Primary Trial Objective: To test whether rapid correction of vitamin D status early in the neonatal period results in greater lean mass in infancy.
It is hypothesized that neonates with low vitamin D status and provided a vitamin D supplement of 400 IU/d will have lower accretion of lean mass by 3 mo of age and thereafter by 12 months compared to infants given 1000 IU/d for the trial; and that this difference would be sustained up to 3 years after the trial at the post-trial follow-up phase.
Secondary Trial Objectives are designed a priori to test:
1. Whether correction of low vitamin D status early in infancy using 1000 IU/d of vitamin D normalizes body composition (lean, fat, bone mass) and biomarkers across infancy; and for the follow-up (also referred to follow-on) study up to 3 y of age;
2. Epigenetic mechanisms related to vitamin D metabolism and body composition that clarify how improving vitamin D status with 1000 IU/d of supplementation regulates the musculoskeletal system.
Tertiary Objective: outside of the trial, establish a healthy reference group with good neonatal vitamin D stores (\>50 nmol/L 25(OH)D) and given 400 IU vitamin D/d with all of the same measurements as the trial since these are not all available (e.g. IGF-1, IGFBP3 using same assays).
The primary end-point is 12 months of age for the trial. The post-trial follow-up (also known as follow-on) study is up to 3 years with data obtained at 2 and 3 years of age.
The secondary outcomes include other variables: body composition including whole body fat mass, whole body bone mineral content, lumbar spine bone mineral content and density along with biomarkers related to vitamin D status, growth and bone.
Neonates from hospitals in the greater Montreal area will be studied. Neonates will be screened for low vitamin D status within 48 h of birth for 25(OH)D to enable rapid entry into study groups. Those with serum 25(OH)D \<50 nmol/L will be randomized to 400 or 1000 IU/d until 1 y of age. Those with 25(OH)D \>50 nmol/L will be provided standard of care (400 IU/d) and form a reference group.
Treatment period: Trial is 1 week to 12 months of age; follow-up to 3 years of age.
Frequency and duration of follow up: 1 week (baseline), 3, 6 and 12 months of age for the trial; then for the follow-on (also referred to as follow-up study) 2 and 3 years of age for the follow-up visits.
At baseline (1 wk), 3, 6, 12 for the trial plus post-trial at 24 and 36 mo of age, infants will be seen at our research unit for measurement of anthropometry, body composition and bone as well as blood sampling, skin pigmentation and surveys for diet, developmental milestones, activity and demographic information.
Anthropometry - All measurements in infancy will be obtained nude and for 24 to 36 mo the child wearing standardized light clothing, dry diaper and no shoes. Weight will be taken using an electronic scale with a dynamic weighing program (Mettler-Toledo Inc., Switzerland). Length (0.1 cm) will be measured using an infantometer until 24 mo of age (O'Learly Length Boards, Ellard Instrumentation Ltd., US) and height at 36 mo will be measured using a stadiometer (Seca Medical Scales and Measuring Systems, US). Head circumference will be measured (0.1 cm) using a non-stretchable tape (Perspective Enterprises, US) to complete the anthropometric panel. Weight-for-age, height-for-age, and BMI-for-age Z-sores will be calculated using WHO software (WHO AnthroPlus, Switzerland). Mother's weight and height will be measured, while she is still breastfeeding, wearing standard clothing using a calibrated balance-beam scale (Detecto, Webb City, MO, USA) and a wall-mounted stadiometer (Seca model 226; new manufacturer Ayrton226 Hite-Rite Precision Mechanical Stadiometer) to calculate BMI.
Body Composition Measurements - Body composition will be assessed using a fan-beam DXA (APEX version 13.3:3, Hologic 4500A Discovery Series, Bedford, MA). Each infant will wear a light sleeper with no metal or plastic components and a diaper and be scanned using the infant whole body software; at 24 and 36 mo standardized light clothing will be worn and scans captured using whole body software. Whole body scans provide lean mass (g and % of weight), fat mass (g and %), BMC and BMD. Lumbar vertebra 1-4 and forearm BMC and BMD will be captured. Mother's body composition will be measured, while she is still breastfeeding, using BIA (Foot-to-foot tetrapolar Tanita TBF-310, Tanita Corp., Tokyo, Japan) as a rapid assessment.
