| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Divisions of Endocrinology, Diabetes, and Clinical Nutrition (C.A.B., B.M., M.K., U.K., J.J.P.), Gynecological Endocrinology and Reproductive Medicine (C.D.G.), and Central Laboratory (P.H.) and Institute for Social and Preventive Medicine (C.S.), University Hospital, CH-4031 Basel, Switzerland
Address all correspondence to: Jardena J. Puder, Division of Endocrinology, Diabetes, and Clinical Nutrition, University Hospital Basel, Petersgraben 4, 4031 Basel, Switzerland. E-mail: puderj{at}uhbs.ch.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Methods: Eight normal weight (body mass index, 21.6 ± 1.9 kg/m2) and nine overweight (body mass index, 30 ± 2.4 kg/m2) young women with normal ovarian function were assessed 15 times throughout the menstrual cycle. Regional fat distribution was assessed using dual x-ray absorptiometry.
Results: Concentrations of highly sensitive C-reactive protein (hs-CRP) changed significantly during the menstrual cycle and were highest in the early follicular phase (P < 0.00001). SHBG concentrations were stable in the follicular phase but increased in the luteal phase (P < 0.00001). During the follicular phase, estimates of insulin resistance had a higher within-subject variance when determined by the homeostasis model assessment index (42%) than when estimated by SHBG concentrations (5%, P < 0.05). During the menstrual cycle, using repeated measurements, hs-CRP correlated inversely to estradiol (ß-coefficient, –0.23, P < 0.0001) and SHBG concentrations (ß-coefficient, –0.83, P = 0.004). Central accumulation of body fat correlated to the mean hs-CRP concentration (r = 0.63, P = 0.007) and the mean homeostasis model assessment index for insulin resistance (r = 0.75, P = 0.001).
Conclusion: We demonstrate that serum concentrations of hs-CRP and SHBG significantly change during the menstrual cycle. We reveal a close link between sex steroids, subclinical inflammation, insulin resistance, and body fat distribution in regularly menstruating women.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Similarly, in premenopausal women, endogenous estrogen has been found to be inversely correlated to IL-6 serum concentrations (11). During the follicular phase of the menstrual cycle with low estrogen and progesterone concentrations, soluble IL-6 receptor concentrations were increased, compared with the luteal phase (12). Also, estimates of insulin resistance change during the menstrual cycle in most studies (13, 14). Recently, low SHBG concentration has been proposed as an additional surrogate marker for insulin resistance in men and anovulatory women (15, 16, 17).
In view of these findings, we investigated the changes of the circulating inflammatory markers hs-CRP, TNF
, and procalcitonin (ProCT) and two estimates of insulin resistance, the homeostasis model assessment index of insulin resistance (HOMA-IR) and SHBG concentrations, during the menstrual cycle and their variance and thus reproducibility in lean and overweight young women. We, furthermore, were interested in the relationship between inflammatory markers, sex steroids, and estimates of insulin resistance throughout the menstrual cycle, and between low-grade inflammation, insulin resistance, and regional body fat distribution in this young and low-risk population.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We studied eight healthy lean female subjects with a body mass index (BMI) of 18–25 kg/m2 and seven healthy age-matched (20–36 yr) overweight females with a BMI of 27–40 kg/m2. Inclusion into the study was based on the following criteria: 1) regular menstrual cycles (26–33 d); 2) not taking oral contraceptives for at least 6 months; 3) not taking any medications or over-the-counter drugs; 4) stable weight for at least 3 months; 5) nonsmoker; 6) no first-degree relative with diabetes; 7) no history of premenstrual syndrome; 8) no minor illness for at least 1 month before commencing the study. Participants were instructed not to exercise vigorously more than 3 h/wk. Subjects were screened by medical history and examination. Women in both groups 1 and 2 had blood drawn on d 21 of their preceding menstrual cycle to confirm ovulation by progesterone levels and for a blood count and clinical chemistry to exclude major illnesses. Asymptomatic urinary tract infection was excluded by a routine urinalysis. The study was approved by the local ethical committee in Basel, Switzerland, and all subjects signed and received a copy of a written informed consent form.
