Low 25-hydroxyvitamin D levels are more prevalent in Canadians of South Asian than European ancestry inhabiting the National Capital Region of Canada

Abstract

The US Institute of Medicine defined serum 25-hydroxyvitamin D (25OHD) cut point values of 30 nmol/L and 40 nmol/L were used to assess the vitamin D status of South Asian and European Canadians of self-identified ancestry living in the National Capital Region of Canada. Serum 25OHD values were measured in the spring and fall of 2012 to represent status during the winter and summer months, respectively. A total of 1238 measurements were obtained from 669 participants (49% South Asian ancestry): some participants were measured only once (spring or fall). Median 25OHD values were significantly higher in participants of European ancestry: 70.8 nmol/L (68.1, 73.5; 95% CI) versus South Asian ancestry: 42.7 nmol/L (40.5, 45.0; P<0.001). Spring vs. fall differences were small for each ethnic group and significant only for those of European ancestry (2.9, CI: 1.0-4.9 nmol/L; P = 0.01). Among participants of South Asian ancestry, 27.3% (fall) and 29.1% (spring) of females had values <40 nmol/L while the percentages for males were considerably higher (36.5% and 44.2%, respectively). The corresponding values for participants of European ancestry were ≤10%, showing that the South Asian participants were less likely to achieve the 25OHD concentrations established by the IOM for optimum bone health. Investigation of the factors related to serum 25OHD levels showed that supplement intake and ethnic background were associated with the biggest differences. Skin color was not a major factor, suggesting that genetic factors are responsible for the observed differences between participants of different ethnic backgrounds.

Fig 1

LSE serum 25OHD values with 95% confidence intervals as a function of ethnic background (A), vitamin D supplement consumption (B), and season (C) by age group. Panel A: values are for Europeans (open) or South Asians (filled). Panel B: those not consuming supplements (open) or supplement consumers (filled). Panel C: spring (open) or fall (filled). Superscripts (a,b,c,y,z) denote significant differences (P<0.05) when compared to other age ranges within South Asian (y, z) or European (a, b, c) ethnic groups. *Different from value for corresponding ethnic group (panel A), supplement status (panel B) or season (panel C; P<0.05).

Fig 2. Relationship between serum 25OHD and BMI as a function of age and ethnic background.

Solid line represents the relationship obtained from the regression model with 95% confidence limits (shaded area). Panels A (N = 173, age range 20-39y), B (N = 284, age range 40-59y) and C (N = 125, age range 60-79y) represent data from participants of South Asian descent. Panels D (N = 233, age range 20-39y), E (N = 271, age range 40-59y), and F (N = 147, age range 60-79y) represent data from participants of European descent.

Fig 3. LSE serum 25OHD values as a function of season, ethnic background and consumption of vitamin D containing supplements (Yes/No).

Values represent mean ± standard error. Open symbols represent participants of European descent and filled symbols represent those of South Asian descent. Triangles represent participants consuming supplements while circles represent participants who did not consume supplement. Symbols with different superscripts differ significantly (P<0.05). The number of participants in each category is indicated below the symbol in parentheses.

Fig 4

Serum 25OHD values as a function of skin color for participants of South Asian (Panel A) and European (Panel B) ancestry. The x axis shows skin color classifications corresponding to Individual topology Angles calculated for each individual [21]: D (dark), B (brown), I (intermediate), T (tanned), L (light), and VL (very light). The straight line represents the least squares regression relationship between serum 25OHD values and skin color obtained from the PROC MIXED model (left axis). The shaded area represents the 95% confidence limits. The bars represent the number of participants falling within the indicated skin color range (right axis).