Objective
Vitamin D deficiency has been linked to the pathogenesis of preeclampsia. Given the demonstrated antiinflammatory function of vitamin D in multiple organ systems including trophoblast cells and placenta, we hypothesized that vitamin D deficiency contributes to the development of preeclampsia through increased inflammation, as indicated by elevated interleukin (IL)-6 concentrations.
Study Design
Plasma samples from a large preeclampsia cohort study were examined in 100 preeclamptic and 100 normotensive pregnant women. Comparisons of vitamin D and IL-6 concentrations used Student t test and χ 2 test or their nonparametric counterparts. A logistic regression model assessed the association among vitamin D, IL-6 concentrations, and preeclampsia risk.
Results
The mean concentration of 25-hydroxyvitamin D was 49.4 ± 22.6 nmol/L in normotensives and 42.3 ± 17.3 nmol/L in preeclamptic women ( P = .01). The median (interquartile range: Q1, Q3) concentrations of IL-6 were 2.0 (1.3, 3.4) pg/mL and 4.4 (2.2, 10.0) pg/mL in the control and preeclampsia groups, respectively ( P < .01). We observed a significant association between IL-6 elevation and preeclampsia (odds ratio, 4.4; 95% confidence interval, 1.8–10.8; P < .01) and between vitamin D deficiency and preeclampsia (odds ratio, 4.2; 95% confidence interval, 1.4–12.8; P = .04). However, there was no association between vitamin D deficiency and IL-6 elevation.
Conclusion
Third-trimester IL-6 elevation and vitamin D deficiency were independently associated with the risk of preeclampsia. We found no evidence to support the hypothesis that vitamin D deficiency alters the pathogenesis of preeclampsia by activation of inflammation as assessed by IL-6 concentration.
Preeclampsia is a multisystemic pregnancy disorder diagnosed clinically by new-onset gestational hypertension and proteinuria. It occurs in 3-5% of all pregnancies worldwide and is a major cause of maternal, fetal, and neonatal morbidity and mortality. Despite recent progress toward the understanding of the pathophysiology of preeclampsia, the disorder remains a challenge with no preventive therapy and effective treatment limited to delivery to terminate pregnancy and the disorder. A current model of the pathophysiology of preeclampsia invokes a 2-stage model : decreased placental perfusion usually secondary to abnormal trophoblastic invasion with consequent failed dilatory remodeling of maternal vessels perfusing the placenta that precedes and results in the clinical manifestations of preeclampsia. Multiple factors have been indicated in the initiation and progression of preeclampsia, including maternal constitutional factors, antiangiogenic factors, endothelial dysfunction, syncytiotrophoblast microparticles, and inflammatory activation.
On the journey of discovering the underlying mechanisms that cause preeclampsia, vitamin D deficiency has been linked with an increased risk of preeclampsia. Maternal vitamin D deficiency is associated with a 5-fold increase in the odds of preeclampsia compared with normotensive controls. Vitamin D is well known for its function in maintaining normal blood concentration of phosphorus and calcium. However, it also has important roles in other cellular responses that could be relevant to preeclampsia. Vitamin D influences inflammatory responses outside of pregnancy. For example, women deficient in vitamin D have higher interleukin (IL)-6 levels after hip fracture repair. In human coronary arterial endothelial cells, pretreatment with vitamin D significantly inhibits the tumor necrosis factor (TNF)-α-induced downstream signaling pathways.
Human trophoblasts both produce and respond to the active form of vitamin D, 1,25-dihydroxycholecalciferol (1,25(OH) 2 D). The concentration of 1,25(OH) 2 D is tightly regulated by the vitamin D activating enzyme 1α-hydroxylase (CYP27B1) and the degradation enzyme 24-hydroxylase (CYP24A1). Both enzymes are expressed in human placenta. The activated 1,25(OH) 2 D mediates its actions through specific vitamin D receptors (VDR), which are expressed in both decidua and trophoblast. Recent pregnancy-related studies indicate that vitamin D inhibits the messenger RNA transcription of inflammatory cytokine genes (TNFα, interferon-γ, and IL-6) in trophoblast cell culture systems. Immune challenge by lipopolysaccharide induces the expression of VDR and CYP27B1 along with cytokines such as IL-6 in placenta; up-regulation of IL-6 is further enhanced in CYP27B1 or VDR knockout mice. Given the demonstrated antiinflammatory function of vitamin D in multiple organ systems including trophoblast cells and placenta, we hypothesized that vitamin D deficiency contributes to the development of preeclampsia through increased inflammation.
