Vitamin D is a pro-hormone and one of the essential vitamins important for good health. Foods naturally rich in vitamin D include fatty fish and liver, although most of the intake for adults and children is achieved through biosynthesis in the skin when exposed to UV-B radiation in sunlight( Reference Holick and Chen 1 ). Vitamin D supplementation has been observed to improve bone health by greatly increasing intestinal absorption of Ca( Reference DeLuca 2 ), which aids in bone mineralization( Reference DeLuca 2 ), increases bone density( Reference DeLuca 2 ) and helps reduce fracture( Reference Bischoff-Ferrari, Willett and Wong 3 , Reference Bischoff-Ferrari, Willett and Orav 4 ). However, there is conflicting evidence which suggests that supplementation may provide little benefit to bone health beyond those who are very deficient( Reference Reid, Bolland and Grey 5 ). Newer areas of vitamin D research include cancer( Reference Ma, Zhang and Wang 6 ), CVD( Reference Parker, Hashmi and Dutton 7 ) and autoimmune disorders( Reference Prietl, Treiber and Pieber 8 ).
An individual’s vitamin D level is most often measured using its inactive but main circulating metabolite 25-hydroxyvitamin D (25(OH)D), as it is representative of vitamin D intake from food and supplements and due to sun exposure( Reference Holick 9 ). The Dietary Reference Intakes currently endorsed by Health Canada are drawn from a 2011 Institute of Medicine report on Ca and vitamin D. The report set the RDA at 50 nmol/l (20 ng/ml), which is the amount estimated to meet the requirements of at least 97·5 % of the population and corresponds to a daily dietary intake of 15 µg (600 IU) for individuals 1–70 years of age( Reference Ross, Manson and Abrams 10 ).
Studies have estimated that subcutaneous synthesis may not be feasible from November to February at latitudes greater than 42°( Reference Webb, Kline and Holick 11 , Reference Terushkin, Bender and Psaty 12 ). As the vast majority of Canada’s population lives north of 42° latitude, and further evidence suggests that this period may extend from October to April in Edmonton( Reference Webb, Kline and Holick 11 ), it is not possible for many Canadians to obtain sufficient vitamin D from subcutaneous synthesis for a significant portion of the year.
Data from the Canadian Health Measures Survey show the prevalence of Canadians aged 6–79 years with serum 25(OH)D<50 nmol/l is 25·7 %, with a significantly higher prevalence among men than women( Reference Whiting, Langlois and Vatanparast 13 ). Research also shows a significantly higher prevalence among non-whites( Reference Whiting, Langlois and Vatanparast 13 , Reference Greene-Finestone, Berger and de Groh 14 ). Using different cut-off points, analysis of the Canadian Multicentre Osteoporosis Study showed that only 2·3 % of Canadians aged 35 years and over fall below 27·5 nmol/l and 59·2 % fall below 75 nmol/l( Reference Greene-Finestone, Berger and de Groh 14 ). Vitamin D supplements are a viable option to compensate for minimal sun exposure. The purpose of the present study was to determine the factors associated with vitamin D supplement utilization among Canadian adults. To our knowledge, this is the first nationally representative analysis on vitamin D supplement usage in Canada.
Methods
The current analysis was based on data from the Canadian Community Health Survey (CCHS) – Health Aging conducted by Statistics Canada in 2008 and 2009( 15 ). The survey targeted the Canadian population 45 years of age and over, excluding residents of the three territories, persons living on Indian Reserves or Crown lands, those residing in institutions, full-time members of the Canadian Armed Forces and residents of certain remote regions. A multistage stratified sampling design was used and a total of 30 865 participants were enrolled in the survey with a response rate of 74·4 %( 15 ). The survey asked the following question: ‘In the past month, how often did you take vitamin D supplements?’ and daily use of vitamin D supplements was affirmed if a participant answered ‘every day’ to the question. Thirty-one participants who did not answer the question were excluded from the analysis. BMI was calculated as [weight (kg)]/[(height (m)]2 and participants were grouped into the following categories: <18·5, 18·5–24·9, 25·0–29·9 and ≥30·0 kg/m2. Current smokers were respondents who reported smoking cigarettes every day at the time of the survey. Former smokers were those who reported smoking cigarettes daily in the past but were not smoking at the time of the survey. Otherwise, participants were classified as non-smokers.
