European Journal of Pediatrics

, Volume 174, Issue 2, pp 183–190

High prevalence of vitamin D insufficiency/deficiency in Dutch multi-ethnic obese children

Authors

    • Department of PediatricsSlotervaart Hospital
    • Diabetes Center, Department of Internal MedicineVU University Medical Center
  • Mariska van Vliet
    • Department of PediatricsSlotervaart Hospital
  • Ines von Rosenstiel
    • Department of PediatricsSlotervaart Hospital
  • Olivier Weijer
    • Department of PediatricsSlotervaart Hospital
  • Michaela Diamant
    • Diabetes Center, Department of Internal MedicineVU University Medical Center
  • Jos Beijnen
    • Department of Pharmacy & PharmacologySlotervaart Hospital
  • Dees Brandjes
    • Department of Internal MedicineSlotervaart Hospital
Original Article

DOI: 10.1007/s00431-014-2378-3

Cite this article as:
Radhakishun, N., van Vliet, M., von Rosenstiel, I. et al. Eur J Pediatr (2015) 174: 183. doi:10.1007/s00431-014-2378-3

Abstract

Vitamin D insufficiency/deficiency is common among non-white children; however, little is known about the prevalence of vitamin D insufficiency/deficiency in non-white obese children living in the Netherlands. Therefore, a retrospective analysis was performed on data from multi-ethnic Dutch children and adolescents 6–18 years who visited the obesity outpatient clinic in 2012–2013. We performed anthropometric measurements, oral glucose tolerance test, and measured 25(OH)D and lipid levels. Vitamin D insufficiency was defined as 25(OH)D levels 37.5- <50 nmol/L and vitamin D deficiency as 25(OH)D <37.5 nmol/L. In total, data from 387 children were obtained (mean age 11.6 years, 41.1 % boys, 10.3 % Dutch native, 25.6 % Turkish, 24.5 % Moroccan, 7.5 % African Surinamese, and 7.0 % West African). The median 25(OH)D level was 34 (range 12–105) nmol/L. In total, 17.8 % were vitamin D sufficient, 24.5 % with vitamin D insufficiency, and 57.6 % with vitamin D deficiency. Obese ethnic children showed the highest (87.5 %) and normal weight white children showed the lowest (20.0 %) prevalence of vitamin D insufficiency/deficiency . Conclusion: Vitamin D insufficiency and deficiency is extremely prevalent in treatment-seeking obese ethnic children. However, there was no evidence of an effect of vitamin D status on various components of the metabolic syndrome in our cohort.

Keywords

Vitamin DCalcitriolPediatric obesityEthnic groupsPrevalence

Abbreviations

25(OH)D

25-Dihydroxyvitamin D3

ALT

Alanine aminotransferase

BMI

Body Mass Index

HDL

High-density lipoprotein

HOMA-IR

Homeostasis model assessment for insulin resistance

IFG

Impaired fasting glucose

IGM

Impaired glucose metabolism

IGT

Impaired glucose tolerance

IR

Insulin resistance

LDL

Low-density lipoprotein

SD

Standard deviation

T2DM

Type 2 diabetes mellitus

WC

Waist circumference

Z-BMI

Body Mass Index standard deviation score for age and sex

Z-WC

Waist circumference standard deviation score for age and sex

Introduction

Vitamin D is obtained from the conversion of provitamin D in the skin to previtamin D3 after sun exposure to ultraviolet-B radiation [6]. The synthesis of vitamin D varies by the skin color; the darker the skin, the higher the amount of sunlight required to produce vitamin D [20]. When people with dark colored skin migrate to countries with a lower intensity and percentage of sunlight, their risk for vitamin D deficiency increases exponentially [31, 24, 20].

Data from epidemiologic studies among adults have found that vitamin D insufficiency/deficiency is associated with the risk of hypertension, cardiovascular disease, type 2 diabetes (T2DM) , and cardiovascular disease-related mortality [39]. In an American study among 411 obese children from 6 to 16 years, 25(OH)D levels (the major circulating form of vitamin D) were lowest in Afro-American , intermediate in Mexican American, and highest in white American children [26].

