AJP - Endo Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Endocrinol Metab 293: E466-E474, 2007. First published May 15, 2007; doi:10.1152/ajpendo.00126.2007
0193-1849/07 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
293/2/E466    most recent
00126.2007v1
Right arrow Alert me when this article is cited
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 ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dadoun, F.
Right arrow Articles by Dutour, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dadoun, F.
Right arrow Articles by Dutour, A.

Effect of sleep apnea syndrome on the circadian profile of cortisol in obese men

F. Dadoun,1,2,3 P. Darmon,1,3 V. Achard,1,2 S. Boullu-Ciocca,1,2,3,4 F. Philip-Joet,5 M. C. Alessi,1,2 M. Rey,5 M. Grino,1,2 and A. Dutour1,2,3

1INSERM, UMR 626; 2Université de la Méditerranée; 3Department of Endocrinology and Nutrition, Hôpital Nord, Assistance Publique-Hôpitaux de Marseille; 4Centre for Clinical Investigations of Marseille, INSERM and Assistance Publique-Hôpitaux de Marseille; and 5Center for the Study of Sleep, Department of Clinical Neurophysiology, Hôpital de la Timone, Marseille, France

Submitted 26 February 2007 ; accepted in final form 30 April 2007


   ABSTRACT
TOP
 ABSTRACT
STUDY DESIGN
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
It has been hypothesized that sleep apnea syndrome (SAS) increases hypothalamic-pituitary-adrenal axis activity and, through increased cortisol levels, participates in the pathophysiology of metabolic and cardiovascular complications. We compared the circadian profiles of cortisol in obese men with [obSAS+; apnea-hypopnea index (AHI) ≥ 20/h] and without SAS (obSAS–; AHI ≤ 5/h). 1) Salivary cortisol (5 samples: before/30 min after dinner, 2100, upon/30 min after awakening) was measured in 15 obSAS+, 19 obSAS–, and 19 normal-weight controls (NWC). 2) Plasma cortisol (every 30 min for 24 h under highly controlled conditions and portable EEG device) was measured in 9 obSAS+, 8 obSAS–, and 10 NWC men. Visceral adipose tissue surface was measured by CT scan. In both studies, obSAS+ and obSAS– men were comparable for age, BMI, waist circumference, and waist-to-hip ratio. First, no difference was found, using ANOVA for repeated measures, between obSAS+ and obSAS– subjects for any salivary cortisol measurement. No correlation was found between salivary cortisol and AHI or nocturnal SaO2. Similarly, obSAS+ and obSAS– men showed no difference in plasma cortisol rhythmicity: 24-h minimum, maximum, and mean, ANOVA for repeated measures, mathematical modeling of cortisol rhythm (COSINOR), and morning secretory peak. Conversely, ANOVA for repeated measures showed decreased cortisol levels in obese vs. NWC men during both the trough (2200–0130) and the peak (0600–0900) independently of SAS status. We show that SAS per se is not associated with any change of the level or of the features of salivary and plasma cortisol rhythmicity and confirm that men with visceral obesity display lower plasma cortisol levels than NWC men.

cortisol; circadian rhythm; obesity; sleep-disordered breathing

SLEEP APNEA SYNDROME (SAS) is highly prevalent in obese subjects, especially in males with abdominal obesity (20). SAS is characterized by repeated episodes of apnea and hypopnea during sleep, and is associated with an increased prevalence of insulin resistance and cardiovascular morbidity and mortality independently of obesity (29). The mechanisms responsible for the increased prevalence of cardiovascular and metabolic complications seen in patients with SAS are poorly understood. Subtle changes in hypothalamic-pituitary-adrenal (HPA) axis activity and/or in peripheral actions of cortisol have been involved in the development of abdominal obesity and insulin resistance (4). Furthermore, HPA axis and sleep display strong interactions: the HPA axis plays a major role in maintaining alertness and modulating sleep, and sleep disturbances alter HPA axis activity (33). Experimental nocturnal awakenings cause significant pulsatile cortisol release (31). Patients with chronic insomnia show elevated cortisol levels, especially during the evening and the early period of nocturnal sleep (37). In addition, experimental sleep deprivation is associated with HPA axis hyperactivity, hypertension, and insulin resistance (32). It has thus been recently hypothesized that SAS, because of subsequent arousal, hypoxia, and/or autonomic activation, may increase HPA axis activity and thereby participate in the pathogenesis of the metabolic disturbances of obesity (7). However, scarce data support this appealing hypothesis. The main purpose of the current study was to evaluate whether SAS is actually associated with an increase of circulating cortisol, which may subsequently drive or worsen metabolic and cardiovascular complications.

To address this question and to assess circulating cortisol during the whole wake/sleep cycle, we performed two separate studies. In the first one, we studied the circadian profile of salivary cortisol during the wake phase in 39 obese men with or without SAS and in 19 normal-weight male controls. In this first study, we used salivary cortisol measurements, because on the one hand, they closely reflect the concentrations of free, biologically active cortisol in the plasma, independently of the rate of saliva secretion; on the other hand, saliva collection is noninvasive and avoids the confounding effect of stressful blood withdrawal.

In the second study, to analyze the nocturnal period and the features of the circadian rhythmicity of cortisol, repeated blood sampling over the whole nyctohemeral period was chosen, because, indeed, salivary collection cannot be performed during sleep. Twenty-four obese men, with (obSAS+) or without (obSAS–) SAS and matched for age and body mass index (BMI), and 12 normal-weight controls (NWC) were studied under concomitant sleep monitoring by use of an ambulatory electroencephalogram device.

