Calcified Tissue International

, Volume 74, Issue 3, pp 255–263

Vitamin D Insufficiency in Greenlanders on a Westernized Fare: Ethnic Differences in Calcitropic Hormones Between Greenlanders and Danes

Authors

    • Dept of Endocrinology and Metabolism CAarhus Amtssygehus, Aarhus University Hospital, Aarhus
    • Faculty of Health SciencesAarhus University, Aarhus
  • M. E. Jørgensen
    • Dept of MedicineQueen Ingrids Hospital, Nuuk
    • Steno Diabetes Centre, Gentofte
  • M. B. Pedersen
    • Dept of MedicineQueen Ingrids Hospital, Nuuk
  • J. C. Hansen
    • Faculty of Health SciencesAarhus University, Aarhus
    • Center for Arctic Environmental Medicine, Dept of Environmental and Occupational MedicineAarhus University, Aarhus
  • L. Heickendorff
    • Dept of Clinical Biochemistry, Aarhus AmtssygehusAarhus University Hospital, Aarhus
  • A. L. Lauridsen
    • Dept of Clinical Biochemistry, AKHAarhus University Hospital, Aarhus
  • G. Mulvad
    • Center for Primary Health Care, Nuuk
  • C. Siggaard
    • Faculty of Health SciencesAarhus University, Aarhus
  • H. Skjoldborg
    • Faculty of Health SciencesAarhus University, Aarhus
  • T. B. Sørensen
    • Faculty of Health SciencesAarhus University, Aarhus
  • E. B. Pedersen
    • Faculty of Health SciencesAarhus University, Aarhus
    • Dept of MedicineHospital of Holstebro, Holstebro
  • L. Mosekilde
    • Dept of Endocrinology and Metabolism CAarhus Amtssygehus, Aarhus University Hospital, Aarhus
    • Faculty of Health SciencesAarhus University, Aarhus
Article

DOI: 10.1007/s00223-003-0110-9

Cite this article as:
Rejnmark, L., Jørgensen, M., Pedersen, M. et al. Calcif Tissue Int (2004) 74: 255. doi:10.1007/s00223-003-0110-9

Abstract

We studied the influence of age, gender, latitude, season, diet and ethnicity on plasma 25-hydroxyvitamin D 25 OHD, PTH, 1,25-dihydroxyvitamin D, vitamin D-binding protein, bone-specific alkaline phosphatase, and osteocalcin levels in 46 Greenlanders living in Nuuk (64°N) on a traditional fare (group A), 45 Greenlanders living in Nuuk on a westernized fare (group B), 54 Greenlanders (group C), and 43 Danes (Group D) living in Denmark (55°N) on a westernized fare. Blood specimens were drawn both summer and winter. Vitamin D insufficiency (plasma 25 OHD <40 nmol/l) was common in all four study groups during summer (23–74%) and winter (42–81%). Compared to groups A and D, vitamin D insufficiency was significantly more frequent in groups B and C. In all groups, summer levels of 25 OHD were above winter levels. Multiple regression analysis revealed a significant effect of ethnicity. Compared to Danes, Greenlanders had higher 1,25-dihydroxyvitamin D levels, but lower 25 OHD and PTH levels despite relatively low plasma calcium concentrations. In addition to ethnicity, 25(OH)D levels were influenced by age, season (summer > winter), and diet (a traditional Inuit diet>westernized diet). Ethnical differences exist between Greenlanders and Danes. Our results suggest that Greenlanders may have an inherent lower “set-point” for calcium-regulated PTH release or an enhanced renal 1,25(OH)2D production. In addition to ethnicity, age, season, and diet were important determinants of vitamin D status. Changes from a traditional to a westernized fare are associated with a reduced vitamin D status in Greenlanders. Vitamin D supplementation should be considered.