In the Netherlands, the largest non-Western ethnic populations originate from Morocco, Turkey, and Surinam. Multiple studies among these adult ethnic groups, both lean and obese, have shown a high prevalence of vitamin D insufficiency/deficiency as well [38]. Two pediatric Dutch studies showed that mean plasma 25(OH)D concentrations were significantly lower in Dutch-Moroccan and Dutch-Turkish children, as compared to white children in a primary school [23] and an outpatient clinic [17]. However, so far, there have been no studies performed on the prevalence of vitamin D insufficiency/deficiency in Dutch ethnic obese children. We hypothesize (1) that in our pediatric treatment-seeking obese cohort, there is a high prevalence of vitamin D insufficiency/deficiency in ethnic children compared to the Dutch native white children and (2) that 25(OH)D levels are associated with Body Mass Index standard deviation score for age and sex (Z-BMI) , Z-waist circumference, and fat percentage.

Materials and methods

This retrospective cohort study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Ethics Review Board of the hospital. In 2012–2013, data of children who were referred to the pediatric obesity outpatient clinic of the Slotervaart Hospital, Amsterdam (the Netherlands) were collected. Children and adolescents with data on anthropometrics and vitamin D status were included in the study. Subjects were excluded if they used glucose- or lipid-lowering drugs (n = 0), were diagnosed with diabetes mellitus (n = 4) or genetic syndromes (n = 1), had an endogenous cause of obesity (n = 4), metabolic rickets (n = 0), malabsorptive disorders (n = 1), or clinically significant hepatic or renal disease (n = 0). All children attending the obesity outpatient clinic who met the inclusion criteria were included in the sample.

Height was measured with a standardized stadiometer. Weight and body composition were measured with the bioimpedance analyzer Tanita BC-418 (Tanita Europe, Amsterdam, Netherlands) which measured weight to the nearest 0.1 kg and calculated estimates of fat percentage and fat (free) mass. Waist circumference was measured as the minimal circumference measurable on the horizontal plane between the lowest portion of the rib cage and iliac crest, with a tape measure, in a standing position after gently exhaling [42]. Pubertal stage was assessed according to the Tanner stages [37]. Blood pressure was measured three times with a 5-min interval on the right arm using a cuff appropriate for the arm circumference after 5 min of rest in seating position (Welch Allyn, NY, USA).

Each child underwent an oral glucose tolerance test (OGTT; 1.75 g/ kg with a maximum of 75 g). In addition, fasting blood samples were drawn for the assessment of 25-hydroxyvitamin D 25(OH)D, the major circulating form of vitamin D [24], total cholesterol, LDL-cholesterol , and HDL-cholesterol; triglycerides; and alanine aminotransferase (ALT) .

According to the latest pediatric guidelines [24], vitamin D status was defined as 25(OH)D levels of:
  • <12.5 nmol/L (5 ng/mL): Vitamin D severe deficiency

  • 12.5–37.5 nmol/L (5–15 ng/mL): Vitamin D deficiency

  • 37.5–50 nmol/L (15–20 ng/mL): Vitamin D insufficiency

  • > 50 nmol/L (20 ng/ ml): Vitamin D sufficiency

Vitamin D was supplemented when 25(OH)D levels of less than 50 nmol/L were found; the results of the efficacy and tolerability of this supplementation in part of this sample will be published elsewhere [29]. BMI and weight circumference (WC) values were standardized using Z-scores (the standard deviation score according to Dutch reference values) [12, 11]. Normal weight was defined as Z-BMI  <2.1, overweight was defined as Z-BMI between 2.1 and 2.3, and obesity was defined as Z-BMI  >2.3 [12]. Impaired glucose metabolism (IGM) was defined as either impaired fasting glucose (IFG, fasting glucose ≥5.6 and <7.0 mmol/L (13) ), impaired glucose tolerance (IGT, 2- h glucose ≥7.8 mmol/L(13)), or both. Insulin resistance was calculated according to the Homeostasis model assessment for insulin resistance (HOMA-IR) : fasting plasma insulin (IU/L) × fasting glucose (mmol/ L)/22.5 and defined as HOMA-IR ≥3.5 [21].

