The data were collected as part of the Oslo Health Study (HUBRO), a cross-sectional population-based multipurpose study conducted in 2000–2001, inviting all individuals in the city of Oslo aged 30, 40, 45, 59, 60, 75 and 76 years. The overall attendance rate was 46% (n = 18,770), varying from 36% in 30 years old to 55% in 59–60 years old individuals. The material and methods, including respondent rates, are described in detail elsewhere [13]. As a substudy of HUBRO, bone mineral density and vitamin D status were measured in the period May 2000 to January 2001. The participants recruited for the substudy were random samples of individuals born in Norway (quoted as Norwegians) aged 45, 60, and 75 years, as well as random samples of individuals born in Pakistan (quoted as Pakistanis) of all age groups in the Oslo Health Study. The participants were invited in a random order throughout the study period regardless of ethnic background. For the current data analysis, we excluded those not having their serum vitamin D metabolites measured, and the oldest age group (75–76 years), as there were only three Pakistanis in this age group.
As previously described [12], of the 1281 Norwegians of age 45–60 who were invited to the substudy, 674 (53%) participated, and 584 (46%) had s-25(OH)D analyzed. Of the 608 Pakistanis of age 30–60 who were invited to the substudy, 237 (39%) participated, and 176 (29%) had s-25(OH)D analyzed. An important reason for these discrepancies was that the data collection for the substudy was discontinued January 15th, 2001, and was thus not performed on those who should have met before, but met after January 15th. However, background characteristics (smoking, BMI, and age) of those who participated in the substudy did not differ from those who met after this date.
In addition, we excluded two Norwegian women and one Pakistani woman due to primary hyperparathyroidism as defined by s-iPTH ≥ 8.5 pmol/L and serum levels of ionized calcium (s-Ca2+) > 1.35 mmol/L, and we excluded one subject due to very low s-iPTH combined with high s-Ca2+ (iPTH < 2.0 pmol/L and Ca2+ > 1.35 mmol/L), which possibly indicate malignancy. In addition, two subjects had unknown PTH status due to missing values, and 33 subjects had unknown s-1,25(OH)2D due to missing values. Thus, the final sample for analysis consisted of a total of 94 Pakistani men and 67 Pakistani women of age 30, 40, 45, and 59–60 years, and 290 Norwegian men and 270 Norwegian women of age 45 and 60 years; in total 721 subjects.
Data collection
All participants underwent a simple physical examination and filled in self-administered questionnaires which included information about various lifestyle factors [14]. A non-fasting blood sample was collected from each participant on the day of attendance. Height and weight were measured in light in-door clothing without shoes.
Blood sample analysis
The serum samples were first stored at -20°C for up to eight weeks at the screening station, and then kept frozen at -70°C until analyzed in the Hormone Laboratory, Aker University Hospital.
S-25(OH)D and s-iPTH (intact PTH) were measured as previously described [3]. S-1,25(OH)2D was measured by competitive radioimmunoassay (DiaSorin, Stillwater, MN, USA). Prior to the 1,25(OH)2D determination, serum lipids and interfering vitamin D metabolites were removed by chromatography on a C18OH column. Cross reaction with 25(OH)D after chromatography is noted to be 0.002%. The intra- and interassay coefficients of variation (CVs) for the s-1,25(OH)2D assay were 7 and 14%, respectively. The limit of detection was 12 pmol/l.
S-Ca2+ was measured using an ion-specific electrode (Ciba Corning Diagnostics, Essex, UK). The interassay CV was 2%. S-Ca2+ levels were adjusted to pH 7.40.
The analyses are not adjusted for time since last meal as additional analyses showed that such adjustment did not influence the levels of any of the metabolites.
Statistical analysis
Statistical tests were performed using the software SPSS for Windows version 14.0. Ethnic differences in s-1,25(OH)2D and s-Ca2+ were assessed by one-way analysis of variance (ANOVA) and linear regression. Associations between metabolites were assessed by linear regression and correlation analysis. We corrected for body mass index (BMI) in order to explore whether a difference in s-1,25(OH)2D could be attributable to differences in BMI. Additional data analyses were performed stratified on ethnic background, gender and age.
Ethics
The study protocol was reviewed by the Regional Committee for Medical Research Ethics and approved by the Norwegian Data Inspectorate. Written informed consent was obtained from the participants.