Keywords

Steroid hormonesVitamin DPeptide hormones PTH/PTHrPBone Turnover MarkersGreenlandEthnicity

Vitamin D insufficiency, as measured by reduced plasma 25-hydroxyvitamin D (25 OHD) levels, is a risk factor for secondary hyperparathyroidism, bone loss, fractures, proximal myopathy and frank osteomalacia [1]. Vitamin D is obtained from dietary sources and from endogenous synthesis in the skin. Dietary sources rich in vitamin D are fat fish and sea mammals [2, 3, 4]. Ordinary meat, milk and eggs contain less vitamin D, and vegetables are void of vitamin D. The endogenous synthesis occurs in the skin under ultraviolet light B (UVB) stimulation [5]. The endogenous production depends on sun exposure, age, clothing, skin pigmentation, and use of sun protection [6, 7]. The exposure to UVB depends again on latitude, solar height, absorption in ozone layer and atmosphere, and reflection from clouds. In Boston (42°N) no previtamin D3 is produced in the period from November to February (inclusive) while in Edmonton, Canada (52°N) no production occurs between October and March [6, 7]. In comparison, previtamin D3 is produced all year in Los Angeles (24°N) and Puerto Rico (18°N). Consequently, circannual variations occur in plasma 25 OHD and PTH levels at higher latitudes in parallel to variations in the degree of sun exposure [8]. In Greenland, protective clothing is customary, and summer is short, with a low solar zenith altitude [7, 9]. Thus, yearly exposure to UVB sunlight is limited. However, the traditional Inuit fare rich in sea mammals contains large quantities of vitamin D [3].

During the last decades, significant cultural changes have occurred in Greenland [10]. Today, many Greenlanders are living on a westernized Danish fare that is low in natural dietary sources containing vitamin D [3, 11, 12]. Furthermore, Danish food is not fortified with vitamin D [13].

Summer is longer in Denmark (DK) than in Greenland and with a higher solar altitude [9]. Thus, exposure to UVB sunlight is more pronounced in DK than in Greenland [7]. Increased hip fracture risk has been reported in Inuit compared to white Caucasians [14], and there have been several reports of vitamin D deficiency and rickets among Inuit living in the Canadian arctic [15, 16]. In addition, lower plasma calcium levels have been reported in Greenlanders living in Greenland compared to Greenlanders and Danes living in Denmark [17]. However, no studies have compared vitamin D metabolism, parathyroid hormone levels, bone turnover, or calcium homeostasis between the two groups.

The aim of the present study was to compare the influence of latitude, season, diet, and ethnicity on vitamin D status, plasma PTH levels, and bone turnover between groups of Greenlanders from Greenland and caucasian Danes.

Subjects and Methods

We studied 188 males and females, aged 22–62 years, living in Denmark (n = 97) or in Greenland (n = 91). By ethnicity 43 were Danes and 145 were Greenlanders (Inuit). We recruited participants by announcement and notices in public institutions and private companies, and from Greenlander clubs in Denmark. Study design and recruitment of participants have been described in detail elsewhere [18]. In brief, we used the following exclusion criteria: 1) unwillingness to participate; 2) history or clinical signs of a disease in the heart, lungs, liver, kidney, brain, or endocrine organs; 3) arterial hypertension; 4) neoplastic disease; 5) daily use of medicine; and 6) alcohol abuse (more than 21 drinks a week for men and more than 14 drinks for women); 7) abnormal laboratory screening including b-hemoglobin, plasma creatinine, plasma aspartate amino transferase, urinary albumin, urinary hemoglobin, urinary glucose, and electrocardiogram.

We performed a dietary interview based on a 3-month recall and focusing on traditional Greenlandic food items, e.g., seal and whale. The diet was characterized as traditional Greenlandic when it contained seal and or whale at least once a week, and as westernized when it contained seal and or whale meat no more than three times a month.

We then allocated all participants to one of four study groups based on place of residence, ethnic status, and diet: Group A: 46 Greenlanders living in Nuuk, Greenland (64°N) on a traditional fare; Group B: 45 Greenlanders living in Nuuk on westernized fare; Group C: 54 Greenlanders living in Denmark (55°N) on westernized fare, and Group D: 43 Danes living in Denmark on Danish (westernized) fare (Table 1). All participants gave their informed consent. The study was performed in accordance with the Declaration of Helsinki II and approved by the regional Ethical Committees.

