Preventing Health Disparities During Pregnancy Through Vitamin D Supplementation

Overview

The purpose of this study is to give all mothers the best chance for a healthy pregnancy through vitamin D supplementation. We will study women of diverse racial/ethnic backgrounds who will receive either the current vitamin D standard of 400 IU/day (in the prenatal vitamin) or 4000 IU/day (dose found in previous pregnancy studies to achieve vitamin D sufficiency). This research is sponsored by the W.K. Kellogg Foundation and the Medical University of South Carolina. The purpose of this study is to examine the effectiveness and infection-fighting properties of the body in relationship to vitamin D levels. This study is being done at the Medical University of South Carolina (MUSC) clinics, and will involve approximately 450 volunteers.

This project, comprised of discrete studies, expert evaluation and translation for public health campaign, promotes racial health equity for pregnant women. A strong racial disparity in the US surrounds vitamin D status: African American women have 20-fold and Hispanic women 2.4-fold greater risk of deficiency than Caucasian women. In South Carolina, greater than 70% of African American, 33% of Hispanic and 12% of Caucasian pregnant women meet the Institute of Medicine's (IOM's) 2010 definition of vitamin D deficiency (1-3). Such disparity on the basis of race/ethnicity represents a serious public health issue. Yet, some-including the IOM-argue that vitamin D deficiency minimally affects maternal and fetal health and that no vitamin D supplementation study has shown improved pregnancy outcome (3, 4). Results of our two recently completed vitamin D trials involving 510 pregnant women throughout pregnancy suggest otherwise (5-7). In the NICHD trial, women supplemented with 400 IU vitamin D/day (the amount in most prenatal vitamins) vs. 2000 or 4000 IU/day not only had worse vitamin D status throughout pregnancy, but combined higher risk of comorbidities of pregnancy (gestational diabetes, hypertensive disorders, infection, preterm labor/birth and periodontal inflammation) (7). In trial 2, women at two South Carolina community health centers were supplemented with vitamin D and again, higher vitamin D status was associated with lower comorbidities of pregnancy risk (6, 8). A novel finding from the NICHD trial was that serum levels of active, hormonal vitamin D (1,25(OH)2D) during pregnancy were optimized at twice the level normally observed in non-pregnant women, levels that would be considered toxic in nonpregnant individuals. At no other time in life does this relationship exist. Is the hormone at such levels driving some vital process during pregnancy? If so, what is that process? There are hints as to why 1,25(OH)2D is elevated in pregnant women. During pregnancy, the maternal immune system undergoes drastic changes for fetal protection. For example, if a solid organ composed of 50% mismatched cell markers (antigens) was transplanted into a new host, it would be rejected within hours; however, a fetus that is 50% mismatched with paternal antigens is protected from immune destruction through a process known as immune privilege. During pregnancy the body increases immune suppressive cells, reduces highly activated natural killer and cytotoxic T-cells with capacity to destroy the fetus and still has the capacity to fight off foreign pathogens. While the mechanisms that orchestrate this complex balance are unknown, we hypothesize that due to vitamin D's ability to regulate immune cells, vitamin D is responsible for many immune alterations associated with fetal immune privilege. We further hypothesize that correction of vitamin D deficiency during pregnancy will protect against immune-mediated disorders such as recurrent miscarriage, preterm birth, bacterial infections, periodontal inflammation and gestational diabetes. This premise, tested and confirmed by this project, will lead to public policy changes and equality in vitamin D status during pregnancy. To test the validity of these hypotheses, the goals of this project are as follows: The central goal of this project is to realize racial equity for all pregnant women and their developing babies in the US. To achieve this goal, there are 3 specific objectives: 1. Determine how vitamin D deficiency leads to immune imbalance and subsequent disparities in pregnancy health outcomes; 2. Determine how vitamin D supplementation prevents such imbalance and disparities; and 3. Translate such findings into public health policy. By comparing women of diverse racial/ethnic backgrounds who receive the current vitamin D IOM standard of 400 IU/day compared to 4000 IU/day \[daily dose shown to achieve optimal production of active vitamin D hormone (1,25(OH)2D)\], the clear effects of vitamin D will be realized. Furthermore, these effects of sufficient vitamin D during pregnancy transcend the racial differences of the women. Study Design: This is a randomized, placebo-controlled clinical trial of 450 pregnant women who will be enrolled between 10-14 weeks of gestation and followed until delivery. Following written, informed consent, each mother will be randomized to receive either placebo or 4000 IU/day vitamin D3 plus the standard prenatal vitamin (containing 400 IU vitamin D3). She will be followed monthly for a total of nine study visits. These visits will include the initial recruitment of mother at 10-14 weeks