The metabolic syndrome was diagnosed when a child was obese (Z-BMI >2.3) and two or more of the following criteria were present [40]: IGT, a triglyceride level ≥95th percentile for age and sex [25], a high-density lipoprotein (HDL) cholesterol level <5th percentile for age and sex [25], and/or a mean (systolic or diastolic) blood pressure ≥95th percentile for height and sex [7]. High low-density lipoprotein (LDL) cholesterol was considered present when levels were ≥95th percentile for age and sex [25], and serum ALT was considered high when levels reached over 30 IU/L [32].

Skin type was assessed according to the Fitzpatrick Skin Type Classification Scale: 1: pale white, 2: white or fair, 3: cream white or quite common, 4: moderate brown or typical Mediterranean skin, 5: brown, Middle Eastern or Asian, 6: deeply pigmented dark brown to black [10]. Ethnicity was defined as Dutch native (mean skin type 2.8) if both parents were Dutch native (white) , and Moroccan (mean skin type 4.0), Turkish (mean skin type 4.0), Indian (mean skin type 5.0), or West African (mean skin type 6.0) in case where both parents originated from that specific country or region. Ethnicity was defined as African Surinamese (mean skin type 6.0) when both parents were from Surinam and originally from African descent. All other ethnicities or children with mixed ethnicities were collected in an ‘other’ group (mean skin type 4.6).

Plasma 25(OH)D was measured by the Architect ci8200 immunoassay analyzer (Abbott, IL, USA), with intra- and inter-assay coefficients of variation of 3.7 and 3.8 %, respectively. The laboratory participated in the Vitamin D External Quality Assessment Schema (DEQAS) . Plasma glucose levels, total cholesterol, triglycerides, and HDL-cholesterol were measured by standardized validated methods (SYNC HRON LX20, Beckman Coulter, USA and MODULAR ANALYTICS EVO solution, Roche, Belgium). LDL-cholesterol was calculated by the Friedewald formula, as applicable. Plasma insulin levels were measured by an immunoluminometric assay (Immulite 200 system, DPC, Los Angeles, USA; intra-assay variation, 3–6 %; inter-assay variation, 3–5 %).

Mean (±standard deviation (SD) ), median (interquartile range), or percentages are shown. Differences in clinical and biochemical characteristics were tested by Student’s t tests, χ2 tests, or ANOVA. When appropriate, further comparisons were made using the Bonferroni post hoc test. Variables with a skewed distribution were log-transformed before analysis. Interaction was tested by including product terms into the regression models, and stratified analyses were performed when P  <  0.05. Linear regression analysis was used to determine the association of 25(OH)D levels with the different continuous variables adjusted for confounders. A P value of <0.05 was considered statistically significant. All analyses were performed with SPSS version 21.0 for Windows (IBM Statistics, Chicago, IL, USA).

Results

Table 1 shows the baseline characteristics of the study population (age range 3–19 years, 41.1 % boys) according to the vitamin D status. The median 25(OH)D level was 34.0 (range 12.0–97.0) nmol/L. Of all children, 24.5 % was diagnosed with vitamin D insufficiency and 57.6 % with vitamin D deficiency. Vitamin D insufficiency/deficiency was most common in obese children (83.7 %) compared to overweight (63.6 %) and normal weight children (66.7 %) (P = 0.038). Children with vitamin D insufficiency/deficiency had a higher Z-BMI , waist-circumference , and fat percentage compared to vitamin D-sufficient children (Table 1).
Table 1

Baseline characteristics stratified by vitamin D status

Characteristics

Total

Vitamin D sufficiency

Vitamin D insufficiency/deficiency

P value

Total, n (%)

387 (100)

69 (17.8)

318 (82.2)

NA

Age, years

11.6 ± 3.2

11.3 ± 3.5

11.8 ± 3.2

0.188

Pubertal stage

2.67 ± 1.55

2.55 ± 1.59

2.69 ± 1.55

0.513

Gender

   

0.687

 Boys

159 (41.1 %)

30 (43.5 %)

129 (40.6 %)