It was previously shown that abdominal/visceral fat deposition is an important risk factor for SAS in obese patients (30) and also that abdominal obesity may alter the circulating levels of cortisol. Low plasma or saliva cortisol concentrations have been described in obese subjects (9, 14) despite a mild to moderate increase in cortisol production, which has been related to an increased metabolic clearance of the hormone (24). Indeed, a negative relationship between plasma or salivary cortisol levels at 0800 and waist-to-hip ratio has been documented in some studies (9, 19) but not in others (21, 26). To avoid potentially confounding effects of body fat repartition, we carefully documented body fat distribution in both studies by anthropometry and by tomodensitometry in obese subjects, and we carefully analyzed, in both studies, the relationship between salivary/plasma cortisol and body fat distribution.


   STUDY DESIGN
TOP
 ABSTRACT
 STUDY DESIGN
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Study 1

Subjects. The study was approved by the local committee for ethics, and informed consent was obtained from all participants. For 1 yr, all obese patients (BMI >30 kg/m2) aged 18–70 yr who were admitted at Marseille's Centre for the Study of Sleep for the detection of SAS were invited to participate in the study. Thirty-nine obese men and 19 normal-weight male volunteers (BMI <25 kg/m2) were recruited. Exclusion criteria for all participants included any endocrine disease, pharmacological treatment with corticosteroids or psychotropic drugs, nightshift working, transmeridian travel during the previous 3 mo, and recent weight loss or gain. Further exclusion criteria were used for the normal-weight controls: use of any pharmacological treatment, presence of any sleep disturbance either diagnosed or reported, a score greater than or equal to 4 on the Epworth Sleeping Scale (15), presence of any of the features of the metabolic syndrome (NCEP-ATPIII criteria, NCEP 2001), and early cardiovascular disease in first-degree relatives.

Sleep study and grouping procedure. In all participants, a polysomnography was performed with continuous recording of electroencephalogram (EEG), electromyogram (EMG), and electrooculogram (EOG), nasal airflow, body position, thoracic and abdominal respiratory efforts, and arterial oxyhemoglobin saturation (SaO2) recorded by a pulse oxymeter. Apnea was defined as the cessation of airflow for at least 10 s; a reduction in the amplitude of the rib cage, and abdominal excursions with a decrease in ventilation exceeding 50% that lasted at least 10 s associated with an SaO2 reduction of at least 4% defined hypopnea. SAS was diagnosed if the average number of apnea and hypopnea per hour of sleep [apnea-hypopnea index (AHI)] was >5. In the present study, obese patients with AHI ≤5 defined the obSAS– group (n = 19); obese patients with severe SAS (AHI ≥20) defined the obSAS+ group (n = 15); obese subjects with AHI >5 and <20 (n = 5) were excluded from part of the analyses performed in this study. In some analyses, obese patients of the obSAS– and obSAS+ groups and obese patients with AHI >5 and <20 were pooled to build an all-obese group (OB, n = 39). Among the normal-weight volunteers, none was diagnosed with SAS (AHI ≤5), so that all were included in the NWC group (n = 19).

Assessment of body fat distribution. Anthropometric parameters (weight, height, BMI, waist, hip, waist-to-hip ratio) were measured in all subjects participating in the study. On a separate day distant by no more than 7 days from the day of saliva sampling, computed tomography using a single cross-sectional scan at the level of the L4–L5 intervertebral disk was also performed in obese patients to measure the surfaces of total, subcutaneous, and visceral abdominal adipose tissue (TAT, SAT, VAT), as previously described (13).

Metabolic assessment. On the morning that followed the collection of the last saliva sample, a blood sample was withdrawn in the fasting state for the measurement of whole blood glucose, total cholesterol, HDL-cholesterol, triglycerides, and serum insulin.

Salivary cortisol circadian profile. Salivary cortisol concentration was assessed from saliva samples collected using cotton dental rolls (Salivettes; Sarstedt, Leicester, UK).

Five saliva samples were collected in the Centre for the Study of Sleep: before dinner (1900), 30 min after dinner (1930), before placement of the polysomnography device (2100), upon awakening (0700), and 30 min later (0730), before breakfast was served. We determined mean (Mean), maximal morning (Max), minimal evening (Min), and {Delta}Max-Min values.

Study 2

Subjects. This study was approved by the local committee for ethics. Informed consent was obtained from all subjects. Among abdominally obese male patients (BMI >30 kg/m2, waist circumference >102 cm) referred to the Centre for the Study of Sleep of Marseilles University Hospital, Timone, for the detection of SAS, we recruited 12 obese male subjects with severe SAS (AHI ≥20), and 12 obese male subjects without SAS (AHI <5), according to the results of a polysomnography (performed as described in study one); each obese subject with SAS was matched with an obese subject without SAS for age (±5 yr) and BMI (±2 kg/m2). This matching procedure was aimed at limiting as much as possible any confounding between group differences, especially with respect to the degree of adiposity/body fat repartition. The study was then performed in the month that followed the polysomnography used for selection. Twelve healthy, normal-weight male volunteers (BMI <25 kg/m2) matched for age were also recruited to build a control group. Exclusion criteria for all three groups included any endocrine disease, pharmacological treatment with corticosteroids or psychotropic drugs, night-shift working, transmeridian travel during the previous 3 mo, and recent weight loss or gain. Further exclusion criteria were used for the NWCs: use of any pharmacological treatment, presence of any sleep disturbance either diagnosed or reported, a score greater than or equal to 4 on the Epworth Sleeping Scale (15), presence of any of the features of the metabolic syndrome (NCEP-ATPIII criteria, NCEP 2001), and early cardiovascular disease in first degree relatives.

Among recruited subjects, eight obese subjects with AHI <5 (obSAS– group), nine obese subjects with AHI '20 (obSAS+ group), and 10 NWCs completed the whole study without any disruption and were considered eligible for data analysis. Especially, subjects who experienced nocturnal awakening (other than microawakening, according to EEG records) or for whom blood collection resumed prematurely because of technical matters were excluded from the study. In some analyses, patients of the obSAS– and obSAS+ groups were pooled to build an all-obese group (OB, n = 17).

