Atopic dermatitis (AD) is a common inflammatory skin disease. It most often develops in early childhood, suggesting that early life factors may be etiologically important. Vitamin D (vitD) seems to play a role in improving immune function and skin barrier, both of which are dysfunctional in AD (Borzutzky and Camargo, 2013). In this study, we aimed to determine the extent to which vitD status in fetal life and early childhood can predict development of AD in early childhood and/or persistence of AD through mid-childhood.
Details of the methods can be found in the Supplementary Material and Methods. Briefly, we based this study on data from a prospective observational pre-birth cohort study in Massachusetts, USA, where women and their children from the general population (unselected for atopic diseases) were followed since pregnancy in 1999–2002 with yearly questionnaires and in-person visits at early and mid-childhood. Our exposures of interest were vitD intake in pregnancy (mean of 1st and 2nd trimesters) and early childhood (median 3.2 years), reported by participants using validated semi-quantitative food frequency questionnaires (Fawzi et al., 2004), as well as 25(OH)D plasma levels in pregnancy (2nd trimester), in umbilical cord blood at birth, and in children in early childhood. We ran models with both continuous and categorical exposure variables. Because of the clinical relevance of low vitD status (e.g. 25(OH)D <25 nmol/L) (Spiro and Buttriss, 2014), and because the three upper categories of maternal 25(OH)D blood levels had similar ORs, exposures were also expressed as dichotomous variables using a cut-off point at 25 nmol/L (=10ng/ml). We examined two outcomes: 1) incident AD in early childhood (birth to 3 years), defined using maternal report of a provider’s diagnosis on annual questionnaires; and among those with early childhood disease, we defined 2) persistent AD in mid-childhood (median 7.7 years) by maternal report of current symptoms in mid-childhood using questions from the validated International Study of Asthma and Allergies in Childhood questionnaire (ISAAC, 1998; Williams et al., 1999)
We examined correlations of exposures using Pearson correlation coefficients (Supplementary Table S1). We tested several covariates for effect modification and confounding, and included the following potential confounders in our final model: child sex, child race/ethnicity, maternal education, parental atopy, and for blood exposures, season of blood draw. We used unadjusted and multivariable-adjusted logistic regression models to estimate odds ratios (OR) of AD.
Of 1418 mother-child participants, 68% of children were white and 71% of mothers had a college education. White children had higher 25(OH)D blood levels and vitD intake in pregnancy and early childhood (Table 1).
Table 1.
Characteristics of 1418 mother/child pairs in Project Viva according to 25(OH)D levels and vitamin D intake
| 2nd trimester 25(OH)D nmol/L n=1089 | Early childhood 25(OH)D nmol/L n=57 | Early childhood vitamin D intake IU/day n=356 | |||||
|---|---|---|---|---|---|---|---|
| Participant characteristics | N (%) | Mean (SD) | p-value* | Mean (SD) | p-value* | Mean (SD) | p-value* |
| Total | 1418 (100) | 59.5 (21.1) | 83.2 (25.8) | 211.6 (102.9) | |||
| Child sex | 0.35 | 0.87 | 0.53 | ||||
| Male | 731 (51.6) | 58.9 (20.1) | 82.7 (27.4) | 215.1 (107.3) | |||
| Female | 687 (48.4) | 60.1 (22.1) | 83.9 (23.7) | 208.1 (98.4) | |||
| Child white | <0.0001 | <0.0001 | 0.04 | ||||
| No | 457 (32.2) | 50.1 (20.3) | 73.6 (26.5) | 197.2 (101.1) | |||
| Yes | 961 (67.8) | 63.3 (20.2) | 92.5 (21.8) | 220.5 (103.2) | |||
| College graduate | <0.0001 | 0.76 | 0.08 | ||||
| No | 413 (29.2) | 53.7 (21.4) | 81.6 (23.6) | 195 (98) | |||
| Yes | 1003 | 61.7 (20.5) | 83.9 (27.1) | 216.9 (104) | |||
| Mother or father atopy | 0.43 | 0.54 | 0.56 | ||||
| No | 576 (40.7) | 58.9 (21.3) | 81 (24) | 207.2 (105.2) | |||
| Yes | 840 (59.3) | 59.9 (20.9) | 85.2 (27.6) | 213.9 (101.8) | |||
| Season 2nd trim blood draw | <0.0001 | ||||||
| Winter | 269 (24.7) | 60.1 (21.7) | |||||
| Spring | 322 (29.6) | 55.7 (20.7) | |||||
| Summer | 242 (22.2) | 63.8 (20.5) | |||||
| Fall | 256 (23.5) | 59.6 (20.6) | |||||
| Season early childhood blood draw | 0.38 | ||||||
| Winter | 12 (21.1) | 76 (27.2) | |||||
| Spring | 12 (21.1) | 77.3 (27.7) | |||||
| Summer | 17 (29.8) | 85 (24.5) | |||||
| Fall | 16 (28.1) | 91.1 (24.6) | |||||
IU: International units, SD: Standard deviation, 25(OH)D: 25-hydroxyvitamin D.
