help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Journal of Clinical Endocrinology & Metabolism , doi:10.1210/jc.2007-1303
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brambilla, D. J.
Right arrow Articles by McKinlay, J. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Brambilla, D. J.
Right arrow Articles by McKinlay, J. B.
Related Collections
Right arrow Male Endocrinology
The Journal of Clinical Endocrinology & Metabolism Vol. 92, No. 11 4224-4229
Copyright © 2007 by The Endocrine Society

Lack of Seasonal Variation in Serum Sex Hormone Levels in Middle-Aged to Older Men in the Boston Area

Donald J. Brambilla, Amy B. O’Donnell, Alvin M. Matsumoto and John B. McKinlay

New England Research Institutes (D.J.B., A.B.O., J.B.M.), Watertown, Massachusetts 02472; and Department of Medicine (A.M.M.), University of Washington School of Medicine, and Geriatric Research, Education, and Clinical Center, Veterans Affairs Puget Sound Health Care System, Seattle, Washington 98108

Address all correspondence and requests for reprints to: Donald J. Brambilla, Ph.D., New England Research Institutes, 9 Galen Street, Watertown, Massachusetts 02472. E-mail: dbrambilla{at}neriscience.com.


   Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
Context: Previous studies of seasonal variation of testosterone and other hormones in men have produced mixed results regarding the number and timing of peaks and nadirs and whether hormones vary seasonally at all. Wide variation in study designs, sample sizes, analytical methods, and characteristics of the study populations may account for the heterogeneity of results.

Objective: The objective of the study was to determine whether serum total, free, and bioavailable testosterone, dihydrotestosterone, SHBG, LH, dehydroepiandrosterone, dehydroepiandrosterone sulfate, estrone, estradiol, and cortisol vary seasonally in men.

Design: Two blood samples were drawn 1–3 d apart at study entry and again 3 and 6 months later (maximum six samples per subject). Hormone levels 1–3 d apart were averaged to reduce short-term intrasubject variation.

Setting: The study population consisted of a community-dwelling population (Boston, MA).

Study Participants: One hundred thirty-four men 30–79 yr old were randomly selected from the respondents to the Boston Area Community Health Survey. One hundred twenty-one men who completed all six visits were included in the analysis.

Main Outcome Measures: In a repeated-measures analysis, 3-month change in hormone levels, measured twice per subject, and in a sinusoidal nonlinear regression with random subject effects, average hormone level in samples 1–3 d apart were measured.

Results: Aside from cortisol, no evidence of seasonal variation in hormone levels was found. The amplitude of seasonal variation was much smaller than total intraindividual variation for all hormones considered.

Conclusions: Seasonal variation is likely an unimportant source of variation clinically and in epidemiological studies of hormone levels.


   Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
SEASONAL OR CIRCAANNUAL variation of levels of testosterone and other reproductive and adrenal hormones in men has received considerable attention from researchers but with inconsistent results (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24). The number and timing of peak and nadir hormone levels has varied widely among studies. In some studies, no seasonal variation was found. Consider testosterone, for example. The timing of annual peak testosterone varied so widely among 15 studies that peaks were found in all seasons and in almost every month of the year (3, 4, 5, 8, 9, 12, 13, 14, 15, 16, 17, 18, 19, 20, 23). No seasonal variation of testosterone was found in three other studies (1, 10, 11). In one case, seasonal variation was found in a cross-sectional analysis of baseline data but not in a longitudinal analysis of follow-up data (17).

Much of the heterogeneity of results may stem from the use of cross-sectional designs in some studies (5, 7, 8, 15, 20, 21, 24), longitudinal designs in others (1, 3, 4, 9, 10, 11, 12, 13, 14, 16, 19, 23), and a mix of the two in one (17); variation in analytic methods even among studies with the same design; wide variation in sample sizes; and differences in age and other characteristics among study populations.

Here we present an analysis of seasonal variation of testosterone of total, free, and bioavailable testosterone (T), dihydrotestosterone (DHT), SHBG, LH, dehydroepiandrosterone (DHEA), dehydroepiandrosterone sulfate (DHEAS), estrone, estradiol, and cortisol using data from a longitudinal study that was initiated in Boston, Massachusetts, in May 2004 (25). The sample size in this study exceeded those in most previous longitudinal investigations. Several steps were also taken during sample collection and data analysis to enhance statistical power and to avoid some of the potential problems that may have affected previous investigations.


