In this study we build on previous Canadian observations indicating high rates of breastfeeding and vitamin D supplementation, and provide new information regarding the dose of vitamin D provided, and mothers' reasons for choosing to supplement or not supplement. Almost 60% of infants had received only breast milk in the week prior to the survey and of these greater than 90% had received vitamin D supplements. This is much higher than the 67% vitamin D supplementation rate reported in a 2007-2008 survey for Canadian women who had exclusively breastfed an infant in the past 5 years . In that study, the supplementation rate in BC was marginally higher than the national average (70%), but still well below the rate in our study. Our rates of vitamin D supplementation are comparable to those reported more recently in one Montreal hospital where 98% of exclusively breastfed (WHO definition) infants had been supplemented with vitamin D at some point during the first 6 months . Data from the Infant Feeding Practices Study II (2005-2007) suggest that US vitamin D supplementation rates are markedly lower than in Canada; 43% of infants were breastfed at 2 months and of these only 10% were receiving vitamin D . However, the American Academy of Pediatrics only began recommending infant supplementation in November 2008, whereas the Canadian recommendation has been present in some form since 1967. Further, breastfeeding rates have been historically higher in Canada than in the US and Canada's higher latitude may have created a greater impetus for infant supplementation in this country. As expected infant supplementation at 2 months was lower amongst those receiving mixed feeds (79%) and lower still in infants receiving only infant formula (20%). In Montreal, an apparently higher 88% of mixed feeders had received supplemental vitamin D; however, this was anytime during the first 6 months . In the US study, amongst mixed feeders, only 5% were receiving vitamin D supplements at 2 months of age . Despite a high proportion of breastfed infants receiving vitamin D supplements, one third were not receiving at least 300 IU/d; mainly because of less than daily supplement administration. Although we report less frequent vitamin D supplementation rates among infants who were receiving mixed breast milk and formula in the week before the survey than among those fed only breast milk (79% versus 91%), fewer of the infants receiving mixed feeding had vitamin D intakes below 300 IU/d vitamin D (22%, versus 33.5% in the fully breastfed group). Further, within the group receiving mixed feeds, as the amount of formula increased infant supplementation dropped; however, the percentage of those receiving less than 300 IU/d remained constant at around 20%. This was not unexpected as formula is fortified with vitamin D and it would take about 700 ml of formula to achieve an intake of 300 IU/d. Consuming less than 700 ml/d also explains why 25% of formula fed infants failed to achieve this intake. Using a stricter cut-point of 400 IU/d, the Montreal researchers reported that 74% of exclusively breastfed infants and 51% receiving mixed feeds achieved an intake of 400 IU/d at 6 months .
Up to a third of infants not achieving 300 IU/d vitamin D and even less achieving the recommendation of 400 IU/d may appear high. However, only 5% of infants were receiving no vitamin D. Further, it is acknowledged that the evidence base used to derive the infant recommendation is limited. Serum 25-hydroxyvitamin D (25OHD) concentration is the best indicator of vitamin D status. Although controversial the US Institute of Medicine recently affirmed a 25OHD of 50 nmol/L as desirable in all age groups including infants . Greer et al.  showed that breastfed infants (n = 9) receiving 400 IU had mean 25-hydroxyvitamin D concentrations of 95 nmol/L after 12 weeks. More recently, infants randomized to 250 or 500 IU per day (n = 20 per group) at birth achieved mean (95% CI) 25OHD concentrations of 139 (114-164) and 151 (126-176) nmol/L, respectively after 6 weeks . Thus, it appears that the recommended intake of 400 IU exceeds the requirements of almost all infants, perhaps by a considerable margin. There have been reports of infant overdosing with vitamin D in the US resulting in the Food and Drug Administration issuing a warning of the potential risk of overdosing infants with liquid vitamin D . In our study only one infant was receiving greater than the upper limit for vitamin D of 1000 IU suggesting this was not a problem.
Among caregivers there was generally good awareness of the need to supplement and why it was important. For example, caregivers indicated that they used a supplement because vitamin D was not present in adequate amounts in breast milk and/or that sunlight exposure was limited or not recommended; and many women who used formula appeared to be aware that supplementation wasn't required. Over 90% of caregivers recalled receiving advice primarily from public health nurses and doctors to supplement with vitamin D, which may explain the high rates of supplementation. In a Seattle study , parents who reported that their child's pediatrician recommended vitamin D were 8 times more likely to provide the supplementation than parents whose child's pediatrician did not. However, only a third of parents recalled receiving any recommendation and of these under half supplemented with vitamin D. In contrast to our study, where < 5% of caregivers thought supplementation was unnecessary, 67% of parents in the Seattle study believed that supplementation was unnecessary because breast milk has all needed nutrition.
Multivariate regression revealed little in the way of predictors of supplement use. There was a non-significant tendency for family incomes less than $40,000 to be associated with lower rates of supplementation. However, vitamin D supplements cost less than $40 for 6 months and cost of the supplements was not given as a reason for not supplementing. Interestingly 80% of caregivers reported giving their infants D-Drops® versus only 16% who supplemented with D-Vi-Sol® . The reason for the popularity of D-Drops® may be their ease of administration requiring only a single drop that can be placed on the mother's breast prior to nursing, versus the need to use a dropper to administer D-Vi-Sol® .
A strength of our study was that we had access to a database that contained the names of nearly all infants born in Vancouver and Richmond over the study period. Also, we sampled an ethnically diverse population where breastfeeding rates are high relative to the rest of North America. Finally, data were collected prospectively at 2 months rather than relying on recall of up to 5 years in one study. In studies of this type, selection bias is always an important consideration. For example, people who choose to participate versus those who do not, may be more educated and of higher socioeconomic status and thus more likely to breastfeed and supplement with vitamin D. Thus a limitation of our study is that we had only a moderate response rate of 56%. Unfortunately, we do not have any data on our non-responders and there are no representative data on pregnant women in Vancouver and Richmond to compare our results with. However, with respect to ethnicity , education  and family income  our sample compares well with 2006 census data for women from Vancouver and Richmond. Further, only half of the non-responders refused participation, while the other half could not be contacted initially or at follow-up. Because the survey was conducted in the summer months it appeared that many women were away on vacations or staying with family outside the area. We acknowledge that our findings cannot be extrapolated to the rest of Canada or even BC. Breastfeeding rates are higher in BC than elsewhere and Vancouver and Richmond have a unique ethnic mix not present in the rest of the province or Canada. Second, we only sampled at 2 months; this was an intentional decision but more data are needed on older infants. In the US, supplementation rates remained relatively constant out to 12 months; however, in this study both breastfeeding rates and infant supplementation were much lower than ours . In Montreal, of all supplemented breastfed infants around a third had stopped taking the supplement by 6 months . More data are needed on older infants especially around the time of introduction of solids and as breastfeeding rates drop with age.