of pregnancy, monthly study visits, delivery study visit, and 18-month post-partum follow-up visit for mother and child. The primary outcome variable is maternal and neonatal health status as a function of maternal and infant vitamin D status. Secondary outcome variables to be analyzed, as a function of maternal vitamin D status during pregnancy will include: T-lymphocyte profile, immune function indicators, neonatal growth, inflammatory cytokine profile, methylation patterns of DNA of both mother and her neonate, 1. Subject Selection Criteria and Recruitment: Any mother (18-45 years of age) who presents to her obstetrician or midwife at the Medical University of SC (MUSC), Charleston, SC; Our Lady of Mercy, Johns Island, SC; or Franklin C. Fetter, Charleston, SC obstetrical facilities within the first 14 weeks after her last menstrual period (LMP) with confirmation of a singleton pregnancy will be eligible for enrollment in the study. Mothers of diverse ethnic backgrounds (African-American, Asian, Caucasian and Hispanic) will be actively recruited. A database will be generated weekly of undelivered patients for the labor and delivery staff continue to alert the staff of impending admissions. 2. Exclusion Criteria: Mothers with pre-existing calcium, uncontrolled thyroid disease, parathyroid conditions, or who require chronic diuretic or cardiac medication therapy including calcium channel blockers will not be eligible for enrollment into the study. Mothers with pre-existing sickle cell disease (not trait only), sarcoidosis, Crohn's disease, or ulcerative colitis may not participate in the study. In addition, because of the potentially confounding effect of multiple fetuses, mothers with multiple gestations will not be eligible for participation in the study. A sub-group of approximately 100 subjects with known diabetes, hypertension, HIV, or morbid obesity (body mass index \> 49) will participate in the study. 3. Subject Enrollment: The potential study subject's primary care provider will make first contact on behalf of the study and refer interested mothers to study personnel for further information and invitation. Upon enrollment into the study, expectant mothers at approximately 10-14 weeks' gestation will be randomized to receive one of two treatment regimens of vitamin D3. They will be randomized to 1 of 2 groups: (1) Group A: 400 IU vitamin D3/day-Standard dose treatment of placebo (0 IU vitamin D3) plus prenatal vitamin (400 IU/day); or (2) Group B: 4,400 IU/day (4,000 IU/2 gummies/day + 400 IU/day in prenatal). If, after 3 months of vitamin D supplementation, a mother is still frankly vitamin D deficient (less than 15 ng/ml), she will receive open label active vitamin D gummies for the remainder of her pregnancy. To attain equal group numbers (n=150) and balanced numbers by racial/ethnic group (African-American, Caucasian and Hispanic, which represent the predominant groups delivering at MUSC), a stratified block randomization will be used. 4. Race/Ethnicity Defined: Each mother will be asked to define the racial/ethnic group to which she belongs: (a) African-American, (b) Caucasian, (c) Hispanic, (d) American Indian, (e) Asian, and (f) other. Because few mothers of American Indian (0.08% of all deliveries) and Asian (1% of all deliveries) descent deliver at MUSC, there is insufficient power to include those groups in this study. In addition, the mother, by self-report, will be asked to define the race/ethnicity of her parents and the father of the baby and his parents. 5. Study Site Population: The Medical University of South Carolina (MUSC) is an urban medical university in Charleston, SC that serves patients predominantly in the Charleston Tri-County area (80%) and 20% from surrounding regions in South Carolina. We perform 2000 deliveries each year. The ethnic composition of our obstetrical population is 57% African American, 36% Caucasian, and 6% Hispanic and 1% Asian. Franklin C. Fetter and Our Lady of Mercy Clinics primarily see Hispanic patients for prenatal care. These Hispanics have prenatal care coverage provided by a Roper Hospital, Roper St. Francis Healthcare, Charleston, SC grant and include approximately 150 deliveries per year at Roper Hospital. MUSC provides care to any woman presenting in labor who requires routine care or who is referred from outlying hospitals due to high-risk pregnancy for which delivery at a Regional Perinatal Center (RPC) is appropriate, regardless of ability to pay. 6. Enrollment and Follow-Up Subjects: The Study Coordinator will be responsible for assisting in screening, enrollment and follow-up. To maintain data quality, the Study Coordinator will review every form to check for basic problems such as missing data. Labor and delivery staff will alert the study coordinator or the Clinical PI's of the admission of any study subjects. Each month the Study Coordinators will generate a report on patient recruitment and retention to be reviewed jointly by the research team. A computer-generated calendar will serve as a reminder to contact recruited patients two days prior to their appointment. Each month the Study Coordinators will generate a report on patient recruitment and retention that will be reviewed by Dr. Wagner. 