
We thank Mr. Jun Yamada from the Data Management Department of Fukuda Hospital for compiling perinatal medical data into an Excel dataset.
Data Availability Statement
The data supporting the findings of this study were obtained from institutional electronic medical records at Fukuda Hospital. Access to these data is strictly regulated under the approval of the institutional ethics committees of Kumamoto University (approval number 2906). Due to ethical and legal restrictions, including the conditions of ethical approval, data use agreements, and the protection of personal information under Japanese law, the underlying individual‐level data cannot be shared publicly or uploaded to a data repository. Secondary use of the data by third parties is not permitted without additional approval from the relevant institutional ethics committees. Aggregated data supporting the findings of this study are included within the article and its Supporting Information.
References
- 1. Monden C., Pison G., and Smits J., “Twin Peaks: More Twinning in Humans Than Ever Before,” Human Reproduction 36 (2021): 1666–1673, 10.1093/humrep/deab029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Ministry of Health, Labour and Welfare (Japan) , “Vital Statistics of Japan: Live Births by Singleton and Multiple Births (Confirmed Data, 2024),” e‐Stat: Portal Site of Official Statistics of Japan. (Serial online), (2025), https://www.e‐stat.go.jp/stat‐search/files?stat_infid=000040320412.
- 3. Santana D. S., Silveira C., Costa M. L., WHO Multi‐Country Survey on Maternal and Newborn Health Research Network , et al., “Perinatal Outcomes in Twin Pregnancies Complicated by Maternal Morbidity: Evidence From the WHO Multicountry Survey on Maternal and Newborn Health,” BMC Pregnancy and Childbirth 18 (2018): 449, 10.1186/s12884-018-2082-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Walker M. C., Murphy K. E., Pan S., Yang Q., and Wen S. W., “Adverse Maternal Outcomes in Multifetal Pregnancies,” BJOG: An International Journal of Obstetrics and Gynaecology 111 (2004): 1294–1296, 10.1111/j.1471-0528.2004.00345.x. [DOI] [PubMed] [Google Scholar]
- 5. Norwitz E. R., Edusa V., and Park J. S., “Maternal Physiology and Complications of Multiple Pregnancy,” Seminars in Perinatology 29 (2005): 338–348, 10.1053/j.semperi.2005.08.002. [DOI] [PubMed] [Google Scholar]
- 6. Kawakami K., Risa S., Saito F., et al., “Pregnancy‐Specific Reference Intervals for Complete Blood Count Parameters in a Japanese Cohort,” Journal of Obstetrics and Gynaecology Research 51 (2025): e70011, 10.1111/jog.70011. [DOI] [PubMed] [Google Scholar]
- 7. Shimokawa R., Saito F., Kawakami K., et al., “Pregnancy‐Specific Reference Intervals for Maternal Biochemical Parameters and Associations With Adverse Pregnancy Outcomes in a Japanese Cohort,” Journal of Obstetrics and Gynaecology Research 52 (2026). [DOI] [PubMed] [Google Scholar]
- 8. Pavord S., Daru J., Prasannan N., et al., “UK Guidelines on the Management of Iron Deficiency in Pregnancy,” British Journal of Haematology 188 (2020): 819–830, 10.1111/bjh.16221. [DOI] [PubMed] [Google Scholar]
- 9. Sekiguchi M., Muromoto J., Ogawa K., Ozawa K., Miyasaka N., and Sago H., “Longitudinal Hematological Change in Uncomplicated Twin Pregnancies: The White Blood Cell Count Decreases Through Pregnancy After the First Trimester,” Taiwanese Journal of Obstetrics & Gynecology 63 (2024): 381–386, 10.1016/j.tjog.2023.12.001. [DOI] [PubMed] [Google Scholar]
- 10. Townsend K., Maric T., Steer P. J., and Savvidou M. D., “Maternal Hemoglobin Drop in Multiple Pregnancy Is Associated With Higher Gestational Age at Birth and Birthweight,” Acta Obstetricia et Gynecologica Scandinavica 103 (2024): 2002–2012, 10.1111/aogs.14907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Liu S., Shi H., He S., Hu L., and Zhao Y., “Differences in Physiological Indicators Between Twin and Singleton Pregnancy,” International Journal of Gynaecology and Obstetrics 169 (2025): 663–670, 10.1002/ijgo.16097. [DOI] [PubMed] [Google Scholar]