Baseline assessment. During the week that preceded the nyctohemeral study, anthropometric and biological characteristics of the patients were collected: weight, height, waist and hip circumferences, and waist-to-hip ratio. The surfaces of TAT, SAT, VAT were measured by CT, as described in study 1. The metabolic profile was assessed as described in study 1.

Twenty-four-hour blood sampling with sleep monitoring. Subjects were advised to respect strictly their usual bedtime, wake time, and mealtime pattern during the week before completion of the 24-h study. All patients were admitted to the Center for Clinical Investigation for 24-h blood sampling and concomitant sleep monitoring. Two hours before the beginning of the 24-h study, a catheter was inserted into an antecubital vein, and patients were subjected to repeated blood sampling every 30 min from 1200 to 1200 the second day. Samples were immediately centrifuged at 4°C, and serum was stored at –20°C until assays. A maximum of 250 ml of blood was removed.

During the daytime of the sampling period, subjects were allowed to freely walk and sit in the research room and watch TV but not to sleep or snack outside scheduled meals. At 2200, subjects were moved to a sound-attenuated, light-controlled room that had a comfortable bedroom-like atmosphere. An inner wall separated the bed area from the sampling area, which received light from a photography-red light bulb, so that subjects could sleep in full dark and were minimally disturbed.

The sleep-wake cycle was recorded on an ambulatory EEG device (Micromed Brain Spy, Venice, Italy) with 12 leads including C3A2, EOG, and EMG for sleep evaluation. Arousal periods concomitant to apnea were carefully recorded.

Assays. Whole blood fasting glucose, total cholesterol, HDL-cholesterol, and triglycerides were measured using automated enzymatic assays (Vitros; Ortho-Clinical Diagnostics, Rochester, NY). Fasting serum insulin was measured using an immunoradiometric assay (Sanofi-Pasteur Diagnostics).

Salivary cortisol was measured using an RIA kit (CORT-CT2, CIS Bio International). Intra- and interassay variability were 3.6 and 6.8%, respectively. Limit of detection was 0.8 nmol/l.

Serum cortisol was measured using an immunoradiometric assay (Immulite cortisol; DPC, La Garenne-Colombes, France) with inter- and intra-assay coefficients of variation of 8 and 10% and limit of detection 0.2 µg/dl. All samples from a given subject were analyzed in the same assay.

Statistical Analysis

All data are given as means ± SE. Statistical analysis was performed using Statview 5.0 for Windows. Between-group comparisons were performed using Student's t-tests or Mann-Whitney tests for two-group comparisons, and ANOVA or Kruskal-Wallis tests for three-group comparisons, when appropriate. ANCOVA followed by Fisher's tests were used to adjust between group comparisons for continuous variables. ANOVA for repeated measures was used to perform between-group comparisons of cortisol levels during time periods that comprised multiple measurements. Correlation between two parameters was analyzed by linear regression. A P value of <0.05 was considered statistically significant.

Analysis of Nyctohemeral Rhythmicity of Plasma Cortisol and Related Statistical Analysis

The nyctohemeral pattern of cortisol was assessed using COSINOR and GraphPad Prism 4.0 software for Windows. The classical sine wave curve modeling (COSINOR) was used to analyze the circadian rhythmicity of plasma cortisol (16), to check for changes of the organization and/or of the phase of cortisol circadian rhythm, that may relate to obesity and/or SAS. The modeling of cortisol data was performed not only per individual, but also per group, because per group modeling allows the use of statistical methods that are more powerfull to identify between group differences. Between group comparisons (OB vs. NWC, and obSAS– vs. obSAS+) was performed using t-tests to compare parameters obtained from the modeling of individual data, and F-tests to compare parameters obtained from the modeling of group data. Since COSINOR sinewave curves do not accurately fit the morning cortisol secretory peak, we also used an alternate model ("peak model"), which is more adequate to describe secretory events and allows asymmetrical ascent and descent (equation: Y = y0 + ymax/[1 + exp(z1 – X)] x {1 + exp[2(X – z2)]}, where y0 is the prepeak baseline plasma cortisol, ymax the maximum increment of the peak, z1 the inflection time point of the ascending leg of the peak, and z2 the inflection time point of the descending leg of the peak). It permits a finer analysis of potential changes of the nadir of cortisol and of the timing and amplitude of the morning cortisol secretory peak that may relate to obesity and/or SAS. It was applied to the time period 2200–1200, and the modeling was performed both per individual and per group. Similar statistical analysis was performed for between-group comparisons of the parameters (y0, ymax, z1, z2) obtained from individual or group modeling.


   RESULTS
TOP
 ABSTRACT
 STUDY DESIGN
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Study 1

Table 1 summarizes the clinical and biological characteristics of the normal-weight and obese men. As expected, BMI, waist circumference, and waist-to-hip ratio were higher in obese (OB group as a whole; obSAS+ and obSAS– groups considered separately) than in the NWC group. Fasting blood glucose, serum insulin, and triglycerides were significantly higher, and HDL-cholesterol significantly lower, in OB than in NWC (data not shown). Conversely, there was no difference between groups for age. Interestingly, obSAS+ and obSAS– groups were found comparable for age, BMI, waist-to-hip ratio, SAT, and metabolic parameters. However, obSAS+ men had significantly more VAT than obSAS– men (277 ± 25 vs. 212 ± 17 cm2, P < 0.05, respectively).


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

  		Table 1. Study 1: clinical and biological characteristics

 
Salivary cortisol: comparison of obese vs. normal-weight men. Obese men displayed diminished Max and {Delta}Max-Min values compared with controls (16.2 ± 1.4 vs. 20.2 ± 1.9 nmol/l, P < 0.05; 13.4 ± 1.3 vs. 17.3 ± 1.8 nmol/l, P < 0.05, respectively). No difference was found between groups for the other values.

