Randomized Trial: Maternal Vitamin D Supplementation to Prevent Childhood Asthma (VDAART)

Asthma is one of the leading causes of morbidity in children with 60% of all cases diagnosed
by age 3. Thus, finding factors that can lead to prevention of this disease would be of great
public health importance. Vitamin D deficiency is highly prevalent among pregnant women, and
thus, represents a potentially modifiable factor for the prevention of disease. Due to the
effect of vitamin D in modulating immune responses, we believe that vitamin D deficiency in
pregnant mothers leads to faulty immune system development in the neonate, predisposing them
to asthma and allergy. We have observational data from two independent birth cohort studies
that higher maternal intakes of vitamin D during pregnancy are each independently associated
with a 50% reduction in risk for recurrent wheezing and asthma in 3- and 5-yr old children.
However, in order to recommend this as a universal treatment to prevent asthma, a randomized,
controlled, clinical trial is necessary. Therefore, we propose a two arm, double-blind,
placebo controlled, randomized, clinical trial of Vitamin D, to determine whether higher
vitamin D intake and levels in the pregnant mother will prevent asthma and allergy in the
child at age 3. We will identify pregnant women who have asthma or allergies or whose partner
has asthma or allergies, from obstetric clinics in the three clinical centers participating
in this trial. We will recruit 870 pregnant women during the first trimester of pregnancy and
randomize them to one of two treatment arms of a 4-year clinical trial: 4000 IU of Vitamin D
in addition to usual prenatal vitamins, n=435; and usual prenatal vitamins alone, n=435. Our
primary specific aim is to determine whether adequate vitamin D supplementation in the
pregnant mother is associated with reduced incidence of asthma in the child during the first
3 years of life. The sub-aims of the study will include (1) To determine whether adequate
vitamin D supplementation in the pregnant mother is associated with reduced secondary
outcomes in the child of (a) allergic sensitization, (b) doctor's diagnosis of eczema and (c)
lower respiratory tract infections during the first 3 years of life; and (2) To determine
whether adequate vitamin D supplementation in the pregnant mother is associated with improved
vitamin D status in the mothers and their offspring through measurement of 25(OH)D levels in
maternal plasma, cord blood, and children's blood at 1 and 3 yrs of age and (3)To determine
whether sufficient vitamin D supplementation in the pregnant mother is associated with
reduced incidence of preterm birth (birth <37 weeks gestation), preeclampsia, gestational
hypertension, and/or Hemolytic anemia, Elevated Liver enzymes, Low Platelet count (HELLP
syndrome) (PB/PE) compared to a usual care control group in VDAART.

VDAART Analysis Plan (Addendum to Protocol)

All analyses of primary and secondary outcomes stated in the VDAART Protocol will be
performed using the intent-to-treat (ITT) paradigm, unless specifically stated.

I. Definition of outcomes.

A. Asthma and recurrent wheezing. The primary outcome of the trial will be development of a
doctor's diagnosis of asthma and/or recurrent wheeze. While asthma is acknowledged to have
its roots in early life, making a definite diagnosis is difficult until the child is older,
and wheezing illnesses may be precursors of an asthma diagnosis. Thus, we will have a
composite definition for the primary outcome.

- Asthma will be defined as a parent's report of physician's diagnosis at any time during
the first three years of life, based on questionnaire responses. For time-to-event
analyses, the time of incident asthma will be determined by the time of first positive
questionnaire response.

- Recurrent wheezing will be defined as at least one report of wheezing in the third year
of life plus any report of wheezing in any of the first two years of life.

Recognizing that wheezing symptoms may be modified by medication use, variable age of onset,
and other variables, the following will be included in the composite outcome:

- children who had at least one wheezing episode (during the 1st 2 years of life), plus an
asthma controller medication (defined as steroid inhalers or nebulizers OR steroid pills
or liquids OR leukotriene modifiers) in the third year

- children who did not report any wheezing in the first 2 years but who report 2 separate
episodes of wheezing in the third year OR are on controller medication (defined as
above) on at least 2 separate reports in the third year OR at least 1 report of wheezing
and at least 1 separate report of controller medication use in the third year

For recurrent wheezing (and the other wheezing phenotypes), the time of incident recurrent
wheezing will be determined by the time of first report of wheezing, for time-to-event
analyses. In the event that the child meets the definition based solely on the use of
controller medication, the first report of use of this controller medication will be used to
determine the time of onset of wheezing for time-to-event analyses.

For these outcomes, because the questions were asked every 3 months, and recognizing that
missing questionnaires occurred, event times will be treated as interval-censored.

B. Eczema. Eczema will be defined as a positive response to parent's report of physician's
diagnosis at any time during the first three years of life based on the questionnaire
responses.

