Vitamin D supplementation in Swiss infants 1

Background: Various recent publications reported clinical manifestations of vitamin D deficiency in infants. Furthermore new research revealed additional properties of vitamin D for bone health and in the prevention of chronic diseases. However, prevalence data on actual supplementation rates are scarce. This study reports the prevalence of vitamin D supplementation in infants in Switzerland and presents risk factors for non-supplementation. Methods: In 2003, mothers of 2861 randomly selected infants aged 0–9 months received a questionnaire on infant feeding, including a question on vitamin D supplementation. The prevalence of vitamin D supplementation was calculated and its dependency on various factors analysed by multiple logistic regression. Results: 64% of the infants had received vitamin D. The regression analysis yielded various significant risk factors for non-supplementation: young maternal age, German language region, Swiss nationality, siblings and breastfeeding. Protective factors were intake of folic acid during pregnancy and professional information on infant feeding. The protective effect of professional information varied significantly by region. Conclusions: Given that the supplementation of vitamin D is recommended for all infants, the supplementation prevalence in Swiss infants is unsatisfactorily low. Various risk factors were identified and a positive impact of professional counselling on the supplementation rate could be demonstrated. In view of the new evidence emerging on additional preventive properties of vitamin D and the resurgence of rickets, the importance of vitamin D for infant health and ways to improve its promotion must be discussed anew.

Vitamin D has been known for the prevention of rickets since the 1930's.Its supplementation has resulted in a major reduction of the clinical manifestations of vitamin D deficiency, such as rickets and growth failure, lethargy and irritability or hypocalcaemic tetany [1][2][3].It is therefore internationally accepted to recommend vitamin D supplementation in infancy.The dose and duration, however, differs from country to country dependent on the risk profile of the concerned population [4][5][6][7][8].The Swiss Paediatric Association recommends 400 µg of vitamin D daily from the infant's second week until the end of the first year [9,10] (SGP).Lately, several publications discuss a resurgence of clinical manifestations, such as rickets and seizures, in risk populations, infants, who are less exposed to sunlight for environmental, cultural or medical reasons or who are fully breast-fed [11][12][13][14][15][16][17][18][19][20][21].These findings imply either a lack of compliance with the recommendations or a need of higher supplementation dose.The later could be explained by the increasing prevalence of breastfeeding as well as a reduction of sunlight exposure and frequent use of sunscreen as preventive measures against skin cancer [5,7,15,16,[22][23][24].These genuinely positive results of health promotion activities may have had unwanted side effects.The prevailing importance of vitamin D supplementation in infancy is accentuated by new research on its preventive properties for bone health after infancy as well as for other diseases, such as diabetes type I, mental disorders or various cancers [6,18,[25][26][27][28][29][30][31][32][33][34][35][36][37][38].While vitamin D deficiency has been reported for risk populations and in various case studies [11,17,19,[39][40][41][42][43][44], only a few studies have investigated the prevalence of vitamin D supplementation in infants [8,39,[45][46][47].The present study investigates for the first time the prevalence of vitamin D supplementation in Switzerland and the possible risk factors for noncompliance to the guidelines.

Ethics approval:
The ethical commission of the university of Basel decided that no ethical approval was necessary, since the data collection was anonymous and no specimens were collected.
Between April and July 2003 183 regional community-based mother-child health services, two hospitals and in the Italian speaking region health authorities were instructed to randomly select mothers, who had given birth within the previous nine months.4114 mothers were sent a questionnaire on infant feeding, including a question on the vitamin D supplementation within the past 24 hours, as well as questions on socio-demographic characteristics and the health of the mother and her child.The return rate was 74% resulting in 3032 records.After exclusion of records with insufficient questionnaire data, missing information on the age of the child or on infant feeding, and records from children older than 11 months, 2868 records were left for analysis.
First a descriptive analysis of the prevalence of vitamin D supplementation was conducted using the Chi2 test to assess differences in prevalence across various subgroups.P-values <0.05 were considered significant.And second, in an explorative approach, various factors, potentially influencing the supplementation of vitamin D, were studied using multivariate logistic regression.Variables, significant in uni-variate testing, as well as variables, known to be associated with vitamin D supplementation, were included in the analysis: season of data collection, age of mother, parity, full breastfeeding, region, education, information on infant feeding, mother's attitude towards primary prevention and sex and birth weight of the child.We further adjusted for potential confounding factors, reported in literature to influence infant feeding praxis, such as body mass index, smoking, allergy of the mother.The results are presented as odds ratios with 95% confidence intervals.The season of data collection was defined as spring/summer for the months May through September (n = 2684) and as fall (n = 235, data from October and November months).Educational status was defined as high, when mothers had reached a degree from a technical college or university, and otherwise as low.The mother's attentiveness regarding her own nutrition was assessed using a question from the Swiss Health Survey 1997 [48].Breast feeding practices mentioned in this article were classified according to the WHO's definitions [49].Mothers were questioned about their source of information on infant feeding.Information on infant feeding was used as a proxy for "information on vitamin D supplementation".Information provided by doctors and/or motherand child-health workers was defined as "professional".If it was provided by relatives and/or friends as "non-professional" and if no information by such persons had been provided as "none".The variable "folic acid intake during pregnancy" was used as a proxy for the mother's attitude towards primary prevention.Potential effect modification by breast feeding was tested for by introducing interaction terms between breast feeding and the covariate of interest, known risk factors for breastfeeding: smoking, nationality, language region and infants' age.Furthermore potential interaction between region and nationality, education, parity and information on infant feeding were tested.Model comparisons were performed using the likelihood ratio test.To demonstrate the interaction seen between region and type of information (LR Test: p = 0,0241), we calculated regional prevalence of vitamin D supplementation adjusted for all other variables of the logistic regression model.The interaction-model was not included into the model presented in this paper (table 3).
A sensitivity analysis was conducted to distinguish the impact of different feeding types on vitamin D supplementation.The model was adjusted for season, maternal age, smoking and language region.The analysis was stratified for infant age groups, since breastfeeding prevalences vary significantly across age groups.