   Subjects and Methods
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
Study participants and recruitment

Subjects for the study were selected from among 2301 male respondents to the Boston Area Community Health (BACH) Survey (26). Subjects for the BACH survey were randomly selected from Boston residents who were 30–79 yr old, using a weighted sampling scheme to recruit approximately equal numbers of Hispanics, non-Hispanic black Americans, and non-Hispanic Caucasians. Recruitment was also stratified on decade of age to provide approximate balance over the target age range.

For the hormone variation study, respondents to the BACH Survey were randomly selected within each of the 15 strata defined by race/ethnicity and decade of age with the goal of obtaining approximately the same number in every stratum. A potential subject who refused or was found to be ineligible was replaced with another randomly selected subject from the same stratum. We also tried to recruit approximately the same number of men each month over the course of a yr.

Men were excluded if they had hypogonadism with known cause, such as treatment for prostate cancer, Klinefelter syndrome, Kallmann syndrome, and orchidectomy; if they were using any medications that alter hormone levels, either as the intended effect or as a side effect; or if they had cirrhosis, liver cancer, other severe liver disease, or kidney disease requiring dialysis. Excluded medications included anabolic steroids, androstenedione, casodex, cimetidine, DHEA, diethylstilbestrol, other estrogens, dutasteride (Avodart), finasteride (Proscar), glucocorticoids (prednisone, cortisone, hydrocortisone, and decadron), ketoconazole, megestrol acetate, opiates (morphine, Percocet, codeine, oxycodone, OxyContin, hydrocodone, etc.), spironolactone, testosterone or any androgen, flutamide, and other medications for prostate cancer. BACH survey respondents who had problems with blood draws, such as hemophilia, or a compromised immune system caused by HIV/AIDS, chemotherapy, radiation, or other conditions were also excluded.

Subjects were enrolled after written informed consent was obtained. The consent form, protocol, telephone scripts, and contact documents were approved by the Institutional Review Board of the New England Research Institutes.

Sample and data collection

Two blood samples were obtained approximately 1–3 d apart (median 2 d) at study entry and again 3 and 6 months later, producing a maximum of six blood samples per subject (Fig. 1Go). Study visits generally took place in the subject’s home. If the subject so requested, the visits took place in other locations, such as the subject’s place of employment or study headquarters at the New England Research Institutes.


Figure 1
View larger version (24K):
[in this window]
[in a new window]

  		FIG. 1. Design of the study of intraindividual variation in hormones in men.

 
At each visit, two blood samples were drawn 20 min apart to reduce the effects of pulsatile secretion on hormone levels. Blood was drawn within 4 h of the subject’s awakening to control for diurnal variation (13, 14, 16, 27, 28). The two samples were placed in an ice-filled cooler for transport back to study headquarters, where they were centrifuged, equal aliquots pooled, and the pooled samples stored in scintillation vials at –70 C. Assays were performed in the Endocrine Laboratory of the Department of Physiology at the University of Massachusetts Medical School (Worcester, MA).

At the first visit in each pair, a brief questionnaire was administered regarding recent changes that could affect hormone levels, including changes in health and behavior, and lifestyle issues, such as smoking and alcohol consumption. Subjects who had started taking medications or had developed conditions that would have made them ineligible at the start were excluded from further participation.

Hormone assays

Total T, cortisol, and DHEAS were measured by RIAs from Diagnostic Product Corporation (Los Angeles, CA) [total T: intraassay coefficient of variation (CV) 4.3%, interassay CV, 9.8%; cortisol: intraassay CV, 3.4%, interassay CV, 6.4%; DHEAS: intraassay CV, 2.5%, interassay CV, 5.2%]. DHEA, estradiol, and estrone were measured by RIAs from Diagnostic Systems Laboratories (Webster, TX) (DHEA: intraassay CV, 2.6%, interassay CV, 5.6%; estradiol: intraassay CV, 2.3%, interassay CV, 7.4%; estrone: intraassay CV, 1.0%, interassay CV, 4.2%). DHT was measured by an RIA that was developed at the Endocrine Laboratory of the Department of Physiology at the University of Massachusetts (29) (intraassay CV, 2.1%; interassay CV, 3.1%). LH and SHBG were measured by chemiluminescent immunoassays from Diagnostic Product Corp. (LH: intraassay CV, 4.2%, interassay CV, 5.5%; SHBG: intraassay CV, 3.1%, interassay CV, 4.1%). Free and bioavailable T were calculated from the Sodergard equation using a constant albumin concentration of 4.3 g/dl (30, 31).

