Alkaline phosphatase (ALP) is an enzyme that catalyzes the hydrolysis of organic phosphate-esters in the extracellular space [1]. Several isoenzymes are produced by the hepatobiliary system, skeletal system, gastrointestinal tract, kidneys and placenta [2]. The placental isoenzyme, a heat stable homo-dimeric protein [3], is produced by syncytiotrophoblasts. It is the primary contributor to the corresponding rise in total ALP during gestation [4], along with elevated bone-related ALP, secondary to amplified bone formation [5, 6]. These two studies were the first to validate that ALP gradually increases during gestation, but failed to reveal associations with adverse outcomes [5, 6]. Nevertheless, several case reports demonstrated associations between elevated ALP with hypertensive disorders [7], gestational diabetes [7, 8, 9], preterm birth [10], placental abruption and stillbirth [11]; as well as in uncomplicated pregnancies [12, 13]. Additionally, elevated mid-trimester ALP, may predict preterm birth [14, 15] and low birth weight [16, 17, 18].

It was also suggested to be associated with arterial stiffness as a predictor of cardiovascular disease in non-pregnant individuals [19] and multiparous women [20].

Previous ALP nomograms demonstrated a gradual increase in total ALP, starting early in the second trimester, and peaking in the third trimester just prior to delivery, with maximum levels approximately 2–3 times higher than non-pregnant references [21, 22, 23, 24, 25, 26, 27]. Other than their small size, these studies were limited by lack of consideration of possible effect modifiers, such as body mass index, mode of conception, neonatal sex and the foe to-placental mass.

In our clinical practice, total ALP is often measured during biochemistry serum analysis, but is seldom used as a potential biomarker for adverse pregnancy outcomes. As maternal physiological adaptation confers substantial alterations in several laboratory tests, it is imperative to have high-quality pregnancy-specific distributions of ALP during pregnancy, on which to base pathological deviations. Therefore, we aimed to establish a large-scale ALP nomogram, and to address the role of possible physiological modifiers, and of total ALP levels during gestation — all as a foundation for further research, to explore associations of divergent ALP with adverse pregnancy outcomes.

This study was a retrospective analysis of ALP values during pregnancy, among healthy women from a variety of ethnicities, with uncomplicated singleton gestations, who delivered in a single tertiary medical center, from August 2007 to December 2012.

Overall, 44,153 women gave birth to singleton, non-anomalous, liveborn neonates at our center during the study period. Among them, 8718 had at least one available ALP measurement during pregnancy. We excluded women with any pre-existing comorbidity (n = 546), preterm or post-term delivery (n = 803), large or small for gestational age (n = 1553), hypertensive disorders (n = 296), gestational diabetes (n = 474), oligo- or polyhydramnios (n = 575), placental complications (n = 191) and chorioamnionitis (n = 4). Additionally, 20 outlier ALP values, all above 1000, were excluded. Accordingly, 5285 women, with 6610 ALP measurements were included.

Demographic, clinical and obstetrical characteristics of the study group, are presented in Table 1, and compared to the population without an available ALP, meeting the same inclusion and exclusion criteria. Although there were some statistically significant differences between the two populations in gravidity (2.7 $\pm$ 1.8 vs. 2.7 $\pm$ 1.7, p = 0.025), parity (2.2 $\pm$ 1.3 vs. 2.3 $\pm$ 1.3, p 0.001), nulliparity (36.1% vs. 31.8%, p 0.0001), birthweight (3213 $\pm$ 313 vs. 3225 $\pm$ 309 grams, p = 0.015) and mode of delivery; they were not clinically relevant.

The distribution of ALP measurements throughout gestation is presented in Table 2, as a nomogram with mean, standard deviation, minimum and maximum range, as well as 2.5th–97.5th percentiles per gestational week. The means and standard errors according to possible modifiers, are presented in Table 3. The mean ALP distribution is graphically presented in Fig. 1, per gestational week and the median, interquartile range and minimum-maximum values, in Fig. 2, per trimester.

Fig. 1.

Nomogram of mean alkaline phosphatase (IU/L) levels, throughout gestation (━━), per gestational week. Linear regression line is presented (___), as well as the reference (- - -) for non-pregnant women in the study population (116.7 $\pm$ 2.3 IU/L).

Fig. 2.

Box-plot nomogram of median alkaline phosphatase (IU/L) levels throughout pregnancy, per trimester week - with 25th and 75th interquartile ranges, as well as minimum and maximum values.

The association of possible physiological modifiers to ALP levels are presented in Table 3. ALP is significantly elevated during pregnancy compared to a non-pregnant state (132.2 $\pm$ 3.0 vs. 116.7 $\pm$ 2.3 IU/L, p 0.001), and during the third and second trimesters compared to the first trimester (166.4 $\pm$ 1.66, 123.7 $\pm$ 2.4 and 95.7 $\pm$ 1.6 IU/L, p 0.001). ALP levels remained unaffected by the modifiers we explored: neonatal sex, neonatal birthweight and maternal age.