Methods
A comparison with Swiss national birth register showed that mothers in the study sample were slightly older and more often of Swiss nationality compared to the national birth register (data not shown [50]).The proportion of primiparous women was also higher in the study sample.
The overall prevalence of vitamin D supplementation was 64% (CI 0.615-0.650).Table 1 presents the prevalence of vitamin D supplementation across different sub-groups.
As to be expected, the prevalence rate was higher during the fall months than during the spring/summer months (72% vs 62%, p = 0.003).The supplementation rate also varied across different sub-groups (table 1).
Young mothers (<25 years) as well as mothers over 35 years gave vitamin D less often, a tendency also seen among multiparous mothers.In addition, prevalence rates differed by language regions and nationality.Swiss infants received vitamin D more often than infants from mothers coming from Balkan countries, but less often than infants, whose mothers came from EU member states or other countries.The prevalence of vitamin D supplementation was lower in the German-speaking region of Switzerland than in the French or Italian region.Further factors with a significant impact on vitamin D supplementation were the source of information mothers received, the intake of folic acid in pregnancy, the infants age and birth weight (univariate analysis).
Table 2 displays the influence of the source of information on the supplementation rate by language region, differentiating professional information into three categories: a) mother-and childcare workers b) information by doctors or c) both sources.Mothers, who had seeked advice from both doctors and mother-and child-care workers, showed the highest supplementation prevalence.
We further investigated the influence of different factors on the supplementation of vitamin D by multivariate logistic regression.The results are summarised in table 3.
The   25 years, Swiss nationality, German language region, the presence of siblings and full breastfeeding.In addition, the analysis revealed protective factors.Mothers, who had taken folic acid in pregnancy and had received professional information on infant feeding substituted vitamin D significantly more often.Certain variables, such as Balkan origin, were no longer negatively associated with vitamin D supplementation after the adjustment.The median duration of exclusive and full breastfeeding in this study were 9 and 17 weeks respectively, while the total breast feeding duration was 31 weeks [50].
Mothers who partially breastfed, which means they complemented breast milk with formula milk, tended to give their infants vitamin D more often than mothers who breastfed fully or exclusively (table 4).While partial breastfeeding showed a positive association with vitamin D supplementation across all age groups, even though non-significant, full and exclusive breastfeeding were negatively associated.Both feeding types were significant in the 2-3 months olds and for exclusive breastfeeding also in the 5-6 months olds.
The observed interaction between language regions and source of information is illustrated in figure 1 showing adjusted prevalences.Compared to the professionally informed, the prevalence of vitamin D supplementaion was consistenly lower in mothers, who were not informed or non-professionally informed.In the German speaking regions the "uninformed mothers" substituted their infants with vitamin D nearly as often as professionally informed mothers had done.Only few studies have been conducted on the prevalence of vitamin D supplementation in infants.A Dutch study reported a supplementation prevalence of 91% at the age of one year [8] and a Finnish study published a prevalence of vitamin supplements of 50% at the age of 2 years [47].The EURODIAB study, investigating the preventive properties of vitamin D in the development of diabetes type I in children, presented a supplementation prevalence varying from 47-97% for 7 different European countries [45].The variation of recommendations in Europe [4,6,[8][9][10] and differences in study settings do not permit a comparison of the results.However, one may state, that compliance with national guidelines varies in Europe and that some countries present an undersupplementation of vitamin D. In Switzerland, given that the Swiss national recommendation on vitamin D supplementation is universal for all infants, the general prevalence of vitamin D supplementation in infants is low, irrespective of the season.Only 62% of the infants had received vitamin D in spring/summer months and 72% in fall months.A possible explanation for the seasonal difference may be the common knowledge of the endogenous production of vitamin D via direct sunlight.Some mothers might restrict vitamin D supplementation to winter months.This mechanism, being commonly known, might also explain, why a higher educational level was not associated with a higher prevalence of vitamin D supplementation.Interestingly, we found a higher prevalence in French and Italian regions than in the Germanspeaking region of Switzerland.The regional differences can only partly be explained by the observed interaction seen between the source of information and the language region.We also found that nationality influences the supplementation of vitamin D. Swiss infants received significantly less often vitamin D than infants, whose mothers came