All samples obtained from a subject were assayed in the same run for each hormone to exclude interassay variation from changes in hormone level within subjects.

Statistical methods

All hormone values were transformed to base 10 logarithms to eliminate positive correlations between intrasubject variation in hormone level and the intrasubject mean (25, 32, 33). Log-transformed values from two visits 1–3 d apart were averaged, reducing the data set to three observations per subject and, more important, improving statistical power to detect changes within subjects by reducing short-term intrasubject variation of hormone levels.

Two approaches to statistical modeling were implemented. First, each set of three observations on a subject was reduced to a pair of 3-month changes in hormone levels. Change over 3 months was modeled as a function of the month in which the 3-month interval started using a repeated measures linear model (34). The reasoning behind this analysis was that 3-month change in hormone level should depend on when the 3-month interval started, if hormones vary seasonally. For the second approach, an annual cycle with one maximum and one minimum occurring 6 months apart was assumed. A nonlinear mixed effects model was used, under which a sinusoidal model of hormone level as a function of day of the year was assumed. The model took the following form:

Formula
where Yij is the average of the log-transformed hormone levels from subject i in visit pair j; Dij is day of the year for the first visit in the pair; B0 is average hormone level across all subjects; B1 is the amplitude of annual variation, measured as the difference between the peak and average log-transformed values; B2 determines when during the year the peak and nadir are reached; b0i is a normally distributed random component that captures variation in average hormone levels among subjects; and eij is a normally distributed error term. Sinusoidal models have been used frequently in previous studies, although rarely with a random effect for differences in mean hormone level among subjects or log-transformed data (1, 2, 6, 7, 8, 10, 13, 22, 23). A test of the null hypothesis that B1 = 0 is a test for seasonal variation. The results were interpreted by calculating 100(10B1 – 1) which is the difference between annual peak and mean hormone levels expressed as a percentage of the mean, after the peak and mean are transformed back to the raw scale. Interpretation of the calculated value is straightforward.

The amplitude of variation as a percentage of the mean was compared with estimates of the total intraindividual SD values of log10-transformed hormone measurements that were derived from this data set previously (25). If s is the intraindividual SD, then 100(10S – 1) is the percentage by which a value in the log scale that is 1 SD above the mean exceeds the mean after that value and the mean have been transformed back to the raw scale. The calculated value provides a reasonably intuitive, although not quantitative, basis for determining the extent to which seasonal variation contributes to total intraindividual variation when both are expressed as percentages of the mean.

Because the observations were obtained over 19 months, we considered two approaches to coding time of the year in the analysis. Under the first approach, the same time point in two successive years received the same code (e.g. d 10 in yr 1 and 2). Under the second approach, time points were numbered consecutively from the start of the study (e.g. d 10 and 375). The two approaches produced very similar results under both the repeated measures and sinusoidal analyses. Therefore, the results of the first approach are presented here.

All model fitting took place using SAS software (SAS Institute, Cary, NC).


   Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
A total of 134 subjects enrolled in the study, 121 of whom were included in the analysis. Eight who completed only the first two visits and three who completed only the first four were excluded. Six of the 11 were dropped from the study after they began taking medications or developed medical conditions that altered hormone levels, four withdrew voluntarily or were lost to follow-up, and one man died. Two other subjects were excluded after a review of their hormone levels and other information indicated that that both were ineligible from the start.

This study cohort was reasonably well balanced on age and race/ethnicity. The cohort included 43 Caucasians, 46 African-Americans, and 45 Hispanics. There were 20 respondents in the first age category (30–39 yr old) and 28 or 29 respondents in each of the other four age categories. As reported previously, the hormone levels in the participants in the study span broad ranges of values (25). The fifth and 95th percentiles of hormone concentration at the first study visit were 6.7 and 26.9 nmol/liter for total T, 153 and 524 pmol/liter for free T, 3.6 and 12.2 nmol/liter for bioavailable T, 202 and 579 nmol/liter for cortisol, 7 and 44 nmol/liter for DHEA, 0.89 and 11.2 µmol/liter for DHEAS, 0.55 and 1.65 nmol/liter for DHT, 63 and 196 pmol/liter for estradiol, 78 and 211 pmol/liter for estrone, 2.05 and 10.6 mIU/ml for LH, and 16.8 and 79.6 nmol/liter for SHBG. The large majority of values were within normal ranges, although a few subjects with extreme values that may have reflected unreported medication use or disease status were included. Even for these subjects, though, values were consistent within subjects.

