Prevalence and predictors of vitamin D deficiency in five immigrant ...

European Journal of Clinical Nutrition (2005) 59, 57–63

& 2005 Nature Publishing Group All rights reserved 0954-3007/05 $30.00 www.nature.com/ejcn

ORIGINAL COMMUNICATION Prevalence and predictors of vitamin D deficiency in five immigrant groups living in Oslo, Norway: the Oslo Immigrant Health Study K Holvik1,2*, HE Meyer1,2, E Haug3 and L Brunvand4 1 Department of General Practice and Community Medicine, University of Oslo, Norway; 2Division of Epidemiology, Norwegian Institute of Public Health, Oslo, Norway; 3Hormone Laboratory, Aker University Hospital, Oslo, Norway; and 4Department of Paediatrics, Ulleva˚l University Hospital, Oslo, Norway

Objective: To study the prevalence of vitamin D deficiency and to identify possible predictors of vitamin D deficiency in five main immigrant groups in Oslo. Design: Cross-sectional, population-based. Setting: City of Oslo. Subjects: In total, 491 men and 509 women with native countries Turkey, Sri Lanka, Iran, Pakistan and Vietnam living in the county of Oslo. Results: Median serum 25(OH)D level (s-25(OH)D) was 28 nmol/l, ranging from 21 nmol/l in women born in Pakistan to 40 nmol/l in men born in Vietnam. Overall prevalence of vitamin D deficiency defined as s-25(OH)Do25 nmol/l was 37.2%, ranging from 8.5% in men born in Vietnam to 64.9% in women born in Pakistan. s-25(OH)D did not vary significantly with age. s-25(OH)D was higher in blood samples drawn in June compared to samples obtained in April, but not significantly for women. Reported use of fatty fish and cod liver oil supplements showed a strong positive association with s-25(OH)D in all groups. Education length was positively associated with s-25(OH)D in women, whereas body mass index (BMI) was inversely associated with s-25(OH)D in women. These two variables were not related to vitamin D deficiency in men. Conclusions: There is widespread vitamin D deficiency in both men and women born in Turkey, Sri Lanka, Iran, Pakistan and Vietnam residing in Oslo. The prevalence of vitamin D deficiency is higher in women than in men, and it is higher in those born in Pakistan and lower in those born in Vietnam compared to the other ethnic groups. Fatty fish intake and cod liver oil supplements are important determinant factors of vitamin D status in the groups studied. BMI and education length are also important predictors in women. Sponsorship: Supported by the Directorate for Health and Social Affairs, Oslo, and Research Forum, Aker University Hospital, Oslo.

European Journal of Clinical Nutrition (2005) 59, 57–63. doi:10.1038/sj.ejcn.1602033 Published online 28 July 2004 Keywords: vitamin D deficiency; immigrants; 25-hydroxyvitamin D

Introduction *Correspondence: K Holvik, Department of General Practice and Community Medicine, University of Oslo, PO Box 1130 Blindern, N-0318 Oslo, Norway. E-mail: [email protected] Guarantor: K Holvik. Contributors: KH performed the data analysis and prepared the manuscript with HM, who was also a responsible organizer of the Oslo Immigrant Health Study. EH was responsible for the blood sample analyses. LB initiated the vitamin D substudy, and contributed to the design of the paper. All coauthors have critically revised and approved the manuscript. Received 17 March 2004; revised 26 May 2004; accepted 17 June 2004; published online 28 July 2004

It is well known that vitamin D deficiency leads to rickets in children and osteomalacia in adults (Holick, 2003). It also seems to be a major risk factor for osteoporosis in elderly people (Lips, 2001). The most reliable marker of vitamin D status is the serum concentration of the main circulating metabolite, 25-hydroxyvitamin D (25(OH)D), which correlates well with the body’s vitamin D stores (Holick, 2003). Its half-life is about 3 weeks and its concentration reflects vitamin D supply from both cutaneous production and diet (Lips, 2001). There is no common agreement of a lower reference limit for the safe level of 25(OH)D, although it has been proposed that levels below 50 nmol/l should indicate