Individuals were grouped into five household income categories (<$CAN 20 000, $CAN 20 000–39 999, $CAN 40 000–59 999, $CAN 60 000–79 999, ≥$CAN 80 000) and three education categories (low (secondary school not completed), medium (secondary school completed), high (post-secondary education)). Other variables included in the analysis were age (45–49, 50–54, 55–59, 60–64, 65–69, 70–74, 75–79, ≥80 years), marital status (married or common law partner, single, separated or divorced or widowed), race (white, non-white), immigrant status (yes, no), alcohol drinking (none, occasional, regular), leisure-time physical activity (yes, no), self-perceived health (poor, fair, good, very good, excellent) and having a chronic condition (yes, no).
We calculated the proportion of daily vitamin D supplement use according to the various factors described above. Logistic regression modelling was used to assess the relationships between each predictor and daily vitamin use after adjustment for other factors in men and women separately. Model parameters were estimated by using the method of maximum likelihood and were tested for significance by using the Wald statistic at α=0·05. The CCHS employed unequal probability sampling and therefore sampling weights were used to make point estimates representative of the Canadian population. Detailed sampling structure information was not included in the publicly released data files and we calculated variance estimates by using adjusted weights that took the average design effect into consideration. The design effect is a ratio of the estimated variance from a complex survey design over a similar estimate from a simple random design, and the average design effect for the survey was calculated and provided by Statistics Canada. All the statistical analyses were conducted by using the statistical software package SAS version 9·3.
Results
Vitamin D supplement use was much higher among women than men aged 45 years and older (Table 1). Overall, the highest prevalence of supplement use was observed in women aged 65–69 years at 48·0 %. Among men, the highest prevalence was in the 70–74 years age group where 25·3 % reported using supplements daily. Supplement use increased with age up to 65–74 years and then declined somewhat thereafter. For women, supplement use was highest in those of normal weight and lowest in the obese, and these differences were not notable in men. Supplement use showed only a modest variation among the income or education groups.
Proportion of Canadians who took a vitamin D supplement daily according to various factors, the Canadian Community Health Survey – Healthy Aging, 2008–2009
* Weighted to the Canadian population.
Current smokers had a lower prevalence of supplement use than former or non-smokers. Those who engaged in leisure-time physical activities and those with chronic conditions had a higher prevalence of supplement use than those who did not and healthy individuals. Participants who were non-white had a lower prevalence of supplement use in both sexes, and female immigrants had a slightly lower proportion of vitamin D supplement use than female non-immigrants, with no similar difference in men.
Logistic regression analysis yielded similar associations where vitamin D supplement use increased with age up to the age of 70–74 years in both sexes (Table 2). Compared with women of normal weight, overweight and obese women had significantly lower odds of supplement use after adjustment for covariates. A similar but non-significant association was observed in men.
AdjustedFootnote * odds ratios (AOR) and 95 % confidence intervals for taking vitamin D supplement daily in relation to various factors, the Canadian Community Health Survey – Healthy Aging, 2008–2009
* Odds ratio estimates for a given independent variable are adjusted for all additional variables that appear in the table.
The odds of vitamin D supplement use decreased with decreasing household income in both sexes, where those from households with incomes less than $CAN 80 000 had significantly lower odds of use than those from households earning $CAN 80 000 or more. A similar trend was seen in education, where men and women who did not complete high school had significantly lower odds of supplement use than those who attended or graduated from a post-secondary institution.
Smokers had significantly lower odds of taking supplements relative to former and non-smokers. Men and women who did not engage in leisure-time physical activities had lower odds of using vitamin D supplements than those who did. Little difference was observed in association with alcohol drinking. Immigrants had significantly lower odds of supplement use, although when stratified by sex it became non-significant in men.
The odds of vitamin D supplement use were higher among participants who rated their health lower than ‘excellent’, and men who rated their health as ‘poor’ had the highest odds of use. The odds of supplement use were significantly higher in both men and women with chronic conditions.
