Adiposity Measures and Vitamin D Concentrations
Adiposity Measures and Vitamin D Concentrations
Data was obtained from the Study of Health in Pomerania (SHIP-1 and SHIP-Trend, conducted in the northeast of Germany) and Health2006 (conducted in Denmark). All investigations were carried out in accordance with the Declaration of Helsinki, including written informed consent of all participants. The survey and study methods of both studies were approved by institutional review boards [SHIP-1 (III UV 73/01) and SHIP-Trend (BB 39/09): ethics committee of the University of Greifswald; Health2006: Ethical Committee of Copenhagen County (KA-20060011) and the Danish Data Protection Agency]. In addition, Health2006 was registered at www.clinicaltrials.gov (Unique ID: KA20060011).
SHIP-1 is the first five-year follow-up of the population-based SHIP-0 cohort. In the baseline examination between 1997 and 2001, a total of 4308 men and women between 20 and 79 years of age participated (response 68.8 %). Out of the 4308 participants, 3300 were re-examined in SHIP-1 between 2002 and 2006. SHIP-Trend is a second population-based cohort in the same study region. In the SHIP-Trend baseline examination between 2008 and 2012, a total of 4420 men and women between 20 and 79 years of age participated (response 50.1 %). Details on sampling methods and study protocols have been reported previously. 25OHD concentrations were measured in all SHIP-1 and a subsample of the SHIP-Trend participants. The SHIP-Trend subsample consisted of the first 1000 study participants without diabetes mellitus. It is thus not truly representative for the whole study population as the subjects included are on average healthier than the whole study population.
Interview and Physical Examination. All participants underwent standardized medical examinations, blood sampling, and an extensive computer-aided personal interview. Data on socio-demographic characteristics and medical histories were collected. Intake of medication was recorded and classified using the anatomical therapeutic chemical classification system (ATC). Participants were defined as physically active, if they reported any physical activity (SHIP-1) or more than one hour of physical activity during summer and winter (SHIP-Trend). During the physical examination, standardized measurements of body height and weight were performed with calibrated scales. Waist circumference and hip circumference were measured using an inelastic tape with the subject standing comfortably with weight distributed evenly on both feet. Waist circumference was measured midway between the lower rib margin and the iliac crest in the horizontal plane. Hip circumference was determined as the greatest circumference between the highest point of the iliac crest and the crotch. BMI was calculated as weight (kg)/height (m). Waist-to-hip ratio and waist-to-height ratio were calculated from the respective measures. Body surface area was calculated according to the formula of DuBois. Additionally, all SHIP-Trend participants were offered a body impedance analysis (BIA). BIA analyses were performed using a Nutriguard M device and NutriPlus software (Data Input GmbH, Darmstadt, Germany). Charges of 5 kHz, 50 kHz, and 100 kHz were applied to measure resistance, reactance and phase angle. Body fat was automatically calculated from the latter measurements within the NutriPlus software. Abdominal VAT and SAT were measured in SHIP-Trend by magnetic resonance imaging (MRI). The MRI was performed on a 1.5-Tesla system (Magnetom Avanto, Siemens Healthcare AG, Erlangen, Germany, software version syngo MR B15), using a body phased-array coil. For the determination of abdominal fat, axial 3D datasets using the 2-point Dixon technique were acquired (matrix: 256 × 176; slice thickness 4 mm/4 mm/3 mm without gap; 3 × 64 slices; in-phase: TE 4.76 ms, TR 7.48 ms; opp-phase: TE 2.38 ms, TR 7.48 ms). The quantification of abdominal VAT and SAT was done using ATLAS (automatic tissue and labeling analysis software), an in-house developed software at the University of Ulm. The software first performs a fully-automated step followed by a manual correction of the results. The manual correction, performed by certified medical students, included setting the upper (left diaphragm) and lower border (bladder) for the abdominal fat analysis, correcting misclassified fat labels and removing fat labels that do not belong to the abdomen (i.e. arms, breast fat and parenchyma, bone marrow of pelvis and spine).