Salivary cortisol: comparison of obSAS– vs. obSAS+, and relationship to indexes of SAS severity. When obSAS– and obSAS+ men were compared, no difference was found for salivary cortisol measurements at any sampling time or using ANOVA for repeated measures. Furthermore, Mean, Max, Min, and {Delta}Max-Min values were also similar in both groups. Because VAT was found significantly greater in the obSAS+ group than in the obSAS– group, and to avoid a confounding effect of this difference in VAT, we also adjusted all the previous comparisons for VAT: no more difference in any salivary cortisol measurement was identified between obSAS+ and obSAS– groups after adjustment. Moreover, in the obese population as a whole (including 5 subjects with 5 < AHI < 20), simple regression analysis evidenced no correlation between any salivary cortisol measurement and AHI, mean nocturnal SaO2, or minimum nocturnal SaO2.

Salivary cortisol: relationship to body fat repartition. Conversely, within the whole OB group, a negative correlation was found between VAT and salivary cortisol concentration at 1900 (simple regression analysis: r = –0.46, P < 0.005). Moreover, obese men with VAT above the median value of the group (235 cm2) showed significantly lower levels of salivary cortisol at 1900 than obese with VAT below this median value (1900 salivary cortisol; obese subjects with VAT >235 cm2: 2.7 ± 0.2 nmol/l vs. obese subjects with VAT <235 cm2: 4.3 ± 0.6 nmol/l, P < 0.05); a similar significant difference was found for mean cortisol level throughout the day (mean 24-h salivary cortisol: obese subjects with VAT >235 cm2: 7.9 ± 0.9 nmol/l vs. obese subjects with VAT <235 cm2: 10.3 ± 1.0 nmol/l, P < 0.05). Among body fat repartition indexes, only VAT appeared to influence salivary cortisol values, since with simple regression analysis no correlation was found between any salivary cortisol concentration and either BMI or SAT. In addition, no difference in salivary cortisol levels (any sampling time, Mean, Min, Max, and {Delta}Max-Min) was found when obese men above vs. below the median BMI, waist circumference, waist-to-hip ratio, proximal thigh circumference, TAT, and SAT of the population (data not shown) were compared.

These data as a whole suggest that SAS has no influence on evening and morning salivary cortisol levels in the wake phase but that visceral fat excess, and not subcutaneous fat excess, and/or global corpulence impacts salivary cortisol levels.

Study 2

As shown in Table 2, waist circumference, waist-to-hip ratio, TAT, SAT, and VAT were higher in the OB, obSAS+, and obSAS– groups than in NWC. These measurements were not significantly different between obSAS– and obSAS+ groups, emphasizing the efficiency of the matching procedure used for patient selection. Despite the matching procedure, VAT was slightly, but not significantly, greater (P = 0.49) in the obSAS+ group than in the obSAS– group. Fasting blood glucose, insulin, and triglycerides were significantly higher, and HDL-cholesterol significantly lower, in OB (and obSAS– and obSAS+ subgroups) than in NWC.


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

  		Table 2. Study 2: clinical, biological, and sleep characteristics

 
Plasma cortisol: comparison of OB vs. NWC men and relationship to body fat repartition. When OB men were compared with NWC men, we evidenced discrete differences in plasma cortisol levels. Mean 24-h plasma cortisol was not different between OB and NWC; neither did ANOVA for repeated measures show any difference in cortisol levels over the 24-h period between OB and NWC. However, the nocturnal minimum (Min) and the morning maximum (Max) of plasma cortisol tended to be lower in OB than in NWC (P = 0.08 and 0.06, respectively). In addition, when two specific time periods were analyzed, i.e., the trough (2200–0130) and the peak (0600–0900) of cortisol secretion, cortisol levels were found significantly lower in OB than in NWC, using both ANOVA for repeated measures (trough F = 4.48, P = 0.04; peak F = 8.19, P = 0.01) and t-tests for the comparison of mean plasma cortisol during these periods (Table 3). Interestingly, mean plasma cortisol during these trough and peak periods was found negatively correlated to BMI (simple regression analysis: trough r = –0.39, P < 0.05; peak r = –0.52, P = 0.007) and to visceral fat (simple regression analysis: trough r = –0.25, P < 0.05; peak r = –0.46, P = 0.03).


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

  		Table 3. Study 2: summary of plasma cortisol data

 
Plasma cortisol: comparison of obSAS+ vs. obSAS– and relationship to indexes of SAS severity. Conversely, when obSAS+ and obSAS– groups were compared with the same methods, no difference in any plasma cortisol measurement was evidenced (Table 3). Because age and VAT were slightly greater in obSAS+ than in obSAS– (even though the differences were not significant: P = 0.20 and 0.49, respectively), and because they may both influence plasma cortisol, we also adjusted all comparisons for age and/or VAT: no difference was found between obSAS– and obSAS+ groups in any adjusted cortisol measurement (24-h mean, morning maximum, trough-period mean, or peak-period mean of plasma cortisol). Furthermore, using simple regression analysis, no cortisol measurement (24-h mean, nocturnal minimum, morning maximum, trough-period mean, peak-period mean) was found correlated to either AHI or to mean and minimal nocturnal SaO2.

To look for potential differences in plasma cortisol levels between obSAS– and obSAS+, during other specific time periods, ANOVA for repeated measures was also applied to every 2-, 4-, 6-, and 8-h periods within the 24-h sampling period and did not evidence any difference between obSAS– and obSAS+ groups (data not shown). The prandial response of plasma cortisol (difference between premeal cortisol and maximal cortisol within the corresponding 2-h postprandial period) to each of three standardized meals (breakfast, lunch, and dinner), and the response of plasma cortisol to awakening (difference between preawakening cortisol and maximal cortisol within the 1-h postawakening period) were also found similar in obSAS– and obSAS+ groups (data not shown). In addition, thorough inspection of individual nocturnal profiles of plasma cortisol and of EEG records did not identify any specific change (especially any increase) of plasma cortisol during or after periods of nocturnal arousal.