 

 Girls

228 (58.9 %)

39 (56.5 %)

189 (59.4 %)

 

Ethnicity, n (%)

   

<0.001

 Dutch native

38 (9.8)

23 (33.3)

15 (4.7)

 

 Moroccan

95 (24.5)

14 (20.3)

81 (25.5)

 

 Turkish

99 (25.6)

14 (20.3)

85 (26.7)

 

 Indian

23 (5.9)

1 (1.4)

22 (6.9)

 

 West African

27 (7.0)

0 (0.0)

27 (8.5)

 

 African Surinamese

29 (7.5)

4 (5.8)

25 (7.9)

 

 Other

76 (19.6)

13 (18.8 %)

63 (19.8 %)

 

Skin type

   

<0.001

 2

13 (3.6 %)

8 (11.6 %)

5 (1.6 %)

 

 3

33 (8.5 %)

15 (21.7 %)

18 (5.7 %)

 

 4

221 (57.1 %)

37 (53.6 %)

184 (57.9 %)

 

 5

48 (12.4 %)

5 (7.2 %)

43 (13.5 %)

 

 6

72 (18.6 %)

4 (5.8 %)

68 (21.4 %)

 

Vitamin D supplementation, n (%)

30 (7.8)

7 (10.1)

23 (7.2)

0.455

Time spend outside, h/ day

2.0 (1.0–3.5)

2.5 (1.5–3.6)

2.0 (1.0–3.5)

0.165

Dairy intake, units/day

1.0 (0.5–2.0)

1.0 (0.5–2.0)

1.0 (0.5–2.0)

0.469

Body mass index, kg/m2

28.9 ± 5.3

27.1 ± 5.0

29.3 ± 5.3

0.002

Z-BMI

3.2 ± 0.67

3.0 ± 0.7

3.2 ± 0.6

0.012

Waist circumference, cm

94.3 ± 14.8

90.8 ± 13.5

95.1 ± 15.0

0.035

Z-WC

3.7 ± 1.2

3.6 ± 1.1

3.7 ± 1.2

0.403

Fat percentage, %

38.0 ± 7.1

35.6 ± 6.7

38.3 ± 7.1

0.023

Systolic blood pressure, mmHg

115.2 ± 11.7

115.6 ± 13.1

115.1 ± 11.4

0.771

Diastolic blood pressure, mmHg

70.0 ± 7.8

70.9 ± 8.6

69.8 ± 7.6

0.297

Hypertension, n (%)

53 (13.7)

11 (15.9)

42 (13.2)

0.560

Fasting glucose, mmol/ L

5.2 ± 0.4

5.1 ± 0.4

5.2 ± 0.4

0.773

2- h glucose, mmol/ L

5.7 ± 1.1

5.7 ± 1.2

5.7 ± 1.1

0.683

HbA1C (DCCT), %

5.4 ± 0.3

5.3 ± 0.2

5.4 ± 0.3

0.159

Impaired glucose metabolism

61 (15.8 %)

13 (18.8 %)

48 (15.1 %)

0.361

Insulin, pmol/ L

100.0 (70.0–146.0)

91.0 (63.5–130.0)

102.0 (72.8–148.0)

0.040

HOMA-IR

3.3 (2.2–5.0)

2.8 (2.1–4.4)

3.4 (2.3–5.0)

0.047

Insulin resistance

177 (45.7 %)

26 (37.7 %)

151 (47.5 %)

0.145

Total cholesterol, mmol/ L

4.26 ± 0.76

4.32 ± 0.78

4.25 ± 0.75

0.436

LDL-cholesterol, mmol/ L

2.66 ± 0.67

2.71 ± 0.66

2.65 ± 0.67

0.546

HDL-cholesterol, mmol/ L

1.17 ± 0.24

1.20 ± 0.22

1.17 ± 0.24

0.384

Triglycerides, mmol/ L

0.9 (0.6–1.3)

0.9 (0.6–1.3)

0.9 (0.6–1.3)

0.654

Dyslipidemia

138 (35.7 %)

24 (34.8 %)

114 (35.8 %)

