Influence of geographical latitude on vitamin D status: cross-sectional results from the BiomarCaRE consortium

Abstract

Even though sunlight is viewed as the most important determinant of 25-hydroxyvitamin D (25(OH)D) status, several European studies have observed higher 25(OH)D concentrations among north-Europeans than south-Europeans. We studied the association between geographical latitude (derived from ecological data) and 25(OH)D status in six European countries using harmonised immunoassay data from 81 084 participants in the Biomarkers for Cardiovascular Risk Assessment in Europe (BiomarCaRE) project (male sex 48·9 %; median age 50·8 years; examination period 1984-2014). Quantile regression models, adjusted for age, sex, decade and calendar week of sampling and time from sampling to analysis, were used for between-country comparisons. Up until the median percentile, the ordering of countries by 25(OH)D status (from highest to lowest) was as follows: Sweden (at 65·6-63·8°N), Germany (at 48·4°N), Finland (at 65·0-60·2°N), Italy (at 45·6-41·5°N), Scotland (at 58·2-55·1°N) and Spain (at 41·5°N). From the 75th percentile and upwards, Finland had higher values than Germany. As an example, using the Swedish cohort as a comparator, the median 25(OH)D concentration was 3·03, 3·28, 5·41, 6·54 and 9·28 ng/ml lower in the German, Finnish, Italian, Scottish and Spanish cohort, respectively (P-value < 0·001 for all comparisons). The ordering of countries was highly consistent in subgroup analyses by sex, age, and decade and season of sampling. In conclusion, we confirmed the previous observation of a north-to-south gradient of 25(OH)D status in Europe, with higher percentile values among north-Europeans than south-Europeans.

Keywords: 25-hydroxyvitamin D; Epidemiology; Europe; Latitude; Population based; Vitamin D.

Fig. 1.

Approximate geographical location of the included cohorts. SHHEC, Scottish Heart Health Extended Cohort; KORA, Cooperative Health Research in the Region of Augsburg; MATISS, Malattie Aterosclerotiche Istituto Superiore di Sanità.

Fig. 2.

Observed monthly variation of 25-hydroxyvitamin D (25(OH)D) status in the study population with complete data on 25(OH)D (n 77 320, including 80 934 observations) by country. The day length on 1st January and 1st July, respectively (using 2019 as an example year), was 4:33 and 20:37 h in Sweden (Umeå), 5:59 and 18:47 h in Finland (Helsinki), 7:05 and 17:30 h in Scotland (Edinburgh), 8:24 and 16:02 h in Germany (Augsburg), 8:46 and 15:38 h in Italy (Monza), and 9:15 and 15:07 h in Spain (Sabadell). (Information on day length was derived from www.timeanddate.com).

Fig. 3.

Percentile distribution of 25-hydroxyvitamin D (25(OH)D) status in the study population (n 81 084, including 84 766 observations) by country (ordered by decreasing geographical latitude) and season of sampling (winter, spring, summer and fall) and based on multiple imputed data sets (n 10). The solid dots represent unadjusted percentile values for each country. The dashed lines are added to facilitate between-country comparisons of each percentile.

Fig. 4.

Differences in 25-hydroxyvitamin D (25(OH)D) status in the study population (n 81 084, including 84 766 observations) by geographical latitude and based on multiple imputed data sets (n 10). The solid line represents median differences and the short dashed lines represent 95 % CI from a multivariable 4-knot restricted cubic spline model (knots at 65·0, 55·8, 48·4 and 41·6°N). The long dashed line represents median differences from a multivariable linear dose–response model. The symbols represent median differences from a multivariable categorical model. The numbers adjacent to the symbols represent the geographical latitude of each cohort or subcohort. The estimates were derived from quantile regression models that were adjusted for the same covariates as the multivariable model in Table 2. The comparator value was set to the northernmost subcohort at geographical latitude 65·6°N. , Sweden; , Finland; , Scotland; , Germany; , Italy; , Spain.