Hormone Concentrations in Older Men Reporting Good Health
Hormone Concentrations in Older Men Reporting Good Health
The study had an observational design recruiting men over the age of 40 years to undergo multiple blood sampling at fixed intervals over 3 months. The study inclusion criterion was based on the question 'In general, would you say your health was excellent, very good, good, fair or poor?' (question 1 from SF-36 questionnaire) at screening, usually by phone. Only men reporting very good or excellent health were eligible for entry. The only exclusion criterion was the use of medications (e.g. 5α reductase inhibitors) interfering with blood sex steroid levels. Participants were offered no reimbursement for participation. Recruitment aimed to fill each age decade with 100 participants.
The participants provided written informed consent, and the study was approved by the Sydney South West Area Health Service Ethics Committee (Concord Hospital) and the Southern Health Human Research Ethics Committee within NHMRC guidelines on ethical conduct in human research.
Men were recruited by local advertising on noticeboards in and around the hospital and in local newspapers or radio throughout the recruitment period. The advertisements sought men who considered themselves in excellent health ('to help us determine why you are so healthy') to make phone contact with the centre to determine their eligibility for the study. Eligible men attending a study centre for standardized medical history and physical examination and to provide nine morning (8–11 am) blood samples during five visits over 3 months. Three blood samples were obtained at 20 min intervals on the first (day 1) and second (day 2) visit and then single blood samples on the third (day 7), fourth (day 30) and fifth (day 90) visits. Men were fasting overnight on the first but non-fasting on subsequent visits. Serum was stored frozen at −20 °C until analysis in batch at the end of the study. Height (nearest 0·1 cm) and weight (nearest 0·1 kg) were measured with participants wearing a light gown and no shoes. Body composition was measured by multi-electrode bioelectrical impedance (IMP5 meter; ImpediMed, Brisbane, Qld, Australia) and by whole body densitometry. Smoking status was classified by the participants into never, ex-smoker and current smokers. For those who had ever smoked, their pack-years of smoking were calculated and for ex-smokers the number of years since cessation was recorded. Obesity was defined as a body mass index (kg/m) over 30 and hypertension as a blood pressure (single measure at rest) with systolic > 140 and diastolic > 90 mm mercury.
Serum T, DHT, E2 and estrone (E1) were quantified within a single run without derivatization as described and calibrated directly against the National Measurement Institute certified reference standard utilized by the Centers for Disease Control T standardization. The assay limits of detection, limits of quantification and within-run and between-run coefficients of variation (%) were T (35 pM, 90 pM, 2·0%, 3·9–6·5%), DHT (10 pM, 0·69 nM, 8·1%, 6·7–13·4%), E2(4 pM, 18 pM, 6·6%, 4·8–8·6%) and estrone (2 pM, 9 pM, 4·7%, 4·6–7·5%). Serum LH, FSH and sex hormone binding globulin (SHBG) levels in baseline blood samples were measured in batch by automated immunoassays (Roche Diagnostics, Australia) with CVs of 1·0–2·0%. Biochemical and haematological profiles used routine auto-analyser methods. Non-protein bound ('free') T (FT) was calculated using the empirical FTZ formula which more accurately estimates FT measured by either centrifugal ultrafiltration or equilibrium dialysis than equations that assume the valid applicability of equilibrium binding theory. As FTZ is a calculated index, rather than a measured hormonal variable, it has no measure of reproducibility. Biochemical analytes (haemoglobin, lipids, renal and liver function tests, iron studies) were measured by standard auto-analyser methodologies.
The primary data analysis was by mixed model analysis of variance taking age and BMI as within-individual, fasting and smoking (never, ex, current) status as between-individual factors and participants as a random effect usingNCSS software (Kaysville, UT, URL wwww.ncss.com). The models were fitted by restricted maximum likelihood with an autoregressive covariance structure allowing for differences between time points. Serial data were also analysed by repeated measures analysis of variance and factorial analysis of variance with appropriate post hoc testing as required in which the serum steroids were the repeated time-dependent variables, whereas all covariates were measured only at baseline.