Vitamin D deficiency in five immigrant groups in Oslo K Holvik et al

58 mild vitamin D deficiency, while levels below 25 nmol/l should indicate moderate vitamin D deficiency (Lips, 2001). Clinical signs of vitamin D deficiency were practically eradicated in Norway in the 1950s, but have re-emerged during the last decades in certain immigrant groups (Brunvand & Nordshus, 1996). Vitamin D deficiency may be exacerbated when moving from low-latitude to high-latitude countries. There are currently about 50 000 immigrants of Asian background living in Oslo, constituting about 10% of the total population in Oslo (Statistics Norway, 2003). Vitamin D deficiency has been demonstrated to be common in pregnant Pakistani women living in Oslo (Brunvand & Haug, 1993; Henriksen et al, 1995; Falch & Steihaug, 2000) as well as in children of immigrant background living in Norway (Brunvand & Nordshus, 1996; Brunvand & Brunvatne, 2001). A recent study from Oslo showed a very high prevalence of vitamin D deficiency and secondary hyperparathyroidism in both men and women with Pakistani background living in Oslo compared to ethnic Norwegians (Meyer et al, 2004). Only 10% of the women and 8% of the men with Pakistani background had s-25(OH)D of 50 nmol/l or more, compared to 86% of men and women with Norwegian background. A high prevalence of vitamin D deficiency among immigrants has also been reported in other Northern European countries, such as Denmark (Glerup et al, 2000; Pedersen et al, 2003) and the UK (Shaw & Pal, 2002; Hamson et al, 2003). However, the evidence is limited and the preponderance of studies are not populationbased but conducted in clinical settings. Asian Indians are the ethnic group most commonly studied. We are not aware of any studies that have evaluated vitamin D status in the adult population of ethnic groups living in Norway other than the Pakistanis. The aims of our study were to describe the prevalence of vitamin D deficiency as well as to identify predictors of vitamin D deficiency in some of the main immigrant groups in Oslo. To do so, we measured s-25(OH)D in five important immigrant groups (Turkey, Sri Lanka, Iran, Pakistan, and Vietnam) that already participated in a large populationbased study (The Oslo Immigrant Health Study). We also analysed background data from this study providing information about risk factors for vitamin D deficiency.

Methods The Oslo Immigrant Health Study was an extension of the Oslo Health Study (HUBRO) 2000–2001. It was a populationbased cross-sectional study conducted under the joint collaboration of the Norwegian Institute of Public Health and the University of Oslo. The study was carried out during the period February–November 2002.

Population All persons between the ages 31–60 y (main cohort) born 1942–1971, and 20–30 y (additional cohort) born 1972–1982 European Journal of Clinical Nutrition

with native countries Turkey (401N), Sri Lanka (71N), Iran (361N), Pakistan (341N), and Vietnam (111N) residing in Oslo (601N) were invited to participate in the Oslo Immigrant Health Study. The study excluded individuals from the five age cohorts that had previously been invited to the main HUBRO study (born in 1970, 1960, 1955, 1940–1941 and 1924–1925). According to the 2001 population register, 7972 individuals in the main cohort (31–60 y) were eligible for participation. Of these, 11 individuals had died and 72 had emigrated prior to the invitation, leaving 7889 eligible for participation. Of the 7607 reached by mail, 3019 gave their written consent and met the criteria of inclusion (completion of at least one question in either of the questionnaires), attaining a final response rate of 39.7%. Participation rates according to country of birth for main cohort are as follows; Turkey 32.7%, Sri Lanka 50.9%, Iran 38.8%, Pakistan 31.7% and Vietnam 39.5% respectively. From the additional cohort (20–30 y) 4116 individuals were eligible for participation. Of these, 60 individuals were either dead or had emigrated prior to the invitation. Of the 3782 reached by mail, 707 (18.7%) participated. The participation rates for the 20–30 y olds are as follows; Turkey 18.3%, Sri Lanka 24.7%, Iran 20.4%, Pakistan 15.4% and Vietnam 15.2%. The dropout problem was similar to that in the Oslo Health Study, which is described elsewhere (Søgaard et al, 2004). All persons included between 3 April and 10 June had their s-25(OH)D analysed. This included a total of 1000 persons (491 men and 509 women) distributed on all five ethnic groups. Mean age was 39 y for men and 37 y for women (P ¼ 0.001).