Median time from awakening to first blood draw was 1.67 h (fifth and 95th percentiles: 0.42 and 3.12 h). Most of the 760 blood draws (92.5%) took place between 0600 and 1200 h, but 37 samples were drawn between 0430 and 0600 h, and the remaining 20 samples were drawn between 1200 and 1625 h. Draw times tended to be fairly tightly clustered within subjects, although they varied considerably among subjects because of work schedules, among other factors.

There was no evidence in the repeated measures analysis that 3-month change in hormone level varied with the month in which the 3-month interval started (P > 0.10 in all cases). The changes were generally very small. Plots of the monthly median and interquartile range for T and cortisol are representative of the low level of variation seen in all hormones in this study (Fig. 2Go). For example, when median 3-month change was determined for each hormone and starting month, 90% of the median changes fell between –0.032 and 0.033 log10; i.e. between a 7.1% loss and a 7.9% gain over 3 months. The median and interquartile range for percent change in hormone level over 3 months for T and cortisol were plotted against the month in which the 3-month interval started in Fig. 3Go. A seasonal rhythm to hormone levels should produce a corresponding seasonal rhythm to plots of 3-month change. For example, if the sinusoidal model is correct, then 3-month change should also vary seasonally with one maximum and one minimum that occur 6 months apart. However, there is little or no discernible seasonal rhythm in either plot. Furthermore, the interquartile ranges imply considerable variation among subjects in the magnitude and direction of 3-month change in any given time interval.


Figure 2
View larger version (32K):
[in this window]
[in a new window]

  		FIG. 2. Median hormone level, with the interquartile range, for each month of the year. Dotted line, the overall median; numbers in the figure are sample sizes.

 

Figure 3
View larger version (31K):
[in this window]
[in a new window]

  		FIG. 3. Median percent change in hormone level over 3 months, with the interquartile range, plotted against the month in which the 3-month interval started. Dotted line, zero 3-month change; numbers in the figure are sample sizes.

 
Aside from cortisol (P = 0.0212), no statistically significant seasonal variation was detected using the sinusoidal model (Table 1Go). Annual peak hormone level exceeded mean hormone level by less than 4% for all hormones. For cortisol, the peak value on the fitted curve occurred in early January but the peak on the fitted sinusoidal model was only 3.7% higher than the mean.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Results of fitting a sinusoidal model of seasonal variation to log-transformed hormone level

 
The measures of total within-subject variation that were obtained from the previously estimated intraindividual SD values were all at least seven times the amplitude of variation from the fitted sinusoidal models for seasonal variation (median 13, range 7.7–68.9; see Table 1Go). It therefore appears that seasonal variation is of low amplitude and makes at most a small contribution to total intraindividual variation.


   Discussion
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
Unlike the sinusoidal model, no specific seasonal pattern of variation was assumed in the repeated-measures analysis of 3-month change in hormone level. The sinusoidal model offered greater statistical power for detecting patterns of variation characterized by one peak and one nadir per year that occur 6 months apart, without requiring specification of the timing of the peak and nadir. The repeated-measures model offered greater statistical power for more complex patterns, such as a pattern involving variation on less than an annual cycle or possibly one in which the annual peak and nadir are something other than 6 months apart.

In this study also, subjects were randomly sampled from a community-dwelling population so that characteristics of study subjects were reasonably evenly distributed over the 1-yr interval of recruitment. The logarithmic transformation stabilized the variance of hormone measurements across a broad range of hormone levels, which improved the fit to the assumption of homogeneous variance that underlies both of the statistical models used in this analysis. Study power was enhanced by pooling equal aliquots of two blood draws obtained 20 min apart and averaging measurements obtained 1–3 d apart to reduce short-term intraindividual variation; drawing all samples within 4 h of each subject’s normal time of awakening to suppress diurnal variation; and assaying all samples from a given subject in the same run to exclude interassay variation from changes of hormone levels within subjects. Because of strong correlations between measurements within subjects, a repeated-measures analysis of 3-month change also provided greater statistical power than would a repeated measures analysis of the sets of three averages per subject.