0.891

ALT, U/ L

21.0 (17–28.0)

21.0 (18.0–28.5)

21.0 (16.0–28.0)

0.647

Fatty liver

84 (21.7 %)

17 (24.6 %)

67 (21.1 %)

0.522

Metabolic syndrome

25 (6.5 %)

2 (2.9 %)

23 (7.2 %)

0.279

Data are expressed as mean ± SD, N (%) or median (IQR) for skewed data (log-transformed prior to analyses). Differences tested by independent samples t test or chi-square. Vitamin D sufficiency >50 nmol/L, insufficiency/deficiency ≤50 nmol/L. Pubertal stage according to Tanner. Skin type according to the Fitzgerald Skin Type Classification Scale

Z-BMI standardized BMI according to age and sex, Z-WC standardized waist circumference according to age and sex, HOMA-IR homeostasis model assessment of insulin resistance, HDL high-density lipoprotein, LDL low-density lipoprotein, ALT alanine aminotransferase, Ethnicity place of birth of both parents, other all other ethnicities or children with mixed ethnicities

Ethnic children had a higher fat percentage compared to non-ethnic (white) children (mean 35.5 SD = 6.9 vs. mean 38.2, SD = 7.1, P = 0.043). There were no differences in age, Z-BMI , Z-waist circumference, or vitamin (D) supplementation use between ethnic and white children. The prevalence of vitamin D insufficiency/deficiency was 87.5 % in ethnic obese children and 20 % in normal weight white children. The median 25(OH)D per ethnic group are shown in Table 2.
Table 2

Median and interquartile ranges (IQR) of 25(OH)D Levels among multi-ethnic children according to subgroups

Variable

Number (%)

Median 25(OH)D

P value

ANOVA

Post hoc

Overall

387

34.0 (26.0–47.0)

Pubertal stage

  

0.324

 

 Ia

142

34.0 (27.0–48.0)

  

 II

37

37.0 (24.0–48.5)

 

1.000

 III

55

31.0 (25.0–44.0)

 

1.000

 IV

73

34.0 (23.5–43.0)

 

1.000

 V

63

33.0 (26.0–42.0)

 

1.000

Ethnicity

  

<0.001

 

 Dutch nativea

40

57.5 (38.8–72.5)

  

 Moroccan

95

36.0 (27.0–47.0)

 

<0.001

 Turkish

99

33.0 (25.0–45.0)

 

<0.001

 Indian

23

28.0 (26.0–33.0)

 

<0.001

 West African

27

26.0 (23.0–36.0)

 

<0.001

 African Surinamese

29

31.0 (25.0–39.0)

 

<0.001

 Other

75

33.0 (27.0–46.5)

 

<0.001

Skin type

  

<0.001

 

 2a

12

60.0 (48.0–68.5)

  

 3

33

42.0 (30.0–68.0)

 

0.757

 4

221

36.0 (26.5–47.0)

 

<0.001

 5

49

29.0 (25.0–36.0)

 

<0.001

 6

67

29.5 (24.3–37.0)

 

<0.001

Season

  

<0.001

 

 Winter

65

27.0 (22.0–30.0)

 

<0.001

 Spring

73

29.0 (24.0–40.0)

 

<0.001

 Summera

120

48.0 (35.3–55.0)

  

Fall

129

33.0 (27.5–43.5)

 

<0.001

Z-BMI

  

0.048

 

Normal weight

21

40.0 (29.5–60.5)

 

0.199

Overweight

11

41.0 (29.0–60.0)

 

1.000

Obesea

355

33.0 (26.0–47.0)

  

Pubertal stage was defined according to Tanner stages. Skin type according to the Fitzgerald Skin Type Classification Scale

Ethnicity place of birth of both parents, other all other ethnicities or children with mixed ethnicities

aDifferences were tested by ANOVA with Bonferroni post hoc analysis. P value for ANOVA and post hoc P value for comparison

There were no differences in median 25(OH)D between boys and girls. Girls had a significantly darker skin type (mean 4.4 SD = 1.1 vs. mean 4.2 SD = 0.9, P = 0.044) and a higher fat percentage (mean 39.8 SD = 6.5 vs. mean 35.3 SD = 7.2, P  <  0.001) compared to boys. Boys had a higher mean Z-BMI (3.3 SD = 0.7 vs. 3.0 SD = 0.6, P  <  0.001) and mean waist circumference standard deviation score for age and sex (Z-WC) (4.2 SD = 1.4 vs. 3.4 SD = 0.9, P  <  0.001). There were no sex-specific differences with regard to age, vitamin D supplementation use, 2 h glucose, insulin, HOMA-IR , total LDL- or HDL-cholesterol , or triglycerides levels (data not shown).

