Impact of two oral doses of 100,000 IU of vitamin D₃ in preschoolers with viral-induced asthma: a pilot randomised controlled trial

Asthma is the most common chronic childhood disease, affecting about 10% of children [1, 2], with preschoolers experiencing the highest rate of emergency department (ED) visits relative to other age groups [2‐4]. Most exacerbations are triggered by viral upper respiratory tract infections (URTIs), particularly in young children [5, 6]. Circulating 25-hydroxyvitamin D (25OHD) has been inversely associated with an increased risk of viral URTIs, asthma severity, inhaled corticosteroid dose, and moderate or severe exacerbations [7‐9], suggesting the potential role for vitamin D in viral infections and asthma. Inadequate dietary intake of vitamin D [10, 11], low use of vitamin D supplements [11, 12], dark skin pigmentation, obesity, and low sun exposure increase the risk of vitamin D insufficiency [13‐15], affecting 40–82% of children living in high-latitude areas [11, 16, 17]. Of concern, is the higher reported rate of vitamin D insufficiency in children with versus without asthma [18‐20]; in Canadian preschoolers with asthma, almost 75% are already vitamin D insufficient in the fall [21].

In a meta-analysis of individual patient data, Martineau and colleagues reported a significantly reduced risk of acute respiratory tract infections in participants receiving vitamin D supplementation [22]. A Cochrane systematic review of nine asthma trials (two adult and seven paediatric) [21], with variable use of inhaled corticosteroids (ICS), reported a statistically significant protective effect of vitamin D supplementation against exacerbations requiring rescue oral steroids or emergency department visits, with high-quality evidence and no heterogeneity [23]. However, the optimal posology of this promising strategy remains to be determined. We have demonstrated a rapid rise in serum vitamin D with the combination of a single bolus dose of 100,000 IU vitamin D₃ and 400 IU daily for 6 months, but no significant group difference at 3 and 6 months compared to a placebo bolus and 400 IU vitamin D₃ daily [21]. We surmised that the daily vitamin D supplementation in both groups attenuated group separation and that a single bolus was inadequate to maintain vitamin D sufficiency during both the fall and winter seasons.

The objective of this study was to determine if two bolus doses of vitamin D₃ supplement_, once in the fall and once in the winter, are associated with a rapid and sustained improvement in serum 25OHD. Second, we wish to provide pilot data on the efficacy and safety of this intervention, before formally testing in an adequately powered trial the efficacy of this strategy to improve health outcomes in preschoolers with viral-induced asthma.

We screened 274 children: 102 were ineligible, primarily due to an insufficient number of URTI (33%), no asthma diagnosis (20%), anticipated difficult follow up (18%), and no recent rescue OCS use (10%). Of the 172 potentially eligible children, 125 declined participation mainly because of the number of blood tests, lack of time to comply to other procedures, and lack of interest. Participants were comparable to non-participants in age and sex (data not shown). There were 47 children randomised to the intervention (N = 23) or control group (N = 24) (Fig. 1).

Most baseline characteristics were similar between groups but some appeared slightly imbalanced, with a greater proportion of male participants, environmental tobacco exposure, use of combination therapy, more school days missed, fewer Caucasians and lower vitamin D dietary intake in the intervention compared to the placebo group (Table 1). With a mean age of 2.9 years, most participants were male, Caucasian, and had persistent asthma and URTI as one of multiple asthma triggers; half were atopic. Participants had experienced significant morbidity in the preceding year and at randomisation, most were prescribed daily ICS with or without adjunct therapy. Barely 13% were taking supplemental vitamin D at baseline at a daily dose of 50–400 UI. The median dietary intake of vitamin D (< 250 IU) was markedly below the recommended 600 IU [34] and did not change significantly throughout the study. Most participants had vitamin D insufficiency (< 75 nmol/L) at randomisation, but none had deficiency (< 25 nmol/L) [35].

The median (25%, 75%) follow up was 7 (6.5, 7.4) months, with 89% retention: one participant from the intervention group withdrew at visit 1, two participants (one each from the intervention and placebo group[s) withdrew at visit 2, before the first and second bolus, respectively (due to fear/inconvenience of blood tests), and two participants from the intervention group withdrew before visit 3 (due to the inconvenience of medical visits). After excluding doses not given due to study withdrawals, full-bolus retention was documented in 94% (85/90) of administered boluses: four patients (two each from the intervention and placebo groups) partially spit five doses. There was no evidence of unblinding among participants, nurses, or physicians (Additional file 3).

There was an overall statistically significant group difference in total serum ∆25OHD over time (57.8 (47.3, 68.2) nmol/L, p < 0.0001), with a significant time (p < 0.0001) and group*time interaction (p < 0.0001) effect (Fig. 2). Ten days following the first and second bolus, the group difference in the adjusted within-patient mean ∆25OHD from baseline was 119.3 (95% CI 105.8, 132.9) nmol/L and 100.1 (95% CI 85.7, 114.6) nmol/L, respectively, whereas the mean ∆25OHD at 3.5 months (6.4 (− 6.9, 19.7) nmol/L) and 7 months (5.3 (− 8.2, 18.9) nmol/L) was not statistically significant. The same held true when comparing crude or adjusted 25OHD serum values at various time points between groups (Fig. 3). All patients in the intervention group became vitamin D sufficient (≥ 75 nmol/L) 10 days after the first and second bolus compared to 48% (p = 0.0003) and 35% (p < 0.0001) of patients in the control group, respectively. Vitamin D sufficiency was maintained at 3.5 months in 63% vs. 39% (p = 0.12) and at 7 months in 56% vs. 36% (p = 0.22) in the intervention versus the control groups, respectively.