C. Lower respiratory tract infections (LRI). Lower respiratory tract infections (LRIs) will
be defined based on parental responses to questionnaires. LRIs will be defined as a parental
report of a physician's diagnosis of pneumonia, bronchiolitis, bronchitis, or croup
(laryngotracheobronchitis). LRIs may occur multiple times over a three-year period in a
child. Information on repeated episodes of LRI will be employed for rate estimation and
comparison.

D. Total and specific IgE. Total and specific serum IgE will be measured from cord blood
samples, and from year 3 samples from the children. Total IgE will be analyzed as a
continuous outcome. Sensitization will be defined as any specific IgE ≥ 0.35 IU.

II. Secondary Analyses.

A. Analyses based on levels. Secondary analyses using baseline and follow-up maternal 25OHD
levels will be performed. We will investigate the effect of baseline and follow-up maternal
levels separately on the development of primary and secondary outcomes in the trial.
Additionally, we will categorize mothers into categories based on joint baseline and
follow-up levels. Mothers who have baseline levels and follow-up levels ≥ 40 mg/dl will be
categorized in the HI-HI group; those who have baseline levels and follow-up levels < 40
mg/dl will be categorized in the LO-LO group; mothers who do not meet these definitions will
be categorized either in the HI-LO or LO-HI groups based on their baseline and follow-up
levels.

B. We also performed metabolomics on the children at age 1 and 3 and eventually age 6. We are
assessing the association with vitamin D, asthma, and other phenotypes such as food
allergies, IgE, and lung function with metabolites and genetic variants.

C. As an ancillary study, we have collected stool samples from the mothers in the 3rd
trimester, and samples from the infants and children at 3-6 months, 1 year, and 3 years. We
will sequence the stool samples for the microbiome, and we will assess the effects of the
early life microbiome on the development of asthma, allergies (including food allergies and
sensitization), growth and development, and obesity.

D. In a secondary analyses, we will determine the effect of 17q21 genotype on the efficacy of
vitamin D supplementation in the prevention of asthma/wheeze. We will compute hazard ratios
for the reduction in asthma/wheeze risk associated with prenatal supplementation, stratified
by rs12936231. rs12936231 is a functional SNP influencing expression of ORMDL3, and given the
role of ORMDL3 as a key sphingolipid biosynthesis regulator, we will subsequently investigate
the relative abundance of sphingolipids between those in the vitamin D Intervention arm and
those in the placebo group, stratified by 17q21 genotype. Relative abundance is quantified by
normalizing the peak area of a metabolite to the median peak area of that metabolite within
the same batch on the day the sample was run. The normalized peak area is then median scaled
and imputed with the minimum across all samples. In this way, we can directly compare
abundances relative to other samples for each metabolite. Relative abundance has no unit of
measure. Finally we will identify interactions between prenatal vitamin D supplementation,
rs12936231 genotype and sphingolipid metabolism in the risk of asthma/wheeze by age three. We
will replicate this analysis in the COPSAC2010 population (ClinicalTrials.gov
Identifier:NCT00856947).

VDAART Year 6 Analysis Plan

PRIMARY ANALYSIS: The null hypothesis of no treatment effect will be tested in a Cox
Proportional Hazards Model of time to physician diagnosed asthma or onset of recurrent
wheeze, adjusted for maternal levels of Vitamin D recorded at 10-18 weeks gestation, allowing
for level by treatment interaction, and for any other relevant baseline covariates (race,
maternal education, and clinical center), using the methods of Tsiatis et al. (Stat Med. 2008
October 15; 27(23): 4658-4677) to maximize precision of estimated treatment effect. The
significance level for the test of no treatment effect will be 0.05.

SECONDARY ANALYSES: Effect estimates from all secondary analyses will be reported as point
estimates with nominal 95% confidence intervals. Estimates will be derived from models
adjusted for baseline maternal levels of Vitamin D as noted above.

Current active asthma, late onset recurrent wheeze, allergic sensitization, any allergic
rhinitis, MD-diagnosed food allergy, food allergy with symptoms, eczema: Analysis is confined
to individuals who provide data after age 5. Loss to followup is reasonably balanced between
arms. Logistic regression will be used to test the hypothesis that there is no effect of
treatment on risk of current active asthma. Separate tests will be conducted for no treatment
effect on risk of late onset recurrent wheeze, on risk of allergic sensitization defined
using specific IgE greater than or equal to 0.35IU, and on risk of allergic rhinitis birth to
age 6. Similar models will be used separately to assess treatment effect on risk of eczema,
active eczema, and food allergies.

Lower respiratory infections: The count of LRI per unit time will be modeled using negative
binomial regression with covariates including treatment arm.

Impulse oscillometry: Analysis of treatment effect on impulse oscillometry components will be
based on data available for 6 year old children, and will use linear modeling with adjustment
for maternal baseline Vitamin D. A secondary analysis will examine treatment effect on trends
in these measures through ages 4-6. This test will be based on mixed effects modeling.