The metabolic syndrome was present in 25 children (6.5 %), two of these children were vitamin D sufficient and 23 were vitamin D insufficient/deficient (P = 0.092). Since sex was a significant effect modifiers in regression analyses of 25(OH)D and cardiometabolic risk factors, further analyses were stratified. 25(OH)D levels were associated with respectively Z-BMI (B −0.091, 95 % CI −0.171 to −0.011, P = 0.026), fat percentage (B −0.010, 95 % CI −0.019 to −0.001, P = 0.036), and HDL-cholesterol (B 0.216, 95 % CI 0.00 to 0.432, P = 0.050) in girls after adjustment for pubertal stage, season, and skin type. There were no significant associations in between 25(OH)D levels and any cardiovascular variable in boys.

Discussion

In the present study, alarmingly, high percentages of vitamin D insufficiency/deficiency were found, being more prevalent in ethnic obese vs. normal weight white children (87.5 vs. 20.0 %). This is the first study to determine vitamin D status in a sample of mainly obese treatment-seeking multi-ethnic Dutch children. A large European study reported vitamin D insufficiency/deficiency in 39 % of the healthy adolescents [14]. Other Dutch studies in both lean adults and children have shown a higher prevalence of vitamin D insufficiency/deficiency in subjects from Moroccan and Turkish descent, as compared to whites [23, 17].

Compared to other studies in obese children, we found extremely high percentages of vitamin D insufficiency/deficiency [1, 31, 26]. To illustrate, Alemzadeh et al. showed a prevalence of vitamin D insufficiency/deficiency of 46 % in obese Hispanic and Afro-American children from an endocrine clinic in Milwaukee (54 % sunshine/year ) [1], and a study performed in Texas (61 % sunshine/year ) reported that 50 % of their obese outpatient clinic cohort from Caucasian, Hispanic, and African-American descent were vitamin D insufficient/deficient [26]. Both studies included children throughout the year to exclude the influence of season, and the distribution of ethnic and white children was quite similar; however, the average sunshine per year in the Netherlands is only 35.8 %. Moreover, the ethnic children from the USA originate from different countries compared to European migrants. A population-based study in Germany reported a prevalence of vitamin D insufficiency/deficiency in 45 % of the immigrant children, of which, most were from Turkish descent [16]. A study among 168 children in Turkey reported vitamin D insufficiency/deficiency in 98 % of the cohort [2]. Surprisingly, they found no differences in vitamin D status between normal weight, overweight, and obese children [2]. The high frequency of vitamin D insufficiency/deficiency might have led to such a narrow range of 25(OH)D concentrations that true differences could not be detected.

In our cohort, 84 % of the obese children showed vitamin D insufficiency/deficiency compared to respectively 64 % of overweight and 67 % of the normal weight children. Even though the group of normal and overweight children was quite small in our sample, this difference was significant and could not be explained by confounding factors. Furthermore, children with vitamin D insufficiency/deficiency had a higher fat percentage and waist circumference compared to vitamin D-sufficient children. The relationship between vitamin D insufficiency/deficiency and obesity is confirmed by several other studies in adults and children [31, 1]. However, its (patho)physiological mechanism remains unknown. It is hypothesized that vitamin D sequesters in the fat cells, causing an intravascular vitamin D deficiency [19]. On the other hand, adipose tissue presents vitamin D receptors, which take part in the regulation of fat cells, and may therefore contribute to the increase of fat mass during vitamin D deficiency [9]. Indeed, Gonzalez-Molero et al. found in a prospective study of 1,226 adults that obesity did not predict vitamin D status, but low 25(OH)D levels (<42.5 nmol/L) were significantly associated with an increased risk to develop obesity in the next 4 years [15]. Future studies are needed to confirm these findings.