Table 1

Characteristics of studied subjects

Inuits in Nuuk (traditional fare)

Inuits in Nuuk (western fare)

Inuits in DK (western fare)

Danes in DK (western fare)

P-value

     

Age (yr)

41 (29–60)

36 (22–61)

41 (27–62)

44 (29–61)

0.15

     

Gender:

     

   No. males

23

15

22

21

     

   No. females

23

30

32

22

0.34

     

No. of blood samples:

     

   No. of pairs1

24

32

44

39

     

   Only winter

2

10

3

3

     

   Only summer

20

3

7

1

     

Plasma creatinine (µmol/l)

70 ± 2c,d

71 ± 2c,d

77 ± 1a,b,d

85 ± 2a,b,d

<0.001

     

DK: Denmark

1 Blood samples were obtained at both summer and at winter time

Post hoc t-test: P < 0.05 compared to a a group A, b group B, c group C, d group D

Measurements and Biochemistry

We obtained blood samples during May to September and November to April in 139 of the participants; in 31 participants, blood samples were only obtained during the summer and in 18 only during the winter (Table 1). Blood samples were stored at −80°C until analyzed. We measured plasma total calcium, creatinine, and albumin by standard laboratory methods. Plasma Ca was adjusted for individual variations in plasma albumin according to the following formula: adjusted plasma Ca (P-Caadj [mmol/l]) = P-Catotal (mmol/l) − 0.00086 * (650 − P-albumin [µmol/l]).

We measured plasma intact PTH by an IMMULITE® automated analyzer (Diagnostic Products Corp, Los Angeles, CA, USA). The CV in our lab was less than 7%. We determined plasma 25 OHD by equilibrium radioimmunoassay (RIA) (DiaSorin Inc., Stillwater, MN, USA). The inter- and intraassay CVs were 13%, and 10%, respectively. We measured plasma 1,25(OH)2D by a competitive radioreceptor assay (Nichols Institute Diagnostics, San Juan Capistrano, CA, USA). The inter- and intraassay CVs were 10%, and 8%, respectively. Plasma vitamin D-binding protein (DBP) was measured by an immunonephelometric method, as described by Lauridsen et al. [19]. The interassay CV was <5%. We calculated the free fraction of 1,25(OH)2D, the free 1,25(OH)2D index, as the molar ratio of 1,25(OH)2D to DBP. We measured plasma osteocalcin by ELISA using an automated instrument (Elecsys, 2010 immunoassay analyzer, Roche Diagnostics, Basel, Switzerland) with antibodies that recognize both intact- (1-49) and N-Mid-osteocalcin (1-43) [20]. The total CV was less than 5%. We determined bone ALP spectrophotometrically using an automated instrument (Hitachi 917, Roche Diagnostics) after lectin precipitation (Boehringer Mannheim, Germany). The total CV was less than 8%. To reduce analytical variation we analyzed PTH, vitamin D metabolites, and biochemical bone markers from each patient in the same run.

Statistics

As both summer and winter samples were obtained in just 78% of the participants, we analyzed data using two approaches: 1) all study participants (n = 188) were analyzed by unpaired analysis; and 2) participants with double samples (n = 136) were analyzed by paired analysis.

We tested all variables for normal distribution, and performed logarithmic transformation of data when indicated. We assessed differences between groups using chi-square tests for categorical variables. For continuous variables, we used a one-way ANOVA, Friedman’s test, a two-sample t-test or Wilcoxons’s signed rank test, as appropriate for the paired analysis. In the unpaired analysis we used a one-way ANOVA, Kruskal-Wallis H-test, a two-sample t-test; or Mann-Whitney U test. We used bivariate correlation analysis [Pearson’s correlation (r) or Spearman’s rho (R)] and multiple stepwise regression analysis to test correlations between variables. We used collinearity diagnostics in the multiple regression analysis to determine whether the independent variables were linearly related to one another. All results are mean ± standard error of the mean (SEM) unless otherwise stated. We performed all statistical analysis using the Statistical Package for Social Sciences (SPSS 8.0) for Windows.

Results

Table 1 shows participant characteristics by study group. Although mean plasma creatinine levels differed slightly among groups, no subjects had renal impairment, as determined by a plasma creatinine level above 115 µmol/l (Table 1). The results of the paired analyses are shown in Figures 1, 2, 3, 4; data from the unpaired analyses are shown in Table 2.