Data collection All participants underwent a clinical examination and filled in a self-administered questionnaire. The clinical examination included anthropometric measures (height, weight, waist and hip circumference) and the collection of nonfasting blood samples. The questionnaire included questions about diet and supplements, physical activity and education level. Physical activity level was defined as sedentary if main activities during leisure time was reading or watching TV, moderate if walking, cycling or moving about in some other way at least four times per week, and high if taking part in physical exercise or competitive sport at least four times per week.

Blood sample analyses Serum samples were kept frozen at 701C until analysed. s25(OH)D was measured by radioimmunoassay (DiaSorin, Stillwater, MN, USA) in the Hormone Laboratory, Aker University Hospital. This assay measures both 25(OH)D3 and 25(OH)D2. The intra- and interassay coefficients of variation (CVs) were 6 and 14–15%, respectively. The limit of detection was 6 nmol/l.


59

Ethics The study protocol was approved by the Regional Committee for Medical Research Ethics and the Norwegian Data Inspectorate.

80 70 60 Serum 25 (OH)D (nmol/l)

Statistical analyses As the distribution of s-25(OH)D was highly skewed, we performed nonparametric statistical tests. The prevalence of vitamin D deficiency was assessed by logistic regression analyses. We defined vitamin D deficiency as having s-25(OH)D below 25 nmol/l, as proposed by Lips (2001). Additional analyses were performed stratified on ethnic groups.

50 40 30 20 10

Results Overall median s-25(OH)D was 30 nmol/l in men and 27 nmol/l in women (P ¼ 0.000, Table 1). Median s-25(OH)D was low (o50 nmol/l) in all ethnic groups and very low (o25 nmol/l) in those born in Pakistan (Figure 1). This applied to both men and women (Table 1). Median s-25(OH)D was significantly lower in women than in men for those born in Turkey (P ¼ 0.000), Iran (P ¼ 0.033), and Pakistan (P ¼ 0.014), but not for those born in Sri Lanka and Vietnam. In all, 37.2% (372 persons) had vitamin D deficiency according to our definition (s-25(OH)D o25 nmol/l), and the prevalence was highest in those born in Pakistan (Table 1). A higher proportion of women than men had vitamin D deficiency (43.0 vs 31.2%, P ¼ 0.000). The lowest s-25(OH)D was seen in women born in Pakistan, of whom 64.9% had vitamin D

Table 1 Serum 25(OH)D levels (nmol/l) according to sex and native country (all ages included)

N Men Turkey Sri Lanka Iran Pakistan Vietnam Men total Women Turkey Sri Lanka Iran Pakistan Vietnam Women total Total

Median (25, 75 percentile) (26, (22, (24, (19, (33,

s-25(OH)D 450 nmol/l (%)

s-25(OH)D o25 nmol/l (%)

87 155 108 94 47

33 30 31 23 40

44) 39) 41) 31) 47)

16.1 8.4 14.8 4.3 14.9

23.0 31.6 28.7 52.1 8.5

491

30 (23, 40)

11.0

31.2

101 155 91 97 65

26 29 27 21 35

33) 40) 36) 27) 45)

7.9 7.7 12.1 2.1 20.0

45.5 34.2 45.1 64.9 24.6

509

27 (20, 37)

9.0

43.0

1000

28 (21, 39)

10.0

37.2

(19, (22, (20, (16, (25,

0 Turkey

Sri Lanka

Iran

Pakistan Vietnam

Native country Figure 1 Serum 25(OH)D levels (nmol/l) according to native country (both sexes included). Boxes range from 25-percentile to 75-percentile, horizontal lines represent median values.

deficiency. In all, 10% of the immigrants had s-25(OH)D levels above 50 nmol/l, defined as sufficient levels (Lips, 2001) (Table 1).

Age s-25(OH)D tended to increase with increasing age, but this was only significant for persons born in Turkey (median difference 11 nmol/l between those 20–30 y and those 51– 60 y, P ¼ 0.041). We did not observe any statistical interaction effect between native country and age on s-25(OH)D (not shown). Season s-25(OH)D was higher in blood samples drawn in June compared to those drawn in April or May (median 34 nmol/l vs 28 nmol/l, P ¼ 0.004). The prevalence of vitamin D deficiency was lower in those drawn in June (24% in June vs 39% in April, P ¼ 0.013). In men, median s-25(OH)D increased from 30 nmol/l in April to 39 nmol/l in June (P ¼ 0.002). In women, the increase was from 26 nmol/l in April to 31 nmol/l in June (not significant).