Spirometry: Spirometric outcomes will be analyzed using linear models with transformations as
needed, adjusting for race, age, sex and height. Bronchodilator response will be dichotomized
at 10% of pre-challenge values, analyzed with logistic regression.

VDAART Asthma and/or Recurrent Wheeze Definitions for Primary Analysis (Time-to-Event)

Physician-diagnosed asthma Report of physician-diagnosed asthma at any time in the first 6
years of life. The time of onset will be the first report of wheeze OR the first report of
use of asthma medication.

Recurrent wheeze - no diagnosis of asthma Caregiver report of wheeze OR use of any asthma
medication on two separate years over the first 6 years. The time of onset will be the first
report of wheeze OR the first report of use of asthma medication.

VDAART Year 6 Analyses - Definitions for Secondary Outcomes

Current Active Asthma Report of physician-diagnosed asthma at any time in the first 6 years
of life, PLUS after the 5th birthday a) report of child's use of any asthma medication (see
below) OR b) report of wheeze.

Recurrent Wheeze at age 6 yrs - no diagnosis of asthma At least 1 caregiver report of wheeze
OR use of any asthma medication prior to the 3rd birthday, PLUS a report of wheeze OR use of
any asthma medication after the 3rd birthday.

Late onset Recurrent wheeze - no diagnosis of asthma, wheeze, or asthma medication use prior
to 3rd birthday A report of wheezing OR use of any asthma medication after the 5th birthday
PLUS either wheezing or use of any asthma medication after the 3rd birthday but before the
5th birthday (between 3yrs and 5 yrs).

Total and specific IgE. Total and specific serum IgE will be measured year 6 samples from the
children.

1. Total IgE will be analyzed as a continuous outcome.

2. Sensitization will be defined as any specific IgE ≥ 0.35 IU to the panel of allergens.

Allergic Rhinitis. This will be a positive response to the question "Since the last time we
talked has a health care provider said that [CHILD] has hay fever, allergic rhinitis, or
allergies involving the eyes or nose?" Any positive response from birth through age 6.

Eczema.

1. Ever eczema - positive response to "…has a health care provider said that [CHILD] has
eczema?" at any time from birth through age 6.

2. Persistent eczema. Ever eczema at any time in the 6 years plus positive response to
"itchy rash which was coming and going but did not completely get better?" after the 3rd
birthday.

Lower Respiratory Infections (LRI). Lower respiratory tract infections (LRIs) will be defined
based on caregiver responses to questionnaires. LRIs will be defined as a caregiver report of
a physician's diagnosis of pneumonia, bronchiolitis, bronchitis, or croup
(laryngotracheobronchitis). LRIs may occur multiple times over a six-year period in a child.
Information on repeated episodes of LRI will be employed for rate estimation and comparison.

Food Allergies.

1. Doctor-diagnosis - any positive response from birth through age 6 to the question "In
the last year, has a health care provider said that [CHILD] has food allergies?

2. Food allergy with symptoms - Doctor-diagnosis of food allergy PLUS any reaction to
specific foods from the grid.

Lung Function.

1. Impulse oscillometry: R5, R20, AX, X5, R5-20, and R5-20% These variables will be used as
continuous variables. The variables are measured at 4, 5, and 6 years of age. The
primary analysis will look at values at 6 yrs of age, and a secondary analysis will
investigate the trajectory (longitudinal analysis) from age 4 through 6 yrs.

2. Spirometry: pre- and post-bronchodilator FEV1, FVC, FEV1/FVC; Bronchodilator Response
(BDR**). Analyses will adjust for race, age, sex, and height.

- Asthma medication: rescue (inhaled beta-agonists or systemic corticosteroids) or
controller medication (inhaled corticosteroids monotherapy, ICS/Long-acting
beta-agonists, or leukotriene modifiers) ** A positive BDR will be defined as a
change of 10% from the baseline FEV1 after inhalation of albuterol

Inclusion Criteria:

- Personal history of asthma, eczema, allergic rhinitis or a history of asthma, eczema,
allergic rhinitis in the biological father of the child

- Gestational age between 10 and 18 weeks at the time of randomization

- Maternal age between 18 and 39 years

- Not a current smoker

- English or Spanish speaking

- Intent to participate for the full 4 years (through Pregnancy and then until the 3rd
birthday of the child)

Exclusion Criteria:

- Not meeting inclusion criteria

- Gestational age greater than 18 weeks

- Presence of chronic medical conditions

- Taking vitamin D supplements containing more than 2000 IU/day of vitamin D3

- Multiple gestation pregnancy (twins, triplets)

- Pregnancy achieved by assisted reproduction techniques (e.g., IUI, IVF)