- 12. Pritchard J. A., “Changes in the Blood Volume During Pregnancy and Delivery,” Anesthesiology 26 (1965): 393–399, 10.1097/00000542-196507000-00004. [DOI] [PubMed] [Google Scholar]
- 13. Lund C. J. and Donovan J. C., “Blood Volume During Pregnancy. Significance of Plasma and Red Cell Volumes,” American Journal of Obstetrics and Gynecology 98 (1967): 394–403. [PubMed] [Google Scholar]
- 14. Shinar S., Shapira U., and Maslovitz S., “Redefining Normal Hemoglobin and Anemia in Singleton and Twin Pregnancies,” International Journal of Gynaecology and Obstetrics 142 (2018): 42–47, 10.1002/ijgo.12506. [DOI] [PubMed] [Google Scholar]
- 15. Shi H., Chen L., Wang Y., et al., “Severity of Anemia During Pregnancy and Adverse Maternal and Fetal Outcomes,” JAMA Network Open 5 (2022): e2147046, 10.1001/jamanetworkopen.2021.47046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Dera‐Szymanowska A., Filipowicz D., Misan N., et al., “Are Twin Pregnancies at Higher Risk for Iron and Calcium Deficiency Than Singleton Pregnancies?,” Nutrients 15 (2023): 4047, 10.3390/nu15184047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Fischer R. L., Bianculli K. W., Hediger M. L., and Scholl T. O., “Maternal Serum Uric Acid Levels in Twin Gestations,” Obstetrics and Gynecology 85 (1995): 60–64, 10.1016/0029-7844(94)00343-c. [DOI] [PubMed] [Google Scholar]
- 18. Fischer R. L., Weisberg L. S., and Hediger M. L., “Etiology of Third‐Trimester Maternal Hyperuricemia in Nonpreeclamptic Twin Gestations,” Obstetrics and Gynecology 97 (2001): 62–65, 10.1016/s0029-7844(00)01106-6. [DOI] [PubMed] [Google Scholar]
- 19. Wei J., Shen N., Shi C., et al., “Exploration of Serum Lipid Levels During Twin Pregnancy,” Journal of Maternal‐Fetal & Neonatal Medicine 36 (2023): 2254891, 10.1080/14767058.2023.2254891. [DOI] [PubMed] [Google Scholar]
- 20. Huang J., Meng X., Li J., et al., “Serum Lipid Reference Values Recommended During a Twin Pregnancy and Evaluating Its Association With Perinatal Outcomes,” BMC Pregnancy and Childbirth 24 (2024): 18, 10.1186/s12884-023-06110-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Wild R. and Feingold K. R., “Effect of Pregnancy on Lipid Metabolism and Lipoprotein Levels,” in Endotext, ed. Feingold K. R., Adler R. A., Ahmed S. F., et al. (MDText.com, Inc, 2000), https://www.ncbi.nlm.nih.gov/books/NBK498654/. [PubMed] [Google Scholar]
- 22. Abbassi‐Ghanavati M., Greer L. G., and Cunningham F. G., “Pregnancy and Laboratory Studies: A Reference Table for Clinicians,” Obstetrics and Gynecology 114 (2009): 1326–1331, 10.1097/AOG.0b013e3181c2bde8. [DOI] [PubMed] [Google Scholar]
- 23. Ichihara K., Ozarda Y., Barth J. H., et al., “A Global Multicenter Study on Reference Values: 1. Assessment of Methods for Derivation and Comparison of Reference Intervals,” Clinica Chimica Acta 467 (2017): 70–82, 10.1016/j.cca.2016.09.016. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Comparison of hematological and biochemical parameters between singleton and twin pregnancies in the 20–29 and ≥ 30 weeks gestational age groups.
Data Availability Statement
The data supporting the findings of this study were obtained from institutional electronic medical records at Fukuda Hospital. Access to these data is strictly regulated under the approval of the institutional ethics committees of Kumamoto University (approval number 2906). Due to ethical and legal restrictions, including the conditions of ethical approval, data use agreements, and the protection of personal information under Japanese law, the underlying individual‐level data cannot be shared publicly or uploaded to a data repository. Secondary use of the data by third parties is not permitted without additional approval from the relevant institutional ethics committees. Aggregated data supporting the findings of this study are included within the article and its Supporting Information.