Laboratory Measurement. Blood samples were taken from the cubital vein of participants in the supine position. In SHIP-1, participants were mostly non-fasting and blood sampling was performed throughout the day. In SHIP-Trend, participants were fasting and blood sampling was performed between 7.30 a.m. and 1.00 p.m.. Serum 25OHD and serum PTH concentrations were measured on the IDS-iSYS Multi-Discipline Automated Analyser (Immunodiagnostic Systems Limited, Frankfurt am Main, Germany). Serum 25OHD concentrations were measured with the IDS-iSYS 25-Hydroxy Vitamin D assay. The limits of detection and quantitiation were 3.6 and 5.5 ng/ml, respectively. For each analyte three concentrations of control material were measured. In SHIP-1, the coefficients of variation were 16.8 % at low, 13.9 % at medium, and 12.0 % at high concentrations of control material. In SHIP-Trend, the coefficients of variation were 11.6 % at low, 9.1 % at medium, and 10.6 % at high concentrations of control material. Serum creatinine levels were determined with a modified kinetic Jaffé method (SHIP-1: Siemens Dimension RxL; Siemens Healthcare Diagnostics, Eschborn, Germany; SHIP-Trend: Dade Behring Inc, Newark, USA). The eGFR was calculated according to the Cockcroft-Gault formula.
The Health2006 study is a population-based cross-sectional study conducted at the Research Centre for Prevention and Health (RCPH). The methodology of the study design has been described previously. Study participants were recruited between June 2006 and May 2008 through the Danish Civil Registration office as a random sample of men and women aged 18–69 years living in the western area of the Capital Region of Denmark. Of 7770 persons invited 3471 participated corresponding to a participation proportion of 44.7 %.
Interview and Physical Examination. Health2006 participants underwent standardized medical examinations, blood sampling, and answered an extensive questionnaire on lifestyle factors and general health. Participants were defined as physically active, if they reported regular sport and exercise or heavy gardening at least three times a week or athletic training. Height and weight were measured wearing light clothes and no shoes. Waist circumference was measured directly on the body surface midway between the lower rib margin and the iliac crest. The hip circumference was measured over light clothing at the widest girth of the hip. BMI was calculated as weight (kg)/height (m). Waist-to-hip ratio and waist-to-height ratio were calculated from the respective measures. Body surface area was calculated according to the formula of DuBois. Additionally, all participants underwent measurement of impedance for determination of %bodyfat and intra-peritoneal fatness (assessed by ultrasound). Ultrasound measures were performed with the participants lying on their back. VAT was the distance in cm to two decimals between the posterior edge of the abdominal muscles and the front of the lumbar spine. SAT was the distance in cm to two decimals between the front edge of the abdominal muscles and the skin. All ultrasound measures were made after a quiet expiration as described by Stolk et al. with Aquila Pie Medical (Esaote Europe, Maastricht, The Netherlands). The ultrasound measures have been validated against CT/MRI. Fat percentage was measured using a foot-to-foot Tanita Body Composition Analyzer (TBF-300, TANITA Corporation of America, Inc., IL, USA).
Laboratory Measurement. Fasting venous blood samples from all participants were taken from the cubital vein of participants in the supine position. Serum 25OHD and serum PTH concentrations were measured on the Cobas e411 (Roche Diagnostics GmbH, Mannheim, Germany). Serum 25OHD concentrations were measured with the Vitamin D3 (OH) assay. The limit of detection was 4.0 ng/ml. For each analyte three concentrations of control material were measured. The coefficients of variation were 18.3 % at low, 12.3 % at medium, and 12.0 % at high concentrations of control material. Serum creatinine levels were determined with a modified kinetic Jaffé method (Siemens Healthcare Diagnostics, Eschborn, Germany). The eGFR was calculated according to the Cockcroft-Gault formula.