Among the inflammatory markers that are increased in preeclampsia, IL-6 has been consistently indicated to be present at higher serum concentrations in preeclamptic patients than in normal pregnant women. Therefore we chose IL-6 as the marker of inflammation in our study. We tested whether the association between vitamin D deficiency and preeclampsia risk is dependent on increased inflammation, as indicated by elevated IL-6 concentrations when vitamin D deficiency is associated with preeclampsia ( Figure 1 ).
Materials and Methods
Subjects
We used banked plasma samples collected as part of a large preeclampsia cohort study and conducted a retrospective study on a total of 100 women with preeclampsia and 100 normotensive pregnant women. Blood samples were collected at ≥24 weeks’ gestation from nulliparous women carrying a singleton pregnancy. The 100 preeclampsia samples were randomly selected from samples collected from women with preeclampsia in a study approved by the University of Pittsburgh Institutional Review Board. These cases were matched by maternal age, gestational age at blood sampling, body mass index, maternal race/ethnicity, and smoking status to a computer-generated randomized control group of 100 uncomplicated pregnancies from the same study.
Preeclampsia was defined as the new onset of hypertension and proteinuria >20 weeks of gestation. Hypertension was defined as systolic blood pressure ≥140 mm Hg and/or diastolic pressure ≥90 mm Hg, and/or an increase of ≥30 systolic and/or an increase of ≥15 diastolic. Proteinuria was defined as urine protein ≥300 mg/24 hours or 2+ dipstick or a protein/creatinine ratio ≥0.3. Among the 100 preeclamptic patients, 3 patients developed blood pressures >30 systolic and/or 15 diastolic without being ≥140 and/or 90. Fifteen also meet at least 1 of the criteria for HELLP syndrome, namely hemolysis (serum total bilirubin concentration of ≥1.2 mg/dL, serum lactate dehydrogenase concentration of ≥600 IU/L, or hemolysis determined by peripheral blood smear), elevated liver function (aspartate aminotransferase ≥70 IU/L), and thrombocytopenia (<100,000/mm 3 ). There were an additional 3 patients who met all criteria for HELLP syndrome.
IL-6 assay
IL-6 was measured by high-sensitivity human IL-6 immunoassay kit (R&D Systems, Abingdon, UK). The minimal detectable concentration was <0.7 pg/mL, and the interassay coefficient of variation was 7%. To ascertain if the highest IL-6 values in preeclamptic women were associated with the lowest vitamin D concentrations, we defined elevated IL-6 as plasma concentrations within the highest quartile of the 100 preeclamptic samples. Concentrations were calculated by averaging values between 2 duplicates. Samples with absorbance readings out of the range of standard curve were repeated with appropriate titration. Two values (1%) were censored due to the lower limit of detection (0.7 pg/mL) and substituted with half of the detection limit for statistical analysis.
Vitamin D assay
Plasma vitamin D concentrations were determined by 25-hydroxyvitamin D [25(OH)D] assay kit (DiaSorin, Stillwater, MN) with interassay coefficient of variation 7-11%. We defined vitamin D deficiency if the serum concentrations were <37.5 nmol/L and vitamin D insufficiency if they were <75 nmol/L. Concentrations were calculated by averaging values between 2 duplicates.