Overall (type 3) p-values. For dichotomous characteristics we used t-test. For characteristics with > 2 categories we used analysis of variance (ANOVA).
By early childhood, 497/1418 (35%) had AD (flow chart, Supplementary Figure S1 online). Odds of incident AD were higher for children of mothers with 25(OH)D <25 vs. ≥25nmol/L (adjusted OR (aOR) 2.74; 95% confidence interval (CI) 1.37, 5.49) (Table 2). Maternal blood 25(OH)D and vitD intake in pregnancy were not well-correlated (r=0.25), and maternal vitD intake during pregnancy was not related to AD (aOR 1.07 for quartile 1 vs. quartiles 2–4; 95% CI 0.82, 1.40) (as reported in Camargo et al., 2007) (Table 2). Analyzing vitD intake from foods and supplements separately did not change the results (data not shown). These inconsistent findings of maternal prenatal 25(OH)D blood levels and vitD intake could reflect that blood 25(OH)D is a better measure of vitD status than intake. However, it is also possible that since 25(OH)D is a marker of sunlight exposure, sunlight exposure itself influences risk; or that dietary questionnaires, even though validated, do not measure vitD intake perfectly. Inconsistencies in previous studies could result from methodological differences, but inconsistencies between dietary studies could also partly be caused by difficulties in teasing out the effects of vitD intake alone, since vitD interacts with other nutritional, environmental and genetic factors.
Table 2.
Associations of 25(OH)D levels and vitamin D intake with atopic dermatitis
| N (cases/eligible)1 | Unadjusted | Adjusted2 | |||
|---|---|---|---|---|---|
| OR (95% CI) | p-value | OR (95% CI) | p-value | ||
| ATOPIC DERMATITIS IN EARLY CHILDHOOD | |||||
|
| |||||
| Second trimester plasma 25(OH)D (nmol/L) | |||||
| Continuous 25(OH)D (per 25 nmol/L decrement) | 378/1089 | 1.06 (0.91, 1.23) | 0.47 | 1.03 (0.88, 1.21) | 0.72 |
| Category of 25(OH)D | 0.04* | 0.03* | |||
| < 25 nmol/L | 22/39 | 2.38 (1.19, 4.76) | 2.46 (1.16, 5.22) | ||
| 25 to < 50 nmol/L | 103/311 | 0.91 (0.63, 1.32) | 0.86 (0.59, 1.27) | ||
| 50 to < 75 nmol/L | 178/526 | 0.94 (0.67, 1.31) | 0.90 (0.64, 1.27) | ||
| ≥ 75 nmol/L | 75/213 | 1.0 (ref) | 1.0 (ref) | ||
| < 25 nmol/L | 22/39 | 2.52 (1.32, 4.81) | 0.01 | 2.74 (1.37, 5.49) | 0.005 |
| ≥ 25 nmol/L | 356/1050 | 1.0 (ref) | 1.0 (ref) | ||