Plasma cortisol: between-group comparisons using mathematical modeling. Because mathematical modeling of cortisol data over a more prolonged time period may identify subtle between-group differences, which could have been overlooked by the previous methods used for comparison, we also applied two mathematical models to cortisol data, using both per individual and per group modeling. Figure 1 shows the measured nyctohemeral profiles (Fig. 1, A and B) and the corresponding modeling curves (COSINOR analysis: Fig. 1, C and D; alternate "peak model": Fig. 1, E and F) of plasma cortisol per study group.


Figure 1
View larger version (47K):
[in this window]
[in a new window]

  		Fig. 1. Measured crude nyctohemeral profiles of plasma cortisol (means ± SE) and corresponding fitting curves obtained from mathematical modeling performed per group. A, C, and E: normal-weight control (NWC) group (n = 10): measured cortisol values, dark blue solid squares; modeling curve, dark blue solid line; 95% confidence interval of the curve, dark blue dotted lines. Obese (OB) group (n = 17): measured cortisol values, red open rhombi; modeling curve, red solid line; 95% confidence interval of the curve, red dotted lines. C: sine wave modeling (COSINOR analysis, 24-h period). The mesor was found significantly lower in OB than in NWC subjects (P < 0.05); no difference was found for other parameters obtained from mathematical modeling. E: alternate modeling of cortisol peak (2200–1200 period). The baseline of cortisol peak (y0) was found significantly lower in OB than in NWC subjects (P < 0.05); no difference was found for other parameters obtained from mathematical modeling. B, D, and F: obese without sleep apnea syndrome (obSAS–) group (n = 8): measured values, green open triangles; modeling curve, green solid line; 95% confidence interval of the curve, green dotted lines. obSAS+ group (n = 9): measured values, purple solid triangles; modeling curve, purple solid line; 95% confidence interval of the curve, purple dotted lines. B: sine wave modeling (COSINOR analysis; 24-h period). D: alternate peak modeling (2200–1200 period). No significant difference was found between obSAS+ and obSAS– groups for either measured or modeled data.

 
Sine wave modeling. Per individual sinewave modeling of cortisol data did not show any statistical difference for mesor (midline estimating statistic of the circadian rhythm), amplitude, or phase (timing of acrophase) between OB and NWC groups or between obSAS– and obSAS+ groups (t-tests; Table 3). Using per group sine wave modeling and the more powerful F-tests for between-group comparisons, we found no difference in the amplitude or the phase (acrophase timing) of plasma cortisol circadian rhythm between NWC and OB, but the mesor was found significantly lower in OB than in NWC (Table 3 and Fig. 1C).

Conversely, per group sine wave modeling did not evidence any statistically significant difference between obSAS– and obSAS+ (Table 3 and Fig. 1D), even though the mesor tended to be lower in obSAS+ than in obSAS– (P = 0.07).

Alternate peak modeling. The alternate model (peak model) provided a significantly better fit of plasma cortisol profiles than sine wave modeling using either per group modeling (F-tests, P < 0.0001 for each group) or per individual modeling (paired t-test, P < 0.0001). When NWC and OB groups were compared using this peak model, the data confirmed that plasma cortisol baseline level was lower in OB than in NWC, using both per individual (y0 = 23.2 ± 2.5 vs. 38.7 ± 6.4 µg/l, P = 0.01) or per group modeling (P < 0.0001; Table 3 and Fig. 1E). Moreover, by use of individual modeling, the increment of the morning peak of plasma cortisol (ymax) was also found significantly smaller in obese subjects (P < 0.05; Table 3), suggesting a flattening of the peak in obesity, which appears to relate to visceral adiposity, since individual ymax were found negatively correlated to VAT (simple regression analysis: r = –0.45, P = 0.03).

Conversely, once again no difference was evidenced between obSAS– and obSAS+ groups, for any of the parameters (y0, ymax, z1, z2) obtained from per individual (Table 3) or per group modeling (Table 3 and Fig. 1F).


   DISCUSSION
TOP
 ABSTRACT
 STUDY DESIGN
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
Our data clearly show that in obese male patients SAS is not associated with an increase of salivary cortisol levels in the wake period or of plasma cortisol levels over the whole nyctohemeral period, including sleep. They also provide evidence that male obese patients display decreased levels of both salivary and plasma cortisol, especially in the trough and peak periods of the circadian rhythm. Cortisol levels appear to be affected by visceral fat excess rather than subcutaneous or global fat excess.

The lack of increase of cortisol levels in SAS patients was rather unexpected. Indeed, numerous pathophysiological conditions that relate to SAS because of disturbed sleep appear to increase the circulating levels of cortisol (7, 33). Provoked nocturnal awakenings appear to cause subsequent pulsatile cortisol release (31). Short-term sleep deprivation of normal-weight volunteers has been shown to increase plasma cortisol during the evening of the following day (18). In nonobese chronic insomniacs, plasma cortisol and ACTH levels were also found increased in the evening and the early part of the night (37). Chronic sleep curtailment has been linked to high evening salivary cortisol levels, increased activity of the sympathetic nervous system, and impaired glucose tolerance (32). It has recently been proposed that SAS may increase HPA axis activity and thereby be involved in the pathogenesis of the metabolic syndrome (7). According to this hypothesis, SAS represents a stress condition that might alter HPA axis control with loss of physiological cortisol secretion. In addition, secondary arousal, hypoxia, and autonomic activation may lead to HPA axis hyperactivity. However, only a few studies have directly assessed the relationship between SAS and the circulating levels of cortisol.