Statistical analyses
Vitamin D concentrations are presented as the means ± SD; IL-6 data are presented as median and interquartile range (Q1, Q3) because the data were substantially skewed. The comparisons between vitamin D and IL-6 concentrations in 2 study groups were conducted using Student t test or Wilcoxon rank sum test as nonparametric counterparts. Proportions of vitamin D deficiency and IL-6 elevation were calculated for both the control and preeclampsia groups, and assessed by χ 2 test. A logistic regression model was used to assess the relationship of vitamin D and IL-6 concentrations with preeclampsia risk. All analyses were performed using software (SAS 9.3; SAS Institute, Cary, NC) assuming statistical significance at P < .05. The prevalence of coexisting IL-6 elevation and vitamin D deficiency in the preeclampsia group was specifically studied. A random association between the 2 variables would yield a frequency equal to the product of high IL-6 frequency and low vitamin D frequency. A difference in the prevalence would indicate a relationship between vitamin D and IL-6 in preeclampsia. With a sample of 100 in each study group, the current study had 80% statistical power to detect a 1.2-fold difference in prevalence of patients with IL-6 elevation and vitamin D deficiency between random vs nonrandom association with the use of a 2-sided test at the .05 significance level.
Results
Demographic and biochemical data are shown in Table 1 . There were no significant differences between normotensive controls and preeclamptic women in maternal age, body mass index, gestational age at blood sampling, ethnicity, or smoking status. All women in this study were nulliparous. Average gestational age at delivery from the preeclampsia group was 4.9 weeks earlier than that of the control group. This disparity was secondary to higher incidence of spontaneous and induced preterm birth in women with preeclampsia.
| Characteristic | All (n = 200) | Nonpreeclamptic (n = 100) | Preeclamptic (n = 100) | P value a |
|---|---|---|---|---|
| Maternal age (y), mean (SD) | 25.7 (5.7) | 25.5 (5.5) | 26.0 (5.9) | .60 |
| BMI (kg/m 2 ), mean (SD) | 26.8 (5.9) | 26.7 (5.6) | 27.0 (6.3) | .76 |
| Gestational age at blood collection (wk), mean (SD) | 34.8 (4.0) | 34.7 (4.0) | 34.9 (4.0) | .68 |
| Gestational age at delivery (wk), mean (SD) | 37.4 (3.8) | 39.9 (1.1) | 35.0 (4.0) | < .01 |
| Maternal race/ethnicity, n (%) | .87 | |||
| Caucasian | 168 (84.0) | 84 (84.0) | 84 (84.0) | |
| African American | 31 (15.5) | 16 (16.0) | 15 (15.0) | |
| Asian | 1 (0.5) | 0 | 1 | |
| Smoking status, n (%) | 1.00 | |||
| Yes | 48 (24.0) | 24 (24.0) | 24 (24.0) | |
| No | 138 (69.0) | 69 (69.0) | 69 (69.0) | |
| Unknown | 14 (7.0) | 7 (7.0) | 7 (7.0) |
a Student t test for continuous variables and χ 2 test for categorical variables.
Table 2 outlines the mean concentrations of maternal vitamin D in the normal pregnancy group and the preeclamptic group. Of the 100 controls, the mean concentration of 25(OH)D was 49.4 ± 22.6 nmol/L. Thirty subjects had severe vitamin D deficiency, with 25(OH)D concentrations <37.5 nmol/L; 54 had vitamin D insufficiency, with 25(OH)D concentrations between 37.5-75 nmol/L; the remaining 16 had normal vitamin D concentrations (>75 nmol/L). In the preeclampsia group, the mean concentration of 25(OH)D was 42.3 ± 17.3 nmol/L. Patients who met at least 1 HELLP criteria were noted to have lower vitamin D concentrations (40.1 ± 14.6 nmol/L). Among the 100 preeclamptic patients, 41% had severe vitamin D deficiency, 54% had vitamin D insufficiency, and only 5% had normal vitamin D concentrations. Plasma 25(OH)D concentrations were 14% lower in women with preeclampsia compared to those in controls ( P = .01). We observed a positive association between vitamin D deficiency and an increased risk for preeclampsia ( P = .03) ( Table 2 ).
| Variable | All (n = 200) | Nonpreeclamptic (n = 100) | Preeclamptic (n = 100) | P value a |
|---|---|---|---|---|
| Vitamin D, nmol/L | ||||
| All, mean (SD) | 45.8 (20.4) | 49.4 (22.6) | 42.3 (17.3) | .01 |
| ≥75, n (%) | 21 (10.5) | 16 (16.0) | 5 (5.0) | .03 |
| 37.5-75, n (%) | 107 (53.5) | 53 (53.0) | 54 (54.0) | |
| <37.5, n (%) | 72 (36.0) | 31 (31.0) | 41 (41.0) | |
| IL-6, pg/mL | ||||
| All, median (interquartile range) b | 2.7 (1.7–6.3) | 2.0 (1.3–3.4) | 4.4 (2.2–10.0) | < .01 |
| Elevated, n (%) c | 32 (16.0) | 7 (7.0) | 25 (25.0) | < .01 |
| Normal, n (%) | 168 (84.0) | 93 (93.0) | 75 (75.0) |
a Student t test or Wilcoxon rank sum test for continuous variables and χ 2 test for categorical variables
b Median and interquartile range = (Q1, Q3)
Median (Q1, Q3) concentrations of IL-6 were 2.0 (1.3, 3.4) pg/mL and 4.4 (2.2, 10.0) pg/mL in the control and preeclampsia groups, respectively ( P < .01). Of the nonpreeclamptic women, 7% were classified into the elevated IL-6 group based on the cutoff we defined, which was the upper quartile of women with preeclampsia ( P < .01). Results are presented in Table 2 .
Associations among IL-6, vitamin D concentrations, and the risk of preeclampsia were analyzed through logistic regression ( Table 3 ). The odds for preeclampsia was 4-fold higher with elevated IL-6 (odds ratio [OR], 4.43; 95% confidence interval [CI], 1.82–10.80; P = .001). The odds of preeclampsia was tripled with vitamin D insufficiency (OR, 3.26; 95% CI, 1.12–9.54; P = .038), and quadrupled with vitamin D deficiency (OR, 4.23; 95% CI, 1.40–12.81; P = .038).
| Variable | OR (95% CI) | P value |
|---|---|---|
| Model 1 (IL-6 only) | ||
| High IL-6 (≥10.0) | 4.4 (1.8–10.8) | < .01 |
| Model 2 (vitamin D, nmol/L only) | ||
| Vitamin D ref: normal (≥75) | .04 | |
| Vitamin D insufficiency (37.5-75) | 3.3 (1.1–9.5) | |
| Vitamin D deficiency (<37.5) | 4.2 (1.4–12.8) | |
| Model 3 (IL-6, pg/mL and vitamin D) | ||
| High IL-6 | 4.4 (1.8–1.1) | < .01 |
| Vitamin D ref: normal (≥75) | .05 | |
| Vitamin D insufficiency (37.5-75) | 3.1 (1.0–9.3) | |
| Vitamin D deficiency (<37.5) | 4.2 (1.4–13.1) | |
| Model 4 (IL-6, vitamin D, and interaction) | ||
| High IL-6 (≥10) | 3.75 (0.19–74.06) | .39 |
| Vitamin D ref: normal (≥75) | .07 | |
| Vitamin D insufficiency (37.5-75) | 2.98 (0.92–9.72) | |
| Vitamin D deficiency (<37.5) | 8.91 (0.84–94.42) | |
| High IL-6 (≥10) vitamin D a | .98 |
We used several strategies to test whether low vitamin D was associated with high IL-6 in women with preeclampsia. First, the prevalence of IL-6 elevation in women with vitamin D deficiency was not significantly different from those who were not vitamin D deficient (27.1% vs 22.0%, P = .56) ( Table 4 ). We then calculated the Spearman correlation coefficient between the ranked vitamin D and IL-6 concentrations (ρ = 0.22, P = .03). The absence of a negative correlation did not support a relationship between high vitamin D and low IL-6 in women with preeclampsia. Furthermore, a linear regression model was applied to assess the association between vitamin D and log-transformed IL-6, and confirmed that high IL-6 concentrations were not significantly related to low vitamin D concentrations ( P = .19) ( Figure 2 ). Likewise, we performed these analyses using all 200 subjects including women with or without preeclampsia as well as 100 control subjects, and did not observe any significant association between IL-6 elevation and vitamin D deficiency.