Methods
Study Design
The study had an observational design recruiting men over the age of 40 years to undergo multiple blood sampling at fixed intervals over 3 months. The study inclusion criterion was based on the question 'In general, would you say your health was excellent, very good, good, fair or poor?' (question 1 from SF-36 questionnaire) at screening, usually by phone. Only men reporting very good or excellent health were eligible for entry. The only exclusion criterion was the use of medications (e.g. 5α reductase inhibitors) interfering with blood sex steroid levels. Participants were offered no reimbursement for participation. Recruitment aimed to fill each age decade with 100 participants.
The participants provided written informed consent, and the study was approved by the Sydney South West Area Health Service Ethics Committee (Concord Hospital) and the Southern Health Human Research Ethics Committee within NHMRC guidelines on ethical conduct in human research.
Study Procedures
Men were recruited by local advertising on noticeboards in and around the hospital and in local newspapers or radio throughout the recruitment period. The advertisements sought men who considered themselves in excellent health ('to help us determine why you are so healthy') to make phone contact with the centre to determine their eligibility for the study. Eligible men attending a study centre for standardized medical history and physical examination and to provide nine morning (8–11 am) blood samples during five visits over 3 months. Three blood samples were obtained at 20 min intervals on the first (day 1) and second (day 2) visit and then single blood samples on the third (day 7), fourth (day 30) and fifth (day 90) visits. Men were fasting overnight on the first but non-fasting on subsequent visits. Serum was stored frozen at −20 °C until analysis in batch at the end of the study. Height (nearest 0·1 cm) and weight (nearest 0·1 kg) were measured with participants wearing a light gown and no shoes. Body composition was measured by multi-electrode bioelectrical impedance (IMP5 meter; ImpediMed, Brisbane, Qld, Australia) and by whole body densitometry. Smoking status was classified by the participants into never, ex-smoker and current smokers. For those who had ever smoked, their pack-years of smoking were calculated and for ex-smokers the number of years since cessation was recorded. Obesity was defined as a body mass index (kg/m) over 30 and hypertension as a blood pressure (single measure at rest) with systolic > 140 and diastolic > 90 mm mercury.
Assays
Serum T, DHT, E2 and estrone (E1) were quantified within a single run without derivatization as described and calibrated directly against the National Measurement Institute certified reference standard utilized by the Centers for Disease Control T standardization. The assay limits of detection, limits of quantification and within-run and between-run coefficients of variation (%) were T (35 pM, 90 pM, 2·0%, 3·9–6·5%), DHT (10 pM, 0·69 nM, 8·1%, 6·7–13·4%), E2(4 pM, 18 pM, 6·6%, 4·8–8·6%) and estrone (2 pM, 9 pM, 4·7%, 4·6–7·5%). Serum LH, FSH and sex hormone binding globulin (SHBG) levels in baseline blood samples were measured in batch by automated immunoassays (Roche Diagnostics, Australia) with CVs of 1·0–2·0%. Biochemical and haematological profiles used routine auto-analyser methods. Non-protein bound ('free') T (FT) was calculated using the empirical FTZ formula which more accurately estimates FT measured by either centrifugal ultrafiltration or equilibrium dialysis than equations that assume the valid applicability of equilibrium binding theory. As FTZ is a calculated index, rather than a measured hormonal variable, it has no measure of reproducibility. Biochemical analytes (haemoglobin, lipids, renal and liver function tests, iron studies) were measured by standard auto-analyser methodologies.
Data Analysis
The primary data analysis was by mixed model analysis of variance taking age and BMI as within-individual, fasting and smoking (never, ex, current) status as between-individual factors and participants as a random effect usingNCSS software (Kaysville, UT, URL wwww.ncss.com). The models were fitted by restricted maximum likelihood with an autoregressive covariance structure allowing for differences between time points. Serial data were also analysed by repeated measures analysis of variance and factorial analysis of variance with appropriate post hoc testing as required in which the serum steroids were the repeated time-dependent variables, whereas all covariates were measured only at baseline.
Source...