Over the 7-month study period, patients experienced an average of 4 URTIs and 2–3 viral-induced asthma exacerbations, with no significant group difference in incidence. URTIs were more frequently reported verbally during monthly contact than documented on diaries. Participants experienced 41 exacerbations requiring rescue OCS (mean: 0.87/patient); more than half of participants made an ED visit and received OCS for asthma. No significant group difference was observed in healthcare utilisation, severity and average duration of asthma episodes, use of β2-agonists, or lost workdays or functional status of caregivers; observations were consistent across data sources (verbal report versus diaries) (Table 2).

Urinary Ca:Cr was normal in all but 9/104 (8.7%) intervention samples (N = 4 patients) and 12/117 (10.3%) placebo samples (N = 8 patients) at baseline or after randomisation, all with borderline abnormal values; no episode of hypercalciuria was associated with hypercalcaemia, decrease in alkaline phosphatase, or elevated (> 225 nmol/L) 25OHD, with one exception where the 25OHD was 231 nmol/L. Six patients in the intervention group had a 25OHD value > 225 nmol/L 10 days after the initial (N = 5 patients: 231–293 nmol/L) or the second (N = 1 patient: 237 nmol/L) bolus; only one episode was associated with borderline hypercalciuria (Ca:Cr 1.02) and none was associated with hypercalcaemia. Overall, 227 adverse health events were reported using MedDRA [36] (intervention, 96; control, 131), the most frequent being infections (N = 88, 39%) and general disorders (N = 47, 21%) (Additional file 4); one serious adverse health event, hospitalisation for pneumonia, occurred in the placebo group.

In this group of high-morbidity preschool children with recurrent viral-induced exacerbation of asthma, two boluses of 100,000 IU vitamin D₃ given once in the fall and once in the winter rapidly raised serum 25OHD and increased overall 25OHD compared to a placebo. However, at 3.5 months and 7 months the residual change from baseline in serum 25OHD was only modest and in both cases was not significantly different from that in the placebo group. This supplemental approach permitted the majority of, but not all, patients in the intervention group to achieve and maintain vitamin D above 75 nmol/L throughout the study.

Children living in countries at high latitude are recommended to take vitamin D supplementation during the fall and winter to maintain serum 25OHD at or above 75 nmol/L for bone health [35]. Despite the recommended average dietary requirement of 600 IU in children ≥ 1 year of age as set by the Institute of Medicine [37], our participants’ dietary intake was less than half the recommended intake, with barely 13% taking a vitamin D supplement, a finding concordant with that of our prior pilot study [21]. The majority of children were already vitamin D insufficient at baseline in the fall, probably as a result of a combination of factors, including low intake, sun protection practices, and skin colour [11, 38]. The lower dietary intake and lower serum 25OHD observed in participants contrasts with that of healthy and population-based Canadian preschoolers [39]. Perhaps the avoidance of milk due to the widespread belief that dairy products increase mucus production [40, 41] or less time spent outdoors because of respiratory symptoms triggered by physical activities or environmental allergies contribute to these findings in our young population in whom half were atopic. Nonetheless, our results indicate that preschoolers with asthma are at high risk of vitamin D insufficiency.

A bolus dose of 100,000 IU vitamin D, once in the fall and once in the winter, is recommended by the French Society of Paediatrics for healthy preschoolers [42] and has been shown to rapidly and safely raise blood 25OHD levels. With recent meta-analyses suggesting a beneficial effect of supplemental vitamin D on reducing the incidence of viral infections [22] and asthma exacerbations [23], a rapid increase in circulating 25OHD may be ideal to efficiently prevent the well-documented paediatric “September epidemic” of viral-induced asthma exacerbations in the Northern hemisphere [43]. Although the optimal posology to prevent asthma exacerbations is unknown, bolus doses previously appeared less protective against respiratory infections than daily doses in children; however, this approach was primarily tested in undernourished children as a means to prevent pneumonia [44, 45]. In contrast, a 2015 meta-analysis of paediatric trials strongly recommended the use of a bolus dose (< 300,000 IU), concluding that daily doses as high as 4000 IU were insufficient to rapidly raise serum 25OHD in vitamin D deficient children [46]. The present trial confirms that each loading dose of 100,000 IU rapidly raises serum 25OHD in predominantly vitamin D insufficient children, but is inadequate to maintain sufficiency in about 40% of children; indeed, only a nominal change from baseline of approximatively 5–6 nmol/L was maintained 3.5 months after each bolus. Our findings of an acute rise in 25OHD following each bolus is concordant with the prior literature [21, 47], as is the return to near baseline levels approximately 3.5 months post bolus [48, 49]. Despite our hope to avoid the need for daily supplementation, two boluses proved suboptimal in many young asthmatic children.