Even with these enhancements to statistical power, little or no seasonal variation was detected in the hormones evaluated. Parameter estimates from both sets of analyses indicate that any such variation is of very low amplitude. Furthermore, for each hormone examined, the amplitude of seasonal variation from the sinusoidal model indicated that seasonal variation made at most a very small contribution to total within-subject variation calculated from the estimated total intraindividual SDs. This comparison actually provides a conservative measure of the contribution of seasonal variation to the total because total intraindividual variation is actually greater, and the contribution of seasonal variation to it is actually less, than the comparison presented here indicates.

The meaning of the statistically significant outcome for cortisol with the sinusoidal model is not clear. Given the large number of statistical tests in this study, the result may simply be a type I error. Previous studies of seasonal variation in cortisol have produced inconsistent results. Statistically significant seasonal variation was found in one study (10) but not in four others (1, 2, 6, 13). In a sixth study, cortisol varied seasonally in young but not elderly men (22). Sample sizes were generally small (median 14; range 6–106), which limited statistical power. Consistent with the results reported here, the amplitude of variation was less than 10% of the mean in all six studies. Thus, even if cortisol does vary seasonally, it does so with limited amplitude.

What then of the heterogeneous results of previous studies? Both cross-sectional and longitudinal designs have been used in studies of seasonal variation. In cross-sectional studies, intraindividual variation must be inferred from intersubject differences. What appears to be seasonal variation in a cross-sectional study may actually be differences among the subjects sampled at different times rather than changes within subjects. Confounding such as this could account for the wide variation in the number and timing of peaks and nadirs among studies that was noted earlier.

Longitudinal studies are more appropriate for studying seasonal variation because they focus directly on seasonal changes in hormone level within subjects (1, 3, 4, 9, 10, 11, 12, 13, 14, 16, 17, 18, 19, 22, 23). Longitudinal studies, however, have also produced inconsistent results, possibly reflecting differences in study methods. Monthly (1, 3, 9, 10, 11, 12, 23), semimonthly (16), and quarterly (4, 13, 19, 22) sampling intervals have been used, which may have affected the precision of estimates of seasonal change. Sinusoidal models were used in some investigations (1, 10, 13, 16, 22, 23) and linear models in others (3, 4, 9, 10, 11, 12, 14, 18). Models that properly accounted for the correlations among repeated measurements on the same subjects were rarely mentioned (3, 12, 18) and were probably rarely used. Failure to take account of the correlations can affect statistical power.

Assay schedule is another important consideration. Suppose that samples are assayed monthly or semimonthly in a study in which each subject is sampled monthly for a year. Then month-to-month differences in hormone levels could be nothing more than interassay variation. This is a potentially serious problem for both longitudinal and cross-sectional studies, given the low amplitude of seasonal variation that has been identified. Fortunately, many investigators have chosen to assay all samples from each subject in the same run (3, 11, 12, 16, 18, 19, 22). However, the processing schedule was not described in sufficient detail in several longitudinal studies, leaving them vulnerable to this problem (4, 9, 10, 13, 14).

Sample sizes have varied widely among previous studies of seasonal variation and have generally been smaller in longitudinal than in cross-sectional studies. Study populations have also varied widely in age and location. Studies have included prepubertal boys (5, 6), young or young and middle age men (1, 2, 3, 8, 9, 10, 11, 12, 16, 17, 18, 19, 23, 24), elderly men (13, 14, 15, 20, 21), or a mix of age groups (7, 22). All of the studies have taken place at mid- to high latitudes.

It is also not clear whether optimal statistical methods have been used to analyze the data in previous studies. As noted earlier, the logarithmic transformation can suppress the positive correlation between variation in hormone level and mean hormone level (25, 32, 33). Failure to transform to logarithms when variation increases with the mean can reduce statistical power (35). Logarithmic transformations have been used in only a few studies of seasonal variation (3, 4, 12), although some investigators (10, 11, 23) have converted the measurements from each subject to percentages of that subject’s mean, which would also stabilize the variance.