Discussion
A strong association was observed between sex and vitamin D supplement use, similar to previous reports( Reference Whiting, Langlois and Vatanparast 13 , Reference Green, Barr and Chapman 16 ). Women as a whole had more than three times the odds of supplement use compared with men and this association was evident in every age group. Previous research has found women to have a significantly higher mean 25(OH)D( Reference Langlois, Greene-Finestone and Little 17 ) and a lower prevalence below thresholds of 30 nmol/l, 40 nmol/l and 50 nmol/l( Reference Whiting, Langlois and Vatanparast 13 ). Research also suggests that the higher vitamin D status of Canadian women may be largely due to their higher prevalence of supplement use as they have less sun exposure, more frequent use of sunscreen and similar dietary intakes of vitamin D to men( Reference Greene-Finestone, Berger and de Groh 14 ).
Another notable finding in the present study was an association between vitamin D supplement use and age. Unlike men, where supplement use increased gradually from middle to old age, supplement use in women was most strongly associated with the 65–74 age group and declined thereafter. This may partially explain the narrowing of sex differences in vitamin D status with increasing age( Reference Whiting, Langlois and Vatanparast 13 , Reference Langlois, Greene-Finestone and Little 17 ), although it is worth noting that a much higher proportion of women than men used vitamin D supplements in every age group. The ability of the skin to produce 25(OH)D decreases with age( Reference Vieth, Ladak and Walfish 18 ), but there is evidence suggesting that conversion of 25(OH)D to 1,25-dihydroxyvitamin D (1,25(OH)2D), the active metabolite, in the kidney also slows with age and may have a neutralizing effect( Reference Vieth, Ladak and Walfish 18 ).
Vitamin D supplement use increased steadily with household income in both sexes, and mineral and vitamin supplement use in Canada as a whole appears to follow this trend( Reference Vatanparast, Adolphe and Whiting 19 ). Focus groups of low-income Canadians have shown that barriers to supplement use in general are similar to those for a healthful diet, including lack of accessibility and lack of knowledge about what to buy and its potential health benefits( Reference Whiting, Vatanparast and Taylor 20 ). Approximately one-third of low-income households in Canada experience food insecurity( Reference Che and Chen 21 ), which has been associated with a significant reduction in vitamin D intake( Reference Vatanparast, Calvo and Green 22 ), and low-income individuals expressed that taking supplements is a low priority when faced with a shortage of food and money( Reference Whiting, Vatanparast and Taylor 20 ). This illustrates a problem with supplements as a whole, and fortification of food may be a more effective method of increasing the vitamin D status of certain groups.
Education followed a similar pattern whereby those with a low level of education were significantly less likely to use vitamin D supplements than those with a medium or high level of education. This is not surprising as education is highly correlated to income and some of the factors identified as being barriers to supplement use, such as lacking knowledge of their specific benefits( Reference Vatanparast, Adolphe and Whiting 19 ), may stem from a lack of education. Being an immigrant was modestly associated with not using vitamin D supplements, which may be due to differences in cultural views towards vitamin supplements.
Among health behaviours, being a current smoker and not engaging in leisure-time physical activities were most strongly associated with not using vitamin D supplements. These two traits are commonly associated with less frequent supplement use as a whole( Reference Robson, Siou and Ullman 23 ) and may be indicative of individuals who do not actively promote their health. Furthermore, women who were overweight or obese had significantly lower odds of using vitamin D supplements than women of normal weight. BMI and other measures of adiposity have been inversely associated with vitamin D levels in several studies( Reference Snijder, van Dam and Visser 24 , Reference Mai, Chen and Camargo 25 ) and it has been hypothesized that obesity may reduce the bioavailability of 25(OH)D, increasing one’s risk of deficiency( Reference Wortsman, Matsuoka and Chen 26 ). It is also possible that low 25(OH)D levels in obese people are explained by a low vitamin D intake.
Data from the National Health and Nutrition Examination Survey (NHANES) in 2003–2006 showed vitamin D supplement use to be 56·3 % for women and 44·0 % for men aged 60 years and older( Reference Gahche, Bailey and Burt 27 ), notably higher than our results of ~45 % and ~23 %. Supplement use in the USA increased significantly from 1988–1994 to 2003–2006, especially in people aged 60 years and over who had comparable use to the general adult population in 1988–1994( Reference Gahche, Bailey and Burt 27 ), perhaps due to increased recommendations since then. Demographic analysis of US users of vitamin and mineral supplements of any kind yielded similar results to ours where age, being female, white and not overweight or obese were associated with general supplement use( Reference Rock 28 ).