For the present analyses all subjects with missing information on serum 25OHD concentration (SHIP-1: n = 63; SHIP-Trend: n = 3425; Health2006: n = 62), on adiposity measures (SHIP-1: n = 10; SHIP-Trend: n = 133; Health2006: n = 56) or confounders (SHIP-1: n = 12; SHIP-Trend: n = 7; Health2006: n = 131) were excluded. Moreover, we excluded all SHIP participants, who reported intake of prescribed vitamin D preparations or parathyroid hormone (SHIP-1: n = 25; SHIP-Trend: n = 7; in Health2006 this information was unavailable), all subjects with suspected hyperparathyroidism, defined as PTH >120 pg/ml, or missing PTH concentrations (SHIP-1: n = 29; SHIP-Trend: n = 7; Health2006: n = 4), with a history of liver disease, extreme alcohol consumption >400 g/day, or missing information on liver disease or alcohol consumption (SHIP-1: n = 67; SHIP-Trend: n = 38; Health2006: n = 8), with renal insufficiency, defined as eGFR <30 ml/min or missing creatinine concentrations (SHIP-1: n = 10; SHIP-Trend: n = 0; Health2006: n = 15), and all pregnant women (SHIP-1: n = 12; SHIP-Trend: n = 0; Health2006: n = 0). This resulted in study populations of 3072 SHIP-1, 803 SHIP-Trend, and 3195 Health2006 participants.
As 25OHD concentrations were measured with different laboratory methods in SHIP and Health2006, the analyses were performed separately for each study. The associations between abdominal VAT, SAT or %bodyfat and 25OHD concentrations were assessed in SHIP-Trend and Health2006. The associations between BMI, waist circumference, waist-to-hip ratio, waist-to-height ratio, or body surface area and 25OHD concentrations were assessed in SHIP-1 and Health2006. Continuous data are expressed as median (1 -3 quartile), nominal data as percentage.
Multivariable linear regression analyses were performed to assess the associations of the adiposity measures (exposure) with serum 25OHD concentration (outcome). The adiposity measures entered the models either categorized in sex-specific quintiles or as continuous variables. We report adjusted mean 25OHD concentrations with 95 % confidence intervals according to quintiles of adiposity measures as well as ß-coefficients with standard errors and p-values for a one unit increase in the adiposity measures. As the association between %bodyfat and serum 25OHD concentration appeared to be non-linear in Health2006, we included %bodyfat as linear and as quadratic term in the regression model. To address confounding, we adjusted all models for sex, age (years), alcohol consumption (g/day), physical activity (yes/no), smoking status (non smoker, ex-smoker, occasional smoker, regular smoker), and month of blood sampling. P-values <0.05 were considered statistically significant. All statistical analyses were performed with SAS 9.1 (SAS Institute Inc., Cary, NC, USA).
Methods
Study Populations
Data was obtained from the Study of Health in Pomerania (SHIP-1 and SHIP-Trend, conducted in the northeast of Germany) and Health2006 (conducted in Denmark). All investigations were carried out in accordance with the Declaration of Helsinki, including written informed consent of all participants. The survey and study methods of both studies were approved by institutional review boards [SHIP-1 (III UV 73/01) and SHIP-Trend (BB 39/09): ethics committee of the University of Greifswald; Health2006: Ethical Committee of Copenhagen County (KA-20060011) and the Danish Data Protection Agency]. In addition, Health2006 was registered at www.clinicaltrials.gov (Unique ID: KA20060011).
SHIP-1 and SHIP-Trend
SHIP-1 is the first five-year follow-up of the population-based SHIP-0 cohort. In the baseline examination between 1997 and 2001, a total of 4308 men and women between 20 and 79 years of age participated (response 68.8 %). Out of the 4308 participants, 3300 were re-examined in SHIP-1 between 2002 and 2006. SHIP-Trend is a second population-based cohort in the same study region. In the SHIP-Trend baseline examination between 2008 and 2012, a total of 4420 men and women between 20 and 79 years of age participated (response 50.1 %). Details on sampling methods and study protocols have been reported previously. 25OHD concentrations were measured in all SHIP-1 and a subsample of the SHIP-Trend participants. The SHIP-Trend subsample consisted of the first 1000 study participants without diabetes mellitus. It is thus not truly representative for the whole study population as the subjects included are on average healthier than the whole study population.