Study design, definition of menstrual cycle phases, and symptoms score
Women in groups 1 and 2 were studied during one menstrual cycle period (i.e. one to two cycles after screening). After an overnight fast of at least 10 h, 30 ml of blood were drawn between 0730 and 0930 h in the morning of d 3, 5, 8–16, 18, 21, 24, and 27 of their menstrual cycle.
Menstrual cycle phases were defined by days from the LH peak: first, each cycle was divided into a follicular period and a luteal period by the serum LH peak. The follicular period was further divided into early follicular phase (d –15 to –9 from the LH peak), midfollicular phase (d –8 to –4), and late follicular phase (d –3 to –1). The luteal period was divided into early luteal phase (d 1–3), midluteal phase (d 4–8), and late luteal phase (d 9–14), respectively. None of the participants developed any clinically apparent infection or other illness during the menstrual cycle that was studied.
Assays
The blood was immediately centrifuged, aliquoted, and stored at –70 C until batch analyses. Serum samples at all time points were assayed for progesterone, estradiol, LH, hs-CRP, TNF, ProCT, SHBG, insulin, and glucose.
Insulin was measured by electrochemiluminescence (Roche-Diagnostics, Rotkreuz, Switzerland). In a reference range of 17.8–173 pmol/liter (2.6–24.9 µU/ml), intraassay coefficient of variation (CV) was 1.90% and interassay CV was 2.60%. The insulin test is highly specific and has no known cross-reactions to proinsulin.
Serum glucose was measured by the hexokinase method. Insulin resistance was estimated by calculating HOMA-IR index [fasting serum insulin (microunits per milliliter) x fasting plasma glucose (millimoles per liter)/22.5] (18).
LH, progesterone, and estradiol were measured by electrochemiluminescence immunoassays (Roche-Diagnostics). The intraassay CVs were 1.80, 2.40, and 3.30%, respectively. The interassay CVs were 5.20, 5.50, and 4.70%, respectively. SHBG was measured by electrochemiluminescence immunoassay (Roche-Diagnostics). The reference range for women was 20–130 nmol/liter. The intraassay CV was 2.70%, and the interassay CV was 5.60%.
hs-CRP was measured automatically by a nephelometric latex immunoassay (Roche-Diagnostics), and TNF was measured by a manual ELISA (R&D Systems, Minneapolis, MN). The sensitivity of hs-CRP was 0.11 mg/liter, and the reference range was less than 0.5–4.71 pg/ml. The intraassay CVs of hs-CRP and TNF were 1.34 and 8.8%, respectively; the interassay CVs of hs-CRP and TNF were 5.70 and 16.70% at the respective cut-offs of the reference range, respectively.
ProCT was measured by an ultrasensitive immunoluminometric assay (ProCa-S; Brahms, Hennigsdorf, Germany). The lower detection limit was 6.0 ng/liter, thus well within the normal reference range of 30 ± 20 ng/liter, whereas the intraassay CV was less than 15%, and the interassay CV was less than 20%. Samples for each subject were run in the same assay, and duplicate measurements for TNF and ProCT were performed for each subject.
Body composition
On d 3 of the menstrual cycle, dual-energy x-ray absorptiometry was performed to determine total and regional lean body and fat mass using an Expert densitometer (Lunar, Madison, WI). Regions of interest (including arms, legs, and trunk) were standardized. Percent body fat and percent lean body mass were calculated. Percent trunk fat was calculated as the ratio of trunk fat tot total fat x 100. Percent extremity fat was calculated as the ratio of total extremity fat (right and left arm fat and right and left leg fat) to total fat x 100. Trunk to extremity fat ratio was determined by dividing percent trunk fat by percent extremity fat (19).