Even if attention is restricted to the longitudinal studies in which samples from each subject were assayed in a single run, results still varied widely. For example, seasonal variation of T level was found in three studies with monthly sampling, with peak T occurring variously in May and June (12), late summer (3), and October (16). In one study with quarterly sampling, mean T was higher in July and October than April and January (19). No seasonal variation was found in a fifth study (11). The issue remains unclear, even after restricting attention to studies with desirable characteristics.

Previous studies of seasonal variation in hormone levels have been consistent in one respect: even when variation was statistically significant, it was of generally low amplitude, compared with total intraindividual variation. For example, estimates of the amplitude of seasonal variation in T were obtained from three studies in which sinusoidal models were fit to cross-sectional data (5, 8, 17), four in which these models were fit to longitudinal data (11, 13, 16, 23), and one longitudinal study in which the results were expressed as the percentage difference between the peak and nadir (3). In seven studies, the amplitude of variation ranged from 2.4 to 11% of the mean (3, 8, 11, 13, 16, 17, 23). Seasonal amplitude was 35% of the mean in the eighth study, but this was a cross-sectional study of 72 subjects with samples collected over a 2- or 3-yr interval (5). The reliability of the latter estimate is unclear.

The results of this study indicate that researchers need not be concerned about seasonal variation in evaluating the results of epidemiological investigations of factors affecting hormone levels. This is especially true of cross-sectional studies in which the variation against which effects are measured consists of both interindividual and intraindividual variation. Interindividual variation is generally much greater than intraindividual variation (25), so seasonal variation would be an even smaller component of total variation in such a study than it is when effects are measured against intraindividual variation alone. Seasonal variation is also not a major concern in interpreting clinical measurements. With the limited contribution of seasonal variation to intraindividual variation, a measurement that is abnormally high or low in one season will very likely be abnormal in another.


   Footnotes
 
This work was supported by Grant AG23027 from the National Institute on Aging of the National Institutes of Health.

Disclosure Statement: A.B.O., D.J.B., and J.B.M. have nothing to declare. A.M.M. consults for GlaxoSmithKline, Johnson & Johnson, Solvay Pharmaceuticals, Amgen, and Quatrx and received grant support from GlaxoSmithKline (June 1, 2004, through May 31, 2008, and July 1, 2007, through June 30, 2010), Ascend Therapeutics (January 1, 2005, through December 31, 2007), Solvay Pharmaceuticals (June 1, 2005, through May 31, 2007) and Ardana Bioscience (July 1, 2007, through June 30, 2010).

First Published Online August 7, 2007

Abbreviations: BACH, Boston Area Community Health (Survey); CV, coefficient of variation; DHEA, dehydroepiandrosterone; DHEAS, DHEA sulfate; DHT, dihydrotestosterone; T, testosterone.

Received June 12, 2007.

Accepted August 1, 2007.