Cod liver oil and fish oil supplements Reported use of cod liver oil or other fish oil supplements showed a strong inverse association with vitamin D deficiency in all groups except men born in Vietnam (Table 2). Those born in Sri Lanka were the most frequent European Journal of Clinical Nutrition


60

Table 2 Prevalence and OR of vitamin D deficiency (s-25(OH)D o 25 nmol/l) related to categories of BMI, physical activity, education length, cod liver oil supplement use and fatty fish intake

Men Cod liver oil use

Fatty fish intake

BMI

Physical activity level

Years of education

Women Cod liver oil use

Fatty fish intake

BMI

Physical activity level

Years of education

Category

N*

Median s25(OH)D (nmol/l)

Prevalence of s-25(OH)D o25 nmol/l (%)

Yes, daily Occasionally No Rarely/never 1–3 /month 1–3 /week Z 4 /week o25 25–29.99 Z 30 Sedentary Moderate High o10 10–12 Z 13

65 116 253 72 151 158 62 179 258 54 200 157 56 85 166 191

42.0 30.0 28.0 26.0 29.0 32.0 34.0 32.0 29.5 27.5 31.0 29.0 32.5 30.0 29.0 31.0

13.8 29.3 36.8 48.6 32.5 25.3 24.2 27.9 33.3 31.5 33.5 29.3 32.1 30.6 33.7 28.8

1.00 2.54 3.60 1.00 0.57 0.35 0.32 1.00 1.30 1.15 1.00 0.75 0.78 1.00 1.02 0.76

(ref.) (1.10–5.89) (1.65–7.85) (ref.) (0.31–1.05) (0.18–0.66) (0.14–0.71) (ref.) (0.83–2.05) (0.56–2.36) (ref.) (0.47–1.21) (0.40–1.54) (ref.) (0.54–1.93) (0.40–1.44)

1.00 2.65 3.54 1.00 0.58 0.34 0.44 1.00 1.39 1.24 1.00 0.85 1.09 1.00 0.95 0.76

Yes, daily Occasionally No Rarely/never 1–3 /month 1–3 /week Z 4 /week o25 25–29.99 Z30 Sedentary Moderate High o10 10–12 Z 13

66 110 269 77 171 153 63 221 185 103 236 151 35 134 158 168

39.5 30.5 22.0 20.0 26.0 26.0 32.0 31.0 25.0 23.0 25.0 28.0 28.0 23.0 27.0 29.0

15.2 31.8 57.6 61.0 45.0 42.5 28.6 34.8 46.5 54.4 48.7 38.4 42.9 56.7 41.1 37.5

1.00 2.37 6.99 1.00 0.60 0.55 0.32 1.00 1.57 2.27 1.00 0.56 0.78 1.00 0.47 0.38

(ref.) (1.06–5.30) (3.37–14.50) (ref.) (0.34–1.06) (0.31–0.99) (0.15–0.68) (ref.) (1.00–2.46) (1.30–3.97) (ref.) (0.36–0.87) (0.37–1.66) (ref.) (0.28–0.79) (0.23–0.65)

1.00 (ref.) 2.94 (1.21–7.11) 7.82 (3.47–17.61) 1.00(ref.) 0.75 (0.40–1.40) 0.62 (0.32–1.18) 0.30 (0.12–0.73) 1.00 (ref.) 1.31 (0.78–2.20) 2.46 (1.25–4.83) 1.00 (ref.) 0.63 (0.38–1.05) 1.22 (0.51–2.91) 1.00 (ref.) 0.48 (0.27–0.87) 0.38 (0.21–0.68)

OR (95% CI)a

OR (95% CI)b (ref.) (1.08–6.50) (1.53–8.18) (ref.) (0.31–1.11) (0.17–0.68) (0.19–1.02) (ref.) (0.83–2.31) (0.55–2.77) (ref.) (0.50–1.44) (0.52–2.26) (ref.) (0.48–1.91) (0.38–1.52)

a

Adjusted for age and native country. Adjusted for age, native country and attendance month as well as the other variables in the table except physical activity level. * The number in the analyses may vary due to missing data. b

cod liver oil supplement users (23% reported daily use), whereas those born in Pakistan were the least frequent users (7% reported daily use).