Table 2

Indices of calcium homeostasis and bone metabolism in Greenlanders (Inuits) and Danes by diet groups, unpaired analyses (mean ± sem)

Inuits in Nuuk (traditional fare) Group A

Inuits in Nuuk (western fare) Group B

Inuits in DK (western fare) Group C

Danes in DK (western fare) Group D

ANOVA P-value

     

SUMMERTIME

     

   Calcium (total) (mmol/l)

2.31 ± 0.02d

2.28 ± 0.02c,d

2.35 ± 0.01b

2.38 ± 0.02a,b

0.01

     

   Calcium (adj.) (mmol/l)

2.37 ± 0.02b

2.31 ± 0.02a,c,d

2.38 ± 0.01b

2.39 ± 0.01b

0.01

     

   PTH (pmol/l)

2.3 ± 0.2b,c,d,(S)

3.2 ± 0.3a,d

2.9 ± 0.2a,d

3.9 ± 0.2a,b,c

<0.001

     

   25OHD (nmol/l)

53 ± 3b,c,(S)

32 ± 2a,c,d

44 ± 2a,b,d,(S)

51 ± 3b,c,(S)

<0.001

     

   No. with vitamin D insufficiency

10 (23%)b,c

26 (74%)a,c,d

24 (47%)a,b,(S)

13 (33%)b

<0.001

     

   1,25(OH)2D (pmol/l)

88 ± 4b,c,d

74 ± 4a,d

69 ± 2a,d

58 ± 2a,b,c,(S)

<0.001

     

   D-binding protein (µmol/l)

4.0 ± 0.1c

3.8 ± 0.8c,(S)

4.3 ± 0.1a,b,d

4.0 ± 0.1c

<0.001

     

   Free 1,25(OH)2D Index (×10−5)

2.2 ± 0.1c,d

2.0 ± 0.1c,d

1.6 ± 0.1a,b

1.5 ± 0.1a,b,(S)

<0.001

     

   Bone-ALP (U/l)

85 ± 5

87 ± 5

86 ± 4

79 ± 5

0.42

     

   Osteocalcin (µg/l)

14.7 ± 1.5c,d,(S)

18.2 ± 1.7c,d

23.1 ± 1.2a,b

26.0 ± 1.5a,b

<0.001

     

WINTERTIME

     

   Calcium (total) (mmol/l)

2.33 ± 0.02d

2.30 ± 0.02c,d

2.35 ± 0.01b

2.38 ± 0.01a,b

0.01

     

   Calcium (adj.) (mmol/l)

2.36 ± 0.02

2.32 ± 0.02c,d

2.38 ± 0.01b

2.39 ± 0.01b

0.02

     

   PTH (pmol/l)

3.2 ± 0.4(S)

3.9 ± 0.3

3.5 ± 0.3

4.1 ± 0.2

0.06

     

   25OHD (nmol/l)

41 ± 3b,c,(S)

29 ± 2a,d

30 ± 2a,d,(S)

38 ± 2b,c,(S)

0.01

     

   No. with vitamin D insufficiency

11 (42%)b,c

34 (81%)a,d

36 (77%)a,d,(S)

22 (52%)b,c

0.01

     

   1,25(OH)2D (pmol/l)

83 ± 4d

81 ± 3d

73 ± 3

71 ± 4a,b,(S)

0.03

     

   D-binding protein (µmol/l)

4.2 ± 0.1c

4.3 ± 0.1(S)

4.4 ± 0.1a,d

4.1 ± 0.1c

0.02

     

   Free 1,25(OH)2D Index (×10−5)

2.0 ± 0.1c

1.9 ± 0.1c

1.7 ± 0.1a,b

1.7 ± 0.1(S)

0.03

     

   Bone-ALP (U/l)

95 ± 4

92 ± 5

94 ± 4

86 ± 5

0.30

     

   Osteocalcin (µg/l)

18.8 ± 1.5c,d,(S)

17.5 ± 1.0c,d

24.1 ± 1.3a,b,d

28.1 ± 1.4a,b,c

<0.001

     

Vitamin insufficiency was defined as plasma 25-hydroxyvitamin D < 40 nmol/l DK: Denmark

Post hoc t-test: P < 0.05 compared to a group A, b group B, c group C, d group D

(S) Effect of season i.e., summer and winter values differ significantly (p < 0.05)

https://static-content.springer.com/image/art%3A10.1007%2Fs00223-003-0110-9/MediaObjects/fig1.jpg
Figure 1

Plasma calcium (albumin adjusted) and PTH levels by study groups in summer and winter (mean ± sem).