Biochemistry Measurements - Capillary blood samples (0.5 ml) will be collected at screening; but venous sampling (1.0 ml: yields \~500-600 μl serum) used thereafter; samples will be centrifuged (2235 x g for 20 min at 4°C) to obtain serum (for biochemistry) and buffy coat (white blood cells for DNA) and stored frozen at -80°C until analysis. One 5 ml sample will be taken from parents at baseline (fasting) for measurement of serum 25(OH)D and buffy coat saved for future epigenetic work. For infant screening and maternal serum, total 25(OH)D will be measured using a dedicated auto-analyzer in the PI's laboratory (25 μl; 150 μl "dead volume" that is recovered, Liaison Diasorin Inc.); this assay will also be used for safety assessments at 3 and 6 mo, but is not to be used in analyzing the trial data outcomes as it does not capture all of the metabolites. Liquid chromatography tandem mass spectrometry (LC-MS/MS by Dr. Jones', Queen's Univ.) will be used to measure of 25(OH)D3, 3-epi- 25(OH)D3, and 24,25(OH)2D for all time-points from baseline to 36 mo. In addition, 1,25(OH)2D (100 μl serum) will be similarly measured using an adapted LC-MS/MS method. Both laboratories are certificated by the Vitamin D External Quality Assessment Scheme and will continue to participate in the National Institute of Standards and Technology quality assurance program. Blood-ionized calcium will be measured immediately using our portable unit (ABL80 FLEX Radiometer Medical A/S, Denmark) and compared to published standards (90). Remaining sample will be used for IGF-1 (20 μl) using Liaison (Diasorin Inc.); PTH (25 μl; Immutopics Inc CAT#60-3100) and IGFBP3 (20 μl; R\&D Systems CAT#SGB300) will be measured by ELISA. Sample for IGFBP3 will be pre-treated with protease inhibitors prior to storage. Plasma total calcium and phosphate (150 μl) and urinary calcium and phosphate:creatinine will be measured in a spot sample collected at each visit during the trial; Beckman Coulter UniCel DxC600 autoanalyzer. We will reserve remaining sample for later measurement of C-telepeptide, propeptide of type 1 collagen (P1NP) as biomarkers related to bone.
Demographic, Dietary and Activity Surveys - At screening/baseline, parents will be asked to complete a demographic survey regarding their anthropometry, ethnicity and race, income and education using the same descriptors as defined by Statistics Canada. Infant dietary intake over the study period will be assessed using 3-day diet records completed by parents after each visit. While infants are breastfed, milk intake will be assessed by test-weighing of the infant before and after breast feeding for a 24-hour period using a portable electronic scale (Tanita Corporation Inc., US). This will provide nutrient intakes to help explain growth. Dietary intake from other foods is documented using household measurement items and recorded on the 3-day record. All nutrient analysis will be completed using the Nutritionist Pro software version 4.7.0 (Axxya Systems LLC, Stafford, TX) and the most recent Canadian Nutrient File database (Health Canada). At 2 and 3 y of age, the Habitual Activity Estimation Scale Questionnaire (HAES) will be completed by parents for a weekday (Tuesday, Wednesday, or Thursday) and weekend day (Saturday) over the past 2 weeks. Parents divide their child's day into 4 segments (wake-up to breakfast, breakfast to lunch, lunch to dinner, dinner to bedtime). For each time interval, the % time spent in each activity level is used to estimate overall level of physical activity. The 4 activity levels as established by the HAES questionnaire are: "inactive" (lying down, sleeping, resting), "somewhat inactive" (sitting, watching television, activities done mostly sitting down), "somewhat active" (walking, playing with toys), and "very active" (activities that make a child "breathe hard and sweat," like running and skipping). Developmental milestones are surveyed at each visit (e.g. walking). In 2017, the parent-administered general development screening tool was added to the post-trial design of the study. We used Ages and Stages Questionnaire (24, and 36 months) to complement the activity survey.
Skin pigmentation and UVB exposure - Skin color (type) for the infant will be established by taking the average of three measurements at each site for constitutive pigmentation at the inner upper arm and facultative pigmentation (UVB exposure) at the forehead, mid-forearm and lower leg using a spectrophotometer (CM-700d/600d, Konica Minolta, USA). Individual typological angle (ITAo) will be calculated with the L\* and b\* values. Using constitutive pigmentation, infants will be classified into skin types (I-III: white; IV-VI: non-white) based on Fitzpatrick descriptions. Sun exposure, winter travel and use of sun block will also be surveyed. Sun exposure is expressed as a percentage of body surface area (BSA) exposed and then sun index calculated for each child by multiplying the percent BSA exposed by the time spent outside (min/d); this index does not include sun block.
Sample Size and Statistical Analyses Sample size: up to 74 infants per trial group. The minimum sample size (n=46) is based on the primary objective, lean mass (effect size of 0.59, SD of 670 g), an allocation ratio of 1:1, power of 80 and alpha of 0.05.
Intent-to-treat analyses. For the primary outcome, one analysis will be conducted on blinded groups after all infants have completed 12 months of study; and a second analysis again at 3 y. We wish to keep all of the staff blinded across the entire trial. All trial data will be audited and analyzed blinded; unblinding of the trial data will occur only after statistical analyses are complete. Baseline characteristics will be expressed as mean (SD or 95%CI) or median (IQR) depending on normality for continuous data or as proportions for ordinal data (e.g. sex). Continuous data for the primary outcome will be examined using mixed-effects regression model accounting for fixed effects (strata for allocation skin color, sex; plus dosage, time and demographic covariates) and random effects (modeled on participant ID); a similar model will be used for other outcomes. For the follow-on study for data at 2 and/or 3 years we will test whether any benefits as observed by the end of the trial were maintained using a similar mixed-effects model with additional time-points added to the model.