   References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Abbaticchio G, de Fini M, Giagulli VA, Santoro G, Vendola G, Giorgino R 1987 Circannual rhythms in reproductive functions of human males, correlations among hormones and hormone-dependent parameters. Andrologia 19:353–361[Medline]
  2. Agrimonti F, Angeli A, Frairia R, Fazzari A, Tamagnone C, Fornaro D, Ceresa F 1982 Circannual rhythmicities of cortisol levels in the peripheral plasma of healthy subjects. Chronobiologia 9:107–114[Medline]
  3. Andersson AM, Carlsen E, Petersen JH, Skakkebaek NE 2003 Variation in levels of serum inhibin B, testosterone, estradiol, luteinizing hormone, follicle-stimulating hormone, and sex hormone-binding globulin in monthly samples from healthy men during a 17-month period: possible effects of seasons. J Clin Endocrinol Metab 88:932–937[Abstract/Free Full Text]
  4. Baker HW, Burger HG, de Kretser DM, Hudson B, O’Connor S, Wang C, Mirovics A, Court J, Dunlop M, Rennie GC 1976 Changes in the pituitary-testicular system with age. Clin Endocrinol (Oxf) 5:349–372[Medline]
  5. Bellastella A, Esposito V, Mango A, D’Alessandro B 1982 Temporal relationship between circannual levels of luteinizing hormone and testosterone in prepubertal boys with constitutional short stature. Chronobiologia 9:123–125[Medline]
  6. Bellastella A, Criscuolo T, Mango A, Perrone L, Sinisi AA, Faggiano M 1983 Circannual rhythms of plasma luteinizing hormone, follicle-stimulating hormone, testosterone, prolactin and cortisol in prepuberty. Clin Endocrinol (Oxf) 19:453–459[Medline]
  7. Bjornerem A, Bjorn S, Oian P, Bentsen GKR 2006 Seasonal variation of estradiol, follicle stimulating hormone and dehydroepiandrosterone sulfate in women and men. J Clin Endocrinol Metab 91:3798–3802[Abstract/Free Full Text]
  8. Dabbs Jr JM 1990 Age and seasonal variation in serum testosterone concentration among men. Chronobiol Int 7:245–249[Medline]
  9. El-Migdadi F, Nusier M, Bashir N 2000 Seasonal pattern of leuteinizing, follicle-stimulating hormone, testosterone and progesterone in adult population of both sexes in the Jordan Valley. Endocr Res 26:41–48[Medline]
  10. Maes M, Mommen K, Hendrickx D, Peeters D, D’Hondt P, Ranjan R, De Meyer F, Scharpe S 1997 Components of biological variation, including seasonality, in blood concentrations of TSH, TT3, FT4, PRL, cortisol and testosterone in healthy volunteers. Clin Endocrinol (Oxf) 46:587–598[CrossRef][Medline]
  11. Martikainen H, Tapanainen J, Vakkuri O, Leppaluoto J, Huhtaniemi I 1985 Circannual concentrations of melatonin, gonadotrophins, prolactin and gonadal steroids in males in a geographical area with a large annual variation in daylight. Acta Endocrinol (Copenh) 109:446–450[Abstract/Free Full Text]
  12. Meriggiola MC, Noonan EA, Paulsen CA, Bremner WJ 1996 Annual patterns of luteinizing hormone, follicle stimulating hormone, testosterone and inhibin in normal men. Hum Reprod 11:248–252
  13. Nicolau GY, Lakatua D, Sackett-Lundeen L, Haus E 1984 Circadian and circannual rhythms of hormonal variables in elderly men and women. Chronobiol Int 1:301–319[Medline]
  14. Nicolau GY, Haus E, Lakatua DJ, Bogdan C, Sackett-Lundeen L, Popescu M, Berg H, Petrescu E, Robu E 1985 Circadian and circannual variations of FSH, LH, testosterone, dehydroepiandrosterone-sulfate (DHEA-S) and 17-hydroxy progesterone (17 OH-Prog) in elderly men and women. Endocrinologie 23:223–246[Medline]
  15. Perry 3rd HM, Miller DK, Patrick P, Morley JE 2000 Testosterone and leptin in older African-American men: relationship to age, strength, function, and season. Metabolism 49:1085–1091[CrossRef][Medline]
  16. Reinberg A, Lagoguey M, Chauffournier JM, Cesselin F 1975 Circannual and circadian rhythms in plasma testosterone in five healthy young Parisian males. Acta Endocrinol (Copenh) 80:732–734[Abstract/Free Full Text]
  17. Reinberg A, Smolensky MH, Hallek M, Smith KD, Steinberger E 1988 Annual variation in semen characteristics and plasma hormone levels in men undergoing vasectomy. Fertil Steril 49:309–315[Medline]
  18. Ruhayel Y, Malm G, Haugen TB, Henrichsen T, Bjørsvik C, Grotmol T, Sæther T, Malm J, Figenschau Y, Rylander L, Levine RJ, Giwercman A 2007 Seasonal variation in serum concentrations of reproductive hormones and urinary excretion of 6-sulfatoxymelatonin in men living north and south of the Arctic Circle: a longitudinal study. Clin Endocrinol (Oxf) 67:85–92[CrossRef][Medline]