Fatty fish intake s-25(OH)D increased significantly by increasing frequency of intake of fatty fish (eg salmon, trout, mackerel, herring) (Table 2). The association between fatty fish intake and s25(OH)D was similar in men and women. There was reduced odds ratio (OR) of vitamin D deficiency with increased frequency of fatty fish consumption in both sexes (Table 2). This was still significant for the highest frequency (43 times/week) after adjustment for possible confounders, including cod liver oil supplement use. The strongest association was present in those born in Sri Lanka and Vietnam (not shown), who consumed fatty fish most frequently. European Journal of Clinical Nutrition

BMI There was a trend of increased OR of vitamin D deficiency with increasing BMI, which was significant in women, also after adjustment for possible confounders (Table 2). This association was strongest in women born in Turkey and Pakistan, who had the highest BMI (mean BMI 29.1 kg/m2 for Turkish women and 27.3 kg/m2 for Pakistani women). For example, obese women born in Turkey had an age-adjusted OR of 5.56 (95% CI 1.38–22.30) of having vitamin D deficiency compared to normal-weight women born in Turkey.

Physical activity The level of physical activity was not associated with vitamin D deficiency, except for women with moderate physical activity level who tended to have decreased OR of vitamin D deficiency compared to sedentary women (Table 2). Moderately active women born in Pakistan had a fully adjusted OR


61 of vitamin D deficiency of 0.15 (95% CI 0.04–0.54) compared to sedentary women born in Pakistan.

Education Increased length of education was associated with increased s-25(OH)D in women but not in men (Table 2). OR for vitamin D deficiency decreased by increasing length of education among women. Significantly reduced OR was present in women born in Turkey and Iran, and Turkish women with 412 y of education had a fully adjusted OR of vitamin D deficiency of 0.18 (95% CI 0.04–0.75) compared to Turkish women with o10 y of education.

Calcium intake We estimated calcium intake from milk, cheese and yogurt, based on reported frequencies of intake as well as standard portion sizes and average calcium content of these foods. It ranged from 0 to 1269 mg/day with a mean of 235 mg/day. It was lowest in men born in Vietnam (mean 117 mg/day) and highest in women born in Turkey (mean 295 mg/day). The prevalence of both being in the lowest quartile of s-25(OH)D and the lowest quartile of estimated calcium intake from milk, yogurt and cheese for men was 1.4% for those born in Turkey, 9.9% for those born in Sri Lanka, 4.2% for those born in Iran, 7.0% for those born in Pakistan and 0% for those born in Vietnam. The corresponding figures for women were 6.7% for those born in Turkey, 6.6% for those born in Sri Lanka, 3.9% for those born in Iran, 13.3% for those born in Pakistan and 13.2% for those born in Vietnam.

Discussion We found a high prevalence of vitamin D deficiency in both men and women born in Turkey, Sri Lanka, Iran, Pakistan and Vietnam living in Oslo. The prevalence was higher among women than men, and it was higher among those born in Pakistan and lower among those born in Vietnam compared to the other ethnic groups. These findings are in accordance with previous Norwegian studies that have reported widespread vitamin D deficiency in Pakistani women (Brunvand & Haug, 1993; Henriksen et al, 1995; Falch & Steihaug, 2000) and in immigrant children of predominantly Pakistani background (Brunvand & Nordshus, 1996; Brunvand & Brunvatne, 2001). However, vitamin D status in other immigrant groups in Norway than the Pakistanis has not previously been studied. To our knowledge, this is the first study internationally that assesses vitamin D status in both men and women of these ethnic minority groups in a community setting. Studies from Denmark and the UK indicate that vitamin D deficiency may be widespread in immigrants living in Western affluent communities, but these have most commonly been conducted in clinical settings (Serhan et al, 1999; Glerup et al,

2000; Wilkinson et al, 2000; Iqbal et al, 2001; Datta et al, 2002; Pedersen et al, 2003). The participation rate was low, 39.7% in the main cohort (31–60 y), and only 18.7% in the additional cohort (20–30 y). Therefore, we cannot exclude the possibility of selection bias. In spite of this, the community-based design is preferable, and to our knowledge this is the first Norwegian population-based study that evaluates vitamin D status in immigrants. Mean s-25(OH)D in ethnic Norwegians in the main HUBRO study was 74.8723.7 nmol/l (Meyer et al, 2004) compared to median s-25(OH)D 28 nmol/l in immigrants, implying that the observed serum levels in the immigrant groups are so low that it is unlikely to be merely due to bias. Also, the differences between the ethnic groups were consistent across the age strata (not shown) and there was no interaction effect between age and native country on s-25(OH)D, suggesting that the low participation rate in the youngest age group did not influence the results substantially. Moreover, previous studies have dealt largely with women. We found a high prevalence of vitamin D deficiency also in men, demonstrating that it is not merely a female problem. Women born in Turkey, Iran and Pakistan had higher prevalence of vitamin D deficiency than men from the same countries, whereas there were no sex differences for those born in Sri Lanka and Vietnam. This implies that factors involved in vitamin D deficiency may differ more between men and women in some ethnic groups than others. There were no significant sex differences in cod liver oil use or fatty fish intake frequency within the ethnic groups. However, there was a marked difference in education (which was strongly inversely associated with vitamin D deficiency in women): Men born in Turkey, Iran, and Pakistan had significantly longer education compared to women from the respective countries, whereas men and women born in Sri Lanka and Vietnam had similar length of education (not shown). We expect that outdoor activity and clothing habits largely account for the differences. There is no general advice for immigrants living in Norway about time spent outdoors or exposure of skin. Veiling or staying indoors may be more common in women from Turkey, Iran, and Pakistan, although such exposure was not accounted for in the questionnaire. However, a currently ongoing study will evaluate time spent outdoors and clothing habits of Pakistani and Turkish women living in Oslo. A study from the United Arab Emirates showed markedly lower s-25(OH)D in Arab women than in European women living at the same latitude, which was mainly determined by clothing. Both education level and total UV score (based on clothing and duration of sun exposure) was lower in the Arab women than in the European women (Dawodu et al, 1998). The question about physical activity was used as a means of studying a possible effect on vitamin D status of moving outdoors, although self-reported physical activity may not be expected to be a good indicator of sun exposure. There was no clear association between physical activity level and European Journal of Clinical Nutrition


62 vitamin D status, although there was reduced OR of vitamin D deficiency in moderately active women compared to sedentary women, which was of borderline significance when taking possible confounders into account. It is well established that s-25(OH)D varies throughout the year in response to the seasonal changes in sunlight exposure (Holick, 1995). All blood samples were collected during the period April through June. The effect of sunlight seemed to be larger for men than for women, as there was a significant increase in s-25(OH)D in blood samples obtained from men, whereas the increase was smaller and not significant for women. This suggests that these women benefit less from the increased sunlight through the spring months than men do, possibly by exposing their skin less to the sun. Use of fatty fish and cod liver oil or other fish oil supplements was independently associated with reduced OR for vitamin D deficiency in all groups. Vitamin D supplement use is well known to influence 25(OH)D levels in blood, particularly when sun exposure is limited (Davies et al, 1999; Brustad et al, 2003). Increasing the intake of vitamin D supplements or vitamin D containing foods may thus be a possible means of reducing the prevalence of vitamin D deficiency. However, a very low proportion used cod liver oil supplements regularly, particularly among the Pakistani women who were the group at highest risk (only 5.5% of Pakistani women used cod liver oil daily). Previous attempts to reduce the prevalence of vitamin D deficiency in pregnant Pakistani women by distributing cod liver oil supplements have been unsuccessful (Brunvand et al, 1996). As cod liver oil supplementation is a typical Norwegian tradition that may be unfamiliar to immigrants with Asian background, encouragement to increase fatty fish intake may be an effective alternative approach. We observed a significant inverse relationship between vitamin D status and BMI in women but not men, and it was strongest in women born in Turkey and Pakistan. A considerably higher proportion of women (20%) than men (11%) were obese. Others have also reported decreased 25(OH)D levels with increasing BMI (Jacques et al, 1997; Semba et al, 2000). This may be due to less sun exposure, or decreased bioavailability of vitamin D from subcutaneous adipose tissue (Wortsman et al, 2000), or through alterations in the vitamin D endocrine system (Bell et al, 1985). Length of education was positively related to vitamin D status in women born in Turkey and Iran, also when controlling for fatty fish and cod liver oil intake. Others have also found 25(OH)D levels to be positively correlated with education level in women (Isaia et al, 2003). Dawodu et al, 1998 did not find education level to be independently associated with 25(OH)D levels in Arab women when sun exposure was controlled for. It may thus be suspected that education level influences sun exposure (such as clothing habits), although socioeconomic background is also known to influence dietary habits. Education level may be a marker of the acquirement of a Western lifestyle. European Journal of Clinical Nutrition

Calcium intake A very low calcium intake would exacerbate the effects of vitamin D deficiency and may be an independent risk factor of rickets and ostemalacia (DeLucia et al, 2003). The combination of low s-25(OH)D and a low calcium intake may be particularly harmful and could be useful in identifying those at high risk. Estimated calcium intake from milk, cheese and yogurt was lowest in those born in Vietnam and highest in those born in Turkey. Possible adverse effects on the calcium balance should be taken into consideration for those who have a very low calcium intake combined with low s-25(OH)D, although levels are not below 25 nmol/l. However, the results should be interpreted with caution as the report of calcium-containing foods is not sufficiently validated and may differ between subjects, and as we have not taken into account other dietary calcium sources than dairy products.

Conclusion Vitamin D deficiency is widespread in both men and women born in Turkey, Sri Lanka, Iran, Pakistan and Vietnam living in Oslo. The prevalence is higher among women than among men, and it is highest in women born in Pakistan. Fatty fish intake and cod liver oil supplement use are strong positive predictors of vitamin D status in all ethnic groups, suggesting that encouragement of an increased intake of fatty fish or fish oil supplements could be a means of improving vitamin D status in these ethnic groups. For women, BMI and length of education are also predictors of vitamin D status.

Acknowledgements The data collection was conducted as part of the Oslo Immigrant Health Study carried out in 2002 as a collaboration between the Norwegian Institute of Public Health and the University of Oslo.

References Bell NH, Epstein S, Greene A, Shary J, Oexmann MJ & Shaw S (1985): Evidence for alteration of the vitamin D–endocrine system in obese subjects. J. Clin. Invest. 76, 370–373. Brunvand L & Brunvatne R (2001): Helseproblemer blant innvandrerbarn i Norge. Tidsskr. Nor. Lægeforen. 121, 715–718. Brunvand L & Haug E (1993): Vitamin D deficiency amongst Pakistani women in Oslo. Acta Obstet. Gynecol. Scand. 72, 264–268. Brunvand L, Henriksen C & Haug E (1996): Vitamin D mangel blant gravide fra Pakistan. Hvordan best forebygge? Tidsskr. Nor. Lægeforen. 116, 1585–1587. Brunvand L & Nordshus T (1996): Alimentær rakitt. En gammel sykdom med ny aktualitet. Nord. Med. 111, 219–221. Brustad M, Sandanger T, Wilsgaard T, Aksnes L & Lund E (2003): Change in plasma levels of vitamin D after consumption of codliver and fresh cod-liver oil as part of the traditional north Norwegian fish dish ‘Mølje’. Int. J. Circumpolar Health 62, 40–53. Datta S, Alfaham M, Davies DP, Dunstan F, Woodhead S, Evans J & Richards B (2002): Vitamin D deficiency in pregnant women from a non-European ethnic minority population—an interventional study. Br. J. Obstet. Gynaecol. 109, 905–908.