| Cord plasma 25(OH)D (nmol/L) | |||||
| Continuous 25(OH)D (per 25 nmol/L decrement) | 243/691 | 1.23 (0.99, 1.54) | 0.06 | 1.11 (0.86, 1.44) | 0.43 |
| Category of 25(OH)D | 0.21* | 0.66* | |||
| < 25 nmol/L | 29/75 | 2.17 (0.90, 5.21) | 1.57 (0.60, 4.08) | ||
| 25 to < 50 nmol/L | 129/344 | 2.07 (0.95, 4.48) | 1.63 (0.73, 3.65) | ||
| 50 to < 75 nmol/L | 76/232 | 1.68 (0.76, 3.70) | 1.44 (0.64, 3.22) | ||
| ≥ 75 nmol/L | 9/40 | 1.0 (ref) | 1.0 (ref) | ||
| < 25 nmol/L | 29/75 | 1.18 (0.72, 1.94) | 0.50 | 0.99 (0.57, 1.72) | 0.98 |
| ≥ 25 nmol/L | 214/616 | 1.0 (ref) | 1.0 (ref) | ||
| Maternal vitamin-D intake (mean of 1st and 2nd trimester) | |||||
| Continuous vitamin D intake (per 100 IU decrement) | 483/1372 | 1.02 (0.96, 1.09) | 0.50 | 1.02 (0.95, 1.09) | 0.59 |
| Category of vitamin D intake (quartiles) | 0.70* | 0.75* | |||
| Q1 | 126/343 | 1.04 (0.76, 1.42) | 1.02 (0.74, 1.42) | ||
| Q2 | 122/343 | 0.99 (0.72, 1.35) | 1.01 (0.73, 1.39) | ||
| Q3 | 112/343 | 0.87 (0.63, 1.19) | 0.87 (0.63, 1.20) | ||
| Q4 | 123/343 | 1.0 (ref) | 1.0 (ref) | ||
| Q1 | 126/343 | 1.09 (0.85, 1.41) | 0.49 | 1.07 (0.82, 1.40) | 0.64 |
| Q2-Q4 | 357/1029 | 1.0 (ref) | 1.0 (ref) | ||
|
| |||||
| PERSISTENT ATOPIC DERMATITIS | |||||
|
| |||||
| Early childhood plasma 25(OH)D (nmol/L) | |||||
| Continuous 25(OH)D (per 25 nmol/L decrement) | 11/57 | 1.17 (0.62, 2.21) | 0.64 | 1.42 (0.62, 3.25) | 0.40 |
| Category of 25(OH)D | |||||
| < 25 nmol/L | 0/5 | NA | NA | ||
| Early childhood vitamin-D intake | |||||
| Continuous vitamin D intake (per 100 IU decrement) | 56/356 | 1.42 (1.05, 1.91) | 0.02 | 1.37 (1.01, 1.85) | 0.04 |
| Category of vitamin D intake (quartiles) | 0.20* | 0.30* | |||
| Q1 | 20/89 | 1.86 (0.85, 4.08) | 1.78 (0.80, 3.99) | ||
| Q2 | 14/89 | 1.20 (0.52, 2.76) | 1.03 (0.44, 2.44) | ||
| Q3 | 10/89 | 0.81 (0.33, 1.99) | 0.88 (0.35, 2.19) | ||
| Q4 | 12/89 | 1.0 (ref) | 1.0 (ref) | ||
| Q1 | 20/89 | 1.86 (1.01, 3.42) | 0.05 | 1.83 (0.98, 3.41) | 0.06 |
| Q2-Q4 | 39/89 | 1.0 (ref) | 1.0 (ref) | ||
CI: Confidence intervals, IU: International units, NA: Not estimable due to low numbers, OR: Odds ratios, 25(OH)D: 25-hydroxyvitamin D, Q: Quartiles
25(OH)D levels: 25 nmol/L = 10 ng/ml, 50 nmol/L = 20 ng/ml, 75 nmol/L = 30 ng/ml
Numbers are for the unadjusted analyses. The adjusted analyses may have lower numbers due to missing values.