The combination of bolus and daily supplementation would appear more promising. Indeed, in our prior pilot study of asthmatic preschoolers [21], the control group, receiving only 400 IU daily vitamin D (with no bolus), had a slow but steady increase from baseline of 20 nmol/L at 3 months, with no further improvement despite ongoing supplementation; yet, only 55% reached 75 nmol/L after 3 months of supplementation, confirming this approach as suboptimal. In contrast, the intervention group in the same trial, receiving a single 100,000 IU bolus combined with 400 IU daily vitamin D₃, displayed a rapid rise of 132 nmol/L in serum 25OHD within 10 days, with a clinically meaningful change from baseline of 27.1 nmol/L at 3 months, but little additional increase to 6 months. Whereas all patients in the intervention group had maintained serum 25OHD at ≥ 75 nmol/L at 3 months, it dropped to 88% by 6 months, suggesting the need for a second bolus. Admittedly, the target serum level for the immune and anti-inflammatory effect observed with vitamin D supplementation by Martineau and colleagues [22, 23] remains to be established. Collectively, the findings of our two pilot studies would support two loading doses of 100,000 IU spaced 3.5 months apart, combined with a daily dose of 400 IU throughout the fall and winter, as an optimal intervention to achieve a rapid and sustained increase in serum vitamin D in preschoolers with asthma.

No significant group difference was observed in the severity of asthma exacerbations or healthcare utilisation. The non-significant trend observed in several outcomes, which was not in favour of vitamin D, is likely random, but may also be explained by baseline group imbalances (despite adjustment); more children in the intervention group displayed several baseline characteristics associated with more severe asthma and lower vitamin D exposure (skin colour and dietary intake) compared to the control group.

The study contributes to accumulating evidence on the safety profile of 100,000 IU vitamin D boluses. Although several participants displayed elevated urinary Ca:Cr, blood calcium levels were normal in all cases, suggesting that hypercalciuria was perhaps due to an unbalanced diet or a non-fasting state. Moreover, all participants with elevated serum 25OHD had normocalcaemia. To date, only five published paediatric trials have tested an oral dose of 100,000 IU vitamin D, with just two reporting circulating 25OHD: two cases (0.3%) with 25OHD > 375 nmol/L occurred in one trial of malnourished Afghan infants [44] and two children (18%) with 25OHD > 250 nmol/L but no associated hypercalciuria or hypercalcaemia were reported in our prior pilot study [21]. In the three small trials reporting serum bone metabolism biomarkers in the paediatric population, no episodes of hypercalcaemia occurred [47, 49, 50]. Yet, additional safety data in larger samples are required.

We acknowledge several trial limitations and lessons learned. Due to its small sample size, this study resulted in baseline group imbalances and provided only preliminary process and efficacy data. However, in line with our prior trials using similar criteria [21, 50], the high morbidity of enrolled patients, evidenced by the frequency of exacerbations requiring rescue OCS in the preceding 12 months (≥ 2/child) and during the study period (0.87/child), confirmed the appropriateness of eligibility criteria and the ongoing health burden, despite ICS therapy. With two thirds of participants already vitamin D insufficient at baseline and the expected further decline during the winter, the pragmatic recruitment without pre-screening for vitamin D status appears justified. Whereas greater efficacy had been suggested in asthmatic patients with the lowest baseline 25OHD [23], randomising children irrespective of baseline vitamin D would provide the opportunity to explore the dose-response association with, and identify the optimal target for, immune and anti-inflammatory effect. The absence of clinically important change in dietary vitamin D intake over the study period suggests no evidence of Hawthorne bias. Tolerance of the 2-mL bolus was excellent with 94% of administered doses completely retained. Importantly, we documented the main outcome in all participants, despite an 11% dropout, with pre-approved consent to obtain drug and medical information at study endpoint. The twofold higher number of URTI and asthma exacerbations documented by monthly verbal versus diary report, suggests the need to facilitate diary completion, perhaps by offering electronic diaries, shown to improve veracity and adherence [51, 52]. Conducted in a single centre with multi-ethnic Canadian preschoolers with asthma, most of whom were vitamin D insufficient at baseline, the observed serum 25OHD response to two boluses of 100,000 IU may not apply to healthy children, other ethnic groups, or those with significantly higher (or lower) baseline vitamin D status or exposure. These lessons learned have been implemented in the ongoing, funded, large, multicenter, triple-blind, placebo-controlled trial testing two boluses of 100,000 IU with daily 400 IU of vitamin D in the same population (NCT 03365687).

The administration of an oral bolus of 100,000 IU vitamin-D₃ in the fall, with a repeat dose in the winter, rapidly and significantly raises overall serum 25OHD in Canadian preschoolers with high-morbidity asthma, despite suboptimal dietary and sun exposure. While sufficient in slightly more than half of children, this strategy appears inadequate to maintain vitamin D sufficiency over 7 months in a notable proportion of this population.