63 Davies PS, Bates CJ, Cole TJ, Prentice A & Clarke PC (1999): Vitamin D: seasonal and regional differences in preschool children in Great Britain. Eur. J. Clin. Nutr. 53, 195–198. Dawodu A, Absood G, Patel M, Agarwal M, Ezimokhai M, Abdulrazzaq Y & Khalayli G (1998): Biosocial factors affecting vitamin D status of women of childbearing age in the United Arab Emirates. J. Biosoc. Sci. 30, 431–437. DeLucia MC, Mitnick ME & Carpenter TO (2003): Nutritional rickets with normal circulating 25-hydroxyvitamin D: a call for reexamining the role of dietary calcium intake in North American infants. J. Clin. Endocrinol. Metab. 88, 3539–3545. Falch JA & Steihaug S (2000): Vitamin D deficiency in Pakistani premenopausal women living in Norway is not associated with evidence of reduced skeletal strength. Scand. J. Clin. Lab. Invest. 60, 103–109. Glerup H, Mikkelsen K, Poulsen L, Hass E, Overbeck S, Thomsen J, Charles P & Eriksen EF (2000): Commonly recommended daily intake of vitamin D is not sufficient if sunlight exposure is limited. J. Intern. Med. 247, 260–268. Hamson C, Goh L, Sheldon P & Samanta A (2003): Comparative study of bone mineral density, calcium, and vitamin D status in the Gujarati and white populations of Leicester. Postgrad. Med. J. 79, 279–283. Henriksen C, Brunvand L, Stoltenberg C, Trygg K, Haug E & Pedersen JI (1995): Diet and vitamin D status among pregnant Pakistani women in Oslo. Eur. J. Clin. Nutr. 49, 211–218. Holick MF (1995): Environmental factors that influence the cutaneous production of vitamin D. Am. J. Clin. Nutr. 61, 638S–645S. Holick MF (2003): Vitamin D: a millenium perspective. J. Cell. Biochem. 88, 296–307. Iqbal SJ, Featherstone S, Kaddam IM, Mortimer J & Manning D (2001): Family screening is effective in picking up undiagnosed Asian vitamin D deficient subjects. J. Hum. Nutr. Diet. 14, 371–376. Isaia G, Giorgino R, Rini GB, Bevilacqua M, Maugeri D & Adami S (2003): Prevalence of hypovitaminosis D in elderly women in Italy: clinical consequences and risk factors. Osteoporos. Int. 14, 577–582.

Jacques PF, Felson DT, Tucker KL, Mahnken B, Wilson PW, Rosenberg IH & Rush D (1997): Plasma 25-hydroxyvitamin D and its determinants in an elderly population sample. Am. J. Clin. Nutr. 66, 929–936. Lips P (2001): Vitamin D deficiency and secondary hyperparathyroidism in the elderly: consequences for bone loss and fractures and therapeutic implications. Endocr. Rev. 22, 477–501. Meyer HE, Falch JA, Søgaard AJ & Haug E (2004): Vitamin D deficiency & secondary hyperparathyroidism and the association with bone mineral density in persons with Pakistani and Norwegian background living in Oslo, Norway. The Oslo Health Study. Bone 35, 412–417. Pedersen P, Michaelsen KF & Molgaard C (2003): Children with nutritional rickets referred to hospitals in Copenhagen during a 10-year period. Acta Paediatr. 92, 87–90. Semba RD, Garrett E, Johnson BA, Guralnik JM & Fried LP (2000): Vitamin D deficiency among older women with and without disability. Am. J. Clin. Nutr. 72, 1529–1534. Serhan E, Newton P, Ali HA, Walford S & Singh BM (1999): Prevalence of hypovitaminosis D in Indo-Asian patients attending a rheumatology clinic. Bone 25, 609–611. Shaw NJ & Pal BR (2002): Vitamin D deficiency in UK Asian families: activating a new concern. Arch. Dis. Child. 86, 147–149. Statistics Norway (2003): Innvandring og innvandrere 2003. Statistics Norway website. www.ssb.no. Søgaard AJ, Selmer R, Bjertness E & Thelle D (2004): The Oslo Health Study: The impact of self-selection in a large population-based survey. Int. J. Equity Health 3, 3. Wilkinson RJ, Llewelyn M, Toossi Z, Patel P, Pasvol G, Lalvani A, Wright D, Latif M & Davidson RN (2000): Influence of vitamin D deficiency and vitamin D receptor polymorphisms on tuberculosis among Gujarati Asians in west London: a case–control study. Lancet 355, 618–621. Wortsman J, Matsuoka LY, Chen TC, Lu Z & Holick MF (2000): Decreased bioavailability of vitamin D in obesity. Am. J. Clin. Nutr. 72, 690–693.

European Journal of Clinical Nutrition