Statistical analysis
The mean values of estimates of insulin resistance and inflammatory markers of both normal-weight and overweight subjects were calculated over the whole menstrual cycle and compared by Student’s t test or Mann-Whitney U test. Associations of these estimates and markers with the different components of the body composition were assessed using bivariate linear regression models. Mean values of the (natural) logarithmically transformed sexual hormone levels, estimates of insulin resistance, and inflammatory markers as well as the symptom scores were calculated for each subject in each menstrual cycle phase and compared by repeated-measures ANOVA. The weight status (normal weight vs. overweight) was added as a covariate. Single values of these laboratory parameters were also expressed in percent of the respective person’s mean value over the whole menstrual cycle and compared by ANOVA with post hoc testing using Fisher multiple comparison tests. The relationship between hs-CRP and TNF, estimates of insulin resistance, sexual hormones, and symptoms in repeated measurements was analyzed by mixed linear models. Variables with a skewed distribution were log transformed for all analyses. Data are shown as means ± SD for normally distributed variables and geometric means and interquartile ranges for not normally distributed variables, respectively. Variance components and intraclass correlations were estimated using a random-effects ANOVA model using the loneway procedure in STATA (Intercooled STATA, version 8, Stata Corp. LP, College Station, TX) (20). For this analysis, time periods without menstrual cycle-specific changes were chosen: the follicular phase for estimates of insulin resistance and the luteal phase for inflammatory markers.
Statistical analyses were done by Statistica for Windows (version 6; StatSoft Inc., Tulsa OK) or Intercooled STATA (version 8, Stata Corp.).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
. Values for estimates of insulin resistance and inflammatory mediators were skewed, and therefore log formation was applied for all statistical analyses. Estimates of insulin resistance and inflammatory markers both were higher in overweight compared with the normal weight subjects. The variances of the serum concentrations of inflammatory markers and estimates of insulin resistance as well as the relationships between inflammatory markers, sex steroid, estimates of insulin resistance, and regional body fat distribution did not differ between normal-weight and overweight subjects. Thus, the results of the two groups together are presented below.
|
Effect of the menstrual cycle on low-grade inflammation and estimates of insulin resistance
In all subjects, hs-CRP and SHBG concentrations changed significantly during the menstrual cycle, independent of the weight status (both P < 0.001). hs-CRP concentrations were highest at the beginning of the menstrual cycle, whereas SHBG concentrations were highest in the luteal phase. The same effect was seen both in normal-weight (P = 0.03 for hs-CRP and P < 0.001 for SHBG, respectively) and overweight (P < 0.001 for hs-CRP and P = 0.01 for SHBG, respectively) probands (Fig. 1). Parallel to this observation, hs-CRP and SHBG also changed significantly over the menstrual cycle when the data were analyzed by expressing a single hs-CRP or SHBG serum concentration of each woman as a percentage of the mean hs-CRP or SHBG serum concentration, respectively, over her menstrual cycle (both P < 0.001). Post hoc analyses of these latter results revealed that SHBG concentrations in all three luteal phases were higher than in all three follicular phases (all P < 0.03) and that hs-CRP serum concentrations were highest in the early follicular phase, compared with all other menstrual cycle phases (P < 0.005, compared with the late luteal phase, P < 0.001, compared with the midfollicular phase, late follicular phase, LH surge, early luteal phase, and midluteal phase). Serum hs-CRP concentrations in the early follicular phase were 1.0 mg/liter (interquartile range, 0.56–2.18 mg/liter), whereas the mean of hs-CRP concentration of all other menstrual cycle phases together was 0.62 mg/liter (interquartile range, 0.29–1.06 mg/liter). In contrast, serum concentrations of TNF, ProCT, and HOMA-IR did not change significantly during the menstrual cycle if the data were analyzed separately for normal-weight or overweight probands or if the two groups were evaluated together (P = 0.66 for TNF, P = 0.42 for ProCT, and P = 0.74 for HOMA-IR in both groups together, respectively).
|
Variation can be expressed by dividing overall variation into between- and within-subject variation and describing the percentage of overall variation attributable to each component. If SD2b is the between-subject variance and SD2w is the within-subject variance, then SD2b/(SD2b + SD2w) is the intraclass correlation and 100 x intraclass correlation is the percentage of variation explained by between-subject variation. The remaining variation is the within-subject variation, which is the combined biological and analytic variation (20).