Adjusted for child sex, child race, maternal education, parental atopy, and season of blood draw (for blood exposures)
Overall (type 3) p-value
Additional analyses: 1) The relationships of maternal 25(OH)D levels, maternal vitamin D intake, and cord blood 25(OH)D levels with AD persistence was also explored, but no associations were found (data not shown). 2) We did a sensitivity analysis because the proportion of children with AD in early childhood was high in comparison with cohorts from other industrialized countries (ISAAC, 1998). In this analysis participants who reported “ever having had AD” at one of the 6 month to 2 year assessments, but not at the 3-year assessment were excluded (n=170), but the results did not change (data not shown).
Cord blood 25(OH)D levels, reflecting late fetal exposure, was not strongly related to early childhood AD (Table 2). To test whether this lack of association could result from selection bias in cord blood donation, we restricted the analysis of maternal 25(OH)D and early childhood AD to participants with available cord blood, and we continued to see an association (aOR 6.51 for <25 vs. ≥25 nmol/L, 95% CI, 1.90, 22.34). If avoidance of low vitD status (<25 nmol/L) is more important in early fetal development, it would explain why 25(OH)D blood level in 2nd trimester but not cord blood was related to AD.
Among 389 children with early childhood AD and follow-up to mid-childhood, 58 (15%) had persistent AD (Supplementary Figure S1). Median follow-up was 4.5 years (2.8–7.4). Lower vitD intake in early childhood was associated with higher odds of persistent AD in mid-childhood (aOR 1.37 per 100 IU decrement; 95% CI 1.01, 1.85), even after additional adjustment for maternal 25(OH)D levels (aOR 1.49; 95% CI 1.06, 2.09). Expressing vitD intake as a categorical variable showed similar results (Table 2). We observed comparable odds for lower childhood 25(OH)D levels (aOR 1.42 per 25 nmol/L decrement; 95% CI 0.62, 3.25), but numbers of early childhood blood samples were small (n=57), and confidence limits wide (Table 2). No other studies have previously explored this association, but several have related childhood 25(OH)D blood levels with the presence or severity of AD, most of which suggest a protective effect of higher 25(OH)D (Chiu et al, 2015; Peroni et al, 2011). Supplementary Figure S2 online illustrates our results.
Few randomized, controlled trials have been published in children. Goldring et al. (2013) and Chawes et al. (2014) did not find effects of late pregnancy vitD supplementation on early childhood AD development, whereas Sidbury et al. (2008) and Camargo et al. (2014) found positive results on AD severity.
Strengths of this study are the prospective design, relatively large sample size, long follow-up period, and our ability to control for multiple socio-economic confounding variables. Limitations include defining AD diagnosis by questionnaires, few early childhood 25(OH)D blood samples, and lack of information about sun exposure and filaggrin mutations.
Overall, this study indicates that maternal prenatal 25(OH)D blood levels below 25 nmol/L are associated with a higher risk of AD in early childhood and that lower intake of vitD in early childhood is associated with higher risk of persistent disease through mid-childhood. However, the lack of association in other analyses (e.g. cord blood 25(OH)D and AD) still questions the role of vitD in AD. Data from ongoing intervention trials (clinicaltrial.gov; identifier, NCT00856947 and NCT00920621) is warranted to establish the extent of vitD’s potentially preventive effect on AD, along with the best timing and dose of an intervention.
Supplementary Material
Acknowledgments
The views expressed in this article do not necessarily represent the views of the US Government, the Department of Health and Human Services or the National Institutes of Health.
FUNDING
Project Viva is supported by grants from the US National Institutes of Health (R01 HD034568, R01 HD064925, K24 HD069408, R01 AI102960, and UG3 OD023286). This work was supported by the Lundbeck Foundation and the Dagmar Marshalls Foundation. The study sponsors had no role in the design of the study; the collection, analysis, or interpretation of the data; the writing of the manuscript; or the decision to submit the manuscript for publication.