https://static-content.springer.com/image/art%3A10.1007%2Fs00223-003-0110-9/MediaObjects/fig2.jpg
Figure 2

Plasma 25-hydroxyvitamin D levels by study groups in summer and winter (mean ± sem). Circannual variations occurred in all four study groups: Group A: P = 0.001, Group B: P = 0.03, Group C: P < 0.001, Group D: P < 0.001.

https://static-content.springer.com/image/art%3A10.1007%2Fs00223-003-0110-9/MediaObjects/fig3.jpg
Figure 3

Plasma 1.25-dihydroxyvitamin D levels by study groups in summer and in winter (mean ± sem). Circannual variations occurred in Group D only (P < 0.001).

https://static-content.springer.com/image/art%3A10.1007%2Fs00223-003-0110-9/MediaObjects/fig4.jpg
Figure 4

Plasma vitamin D-binding protein levels and the free 1,25(OH)2D index by study groups in summer and winter (mean ± sem). Circannual variations in the free 1,25(OH)2D index occurred in Group D only (P = 0.002).

Plasma Calcium

We observed no significant difference between summer and winter levels of albumin-adjusted plasma calcium in any of the four study groups (Table 2). Levels differed among groups in summer (P = 0.02 by ANOVA, paired analysis), as Greenlanders in Nuuk on a westernised fare (group B) had albumin-adjusted plasma Ca levels below the levels of the other three study groups (Fig. 1 and Table 2). Additionally, according to the unpaired analysis, plasma calcium levels in group B were below the levels of group C and D at wintertime (Table 2).

Plasma PTH

We found significant seasonal variations in plasma PTH levels in subjects living in Nuuk, i.e., in group A (Table 2) and B (Fig. 1), with the highest levels during winter. Furthermore, at summertime Greenlanders living in Nuuk on a traditional fare (Group A) had lower PTH levels and Danes (Group D) had higher PTH levels than the subjects in the other study groups (Table 2 and Fig. 1).

Vitamin D Metabolites

We observed circannual variations in plasma 25OHD levels in all four study groups, and plasma levels differed among groups during both summer and winter (Fig. 2 and Table 2). In both summer and winter, plasma 25OHD levels in groups B and C were below the levels of groups A and D (Fig. 2 and Table 2). Moreover, in summer the level in group B was below that in group C (Table 2, Fig. 2). As shown in Table 2, vitamin D insufficiency (<40 nmol/l) was most prevalent in Greenlanders on a westernized diet. However, frank vitamin D deficiency was uncommon as plasma vitamin D levels below 10 nmol/l were found in only three subjects (all in group B) during winter.

Seasonal variation in plasma 1,25(OH)2D levels occurred in Danes only (Group D, Fig. 3 and Table 2) with the highest levels during winter. However, plasma 1,25(OH)2D levels differed significantly among groups in both summer and winter. Greenlanders living in Nuuk on a traditional fare (Group A) had higher, whereas Danes (Group D) had lower, summer 1,25(OH)2D levels than subjects in the other three groups. In winter, Danes (Group D) had lower 1,25(OH)2D levels than Greenlanders living in Greenland (Groups A and B).

DBP levels exhibited significant seasonal variations in group B only (Fig. 4 and Table 2). At both summer- and at wintertime, Greenlanders living in Denmark (Group C) had higher DBP levels than group A and D (Fig. 5 and Table 2). Moreover, at summertime DBP levels were higher in group C than in group B (Fig. 4 and Table 2).

https://static-content.springer.com/image/art%3A10.1007%2Fs00223-003-0110-9/MediaObjects/fig5.jpg
Figure 5

Plasma bone alkaline phosphatase (bone-ALP) and osteocalcin levels by study groups in summer and winter (mean ± sem). Circannual variations in plasma osteocalcin levels occurred in groups C (P = 0.04) and D (P = 0.005).

Similar to plasma total 1,25(OH)2D levels, the free 1,25(OH)2D index exhibited seasonal variations in Danes only, with highest values during winter (Fig. 4 and Table 2). Moreover, between-groups differences existed in both summer and winter. At both seasons, the free 1,25(OH)2D index was lower in Greenlanders living in Denmark (Group C) than in Greenlanders living in Greenland (Groups A and B). In addition, in summer, the free 1,25(OH)2D index was lower in Danes (Group D) than in Greenlanders living in Greenland (Groups A and B), and in winter it was lower in Danes (Group D) than in Greenlanders living on a traditional fare in Greenland (Group A) (Fig. 4).