To our knowledge, four studies report measurements of spontaneous plasma cortisol levels performed in patients with SAS and in control subjects without SAS (6, 10, 12, 17). Only one study, in which a single morning blood sample was withdrawn at 0800, actually found increased morning plasma cortisol in patients with SAS; it also suggests that continuous positive airway pressure therapy could reverse this abnormality (6). However, other studies do not lead to the same conclusions (10, 12, 17). Indeed, Grunstein et al. (12), performing a single plasma cortisol measurement at 6:00 AM, found no increase in morning cortisol levels in sleep apnea patients compared with control subjects. More recently, Lanfranco et al. (17) have shown that basal morning plasma ACTH and cortisol levels and 24-h urinary free cortisol were similar in obese men with or without SAS and in normal-weight controls but that ACTH response to corticotropin-releasing hormone was higher in obese patients with SAS than in obese patients without SAS; however, no nyctohemeral exploration was performed during sleep. Besides, such studies should be considered with caution, since they were not controlled for BMI or fat distribution. Obese subjects with SAS usually have a greater degree of abdominal and/or visceral fat excess (30). Because visceral fat accumulation is associated with insulin resistance, dyslipidemia, increased secretion of deleterious adipokines, and decreased release of beneficial adipokines, some of the discrepancies found between subjects with or without SAS may relate to dissimilar body fat distribution, rather than to SAS status per se. We show here that obSAS+ and obSAS– men with comparable anthropometry and abdominal fat distribution exhibit similar nyctohemeral patterns of plasma cortisol. It is thus unlikely that increased cortisol levels participate in worsening the metabolic complications of obese patients with SAS. Indeed, multiple other mechanisms related to hypoxemia, sleep fragmentation, or both may provide a consistent pathophysiological link among SAS, metabolic disturbances, and cardiovascular diseases, such as increased sympathetic/parasympathetic balance, increased proinflammatory cytokines, endothelial dysfunction, and/or hemostatic alterations (23, 36, 38).

Our results demonstrate that flattened curves of salivary and plasma cortisol are especially found in obese subjects with documented excess of visceral fat, in whom both evening and morning cortisol levels are low. Using repeated measures of salivary cortisol, Rosmond et al. (27) previously showed that overweight middle-aged men with increased waist-to-hip ratio have greatly attenuated cortisol morning peak and circadian rhythmicity, blunted dexamethasone suppression, and stress-related disturbances of cortisol secretion; they suggested that such perturbations of the HPA axis may constitute a pathway to type 2 diabetes in patients with genetic susceptibility (28). In the same way, Putigano et al. (25) found that obese women with increased waist-to-hip ratio displayed diminished 0800–2400 salivary cortisol ratio but had similar diurnal concentrations of plasma cortisol and an unaltered diurnal rhythmic pattern compared with normal-weight controls. Other studies have documented a negative relationship between 0800 salivary cortisol levels and waist-to-hip ratio (9), but this relationship is not always found (21). Conversely, fewer studies have pointed out a positive correlation between salivary cortisol response to awakening and waist-to-hip ratio, and this only in men with mild overweight (34, 39). The lack of direct measurements of regional fat distribution may explain some of the discrepancies found in former studies that assessed the relationship.

The pathogenic mechanisms involved in the changes of the HPA axis activity in abdominal obesity remain unclear and may result from a combination of environmental, such as exposure to repeated or chronic stress, and genetic factors (4). The increased metabolic clearance of cortisol described in visceral obesity could explain decreased salivary levels (24); increased clearance may be caused by altered function of hepatic enzymes, since inactivation of cortisol by the 5{alpha}-reductase is increased (1) while regeneration of cortisol from cortisone by the 11beta-hydroxysteroid dehydrogenase type 1 (11beta-HSD1) is impaired in the liver of obese subjects (26). Another hypothesis suggests that the decrease of cortisol output and the flattened diurnal secretory pattern of cortisol may be caused by chronic stress, and a subsequent "burn-out" of central regulatory systems (4, 27). Both of these phenomena could be viewed as adaptive processes, aimed at protecting against an initial increase of cortisol production either of primary or environmental origin. However, these hypotheses do not satisfactorily explain the paradox of the concurrence of low circulating cortisol levels with the clinical and metabolic features of hypercorticism frequently seen in abdominal/visceral obesity. Whatever the cause of the decreased level of circulating cortisol, recent data suggest that increased peripheral actions of cortisol caused by changes of intracellular metabolism and/or glucocorticoid receptor sensitivity may participate in the paradox. Differences in glucocorticoid intratissue metabolism leading to increased local exposure to cortisol might explain the paradox of the presence in abdominal obesity of Cushing's features despite low cortisol concentrations. Indeed, glucocorticoid receptor expression is upregulated in visceral adipocytes, leading to an amplification of glucocorticoid signaling (5). Furthermore, an increased regeneration of cortisol from cortisone driven by the 11beta-HSD1 in the adipose tissue could play a pivotal role by amplifying local glucocorticoid action (8, 22). Indeed, increased cortisol action in VAT may lead to adverse metabolic and cardiovascular consequences through changes in the release of free fatty acids and/or adipokines such as adiponectin, proinflammatory cytokines, and/or PAI-1 (2, 3, 11). One cannot rule out that products released by VAT in response to increased local cortisol action may also alter the central command of the HPA axis and decrease cortisol output through a specific adipose-hypothalamic-pituitary feedback loop. Indeed, some cytokines produced by VAT have been shown to influence the central component of the HPA axis (35). Besides, a local increase of cortisol action inside the central hypothalamic and extrahypothalamic regulatory regions of the HPA axis, similar to that seen in VAT, could also increase cortisol feedback action, participate in decreased cortisol output, and generate the pattern of central burn-out described by others (4, 27). The lack of a stimulating effect of sleep disturbances caused by SAS upon cortisol secretion may possibly be a consequence of this pattern of central burn-out. Indeed, this remains to be demonstrated.


   GRANTS
TOP
 ABSTRACT
 STUDY DESIGN
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by grants from the Assistance Publique-Hôpitaux de Marseille, AORC 43/2001 (Marseille, France), and Association Régionale d'Aide pour la Respiration à Domicile (Aubagne, France).


   ACKNOWLEDGMENTS
 
We thank Drs. G. Bechis and J. P. Rosso for help in biochemical measurements.