ABBREVIATIONS
- AD
Atopic dermatitis
- aOR
Adjusted odds ratios
- CI
Confidence intervals
- 25(OH)D
25-hydroxyvitamin D
- Q
Quartile
- vitD
vitamin D
Footnotes
This work was done in Boston, Massachusetts, USA
CONFLICTS OF INTEREST
Dr. Asgari has received grants from Valeant Pharmaceuticals and Pfizer Inc., and Dr. Gold has received grants from the NIH. All other authors report no conflicts of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- Borzutzky A, Camargo CA., Jr Role of vitamin D in the pathogenesis and treatment of atopic dermatitis. Expert Rev Clin Immunol. 2013;9:751–60. doi: 10.1586/1744666X.2013.816493. [DOI] [PubMed] [Google Scholar]
- Camargo CA, Jr, Rifas-Shiman SL, Litonjua AA, Rich-Edwards JW, Weiss ST, Gold DR, et al. Maternal intake of vitamin D during pregnancy and risk of recurrent wheeze in children at 3 y of age. Am J Clin Nutr. 2007;85:788–95. doi: 10.1093/ajcn/85.3.788. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Camargo CA, Jr, Ganmaa D, Sidbury R, Erdenedelger Kh, Radnaakhand N, Khandsuren B. Randomized trial of vitamin D supplementation for winter-related atopic dermatitis in children. J Allergy Clin Immunol. 2014;134:831–835. doi: 10.1016/j.jaci.2014.08.002. [DOI] [PubMed] [Google Scholar]
- Chawes BL, Bønnelykke K, Jensen PF, Schoos AM, Heickendorff L, Bisgaard H. Cord blood 25(OH)-vitamin D deficiency and childhood asthma, allergy and eczema: the COPSAC2000 birth cohort study. PLoS One. 2014;9:e99856. doi: 10.1371/journal.pone.0099856. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chiu CY, Huang SY, Peng YC, Tsai MH, Hua MC, Yao TC, et al. Maternal vitamin D levels are inversely related to allergic sensitization and atopic diseases in early childhood. Pediatr Allergy Immunol. 2015;26:337–43. doi: 10.1111/pai.12384. [DOI] [PubMed] [Google Scholar]
- Fawzi WW, Rifas-Shiman SL, Rich-Edwards JW, Willett WC, Gillman MW. Calibration of a semi-quantitative food frequency questionnaire in early pregnancy. Ann Epidemiol. 2004;14:754–62. doi: 10.1016/j.annepidem.2004.03.001. [DOI] [PubMed] [Google Scholar]
- Goldring ST, Griffiths CJ, Martineau AR, Robinson S, Yu C, Poulton S, et al. Prenatal vitamin d supplementation and child respiratory health: a randomised controlled trial. PLoS One. 2013;8:e66627. doi: 10.1371/journal.pone.0066627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- International Study of Asthma and Allergies in Childhood (ISAAC) steering committee. Worldwide variation in prevalence of symptoms of asthma, allergic rhinoconjunctivitis, and atopic dermatitis: ISAAC. Lancet. 1998;351:1225–1232. [PubMed] [Google Scholar]
- Peroni DG, Piacentini GL, Cametti E, Chinellato I, Boner AL. Correlation between serum 25-hydroxyvitamin D levels and severity of atopic dermatitis in children. Br J Dermatol. 2011;164:1078–82. doi: 10.1111/j.1365-2133.2010.10147.x. [DOI] [PubMed] [Google Scholar]
- Sidbury R, Sullivan AF, Thadhani RI, Camargo CA., Jr Randomized controlled trial of vitamin D supplementation for winter-related atopic dermatitis in Boston: a pilot study. Br J Dermatol. 2008;159:245–7. doi: 10.1111/j.1365-2133.2008.08601.x. [DOI] [PubMed] [Google Scholar]
- Spiro A, Buttriss JL. Vitamin D: An overview of vitamin D status and intake in Europe. Nutr Bull. 2014;39:322–350. doi: 10.1111/nbu.12108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams H, Robertson C, Stewart A, Aït-Khaled N, Anabwani G, Anderson R, et al. Worldwide variations in the prevalence of symptoms of atopic dermatitis in the International Study of Asthma and Allergies in Childhood. J Allergy Clin Immunol. 1999;103:125–38. doi: 10.1016/s0091-6749(99)70536-1. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.