Biochemical Bone Markers

Plasma bone-ALP levels did not differ significantly among groups. However, the paired analysis revealed a significant seasonal variation in subjects living in Denmark (Group C: P = 0.04 and Group D: P = 0.04) with lower levels in summer than in winter (Fig. 5).

Plasma osteocalcin levels differed significantly between summer and winter in subjects living in Denmark irrespective of ethnicity (Groups C and D) (Fig. 5 and Table 2), with highest values during winter. Additionally, subjects living in Denmark (Groups C and D) had higher plasma osteocalcin levels during winter and summer than subjects living in Greenland (Groups A and B). Moreover, within Denmark, osteocalcin levels were higher in Danes (group D) than in Greenlanders living in Denmark (group C).

Relations Among Measured Indices

Table 3 shows the results of multiple regression analyses. Significant independent predictors of 25 OHD were age, season, diet, and ethnicity (P < 0.001), whereas gender and latitude did not reach statistical significance in the model. The positive partial regression coefficient between 25OHD and age (β = 0.687, P < 0.001) indicates increased 25OHD levels with increasing age. Similarly, the positive partial correlation with ethnicity (β = 9.254, P < 0.001) indicates higher 25OHD levels in Danes than in Greenlanders. On the other hand, the negative correlations with season (β = −10.975, P < 0.001) and diet (β = −11.601, P < 0.001) indicates lower 25OHD levels in winter than in summer, and lower 25OHD levels in subjects living on a westernized rather than a traditional diet. Significant independent predictors of plasma PTH levels were ethnicity (with higher levels in Danes than in Greenlanders, P < 0.001), plasma 25OHD levels (inverse correlation, P < 0.001) and age (increased levels with increasing age, P = 0.003). Ethnicity was the only independent predictor of the free 1,25(OH)2D index, with higher levels in Greenlanders than in Danes (P = 0.002). Finally, plasma levels of biochemical markers of bone turnover were directly correlated. However, ethnicity was found to influence plasma bone ALP and osteocalcin levels differently. Higher plasma osteocalcin levels were found in Danes than in Greenlanders (P < 0.001), whereas plasma bone ALP levels were higher in Greenlanders than in Danes (P = 0.001). Finally, bone ALP levels were inversely correlated with gender (β = −6.926, P = 0.03), indicating higher bone ALP levels in males than in females (Table 3).

Table 3

Multiple regression analyses on indices of bone turnover

Dependent varaible

Model summary

β-coefficients for significant predictors (95% CI)

P-value

    

25-hydroxyvitamin D1)

r = 0.61

Age

β = 0.687 (0.531; 0.842)

<0.001

    

P < 0.001

Season

β = −10.975 (−13.971; −7.980)

0.001

    

Diet

β = −11.601 (−15.446; −7.735)

<0.001

    

Ethnicity

β = 9.254 (5.668; 12.840)

0.001

    

PTH2)

r = 0.34

Ethnicity

β = 1.105 (0.577; 1.454)

0.001

    

P < 0.001

25(OH)D

β = −0.031 (−0.044; −0.018)

0.001

    

Age

β = 0.035 (0.012; 0.057)

0.003

    

Free 1,25(OH)2D index3)

r = 0.19

Ethnicity

β = −0.212 (−0.344; −0.080)

0.002

    

P = 0.002

    

Osteocalcin4)

r= 0.49

Bone-ALP

β = 0.122 (0.089; 0.156)

<0.001

    

P < 0.001

Ethnicity

β = 7.319 (5.213; 9.426)

<0.001

    

Bone-ALP4)

r = 0.43

Osteocalcin

β = 1.254 (0.903; 1.606)

<0.001

    

P < 0.001

Ethnicity

β = −16.608 (−23.567; −9.460)

0.001

    

Gender

β = −6.926 (−13.187; −0.665)

0.03

    

Plasma levels of calcitropic hormones and biochemical markers of bone turnover were investigated as dependent variables with season (1 = summer, 2 = winter), diet (1 = traditional Greenlandic fare, 2 = westernized fare), ethnicity (1 = Greenlandic, 2 = Dane), latitude (1 = Greenland, 2 = Denmark), Gender (1 = male, 2 = female), age, and plasma PTH, 25-hydroxyvitamin D 25(OH)D, free 1,25(OH)2D index, calcium (adj.), creatinine, osteocalcin, and, bone-ALP levels as potential independent predictors.