   FOOTNOTES
 

Address for reprint requests and other correspondence: A. Dutour, Service d'Endocrinologie, des Maladies Métaboliques et de la Nutrition, Hôpital Nord, Chemin des Bourrely, 13915 Marseille Cedex 20, France (e-mail: anne.dutour{at}ap-hm.fr)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


   REFERENCES
TOP
 ABSTRACT
 STUDY DESIGN
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Andrew R, Phillips DI, Walker BR. Obesity and gender influence cortisol secretion and metabolism in man. J Clin Endocrinol Metab 83: 1806–1809, 1998.[Abstract/Free Full Text]
  2. Andrews RC, Walker BR. Glucocorticoids and insulin resistance: old hormones, new targets. Clin Sci 96: 513–523, 1999.[CrossRef][ISI][Medline]
  3. Ayachi SE, Paulmyer-Lacroix O, Verdier M, Alessi MC, Dutour A, Grino M. 11Beta-hydroxysteroid dehydrogenase type 1-driven cortisone reactivation regulates plasminogen activator inhibitor type 1 in adipose tissue of obese women. J Thrombosis Haemostasis 4: 621–627, 2006.[CrossRef]
  4. Björntorp P, Rosmond R. Obesity and cortisol. Nutrition 16: 924–936, 2000.[CrossRef][ISI][Medline]
  5. Boullu-Ciocca S, Paulmyer-Lacroix O, Fina F, Ouafik L, Alessi MC, Oliver C, Grino M. Expression of the mRNAs coding for the glucocorticoid receptor isoforms in obesity. Obes Res 11: 925–929, 2003.[ISI][Medline]
  6. Bratel T, Wennlund A, Carlstrom K. Pituitary reactivity, androgens and catecholamines in obstructive sleep apnoea. Effects of continuous positive airway pressure (CPAP). Respir Med 93: 1–7, 1999.[CrossRef][ISI][Medline]
  7. Buckley TM, Schatzberg AF. On the interactions of the hypothalamic-pituitary-adrenal (HPA) axis and sleep: normal HPA axis activity and circadian rhythm, exemplary sleep disorders. J Clin Endocrinol Metab 90: 3106–3114, 2005.[Abstract/Free Full Text]
  8. Desbriere R, Vuaroqueaux V, Achard V, Boullu-Ciocca S, Labuhn M, Dutour A, Grino M. Increased expression of 11beta-hydroxysteroid dehydrogenase type 1 mRNA in both visceral and subcutaneous adipose tissue of obese patients. Obes Res 14: 794–798, 2006.[Medline]
  9. Duclos M, Marquez Pereira P, Barat P, Gatta B, Roger P. Increased cortisol bioavailability, abdominal obesity, and the metabolic syndrome in obese women. Obes Res 13: 1157–1166, 2005.[ISI][Medline]
  10. Entzian P, Linnemann K, Schlaak M, Zabel P. Obstructive sleep apnea syndrome and circadian rhythms of hormones and cytokines. Am J Respir Crit Care Med 153: 1080–1086, 1996.[Abstract]
  11. Fallo F, Scarda A, Sonino N, Paoletta A, Boscaro M, Pagano C, Federspil G, Vetto R. Effect of glucocorticoids on adiponectin: a study in healthy subjects and in Cushing's syndrome. Eur J Endocrinol 150: 339–344, 2004.[Abstract]
  12. Grunstein RR, Handelsman DJ, Lawrence SJ, Blackwell C, Caterson ID, Sullivan CE. Neuroendocrine dysfunction in sleep apnea: reversal by continuous positive airways pressure therapy. J Clin Endocrinol Metab 68: 352–358, 1989.[Abstract]
  13. Jensen MD, Kanaley JA, Reed JE, Sheedy PF. Measurement of abdominal and visceral fat with computed tomography and dual-energy x-ray absorptiometry. Am J Clin Nutr 61: 274–278, 1995.[Abstract/Free Full Text]
  14. Jessop DS, Dallman MF, Fleming D, Lightman SL. Resistance to glucocorticoid feedback in obesity. J Clin Endocrinol Metab 86: 4109–4114, 2001.[Abstract/Free Full Text]
  15. Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep 14: 540–545, 1991.[ISI][Medline]
  16. Jusko WJ, Slaunwhite WR Jr, Aceto T Jr. Partial pharmacodynamic model for the circadian-episodic secretion of cortisol in man. J Clin Endocrinol Metab 40: 278–289, 1975.[Abstract]
  17. Lanfranco F, Gianotti L, Pivetti S, Navone F, Rossetto R, Tassone F, Gai V, Ghigo E, Maccario M. Obese patients with obstructive sleep apnoea syndrome show a peculiar alteration of the corticotroph but not of the thyrotroph and lactotroph function. Clin Endocrinol (Oxf) 60: 41–48, 2004.[CrossRef][Medline]
  18. Leproult R, Copinschi G, Buxton O, Van Cauter E. Sleep loss results in an elevation of cortisol levels the next evening. Sleep 20: 865–70, 1997.[ISI][Medline]
  19. Ljung T, Holm G, Friberg P, Andersson B, Bengtsson BA, Svensson J, Dallman M, McEwen B, Björntrop P. The activity of the hypothalamic-pituitary-adrenal axis and the sympathetic nervous system in relation to waist/hip circumference ratio in men. Obes Res 8: 487–495, 2000.[ISI][Medline]
  20. Malhotra A, White DP. Obstructive sleep apnoea. Lancet 360: 237–245, 2002.[CrossRef][ISI][Medline]
  21. Pasquali R, Biscotti D, Spinucci G, Vicenatti V, Genazzani AD, Sgarbi L, Casimirri F. Pulsatile secretion of ACTH and cortisol in premenopausal women: effect of obesity and body fat distribution. Clin Endocrinol 48: 603–612, 1998.[CrossRef][Medline]
  22. Paulmyer-Lacroix O, Boullu S, Oliver C, Alessi MC, Grino M. Expression of the mRNA coding for the 11beta-hydroxysteroid dehydrogenase type 1 in adipose tissue from obese patients: an in situ hybridization study. J Clin Endocrinol Metab 87: 2701–2705, 2002.[Abstract/Free Full Text]