There was a consistent inverse relationship between self-perceived health and vitamin D supplement use. Self-perceived health appears to be a product of many factors including socio-economic status, maintaining a healthy weight, exercise, not having a chronic condition and not smoking( Reference Mackenbach, van den Bos and Joung 29 ) which, conversely, were found to be positively associated with supplement use in the current analysis.
The current RDA taken from the 2011 Institute of Medicine Report only consider bone health outcomes such as the risk of fracture and osteomalacia( Reference Ross, Manson and Abrams 10 ). For conditions such as autoimmune disorders, cancer and CVD, the report cited insufficient evidence for effect and causality and a lack of randomized clinical trials as the rationale for excluding most of the considered outcomes( Reference Ross, Manson and Abrams 10 ). Research into the benefits of vitamin D supplementation has been somewhat inconsistent, which may be due to using too low a dose. Health Canada currently recommends that individuals over the age of 50 years take a daily 10 µg (400 IU) vitamin D supplement and an earlier report shows that ~70 % of Canadians who use vitamin D supplements use ≤10 µg (≤400 IU)( Reference Whiting, Langlois and Vatanparast 13 ). However, several meta-analyses( Reference Bischoff-Ferrari, Willett and Orav 4 , Reference Bischoff-Ferrari, Willett and Wong 30 ) have found significant differences between trials using low doses (<10 µg (400 IU)) v. high doses (>17·5 µg (>700 IU)) of vitamin D to improve bone health. In their meta-analysis of over 63 000 participants, Tang et al.( Reference Tang, Eslick and Nowson 31 ) found that the addition of vitamin D to Ca supplementation did not provide significant further reduction in rates of bone fracture, but there was a significant difference between trials using higher doses and those using lower doses. The authors comment that the much larger number of low-dose trials may have attenuated the associations observed in the high-dose trials when combined into one overall measure. Several other meta-analyses have found that vitamin D supplementation provides little to no benefit for fractures( 32 , Reference Avenell, Mak and O’Connell 33 ) and bone mineral density( Reference Reid, Bolland and Grey 5 ).
Reid et al.( Reference Reid, Bolland and Grey 5 ) contend that many trials compare a combined Ca and vitamin D intervention against a placebo and so are not designed to isolate an independent effect, citing the trial of Tang et al. who found no additional benefit when adding vitamin D to Ca. Reid et al. conclude that current vitamin D supplement recommendations are likely only appropriate for particularly deficient groups such as elderly individuals with little sunlight exposure and not for the general population over 50 years old.
The present study has several limitations that must be considered. Participants might have taken a vitamin supplement daily for the past month or more but were unaware which supplement it was. Recall bias may be less of a factor in the current study as participants were scored as supplement users only if they took vitamin D supplements daily, which may be more easily recalled than a less frequent practice. Participants who took vitamin D less frequently than daily were scored as negative and the doses taken were not recorded, so some of the diversity in use may not be captured in the study. Furthermore, the study did not include Aboriginal peoples living on reserve who are at higher risk of vitamin D deficiency. Some participants may be unaware they are taking vitamin D daily due to it being combined with Ca or because they accepted supplements from a health-care provider without reading the label. As with any cross-sectional analysis, it is not possible to make casual inferences on the associations observed herein. Furthermore, a number of comparisons were made in the analysis and there may be an increased possibility of observing significant associations.
Conclusion
In this representative sample of Canadians aged 45 years and over, vitamin D supplement usage was consistently higher in women and increased with age in men. Similar to previous reports, vitamin D supplement use was associated with higher education and household income. Supplement use was also associated with engaging in leisure-time physical activities, never smoking, a low level of self-perceived health and having a chronic condition.
Acknowledgements
Financial support: This study was partially supported by the Norwegian Research Council (project number 201895/V50). The Norwegian Research Council had no role in the design, analysis or writing of this article. Conflict of interest: None. Authorship: Y.C. designed and conducted research and statistical analysis. D.M. wrote the manuscript and M.X. and Y.C. edited and consulted. Ethics of human subject participation: This study used a public released data file and ethics approval is not required for the current analysis.