  19. Smals AG, Kloppenborg PW, and Benraad TJ 1976 Circannual cycle in plasma testosterone levels in man. J Clin Endocrinol Metab 42:979–982[Abstract/Free Full Text]
  20. Svartberg J, Jorde R, Sundsfjord J, Bonaa KH, Barrett-Connor E 2003 Seasonal variation of testosterone and waist to hip ratio in men: the Tromso study. J Clin Endocrinol Metab 88:3099–3104[Abstract/Free Full Text]
  21. Tancredi A, Reginster JY, Luyckx F, Legros JJ 2005 No major month to month variation in free testosterone levels in aging males. Minor impact on the biological diagnosis of ‘andropause.’ Psychoneuroendocrinology 30:638–646[CrossRef][Medline]
  22. Touitou Y, Sulon J, Bogdan A, Reinberg A, Sodoyez JC, Demey-Ponsart E 1983 Adrenocortical hormones, ageing and mental condition: seasonal and circadian rhythms of plasma 18-hydroxy-11-deoxycorticosterone, total and free cortisol and urinary corticosteroids. J Endocrinol 96:53–64[Abstract/Free Full Text]
  23. Valero-Politi J, Fuentes-Arderiu X 1998 Annual rhythmic variations of follitropin, lutropin, testosterone and sex-hormone-binding globulin in men. Clin Chim Acta 271:57–71[CrossRef][Medline]
  24. Wisniewski AB, Nelson RJ 2000 Seasonal variation in human functional cerebral lateralization and free testosterone concentrations. Brain Cogn 43:429–438[Medline]
  25. Brambilla DJ, O’Donnell AB, Matsumoto AM, McKinlay JB 3 August 2007 Intra-individual variations in serum testosterone and other reproductive and adrenal hormone levels in men. Clin Endocrinol 10.1111/j.1365-2265.2007.02976.
  26. McKinlay JB, Link CL 2007 Measuring the urologic iceberg: design and implementation of the Boston Area Community Health (BACH) Survey. Eur Urol 52:389–396[CrossRef][Medline]
  27. Bremner WJ, Vitiello MV, Prinz PN 1983 Loss of circadian rhythmicity in blood testosterone levels with aging in normal men. J Clin Endocrinol Metab 56:1278–1281[Abstract/Free Full Text]
  28. Diver MJ, Imtiaz KE, Ahmad AM, Vora JP, Fraser WD 2003 Diurnal rhythms of serum total, free and bioavailable testosterone and of SHBG in middle-aged men compared with those in young men. Clin Endocrinol (Oxf) 58:710–717[CrossRef][Medline]
  29. Longcope C, Franz C, Morello C, Baker R, Johnston Jr CC 1986 Steroid and gonadotropin levels in women during the peri-menopausal years. Maturitas 8:189–196[CrossRef][Medline]
  30. Sodergard R, Backstrom T, Shanbhag V, Carstensen H 1982 Calculation of free and bound fractions of testosterone and estradiol-17ß to human plasma proteins at body temperature. J Steroid Biochem 16:801–810[CrossRef][Medline]
  31. Vermeulen A, Verdonck L, Kaufman J 1999 A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 84:3666–3672[Abstract/Free Full Text]
  32. Derby CA, Zilbert S, Brambilla D, Morales KH, McKinlay JB 2006 Body mass index, waist circumference and waist to hip ratio and change in sex steroid hormones: the Massachusetts Male Aging Study. Clin Endocrinol (Oxf) 65:125–131[CrossRef][Medline]
  33. Feldman HA, Longcope C, Derby CA, Johannes CB, Araujo AB, Coviello AD, Bremner WJ, McKinlay JB 2002 Age trends in the level of serum testosterone and other hormones in middle-aged men: longitudinal results from the Massachusetts Male Aging Study. J Clin Endocrinol Metab 87:589–598[Abstract/Free Full Text]
  34. Fitzmaurice G, Laird N, Ware J 2004 Applied longitudinal analysis. New York: John Wiley
  35. Seber, GAF, Lee AJ 2003 Linear regression analysis. New York: John Wiley



This article has been cited by other articles:


Home page
 Eur J EndocrinolHome page
B Lapauw, S Goemaere, H Zmierczak, I Van Pottelbergh, A Mahmoud, Y Taes, D De Bacquer, S Vansteelandt, and J M Kaufman
The decline of serum testosterone levels in community-dwelling men over 70 years of age: descriptive data and predictors of longitudinal changes
Eur. J. Endocrinol., October 1, 2008; 159(4): 459 - 468.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Brambilla, D. J.
Right arrow Articles by McKinlay, J. B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Brambilla, D. J.
Right arrow Articles by McKinlay, J. B.
Related Collections
Right arrow Male Endocrinology

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals