Preeclampsia, which uniquely manifests during pregnancy after 20 weeks post-gestation, is a severe disorder affecting various systems (American College of Obstetricians and Gynecologists; Task Force on Hypertension in Pregnancy, 2013). As a leading cause of morbidity and mortality during pregnancy around the globe, preeclampsia induces severely elevated blood pressure and proteinuria. Therefore, it is essential to monitor and control the blood pressure of patients in the clinical treatment of preeclampsia (Magee et al., 2008; Abalos et al., 2014), in which anti-hypertensive drugs are widely used (Belfort et al., 1990).

Nifedipine, a blocker of Ca²⁺ channels (Fenakel et al., 1991), is commonly prescribed to preeclampsia patients as a first line anti-hypertensive drug (Fenakel et al., 1991; Magee et al., 2008; Duley et al., 2013). Nifedipine improves renal blood flow, thereby decreasing vascular resistance. It also yields increased urine output by suppressing anti-diuretic hormones. Hence, nifedipine has become a potent drug to effectively control maternal hypertension (Giannubilo et al., 2012). More importantly, it has been proven that nifedipine is safe with patients, in particular pregnant women (Clark et al., 2015). Treatment with nifedipine during the third trimester can effectively alleviate high blood pressure without any serious neonatal or maternal adverse effects (Childress and Katz, 1994).

Recent findings have implicated vitamin D (VD) in the pathogenesis of pregnancy-related hypertension and/or preeclampsia as a potential risk factor (Baker et al., 2010; Robinson et al., 2013; Adela et al., 2017). For example, VD deficiency increases risks of both mild and severe preeclampsia (Baca et al., 2016), causing the incidence of preeclampsia to increase as much as fivefold in Bangladesh population (Ullah et al., 2013). In women at high risk for preeclampsia, maternal VD level in the second trimester was found to be negatively correlated with risk of early-onset preeclampsia and preterm birth at <35 weeks (Gernand et al., 2016; Zhao et al., 2017). In line with these reports, therapies aiming at supplementing VD have yielded promising results in experimental animal hypertensive/preeclampsia models. In a rat hypertensive model without preeclampsia features, VD supplementation improved factors associated with preeclampsia, such as AT1-AA-induced downstream targets, and reduced blood pressure (Faulkner et al., 2017). In a rat model of preeclampsia, VD supplementation restores angiogenic balance and decreases tumor necrosis factor-α (TNF-α), and importantly, reduced blood pressure (Song et al., 2017). Taken together the abovementioned studies likely supported the potential of VD as an anti-hypertensive agent in the treatment of preeclampsia. However, to the best of our knowledge, the effect of VD hasn’t been investigated in a clinical setting.

Recently, the inflammatory process was implicated in the pathophysiology of preeclampsia (LaMarca et al., 2016). In addition, since development of preeclampsia was reported to be caused by defective trophoblastic invasion, probably through increased pro-inflammatory cytokine TNF-α and decreased IL-10 (Royle et al., 2009), and VD was also reported to act as an immune modulator by downregulating TNF-α and upregulating IL-10 (Mora et al., 2008; Akbar and Zacharek, 2011). Therefore in this current trial, we planned to supplement VD into oral nifedipine therapy among women with preeclampsia, and evaluate the treatment outcomes, especially time needed to lower blood pressure and time before another hypertensive crisis, as well as assess potential adverse effects.

From January 2011 to December 2016, 683 primigravid women (age 31–37) with preeclampsia and singleton pregnancy were enrolled in the current trial, among which 81 participants were later excluded. The remaining 602 eligible participants were assigned to two treatment groups in a random manner (Figure 1). 298 patients received nifedipine+VD orally, and 304 patients received oral nifedipine+placebo, a single dosage every 15 min until their blood pressure were equal to or lower than 150/100 mmHg. Anthropometrical parameters, such as maternal age, gestational age at term, heart rate, BMI, and systolic and diastolic blood pressures, are presented in Table 1. The difference between patients from two treatment groups was not statistically significant.

Table 2 summarized the primary endpoints of the two groups. In the group given nifedipine+VD the time required to control hypertension was 41.8 ± 18.3 min; whereas it was 61.1 ± 15.9 min in the group given nifedipine+placebo, markedly longer compared to the experimental group (p = 0.013). Further, group the time before another hypertensive crisis after blood pressure was controlled in the nifedipine+VD group was 8.1 ± 2.2 h, statistically longer than 4.8 ± 2.6 h in the nifedipine+placebo (p = 0.017).

Secondary endpoints following both treatments were also assessed. The number of dosages needed to effectively control blood pressure was greatly lower in patients who received nifedipine+VD compared to nifedipine+placebo group, as demonstrated in Figure 2 (all dose categories p < 0.05). Further, both maternal and neonatal adverse effects, were evaluated at the end of the trial as summarized in Table 3. No statistical difference with regard to maternal adverse effects was observed between the two treatment groups, including vomiting, dizziness, nausea, mild headache, chest pain, maternal tachycardia, hypotension, or shortness of breath. Neither was there any difference in terms of neonatal parameters, such as birth weight and newborn Apgar scores. Moreover, delivery parameters, including placenta weight, mode of delivery, incidences of intensive care unit stay of mother or newborn were also analyzed (Table 4) and yielded no significant difference between the two treatments. No neonatal death was observed in the study.

Serum levels of TNF-α and IL-10 were examined in all patients from both the treatment groups. As shown in Figure 3, in nifedipine+VD group, serum level of TNF-α was significantly reduced (Figure 3A), whereas that of IL-10 was significantly elevated (Figure 3B).

Based on the reported function of VD against hypertension, we included VD in our clinical trial among women with preeclampsia, to examine whether VD could synergize with oral nifedipine to alleviate hypertension in preeclampsia. Our findings demonstrated that use of oral nifedipine and VD combined greatly improved the primary as well as the secondary endpoints, compared to nifedipine and placebo. Specifically, in the nifedipine+VD group, the time to control hypertension was reduced, time before another hypertensive crisis was delayed, and number of dosages needed to control hypertension was lower. Furthermore, we did not observe any serious adverse effects, either maternal or neonatal, following conjunctional use of VD, demonstrating the clinical safety of VD in pregnant patients.

Controlling severe range blood pressure in the management of preeclampsia is essential for reducing both maternal and fetal complications (Rey et al., 1997; Brown et al., 2000; Shi et al., 2015). Currently, nifedipine is commonly used for hypertensive management in patients with preeclampsia (Magee et al., 2008; Duley et al., 2013). Compared to other anti-hypertensive drugs, oral nifedipine is safer, less expensive and more effective in achieving comparable anti-hypertensive effects in preeclampsia (Fenakel et al., 1991; Aali and Nejad, 2002).

Despite of the fact that drugs for alleviating hypertension are widely available, patients with preeclampsia are often faced with the dilemma that effective anti-hypertensive compounds may not be safe for pregnant individuals, particularly for oral medications. Therefore, an oral agent with proved safety and efficacy to assist oral nifedipine therapy will definitely benefit the treatment of severe preeclampsia. VD is a commonly taken as dietary supplement, and reportedly helpful in a number of pregnancy-related complications such as gestational diabetes mellitus (Li and Xing, 2016).

Moreover, the mechanism accountable for the therapeutic actions of VD in preeclampsia remains largely unclear. A recent study has proposed that, in preeclampsia affected pregnancy, deficiencies in VD and long chain polyunsaturated fatty acid metabolism might disturb the one carbon cycle, leading to altered feto-placental growth and development in preeclampsia (Nandi et al., 2017). On the other hand. It was suggested by Reslan colleagues that systemic effects of preeclampsia may be linked to reduced calcium absorption in VD-deficient individuals (Reslan and Khalil, 2010). In addition, since development of preeclampsia was reported to be caused by defective trophoblastic invasion, probably through increased pro-inflammatory cytokine TNF-α and decreased anti-inflammatory IL-10 (Royle et al., 2009). In this context, VD was also reported to act as an immune modulator by downregulating TNF-α and upregulating IL-10 (Mora et al., 2008; Akbar and Zacharek, 2011). Consistent with previous studies, we also found that in patients from the nifedipine+VD group, serum level of TNF-α was significantly downregulated, whereas serum level of IL-10 was greatly upregulated, compared with patients from the nifedipine+placebo group. In this context, the previously established rat model of preeclampsia (Song et al., 2017) could serve as a useful platform to further investigate the mechanism explaining the shorter duration and fewer doses of the combinational therapy to achieve blood pressure control observed in human patients in our current study.

Nevertheless, this study has some limitations. First, the data of adverse maternal effects was based on patient recall rather than objective measures, which limited the significance and accuracy of this data. Second, dose effect of VD on hypertension and cytokine regulation was not adequately investigated in the current study, which also limited the understanding of VD action on the inflammatory process. Third, the expectant management of preeclampsia patients with severe features beyond 34 weeks is may not be the standard of care in all institutions, which limited the generalizability of the data. Last, the study population was entirely comprised of primips without preexisting hypertension or diabetes mellitus, who were treated at an atypical blood pressure threshold of severe hypertension, which poses another potential issue in the generalizability of the study.

In summary, our data from the current randomized, double blinded and placebo-controlled clinical trial is the first report on the potency and safety of VD serving as a potential adjuvant to oral nifedipine to enhance the efficacy of therapies against hypertension in women with preeclampsia. Importantly, this beneficial action of VD is likely mediated by downregulating TNF-α and upregulating IL-10. Future study is warranted to investigate the molecular mechanism of VD action on downregulating TNF-α and upregulating IL-10 during preeclampsia.

Did the experiments and analyzed the data: D-DS, YW, J-JG, and LZ. Wrote the manuscript and designed the study: NW.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.