Table 2 lists the variance components for the serum inflammatory markers and the estimates of insulin resistance. For this analysis, time periods without menstrual cycle-specific changes were chosen: the follicular phase for estimates of insulin resistance and the luteal phase for inflammatory markers. After log-transformation, a random effect ANOVA estimated the intraclass correlation for SHBG to be 0.95 (95% of the variance explained by between-subject variance and 5% by within-subject variance), significantly lower than the estimated intraclass correlation for HOMA-IR (0.58, P < 0.05). Thus, the within-subject variance of SHBG was significantly lower than for HOMA-IR. The intraclass correlation of TNF and hs-CRP was not statistically different from each other (P = NS) but were both higher than the intraclass correlation of ProCT (both P < 0.05).
|
There was a significant relationship between the log-transformed values of inflammatory markers, serum estradiol, and SHBG concentrations during the menstrual cycle as analyzed by random-effects ML regression (Table 3). Thus, a 100% increase in serum estradiol concentration was associated with a decrease in serum hs-CRP concentration of 23%. Progesterone did not influence hs-CRP concentrations (P = 0.15). These results were similar after adjusting for the following covariates, respectively: changes in weight, serum estradiol, SHBG, and TNF concentrations (data not shown).
|
The log-transformed mean hs-CRP concentration correlated well with the amount of total fat (n = 15, r = 0.53, P = 0.024) and the trunk to extremity fat ratio (n = 15, r = 0.63, P = 0.007; Fig. 2), even after adjusting for BMI (P = 0.04). The percentage of tissue fat correlated with the mean hs-CRP concentrations (n = 15, r = 0.55, P = 0.03) as well as the hs-CRP concentrations in the early follicular phase, although the latter did not reach statistical significance (n = 15, r = 0.5, P = 0.06).
|
). |
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In premenopausal women, sex steroids have been shown to influence proinflammatory mediators. In accordance with our data, Gorai et al. (12) observed a rise of soluble IL-6 receptor at times of low estrogen and progesterone concentrations, i.e. the follicular phase of the menstrual cycle. Similarly, another study has found an inverse correlation between endogenous estrogen concentrations and serum IL-6 levels in the follicular phase (11). In our study, we could not find an effect of the menstrual cycle or of estradiol on TNF. This could be due to the predominantly local reported effects of TNF (4) and its short half-life (22). Indeed, the effects of estrogen on TNF levels are controversial (4, 23, 24, 25).
Estrogen has been shown to decrease unstimulated and stimulated serum concentrations of cytokines like IL-6 and its receptor and TNF (4, 9). Estrogen in vivo also suppresses the secretion of cytokines in blood or bone marrow cells (5, 10). On a molecular level, estradiol in vitro suppresses IL-6 production by bone and marrow stromal cells (26). Thus, the effect of estrogen on bone metabolism is thought to be modulated through bone-resorbing cytokines like IL-1, IL-6, and TNF (10, 24, 26). Accordingly, cytokines seem to be the causative agents underlying the bone loss induced by estrogen deficiency (3, 24). Accordingly, in parallel with the rise in certain proinflammatory mediators, bone resorption markers are elevated in the follicular phase of the menstrual cycle and correlate inversely to estrogen concentrations (11, 12, 27).
In addition, menstruation per se results in a local increase in inflammatory mediators and thus contributes to elevated serum levels of inflammatory mediators in the early follicular phase (28).