7. Follow-up Infant Visit: The Study Coordinator will follow up by phone several times from birth to 18 months in order to verify subject's contact information. The infant (accompanied by mother) follow-up visit will take place at MUSC's CTRC Outpatient Clinic around 18 months postpartum. The infant will have blood drawn for vitamin D and will have a dental assessment of the primary teeth, including digital photographic images. With the assistance of Dr. Thomas Hulsey, epidemiologist and co-investigator of this application, sample size calculations were made. All outcome markers are ratio-scaled and normally distributed. Because of the possibility of premature delivery and patient attrition during this study, enrollment of additional subjects over that minimally needed, will provide a cost effective opportunity to conduct additional analyses examining changes in vitamin D associated with diet, seasonality, and maternal medical complications. As one of the major goals of this study is to explore maternal health disparities by ethnicity, the two supplemented groups (400 and 4,400 IU/day) will be balanced by ethnicity (equal numbers of Caucasian, African American, and Hispanic). This sampling distribution will be accomplished by stratified randomization. This sampling frame was selected to provide: (1) sufficient numbers to examine associations with covariates; (2) the volume of initial and follow up samples will be manageable by our laboratory; (3) over-samples to compensate for losses to follow up; and, (4) stabilizes the sophisticated statistical analysis techniques (repeated measures). The sources of research material will be maternal interview, history and physicals at clinic visits, the questionnaires, the lab reports for blood and urine analyses, the dental periodontal exam results, the dental salivary occult blood results, the vaginal swab results, the placental pathology report, the medical records of each mother and her infant following delivery, and the photographic teeth results on the infants. The material/data that is routinely found in a maternal health record during pregnancy (including information on past pregnancies and disease history) will be utilized in the study. There are additional materials and data from questionnaires, blood, urine, saliva, infant teeth, placental tissue and vaginal samples that will be obtained for research purposes only. A master list will be stored with study files accessible for clinical intervention only. The PI and study staff will have access to the master list. All archived samples and samples provided to collaborating laboratories will be identified with study subject numbers only. The DNA samples collected will have no recontact. The study subject's primary care provider(s) will make first contact on behalf of the study so there is no appearance of a breach in confidentiality. Once a subject has expressed interest in this research study, to obtain informed consent, one of the study coordinators will explain to each subject in lay language the purpose, benefits and risks of the investigation. Consent will be given following an informative discussion period and either by reading the Informed Consent to the subject or by allowing the subject to read the consent, then reviewing it with the subject. All study coordinators have taken the Research Coordinator course and have passed their Miami CITI. Every attempt will be made to conduct this study in full compliance for the protection of the subjects. For those women who are not fluent in English and whose first language is Spanish, a Spanish version of the consent form will be used (translated by BiLingo from IRB approved English consent and submitted as an amendment to IRB). Following the subject's verbal agreement to participate in the study, the subject will sign the informed consent; the study coordinator also will sign the consent. A copy of the signed, written informed consent will be given to the study subject. The original informed consents will be placed in locked research file in study coordinators' office, 20 Ehrhardt St, #4. A copy of the informed consent also will be placed in the subject's study binder (located in locked cabinets at same location). A study coordinator note will be placed in Epic identifying an MUSC woman as participating in a vitamin D and pregnancy RCT. All of the investigators have conducted several clinical trials; they are well versed in the issues of patient/subject confidentiality and have completed the Miami CITI course. They are nationally and internationally known in their fields. All study coordinators have completed the Miami CITI and the Research Coordinator course at MUSC. The endeavors of the PI, co-investigators and study coordinators to maintain subject confidentiality will be ensured. All maternal blood, urine, and vaginal samples as well as infant blood and placental tissue will be labeled with the patient's name and medical record number to ensure proper tracking. Each study subject (and infant) then will be assigned a specific study number without any reference to the woman or infant's name, which will be used to enter all data into the web database. We will follow all HIPAA guidelines to maintain protection of patient/subject information. With these precautions in place, there is a remote risk that the samples and data could be linked to the woman and her infant; however, the study number and the name will be kept in a separate file in the locked offices of the study coordinators. The data will be entered in a secured database with only the study number entered, thus ensuring that the number and data will be kept separately