Interview and Physical Examination. All participants underwent standardized medical examinations, blood sampling, and an extensive computer-aided personal interview. Data on socio-demographic characteristics and medical histories were collected. Intake of medication was recorded and classified using the anatomical therapeutic chemical classification system (ATC). Participants were defined as physically active, if they reported any physical activity (SHIP-1) or more than one hour of physical activity during summer and winter (SHIP-Trend). During the physical examination, standardized measurements of body height and weight were performed with calibrated scales. Waist circumference and hip circumference were measured using an inelastic tape with the subject standing comfortably with weight distributed evenly on both feet. Waist circumference was measured midway between the lower rib margin and the iliac crest in the horizontal plane. Hip circumference was determined as the greatest circumference between the highest point of the iliac crest and the crotch. BMI was calculated as weight (kg)/height (m). Waist-to-hip ratio and waist-to-height ratio were calculated from the respective measures. Body surface area was calculated according to the formula of DuBois. Additionally, all SHIP-Trend participants were offered a body impedance analysis (BIA). BIA analyses were performed using a Nutriguard M device and NutriPlus software (Data Input GmbH, Darmstadt, Germany). Charges of 5 kHz, 50 kHz, and 100 kHz were applied to measure resistance, reactance and phase angle. Body fat was automatically calculated from the latter measurements within the NutriPlus software. Abdominal VAT and SAT were measured in SHIP-Trend by magnetic resonance imaging (MRI). The MRI was performed on a 1.5-Tesla system (Magnetom Avanto, Siemens Healthcare AG, Erlangen, Germany, software version syngo MR B15), using a body phased-array coil. For the determination of abdominal fat, axial 3D datasets using the 2-point Dixon technique were acquired (matrix: 256 × 176; slice thickness 4 mm/4 mm/3 mm without gap; 3 × 64 slices; in-phase: TE 4.76 ms, TR 7.48 ms; opp-phase: TE 2.38 ms, TR 7.48 ms). The quantification of abdominal VAT and SAT was done using ATLAS (automatic tissue and labeling analysis software), an in-house developed software at the University of Ulm. The software first performs a fully-automated step followed by a manual correction of the results. The manual correction, performed by certified medical students, included setting the upper (left diaphragm) and lower border (bladder) for the abdominal fat analysis, correcting misclassified fat labels and removing fat labels that do not belong to the abdomen (i.e. arms, breast fat and parenchyma, bone marrow of pelvis and spine).
Laboratory Measurement. Blood samples were taken from the cubital vein of participants in the supine position. In SHIP-1, participants were mostly non-fasting and blood sampling was performed throughout the day. In SHIP-Trend, participants were fasting and blood sampling was performed between 7.30 a.m. and 1.00 p.m.. Serum 25OHD and serum PTH concentrations were measured on the IDS-iSYS Multi-Discipline Automated Analyser (Immunodiagnostic Systems Limited, Frankfurt am Main, Germany). Serum 25OHD concentrations were measured with the IDS-iSYS 25-Hydroxy Vitamin D assay. The limits of detection and quantitiation were 3.6 and 5.5 ng/ml, respectively. For each analyte three concentrations of control material were measured. In SHIP-1, the coefficients of variation were 16.8 % at low, 13.9 % at medium, and 12.0 % at high concentrations of control material. In SHIP-Trend, the coefficients of variation were 11.6 % at low, 9.1 % at medium, and 10.6 % at high concentrations of control material. Serum creatinine levels were determined with a modified kinetic Jaffé method (SHIP-1: Siemens Dimension RxL; Siemens Healthcare Diagnostics, Eschborn, Germany; SHIP-Trend: Dade Behring Inc, Newark, USA). The eGFR was calculated according to the Cockcroft-Gault formula.
Health2006
The Health2006 study is a population-based cross-sectional study conducted at the Research Centre for Prevention and Health (RCPH). The methodology of the study design has been described previously. Study participants were recruited between June 2006 and May 2008 through the Danish Civil Registration office as a random sample of men and women aged 18–69 years living in the western area of the Capital Region of Denmark. Of 7770 persons invited 3471 participated corresponding to a participation proportion of 44.7 %.
Interview and Physical Examination. Health2006 participants underwent standardized medical examinations, blood sampling, and answered an extensive questionnaire on lifestyle factors and general health. Participants were defined as physically active, if they reported regular sport and exercise or heavy gardening at least three times a week or athletic training. Height and weight were measured wearing light clothes and no shoes. Waist circumference was measured directly on the body surface midway between the lower rib margin and the iliac crest. The hip circumference was measured over light clothing at the widest girth of the hip. BMI was calculated as weight (kg)/height (m). Waist-to-hip ratio and waist-to-height ratio were calculated from the respective measures. Body surface area was calculated according to the formula of DuBois. Additionally, all participants underwent measurement of impedance for determination of %bodyfat and intra-peritoneal fatness (assessed by ultrasound). Ultrasound measures were performed with the participants lying on their back. VAT was the distance in cm to two decimals between the posterior edge of the abdominal muscles and the front of the lumbar spine. SAT was the distance in cm to two decimals between the front edge of the abdominal muscles and the skin. All ultrasound measures were made after a quiet expiration as described by Stolk et al. with Aquila Pie Medical (Esaote Europe, Maastricht, The Netherlands). The ultrasound measures have been validated against CT/MRI. Fat percentage was measured using a foot-to-foot Tanita Body Composition Analyzer (TBF-300, TANITA Corporation of America, Inc., IL, USA).