Insulin resistance also changes during the menstrual cycle and is thought to be increased in the luteal phase (13, 14). However, this view has been challenged because the two studies that measured insulin resistance by the euglycemic hyperinsulinemic clamp could not demonstrate a change during the menstrual cycle (29, 30). We could not find any change of insulin resistance during the menstrual cycle using the HOMA-IR index. This might be due to the small number of probands in our study. However, HOMA-IR might not reflect insulin resistance accurately enough in healthy young subjects to be able to detect small changes (18, 31). Another reason for the lack of changes of HOMA-IR could be the high within-subject variation of this index (17, 32). Indeed, we found in our own lean and overweight subjects that in the follicular phase, a time period without menstrual cycle-specific changes of insulin resistance, the within-subject variance was high for HOMA-IR, and markedly higher than the within-subject variance for SHBG. Thus, the reproducibility of SHBG was superior to the one of HOMA-IR. Recently, low SHBG concentration has been proposed as a surrogate marker for insulin resistance in men, postmenopausal women, and anovulatory women with the polycystic ovary syndrome (15, 16, 17). In women with regular menstrual cycles, SHBG concentrations in the follicular phase correlated well with the insulin resistance calculated by the clamp (33). A recent study has shown that the variation of SHBG is similar in insulin-sensitive controls, compared with insulin-resistant patients with type 2 diabetes (17). In contrast, the within-subject variation of HOMA-IR seems to be even larger in insulin-resistant subjects with the polycystic ovary syndrome or type 2 diabetes (16, 32). In the evaluation of the within-subject variation of SHBG, menstrual-cycle-specific changes have to be taken into consideration. Indeed, in accordance with previous data, we found that SHBG concentrations change during the menstrual cycle with an increase during the luteal phase (34).
In summary, HOMA-IR is very good for the overall estimate of insulin resistance. However, for serial monitoring of changes in insulin resistance in a given woman on treatment or during weight changes, SHBG concentrations, measured in the follicular phase, might be more reliable due to their low within-subject variance. Regarding changes in inflammatory markers, both hs-CRP and TNF showed a small within-subject variance in the luteal phase, i.e. a period during which both markers had no menstrual cycle-specific changes. Recently, a high within-subject variance of hs-CRP has been shown in patients with coronary artery disease and healthy subjects including premenopausal women (35). In these studies, the observed variance might be due to large functional assay variabilities, menstrual cycle-specific changes of hs-CRP, or inherent fluctuations of the inflammatory status of the patients studied. In our healthy subjects, both hs-CRP and TNF were superior for the assessment of subtle changes in noninfection-related, low-grade inflammation, compared with highly sensitive ProCT measurement.
Low-grade inflammation and central obesity are both strong determinants of insulin resistance (1, 22, 36). CRP is known to correlate with both total body fat and central accumulation of body fat (37). Most of these data has been accumulated in high-risk subjects. However, we could demonstrate in a relatively small number of young, low-risk, regularly menstruating women that central accumulation of body fat was related both to hs-CRP and insulin resistance. However, due to the small number of probands studied, we might have missed or overestimated certain interactions among subclinical inflammation, insulin resistance, and sex steroids.
In summary, we demonstrated that serum concentrations of hs-CRP and SHBG significantly change during the menstrual cycle in normal-weight and overweight women and that hs-CRP concentrations are highest in the early follicular phase, whereas SHBG concentrations were highest during the luteal phase of the menstrual cycle. Both hs-CRP and TNF have an excellent reproducibility. However, the reproducibility of HOMA-IR was significantly inferior to the one of SHBG in nonmenstrual cycle-specific periods. During the menstrual cycle, hs-CRP correlated inversely to estradiol and SHBG concentrations and positively to TNF levels independent of other covariates. In these young, normal-weight and overweight, regularly menstruating women, we revealed a close link among sex steroids, symptoms, low-grade inflammation, insulin resistance, and body fat distribution.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online March 29, 2005
Abbreviations: BMI, Body mass index; CV, coefficient of variation; HOMA-IR, homeostasis model assessment index of insulin resistance; hs-CRP, highly sensitive C-reactive protein; ProCT, procalcitonin.