We constructed an ALP nomogram among healthy women during the course of an uncomplicated pregnancy. Our main results indicate that ALP gradually increases during pregnancy, starting from early gestation and that other than gestational week, physiological parameters of maternal age, fetal sex and neonatal birthweight, did not affect the levels.

In accordance with our results, previous studies reported similar trends for both total and heat stable ALP, throughout pregnancy [21, 22, 23, 24, 25, 26, 27]. In a longitudinal study among 52 women, Larsson et al. [24] evaluated 25 common biochemical indices, including ALP. Similar to our results, they demonstrated rising ALP levels through gestation. However, their cohort was smaller and less homogenous, included preterm deliveries and small for gestational age newborns, as well as mothers who were smokers and used medication. Abbassi-Ghanavati et al. [25] established a normal reference range during pregnancy for multiple blood components, derived from a comprehensive literature review. Specifically, for ALP, they reported the upper and lower limits during a non-pregnant state and per trimester, based on 5 prior publications [22, 23, 24, 27, 33], again demonstrating gestational rise of ALP. Selected comparisons of ALP levels between the available results of two studies [24, 25] and our own, are presented in Table 4 (Ref. [23, 24]).

Okesina et al. [21] in a longitudinal study of 67 pregnant women, determined the rise in total ALP and demonstrated that both placental and bone isoenzymes contribute to this increase. Placental ALP appears in serum as early as 13 weeks and becomes a significant contributor beyond 16 weeks, whereas bone ALP is a significant contributor at 32 and 38 weeks.

This is the first study to demonstrate that other than gestational age, possible modifiers of fetal sex, birthweight (as a surrogate for the fetal-placental mass) and maternal age did not affect ALP. This observation contradicts a small-scale study by Gol et al. [26], who compared total and placental ALP, according to fetal sex, among 30 women. Unlike our results they concluded that fetal sex affects ALP levels, across all gestational ages, with ALP significantly higher in mothers carrying a female versus a male fetus.

Establishing a normal reference range is of paramount importance, especially during pregnancy. Maternal adaptation to pregnancy induces profound physiological changes, in all organ systems, resulting in substantial changes in multiple blood analytes. An appropriate, robust, gestational nomogram is important to truly identify normal and pathological deviations. As there are only limited nomograms for pregnancy, and those for ALP are minimal, we constructed a large-scale, week-specific nomogram. Our planned research will use this nomogram to evaluate the association between abnormal ALP and pregnancy complications, all of which were not included in the current study.

We speculate that ALP is released following placental damage. Accordingly, higher than normal ALP may be associated with adverse pregnancy outcomes. This pathogenesis was occasionally supported by compatible histological findings associated with elevated ALP and pregnancy complications, such as placental infarcts [10, 11], abnormal syncytial and villus formation [12], chronic histiocytic intervillositis [34], and uteroplacental vascular disease [35]. Alternatively, others have suggested a different pathway through which placental ALP rises insufficiently in pregnancies complicated with fetal growth restriction and preeclampsia [36, 37], possibly reflecting decreased placental mass or malfunction, and relative ALP deficiency. Interestingly, there is a possible role in this mechanism to insufficient plasma level of the ectopic mineralization inhibitor inorganic pyrophosphate, which is a substrate of alkaline phosphatase [19].

The current study is not without limitations, mainly due to its retrospective design and the lack of some data (body mass index, mode of conception and smoking status), which may theoretically impact ALP levels. Additionally, we analyzed total ALP only and not the direct placental fraction. However, previous studies have demonstrated that placental isoenzymes have a half-life up to 7 days [38], which is the longest of all isoenzymes, and as such, have the most impact on the total ALP [5, 21]. As we included only healthy women, without any chronic or gestational morbidity, we safely assume that the rising levels of ALP represent the placental component and not a pathological source, from either maternal or fetal origin. Comparing the included and excluded population, i.e., those without an ALP measurement, demonstrated that the excluded population was similar to those included, confirming that there was no selection bias.

In conclusion, normal ALP levels associated with pregnancy are higher than in the non-pregnant state, gradually increasing from the end of the first trimester. This should be accounted for when evaluating lab results obtained for pregnant woman. In future studies, this nomogram can serve as the basis for comparison between a normal population and those with various pregnancy complications, to determine how ALP is affected during pathological events, the corresponding placental histopathology and whether ALP has a predictive role for adverse pregnancy outcomes.

NA—Conceptualization; Data curation; Investigation; Methodology; Validation; Supervision; Visualization; Writing - review & editing. Final approval. KNS—Investigation; Methodology; Validation; and data analysis and interpretation. LS, OSA, AB, EK—Data curation, Review and editing, Final approval. EH—Conceptualization; Data curation; Investigation; Methodology; Project administration; Supervision; Validation; Visualization; Roles/Writing - original draft; Writing - review and editing, Final approval. All authors agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

The study was approved by the Institutional Review Board of Rabin Medical Center (Approval No. 0121-13-RMC). Informed consent was waived due the retrospective design of the study.

The authors thank Faye Schreiber, MS for editing the manuscript. Also, we thank all the peer reviewers for their opinions and suggestions.

This research received no external funding.

The authors declare no conflict of interest.