Laboratory Measurement. Fasting venous blood samples from all participants were taken from the cubital vein of participants in the supine position. Serum 25OHD and serum PTH concentrations were measured on the Cobas e411 (Roche Diagnostics GmbH, Mannheim, Germany). Serum 25OHD concentrations were measured with the Vitamin D3 (OH) assay. The limit of detection was 4.0 ng/ml. For each analyte three concentrations of control material were measured. The coefficients of variation were 18.3 % at low, 12.3 % at medium, and 12.0 % at high concentrations of control material. Serum creatinine levels were determined with a modified kinetic Jaffé method (Siemens Healthcare Diagnostics, Eschborn, Germany). The eGFR was calculated according to the Cockcroft-Gault formula.
Exclusions
For the present analyses all subjects with missing information on serum 25OHD concentration (SHIP-1: n = 63; SHIP-Trend: n = 3425; Health2006: n = 62), on adiposity measures (SHIP-1: n = 10; SHIP-Trend: n = 133; Health2006: n = 56) or confounders (SHIP-1: n = 12; SHIP-Trend: n = 7; Health2006: n = 131) were excluded. Moreover, we excluded all SHIP participants, who reported intake of prescribed vitamin D preparations or parathyroid hormone (SHIP-1: n = 25; SHIP-Trend: n = 7; in Health2006 this information was unavailable), all subjects with suspected hyperparathyroidism, defined as PTH >120 pg/ml, or missing PTH concentrations (SHIP-1: n = 29; SHIP-Trend: n = 7; Health2006: n = 4), with a history of liver disease, extreme alcohol consumption >400 g/day, or missing information on liver disease or alcohol consumption (SHIP-1: n = 67; SHIP-Trend: n = 38; Health2006: n = 8), with renal insufficiency, defined as eGFR <30 ml/min or missing creatinine concentrations (SHIP-1: n = 10; SHIP-Trend: n = 0; Health2006: n = 15), and all pregnant women (SHIP-1: n = 12; SHIP-Trend: n = 0; Health2006: n = 0). This resulted in study populations of 3072 SHIP-1, 803 SHIP-Trend, and 3195 Health2006 participants.
Statistical Analyses
As 25OHD concentrations were measured with different laboratory methods in SHIP and Health2006, the analyses were performed separately for each study. The associations between abdominal VAT, SAT or %bodyfat and 25OHD concentrations were assessed in SHIP-Trend and Health2006. The associations between BMI, waist circumference, waist-to-hip ratio, waist-to-height ratio, or body surface area and 25OHD concentrations were assessed in SHIP-1 and Health2006. Continuous data are expressed as median (1 -3 quartile), nominal data as percentage.
Multivariable linear regression analyses were performed to assess the associations of the adiposity measures (exposure) with serum 25OHD concentration (outcome). The adiposity measures entered the models either categorized in sex-specific quintiles or as continuous variables. We report adjusted mean 25OHD concentrations with 95 % confidence intervals according to quintiles of adiposity measures as well as ß-coefficients with standard errors and p-values for a one unit increase in the adiposity measures. As the association between %bodyfat and serum 25OHD concentration appeared to be non-linear in Health2006, we included %bodyfat as linear and as quadratic term in the regression model. To address confounding, we adjusted all models for sex, age (years), alcohol consumption (g/day), physical activity (yes/no), smoking status (non smoker, ex-smoker, occasional smoker, regular smoker), and month of blood sampling. P-values <0.05 were considered statistically significant. All statistical analyses were performed with SAS 9.1 (SAS Institute Inc., Cary, NC, USA).
Source...