The lowest value of 25(OH)D in our cohort was 12 nmol/L; 25(OH)D levels this low have been reported in several case reports of children with rickets in the Netherlands [33, 8, 4]. Although these children were all younger than 4 years, rickets can occur until the growth plates are closed. However, the development of rickets is dependent not only on the severity of the vitamin D deficiency but also on the duration of the deficiency, the rate of the child’s growth, and the dietary calcium content [27]. We did not measure calcium levels, however daily dairy intake did not differ between the vitamin D-sufficient and vitamin D-insufficient /deficient group.

Besides rickets, vitamin D deficiency may have other serious health consequences. A meta-analysis showed a significant higher risk for cardiovascular disease (relative risk 1.42, 95 % confidence interval 1.19–1.71) and related mortality (relative risk 1.38, 95 % confidence interval 1.21–1.57) in adults with 25(OH)D concentrations between 20 and 60 nmol/L [39]. In children, no associations between vitamin D status and cardiometabolic risk factors have been reported [5, 36], while others did find significant associations between vitamin D and hypertension [41], insulin sensitivity [1], and the metabolic syndrome [31]. Almost all children with the metabolic syndrome in our cohort were vitamin D insufficient/deficient (92 %); however, this was not significant probably due to the low prevalence of the metabolic syndrome overall.

In the present study, low serum concentrations of 25(OH)D were significantly dependent on weight status, ethnicity, season, and skin type. Surprisingly, only in girls, small though significant associations were found between 25(OH)D levels and respectively Z-BMI , fat percentage, and HDL-cholesterol . Evidence for sex-dependent associations between weight measures, cholesterol, and vitamin D status has been reported by others [30, 22]. Rajakumar et al. studied 237 children (mean age 12.7, SD = 2.2, 47 % black, 47 % obese, 43 % male) and found significant associations between vitamin D deficiency and adiposity indices in white females, but not in black females or white/black males [30]. Another study from the USA in 58 multi-ethnic obese children found significant associations between total body fat (measured by DEXA scan) and 25(OH)D levels as well [18]. However, no significant effect of sex was found [18]. The smaller sample size or differences in study population could have caused these different findings.

Besides the strong points of this study (inclusion of a large, multi-ethnic cohort), several limitations should be acknowledged. These include the lack of a normal weight control group, which precludes generalization of our results. Second, due to the cross-sectional nature of this study, a causal link cannot be made. Third, this is an outpatient clinic cohort and, therefore, not similar to the overweight and obese children from the general population. Fourth, measurements of parathyroid hormone were not performed, so vitamin D deficiency due to hyperparathyroidism could be missed. However, the treatment of these children would still be vitamin D supplementation, which was already given by the pediatrician [29]. Fifth, only total 25(OH)D levels were measured, while the bio-available (free) vitamin D (the fraction not bound to vitamin D-binding protein) correlates stronger with measures of bone mineral density in adult subjects [34, 3, 28]. Although this relation has not been determined for obese children, it would have been beneficial to determine the prevalence of vitamin D insufficiency/deficiency according to free 25(OH)D concentrations as well. Finally, even though bio-impedance analysis is a safe and easy way to measure body composition, it has shown to be less accurate in obese children compared to the DEXA scan [13, 35].

In conclusion, vitamin D insufficiency/deficiency is extremely common in this cohort consisting of mainly obese treatment-seeking multi-ethnic Dutch children. Almost 88 % of the obese ethnic children had either vitamin D insufficiency or deficiency compared to 20 % in the normal weight white children. However, in our cohort, there was no evidence of an effect of vitamin D status on various components of the metabolic syndrome. Pediatricians should be aware of the high prevalence of vitamin D insufficiency/deficiency to prevent severe vitamin D deficiency in these risk groups.

Conflict of interest

The authors declare no conflict of interest.

Copyright information

© Springer-Verlag Berlin Heidelberg 2014