β: regression coefficient; r: correlation coefficient

1) In addition to shown significant independent predictors, gender and latitude were included in the model, showing no significant partial correlation

2) In addition to shown significant independent predictors, gender and plasma calcium and creatinine levels were included in the model, showing no significant partial correlation

3) In addition, age and gender as well as plasma levels of calcium, creatinine and PTH were included in the model, showing no significant partial correlation

4) In addition, age, gender and the free 1,25(OH)2D index as well as plasma levels of PTH and 25-hydroxyvitamin D were included in the model, showing no significant partial correlation

Discussion

We have shown that age, season, ethnicity and diet influence indices of calcium homeostasis and bone metabolism as measured by calcitropic hormones and biochemical markers of bone turnover.

Effect of Season and Latitude

Although the bivariate analyses showed differences between summer and winter levels in several measured indices, only plasma 25OHD levels were influenced by season in the multiple regressions analysis (Table 3). Plasma 25OHD levels varied significantly between summer and winter in all four study groups, with higher levels in summer than in winter. Thus, although summer is short in Greenland with a low solar zenith (64°N), endogenous vitamin D synthesis due to sun exposure may contribute significantly to plasma vitamin D levels [8].

Most surprisingly, latitude was not an independent predictor of 25OHD levels. Thus, although the length of the summer period varies with latitude, the shorter summer period in Greenland compared to Denmark does not, to any major degree, reduce P-25-hydroxyvitamin D levels. However, as season independently predicted plasma 25OHD levels, sun exposure is important. This may imply that existence of a period with sun exposure is more important than the actual length of the summer. The lower 25OHD levels in summer in Greenlanders living in Denmark compared to Danes may be explained by a reduced vitamin D production in the skin due to the increased skin pigmentation or to other clothing or sun-bathing habits.

Effect of Diet

In the multiple regression analysis, after inclusion of the effect of other potential variables, diet was a significant independent predictor of plasma 25OHD levels. The negative partial correlation coefficient implies that a traditional Inuit fare compared to a westernized fare is associated with higher 25OHD levels This finding agrees with the fact that the traditional Greenlandic fare is rich in vitamin D [2, 21, 22]. As shown in Figure 2, a change from a traditional to a westernized diet is associated with a decrease in plasma 25OHD levels, causing vitamin D insufficiency in a large proportion of Greenlanders living on a westernized fare (Table 2).

Effect of Ethnicity

We found that ethnicity independently influenced all measured indices of calcium homeostasis (25OHD), PTH, and the free 1,25(OH)2D index) and bone metabolism (bone ALP and osteocalcin levels), indicating that genetic differences may exist between Greenlanders and Caucasian Danes. Similarly, ethnicity has been shown to influence the occurrence of different diseases [23, 24]. Thus, as ethnical differences exist, pathogenetic mechanisms known to cause bone metabolic diseases in Caucasians cannot uncritically be applied to the Inuit population.

Our data suggest that further studies should focus on whether ethnical differences exist in the Ca-sensing receptor [25, 26], as low plasma Ca levels seem to be associated with relatively low plasma PTH levels in Greenlanders compared to Danes (Fig. 1). This finding may suggest a lower “set-point” for calcium-regulated PTH release in Greenlanders compared to Danes. Furthermore, we revealed higher 1,25(OH)2D concentrations in Greenlanders than in Danes (Tables 2 and 3).