from the woman and infant's names. All this information will be kept confidential, and when reported in scientific journals, the information will not have any identifying information regarding study subjects. Should any untoward reaction occur, subjects would receive treatment at the Medical University of South Carolina. The study will cover the costs of laboratory tests; however, additional costs of treatment will be the responsibility of the subject as described in the informed consent. This study meets the guidelines of clinicaltrials.gov as a clinical trial requiring a Data Safety and Monitoring Plan and a Data and Safety Monitoring Committee. Two of the members of the Data and Safety and Monitoring committee are scientists external to MUSC well known in the field of vitamin D metabolism and calcium metabolism. Two of the members are physician scientists external to MUSC, who are also well known in the field of calcium and vitamin D metabolism. A fifth member will be Dr. Tom Hulsey, epidemiologist at MUSC with considerable experience serving on the DSMC of other clinical trials. Dr. Hulsey was the chair of the DSMC of the two prior vitamin D supplementation trials conducted by this study team. He will maintain the database of the study, follow HIPAA guidelines, and conduct ongoing interim analyses to ensure that the risk: benefit analyses remain in favor of benefit to the subjects. Creation of a Data and Safety Monitoring Board (DSMB): This grant application meets NIH policy and Guidelines on the inclusion of a Data and Safety Monitoring Plan for clinical trials mL and http://grants.nih.gov/grants/guide/notice-files/not98-084.html). MUSC as the Institute and Center (IC) of this grant will have a Data and Safety Monitoring Committee (DMSC) in place as outlined in this grant. Two of the members of the DSMC are scientists external to MUSC well known in the field of vitamin D and calcium metabolism. Two of the members are physician scientists external to MUSC who also are well known in the field of calcium and vitamin D metabolism. A fifth member will be Dr. Tom Hulsey, epidemiologist at MUSC with considerable experience serving on the DSMC of other clinical trials (current and completed). He will maintain the database of the study, follow HIPAA guidelines, and conduct ongoing interim analyses to ensure that the risk: benefit analyses remain in favor of benefit to the subjects. B. Conduct ongoing monitoring of interventional trial by those who have appropriate expertise to accomplish the trial's mission: The MUSC clinical lab is required to notify the PI of any critical lab values. The Study Coordinators will be responsible for checking laboratories of the women within 72 hours of their reporting and then entering the laboratory data into the computer. They will be provided with a set of normative laboratory values as a reference. The Clinical PI will review values during the weekly study meeting or will be notified if a value falls outside of the referent value range. All data will be verified independently by the Data Processing Center under the direction of Dr. Hulsey. The DSMC will be notified via telephone and fax with source documents and adverse report sheets if a subject's value falls outside the referent value range. In addition, all adverse events will be reviewed quarterly by the DSMC, whose report will be forwarded to the Kellogg Foundation on a quarterly basis and the IRB yearly. The Investigators will report all Serious Adverse Events by telephone to the IRB and the DSMC; in addition, the IRB, the DSMC, and Kellogg Foundation will receive a written report within 10 days of the Clinical Investigators' knowledge of the Event. In addition, the investigators will generate a quarterly report to the DSMC regarding subject enrollment, subject completion, adverse events and serious adverse events. The DSMC will review the report and a summary letter with their findings sent to the IRB. Interim Data Analyses \& Monitoring: Interim analyses for the evaluation of safety and efficacy will be conducted based on the recommendations of the Data Safety \& Monitoring Board. The DSMC will serve to monitor for safety and efficacy. DSMB reports produced by Dr. Hulsey's team will include summary statistics: on mother recruitment (expected vs. actual), data form quality (completion and timeliness of forms); tracking of data editing; demographics of the randomized mothers; aggregate safety; aggregate efficacy; and related information. The DSMB also will monitor the trial from the standpoint of futility using the techniques of Lan and Wittes (119). First, differences in losses to follow-up and missed visits will be compared between the six comparison groups to detect any potential bias (400 IU group stratified by race: African American, Hispanic and Caucasian; 4000 IU group also stratified by race). The biggest problem with unbalanced losses will be the loss of statistical power. We will aggressively attempt to avoid any losses by monitoring adherence to visits. In the event that there are differential losses between ethnic groups or supplementation groups, we will conduct statistical tests to ascertain whether these are random or biased. Random losses will be reported but should not affect outcomes. Biased, or non-random, losses will be assessed on their potential impact on differences in outcomes. If necessary, we will conduct a propensity score analysis to control for