from EU member states.Also important were other socio-demographic factors, such as young maternal age (<25 years) and parity.Having a sibling reduced the probability of receiving vitamin D by 38%.Exclusively or fully breastfed infants have been reported to be at higher risk for vitamin Discussion D deficiency [5,13,16,20,21,[51][52][53][54][55].This observation is often explained by the small amount of vitamin D in breast milk, especially in vitamin D deficient, lactating mothers.Our study shows, that exclusive or full breastfeeding itself poses a risk for vitamin D non-supplementation, while partial breastfeeding does not, on the contrary.The recent breastfeeding promotion may have led to the belief that the exclusive breastfed infant is in need of no further supplements.The association of breastfeeding with vitamin D supplementation and discontinuation of vitamin D in older infants would best be analysed by a longitudinal study design.The supplementation prevalence showed a significant association with the infant's age, although vitamin D is recommended from the 2 nd week until the end of the first year.This association did not change after adjustment for potential confounders.Infants aged 6 months and older had a significant higher risk of non-supplementation than younger infants.Possibly, the professional as well as the parental attention on preventive measures, such as vitamin D supplementation, slackens once an infant thrives and gets older.This speculation correlates with the fact, that infants with low birth weight and health problems are better substituted than healthy infants.Furthermore, mothers might stop vitamin D supplementation when infants are weaned and formula milk, fortified with vitamin D, replaces breast milk.Apart from the risk factors, the multiple logistic regression yielded factors positively influencing the supplementation of vitamin D, such as the mother's positive attitude versus preventive measures (use of folic acid in pregnancy) and professional information on infant feeding.Information on infant feeding in general, although not specific for vitamin D, showed a positive influence on the supplementation rate.Certainly, the non-specificity of the question limits the interpretation of the results.The highest prevalence rates could be seen in infants, whose mothers had received professional information.After further differentiating the professional source, we observed that the prevalence increased with the level of source authority and the number of professional sources a mother consulted.
The surprisingly high prevalence in the "non-informed" mothers can be explained by other sources of information, such as print media or internet, which were not covered by the questionnaire.These sources, however, are open to all mothers.We can only hypothesize, that women, who solely rely on information by non-professionals, will seldom consult other media for professional information.
The study was primarily designed to study infant feeding in a representative study sample in Switzerland.The generalisability of the study is limited mainly by the higher prevalence of primiparous mothers and the under-representation of non-Swiss mothers.In addition, we probably reached mothers, capable of reading and understanding at least one of the national languages in which the questionnaires were available.These groups, however, supplemented their children more often, than did multiparous or Swiss mothers, resulting in an overestimation of the supplementation prevalence.On the other hand, the vitamin D supplementation rate is known to be higher in winter-months and the study only collected spring-fall data (May -November), thereby possibly underestimating the vitamin D supplementation.
Investigating the prevalence of vitamin D supplementation in infants was a first step to assess a potential vitamin D deficiency in Switzerland.Further investigations on actual vitamin D levels in Swiss infants and children are needed.In view of the new evidence of preventive properties of vitamin D, a discussion on the importance of vitamin D supplementation for public health needs to be initiated.
The analysis showed a low prevalence of vitamin D supplementation in Switzerland and thereby a serious non-compliance to national recommendations.So far, vitamin D deficiency in infants is not a priority health problem in Switzerland.However, given the importance of vitamin D for the infant's bone health and possibly other chronic diseases, the results are worrisome.The study revealed certain risk factors, such as age, multi-parity or breastfeeding, which can easily be addressed in counselling settings.The regional and national differences need to be further investigated.Professional information on infant feeding and mother's attitude towards primary prevention were positively correlated with vitamin D supplementation.This leads us to the conclusion, that the supplementation rate of vitamin D can be influenced by promotion and repeated professional information on the need and benefit of vitamin D. The prevailing importance of vitamin D for Conclusion infants' health and better ways to promote the supplementation of vitamin D must be discussed in the public health community.Med Wochenschr (1871-2000) Swiss Med Wkly (continues Schweiz Med Wochenschr from 2001) Editores Medicorum Helveticorum

Table 3
1 mulitvariate logistic regression adjusted for season, maternal age, smoking and region