Received February 2, 2005.
Accepted March 21, 2005.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
receptors circulate in proportion to leptin levels during the menstrual cycle in lean but not in obese women. Eur J Endocrinol 143:235–241[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  T. Sir-Petermann, M. Maliqueo, E. Codner, B. Echiburu, N. Crisosto, V. Perez, F. Perez-Bravo, and F. Cassorla Early Metabolic Derangements in Daughters of Women with Polycystic Ovary Syndrome J. Clin. Endocrinol. Metab., December 1, 2007; 92(12): 4637 - 4642. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Shroff, A. Kerchner, M. Maifeld, E. J. R. Van Beek, D. Jagasia, and A. Dokras Young Obese Women with Polycystic Ovary Syndrome Have Evidence of Early Coronary Atherosclerosis J. Clin. Endocrinol. Metab., December 1, 2007; 92(12): 4609 - 4614. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O Rogowski, I Shapira, A Shirom, S Melamed, S Toker, and S Berliner Heart rate and microinflammation in men: a relevant atherothrombotic link Heart, August 1, 2007; 93(8): 940 - 944. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Campbell and A. McTiernan Exercise and Biomarkers for Cancer Prevention Studies J. Nutr., January 1, 2007; 137(1): 161S - 169S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Muller and M. Tamm Biomarkers in acute exacerbation of chronic obstructive pulmonary disease: among the blind, the one-eyed is king. Am. J. Respir. Crit. Care Med., October 15, 2006; 174(8): 848 - 849. [Full Text] [PDF]
|
|
|
|
 |
|
 D M Wunder, M Yared, N A Bersinger, D Widmer, R Kretschmer, and M H Birkhauser Serum leptin and C-reactive protein levels in the physiological spontaneous menstrual cycle in reproductive age women. Eur. J. Endocrinol., July 1, 2006; 155(1): 137 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Moriconi, M. Kraenzlin, B. Muller, U. Keller, C. P. G. Nusbaumer, S. Stohr, M. Tamm, and J. J. Puder Body Composition and Adiponectin Serum Concentrations in Adult Patients with Cystic Fibrosis J. Clin. Endocrinol. Metab., April 1, 2006; 91(4): 1586 - 1590. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. K. Koh and B.-K. Yoon Controversies regarding hormone therapy: Insights from inflammation and hemostasis Cardiovasc Res, April 1, 2006; 70(1): 22 - 30. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. N. Simhan and L. M. Bodnar Prepregnancy Body Mass Index, Vaginal Inflammation, and the Racial Disparity in Preterm Birth Am. J. Epidemiol., March 1, 2006; 163(5): 459 - 466. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Ibanez, C. Valls, and F. de Zegher Discontinuous low-dose flutamide-metformin plus an oral or a transdermal contraceptive in patients with hyperinsulinaemic hyperandrogenism: normalizing effects on CRP, TNF-{alpha} and the neutrophil/lymphocyte ratio Hum. Reprod., February 1, 2006; 21(2): 451 - 456. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Puder, S. Varga, M. Kraenzlin, C. De Geyter, U. Keller, and B. Muller Central Fat Excess in Polycystic Ovary Syndrome: Relation to Low-Grade Inflammation and Insulin Resistance J. Clin. Endocrinol. Metab., November 1, 2005; 90(11): 6014 - 6021. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Puder, B. Muller, and U. Keller Letter re: The Biological Variation of Testosterone and Sex Hormone-Binding Globulin (SHBG) in Polycystic Ovarian Syndrome: Implications for SHBG as a Surrogate Marker of Insulin Resistance J. Clin. Endocrinol. Metab., July 1, 2005; 90(7): 4419 - 4420. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