any statistically significant differences in non-random losses. Following that, differences in compliance to supplementation will be compared as an 'intent-to-treat' design. Any significant deviation from protocol will be considered in subsequent analyses. In this section, we will report on: (1) the deviations from protocol; (2) outcomes comparisons using an efficacy design to test for true differences associated with subjects who adhered to protocol; and, (3) outcomes comparison using the intent to treat (effectiveness) design to test for actual differences that might be expected in a non-controlled intervention. This will help determine whether this type of supplementation could be expected to be effective if applied to the general population. Next, vitamin D levels at study entrance (10-14 weeks' gestational age) will establish the baseline for each supplementation group. It is expected that at baseline there should not be any differences between groups. Because we do expect differences between ethnic groups (which accounts for the stratification), mean vitamin D at baseline will be compared by two-way analysis of variance (ethnicity by experimental group). In step 3, using repeated measures analysis of variance, the changes in vitamin D will be described separately for each ethnic group within each of the supplemented groups. This will depict separately for each ethnic group the changes in vitamin D over time as a function of the level of supplementation. This method will reveal how soon after supplementation the changes in blood (from baseline) are detected, the linear change in blood vitamin D over time and, when and if a peak level (or plateau) was attained. Adjusted differences will be examined by use of multiple regression with and without the inclusion of other covariates. Regression coefficients provide the formula for the linear association between vitamin D in blood and time, separately for each ethnic group. In the event that there is a non-linear change over time, appropriate data transformations will occur as necessary. This step also will provide information on the magnitude of importance that covariates play in changes in vitamin D over time within each ethnic group. As an example, we may find that seasonality plays a significant role in changes in vitamin D for Caucasian women but the association is not as strong for African American women. Further, these associations may become similarly unimportant at the higher levels of supplementation. The next series of analyses will compare the differences in vitamin D between supplemented groups over time between ethnic groups. This will be accomplished by use of a two-way (ethnicity and supplementation group) repeated measures analysis of variance. This technique will detect any significant differences between main effects (ethnic groups and supplementation groups) and interaction effects (ethnicity by supplementation) in changes in vitamin D over time. Multiple regression also will be used to examine the changes with and without the use of potential confounders. In secondary analyses, multi-level mixed-effects models will be used to estimate the average individual monthly rate of change in 25(OH)D, compare this rate between dose groups, and explore the effects of covariates on the rate of change. These models include fixed effects for dose group, time, and the group-time interaction, and a random intercept effect, with additional covariate effects as required. Time will be considered a continuous variable, measured in months rather than assuming structured visit occurrences. An unstructured covariance matrix will be assumed. The multi-level modeling approach also will be used to evaluate the longitudinal association between 25(OH)D and the safety parameters calcium, iPTH (log-transformed), phosphorus, and urinary calcium, creatinine, and calcium: creatinine levels. The cumulative occurrence of pregnancy complications such as preterm labor and infection will be compared between dose group levels using logistic regression. All analyses will be performed using SAS (SAS Institute Inc., Cary NC). Participant attrition and missing data: Because the primary endpoint of the study is change in 25(OH)D from baseline to delivery, the primary analysis will be restricted to participants who have remained in the study until delivery and provided a blood sample within 6 weeks prior to delivery, at delivery, or at the post-delivery visit (completers-only analysis). Typically, multiple imputations would be used to impute missing values in support of the favored intention-to-treat analytic approach. Because the multiple imputation model for this analysis would have required variables also measured in the final blood sample, however, it cannot be used to impute cases with a missing final blood sample. Thus, to assess the primary findings' robustness to assumptions about the missing data, we will perform a sensitivity analysis. Missing data will be imputed under the following conditions: cases with missing endpoints that experienced no change in both groups; experienced the group-specific median change observed in completers; experienced no change in the 400 IU group and minimal change in the 4000 IU group. In the secondary analyses using multi-level mixed-effects models for longitudinal modeling, all available data points will be used, as it is not necessary to delete cases with missed timepoints when using this approach.