As 1,25(OH)2D is a major determinant of intestinal calcium absorption, the higher levels in Greenlanders may partly compensate for a relatively low dietary Ca intake in parallel with a relatively high frequency of lactose intolerance in Greenlanders compared to Danes [17, 27]. However, the biological mechanism causing higher 1,25(OH)2D levels in Greenlanders compared to Danes is not obvious. Correction for differences in the vitamin D binding protein concentration did not cause major changes in the distribution of the free fraction among the four study groups (Figs. 3 and 4). Synthesis of 1,25(OH)2D in the renal tubules from 25OHD is catalyzed by the 25-hydroxyvitamin D-1α-hydroxylase. Key stimulators of this enzyme are high plasma PTH and low phosphate and calcium concentrations [28]. Greenlanders had lower plasma PTH levels than Danes. Thus, one explanation could be the lower plasma Ca concentrations in Greenlanders compared to Danes (Fig. 1). In addition, a higher intake of 1,25(OH)2D through food sources in Greenlanders compared to Danes may account for the observed differences in plasma 1,25(OH)2D concentrations [29, 30]. Finally, genetic selection may explain our finding of higher plasma 1,25(OH)2D levels in Greenlanders than in Danes. Due to a low endogenous 25OHD production in Greenlanders, caused by dark skin pigmentation and limited UVB exposure, evolution may have selected individuals with a relatively high 25-hydroxyvitamin D-1α-hydroxylase activity [31]. If so, this may also explain the relatively low plasma calcium and PTH concentrations in Greenlanders compared to Danes, as 1,25(OH)2D directly reduces PTH synthesis [32].

Effect of Age and Gender

Our finding of increasing plasma 25OHD levels with increasing age is surprising, as the endogenous synthesis of 25OHD in the skin normally decreases with age [33, 34]. However, our finding may be due to a quantitatively higher intake of dietary sources rich in vitamin D in the elderly compared to the younger age groups [35]. Our finding of an inverse association between age and plasma PTH levels is in accordance with other studies [36, 37, 38]. Moreover, our finding of an inverse correlation between gender and bone ALP (Table 3), indicating higher bone ALP levels in men than in women, is consistent with previous studies showing that before the age of 50 bone ALP levels are higher in men than in women [39].

Limitations to Study

Our Greenlandic population was divided into diet groups based on the content of traditional Greenlandic food items, i.e., the diet was characterized as traditional fare when it contained traditional items at least once a week. However, seasonal variations in the amount and composition of the Inuit diet have been reported with a higher intake of traditional food items during summer than during winter [35]. Thus, the seasonal variations in 25OHD levels observed in Greenlanders living on a traditional fare may partly be explained by a higher dietary vitamin D content during summer than during winter. Nevertheless, our finding of seasonal variations in plasma 25OHD levels in Greenlanders living on a westernized fare implies an effect of sun exposure at the latitude of Nuuk, Greenland. Similarly, another limitation to our study is that we did not measure the extent of outdoor activities since differences in lifestyle (clothing/sunbathing/exposure habits) may affect the results.

Summary

Calcium homeostasis and bone metabolism is influenced by ethnical differences. Greenlanders compared to Caucasian Danes have lower plasma PTH and calcium levels with concomitantly higher plasma 1,25(OH)2D levels. Compared to Danes, Greenlanders may have an inherent lower “set-point” for calcium-regulated PTH release or an enhanced renal 1,25(OH)2D production. In addition to ethnicity, age, season, and diet are important determinants of vitamin D status. In Greenlanders living in Greenland on a traditional fare with high vitamin D content, vitamin D status is comparable to Danes despite the shorter summer period in Greenland compared to Denmark, whereas vitamin D levels are reduced in Greenlanders living on a westernized fare. However, although most pronounced in Greenlanders living on a westernized fare, vitamin D insufficiency is common in both Danes and Greenlanders. Thus, vitamin D supplementation should be considered in both populations.

Acknowledgements

We thank secretary Inge Lise Walhovd, chief laboratory technician Lisbeth Mikkelsen and technicians Rikke Andersen, Elsebeth Fibiger, Dorthe Jensen, Jane Hagelskjaer Knudsen, Gitte Paulsen, Kirsten Tønder, Ellinor Hansen, Donna Lund, Lisbeth Flyvbjerg, and Charlotte Beck Gylling, Holstebro and Aarhus, and chief laboratory technician Inger Lise Kleist, Nuuk, for technical assistance. We thank Efraim Hans Olsen, Gunnar Pallisgaard, and Eivind Thorling for inspiring discussions and advice. The study was supported by grants from The Commission for Scientific Research in Greenland, Jyllands Postens Foundation, the Ellen Hansen Foundation, Eli Lilly’s Foundation for Research in Osteoporosis, Centre for Clinical Pharmacology at The University of Aarhus, Department of Experimental Clinical Research at The University of Aarhus, The Government of Greenland, and Aarhus University, Denmark.

Copyright information

© Springer-Verlag 2003