  23. Punjabi NM, Polotsky VY. Disorders of glucose metabolism in sleep apnea. J Appl Physiol 99: 1998–2007, 2005.[Abstract/Free Full Text]
  24. Purnell JQ, Brandon DD, Isabelle LM, Loriaux DL, Samuels MH. Association of 24-hour cortisol production rates, cortisol-binding globulin, and plasma-free cortisol levels with body composition, leptin levels and aging in adult men and women. J Clin Endocrinol Metab 89: 281–287, 2004.[Abstract/Free Full Text]
  25. Putigano P, Dubini A, Toja P, Invitti C, Bonfanti S, Redaelli G, Zappulli D, Cavagnini F. Salivary cortisol measurement in normal-weight, obese and anorexic women: comparison with plasma cortisol. Eur J Endocrinol 145: 165–171, 2001.[Abstract]
  26. Rask E, Walker BR, Soderberg S, Livingstone DEW, Eliasson M, Johnson O, Andrew R, Olsson T. Tissue-specific changes in peripheral cortisol metabolism in obese women: increased adipose 11beta-hydroxysteroid dehydrogenase type 1 activity. J Clin Endocrinol Metab 87: 3330–3337, 2002.[Abstract/Free Full Text]
  27. Rosmond R, Dallman MF, Björntorp P. Stress-related cortisol secretion in men: relationships with abdominal obesity and endocrine, metabolic and hemodynamic abnormalities. J Clin Endocrinol Metab 83: 1853–1859, 1998.[Abstract/Free Full Text]
  28. Rosmond R. Stress induced disturbances of the HPA axis: a pathway to type 2 diabetes? Med Sci Monit 9: RA35–RA39, 2003.[Medline]
  29. Shamsuzzaman AS, Gersh BJ, Somers VK. Obstructive sleep apnea: implications for cardiac and vascular disease. JAMA 290: 1906–1914, 2003.[Abstract/Free Full Text]
  30. Shinohara E, Kihara S, Yamashita S, Yamane M, Nishida M, Arai T, Kotani K, Nakamura T, Takemura K, Matsuzawa Y. Visceral fat accumulation as an important risk factor for obstructive sleep apnoea syndrome in obese subjects. J Intern Med 241: 11–18, 1997.[CrossRef][ISI][Medline]
  31. Spath-Schwalbe E, Gofferje M, Kern W, Born J, Fehm HL. Sleep disruption alters nocturnal ACTH and cortisol secretory patterns. Biol Psychiatry 29: 575–584, 1991.[CrossRef][ISI][Medline]
  32. Spiegel K, Knutson K, Leproult R, Tasali E, Van Cauter E. Sleep loss: a novel risk factor for insulin resistance and type 2 diabetes. J Appl Physiol 99: 2008–2019, 2005.[Abstract/Free Full Text]
  33. Steiger A. Sleep and the hypothalamic-pituitary-adrenocortical system. Sleep Med Rev 6: 125–138, 2002.[CrossRef][ISI][Medline]
  34. Steptoe A, Kunz-Ebrecht SR, Brydon L, Wardle J. Central adiposity and cortisol responses to waking in middle-aged men and women. Int J Obes 28: 1168–1173, 2004.[CrossRef][ISI][Medline]
  35. Turnbull AV, Rivier CL. Regulation of the hypothalamic-pituitary-adrenal axis by cytokines: actions and mechanisms of action. Physiol Rev 79: 1–7, 1999.[Abstract/Free Full Text]
  36. Vgontzas AN, Bixler EO, Chrousos GP. Metabolic disturbances in obesity versus sleep apnoea: the importance of visceral obesity and insulin resistance. J Intern Med 254: 32–44, 2003.[CrossRef][ISI][Medline]
  37. Vgontzas AN, Zoumakis M, Bixler EO, Lin HM, Prolo P, Mastorakos G, Vela-Bueno A, Kales A, Chrousos GP. Chronic insomnia is associated with nyctohemeral activation of the hypothalamic-pituitary-adrenal axis: clinical implications. J Clin Endocrinol Metab 86: 3787–3794, 2001.[Abstract/Free Full Text]
  38. von Känel R, Dimsdale JE. Hemostatic alterations in patients with obstructive sleep apnea and the implications for cardiovascular disease. Chest 124: 1956–67, 2003.[CrossRef][ISI][Medline]
  39. Wallerius S, Rosmond R, Ljung T, Holm G, Björntorp P. Rise in morning saliva cortisol is associated with abdominal obesity in men: a preliminary report. J Endocrinol Invest 26: 616–619, 2003.[ISI][Medline]



This article has been cited by other articles:


Home page
 Am. J. Physiol. Endocrinol. Metab.Home page
G. Carneiro, S. M. Togeiro, L. F. Hayashi, F. F. Ribeiro-Filho, A. B. Ribeiro, S. Tufik, and M. T. Zanella
Effect of continuous positive airway pressure therapy on hypothalamic-pituitary-adrenal axis function and 24-h blood pressure profile in obese men with obstructive sleep apnea syndrome
Am J Physiol Endocrinol Metab, August 1, 2008; 295(2): E380 - E384.
[Abstract] [Full Text] [PDF]

Home page
 Eur Respir JHome page
N. Burioka, S. Koyanagi, M. Endo, M. Takata, Y. Fukuoka, M. Miyata, K. Takeda, H. Chikumi, S. Ohdo, and E. Shimizu
Clock gene dysfunction in patients with obstructive sleep apnoea syndrome
Eur. Respir. J., July 1, 2008; 32(1): 105 - 112.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
293/2/E466    most recent
00126.2007v1
Right arrow Alert me when this article is cited
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 ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dadoun, F.
Right arrow Articles by Dutour, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dadoun, F.
Right arrow Articles by Dutour, A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2007 by the American Physiological Society.